<%BANNER%>

Embedding the Refugee Experience: Forced Migration and Social Networks in Dar es Salaam, Tanzania


PAGE 1

EMBEDDING THE REFUGEE EXPERIENCE : FORCED MIGRATION AND SOCIAL NETWORKS IN DAR ES SALAAM, TANZANIA By ROOS WILLEMS A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2003

PAGE 2

Copyright 2003 by Roos Willems

PAGE 3

This document is dedicated to the refugees in Dar es Salaam, Tanzania.

PAGE 4

ACKNOWLEDGMENTS Many persons have contributed towards the success of my doctoral study. I would first like to thank the members of my advisory committee, namely, Professors Anthony Oliver-Smith, Paul Magnarella, Goran Hyden and Maria Grosz-Ngat, for their scholarly advisement. Dr. Oliver-Smith, my advisor, deserves special mention for his guidance and help from the inception of proposal writing to the production of this dissertation, as does Dr. Maria Grosz-Ngat. The staff at the Department of Anthropology, Karen Jones, Patricia King and Salena Robinson, deserve my deep appreciation for their help with routine administrative activities. I owe special thanks to the Department of Anthropology and the Center for African Studies at the University of Florida for the various graduate assistantships that made it possible for me to complete my doctoral studies, and to the Wenner-Gren Foundation for Anthropological Research which funded my field research. In Tanzania, grateful acknowledgements are due to Dr. Masanja of the Department of Sociology and Dr. Kamanga, Director of the Center for the Study of Forced Migration at the University of Dar es Salaam, for sharing their knowledge and facilitating numerous local contacts with institutions and persons alike. Arjun Jain at UNHCR, Dar es Salaam, was indispensable in providing information and facilitating contacts with potential informants, as was Dr. Sinje of Umati at Upanga. I am very grateful to the many forced migrants in Dar es Salaam who were willing to share their experiences with me and to introduce me further into the refugee iv

PAGE 5

communities, as well as to the staff of government offices and humanitarian agencies in Dar es Salaam. While apologizing for not mentioning all the names of individuals, I would like to express my appreciation for my research assistants Denis, Xavier, Mopetane, Gaspard, Bati and above all, to Justine Buhanda for her untiring and ever efficient endeavors, without whom this research project would have been so much more difficult to complete. Karen, Jasper and Suzan deserve special mention for providing me with accomodation and very much needed social contacts during my stay in Dar es Salaam, as do Frieda and Eva for their companionship and encouragements. Special thanks for friendship and logistical support go to Lilian Osaki, as well as to Rob Duijs, whose frank comments on earlier writings were very helpful. Also very much appreciated were the insights and suggestions from the members of our anthropology writing group, namely, Fatma, Elli, Antoinette, Mercedes and Ermitte, as well as insightful conversations with numerous other fellow-students. I thank them all. v

PAGE 6

TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES.............................................................................................................xi LIST OF FIGURES.........................................................................................................xiii ABSTRACT.......................................................................................................................xv CHAPTER 1 CONTEXTUALIZING RESEARCH TOPIC AND CHOICE OF METHODS..........1 1.1 Contextualizing the Research Topic.......................................................................3 1.1.1 Anthropological Concerns............................................................................5 1.1.1.1 Anthropology and forced migration...................................................6 1.1.1.2 Anthropology and social networks.....................................................7 1.1.1.3 Forced migration and social networks...............................................9 1.1.2 Refugee Studies and Gender Studies..........................................................12 1.1.2.1 De-essentializing the "refugee-experience".....................................13 1.1.2.2 Gender and forced migration............................................................17 1.1.2.3 Engendering forced migration..........................................................19 1.1.3 The Social World of Refugees...................................................................20 1.1.3.1 Coping with loss...............................................................................21 1.1.3.2 Social networks................................................................................23 1.1.3.3 Gender, nationality and age..............................................................26 1.2 Choice of Methods................................................................................................31 1.2.1 Collecting Data...........................................................................................32 1.2.2 Collecting "Voices"....................................................................................35 1.2.3 Representativeness.....................................................................................38 1.3 In the Field............................................................................................................40 1.3.1 Methods Used.............................................................................................41 1.3.2 Selecting Respondents................................................................................44 1.3.3 Locating Informants...................................................................................47 1.4 Structure of the Dissertation.................................................................................54 2 SITUATING THE DISSERTATION PROJECT AND RESEARCH SITE..............58 2.1 Background to the Research Project.....................................................................58 vi

PAGE 7

2.1.1 The Refugee Producing Countries..............................................................59 2.1.1.1 The historical interconnectedness of the Great Lakes countries......60 2.1.1.2 The colonial impact in Congo, Rwanda and Burundi......................62 2.1.1.3 From independence to the early 1990s.............................................66 2.1.1.4 From the early 1990s onward...........................................................69 2.1.2 The Refugee Receiving Country................................................................76 2.1.2.1 Nyerere's Tanzania...........................................................................77 2.1.2.2 Africa's "open door" policy in Tanzania..........................................79 2.1.2.3 The change in climate......................................................................80 2.1.3 Rwandese, Burundese and Congolese refugees in Tanzania......................83 2.1.3.1 The refugee camps in Western Tanzania.........................................85 2.1.3.2 Urban refugees in Tanzania..............................................................89 2.2 The Research Site: Dar es Salaam........................................................................91 2.2.1 Governmental and Political Structures.......................................................94 2.2.1.1 The 1998 Refugee Act and the Refugee Department.......................94 2.2.1.2 The Ten House System (Nyumba Kumi)..........................................96 2.2.1.3 The Center for the Study of Forced Migration.................................98 2.2.2 Humanitarian Organizations.....................................................................100 2.2.2.1 United Nations High Commissioner for Refugees.........................100 2.2.2.2 Umati..............................................................................................102 2.2.2.3 Other organizations........................................................................103 2.2.3 The Social Context...................................................................................106 2.2.3.1 Level of Contact.............................................................................110 2.2.3.2 Intermarriage..................................................................................112 2.2.3.3 Refugee policy................................................................................113 2.2.3.4 Likes, dislikes and differences.......................................................114 2.2.3.5 Summarizing findings....................................................................116 2.3. Conclusion.........................................................................................................117 3 HEADING FOR DAR ES SALAAM......................................................................125 3.1 Localizing Voices...............................................................................................125 3.1.1 Refugees in Post-Independence Dar es Salaam........................................126 3.1.2 Demographics of the Sample....................................................................129 3.1.2.1 Religion, marital status, and children.............................................130 3.1.2.2 Education level and occupation......................................................133 3.1.2.3 Arrival in Tanzania, duration of stay in Dar es Salaam and UNHCR registration.............................................................................137 3.1.3 Introducing the Protagonists.....................................................................142 3.2 Leaving Home....................................................................................................148 3.2.1 Fleeing Wars.............................................................................................149 3.2.2 Personal Loss............................................................................................150 3.2.3 Saving One's Life.....................................................................................151 3.2.4 Political Reasons......................................................................................152 3.3 Avoiding the Refugee Camps.............................................................................154 3.3.1 "Miserable" Camp Life.............................................................................156 3.3.2 Security Concerns.....................................................................................157 vii

PAGE 8

3.3.3 Joining Relatives or Compatriots.............................................................159 3.4 Arriving in Dar Es Salaam..................................................................................160 3.4.1 Locating Relatives or Other Compatriots.................................................162 3.4.2 Tanzanian Friends....................................................................................162 3.4.3 UNHCR....................................................................................................163 3.5 Prospects of Returning Home.............................................................................164 3.5.1 The Return Home.....................................................................................165 3.5.2 Alternatives to Repatriation......................................................................166 3.6 Emerging Profile(s)............................................................................................168 3.6.1 Gender......................................................................................................168 3.6.2 Nationality................................................................................................169 3.6.3 Age Groups...............................................................................................170 4 ADAPTING TO THE NEW LIFE...........................................................................178 4.1 Local Interactions...............................................................................................180 4.1.1 "Too many questions"..............................................................................183 4.1.2 Language..................................................................................................184 4.1.3 "Lack of respect"......................................................................................188 4.1.4 Religion....................................................................................................191 4.2 Economic Endeavors..........................................................................................194 4.2.1 Permits......................................................................................................195 4.2.1.1 Applying for a permit.....................................................................196 4.2.1.2 Marrying a Tanzanian citizen.........................................................196 4.2.1.3 "Adoptive" fathers..........................................................................198 4.2.1.4 Staying out of prison......................................................................198 4.2.2 Kupata riziki ("Making ends meet").........................................................200 4.2.3.2 Self-Employment...........................................................................200 4.2.2.2 Trading activities............................................................................203 4.2.2.3 Paid employment............................................................................205 4.3 Major Pre-occupations........................................................................................208 4.3.1 Education..................................................................................................210 4.3.2 Sexual Exploitation..................................................................................212 4.3.3 Aspiring to Leave Tanzania......................................................................214 4.3.4 Worrying about Lost Relatives.................................................................215 4.4 Social Worlds......................................................................................................217 4.4.1 Meeting People.........................................................................................218 4.4.1.1 Meeting Tanzanians.......................................................................218 4.4.1.2 Meeting fellow-refugees................................................................220 4.4.1.3 Socializing with friends..................................................................223 4.4.2 Meeting Places..........................................................................................225 4.4.2.1 The neighborhood..........................................................................226 4.4.2.2 The church/mosque........................................................................227 4.4.2.3 Associations...................................................................................228 4.5 Concluding Remarks..........................................................................................232 viii

PAGE 9

5 DE-CONSTRUCTING SOCIAL NETWORKS......................................................237 5.1 The Average Personal Social Network...............................................................239 5.1.1 Gender, Nationality and Age....................................................................241 5.1.2 Other Demographic and Socio-Economic Attributes...............................243 5.1.3 Place of Residence, Lenght of Relationship and Frequency of Contact..245 5.1.4 Ways of Meeting Alters and Types of Support Received........................248 5.1.5 Summarizing Findings..............................................................................250 5.2 Homophily..........................................................................................................252 5.2.1 Same Gender............................................................................................252 5.2.2 Same Nationality......................................................................................254 5.2.3 Same Age Category..................................................................................255 5.2.4 Summarizing Findings..............................................................................255 5.3 Disaggregating Ego............................................................................................257 5.3.1 Gender, Nationality and Age Groups.......................................................257 5.3.2 Attributes of Relationships.......................................................................262 5.3.3 Alters by Ways of Meeting and Types of Support Provided....................266 5.3.4 Summarizing Findings..............................................................................268 5.4 Disaggregating Alters.........................................................................................270 5.4.1 Ways of Meeting......................................................................................270 5.4.2 Types of Support Provided.......................................................................272 5.4.3 Summarizing Findings..............................................................................274 5.5 Conclusions.........................................................................................................276 6 EMBEDDING THE REFUGEE EXPERIENCE.....................................................290 6.1 Gender.................................................................................................................292 6.1.1 Women as "Political Objects"..................................................................294 6.1.2 Women as "Apolitical Subjects"..............................................................299 6.1.3 "You are who you know."........................................................................302 6.2 Nationality..........................................................................................................307 6.2.1 Economics................................................................................................311 6.2.2 Politics......................................................................................................317 6.2.3 History......................................................................................................322 6.3 Age......................................................................................................................330 6.3.1 Persons younger than 25 years of age......................................................334 6.3.2 Between 26 and 35 years of age...............................................................335 6.3.3 Between 36 and 45 years of age...............................................................338 6.3.4 Persons older than 45 years of age...........................................................340 6.4 Concluding Remarks..........................................................................................343 7 REFLECTIONS, PREOCCUPATIONS AND CONSIDERATIONS.....................348 7.1 "To aid or not to aid?".........................................................................................350 7.2 Positionalities......................................................................................................356 7.3 Urban Refugees, Social Networks and Forced Migration..................................360 7.4 Epilogue..............................................................................................................363 ix

PAGE 10

APPENDIX A MAPS.......................................................................................................................365 B TANZANIAN PERCEPTIONS ON REFUGEES IN DAR ES SALAAM.............372 C SOCIAL NETWORKS SURVEY............................................................................376 LIST OF REFERENCES.................................................................................................385 BIOGRAPHICAL SKETCH...........................................................................................413 x

PAGE 11

LIST OF TABLES Table page 2-1. Level of Contact with Refugees ............................................................................118 2-2. Tanzanian Perceptions on Prevalence of Refugee Nationalities ...........................119 2-3. Tanzanian Perceptions on Refugee Occupations .................................................119 2-4. Tanzanian Perceptions on Intermarriage with Refugees .......................................120 2-5. Tanzanian Perceptions on Refugee Policies ..........................................................121 2-6. Level of "Don't know" Replies ..............................................................................122 2-7. Refugee Characteristics Liked ..............................................................................122 2-8. Refugee Characteristics Disliked ..........................................................................123 2-9. Perceived Differences between Tanzanians and Refugees ...................................124 3-1. Composition of Sample: Gender, Nationality, and Age Group.............................172 3-2. Education Level Compared to Camp Population ..................................................172 3-3. Demographic Characteristics of Sample ...............................................................173 3-4. Education Level and Professional Occupation ......................................................174 3-5. Characteristics of Stay in Tanzania and Dar es Salaam ........................................175 3-6. Reasons for Leaving Home and Heading for Dar es Salaam ................................176 3-7. First day in Dar es Salaam and Prospects of Returning Home .............................177 4-1. Comparing Life at Home and in Exile ..................................................................235 4-2. Interactions with and Perceptions of Tanzanian Nationals ...................................236 5-1. Ego's alters disaggregated by Gender, Nationality and Age..................................279 5-2. Social Network Members by Gender, Nationality and Age ..................................280 xi

PAGE 12

5-3. Demographic and Other Socio-Economic Attributes ............................................281 5-4. Relationship attributes of Social Network Members ............................................282 5-5. Ways of Meeting Network Members and Types of Support Received .................283 5-6. Homophilous relationships for sex, nationality and age group .............................284 5-7. Social networks by gender, nationality and age category .....................................285 5-8. Relationship attributes by gender, nationality and age category ...........................286 5-9. Ways of meeting and types of support received ....................................................287 5-10. Alters by ways of meeting gender, nationality and age category .......................288 5-11. Alters by type of support gender, nationality, and age category .........................289 xii

PAGE 13

LIST OF FIGURES Figure page 3-1: Religion by gender, nationality and age.................................................................131 3-2: Marital status and living arrangements..................................................................131 3-3: Number of Children in Dar es Salaam...................................................................132 3-4: Education levels.....................................................................................................133 3-5: Employment activities at home before leaving......................................................134 3-6: Employment activities after arrival in Dar es Salaam............................................135 3-7: Year of arrival in Tanzania.....................................................................................137 3-8: Number of years resided in refugee camp in Tanzania..........................................138 3-9: Length of Stay in Dar es Salaam............................................................................139 3-10: Registered with UNHCR........................................................................................141 3-11: Reasons for leaving the home country...................................................................148 3-12: Motives for choosing Dar es Salaam instead of the refugee camps.......................155 3-13: Accommodation upon arrival in Dar es Salaam.....................................................161 3-14: Prospects on Returning Home................................................................................164 4-1: Comparing life at home and in Dar es Salaam.......................................................179 4-2: Why establishing relationships in Dar es Salaam is difficult.................................181 4-3: Major worries in Dar es Salaam.............................................................................209 5-1: Homophilous ties for gender..................................................................................253 5-2: Homophilous ties for nationality............................................................................254 5-3: Homophilous ties for age category.........................................................................255 xiii

PAGE 14

5-4: Pre-Dar es Salaam alters by gender........................................................................258 5-5: Post-Dar es Salaam alters by gender......................................................................258 5-6: Pre-Dar es Salaam alters by nationality.................................................................259 5-7: Post-Dar es Salaam alters by nationality................................................................259 5-8: Pre-Dar es Salaam alters by age group...................................................................261 5-9: Post-Dar es Salaam alters by age group.................................................................261 5-10: Preversus post-Dar es Salaam alters....................................................................262 5-11: Pre-Dar es Salaam alters by place of residence......................................................262 5-12: Post-Dar es Salaam alters by place of residence....................................................263 5-13: Pre-Dar es Salaam alters by length of relationship................................................264 5-14: Post-Dar es Salaam alters by length of relationship...............................................264 5-15: Pre-Dar es Salaam alters by frequency of contact..................................................265 5-16: Post-Dar es Salaam alter by frequency of contact..................................................265 5-17: Pre-Dar es Salaam alters by way of meeting..........................................................267 5-18: Post-Dar es Salaam alters by way of meeting........................................................267 5-19: Types of Support received from pre-Dar es Salaam alters.....................................268 5-20: Types of Support received from post-Dar es Salaam alters...................................268 5-21: Pre-Dar es Salaam alters by ways of meeting........................................................270 5-22: Post-Dar es Salaam alters by ways of meeting......................................................271 5-23: Pre-Dar es Salaam alters by type of support..........................................................272 5-24: Post-Dar es Salaam alters by type of support.........................................................273 xiv

PAGE 15

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy EMBEDDING THE REFUGEE EXPERIENCE: FORCED MIGRATION AND SOCIAL NETWORKS IN DAR ES SALAAM, TANZANIA By Roos Willems December, 2003 Chair: Anthony Oliver-Smith Major Department: Anthropology Due to the tragic events in the countries of the Great Lakes region during the 1990s, more than one million refugees have sought a safe haven in Tanzania. Increasing numbers among these, estimated to be in the tens of thousands, resist going to one of the refugee camps and head for Dar es Salaam instead, even if it means foregoing any material assistance from humanitarian agencies and legal residence status from the national authorities in the country of asylum. Living scattered across the Tanzanian metropolis, these forced migrants show their resilience under adverse circumstances by coping with a new language, economic challenges and unfamiliar social environments. In this process, they rely heavily on their personal social networks, members of which provide them with financial and material assistance as well as emotional support in the form of friendship, advice and companionship. This process of rebuilding their shattered social worlds is analyzed in terms of the gender, nationality and age group of xv

PAGE 16

the refugees. The central thesis is that there is no essential refugee experience, and that the coping strategies of refugees are embedded in the social, cultural, economic, political and historical background of the social groups, be they gender, nationality or age groups, that the individual refugees are member of. The principal methodological tool applied to the analysis of the data collected during fieldwork sessions in 2000, 2001 and 2002, is social network analysis. Its findings are contextualized qualitatively by the accounts, life histories and interviews with informants of Congolese, Burundese, Rwandese and Tanzanian origin. Social network analysis is proposed as the tool to situate individual persons' agency and interactions at the micro-level within the structures and infrastructures (including events at the regional, national and international level) constituting the macro-framework. xvi

PAGE 17

CHAPTER 1 CONTEXTUALIZING RESEARCH TOPIC AND CHOICE OF METHODS This research project focuses on the self-settlement strategies of forced migrants in an urban context, and explores how individual social networks are deployed in the process of coping with displacement. The study is situated in Dar es Salaam, Tanzania, and concentrates on urban refugees of Congolese, Burundese and Rwandese nationality. Reconstructing and deploying one's personal social networks becomes an essential part of survival in a situation where the policy of host governments and international agencies is to not provide the self-settled urban refugee with material assistance, as is the case in Tanzania. The variations in the ways in which Congolese, Burundese and Rwandese refugees, men and women, young and old, adapt to the new life in exile while reconstructing their social worlds are subsequently contextualized in a larger historical, socio-political and economic framework of the respective social groups under study. By April 2002, Tanzania was hosting approximately 512,000 refugees assisted by the United Nations High Commissioner for Refugees [UNHCR], the overwhelming majority of whom originated from Burundi (69%), Congo-Kinshasa (25%), and Rwanda (5%) (UNHCR 2002a).1 Tanzanian refugee policy stipulates that all assistance to refugees is to be provided only in the "designated areas" in rural Western Tanzania. In 1 The Tanzanian government estimates that the total number of refugees in the country (UNHCR assisted and self-settled) is as high as one million. This study will concern itself only with refugees of the above mentioned three nationalities. The remaining 1% of refugees in Tanzania comprise a/o Somali's, Ugandans, Ethiopians, Sudanese. 1

PAGE 18

2 the wake of the ever increasing rate of urbanization on the African continent,2 a growing number of the refugees in Tanzania are urbanites, who prefer to live in the city, even when this means foregoing the material assistance offered by the international agencies. Being clandestine residents in Dar es Salaam, no official statistics exist on their number. Congolese, as well as Burundese and Rwandese, have a long history of finding refuge in Tanzania due to the crises and political tensions that have reigned in their respective home countries over the past four decades. This study, contrary to others on the same topic,3 will focus specifically on "first-generation" refugees, those who have vivid memories of life in the home country and are still in the first few years after arriving in the host country. Despite the enormous difficulties encountered in building up a livelihood in a foreign city, close to all of the refugees interviewed in this study were categorically resisting the official Tanzanian policy of indiscriminately allocating all refugees to camps in Western Tanzania. My argument is that the coping strategies of individual refugees are embedded in the larger historical, cultural, social, political, economic and national frameworks that constituted their lives before the flight. Dar es Salaam, at this point in time, provided the perfect venue for a cross-cultural analysis allowing a comparison between Congolese, Burundese and Rwandese refugee men and women, young and old. This first chapter will contextualize the research project both theoretically and practically by situating the 2 According to UNDP, in 1999, 33.5 % of the Sub-Saharan African population lived in urban centers, compared to 46.5% worldwide. On the basis of a higher growth rate for urban Africa, by 2015 its urban population will comprise 43.3% of the total, increasingly closer to the global average of 53.2% projected for 2015 (UNDP 2001). 3 Sommers (2001) in his study of urban refugees in Dar es Salaam, concentrates solely on "young men . Burundian nationals who have little or no memory of Burundi," alternatively called "second Burundi refugee generation" (7) by the author.

PAGE 19

3 research topic within the literature, documenting my phenomenological approach to fieldwork and methods followed by their subsequent application and consequences thereof in the field setting. 1.1 Contextualizing the Research Topic The goal of this research project is a cross-cultural comparison of the ways in which urban refugee men and women of different nationalities and ages cope with a situation of forced migration and rebuild their personal social networks. Before venturing out into any disciplinary concerns, however, I would like to elaborate on the terms "refugee" and "forced migrant." The legal definition of refugees stipulates them to be persons having crossed international borders in fear of personal persecution, and as such entitled to protection from the UNHCR. The 1951 UN Convention defines a refugee as any person: . who owing to a well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable or, owing to such fear or for reasons other than personal convenience, is unwilling to avail himself of the protection of that country. . (UNHCR 2000a: 6) In 1969, the Organization of African States [OAU] enlarged that definition (as applied in African countries) to include persons: . who owing to external aggression, occupation, foreign domination or events seriously disturbing public order in either part or the whole of his country of origin or nationality, is compelled to leave his place of habitual residence in order to seek refuge in another place outside his country. . (82) Over the past decade, a debate has been raging as to whether to enlarge the international definition of a refugee even further by including "internally displaced persons" (IDPs) as equally entitled to the protection of the UNHCR (Lee 1996). However, as this would, according to some observers, challenge the sovereignty of the

PAGE 20

4 nation-state--because IDPs by definition never crossed the borders of their homeland--it is unlikely that this debate will draw to a close any time soon. Nevertheless, recognizing that both refugees and IDPs suffer the same experience of "uprootedness" and of having forcibly left their homes, the term "forced migrant" emerged to encompass both categories and create a wider focus for both academic and policy makers' interests (Adelman 2001). As such, "a forced migrant can be defined roughly as someone who is forced to leave his or her home because of a real or perceived threat to life or well-being" (Reed 1998:1). It is my understanding that in the field of Diaspora Studies, the term "forced migrant" is reserved for those who had no say whatsoever in their "migration," such as for example, the Africans subjected to the transatlantic slave trade. This is not the guiding principle in the area of refugee studies, where although named forced migrants, these individuals are recognized to have taken the decision to leave.4 For the purpose of this dissertation the terms "refugee" and "forced migrant" will be used interchangeably.5 In this research project, I focus solely on Congolese, Burundese and Rwandese nationals who sought refuge across international borders, namely in Tanzania, for reasons ranging from personal political persecution and situations of war to a motivation based on a lack of economic opportunities due to the disturbances created in day-to-day life by politically motivated (violent) events. In recognition of the variety of conditions under which the refugees from the Great Lakes 4 . depending on their financial possibilities. This point was brought home to me only when one of the Congolese informants was recounting how in 1996 when he left, many Congolese were unable to pay the US$20 boat fare to cross Lake Tanganyika and find refuge and safety in Tanzania. Umutesi's detailed account of her year long flight across Zaire highlights more than once that the quest for safety is an expensive one (2000). 5 In the same line, the Center for the Study of Forced Migration at the University of Dar es Salaam in Tanzania concerns itself almost exclusively with studies of refugees.

PAGE 21

5 region flee their homes and seek asylum abroad, Tanzanian authorities apply different procedures for different nationalities. As such, Congolese and Burundese asylum seekers were recognized as prima facie refugees under the 1969 OAU Protocol, while Rwandese asylum seekers, at the time of research, were expected to undergo an individual interview before being accorded refugee status, in line with the 1951 UN Convention. Ultimately, however, there was no difference between the two groups with regard to their rights as refugees in Tanzania or the challenges that refugee life posed upon them. 1.1.1 Anthropological Concerns In anthropological terms, refugees are people who have undergone a violent separation and--unless or until they are "incorporated" as citizens into their host state or have returned to their state of origin--find themselves in "transition," or in a state of "liminality" (Harrell-Bond & Voutira 1992). In Turner's words: "Liminal entities are neither here nor there; they are betwixt and between the positions assigned and arrayed by law, custom, convention, and ceremonial" (1969:95). Refugees have to deal with chaos. They have been ripped from their established lives; often their families are dismembered and their support systems destroyed. Many of them are struggling with a new and unfamiliar social and physical environment, in which the belief systems or styles of action they brought into refugee status do not work well. It is in this sense that they are "betwixt and between" their old and new lives, not knowing what to expect next. Gold rightly points out that, while voluntary immigrants have frequently made a conscious decision to accept the indignities of dealing with an alien culture, refugees [as involuntary migrants] often cling to their traditional roles and values and may resist adaptation. (Gold 1992:18)

PAGE 22

6 As the following chapters will apply illustrate, the Congolese, Burundese and Rwandese refugees in Dar es Salaam are not inclined to adapt to the culture of their host country, except in instances where not adapting could possibly have consequences endangering their personal safety, such as the use of Swahili as it is spoken in Dar es Salaam (see chapter 4). In a situation of forced migration, which is ultimately a transient one, given that the majority of respondents arrived in Dar es Salaam fairly recently and aspire to repatriate as soon as the conditions at home allow it (see chapter 3), the emphasis will be on coping rather than adapting. 1.1.1.1 Anthropology and forced migration Hansen and Oliver-Smith (1982) point out that when people migrate across national borders as forced migrants, they do not give up their understandings of what various individual, family, and social behaviors mean. On the contrary, "our responses to change are always innovative but are set within the parameters specified by our perceptual frameworks, and these are a function of our past" (Oliver-Smith 1986:16). Refugees will choose from alternative behaviors those that will allow them to maintain their basic conceptions of what life is amidst the turmoil and disruptions, a process possibly (but not necessarily) leading to socio-cultural change through innovative behaviors (Geiger 1993). It is in this sense that Harrell-Bond and Voutira stated that: "of all the disciplines involved in the study of human behavior, we contend that anthropology has the most to contribute to the study of refugees" (Harrell-Bond and Voutira 1992: 6). Just as anthropology may contribute to the study of refugees--and, on a more practical note, to the designing of refugee relief efforts as well (Waldron 1988)--the study of refugees may also contribute to the development of anthropological theory. Anthropology should not just be "the comfortable anthropology of social organization," it

PAGE 23

7 should include as well "the painful anthropology of disruption and despair" (Davis 1992:149). The particular circumstances of refugee situations allow us to uncover certain socio-cultural dynamics and realities that could be hard to distinguish or identify in "normal" life. At the same time, I would like to point out that these socio-cultural dynamics and realities are not always generally applicable. Rather, members belonging to one or the other social groups (i.e., men versus women, young versus old, Rwandese versus Burundese or Congolese) will act following the cultural norms and values governing the pre-flight societal positions of their respective gender, nationality or age groups. As such, studying forced migrants through an anthropological lens may contribute to the further development of social science theory. 1.1.1.2 Anthropology and social networks Social networks as a methodology emerged in the 1960s out of the Rhodes-Livingstone Institute (Lusaka) in the wake of the ever increasing urbanization rate in the Copperbelt area in Zambia.6 While urban anthropology itself, as a "self-labeled body of research" (Sanjek 1990:151), saw the light in the 1960s, only in the 1980s did it rear its head in the anthropological literature.7 Rendering all but obsolete the traditional kit of anthropological research tools designed for small-scale rural communities, the upsurge in urban anthropology interest induced many researchers to look into the direction of social network analysis as a more appropriate methodological approach. Indeed, as Hannerz (1980) pointed out: "urbanites especially . typically do not draw their sustenance directly out of the earth, but to a great extent from their dealings with one another" (57). 6 While Bott (1957) carried out research on social networks in the 1950s, she did not as such belong to a school of thought on social network analysis. 7 See Sanjek 1990 for a literature overview.

PAGE 24

8 And yet, while a 1988 survey of social network researchers found that prior to 1970, 60% of the scholars most influential to the "social network community" were anthropologists (Johnson 1994: 142), it appeared that by 1988, only 20% of the top positions within the field were anthropologists, the majority of whom specialized in cognitive anthropology. The key to the contribution of anthropologists to the development of the concept of social network appears to have been the general anthropological interest in social structure in the 1950s and 1960s. As two major contemporary proponents of social network analysis at the time put it: . the study of social networks has arisen out of the dissatisfaction of a number of social anthropologists with the conventional form of structural-functional analysis. Structural-functional analysis by definition views societies as essentially static [and thus] is inadequate to cope with the study of rapidly changing complex societies. (Boussevain and Mitchell 1973: vii) It has been suggested that when the techniques of social network analysis began to lead a life of their own, not necessarily linked to a particular social theory (Scott 2000: 7-37),8 more and more anthropologists began to see scientific rigor as a hindrance, impractical and unnecessary for anthropological research on social networks (Walsh and Simonelli 1986). Notwithstanding the alleged historical unpopularity of social network analysis among the majority of anthropological researchers, I believe that this method is extremely appropriate for studying groups of people faced with tremendous changes, because it allows for establishing the linkages between micro and macro level events. One anthropological method, occasionally called situational analysis, is defined as the "intellectual isolation of a set of events from the wider social context in which they occur 8 A development largely due to the availability of computer systems and software packages able to digest ever increasing amounts of data.

PAGE 25

9 in order to facilitate a logically coherent analysis of these events" (Mitchell 1987:7). This methodology was first applied by Gluckman in his account of the opening ceremony for a bridge in Zululand, showing how elements of the larger social order are at work in individual situations and need to be incorporated in the final analysis (Rogers and Vertovec 1995: 5-11; Marx 1990:192). Applying this technique to social networks allows uncovering the ways in which the composition of an individual's personal network varies according to the larger world this person belongs to. The key focus of this anthropological dissertation project is to contextualize the daily actions (agency) and interactions (social networks) of the Congolese, Burundese and Rwandese urban refugees (men and women, young and old) in Dar es Salaam by situating them into the larger framework of their respective historical, socio-cultural, political and economic backgrounds (structure). Leaving the static principles at the basis of structural-functionalism far behind it, this approach intends to contribute to the anthropological understandings of the dynamic complexities of contemporary human existence. 1.1.1.3 Forced migration and social networks An important, if not the most important, link in today's world, between forced migration and social networks is the issue of globalization. The current era of continuing improvements in transport and communication has seen the migration of ever increasing numbers of people. Castles estimated that in 1992 over 100 million people worldwide lived in a country other than that of their birth, and that some 20 million of those, or 20%, were refugees (1998:179). Academic, and more particularly, social science theorists could not have been left untouched by this reconfiguration of social life (Amit-Talai 1997), and therefore a growing number of authors are reformulating the classical

PAGE 26

10 anthropological (and sociological) perception of "community as a place" to one that sees communities as social networks (Wellman 1999: xiv). At the same time, within the field of international migration, the realist perspective sees refugees as fundamentally distinct from immigrants, while the nominalist perspective ponders "whether the category is a social construction masking similarities with immigrations" (Hein 1993:43). Whereas in earlier research on refugees, the starting point was a simple dichotomy with immigrants constituting an economic form of migration and refugees a political form, it has been found more recently that "refugees share some migratory experiences with immigrants, particularly prior [to] migration and [concerning their] reliance on social networks" (46, emphasis added).9 Despite theoretical recognition of the close link between forced migration and social networks, empirical research, in terms of the collection of actual social network data, seems to have lagged behind. The existence of social networks and their importance to the survival strategies of forced migrants have been mentioned by only a few researchers who have done major research projects on urban refugees in Tanzania. Malkki (1995a), for one, remarks that the Burundi urban refugees in Kigoma relied on networks of their own making (46), however, she did not undertake any systematic analysis of these networks. Sommers (1994) noticed how the group of Burundi refugees in Dar es Salaam he studied were able to obtain employment through networks, yet again, no further analysis was ventured. Outside of Tanzania, there have been only a few small-scale projects looking into the role of social networks in a refugee situation (e.g., Koser 9 In their last State of the World's Refugees UNHCR concedes that "while the categories are neatly distinguished on paper, in reality the boundaries between them are far from clear and the connections between them many" (UNHCR 2000b: 280).

PAGE 27

11 1997; Williams 1993). Williams found, in her small-scale study among Angolan self-settled refugees in rural Zambia, that the refugees formed social relationships with non-kin individuals, Zambian or Angolan, to a much larger extent than was previously assumed. In addition, she discovered that those social networks were the channels par excellence through which the refugees were able to rebuild their livelihoods in a new and unfamiliar environment. At a more general level, Van Duijn et al. (1999) draw attention to the fact that "the idea that people invest in one another is also one of the basic assumptions in social capital theory . It is assumed that people invest in each other to gain future access to different resources" (191). Likewise, Wellman (1999) argues that in situations of uncertainty, such as living as clandestine refugees in a foreign country, social networks seem "to be especially important" (xviii). Notwithstanding, the existing body of research on urban refugees in Africa (e.g., Kibreab 1996; Kpanou 1989; Weaver 1988; Goitom 1987; Karadawi 1987; Kuhlman 1990), while recognizing that the main concern of urban refugees is to find (informal) paid employment, pays little attention to the mechanisms through which the refugee actually seeks and finds financial and other support. My study intends to illustrate that the Congolese, Burundese and Rwandese urban refugees in Dar es Salaam depend on their social networks to either find gainful local employment in the informal sector as hair-dressers, tailors, daily laborers, etc. or receive financial support from friends and relatives, whether from the other side of town or from the other side of the world. The reliance of these urban refugees on their personal relationships for survival in the absence of any institutional structures providing assistance may be constructed through Hyden's concept of the "economy of affection."

PAGE 28

12 It must be said from the outset that the economy of affection has nothing to do with fond emotions per se. Rather, it denotes a network of support, communications and interaction among structurally defined groups connected by blood, kin, community or other affinities . To be sure, the economy of affection is most prevalent in the rural community but it is an integral part of the society at large. Its influence stretches right from the grass-roots to the apex of society. (1983:8-9) According to Hyden, the economy of affection may be considered an alternative economy that fulfills an "extremely important welfare function" particularly in African countries that have experienced a breakdown of the state (10-11), but even more so in refugee situations. "In no other situation has the economy of affection proved itself more important as a mechanism for basic survival than where endless flows of refugees have suddenly overwhelmed a country" (12). Even as Dar es Salaam has not been overwhelmed by endless flows of Congolese, Burundese and Rwandese refugees, it is certainly true that in the absence of any assistance from institutional structures (whether material assistance or residence permits), the economy of affection as a support network is of the utmost importance to the refugees' survival. The following chapters will illustrate, however, that Dar es Salaam's urban refugees rely not only on their social network members from before the flight, but that newly established ties in exile are equally essential with regard to both material assistance and emotional support. 1.1.2 Refugee Studies and Gender Studies Very often in the refugee studies literature, the emphasis is on "the" refugee and "the refugee experience" (Ager 1999), and too little attention goes to the refugee as a cultural, gendered person with his/her own particular historical, political, economic and social background. With regard to the gender studies literature, it is not so long ago that one researcher, while conceding that gender is an organizing principle in the social, economic, political structures and the historiography of society, underlined that, at the

PAGE 29

13 same time, the cultural construction of gender is embedded in its social, economic, political and historic contexts (di Leonardo 1991). At the same time, Malkki challenged the assumption, recurrent in the refugee literature, that "to become uprooted and removed from a national community is automatically to lose one's identity, traditions and culture" (1995b: 508). The term "de-essentializing the refugee-experience" may be interpreted in more than one way. One interpretation is based--in line with Malkki--on the assumption that different refugee populations will experience what it is to be a refugee in different ways because of their respective socio-cultural and historical backgrounds. The second interpretation of the term relates to the assumption that the experiences that refugees go through are not exclusive to forced migrants that crossed international borders, but are shared by persons who were uprooted and displaced from their homes for a variety of possible reasons (e.g., economic, political or related to development), and who are often referred to as internally displaced persons [IDPs]. This research project, however, concentrated exclusively on the first interpretation, namely, the various ways in which different refugee populations experience life in exile. 1.1.2.1 De-essentializing the "refugee-experience" In the decades following the second world war, refugees became the focus for the development of a vast and complex network of institutionalized assistance composed of host governments, UNHCR and other UN organizations, and also non-governmental organizations which were assigned or assumed responsibility to deal with their material needs (Harrell-Bond and Voutira 1992; Zolberg et al. 1989; Adelman and Sorenson 1994). In the wake of this development, a wide range of historically generated Western conceptions about refugees, rights, nations and international aid has been unconsciously

PAGE 30

14 incorporated into large parts of the current domain of refugee studies (Malkki 1995b). Indra (1998), rightly, points out that: Researchers often unproblematically used highly stereotypical social issues and bureaucratically generated representations of "the" culture, experiences, and goals of refugees that would never have been acceptable in contemporary ethnographic treatments. . (xii) Consequently, the refugee was given an as it were "bureaucratic identity" (Zetter 1991) and a "visual image" accompanied by a peculiar kind of speechlessness in the face of the national and international organizations whose object of care and control they are. Their accounts are disqualified almost a priori, while the languages of refugee relief, policy science, and "development" claim the production of authoritative narratives about the refugees. (Malkki 1996: 386) This passive image of "the" refugee was compounded by the discovery of the so-called dependency syndrome among refugees due to prolonged residence in camps, where the bulk of studies on refugees takes place (Harrell-Bond 1986; Van Damme 1995). The very configuration of the international refugee aid regime in the form of refugee camps--especially those where access and exit are restricted by the authorities of the host country--requires the refugees to be not only completely dependent on the camp management for things such as food and medical assistance, but in addition, local housing arrangements and configurations are routinely disregarded by international agencies for the sake of efficiency (Cuny 1977; Ellis and Barakat 1996). Comparative studies into the living conditions of self-settled refugees outside of refugee camps, on the other hand, have shown them to be materially less well off than camp refugees, but much stronger and better healed at the psychological level (Harrell-Bond 1986: Hansen 1979, 1981). Studies among Mayan Guatemalan refugees in Mexico in camps (Light 1992) and urban areas (Bottinelli et al.: 1990) showed that, regardless of

PAGE 31

15 gender, the coping process of men and women refugees is accelerated by the opportunities to regain independent livelihoods, or to feel in charge of their own lives allowing for a renewed self-esteem and a view on the future. Clearly, the specific conditions in which refugees are to cope with their situation of forced migration not only influence the process of adaptation and the extent to which they are able to exercise agency, but play a determinative role in the multiple refugee experiences. Even as the majority of research studies among refugees has taken place among camp refugees, it is estimated that in Africa they make up only 20 to 25% of the total refugee population (Kuhlman 1994). Refugee camp populations nevertheless continue to attract the most attention not only from the international aid regime, but also from the news media and academics (Kuhlman 1994; Malkki 1996; Zetter 1991). Harrell-Bond (1986) and Hansen (1979, 1981, 1992) were among the first to compare the situation of camp refugees with rural self-settled refugees (Chambers 1979), but the debate on whether spontaneous self-settlement or organized camps is the better option for refugees rages on (Kibreab 1991; Black 1998a, 1998b; Crisp and Jacobson 1998; Harrell-Bond 1998). Whereas the humanitarian perspective has pointed out the psychological consequences of living in refugee camps for prolonged periods of time (i.e., dependency syndrome, see earlier), at the political level, there is an increasing tendency among the authorities of refugee-receiving countries to invoke reasons of national security to confine the refugees to a restricted or "designated area" (Kibreab 1991). This was also the case in Tanzania, where until the mid-1990s, the majority of refugees were allocated plots of land in settlements and encouraged to become self-sufficient while a substantial minority were issued with work permits, allowing them to reside in one of the country's urban centers. The influx of one million Rwandese refugees in the wake of the 1994

PAGE 32

16 genocide, together with armed ex-military and other militant groups had a drastic effect on Tanzania's refugee policy, and from then on, all refugees were required to reside in "designated areas," a euphemism for the refugee camps in the Western provinces of the country. The afore-mentioned studies reported that Africa's rural self-settled refugees usually join their co-ethnics living just across the border in order to cope with the absence of assistance from the host government or international agencies (Hansen 1992, 1979; Nelder 1979; Leach 1992). Increasing urbanization rates across the African continent, however, have changed the rural/urban ratio of its populations, hence of its refugee populations. In the mid-1980s, one researcher estimated that probably close to 30% of Africa's refugee population are urban, and are likely to choose to reside in urban centers while in exile (Rogge 1986). This research project focuses specifically on African urban self-settled refugees and their coping strategies in the absence of the possibility of joining co-ethnics or of receiving humanitarian assistance. Whereas assisted urban refugee populations have received some attention in the literature,10 the increasing numbers of self-settled urban refugees have been largely neglected in academic research (Kibreab 1996). In addition to Kibreab (1996), there is Sommers' study among young male Burundi refugees in Dar es Salaam done in the early 1990s (1993; 1994; 2001). It should be noted however that his research project concentrated on only a particular segment of the Dar es Salaam refugee population, namely second generation Burundi refugees--children from the 1972 Burundi refugee 10 E.g., Ethiopians and Eritreans in Khartoum who received assistance until 1986 (Rogge 1986; Goitom 1987; Karadawi 1987; Weaver 1988), and Salvadorans in Costa Rica (Basok 1988, 1993; BIT and UNHCR 1987).

PAGE 33

17 influx in Tanzania--who grew up in one of the rural settlement in Western Tanzania and migrated to an urban setting as young, mostly male, adults. They share with Tanzanian rural young men, who equally left their villages trying to "find life" (kutafuta maisha) in the big city, a large part of their experiences. From the above, it appears that a large part of the refugee literature of the past decades has been dominated by the stereotyped image of a refugee as a passive recipient of humanitarian assistance in a rural refugee camp. While some researchers studied self-settled rural refugees or assisted urban refugee populations, a lacuna exists in the area of urban, self-settled refugees. And yet this is a population which has known a tremendous growth in numbers over the past decade due to increasing rates of urbanization11 as well as the ever restrictive policies of host governments. This increase in numbers of urban refugee populations worldwide had equally led to a reconfiguration of their composition, traditionally typified as young, single and male (UNHCR 2001c). Taking into account variables such as gender, nationality and age will demonstrate the variety of the refugees' experiences and ways of coping in exile, and de-construct the essentialist definition of "the" refugee and "the" refugee experience. 1.1.2.2 Gender and forced migration The anthropology of women, emerging in the 1960s and 1970s, intended to provide a counterbalance for the male bias, prevalent in social studies at the time, by studying women (e.g., Amadiume 1987; Hafkin and Bay 1974; Stoler 1989, 1995; Nederveen 1990), an approach which was later termed--somewhat derogatorily--the "add-women-and-stir-method" (Moore H. 1988:3). Many of these ethnographic studies were based on 11 In 1999, Burundi was reported to have a 8.9% urbanization rate compared to 14.5 % and 39.3% in Rwanda and the Congo respectively (UNDP 2000).

PAGE 34

18 the concept of the "universal presence of the patriarchal system" prevailing among feminists and anthropologists at the time. Even as colonialism and capitalist penetration were not considered solely responsible for inegalitarian gender relations in Africa (Leacock and Etienne 1980), studies confirmed that male-oriented Christian ideologies were often reflected in official colonial state policies (e.g., Hunt 1990; Mbilinyi 1990). The emphasis of the feminist anthropologists of the 1980s was on the sexual division of labor, the decision-making process within the household, and the deficiencies of the development strategies in terms of gender (Holmboe-Ottesen and Wandel 1991; Spring 1992; Jahan 1995). They failed, however, to explicitly take into account gender relations, as well as the experiences of gender by local men and women, and their agency. While exploring the ways in which economics, kinship and social and political organization are structured by gender (e.g., Oppong 1983; Amadiume 1987; Leach in Croll and Parkin 1992; Mikell 1997), they stopped short of asking how gender itself is constructed through culture. The anthropology of gender, emerging from the late 1980s onward, specifically concentrated on the study of gender identity and its social (West and Zimmerman 1991) and cultural construction (Moore H. 1988). It explores the relationship between discursive constructions of women and men and their political economic realities, whereby both women and men take advantages of the opportunities produced by social and material changes in their lives and endlessly renegotiate the boundaries of gender (e.g., Hodgson and McCurdy 2001). Up until recently, the focus on gender in the field of refugee studies followed the earlier reorientations within the field of gender studies rather closely. The substantial increase in research on women refugees focused mainly on a range of conventionally-defined women's issues, such as health and diet, basic protection issues, women's

PAGE 35

19 employment, education, child care, language training and sexual violence (e.g., Martin 1992; Martin and Copeland 1989; Ager, Ager and Long 1995; Daley 1991). The result was a growing awareness among refugee agencies of the gender specific problems women refugees are exposed to. As a result, UNHCR and other agencies started issuing and distributing explicit guidelines, policy documents and training materials to sensitize all refugee workers of the necessity to take into account the specific needs and vulnerabilities of women refugees (UNHCR 1991; 1990).12 At the theoretical level, however, women's issues continue to remain quite peripheral to the mainstream discourse of academic refugee-studies: "the work is on refugees who happen to be women rather than on women and men who happen to be refugees" (Indra 1989:224). Gendering the refugee experience should not involve additional research on women only, rather it means looking at how both men and women--be it in the refugee camps or among self-settled refugees--discursively (re)construct social realities and renegotiate gender roles and boundaries in the process of coping with forced migration. 1.1.2.3 Engendering forced migration The process of coping with loss resulting from forced migration will in the first instance give rise to an increased importance attached to ethnicity, and traditional (usually patriarchal) cultural values and familiar rituals. The increased ethnic awareness of forced migrants--especially in refugee camps where the feeling of alienation is even more pressing--have been illustrated amply by researchers (e.g., Phillips 1996, Malkki 12 Contrary to earlier contentions, UNHCR currently confirms women refugee and girls to make up 50% of the refugee population dispelling previously held assumptions that over 80% of most refugee populations is female.

PAGE 36

20 1995a, Harrell-bond 1986, Camino and Krulfeld 1994). In contrast, Turshen and Twagiramariya (1998) have argued that war also destroys the traditional patriarchal structures of society that confine and sometimes degrade women, and that in the very breakdown of morals, traditions, customs and community, war also opens up and creates new beginnings. They illustrate their argument referring to specific case studies in Chad, Rwanda, Sudan and Mozambique where war, genocide and conflicts forced significant changes in gender roles. As a result of the increasing economic role of women, their power within the home increased significantly. The configuration of the personal social networks of refugee men and women respectively13 should prove to be an appropriate site to explore the agency exercised by members of both genders in the process of renegotiating socio-cultural boundaries, especially when compared to the pre-flight situation in the respective countries. In addition to exploring changing gender dynamics as a result of the situation of forced migration, I intend to look at the role played by nationality and age as well. The context for selecting nationality as well as age, in addition to gender, as the main parameters of the research project is presented in the following section. 1.1.3 The Social World of Refugees The legal nature of the definition of a refugee and the fact that s/he is considered an aberration in a world order of nation-states (Ager 1999:90) has given rise to an impressive body of literature with regard to the legal, political and human rights aspects of the matter (e.g., Campbell and Van Arsdale 1994; Hathaway 1991, 1995; Lee 1996; Warner and Hathway1992). Furthermore, the political, and often also ethnic, nature of 13 The unit of analysis of this study is the individual man or woman, not the household.

PAGE 37

21 the causes that lie at the source of many of today's refugee flows have triggered an avalanche of studies on how refugees politically and ethnically (re)organize themselves in exile, again mainly in refugee camps (e.g., Malkki 1996; Phillips 1996). Less attention has been paid in the literature, however, to the refugee as a social being uprooted from his or her "home" (Xenos 1993).14 Allowing for the fact that "home" is not just a geographic notion, but maybe more importantly so, a set of social relations, Marx defines a refugee as "a person whose social world has been disturbed" (1990:190). However, in order to better comprehend the process of rebuilding one's social world in exile, it is necessary first to conceptualize how people deal with loss. 1.1.3.1 Coping with loss Marris analyzes the process of coping with loss in terms of three elements: "conservatism, bereavement and innovation" (1986:2). The element of conservatism is based on the assumption that the impulse to defend the predictability of life is a fundamental and universal principle of human psychology. Predictability is related to control, and according to Marris "we cannot control the physical world without discovering its laws, but we can control our social world, more or less, by imposing laws upon it" (16). When confronted with a situation of a loss, people will look for a continuity of the meaning of life as they have known it so far, hence the deep-seated impulse to defend the validity of what they have learned, because without it they would feel lost. Because meanings are learned in the context of specific relationships and circumstances from early childhood onwards, the continuity of this context represents for an individual his/her identity (5-11). When there is a drastic change in an individual's 14 See also Portuous and Smith's Domicide and the Global Destruction of Home (2001).

PAGE 38

22 relationships or circumstances (e.g., the loss of a close relative, or the event of forced migration) this will affect the individual's sense of identity, in that the continuity between past and present is threatened, and life has become unpredictable. The second element of Marris' process of loss, bereavement, arises not from the loss of others (e.g., the death of a close relative, or the loss of one's socio-cultural environment), but from the loss of self, in that the relationships fundamental to one's identity have been abruptly altered and meaning now needs to be reformulated. As the meaning of life is defined by the particular experiences of each individual, those experiences need to be treated with respect, and so where changes imply loss, this loss must be grieved for. Grief . is the expression of a profound conflict between contradictory impulses to consolidate all that is still valuable and important in the past, and preserve it from loss; and at the same time, to re-establish a meaningful pattern of relationships, in which the loss is accepted. (31) The contradictory impulses within the process of grieving, which seek to reconnect past and future, ask for a predictable set of "mourning behavior," as formulated in mourning customs and rituals, in order to protect the mourner from the complexity of the issues involved (92). The necessity for a mourning ritual applies not only in situations where one has lost a close relative through death, but also when a drastic change in (living) circumstances occurs. It is not difficult to see in Marris' example of bereavement in a situation of slum clearance in Lagos, Nigeria (43-58), the analogy with a situation of forced migration "as in grieving for the dead, all the purposes and understanding inherent in those surroundings have to be retrieved and refashioned so that they still make sense of life elsewhere" (57).

PAGE 39

23 The conservative impulse in itself is not incompatible with innovation, since it "seeks to consolidate those skills and attachments, whose secure possession provides the assurance to master something new" (19). Marris says that "we can readily accommodate to changes, so long as they can first be assimilated to our existing patterns of thought and attachment" (104). People organize their lives as a function of hopes and goals that are in line with conventional expectations, and when they realize that these goals have become unattainable in the conventionally prescribed set of actions, frustration arises. Only in this situation of frustration, will people innovate and initiate changes which might affect the predictability of their lives. And so we may expect Congolese, Burundese and Rwandese refugees in Dar es Salaam to adapt to the new life and rebuild their social worlds following existing patterns of behavior up to the point where these behaviors no longer generate the expected outcomes and changes have to be accommodated to. Having conceptualized the process of coping with loss as a result of forced migration, we need to de-construct the concept of social worlds and establish a methodological tool with which to conduct the research. 1.1.3.2 Social networks The refugees' social worlds may be defined as "the sum of all the [refugees'] relationships and of the forces impinging on them at any moment" and "can be explored through the social networks [of the refugees] and the changes they undergo" (Marx 1990:189) during the process of forced migration and adaptation to life in exile. The idea of using social networks as a tool to analyze the dynamic forces that play at the macro-level is not new. One of the forerunners of social network analysis and member of the Rhodes-Livingstone Institute (see earlier) stated from the very beginning that the wider social context is reflected in the composition of the personal social network and that

PAGE 40

24 each personal network, therefore, will be unique though obviously influenced by such factors as the social position of the persons concerned and the social situation in which he is placed [and] the stage of the life-cycle he is at. (Mitchell 1969:43) As such, the composition of social networks will not only be affected by a situation of forced migration, but will also be influenced by the societal position of the individual in terms of his/her belonging to various types of social groups, e.g., men or women, different nationalities and age categories (see more in next section). Marx's "impinging forces" are structures or macro-elements of a historical, political, economical or socio-cultural nature related to the functioning, societal role and status of the (members of) respective social groups. These forces will affect the behavior, albeit the agency, displayed by individual refugees as members of the various social groups. It is in this respect, that I intend to disaggregate the social network data collected in order to uncover multiple patterns of coping with forced migration and rebuilding of social networks, thereby ultimately de-essentializing the refugee experience and embedding it in several layers of larger contexts at the macro-level. Approximately two decades into social network analysis, researchers cautioned that "it is important that network researchers consciously and clearly specify what they need to know about networks and what they mean theoretically by relation" (McAllister and Fischer 1983:87). The relationship (or relation) is the social process that ultimately links one with his/her social network members. Mitchell defined potential network members as that category of people who "in terms of the general norms or values of the community might be expected to provide ego with some type of service or support" (1969:43), and for the potential relationship to become a link in the personal network "some social exchange or transaction which converts the possible into an actual social linkage" (Ibid.) needs to take place. Applying this principle to the topic of this research project, I

PAGE 41

25 postulate that a relationship between an urban refugees in Dar es Salaam and one of his/her social network members consists of the actual providing of some type of support in the wake of the situation of forced migration. Support may include a variety of resources such as "emotional help, personal service, material assistance, financial aid, social brokerage, and empathetic understanding" (Wellman 1981:184). The same author found that While . sociability [does] not provide supportive resources directly, [its] existence affirms self worth and continued network membership. However, we are also finding non-supportive "structurally embedded" strands to be common. In such instances, two persons interact with one another only because the larger work or kinship structure requires it. (Ibid.: 183-4) In view of the fact that Wellman's research largely took place in stable and well-structured communities under ordinary circumstances, I propose that "structurally embedded" strands may not be as common in crisis situations, especially one of displacement. When one's survival is on the line, it is more likely than not that an individual will attach more importance to and invest more effort in those ties that he/she expects to be beneficial in the short term, whether in terms of receiving material or immaterial support. Thus, for the purpose of this study, social network members are considered those individuals that have provided the urban refugee in Dar es Salaam with some type of support from the moment of flight from home onward to the time of the research.15 The body of research on support networks of the past two decades--none of which took place in a situation of displacement or forced migration, and few outside the Western hemisphere--have uncovered a number of common patterns. A first finding was 15 The name elicitor was formulated conform this definition and will be elaborated on further in the methods section of this chapter.

PAGE 42

26 that strong ties or ties between individuals with common characteristics (also called homophilous ties) are more important conduits of social support than weak ties or those between individuals with dissimilar characteristics (also called heterophilous ties) (Wellman and Wortley 1990; Lin, Woelfel and Light 1985). Secondly, kin appeared as a primary source of support while residential proximity proved essential in support transactions involving material aid (Fischer 1982; Antonucci and Akiyama 1987; Wellman and Wortley 1990). Lastly, at the network structure level, it seemed that individuals who are embedded in dense, homogeneous networks receive more social support in emergency situations than do individuals in wide-ranging networks (Marsden 1988; Smith-Lovin and McPherson 1993; Beggs, Haines and Hurlbert 1996). My findings on the social networks of Dar es Salaam's urban refugees evaluate whether all of these patterns hold true in an East-African context of forced migration in an urban setting. 1.1.3.3 Gender, nationality and age As mentioned earlier, I decided to focus on three social parameters to disaggregate the collected data further, aiming to explore whether certain coping patterns, including social networking, are different among specific social groups: men versus women, Congolese versus Burundese versus Rwandese, and refugees belonging to different age groups. Gender The study of gender in an African context, as mentioned earlier in this chapter, has demonstrated the centrality of gender as an organizing principle of society as it is elsewhere. On the one hand, accounts of African women as the primary economic providers for their households in their roles as farmers and traders and descriptions of the complicated gender and generational rights and responsibilities for household production, consumption, and

PAGE 43

27 distribution have challenged prevailing Euro-American understandings of marriage and households as partnerships in which men are the main providers and household resources are pooled for consumption and distribution. (Hodgson and McCurdy 2001:5) On the other hand, it should also be noted that Christian male-oriented ideologies as well as the colonial administration had profound effects on urban populations, through, for example, the organizing of training sessions on "how to become the perfect housewife" (Hunt 1990). Karen Hansen suggests that it is the visibility of women's economic work in the fields and markets, together with the assumption that domesticity in Africa consisted mainly of fragments of Western-style idealizations, that made scholars neglect "both African reformulations and appropriations of elements of Western-derived domesticity, and their variously constituted indigenous counterparts" (1992:14). While fully subscribing to Hansen's contention that the domestic/public divide is not exclusively organized by gender, and "gender relations are not only or even primarily negotiated across a politico-jural/domestic divide" (17), I propose that local or popular perceptions to the contrary may play an influential role in the coping strategies of Dar es Salaam's refugees. Particularly with respect to the establishing of new ties in the process of rebuilding one's social networks, my findings suggest that popular perceptions of gender, of what it implies to be a man or a woman, constitute an essential element of life in exile. Nationality The second parameter I choose to define distinct social groups is nationality. In the literature, African identity has traditionally been closely linked to ethnicity rather than nationality. Nevertheless, from independence onward, integrationist policies aimed at overcoming ethnic differences and oppositions at the national level, such as e.g., Mobutu's retour l'autenticit during the 1970s, generally succeeded at creating a nationalist identity without excluding the ethnic one. Biaya describes this

PAGE 44

28 nationalist identity "like a garment that can be worn or discarded, depending on the circumstances" (2001:59). In a refugee situation, for example, individuals are likely to become more acutely aware of the important role their respective nationalities may play in their lives as exiles. The act of applying for asylum in a third country is usually, albeit not always, based solely on an individual's being a citizen of a particular nation, as is the evaluation of that application by the authorities of the asylum country. Manipulating one's nationality hence becomes a strategy of survival. When, for example, the Tanzanian government decided to forcibly repatriate the hundreds of thousands of Rwandese refugees from its territory in 1996, more than twelve thousand Rwandese refugees managed to leave the Rwandese refugee camps and register as Burundese refugees in camps established exclusively for Burundese refugees (Rutinwa 2002:Internet). They stayed on in the camps as Burundese even after the repatriation exercise, because from 1997 onward, individualized refugee status procedures became mandatory for Rwandese, while Burundese continued to be given prima facie refugee status. Only when it appeared that over 99% of these applications were successful, did the twelve thousand Rwandese refugees re-declare themselves Rwandese (Ibid.). As will be described in the following chapters, the urban self-settled refugees in Dar es Salaam apply the same survival strategy of "switching" nationalities as camp refugees when the need arises: Rwandese may pretend to be Congolese, Burundese pose as Tanzanians, etceteras. Apart from the obvious political implications, I propose that nationality equally plays a role in the ways in which refugees adapt to the situation of forced migration socio-culturally, on condition that the individual actually grew up and was socialized in

PAGE 45

29 the context of his or her home country.16 Each country with its specific historical, economic, political, legal, social and (multi)cultural conditions constitutes a unique macro-level framework in which the socialization of its nationals takes place. Without falling into a determinative, structuralist view of human life or neglecting the role of agency, certain national patterns of behavior geared towards survival are likely to emerge. Even as these patterns differentiate to a larger or lesser extent across socio-economic classes, gender, ethnic or political affiliations or religious congregations, there exist areas of common norms, values and behaviors often closely linked to perceptions of one's national identity. Therefore, disaggregating the collected data by nationality should bring to light existing commonalities among refugees from either Congolese, Burundese or Rwandese nationality. Age In the wake of Evans-Pritchard's research on the Nuer in the 1930s, numerous ethnographic studies on age in Africa took place in East-African, traditional societies and focused largely on age-sets17 (e.g., Spencer 1976; Foner and Kertzer 1978; Bonte 1985; Bernardi 1985). However, age plays a major role as well in African societies not structured by age class systems. In the introduction to his edited volume The Politics of Age and Gerontocracy in Africa (1998), Aguilar states that: the passing of age as a universal, human and biological process can only be fully understood by looking at its cultural--and therefore localized--processes of construction, perception, change and adaptation. Perceptions of age do not arise from a given definition but by sociality, from daily and habitual human activities . With Africa at large in mind, it is possible to argue that older people are given 16 This argument is operationalized in the selection criteria of respondents (see later in this chapter). 17 Whereas the term age-set is used to refer to a set of persons who were initiated together as youth, and who will go through the different stages of the age-system together, the term age-grade is used to refer to the respective stages of the life cycle (Spencer 1976)

PAGE 46

30 some kind of recognition, and that old age seems to be regarded as an attribute rather than a hindrance. (9) At the same time, the functionalist perspective of the majority of the studies on age class systems allowed little, if any, space to incorporate the influence of social change on the formal structures of age in these societies. The changing perceptions of age among growing numbers of individuals with different ethnic affiliations in Africa's urban centers are insufficiently documented in the literature. Increasing education levels, for example, among the younger (more urbanized) generations, both male and female, are known to challenge the traditional authority of the older generations in particular situations. And yet, even in the 1990s . mzee is [still] used in East Africa to show respect for older men [and women]. Elderhood, however, is not always associated with an active political role, but sometimes with the non-threatening role of advisor and councilor. (Aguilar 1998:22) The daily confrontations and contestations between young and old, whether men or women, are part and parcel of the ever changing social structures of current day African societies, particularly, but not exclusively, in the urban context. As with gender, the social roles that come with age are cultural constructs embedded in larger economic, political and historical structures. As such, persons of different generations are more likely than not to display different socio-culturally inspired responses in a particular situation. It is this logic that made me decide to add age as the third social parameter to this study of forced migration and social networks. Arriving in Dar es Salaam, Congolese, Burundese and Rwandese refugees have left behind a substantial part of their social worlds. Relatives and close friends from home are not always reachable through telephone or e-mail, even letters may not arrive at their destination depending on the severity of the situation back home. In addition, they

PAGE 47

31 experience how difficult it is to locate compatriots in a city of three million inhabitants "where the majority live in unplanned areas . [and] where most streets and houses are not named or numbered" (Kironde 2000:53). To complicate matters even further, there exists a lot of suspicion and distrust among nationals from the same country usually for political and/or ethnic reasons. And last but not least, for lack of legal documents, the urban refugees have to be wary of being arrested by the Tanzanian Immigration Police at any time. Reconstructing one's social world in these particularly challenging circumstances is a task tackled in multiple ways as we shall see in the following chapters. But first, I would like to present the arguments underlying my choice of methods. 1.2 Choice of Methods Before going from theoretical considerations about the topic in the various disciplinary fields to the actual geographical research site, important matters concerning the phenomenological and methodological options of the researcher need to be addressed. Traditionally, field experience was a conduit sine qua non for a piece of research to be considered anthropological, based on the perception that only participant observation allows the researcher to observe and write down in a scientifically objective manner the "other" culture or society waiting to be discovered (Gupta and Ferguson 1997). Increasingly, however, the inevitability of the subjectivity or "social, cultural and political location" (5) of both researcher and those being researched came to be recognized. Whereas nowadays "participant observation continues to be a major part of positioned anthropological methodologies, . it is ceasing to be fetishized" (17) and is usually complemented by a vast array of quantitative and qualitative methods. In addition, traditional anthropological fieldwork usually took place in village communities where participant observation was considered to be "the" data collecting

PAGE 48

32 tool par excellence. This study, on the other hand, takes place in an urban setting and among individuals who do not form "a community" in the traditional anthropological sense of the word. In order to hide their refugee status from the Tanzanian authorities--occasionally even hiding from their own countrymen for political reasons--the urban refugees in Dar es Salaam are forced to live a clandestine life.18 For the researcher, this means that relying mainly on participant observation by talking and living with the members of the refugee community is not really a viable option, and the necessity of additional data gathering tools ensues. 1.2.1 Collecting Data The goal of this research project, linking the adaptation processes, including the (re)configuration of forced migrants' social networks, to the larger worlds surrounding them, requires the collection of both quantitative and qualitative data. In view of the quantitative nature of the social network data, a myriad of qualitative collecting methods is called for in order to assure the correct interpretation and contextualization of the quantitative data. Within the field of social network analysis, there are two basic concepts of what constitutes a social network, namely socio-centric or "whole" networks, and ego-centric or personal networks (Johnson 1994: 117). Socio-centric networks represent the ties between all members, two by two, of a particular group of persons. This set of relations between a limited number of people is often visualized in a matrix, whereby all the members of the group are laid out on both the X and Y-axis. The number on the position 18 Notwithstanding, sharing similar experiences with regard to the flight from the home country and carving out a new life in an unfamiliar environment, they form what Malkki has called an accidental community of memory consisting of individuals with "accidental sharings of memory and transitory experience" (Malkki in Gupta and Ferguson 1997: 91).

PAGE 49

33 where a person on the X-axis and another person on the Y-axis meet usually indicates the intensity or another tie-attribute of the relationship between those two persons. An ego-centric network, on the other hand, is the set of relations one person--called ego--has with a certain number of other individuals--called alters. This type of network is often visualized as a star, whereby the length of the line between ego and one of his/her alters indicates the intensity or another attribute of a certain type of social relationship between these two persons. Whereas the bulk of the research into the statistical analysis of social network analysis concentrates on socio-centric networks (Freeman et al. 1989; Wasserman and Faust 1994), ego-centric networks as well received a substantial amount of attention with regard to the fine-tuning of data-collection and statistical analysis techniques (e.g., Straits 2000; Brewer and Webster 2000; Brewer 2000; Van Duijn et al. 1999; Louch 2000). The clandestine nature of the life of the Congolese, Burundese and Rwandese refugees not permitting research on socio-centric networks, this study therefore concentrates solely on ego-centric social networks. Traditionally, ego-centric or personal networks have proven to be excellent tools to engage in the study of social support in fields such as clinical and community psychology, as well as sociology and psychology (Wasserman and Faust 1994: 42; Wellman 1981; Lin, Woelfel and Light 1985, Smith-Lovin and McPherson 1993; Schweizer, Schness and Berzborn 1998; Wellman and Gulia 1999). Applications in the field of disaster studies have been far fewer (e.g., Beggs, Haines, and Hurlbert 1996; Haines, Hurlbert and Beggs 1996). However, it is essential at this point to draw attention to the difficulties raised by many social researchers of doing cross-cultural studies using methodologies that are based on purely Western concepts and instruments (Ahearn 2000). In an attempt to limit

PAGE 50

34 as much as possible the collection of inaccurate data by using culture-laden concepts, the questions pertaining to social networks asked in this study focused exclusively on: (1) the attributes of ego and his/her network members (e.g., gender, age, nationality, educational level, marital status, profession, etc.); (2) the type of tie (e.g., kinship, friendship, acquaintance, emotional support, financial support, etc.); and (3) the intensity of the tie (length of relationship, frequency of contact and place of residence). After an initial period in the field setting, I decided to reformulate the principal question eliciting the social network data, also called name generator, from the more generally used, abstract "Could you tell me who are the most important people in your life?" to the more practical and concrete: "Could you give me the names of ten persons who have helped you from the time you left your home and during your life in Dar es Salaam? They may include persons from any nationality, men and women, no matter where they live, and who helped you in any way, be it financially, emotionally, information-wise, socially, neighbourly, etceteras." In Swahili, the term hali na mali ("situation and riches") was used to indicate that the term "help" was to be interpreted in both its material as well as immaterial sense. This reformulation of the name generator was necessary because replying to its first version, respondents would start listing persons who had been important to them at one point in their life long before the current situation was forced upon them (e.g., during childhood), and with whom they had since lost contact. Considering that this is a study on "forced migration and social networks" it was primordially important to guide people to the current phase of their lives to elicit information on the persons playing an important role at that particular point in time. It should be mentioned at this point, that because of the sensitive and precarious situation of the undocumented urban refugees, I never insisted on collecting the real names of

PAGE 51

35 respondents or their social network members. However, on more than one ocassion, the informant, after initially introducing him or herself to me under a pseudonym, decided to reveal his/her real name once a relationship of trust had been established. "It has been said, rather facetiously, that qualitative research is what you do when you don't have enough data or don't understand statistics" (Ahearn 2000:15). Clearly, would it not be more precise to state that instead of focusing on the elusive goal of total objectivity, qualitative research methods implicitly recognize subjectivity through their reliance on other people's voices collected through narratives, interviews, life histories, focus groups and participant observation? In addition, certain types of data cannot be ethically or methodologically quantified, such as narratives. It is in this sense that quantitative and qualitative research methods should be considered complementary instead of each other's opposite. The premise on which qualitative research is based is "the value of the insider's view: to understand the world from their perspective, rather than categorizing their experiences out of context or from the outsiders' view" (Omidian 2000: 42).19 The qualitative data or "voices" collected in this study through interviews, (partial) life histories, media articles, and other published materials represent an avenue towards uncovering and understanding the multiple interpretations and meanings of refugee life in Dar es Salaam, and constitute the larger part of the contextualization of the data collected. 1.2.2 Collecting "Voices" The "Kuhnian disrobing of 'timeless' scientific authority" (di Leonardo 1991: 2) in the 1960s was one of the major elements of feminist criticisms of traditional social 19 What anthropologists have traditionally called the emic versus the etic view.

PAGE 52

36 science methodology. It was based on a "general rejection of positivism, its claim that science is value neutral, and [the false conviction] that the scientific method protects against contamination of findings by subjectivity" (Jayaratine and Stewart 1991: 89). Feminist researchers at the time blamed quantitative methods for concealing women's real experience, and subsequently became proponents of qualitative methods as methods which permit women to express their experience fully and in their own terms.20 The resulting body of literature on qualitative methodology was generally based on the actual experiences of the feminist researchers and on the theoretical and methodological reflections that ensued thereof (Stanley and Wise 1990). Even as it was feminist scholarship that gave the impetus to the development of this body of literature, its findings are extremely useful to other social science research as well. Feminism aimed to give a voice to women who were unheard in a traditionally white, male and Western world of research, yet at the same time, attention was drawn to the unheard voices of people--women and men--who were not white, not Western, or simply did not belong to the privileged layers of society. It is in this sense of giving voice to the unheard ones, that I consider it worthwhile to have a closer look at the feminist reflections on the use of qualitative methodologies. The most self-evident way of giving a voice to those unheard is through personal narratives, be it life histories, structured or unstructured interviews, etc. Personal narratives allow the expression of people's experiences and are, as such, a very popular research method among social scientists. They have produced a wealth of information on 20 Nowadays, there is a growing consensus among feminist researchers that both quantitative and qualitative research can result in erroneous and misleading results, and that "there can be no single, prescribed method or set of research methods consistent with feminist values" (Jayaratine and Stewart 1991: 100).

PAGE 53

37 human life and experience previously omitted or overlooked: "what could be truer [sic], after all, than a subject's own account of what he or she has lived through?" (Scott 1992: 24). Yet, it is exactly this reliance on "experience" as the undisputed evidence that Scott subsequently puts into question. She argues that experience itself is a subjective interpretation of events, and that the "project of making experience visible" obscures "the workings of the ideological system itself" (25). Social scientists therefore need to include into their analyses the discursive construction of differences and categories that subjects utilize to describe their experiences.21 Canning (1994), on the other hand, goes a step further and refuses to accept that discourses are completely determinative of people's experiences in the sense that "subjects do have some kind of agency, even if the meanings they make depend on . the discourses available to them" (377). Moreover, discourses not merely constitute the domain in which subjectivities emerge, they also supply the actual arguments for people to challenge their uncontested applicability and to endeavor changes (395-6). Applying the above reflections to this particular study of urban refugees in Dar es Salaam, it becomes clear that the refugees' personal narratives need to be contextualized through analyzing other texts as well, such as the discourses of refugee agencies, of the Tanzanian government and local authorities, of the media, and of Dar es Salaam's Tanzanian residents. Because of the international nature of the research setting and the politico-historical laden-ness of the refugee situation, time must also be spent on the political, economic and socio-cultural history of the region. In addition, there will be ample illustrations of Canning's argument, namely that the refugees in this study exercise 21 This emphasis on representation or discourse, as opposed to the pursuit of a discernible, retrievable historical "reality" is also known as the "linguistic turn"(di Leonardi 1992:22).

PAGE 54

38 agency through re-interpreting the existing discourses--of (inter)national agencies as well as other institutions--to their advantage in the pursuit of individual and personal goals. 1.2.3 Representativeness It is one thing to contextualize the narratives of the persons interviewed through discursive analysis (i.e., locating the positionality of the interviewee) however, it is equally important to take into account the positionality (role and setting) of the interviewer and the interpretation thereof by the interviewee (Gluck and Patai 1991). This point is beautifully illustrated in Knudsen's experience as a Canadian researcher among Vietnamese refugees in South East Asian refugee camps (1991; 1990; 1983). His study draws attention to the role of the investigator in the construction of the refugees' life history stories and of the transit refugee camp itself.22 He found that a refugee will "construe a life history in an attempt to reduce the uncertainty of his/her life course and simultaneously secure his/her rights towards the future" (1990: 123). These survival strategies led the refugees to construct life histories that forewent personal past experiences in order to fulfill the ascribed role of refugees, especially where it concerns refugees who will resettle permanently, and who thus tend to "perceive the present camp sojourn in itself as a liminal phase between the past and the future" (125). Eventually, to some camp dwellers, the story which was first meant to serve as a strategic construct, started to live its own life, the person no longer able to split the story s/he constructed from the reality invented. Knudsen's role as a Canadian investigator was perceived by the refugees as "related to" the Canadian immigration officials who were to decide on the refugees' individual immigration statuses. Ultimately, in the refugees' life histories past 22 Because the refugees residing in the transit camp were all in the process of being resettled or seeking resettlement to Western countries.

PAGE 55

39 and present were subconsciously negated in favor of the future due to the specificities of the refugee transit camp and the nationality of the investigator. Langness and Frank (1985) have pointed out that this phenomenon is not exclusive to refugees: ". . the very act of asking questions may make interviewees "recreate" themselves in the sense that, in looking for a coherence in their life histories, they often tend to reinterpret the past through the present" (107). In addition, the narrating of the story in itself can be empowering, validating the importance of the speaker's life experiences (Oliver-Smith 1986; Gluck and Patai 1991). For example, with regard to the traumatic experiences of the Mayan Indian widows, Zur (1997) found that the women resisted the official version of events by producing an alternative unofficial discourse: As they uncover memories, the women rework, relive, make sense of, come to terms with and integrate the traumatic events which dramatically changed their lives, thus giving a boundary to their suffering. Together they attempt to construct a narrative of the past which makes sense to them. (66) In recounting to each other the memories they had of the events that occurred, the women were primarily concerned with how this past related to their present lives, not with the accuracy of the events that actually had transpired. The same coping strategy was applied by Burundi refugees in the Tanzanian refugee camps when constructing their own "mythico-histories" about the events that forced them into exile (Malkki 1995a). Whereas the use of life histories, memories and oral histories will assist the researcher in being "more sensitive to situationality and variability" (Indra 1998: 19), informants' re-interpreting of the past in order to give meaning to the present, on the other hand, does raise the question of the "representativeness" of the narratives collected. The important questions at this point are: how does one determine what or who is "representative" and who decides which norm to apply? Or in other words: should

PAGE 56

40 researchers apply themselves "to produce general ethnographic descriptions of people's beliefs or actions" by "smoothing over contradictions, conflict of interests, doubts, and arguments, not to mention changing motivations and historical circumstances" (Abu-Lughod 1993: 9)? Susan Geiger (1992) cautions that a methodology "that features as major conceptual organizers the positional markers of marginality and representativeness automatically privileges certain voices and obstructs others through the very framework imposed" (309). In order to counter that tendency, she advocates the concept of "multiple truths" to be the leitmotiv for researchers intending to describe people's lives. Subscribing to Geiger's advice, I intend to give voice to a whole range of individuals, who are likely to have contradictory experiences and interpretations thereof, despite of all being termed "urban refugees." 1.3 In the Field With the aim to take into account all of the above considerations in order to be attentive to "the different forms of knowledge available from different social and political locations" (Gupta and Ferguson 1997:37), the methods applied in this study include--in addition to participant observation and ethnographic interviewing of the refugees--the following: administering a survey to a number of urban refugees in Dar es Salaam, administering another survey to a group of Tanzanian residents of Dar es Salaam, interviews with government officials and staff members of the international and national organizations concerned with the refugee population in Dar es Salaam, collecting relevant newspapers articles and analyzing government documents as well as other official and academic publications with regard to both Tanzania's refugee populations and the situations in the three major refugee producing countries, i.e., Congo, Rwanda and Burundi. In addition to determining which methodological tools to apply, the selection

PAGE 57

41 criteria for respondents needed to be thought out as well as strategies to locate informants and convince them to participate in the survey and share their experiences with an outsider. 1.3.1 Methods Used Collecting quantitative and qualitative data as mentioned above was done through a myriad of methods applied to different groups of people and through different media. In order to complement participant observation and ethnographic interviewing of Congolese, Burundese and Rwandese informants, I decided to concentrate on the following six data collecting activities. (1) A Refugee Survey was developed in French23 (see copy in appendix c) soliciting information about the events causing the respondent to flee his/her home country, the motivations for going to Dar es Salaam instead of heading for one of the refugee camps, and the first few days after arrival in Dar es Salaam. Included were also questions on the perceived main differences between his/her life at home and in Dar es Salaam and the motivation to (eventually) repatriate or remain in exile. Finally, detailed information is requested about ten social network members who helped the respondent from the time of the flight onward, and individuals' opinions regarding certain aspects of life in Dar es Salaam, such as religious activities, establishing relationships with Tanzanians, encounters with the Immigration police, etc. This survey was administered to 300 refugees in total: 100 of each nationality (Congolese, Burundese and Rwandese), equally divided between men and women. In addition, 30% of the respondents were between 20 23 The main European language used by nationals from the Democratic Republic of Congo, Rwanda and Burundi is French. In Congo, French is the language of tuition in schools from the 4th year in primary school onward. In Rwanda and Burundi French becomes the only language of tuition from the 1st year of secondary school onwards.

PAGE 58

42 and 25 years of age, 30% between 26 and 35, another 30% between 36 and 45 and 10% older than 45 years. The survey was administered by a research assistant in the mother tongue of the respondent. I strived to have a balanced sample as large as possible in order to increase the statistical significance of my findings. Three hundred was found to be the maximum numbers of respondents that could be located and surveyed in the time space of nine months.24 (2) A Tanzanian Survey was developed in English, later translated in Swahili (see copy in appendix B), and solicited information regarding the Perceptions on Refugees among the Tanzanian population in Dar es Salaam. It recorded the level of acquaintance the respondent had with refugees, the respondent's personal opinion on refugee policy vis--vis urban refugees and his/her perception of the professional activities of urban refugees according to gender and nationality. Equally noted were existing perceptions on the proportionate numbers of refugees in Tanzania and in Dar es Salaam and how the respondent would react to the marriage of his/her son or daughter with a refugee. Ultimately, the questionnaire solicited perceived positive and negative characteristics of refugees according to gender and nationality, as well as perceived differences between Tanzanian men and women on the one hand and refugee men and women of the three respective nationalities on the other hand. This survey was administered by an experienced Tanzanian social researcher contacted through the University of Dar es Salaam. A total of 216 respondents were interviewed, equally divided between men and women. Other parameters applied to have a balanced composition of the sample were the 24 See Bloch 1999 for an interesting overview of the methodological considerations of surveying refugees.

PAGE 59

43 socio-economic status of the place of residence, age, religious affiliation and, the level of education. (3) Most government officials interviewed belonged to the Department for Refugee Affairs at the Tanzanian Ministry of Home Affairs, a few to the Immigration Department in the same Ministry. The only international agency concerned with urban refugees in Dar es Salaam is the UNHCR. Many interviews were held with different officials in the Protection unit, Community Services unit and External Relations unit. Among the national agencies contacted were Umati, Tanganyika Council for Refugee Services, the Legal and Human Rights Office, the local branch of Amnesty International and the Center for the Study of Forced Migration at the University of Dar es Salaam. (4) National newspapers browsed were: The Guardian, The Daily Mail, and The African (dailies), and The East African, The Sunday Observer, and The Business Times (weeklies) over a period of approximately 3 years, from 1999 onward. (5) Government documents include the 1998 Refugee Act, the 1995 Immigration Act and others. Among the academic publications are a number of extremely interesting Master's theses from the University of Dar es Salaam in both Sociology and Political Science focusing on certain refugee matters relevant to Tanzania. One very important official publication was the preliminary policy review by five scholars, affiliated with the Center for the Study of Forced Migration at the University of Dar es Salaam, done at the request of the Prime Minister's office. Other were the UNHCR Newsletters, and publications by various other refugee agencies. Having determined and developed the above listed methods as the ones to be used in the field, the time has come to establish the selection criteria for respondents and to locate the informants.

PAGE 60

44 1.3.2 Selecting Respondents Before going into the ethnography of the refugees of Congolese, Burundese and Rwandese descent residing in Dar es Salaam, it seems advisable to first have a closer look at the label "urban refugee." From the literature there appears to exist an increasing diversity in the composition of urban refugee populations, as well as in their motives to head for the urban centers instead of the refugee camps. In one of the earliest references to urban refugees as a separate category of forced migrants, albeit one continuously increasing in numbers, Gould (1974) defines them as "educated, skilled, often politically active, and involved with freedom fighters of governments in exile" (416). Subsequent studies however, the majority of which took place among Eritrean and Ethiopian refugees in Sudan's capital, Khartoum in the 1980s and 1990s, showed that over time the proportion of non-urbanites became more important, and that their reasons for preferring to reside in Khartoum instead of the refugee camps covered a whole range of motives: search for education, rural to urban migration, ambitions to emigrate, etc. (Weaver 1988; Goitom 1987; Karadawi 1987; Kibreab 1996). In addition, increasing rates of urbanization on the African continent have had a determinative influence on the rural/urban ratio of Africa's refugee populations. Contrary to the situation in the 1970s, the current profile of the African urban refugee is far more diversified. The diversity in the composition of the refugee populations of Congolese, Burundese and Rwandese descent residing in Dar es Salaam, and their motives for going there in the first place instead of one of the refugee camps will be presented in one of the following chapters. Given the long history of refugee influxes from the Great Lakes region in addition to migration flows and inter-marriage patterns among neighbors, I estimated that defining the selection criteria of respondents deserved thorough

PAGE 61

45 consideration. After studying the existing situation through consultation with Congolese, Burundese and Rwandese friends "on the ground" as well as going through the existing, albeit scant, literature, I arrived at the following three general selection criteria. In view of the fact that this research project focuses on the reconstruction of individuals' personal social networks in the wake of their forced migration, the first criteria was that respondents should have a maximum of ten years, yet a minimum of one year of residence in Dar es Salaam.25 This selection criterion thus excludes those Congolese, Burundese and Rwandese who sought refuge in Dar es Salaam in the 1960s and 1970s, quite a few of whom are now Tanzanian nationals, particularly among the Rwandese. Secondly, interviewees should have recollections of life at home as an adult, i.e., older than 15 years of age at the time of leaving home, and these recollections should be no older than 10 years in order for the individuals to be able to compare social relations at home and in the country of asylum. This second criteria excludes many Burundi refugees who spent over 20 years, sometimes 30 years, in exile in (then) Zaire, and who came to Tanzania only because of the wars in 1996 and 1998 (see chapter 2). The third criteria pertains to the motive for leaving the home country, and relates to the principle of forced migration. In addition to persons deciding to leave home for political reasons or the dangers of war, I decided to also include individuals who came to Dar es Salaam for lack of economic opportunities at home as a result of the political situation in the Great Lakes region.26 Having established these three general selection criteria for 25 E.g., someone who spent 5 years in a Tanzanian refugee camp, and came to Dar es Salaam only 6 months ago does not respond to this criteria. 26 E.g., a Rwandese university lecturer lost his job of many years to a "returnee" from Uganda who was more closely related to the current regime in Kigali than he was.

PAGE 62

46 both respondents and informants, the next step was to further fine-tune the sampling method with regard to the demographic parameters of the population. Chambers (1979) estimated that in 1976 only 1.2% of African refugees were urban, the majority of whom unassisted (382). By 1986 however, Rogge speculates that "although there are still no specific data on actual numbers of urban refugees for the continent as a whole, the figure may now be as high as one third the total" (8). If this holds true, there might be as many as one hundred thousand refugees residing in Dar es Salaam. Notwithstanding the general consensus on the fact that urban refugee populations are steadily increasing in size, there remains a lot of conjecture on their demographic structure. Contrary to widely held assumptions that urban refugee populations are comprised predominantly of young, single males (Sommers 2001; Goitom 1987; UNHCR 1997c), recent statistical information shows that "this is not the case in many cities where substantial numbers of refugees are to be found" (UNHCR 2001c: Internet) as is the case in Dar es Salaam.27 Based on UNHCR's finding that larger urban refugee populations are more diversified demographically and in order to avoid a probable bias resulting from free snowball sampling in a clandestine population, I predetermined the following three demographic parameters.28 The first one was nationality: the aim was to locate one hundred respondents of each nationality, 27 Given that young males, for reasons of language and culture norms, are more open to contact with outsiders than women or older men, it is easy to see how this perception came about. The description of the sampling method used all too often missing in the write-up (Sommers 2001; Goitom 1987), the reader is left to speculate on the level of its randomness. This holds especially true for clandestine populations where snowball sampling appears the only indicated method of locating respondents. It is in this context that I decided that without some pre-established demographic parameters, the likelihood of a skewed sample (i.e., one that reflects the demographics of the initial contacts) would be all too high. 28 In addition to pre-setting the three demographic parameters, and being dependent on the snowball method to locate informants, I considered it counterproductive to an unbiased analysis of people's social networks to survey more than two individuals listed among one respondent's social network members, lest there should be too much repetition in the alters listed.

PAGE 63

47 Congolese, Burundese and Rwandese, totaling three hundred persons. Secondly, gender was to be equally represented in each group of one hundred respondents, i.e., fifty adult men and fifty adult women. Thirdly, four age categories were defined. The principle underlying the defining of the four age categories was based on marital and child status, because both marital and child status compliment the role of age in the societal status of an individual in an African context. Persons younger than 25 years of age are generally considered to be not yet married and childless, while persons between 26 and 35 years of age are much more likely to be married as well as have young children. The next generation, the 36 to 45 years olds, more frequently have larger families consisting of adolescent children generating pre-occupations of a different nature than the two younger age categories. The oldest age group in this study consists of persons over the age of 45, the majority of whom will likely have children of an adult age.29 Having established who I wanted to survey and interview, the next step was to locate informants. 1.3.3 Locating Informants One of the most frequently asked questions, when I explain to colleagues and friends the topic of my research and the fact that the Congolese, Burundese and Rwandese refugees are illegal residents in Dar es Salaam, is: "Well, if these people are in hiding, how are you able to contact them and why would they agree to talk to you?" As it is an extremely pertinent question in this particular research situation, I considered it worthwhile to elaborate at some length on my experiences in the process of locating informants and gaining their trust. 29 For the final number of respondents surveyed per nationality, gender and age, refer to Table 3-1 in chapter 3.

PAGE 64

48 "Trust" is an issue of the utmost relevance to the process of collecting data among refugee populations, and particularly so with regard to the overwhelming majority of the urban refugees in Dar es Salaam. The concept of "trust" in Daniel and Knudsen's Mistrusting Refugees (1995) raised a good deal of dust in the academic world of refugee researchers (Bisharat 1997; Schwartz 1997; Geuijen 1997; McGovern 1998; Voutira and Harrell-Bond in Daniel and Knudsen 1995: 207-224). While I agree that assuming "trust" as an a priori state of humanity does not really have a scientific basis, it is impossible to deny that becoming a refugee means having to deal with a huge increase in the lack of predictability of day-to-day life as it was known to them before the flight. Or as Bisharat puts it: ". . trust . is shattered when refugees are created by systematic violence and . mistrust . replaces it and stubbornly persists, even when the conditions that originally fostered the mistrust have abated" (1997: 664). The high level of "mistrust" existing in refugee camps and settlements, is even compounded in a situation where refugees opt to self-settle elsewhere, thereby contravening the explicit official policies of the country of asylum, as is the case for the refugees in Dar es Salaam. Outside of the officially designated areas in Western Tanzania, refugees in Tanzania run the permanent risk of being arrested and, as illegal immigrants, may even be deported to their home country by the Tanzanian Immigration officials. In such circumstances information becomes a powerful weapon, that, once released, may be turned against the refugee who lost all control over it. Suspicion runs high, especially vis--vis strangers, such as social researchers who go around asking all sorts of questions and spreading the word that they are particularly interested in talking to refugees.

PAGE 65

49 During the first months of predissertation research in the Spring of 2000, I was fortunate enough to meet with a person closely linked to the University of Florida, who at the same time happened to be married to a Congolese History professor teaching at the University of Dar es Salaam. Through her, I met my first Congolese assistant, who was not a refugee but a legal resident of Tanzania. My assistant and his family were well integrated and respected in the Congolese community in Dar es Salaam, which consists not only of refugees, but also of academics, businessmen, diplomats, etc. The ward of Kinondoni30 was by that time already locally known as the "place where all the Congolese live,"31 and it was here that I spent many hours during the predissertation phase of my research project talking mainly to Congolese men and women. Most of them had come to Dar es Salaam in the 1990s, some registering as refugees with UNHCR, others accepting being undocumented as an unavoidable fact of refugee life. Congolese informants were located using the snowball method, and contacts were fairly easily established due mainly to two reasons: the link to the parents of my assistant, who were well-established in the community and widely respected, and, secondly, I surmise, my Belgian nationality. Despite a violent colonial history and the, at times, ambiguous postcolonial relationship between the two countries, I gathered from contacts with Congolese informants that at the individual level there still exists a certain level of affinity towards Belgium and Belgian nationals. Speaking French was an unquestionable, even indispensable, asset, as many of the educated Congolese refugees expressed their joy at the opportunity of engaging in conversations and discussions in 30 Centrally situated between Manzese and Mwenge on the Dar es Salaam map 1-7 in appendix a. 31 In 2002, a popular Swahili song by a Tanzanian singer refers to Kinondoni as "Congo."

PAGE 66

50 French. At the same time, few, if any of the Congolese informants appeared at any time uncomfortable walking around with a white person, thus attracting attention in a popular neighborhood such as Kinondoni. I was able to meet with people in their homes, at their place of work, in public places such as local bars, or on the street. By the time I returned to Dar es Salaam in the Spring of 2001, my research proposal included refugees of Burundese and Rwandese descent as well for the purpose of cross-cultural comparisons. A Rwandese research assistant was recruited through contacts at the University of Dar es Salaam, a Burundese assistant through the French Cultural Center (Alliance Franaise) in Dar es Salaam. The Rwandese young man's father had sought refuge in Tanzania in 1959, indicative of his Tutsi background, and had since acquired Tanzanian nationality. The young man having grown up in Tanzania spoke only English and Swahili, and never having visited Rwanda, he mastered neither Kinyarwanda nor French. After numerous fruitless attempts to convince some Rwandese refugees of the 1990s to participate in the survey and eventually be interviewed by me, we agreed this line of action was not going to work. At the same time, the Burundese assistant had little success as well, claiming that the political tensions between Burundi and Tanzania at the time32 had frightened his Burundese refugee friends to the extent that they categorically refused to have any part of the research project, let alone meet with the researcher. Clearly, contacts with the Rwandese and Burundese refugee populations in Dar es Salaam were not going to develop as smoothly as with the Congolese. 32 In August and September 2001, the press had reported incursions of the Burundi army into Tanzanian territory claiming Tanzania was supporting the Burundi rebels. The reaction of president Mkapa was to forthwith announce the involuntary repatriation of all Burundese from Tanzania. This announcement caused a great deal of unrest and anxiety among Dar es Salaam's Burundese residents.

PAGE 67

51 Over time, continuing the meeting and interviewing of Congolese men and women, another Congolese research assistant was recruited, a young woman who at one time was a teacher at the Groupe Interscolaire des Grands Lacs33 through which she had established contacts with the Burundese and to a lesser extent the Rwandese community in Dar es Salaam. Having learned by now, that for a variety of reasons that will be elaborated on in the following chapters, it would be even more difficult to gain access to the Rwandese refugees than the Burundese, I decided to concentrate on the last ones first.34 After a few initial conversations and interviews with several Burundi young men35 through my assistant's contacts from the school, one of them was recruited with the sole purpose of locating Burundi respondents according to the selection criteria set out earlier. We devised a routine whereby initial contact was made by the Burundese assistant inquiring about the willingness of the person to participate in the survey, and if positive, subsequently setting up an appointment with the Congolese research assistant who by now was well trained in administering the survey and eliciting the social network data. 33 This is a school founded by Congolese, Burundese and Rwandese nationals in Dar es Salaam, where close to 300 children at primary and secondary school level are instructed the French/Belgian school curriculum in French by around twelve (often professional) teachers, often themselves refugees from the Great Lakes region. 34 Another possible avenue to the Burundi refugee community in Dar es Salaam through a local NGO Neighbors without Borders, founded by a Burundi mzee, did not work out either. The several persons I was referred to by the Director of this NGO, turned out to be political activists and members to one of the Burundi rebel groups, who wanted to check out whether I was really a social science researcher, but who were further unwilling (or unable) to introduce me to potential respondents and/or informants. 35 Young men are always the easiest point of entry into any African refugee community in my experience.

PAGE 68

52 The third step was to establish a place and a time, determined by the Burundese respondent, to meet with me for an interview.36 Usually, both the Burundese and Congolese research assistants accompanied me on the first meeting. After explaining to the informant at length (in either French or Swahili depending on his/her preference) who I was and why I wanted this meeting and ask all sorts of questions, I went over the social networks data, trying to slowly establish contact and gain trust. With a limited number of Burundese informants, I made arrangements for further conversations to which neither assistant would be present. Participant observation, or observation tout court was not really a viable research method in these circumstances. Most of the Burundese preferred to meet with me in public places, often in a ward different from the one they resided in, obviously in an attempt to not attract the attention of their Tanzanian neighbors by taking advantage of the anonymity a big city such as Dar es Salaam with its three million inhabitants has to offer. By the time I completed six months of research, in December 2001, I had succeeded in reaching the research goals established for the Congolese and Burundese urban refugees, but it was still unclear whether I would be able to accomplish the same among the Rwandese forced migrants.37 In the Summer of 2002, I returned to Dar es Salaam for the final three months of research focusing this time almost exclusively on the Rwandese refugee population. Fortunately, during my absence my Congolese assistant and two other Congolese young men conversant in Kinyarwanda, had persevered in trying to not just establish contacts, 36 In accordance with Spradley's definitions, a respondent is a person who participates in a survey, an informant agrees to ethnographic interviewing and more frequent personal contacts with the ethnographer. 37 A few months earlier, I had made another attempt at recruiting another Rwandese assistant. This young woman appeared very enthusiastic, but it turned out that suspicions run so high among the Rwandese--as I was to learn later--that she refused to give me even her mobile phone number, let alone any other indication as to how I would be able to contact her, thus impeding any fruitful collaboration.

PAGE 69

53 but more importantly, convince their Rwandese frres et soeurs ("brothers and sisters") that participating in the survey or agreeing to talk to me, would not put them in any danger of being discovered by the Tanzanian authorities.38 In view of the Rwandese reluctance to divulge any information on paper about their social networks for fear of the Tanzanian police and the International Crimes Tribunal for Rwanda (ICTR) in Arusha, I had to change the routine used so far. One of the Congolese middle persons would arrange for a meeting, at which my main goal was to convince the Rwandese of my "real" identity by pulling out not only a letter of introduction by the Department of Anthropology (that I had not had to use even once until then) but also my Gator-1 student card. The first Rwandese persons I met were, as usual, young men. Most would agree to meet with me on a second or third occasion for ethnographic interviewing, but in the absence of any of the Congolese assistants. When a relationship of trust was finally established, most would also agree to provide the social network data, especially when it was emphasized that no real names needed to be revealed. This new routine worked well for young men and women in their twenties and early thirties, but it appeared after a while that, unfortunately, despite numerous attempts older Rwandese would continue to categorically refuse to meet with me. They would however, accept to be surveyed on the social network data and a few other questions by the female Congolese assistant, because she was notre soeur ("our sister").39 Ultimately, I succeeded in collecting the Rwandese 38 The specificities of the pre-occupations of the Rwandese refugees in Dar es Salaam are complex and many, and will be elaborated on later. Basically, Rwandese refugees in Dar es Salaam, many of whom (linked to) former political figures, are afraid of being accused of participating in the master-minding of the 1994 genocide. 39 One was even interviewed and recorded on tape by a Congolese frre. This Rwandese refugee agreed to the interviewing and to the taping, but he was too afraid to meet with me personally, as there was the slightest chance that "as a white person" I should be connected to the UN International Crimes Tribunal for Rwanda in Arusha.

PAGE 70

54 social network data on Rwandese respondents of all ages, yet had frequent ethnographic interviews with men and women between 20 and 35 years of age only. Clearly, not every single Congolese or Burundese refugee contacted agreed to talk to me, yet, contrary to the Rwandese, the former appeared to base their decision on individual and personal motives rather than the structural or collective reasons influencing the decision of the latter. On the basis of feedback from my research assistants and a number of informants, I learned that whereas the Congolese perceived me to be the representative of an international NGO, sometimes even of UNHCR, looking for a project to assist the refugees, the Burundese usually saw me as a possible gateway to a third country, while the Rwandese would only grudgingly accept that I was not in any way linked to the International Crimes Tribunal for Rwanda in Arusha or a journalist bent on publishing a scoop at their expense. The process of locating informants for the purpose of this research process and the different types of problems that I had to confront, are in itself an argument in favor of one of the main theoretical stances this research project is built on, namely the de-essentializing of the refugee-experience. 1.4 Structure of the Dissertation I would like to conclude the first chapter by a brief outline of the remaining chapters of the dissertation. Having presented the theoretical and methodological concerns of this research project in the first chapter, the second chapter aims to contextualize the research topic and research site. The historical overview of the political events that took place in the three refugee producing countries under study, Congo, Burundi and Rwanda and that are at the root of the current refugee out-fluxes, aims at giving the reader a sense of where the refugees are coming from, and what they have had to endure before deciding to leave their homes behind. The section on Tanzania situates

PAGE 71

55 the changes in the refugee climate that have taken place in the country since independence, and that have had direct effects on the daily lives of its refugee populations, in a more global framework. The second part of chapter 2 describes the research site, Dar es Salaam, giving particular attention to the political and humanitarian infrastructure of potential relevance to the urban refugee population. In addition, a survey on the perceptions of Dar es Salaam's residents on the refugee population paints an interesting picture of the stereotypes and public images circulating among the refugees' hosts, as well as their attitudes towards the former. Having set the stage in chapters one and two, chapter 3 relates to the actual Congolese, Burundese and Rwandese refugee populations in Dar es Salaam. The findings of the survey administered to three hundred refugees give an overview of their demographic and socio-economic composition not only by nationality, but also by gender and age category. From all the informants interviewed, I selected twelve on the basis of gender, nationality and age, whose stories and accounts personalize and illustrate the findings of the survey throughout the following chapters. This chapter also presents the refugees' motives for fleeing their homes and for opting to go to Dar es Salaam instead of one of the refugee camps closer to the home country. Having found a safe haven, the next major challenge to the refugees is to adapt to the new life in an unfamiliar environment and to make ends meet in the absence of any material assistance from humanitarian agencies. Chapter 4 relates to difficulties experienced with regard to the use of language and the confrontations with Tanzanian residents' prejudices towards refugees. The lack of any official residence or work permit not only requires a clandestine existence in order to avoid arrest by the Tanzanian Immigration authorities, but it also complicates the quest for daily survival. And yet,

PAGE 72

56 refugees generally manage somehow to make ends meet, school their children, and at times, form grass roots associations for a variety of reasons. In the process, they meet with Tanzanians, fellow refugees and other Dar es Salaam residents, many of whom become supportive members of their social networks. The purpose of chapter 5 is to de-construct the refugees' personal social networks and to explore the extent to which gender, nationality and age play a role in the ways of meeting network members and the types of support provided. Through the use of concepts such as homophily, I attempt to uncover the variations that exist between women's and men's personal social networks, those of Congolese versus Burundese or Rwandese, and those of the different age groups. Throughout, I pay particular attention to the effect of the situation of forced migration on the ways refugees establish relationships with social network members. The analysis of the social network data is generally quantitative, the contextualization of its findings being the topic of the following chapter. Embedding the refugee experiences forms the main goal of chapter 6. The findings of the previous chapters, namely the variations across gender, nationality and age groups in the way of coping with forced migration and the reconstruction of personal social networks, are linked to the relevant macro frameworks. Going back to the premise that people react to change and are innovative as a function of their previously acquired behavioral repertoires and cognitive perceptions, a range of political, economic, historical, social and cultural arguments illustrate the embeddedness of the individual refugee experience according to the respective social groups the individual belongs to.

PAGE 73

57 The final chapter and conclusion delves into a number of policy implications related to the findings of the above research project, as well as methodological pre-occupations and theoretical considerations.

PAGE 74

CHAPTER 2 SITUATING THE DISSERTATION PROJECT AND RESEARCH SITE The previous chapter placed the research topic in the existing bodies of literature of cultural anthropology, refugee studies and gender studies with the aim of investigating the ways in which it may contribute to these fields of study. In addition, the choice of methods used was contextualized by perusing the work of a number of feminist scholars and locating points of similarities with the study of forced migration. However, before presenting the actual experiences of the urban refugees in Dar es Salaam, it will prove useful to provide a context to the research project, both in time and in place. The brief historical overview of the four countries involved, the Democratic Republic of Congo, Rwanda and Burundi as the refugee producing countries and Tanzania as the refugee receiving country situates the events at the root of the current refugee situation in the Great Lakes region and Tanzania. The second part of this chapter provides a short description of the actual research site, the city of Dar es Salaam, and presents the existing official, humanitarian and social structures in place that are of relevance to the urban refugees. 2.1 Background to the Research Project The structure of this section is unavoidably based on the concepts of the nation-state and its corollary "the national/citizen" as categorizing tools. One should, however, particularly in the context of this research project, be aware of the relatively short history 58

PAGE 75

59 of the nation-state1 and of its currently diminishing independence, whether economically, socially, culturally or politically, in today's era of globalization. In this context of closely interacting and localizing processes and forces, it is obvious that the distinction between domestic and international politics is meaningless . Even to the extent that the nation-state maintains its authority, it itself is becoming increasingly 'internationalized' or 'globalized' . (Collinson 1999: 5) An overview of the current debates on both these points being far beyond the scope of the dissertation project, it is nevertheless important to mention their relevance with regard to the current situation in the Great Lakes Region.2 In the following sections, I briefly present the historical interconnectedness of the three refugee producing countries, and after a short overview of the colonial and post-colonial period, continue to describe in brief the actual events that eventually led to "Africa's First Continental War" (Weiss 2000:1). Subsequently, the recent changes in Tanzania's refugee policy are situated in a more global context, followed by a short history of the refugee flows into the country and an overview of the current situation. 2.1.1 The Refugee Producing Countries The dramatic events of 19943 in Rwanda triggered an avalanche of publications (books, articles, web sites, etc.) on the situation in the region of the Great Lakes, written by political scientists, historians, journalists, aid workers, politicians and many others, all trying to make their point known to the international world. Jan Vansina, the most eminent historian on the Great Lakes region recently cautioned that: 1 Especially in Africa (e.g., Davidson 1992). 2 For a map of the Great Lakes Region of Africa, please refer to map 1-1 in appendix a. 3 The genocide of between five hundred to eight hundred thousand Tutsi and moderate Hutu.

PAGE 76

60 Indeed, most of the accessible sources (especially those generated by the media) are not reliable, in part because of individual bias or even incompetence, but mainly because the sources are all part of an intense propaganda war. This campaign is intended to generate support for one side or the other, whether it be to raise morale, to facilitate the procurements of arms, or to generate outside diplomatic cover. Furthermore, atrocities and the violent emotions they generate make it nearly impossible for any observer to retain a "neutral" stance or to avoid "demonizing" one side or the other. (1998:37) The implications for the historiography of the region are clear: Hence, it is not surprising that all sides in today's Central African crisis reinvent history. They use portions of the existing historiography written by respected authors . to derive novel interpretations of supposedly well established facts. Their historical arguments range not only over the last thirty years but back into the colonial period and as far back as the early Iron Age. (1998:38-9) Vansina's well-formulated considerations concerning the origins of Hutu and Tutsi are not new to the history of the Great Lakes region, but in the wake of the recent events they have taken on a whole new dimension. Even before the 1994 events, MacGaffey, reflecting on (the mythologization of) the history of the ethnic groups of the region,4 had concluded that "[t]he present, we discover, is mother to the past" (1993:15). 2.1.1.1 The historical interconnectedness of the Great Lakes countries Leaving the above considerations largely as reflections on the controversies surrounding the origins, political history and ethnic parameters of the Hutu and Tutsi populations in Rwanda and Burundi,5 the following overview of the region's history will concentrate on its historical interconnectedness. The Great Lakes Region consists of (parts of) five countries: Rwanda, Burundi, Eastern Congo, South Western Uganda and North Western Tanzania (map 1-1 in appendix a). There are four major lakes in the area: 4 One well-illustrated example of this phenomenon are the mythico-histories of the Burundi Hutu refugees in the Tanzanian refugee camps (Malkki 1995a). 5 Giving a complete overview of the controversies surrounding the origins of the Hutu and Tutsi populations in Rwanda and Burundi falls beyond the scope of this dissertation project and the reader is referred to the publications of experts on the matter: e.g., Lemarchand, Vansina, Reyntjens, Chrtien, etc.

PAGE 77

61 Lake Victoria, Lake Edward, Lake Kivu and Lake Tanganyika, hence its name. Its geography is dominated by the Great Rift Valley in which the huge lakes are located and most parts of the area lie at an altitude of between one and two thousand meters, combining a pleasant, moderate climate and lush green, fertile hills (Chrtien 2000: 11-12). Linguistically, the myriad of languages in the region can be divided into four or five families of, often to a lesser or larger extent, mutually intelligible vernaculars (Chrtien 2000:34; Mworoha 1987:188). Recent archeological and linguistical research has established the presence of cattle throughout the region from more than 2,000 years ago6 accompanied by a high level of interaction between its ancient Early Iron Age societies (Schoenbrun 1998).7 Over time, a variety of political systems saw the light in the region, but notions of nobility, kingship and courtly life took firm ground by the beginning of the second millennium only, with "the invention of concepts of hereditary nobility and the link that was created between nobility and the institutions of instrumental power over people" (185).8 By the beginning of the 19th century, many of the kingdoms in current-day Rwanda, Burundi, Uganda and Eastern Congo had developed into states with political systems based on centralized power. These states were not isolated entities, but 6 Contrary to the assertions of earlier historians that cattle--considered "the" hallmark of the Tutsi pastoralists--was introduced into the region by the "Hamitic invaders" of the 15th century (see more on this issue later in this chapter). 7 Linguistic research had shown earlier that the region was populated during the Bantu migrations from West Africa between 5,000 and 3,000 years ago. More information on Greenberg and Guthrie's linguistic theories and on how archeological finds eventually helped shape the Bantu migration theories may be found in e.g., Gailey 1970; Chrtien 2000; Schoenbrun 1998, Collins 1968. 8 Some oral history sources on the royal dynasties of the Great Lakes region contend to be able to go back as far as the 15th century. Scientific research however, was able to confirm the reliability of the data provided up until the beginning of the 18th century only (Chrtien 2000: 117-172). The names of the kings of no less than fifteen royal dynasties in the Great Lakes region between 1500 and 2000 are listed in the appendices of Chrtien's L'Afrique des Grands Lacs: Deux mille ans d'histoire (2000).

PAGE 78

62 constituted a communicating network of diplomatic relationships, alliances, dynastic marriages, commercial exchange and military activities throughout the region, incorporating not only present-day Rwanda, Burundi and Uganda, but also what is now Eastern Congo and Western Tanzania (Ajayi 1998:100). In addition, regional migrations took place as early as the 18th and 19th century, usually from the more populated kingdoms of Rwanda and Burundi into sparsely populated Western Tanzania (Roberts 1968: xi, 120; Daley 1993:19-20; Eger 1979:30). At the same time, Arab and Swahili traders from Zanzibar and Bagamoyo on the Tanzanian coast introduced firearms into the region in exchange for slaves and ivory, thereby ultimately opening the way for European explorers, as well as Muslim and Christian missionaries. The stage was set for the military incursions and political and economic impositions of the colonial powers that were to have a devastating effect on the societal landscape of the Great Lakes region. 2.1.1.2 The colonial impact in Congo, Rwanda and Burundi Congo was the private property of the Belgian king Leopold II from 18859 until 1919, when it was taken over by the Belgian government under mounting international outrage over the king's exploitation of the Congo Free State.10 Rwanda and Burundi, on the other hand, were first colonized by Germany together with Tanganyika, as the mainland of Tanzania was called. After World War I, Germany was punished and its colonies redistributed: Tanganyika to Great Britain and Rwanda and Burundi to Belgium. Whereas in the Congo in general, the Belgians applied a system of direct rule by removing the legitimate chiefs and appointing the "chiefs of the whites" (Lemarchand 9 This is the year the Berlin Conference took place, where the African continent was divided between the European powers. 10 See Adam Hochschild's King Leopold's Ghost (1998) for a fascinating account of this time period.

PAGE 79

63 1970: 66), in Ruanda-Urundi (as it was then called) as well as in the Kivu region in Eastern Congo, they applied the principle of indirect rule as had the Germans before them. The decision "to follow the traditional lines of authority rather than to create anew an administrative structure" as they had in the rest of the Congo, was based on the existence of "micro-kingdoms which tended to hold up rather well against the administrative, colonial onslaught" in both the Kivu region and Ruanda-Urundi (Prunier 2000:146). In addition to political impositions, the colonizers introduced a system of ethnic categorization in all three countries. Making rigid categories out of ethnic groups that before had somewhat fluid boundaries11 had much less far-reaching effects in the Congolese ethnic mosaic than it did in the two small kingdoms of Rwanda and Burundi. These were considered by the colonizers to be inhabited by the same three ethnic groups: Hutu, Tutsi and Twa. Two of these, the Hutu and Tutsi, were categorized, not by language, religion or culture (which they shared), but by perceived differences in status and physical characteristics. The Tutsi were seen as the ruling minority, the Hutu majority as the agricultural vassals of the rules, and the few Twa as the remaining hunter-gatherers from the rain forest. Each group was categorized on the basis of certain physical traits considered markedly different from the others', and scientists of the time went as far as to put together tables with height, weight, nasal and facial index per group (Codere 1993: 96). These ethnic labels eventually became the basis of the German and the Belgian colonial administration and "the listing of ethnic labels on identity cards 11 In his Hutu and Tutsi: Primordial Illusions (1994: 6-16) Lemarchand argues that "status, not ethnic identity, was the principal determinant of rank and privilege" (10), and that "depending on luck and individual ability, some Hutu elements could end up with considerable more influence, wealth and social recognition than the average Tutsi" (11).

PAGE 80

64 issued to all residents served to eliminate the previous flexibility in identity" (Longman 2001:169).12 Another colonial inheritance was the "Hamitic Hypothesis" that explained the "separate origins of the Tutsi," who were seen as "a civilizing Caucasian influence in Negro Africa" (Mamdani 2001:47). According to the Hamitic Hypothesis the Tutsi migrated from the Ethiopian Highlands around the 15th century and "conquered" the sedentarist Hutu farmers who had inhabited the region since the Early Iron Age Bantu migrations. The Tutsi further introduced pastoralism to the "uncivilized peasantry" and a royal state court (Ibid.).13 This myth was the basis for, and at the same time, explanation of the colonial perception of the Tutsi as "Nilotes" being superior to other African groups defined as "Bantu."14 The Church was instrumental in co-opting the Tutsi population into the colonial system of indirect rule by giving Tutsi children preference over Hutu children in the mission schools, the only avenue for Africans at the time to acquire a Western education and enter into the colonial administration (Gahama 2001:243-260; Rumiya 1992:159-168). Up until shortly before independence, the Church "posed as the 12 Umutesi, a Rwandese woman, recalls that in the few years preceding the 1994 genocide some children "who looked like Tutsi" were chased from school by fellow students and "were not to come back unless they had the identity papers (livret d'identit) delivered by the Germans to their fathers or grandfathers. That paper was the only document, according to the students, that could give authentic information as to the ethnic identity of the Rwandese." (2000: 12). In the same line, other observers remarked that the very fact that the Interahamwe relied on individuals' identity cards in order to decide whom to kill, proves that Hutu and Tutsi are physically indistinguishable. 13 For a comprehensive overview of the other theories circulating at one time or another regarding the origins of Hutu and Tutsi, see Mamdani 2001 (41-75). 14 Whether Hutu in Rwanda and Burundi or other ethnic groups in Eastern Congo. It is interesting to note that contemporary scholarship has (re)classified the Tutsi as a Bantu group (Nzongola-Ntalaja 2001). In fact, "[t]he terms 'Bantu' and 'Nilotic' refer exclusively to a linguistic distinction and it is only within a linguistic context that they are scientifically warranted. They cannot possibly be justified if extrapolation is made from this very context to other than linguistic features such as, for instance, putative historical, physiological or social attributes" (Goyvaerts 2000:302).

PAGE 81

65 strongest advocate of Tutsi supremacy, largely on grounds of political expediency" (Lemarchand 1970:73). In Congo, the Church equally had the monopoly in education through its mission schools and seminars providing virtually the only route for a Congolese to become an volu.15 These educated Congolese over time began to constitute a separate "class . different from the untutored mass" (Anstey 1970:202). When their growing political aspirations remained largely unmet by the Belgians, they eventually turned to revolt. By that time, however, there were only a handful university graduates and few Congolese had been trained to take over the administration in the national government offices. This level of unpreparedness and lack of intellectual capital at the time of independence was to set the Congolese stage for a long time to come. In Rwanda and Burundi, when the educated Tutsi began to rally for independence in the 1950s, the colonial administration suddenly decided to democratize the countries and give the political power to the Hutu majority. Electoral processes were introduced at a time when the Hutu educated elite in Rwanda and Burundi constituted a very tiny minority, and with little fundamental change in the traditional social structure, the stage was set for a turbulent era of independence (Lemarchand 1970:81-83). Although the two countries shared many cultural characteristics, ethnic relations in Burundi, for historical reasons,16 were more fluid and not reducible to a simple Hutu-Tutsi split, hence their 15 An volu (literal translation: developed or advanced person) was a term given to those Congolese who appeared to have given up the traditional African lifestyle and assimilated the lifestyle of the "whites": i.e., living in a nuclear family, in a house in town, and exercising a paid job requiring a certain level of Western education and fluent French. They were never to aspire, however, to be completely accepted as a member of the white community. Representative figures were the clerk, the male nurse, and the teacher (Anstey 1970:195). 16 See Lemarchand (1970) for a detailed analysis and comparison of the historical backgrounds to the political situations in the respective countries.

PAGE 82

66 "radically different political trajectories, with Rwanda acceding to independence as a Hutu-dominated republic and Burundi as a constitutional monarchy" (Lemarchand 1994:1). 2.1.1.3 From independence to the early 1990s In Congo (map 1-2 appendix a), the democratic election of Patrice Lumumba as the country's first Congolese prime minister raised quite a few eyebrows because of his socialist ideologies and stern condemnation of the era of Belgian colonization.17 Amidst the ethnic turmoil of the early 1960s and the secession attempts by some of the richer Congolese provinces, Lumumba was assassinated18 and a young army sergeant Joseph Mobutu was considered more trustworthy by the Western powers to rule the Central African country. Zaire (as Congo was renamed) enjoyed a relative political stability under Mobutu's thirty three years of rule. However, the lack of skilled and experienced Congolese cadres to manage the vast country had devastating effects on its economic and infrastructural development, especially from the mid 1970s onwards after Mobutu's nationalization exercise (retour l'autenticit). The crisis continued to deepen in the 1970s and 1980s (Nzongola-Ntalaja 1986), and while the state withdrew further and further to Mobutu's immediate entourage, the 17 This section has been intentionally kept to a minimum and is meant only to give the reader unfamiliar with the region some background information as to the events that took place in the three countries in the course of the 1990s, ultimately leading to the current situation in the Great Lakes region. 18 De Witte (1999) gives an elaborate account of the time era and is the first to document the direct involvement of the Belgian military and senior members of the Belgian governement in the murder of Patrice Lumumba. It appears that even the Secretary General of the United Nations at the time publicly expressed disdain for and failed to recognize the legitimacy of Congo's first democratically elected Prime Minister.

PAGE 83

67 Zairian people gained an international reputation as dbrouillards,19 engaging increasingly in urban "popular" economic activities, i.e., the informal sector (Willame 1997). Biaya (2001) describes the development of a parallel society where "various associations and local NGOs blossomed and set objectives to resume control of what the state had abandoned or was unable to control, such as local universities, regional development councils, and NGOs" (51). In Eastern Zaire, in particular, there was a proliferation of independent churches as part of citizens' response to the withdrawal of the state and its infrastructures. These churches were increasingly involved in development programs based on local initiatives, at times even setting up "parallel networks of schools, health and agricultural services, and courts" (Newbury in Nzongola-Ntalaja 1986: 105). They complimented the "enormously important role of the [Catholic] Church" in "a vacuum from the point of view of the basic social and administrative services which a normal government is supposed to provide" (Prunier 2001: 155). In Rwanda (map 1-3 appendix a), from 1959 onward, increasing Hutu political violence aimed to dismantle Tutsi power at the local administration level and ultimately triggered a transfer of governmental power from a Tutsi to a Hutu elite in 1962 (Mamdani 2001: 103-131). The change of regime went hand in hand with multiple massacres of thousands of Tutsi, producing refugee outfluxes of tens of thousands of Tutsi to Burundi, Uganda, Tanzania, and Zaire. Rising unemployment among increasing numbers of educated Hutu and the 1972 events in Burundi (see later) created ethnic 19 A popular joke of the time was that Article 15 of the Zairean constitution advised Congolese citizens to Dbrouillez-vous. Se dbrouiller means "to fend for oneself." A dbrouillard is a resourceful person coping by him/herself.

PAGE 84

68 tensions that led to the 1973 military coup of Habyarimana (a Hutu) inducing more Rwandese Tutsi to seek refuge abroad. Habyarimana was re-elected in 1983 as the president and stayed in power until his assassination in 1994. In Burundi (map 1-3 appendix a), before independence "caste cleavages were more numerous and fluid, and less consistent with ethnic differences" than was the case in Rwanda (Lemarchand 1970: 89). The effect of the 1959-1962 events in Rwanda, however, made the Tutsi monarchy cling to power yet more strongly--despite several Hutu victories in democratic elections--and eventually developed into a military dictatorship. In 1972, a Hutu uprising led to Tutsi massacres in Bujumbura and the Southern provinces. The response of the Tutsi regime was swift: over a hundred thousand Hutu were massacred, many of them belonging to the educated elite. The result was an outflow of hundreds of thousands of Hutu refugees to Rwanda, Tanzania, and Zaire. After a series of military coups following the 1972 genocide, Buyoya (a Tutsi) seized power in 1987 in an attempt to restore order. When another Hutu uprising involved more massacres of Tutsi in the North of the country in 1988, the devastating reply from the Tutsi dominated army resulted in twenty thousands persons killed and ever increasing Hutu refugee flows to the neighboring countries (Braeckman 1996:127-201). A number of researchers have pointed out that the competition for strategic minerals and land is the main cause of the war in Burundi rather than ethnic hatred (Balati 2001; Ngaruko and Nkurunziza 2000). Nyerere, the facilitator of the Arusha Peace Negotiations once said: "The strife in Burundi is cast in ethnic terms. It is not, it is really about resources" (Daily News June 9, 1997).20 20 More on this specific topic follows in chapter 6.

PAGE 85

69 2.1.1.4 From the early 1990s onward Lemarchand (1997) rightly proposes that "much of the Hutu-Tutsi confrontations in Rwanda and Burundi is indeed reducible to the polarization of group identities that has accompanied the movement of refugee populations from one state to the other"21 (Internet). Mamdani (2001), however, ventures out further than the Rwanda-Burundi or Hutu-Tutsi dichotomy, locating the very beginning of the Great Lakes region events of the 1990s in Uganda. It is estimated that by 1990, the number of Tutsi refugees in Uganda that had fled Rwanda between 1959 and 1964 numbered two hundred thousand, most of whom were living an uncomfortable and insecure life in the refugee camps. By the early 1980s, several thousand of the second generation refugees had joined the National Resistance Army's (NRA) guerrilla war against Uganda's president Obote, and assisted Museveni in eventually taking over power in 1986. In appreciation, some of these Rwandese guerrilla fighters (the most notorious of whom is undoubtedly Kagame22) acquired important posts in the Ugandan government. In 1990, a proposal from president Museveni to allow non-Ugandans with a ten year residency in Uganda to apply for Ugandan citizenship, was voted down in parliament in favor of a citizenship law that required proof of "indigeneity," i.e., having proof of forefathers residing in Uganda during colonial times. For the Rwandese Tutsi refugees, "the die had been cast" (182). Having no future prospects in Uganda, they decided to return home and demand their place in Rwanda. 21 I.e., Rwandese Tutsi refugees to Burundi and Burundese Hutu refugees to Rwanda. 22 Rwanda's current President.

PAGE 86

70 The Rwandese Patriotic Front (RPF) invaded Rwanda in 1990, reportedly with Ugandan assistance (164-183). In Burundi, in the meantime, Buyoya was pressurized by the international community into organizing democratic elections. While higher education systems and the administrative government offices had slowly been opened to Hutu participation,23 Buyoya remained firm in his refusal to accept Hutu officers in the national army. In June 1993, Melchior Ndadaye was elected as the first Hutu president of Burundi, promising "in French democracy, in Kirundi, the liberation of a people suppressed for too long" (Braeckman 1996:152 own translation). The majority of the provincial governors and local administrative officers were promptly replaced, fueling the feeling of increasing resentment and feelings of insecurity in Tutsi circles. In the wake of increasing pressure on the government to integrate former Hutu rebels into the Tutsi dominated national army, the democratically elected president Ndadaye was assassinated by this very army in October 1993. Civil war ensued involving the massacre of thousands of Tutsi civilians. Subsequent repression on the Hutu population by the army led to tens of thousands being killed, and several hundreds of thousands displaced or in exile. From 1993 onwards, Burundi has been prey to what observers have called a "gnocide compte-goutte" (literal translation: genocide drop-by-drop) because of the systematic usage of terror by Tutsi government forces to intimidate the Hutu population (Reyntjens 1999:37). By 1997, 5% of the population, or three hundred thousand persons, were interned in government camps, deemed "necessary" by the Burundese government 23 Upon until that time Burundese Hutu were systematically discriminated against in the educational system making it sheer impossible for them to participate in national, and to a lesser extent local, government structures.

PAGE 87

71 "in order to isolate the rebels" (Buyoya in Reyntjens 1997:190). By 2000, close to seven hundred thousand Burundese, more than 10% of the population, were internally displaced (map 1-5 in appendix a). While Hutu rebels continued to launch attacks throughout the country, the Tutsi government forces remained directly and indirectly involved in the killing of Hutu civilians, all parties truly "caught in the spiral of violence" (USRC 1995). In Rwanda, in the meantime, the RPF was slowly making its way towards the capital Kigali, while the international community was putting pressure on the Hutu government to share power with the Tutsi population. At the same time, Hutu refugee flows from the RPF controlled areas in the North were pouring into Kigali, and rumors abounded about atrocities allegedly committed on the Hutu population by the advancing RPF forces (Umutesi 2000:19-51). The events of October 1993 in Burundi and the pouring into Rwanda of several hundreds of thousand Burundese Hutu refugees contributed further to a general climate of fear (Mamdani 2001:191). When on April 6th, 1994, the plane carrying the Rwandese president Habyarimana was shot down by unknown assailants,24 the Rwandese 1994 genocide started. More than half a million Tutsi and moderate Hutu were killed by the Interahamwe militia and the Rwandese Armed Forces (FAR), albeit assisted by parts of the Rwandese civilian Hutu population later to be known as the gnocidaires.25 26 27 By the time the RPF reached Kigali in July 24 It has not been established beyond doubt until now, who was responsible for this action. 25 Many have tried to answer the question: "How was this possible ?" Fear was most likely the primordial instinct played upon by the extremist Hutu (with the help of Radio Mille Collines) to convince the illiterate peasant villager to kill his lifelong Tutsi neighbors. "They hear over and over that the Tutsis are out to kill them, and that is reality. So they act not out of hate as fear. They think they have only the choice to kill or be killed" (Mamdani 2001:191). Mukagasana's chilling account of the events is consistent with this analysis in her book La mort ne veut pas de moi (1997) written as a Tutsi woman survivor of the genocide. See Des Forges' (1995) and Newbury C. (1998) for an analysis of the role of the state in shaping ethnic consciousness, and Jefremovas (1995) and Longman (2001) for a contextualization of the genocide beyond the ethnic framework.

PAGE 88

72 1994, more than one million Rwandese were internally displaced and close to three million crossed the borders seeking refuge in Tanzania and Zaire. Soon, it became apparent that the Interahamwe and other gnocidaires were hiding among these refugee populations. This complication eventually discredited the UNHCR and other refugee agencies for not being able to distinguish between civilian refugees and the gnocidaires, thereby jeopardizing the security of the civilian refugee population (UNHCR 1997a). In 1996, the Tanzanian government, worried about the large numbers of armed former military and militia members hidden among the civilian refugee population, decided to forcibly repatriate the one million Rwandese refugees in Tanzania. The Zairian president Mobutu, on the other hand, who had been a close friend of the assassinated Rwandese president Habyarimana, was happy to provide shelter for the Rwandese militia much to the exasperation of the new Tutsi government in Kigali (see later). After the return of the majority of the 1994 refugees in 1996, the war in Rwanda was considered over. However, notwithstanding the international goodwill initially accorded to the new government in view of the genocide, increasingly observers saw a dictatorship emerging that relied on arbitrary arrests, disappearances, killings of unarmed civilians and unlawful detentions28 (AI 1997a, 1998, 2001, 2002; Reyntjens 1999:25-37; HRW 2000b). 26 Both Mamdani and Lemarchand (1997) relay how the Burundi refugees played an active role in the Rwandese ethnic killings and some of the most gruesome tortures "as if they were settling old scores, even across the border" (Mamdani 2001:205-6). 27 Jefremovas (1995) estimates that maximum 1 to 2% of the Rwandese Hutu population, 75,000 to 150,000 individuals, joined in the killings. As such, the 4.5 million displaced people, 75% of the total population had unjustly been accused of complicity, even as "many of these millions . .have to answer for . cowardice" (28). 28 It should be mentioned that increasing tensions between the Tutsi genocide survivors and the Tutsi old diaspora, occasionally led to open confrontations. The departure into exile of an increasing number of Tutsi

PAGE 89

73 By 1996, the refugee camps in Zaire had been turned into military bases from which regular cross-border incursions were launched in order to destabilize the new Rwandese Tutsi government, a situation condoned, albeit supported, by Zairian government officials because of the alliance that existed between Habyarimana and Mobutu. At the same time, attacks against ethnic Tutsi in the Kivu provinces29 in Eastern Zaire, instigated by the Interahamwe militia and the former Rwandese military, also called the ex-FAR (Forces armes rwandaises), multiplied. They destabilized the region to such an extent that in October 1996 the provincial authorities demanded that the Banyamulenge leave the country altogether. This attempt to expel the Banyamulenge from their homes in Zaire, was "a gift from the heavens" for the Rwandese government, now in a position to defend its cross-border advances as preventive actions against another genocidal attack on a Tutsi community (Weiss 2000:3).30 The Rwandese army invaded the Hutu refugee camps and attacked the Zairian army, causing the flight westwards of the ex-FAR, the Interahamwe and thousands of Hutu civilians, while the (e.g., military men, journalists, students and administrators from voluntary organizations) testified to this mounting tension (Reyntjens 2000: 5-10). 29 Since precolonial times, there had been many migrations into different parts of the Kivu provinces in Eastern Zaire by Rwandese Tutsi and Hutu, commonly called the Banyarwanda. One of these groups in South Kivu (consisting of mainly Tutsi generally said to have arrived in the region in the 1880s and thus legally Zairean nationals) for electoral reasons decided to assume a territorial identity (Banya-mulenge Sons of the mountain Mulenge) rather than an ethnic one (Banya-rwanda Sons of Rwanda). This move was interpreted by the local Zairean politicians as suspicious and intended to cover up their "real" Rwandan nationality (Mamdani 2001:234-253). 30 Rwanda had pleaded with the international community time and again, yet to no avail, that the Hutu military and Interahamwe be separated from the civilian refugees.

PAGE 90

74 vast majority of Rwandese refugees had little choice but to walk back into Rwanda (see map 1-4 in appendix a).31 In order to portray their actions as something other than an attack against a sovereign state, Rwanda and Uganda32 were eager to find Zairian allies against Mobutu. They found one in the person of Kabila. With the help from Rwandese and Ugandan troops, Kabila reached Kinshasa by May 1997 meeting very little resistance from the Zairian government soldiers who preferred to flee or join the ranks of the rebels rather than fighting for a dictator hated by the Zairian people (Manahl 2000). Unfortunately, Kabila was not able to rally the major political forces of the country around him, nor to contain the deteriorating security situation in Eastern Zaire (now renamed the Democratic Republic of Congo). On the Kivu-Rwanda border, the Interahamwe and former Hutu military continued their operation against the Rwandese government, while further North, various Ugandan insurrectionist forces continued to use Congo as a base from which to attack Uganda (Weiss 2000:8). In August 1998, Rwanda, Uganda and several Congolese army units took control of large segments of Eastern Congo. Before long, Zimbabwe, Angola and Namibia joined the Congolese side, while Burundi army units were observed fighting on the Rwanda-Uganda side of the conflict.33 Apart from the Congolese rebel groups cooperating with or backed by the Ugandan and Rwandese armies, another 31 UNHCR estimated that in the process of breaking up the camps and the flight west ward close to two hundred thousands persons "remained unaccounted for, with many dying of natural and violent causes" (UNHCR 1999c: Internet) 32 Uganda joined the Rwandese invasion because for some time, Zairean territory along the Ugandan border has been used as a rear base by anti-Museveni forces (Weiss 2000:5). 33 Burundi's Tutsi forces joined the war because Burundi Hutu rebels joined hands with the Rwandese Hutu in exile and used the DRC as an rear base to launch attacks against the Burundi government. An elaborate overview of the motivations of any of the other countries to join Africa's First Continental War can be found in Braeckman's L'enjeu Congolais: L'Afrique centrale aprs Mobutu (1999).

PAGE 91

75 Congolese rebel group opposing the invaders had been gaining strength in the meantime. The Mayimayi34 blamed the Banyamulenge for the situation in Eastern Congo, and targeted all Congolese seen as linked in one way or another with the Rwandese and Ugandan occupying forces. In view of the fact that Eastern Congo is extremely rich in mineral deposits, the assertion posited by certain researchers that "the continuation of seemingly "senseless" civil wars is sometimes linked to the "rational pursuit of economic goals" by the warring factions" (Mwanasali 2000: 145) is most applicable to the Congolese current situation. A 2002 United Nations report on the pervasive plundering of Congo's natural resources has identified three elite networks consisting of "political and military elites and business persons" (6) many of whom it refers to by name. One of these networks, for example, is estimated to have "transferred ownership of at least US$ 5 billion of assets from the State mining sector to private companies under its control in the past three years" (7). In another example, a Canadian national was recently arrested in Brussels, Belgium, carrying US$ 500,000 worth of gold bullion bars in his backpack. The arrested man was reported to have been making such trips bi-weekly, carrying each time the same amount of gold from Congo to Belgium, the UK, the US and Switzerland for the past four years (IRIN 2002b: Internet). The human costs suffered by Eastern Congo's population in Africa's First Continental War are enormous. A recent study (IRC 2001) estimated that between 34 The Mayimayi have a long history in the province of Kivu in Eastern Congo going back to the 1960s. Initially siding with Kabila against Mobutu, they switched sides in view of Kabila's perceived ties to Rwanda. Holding out against the other rebel movements in Eastern Congo--who are either Rwanda or Uganda backed--the Mayimayi are generally considered the only Congolese militant group that has not sold out to foreign powers. Their popularity among the population however, suffers greatly from their ruthless, often cruel and sadistic, modus operandi (Custers 2001).

PAGE 92

76 August 1998 and April 2001, more than 12% of the region's total population of twenty million died of causes related to the war. With regard to the whole region, USRC estimated that by the end of 2001, close to 4 million persons from Congo, Rwanda and Burundi were uprooted. Two thirds of these were internally displaced and, of the other third, close to 1 million found refuge in Tanzania.35 2.1.2 The Refugee Receiving Country Contrary to the precolonial centralized states in the Great Lakes region, the territory of present-day mainland Tanzania knew "a bewildering series of . [migrations] of Bantu, Nilotic, Nilo-Hamites, and Hamites which resulted in the present highly fragmented populations and cultures of this area" (Gailey 1981:210). From the late 18th century onward, the influence of the Omani Arabs--during the previous millennium limited to the East African coastal towns of Mombassa, Zanzibar and Kilwa--began to extend to the hinterland through an expansion of the Arab and Swahili36 long-distance trade. The impetus for the extension of the long-distance trade37 came from increased contact and trade agreements between the Sultan of Zanzibar and American and European states. The major social, economic and cultural impact of the long-distance trade on the societies in the hinterland was the spread of Islam and the Swahili language. 35 Roughly 500,000 refugees receive UNHCR assistance, another 470,000 (mainly Burundi) refugees have selfsettled in Tanzania's rural Western provinces (USRC 2002:56). 36 The people occupying the narrow coastal strip of East Africa stretching from Mozambique in the South to Somalia in the North, and its adjacent islands are referred to as the Swahili. For centuries, this region was a zone of intercontinental trade between the people of the Arab world, Persia, India, China, Eastern and Central Africa. "In the process, small communities of foreigners settled in the region, intermingled with and were absorbed into the local population. Consequently, under the influence of foreign settlers the indigenous population converted to Islam and Africanized the religion into local temper and taste" (Mazrui and Shariff 1994: 3). Their language, Swahili, is a Bantu language with a Bantu grammatical structure, although it has incorporated many words from Arab over the past centuries and still continues to do so. 37 Traded goods were slaves, ivory, gum copal and cloves in exchange for sugar, beads, brassware, gums, gunpowder and cotton cloth (Ajayi 1998:86).

PAGE 93

77 The travels of the abolitionist David Livingstone were a major factor in drawing European attention to the "inhumane character of the Arab slave trade" in the interior of East Africa, and the latter half of the 19th century saw a boom in the establishment of missionary colonies for the freed slaves paving the way for military conquest (Ajayi 1998:99). After World War I, the Germans, who had colonized Tanganyika (the name for mainland Tanzania) from the end of the 19th century onwards, were forced to relinquish authority to the British, who governed the country through indirect rule until 1961. Colonial policies "accentuated regional differentiation and rivalry," with the aim of weakening any territorially organized opposition to the colonial government (Mazrui 1990:232-235). One man, however, was able to coordinate and unite the many resistance initiatives and in 1954, the Tanganyika African National Union (TANU) was born under the leadership of Julius Nyerere, who was subsequently to become independent Tanganyika's first president in 1961.38 2.1.2.1 Nyerere's Tanzania Tanzania is one of the few African countries that have not known civil strife either before, during or after independence, unlike many of its neighbors. Observers have attributed the country's stability to the vision of Julius Nyerere but also to the territory's considerable ethnic complexity of around one hundred and twenty ethnic groups, only one of which counts more than one million members.39 From the beginning, Nyerere's goal was national unity, based on a state ideology (the Ujamaa socialism) and one 38 Zanzibar and Tanganyika united in 1964 to form present-day Tanzania. 39 The total population of current-day Tanzania is 30 million.

PAGE 94

78 national African language, Swahili (Forster et al. 2000: 102-105). The Ujamaa philosophy was based on the idea that the solidarity of Africa's traditional societies could be brought up to the national level, creating a new national culture.40 With numerous ethnic groups, each one having its own histories and myths, traditions and heroes, "it was the political system itself which came to constitute the main base of the national culture" (Lange 1999: 42). All these factors have contributed to the remarkably low role the ethnic factor has played in Tanzanian politics over the past four decades, contrary to the situation in many of the other countries in the region. One observer points out that: For Julius Nyerere of Tanzania, . socialism was first of all an attitude of mind. The key oppositions, for Nyerere, were not between rival economic systems or modes of production, but between conflicting moral orientations: selfishness versus sharing; exploitation versus solidarity; individual acquisitiveness versus communal mutuality. Socialism, for Nyerere, was the rejection of selfishness; a capitalist, in contrast was defined as "the man who uses wealth for the purpose of dominating any of his fellows" (Nyerere 1968). (Ferguson 1993: 83) Nyerere applied his philosophy not only at the national, but at the international level as well by extending his country's hospitality to those seeking refuge from armed struggles against colonialism, racial domination and apartheid. In 1983, he received the Nansen Award from the Nansen Medal Committee41 chaired by the UNHCR "in recognition for his outstanding services to refugees" (UNHCR 1999a:10). 40 The overall rate of success of the Ujamaa exercise was put into question by several observers, notably by Goran Hyden in his Beyond Ujamaa in Tanzania: Underdevelopment and an Uncaptured Peasantry (1980). 41 The medal is named after the Norwegian explorer Dr. Nansen, pioneer of international humanitarian aid, who became the first League of Nations High Commissioner for Refugees in 1921. One of his greatest achievements was the introduction of a travel document for displaced people commonly referred to as the "Nansen passport" (UNHCR 1999a:11).

PAGE 95

79 2.1.2.2 Africa's "open door" policy in Tanzania The period from the 1960s onwards until the late 1980s is sometimes described as the "golden age" of asylum in Africa with many states adopting an "open door policy" (Rutinwa 1999:5). While in Southern Africa, most independent nations received thousands upon thousands of refugees fleeing racism, colonialism and apartheid (from countries such as South Africa, Angola, Mozambique and Namibia), in Central Africa, Rwandese and Burundese refugees escaping internal political conflicts readily found refuge in Tanzania, Uganda and Zaire. Later, a number of the refugee receiving countries in Southern Africa (e.g., Botswana, Lesotho and Swaziland) succumbed to intense pressure from South Africa and evacuated the South African refugees from their territory. Other countries, such as Tanzania, Zimbabwe and Zambia, remained committed to the norm of non-refoulement by accepting the evacuated South Africans (6). During this area, African host countries, in collaboration with UNHCR and the international community, ensured not only the refugees' right to protection but attended to their basic needs as well. Even if refugees were mostly protected in camps, "these camps were sufficiently large to enable refugees to acquire land and engage in economic activities which earned them a dignified living" (7). Tanzania, of all African countries, was one of the most generous countries to refugees, allowing them full access to social welfare, the education system, as well as the job market, both in the private and public sectors. They were offered land and even Tanzanian citizenship (Gasarasi 1990). Many former refugees went on to rise to very high positions in the civil as well as the diplomatic services. The underlying philosophy for these policies was Nyerere's strong conviction that Tanzania "could only sustain genuine freedom if the neighboring countries are peaceful" (UNHCR 1999a:9).

PAGE 96

80 Nyerere's philosophy is best exemplified in his attempt in 1983 to ban the Swahili word for refugees, wakimbizi and replace it by the term wageni wakazi. The word wakimbizi comes from the verb kukimbia which means "to run away from" implying negative connotations such as cowardice and rootlessness. The term wageni wakazi may be translated as "guest-workers," emphasizing the positive contributions refugees can make to the national economy of the country of asylum and their sense of self-sufficiency and economic independence.42 2.1.2.3 The change in climate The economic crisis of the 1980s and the IMF's and Worldbank's austerity measures that followed in its wake,43 together with the massive refugee influxes after the 1994 Rwandese genocide, are only some of the reasons behind the drastic changes in Tanzania's refugee policy from the late 1980s onward. In the aftermath of the "liberation" of the former Soviet Union and other East European countries, European countries fearing an immense influx from the East, changed their asylum and migration policies to become even more restrictive (Escalona and Black 1995) much to the agony of human rights advocates (Dacyl 1996). This "downward standard-setting led by the migrantand refugee-receiving countries of the North" all but accelerated "the erosion of an international commitment to the refugee regime" (Collinson 1999: 21). In 1995, after receiving international criticism for closing the borders with Rwanda and Burundi to prevent further refugee influxes, the then-Tanzanian Minister for Foreign Affairs said that 42 Despite all good intentions, the new Swahili term for refugees did not hold longer than two years because "it was felt at the annual conference of refugees, that the word was unacceptable because the fact remained that persons who had fled from their respective countries due to war were [still] refugees" (Kasomangala in UNHCR 1999a:9). 43 In fact, it was Nyerere's initial refusal to accept IMF conditions for further loans that lay at the basis of his stepping down as president in 1985, after 24 years of ruling the country.

PAGE 97

81 "it was a double standard to expect weaker countries to live up to their humanitarian obligations when major powers did not do so" (Rwegasira in Rutinwa 1999: 17). Following the fall of the Berlin Wall and the end of the Cold War in 1989, African countries were strongly encouraged to democratize, institute multipartism and organize general elections. At the same time, the profile of the 1960s and 1970s refugee as a "freedom fighter" (e.g., from South Africa, Angola, Mozambique) had drastically changed in the 1980s and 1990s with virtually all refugees coming from independent countries engaged in internal (often ethnic) conflicts, and thus not conducive to the formerly existing feelings of sympathy. These negative dispositions were fueled further by the negative economic and environmental effects on already impoverished host populations44 of ever larger groups of refugees, ultimately giving rise to xenophobic feelings. This happened at a time when most of Africa is democratizing and governments are compelled to take into account public opinion in formulating various policies. The result has been the adoption of anti-refugee platforms by political parties which result in anti-refugee policies and actions by governments. (Rutinwa 1999:1) Another important reason for the change in refugee perceptions and policies in Tanzania lies in a heightened sense of insecurity both at the local and the national level. The media are rife with stories blaming refugees of committing violent crimes and overcrowding prisons, to retarding national development: Growing waves of crime and violence perpetrated by refugees in Ngara district have retarded economic development because Tanzanians are no longer certain of their security. . (Ngara District Executive Director Esther Mbigili in The African, Jan 22, 2000) The situation reflects how pathetic the security in the region is . the influx of refugees in Kigoma fuels crime in the region, leading to the overcrowding of 44 Tanzania is still one of the poorest countries worldwide: 156th out of 174 (UNDP 2000).

PAGE 98

82 prisons . Incidents of highway robbery using fire arms and home breaking which were not very common before, now are noticeably increased . (Kigoma Regional Police Commander in The Daily News, Feb 26, 2000) An unfortunately common side effect of refugee influxes (equally observable in other African countries) is that among the civilian refugees pouring into the country of asylum, are hiding former military, rebels, or people otherwise engaged in the armed conflicts in the home countries, who possess weapons. While some will use their weapons to commit crimes in the country of asylum, others will sell theirs in an attempt to complement the meager rations provided by international aid agencies. Prices are usually deflated because of the supply-demand effect. Landau (2001) reports that "AK-47s and other sub-machine guns are readily available in the refugee-populated areas [in Tanzania] for between $10 to $20" (23-footnote). Whereas, undoubtedly, there is an increase in crimes in heavily refugee populated areas, inducing a feeling of heightened insecurity at the local level, one should not underestimate the extent to which the situation is blown out of proportion in the media. Refugees are regularly used as scapegoats by local and national officials for existing infrastructural shortcomings and professional incompetencies. At the national level, the refugee is seen as an additional threat to Tanzania's state sovereignty already weakened by the increasingly authoritative power of international organizations. The IMF and Worldbank dictate to a large extent how the economy should be run while in the refugee-populated areas, residents have reportedly become more inclined to turn to international actors than to the state (Landau 2001). Conceding that refugees are part of more complex migrant diaspora networks45 that crosscut 45 A topic that is currently underresearched, as Van Hoywegen rightly points out. "With respect to Africa, there is . an exaggerated focus on crisis related displacement and less so on labor migration. . It is . important to see how refugee and migrant labor networks interact" (2001: 10). The extent of these interactions became increasingly clear to me while interviewing informants, and discovering that many

PAGE 99

83 international borders often operating on the margins of the law (MacGaffey and Bazenguissa-Ganga 2000) and bearing in mind the role these trade networks play in the warlord politics of many of Africa's current conflicts (Reno 1998), one easily understands why "[r]efugees have become, according to the authorities, a security concern, threatening the peace and stability of the Tanzanian nation" (Van Hoyweghen 2001:1). From the above, it has become clear how local, national, international and global elements and events contributed to the current refugee climate in Tanzania. 2.1.3 Rwandese, Burundese and Congolese refugees in Tanzania The history of Rwandese and Burundi refugees in Tanzania, sadly enough, goes back many years.46 Since independence, Rwandese Tutsi47 and Burundese Hutu have been coming to Tanzania in multiple, large waves in 1959, 1964, 1973 and 1972, 1988, 1993 respectively. Until 1996, however, there were no major refugee influxes from Congo (then Zaire), as only a limited number of individuals sought refuge in Tanzania as political opponents of Mobutu and his regime.48 When war broke out, in 1996 and again in 1998, large refugee flows poured into Tanzania not only of Congolese, but also of Rwandese and Burundese Hutu "uprooted" from the refugee camps in Eastern Congo, where they had fled to from their home countries in 1994 and 1993 respectively. Some refugees--although genuinely refugees--were part of larger migration networks to other African countries, but also Europe and Canada. 46 It should be mentioned that refugees from other countries (South Africa, Mozambique, Uganda, Sudan, Somali to name just a few) have been equally welcome on Tanzania over the past, however their history falls beyond the scope of this study. 47 In 1994, the year of the genocide and regime change from Hutu to Tutsi, former Tutsi refugees from countries all over the world returned to Rwanda in large numbers, while at the same time about 1 million Rwandese Hutu sought refuge in Tanzania and 2 million in Congo (then Zaire). 48 According to Eger (1979), of the three thousand Congolese refugees that fled to Tanzania during the Mulele uprisings of 1964, only several hundred remained by the late 1970s (Mulele was a close ally and friend of the assassinated Patrice Lumumba).

PAGE 100

84 of these Rwandese camps were actually attacked in 1996 by Rwandese government troops in their quest to eliminate the Interahamwe militants and other gnocidaires.49 Up until the time of writing, despite several--eventually aborted--repatriation exercises for all three nationalities, new arrivals keep trickling in due to the unabated conflicts raging in the Great Lakes region, albeit in another form than the full-scale wars of 1993, 1994, 1996 and 1998. Rutinwa (1999) mentions the growing trend in the African context of states refusing entry to asylum seekers or returning refugees to their countries of origin before the situation is considered safe to return. Tanzania as well, after receiving hundreds of thousands of refugees from Burundi in 1993 and Rwanda in 1994, closed its borders with those two countries in 1995, while the Minister for Foreign Affairs openly declared: "We are saying enough is enough. Let the refugees go home and no more should come" (11). Not only are refugees refused asylum, thereby contravening international refugee law, scores of Rwandese, Burundese and Congolese refugees have at times been forcibly repatriated. In 1996, both Zaire and Tanzania forcibly repatriated 700,000 and 540,000 Rwandese refugees respectively (AI 1997b), and in 2000 Amnesty International reported that "standards of refugee protection in Tanzania have dropped dramatically since late 1999 as scores of Burundese and Rwandese refugees have been forcibly returned to their countries" (AI 2000b: Internet). UNHCR as well mentions that during 2001 several 49 The larger part of the 2 million Rwandese Hutu refugees in Congo (then Zaire) were thus forcibly repatriated (see also earlier in this chapter), yet several hundreds of thousands preferred to flee to Western Zaire instead. These groups, suspected to be involved in the genocide, were pursued by the Rwandese government troops (and collaborating Congolese troops) through thousands of miles of dense forests. No exact numbers exist, but several tens of thousands persons are assumed to have vanquished. Umutesi's Fuir ou Mourir au Zaire (2000) recounts the personal experiences of the author during this ordeal. See map 1-4 in appendix a.

PAGE 101

85 hundred refugees were "rounded-up" when outside of the "designated areas" and were "forcibly relocated . to their home countries" (UNHCR 2002b: 4). Consequently, [r]ound-ups, conjoined with a general nervousness on the part of the refugees about the sustainability of their asylum in Tanzania, appears to have resulted in several thousand refugees moving to other countries in the region usually Uganda, but also Mozambique, Malawi and Kenya. (Ibid.) Notwithstanding the difficult circumstances for refugees in Tanzania, the country continues to host the largest refugee population of Africa, and the second largest worldwide.50 Over the period 1998-2001 alone, Tanzania's refugee population increased by 30%, adding to the "considerable hardening of attitudes by Tanzania to refugees" (USRC 2002:101).51 2.1.3.1 The refugee camps in Western Tanzania By the end of 2001, Tanzania hosted over half a million UNHCR registered refugees living in twelve refugee camps along the Tanzanian border with Congo, Rwanda and Burundi (see map 1-6 appendix a). UNHCR is the agency receiving funding for refugee programs from the international community, with which it subcontracts governmental, national or international agencies, called "Implementing Partners." In Tanzania, around twenty implementing partners are operating in the refugee camps (UNHCR 2002b:16). Each one is responsible for a specific program or assistance project in their respective areas of expertise, while UNHCR retains the responsibility of the overall coordination and the monitoring of the international funds through its five offices 50 Afghani refugees in Pakistan still numbered over 2 million in 2001. 51 Whereas 2001 saw the repatriation of 6,500 refugees to Rwanda and Burundi, there was at the same time a continued influx of new arrivals numbering 34,000 of mainly Congolese, Burundese and Rwandese nationals (UNHCR 2002b).

PAGE 102

86 in the region.52 Each refugee camp, on average, has a population of between thirty five and fifty thousand persons and is run by a representative from the Refugee Department at the Tanzanian Ministry of Home Affairs. There have been many problems in Tanzania's refugee camps over the past years (e.g., Express 1999; USRC 1999; Long 1999), but the following three have attracted the most attention: cuts in already low food rations, the worsening security situation, and the high incidence of sexual violence. Funding shortages experienced by UNHCR at the beginning of 2001, forced the agency to reduce their implementing partners' budget by 20% (UNHCR 2002b:3). The most severe consequence was a cut in the refugees' food rations to 80% of the required level, causing unrest in the camps. The Daily News reported that "police last week forced to intervene and stop about 3,000 refugees from the DRC, who had fled Lugufu camp, from further advancing towards Kigoma town in search of food" (March 14th, 2001:1). In the Guardian, it was written that in Kibondo "the reduced food rations in the camps have led to marauding hungry refugees who take food by force . villagers are afraid to do farm work and the women afraid to venture out and risk rape . ." (May 17th, 2001:1). Increasing security concerns as perceived by the Tanzanian population have already been mentioned, yet too often it is easily forgotten that the refugees themselves experience feelings of heightened insecurity as well. The most striking example was the assassination of the brother (a Hutu intellectual) of the president of one of the Burundi rebel parties participating in the Arusha peace talks, and four of his family members on 52 In addition, and in accordance with its mandate, UNHCR remains solely responsible for the legal protection of the refugees and any protection-related matters, e.g., selection of eligible refugees for resettlement in third countries (more in chapter 3).

PAGE 103

87 February 24th, 2000.53 However, the situation is said to have improved recently through the increased funding by UNHCR of several hundred Tanzanian policemen in and around the refugee camps and the appointment of refugee watchmen and women, who patrol the camps and report incidents to the police (UNHCR 2002c; 2001a). With regard to the high incidence of sexual violence54 (HRW 2000a), there is increasing recognition of the fact that "minors [are] involved in such crimes, not only as victims, but as offenders" and, of "the need for strengthening community values" to counter the problem55 (UNHCR 2002b:5). Fortunately, Human Rights Watch recently announced that significant improvement had been made in this area through the hiring of "two sexual and gender based violence assistants to follow up on cases of violence against women, two Tanzanian lawyers to assist women in taking their case to court, and an international security liaison officer to train police deployed in the camps" (HRW 2000c: Internet). Notwithstanding the efforts made by national and international refugee agencies to improve the conditions in the refugee camps with regard to (amongst others) food rations, security and sexual violence, all of these issues induce growing numbers of refugees to disregard the international assistance provided and head for Dar es Salaam instead (more in chapter 3). 53 Countless other allegations of nightly incursions of bands of armed men into the refugee camps and the subsequent disappearing of refugees were recounted to me by informants (more in chapter 3). 54 Leading one Tanzanian journalist to advocate the establishment of a separate camp for refugee women (The Guardian, Sept 28, 2000, p7). 55 Umutesi (2000) notes from her experience as a refugee in one of the Rwandese refugee camps in Congo (then-Zaire) that "[t]here were voices in the international community ascribing the increasing number of single mothers in the camps to a well-designed Hutu plan to regain power in Rwanda in the future . It was parental authority in crisis. Parents were no longer obeyed because children were no longer economically dependent of them" (96, own translation). See also Turner (1999) for a description of how socio-cultural values are subverted by the workings of international aid programs in refugee camps.

PAGE 104

88 In an area described as "Tanzania's impoverished North West, where basic social services barely meet the needs of local residents" (USRC 2002: 101), the impact of such large refugee populations and the accompanying influx of enormous amounts of money56 and hordes of expatriates cannot but have devastating effects on local structures. Compared to the total Tanzanian population of thirty million, the refugee ratio is one in sixty. Yet in Western Tanzania, where the overwhelming majority of refugees resides, the ratio becomes one in three compounding the overstretching of the local infrastructure. Fortunately, donors responded strongly to the government's request for support in this area and long-neglected infrastructure was upgraded with minimal required contribution from host communities (Whitacker 1999: 8-10). At the same time, money poured into the area through the purchase by expatriates of imported western comforts, and as a result of the wages paid by the refugee agencies to local Tanzanians, which were of a level unheard of even in the capital of Dar es Salaam (Waters 1999:147-148). These large injections of cash into the local economy had many unintended effects, e.g., "thousands of Tanzanian women from urban and rural areas flock to Western Tanzania, Kigoma, for sex work because of the high salaries paid out to aid workers" (The Guardian, Nov 26, 1999, 7). Whitacker, after a detailed study, concluded that: It is not possible to say whether host communities in Western Tanzania as a whole gained or lost as a result of the influx of refugees and relief resources. The situation created both positive and negative opportunities for local hosts. (1999:14) As briefly mentioned before, one more unexpected consequence of the way the international aid system is set up, was the fact that local Tanzanian residents increasingly turned towards international agencies operating in the region with their demands, 56 E.g., foreign aid to Ngara (in Western Tanzania) alone amounted to 4.8% of Tanzania's gross national product in 1995 (Waters 1999:151).

PAGE 105

89 expectations and accusations, instead of local government structures (Landau 2001). While recognizing that such practices exist elsewhere in Tanzania as well, the author of a recent study argues that "their prominence and significance in refugee-populated areas demands careful consideration" (1). The effects of setting up an artificial infrastructure to accommodate over half a million refugees in a poor, rural area such as Western Tanzania clearly ripple well beyond the lives of the refugees and the local populations.57 2.1.3.2 Urban refugees in Tanzania UNHCR had an assistance program for the urban caseload since the office was established in 1964. Until 1995, urban refugees in Tanzania were provided with work permits by the Tanzanian government enabling them to engage in income generating projects, albeit assisted by UNHCR in the form of grants or soft loans (UNHCR 2001b). Tanzanian refugee policy changed drastically in the wake of the 1994 Rwandese genocide and the subsequent influx into Tanzania of over 1 million refugees. Currently, only a very limited number of refugees can obtain permits to reside in Dar es Salaam on the basis of either medical or political reasons, or because they are "in transit" to a third country. The so-called "medical cases" are refugees who have been referred for medical reasons by the camp health authorities to come to Dar es Salaam for treatment in the Muhimbili Hospital. The "political cases" are refugees who for security reasons cannot reside in one of the refugee camps, even in the Mkungwa camp, a camp specifically established to house "security-cases," e.g., refugees of mixed parentage Hutu-Tutsi, Congolese ex-military that collaborated with the Rwandese occupation forces, etceteras. 57 For a more indepth analysis of the effects of the refugee situation on the local population of Western Tanzania, the reader is referred to authors such as e.g., Landau (2001), Whitacker (1999), Waters (1999).

PAGE 106

90 The situation of some refugees is considered to be so politically sensitive that for their own security the Tanzanian government allows them to stay in Dar es Salaam.58 The final category consists of those refugees residing in Dar es Salaam on transit, pending resettlement in a third country, family reunification or repatriation. All three categories of people are accommodated by UNHCR in four different guest houses and receive an allowance of TS 60,000 per month (US$60) covering all other living expenses. UNHCR's current strategy is to limit as much as possible the number of urban refugees for budgetary reasons, on the grounds that each urban refugee costs $400 per month and that there have been too many abuses of the system in the past. While in August 2001, UNHCR assisted 214 urban refugees, by July 2002 only 35 refugees received UNHCR assistance, notwithstanding the continuous increase of the total (urban) refugee population during the same period. In addition to the UNHCR assisted urban refugees, there are several hundred refugees--mainly but not solely Burundese and Congolese--who obtained permits to reside in Dar es Salaam from the Tanzanian Ministry of Home Affairs, but whom UNHCR have decided not to assist financially claiming budgetary constraints. Despite their official, strict stance on urban refugees, both the Tanzanian Government and the UNHCR office acknowledge a continuous increase in the number of undocumented and unassisted urban refugees "the number of [which] is not known" (UNHCR 2001b:1)59 but which can safely be assumed to run at least into the tens of 58 E.g., a journalist working for the Rwandese-backed rebel forces in Congo, who at a later stage, filmed their atrocities and distributed them to the international press. 59 Contrary to earlier assumptions (Sommers 2001; UNHCR 1997c), recent statistical information on urban refugees in other parts of the world has revealed that it does not hold true that the majority of refugees in urban areas are usually young, single males, especially "in many cities where substantial numbers of refugees are to be found" (UNHCR 2001c: 2-3) as is the case in Dar es Salaam.

PAGE 107

91 thousands. These forced migrants mostly originate from Uvira, Bukavu and Goma in Eastern Congo, Bujumbura in Burundi and Kigali in Rwanda (more detail in chapter 3). Although urbanites, their home towns have populations that rarely exceed several hundred thousand inhabitants, unlike the vibrant cosmopolitan city Dar es Salaam has become over the past few years. 2.2 The Research Site: Dar es Salaam Conceived of by the Arabs in the latter half of the 19th century, Dar es Salaam's (Haven of Peace in Arabic) port and center were constructed by the Germans who, in 1891 moved Tanganyika's capital here from Bagamoyo.60 The population of barely 5,000 consisted of European colonizers, Asian merchants and a majority of African inhabitants, most of whom were Zaramo, the ethnic group on whose land the city was built (Swantz 1970:408). In the wake of the increasing rural-urban migration during the British colonial era, many Zaramo villages were absorbed by the immigrants' shanty towns (Iliffe 1979: 386). At independence, Dar es Salaam's population was around half a million people or less than 5% of the total population of Tanganyika (Ngware and Kironde 1996:8). Even after the Tanzanian capital was relocated to Dodoma in October 1973 for geographical and ideological reasons (Lugalla 1995:33), Dar es Salaam remained the de facto capital. Currently accommodating 3 million inhabitants, the population is likely to continue to grow at 8 to 10% per year further contributing to the poor planning and the uncontrolled spatial growth of the city (see map 1-7 appendix a). Since the trade liberalization and the retrenchment programs of the late 1980s and the 1990s, three 60 Bagamoyo is a port town approximately 60 kms to the North, that used to be the point of departure of the slave caravans to the interior around and beyond the Great Lakes region during the 19th century.

PAGE 108

92 hundred thousand Tanzanians move to Dar es Salaam every year (Ngware and Kironde 1996:125), increasing at a tremendous rate the settlements of spontaneous slums, a phenomenon unknown during Nyerere's socialist era. Evaluating the effects of the IMF's Structural Adjustments Programs, Lugalla (1997) describes Dar es Salaam as a place where: . there are frequent water cuts which sometimes leave areas dry for more than a week. There are electricity blackouts, telephones which maintain an eerie silence, inadequate parking spaces, overflowing sewage, congestion of vehicles which do not observe traffic regulations, hospitals without medicine, roads with pot holes, pick-pockets and gangs of armed robbers, and streets without light but with the pungent smell of uncollected garbage. There are more beggars, disabled, street-children, hawkers, cows and goats, all of which contribute simultaneously to traffic jams. (Internet) At the same time, [m]ulti-story buildings are changing the urban geography of Tanzania. The winds of modernization and dependency have increased their speed to the extent that even the Sheraton Hotel chain has found a home in Tanzania. Luxurious buildings have mushroomed in beach zones like Msasani, Mikocheni, Kawe, Mbezi and Tegeta in Dar es Salaam . The number of luxurious air-conditioned four-wheel vehicles fitted with telephones, video and television sets has increased. (Internet) To this, the last few years have added to Dar es Salaam's landscape supermarkets, international food and clothes chains (such as Woolworth and Subway), countless Internet cafes and, last but not least, mobile phones.61 Lugalla's conclusion rings very true, namely that "the SAPs have not meant the same thing to everyone; wealth and poverty are not isolated trends but rather two sides of the same coin" (1997: Internet). It should be emphasized here that the fast growing discrepancy between rich and poor is a reasonably new phenomenon in Tanzania, unlike in many other countries. Whereas it is 61 Virtually every fourth person on the bus has a mobile phone. While personal means of transportation in the form of a private car is still out of the financial reach of the majority of Dar es Salaam's residents, mobile phones appeared to have been to many an affordable answer to the inadequacies of the traditional landline net.

PAGE 109

93 sometimes condescendingly said by nationals from non-socialist African countries that under Nyerere "Tanzanians used to share their poverty,"62 the Ujamaa policies of the 1970s did result in a more or less equal living standard for all Tanzanians. Unfortunately, its main pillars (i.e., an increased access to education and health services, and guaranteed employment even at subsistence wages) were abandoned in the latter half of the 1980s when Nyerere's successor, Ali Hassan Mwinyi, under international pressure, introduced the country to the principles of liberal economics. The full-blown liberalization of the economy being a recent phenomenon (early 1990s onwards) today's average adult Tanzanian grew up under Nyerere's socialism, and was inculcated by the regime's suspicions vis--vis Western and Westernized countries' attempts to discredit its socialist principles and introduce the capitalist market economy. The subsequent, rather abrupt, discontinuation of the old egalitarian values and the rapidly widening gap between the few rich and the many poor--a phenomenon strikingly visible in Dar es Salaam as the country's largest urban center and main gateway to the outside world--have contributed to the old feelings of suspicion to anything "foreign" and created a general sense of resentment among ordinary Tanzanians towards non-Tanzanians. In addition to further elaborating upon these elements that constitute the social framework that refugees in Dar es Salaam will unavoidably be confronted with, the following sections also give an overview of the political and humanitarian infrastructure present in Dar es Salaam and which may be relevant to the lives of many of the Great Lakes region's forced migrants. 62 Les Tanzaniens se partagaient leur pauvret. Interview with Congolese refugee, May 20, 2000.

PAGE 110

94 2.2.1 Governmental and Political Structures Some of the following political institutions or practices were established in the wake of the current refugee situation in the country (e.g., the 1998 Refugee Act and the Center for the Study of Forced Migration), others are inheritances from the Nyerere era and have gained new meanings over time (such as the Ten House system). 2.2.1.1 The 1998 Refugee Act and the Refugee Department During the golden days of the "open-door-policy," the Government of Tanzania issued a notice stipulating that persons from certain nationalities entering Tanzania after a certain date automatically acquired refugee status. As such, Congolese entering Tanzanian territory after June 1st, 1964, Rwandese after January 1st, 1961, and Burundese after April 29th, 1972 automatically fell under the Refugee Act of 1965 and were exempted from the provisions in the Immigration Act of 1972 (GOT 1973; 1966). Today this de jure automatic recognition of the refugee status of nationals from certain countries no longer applies. The 1998 Refugee Act insists that all persons who enter Tanzanian territory wishing to claim refugee status must "immediately and not later than seven days after entry . present or report to the nearest authorized officer . and apply for recognition as a refugee" (GOT 1998: section 9(1)). Upon being recognized as a refugee by a representative of the Ministry of Home Affairs, the person will be assigned to reside in one of the "designated areas" (section 17). Failure to comply means that this person shall be treated as an illegal alien falling under the Immigration Act of 1995 instead of the Refugee Act, thereby becoming eligible for forceful repatriation to his/her country of origin.63 63 In addition, the maximum penalty for not having an official permit to reside in Tanzania is much higher under the 1995 Immigration Act, namely 3 years (GOT 1995: Section 31) versus 6 months under the 1998

PAGE 111

95 Consequently, those persons arriving in Dar es Salaam who choose to approach the UNHCR office in order to apply for refugee status, are immediately referred to the Refugee Department at the Ministry of Home Affairs. In practice, refugees of Congolese and Burundese origin enjoy a de facto automatic recognition and acquire without further complications the status of prima facie refugees under the 1969 OAU Protocol. They receive a letter from the Refugee Department recognizing them as refugees and assigning them to one of the refugee camps, unless there is a particular reason to allow them to reside in Dar es Salaam, in which case they may become eligible for UNHCR assistance outside the refugee camps. Individual Rwandese asylum seekers, on the other hand, are required to submit their request for asylum through the Refugee Department at the Tanzanian Ministry of Home Affairs to the National Eligibility Committee (NEC). The NEC consists of eight senior officials from various ministries, and formulates its recommendations on each applicant's refugee status to the Minister of Home Affairs, who has the final authority to approve or reject. UNHCR is a non-voting member of this committee, having only an advisory function (GOT 1998: sections 6, 7). However, the Tanzanian government decided recently that granting Rwandese asylum seekers refugee status is in contradiction with the Tripartite Repatriation Agreement signed by the Rwandese government, the Tanzanian government and UNHCR. As a result, Rwandese asylum seekers were, at the Refugee Act. Interview with Mr. Mgonja, Immigration Department at the Tanzanian Ministry of Home Affairs, on November 21st, 2001.

PAGE 112

96 time of this research, forthwith referred to one of the designated areas in Western Tanzania pending their decision to repatriate.64 Of the Congolese, Rwandese and Burundese refugees who apply for refugee status in Dar es Salaam and who are referred to one of the refugee camps, only the most destitute accept to go. The overwhelming majority decide to remain in Dar es Salaam notwithstanding the implications of leading a clandestine life in unfamiliar surroundings.65 In search of a roof over their heads and initial contacts with Dar es Salaam's Tanzanian residents, these refugees are soon to learn the implications of a system that dates back several decades and that was implemented for reasons totally unrelated to any refugee situation. 2.2.1.2 The Ten House System (Nyumba Kumi) During the few years following independence, the new Tanzanian government abolished the traditional system of hereditary chieftainships in the quest for a unifying national political system at the local level. By the mid 1960s, TANU ten-house cells were established as the most basic political unit, with each cluster of roughly ten households constituting a cell and electing its own leader (Moore 1986:139-167). Until the recent introduction of multipartism,66 Ten House leaders (Mjumbe wa Nyumba Kumi) were representatives of the CCM67 responsible for being aware of what happened in the 64 Interview with Nzuki, Senior Legal Officer in the Refugee Department, June 30, 2002. As a result of this policy, Rwandan refugees were increasingly trying to leave Tanzania for other countries fearing impending compulsory repatriation (IRIN 2002a). 65 The motives that are at the basis of these decisions will be discussed in detail in chapter 3. 66 I was told by officials at the Ministry of Home Affairs, that the system does not work as flawlessly anymore these days, when neighbors are members to different political parties. 67 In 1977, the Chama cha Mapinduzi (CCM) was born out of the Tanganyika African National Union (TANU), the party that had led Tanzania to independence from 1954 onwards (Mazrui 1990: 443).

PAGE 113

97 ten houses under their authority, and for reporting any suspicious people or activities to the CCM Ward Secretary or Katibu Kata (Lugalla 1995: 87). Thereafter, the information traveled further upward to the District Commissioner, Regional Commissioner and finally the national government. The Ten House system, representing "an additional form of social control" (Moore 1977:117), befitted the political ideology of Nyerere's time "designed to make the citizen more security conscious" and expecting him or her "to be all ears on behalf of the state" (Okema 1996: 17).68 Renting accommodations in this system is no longer solely a private transaction between two individuals. The landlord of the premises has the obligation to inform his mjumbe (Ten House leader) of the identity of his/her new tenant, and the mjumbe is expected to pass on the information to the higher levels, and eventually the Immigration police. Obviously, loopholes exist. On the one hand, it is not difficult through monetary means to "convince" certain persons to pass on incorrect information, on the other hand, many informants spoke with appreciation of the humanitarian empathy on the part of their landlords, neighbors, mjumbe or others. These types of situations together with the common observation among the urban refugees that "all Tanzanians are police informants" will be further elaborated upon and contextualized in chapter 4. The following institution under discussion may at first sight seem less relevant to the lives of the urban refugees in Dar es Salaam in the short term. However, as a kind of national think tank on refugee policy matters, some of the recommendations of this institution--if and when accepted by the political authorities--could play an important role in the long run with regard to the local integration of the forced migrants. 68 This is the era in which Tanzania's adoption of socialism and its relations with Eastern countries represented a source of anxiety to the West in the context of the war against communism.

PAGE 114

98 2.2.1.3 The Center for the Study of Forced Migration Founded in 1995 and operational since 1998, the Center for the Study of Forced Migration (CFSFM) at the University of Dar es Salaam, is an interdisciplinary institute with outside links to universities in Kenya and Uganda, to the Oxford Refugee Studies Program and Georgetown University in Washington, DC (UDSM 2001: Internet). It was conceived to initiate activities in the areas of research, training and outreach. Research and training initiatives are being implemented to the fullest. Publications on the refugee situation in the Great Lakes Region by affiliated scholars in international journals have been many, and workshops and courses are organized on a regular basis involving students as well as middle level and senior personnel in government, media or other institutions dealing with refugees and internally displaced persons, e.g., the East African School on Refugee and Humanitarian Affairs (EASRHA n.d.). The outreach activities and dissemination of information relevant to forced migration have yet to be fully developed for lack of funding. The intention is to disseminate information on refugee matters not only to the general public, but to the refugee population as well. Currently, extremely few refugees in Dar es Salaam know of the Center's existence. Occasionally, a letter is received from a refugee requesting legal assistance in appealing the decision from the Ministry of Home Affairs rejecting his/her refugee status. When the Center staff is of the opinion that the Minister in his decision overlooked certain elements which may have changed the final outcome, they may agree to represent this person and take up his/her appeal. However, these instances up until now have been extremely rare.69 69 Interview with Dr. Khoti Kamanga, Director of the CFSFM, and Professor in Law, August 9th, 2001.

PAGE 115

99 A major policy-oriented project the Center recently was involved in, is the Review of Refugee Related Policies and Laws in Tanzania, a study requested by the Tanzanian Prime Minister's Office and financed by the European Development Fund. All five contributors, being affiliates of the Center, have critically evaluated Tanzania's current policies and practices, and formulated recommendations for a future review in their respective areas of expertise. While Rutinwa's contribution on Refugee Admission and Eligibility Procedures denouncing the shortcomings of the 1998 Refugee Act's appeal procedures is important to the general refugee population, Musoke's recommendations in his contribution on Employment Opportunities for Refugees in Tanzania are particularly relevant to the urban refugee population (CFSFM 2001). He suggests, as has UNHCR in the past, to have an additional category of work permits created especially for refugees, the three existing categories of work permits being generally unaffordable to the average refugee. This new category of work permit would be free for the first year, while the fee of the subsequent yearly renewals would be modest and affordable. At the time of writing the dissertation, this document was still in the process of being distributed for comments to those local government structures concerned with refugees, after which it will be submitted to parliament for more discussions. It is unclear at this point in time when and if a major refugee policy review in Tanzania will take place. The current refugee policy not only governs the workings of the humanitarian refugee agencies in Tanzania, it implicitly delimits the possible range of initiatives from humanitarian agencies concerned with refugees in Dar es Salaam, as the following sections will show.

PAGE 116

100 2.2.2 Humanitarian Organizations One international agency, the UNHCR and a limited number of national non-governmental organizations assist refugees in Dar es Salaam. Of the Tanzanian NGOs, only one is subcontracted by UNHCR as an implementing partner and receives international funds to finance their refugee project. The other NGOs are contacted by a small number of the urban refugees on an occasional basis and provide assistance on purely humanitarian grounds because none has an official budget line for refugees in Dar es Salaam. 2.2.2.1 United Nations High Commissioner for Refugees Apart from the main office in Dar es Salaam, UNHCR has five sub-offices in Western Tanzania, the area where the eleven refugee camps are located, and employs close to two hundred staff, 20% of whom are international officers. Around twenty subcontracted non-governmental organizations are operational in the camps. Together with five other UN agencies (i.e., WHO, UNICEF, FAO, WFP and UNFPA), they implement support activities in their respective areas of expertise. The Refugee Department of the Ministry of Home Affairs is responsible for the co-ordination of all refugee-related matters in Tanzania. The 2002 budget is in the order of US$ 30 million, of which more than 90% goes directly into the assistance programs (UNHCR 2002c). Apart from financing activities in the refugee camps (e.g., food, health, education, sanitation, crop production, shelter, etc.), UNHCR has committed itself to "providing tangible benefits to the local populations" in the refugee areas "through the strengthening of local institutions, transport networks and infrastructure," as well as maintaining "vigorous environmental protection programs," e.g., soil erosion, and depletion of wood and water resources (UNHCR 2002c:Internet; UNHCR 1999b).

PAGE 117

101 All of the above activities taking place in Western Tanzania, the UNHCR Branch Office in Dar es Salaam is mostly involved with matters of an administrative/financial (e.g., fundraising, budget control, etc.) and political/diplomatic nature (such as liaising with the Tanzanian government and other diplomatic missions). The Protection and Community Services Unit, however, do receive visits from individual refugees with regard to medical referrals, permits to remain in Dar es Salaam and requests for resettlement in a third country. The medical referrals from the refugee camps and the politically sensitive cases who have the permission to stay in Dar es Salaam are catered for with respect to accommodation and payment of allowances by one of UNHCR's implementing partners, a Tanzanian NGO, Umati. Apart from these assisted refugees, there are still considerable numbers of other refugees who seek to contact the UNHCR office in Dar es Salaam for a variety of reasons, the most frequent of which is the request for resettlement. According to UNHCR principles resettlement to a third country should be considered for the following categories of refugees only: those refugees subject to a threat to physical safety or human rights in the country of refuge, survivors of violence and torture, persons with specific medical needs, women-at-risk, family reunification, unaccompanied minors, elderly refugees and lastly, refugees without local integration prospects (1997b). In Tanzania, procedures stipulate that the identification of refugees eligible for resettlement is the primary responsibility of the Protection Officers at Sub Office level, i.e., in the refugee camps. Once identified by a Protection Officer, the eligible refugee has a first interview in the camp with the UNHCR Resettlement Officer visiting from the Dar es Salaam office. If positive, the file is forwarded to UNHCR Geneva for final in-house approval before being submitted to the immigration authorities

PAGE 118

102 of a third country (usually Canada, the USA, Australia, or a West European country) who will take the final decision independently from UNHCR. The quotas are very low: in 2001 only 534 refugees, equivalent to 0.1 % of Tanzania's total refugee population, was resettled in a third country (UNHCR 2002b:3). The process is not only frustratingly long, many of the refugees are unaware of the very strict selection criteria and procedures of both UNHCR and the resettlement countries and consider it "their right" as refugees to be resettled. Feeling cut off from the rest of the world for lack of modern means of communication and tired of waiting, a considerable number of refugees leave the camps and head for Dar es Salaam, expecting in vain, that it will be easier to "push their case through" by knocking on the doors of the UNHCR's Head Office. Apart from these resettlement-seekers, UNHCR Dar es Salaam also receives new arrivals, referring them in the first instance to the Refugee Department in accordance with the 1998 Refugee Act. The majority of these refugees, if they accept to go to the designated area to which they are assigned, will be assisted by Umati to make the three day train ride covering the 1,200 kms between Dar es Salaam and Kigoma. 2.2.2.2 Umati Since 2001, UNHCR has subcontracted Umati to administer the assistance package to the officially registered urban refugees. Before 2001, the Tanzania-Mozambique Friendship Association [TAMOFA] was the implementing partner for urban refugees, but years of corruption charges by the refugees towards TAMOFA staff and its mismanagement of UNHCR's funds led to the decision in 2001 to continue the urban refugees program with the Chama cha Uzazi na Malezi Bora Tanzania (Organization for Reproduction and Family Planning) [Umati] instead. Umati is the oldest and largest NGO

PAGE 119

103 in Tanzania and it is also one of UNHCR's implementing partners in the refugee camps, with its activities concentrated mainly in the health sector. In Dar es Salaam, the Umati office concerned with urban refugees consists of a Project Director (who is also a Medical Doctor), a Senior Counselor, a Social Services assistant and several support staff. They are responsible for monitoring the evolution of the medical cases, who make up the bulk of the urban refugee caseload, and for following up their medical treatment at Muhimbili Hospital or other medical facilities in Dar es Salaam. Apart from the medical follow-up, Umati ensures that the assisted urban refugees are lodged in four different guest houses located in the Ilala and Buguruni wards. Twice a month, refugees are summoned to come to the Umati office in Upanga to collect the living allowance for themselves and any family members residing with them: TS 2,000 ($2) per day per adult and TS 1,000 ($1) per child of under 12 years of age. Payments are disbursed only after explicit approval by the UNHCR office of the list of names of people to be paid. While each refugee receives only $60 per month cash in hand, the monthly cost to UNHCR for one urban refugee currently amounts to approximately $40070 because of the financing of the whole UMATI infrastructure.71 2.2.2.3 Other organizations Caritas is one of the organizations that occasionally assist urban refugees (on condition that they have a letter from UNHCR confirming their refugee status) yet does not receive any funds from UNHCR. During the Arusha Peace Negotiations on Burundi, Caritas financially assisted one particular group of Burundi politicians and their families, 70 Personal communication from the UNHCR Office in Dar es Salaam.

PAGE 120

104 who had been transferred to Dar es Salaam for security reasons by the Tanzanian Ministry of Home Affairs in the wake of the assassination in February 2000 of the brother of a senior Burundi opposition politician (see earlier).72 As soon as the Burundi Government of Transition was installed, on November 1, 2001, Caritas discontinued the assistance and the Refugee Department no longer extended their biweekly residence permits for Dar es Salaam. However, this group of Burundi refugees disregarded the recommendation that they should return to the camp awaiting repatriation, claiming continued insecurity both in the camps and in Burundi, and joined the tens of thousands of clandestine urban refugees trying to cope on their own. Apart from the exceptional assistance given to this particular group of Burundese refugees, Caritas occasionally assists refugees who decide to go (or return) to one of the refugee camps in Western Tanzania, but who are too destitute to do so. The Caritas leitmotiv is that it is preferable to assist those refugees unable to survive in Dar es Salaam to go to one of the camps, and thus avoid creating another "social case" potentially up to mischief out of sheer poverty and despair. After ascertaining that the request is genuine (i.e., that this person is not one of the many merchants and petty traders commuting between Dar es Salaam and Kigoma on business) and clearance is obtained from the Refugee Department who issues the travel permit, Caritas provides the train ticket and some pocket money for food during the three day train ride. 71 At an average of 200 urban refugees, the total budget easily reaches 1 million dollars per year. In comparison, the infrastructure for 500,000 camp refugees requires only 24 million dollars. This is US$48 per person annually or 1% of the budgetary cost of an urban refugee. 72 Under international law, refugees are not allowed to engage in political activities in the country of asylum, lest they loose their refugee status. However, in the case of the Burundi Peace Negotiations, the Tanzanian government decided it was in their best interest to facilitate the participation of the opposition leaders in order to arrive to a peace agreement and refugee repatriation as soon as possible. Interview Nzuki, Refugee Department at the Tanzanian Ministry of Home Affairs, June 30, 2002.

PAGE 121

105 While Tanganyika Christian Refugee Services [TCRS]73 has been active in the refugee settlements in Western Tanzania for decades, where they are one of UNHCR's implementing partners, in Dar es Salaam there are no budgetary provisions for any assistance program for urban refugees. However, one of the smaller TCRS offices in Mikocheni, receiving around 20 refugees twice a week, occasionally distributes material aid, such as clothes or shoes, to refugee mothers with children on purely humanitarian grounds, while other refugees are often counseled at length to go (back) to the camps. Both the Legal and Human Rights Center and the Amnesty International office in Dar es Salaam occasionally receive visits from (mainly Congolese) refugees seeking not material aid, but legal counseling. Their complaints pertain specifically to the slow bureaucracy at the UNHCR branch office, when for example refugees who have acquired sponsorship in a third country, be it for studies or residence purposes, approach the UNHCR office to request a so-called "Protection Letter" confirming their refugee status, and subsequently are being kept waiting for several months. It is the understanding of many refugees that once sponsorship is acquired, the only thing standing between them and a new life in a third country is this particular document issued by UNHCR. Often, it is all too easily forgotten that the immigration authorities of the third country in question have to approve a visa first. In addition, once the visa issued, when it concerns refugees, the third country through their respective Embassies will request UNHCR to arrange for the necessary travel formalities. The complexity of the whole process is often underestimated by the refugees eager to start a new life "elsewhere," and the length of its duration creates a lot of frustration. Hence, their request for assistance to the Legal and 73 While a local NGO in name, TCRS receives the bulk of its funding and a limited number of international staff members from the Lutheran World Federation.

PAGE 122

106 Human Rights Office and Amnesty International office for "infringement of their rights," while in fact, other than alerting the UNHCR office of the refugees' frustrations, there is little, if anything, these offices can actually do.74 2.2.3 The Social Context As will be documented at length in chapter 3, the majority of urban refugees residing in Dar es Salaam have little, if any contact with one of the organizations described in the previous sections. However, each one will have contacts with the Tanzanian inhabitants of Dar es Salaam, as their neighbors, landlords, shopkeepers, fellow passengers in the bus, the parents of the friends their children play with in the streets, etc. After interviewing a limited number of the refugees, I felt that Tanzanian attitudes and prejudices played an important role in the refugees' social lives and in the process of their adaptation, albeit integration, to life in Dar es Salaam. I subsequently developed a survey to be administered to a sample of 216 Tanzanian Dar es Salaam residents gauging their perceptions and opinions on a limited range of topics related to refugee matters and the cross-cultural confrontations of everyday life, with a particular emphasis on the different perceptions of refugees men and women by nationality, i.e., Congolese, Burundese and Rwandese.75 I would like to emphasize at this point that the purpose of this survey was not to "discover" the personalities or characteristics of the three largest refugee groups in Tanzania. The reason for administering this set of questions to a sample of Dar es Salaam residents was to explore and document the images and public opinions circulating 74 Interviews with the Legal and Human Rights Center on November 28, 2001 and Amnesty International on June 21, 2002. 75 A copy of the survey form in English and in Swahili can be found in appendix b.

PAGE 123

107 on refugees. Popular perceptions are often based on hearsay, gossip and rumors, but possibly also on personal experiences by individuals, or singular events published in the national and/or local media. Whatever their source, they constitute an important part of the social context that the urban refugees have to cope with, because human existence is not only about economic survival. With regard to the particular situation in Dar es Salaam, the urban refugees are forced to live clandestinely and to keep a low profile and thus not inclined to form some type of supportive communities with other compatriots. As such, the daily interactions with local Tanzanians, including those with officials or other staff members at governmental as well as non-governmental organizations, gain a very prominent role in the reconstruction of their social worlds and networks. The existence and nature of these public images and stereotypes influence the social integration of urban refugees and subsequently influence the process of coping with the situation of forced migration. In order to have a sample as representative as possible for the Dar es Salaam population,76 the following five parameters were defined in consultation with the Department of Sociology at the University of Dar es Salaam: gender, religion, age category, ward of residence and level of education. The sample of 216 individuals consists of as many men as women, and as many Muslims as Christians. Three age categories were established: 25 years or younger, between 26 and 45 years of age, and older than 45. Considering it unfeasible to establish the socio-economic status of each respondent within the time frame set out for this survey, I decided that it would be useful that the same number of informants be interviewed in three wards of different socio76 For lack of a recent census, as the last census took place in 1988.

PAGE 124

108 economic status: Mikocheni where most residents belong to the higher socio-economic strata, Sinza representing the middleground, and Manzese, one of the more populous wards with residents of a lower socio-economic status. The last parameter was the level of education: out of every 6 persons in the sample (e.g., Christian women older than 45 residing in Sinza) two are secondary school graduates (possibly university graduates), three completed primary but not secondary school, and one out of every six never completed primary school. All the interviewing was done in Swahili by an experienced Tanzanian researcher from the University of Dar es Salaam.77 Locating respondents was done as randomly as possible, namely by walking around in the streets of the three wards (Mikocheni, Sinza and Manzese)78 and addressing people in the markets, in the shops, sometimes in an office, etc. Regarding the level of representativeness of a sample of 216 persons for a population of 3 million people, according to statistical theory, we can say at a 95% confidence level that the margin of error in estimating the frequencies resulting from the survey are no greater than 7% (Agresti and Finlay 1997: 135-138). For example, if the results of the survey show that out of 216 respondents, 122 (or 56%) have personal and/or regular contact with refugees, we can deduce with 95% confidence that the percentage of Dar es Salaam residents having personal and/or regular contacts with refugees lies between 49 and 63% (or 56 +/-7%).79 77 The researcher being a woman mzee (i.e., of the age where one has adult children) no cultural sensitivities hampered her interviewing both women and men of all ages. 78 On map 1-7, Mikocheni is situated between Mwenge and Oyster Bay. 79 Building on this fact, I decided to consider only variations of more than 14% across subgroups of the sample significant enough to mention and/or discuss in the text.

PAGE 125

109 A sixth parameter of relevance for this type of survey could have been the region of origin of the respondent, because it is highly likely that this would influence the probability of the respondents' being familiar at a personal level with refugees and their attitudes towards them. This parameters was not pre-determined as it would have rendered administering the survey and locating respondents extremely time-consuming, in addition to the fact that there exist no recent or reliable estimates of the proportions of the population of Dar es Salaam with respect to their region of origin. From the results of the survey, it appeared later on that 35% of the respondents were from the Coastal Region, 20% from one of the Northern provinces, 18% from Central Tanzania, 14% from the South and 13% from one of the Western regions. Predictably, 75% of those from the Coastal region were Muslim, whereas 80% of the Northern Tanzanians in the sample (mainly from Moshi and Arusha) were Christian. Regarding the other regions, the proportion of Muslims is between 38% (West) and 50% (Central), tallying with the decreasing influence of Islamization as one travels from the Coast further inland to Lake Tanganyika.80 While the first questions of the survey gauge at the level of personal and/or regular contacts between the Tanzanian and the refugee population in Dar es Salaam (Table 2-1 and 2-2), the proportion of "don't know" answers to the next questions possibly functions as an indicator of the (in)visibility of certain categories of the refugees, by gender and by nationality. 80 The sizes of the different subsamples are listed in Table 2-1 at the end of this chapter.

PAGE 126

110 2.2.3.1 Level of Contact On average 56% of the respondents in the sample have personal and/or regular contacts with refugees as friends, relatives, neighbors, etc. (Table 2-1). Looking at this percentage a little closer and disaggregating it by gender, religion, age, socio-economic status, education level, it becomes clear that there is a lot of variation in the ratio of respondents having personal and/or regular contacts with refugees, ranging from minimum 22 to maximum 81%. Unsurprisingly, the level of personal contacts with refugees appears to be most heavily influenced by the region of origin of the respondents (minimum 17 and maximum 100%). Overall, most striking is the consistently higher percentage of Tanzanian women having personal contacts with refugees (70%) compared to Tanzanian men (43%). In any of the categories listed, the higher percentage of respondents having regular contacts with refugees are women. In the same line, we see that among the respondents originating from Western Tanzania, where the bulk of the refugee population lives, 75% of the men and 100% of the women are in regular contact with refugees. At the same time, only 29% of the men and 50% of the women from Tanzania's remote and economically underdeveloped regions in the Southern and Central parts of the country, say they have personal and/or regular contact with refugees.81 No significant variations in the level of contact appear between Muslim and Christian populations (58 versus 55%). However, substantially larger proportions of older people (63%) and middle-aged respondents (60%) reportedly have contacts with refugees compared to the youngest respondents (43%). The disaggregated percentages 81 Sample sizes for these populations are rather small (n=29 and 31 respectively), nevertheless, the outcomes appear consistent and logical.

PAGE 127

111 for the different levels of education are consistent with the trend observed in comparing residents in wards of different socio-economic status. The more educated and of a higher economic status, the higher the percentage of respondents confirming personal/regular contact with refugees: 65 and 64% respectively compared to 47% of the respondents below primary school level, and 44% of Manzese residents. Whether this is the result of a higher mobility and cosmopolitan attitude among more educated Tanzanians and those of a higher socio-economic status or, of an on average higher level of education and socio-economic status among the refugees in Dar es Salaam facilitating contact with their peers, is impossible to deduce from these findings. When asked which nationality represents the largest group of refugees in Tanzania, no significant variations appear: roughly as many Tanzanian men and women quote either one or the other nationality as the largest refugee population in the country (Table 2-2). With regard to the refugees' presence in Dar es Salaam, however, Congolese are cited by significantly larger numbers of respondents than either Burundese or Rwandese. This perception may be in line with the actual situation on the ground or reflecting the degree in which certain refugee nationalities manage or intend to blend in and keep a low profile. Along the same line, regarding the perceptions of the professional occupation of the urban refugee by gender and nationality (Table 2-3), the percentage of people who have no opinion whatsoever on the matter is much higher where it concerns Burundese men and women (38 and 40% respectively) and Rwandese (34 and 36% respectively) than when it does Congolese (11 and 13%). This trend is again confirmed when looking at the percentages of people who replied "I don't know" to questions asking about likable and dislikable characteristics of the refugees by nationality, and about the perceived differences between men and women (Table 2-6). For each question, Congolese

PAGE 128

112 consistently appear the best known, compared to the two other nationalities. These results are all the more striking considering that historically larger numbers of Burundese and Rwandese refugees have been given refuge in Tanzania since the early 1960s and 1970s, whereas the Congolese came to Tanzania in large numbers only since the late 1990s. 2.2.3.2 Intermarriage The issue of intermarriage between Tanzanian citizens and refugees in Tanzania is a contentious one, and, according to some, the legislature on this point is gender discriminative. Tanzanian law stipulates that if a Tanzanian citizen can show proof to the Ministry of Home Affairs that he/she can provide for his/her spouse of another nationality, that spouse is automatically issued with a residence permit. This principle is generally applied without a problem where Tanzanian men marry non-Tanzanian women. However, in practice, it is almost impossible for a Tanzanian woman to demonstrate convincingly that she can financially provide for her non-Tanzanian husband. There have been cases in the past where the refugee husband and his Tanzanian wife were given notice by the Ministry of Home Affairs to go to one of the refugee camps. Recently, human rights lawyers have taken these cases to court, accusing the Tanzanian government of gender discrimination and "outdated legislation" (The Guardian, May 25, 2000, 1). The decision was still pending at the time of writing and the issue of intermarriage with refugees remains sensitive. Table 2-4 lists the replies to the questions "How would you feel if your daughter/son were to marry a refugee man/woman?" disaggregated by gender, religion, age, socio-economic status, education level, and having personal contacts with refugees. On average, two thirds Dar es Salaam residents would dislike their daughters (67%) or

PAGE 129

113 sons' (61%) to marry a refugee. Across categories of gender, religion, age, socio-economic status, educational level or having/nor having regular contacts with refugees, there is no significant difference between the level of intolerance against bringing a refugee man or a refugee woman into the family. However, whereas more than one in three men would accept their daughter's (34%) or son's (42%) marriage to a refugee, only one in every six women would (10% and 15% respectively). Interestingly, acceptance of intermarriage appears to be lower among Tanzanians regularly in touch with refugees (particularly for daughters, 16%) than among respondents that do not have regular and/or personal contacts with refugees (31%). This finding is very likely related to the numerous stories abound about refugee men abandoning their Tanzanian wives and children, obliging the mother's family to take care of them. Better aware of the vulnerable position of women, especially those with children and abandoned by the father and husband, Tanzanian women are twice as likely than men to be opposed to their daughters' marrying a refugee rather than merely disliking the idea.82 2.2.3.3 Refugee policy Another contentious issue regarding refugees, and urban refugees in particular, concerns the possibility of obtaining a residence and work permit for Dar es Salaam instead of being subject to forceful transportation to one of the refugee camps as the 1998 Refugee Act stipulates. Despite the fact that the policy to send all refugees to designated areas is rather recent and the advocacy by Tanzanian academics and UNHCR of an 82 I was told by informants that this reluctance includes not only refugees, but any non-Tanzanian, because of the impossibility, in case of abandonment, to contact the family-in-law and demand some type of support for the children.

PAGE 130

114 approach "of letting refugees take care of themselves and of letting them contribute to the national economy" (UNHCR 2002a: 5), there does not appear to be a political will, possibly for electoral reasons, to open a debate on the topic. Not surprisingly then, three quarters of the respondents, regardless of their level of personal/regular contacts with refugees (74 versus 75%), are of the opinion that refugees should be forced to go to the camp and depend on international aid (Table 2-5), a trend consistent across all categories of respondents. 2.2.3.4 Likes, dislikes and differences A glance at the percentages of respondents who replied "I don't know" to the questions probing them for characteristics of refugees that they either liked or disliked, reveals an all too common human characteristic, namely that it is easier to come up with a negative comment about "others" (62%) than a positive one (18%) (Table 2-6). This trend is confirmed in the replies to questions comparing Tanzanian men and women with men and women of each of the three refugee nationalities where only between 4 to 19% of the respondents came up with a comparison in favor of the other nationality. The only exception to this are Burundese women: 64% of the 66 Tanzanian respondents compared them favorably to Tanzanian women describing them as polite, friendly, quiet, stable (wapole, wavumilivu) and hard workers (Table 2-9). Disaggregating the likes (Table 2-7), dislikes (Table 2-8) and perceived differences (Table 2-9) by refugee nationality, it becomes clear that each of the three nationalities involved has a distinctly different image among Dar es Salaam's Tanzanian residents, allowing for some overlapping between Burundese and Rwandese refugees. Congolese are liked because they are considered charming (44%), and friendly and generous (52%),

PAGE 131

115 whereas Burundese are appreciated as hard workers (68%) and Rwandese are commented for valuing education (36%) and religion (25%) (Table 2-7). Probed for the characteristics disliked, between 10 and 19% of the Tanzanian respondents criticize refugees for their involvement in crimes, the only characteristic that more than 1 in 10 respondents consider shared by all three of the refugee nationalities (Table 2-8). The majority of respondents have a distinctive image of the refugees by nationality: Congolese are perceive as cheating conmen (45%), Burundese as aggressive troublemakers (52%), and Rwandese as tribalists (53%). Interestingly, the characteristics perceived to be typical for the Congolese, i.e., cheating (45%), sexual misbehavior (15%), and a penchant for the easy life (10%), are rarely mentioned with regard to the Burundese and Rwandese. In the same line, only 7% of the respondents mention aggressiveness (considered typical of Burundese) as a characteristic commonly found among Rwandese, while only 13% indicate racism and tribalism (perceived a common characteristic of Rwandese) as typical of the Burundese. The percentage or respondents ascribing both qualities to Congolese are even fewer (4 and 0% respectively). In short, these findings confirm the existence of distinctive public images among Dar es Salaam's Tanzanian residents for each one of the three different refugee nationalities. Further illustration of this observation is to be found in the perceived differences between Tanzanian men and women and both genders of Congolese, Burundese and Rwandese origin (Table 2-9). While Congolese men are perceived to mainly differ from Tanzanian men with regard to their penchant for expensive clothing and a luxurious life (44%) and to a lesser extent their sexual (mis)behavior (24%), Congolese women are considered different from Tanzanian women for the very same reasons but at different

PAGE 132

116 rates: 25 and 40% respectively. There is little difference between the general perception of Rwandese men and women, who are thought of as proud and keeping to themselves (52 versus 61%). The most striking difference is found between Burundese men and women: while Burundese men are perceived as different from Tanzanian men due to aggressive and hostile behavior, Burundese women are the only category to largely compare favorably with their Tanzanian counterparts. 2.2.3.5 Summarizing findings The first finding of this survey among Dar es Salaam's Tanzanian residents of all walks of life is that the level of personal and/or regular contacts with refugees is much higher among women (70%) than among men (43%), a trend found to be consistent across religion, age, socio-economic status, education and region of origin (Table 2-1). Secondly, the findings of the survey confirm the current refugee-unfriendly climate among the Tanzanian residents of Dar es Salaam: while three quarters of all respondents feel all refugees should forthwith be sent to one of the refugee camps in Western Tanzania (Table 2-5), two thirds expressed their dislike to the idea of having a refugee son or daughter-in-law (Table 2-4). A third important finding purports to the distinctively different public images and stereotypes circulating among Dar es Salaam's Tanzanian population (and probably beyond) regarding the three refugee nationalities (Tables 2-7, 2-8 and 2-9). Common images of refugees as armed thieves and troublemakers exist, but are largely overshadowed by the more general consensus of "charming but dishonest Congolese," "hardworking but aggressive Burundese" and "religious and educated but tribalist Rwandese."

PAGE 133

117 2.3. Conclusion This chapter gave an overview of the historical facts leading up to the events of the 1990s that are at the root of the huge influxes of refugees from the Great Lakes region into Tanzania. In addition, I have tried to give the reader a first sense of the refugee situation in Tanzania, i.e., its effects on local structures in Western Tanzania, the major problems in the refugee camps and the existing infrastructures relevant to refugees in Dar es Salaam, including circulating stereotypes and public images. The stage set, the following chapter endeavors to familiarize the reader with certain aspects of the refugee population in Dar es Salaam. Who are they? Why do they come to Dar es Salaam? What are their hopes and ambitions with regard to their current situation?

PAGE 134

118 Table 2-1. Level of Contact with Refugees (in %). Categories Women in % Men in % Total in % Sample Size n= Religion Christian 69 41 55 108* Muslim 72 44 58 108* Age Category <25 64 31 47 72* 26-45 75 44 60 72* 46< 72 53 63 72* Wards of Different Socio-Economic Status Mikocheni Higher 72 56 64 72* Sinza Middle 81 42 61 72* Manzese Lower 58 31 44 72* Education Level Below Primary 72 22 47 36* Completed Primary 69 39 54 108* Secondary and Beyond 73 58 65 72* Region of Origin Coastal 69 49 60 74 Northern 57 50 52 44 Central 65 17 42 38 Southern 50 29 39 31 Western 100 75 93 29 Total Average 70 43 56 216 : Pre-established sample levels.

PAGE 135

119 Table 2-2. Tanzanian Perceptions on Prevalence of Refugee Nationalities (in %). Which Nationality represents the largest group of refugees . In Tanzania ? In Dar es Salaam ? Congo Burundi Rwanda 29 27 32 49 17 24 Other Don't Know 1 11 0 10 E.g., 17% of informants think that the Burundese refugees represent the largest group in the total refugee population of Dar es Salaam. Table 2-3. Tanzanian Perceptions on Refugee Occupations (in %) Congolese Burundese Rwandese Professions Men Women Men Women Men Women Trading (biashara) 29 53 41 43 53 53 Music 33 21 2 0 0 0 Paid Employment 1 1 9 6 7 5 Tailor 15 2 1 0 1 2 Barber/Salons 7 9 1 0 2 1 Missiontown 4 1 1 1 1 0 Agriculture/Herding 0 0 7 7 1 1 Other 0 0 0 2 1 1 Don't Know 11 13 38 41 34 37 E.g., 15% of Survey respondents perceive that Congolese refugee men most commonly exercise the profession of tailor, while 7% perceive that most commonly Burundese refugee men and women profess an activity related to agriculture.

PAGE 136

120 Table 2-4. Tanzanian Perceptions on Intermarriage with Refugees (in %). How would you feel if . your daughter were to marry a refugee man ? Your son were to marry a refugee woman ? I would . Dislike In % Accept In % Not care In % Dislike In % Accept In % Not care In % Gender Women 72 10 17 69 15 15 Men 62 34 2 53 41 6 Religion Christian 65 20 12 59 27 13 Muslim 69 24 7 63 30 7 Age <26 60 25 13 54 32 14 26-45 65 21 13 58 29 13 45< 76 21 3 72 24 4 Wards' Socio-Economic Status Mikocheni Higher 69 21 10 63 28 10 Sinza Middle 72 14 10 63 22 14 Manzese Lower 60 32 8 58 35 7 Education Level Below Primary 86 9 6 78 11 8 Completed Primary 58 31 10 56 35 9 Secondary and Beyond 72 17 10 61 26 13 Having Personal Contacts with Refugees ? Yes 71 16 12 64 24 12 No 62 31 5 57 34 7 Average 67 22 9 61 28 10 E.g., 63 % of the informants who reside in Sinza would dislike the fact of their son marrying a refugee woman.

PAGE 137

121 Table 2-5. Tanzanian Perceptions on Refugee Policies (in %). Policy Option Refugees should be allowed to exercise their profession and be given a residence permit Refugees should be forced to go to the camp and depend on international aid Knows Refugees Personally ? Y (n=122) N (n=94) Total Sample Y (n=122) N (n=94) Total Sample Gender Women 17 25 19 78 72 76 Men 24 13 18 70 76 73 Religion Christian 27 16 22 68 76 71 Muslim 13 18 15 81 73 78 Age <26 17 13 15 79 76 78 26-45 23 21 22 70 76 72 45< 18 19 18 76 70 74 Wards of Different Socio-Economic Status Mikocheni Higher 17 11 15 72 77 74 Sinza Middle 11 21 15 84 61 75 Manzese Lower 34 18 25 66 83 75 Education Level Below Primary* 24 16 19 77 53 64 Completed Primary 26 20 23 67 78 72 Secondary and Beyond 11 12 11 83 84 83 Total Average 20 17 19 74 75 75 E.g., 78% of the respondents residing in Sinza and having personal and/or daily contacts with refugees feel that they should all live in the camps and depend on international aid. 31% of the respondents below the primary school level, and not having regular or personal contact with refugees did not state an opinion regarding which refugee policy they favor.

PAGE 138

122 Table 2-6. Level of "Don't know" Replies (in %). Question Congolese Burundese R wandese Average Which quality or characteristic do you like most among . ? 78 84 83 82 Which quality or characteristic do you dislike most among . ? 27 49 38 38 What do you see as the main difference in character or lifestyle between Tanzanian men and . men ? 35 65 52 51 What do you see as the main difference in character or lifestyle between Tanzanian women and . women ? 38 69 54 54 E.g., When asked for the most disliked characteristic of Burundese refugees, 49% of the survey respondents did not have an opinion. Table 2-7. Refugee Characteristics Liked (in %) Quality/Characteristic most liked Congolese (n=48) Burundese (n=34) Rwandese (n=36) Mix easily (wanachanganyika na watu); Friendly (wana upendo); Generous (wakarimu). 52 15 11 Charming (wachangamfu); Smart (wastaarabu) 42 0 6 Hardworking (wafanyakazi hodari) 0 68 14 Calm (wanatulia); Patient (wavumilivu) 0 12 3 Value education (wanapenda elimu) 0 0 36 Religious (wanapenda dini) 0 0 25 Other 6 6 6 E.g., 42% of the respondents replying question 1 of table 2-7 said that the characteristic/quality of Congolese that they liked most was their "charm" and "being smart."

PAGE 139

123 Table 2-8. Refugee Characteristics Disliked (in %). Quality/Characteristic of . most disliked Congolese (n=157) Burundese (n=111) Rwandese (n=135) Armed thieves (wezi wa silaha); robbers (ujambazi); smuggling (uporaji); law breakers (wahalifu) 19 10 17 Quarrelsome (ugomvi); rude (wakorofi); disrespectful (hawana heshima); bad characters (tabia mbaya) 5 17 3 Cheating (matapeli); cunning behavior (ujanja); conmen (wajanja); liars (walaghai) 45 2 3 Aggressive (wakali, hasira); sadistic, cruel (wakatili); hot tempered (hawatulii); troublemakers (fujo, vurugu) 4 52 7 Racism (ubaguzi); tribalism (ukabila) 0 13 53 Don't mix with others/Tanzanians (kujitenga); self-centered (ubinafsi); proud (majivuno) 2 6 11 Indecent dressing (mavazi yasiyo na heshima); sexual misbehavior (wahuni, kubaka); unstable marriages 15 0 2 Like easy life, luxury (wanapenda starehe, anasa); use heavy make-up (kujipamba); bleach skin (kujichubua) 10 0 1 Other 0 0 3

PAGE 140

124 Table 2-9. Perceived Differences between Tanzanians and Refugees (in %) Perceived Differences between . Congolese Burundese Rwandese and Tanzanian . Men (n=140) Women (n=134) Men (n=75) Women (n=66) Men (n=104) Women (n=99) Drugs (bangi); armed thieves (wezi wa silaha) 2 0 1 0 2 0 Show-off (maringo); disrespectful (hawana heshima); argumentative (wabishi) 7 8 5 2 4 5 Cheating (matapeli); dishonest (waongo); fraud (walaghai) 5 3 8 0 4 3 Like fighting, heavy drinking, Beat their wives, hostile 5 1 41 3 12 0 Racism (ubaguzi); tribalism (ukabila) 0 0 1 3 7 6 Don't mix with others/Tanzanians (kujitenga); self-centered (ubinafsi); proud (majivuno) 9 13 16 25 52 61 Indecent dressing (mavazi ya kihuni, marembo); sexual misbehavior (wahuni, kubaka) 24 40 4 3 1 5 Easy life (starehe); dress up/ expensive clothing (kujipamba); bleach skin ( kujichubua) 44 25 3 0 0 0 Positive Comments 4 8 15 64 18 19 Other 0 2 6 0 2 1

PAGE 141

CHAPTER 3 HEADING FOR DAR ES SALAAM Tukaendelea kukimbia, tukajionana kama vile jamaa ya Israel. "We continued to flee, we compared ourselves to the people of Israel." Congolese refugee, Dar es Salaam, September 29, 2001. Having set the stage in the preceding chapters, the time has come to meet some individual urban refugees of Dar es Salaam. Based on the findings of the survey data as well as the testimonies of the Congolese, Burundese and Rwandese women and men, young and old, slowly a profile emerges. At the same time, variations across gender, nationality and age surface, some easily explainable by the different contexts of the conflicts in the Great lakes countries, others appearing less clear at this point and which I intend to contextualize at a later stage. The sections in this chapter explore the composition of the urban refugee sample in terms of demographics as well as their motivations to leave home, to avoid the refugee camps and to head for Dar es Salaam. 3.1 Localizing Voices After a brief historical overview of the refugee flows from the Great Lakes region into post-independence Dar es Salaam, I move on to the current situation by first presenting the sample of this study and subsequently zooming in on individual accounts. The quantitative overview of the three hundred surveyed respondents1 aims to give the reader an impression of the demographics and other types of information about Dar es 1 The methodology used and selection criteria applied are described in chapter 1. 125

PAGE 142

126 Salaam's refugee population, such as: the year of their arrival in the country, length of residency in the city, level of contact with the UN refugee agency, etceteras. During my research, I interviewed more than forty informants out of the three hundred respondents. However, for the purpose of individualizing the material presented, I decided to concentrate on the ethnographic accounts of a limited number of persons only. Twelve men and women were selected as a function of their representativeness of the larger sample. They are between the ages of 22 and 55, and each of the three nationalities under study is represented by two men and two women.2 I will introduce each of the twelve protagonists separately at the end of this introductory section. 3.1.1 Refugees in Post-Independence Dar es Salaam A recently published study on Nyerere's role in the Cuban intervention in Congo in the 1960s, recalls how by 1964 the CIA was expressing its concern that Dar es Salaam "has become a haven for exiles from the rest of Africa. It is full of frustrated revolutionaries plotting the overthrow of African governments, both black and white" (Kelley 2002: Internet). Not taking into account the numerous exiles from South Africa, Zimbabwe (then-Rhodesia), Mozambique and Angola at the time, this quote is more pertinent to the Congolese revolutionaries than the Burundese or Rwandese refugees in light of the Katangese Secession wars of the early 1960s.3 After the assassination of Lumumba in 1961, several of his political allies took up arms against the newly established Mobutu regime with the help of China and the Soviet Union. Ultimately 2 These accounts are complemented occasionally by short excerpts from the interviews with other informants, who are identified by age, gender and nationality only where relevant. 3 Katanga, now Shaba province, in Southern Congo is a region extremely rich in minerals and other natural resources where, at the time, large foreign mining companies (e.g., the Gcamines) were very influential political players.

PAGE 143

127 defeated by the US backed Congolese government troops in 1965, many of these rebels took refuge in Tanzania, and Dar es Salaam. The most famous among them was Laurent Kabila,4 who ran a business in Dar es Salaam selling gold mined in Eastern Zaire from the early 1980s onwards (FOF 1997: Internet). But there were others, such as Ernest Wamba dia Wamba, also an opponent to the Mobutu regime, who taught History at the University of Dar es Salaam for many years.5 Many other political opponents of the Mobutu regime found their way to Dar es Salaam over the following decades, but never in large numbers.6 Only since the full scale wars of 1996 and 1998 have there been large influxes of Congolese refugees into Tanzania, a situation very different from the Burundese who have been seeking refuge in the country in several waves from 1972 onwards. During the 1970s and 1980s, these Burundese refugees were relocated to three settlement areas in Western Tanzania, where they were allocated land by the Tanzanian government and provided agricultural tools by the international aid organizations. These refugees were, and still are to a certain extent, mainly villagers and agriculturists, due to the discriminatory government policies against the Hutu in Burundi in the areas of education and professional training.7 A recent study shows that in the early 1990s, however, a 4 His son, the current president of the Democratic Republic of Congo, Joseph Kabila, was raised and educated in Tanzania. 5 In 1998, Dr. Wamba dia Wamba became the head of the Rassemblement congolaise pour la dmocratie, (RCD) that took up arms against the then president Laurent Kabila. 6 No study was conducted on the Congolese refugee population in Dar es Salaam, or even in Tanzania, until 1993 by Benyi Bakaji. This Congolese author was refused a research permit to do research among her compatriots in the refugee camps, preventing her to venture out far from theoretical considerations and a one page survey administered to a group of 50 Congolese young men in Dar es Salaam (Bakaji 1993). 7 The 1972 genocide, in which hundreds of thousands of people died, was, in fact, an attempt by the Tutsi government to eliminate the Hutu intelligentsia, as were subsequent outbursts of violence in the 1980s and 1990s. See also chapter 2.

PAGE 144

128 growing number of disenchanted youth who grew up in one of the Tanzanian rural settlements, found their way to Dar es Salaam in search of urban excitement (Sommers 2001; 1994; 1993), in addition to growing numbers of new arrivals straight from Bujumbura and other urban centers in Burundi. Whereas the Burundese refugees have always constituted the majority of the refugee population in Tanzania during the past decades, the very first refugee influx newly independent Tanzania experienced came from Rwanda in 1961. In the wake of the 1959-1961 "social revolution," whereby the Belgian colonial authorities decided almost overnight that the Hutu were to rule the newly independent country, a Tutsi exodus took place to the neighboring countries, and tens of thousands of Rwandese Tutsi found their way to Tanzania (Daley 2001). These refugees were the first to be relocated in the Tanzanian settlement schemes, a policy that would later be applied to the much larger Burundi influxes of the 1970s.8 The striking difference between these two groups of refugees is that a far larger proportion of Rwandese refugees applied for Tanzanian citizenship, when Nyerere's government issued the 1980 citizenship decree. In the course of one decade (1980-1990), 10,047 Rwandese9 applied for and received Tanzanian citizenship compared to only 61 Burundese.10 Acquiring Tanzanian citizenship opened many doors for these former Rwandese refugees, now no longer confined to the settlement areas in Western Tanzania. 8 In 1977, Tanzania hosted a total of 35,000 Rwandese compared to 130,000 Burundese (Van Hoyweghe 2001). 9 However, not all of these Rwandese applicants were refugees (Gasarasi 1990). 10 One could argue that the length of their stay in Tanzania was a factor in the decisions made by refugees of either nationality. However during the next decade 1991-2000, only 550 more Burundese applied for Tanzanian citizenship versus 179 Rwandese (Based on statistics from the Tanzanian Ministry of Home Affairs Courtesy of Mrs. Mukuma, Head of the Citizenship Department).

PAGE 145

129 Many headed for the urban centers, e.g., Dar es Salaam, and established successful careers as university professors or business men, while others obtained influential positions in the government and security services. In 1994, in the wake of the genocide and the subsequent regime change in Kigali, the majority of these Tutsi Rwandese returned home, leaving their Tanzanian lives of the past three decades behind them. Of the one million Rwandese Hutu, who, accused of participating in the genocide, sought refuge in Tanzania in 1994, the large majority was repatriated in 1996. The forced nature of this repatriation, however, encouraged many Rwandese camp refugees to head for Dar es Salaam in search of a safe hiding place from possible future repatriation exercises.11 In short, all three refugee nationalities have a long history in Tanzania, yet it is only recently that increasing numbers have been leaving or avoiding the camps and settlement areas to head for Dar es Salaam,12 for reasons that will be explored further on in this chapter. 3.1.2 Demographics of the Sample As explained in the first chapter, the composition of the sample was based on three parameters: gender, nationality and age. Both genders are equally represented, 150 men and 150 women, as are the three nationalities, 100 individuals each (Table 3-1 at the end of this chapter). The number of people in each age category respond to a large extent to the initial goals of having 30% of respondents in the three youngest age categories (90 11 Only recently, the Tanzanian government decided to revoke the refugee status of all Rwandese refugees still in the country after December 31, 2002. The Tanzanian Minister for Home Affairs declared that any Rwandese refugee not having returned home by the given deadline "should find another country to stay in" (The Guardian, December 23, 2002: 6). 12 During my pre-dissertation research in 2000, government officials, humanitarian workers as well as academics denied that there were any urban refugees in Dar es Salaam "apart from a few Congolese musicians." During 2001 and 2002, however, I did not meet a single official who did not acknowledge the presence of tens of thousands of Great Lakes refugees in Dar es Salaam.

PAGE 146

130 each) and 10% in the category of older people. The second age category, consisting of persons between 26 and 35 years of age, is slightly over-represented (95) whereas both the persons of between 20 and 25 years of age (88) and of between 36 and 45 years of age (86) are slightly underrepresented. In addition, while the number of persons in the oldest age category is approximately on target, 31 instead of 30 persons, women make up its majority: 19 women compared to 12 men. Considering all the efforts that were made to compile a sample of a given composition within in a limited period of time, these differences appear to indicate certain existing tendencies in the larger urban refugee population, namely that it consists of proportionately more persons between 26 and 35 years old, while among refugees over the age of 45 women represent the majority.13 3.1.2.1 Religion, marital status, and children On average, more than half of the respondents are Roman-Catholics (56%), a quarter is Pentecostal (24%) and around one in ten is Muslim (11%).14 No significant differences with regard to religious affiliation appear to exist between men and women or across age categories. Yet, where Congolese count more Pentecostal followers than the two other nationalities (34%), Burundese count distinctly more Muslims (25%) and Rwandese more Roman Catholics (67%). Overall, however, Roman Catholics make up the obvious majority in each of the social categories. Their religious affiliations set the refugees from the Great Lakes region quite apart from the large majority of Dar es Salaam residents who are Muslims. Whether this situation raises any additional difficulties in the adaptation process will be discussed in the next chapter. 13 The margin of error at a 95% confidence level for a sample of 300 is situated between 5 and 6%, as such the following sections discuss or mention only differences between subgroups of the sample that exceed 10%. 14 Statistical information in this section is compiled in Table 3-3.

PAGE 147

131 01020304050607080AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Roman-Catholic Pentecostal Muslim Other Figure 3-1: Religion by gender, nationality and age On average, approximately the same proportion of urban refugees are unmarried (46%) as married (44%), while 10% is either divorced or widowed. Obviously, age plays an important role in one's marital status: whereas the majority of the respondents under the age of 35 years old are unmarried and not living with a partner, approximately half of the persons over 36 years of age are married and living with their spouse. The other half of the two oldest age groups consists of single headed households: partly married and not living with spouse (especially 36 to 45 years olds), partly widowed or divorced (especially those over 45 years of age). 020406080100AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Married, living with spouse Married, not living with spouse Unmarried, lives with partner Unmarried, not living with partner Widowed Divorced Figure 3-2: Marital status and living arrangements Comparing across gender, it appears that male respondents are much more likely to be unmarried and not living with a partner than females (54 versus 33%), while more women than men are widowed (13 versus 2%) or divorced (4 versus 0%). This trend is a

PAGE 148

132 logical result of the gender difference in the average marriage pattern, whereby women usually marry at a younger age than men.15 020406080100AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Childless One child Two children Three children Four and more Figure 3-3: Number of Children in Dar es Salaam On average, one in two respondents (48%) is childless or has none of his/her children in Dar es Salaam, one in three (32%) has one or two children with them, while only one in five (20%) has three or more children residing in Dar es Salaam. Again, age plays an important role in the number of children one has. The overwhelming majority of respondents in the youngest age category are childless (84%) compared to only one in six (13%) in the oldest age category. Of the oldest refugees, 61% have children, usually adults, that reside elsewhere.16 Comparing across genders, significantly more refugee men than women are childless (64 versus 33%), while women are twice as likely than men to have one to two children (41 versus 21%). This trend confirms that, concurrent with prevalent marriage patterns, women start having children at a much earlier age than men. 15 While the inverse association of female education with age has been well documented, it appears that in Africa, a sharp break in the trend occurs only at the secondary school level (Westhoff 1992: 7-8). 16 Of all the respondents with children residing outside of Dar es Salaam, one third has children that remained at home, others reside elsewhere in Tanzania, while one in ten does not know the whereabouts of their children outside of Dar es Salaam. Every one in fifteen respondents has children residing in Canada, Finland or the Netherlands.

PAGE 149

133 3.1.2.2 Education level and occupation On average, one in two respondents has a secondary school education (54%), one in five reached only the primary school level (19%), yet one in six (14%) is a university graduate. 17 As urbanites, the refugees in Dar es Salaam are considerably more educated than their compatriots in the refugee camps: barely 1% of the camp refugees has a university degree, compared to between 8 to 18% of the urban refugees depending on nationality (Table 3-2).18 010203040506070AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< No Education Primary Education Secondary Education Post-Secondary Education Figure 3-4: Education levels Disaggregating the sample data by nationality, the Burundese urban refugees, while more educated than their compatriots in the camps, count significantly less high school and university graduates (39 and 8% respectively) than do the Congolese (63 and 18% respectively) and Rwandese (59 and 16% respectively). Across gender as well, there are differences between categories as men are distinctly more educated than women especially at the post-secondary level where men outnumber women three to one, while 17Statistical information is based on Table 3-4, unless indicated otherwise. 18Based on statistical information from Muyovosi, Mtabila 2, Kanembwa, Karago, Mtendeli, and Nduta camps, representing a total of 95,114 Burundese refugees, and Nyarugusu, Lugufu 1 and Lugufu 2 camps representing a total of 47,332 Congolese refugees. Unfortunately, no similar information for the Rwandese camp refugees was obtainable (Courtesy of UNHCR Dar es Salaam Mr. Kwakye, Senior Program Officer).

PAGE 150

134 twice as many women as men have not reached the secondary school level.19 With regard to age, the data reveal that older respondents have a lower education level than the younger ones. One third of the respondents older than 45 years of age did not reach a primary school level (32%), three times as many as in the two youngest age groups. The fact that the youngest respondents count only half as many post graduates as the respondents between the ages of 26 and 45, is indicative of the condition of universities and schools of higher education in war zones disrupting the usual time frame of one's education, rather than of a reversal of trends. This situation is equally reflected in the high average percentage of respondents who reportedly were students at the time of leaving home (36%) (Fig. 3-5). Of the youngest respondents close to three in four (73%) were students before coming to Dar es Salaam. Only one in twelve (8%), however, were able to continue their studies in exile. 01020304050607080AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Self-Employed No Occupation Trade Paid Employmnet Housewife Student Agriculture Figure 3-5: Employment activities at home before leaving. While, in general, one's educational level is often indicative of one's occupation, in a situation of forced migration, this direct link is blurred by the lack of employment opportunities as a result of the restrictive refugee policies of the asylum country (see chapter 2). The occupations held by respondents at home, before coming to Dar es 19 Even among this generally more educated urban population, the incidence of an average lower level of education among women together with the fact that significantly fewer women in the sample are unmarried, confirms the previously noted inverse association between education and age at marriage.

PAGE 151

135 Salaam, are, in order of importance: students (36%), trade (18%), paid employment (16%), and self-employment (12%). Compared to the professional occupations held by respondents after coming to Dar es Salaam, the proportion of self-employed tripled (39%) to become the most important one. While roughly the same number of people as before engages in trade activities (20%), only half as many as before have paid employment (8%). One in four (24%) contends to have no occupation. 0102030405060AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Self-Employed No Occupation Trade Paid Employmnet Housewife Student Agriculture Figure 3-6: Employment activities after arrival in Dar es Salaam Self-employment in this study includes a wide variety of professional activities. However two-thirds of the self-employed respondents are engaged in only two types of activities in Dar es Salaam. Hairdressing and the braiding of hair20 is the activity exercised by 16% of all respondents, followed closely by the missiontowns (10%). Missiontown21 is a typically Dar es Salaam term, translatable to "middleman" in English. They are persons, usually young men, who basically bring sellers and buyers of any type of goods or services (e.g., landlords and tenants) together and make a living of the commission. Whereas this activity does not require a start up capital (contrary to trade), the person intending to become a missiontown needs excellent communication skills and the ability to make personal connections easily. The remaining third of self-employed 20 Tressage in French, kusuka nywele in Swahili. 21 In French often called commissionaire. There exists no Swahili translation for the term.

PAGE 152

136 respondents earn a living as tailors, carpenters, mechanics, plumbers, etc. Not allowed to start up their own workshop for lack of the necessary legal documents, self-employed refugees (except for the missiontowns whose "office" consists solely of a cell phone) need to locate a Tanzanian-owned shop in their area of expertise, and agree with the owner to work on commission rather than receiving a fixed salary at the end of each month. Other respondents, mostly older women, manage to make a living by preparing mandazi or chapatis, which are local types of pastry, or other food items, such as smoked fish, which they subsequently sell to passerby's in the street. A few respondents self-defined as photographers, preachers or prostitutes. The proportion of self-employed respondents decreases significantly with age: 51% in the youngest age group compared to only 13% of the oldest respondents. Trading activities, on the other hand, include the buying and selling of any type of goods, from precious stones and African art objects (e.g., Central African masks are popular with tourists) to women's wear, including the West and Central African fabrics (such as the basin or the superwax) which are very different in quality, fashion and price range from the ones produced locally in Tanzania (such as the kitenge and the kanga). Whichever the nature of the goods involved, trading activities unavoidably need a certain level of start up capital, undoubtedly the reason why 30 to 39% of the older age categories are able to engage in trade, compared to only 2% of the youngest respondents. Under paid employment, as a category, I included not only jobs in the formal sector (e.g., pharmacist, shop attendant, school teacher or receptionist) but any type of activity for which a respondent receives some type of salary: e.g., housegirls, private teachers, drivers, and even a football trainer. Only 8% of all respondents enjoy some type of paid employment, and there appear no significant differences across gender, nationality or age

PAGE 153

137 groups. Among the unemployed, we find twice as many men as women (31 versus 16%), however when including the percentage of women who self-defined as housewives (12%), the gender difference among unemployed respondents becomes negligible. 3.1.2.3 Arrival in Tanzania, duration of stay in Dar es Salaam and UNHCR registration On average, one in two of the respondents (52%) arrived in Tanzania between 1996 and 1998, the respective years of the first and the second Congolese war (see chapter 2).22 Close to one in three (29%) arrived after 1999, and one in five (19%) before 1995.23 Only 3% of the Congolese respondents arrived in the country before 1996, compared to as many as one in four of the Burundese and Rwandese refugees (28%). These data are in direct correlation with the time of commencement of the conflicts in the respective countries: Burundi 1993, Rwanda 1994 and Congo 1996 (see chapter 2). From 1999 onwards, the intractable situation in the Great Lakes region has ensured a steady stream of refugees from all the three nationalities into Tanzania. 020406080AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< 1999-2001 1996-1998 1992-1995 Figure 3-7: Year of arrival in Tanzania There is no significant difference in these trends between men and women, or accross age categories, except that the youngest respondents seem to have arrived later. However, this 22The statistical data in this section are compiled in Table 3-5, unless indicated otherwise. 23 These percentages are representative only for that segment of the refugee population that arrived in Tanzania maximum 10 years ago, as that was one of the criteria to select respondents (see chapter 1).

PAGE 154

138 is undoubtedly more a result of one of the selection criteria used (i.e., memories of life at home as an adult (see chapter 1), than a genuine representative trend among the urban refugee population. Of all respondents, the majority (76%) never resided in one of the refugee camps in Western Tanzania. Comparing nationalities, only 10% of the Congolese ever resided in one the Tanzanian refugee camps, compared to 39% of the Rwandese and 22% of the Burundese respondents.24 As will be explained in the following chapters, the reason for this phenomenon is closely related to a more pronounced tendency among Congolese, compared to the two other nationalitites, to rely on their social networks rather than on international aid programs. With regard to age, considerably more older (39%) than younger respondents (14%) experienced a number of years in the refugee camp before heading for Dar es Salaam. Overall, however, the overwhelming majority in any of the social groups never resided in a refugee camps. 020406080100AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Never resided in refugee camp One to two years More than two years Figure 3-8: Number of years resided in refugee camp in Tanzania Nine out of every ten respondents (88%) arrived in Dar es Salaam less than 5 years ago, i.e., from 1996 onward. Concurrent with the information on the year of arrival in 24 The number of Rwandese and Burundese refugees with actual camp experience is higher still, as many lived in one of the Congolese refugees camps before coming to Tanzania (see chapter 2).

PAGE 155

139 Tanzania and on the number of years resided in the refugee camp, Fig 3-9 indicates that the Congolese overwhelmingly head straight for Dar es Salaam from the moment of arriving in the country. This is contrary to the tendency among Burundese and Rwandese, quite a few of whom have spent time in the refugee camps before heading for Dar es Salaam. A major reason at the root of this trend is that the Congolese refugee population is much more urbanized than the population in the other two countries, and thus less likely to accept the rural living conditions of the refugee camps in Western Tanzania. 010203040506070AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< One to two years Three to five years Six to ten years Figure 3-9: Length of Stay in Dar es Salaam Looking at each country's respective urbanization rates, the sample shows that the refugee populations in Dar es Salaam comprises three times as many urbanites than each of the national averages. Whereas, for example, the national urbanization rate for Congo was 30% in the year 2000 (UNDP 2001), nine in ten Congolese refugee respondents reported being prior inhabitants of one of the three major urban centers in Eastern Congo: Uvira (42%), Bukavu (33%) and Goma (15%). A similar trend is observable in both Burundi and Rwanda. Even as only 8% of the Burundese population is urbanized (Ibid.), 39% of the Burundese respondents in the sample resided in the capital Bujumbura before

PAGE 156

140 the flight.25 The same holds true for the Rwandese refugees in Dar es Salaam, among whom 34% originate from the capital Kigali, compared to a national urbanization rate of only 6% (Ibid.). Whether at the national level, or at the level of the urban refugee sample, the Congolese urbanization rate is three times higher than the one for Burundi and Rwanda, leading to a much higher tendency among the former to head directly for the city of Dar es Salaam instead of one of the refugee camps.26 On average, only one in three respondents (36%) is registered with UNHCR, whether at the camp level or with the UNHCR office in Dar es Salaam, yet there are significant differences across nationalities and age categories. Twice as many Rwandese respondents (56%) are registered with UNHCR than Congolese (22%) or Burundese (31%). The following chapters will demonstrate that this is related to the fact that persons who do not have access to informal support systems through their social networks, i.e., the Rwandese urban refugees in Dar es Salaam, are more likely to seek forms of formal support, i.e., through UNHCR. With respect to age, respondents older 25 These data show that the Burundese refugees who have arrived in Dar es Salaam over the past 10 years are considerably more urbanized than may have been the case in the past. The first--and so far only--research study on Burundese refugees in Dar es Salaam focused solely on the rural-urban migration of the children of the 1972 refugees, who had grown up in the settlements in the 1970s and 80s, and decided, as young adults to head for Dar es Salaam. The findings of my research projects contradicts the author's contention that "nearly all the Burundi refugee youths in Dar es Salaam were raised in the refugee settlements" (Sommers 2001:348). 26 The following is a typical illustration of the refugees' urban background by a 22-year old Congolese respondent. As her parents before her, this informant was also born in the village, but moved to Uvira permanently at the age of nine. Having regularly visited her mother's and father's villages (les villages natales) during school holidays, she knows the village and vividly expresses her preference for the comforts of urban life: "In town, we were studying well, school was going along nicely. But in the village, in order to have anything to eat, you have to go to the fields (shambani) and work hard to cultivate (kulima). Then when you return to the house, you have to cook food. In [the village] school you sit on the floor, there are no chairs. We had to write on slates, there were no notebooks. . In town, there are no problems, but in the village there are many, many problems. You have to work very hard to have something to eat, there is no electricity. . You cannot even buy food, what you cultivate is what you eat, and working the fields is hard, hard work. You plough, you are so tired, and then you still have to bathe before going to school."

PAGE 157

141 than 35 years of age are twice as likely to register with the refugee agency than the those in the two youngest age groups (from 26 to 52%). At the root of this trend are most likely the greater familial burdens of respondents in these age groups, as well as the higher level of political involvement of the older generations encouraging them to seek UNHCR protection. 0102030405060AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< UNHCR registered Figure 3-10: Registered with UNHCR The above overview provides a first description of the characteristics of the urban refugee population from the Great Lakes region in Dar es Salaam. Many differences between nationalities have surfaced, some more striking than others: e.g., religious affiliation, education levels, the times and numbers in which nationals of each of the three countries came to Tanzania, and more in particular Dar es Salaam. Gender, on the other hand appears to play a more important role with regard to the marital status and family composition, education level and professional activities exercised. Age has an influence on family composition, employment activities and the level of experience with refugee camps and humanitarian agencies. Some of these variations are directly related to the nature of the respective conflicts, and have as such been briefly contextualized in the text, others are not so easily explainable and I will come back to them in the following chapters. But first, I would like to introduce the twelve persons whose life histories,

PAGE 158

142 experiences and reflections will document, enliven and hopefully visualize what it means to be a refugee in today's Dar es Salaam. 3.1.3 Introducing the Protagonists The Congolese contingent consists of Melanie (22 years of age), Dieudonn (30), Bonaventure (28) and Euphrasie (55).27 They all arrived in Dar es Salaam recently, i.e., after 1998 and none have ever resided in a refugee camp. The three younger people are urbanites, whereas Euphrasie lived in the village most of her life. The two men left Congo because they were personally targeted, one by the new Rwandese/Ugandan regime, the other by the Mayimayi rebel forces for collaborating with the new authorities/occupiers. At the age of 19, Melanie fled her parental home in Uvira for the second time (the first time being in 1996 during the first Congolese conflict) when in 1998 the Rwandese and Ugandan military invaded Congo. A stranger helped her to head for Dar es Salaam where she was hoping to meet with a relative. She currently works as a housegirl, and is appreciated by her employers as a hardworking and honest young woman. She has not heard from her parents or siblings since 1998, and her biggest wish is to return to school and obtain her secondary school diploma. Dieudonn is originally from Kinshasa, but was stationed in Eastern Congo as a policeman when the second war broke out in 1998. As a civil servant, he initially continued his work under the new Eastern Congolese government headed by the Rwandese and Ugandan military, until his house was burned down and his family violated by the Mayimayi forces. In May 2000, he came to Dar es Salaam with his 27 All of the names used to name refugee informants are psuedonyms.

PAGE 159

143 family. He occasionally works as a French teacher, while in the process of applying for refugee resettlement in a third country. Bonaventure was a medical student in Bukavu at the time of the 1998 invasion. Being among those students advocating that the Rwandese and Ugandan military authorities should not be allowed to enter their campus, he was quickly branded as a wanted person and fled in 1999 via Uvira and Bujumbura. Bonaventure is well integrated in his church community and held in great respect by his friends and relatives alike. It did not take him long to learn the trade of electrician in order to be able to earn a living. Bonaventure's communication skills and knowledge of Kinyarwanda proved a major asset for locating informants of Burundese and Rwandese nationality. His natural inquisitiveness contributed greatly to many hours of fruitful discussions about refugee life in Dar es Salaam. Euphrasie divorced her Tanzanian husband in the beginning of the 1980s, and subsequently provided for herself and her children through agricultural activities. In 1998, when many of her relatives where killed in the Rwandese/Ugandan invasion, she decided to come to Dar es Salaam, and found shelter with her ex-husband and his new Congolese wife. Although currently partly paralysed and unable to move about without a walking stick,28 she was able to earn a living as soon as she arrived by selling smoked fish. Her only wish is to get better, return home and enjoy her old days in the company of those members of her family that are still alive. The Burundese are represented by Pierre (30), Amlie (51), Aristide (36) and Christine (25). While Amlie and Aristide were among the 1972 Burundese refugees in 28 For reasons unrelated to the conflict or her flight to Dar es Salaam.

PAGE 160

144 Tanzania, Pierre spent four years in Congo in the mid-1990s. All three returned to Burundi at one point, but were forced to leave their home country again at a later stage. Christine, who is of mixed Hutu/Tutsi parentage, and Aristide are both married to Tanzanians. After the 1993 assassination of his father, who was a senior member of the Ndadaye government, his mother decided to take Pierre (then 22 years old) and his two younger siblings from Bujumbura to one of the refugee camps near Uvira in Eastern Congo, where they lived for three years. When in 1996, the Rwandese government troops invaded Congo and started attacking both Burundese and Rwandese Hutu refugee camps, they returned to Bujumbura. Unfortunately, Pierre's father's political legacy made him a target of the Buyoya government and in 1997, he came to Dar es Salaam. He has been working as a tailor since his arrival, and is very worried about being discovered by the Tanzanian authorities. Amlie came to Tanzania for the first time in 1972, after the assassination of her husband. She and her children returned to Burundi soon afterwards, when life in Kigoma became too difficult. She raised her children by herself through agricultural and market activities, and regrets not having been able to school them all. In 1993, the war drove her to leave her home and join a relative in Dar es Salaam who had been living there since the 1970s. As Euphrasie, Amlie regrets very much having to confront the daily struggle of life in a foreign country instead of enjoying the comforts of her own home in her old age. Aristide also came to Tanzania for the first time in 1972, as a 4-year old. He grew up in one of the refugee settlements in Western Tanzania, but at the age of 14 he was sent to Congo (then-Zaire) to live with Burundese relatives and go to secondary school. Upon

PAGE 161

145 obtaining his degree in 1990, he returned to Burundi and started studying at the university of Bujumbura. The 1993 war forced him to leave, but eager to further his studies, he came to Dar es Salaam and was able to obtain a scholarship from the Ndadaye Foundation. He is currently married to a Tanzanian and works in a marketing company. Christine's Tutsi mother and Hutu father were killed during the events of 1993, and as a 17-year old she left the village and joined relatives in Bujumbura. Yet, as a witness to the assassination of her parents, she continued to be sought by the authorities and with the help of a Ugandan friend came to Dar es Salaam in 1996. Her friend abandoned her, and soon after she married a Tanzanian man twice her age. She currently has two children, is very unhappy in her marriage and feels generally isolated and frustrated for lack of economic and social opportunities. The Rwandese representatives are Daniel (28), Cathrine (24), Jeanne (31) and Charles (35). They are all in their twenties and thirties, because as explained in chapter 1, I was not able to meet, let alone interview, any Rwandese man or woman older than 36 years of age.29 Daniel, Jeanne and Cathrine left Rwanda in 1994 because of the genocide, and only Cathrine returned home temporarily. Both she and Charles estimated however that the reigning climate of insecurity had become intolerable and decided to head for Dar es Salaam in 1998 and 2000 respectively. Daniel, after six years of roaming around the refugee camps in Eastern Congo and a brief stay in one of the Tanzanian refugee camps, also arrived in Dar es Salaam in 2000. 29 As described in detail in chapter 1, older Rwandese men and women did agree to meet with my Congolese assistant whom they regarded as their 'sister' (notre soeur) and are therefore included in the survey but not in the ethnographic interviews.

PAGE 162

146 Jeanne is a Tutsi woman who survived the genocide together with her newly born baby while her husband perished. She came to Tanzania in 1994, married a Tanzanian man as his second wife and has had several ventures as a skilled business woman. She went back to visit her mother in Rwanda in 1998, but is determined to make a life in Tanzania. The memories of the horrors she lived through in 1994 prevent her from returning permanently to the country of her birth. Cathrine was 16 at the time of the genocide. She lost her Tutsi father, Hutu mother and two sisters. One year as an unaccompanied minor in one of the Ugandan refugee camps right after the genocide, hardened her in her determination never to set foot in another refugee camp again. She paid her own way from Kigali to Dar es Salaam in 1998 by braiding hair, and has a two year old daughter by a Tanzanian man that she provides for by herself. Daniel was 20 years old at the time of the genocide, and his father being a senior member of the Habyarimana regime, the whole family fled to Congo (then-Zaire) under suspicions of being part of the master-minding of the genocide. Hunted down by the Rwandese government troops throughout Eastern Congo (see chapter 2), both his father and mother and two out of four siblings succumbed to the hardships. Daniel reached one of the refugee camps in Western Tanzania in 2000, together with the two youngest siblings, but after only six months decided it would be better to try his luck in Dar es Salaam. Recently, his twelve year old sister and eight year old brother were adopted by an old Congolese friend of his father's and are currently living in Kinshasa. Daniel tries to get by as a missiontown and has a child with a Tanzanian woman. Charles is a civil engineer and together with two business partners managed his own construction company in Kigali. Its success aroused the jealousy of certain persons

PAGE 163

147 (suspected by him to be 1994 Tutsi returnees from Uganda or Tanzania) well connected with the current regime. When one of Charles' associates died in an accident and the other (was) disappeared, he decided it was time to leave. He currently earns a living as a computer science teacher at the secondary school level and lives with his young wife and new born baby in a room, close to the Tanzanian army barracks. All of the above are Roman-Catholic, except Christine who is Muslim while Amlie and Dieudonn are Pentecostal. Only few of the women (Amlie, Euphrasie and Christine) have a rural background, growing up and living in the village rather than in the city. Young or old, all women requested to be interviewed in Swahili, stating that their French "wasn't good enough," while the men, each one a high school or university graduate, preferred to converse in French.30 Recounting the experiences of these protagonists, it is my intention to personalize the "refugee experience" while nevertheless taking into account the issue of representativeness. Therefore, the introduction of each of the following sections (leaving home, avoiding the refugee camps, heading for Dar es Salaam and prospects on returning home) gives an overview of the survey results pertaining to each section's topic. The body of the respective sections consists of a number of interview excerpts selected on the basis of their representativeness and power to illustrate the most frequent answers of the respondent to the survey's questions.31 30 All interviews were conducted by the author in Dar es Salaam and translated by her. 31 All questions were open-ended and often respondents gave multiple answers. These answers were subsequently divided into a limited number of categories. Those categories of answers are presented in the tables at the end of this chapter and visualized in the text. They are listed in order of frequency at the level of the whole sample of three hundred respondents. Each set of answers was then disaggregated by gender, nationality and age category.

PAGE 164

148 3.2 Leaving Home On average, the majority (70%) of the respondents quoted very general reasons for deciding to leave their respective home country: namely, the civil war, the ethnic war, the massacres or the genocide (Table 3-6). This answer was more prevalent among Congolese and Burundese (75 and 81% respectively) compared to Rwandese (53%). The second most common reply (28%) referred to more specific, personal losses suffered, such as the killing or arrest of relatives, the looting and raping by military troops or the destruction of one's home. 0102030405060708090AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< War, massacres Personal loss Personal insecurity Political reasons Other Figure 3-11: Reasons for leaving the home country Thirdly, one in five respondents (21%) declared that his/her own life was in danger due to the reigning climate of insecurity, while only one in ten (9%) mentioned political motives to leave home. However, more than twice as many Rwandese respondents as Congolese or Burundese quoted personal insecurity as a reason for leaving home (34% versus 16 and 13% respectively). As described in the previous chapter, contrary to the full-blown conflicts in Congo and Burundi, the situation in Rwanda is not one of outright war, but rather characterized by arbitrary arrests and disappearances related to the settling of scores in the aftermath of the 1994 genocide. It is in these circumstances that a

PAGE 165

149 significantly larger proportion of the Rwandese respondents feel personally threatened by the situation in the home country. 3.2.1 Fleeing Wars When the First Congolese war broke out in 1996, Melanie and her family left her hometown of Uvira for about one week and then returned home. I asked her what was different in 1998, during the second Congolese war, that made her decide to leave Congo altogether. She explained that: In the first war, they invaded [Uvira] at night and started killing people. The others ran to Makobola. [There] we were told about the continuing fighting in Uvira, and also when things calmed down. We returned home, hid in our house for two weeks until the war was over and then went on to live well (kuishi vizuri). But during the second war, they made us run (wametukimbiza) with the bombing, and we did not know anyone to show us where to go . Houses were demolished, burned down by the military. Many, many people died. From Melanie's account, it appears that the 1998 invasion of Rwandese and Ugandan troops, contrary to the 1996 Congolese rebellion aimed at overthrowing Mobutu, was of a whole different nature and struck heavily, taking the civilian population completely by surprise. Melanie recalls that it all happened on a Tuesday: At that moment I was at the river, doing laundry. I heard the gunshots (masasi)32 and wanted to run home. But the neighbors told me not to go home, they said they had seen my mother and the others leave. So I left the clothes there and then. Fortunately, in the pocket of one of my father's trousers I found some money. I didn't go look for my mother, I just ran. Everybody went his own way (kila mtu kwako). Then I saw people boarding a boat, while the sound of guns became stronger. I jumped in, hiding below, thinking that with God's help, my uncle (mjomba mother's older brother) in Dar es Salaam could help me. 32 This word masasi is not a standard Swahili word, but was used by all of the informants to describe gun fire. It is most probably derived from the word risasi which means bullets.

PAGE 166

150 Up until the time of the interview in 2001, Melanie had not received any news from her parents or brothers and sisters, despite several requests to tradesmen and women visiting her home area. Amlie, recalls how even between 1972 and 1993, life in Burundi was hard and dangerous. They said that things had calmed down but there was still trouble. They took people in secret, and then killed them. [For example] if you were two men, three others might call you to come to the village hall, you'd go and never come back. We always thought they'd been taken to Bujumbura. We continued to live [in Burundi] because we had experienced ourselves [in 1972] how much suffering there was in Kigoma [Tanzania]. That's why we had to get used to living [in Burundi]. But in 1993, when the war broke out, the situation became so bad that: we had to sleep in the forest because of the dangers. We couldn't even sleep in our own house anymore. In the end, Amlie decided that enough was enough and left her home in November 1993. 3.2.2 Personal Loss Contrary to the motivation to leave for reasons of general unrest or war in order to avoid being hurt or killed, many respondents had already suffered personal losses at the time of the flight, such as the assassination of relatives, the burning down of their houses, etceteras. In addition, military and rebel factions often went on the rampage looting and raping the civilian population. This happened to Dieudonn's family. After the regime change in 1996, Dieudonn was transferred to Eastern Congo as a police officer. In 1998, the Congolese rebel forces, the Rassemblement congolais pour la dmocratie (RCD) in close collaboration with the Rwandese and Ugandan government forces occupied the region, and all Congolese government officials and civil servants were

PAGE 167

151 instructed to either cooperate or leave. Dieudonn decided to stay on. However, after an unpleasant and potentially dangerous encounter as the local Chief of Police with a number of Ugandan military intent on taking stolen gold with them to Uganda, he had to leave Bni. He requested a transfer and moved to Uvira. Not only was Dieudonn, as a government official, considered by the Congolese Mayimayi to be collaborating with the Rwandese and Ugandan occupying forces (see chapter 2), but, in addition, his wife was rumored to be a Tutsi of Burundese descent. So I went to Kivu to work there. It came as a big surprise that the Mayimayi were there, it seems they had left their villages to come to Uvira. They came to my house, my wife was at home and the child also. They took the child, and killed it. They took my wife . made love to her (sic, ont fait l'amour avec elle). They took the child . and threw it in the toilet. A neighbour heard my wife cry, she came and asked these people to leave my wife alone. But these people before going they took some petrol which they threw in the house and lit it. This woman took my wife with her. Myself, I was at the prison of Uvira, I was the officer of watch, when I was informed. At that moment I was too hurt, and I decided that it would be better to leave altogether. Because if I stayed I would be edgy and might do things that would not please the authorities, and so I decided to leave. . Dieudonn's describes the gang rape of his wife as "a love making" apparently unable to use the word "rape" in connection with his wife. After arrival in Dar es Salaam, UNHCR arranged for Dieudonn's wife to receive medical treatment and counseling for a period of six months. 3.2.3 Saving One's Life Since the RPF took over the regime in 1994, Rwanda has not known war or civil strife, yet its high level of arbitrary arrests and disappearances lies at the basis of the steady trickling in of Rwandese refugees into Tanzania during the late 1990s. Charles, as a civil engineer, founded a small construction company in Kigali together with two associates. Business went well and the company executed quite a few construction and rehabilitation projects. In 2000, his company won a bid for a project worth around

PAGE 168

152 US$100,000 for the rehabilitation of a primary school building financed by the Rwandese Ministry of Public Works. This, Charles suspects, aroused the jealousy of some well connected persons. Soon afterwards, one of my associates was assassinated on the street. They (sic) didn't allow us to bury him in the appropriate manner, and his body just stayed on the road. Two months later, I had gone with my other associate somewhere to have a drink. Two men came and asked him to go with them because they wanted to talk to him. He never came back, we waited and waited. We never found him again. We looked everywhere, went to the prisons, hospitals, everywhere. . Yes, the two men were dressed in civilian clothes, but that does not mean they are not military. After that, I as well, started to receive threats. They wanted to imprison someone, well, they came for me first, telling me I had to knock on his door. They told me that seeing me, as he was a neighbor, he would open his door, at which point I would have to step aside. So we went. This time, they were wearing their military clothes. It was night, I called out this person's name, and while he was opening the door, they told me to go home. I went home, and this person was imprisoned . I understood then that one day it would be my turn. Charles throughout his account uses the impersonal "they" to indicate the persons terrorizing the civilian population, only slightly alluding, in the course of the conversation, to the fact that "they" are part of the Tutsi military and/or the FPR government, and may possibly be Tutsi returnees from Uganda or Tanzania. 3.2.4 Political Reasons When refugees quote political motives as the reason for leaving home, it is very often not one's own political activities that expose one to danger. Pierre recounted how his father was a senior advisor to President Ndadaye, Burundi's first democratically elected president, whose assassination in October 1993 plunged the country into civil war. When her husband, a Hutu intellectual and widely known politician, "was disappeared" later that year, Pierre's mother decided to bring her 22 year old son and other children to safety and fled to Eastern Congo (then Zaire). They lived in one of the refugee camps near Uvira until 1996, when the first Congolese war broke out. Pierre

PAGE 169

153 recalls how the Congolese Banyamulenge together with the Rwandese Tutsi military were pursuing and attacking the Rwandese Hutu refugees in an attempt to neutralize the Interahamwe and ex-FAR (see chapter 2). However, at the same time, the invaders also accused the Burundese Hutu refugees in Eastern Congo of collaborating with their Rwandese Hutu brothers, and indiscriminately attacked their refugee camps as well. Pierre's mother decided that "instead of dying among the Congolese, it was better to die at home" (au lieu de mourir ici au Congo, tant mieux aller mourir chez nous) and thus they returned to Bujumbura in 1996. Pierre, then 25, returned to secondary school (lyce) to continue his studies. Only two weeks after he restarted taking his classes: The school's Principal asks me to come to his office, after checking whether I am such and such (ce petit). When I went into his office, I found four armed military present, and I also saw a car outside. The only word they said to me was "patangaa," in other words I should get into the car. I had no choice, I had to obey their orders, they were very well armed. I got into the car and they took me some place North of the capital, more precisely to Kamenge, to the camp that is situated there and is called Kishoto. So I stayed there and I was treated very badly, I was hit with sticks, I was alone in the prison cell. I was beaten . I don't know how to explain this to you. I was there more than a week without my mother knowing where I was. What did my mother do in the meantime? The one person who went to inform my mother of what happened, was a classmate of mine, her name is Magnifique . As soon as she learned what happened, my mother went to see the Principal. Words were exchanged. My mother told him that "If my son disappears, if he dies, then his death will be on your head, you will be responsible for his death." The Principal was a Tutsi. My mother is a Hutu but she looks like a Tutsi . The Principal replied that he did not know where I had been taken to, and that he only knew that four military came accusing me of arms trafficking from Congo to Kibira. Kibira is the headquarters (tat-major) of the [Burundese] rebels . He told my mother that the only thing she could do was to go to the Human Rights Office . My mother immediately went there, and they wasted no time launching a search . Thus they found out that I was in Kishoto. They called the Commander-in-Chief and requested I be released. . I was transferred to another camp [prison] in the center of Bujumbura, where I stayed another week. .. After pulling more strings in high places, the Human Rights Office was finally able to have me released, but it was a conditional release. The Commander told my mother that she should evacuate me . That she was lucky to see me again, and should ensure now that I not stay [in the country]. .

PAGE 170

154 We learn from Pierre's account that if his father had not had such a high political profile, he would not have attracted the attention of the Burundese Tutsi military authorities and be unjustly accused of arms trafficking. Pierre came to Tanzania by himself, his mother and two younger brothers still live in Bujumbura. Once the decision to leave is taken, and the country of asylum, i.c. Tanzania, reached, the refugees have the option to go to one of the refugee camps in Western Tanzania . or not. 3.3 Avoiding the Refugee Camps On average, more than half of all respondents (55%) mention intolerable living conditions in the refugee camps as the reason why they came to Dar es Salaam. Whether considering insufficient food rations, the prevalence of illnesses, previous camp experience or the general perception of camp life as miserable, there appears to be no difference across gender or nationality in respondents' determination to avoid the camps. Respondents younger than 25, who are largely unmarried and childless, are only half as likely (42%) to quote the living conditions in the refugee camps as a motivation for heading for Dar es Salaam compared to those between 36 and 45 years of age (70%), of whom a large majority (81%) has families to take care of (Table 3-3)33 and a higher number of whom have had previous living experience in a refugee camp (Table 3-5). 33 While it is true that 87% of the respondents over 45 years of age also have one or more children, these are usually already adults whose needs are no longer the responsibility of their parents.

PAGE 171

155 01020304050607080AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Miserable camp life Insecurity in camps Friends/relatives in DSM Impossible to exercise profession in camp Other Figure 3-12: Motives for choosing Dar es Salaam instead of the refugee camps. Security concerns for politico-ethnic reasons motivate one in three of all respondents (31%) to avoid the refugee camp. Rwandese and Burundese respondents appear significantly more concerned with the insecurity in the refugee camps than Congolese (35 and 34% respectively versus 23%), due to the ethnic roots of the conflicts in those two countries (see chapter 2). Among 26 to 35 year old men, I was told by informants, an additional important motive for avoiding the refugee camps (41%) is the fear to be (forcibly) recruited into one of the rebel factions. Almost one in every four respondents (24%) decided to by-pass the refugee camps because they had a relative, friend or family friend in Dar es Salaam, a trend especially outspoken among Congolese (33%) and the youngest age group of respondents (34%). The following chapters will illustrate that while Congolese respondents enjoy strong, supportive networks providing them with large informal support systems, the youngest respondents are socio-culturally more easily incorporated as dependent relatives. Other motivations (6%) to head for Dar es Salaam included "I have my own means" prevalent among Rwandese (14%) and the oldest age group (16%), and "the camp prevents me from being in contact with others" quoted mostly by Congolese (15%)

PAGE 172

156 and 26 to 35 year olds (12%). The key to these last variations across age and nationality will equally emerge in the following chapters. 3.3.1 "Miserable" Camp Life Since she became paralyzed and no longer able to earn a living, Euphrasie unwillingly came to depend on other persons. Her ex-husband and his wife provide her with a room in their house, but for all other necessities of life she must rely on her 20 year old twin sons (both jobless and not bringing in any money to even buy food) or on handouts from neighbors, whether Congolese or Tanzanian. Daily survival is a frustrating struggle for Euphrasie, and yet when I gently suggested that with regard to her basic needs, she might be better off going to one of the refugee camp, she vehemently declined: No, I cannot go to the camp . There are many diseases . Even if you are of a respectable age, you are not treated with respect . Not a single member of my family is in the camp . People who have been there and left to come here, told me that a person may have children and yet receives only one cup of flour. How will a whole family eat from one cup of flour? Other Congolese have returned home because of this misery. They said it is better to die from the bullets in Congo than from a miserable life there. Those that live well in the camp are the ones that fled the war with money. They trade in the camp, eat well and can afford to have their children study. . But even they say it's no fun . People [in the camp] don't eat well, they cannot eat to their satisfaction (hawashibi). Misery everywhere. I am an old woman, and I don't have anyone to rely on in the camp. If I had gone to live in the camp, I would have died by now. Euphrasie based her decision to avoid the refugee camps on the hearsay and testimonies of others. One of these, a Congolese informant, recounted that during his 96-98 stay in a Western Tanzania camp, one of his children had died, in his opinion, "due to lack of medical attention" from the camp's infrastructure. This person, together with his family members, had repatriated to Congo in 1998, when after a few months the Rwandese/Ugandan invasion took place, prompting them to leave their home again. This

PAGE 173

157 time, however, husband and wife and three children headed straight for Dar es Salaam, determined never to set foot in the refugee camps again. 3.3.2 Security Concerns While the particularities of the Rwandese and Burundese political conflicts are pretty straightforward,34 the Congolese situation is more complicated because of the multitude of militant factions. From 1998 onwards, the Rwandese and Ugandan invaders took over the state infrastructure in Eastern Congo by appointing their sympathizers into official positions. In the beginning, a number of Eastern Congolese citizens accepted the new rulers and collaborated with the new regime by simply continuing to do their job. However, the Congolese rebel group Mayimayi always considered their collaborating compatriots as traitors, whether in their official capacity in Congo or as refugees in Tanzania. Quite a few informants insisted that the camps are infiltrated with Mayimayi militants, posing as refugees of Babembe origin. Their aim is to eliminate all Congolese rumored to have a link to the Banyamulenge (e.g., by having a Tutsi appearance such as height or facial features) or a link to the Rwandese and Uganda occupying forces (albeit by simply exercising a public function, as Dieudonn did). After arriving in Dar es Salaam in 2000, Dieudonn received UNHCR assistance because of the medical condition of his wife, but as soon as her medical treatment was considered completed, the assistance was cut off and they were told to go to the refugee camp. In Dar es Salaam, I was [first] in a hotel. I stayed there for one day, and from there I went to the Ministry, no, I first went to the UN. UN [UNHCR] told me to go to 34 Although not as straightforward as CNN would like us to believe, Hutu versus Tutsi. In both Rwanda and Burundi, there exist affiliations along regional and political lines within the Hutu and Tutsi populations as well, compounding the political complexity and social tensions in both countries.

PAGE 174

158 the Ministry [of Home Affairs]. At the Ministry they gave me forms to fill. I did so and they sent me to the UN, and the UN decided to provide assistance for the next six months. They also helped the wife, a little bit, they didn't really treat her much, but at least they did. After that, they decided that I should go to the camp. However in my statement I had clearly mentioned that there are these Babembe, tribes like the Babembe, the Bafira, and Bafoulro and among those three tribes are a lot of Mayimayi. . So they said I should go to the camp Mkungwa, it's a camp of protection, where there are Rwandese, Burundese, Rwandese Tutsis and Burundese Hutus. But I refused to go there either. I wasn't going to live there with these people, because you can never be sure that there is no infiltration of the Babembe and likes. That is why I refused to go. But the UNHCR didn't accept my refusal and cut the assistance I was receiving. They told me to go fend for myself (se debrouiller), and so I did . Another Congolese refugee was more fortunate, in the sense that his claim of personal insecurity in the refugee camps was accepted by both the Ministry of Home Affairs and the UNHCR office.35 He told me that: If I had approached the UNHCR office upon arrival in Kigoma, they would have taken me immediately to one of the camps. Because of the many conflicts between the Congolese, especially us, who have worked together with the rebellion, are considered to have "dirty hands" (les mains sales) by the local population. As the camps are full with Mayimayi militia . Let's say that 70% of all Babembe are Mayimayi . I said to myself: "No, my life will be in danger. It would be better to go directly to Dar es Salaam." That the refugees' fears for their personal security are not imaginary is illustrated by the assassination of a Burundese medical doctor, his wife and three children in the Mtabila refugee camp in February 2000.36 Nobody was ever charged with the murder, which was presumed to be of a political nature, based on the fact that the brother of the victim was the head of one of the Burundese rebel factions participating in the Arusha Peace Talks. The victim's sister-in-law who resided in the camp at the time the assassination took place, describes the situation. 35 This refugee was ultimately resettled in Norway. The major principles of refugee resettlement programs and procedures are described in chapter 2. 36 See also chapter 2.

PAGE 175

159 How were they able to kill le docteur Jean? There were people with fire arms inside the camp . they didn't even bother to hide. Everyone knew the houses/huts (blinds), where those people having arms lived. Even the security forces had that information, whether the sungu sungu37 or the police officers . But the Tanzanians don't want it to be known publicly that there are armed gangs in the camps . The armed gangs come into the camp with a lot of money . or they come with many cows, which they sell . and use this money to bribe the police and other security officers . even the local authorities. . After the death of the doctor, lists were posted of all the persons that they were going to kill. We were on the second list. She further explained that the armed gangs (bandes armes) are financed by the government in Bujumbura, and sent to eliminate the Hutu intellectuals in the refugee camps. The brutal manner in which her brother-in-law and his family were killed was exceptional, she said: "Usually when they want to kill someone, they take him elsewhere. They kill him outside of the camp, making it seem like an accident." She conceded that recently the security situation in the camps had somewhat improved, but not to the extent that she would be willing to return to the refugee camp.38 3.3.3 Joining Relatives or Compatriots In addition to the reasons for avoiding the refugee camps, there is one important pull factor, namely the presence of a relative or friend in Dar es Salaam, whom refugees count on for accommodation and other assistance upon arrival. Melanie, for example, as we saw in a previous section, came to Dar es Salaam intending to locate her uncle. Yet things did not work out as planned and, fortunately for her, strangers took pity on her, both on the train from Kigoma and upon arrival in Dar es Salaam. 37 The sungu sungu are a type of unarmed citizen guard, they consist of refugees recruited by the Tanzanian police officers stationed in the camp. Their responsibilities are limited to patrolling and reporting suspicious persons or incidents to the police forces. It is a traditional Tanzanian form of civil watch, which was also instituted by the Tanzanian government in Dar es Salaam in the early 1990s in an attempt to increase urban security (Kironde 2000:54). 38 Substantial efforts were made by UNHCR recently to improve the security situation in the camps (UNHCR 1997a; 2001a), but clearly not all refugees are convinced that they are sufficient.

PAGE 176

160 We left Kigoma at 3pm and spent three days on the road. Only God was there . I didn't pay for the train, I paid $20 for the boat and with the Zairian money I was left with I bought food. In the train, I hid without paying until Dar es Salaam. When they came to check the tickets, I hid in the toilet . I came with someone, I met him on the train, I explained everything to him. When I arrived here in Dar es Salaam, I couldn't find my uncle. I explained to an old woman, she told me I could stay in her house. She said that she would help me, I stayed at her place. At a later stage, Melanie did find her uncle's residence, but was told by his neighbors that he had long since left and his current whereabouts were unknown. Bonaventure was more fortunate. He, as well, had never before been to Dar es Salaam, but friends in Bukavu, who were aware of his problems, had given him a telephone number to call upon arrival. I took the train to Dar es Salaam, but got off at Morogoro. That's what I had been advised to do because I didn't have a ticket . I took the bus, arrived at MwembeChai [in Dar es Salaam], made the telephone call and someone answered. It is this brother39 who came to pick me up. Once the decision is taken to come to Dar es Salaam instead of going to one of the refugee camps, practical arrangements are in order. Most informants' point of arrival in Tanzania is Kigoma, from where there are daily trains to Dar es Salaam. The train ride takes three full days and covers around 1,200 kilometers. From Melanie and Bonaventure's accounts, we gather that more often than not, the refugees try to avoid paying the $15 train ticket in an effort to save money, not knowing what other problems are waiting for them in Dar es Salaam, especially during the first few days. 3.4 Arriving in Dar Es Salaam On average, respondents turned to five different possibilities to find accommodation and/or assistance upon arrival in Dar es Salaam. One in four joined a relative or compatriot friend (26%) or simply rented a room in a guest house (25%). In 39 Here used in the meaning of compatriot.

PAGE 177

161 addition, as many respondents (16%) turned to UNHCR as to a religious institution upon arrival, such as a church or mosque. Only one in ten relied on assistance from Tanzanian friends (Table 3-7). No significant differences appear when comparing women with men, while there are a few variations with regard to nationality and age. More than one third of the Burundese respondents (37%) found accommodation upon arrival with relatives or friends from home, far more than the Congolese (23%) or Rwandese (20%). Unlike the latter, the Burundese refugee population has a long history in Tanzania, and over the past three decades growing numbers of them have integrated locally and are now in a position to assist newly-arriving relatives and other compatriots. 0510152025303540AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Relative, compatriot Guest house Church, Mosk UNHCR Tanzanian friend(s) Unknown samaritan Other Figure 3-13: Accommodation upon arrival in Dar es Salaam Among the oldest respondents, a small majority relied on the help of Tanzanian friends with regard to initial accommodation (24% versus 10% as average). Listening to informants, I learned that older refugees (>45 years) are much more reluctant to leave their homes and it is very likely that having Tanzanian friends is an important element in the decision to come to Dar es Salaam in the first place as it allows a respectful visitor's status rather than a refugee status which is considered degrading (more in chapter 4).40 40 This issue will be further contextualized in chapter 6.

PAGE 178

162 3.4.1 Locating Relatives or Other Compatriots Locating someone in Dar es Salaam, a metropolis with three million inhabitants, is not simply a matter of having a street address or a telephone number, as most of the streets do not have a name, and not everyone has a telephone, which, in addition, may or may not be working (see also chapter 2). Bonaventure, in the previous section, was very fortunate to have located his contact so easily, contrary to Melanie who ultimately failed to join her uncle. All Pierre had to go by was the name of an old friend of his father's, who had fled from Burundi to Kigoma in 1993. This one, he came with his money, so he is known there, he has a big house there [Kigoma]. My mother told me before leaving [Bujumbura]: "Just as you arrive there, you ask for the house of Mr. Y and they will show you." . When Mr. Y saw me, he told me that he was convinced that I had died: "Because I had heard that you were imprisoned by the military, I really thought you would die" . He gave me the money for the Kigoma/ Dar es Salaam train fare . and told me: "Upon arrival in Dar es Salaam, you have to go to the house of Mr. X and once you arrive there, you have to ask for Jacques . When you tell Jacques that you are the son of R., he will receive you well." . So when I arrived at his place, Jacques said: "There is no problem, since you are here now and because I know your family, there is no problem. You will stay here." Pierre was very fortunate in the sense that, thanks to his father's high level profile and connectedness, both Mr. X and Mr. Y (two prominent members of the Burundese and Congolese refugee communities in Kigoma and Dar es Salaam respectively) were willing and able to help him along. Other refugees, of a more modest background, need to rely on personal contacts, albeit friends from a distant past. 3.4.2 Tanzanian Friends Euphrasie, for example, was married to a Tanzanian man with whom she had five children. They used to live in Eastern Congo until they divorced at the end of the 1970s, and he returned to Tanzania. She was very fortunate in being able to locate him upon arrival in Dar es Salaam as they had had little or no contact in the meantime.

PAGE 179

163 When I arrived, I asked around and this one woman said "Yes, I know him." She took me to this place, Mabibo. Since [living together in] in Congo, we had been separated, but I am very grateful that he gave me a place to sleep. If he hadn't given me accommodation, I would have had to rent a place. Renting a room in a guest house is an option most refugees would like to avoid as it is the most costly option and a drain on quickly dwindling resources. One in every three respondents, therefore, first tried their luck with a religious institution, be it the church or mosque, others thought UNHCR would assist them. 3.4.3 UNHCR Even as significantly more Rwandese approached the office of the UNHCR in Dar es Salaam, in an attempt to find assistance with regard to accommodation and other material needs, they were not successful. A substantial number of the Rwandese respondents bitterly related that they were literally "chased away from UNHCR" (le HCR nous a chasss) and ended up renting a room or spending the night in a church. Others "passed the night at the gates of the UNHCR office," and ultimately had no choice but to "go look for people from home" (mes frres). The rationale for UNHCR to "chase" the refugees away is the Tanzanian official policy that all refugees should relocate to a refugee camps in order to receive humanitarian assistance (see also chapter 2). Avoiding the refugee camps and arriving in Dar es Salaam is just the beginning of life in exile. Yet, before exploring in detail the ins and outs of the adaptation process that the urban refugees go through to make a new life, let us have a cursory look at respondents' prospects with regard to their stay in Tanzania, the return home, albeit, the aspiration of going to a third country. Clearly, important decisions such as when to return home or whether to return home at all are not only influenced by the individual's

PAGE 180

164 past experiences, they will at the same time affect the efforts made and success achieved in carving out a new existence in exile. 3.5 Prospects of Returning Home On average, more than one in two respondents (56%) is simply waiting for the hostilities to cease in order to return home, yet important differences between nationalities appear (Table 3-7). While three quarters of Congolese (76%) intend to repatriate as soon as the war is over and peace reigns, only one in three Rwandese (36%) accepts this as a sufficient condition for going home. Rwandese refugees, in particular, (48%) and Burundese to a somewhat lesser extent (43%) wish to see democracy installed, human rights respected or ethnic equality established especially in the armed forces, currently dominated by the Tutsi. Because many Congolese perceive the current conflict in their country as an invasion of foreign forces, contrary to the politico-ethnic nature of the conflicts in the two other countries, they see the return of peace as a sufficient condition for returning home. The Rwandese refugees, on the other hand, are the most politicized group (for reasons that will be explained later) and therefore very sensitive to the political conditions of their possible return home. 01020304050607080AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Return home if peace Return home if democracy Wish to go to a 3rd country Wish to remain in Tanzania Here, refugee rights are not respected Other Figure 3-14: Prospects of Returning Home

PAGE 181

165 One in six respondents, on average, does not wish to return to the home country at all and aspires instead to go to a third country, while one in ten reportedly prefers to remain in Tanzania rather than ever returning home. Congolese appear to be most opposed to the idea of remaining in Tanzania, while Burundese are the least inclined to have the ambition to resettle in a third country.41 3.5.1 The Return Home The overwhelming majority of Congolese respondents, such as for example, Melanie and Dieudonn expressed their wish for peace in Congo which would enable them to return home: "What would convince me to go home, is there would be peace again. Really, when that it the case, I want to go back." The situation in Burundi and Rwanda, however, is somewhat more complicated as Daniel, who left Rwanda at the time of the genocide, explains: The big problem is the settling of scores (reglement de comptes). I come from a well-to-do family, my father was a senior executive, you know, like the ones that were minister or Director General of parastatals [senior civil servants], their names are known. I could go back in principle, but I'm really afraid of the settling of scores. . If I were a girl, maybe it could work.42 But now, for them, it is in the head of the Tutsi, that all Hutu have killed. When I left the country, I was 23 year old, an adult person, somebody. So they think: "You, who left as an adult, you must certainly have killed." So, they could kill me too. . I have heard of others who returned home only to flee again. In the beginning, they told us: "Don't worry, go home, we'll assure your safety." But it is not true. When people go home, they are captured, especially the men . Those that returned [in 1996] were the peasants, mainly women and children, children under the age of 15 that is. But many adult men remained. . And then, return home to live where, all the houses have been confiscated. So what, I return home and then I have to live in a camp in my own country? No. 41 Both of these tendencies will be contextualized in more detail in chapter 6. 42 This remark will be situated in its socio-cultural context in chapter 6.

PAGE 182

166 When asked whether the pending installation of a transitional government in Burundi in November 2001, could convince him that the time to return home had come, Pierre disagreed. I could accept on condition that I see that the army is really mixed now. Without that, there will always be problems. In fact, it is the army that is the problem over there. If they could mix the army, that would be 80% of the peace in Burundi. But now, no, there is not really peace, because people continue to die. Even the other day, I heard on the radio that the rebels, in Ruhigi or some place, captured 300 school pupils. The other day, they took 50. And the others, they continue the killings. In Bujumbura rurale people are still dying. There is no real peace. While Daniel would consider going back home if there was a regime change (i.e., Hutu instead of the current Tutsi minority), Pierre is very cautious about the actual conditions on the ground accompanying the recent installation of a transition government in Burundi. 3.5.2 Alternatives to Repatriation Even as most respondents hope to return home at some point, a substantial minority of one in four respondents, ponders other possibilities, such as going to a third country or staying on in Tanzania. Bonaventure, for example, came to Dar es Salaam with the intention to continue beyond Tanzania. I don't have the intention to stay in Tanzania . and from Dar es Salaam there is always the possibility to go to South Africa, or other countries, such as Mozambique . My ultimate objective is to go to a francophone country, where I can find work more easily because I don't speak English . such as Madagascar . or Senegal. Unfortunately, things proved to be much more complicated than he imagined, one of the major stumbling blocks being obtaining an international passport.43 Contrary to the Congolese, among whom only one respondent considered staying on in Tanzania, quite a 43 More on this particular issue follows in the next chapter.

PAGE 183

167 few Burundese and Rwandese would prefer this option to returning home. Christine, for example, who is married to a Tanzanian with whom she has two small children, stated that: I cannot go back to Burundi because I no longer have my parents . If you don't have a home, were will you return to? And then you know that at night when you sleep, they will come to kill you. . The problems that made me leave are still there. [Besides] for the moment I have my own family that I need to take care of. In the same line, many Rwandese informants, some of whom lost their whole family during the genocide resolutely vowed never to return home, such as Jeanne. I don't like to talk about over there [Rwanda]. Friends of mine [continued to] live there, but they do so with a heavy heart (roho ngumu). When I think about those things (the genocide) I know I couldn't live there. I would be so afraid that someone will still come [and hurt me]. Jeanne, as well, is married to a Tanzanian man with whom she has one child. As a Roman-Catholic, it was hard for her accepting to be the second wife in a Muslim marriage. However, she felt there was no other long term alternative to the looming prospect of a potentially forced repatriation of all Rwandese refugees from Tanzania.44 From the above survey results and protagonists' testimonies, we are able to get a glimpse of the demographic and other socio-economic characteristics of the Great Lakes refugee population in Dar es Salaam, while at the same time their motivations for and coping strategies applied to the heading for Dar es Salaam have become clearer. A general profile emerges, even as important differences across gender, nationality and age should be acknowledged. 44 One year after this interview, the Tanzanian government revoked the refugee status of all Rwandese men and women, and it was only through her marriage that Jeanne was able to remain in the country legally.

PAGE 184

168 3.6 Emerging Profile(s) The following is a profile of the average refugee, who fled his/her country in the Great Lakes region sometime during the past ten years to come to live in Dar es Salaam. As a Roman Catholic (or to a lesser extent Pentecostal), s/he is most likely to be single and childless (even as substantial minority is married and has one or two children). In general, s/he is vastly more educated than his/her compatriots in the refugee camps and three times as likely to originated from an urban background. S/he is most likely to have left home after 1996, the year of the first Congolese war, and had been living in Dar es Salaam for five years or less at the time of this study. The refugee camps reportedly were avoided because of the difficult living conditions and to a lesser extent for security reasons, while, at the same time, it is not very likely that s/he ever set foot in a refugee camp or registered with UNHCR. Upon arrival in Dar es Salaam, s/he found shelter with compatriots or in a guest house (or, to a lesser extent, with a religious institution or UNHCR). Whereas s/he was a student, trader or employee at home, s/he turned to self-employment and, to a lesser extent, to trade activities, to make ends meet in exile. S/he hopes to be able to return home as soon as peace is re-established and preferably certain political conditions are satisfied. 3.6.1 Gender The most important difference between male and female refugees surfacing at this point are of a demographic nature, as women are more likely to be married, have more children and to be widowed earlier than men. In addition, they generally have a lower level of education than their male counterparts. With regard to the motivations to leave home, avoid refugee camps and head for Dar es Salaam, as well as any future aspirations

PAGE 185

169 regarding the return home, there appear no significant differences between refugee men and women respondents. 3.6.2 Nationality Comparing across the three nationalities, certain variations in trends emerge, however rarely to the extent that they turn the overall average refugee profile upside down. The following will concentrate exclusively on these differences without the intent of downplaying the relevance of the commonly shared characteristics as listed above. The Congolese respondents include a significant minority of Pentecostal believers, and are three times as likely to be urbanites than any of the other two national groups. Almost all arrived in Tanzania after 1996, and extremely few ever stayed at one of the Tanzanian refugee camps. Instead, they headed straight for Dar es Salaam with the general objective to join compatriots or other friends, quoting a lack of communication facilities in the camps. They are more preoccupied with the return of peace rather than political reform at home, and almost no Congolese respondent considers remaining in Tanzania an option preferable to returning home. The Burundese respondents include a significant minority of Muslims adherents, and have the lowest education levels of all three nationalities. A larger proportion of Burundese respondents arrived in Tanzania earlier than the Congolese, but it is only very recently that they have been coming to Dar es Salaam in large numbers. A minority resided in one of the Tanzanian refugee camps for a number of years, others found shelter elsewhere in Tanzania. They are the most likely to find accommodation with relatives or compatriots once the decision to head for Dar es Salaam is taken, but are the least aspiring to continue further, i.e., go to a third country. They are also more concerned

PAGE 186

170 with insecurity in the refugee camps and political conditions for the return home than their Congolese fellow refugees. A substantial proportion of the Rwandese respondents as well arrived in Tanzania earlier than the Congolese, but contrary to the Burundese, more spent a number of years in one of the refugee camps before coming to Dar es Salaam. Of the others, a minority cited sufficient financial means as a reason for avoiding the camp. A majority is registered with UNHCR, where a substantial number of them turned to for assistance after arriving in Dar es Salaam. A larger proportion than among the other two nationalities feels personally threatened and individually targeted by the current situation at home as well as in the refugee camps, and a substantial minority of one in three does not wish to return home at all. A number of the national differences are readily explainable by the nature of the conflicts in the respective countries, such as the time of arrival in the country of asylum and the political pre-occupations of the Burundese and Rwandese vis--vis the Congolese respondents. Other variations, however, are not so easily explained and the key to understanding and contextualizing them is the topic of the following chapters. 3.6.3 Age Groups From comparing the data and testimonies from persons belonging to the four different age groups, it becomes clear that as age increases, respondents are more likely to be married and have a higher number of children while enjoying a lower education level, and being more involved in trade activities and less in self-employment. Respondents' experience with UNHCR also increases with age, whether through more years of living in one of the refugee camps or being registered with UNHCR.

PAGE 187

171 While the youngest respondents (<25 years) were overwhelmingly students at the time of leaving home, and more likely to head for Dar es Salaam with the intent of joining relatives and compatriots, the oldest (>45 years) are more likely to avoid the refugee camps because they reportedly have their own means, and able to find accommodation with Tanzanian friends. Respondents of the second age group (26 to 35 years), on the other hand, are significantly more worried about the security conditions in the refugee camps and also prefer to come to Dar es Salaam because it provides more possibilities to communicate with others across the globe. And the third age category (36 to 45 years) express significantly more concern about the difficult living conditions in the refugee camps, while, unlike the other age categories, a substantial minority of them arrived in Tanzania before 1996. Some of the above variations are closely related to the demographic characteristics of the respective age groups, such as the different occupations held before arriving in Dar es Salaam, or the average increasing/decreasing familial burden and responsibilities depending on age. Other variations however, appear not to be so easily explainable, and I intend to contextualize them in depth at a later stage (see chapter 6) in combination with the results of the comparative social network analysis. The adaptation process to the new life, i.e., social and cultural adjustment, language, how to make ends meet, and where to meet new friends, is the topic of the next chapter.

PAGE 188

172 Table 3-1. Composition of Sample: Gender, Nationality, and Age Group. Nationality and Gender Between 20 and 25 year Between 26 and 35 year Between 36 and 45 year Older than 45 years Total Sample Congolese Women 14 15 14 7 50 Men 14 17 15 4 50 Burundese Women 14 16 13 7 50 Men 16 17 14 3 50 Rwandese Women 15 15 15 5 50 Men 15 15 15 5 50 Total Sample 88 95 86 31 300 Table 3-2. Education Level Compared to Camp Population (in %). Education Level Below Primary Primary Secondary Post-Secondary Congolese Camp 36 26 37 1 Urban 7 12 63 18 Burundese Camp 64 29 6 1 Urban 23 30 39 8 Rwandese Camp N/A* N/A N/A N/A Urban 9 16 59 16 N/A: No data available.

PAGE 189

Table 3-3. Demographic Characteristics of Sample (in %). Demographic parameters Women n=150 Men n=150 Cong. n=100 Buru. n=100 Rwan. n=100 <25y n=88 26-35y n=95 36-45y n=86 >45y n=31 Average N=300 Religion Roman-Catholic ied911108117813139 59 52 52 48 67 59 58 51 52 56 Pentecostal 21 26 34 19 18 23 22 28 19 24 Muslim 11 11 4 25 4 11 12 8 16 11 Other/Not Specif Marital Status Married, living w/ spouse 33 32 30 34 33 8 28 56 48 32 Married, not living w/ sp. 13 11 9 11 16 3 14 20 10 12 Unmarried, lives w/part ner4113104102wed 1329490 113327ced 405300 14103 Unmarried, not living w/ p. 33 54 46 45 41 89 52 6 0 44 Wido Divor Number of Children in Dar es Salaam Childless 33 64 52 45 48 84 54 19 13 48 One child 18 9 8 17 16 10 18 14 10 14 Two children 23 13 19 17 18 5 17 33 20 18 Three children 13 8 10 9 13 1 8 20 20 11 Four and more 13 6 11 12 5 0 3 14 37 9 Have Children not residing in Dar es Salaam 21 21 28 14 20 5 18 26 61 21 173

PAGE 190

Table 3-4. Education Level and Professional Occupation (in %). Women Men Cong. Buru. Rwan. <25y 26-35y 36-45y >45y Average n=150 n=150 n=100 n=100 n=100 n=88 n=95 n=86 n=31 N=300 Education Level Beyond Secondary Educ. 7 21 18 8 16 8 17 19 10 14 Secondary Education 50 57 63 39 59 65 55 29 54 Primary Education 24 15 12 30 16 17 20 17 29 19 No Education 19 7 7 23 9 10 8 14 32 13 Professional Occupation Self-Employed DSM 36 43 34 42 47 51 40 36 13 40 Before ore 373836 7205 SM4121257811678 SM120567257166ore 15055131 131208SM2343181003 SM0000000000ore 834490 012236 10 13 7 17 11 14 12 9 13 12 No Occupation DSM 16 31 23 22 21 28 21 20 26 23 Bef Trade DSM 30 14 22 22 17 2 22 30 39 20 Before 21 11 25 18 10 1 16 30 36 18 Paid Employment D Before 13 19 13 17 19 6 13 27 26 16 Housewife D Bef Student D Before 30 43 43 31 35 73 39 8 3 36 Agriculture D Bef 50 174

PAGE 191

Table 3-5. Characteristics of Stay in Tanzania and Dar es Salaam (in %). Women n=150 Men n=150 Cong. n=100 Buru. N=100 Rwan. n=100 <25y n=88 26-35y n=95 36-45y n=86 >45y n=31 Average N=300 Year of Arrival in Tanzania Between 1999-2001 29 29 27 27 32 48 31 10 19 29 Between 1996-1998 54 49 70 45 40 51 47 55 58 52 Between 1992-1995 17 22 3 28 28 1 22 35 23 19 # of Years Resided in Tanzanian Refugee Camp Never Resided in Camp 77 76 90 78 61 86 81 66 61 76 One to two years 14 18 9 8 31 9 15 21 26 16 More than two years 9 6 1 14 8 5 6 13 13 8 Length of Stay in Dar es Salaam One to two years 42 44 34 48 46 58 41 29 42 43 Three to five years 48 42 64 31 40 42 42 51 45 45 Six to ten years 10 14 2 21 14 0 17 20 13 12 Registered with UNHCR in Camp or in Dar es Salaam 37 35 22 31 56 26 27 51 52 36 175

PAGE 192

Table 3-6. Reasons for Leaving Home and Heading for Dar es Salaam (in %). 176 Women n=150 Men n=150 Cong. n=100 Buru. N=100 Rwan. n=100 <25y n=88 26-35y n=95 36-45y n=86 >45y n=31 Average N=300 Reasons for Leaving my Home Country* 1. Civil/ethnic war, massacres/genocide 75 65 75 81 53 36761 65265 72 72 67 65 70 2. Murdered/arrested relatives, looting, raping, home destroyed 31 25 30 20 33 33 23 24 35 28 3. Insecurity, to save my life 17 19 16 13 34 16 21 27 19 21 4. Political Reasons 7 12 15 5 8 5 12 10 13 9 5. Other (E.g., Human Rights Abuse, Followed relatives) Choosing Dar es Salaam over the Refugee Camp* 1. Lack of food, too many illnesses, camp life miserable, previously lived in camp 54 56 48 58 60 1056710 147338 42 54 70 58 55 2. Camp lacks security, ethnic reasons 26 35 23 35 34 19 41 30 32 31 3. Relative, friend or friend of the family residing in DSM 28 20 33 20 19 34 19 20 23 24 4. Not possible to exercise my profession in the refugee camp 5 Other reasons 1 14 15 2 14 5 12 5 16 6 Multiple answers allowed

PAGE 193

177 Table 3-7. First day in Dar es Salaam and Prospects of Returning Home (in %). Women n=150 Men n=150 Cong. n=100 Buru. N=100 Rwan. n=100 <25y n=88 26-35y n=95 36-45y n=86 >45y n=31 Average N=300 Upon arrival in DSM, where did you go to find accommodation ? 1. Relative, compatriot friend 21 31 23 37 20 29 28 24 17 26 2. Guest house 25 24 17 27 28 20 30 24 21 25 3. Church, Mosque 17 16 22 13 14 21 15 13 17 16 4. UNHCR 19 14 16 10 23 10 15 26 14 16 5. Tanzanian friend(s) 10 10 10 6 14 10 7 9 24 10 6. A stranger helped her me8410711044760120001001 7. Ot Conditions for returning home or possible other prospects* 1. Return home if peace/security at home, when war ends 61 51 76 56 36 12711413 148869581424 39612649502072123 61 54 53 55 56 2. Return home if democracy, human rights, ethnic equality 31 40 16 43 48 25 39 41 42 36 3. No return, wish to go to a third country 11 19 15 8 22 20 15 13 6 15 4. No return, wish to remain here in Tanzania 5 No staying in Tanzania where refugee rights are not respected her 6. Ot Multiple answers allowed

PAGE 194

CHAPTER 4 ADAPTING TO THE NEW LIFE C'est a la vie d'un rfugi . Ici, moi, je dirais que je ne vis pas, j'existe comme une pierre. "This is the life of a refugee . For myself, I can say, I don't have a life here, I exist like a stone." Rwandese refugee, Dar es Salaam, May 23, 2002. Once arrived in Dar es Salaam, the refugees have a long process of adaptation to the new environment ahead of them: different language, different culture, different neighborhood, different economic infrastructure, etceteras. Above all, they have to cope with a change in societal status from "locals" to "strangers" and uninvited ones at that.1 The following sections are organized in the order of the priorities, or sequence of pre-occupations as perceived by the respondents. In each section, a multitude of testimonies is included in order to let the protagonists and other informants speak for themselves to the largest extent possible. When asked to compare life at home before the flight with the new life in Dar es Salaam, a majority of the respondents (58%) reminisced about how good life was at home, and how much they regretted the disruption of their social worlds. Replies varied from "At home I lived with my family," "At home people live(d) in harmony," "At home, life was better because people helped one another," to "Here, I'm alone, I'm (like) an orphan" (Table 4-1). This feeling is stronger among Congolese compared to Burundese 1 The Swahili word mgeni can be translated as "visitor" as well as "stranger" and stands in opposition to the word mwenyeji which means literally "the owner of oneself" and is used to refer to "someone from here, someone who knows." 178

PAGE 195

179 and Rwandese (69 versus 46 and 49% respectively) as the former are the most recent arrivals in Tanzania and Dar es Salaam, whose lives were disrupted in a more abrupt manner than those of the latter in Rwanda and Burundi, two countries that have know civil and internal strife intermittently for decades (see chapter 2). 01020304050607080AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Life at home was good Here, no job opportunities Here, life is expensive/difficult Here, we need permits Here, there is peace Other Figure 4-1: Comparing life at home and in Dar es Salaam The next three most frequently cited differences between life at home and in Dar es Salaam belong to the economic realm. More than one in three persons (between 42 and 36%) complained about "the lack of job opportunities and money," "how expensive life in Dar es Salaam is," and the fact that as non-citizens, they "need permits to exercise any (economic) activities." Only a few variations emerge across gender, nationality and age categories, such as that Rwandese appear to be only half as likely to worry about obtaining the necessary permits as are Congolese and Burundese (22 compared to 46 and 40% respectively). This may be an indication that they are in a financially better position to buy the necessary work permits from the Tanzanian authorities.2 With regard to age, those age categories representing the economically most active segment of the population consider the work permit issue almost twice as important as the youngest and oldest age categories (46 and 43% versus 20 and 29% respectively). 2 This issue will be elaborated on further in this chapter.

PAGE 196

180 The last important difference between life at home and in Dar es Salaam, as perceived by the respondents, is of a more political nature. One in four respondents feel that, contrary to the home country, "in Dar es Salaam, there is peace" and they "need not fear being killed" (26%). Substantially more Rwandese (48%) than Congolese and Burundese (16 and 15% respectively) made this observation. This trend is compatible with the findings presented in the previous chapter, namely that the Rwandese respondents' motives to leave their home country are closely related to the perception of personal insecurity more than any of the other categories of respondents. 4.1 Local Interactions Contrary to many migrant and/or refugee communities, the urban refugees in Dar es Salaam do not have the possibility to form enclaves because the lack of a residence permit for Dar es Salaam, and thus illegal status, requires a clandestine existence. Living scattered in a metropolis of three million inhabitants has a profound effect on the (re)composition of their social networks disrupted by the flight. No longer embedded in a familiar social environment and feeling "like an orphan" the refugees slowly begin to make efforts towards establishing new ties and social relationships. The majority (54%) of the relationships established after arrival in Dar es Salaam are with Tanzanians (see chapter 5), a social group the refugees have had little experience with before.3 Hence, 93% of all respondents state explicitly that establishing relationships in Dar es Salaam is difficult (Table 4-2), a trend remarkably consistent across gender, nationality and age. Even the oldest respondents, albeit somewhat less strongly convinced, are overwhelmingly (83%) in agreement on this point. 3 Chapter 5 will illustrate that only 8% of social relationships established before coming to Dar es Salaam are with Tanzanians.

PAGE 197

181 When probed to elaborate on why it was difficult to have social relationships with people in Dar es Salaam, a majority (58%) of respondents felt that "Tanzanians ask too many questions." Rwandese complain hardest about this (85%), Burundese somewhat less (56%) while Congolese appear the least troubled in this respect (34%) (Table 4-2).4 The second most important reason at the root of the difficulties experienced in establishing new ties in exile is perceived to be the language (29%), and again the significant variations appear between nationalities, not gender nor age. While 38% of the Rwandese and 31% of the Congolese struggle with the language, only 17% of the Burundese feel language is a problem.5 0102030405060708090AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Too many questions Because of language Tz minimize foreigners Cultural differences Tanzanians don't like us Other Figure 4-2: Why establishing relationships in Dar es Salaam is difficult The third observation (15%), namely that "Tanzanians minimize foreigners," is cited by a significantly larger proportion of the oldest respondents (26%) compared to any of the other categories.6 I was told by informants that this perceived lack of respect by Tanzanians is symbolized in the frequent use of the word mkimbizi ("refugee" in Swahili). A large majority of all respondents (66%) reported being addressed as 4 The socio-cultural and historical roots for this trend will be explained in detail in chapter 6. 5 The different aspects of these differences are elaborated on later in this section, as well as in chapter 6. 6 The socio-cultural dynamics of age and the African gerontocratic system are explained in chapter 6.

PAGE 198

182 mkimbizi "daily or regularly" regardless of gender, nationality or age (Table 4-2). All informants, without exception, objected strongly to being called mkimbizi for reasons that will be explained later on in this chapter. Other often quoted reasons perceived by the respondents to be responsible for rendering difficult social relationships with the local population, differ considerably by nationality. "Cultural differences" are cited by 40% of Rwandese respondents (compared to only 14% on average) and that "Tanzanians are hypocrites" by 24% of the Congolese (included in the category of "Other") (Table 4-2).7 In view of the fact that the above national variations will be explained in their respective historical and socio-cultural contexts in chapter 6, the following sections in this chapter will concentrate on illustrating the trends shared by the majority of Dar es Salaam's urban refugees from the Great Lakes region. The last point in the section on local interactions pertains to religion.8 Contrary to my expectations, during the survey, religion was never explicitly mentioned as an obstacle to establishing relationships with Dar es Salaam's Tanzanian residents, the overwhelming majority of whom is Muslim compared to only 10% of the respondents (see chapter 3). Yet, from the informants' experiences, I gathered the sense that there may be differences in the level of hospitality between the two religious communities, the reasons of which I subsequently explored to some extent. 7 This variations as well are discussed in chapter 6. 8 Gozdziak and Shandy recently pointed out that religion and spirituality, despite their importance "as a source of emotional and cognitive support, a form of social and political expression and mobilization, and a vehicle for comunication building and group identity" have generally been understudied, even neglected, by scholars in the field of refugee studies (2002:129).

PAGE 199

183 4.1.1 "Too many questions" The observation that "Tanzanians ask too many questions" has its historical roots in the state policies implemented during the Nyerere era, which were "designed to make the citizen more security conscious" and expected him or her "to be all ears on behalf of the state" (see chapter 2). From the following account, it would appear that this situation is rather particular to Tanzania. Pierre compares the situation in Congo, where he was commuting between the refugee camp and the school in down town Uvira on a daily basis for four years, with his experience of the last five years in Dar es Salaam. There is a big difference. In Uvira, even though the situation wasn't good, we didn't have the same fear to make us hide [as here]. You could tell anyone that you are a refugee, even if you were with, let's say a military person or the governor himself, if someone asks you, you simply reply: "Yes, I am a refugee." That is not the case here. Here, they will use any means possible to find out whether you are a refugee. Of late, they have even started to talk about denying us our right of asylum (dire que nous ne sommes pas des rfugis). So you have to hide really well in order not to be captured by immigration. Because people here are really, I don't know how to say . They can come and talk to you. You, you think that you are having a usual conversation with someone. He, on the contrary, he came to probe (faire des sondages), to know "You, who are you? How are you? Where do you come from? Which province do you originate from? Where do you come from?" And if he suspects that you are a refugee, he won't tell, you'll just continue to have a good conversation. Then, one day, you'll just see the police coming to take you. Lately, not even two weeks ago, they took a young woman, a Congolese, on the other side from where I work, the immigration people did. The people around her (entourage) had gone to the police to pass on the information that next to them lives a Congolese refugee. Here, they say, even if you came before the events [of 1993], they say you are a refugee." Not only neighbors, but basically any passerby is considered a possible informant. Many a time, it was pointed out to me during interviews that "all street vendors and bar maids are paid police informants" or polisi wa upelelezi ("plain clothes policemen"). This may be an exaggerated statement, yet it is a well-known fact that many Tanzanian citizens complement their meager salaries by gathering information on foreigners in the hope of

PAGE 200

184 selling it to a government official or police officer. One Congolese informant stated it succinctly, namely that: The polisi wa upelelezi, they are everywhere. And if you don't get along well with your neighbors, you will have problems. That's the way it is. The current atmosphere of an emerging xenophobia has aggravated the situation even more, as refugees are portrayed very negatively in the media and singled out as criminals (see chapter 2). To complicate matters further, in addition to circumventing questions and providing satisfactory answers, many refugees struggle with the local Swahili. 4.1.2 Language The Swahili spoken in the Great Lakes region is considered a dialect compared to the Standard Swahili of the Tanzanian Coast. The implementation of Swahili as the language of instruction in all Tanzanian primary schools during the Nyerere era ensured not only that all Tanzanian citizens spoke a national, African language, but it also contributed to its standardized use.9 Zanzibari Swahili was officially chosen at independence as the Standard Swahili to be spoken in the whole country, and after four decades, even the majority of the inhabitants of the smallest villages speak a Swahili--albeit with regional accents--that respects the same grammatical rules and uses virtually the same vocabulary. In a country like Tanzania, where most people do not possess a national identity card, it is often said that "Swahili is the identity card of the Tanzanian." In other words, someone who speaks (or manages to give the impression of speaking) "good Swahili" 9 The (dis)advantages of maintaining Swahili as the language of instruction in Tanzanian schools have increasingly become an issue of national debate (Roy-Campbell 2001).

PAGE 201

185 will, in every day life, not be subjected to extensive questioning and be considered "a real Tanzanian." Charles, for example was told upon arrival that: When you know Swahili, people will start thinking you are a Tanzanian. I was told to learn the Swahili, so that when I knew it really, really well, I'd be able to pass for a Tanzanian. According to Daniel, this is easier said than done: This accent, it is not easy to learn (acqurir) because it is the English accent. They use many English words, and the accent is not easily acquired. I've been here for two years, and everybody in my surroundings is Tanzanian, and still it is difficult. As soon as I speak, people say: "No, no, where are you from?" I don't hide, I say that I am from Rwanda. Bonaventure feels more confident, and has developed a knack for imitating regional accents. For me, in order for a Tanzanian to detect that I am not Tanzanian, he will have to try real hard, because I put a lot of effort into speaking well. But then, they still have this habit of asking "Where are you from ?" so as to place your Swahili. We try to avoid that type of question, because you are not going to tell just anyone "Me, I'm not Tanzanian. I'm Congolese." Often, in these circumstances, we say that we originate from Kigoma, and it may work. Since I travel a lot, for example to Arusha, I try to imitate the Arusha accent, or pose as a Kenyan, because I know their ways. Contrary to the situation in Tanzania, and despite being one of the four official languages of the Democratic Republic of Congo and the lingua franca of Eastern Congo, Swahili was never instituted as a standardized language of instruction, whereas French was. As a result, through popular usage, the Congolese version diverged further from the Tanzanian Swahili, not only through the import of grammatical rules from other regional Bantu languages, but especially through the frequent use of French words. For example, in the following excerpt from one of the interviews in Swahili, all French imports are underlined and all non-Swahili conjugations or those not respecting Standard Swahili grammar are in italics:

PAGE 202

186 Wakatueleza twende mucamp de rfugis Camp ya Nyangezi iko kutrente kilomtres na Bukavu. Tukabaki hapo mais finalement hatukukuwa tunapata chakula kizuri, maisha ikawa magumu. Mzee hakupata tena dawa, baridi, akaugua rheumatism Pia alikuwa anaugua tension akawa oblig kutuacha. Tukaona vie ya pale inakuwa ngumu, haupate chakula rgulirement Mwaka nonante quatre bamilitaires ya Kagame banatoka ngambo ya Rwandais wanakuja kuua watu mucamp ya Nyangezi. Finalement tukaona maisha inakuwa difficile Tukatoka, tukaenda kutafuta nyumba kidogo mucit kwa sababu ya inscurit Eliminating all use of French words and replacing them with their Swahili equivalents requires constant concentration on what one is saying even in the most menial daily activities. A small mistake can have disastrous consequences, as illustrated by Melanie's experience: I was on my way to the shop, and I said to the shopkeeper "Give me savon" [French for soap] instead of sabuni [the Swahili word for soap]. The people who heard me said "That one, isn't she Mzaire [Swahili for a Congolese person] ? She's a refugee." . So they went to call the police to arrest me. The police came and mama Joli [Melanie's employer] came to talk to them to let me go. Contrary to the Congolese Swahili, the Burundese Swahili is closer in accent to one of the regional variants of Tanzanian Swahili, because of the historical ties with the Waha of Western Tanzania, who speak a language very similar to Kirundi. As such, Burundi refugees whose accent is being questioned by Tanzanians manage rather easily to pass as Waha from Kigoma. Pierre has noticed that: People from here [Dar es Salaam] know that people from Kigoma, their Swahili doesn't sound like the Swahili from here. The Swahili from there [Kigoma] is very much like the Swahili from Burundi. Another Rwandese informant observed that: Burundese Hutu, there are many here, and they always manage to pass as Tanzanians, because their Swahili passes as a Tanzanian Swahili. The Rwandese Swahili is not good at all.

PAGE 203

187 In fact, in Rwanda before 1994,10 Swahili was not widely spoken. Thus, unlike Congolese and Burundese nationals, who were speaking a Swahili dialect before their arrival in Dar es Salaam, many Rwandese respondents had no prior knowledge. Jeanne remembers that: I had many problems [in the beginning] . The Swahili, I learned it here, I didn't know it [before coming]. Even you may speak better than I did. And because I didn't know it, I spoke/talked very little, and so it took a long time to learn. Consequently, the Rwandese refugees are the ones that struggle hardest with the language. This is reflected in the fact that more than twice as many Burundese compared to Rwandese manage to pass as Tanzanians (17 versus 8%) and that Rwandese wishing to conceal their nationality will twice as likely resort to pretending to be Congolese than Tanzanian (15 versus 8%) (Table 4-2). Cathrine, who lost both parents and her two sisters during the genocide, often poses as a Congolese. She explains why it is better to pretend to be a Congolese refugee than a Rwandese. If I tell people that I am a Rwandese, they will be afraid of me. They think maybe I can kill. They know that Rwandese are persons that kill each other at home. We kill each other. So here, we say that we are Congolese so that people are not afraid of us. While claiming that one originates from Kigoma, when found out with a strange accent in Swahili, is a workable option, especially for the Burundese, it is not the perfect solution. Christine says that: Even if you say you are from Kigoma, they can still capture you. Because for them, people from Kigoma may come from either Burundi or Congo. [In other words] they have doubts about their nationality (wasiwasi ya urahia) . Sometimes, even people that are really from Kigoma are arrested, they are being told: "You people, you are Burundese." Yes, that is why you get scared to communicate/make contact 10 Many Rwandese complain that since 1994, the Swahili and English of the Tutsi returnees from Uganda and Tanzania superimpose on Rwanda's traditionally official languages of Kinyarwanda and French. Currently, Rwanda recognizes all four of the above languages as its official languages.

PAGE 204

188 with people. Even if you meet a Burundese, you will hide, so he won't know [that you are too]. But inside my heart, it hurts. Language as a primary tool of communication may turn into an impediment for establishing new relationships in particular situations, such as the one of undocumented refugees in an increasingly xenophobic country of asylum. Even as many of the refugees eventually learn how to cope with the local Swahili variant, having to be alert and concentrate on one's use of language and choice of words all the time, compounds to an already stressful situation. 4.1.3 "Lack of respect" Once the practical hurdles of communicating with some Tanzanian residents and avoiding close contact with others eager to denounce "illegals" to the authorities, are overcome, another hurdle of a more emotional nature emerges. Refugees need to cope not only with their undocumented situation in exile, but also with the loss of their societal positions at home. The insecurity of this situation has made them very sensitive to the ways they are perceived by others, and one Tanzanian custom in particular, is considered very offensive by near all respondents. Dieudonn, for example, does not generally have a problem with the Tanzanians, but one of their habits really troubles him. The Tanzanians are not bad people, no, they are not bad, but they humiliate the wakimibizi (refugees). When you are a refugee, you are put down (minimis). Whenever he sees you, just like that, he will say "Ah, wewe mkimbizi." (Hey, you refugee). While one in six of the respondents manages to pass for a Tanzanian national, the overwhelming majority states being called mkimbizi, a term perceived as matusi (insulting, offending), on a regular to almost daily basis, regardless of gender, nationality or age (Table 4-2). The word mkimbizi, as mentioned before, comes from the Swahili

PAGE 205

189 verb kukimbia ("to run away") implying negative connotations such as cowardliness, homelessness and dependency.11 Tacoli (2001) has pointed out that "in many parts of Africa, links with home areas are considered an essential part of one's identity" (148). In other words, an individual derives a large part of his/her identity from being a member of his/her family, especially parents12 and grandparents. Both self-respect and the respect received from others are dependent on belonging to and identification with one's familial roots. In East-Africa, these familial links are generally symbolized in the concept of nyumbani (home). Nyumbani literally means "in the house," but it is more than that, as one Congolese informant13 describes: Nyumbani is where you are born, . where the villagers/inhabitants know you and your family, all families from high to low. Yes, that is the meaning of nyumbani. Where you know all the ways because you are born there. . My home is Congo, it is not here, because here, I cannot go see my mother, my mothers14 (bamama) and my uncles and aunts. .. Home is over there, where the graves of our grandmothers and grandfathers are, of our ancestors. The implications are manifold, because: In many African . refugee-hosting societies, territorially anchored identity is a sine qua non (sic) for the enjoyment of rights to property, equal treatment before the law, freedom of movement and residence, political participation, etceteras. (Kibreab 1999: 407). 11 In an ultimately failed attempt, Nyerere in 1983, suggested the term wageni wakazi, loosely translatable as "guest workers," to underline the refugees' contributions to the national economy of Tanzania (see also chapter 2). 12 A Kenyan friend of mine had for over a decade made every effort to conceal from his colleagues at work that he had been orphaned at the age of thirteen, because "It is shameful and people would no longer respect me" (Personal Communication). 13 Swahili is the lingua franca of Eastern Congo, therefore this Congolese informant's focus on the Swahili word nyumbani instead of the Lingala word mboka. 14 Mother's sisters are called mothers as well, the older sister is called mama mkubwa, the younger one mama mdogo.

PAGE 206

190 In this context, a refugee, as one "outside his physical context," becomes subjugated to "the original occupiers" who "have the right to exclude or deny entry to outsiders," and, if allowing entry to the refugees, "can impose conditions" (387). When these conditions explicitly imply denying refugees any access to resources, as is the case in Dar es Salaam, it becomes clear how the word mkimbizi has acquired its negative connotations, such as a dependent, good-for-nothing with no place, let alone status, in society. One Congolese informant recounts: They [Tanzanians] minimize us by calling us mkimbizi. For them it is another word for a worthless person (un ne-vaut-rien). That is what hurts me in my heart (me fait mal au coeur). I know that I am a refugee, but the word should be used in its proper sense, not to minimize someone. In addition, informants perceive that it is often children or youngsters hanging out in the streets, who shout after the refugees. According to them, this makes matters even worse in the African context of absolute respect for one's elders.15 Even 22-year old Melanie indignantly complains that: You cannot come across a Congolese child that uses insulting language or addresses you with such contempt. [At home] that child knows your mother, father, brother and considers them as his own family. He cannot insult you like that. In other words, geographical displacement is compounded by societal dislocation due to the loss of status and the perceived lack of respect by the host population. Kibreab (1999) succinctly puts it as follows: As long as territorially anchored identity remains the basis of membership and of apportioning rights, involuntary displacement by uprooting people from their places of origin undoubtedly constitutes gross deprivation and loss of "some part of one's very humanity" (407). 15 Disaggregating for age, the percentage complaining of the Tanzanians minimizing attitude steadily increases with the respondent's age: 10% of the youngest respondents, 13% of the ones between 26 and 35 years of age, 16% of the between 36 and 45 year olds, up to 26% of the oldest age category (Table 4-2).

PAGE 207

191 4.1.4 Religion As mentioned before, the large majority of the host population in Dar es Salaam is Muslim, contrary to the refugee populations from the Great Lakes region, who are overwhelmingly Christian. No exact statistics exist, but an estimated 60 to 75% of Tanzanians in Dar es Salaam adhere to Islam. When I asked Aristide, whether he had noticed a difference in the way Muslims and Christians perceive the refugees, he replied in a very general manner: There is no difference because one thing that these people [Tanzanians] have managed to do is going beyond their tribes . Tribalism is not dominant, nor is religion. This statement finds confirmation in the survey administered to a sample of 216 Tanzanian residents, that indeed found very little differences between the perceptions of the followers of the two religions with regard to the refugee population (see chapter 2). Considering however that, on average, at least one in six of the respondents approached their church or mosque for assistance (Table 3-7), I wondered whether the official Tanzanian ideology was reflected in the support provided to the refugees by these religious institutions or their respective adherents. If this proved not to be the case, religion would have to be considered as affecting the establishment of social ties between the refugees and a large segment of the host population. The following is an excerpt of a discussion between Christine, a Muslim woman, and a female Congolese informant who is a member of the Pentecostal church. The initial question was whether Muslim refugees seek support from the mosque to the same extent that Christian refugees go to a church when in need. Christine: It is different. Those who go to the Church may be helped, but for Muslims it is different. Those that guard the mosque, they don't have means . You see, the Christians have the wazungu [Here: missionaries from Europe and

PAGE 208

192 US] from outside, they are their benefactors. But the people of the mosque don't have the means to sponsor you . There is the Tanzanian Muslim who will listen to you in peace if you are a Muslim. He can give you food, but [it is true] he won't like to hear that you are a mkimbizi. [And yet] the will to help someone doesn't come from the religion . Even at the mosque they will help you, except they fear that one day you may be arrested. Congolese woman: So Muslims do not have the love of unity [solidarity]? Myself, I go to the Church in Sinza, the Pastor himself tells the refugees to stay behind after Mass. He wants to know who is living without papers in this country, so that he can defend them [should they be arrested]. But for you people in the mosque, it's a whole different thing? Christine: We at the mosque, we pray and then we go home. Sometimes, you may receive a little food if you need some. Congolese woman: But if for example you are arrested, is there solidarity between you [the Muslim community] to go see the mosque elders for help [to intervene with the authorities]? Christine: No, we cannot. The fact that less support appears to be forthcoming from the mosque than from the church should be contextualized in the larger Tanzanian picture. Muslims are over-represented among the economically disadvantaged and underrepresented among the politically influential layers of Tanzanian society. In fact, a Muslim female compatriot of Christine has the same reflection. She compares the Tanzanian situation with the one in Burundi: You know, at home, there are few Muslims, especially among influential persons. They are like almost not there, only 2% of Burundese are Muslim, 98% are Christian, whereas here almost everyone is Muslim . they may reach even 60%. And even those who do their best and study hard, will not progress because they are Muslim [alluding to discrimination]. The difference is that you may hear of a church where they help Burundese refugees, but as it is not my religion I cannot go there, that would be difficult. For example, they can go as far as Bagamoyo [60 kms north of Dar es Salaam] and be helped by the Christians there. But you [as a Muslim], if you were to go there, you wouldn't be helped, instead they would try to convert you. And that is not possible because each person has his own faith.

PAGE 209

193 Thus poverty and the Muslim perception of being discriminated against may play a larger role in the relationships between Muslim Tanzanians and the refugees, especially the Christians, than religion itself. Keeping this in mind, Amlie's (Pentecostal) lamentations, following the question whether relations with Tanzanian Christians are different than with Tanzanian Muslims, can be put in an appropriate perspective. There is a difference. Muslims, even when you live with them, they don't like to see someone do well (maendeleo ya mtu). Let them not see you well-dressed or your child. He will move everything upside down to do something bad to you like witchcraft/sorcery.16 Christians are better, they don't have these thoughts . Well, a few may, but not to the same extent like those Muslims. The Muslims exaggerate, they will do their utmost [to hurt you]. A Christian considers and loves you as his Christian brother or sister. It is only when he doesn't have the means to give you something, for example food, that he won't. Not like that Muslim who refuses to cook meat or fish, and instead will have you tortured by witchcraft at night. We have a difficult life here. A Muslim can come to your place to watch how you are living, what you eat, whether you have a bed or not, and then he will laugh [at you] and go tell his friends. They will start despising you, and insult you as well, sometimes even fight you. Yes, that is the way they are, very provocative. Very different from Christians. From the above, it has become clear that it is not religious affiliation in itself that brings potential friction in social relations. Ultimately, it is issues such as poverty, illiteracy, economic disadvantage, lack of political power and the perception of being discriminated against because of one's religious affiliation that play a considerably more important role in the relationships between people of different faiths. The experiences recounted by the Congolese, Burundese and Rwandese protagonists and others testify to the difficulties the Great Lakes refugees encounter in Dar es Salaam at the social level. Many arrive by themselves without contacts and struggle to find their way in an unfamiliar city, yet all are confronted with the reality of 16 Several informants assured me that witchcraft is not playing a larger role in their lives in Dar es Salaam compared to life at home before the flight. For the enhanced role witchcraft has been playing in Tanzanian society over the past decade see Sanders 2001.

PAGE 210

194 being strangers, and illegal ones at that. The trust in the predictability of life as they knew it before the flight, shattered by the necessity to leave homes, relatives and friends behind, cannot easily be restored in a situation where the refugees have to be wary of each friendly stranger, lest they run the risk of being denounced to and arrested by the immigration authorities, albeit deported to their home country. In addition, the lack of respect by their new environment for the losses undergone, through the persistent use of the word mkimbizi and the current generally refugee unfriendly climate in Dar es Salaam, compounds further to the frustration of being in a situation of forced migration per se. At the same time, in the midst of this traumatic turmoil, when the refugees are trying to cope with emotional and social losses, matters of very different nature unrelentlessly ask for their immediate attention, namely, the necessity to foresee daily needs in the absence of any humanitarian aid or legal documents. 4.2 Economic Endeavors After decades of Nyerere's socialism, Tanzania underwent radical economic changes under the subsequent presidents Mwinyi (1985 to 1995) and Mkapa (1995 until now). The liberalization of the economy and the IMF-led restructuring had profound effects on the lives of ordinary Tanzanians, and not all for the better. It is estimated that while currently 700,000 people enter the labor market annually, only 30,000 of these find work in the formal sector (Kironde 2000: 27).17 In Dar es Salaam, half of the economically active population works in the informal sector, and three quarters of these are estimated to be self-employed (Ibid.). As self-employed, they are extremely sensitive 17 Because of recent massive redundancies in the governmental and parastatal sector, and the fact that many industries close down or work under capacity, it is expected that increasingly fewer persons will be able to join the formal work force (Kironde 2000: 27).

PAGE 211

195 and vulnerable to the turbulences of an economy in transition, and things are not looking good. One citizen sums up the situation of the past decade: In 1991, business was very good . By 1995, things began to decline. 1996 wasn't so good, and 1998 was just mediocre, but better than today . Today, business is extremely bad for everyone . because we're all locked together by one thing called "money." And the circulation of money--even at the national level--the nation itself is sticking (kwama) and is bemoaning the fact that there isn't any money . We're all ruled by the circulation of money (Sanders 2001:168). As described in the previous chapter, a majority of the urban refugees (60%) are also engaged in self-employment or trading activities (Table 3-4). The success of either type of activity depends on customers' propensity to spend money, hence their often heard complaint--in line with the Tanzanian quote above--that "there isn't enough money to go around here." Rendering the challenge of making a living under these circumstances even more insurmountable than for the Tanzanian residents of Dar es Salaam is the refugees' eternal problem of residence and/or work permits, without which they are "fair game" to any police officer or immigration official. 4.2.1 Permits In order to obtain a work or residence permit, refugees have applied a multitude of strategies, such as, following the official application procedure, intermarrying Tanzanian citizens or locating Tanzanian "adoptive" parents. If all has failed, the only option to remain out of prison is to pay off the arresting officials. The following testimonies illustrate the different coping strategies applied to what is one of the foremost pre-occupations of the urban refugees (Table 4-1 and Figure 4-1).

PAGE 212

196 4.2.1.1 Applying for a permit Aristide remembers how, in 1995, it was still reasonably easy for a refugee to obtain the work permit for Dar es Salaam. In order to apply for a work permit to be able to teach French, I went to the Ministry of Home Affairs, they gave me the work permit without a single problem, I don't know how God helped me. I went without a problem . in 1995 . I didn't even pay a single cent. . I think they weren't strict [at that time]. [Also] teaching is a job that doesn't give much money and in addition there are few Tanzanians that can teach French. Since 1995 however, Tanzanian policy stipulates that refugees need to apply for work permit that costs $600 yearly. Considering that the average Tanzanian salary is less than $100,18 this means that applicants are expected to spend half of their yearly salary to be allowed to work. Clearly, a work permit is not affordable on a local salary, and those very few refugees having work permits are usually sponsored by close relatives in Europe or the USA with well-paying jobs. 4.2.1.2 Marrying a Tanzanian citizen An alternative method to obtain a Tanzanian residence and/or work permit is by marriage to a Tanzanian citizen. However, in practice, while residence permits are issued fairly easily to non-Tanzanian women married to Tanzanian men, non-Tanzanian men married to Tanzanian women have a hard time obtaining the same legal document. One of the Congolese informants came to Tanzania in 1995, subsequently married a Tanzanian woman and has been struggling ever since to receive the resident's permit. At one point, he received a letter from the Ministry for Home Affairs telling him to go to one of the refugee camps and take his Tanzanian wife with him, while UNHCR in another 18 This is an estimate for Dar es Salaam, salaries in the rest of Tanzania are even lower.

PAGE 213

197 letter had stated to him that "your marriage gives you the right to reside anywhere you want" in this country. The contradictory statements of the relevant official agencies, have left the man in an (il)legal limbo. Up until now, I'm still here illegally. The temporary permit that had been given to me [by the Tanzanian Refugee Department] is no longer extended . At the legal level, there is even a trial pending at the court . Regarding the problem of my residency, work permit . Really, there is no local integration and what about protection? At any moment, I could be arrested. And what about the future? I have a family to provide for, my children need to go to school. The court case he is referring to was initiated by the lawyers of the Legal and Human Rights Center, who sued the Tanzanian government for gender discrimination in the application of the law pertaining to residence permits for non-Tanzanian spouses.19 The court's decision was still pending at the time of writing. According to some informants, there are considerably more marriages between refugee men and Tanzanian women than between refugee women and Tanzanian men. A Burundese, female friend of Christine's explains that: Our men marry here in order to find a way to make ends meet more easily. Although we, as women, may be married to Tanzanians, our husbands give us a hard time, call us mkimbizi . More men, whether Burundese and Congolese marry Tanzanian women [than vice versa]. [A refugee man] who is without documents, he'll decide to marry a local woman. For example, those Congolese with a salon, they'll marry a Tanzanian so that if they are arrested, their wife will be there to defend them. At the same time, our brothers tell us "We live with her [our Tanzanian wife], but the day we go back home, she's not coming along." Marriage to a Tanzanian woman, even as it does not produce the necessary permit to the non-Tanzanian husband for reasons explained above, does provide a social connection that has frequently proven useful in dealings with the Tanzanian authorities. Often, so I was told, it was possible to convince the arresting authorities to release the refugee by 19 The Guardian, May 25, 2000.

PAGE 214

198 involving his/her Tanzanian friends and/or relatives. The same principle of social and/or familial local connections lies at the root of an alternative strategy, namely finding an "adoptive" Tanzanian father. 4.2.1.3 "Adoptive" fathers Wealthier refugees, so I was told, often resorted to locating an older, needy Tanzanian man and paying him to testify to being their father, in order to obtain the necessary documents from the authorities. However, Christine warns, this modus operandi is not without difficulties. Refugees agree with a poor man from here, one that doesn't have any business or money. But once you have used him, and you have a job or a business, for him to extort money (escroquer) from you is very easy. If you have money or a car, you think that man is on your side, but then you find out he's doing everything possible to have you arrested. In other words, the scheme may backfire when the Tanzanian starts initiating the arrests by informing the authorities. Each time, the refugee will have no option but to have the Tanzanian testify on his behalf as "his relative," a gesture necessarily handsomely rewarded lest the refugee goes to jail. 4.2.1.4 Staying out of prison Being arrested for lack of the necessary documents is a constant worry, hence the refugees' endless schemes and attempts to obtain some type of legal document. While only 9% of the respondents have at one point or another been really arrested and taken to court (Table 4-2), many more are regularly harassed by the Tanzanian immigration officials or police officers. For lack of Tanzanian relatives or friends, usually a little "gift" may suffice to solve the problem. Charles recalls that: One day, the landlord took me to the police. . Apparently someone had informed the police that I was there. And as I had never gone to report my presence to the

PAGE 215

199 police,20 they had threatened the landlord, that's what he told me. So he was forced to take me to the police station to explain why I was here. So he comes in the morning "Come, let's go to the police station." We arrive there, they ask me to show all my documents, the [UNHCR] letter that shows that I am a Rwandese refugee. They told me I do not have the right to reside here, in Dar es Salaam, so I explained to them that my wife was pregnant and ill . and that really, I couldn't go to the camp, and that I was waiting for the reply from UNHCR. They understood, they let me go . Of course not without asking for money, how can they leave you alone without asking for money? One of Charles' compatriots says he saves his hard earned money for just such an occasion: I knew a bit of Swahili, now it is better. But even when I speak to a Tanzanian from the accent he'll immediately tell I am Rwandese. Unfortunately, I met with the police and immigration officers, they arrested me, not once, but four times, because I don't know Swahili well. [Q: Were you taken to prison ?] No, when I have worked, I save my money, and then, when I am arrested, I give them that money. It happens that refugees are warned beforehand by friends that immigration officials are looking for them. Aristide, in 1997, through funding by the Ndadaye Foundation, was taking classes for a degree in Commercial Diplomacy when, those guys who were handling our documents, I don't know why they went to the Ministry of Home Affairs. They mishandled some documents, they were trying to find the registration of the Foundation here, so they had to disclose everything. So the Immigration Department came, up to the class. Fortunately, I got the information that "these people are looking for you now" a day beforehand. So the next day, I didn't go to the college, but I heard that they came to the class, looking for me. I wasn't there, they looked in different corners and rooms and I wasn't there. I didn't go to the college for three days. The above testimonies illustrate why not having a residence and/or work permit is considered such an important issue by respondents (Table 4-1), as it not only drains already meager resources, yet it also affects them in their endeavors to go about their daily activities. 20 Tanzanian law stipulates that all non-Tanzanians entering the country must report to the nearest local authorities within 7 days after their arrival (see chapter 2).

PAGE 216

200 The two other concerns of an economic nature that were quoted by respondents as major differences between life at home and in exile are "life in Dar es Salaam is expensive" and "job opportunities or opportunities to make money are scarce." While men are worried in significantly larger numbers about making money than women (48 versus 36%), significantly more women than men complain about the high cost of living (49 versus 34%) (Table 4-1). In the end, however, both pre-occupations constitute two sides of the same coin, namely, how to make ends meet. 4.2.2 Kupata riziki ("Making ends meet") The previous chapter already alluded to the fact that the current pattern of employment activities of Dar es Salaam's refugees are not reflective of their respective education levels and/or former professional experiences. Rather, for lack of any material assistance from humanitarian agencies or procedural leniency from the government with regard to residence and/or work permits, the urban refugees have necessarily learned to take advantage of any type of opportunity to earn an income. In all, three quarters of the respondents reported being able to make ends meet through either self-employment, trading activities or paid employment (Table 3-4). 4.2.3.2 Self-Employment As mentioned before, two out of every five respondents is currently self-employed, a tripling of the proportion that was self-employed before the flight (Table 3-4). Some refugees found ways to exercise the same profession as they did at home, others were forced to learn a (new) trade. Cathrine, for example, braids hair, a profession with which she has been able to provide for herself since 1996. Upon arrival in Dar es Salaam, she almost immediately found a salon where she could exercise her profession at

PAGE 217

201 a fixed rate of commission. The relatively high cost of life, however, encouraged Cathrine to keep her eyes open for a better opportunity to come along. My first job was at MwanaNyamala. There I got to know a Tanzanian woman,21 she advised me that I couldn't go on staying in the guest house "You will spend all your money there, how will you live ?" She helped me, and I went to live with her in her house, her name is Leila. While I was working there [in MwanaNyamala] I didn't get a salary, I was trying to find a job elsewhere, but could not find one. Leila found me the job in Mwenge (where Cathrine works now). After two or three months of working in Mwenge, I was able to move out and rent a room. Here, the rent for a room with water and electricity is 10,000 shilling22 monthly. But as I didn't have that kind of money,23 I looked for a room of between 4,000 and 5,000 without water or electricity. That's where I live now . In Mwenge, also I don't receive a salary. When you braid someone's hair, you split the money fifty-fifty with the owner of the salon . You can do hairbraiding for 15,000, 10,000 or 5,000, it depends. . On a weekly basis, Saturday and Sunday included, you can have two or three clients. Hairbraiding one person may take up to 8 or even 10 hours, limiting the number of persons a hairdresser can braid in one day. The price obviously depends on the complexity of the hairdo and the time spent.24 Coming up with the yearly rent of a room is a major obstacle, and often appeal is made to others for a loan. Bonaventure, for example, was fortunate in meeting fellow church-goers willing to extend a helping hand in more than one way. Not only did they lend him the yearly rent for a room, they also assisted him by making the necessary contacts for a job, on condition that he was prepared to learn a new trade.25 21 Cathrine calls her dada, a Swahili term which literally means sister, but is often used to indicate a female friend of around the same age as oneself. 22 At the time of doing research, one thousand Tanzanian Shilling was approximately one US$. 23 It is customary in Dar es Salaam for tenants to be asked by the landlord to pay the rent of 12 months up front, whether it concerns a room with a monthly rent of TS 5,000 (US$5) or a diplomatic mansion, the rent of which can easily reach US$3,500 monthly. 24 I learned that Tanzanian women of certain means, e.g., photo models, easily spend up to 35,000 or 40,000 shillings to have their hair styled. 25 At the moment of leaving Congo, Bonaventure was pursuing a degree in nursing.

PAGE 218

202 To pay for the room, fortunately I could count on some friends to help me. They are persons that I met through the faith [at church], I explained to them my problem, and they understood. They gave me the money to pay the yearly rent. I have been on good terms with these persons, and they also helped me a lot in other ways. It is through them that I met an old man (un vieux), an electrical engineer, who taught me how to do the [electrical] wiring of a house. Thanks to that, I was able to earn 2,000 shillings per day, sometimes even 3,000. Pierre, as well, had to learn a trade upon arrival in Dar es Salaam as he had been concentrating solely on obtaining his secondary school diploma throughout the troubles in Congo and Burundi. His friend Jacques was able to help out. As soon as I arrived, I started to learn the trade [of tailor] in his workshop (atelier) . I started to make trousers, shirts, and skirts. When I had learned to do all this, I told Jacques: "Well, the time has come to give me something, because now it is no longer a matter of learning." Until then he was providing lunch only. Usually, in other tailor shops, you get 40%. So, if you make a blouse for 10,000, he [the shop owner] has to give you 4,000. Jacques proposed to give me 2,000, or 20%. I accepted. Among my clients I had made some friends, such as Ashura and another woman. I asked them to look for another job for me, because what I received from Jacques was insufficient . Ashura found me a job in the tailor shop of a Tanzanian. She asked me if I was sure I wanted to change, because even if Jacques was giving me only 2,000, at least at the end of the month I received the correct amount of money. [Where I am] now, I received only parts of the money that is owed to me. That tailor, instead of 20,000 she pays me 5,000 or 10,000 . So instead of bettering myself, I still have problems. But I have accepted that that is the reality of life. Man (l'homme) necessarily has to go through the difficulties of life. Other urban refugees may be of an age where it is not so easy to learn a trade or find someone willing to teach them. Many of the older female respondents, usually divorced, like Euphrasie, or widowed, like Amlie, try to get by preparing and selling food on the street to passerby's. Euphrasie, for example, who used to be an agriculturist at home, learned how to make a living out of selling fried fish that she bought fresh at the fish market. Those people who sell fish, do so in open air. So you observe them to learn how they set the price. The first day, I asked someone to buy fish for me, and I repaid him afterwards [after frying and selling the fish]. The second time I went there, same thing. After that, I decided to buy myself and participate in the bidding . It

PAGE 219

203 went well, I told someone that I did not understand their way of buying. He couldn't refuse [to explain me]. They have a kind heart here in doing business. He took me with him and asked: "What price are you looking for ?" So, you tell him how much money you have, maybe 5,000 or 3,000 shillings. He buys for you the quantity for the money you have. He tells you "Bring the bucket," and pours the fish . [After frying the fish] I put a table outside, and piled the fish on it. When someone comes by, you tell him the price and he goes with it. I was selling from home right here . Some days, with a good buying price, you can make up to 1,000 shillings of profit.26 There was this one woman over there in MwanaNyamala, who fried fish for a hotel. We agreed that I would bring her the fresh fish, I did that, she fried the fish herself and then gave me my money . That woman was a Tanzanian. Throughout people's testimonies on their coping strategies to make ends meet, it becomes increasingly clear that despite previously cited frustrations with Tanzanians in general, a certain level of cooperation, albeit friendship, develops with Dar es Salaam's Tanzanian residents at the individual and personal level. Cathrine received accommodation from and obtained a good job opportunity through her Tanzanian friend Leila. Bonaventure was assisted by his Tanzanian friends from church with regard to the payment of the yearly rent, and ultimately received training and a paying job from a Tanzanian electrical engineer. And finally, Pierre as well, was able to change jobs through his Tanzanian connections while Euphrasie was introduced to the trading routines of the fish market by a Tanzanian passerby. All but Euphrasie were able to circumvent the officially required work permits or business licenses by making arrangements with established Tanzanian businesses or workshops and work on commission, as do the majority of self-employed refugees. 4.2.2.2 Trading activities In trading activities as well, Tanzanian contacts are indispensable. Jeanne, as the second wife of a Tanzanian man, does not have to worry about residence or work 26 Most respondents considered 1,000 shillings ($1) the minimum amount required to provide for the daily meals of a family of between three to five persons.

PAGE 220

204 permits. In addition, her husband provides her and their child with two rooms, inclusive of running water and electricity, and a housegirl. Notwithstanding, Jeanne endeavored to found her own business, but experienced how difficult things can get. I opened up a salon. As soon as it opened, I was robbed. I had four dryers, they broke in and stole everything. [Before that] I had been doing business, commuting between Zanzibar and Dar es Salaam, that was before these troubles with the CUF started27 . So I took money from the bank to buy dryers, and then they were stolen. Also, I knew this one woman who goes to Kampala and Dubai. I had asked her to bring me things, like handbags and shoes. I would put those in the salon to sell and make sufficient money so as not to have a single problem anymore. That woman went to Dubai, and never came back. She had my money, fortunately, it wasn't too much, it was 350,000 shillings. So I bought other things to put in the salon. One morning I wake up, they had broken in and stolen everything. Since then, I did not work again. I just stay here with the help from that man, my husband . When I need something, I get it. I have been trying to convince him to give me some capital (mtaji) so that I can restart my activities . I was robbed last year, for about one million shilling. Opening the salon was easy, [but] then everything was lost. That's how it goes. You may think you know someone, you take him in, but there is no certainty. The last sentence alludes to the fact, that Jeanne suspects one of her Tanzanian employees to have staged the robbery. Her remark that US$350 is not a lot of money, however, illustrates her socio-economic background. Unlike many of the other respondents struggling to make a dollar a day, Jeanne belongs to a different group of refugees, namely those that came with enough money to start their own businesses and engage in trade. One Rwandese informant, who lived in the Rwandese refugee camp between 1994 and 1996, told me that: After the forced repatriation of the Rwandese in Tanzania in 1996, quite a few [Rwandese] stayed in the refugee camps of the Burundese.28 After two or three years, they came to Dar es Salaam and then registered as if they just arrived. They have means, because they have been doing business in the camps. If you look at 27 In February 2000, major political riots took place in Zanzibar. Around 100 demonstrators were shot by the police, and thousands fled from the islands of Pemba and Zanzibar to Kenya. Most have since returned. 28 See also chapter 1.

PAGE 221

205 the majority of Rwandese refugees over 40 years of age [here in Dar es Salaam], you will see that they are businessmen. As mentioned before (Chapter 1), it was unfortunately this category of refugees that refused to meet with me and be interviewed. Nevertheless, Jeanne as well, confirmed at the end of our conversation, that "many of the refugees here in Dar es Salaam, they are rich, they have means." These remarks, coming from two Rwandese informants, pertain mainly to the Rwandese refugees. As explained before, the older Rwandese refugees are generally suspected to have been part of the masterminding of the genocide, as members of the Hutu intelligentsia and the politically and economically powerful elite under the Habyarimana regime. The fact that they have sufficient means to pay the $600 annual fee for a residence/work permit, also explains why they are to a far lesser extent than the two other nationalities pre-occupied with the lack of official documents (Table 4-1). 4.2.2.3 Paid employment The third type of activities most frequently exercised by the respondents I categorized under paid employment (Table 3-4). They include formal employment (such as Aristide who is a marketing executive), informal (such as Melanie who is a housegirl) and part-time activities (such as Dieudonn teaching private French courses). Melanie, who was in her last year of secondary school when the war broke out in 1998, became a housegirl for a Congolese woman she met through the old lady who provided her with accommodation on her first day after arrival in Dar es Salaam. I explained to her [mama Jolie] that I am Congolese and asked her to help me. She helped me trying to find my uncle, but to no avail. People said he was there [before] but he left in the meantime . So I started working for her . I wake up in the morning, bathe her little son, he goes to school, I dress him, scrub the floor, wash the dishes, fetch water and go to work in the garden/field (shamba). When I come back, I eat. I also wash the clothes of her son.

PAGE 222

206 Dieudonn stated bitterly that he "does not work," he only "gets by" (je me dbrouille) through teaching French to individual pupils. I teach two hours a day, but not every day, only three times a week, thus six hours . Some are secondary school students, others primary school students . .For example, the child of mama Sebastien (a neighbor) I teach her from 10 till 12, then I leave there and go to another girl who I teach from 12 till 14h. . So each child receives six hours of schooling per week, but I have more than one student, so all in all I teach 14 hours per week . I ask a monthly fee . Myself I am not expensive, I ask only TS15,000 per month, so that I can have many clients. You know there are other Congolese that teach per hour, and they ask at least TS2,500 per hour, that's too expensive. I prefer to be paid per month. Aristide is the only one who has a job in the formal sector, but even there he experiences problems related to the implications of being a refugee in another country. Just two months after I completed the studies, I managed to get a job . I studied for three years, so in 2000 I completed. I worked with a private company--the government doesn't employ anymore these days--got a job as a sales representative, I worked for one year . The following year, I changed jobs, started working with a certain media industry, and there I even forgot that I was a refugee, I was completely integrated, got married, life went on, as other Tanzanians, except that for me there is no room for expansion. You see, you just eat and sleep. Other young men like me in the same company think of furthering their studies, for me that room [possibility] was not there. It is strongly closed . You see when you are here, you have your uncle who is in the ministry, you have your uncle there, you have exposure. Myself, I have no exposure, I am the father of myself, the mother, everything of myself. You see in Africa that's how we are. So my exposure is limited. I'm not much exposed to different opportunities. I remember that I started work at the same time as three of my colleagues. All of them are now taking their studies, higher further studies. And surely when they come back--two are abroad, one is here--with their Master's degrees, they'll be driving a very nice car, life will be OK . They will just be waiting for Jesus to come back and finish this world. Life will be OK [for them]. Aristide puts the refugees' frustrations into words very well when relating their job experiences. Their studies interrupted, Bonaventure and Pierre had no other choice but to learn a trade (electrician and tailor respectively) in order to make ends meet. Both Cathrine and Melanie also regret very much not being able to continue studying and being stuck in activities lacking in future prospects (hair dresser and housegirl

PAGE 223

207 respectively). And Dieudonn expresses his bitter frustration with the current situation by stating that he "does not really work," instead he "is merely getting by through fending for himself (se dbrouiller). Even the only informant enjoying a well-paid position in the formal sector, Aristide, feels limited in his ambitions because being a stranger in another country he lacks the social and familial connections necessary to further his career. Aristide's description illustrates vividly Turner's concept of "liminality," of being "being betwixt and between the positions assigned and arrayed by law, custom, convention, and ceremonial" (1969:95). Despite his advanced degree, perfect Swahili, marriage to a Tanzanian citizen, and obtaining a job as a marketing executive in the formal sector, he is unable to feel fully integrated in Tanzanian society and remains "betwixt and between." The multiple ways devised by the urban refugees to obtain work permits and make a living document their innovative approach to deal with an insecure and unpredictable situation (Marris 1986). And yet, it should be noted also, that whether it concerns "adopting a Tanzanian father," learning a trade, or bribing an immigration official, these behaviors are not (completely) unfamiliar to the refugees. They are behaviors that incorporate the cultural norms and values that were prevalent as well during life at home before the flight: e.g., the importance of kinship ties to have a place in society, the flexibility to adapt one's occupational activities in order to survive under adverse circumstances, and the deployment of a range of alternatives to cope with hostile of threatening state structures. Yet, while "set within the parameters" of the refugees' "perceptual frameworks" from before the flight (Oliver-Smith 1986: 16), these behaviors are innovative in the sense that there never was a need to devise or apply any of these strategies at home.

PAGE 224

208 In the end, however, despite the amazing range of innovative initiatives displayed by the refugees regarding obtaining legal documents and providing for their daily needs, the circumstances particular to their situation of forced migration in Dar es Salaam, prohibit any real local integration, even among the most educated ones such as Aristide, and as a result the refugees are stuck in Turner's phase of liminality. While managing to make ends meet through their inventive resourcefulness and working ardor, the lack of future prospects or possibilities to regain control and re-establish some sort of predictability and continuity of life are experienced as suffocating. One of the Rwandese informants sums up his frustrations: "Such is the life of a refugee. .. I don't have a life here, I exist like a stone." 4.3 Major Pre-occupations With the objective in mind to prioritize the major pre-occupations of life in exile as perceived by the respondents, I included in the survey the question "What is your major worry here, in Dar es Salaam?" (Table 4-1). Close to half of all respondents (43%) expressed the desire to see peace at home so that they would be able to return, while one in four (26%) worried most about their own or their children's education. Material needs (such as: money, a job, food and accommodation), aspirations to leave Tanzania for a third country, and the uncertainties of the situation of lost relatives or relatives left behind, were foremost on the minds of approximately one in every six respondents (18, 17 and 15% respectively).

PAGE 225

209 0102030405060AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Peace at home Continuing studies Money, job, food, etc. Leaving Tanzania Lost relatives Figure 4-3: Major worries in Dar es Salaam No significant differences appear between men and women's major pre-occupations. Comparing across nationalities, however, we see that especially the higher educated nationalities, the Congolese and Rwandese, considered lack of education opportunities a major worry in significantly higher numbers (30 and 40% respectively) than the Burundese 11%). The youngest generation, as well, of whom 73% self-defined as students at the time of leaving home, are first and foremost preoccupied with continuing their studies.29 The eldest age group, on the other hand, are significantly more worried about lost relatives (30%), as are the Rwandese (28%). During the mass exodus from Rwanda in 1994, more than 3 million people out of a population of 7 million left their homes. While most of them have returned in the meantime, a significant number either perished in Eastern Congo30 or continued to flee to other countries in Central and West Africa (see chapter 2). In view of the fact that previous sections already extensively covered people's views on returning home (see chapter 3), as well as the difficulties related to providing 29 In the same respect, the majority of their second reply to the question comparing life at home and in exile (49%), pertained to the lack of someone to pay for their studies rather than a lack of job opportunities (Table 4-1). 30 UNHCR at one point estimated the casualties to number close to two hundred thousand (UNHCR 1999c:Internet). See also map 1-4 in appendix a.

PAGE 226

210 for one's daily needs, the following will concentrate on the refugees' views and initiatives with regard to education, strategies to leave Tanzania for a third country and attempts to get in touch with, or obtain news from or about relatives. In addition, I decided to include experiences related to sexual exploitation and harassment as an important aspect of women's daily quest for food, accommodation and money. Even as this issue was not listed by respondents as one of their major pre-occupations, it is clear that young, unaccompanied and undocumented women in a large unfamiliar city run considerable risks in this respect when trying to make ends meet. 4.3.1 Education During the first years of independence under Nyerere, major succesfull efforts were made to make the Tanzanian educational system accessible to all citizens, and literacy rates soared to become among the highest in Africa, for both men and women. However, due to an increasing lack of resources, crisis set in from the 1970s onward. Whereas in 1961, 36% of all primary school graduates continued into secondary school, by 1987 this percentage had fallen dramatically to 3.8% (Roy-Campbell 2001: 81). Daniel lamented the situation in Tanzania by comparing it to Rwanda and Congo. It is the socialist system that played an important role here . they weren't given the opportunity to study, to better themselves through education (promouvoir dans les sense d'ducation) . One day, I was talking to a student in Makongo, he's in Form Three. Him, he considers himself really well educated. He says: "Next year, I will finish Form Four, and then look for a job." For him, that's it, being educated is reaching Form Four.31 Contrary to other countries, for example, in Congo, if you don't have your Diplme d'Etat,32 you're nobody. Here, with a degree from Form Four, you can have a job, because for them, it stops at the level of Form Four. So, really, compared to Congo and Rwanda, and even compared to Burundi, with regard to education, here, they're done for (c'est foutu). 31 Equivalent to a Primary School degree. 32 Equivalent to a Secondary School degree.

PAGE 227

211 This conception of the importance of education in general and of the low level in the Tanzanian school system in particular is prevalent among the urban refugees from the Great Lakes countries. Compounded by the fact that their children would have to cope with a different language of instruction and different curriculum (English or Swahili instead of French) when entering the Tanzanian school system, the situation prompted the parents to put their heads together and in 1999, the Systme Scolaire des Pays des Grands Lacs was born, also commonly known as "the Congolese School." The school follows the French/Belgian curriculum and has a primary and secondary school section, comprising in all about 267 pupils33 (at the time of research) who receive classes from around twelve, often professional, instructors. The school faces serious financial problems because the parents, who are mostly refugees, have difficulties raising enough money to pay the tuition fee of 15,000 shillings per child per month. The result is that teachers regularly receive only partial salaries. Charles teaches computer science at the school: like 12 or 13 hours [weekly]. . At the secondary level, there are approximately 10 instructors . I'm the only Rwandese, the others are Burundese and Congolese . .The students are about half Congolese, half Burundese. There are also a few Rwandese [students] but not many. The fact that the major financial sponsor of the school34 is a Congolese politician, who reportedly staunchly supports the Mayimayi rebels, has in the past created tensions with the Congolese Embassy. The Congolese Ambassador at the time, was said to be someone closely befriended to the then-president of Congo, Laurent Kabila, whom the 33 As is customary in Dar es Salaam, the students wear the school's uniform. This makes them recognizable to any passerby in the street as pupils of the Congolese School, apparently without problems. 34 At times, the Belgian and French Embassies have provided assistance as well, e.g., in the form of computers and other electronic materials, or schoolbooks, etc.

PAGE 228

212 Mayimayi rebels sought to remove from power. Several informants I spoke to at the school recounted how one day, Tanzanian Immigration officials raided the school at the instigation of the then Congolese Ambassador, and all instructors were arrested and escorted to prison. Fortunately, the next day the problem was resolved and courses resumed. Notwithstanding the political complications, the initiative shows the extreme importance the refugees attach to their children's education and the quality thereof. Some informants suggested that the lack of secondary schools in the refugee camps encourage refugees with school-going children to come to Dar es Salaam. Several newspaper articles illustrate the length to which refugee parents (Congolese in particular) are prepared to go to have their children schooled. In view of the fact that UNHCR funds only primary schools in refugee camps, refugee parents were reported to be selling part of their already meager food rations in order to be able to pay a number of secondary school teachers to instruct their children.35 4.3.2 Sexual Exploitation In the daily struggle for food, accommodation and money, it happens that young women, especially when unaccompanied by their parents or other older relatives, end up in prostitution. The only self-professed prostitutes I met during this research was a group of four Burundese women between 24 and 29 years of age. They were happy to participate in the survey, as well as to meet with me and talk about their lives at home and in Dar es Salaam, but I never succeeded in convincing them to be tape-recorded during ethnographic interviewing. Cathrine, I was told by other informants, frequently 35 The Guardian and The Daily Mail, November 2, 1999.

PAGE 229

213 complemented her hair braiding earnings by making trips to Morogoro. There she was rumored to "keep company" with white, male tourists, sometimes for weeks on end (leaving her two-year old son by a Tanzanian man with her Congolese boyfriend in Dar es Salaam), but she never mentioned this to me personally. Bonaventure observed that Congolese young women often "change" when they come to Dar es Salaam. These Congolese girls,36 when they arrive here in Dar es Salaam, they change completely. I don't know why. They have this tendency to forget their education. As a Christian, I have the habit of visiting people, not because they are my relatives or friends, but so that I can have some idea of their behavior. Especially, girls that are far away from their family, do they continue to behave according to the education they received from their parents? Some of them have really shocked me. They say "We didn't come to Dar es Salaam to get married, we came to look for a life (kutafuta maisha)." So going by what they say, I learned that our Congolese sisters no longer want to be married, instead their goal is to have money, and that at a time that there is too much risk involved in their way of looking for money. Because they run the risks of diseases [refering to AIDS]. Other young women, who are not looking to make money through sexual activities, may be vulnerable to unwanted male attention. Amlie, at one point, was almost evicted from her house, because one of her daughters refused to "accommodate" the son of their landlord. I have three daughters. There is this one young man, who again and again tries to get one of my daughters, and refusing him was the source of troubles. But my daughter refused him. One day, she woke up at 6 am to go fetch water. And he tried to rape her. All the people [neighbors] were awoken by the noise and the shouting. They asked "What's happening ?" My daughter explained it to them. She had already explained to me before that for a long time that young man had been asking her to make love. They struggled hard but the boy was overtaken. She had hit him with a wooden stick really hard. This boy was the son of our landlord. His father and the boy himself, told us to leave their premises at once. 36 Girls (filles) is the term used to describe women who are not (yet) married. The use of the term is not indicative of a woman's age, rather of her marital status.

PAGE 230

214 Fortunately, with the help of some Tanzanian neighbors and friends (including the local Ten House leader), Amlie succeeded in convincing the landlord to let them stay on in their room. Whether voluntarily engaged in income generating sexual activities or not, refugee women in an urban context are particularly vulnerable to sexual violence. Recently, Human Rights Watch (2002) published a report on numerous such incidences in neighboring Kampala (Uganda) and Nairobi (Kenya). There is little doubt that the situation in Dar es Salaam should not prove to be very different.37 4.3.3 Aspiring to Leave Tanzania Often the desire to leave Tanzania is closely linked with the aspiration to resettle in or migrate to a Western country, such as Canada, Australia, the USA or one of the Western European countries. UNHCR has a refugee resettlement program, but its selection criteria and procedures are very strict (see chapter 2). Only those few refugees whose security is considered to be acutely in danger in the refugee camps and to a certain extent in Dar es Salaam, are eligible for the resettlement program. UNHCR policy stipulates that in principle, all applications for resettlement need to be initiated from one of the refugee camps. Dieudonn, however, decided that, rather than going to the refugee camp, he wants to try to relocate to another country without UNHCR assistance. The idea of resettlement came before my [UNHCR] assistance was cut . but after I had been accepted by HCR. That is why I wrote to any country, and Australia agreed to accept my application . I decided not to go through HCR because they will complicate the process and tell me to go to the camp . so I never asked UNHCR assistance to apply for resettlement. 37 However, it was never my intention to actively pursue this type of information, hence only a few spontaneous accounts.

PAGE 231

215 The chances of being accepted for refugee resettlement by a Western country without referral by UNHCR are next to non-existent, and so most urban refugees desiring to leave Tanzania seek other avenues to do so. A Rwandese informant recounts that: Many friends of mine left, the majority to the Netherlands. Those [Rwandese] who come to Dar es Salaam have money, and they want to go elsewhere, like Europe. With KLM it's easy, they travel on a Tanzanian passport38 . Once they arrive in Europe, it's easy for the Rwandese, they just throw away the passport and apply for asylum. [They do it like this] because there is no other way to leave here . I know quite a few people who left here, some of whom arrived in the USA . I remember one guy, his brother is a US resident. He left here on a [Tanzanian] passport with a visa for Cuba, because that [visa] is easily obtainable here . Instead of going to Cuba, he traveled to the Bahamas. He was well prepared, and knew that there are agreements between the two governments so that Tanzanians don't need a visa [for the Bahamas], only a letter from the Ministry of Foreign Affairs. So he stayed there [in the Bahamas], and telephoned his brother in the USA. A lawyer went down to the Bahamas to get him [and bring him to the US] . Yes, when you have money, the means exist. Not everyone is successful, and I met quite a few refugees who tried to get to Europe on a Tanzanian or falsified passport and/or visa, but who had been found out upon arrival and sent back to their point of departure, where they arrived disillusioned and broke. 4.3.4 Worrying about Lost Relatives As mentioned in the introduction, both Rwandese and older respondents worry about lost relatives in significantly larger numbers than any of the other categories of respondents. Euphrasie, for example, has two of her seven children with her in Dar es Salaam, but when I asked if she had had news from the others since 1998, she replied: I can't lie, they haven't written to me yet. . Them, I don't know where they are. Also I haven't seen them arrive here yet. People tell me some are in Kenya, others in Malawi, but we haven't yet seen each other, and a letter I have yet to receive. 38 According to my informants, anyone can 'buy' a Tanzanian passport from the Immigration Authorities at the Ministry of Home Affairs for a price ranging from $30 to $300 depending on the delivery period.

PAGE 232

216 Often relatives lost each other in the fracas and confusion of the unprepared flight from home. Several Congolese informants who left Uvira in the 1998 refugee flows, such as Melanie, had since lost all contact with parents, children, brothers or sisters. Rwandese frequently lost touch with relatives because of the genocide and the subsequent mass exodus of millions of Rwandese Hutu to Congo and Tanzania (see chapter 2). When I asked Daniel, whose parents and two of his four siblings succumbed while fleeing the Rwandese army into Congo, whether he still had other relatives alive, he was not sure: I think so, but I haven't received any news. When I left [Rwanda in 1994], they were all there in Kibumba, when we fled, they were there all of them. But then later on . we parted ways. Maybe they're still there [alive], but I don't believe they are in Rwanda. If they are still alive, they must be in Congo, in Chad, yes, they [Rwandese refugees] went all the way to Chad. Even the Central African Republic. Others went to Congo Brazzaville. 39 The International Committee for the Red Cross [ICRC] has a tracing program aimed at reuniting lost relatives, yet they need to have at least some indication as to the possible whereabouts of those lost. Unfortunately, this information is often not available, while, at the same time, informants said they were afraid to approach the ICRC office in Dar es Salaam and disclose their illegal status. Instead, people rely on, more often verbal than written, messages and news carried by more recently arrived persons from home or commuters. Many a truck driver between the port of Dar es Salaam and Eastern Congo, Rwanda or Burundi is the carrier of messages, letters, even sums of money back and forth between relatives and friends. 39 See map 1-4 in appendix a for an overview of the escape routes of the Rwandese and Burundese refugees in 1996.

PAGE 233

217 The above descriptions and selected testimonies aimed to give the reader an impression of how the urban refugee in Dar es Salaam adapts to his/her new life in exile. Whether with regard to local interactions, economic endeavors, educational needs, sexual exploitation, aspirations to migrate or worrying about lost relatives, the problems encountered are manifold and obstacles plentiful. In the face of this multitude of challenges and difficulties, that at times appear insurmountable, it helps to have a listening ear at one's disposal or a shoulder to cry on every now and then so as to be able to release some of the tensions and frustrations. In addition, the refugees often depend on material as well as immaterial support in the form of companionship or advice from friends or others that can bridge them over occasional particularly harsh periods. The purpose of the following section therefore is to shed some light on the refugees' social worlds in exile. 4.4 Social Worlds Having defined a refugee as "a person whose social world is disturbed" (1990:190), Marx asserts that after a certain period, the initially "helpless refugees . gradually acquire some power of their own, and . integrate in society" (198). However, real integration in society may not be an achievable goal for Dar es Salaam's undocumented, hence "illegal," refugees. Nonetheless, they will inevitably make the necessary efforts to rebuild their disrupted social worlds by establishing new ties which may to a certain extent replace the ones lost in the wake of the flight from home. The next chapter will illustrate in detail that more than half of the refugees' social network members are persons with whom they have established relationships after arrival in Dar es Salaam.

PAGE 234

218 The following sections in this chapter intend to illustrate and document the refugees' experiences, constraints and strategies in rebuilding their social worlds.40 4.4.1 Meeting People One of the Rwandese informants, who considered himself a keen observer of human behavior, stated confidently to be able to recognize any of the four nationalities by appearance only. Me, when I see a person, I can tell if it is someone from here, if it's a Burundese, if it's a Tanzanian, if it's a Congolese, if it's a Rwandese. Just by looking at them. By the nose, I will know whether it is a Tutsi, or a Hutu. There is a difference in appearance, in the way they walk. By the way they walk, me, I know, this one is Tanzanian, Congolese, that one is a Rwandese brother, that one a Burundese. For example, a Burundese walks as someone who is about to fall over. One would say he is extremely tired. A Congolese, when he walks, you will think the street is his, you'd say he's walking on eggs, he's afraid to crash the eggs. A Tanzanian is suspicious about everything, he is like someone who doesn't feel like living anymore. And the Rwandese, he will be looking 100 meters ahead, it is someone who is afraid that maybe someone will catch him, so he's really careful when walking. The above statement, while amusing at first, illustrates very well the extent to which the urban refugees are preoccupied with the nationality of others even before they have actually met them, because of all the implications such an encounter might possibly entail. 4.4.1.1 Meeting Tanzanians Euphrasie is one of the respondents who count many Tanzanian friends among their network members in Dar es Salaam. Despite the fact that she can no longer venture far from her house, because of her handicap, Euphrasie says she gets along great with the 40 Where relevant reference will be made to the composition of the protagonists' set of ten network members that each respondent was requested to list during the survey (see chapter 1)

PAGE 235

219 "locals." When I asked her if, apart from the Tanzanian neighbor mentioned in her survey form, there had been other Tanzanians who had helped her, she replied: Many, many, there wasn't one who didn't help me . If you have a bad mouth people will hate you, but if you have a good mouth you will be liked. If I had had a bad mouth, I would have died by now. I haven't a single relative here, and yet I am alive. They took two women because they were refugees, and imprisoned them, but I'm still here. Nobody, even a single person has called me mkimbizi, young or old. I talk well to people, that's all. They tell me: "Bibi,41 the day that you leave here to go back home, we will be sad." I say: "Why?" They say: "Since you are here, we have gotten so used to you." Any passerby asks me how I am. He chats with me while he knows I am here without any papers. And yet, never have I been arrested because I know how to talk to people . If you live well with people, they won't pursue you, ever. I am an old woman (mzee), who will pursue me? Whether man or woman, who would want to investigate my life? Charles, as well, expressed his appreciation for the Tanzanian willingness to help, even from strangers in the street. [Arriving in Dar es Salaam] I didn't know anyone here. After one week, I bumped into this Tanzanian man from Kigoma. He spoke Kirundi, he approached me and so we talked . I think he must have seen I was really lost. He called me, and asked: "Where are you from?" I told him: "I am from Rwanda." And because he knew our mother tongue he started talking to me. He asked me what had happened, I told him the whole story. Then he said: "OK, we will help you." Later on, I went to see him to inquire where the UNHCR office was. He said: "Very good. Wait a moment, I will look for someone to bring you there." Bonaventure has experienced that, at times, Tanzanians were more helpful to him than his own compatriots. There was this one woman at the Ministry of Education, who took French language courses with me for about two months. She appreciated me, and told me: "If you have your degree here, why not submit it to the Ministry of Education so that you can get a seal allowing you to teach." I told her, I had my degree with me, but no residence permit. She said: "Just show me your degree, I will look for ways to find you an employment." Because I had the original degree with me, she told me to go to the Congolese Embassy42 to have one photocopy validated, as she couldn't take 41 Bibi means grandmother, and is used as an affective term of respect to persons who age-wise could be a grandmother. 42 Unlike in most refugee situations, the Congolese refugees from the Kivu region are not persecuted by their own government, because they are fleeing the Rwandese/Ugandan occupying forces in Eastern Congo. As such, they may approach the Congolese Embassy in Dar es Salaam.

PAGE 236

220 the original with her. I went to the Embassy and explained everything to them . It's a small thing, all they need to do is telephone Kinshasa and confirm the validity of my degree. But they refused, then started to ask me money, US$10. I told them: "I am a refugee, I don't have that kind of money." They told me to write a letter to the Embassy requesting assistance. Up until today, I never received a reply. This really shocked me . Until now, nothing moved, if you go to the Embassy, you'll find my file sitting there. It appears that despite certain frustrations with Tanzanians at the general level (see earlier in this chapter), most respondents appreciate their hospitality and generosity at the individual level. Aristide reflects that: I believe even my compatriots, the Burundese, they are not ready to understand other people the way these people do. Suppose there were as many Tanzanians in Burundi, I don't think the Burundese could understand them the way we are understood here, I doubt it. The above testimonies express a high level of appreciation and gratitude on the part of the urban refugees vis--vis individual Tanzanian men and women. While there exists a certain level of frustration and discontent with Tanzanians who "ask too many questions" and who "minimize us as refugees," it is at the moment of establishing ties at the personal level with individual Tanzanians that an understanding is born and often mutual respect found. While more than half of the refugees' newly found social network members are of Tanzanian nationality, a substantial minority is either Congolese, Burundese or Rwandese, in other words, more likely than not, fellow-refugees (see also chapter 5). 4.4.1.2 Meeting fellow-refugees Some respondents, such as Cathrine and Jeanne, would prefer never again having anything to do with their compatriots. Both of Tutsi ancestry, it is easy to understand why they feel like this. Jeanne says she is always afraid because: Those that did the killings, many of them don't want to return. Some, you may bump into them [here] even as they try to hide. It is a problem when he knows that

PAGE 237

221 you know him, he will kill you . I don't understand how a person can come, kill you, steal the things and leave you. A person who was your neighbor, you greeted each other, then he comes and kills you. Those [perpetrators] who are here, try their very best to hide. So even, if you tell a Rwandese: "I know a compatriot of yours," he will not allow him to come [to his home], or even to meet him. Cathrine, from her side, made a clean break with her past. In her list of ten network members, she mentioned not a single Rwandese. She says she prefers Tanzanians and Congolese friends to Rwandese because I see that Tanzanian women are better than Rwandese women . Because if you explain your problem to a Tanzanian, she will show pity and sympathy. Rwandese have a very bad spirit (roho mbaya sana). Whereas Jeanne and Cathrine actively avoid their compatriots, Christine would wish to connect more with her Burundese countrymen. But, she says, she is afraid to move about too much and talk to people lest she be recognized as a refugee and reported to the police. Her fear is based on stories that even documented refugees have problems with the immigration authorities. For example, there is Mr. Faraja who used to live in Burundi and then came here. He is now a Tanzanian citizen and has every document to proof that he is a national. Many Burundese arrive at his house to visit him, and every day, immigration comes to pursue him. The authorities told him that if he continues to receive another Burundese, they will take him by force. There is another Congolese, who has a UNHCR protection letter, but still they continue to harass him, and watch him to see who his visitors are. Thus even if Christine was to overcome her fears and connect with compatriots, she would still be afraid to be found out through associating with them publicly. Unsurprisingly then, she did not list a single Burundese friend met after arriving in Dar es Salaam. On the contrary, her two only friends in Dar es Salaam are a 50 year old Tanzanian man, who often acts as mediator in her frequent marital disputes, and a 37 year old Tanzanian female neighbor.

PAGE 238

222 Charles' social world centers around his place of employment as each of the ten network members he listed are connected in one way or another with the Congolese school. Eight of them originate from one of the Great Lakes countries, but only one is a Rwandese compatriot. It is also through fellow-refugees rather than through compatriots that Charles learned of the Congolese school and the possibility of employment. I had gone again to Tamofa43 to see if there had been any changes, and there I met a Burundese. We talked, I explained my problems to him . I asked him for advice. He suggested I go to the Alliance Franaise,44 maybe to teach French . So I looked for their office . but when I had found it, they told me that there wasn't a vacancy. But there, at the Alliance, I met a Congolese, we talked. It was the first time since my arrival [in Tanzania] to converse in French . and I really enjoyed it . it had been approximately three months . He invited me at his place, and told me [about] the Congolese school, and that the Director was his brother,45 so it was absolutely no problem to introduce me to him. Refugees of different nationalities and different dispositions have different attitudes towards meeting with or seeking out compatriots. One Congolese informant related how, upon arrival at the railway station in Dar es Salaam, he asked a taxi driver to bring him to "where all the Congolese live." He was driven straight to Kinondoni (see chapter 1). Neither Burundese nor Rwandese informants ever expressed the same open eagerness as most Congolese to meet just any of their compatriots, whether out of fear or suspicion.46 Even as the refugees establish new relationships with Tanzanians and fellow-refugees 43 The predecessor of Umati (see chapter 2). 44 The Alliance Franaise used to be a rather important meeting site for a significant number of Congolese, Burundese and Rwandese refugee intellectuals who could afford the $10 annual subscription fee. Following many heated debates and political exchanges, however, it is currently formally forbidden to talk Great Lakes politics on its premises. 45 Here used in the meaning of compatriot, not biological brother. 46 This issue will be further elaborated on in chapter 6.

PAGE 239

223 alike, it is not with all social network members that they will be befriended in the same manner. In addition, occasions and opportunities to "just hang out with friends" are few. 4.4.1.3 Socializing with friends Both Dieudonn and Melanie list Congolese compatriots as their closest friends in Dar es Salaam. Dieudonn's closest friends are of the same age as he is. With the first, whom he knew from before arriving in Dar es Salaam, he has daily contacts. With the second, whom he met through Umati when receiving UNHCR assistance, he meets twice a month. Contrary to the Rwandese and Burundese, the Congolese generally do not perceive the same need to hide and they even have a football team. Dieudonn is one of its players, and regards the football playing more as way of socializing with other Congolese than anything else. In the weekends I don't work, I watch television and play football. We play every Saturday, I am one of the players. The Congolese football team exists since 85. We play against Tanzanian teams, also white teams. It's just playing for fun, we don't do any training during the week. It's only football, there are no other Congolese sports teams. Melanie met her two girl friends through the neighborhood after arriving in Dar es Salaam and has daily contacts with both of them. Unlike most young women of her age in other parts of the world, she does not have many opportunities to socialize with girls and boys of her own age. She explains: On Sunday, I go to the church. Saturdays, they (her employer's family) go to visit relatives, I stay at home and prepare juice or go work in the field. [Tanzanian friends,] I am not allowed to socialize with because they are very rude and wrong-headed. But I have two Congolese girl friends. Mama Jolie and mama Yannick (her employers) refused at first that I see them as well, but I explained that they are very polite. And so now I have a friendship with them, because they are not rude or strong headed like the others . Boys try to see me, but I don't want to see them. They want to make love, but I refuse. We just greet each other. That's all.

PAGE 240

224 Pierre, as one among many other informants, says he lacks the money to go out in the evenings or the weekend to socialize with friends. He regrets not being able to enjoy Dar es Salaam's lively music scene. What I have noticed here, is that Tanzanians and Congolese have the same . character . when it comes down to having a good time (le sense de l'ambiance). They are so alike, because they [Tanzanians] like music, and so do the Congolese. A Congolese cannot stay in a place where there is no music.47 In Burundi, it is very different, you cannot find people like here, its too calm . The Hutu . when they arrived over there in Congo, they really learned all about having a good time (ambiance), the music, the drinks, enjoying a beer. But, for the Tutsi,48 you will find a place where they're sitting around having a beer, but there is no music. Everything will be there, except the music. Receiving friends in his room could have been an alternative option. But, as a result of his continuous endeavors to hide, all of Pierre's neighbors think he is a Tanzanian, and he is afraid to invite any non-Tanzanian to his place.49 Like his compatriot Christine, Pierre did not mention a single Burundese among the network members whom he met after arrival in Dar es Salaam. Rather, he listed as his closest friends in Dar es Salaam, namely those he goes to talk to when he feels depressed, two Tanzanian persons. 47 The music scene in Dar es Salaam is heavily influenced by different styles of Zairean/Congolese music, and the most popular music bands in town are Congolese. Dieudonn explains that: "Tanzanians have this image that all Congolese are musicians, and myself, I have been asked many times whether I am a musician." Very recently, however, the work permits of these Congolese musicians were revoked, reportedly in a Tanzanian attempt to do away with the competition because Tanzanian music bands were never able to draw the same large crowds at concerts as the Congolese bands. 48 From this remark, it would appear that while the minority, Tutsi manage to dominate not only the political and economic scene, but also the cultural expressions of the social scene. 49 Contrary to average Western style accommodation, the rooms that most of the refugees rent offer almost no possibility of privacy. A rented room will typically be situated in a house consisting of a corridor with five to six rooms located on each side. While the shared bathroom facilities are situated in the court yard behind the house, there is no separate kitchen. Because cooking takes place in the corridor twice a day, there are always persons around observing the comings and goings of the other tenants and their respective visitors.

PAGE 241

225 Not having the same nationality in itself, does not impede the establishment of real friendship. However, the strategy of pretending to be Tanzanian does affect the depth of the relationship. One of Pierre's compatriots explains: I pretend to be a Tanzanian out of security reasons. But I am not a Tanzanian. And here, in Dar es Salaam, I don't have close friends who really know me . If you become too close . they will really want to know you . And they'll start digging, into your inner self (undani) . So, yes, superficially, we understand each other well (tunaelewana ya juujuu) [with the Tanzanians]. From the testimonies it appears that reconstructing social networks through meeting people and establishing new relationships is a process fraught with complications for the urban refugees in Dar es Salaam. At the same time, it becomes clear that different people have different motives to relate to specific groups of individuals. Congolese appear to seek out compatriots with the same fervor as many Rwandese are avoiding theirs, while Burundese are particularly weary of hiding their true nationality and pass as Tanzanians. Leaving the respective motivations for these modi operandi to be elaborated on and contextualized in the following chapters, I first wish to complete the overview of the process of meeting people by briefly exploring the meeting places. 4.4.2 Meeting Places The survey results (see chapter 5) show that only three ways of meeting persons account for two thirds of the new ties the urban refugees establish in Dar es Salaam: the neighborhood, the church or mosque and mutual friends. At times, informants would mention the existence of associations based on nationality, and as these groups would often consists of groups of friends, I discuss them as an additional meeting site.

PAGE 242

226 4.4.2.1 The neighborhood While close to one in three of the refugees' new network members were met through the neighborhood, these neighbors are not necessarily aware of the true nationality of the refugees. Burundese informants gave the impression to conceal their true nationality to a greater extent, yet both Congolese or Rwandese as well exercise prudence in contacts with their neighbors. Bonaventure and Cathrine, for example, have confided only in some of their neighbors. Bonaventure says that: Most neighbors don't know that I am Congolese. Well, some do, that is true. But so far I have not had the problem that someone comes to my door to confront me by saying: "You, aren't you Congolese ?" Cathrine has her reasons to not confide in all of her neighbors. Many know that I am Rwandese, some think that I am Tanzanian. I don't tell them [differently] because I am afraid of Immigration. Both Bonaventure and Cathrine are single and live by themselves in a room. As long as they do not receive compatriot visitors and speak their mother tongue, there should not be a problem. However, things may become more complicated for families. One of the Burundese informants, who grew up in Tanzania in the 1970s and 1980s, explained that he had been able to convince his neighbors that he, his wife and two children are Tanzanian through speaking Swahili all of the time. As a result, his children do not know any Kirundi. I asked him if that would not pose a problem at the time of returning home. He concurred: [They'll have] big problems. Even me, I can already see it. A child that doesn't know the language of his home is lost . [Even inside the house] we cannot speak Kirundi because of our security. If we were to speak it, we'd be caught by the first word. And so, we just cannot speak it.

PAGE 243

227 As mentioned before, Rwandese respondents reported many more problems acquiring a certain level of Swahili, and most do not even attempt to pass as Tanzanians. As one of the Rwandese informants grudgingly explains: [My neighbors] know that I am not a Tanzanian . [It is not a problem] because I don't have a close relationship with them. I do my work [as a tailor], go to church on Sunday, come back and stay . I have no choice but to keep good relations, How else will I manage ? Thus, while neighbors are often mentioned by the respondents as supportive and helpfull to a certain extent (see chapter 5), the testimonies show that this does not mean that relationships are necessarily intimate or indicative of real friendships. 4.4.2.2 The church/mosque A majority of the respondents reported going to church/mosque weekly. Among the Congolese and the oldest refugees, however, a substantial minority frequents their church/mosque on a daily basis. Despite such a high frequency of church/mosque visits, only few respondents reported the church/mosque as a place to meet friends. Amlie, for example, is a devout Pentecostalist. She goes to church every Sunday, Tuesday and Friday. Half of the network members listed as met after arriving in Dar es Salaam, she came in contact with through the church. She received not only emotional support from them, but money, food and clothes for herself and her children as well When I asked her how she had found this church community, she explained that: First of all, I am a Christian, so I have to look for my Church. I live in Mburahati, but I used to go pray in the Church in Ilala. There I met a pastor, a Mchagga,50 whose name is Itiari, who helped me with the bus fare. Good, so I stayed with these believers. Some asked me how I lived/managed to get by. So I explained to them and they took pity on me and showed mercy. They gave me sugar, flour, and money as well . After I met Siriro,51 he started to give me help, and I became 50 A member of one of the largest ethnic groups in Tanzania, the Chagga from the Kilimanjaro region. 51 A Burundese pastor, refugee in Dar es Salaam since 1993.

PAGE 244

228 interested in joining him in his church . My [old] church was Pentecostal and this Full Gospel (her new Church) is Pentecostal as well. It's because he gives me the bus fare . So today I pray together with him. Amlie met her compatriot Siriro through a mutual friend, and as he paid for her children' school fees, she is eager to show him her appreciation by becoming a member of his church.52 Amlie was fortunate in finding this much support from church members, and she is one of the small minority of respondents who consider it a place to meet friends. 4.4.2.3 Associations A majority of respondents stated not knowing or ever having heard of any Congolese, Burundese or Rwandese associations, and only a small number reported ever having been a member of any such association. Pronounced differences appear across nationalities as only few Congolese respondents have never heard of any association bringing together persons from home, compared to most Burundese and almost all of the Rwandese respondents. That fact that only half of the Congolese who are aware of the existence of Congolese associations, are or were at one point an actual member is indicative of a general lack of satisfaction with the associations' workings. The Congolese Community of Tanzania (CCT) is an officially registered association and was founded in the mid-1980s under the acronym TAZAFA (Tanzanian-Zairian Friendship Association). Bonaventure, at one time, considered becoming member. I know of this Congolese association, by the name of CCT, but then I heard of the conflict between them and the Congolese Embassy. Twice I went to their office 52 Another decisive factor in Amlie's choice of churches seems to be that both pastors provided her with the bus fare. The standard rate of bus fare in Dar es Salaam is 150 Tanzanian Shilling (15 US dollar cents). Yet because Dar es Salaam is such a spread out metropolis (see map 1-7 in appendix a), often passengers need two or three buses to get to the point of destination. As such, a trip across town and back for one person can easily cost close to 1,000 Tanzanian Shilling (1 US$) or the equivalent of the evening meal for a whole family (see earlier in this chapter).

PAGE 245

229 [CCT] to obtain a copy of the statutes, but I never found anybody there. Yes, I considered becoming member, because that association is a national one . that's why I wanted to join. Other Congolese informants as well complained of the inefficient workings of the CCT, and of political bickering. The association's funds consist of the monthly membership fees, and are supposedly allocated on the basis of individual needs, e.g., high medical bills, or the organizing of a funeral. However, as a result of persistent allegations of favoritism on the basis of ethnic affiliation in the allocation of these funds, many Congolese renounced their membership in the CCT. Recently, there has been a proliferation of smaller Congolese associations which are formed strictly on the basis of ethnic affiliations. These have fewer members (often around thirty) and reportedly function much more efficiently than the larger CCT. Usually these associations are not officially registered with the Tanzanian authorities, but in addition, they face obstruction from the Congolese Embassy. Bonaventure explains that: The problem is that the Embassy here doesn't want people from the Kivu region to have an association. They say that the Kivusiens [inhabitants of the Kivu region in Eastern Congo] are plotting to become independent . The Kivusiens, on the other hand, know that the [Congolese] government has the power to beat the Rwandese and [could] chase them away, but [that it] is not doing so because of negligence. And this is exactly, what Kivusiens in Kinshasa have been telling the government publicly. That's why, the Ambassador here, who is from the same region as the president, looks for ways to discourage the Kivusiens from organizing themselves. I remember that our organization, Cyabata, was doing well and growing in members, and one day the Embassy itself was telling the [Tanzanian] Immigration people that "These people shouldn't be allowed to organize meetings, . because what they do there is purely politics." In view of all these constraints, whether from the Tanzanian authorities or from their own compatriots, it is not surprising that ultimately only a small minority of the Congolese respondents remained members of an association despite multiple local initiatives.

PAGE 246

230 Burundese refugees in Dar es Salaam, on the other hand, do not found associations on the basis of ethnic affiliations, because they are overwhelmingly Hutu.53 Most Burundese associations I learned of, are small and have economic objectives. I was told of women's associations, who, for example, raise money to buy knitting materials and subsequently pool their efforts in trying to find customers for their products. Other Burundese associations concentrate on young people.54 One of these, so I was told by its chairman, organizes painting lessons for its members on Sundays. Their creations are geared towards sale on the tourist market, and they pool their efforts in looking for and making contact with Tanzanian shops with a large tourist clientele. One day, as I was interviewing a group of Burundese women, it appeared that almost each and every one of them had their own women's association. In addition, most of them had just returned from a Conference in Arusha organized by UNIFEM and UNHCR, aiming to involve women in the Burundi peace process. Some of the informants had brought their certificates of participation to show me, and were lobbying fervently that I should "use my international contacts" to raise funds for their respective women's associations. Puzzled by the high number of small and recently founded women's associations, I asked one of my female Burundese informants what could be the cause of this proliferation. Yes, all these little groups of women . A lot of meetings were organized [by UNHCR and UNIFEM] . and looking for women's associations to participate. So the women each founded their own association and, so all were invited to meetings in Arusha or Dar es Salaam as representatives (prsidente) of their respective 53 I met only one Burundese Tutsi woman refugee in Dar es Salaam during the whole research period. She was the widow of a Hutu politician and after his assassination in 1993 she feared for her children's lives. 54 Association des Jeunes. The word les jeunes in this African context, is used to indicate young unmarried adults. Only when married, a jeune becomes a man or a woman regardless of age.

PAGE 247

231 associations . Some time ago, they were all members of the same association Dawa . one that has been there for a longtime . but then because only the representative could participate in these big meetings, they decided to have their own little associations. . Because, here, in meetings, they give you money to participate.55 The older association, Dawa, that this informant considers a genuine women's association, counts up to fifty members, and its chairwoman is a longtime Burundese resident of Dar es Salaam. According to the same informant, Burundese men do not form any associations because large gatherings of Burundese men would arouse too much suspicion among Tanzanians.56 In all, only half as many Burundese respondents as Congolese are currently member of an association. The Rwandese appear to have the lowest number of national associations, however from one informant's account they appear the most effective towards assisting their members. Daniel mentioned being member of one. Yes, we have an association (une mutuelle) that meets every Sunday . It's mixed, both Hutu and Tutsi can participate, there is no discrimination, you just need to be Rwandese. We are in all 87 members, not all of them come to each meeting, men, women and children. I believe it has been existing for a longtime, from before 96 . In principle, each one contributes 500 Tanzanians shilling (50US cents) per meeting, but sometime, it happens that you don't have the money. And that's OK, we understand the situation, because we're in it ourselves. We don't demand a contribution . We don't talk politics, because you know, we're in a foreign country, and one that is very sensitive with regard to politics. You know, in such an association, I'm sure there are one or two Tanzanians, certainly. But we don't know that who are the Tanzanians, so we don't talk politics. Also, we talk Swahili because some members don't speak Kinyarwanda . others don't speak French, but we all speak Swahili . The chairman came a longtime ago, before 94, not as a refugee but as a businessman. . I don't know if the association is registered, it better be because otherwise sooner or later we'll have troubles . For the 55 Sic. In addition to a per diem, paid when meetings take place for example in Arusha or Nairobi, it is customary in Tanzania to pay a sitting allowance (of up to US$15 per day) to persons who participate in meetings or even workshops organized by international organizations. 56 Through other refugee informants, I learned that there are regular secret political meetings among Burundese refugee men in Dar es Salaam. However, not a single participant would openly acknowledge these meetings to me.

PAGE 248

232 Rwandese, this is the only association I know, I know of another Congolese one, and the Burundese have several. But for the Rwandese, I believe it is the only one. From the above accounts and the survey findings, national associations appear to play a role in the lives of only a limited number of Dar es Salaam's refugees. In addition, the motivations to form an association are of an economic nature, rather than for social reasons. The smaller Congolese associations based on ethnic affiliation function to provide financial support in instances such as funerals or hospital expenses for the chronically ill. The gender and age based Burundese associations on the other hand, aim to generate income for their members through a number of collective efforts, while the Rwandese association uses its members' contributions to meet particular needs, such as, for example, when they gave Daniel the money to pay for the yearly rent on his room. 4.5 Concluding Remarks International agencies concerned with refugees often address the process of dislocation and displacement that refugees go through in terms of trauma (World Health Organization 1996) or problems of mental health (Miserez 1988). These organizations usually focus on the extreme upheavals and violent experiences refugees may have been victim of (e.g., torture, rape, etc.) and formulate necessary advice to refugee workers on how to recognize and deal with psychological stresses and mental illnesses. However, as one study points out, "while their experiences may place them at risk, most refugees do not become mental health casualties" (Beiser, Turner and Ganesan 1989: 193). It is in the same line that another researcher regrets the exclusively negative attention to refugee health in the literature (Muecke 1992). She suggests the concept of resilience to be the central focus in researching refugees' lives as it refers to "social competence or other types of functional adequacy despite losses and stressors" (520).

PAGE 249

233 The recognition of the fact that the majority of the urban refugees in Dar es Salaam manage to regain a sense of autonomy constituted the silver lining of this chapter. The situation of being in a state of liminality (Turner 1969), or being "betwixt and between" their old lives and the new ones is compounded by the illegal status of the overwhelming majority of the urban refugees in Dar es Salaam. Without the necessary documents--a lack necessitating a kind of hide-and-seek with Tanzanian Immigration officials and police officers, albeit neighbors and passersby in the street--there is not the slightest glimmer of hope for local integration. Yet despite these adverse living conditions the refugees continue to design strategies and find ways to adapt to the new life. Their determination to stay on in Dar es Salaam instead of going to the refugee camps or returning home is indicative of the fact that the majority ultimately succeeds in making ends meet. Whether it concerns learning a trade, setting up a small business, engaging in trading activities (often through former contacts) or performing some type of paid employment, the legal implications of being without a necessary residence and/or work permit are dealt with in a creative array of ways, more often than not thanks to multiple contacts and relationships with the Tanzanian population. And yet, while Tanzanians are generally appreciated for their hospitality and support at the individual level, quite a few refugees continue to hide their real nationality prohibiting the development of true friendships. At the same time, the accounts testify to the fact that not all respondent were eager to connect with their compatriots for a variety of (both political and ethnic) reasons whether through national associations or one of the other meeting sites. Marx defined social worlds as "the sum of all the migrants' [and refugees'] relationships and of the forces impinging on them at any moment" (1990:189). Following Gluckman, he argues that "complex social forces" at the macro-level are at

PAGE 250

234 work in "social situations" at the micro-level (192). In addition, "a social world is not confined to a particular place or limited by territorial boundaries. Some of the relationships may be very important but physically distant, while others may be almost insignificant although located close by" (194). This means that the social worlds of the refugees, which are absolutely indispensable in the process of coping with the situation of forced migration and redefining oneself through one's relationships (Marris 1986), extend beyond the relationships established with persons met after arrival in Dar es Salaam. Following Marx, I propose that the most appropriate method to fully explore the urban refugees' social worlds, including "the forces impinging on them," is by analyzing and de-constructing their social networks and contextualizing the findings in the larger macro-worlds. These are the goals of chapter 5 and 6 respectively.

PAGE 251

Table 4-1. Comparing Life at Home and in Exile (in %). Questions/Answers Women n=150 Men n=150 Cong. n=100 Buru. N=100 Rwan. n=100 <25y n=88 26-35y n=95 36-45y n=86 >45y n=31 Average N=300 Comparing life at home and in Dar es Salaam* 1. Life at home was good, people loved one another, social harmony 62 54 69 46 49 60 57 57 58 58 2. Here, no job opportunities, no money, nobody pays studies 36 48 44 38 44 49 38 44 32 42 3. Her, life is difficult and/or expensive 49 34 39 41 34 43 40 40 45 41 4. Here, we need permits, as strangers we are not free 29 43 46 40 22 20 46 43 29 36 5. Here, there is peace, I'm not afraid of being killed 27 26 16 15 48 30 32 28 29 26 6. Other 5 12 2 15 8 7 6 14 3 9 What is your major worry here in Dar es Salaam?* 1. Returning home, peace at home 47 40 39 48 52 33 48 53 41 43 2. Continuing my studies/my children's studies 27 25 30 11 40 48 32 13 9 26 3. Money, a job, food, accommodation 17 18 23 18 14 21 19 16 14 18 4. Leaving Tanzania 13 19 19 11 16 18 28 12 6 17 5 Lost relatives 15 14 16 2 28 10 16 18 30 15 235 Multiple answers possible

PAGE 252

236 Table 4-2. Interactions with and Perceptions of Tanzanian Nationals (in %). Questions/Answers Women n=150 Men n=150 Cong. n=100 Buru. N=100 Rwan. n=100 <25y n=88 26-35y n=95 36-45y n=86 >45y n=31 Average N=300 Establishing relationships at home and in Dar es Salaam 1. It is the sa me 51630221133me3453281134 2. It is easier here than at ho 3. It is more difficult here 92 95 89 94 98 90 97 98 83 93 Why ?* 1. Tz ask too many questions 55 62 34 56 85 57 65 54 55 58 2. Because of the language 31 27 31 17 38 26 37 24 23 29 3. Tz minimize foreigners 13 15 9 13 21 10 13 16 26 15 4. There are cultural differences 12 16 1 1 40 10 12 19 16 13 5. The Tz do not like us 11 9 15 6 8 14 10 9 0 10 6. Other 27 24 47* 19 11 23 25 27 25 27 Are you sometimes called mkimbizi (refugee)? How often ? 1. Daily to regularly 71 63 72 63 64 68 67 64 65 66 2. On a regular basis 27 27 43 39 21 29 26 29 24 27 3. Never, I pass for a Tanzanian 10 17 13 17 8 10 12 15 21 13 4. Generally, I pass for a Congolese 490015 1142767111485 1351149 Have you ever been arrested ? Multiple answers possible

PAGE 253

CHAPTER 5 DE-CONSTRUCTING SOCIAL NETWORKS "Iminozi niyo idahura." Only mountains don't meet. (Kirundi Proverb) Burundese female refugee, Dar es Salaam, August 2001 The testimonies of the previous chapter have indicated the importance of social contacts and relations in the adaptation process of the Great Lakes refugees to their new life in Dar es Salaam. Even as the urban refugees live scattered all over the city of Dar es Salaam in an effort to hide their clandestine and undocumented status, as social beings they are continuously meeting other individuals and establishing new relationships. Whether involving compatriots, fellow-refugees or nationals of the country of asylum, many of these newly established relationships become part of the refugees' personal social networks.1 At the same time, part of the relationships that had been established before (or during) the flight from home continue to play a significant role in the refugees' social worlds in Dar es Salaam. In view of the fact that the refugees are of different genders, nationalities and age groups, I propose that the cultural norms and societal roles ascribed to these respective social groups, will play a role in their members' coping strategies, i.e., here, the composition and reconstruction of their (supportive) social networks. In order to gauge these socio-culturally inspired responses to forced migration in an urban setting, it proved 1 The difference between personal or ego-centric social networks and socio-centric networks is explained in chapter 1. 237

PAGE 254

238 necessary to apply an in depth quantitative analysis of the social network data generated through the survey of the three hundred respondents, and disaggregate them across gender, nationality and age. The first section of this chapter, however, is an overview of the "average" personal social network of an urban refugee from the Great Lakes region, who arrived in Dar es Salaam more than one but no more than ten years before the time of the research project.2 The composition of the "average" social network will be presented in terms of demographic and relationship attributes, disaggregated by the moment of establishing the relationship.3 The next section deals with homophilous relationships in order to investigate to what extent belonging to the same gender, nationality or age group plays a role in the composition of the respondents' social networks. Complementary to this analysis, the third section disaggregates the social network data by gender, nationality and age category of ego, with the aim of discovering similarities and differences in the network building strategies between the social groups. Lastly, the disaggregation of the same data for the gender, nationality and age group of alter, provides additional information with regard to certain socio-cultural patterns and societal roles of the respective social groups. The findings will ultimately be contextualized in the larger socio-cultural, economic, historical and political framework in chapter 6. 2 As explained in chapter 1, these were among the selection criteria of the respondents to be surveyed. 3 The term 'ego' is used to indicate the respondents and 'alter' to refer to ego's network members. Network members whom ego met before arriving in Dar es Salaam will for the sake of brevity often be referred to as 'pre-Dar es Salaam alters', network members met after arriving in Dar es Salaam as 'post-Dar es Salaam alters'.

PAGE 255

239 5.1 The Average Personal Social Network As mentioned in chapter 1, the name generator or question eliciting the social network data of the three hundred respondents was formulated as follows: "Could you give me the names of ten persons who have helped you from the time you left your home as well as during your stay in Dar es Salaam? They may include persons from any nationality, men and women, no matter where they live, and who have helped you in any way, be it financially, emotionally, information-wise, socially, neighborly, or otherwise."4 The formulation of the name generator does not mean to imply that the respondents only received support, and that they did not provide any support to others. As we shall see, a substantial amount of the assistance provided to ego comes from fellow-refugees, implying that ego him/herself equally supports his/her alters whenever and however possible. Translated into social network analysis terms, we say that the name generator contained a numerical limit ("names of ten persons"), a time frame ("since the time you left your home"), and was content-based ("persons who provided support both material as well as immaterial"). No spatial boundaries nor affective specifications were included in the name generator. The survey resulted in a database of the attributes and other particulars of 2,921 persons (Table 5.1). Each of the three hundred respondents was asked to provide information on ten members of his/her personal social network. However, some respondents included the names of humanitarian aid agencies (such as UNHCR, Caritas, 4 In Swahili, the expression hali na mali was used to indicate that the word "help" was to be interpreted in its material as well as immaterial sense.

PAGE 256

240 Care, etc.)5 on their list, while few others stated that they were unable to come up with the names of ten individuals. On average, each respondent listed 9.77 alters. There was no difference in response rate between men and women, and only a negligeable difference between nationalities: Congolese 9.67, Burundese 9.76 and Rwandese 9.89. The oldest respondents had the lowest response rate, 9.52 compared to 9.88 among the youngest group of egos. Table 5.1 provides an overview of the 2,921 alters constituting the social networks of the urban refugees under study, and of whom ego met 1,693 after arriving in Dar es Salaam. They are categorized by gender, nationality and age category. Whereas gender and nationality of each alter is provided, ego was sometimes not able to provide or unsure about alter's age. This is the case for 367 alters, whose age is listed as unknown. In order to allow for effective and efficient comparison of frequencies across categories of both egos and alters, I decided to use only percentages in all of the subsequent tables in this chapter. The respective sample sizes are included in the first line of each table. The following tables are discussed one by one in the body of this section: table 5-2 presents the average network members categorized by alter's gender, nationality and age. Table 5-3 gives an overview of the marital status, domestic arrangements, number of children, level of education and professional occupation of the 2,921 alters. Table 5-4 focuses on relationships attributes, such as the moment the relationship was established, alter's place of residence, the duration of the relationship and frequency of contact, while table 5-5 lists the ways in which ego met alter and the types of support provided by the latter to the former. In addition to the average frequencies (n=2,921), each table differentiates 5 Often these respondents had stayed at one of the refugee camps in Western Tanzania for a shorter or longer period of time.

PAGE 257

241 between alters met before (n=1,228) and after arriving (n=1,693) in Dar es Salaam. The 1,228 persons ego met before arrival in Dar es Salaam include--in addition to friends and relatives from home--any individual that ego reportedly received support from between the moment of leaving home and the moment of reaching Dar es Salaam. While on average, ego counts among his/her ten network members four persons (42%) met before and six persons (58%) met after having arrived in Dar es Salaam, there is substantial variation between respondents. As many as one in twelve respondents (8%) mentioned not one single supportive alter met before arrival in Dar es Salaam, but only one in thirty (3%) stated having received no assistance whatsoever from any person met after arriving in Dar es Salaam. The majority of survey respondents (50%), however, provided the names of between 4 to 7 persons known from before arriving in Dar es Salaam, complemented by between 3 to 6 persons met afterwards. 5.1.1 Gender, Nationality and Age A first observation is that the urban refugees in Dar es Salaam on average count more men than women among their social network members, 58 versus 42% respectively.6 Bearing in mind that a sample of three hundred respondents generates a margin of error of 6%, we may state with 95% confidence that between 52 and 64% (58+/-6) of all alters providing or having provided some type of support to the refugees in Dar es Salaam since the time they left their respective home countries are men. When disaggregating the data between alters that ego knew before and after arriving in Dar es Salaam, we see that the proportion of women providing some type of support is 6 All data in this section are compiled in Table 5-2.

PAGE 258

242 significantly higher among alters s/he met after arrival in Dar es Salaam (48%) than among those known from beforehand (33%). With regard to nationality, it appears that roughly one person out of every three alters is of Tanzanian nationality (35%), the second of the three alters is Congolese (27%), while the third originates from either Burundi (20%) or Rwanda (16%). These proportions change considerably when disaggregated by the moment/place of establishing the relationship. Whereas only 8% of the alters ego knows before arriving in Dar es Salaam are Tanzanians, this percentage increases to 54% of all alters met after arrival. Not surprisingly, the proportion of alters originating from the Great Lakes region evolves in the opposite direction: 89% (33, 30 and 26% respectively for Congolese, Burundes and Rwandese) before versus 45% (23, 13 and 9% respectively) after arrival in Dar es Salaam. The 2% or 60 alters who are not of Tanzanian, Congolese, Burundese or Rwandese origin are of the following nationalities: Belgians (12), Kenyans (11), French (9), Ugandans (4), Zambians (3), Camerounians (3), US Americans (3) and 15 more nationals from Brazil, Canada, Ethiopia, Ghana, Iran, Lebanon, Liberia, Mali, the Netherlands and the UK. Close to one in two alters in the average network are of the same nationality as ego (49%). Yet again, we find that this proportion changes when disaggregating the average network across alters met before and after arriving in Dar es Salaam to 77 and 31% respectively. On average, three out of every five persons providing some type of support to the refugees are between 26 and 45 years of age (34 and 26% respectively). Comparing before and after, we see that--while the proportion of alters younger than 35 years of age remains more or less the same--the proportion of persons older than 35 years of age

PAGE 259

243 decreases from 50 to 36%. At the same time, however, the percentage of alters whose age ego does not know increases from 5 to 18%, thus rendering difficult an in-depth comparison of the proportion of the different age groups among alters met before and after arriving in Dar es Salaam. 5.1.2 Other Demographic and Socio-Economic Attributes While on average, roughly one third of the alters are single (35%), not living with a partner (36%) and childless (34%), the majority are married (59%), living with a partner (58%) and have one or more children (63%).7 Comparing the alters met before and after arrival in Dar es Salaam, there appears to be a slight increase in the proportion among post-Dar es Salaam alters of single persons (from 32 to 38%), of those not living with a partner (from 33 to 39%) and of childless persons (from 31 to 37%). These increases are balanced out by concurrent decreases in the respective proportions of married persons (from 63 to 55%), persons living with partners (from 62 to 54%) and those having one or more children (from 66 to 59%) among alters met in Dar es Salaam. Compared to the "don't know" percentage of alter's age, the respective proportions of alters of whom ego does not know the marital status (1%), domestic arrangment (6%) or child status (3%) remains relatively small and changes very little when disaggregated for the moment of establishing the relationship. This is clearly not the case with regard to alter's educational level. Ego appears unable to mention the level of education of as many as one third of his/her alters (36%) and this percentage increases to almost half (48%) when considering only those alters met after arrival in Dar es Salaam. During the survey, I noticed that it was often persons 7 All data in this section are compiled in Table 5-3.

PAGE 260

244 of a lower level of education who were unable to provide their alters' educational level. However, that in itself does not explain the large percentage of "don't know's." More likely, alter's educational level is not directly relevant to his/her relationship with ego, contrary to, for example, occupation and place of residence (see later). Both of those types of information are observable facts and likely to be more determinative of the establishment of a new relationship or the continuation of an old one than the level of education. The large percentage of missing information on the alters' respective educational levels make it unfortunately very conjectural to draw major conclusions when comparing alters met before and after arrival in Dar es Salaam. Contrary to the lack of knowledge related to the educational level, ego appears to be much more aware of the professional occupation of his/her network members. The low percentage of "don't know"s (6%) confirms the above observation that occupation is a more visible type of information and likely to be of greater relevance to (the establishment of) the relationship. On average, close to half of all alters are engaged in either trade (27%) or are self-employed (21%), as e.g., hairdressers, tailors, food vendors. Whereas 15% of the alters receive payment for their services from a third person, whether in the formal or informal sector (e.g., housegirls, drivers), one in ten (10%) of all alters is a student,8 and nearly one in ten a housewife (9%). Comparing the proportions of the different categories of occupations among alters met before and after arrival in Dar es Salaam, there is a steep increase in the self-employed alters (from 14 to 26 %), offset by a decrease in the number of alters engaged in paid employment (from 18 to 12%) and of students (from 12 to 7%). Given that nearly half of all alters are refugees themselves, 8 Many of these students, however appear to reside outside of the region, in countries such as Canada (15), Belgium (9), France (9), South Africa (8), Kenya (6), and others.

PAGE 261

245 it is not surprising to observe trends similar to those of the respondents' professional occupations before and after the flight (see chapter 3). 5.1.3 Place of Residence, Lenght of Relationship and Frequency of Contact On average, ego met 58% of his/her current alters after arrival in Dar es Salaam, and, looking at the place of residence, we see that a more or less concurrent 66% of all alters are currently residing in Dar es Salaam.9 An additional one in fifteen alters lives in Tanzania, but outside of Dar es Salaam: 5% in Western Tanzania (including the refugee camps) and a mere 2% elsewhere in the country. Disaggregating alters met before and after arrival in Dar es Salaam, shows that more than nine out of every ten alters (93%) met after is a resident of Dar es Salaam. At the same time, every two in five alters met before arrival in Dar es Salaam is also currently residing somewhere in Tanzania: Dar es Salaam (28%), Western Tanzania including the refugee camps (10%) and elsewhere in Tanzania (3%). This high frequency of persons residing in Tanzania among pre-Dar es Salaam alters is explained by the fact that the majority of refugees traveled overland from their respective homelands to Kigoma, and onward from Kigoma to Dar es Salaam. Some spent a certain period of time in one of the refugee camps, others temporarily self-settled in one of the towns in Western Tanzania (or, to a lesser extent in one of the other regions of the country) before heading for Dar es Salaam. During this journey, which may have taken from a few days to a few years, the refugees currently residing in Dar es Salaam established relationships throughout the country with Tanzanian nationals, fellow-refugees or persons of other nationalities who, at one time or another, provided them with some type of support. At the same time, an average of 15% of all alters reside 9 All data in this section are compiled in Table 5-4.

PAGE 262

246 in ego's home country. This proportion increases to one in three of all pre-Dar es Salaam alters (34%), while only 1% of ego's post-Dar es Salaam alters are currently living in his/her home country. The respective average proportions of alters currently living in one of the three countries of the Great Lakes region under study are more or less the same: Congo (7%), Burundi (5%) and Rwanda (5%). The slight preponderance of Congolese residents is the result of the double flight of a number of the Burundese and Rwandese respondents, who initially fled to Congo, in 1993 and 1994 respectively, and resided there for several years until the wars of 1996 and 1998 forced them to relocate to Tanzania. This period of their lives was included in the question ". . since you left your home?," thus generating names of Congolese residents providing them with some type of support during that period. This observation is further confirmed when disaggregating the average frequencies of alters met before and after arrival in Dar es Salaam. The proportion of Congolese residents exceeds that of Burundese and Rwandese residents alike, among pre-Dar es Salaam alters (17% versus 11 and 10% respectively). Almost one in every five alters met before arrival in Dar es Salaam (18%) currently resides in a country other than Tanzania, Congo, Burundi or Rwanda. The bulk of these 260 persons (9% of all alters) lives in South-Africa (45), Canada (27), Belgium (26), Kenya (23), France (21), UK (17), USA (17) and Zambia (14). Another 40 persons live in thirteen other African countries, 20 persons in Europe and the remainder (10) is spread out over Australia (4), Asia and the Middle East (3), Latin-America (2) and New-Zealand (1). The social networks of the three hunderd surveyed respondents are truely international, including each and every single continent worldwide.

PAGE 263

247 Telephone as a relatively expensive means of communication in Tanzania10 and the mail delivery through lorry drivers11 being a geographically limited option, other means of communication are used to stay in contact with alters worldwide. Internet-cafs can now be found on virtually every street corner in Dar es Salaam, charging fees affordable to many of the urban refugees.12 The internet might not be the ideal means of communication with people at home depending on the local availability of internet services. However, it certainly encourages easy, cheap and regular contacts with the many relatives and friends among ego's alters who currently live in a Western country, and whom the refugees often depend on for financial help (see later in this chapter). While on average, ego has known half of his/her alters (49%) for a period longer than five years, a distinct difference appears when comparing alters met before Dar es Salaam (93%) and after (16%). The fact that the respondents' average length of residence in Dar es Salaam is around 4 years (see chapter 3) explains why, on average, ego has known only few alters met in Dar es Salaam for a period longer than five years (16%). Consequently, on average, the length of the relationships established between ego and alters in Dar es Salaam is fairly short: 29% between two and five years, 41% between one and two years and 14% were met less than one year before the time of the survey. 10 Particularly international rates are exorbitantly high. 11 These commute between the port of Dar es Salaam and the landlocked countries/regions of Eastern Congo, Burundi and Rwanda. 12 Even the refugee camps are presently connected to the virtual world. In the Spring of 2001, an American foundation initiated the installation of an Internet connection and accompanying Internet Learning Centers in a number of the refugee camps in Western Tanzania. This was a major challenge in an area where there was "no electricity, the telephone system could barely support e-mail and . [there was] no Internet service provider in the district" (UNHCR 2002b:11; Lewis 2003).

PAGE 264

248 The frequency of contact ego has with his/her alters is strongly influenced by whether the relationship was established before or after arrival in Dar es Salaam as well. While on average, ego has daily to weekly contact with 55% of his/her alters, this percentage is reduced by half (28%) when considering only alters met before Dar es Salaam while it increases by half (77%) for alters met afterwards. With every one in two (52%) pre-Dar es Salaam alters, ego has contact less than once a week, usually by letter, telephone or email. On the other hand, all contact was lost with every tenth alter (11%) who provided some type of support during the situation of forced migration. This percentage doubles (20%) when considering only those alters met before arriving in Dar es Salaam. This group of network members often include good Samaritans who helped ego during the flight, e.g., by providing accommodation or sharing food. 5.1.4 Ways of Meeting Alters and Types of Support Received On average, the five most important ways of meeting alters are through the neighborhood (22%), through a mutual friend (19%), as a relative or in-law (18%), through a religious congregation (10%) and through one's daily activities, such as work or school (9%).13 Comparing between alters met before and after arriving in Dar es Salaam, significant changes in importance appear among four of the above ways of meeting people. Whereas relatives make up 40% of the supportive pre-Dar es Salaam alters, they represent only 3% of the alters met after.14 On the other hand, the neighborhood, mutual friends and the church or mosque gain in importance as ways to meet people and establish new relationships. The neighborhood generates more than 13 All data in this section are compiled in Table 5-5. 14 Spouses as well as members of the family-in-law were categorized as relatives.

PAGE 265

249 twice as many alters for ego in Dar es Salaam compared to before (29 instead of 13%), and so do mutual friends (25 instead of 10%) and especially the church and mosque (14 instead of 3%). Together, these three avenues of meeting people generate more than two thirds (68%) of all of ego's relationships established in Dar es Salaam, compared to only 26% in the pre-Dar es Salaam era. When asked the question "In what respect did this person help you ?," not seldom did respondents give multiple answers. For exemple, a neighbor can provide ego not only with friendship and advice but with material assistance as well, such as water and cooking oil. Multiple answers with regard to the type of support received were given for 24% of the alters listed. On average, ego receives friendship, advice, companionship and emotional support from two thirds of all his/her alters (66%), and some type of material assistance such as money (21%), food (11%), accomodation (8%), clothes (3%), or other items (2%) like household articles, medicines, merchandise, etc. from close to half of all his/her alters (47%). Among "a specific deed" (6%) are categorized instances such as when an alter helped ego to find a relative residing in Dar es Salaam, or when a network member drove ego's sick child to the hospital. Among the alters known before arriving in Dar es Salaam, there are military officers who smuggled ego out of the home country, while other alters are currently guarding ego's personal properties (e.g., house, car, fields) left behind. "Providing news from home" is equally appreciated by the respondents as a type of support, in that it helps alleviate to a certain extent the worries about relatives and friends that remained behind (see chapter 3). Mainly pre-Dar es Salaam alters bring news from home (6%). Only a small percentage of alters was reportedly able to assist ego in

PAGE 266

250 finding a job or establishing self-employment, whether among alters met before (2%) or after arrival in Dar es Salaam (3%). 5.1.5 Summarizing Findings From the data on the "average" social network of Dar es Salaam's urban refugees who arrived from one of the Great Lakes countries less than ten but more than one year ago, significant differences arise when disaggregating alters by the moment of establishing the relationship, i.e., before and after ego's arrival in Dar es Salaam. Among the supportive network members that ego met after arriving in Dar es Salaam are significantly larger proportions of women and of Tanzanians than among pre-Dar es Salaam alters. While the increase in the proportion of Tanzanian alters is obviously due to the geographical location and the fact that the refugees lived scattered across the city and not in ethnic enclaves, the reasons for the significant increase in the number of female alters is not so easily explainable at first sight and will be contextualized in chapter 6. No significant variations appear between preand post-Dar es Salaam alters with regard to age, marital status, domestic arrangements and the number of children, except for the increase in the number of alters of whom ego does not know the age. The characteristic that appears to be the least relevant to ego when establishing a relationship with alter, appears to be the latter's education level, illustrated by the extremely high number of "don't know" replies with regard to the education level of post-Dar es Salaam alters. The variations in the occupations of preversus post-Dar es Salaam alters approaches those between the jobs respondents held before leaving home and in Dar es Salaam, namely an increase in self-employment and decrease in paid employment while

PAGE 267

251 there is virtually no change in the number of persons engaged in trade activities. The major difference is that only 6% of alters are unemployed compared to 24% of the respondents. This is not so surprising in view of the fact that ego was asked to list ten supportive alters, i.e., persons who could afford to provide ego with some type of assistance. Among post-Dar es Salaam alters, Dar es Salaam residents make up the overwhelming majority, as do alters that ego has known for less than 5 years and with whom s/he has daily to weekly contacts. Among pre-Dar es Salaam alters, the reverse trends are observable for obvious reasons: the geographical location, average length of residence of the respondents in Dar es Salaam and physical proximity inhibiting frequent face-to-face contact. Whereas relatives make up close to half of all pre-Dar es Salaam network members, once in Dar es Salaam the three most important ways of meeting alters are the neighborhood, through mutual friends or via one's religious congregation, whether church or mosque. With regard to the type of support provided, more than twice as many pre-Dar es Salaam alters provide ego with financial support compared to post-Dar es Salaam alters. From the latter, however, ego appears to receive significantly more emotional support (including advice, friendship, companionship, etc.) than from the former. These observations apply to the whole sample, combining refugee men and women, Congolese, Burundese and Rwandese, young and old. In order to explore the different socio-culturally inspired ways in which members of various social groups cope--the ultimate goal of this dissertation project--the subsequent sections will disaggregate the social network data for gender, nationality and age group of ego, albeit alter, starting with rates of homophily.

PAGE 268

252 5.2 Homophily As explained in the first chapter, studies have shown that, as homophilous relationships are indicative of strong ties, they represent more important conduits of social support than either weak ties or heterophilous ties.15 In a homophilous relationship, ego and alter may share common characteristics such as, e.g., gender, age, nationality, marital status, child status, religion, professional activity, etc. (McAllister and Fischer 1983:83). For the purpose of this study, only the extent to which Dar es Salaam's urban refugees share the same gender, nationality and age with their supportive alters is explored and compared to results from other studies, where available. In addition, I disaggregate the rate of homophily across ego's gender, nationality and age, and compare the respective rates of homophilous relationships with alters met before arriving in Dar es Salaam to those with alters met after. All data are compiled in Table 5-6. 5.2.1 Same Gender On average, respondents share their gender with 64% of their alters,16 and this proportion does not change when considering separately alters met before or after arriving in Dar es Salaam. However, when disaggregating ego for gender, it becomes clear that female respondents count significantly17 fewer women among their supportive alters (55%) than male respondents do men (71%). Across age categories, the proportion of alters with whom ego shares the same gender is approximately the same (between 63 15 E.g., Haines et al. 1996:254; Lin et al. 1985:249; Beggs et al. 1996:217; Marsden 1988:58. 16 This is very similar to the 62% resulting from a study by Marsden (1988) based on the 1985 General Social Survey (GSS). 17 "Significant difference" is defined for the purpose of this study as a difference of 10%, given that the margin of error at a confidence level of 95% for a sample of three hundred lies between 5 and 6%.

PAGE 269

253 and 66%), and across nationalities as well some variations appear (between 59 and 69%), but none as significant as when comparing men with women. Comparing rates among preand post Dar es Salaam alters, it appears that particularly men establish a proportionally larger number of relationships with persons of the opposite sex in the situation of forced migration, thereby reducing their initial high rate of homophily. The opposite pattern emerges among women who count more alters of the same sex among post-Dar es Salaam alters than among their pre-Dar es Salaam alters, thereby increasing their initially lower homophilous rates. The proportion of female alters of women respondents increases from 47 to 62%, that of men's male alters decreases from 79 to 67% when comparing relationships established before and after arrival in Dar es Salaam. Across nationalities, there are no significant variations in homophilous rates when comparing preand post Dar es Salaam alters. 020406080100AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< All alters Before After Looking at age groups however, it is only among the oldest respondents that the proportion of homophilous ties changes significantly in the situation of forced migration, namely an increase from 55 to 71% among preand post-Dar es Salaam alters respectively. This finding is likely related to the fact that the majority of the respondents older than 45 years of age are women (see chapter 3), assumed to follow the trend among female egos, i.e., an increase in the proportion of same gender network members. Figure 5-1: Homophilous ties for gender

PAGE 270

254 5.2.2 Same Nationality On average, respondents share their nationality with half (50%) of their alters. There is a negligible difference between women (49%) and men (51%), while Congolese and the oldest respondents have the highest proportion of alters with the same nationality (55%), compared to the smallest proportions among Rwandese (45%) and the youngest respondents (46%) respectively. On average, the rate to which ego shares the nationality of his/her supportive alters decreases by half when comparing the network members met before Dar es Salaam (77%) with those met after (31%). 020406080100AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< All alters Before After Figure 5-2: Homophilous ties for nationality While the proportion of homophilous relations with regard to nationality is more or less the same among alters met before Dar es Salaam across categories of gender, nationality and age (between 72% and 82%), pronounced differences arises when focusing solely on post-Dar es Salaam alters. Congolese, as well as Rwandese, share the nationality of approximately four in five of the alters they met before arriving in Dar es Salaam (79 and 80% respectively), yet among the post-Dar es Salaam alters, the latter's rate drops to 22% compared to only 39% for the former. This means that Congolese count twice as many compatriots among the persons they establish relationships with in Dar es Salaam than do Rwandese, while the Burundese average in-between (32%).

PAGE 271

255 5.2.3 Same Age Category The average frequency of alters that are in the same age group as ego, is lower than the respective homophilous rates for nationality and gender (34%) and hardly changes when comparing pre and post Dar es Salaam alters (36 and 33% respectively). On average, the three nationalities have approximately the same rate (between 34 and 35%). Men tend to have somewhat more age mates among their alters (38%) than women (31%), but the largest difference exists between the youngest category of respondents (24%) and the 26 to 35 year age group (44%). Very little variation appears, however, across categories when comparing rates among alters met before or after Dar es Salaam. 01020304050AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< All alters Before After Figure 5-3: Homophilous ties for age category 5.2.4 Summarizing Findings The average rate of relationships between persons of the same gender are consistently higher than those for same nationality and same age. This indicates that sharing the same gender is a more important element in establishing and/or maintaining a relationship, than sharing the same nationality or age, regardless of whether the respondent is male or female, Congolese, Burundese or Rwandese, and young or old. In addition, we observed that the homophilous rate for gender increases significantly among women, while decreasing significantly among men in a situation of forced migration. This indicates a substantial increase in the proportion of women among the refugees'

PAGE 272

256 supportive network members in Dar es Salaam, a finding which will be explored and contextualized in the following chapter. The homophilous rate for nationality decreases tremendously across gender, nationality and age groups due to the geographical displacement of the forced migrants rather than a conscious choice on the part of ego, and as we saw in the previous section, the majority of relationships established in Dar es Salaam are with Tanzanians. Some nationalities appear to enjoy more homophilous relationships in Dar es Salaam than others, indicating stronger and more supportive ties and suggesting increased access to more informal support sources (see chapter 1). Congolese have the highest proportion of homophilous relationships compared to the Burundese respondents, while the Rwandese have the least supportive networks. Beggs et al. (1996) have argued that individuals embedded in more homogenous networks are less likely to seek formal support than those in more diversified networks, because of higher access to informal sources of support through their personal social networks. Their thesis seems to be confirmed when comparing the proportion of respondents that have sought assistance from UNHCR across nationalities (Table 3-5). More than half of the Rwandese respondents are registered with UNHCR (56%), compared to only 22 and 31% for the Congolese and Burundese respectively. Even as UNHCR provides only extremely few of the registered refugees with material assistance, their commonly called "protection letter," certifying the bearer's refugee status, is a necessary prerequisite to assistance from the other humanitarian agencies in Dar es Salaam (see chapter 2). The underlying reasons as to why there is a difference between nationalities in the rate of homophilous ties and the composition of networks with regard to access to informal support sources will be explored and contextualized in the following chapter.

PAGE 273

257 5.3 Disaggregating Ego In order to explore to what extent the respondents build relationships and receive support from persons of a different gender, nationality or age group than their own, this next section compares the composition of the networks of male and female respondents, of each of the three nationalities and each of the four age groups. In addition, separate frequencies are provided for network members met before arrival in Dar es Salaam and those alters met afterwards. 5.3.1 Gender, Nationality and Age Groups The following section elaborates on some the characteristics of ego's alters, namely gender, nationality and age group, that were touched upon briefly as homophilous relationships in the previous section. However, unlike gender, nationality and age groups have more than two mutually exclusive categories. In order to fully explore the trends in network composition across ego's gender, nationality and age group, I decided it would be useful to present the following disaggregated information in detail.18 Gender. From the data on the "average" social network, we gathered that--although still slightly in the minority--the proportion of women increases with 45% among alters met after arriving in Dar es Salaam compared to among pre-Dar es Salaam alters (i.e., from 33 to 48%). This trend is consistent for all categories, albeit it is most outspoken among male and Rwandese ego's. In both of these two categories, the proportion of women among post-Dar es Salaam alters increases by 57% (from 21 to 33%, and from 28 to 49% respectively), compared to only a 32% increase among female respondents and a 18 All data in this section are compiled in Table 5-7, unless indicated otherwise.

PAGE 274

258 mere 21% increase among Congolese egos.19 The increase in female network members across the different age groups is more or less constant, varying between 40 to 44%. 020406080100AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Women Men 020406080AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Women Men Figure 5-4: Pre-Dar es Salaam alters by gender Figure 5-5: Post-Dar es Salaam alters by gender Nationality. The decreases in the proportions of Congolese, Burundese and Rwandese alters after arriving in Dar es Salaam and the concurrent substantial increase in the proportion of Tanzanians in the "average" social network are consistent across all categories. Whether ego is male, female, Congolese, Burundese, Rwandese, young or old, s/he will count many more Tanzanians and far fewer persons from the Great Lakes region among his/her post-Dar es Salaam alters (Fig 5-7) than among his/her supportive network members met before arriving in Dar es Salaam (Fig 5-6). The extent of the decrease and increase of the respective proportions of the different nationalities does not differ between male and female ego's, and is consistent as well across age categories except for the 26 to 35 year old respondents. Respondents in 19 These trends will be fully contextualized in chapter 6.

PAGE 275

259 this age category count substantially more Congolese (42%) and significantly fewer Rwandese (14%) among their pre-Dar es Salaam alters than any of the other categories, while having significantly fewer ties with Tanzanians in Dar es Salaam (44% compared to between 57 and 61%). 020406080100AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Congol Burund Rwand Tanzan Other 020406080AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Congol Burund Rwand Tanzan Other Figure 5-6: Pre-Dar es Salaam alters by nationality Figure 5-7: Post-Dar es Salaam alters by nationality With regard to respondents of different nationalities, the most striking is the overall decrease in alters originating from the Great Lakes region. The changes related to the frequencies of relationships between compatriots as a result of forced migration have already been discussed in the previous section, as homophilous ties. Yet, from the remaining data we gather that Congolese are present in Burundese and Rwandese networks in significantly larger numbers than vice versa regardless of the moment the relationship or tie was established. Burundese egos count between 12 and 17% Congolese among their alters, roughly the same as Rwandese who have between 11 and 14% Congolese alters. Congolese egos, on the other hand, have very few Burundese

PAGE 276

260 (4%) and even fewer Rwandese (between 4 to 1%) supportive alters. The same trend is observable for Rwandese egos' Burundese alters (between 2 and 5%) and Burundese egos' Rwandese alters (between 3 and 2%). The most obvious explanation for this phenomenon would be the larger population of Eastern Congo compared to both Burundi and Rwanda: 23 million versus 6 and 8 million respectively (see chapter 2). However, other possible reasons of a more socio-cultural and historical nature will be explored in depth in chapter 6. Age. Even as the age of 12% of alters was not known to ego, it still remains worthwhile to look for trends among the 88% of alters whose age was reported. The frequencies of alters of the four age categories does not show significant variations across gender or nationality. The most significant differences in the proportion of alters according to age group are found when disaggregating for ego's age group. Most of these have been discussed in the previous section, as homophilous ties. From the additional data it appears that egos in the three oldest age categories count far fewer alters of the youngest age category among their alters than vice versa (namely, between 3 and 11% compared to between 10 and 38%), indicating that relatively less support, material or immaterial, is forthcoming from alters under the age of 25 compared to others. This phenomenon is not only linked to a weaker financial position of the youngest persons, but finds its roots in a social system based on the principle of gerontocracy (more in chapter 6). Comparing prewith post-Dar es Salaam alters, there appears to be no significant variation in the proportions of alters according to age, apart from the increase in the numbers of alters whose age ego is unaware off.

PAGE 277

261 01020304050AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< <25 26-35 36-45 >45 Don't know 01020304050AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< <25 26-35 36-45 >45 Don't know Figure 5-8: Pre-Dar es Salaam alters by age group Figure 5-9: Post-Dar es Salaam alters by age group Summarizing the above findings on the effects of the situation of forced migration on the composition of the urban refugees' social networks, it appears that with regard to age there is no change when comparing prewith post-Dar es Salaam alters. As far as nationality is concerned, the geographical relocation to Dar es Salaam is obviously at the roots of the higher proportion of Tanzanian nationals among the alters (even as the 26 to 35 year age group slightly diverges from the general trend). The situation of forced migration appears to considerably influence the gender composition of the urban refugees' social networks however. The reason for this overall increase of women among post-Dar es Salaam alters is not readily explainable and will be further explored in chapter 6. Other differences in the composition of the social networks across the various social categories, regardless of the effects of the situation of forced migration, will equally be contextualized in the next chapter, such as, why Rwandese and men show a

PAGE 278

262 significantly higher increase in female network members compared to Congolese and women, why Congolese are represented in higher number among the social network members of Burundese and Rwandese egos than vice versa, and why younger alters are underrepresented among the social network members of the older egos. 5.3.2 Attributes of Relationships Whereas on average, ego reported having received some type of support since leaving home from more alters met after arriving in Dar es Salaam (58%) than from those met before (42%), no significant variations appear when comparing networks across ego's gender and nationality.20 With age however, there appears a decreasing propensity to establish new relationships when one is away from one's familiar surroundings. 020406080AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Before After 0204060AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Dar es Salaam Home country Elsewhere in Africa Global (excl.Africa) Figure 5-10: Preversus post-Dar es Salaam alters Place of Residence. With regard to alter's place of residence, there is very little variation across ego's gender, nationality or age group, for post-Dar es Salaam alters, as the overwhelming majority of them resides in Dar es Salaam (between 90 and 95%). Figure 5-11: Pre-Dar es Salaam alters by place of residence 20 All data in this section are compiled in Table 5-8, unless indicated otherwise.

PAGE 279

263 050100AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Dar es Salaam after Home country after Elsewhere in Africa Global (excl.Africa) Figure 5-12: Post-Dar es Salaam alters by place of residence With regard to pre-Dar es Salaam network members (Fig 5-11), women and men, as well as Congolese, Burundese and Rwandese respondents follow the average trend of having most pre-Dar es Salaam alters residing in the home country. With regard to egos of different age groups, only those younger than 25 and between 36 and 45 years of age have the majority of their supportive pre-Dar es Salaam alters live at home. While the 26 to 35 year old respondents have a majority of their pre-Dar es Salaam network members reside in other African countries--suggesting a higher rate of mobility across the continent--the oldest egos count the highest percentage of Dar es Salaam residents among their pre-Dar es Salaam alters (41%). This last finding is consistent with the fact that this category of oldest respondents found accommodation with Tanzanian friends upon arrival in Dar es Salaam to a far larger extent than any of the other respondents (see chapter 3).21 Length of the Relationship. Given that the overwhelming majority of ties established before arrival in Dar es Salaam were established more than five years ago (between 89 and 98%), there is little variation across categories with regard to pre-Dar es Salaam alters (Fig 5-13). Regarding post-Dar es Salaam alters, there are no significant differences between male and female egos with regard to the length of the relationships 21 Both trends are fully contextualized in chapter 6.

PAGE 280

264 they have with their social network members, contrary to the different nationalities or age groups. 050100AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< < 1 year 1 to 2 years 3 to 5 years > 5 years Figure 5-13: Pre-Dar es Salaam alters by length of relationship 0204060AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< < 1 year 1 to 2 years 3 to 5 years > 5 years Figure 5-14: Post-Dar es Salaam alters by length of relationship Disaggregating by ego's nationality, as the most recent arrivals, Congolese count significantly more persons among their post-Dar es Salaam alters that they have known for a period of only between one and two years (49 versus 35 and 37%). Burundese, who arrived the earliest in Dar es Salaam, logically count significantly more alters whom they have known for more than five years (25 versus 12 and 15%). Notwithstanding these comparative differences, across all three nationalities the bulk of ties with post-Dar es Salaam alters is one to five years old, consistent with the fact that the majority of respondents arrived in Dar es Salaam less than 5 years ago (Chapter 3). By the same logic, the youngest respondents, who arrived most recently have significantly more alters (52%) that they met only one to two years ago, while the 36 to 45 year old egos, of whom a larger number arrived in Dar es Salaam earlier, have

PAGE 281

265 significantly more ties that are 3 to 5 years old (37%) than any other age group. In other words, the social network data are consistent with the information presented in chapter 3. Frequency of Contact. Comparing across categories the proportions of post-Dar es Salaam alters (Fig 5-16), ego has contact with on a daily, weekly or monthly basis, there appear to be few trends divergent from the average frequencies. 010203040AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Daily Once/twice a week Once/twice monthly Less than monthly No communication 0204060AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Daily Once/twice a week Once/twice monthly Less than monthly No communication In addition, there are no gender differences in the level of contact ego has with pre-Dar es Salaam alters, while variations across nationality or age are not significant. Figure 5-15: Pre-Dar es Salaam alters by frequency of contact Figure 5-16: Post-Dar es Salaam alter by frequency of contact From the above, it appears that the situation of forced migration has distinct, yet easily explainable, effects on the composition of the social networks of Dar es Salaam's urban refugees with respect to place of residence of alters, length of relationship and frequency of contact. Unsurprisingly, contrary to the pre-Dar es Salaam alters, the majority of post-Dar es Salaam alters reside in Dar es Salaam, and most of them ego has known for less than five years (allowing for some variation across nationalities depending on their time of arrival in Dar es Salaam). In view of this, ego logically had more daily

PAGE 282

266 and weekly contacts with his/her post-Dar es Salaam alters, compared to monthly contacts with his/her pre-Dar es Salaam alters, few of whom reside in Dar es Salaam. However, the reason as to why the age group of 26 to 35 year olds have a majority of their pre-Dar es Salaam alters live outside of Tanzania or the home country--contrary to the other age groups--remains one of the topics to be further explored in chapter 6. 5.3.3 Alters by Ways of Meeting and Types of Support Provided Having presented various demographic aspects of the refugees' social networks' composition disaggregated for gender, nationality and age, I now turn to two of its more dynamic aspects, namely, the ways of meeting network members and the types of support provided.22 Ways of Meeting. As mentioned before, there is an important shift, on average, from "being relatives" as the most frequent way through which ego knows his/her pre-Dar es Salaam alters (40%) to "being neighbors," "through a mutual friend," or "through the church/mosque" as ways of meeting people and establishing ties after arrival in Dar es Salaam (68%). This shift is consistent among both male and female egos, whether Congolese, Burundese or Rwandese, young or old. With regard to pre-Dar es Salaam alters, we observe that Congolese establish more than twice as many supportive ties through the church or mosque (22%) than do either Burundese (8%) or Rwandese (11%). This trend is explained by the fact that church communities have played an exceedingly important role in the social, economic, and political every day life in Eastern Congo for decades, suggesting a pre-flight behavior among Congolese of seeking support from church communities to a much larger extent 22 All data in this section are compiled in Table 5-9, unless indicated otherwise.

PAGE 283

267 than among Burundese or Rwandese (see chapter 2). Religious institutions as ways of establishing supportive ties increase in importance with age as well, both among preand post-Dar es Salaam alters, quadrupling from 7 to 28% among the oldest respondents. 0102030405060AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< As neighbor Through a friend Is a relative Church/Mosk Work/School Through a relative Professionally Other Figure 5-17: Pre-Dar es Salaam alters by way of meeting 010203040AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< As neighbor Through a friend Is a relative Church/Mosk Work/School Through a relative Professionally Other Figure 5-18: Post-Dar es Salaam alters by way of meeting Types of Support Received. The trends in the proportions of network members providing certain types of support when comparing preand post Dar es Salaam alters are consistent across ego's gender, nationality or age group, namely that post-Dar es Salaam alters provide significantly more emotional support (inclusive of advice, companionship, friendship, etceteras), but less financial support than pre-Dar es Salaam alters. Across nationalities we observed that while Rwandese receive financial support from significantly more pre-Dar es Salaam alters (46%) than either Congolese (31%) and especially Burundese (24%), they receive food, accommodation and clothes from consistently fewer alters (13%) in Dar es Salaam than do Burundese (24%) and especially

PAGE 284

268 Congolese (31%). No variations of this significance appear when comparing support from preversus post-Dar es Salaam alters across gender of age group of ego. 020406080AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Emotional support Financial support Food, clothes, accomodation Specific deed Helped find a job Other material help 020406080100AverageWomenMenCongoBurundRwand<25y26-35y36-45y45y< Emotional support Financial support Food, clothes, accomodation Specific deed Helped find a job Other material help Figure 5-19: Types of Support received from pre-Dar es Salaam alters Figure 5-20: Types of Support received from post-Dar es Salaam alters 5.3.4 Summarizing Findings Many of the composition characteristics of the "average" social network presented earlier in this chapter are shared by female and male respondents, Congolese as well as Burundese and Rwandese, young and old. There is the increase in the proportion of female and Tanzanian network members, the majority of post-Dar es Salaam alters residing in Dar es Salaam, the bulk of ties in Dar es Salaam being younger than five years old, different ways of meeting new network members compared to the pre-flight situation and the fact that most financial help is provided by network members that ego knew from before arriving in Dar es Salaam. A few variations in trends are easily explainable as directly related to the specifics of the respective conflicts, e.g., the fact that the Congolese have significantly more

PAGE 285

269 younger ties with post-Dar es Salaam alters than Burundese is directly linked to the fact that the latter, on average, arrived in Dar es Salaam at an earlier date than the former. Other trends result from specific infrastructural conditions in the home country, such as the fact that Congolese are significantly more likely to meet network members through church communities, who play a pivotal role in the functioning of society at home. One outcome is directly related to the selection criteria of the respondents, namely that the youngest respondents have significantly more younger relationships with network members met in Dar es Salaam. This observation is the result of the fact that selection criteria stipulated that respondents should have memories of life at home as an adult. Hence, most of the respondents younger than 25 years of age arrived in Dar es Salaam only a few years ago. However, a number of questions remain unanswered. Why do Burundese and Rwandese egos have ties with a significantly larger proportion of Congolese than vice versa? Why do Rwandese receive significantly more financial assistance from network members met before arriving in Dar es Salaam? What is the background to the fact that alters of the youngest age group (i.e., younger than 25) are significantly underrepresented in relationships with older respondents (i.e., older than 25)? Why do respondents in the 26 to 35 year age group have significantly more supportive network members living elsewhere in Africa (i.e., outside of Dar es Salaam and the refugee camps)? Why should egos of between 36 and 45 years old have more older ties, or in other words, why did they arrive in Dar es Salaam earlier than any of the other age groups? And lastly, why do the oldest respondents (>45) have significantly more pre-Dar es Salaam alters residing in Dar es Salaam than any of the other age groups? All of these questions are contextualized and answers suggested in chapter 6. But first, I intend to compliment the

PAGE 286

270 above findings by disaggregating the social network data by the gender, nationality and age group of the alters, given that close to two thirds of all alters are fellow-refugees of the respondents (Table 5.2). 5.4 Disaggregating Alters The previous section compared the different ways ego meets alters and the types of support s/he receives from them according to the characteristics of ego, namely ego's gender, nationality and age group. However, it should prove useful for the purpose of this research project, to explore the characteristics of alters in function of the ways of meeting and the type of support provided to ego. In other words, who do the urban refugees in Dar es Salaam meet through the church or the neighborhood, and who provides them with money, food, accommodation or friendship in terms of gender, nationality and age? 5.4.1 Ways of Meeting 0102030405060WomenMenTanzanCongoBurundRwand<25y26-35y36-45y45y< As neighbor Through a friend Is a relative Church/Mosk Work/School Through a relative Professionally Other Figure 5-21: Pre-Dar es Salaam alters by ways of meeting The average frequencies listed in the first two lines of Table 5-10 give an indication of the relative proportions of preand post Dar es Salaam alters met through each of the ways of meeting.23 Concurrent with earlier findings in this chapter, more post-Dar es 23 All data in this section are compiled in Tables 5-10.

PAGE 287

271 Salaam alters were met through the neighborhood (29%), through mutual friends (35%) and the church/mosque (14%), while relatives make up the majority of pre-Dar es Salaam supportive alters (40%). 01020304050WomenMenTanzanCongoBurundRwand<25y26-35y36-45y45y< As neighbor Through a friend Is a relative Church/Mosk Work/School Through a relative Professionally Other Figure 5-22: Post-Dar es Salaam alters by ways of meeting Gender. None of the presented ways of meeting appears to generate significantly more male than female network members in the situation before arriving in Dar es Salaam. In Dar es Salaam, however, significantly more female than male alters are met through the neighborhood (35% versus 23%). This is not very surprising, as women, especially those with (small) children are much more likely to spend the day working at home or in the neighborhood, thus facilitating contact between women. Nationality. Whereas before the flight, respondents met more Tanzanian alters through mutual friends than any other way of meeting (26%),24 after arrival in Dar es Salaam, it is the neighborhood that brings ego in contact with the majority (45%) of his/her Tanzanian alters. New network members from the Great Lakes countries, however, are significantly more likely to be met through mutual friends (between 28 and 40%) than any other way of meeting. 24 Informants often mentioned business relations with Tanzanian traders commuting between Tanzania and the Great Lakes region.

PAGE 288

272 The next three major ways of meeting Congolese, Burundese or Rwandese alters are, in order of importance: church/mosque (between 15 and 17%), work/school (between 12 and 14%) and, only then, the neighborhood (between 5 and 11%). This finding concurs with the fact that the urban refugees live scattered across the many wards of the city and not in enclaves (Map 1-7 in appendix a). As such, they will only rarely meet fellow-refugees through the neighborhood, and rather depend on mutual friends or public spaces to meet new network members originating from the Great Lakes region. Age. Both the neighborhood and mutual friends are two major ways of meeting alters of any age in Dar es Salaam. Younger people however, are also easily met through work/school (26%) and alters older than 35 through the church/mosque (20 and 19% respectively). As mentioned before, older respondents reported going to church in larger numbers and higher frequencies than persons in the two younger age groups (see chapter 4). Even though few of them thought of the church/mosque as a place to meet friends, it is obvious from the data that religious congregations are an important source of supportive network members, especially of their own age. 5.4.2 Types of Support Provided 010203040506070WomenMenTanzanCongoBurundRwand<25y26-35y36-45y45y< Emotional support Financial support Food, clothes, accomodation Specific deed Helped find a job Other material help Figure 5-23: Pre-Dar es Salaam alters by type of support On average, both preand post-Dar es Salaam alters provide ego with more emotional than financial support, however, the difference in proportion of alters is

PAGE 289

273 significantly greater among the latter (73% versus 15%) than the former (56% versus 34%).25 020406080100WomenMenTanzanCongoBurundRwand<25y26-35y36-45y45y< Emotional support Financial support Food, clothes, accomodation Specific deed Helped find a job Other material help Figure 5-24: Post-Dar es Salaam alters by type of support Gender. No significant variations appear in the proportions of male versus female alters providing different types of support, except regarding material assistance other than money. Female network members are twice as likely to provide ego with food, clothes and/or accommodation in Dar es Salaam, than male network members (29% versus 16%). This finding results from the fact that women are the ones in charge of domestic matters rather than men in the regional socio-cultural context. Nationality. Alters that are particularly instrumental in providing ego with food, clothes and/or accommodation (62%) are Tanzanians with whom a relationship existed before arrival in Dar es Salaam. As mentioned before, these more often than not were traders plying their merchandise in the countries of the Great Lakes region, or others who had resided in the Great Lakes region for a number of years. Not seldom had these Tanzanian nationals fled to their home country together with the very persons who had helped them integrate (whether in Congo, Burundi or Rwanda) and to whom they were happy to extent the same level of hospitality. 25 All statistical data in this section are compiled in Tables 5-11.

PAGE 290

274 Rwandese alters are by far the category of network members providing financial support in large numbers (48%). This finding tallies with the observation that Rwandese refugees receive money from significantly larger numbers of alters than other refugees. It indicates that Rwandese refugees have ties with more economically well-to-do persons to a far larger extent than either Burundese or Congolese do.26 This finding also concurs with observations from informants in the previous chapters, namely that Rwandese refugees in general dispose of more financial resources than either Congolese or Burundese, allowing them to engage in trade in larger numbers and not worrying about residence and work permits, which they can easily afford to pay for. Age. The data show that with age more alters increasingly provide financial support to ego, to the extent that pre-Dar es Salaam alters over the age of 45 manage to provide as much financial as emotional help (47%). This trend indicates that a person's financial capacity increases with age, a phenomenon closely related to the general African system of gerontocracy. The complete socio-cultural background of the issue is one of the topics of chapter 6. 5.4.3 Summarizing Findings Gender. In a socio-cultural context where women are much more likely to spend the day near the home--whether taking care of children and household tasks or engaging in petty trade and selling food stuff--it is not surprising to find that the neighborhood is the major source of relationships with female network members. In the same logic, where women are the ones in charge of domestic matters, female alters will logically be 26 Contextualization of this observation will take place in chapter 6.

PAGE 291

275 in a better position to provide support in the form of food, clothes and/or accommodation than male network members, who have more access to financial sources. Nationality. When disaggregating ways of meeting by the nationality of alters, there emerges a stark difference between Tanzanian alters and network members originating from the Great Lakes region. When still at home, in either Congo, Burundi or Rwanda, relationships with Tanzanian nationals were most likely established through mutual friends, while supportive alters of Congolese, Burundese and Rwandese nationality from before the flight are largely relatives. In the situation of forced migration, it is fellow-refugees who are the most likely to be met through mutual friends, while the neighborhood becomes the major source of Tanzanian alters. Mutual friends, therefore, appears to be the most prominent way of meeting non-nationals: i.e., Tanzanians in the Great Lakes region and Congolese, Burundese and Rwandese nationals in Tanzania. As mentioned before, this is particularly so in situations where non-nationals do not live in enclaves, as it is the case in Dar es Salaam. As "locals," Tanzanians are also in a better position to provide the refugees with food, clothes and/or accommodation, but appear inclined to do so to a much larger extent only when the relationship with ego dates back from before the flight. Rwandese alters, in view of their financial ability to support ego, generally appear to belong to an economically advantaged layer of society. Age. The youngest alters are less easily met through mutual friends in Dar es Salaam and more through the neighborhood and at school. They are also the least well placed to provide others with either financial or other material support, such as food, clothes and/or accommodation. Respondents between 26 and 35 years of age, on the other hand, seem to be the most mobile age group, and are easier to connect to through

PAGE 292

276 mutual friends than any other age group. They also show signs of increasing financial autonomy in the sense that they are in a better position than the youngest alters to provide material assistance to ego. Older alters (over 35 years of age) in Dar es Salaam are increasingly met through the neighborhood rather than mutual friends, indicative of a decreasing level of mobility after the age of 35. Both among preand among post Dar es Salaam alters the proportion of persons providing ego with financial and other material support steadily increases with age, a pattern compatible with the African system of gerontocracy (see more in chapter 6). 5.5 Conclusions The above analyses have outlined the "average" social network of Dar es Salaam's urban refugees as well as variations between social groups regarding network composition, relationship attributes, ways of meeting and types of support received. Complimentary to these, I explored which alters were more likely to be met through certain ways or to provide which types of support. This was done in order to get an increasingly clearer picture of the cultural norms and societal positions ascribed to and governing the role of the respective social groups under study in this dissertation. Many, but not all, of the variations emerging from the social network analysis have been contextualized in this chapter and explained as a function of the information readily available in one of the previous chapters. The high proportion of Tanzanian nationals and of Dar es Salaam residents among post-Dar es Salaam alters evidently result from geographical factors, while the average length of relationships fluctuates according to the respective average times of arrival of the three national groups in Dar es Salaam. Considerably fewer persons among the network members are unemployed when compared to respondents--despite the fact that

PAGE 293

277 close to two thirds of all alters originate from one of the Great Lakes region and are largely assumed to be fellow-refugees--because of the nature of the name generator soliciting information on persons able to provide ego with some type of support. By and large, there are few significant differences in the demographic composition with regard to marital status, living arrangement, child status, educational level and professional occupation between the respondents and their network members. The larger proportion of Congolese network members in Burundese and Rwandese networks than vice versa (particularly among post-Dar es Salaam alters) is most probably directly related to the fact that there are more Congolese refugees residing in Dar es Salaam than Burundese and Rwandese. Even as the Burundese make up the majority of the registered refugee population in Tanzania, the findings of chapter 3 already suggested that the Congolese head directly for Dar es Salaam in larger numbers than the other two nationalities27 because of higher national rates of urbanization but also because they appear embedded in larger supportive networks. The higher frequency of Congolese among pre-Dar es Salaam network members results from the fact that a substantial number of the recent Burundese and Rwandese arrivals in Dar es Salaam spent a period of time in one of the refugee camps in Eastern Congo (before being forced to flee in 1996 and 1998) where they were more free to have contacts with the local population than is the case in the Tanzanian refugee camps (Chapter 4). Overall, pre-Dar es Salaam network members consist largely of relatives, among whom mostly men are instrumental in providing financial support to the urban refugees. Among post-Dar es Salaam alters, women play a significantly more important role in 27 In addition, the majority of the Tanzanian survey respondents estimated the Congolese to be the largest refugee population in Dar es Salaam (see chapter 2).

PAGE 294

278 providing accommodation, food, clothes and other material support to the refugees in view of the fact that the household is a female domain in the regional socio-cultural context. A corollary of this same principle, namely that women will spend more of their time closer to home than men, lies at the root of the observation that the neighborhood is the major source of female network members among post-Dar es Salaam alters. The finding that significantly more Tanzanian network members are met through the neighborhood than either Congolese, Burundese and Rwandese confirms that the refugees live scattered across the metropolis of three million inhabitants and not in ethnic enclaves. In these circumstances, it is not surprising that having mutual friends necessarily became the most important way of meeting compatriots or other fellow-refugees. Yet, a number of questions, particularly but not solely with regard to differences across gender and age groups, remain unanswered and need a larger framework to be fully understood. For example, why the proportion of women among post-Dar es Salaam network members is so much higher than among pre-Dar es Salaam alters, or why Congolese appear to be embedded in larger, more supportive social networks, thus less reliant on formal sources of support, than Burundese and especially Rwandese, or why the oldest group of alters provides more financial support while the 26 to 35 years old respondents appear to be connected to a geographical area beyond Dar es Salaam and the home country to a much larger extent than any other age group. The following chapter endeavors to present the cultural, economic, historical, political and social context in which the refugees' respective experiences and strategies are embedded.

PAGE 295

279 Table 5-1. Ego's alters disaggregated by Gender, Nationality and Age Nationality Gender >25y 26-35y 36-45y <45y Age not known Total/ Gender/ Nation. Total/ Nationality Tanzanian Women 58 147 103 73 121 502 55 137 93 71 85 471 Men 59 135 104 109 99 506 1008 55 113 89 93 85 435 906 Congolese Women 33 144 79 33 26 315 21 72 40 13 15 161 Men 51 178 137 83 36 485 800 28 87 53 34 29 231 392 Burundese Women 44 85 47 21 20 217 18 42 18 4 15 97 Men 38 124 110 59 33 364 581 15 39 23 17 26 120 217 Rwandese Women 21 81 40 19 7 168 9 48 10 4 4 75 Men 23 94 92 79 16 304 471 2 29 19 13 11 74 149 Other Afr. Women 1 1 6 0 2 10 1 1 2 0 1 5 Men 1 3 5 2 5 16 26 0 1 4 2 4 11 16 Other nat. Women 0 4 2 0 2 8 0 3 0 0 2 5 Men 0 7 5 13 1 26 34 0 4 1 3 1 9 14 Total/Age Women 157 462 275 146 178 1218 104 303 163 92 152 814 Men 173 541 453 345 190 1701 101 272 188 162 156 879 Total Age 330 1003 729 492 367 2921 205 575 351 254 308 1693 In italics: network members met after arriving in Dar es Salaam

PAGE 296

280 Table 5-2. Social Network Members by Gender, Nationality and Age (in %). Demographics of Alters whom Ego met (in %) Before arrival in Dar es Salaam n=1,228 After arrival in Dar es Salaam n=1,693 Total Average N=2,921 Gender Female 33 48 42 Male 6752 58Total100100 100 Nationality Tanzanian854 35Congolese3323 27Burundese3013 20Rwandese269 16Other nationality31 2 (Same Nationality as Ego)(77)(31) (49)Total100100 100 Age Younger than 251012 1126 to 35 years3534 3436 to 45 years3121 26Older than 45 years1915 17Don't Know518 12Total100100 100

PAGE 297

281 Table 5-3. Demographic and Other Socio-Economic Attributes (in %) Other Attributes of Ego's Alters met In %Before arrival inDar es Salaamn=1,228After arrival in Dar es Salaam n=1,693 TotalAverageN=2,921 Marital Status Single3238 35Married6355 59Divorced13 2Widowed34 3Don't Know10 1Total100100 100Domestic Arrangement Living with partner6254 58Not living with a partner3339 36Don't know57 6Total100100 100Number of Children Childless3137 34One children1013 12Two/Three children2625 26Four/Five children1611 13Six or more children1410 12Don't know34 3Total100100 100Level of Education Below Primary43 4Primary School1212 12Secondary School3529 32Beyond Secondary278 16Don't know2248 36Total100100 100Professional Occupation Trade2628 27Self Employed1426 21Paid Employment1812 15Student127 10Housewife711 9Church Official33 3Agriculture31 2Government Official32 2No occupation66 6Don't know84 6Total100100 100

PAGE 298

282 Table 5-4. Relationship attributes of Social Network Members (in %) Relationship attributes of Alters metBefore arrival in Dar es Salaamn=1,228After arrival in In Dar es Salaam TotalAverageN=2,921 In %N=1,693 When did you first meet this person? Before arriving in Dar es Salaam1000 42After arriving in Dar es Salaam0100 58Total100100 100 Place of Residence Dar es Salaam2893 66Western Tz (Incl. Refugee Camp)101 5Elsewhere in Tanzania32 2Congo170 7Burundi111 5Rwanda120 5Elsewhere in Africa81 4Elsewhere worldwide101 5Don't Know11 1Total100100 100Ego's Home Country341 15 How long have you known this person ? Less than one year014 8Between one and two years241 24Between two and five years529 19More than five years9316 49Total100100 100 How often do you have contact with this person ? Daily1135 25Once/Twice weekly1742 30Once/Twice Monthly3016 22Less than once monthly223 12No Communication204 11Total100100 100

PAGE 299

283 Table 5-5. Ways of Meeting Network Members and Types of Support Received (in %). Relationship attributes of Alters metIn %Before arrival in Dar es Salaamn=1,228After arrival in In Dar es Salaam TotalAverageN=2,921 N=1,693 How did you meet this person ? 3 19 Was/is my neighbor1329 22Through a mutual friend1025 19Is a relative *403 18In church/the mosque314 10Through my work/school910 9Through a family member86 7In his/her official capacity55 5Other128 10Total100100 100 In what respect did this person help you ?** Emotional support, friendship, advice56 73 66Financial support3415 21Food1013 11Accommodation87 8A specific deed74 6Clothes42 3News from Home61 3Helped me find a job/work23 2Other material help22 2 : Includes Family-in-Law **: Multiple answers possible.

PAGE 300

284Table 5-6. Homophilous relationships for sex, nationality and age group (in %) Ego Women MenCongol.Burund.Rwand.< 26 y26-3536-4545
PAGE 301

285Table 5-7. Social networks by ego's and alters' gender, nationality and age category (in %) Alter \ Ego Women Men Congol. Burund. Rwand. < 26 y 26-35 36-45 45< Average Women before 47 21 38 34 28 35 33 32 35 33 after 62 33 46 49 49 50 47 46 49 48 Men before 53 79 62 66 72 65 67 68 65 67 after 38 67 54 51 51 50 53 54 51 52 Congolese before 32 35 79 12 11 26 42 30 32 33 after 23 24 39 17 14 23 27 21 17 23 Burundese before 30 29 4 73 2 29 30 30 30 30 after 12 14 4 32 5 14 15 10 12 13 Rwandese before 29 24 4 3 80 35 14 31 29 26 after 10 7 1 2 22 5 13 9 10 9 Tanzanian before 7 10 10 10 5 7 11 6 9 8 after 53 55 55 47 58 57 44 58 61 54 Other Nat. before 2 2 3 2 2 3 3 3 0 3 after 2 0 1 2 1 1 1 2 0 1 < 25 before 8 12 8 14 8 22 11 3 5 10 after 8 16 12 16 9 25 8 3 3 12 26 to 35 before 33 36 42 32 31 38 44 29 20 35 after 34 33 30 32 39 34 44 28 17 34 36 to 45 before 31 31 31 30 32 21 28 40 33 31 after 21 20 25 17 20 14 17 30 32 21 > 45 before 22 18 17 16 27 12 14 24 35 19 after 15 15 18 10 16 10 12 18 36 15 Don't know before 6 3 2 8 2 7 3 4 7 5 after 22 16 15 25 16 17 19 21 12 18 E.g., Respondents between 26 and 35 years of age count 44% of Tanzanians among alters they met after arrival in Dar es Salaam.

PAGE 302

* : This includes 15%, who live in Kigoma and the refugee camps, compared to only 8 and 5% of Congolese and Rwandese refugees. **: This includes 10% who reside in Congo, compared to only 3% of the alters of Burundese respondents. 286Table 5-8. Relationship attributes by ego's gender, nationality and age category (in %) Ego Women Men Congol. Burund. Rwand. < 26 y 26-35 36-45 45< Average % of all alters met before 39 45 41 48 38 35 43 46 51 42 after 61 55 59 52 62 65 57 54 49 58 Dar es Salaam before 30 26 31 32 22 31 22 26 41 28 after 93 94 94 90 95 94 93 92 93 93 Home Country before 37 31 39 28 37 38 33 34 27 34 after 1 2 1 3 1 1 2 1 3 1 Elsewhere in Africa before 25 31 22 31* 25** 20 38 27 24 21 after 5 3 3 6 3 4 5 5 3 4 Out of Africa before 8 12 8 8 16 11 7 13 8 10 after 1 1 2 1 1 1 1 2 1 1 < 1 year before 0 0 0 0 0 0 0 0 0 0 after 14 13 10 18 14 18 17 6 7 14 1 to 2 years before 2 2 3 2 2 3 2 2 1 2 after 38 44 49 35 37 52 35 30 47 41 2 to 5 years before 5 6 7 5 3 4 9 3 1 5 after 32 26 29 22 34 26 25 37 30 29 > 5 years before 93 92 90 93 95 93 89 95 98 93 after 16 17 12 25 15 4 23 27 16 16 Daily before 10 11 14 10 7 15 9 8 13 11 after 34 35 32 38 34 39 35 30 32 35 Once/twice weekly before 17 17 12 16 23 15 11 21 30 17 after 41 40 40 31 48 40 39 41 38 42 Once/twice monthly before 31 30 24 30 36 34 29 29 28 30 after 18 16 18 18 13 17 17 16 19 16 < Once monthly before 22 17 27 16 16 19 21 20 14 22 after 3 3 5 3 2 2 3 5 6 3 No communication before 20 25 23 28 17 17 30 22 15 20 after 4 4 3 7 2 2 6 6 5 4

PAGE 303

Other before 33332 24332 after 53554 44332 *: Multiple answers possible 287Table 5-9. Ways of meeting and types of support received* by ego's gender, nationality and age category (in %). WomenMenCongol.Burund.Rwand. < 26 y26-3536-4545
PAGE 304

288Table 5-10. Alters by ways of meeting by alter's gender, nationality and age category (in %). Alters by Category Neigh-borhood N=647 Through FriendsN=547RelativesN=535Church/ MosqueN=282Work/ School N=274 Through RelativesN=203Profess-ionalyN=139OtherN=294Sample sizeN=2,921 Average before 13 104039 85121,228 after 29 3531410 6581,693Women before 15 84239 7412408 after 35 2231311 448812Men before 13 113839 9511820 after 23 283169 858881 7 11914102 Tanzanian before After 13 26 19 16 2 5 3 45 17 445906Congolese before 8 844511 6412408 After 11 2851714 7611392Burundese before 21 350 104364 After 11 1512 85217Rwandese before 12 838 153225 401513 514916 11311 101125 936 8 1212 1 50 811 After 230 76 < 26 years before 1 After 26 to 35 years before 30 4826 2720511 412842725 31411 657516 9406 79378 After 26 27224 7351> 45 years before 8 74566 9238 After 28 224193 810254 2011 33 13 8 910 After 36 to 45 years before 5 30 75 810 56

PAGE 305

289Table 5-11. Alters by type of support provided* by alter's gender, nationality, and age category (in %) Alters by Category Emotional Sample Other Support Job related Info Specific Food, Acco, Clothes Financial Size deed support Support. N=2,921 N=51 2 N=66 2 N=243 N=650 22 N=622 N=1668 Average before 56 34 13 1,228 After 73 15 22 5 32 2 1,693408 Women before 56 29 28 10 2 After Men before 7456 1137 2920 3 81215 820 22 21 After Tanzanian before 7346 1924 16 7 3 2 881 62 14 7 0 102 After 7013 2814 234 206 9 12322 149 2525 5 4 3 906 Congolese before 60 13 21 408392 After Burundese before 7660 1729 11 21 364217 After 77 13 15 Rwandese before 51 48 10 19 After < 26 years before 85 17 12 3 0 3 66 83 25 8 10 12 14 5 2 2 1 1 125 205 After 26 to 35 years before After 504118258 9 35 3 6279 2814 2621 7 1 2 427 3 2 2 575 36 to 45 years before 9 24 2 378 After 66 18 31 351238 > 45 years before After 4768 4720 2632 2 254 *: Multiple answers possible.

PAGE 306

CHAPTER 6 EMBEDDING THE REFUGEE EXPERIENCE The previous chapters illustrated how Dar es Salaam's urban refugees experienced conditions of war, hunger, fear, the death of family and loved ones, persecution by a hostile authority, loss of status and a familiar place in society, of livelihoods and homes, and of a future. They have also experienced the uncertainty of the flight, and the mixed relief, confusion and ambivalence of survival in a foreign land among strangers, while struggling with the implications of their illegal status. Before becoming refugees, however, they had "lives in which there was peace, stability, enough food to eat today and tomorrow, a place in society and a future for themselves and their children" (Geiger 1993:68). As historical human beings with a "wealth of prior learning upon which they will rely as much as possible and build upon when coping with new experiences," refugees will adapt to new circumstances by exploring alternative behaviors, and altering old ones only when they no longer produce the desired results (Geiger 1993: 76). The process of coping with loss, not only for refugees but for all human beings, consists of assimilating "new experiences by placing them in the context of a familiar, reliable construction of reality. This structure in turn rests not only on the regularity of events themselves, but on the continuity of their meaning" (Marris 1986:8). This continuity of meaning will be looked for by applying familiar strategies, whose outcomes the refugees consider predictable. It is only when these strategies no 290 longer generate the expected outcomes that people will initiate changes (innovate) but

PAGE 307

291 these changes are always "set within the parameters specified by our perceptual frameworks, and these are a function of our past" (Oliver-Smith 1986:16). The three previous chapters illustrated the different ways in which the Congolese, Burundese, and Rwandese men and women, young and old, deal with their situation of forced migration. Some of the differences in coping strategies between groups are directly related to the nature of the respective political conflicts and their historical roots, and were easily contextualized, others were not. Subscribing to Oliver-Smith's contention that these remaining, hitherto unexplained, variations can be understood only by looking at the historical backgrounds of the respective social groups, I aim to do just that in the following sections. The first section will demonstrate how regional socio-cultural gender perceptions, i.e., the political status of women in Africa, play a role in the innovative manner in which refugee women and men rebuild their social networks in the situation of forced migration. Comparing across nationalities, the second section illustrates that the historical, economic and political backgrounds of the three refugee producing countries provide the refugees with different sets of coping strategies. Regional cultural perceptions about age, related to the traditional East African age system societies, are equally at the basis of the variety in coping and networking strategies applied by the four different age categories. By using gender, nationality and age categories as analytical tools, the experiences of the Congolese, Burundese and Rwandese refugee men and women, young and old are truly embedded in their respective larger historical, economic, political and socio-cultural frameworks.

PAGE 308

292 6.1 Gender As argued at length in chapter 1, the anthropology of gender--contrary to the anthropology of women concerned with the collection of data on women, or feminist anthropology focusing on gender as the organizing principle of human social life--aims to explore the relationship between the discursive constructions of women and men and their political economic realities. This dynamic approach conceptualizes both men and women as taking advantage of the opportunities produced by social and material changes in their lives and endlessly renegotiating the boundaries of gender. As such, gender constructs are not only "constituted in the culturally patterned rhythms of everyday life," but, in addition, [h]istorical events and altered material circumstances, perceived and acted upon with the cultural categories of a given society, [will] influence the appropriation of gendered personhood and consequently the cultural constructs of gender. Gender constructs, as all ideological constructs, should thus be expected to change. (Grosz-Ngat 1989:168) Thus, constructs of gender, of what it means to be a woman or a man, change across space and across time, depending on specific historical circumstances and altered living conditions. Traditional African societies were constructed upon a corporate base that emphasized kinship, also called the corporate group, whereby the importance of maintaining the well-being of the social group superseded that of the individual (Mikell 1997:10). In view of the fact that family groups were the basic units of the community, and the right to political participation derived from membership in these units, both male and female members of high-status lineages or clans enjoyed political privileges (O'Barr and Firmin-Sellers 1995:192). As such,

PAGE 309

293 African women elites and aristocrats . emerged as political actors primarily because of their statuses and roles within certain families, whether as daughters, sisters, and mothers, or as wives and in-laws. (Mikell 1997:11) In this African context, belonging to elite kin groups was determinative of political participation to a much larger extent than gender was. The colonial period and the importation of Western gender constructs had a profound effect on the traditional principles governing political life and gender constructs in African societies. Most of the political theory which underlies Western liberal democracy and liberal democratic theory, has its roots in the separation of the public and the private (or domestic) spheres. Those individual citizens active in the public sphere were implicitly assumed to be male heads of households, and most analyses relegated women to the private sphere, where they were subsumed within a household headed by the individual male (Moore 1988: 21-4). Political theory, while appearing gender neutral--by maintaining a division between private and political life as central to liberal democracy--maintains a division between men and women, in which only men can be abstract individuals and thus eligible to become citizens (Pateman 1989). The political is therefore defined as masculine in a very profound sense, and the political construct of the nation-state conceptualized as a fraternity (Andersen 1991:7), where women's activities were considered politically irrelevant. The implicit gender ideology of the colonial administration, even where colonial actions appeared to free women in domestic issues (primarily marriage), categorized women simply as men's wives, thereby separating them from the larger kingroup. The concurrent introduction of a market-oriented economic model (coupled with the male-oriented Christian ideology) involved only men in the cash economy while relegating women to a subsistence role, resulting in a significant loss of status (Leacock and Etienne

PAGE 310

294 1980:5-22; Moore H. 1988:42-127). After independence, most African politicians, fearing ethnic conflicts, turned their back on traditional chieftaincy and women's public roles. In doing so, they continued to perpetuate the Western dichotomy between male and female spheres of action, and with it, the mechanics of women's subordination (O'Barr and Firmin Sellers 1995; Mikell 1997), a situation which led some academic observers of the 1980s and 1990s to speak of "African women's withdrawal from politics" (Drew 1995:1). The above, necessarily brief, historical overview provides the background of historical events and changed conditions in the lives of African men and women during the past century, and against which new cultural constructs of gender emerged. Even as traditional gender constructs were not simply replaced by the Western version, the altered political economies of African men and women unavoidably altered people's perceptions of what it entails to be a woman or a man, and gender boundaries were renegotiated both discursively and through daily praxis. It is in this perspective of discourse and praxis, that I undertook to explore the interplay between the gender perceptions and dynamics of the urban refugees in Dar es Salaam and the region, and their coping with a situation of forced migration. 6.1.1 Women as "Political Objects" My quest for an answer to the question what role gender plays in the coping strategies of Dar es Salaam's urban refugees was a long one. During the pre-dissertation phase of this research project, I asked informants what appeared to me--as a social scientist--to be a perfectly simple question, namely: "What are the differences between a Congolese man and a Congolese woman?" with the intent to derive from the answers follow-up questions geared towards gender specific problems of adaptation in the

PAGE 311

295 situation of forced migration. However, informants' frequently puzzled looks and replies such as "Do you mean biologically?" as well as decided statements that "There is no forms of individual entrepreneurial initiatives, the remainder of this section will focus exclusively on the interpretation of the perceptions on women. difference whatsoever" brought home the realization that a situationally more appropriate approach was called for. I subsequently reformulated the above question and instead asked informants to compare women (and men) from their respective nationalities to Tanzanian women (and men) with a view to discerning the strategies applied by the former to life as a refugee in Dar es Salaam compared to the daily praxis of the latter who are at home.1 It was an open question eliciting perceived differences with regard to any aspect of life, but the answers given were remarkably uniform in nature across nationality and gender. Both Tanzanian men and women were considered "lazier," "less ambitious" and "more promiscuous" than refugee men and women, but to different degrees. While Tanzanian men were mostly blamed for their laziness and lack of ambition in comparing them with refugee men,2 comparisons between Tanzanian women and refugee women centered primarily on the former's presumed promiscuity. With regard to comparing refugee men and Tanzanian men, Bonaventure is of the opinion that: Congolese men are strong and dynamic, and also intelligent because when you give him a job, he knows [how to do it] . abroad, he has no time to loose, thus he earns an income in order to survive. It's the opposite [for the Tanzanian] . they don't have this love for working. 1 In this particular context, the term Tanzanian was used to refer to persons with the Tanzanian nationality, but who, due to being residents of Dar es Salaam, belong to what is colloquially refered to as the Coastal culture. 2 While the perception on Tanzanian men is an intriguing one, and may be linked to the fact that--contrary to Congo, Burundi and Rwanda--Tanzania was under a socialist regime for decades, discouraging most

PAGE 312

296 His female compatriot Euphrasie concurs: You can have a [Tanzanian] healthy, strong man, but every evening he stays in this one place where he looks for beer. He's not thinking "I have to look for something to do because without work, how will I get money." No, he won't. Many of the men here, when they leave the bar, they'll ask their wife at home "What's for dinner?" They're very different here. Daniel, as well, perceives a lack of ambition among his male hosts. [the Rwandese] are much more hardworking than the Tanzanians. People here are too passive, they're drawn upon themselves, they don't want to see what's out there, meet the world, meet other people . The Rwandese is not like that, he likes to venture out. Only Aristide perceives a difference in marital attitudes as well as in work ethics: First of all, a Burundese man is different from a Tanzanian vis--vis women. Towards his wife, the Burundese man is more faithful than the Tanzanian man. It's the culture. Burundese culture expects fidelity, but here this is rare. Something which is frequent is what they call the nyumba ndogo ("small house"), when you are married . and you have a wife here, and another one across town. That's the first difference. The other one . is that the Burundese man is not lazy whereas many people here are not ambitious . He wants things, but he wants them to happen by themselves. Whereas most Congolese, Burundese and Rwandese informants perceive their male compatriots to be much more hardworking than Tanzanian men in general, only few indicate to have observed a difference in sexual and marital behavior. This is very different from the answers with regard to refugee women and Tanzanian women, the majority of which focused solely and specifically on women's sexuality. Melanie, for example, assured me that: [a]t home in Congo, when a woman marries she knows that one is my husband, when you go to the wedding everyone will be there, and everybody knows. All people know that these are husband and wife who are assumed not to love someone else outside the marriage. But here it is very different. A woman can have a husband and yet go to a party with another husband/lover from outside the marriage. Pierre feels that:

PAGE 313

297 Tanzanian women are more unfaithful, than Burundese women . When I read the newspapers, I find that women here have a bad character, they are unfaithful. Daniel, as well, sees differences between women from home and in Dar es Salaam. description of the custom included the advice to the young girl "that if one husband doesn't bring home money, at least one of the two others will." Yes, really, there are many, many differences . There are things that I have learned about here, that are unbelievable, things that I have never seen or heard but here. For example, a Tanzanian girl can easily have two husbands. And she doesn't hide it, it's done in the open . For her it's normal, it's not a secret . It's not legal in the government's eyes, but she feels it's a legal thing to do, it's normal. It is very strange to me. They even have a custom here authorizing a woman to have three husbands . I believe it is called the send-off, when the bride-to-be is given advice about marriage. Yes, yes, there are tribes that legalize this custom . they use the example of the cooking pot (la casserole). In the village, the cooking pot needs three stones to remain stable [on the fire]. So, they compare the woman to the cooking pot. "This is you, the cooking pot, and in order for you to be stable you need at least three husbands" . These are things unheard of in Rwanda. One of Daniel's compatriots explains how things are in Rwanda: Tanzanian women . wear trousers, mini skirts. It's very, very difficult to see a Hutu woman wear mini skirt and sit together with her brother-in-law or her father-in-law . So, for example, in Rwanda, I have my wife, she cannot just chat with my elder brother. No, if my elder brother is in the sitting room, she cannot come in, she will stay in the kitchen . If I need her, I will call her, she'll come and sit on her knees next to me. But here, it's not like that, it's not like that at all. The overwhelming majority of replies to the question to compare women3 across nationalities focused on the perceived promiscuity of Tanzanian women contrasting starkly with the image of the pure, faithful woman at home, one completely subservient to the husband and other male relatives. Regardless of the fact that Tanzanian informants confirmed the existence of the cooking pot ritual among Coastal people,4 it was most remarkable that underlying the emphasis on women's comportment was a preoccupation 3 Other replies will be discussed in the following section. 4 Aristide, being married to a Tanzanian woman, explains this custom as a result of increasing poverty. His

PAGE 314

298 with female sexuality in both male and female refugee informants' replies alike, whereas male sexuality was hardly mentioned. 5 See for exemple Nurrudin Farah's Maps (1986) in which "a crisis of gender and sexual identities . parrallels and intersects with the socio-political manifestations of disorder (Cobham 1992:43). The construction of the nation-state as a fraternity of men, relegating women to the non-political, private domain (see earlier) was complimented by the end of the eighteenth century by a construction of middle-class norms and sexual behavior (Parker et al. 1992). The colonial state's increasing reliance for the sake of good governance on sexual prescriptions by race, gender and class "reverberated in the metropole as they were fortified on colonial grounds" (Stoler 1989:652). Ultimately, the control of women as "national reproducers," related to their actual biological role as bearers of children, became inextricably linked with conceptions of "the nation" (Yuval-Davis 1993:628). In Africa, as well, the idea of a "gendered construction of nationhood" (Grosz-Ngat 1997:15) was increasingly accepted.5 As women's national importance is suggested to be based on their reproductive roles, they have become representatives of purity, and "only pure and modest women can re-produce the pure nation; without purity in biological reproduction the nation clearly cannot survive" (Mayer 2000:7, italics added). Consequently, "women's bodies have become microcosms of their respective countries reflecting the economic, political, cultural and religious characteristics of their homeland" (Palmer 1998:187). Women's sexuality and the discourse related to the control of women's sexuality are the daily praxis through which national identity is "maintained, communicated and understood, twenty-four hours a day, three hundred and sixty-five days a year, in a variety of different places, countries and continents" (180).

PAGE 315

299 From this vintage point, the testimonies about Congolese, Burundese and Rwandese women being more faithful and thus purer than the women of the asylum women trade, they have an eye for it. Some time ago, even the Congolese woman was leading a poor life, she dressed badly, she was living low. country, become part of a nationalist discourse subscribed to and explicated by both men and women. As Mayer indicates: "It is important to recognize that [not only men, but] women too, participate culturally in reproducing the nation, defending the "moral code" and partaking in controlling the Other" (2000:6). The overt criticism of the Other--i.e., here Tanzanians--appears to be more indicative of the refugee men and women's attempts to cope with the loss of home and all that this entails, rather than an actual judgment of the sexual, perceived promiscuous behavior of their female hosts. Yet, even as this particular strand of discourse treats both Tanzanian and refugee women as "political objects," a general sense of acknowledgment of women's agency in other areas emerged as well from the interviews. 6.1.2 Women as "Apolitical Subjects" A number of informants were quick to point out, still in response to the question to compare refugee women with Tanzanian women, that Congolese, Burundese and Rwandese women are much better in finding ways to make money than their Tanzanian sisters. At the same time, listening carefully to the voices, it appears that the situation is an ever changing one. Euphrasie, for example, in emphasizing that Tanzanian women have a lot to learn from her Congolese compatriots, recalls that it was not long ago that her women compatriots were in the same situation as their Tanzanian sisters. You know, these Tanzanian women don't know how to do business, they're learning it just now because of us, the Congolese (Kongomani). They are just waiting for their husbands to bring them something to eat. But now, they are seeing how Congolese women are doing business, and getting money with which to buy clothes . They really don't know how to do business [here], but in Congo all

PAGE 316

300 Amlie, as well, points out that in her view Burundese women are more self-sufficient, but that things appear to be slowly changing in Tanzania as well. 6 Primary School. Those that wait for their husbands to bring food home, are many here indeed . I've seen that a lot here. But, we Burundese [women], we labor hard to earn money . Then you can buy what you need without waiting for your husband. And one who succeeds in making money, will also give money. . You may buy this shirt for your husband, but here I have never seen it done. . They just wait for their husband to bring home money . Although, these days, women here are starting to wake up. Jeanne wonders whether it is education playing a role in the perceived differences. What I have seen is that Rwandese women like to develop themselves, not like here. Many women here think only of marriage, that's it. She stays at home and raises the children. And then, many haven't gone to school . They finish class 7 (darasa la saba)6 and get married. So they don't know a thing, they stay at home and raise the kids . But in Rwanda . [women] will look for ways to earn money for themselves, without depending on their husband. Contrary to the criticism of Tanzanian men by both refugee men and women, Tanzanian women apparently are not considered lazy per se. Even as the refugee women are constructing themselves as superior to Tanzanian women in work ardor, ultimately, it is the increasing levels of education that are (considered to be) at the root of their (recently acquired) level of self-sufficiency, whereas Tanzanian women are only slowly beginning to acquire more economic autonomy largely due to the deplorable state of the Tanzanian education system (see chapter 4). Aristide concedes that, in general, "because of changes in the world itself, women are getting education, employment, good salaries. Things have changed, women have stopped being dependent." However, he is quick to point out a number of additional reflections:

PAGE 317

301 There is a survey that was done that shows that these ladies forget that God has created them women . I don't know in Europe, but in the African context, [as a woman] you are married, you are a minister in the office, but when you come back you are a wife. Even if the husband is a shoe polisher, the minister had to respect contemporary political and violent conflicts as perpetrators (e.g., Moser and Clark 2001; African Rights 1995). For the purpose of this dissertation, however, I am concerned only with popular gender perceptions. him as her husband. This has brought about conflicts in families, and there is a study that has been done that [shows that] those [educated ones] who get married are likely to divorce . It is because they forget that they are women . If this woman who is respected as a woman, now she wants to be respected as a man, this can't go. She can become whatever she wants to become, minister, president, mayor, but when you are at home I will respect you as my wife, and you will respect me as your husband. No matter how small might be my status . There were some intellectual ladies here, but . they forget that they are women, and that cannot be tolerated by men, especially African men. What emerges from his account is the conviction that despite more education and growing economic autonomy resulting from increasing entrance to important positions in the private as well as the public sector, women remain subjugated to men at the decision-making level. The often heard expression "she is a woman who takes decisions like a man" illustrates the cultural norm that women should never aspire to exercise the same level of self-determination as men.7 In other words, economic independence does not translate into political autonomy, whether in the public or private domain.8 The fact that all women are ultimately considered "apolitical beings"9 was brought home to me during a conversation with a Tanzanian Immigration official. Apparently, French and Belgian aid workers had encountered serious problems in the Rwandese and 7 In this respect, most married female informants and respondents had to obtain the permission of their respective husbands in order to participate in the survey or ethnographic interviewing. 8 Hansen mentions in the introduction of her book African Encounters with Domesticity (1992) that: "African women in many countries still face an uphill battle to . reduce entrenched gender inequities in many aspects of life. The slowness with which discriminatory laws concerning marriage, inheritance, employment, and housing have been reformed bear dramatic evidence of the blurred line between the domestic and the public, turning the household into a political arena" (8). 9 This statement is in no way intended to downplay the political roles women played in pre-colonial Africa, and many historical studies have documented the detrimental influence of colonialism on the status and role of women in African society. In addition, recent studies have pointed out the active role women play in

PAGE 318

302 Burundese refugee camps in Western Tanzania a few years earlier, hence his assumption that my Belgian nationality would constitute an impediment for contacting and gaining dominated state structures and political institutions, they maintain their autonomy by avoiding state authority and participating in the unofficial (economic) sector (1989:1-19). the trust of the refugees in Dar es Salaam. Upon my assertion that this was certainly not the case, his reaction was pertinent: "Of course not, you're [just] a woman." On a second occasion, an eminent political science scholar who spent the larger part of his academic career in Tanzania, wondered if I had ever been approached by the Tanzanian intelligence officers (Usalama wa kitaifa) because of meeting with and interviewing so many refugees, some of whom were known political activists. We arrived at the same conclusion as the Immigration official, namely that as I was "just a woman" it was considered highly unlikely by the local authorities that I be involved in or associated with any type of political activity. And thirdly, a number of male refugee informants stated that they had agreed to meet with me only because I was a woman, and that they would have refused to talk to a male researcher, even as a few jokingly added: "Well, we know that some of the best spies are women, take for example this Mata Hari." Yet, even as "apolitical subjects,"10 women--Tanzanian, as well as Congolese, Burundese and Rwandese--are able to exercise agency in a number of domains as illustrated in the above interview excerpts and as will follow from the interpretation of the social network data in the next section. 6.1.3 "You are who you know" The survey regarding Tanzanian perceptions on refugees revealed that a much larger proportion of Tanzanian women than men had personal and/or regular contacts 10 Parpart and Staudt point out that because African women have limited access to and leverage in male-

PAGE 319

303 with refugees from the Great Lakes region in Dar es Salaam (see chapter 2). This gender difference in level of contact with refugees was persistent across age, religion, education I intend to bring in another more socio-culturally inspired explanation for the gender difference later in this section. level, residential area and area of origin (see Table 2-1). On average 70% of Tanzanian women stated having regular and/or personal contact with refugees, contrary to only 43% of Tanzanian men. One of the Congolese informants11 has observed the same phenomenon. He feels that: Tanzanian men are a bit grim (svre), they don't relate as well to the refugees as the [Tanzanian] women. As for women, I know that there are refugees who are married to Tanzanian women, but not the men. Why is that ? Maybe the men think that . allowing refugees to marry Tanzanian women, means they will loose their place. Tanzanian [men] don't relate to refugees easily because they are afraid that when the refugees marry their sisters, they will integrate, and this means that they themselves might loose their place, their job. But the women accept and don't really denigrate the refugees. This informant's interpretation of the reasons why Tanzanian men are more opposed to contact with refugees is most likely inspired by his personal experiences, namely his frustrations of not being able to obtain the work permit and exercise a professional activity despite his university education.12 However, his reflection that Tanzanian women are more receptive to refugees is confirmed again from the social networks analysis, just as it was by the survey administered to Tanzanian residents. As already mentioned in chapter 5, the average proportion of women among post-Dar es Salaam alters increased by 45% when compared to pre-Dar es Salaam alters, namely from 33 to 48% (Table 5-2). When disaggregating data by nationality, it appears that women actually make up the majority of Tanzanian 11 This is the same informant, married to a Tanzanian woman, that complained bitterly about the impossibility of obtaining a residence permit as a non-Tanzanian male spouse in chapter 4. 12

PAGE 320

304 post-Dar es Salaam alters: 471 women versus 435 men (Table 5-1). Even as the female majority is a small one in absolute terms, it gains in importance when contextualized and compared to other social support studies, which consistently show a predominance of men among supportive alters (see chapter 1). Looking at the gender proportions of post-Dar es Salaam alters among the other nationalities, we notice that men still make up the majority of Congolese and Burundese post-Dar es Salaam alters and around half of the Rwandese. However, when comparing these gender proportions with the ones among pre-Dar es Salaam alters, a clear trend appears. Regardless of the nationality of the alter, the proportion of women is consistently higher among post-Dar es Salaam alters than among supportive alters met before arriving in Dar es Salaam (Table 5-7). In addition, we know that it is both refugee women and men who have established an increasing number of ties with female social network members after arriving in Dar es Salaam because of the significant changes in the rate of homophilous relationships (Table 5-6). The rate of homophilous ties among women increases from 47 to 62% when comparing prewith post-Dar es Salaam alters, while the homophilous rate among men decreases from 79 to 67%. In other words, both male and female refugees count a significantly higher number of women among the social network members with whom a relationship was established after arriving in Dar es Salaam, regardless of the latter's nationality. One possible explanation for this phenomenon could be the increased economic or financial capacity of the refugees' female post-Dar es Salaam alters. Since this is the reason as to why men have been found to provide more support than women in other social support studies, it would be a logical explanation. However, a closer look at the

PAGE 321

305 types of support provided by male and female preand post Dar es Salaam alters respectively demonstrates that this is not the case (Table 5-11). While 29% of the female pre-Dar es Salaam alters provide the refugees with financial support, this proportion decreases to 11% among post-Dar es Salaam alters. Both among preand post-Dar es Salaam alters, men remain the more important, albeit decreasing, source of financial support, from 37 to 19%. I argue that it is the popular perception of the woman as an "apolitical being" that in the situation of forced migration has facilitated an increased contact between refugees and women (particularly Tanzanian but also of any of the refugee nationalities under study), who ultimately became members of the refugees' social networks in Dar es Salaam. In view of the fact that (often national) politics are at the basis of a refugee's decision to leave home and seek asylum, it is not surprising that politics and political perceptions continue to play an important role in his/her life in exile.13 Whether it regards Tanzanian officials, compatriots or fellow refugees, men are considered to be more politicized than women. One example in this regard, is Daniel's remark in chapter 3, where he states that if he were a woman, he would not be afraid to go home to Rwanda and face the settling of scores. Yet being a man, he feels he will automatically be suspected of having actively participated in the genocide and be killed.14 Men are generally considered to be the political representatives of their respective nationalities or political affiliations, and are therefore much more likely to raise suspicion among persons 13 See also chapter 4 for examples on the ways in which the urban refugees for a variety of reasons and depending on their capabilities pretend to be of another nationality than their own. 14 Despite the fact that a significant number of Rwandese women were found to be actively involved in the killings of the genocide (African Rights 1995).

PAGE 322

306 of other nationalities or opposite political affiliations. This situation impedes the easy establishment of new relationships between men, for lack of its basic ingredient, "trust." Whereas the presented data do not include specific information on the strength of the ties ego enjoys, studies have shown that strong ties tend to occur among persons with similar attributes, and in addition, strong homophilous ties have been proven to be "more effective channels of social support . in crisis situations" (Beggs et al. 1996: 217). Ultimately, the fact that women's social networks consist of an increased number of strong, homophilous ties in the situation of forced migration decreases their level of dependency on men, whether with regard to material or immaterial support. At the same time, the decreasing number of strong, homophilous ties of refugee men attributes to their increasing level of dependence of women when compared to the pre-flight situation. It is as Turshen and Twagiramariya (1998) point out: War also destroys the patriarchal structures of society that confine and degrade women. In the very breakdown of morals, traditions, customs, and community, war also opens up and creates new beginnings . Paradoxically, war has opened up new spaces for women (20). In the recomposition of their social networks during exile in Dar es Salaam, refugee women demonstrate an increased level of independence from men, whereas refugee men experience an increased level of dependence on the support provided by female network members. The new situation does not constitute a reversal of roles--men continue to make up the majority of supportive alters--however it does entail a reconfiguration of gender in that the power dynamics and relations between men and women are affected (Hodgson and McCurdy 2001). One Congolese woman describes the change in the relationship with her husband: [At home] I was doing my own little things, . but here we have to help each other . I need to put in my efforts so as to help him, so as to pay for the house, food

PAGE 323

307 and the children's school fees. So these days our relationship is one where we help each other, whereas before it was him who did everything, in Congo. From the above findings, based on both qualitative and quantitative data, we find that historical events and altered material circumstances influence the construction of gender, but always within the limits of people's perceptual frameworks. It is because they are considered "apolitical beings" in the contemporary African cultural context that the refugee women have been able to gain a certain level of independence from men, unknown before the flight from home. Whether the social and cultural change among Dar es Salaam's urban refugees brought on by the situation of forced migration will persist and bring about a permanent change in gender perceptions is impossible to say at this moment in time. In view of the fact that refugee situations are, in principle, of a transient nature and that the urban refugees do not form a community, neither in their place of exile nor at home (after the repatriation), it is safe to assume that a major part of the process of reformulating cultural gender constructs permanently will depend on the effects of the war (crisis) on the local gender dynamics and perceptions in the refugees' respective countries of origin. 6.2 Nationality Benedict Anderson defined the nation as an imagined community "because the members of even the smallest nation will never know most of their fellow-members" (1991:6). Feelings of national identity then are the result of nation-building policies based on the "instilling of nationalist ideology through the mass media, the educational system, administrative regulations, and so forth" (114). In addition, Smith (1986) documented that national ties are also stimulated through the (re)invention of historic memories, myths, symbols and ceremonies reflecting a common heritage. In African

PAGE 324

308 countries as well, nation building policies were part and parcel of the post-independence governments, with projects such as e.g., Mobutu's Retour l'autenticit (Zairianization) 15 Allowing for degrees in variation due to class, gender and ethnic differences. and Nyerere's implementation of Swahili as a national language, while both Rwanda and Burundi had been historical entities as centralized kingdoms for centuries (see chapter 2). These nationalist projects allowed "the idea of a cultural community to emerge" (Palmer 1998:180, italics added) in which "an historically transmitted pattern of meanings [is] embodied in symbols, a system of inherited conceptions expressed in symbolic forms, by means of which men communicate, perpetuate, and develop their knowledge about and attitudes to life" (Geertz 1973:89). As social constructs, cultural conceptions are thus embedded in the historical, economic and political structures of the cultural community. Based on this definition, we may postulate that nationals of each of the three refugee-producing countries under study dispose of similar patterns of meanings and conceptions about life on the basis of being a member of their respective nations of origin, where they shared a common history and dealt with the same economic, political and social structures and infrastructures as their fellow country men and women.15 Culture thus establishes a "more or less homogenous set of beliefs and assumptions by means of which everyone can project their perceptions and expectations onto other people without thinking about it" (An-Na'im and Hammond 2002:22). The question emerges as to what happens when persons of a particular cultural community (here of the nation) find themselves outside of their usual, national surroundings, such as in a situation of forced migration? Swindler (1986) proposes an "image of culture as a "tool

PAGE 325

309 kit" of symbols, stories, rituals, and world-views, which people may use in varying configurations to solve different kinds of problems" (273). She postulates that People do not readily take advantage of new structural opportunities which would require them to abandon established ways of life. This is not because they cling to cultural values, but because they are reluctant to abandon familiar strategies of action for which they have the cultural equipment (281). In other words, the Congolese, Burundese and Rwandese refugees who arrived in Dar es Salaam fairly recently, are likely to deploy strategies, similar or identical to the ones that ensured survival at home and whose outcomes were considered predictable. As such, I propose that the agency displayed by the refugees under study in confronting the daily quest for survival in exile is culturally defined by the historical, political and economic background of their respective home countries. When interviewing staff members from humanitarian organizations as well as Tanzanian government agencies who are in daily contact with refugees of many nationalities, I was struck by the recurrent images of each of the three refugee nationalities under study. While there exists a general stereotypical image of "the" refugee as a poor and uneducated person, these Tanzanian informants expressed distinct perceptions of the respective socio-cultural ways in which nationals from Congo, Burundi and Rwanda display agency. One Tanzanian UNHCR staff member puts it as follows: You can easily notice differences . in the way people present their problems. I would rank the Burundese as the least provocative . I find it a lot easier dealing with Burundese generally, followed by Rwandese. And for Congolese . I do feel that when a Congolese wants to pursue something, they will do it, no matter how difficult . Burundese, at times, I feel they should be more assertive. The Rwandese are more assertive, but not as pushy as the Congolese. An Umati counselor, who deals mostly with Congolese and Burundese refugees, shares the same opinion:

PAGE 326

310 The problem I have noted especially among Congolese is that they are really harsh [sic] . The Burundese, they are kind, they don't want quarrels . They accept what you explain to them. But for the Congo people, you can explain to them but they will refuse completely, and say "This is my right, and you have to do this" . This principle is compatible with Romer's rule, namely that people usually wish to change just enough to maintain what they have. So that is the difference. I noted many observations similar to the ones above from persons in daily contact with the refugees from the Great Lakes region--and even from the refugees themselves--all acknowledging pertinent differences in the cultural expression of the refugees' display of agency. It is my intention to link elements of the respective larger national orders (structure) with these cultural differences emerging from individuals' actions (agency) and interactions (social networks).16 The following sections focus on three domains of the larger national orders, namely economic organization, political infrastructure and historical background. Together they provide a framework for embedding the various strategies used by the Congolese, Burundese and Rwandese refugees in Dar es Salaam to cope with the situation of forced migration as well as the ways in which they re-construct their social networks, as outlined in the previous chapters. The underlying principle is that when people have acquired the behaviors and norms necessary for survival in one particular context, they will (initially) continue to display the same pattern of behavior in a different context in the conviction that it will lead them to the same predictable outcomes as before.17 16 See also chapter 1. 17

PAGE 327

311 6.2.1 Economics When asked if there was a difference in the coping strategies of her compatriots the Arab Emirates continued to flourish in the absence of Zairean state structures, thanks to a decreased level of state control in the form of taxes, import and export restrictions, etceteras (Mwanasali 2000). compared with those of Congolese fellow-refugees, Christine's Burundese lady-friend observed the following: There are many Congolese here [in Dar es Salaam], but there are many Burundese in the camps because Burundese fail to succeed here, unlike the Congolese, who know how to fend for themselves. They play music, open up a saloon or a tailor shop, or do business, more than the Burundese. The Burundese mind isn't as probing as the Congolese . The Burundese . they don't know how to pursue. And many of the Congolese have studied, they know that if you have a good education (mali za shule), you will know how to make money instead of waiting for a job from the government. For a better understanding as to why Congolese appear, to this and other informants, to display more initiative in coping with the situation of forced migration in Dar es Salaam than Burundese refugees, the above statement will be contextualized in the economic structures and organization of the two countries, Congo and Burundi. From the early 1970s onward, the Zairian central government in Kinshasa had done nothing in the way of public spending in its Eastern provinces, especially the Kivu region. The rebellions of the 1960s (see chapter 2) had made president Mobutu averse to even build roads in the region because they "would only be used by rebellious soldiers to drive their lorries to Kinshasa and carry out a coup" (Prunier 2001:151). The Zairian state all but completely withdrew from the region, leaving the Kivusiens to "fend for themselves" (se dbrouiller). However, informal trade networks from the Eastern part of Congo to the markets in Eastern and Southern Africa, operating "on the periphery of the existing state structures"18 (Newbury D. 1986: 90), ensured the region of a steady source 18 In fact, the historical trade networks in the Great Lakes region reaching out to Kenya and Tanzania, even

PAGE 328

312 of foreign exchange through the smuggling of diamond, gold, coffee and ivory. The fortunate presence of vast deposits of strategic natural resources for which there was a high demand on the world market (Mwanasali 2000) provided sufficient capital to import vehicles, spare parts and fuel, consumer goods, medical products and consumer materials to the Kivu region (MacGaffey 1992). Individual initiatives not only reinvigorated a state-less economy, they also have been dealing with the complete breakdown in public health services and public education through private projects. As such "health care and adequate education, otherwise almost non-existent, [are] available for those who can find the money" (253). This state-less system is necessarily based on personal ties: "Traders, petty producers, truckers, and retailers have elaborated ties of clientage into networks, extending sometimes over enormous distances, to organize unofficial systems of distribution and marketing that, in large measure, provide the food supply for cities and towns" (254). However, as MacGaffey points out: "the second economy certainly does not bring equity. Access to the necessary resources for participation is extremely uneven and intensely competitive" (256). At the same time, having defeated the state through popular initiatives, the Congolese people remain with little respect for state structures or its attempts to implement legislation and "widespread evasion and disregard for the law" is rampant (247). The situation in Burundi is very different. Ngaruko and Nkurunziza (2000) argue that in Burundi, the absence of strategic natural resources--which are abundant in neighboring Congo--is the root cause of the "predation" of the state's resources by an

PAGE 329

313 elite. For these individuals,19 "the control of the public sector as a source of accumulation, the army as the guarantor of this control and the education sector as a and social activity" and "where the president and other powerful actors use offices and material rewards to build a dominant coalition of supporters (Newbury 1992: 199). means of accessing it are crucial factors" (384). Not only does "public employment represent 80% of full-time employment in the modern sector" (388), in addition, wages are relatively high when compared to the average Burundese salary. While the agricultural sector, the backbone of the nation and source of income to the majority of the Burundese people, particularly Hutu, hardly receives much needed financial government resources, "almost all large private firms belong to former high-ranking civil servants" (387). From the above, we gather that the economic vacuum left by the Congolese state ultimately gave rise to a booming second economy20 built on the existence of informal trade networks, the presence of large deposits of strategic natural resources, the "ingenuity of local entrepreneurs" (Mwanasali 2000:140) and the "nurturing of social networks for mutual support" (Newbury C. 1986: 100). In Burundi, on the other hand, there was no economic space in which individual or private initiatives could develop or flourish due to the suffocating hold on the economy of a bureaucratic elite bent on keeping total control of the state at all costs. Both situations have been in existence for decades and have given rise to different cultural "tool-kits" among Congolese versus Burundese (or Rwandese21) nationals. Whereas the average Congolese in Eastern Congo 19 In Burundi, overwelmingly Tutsi. 20 Following C. Newbury (1986), defined as extralegal or legal forms of economic production and exchange which do not enter into national statistics. 21 The situation in Rwanda being similar to the one in Burundi, its state system has been described by analysts as one where "those in power try to control and direct from above many (or all) aspects of political

PAGE 330

314 was never able to rely on formal government structures, s/he developed the art of nurturing strong, large personal social networks that function as safety nets in case of 22 The Daily News, March 14, 2001, p.1. Refugees in push to flee 'starvation' by Pudengiana Temba. unpredictable or crisis situations. Yet, because of the intense competition and the lack of formal regulations and laws guaranteeing one's rights as a citizen, s/he also had to acquire a level of assertiveness that would ensure access through the necessary personal ties to a limited supply of resources. A generally high level of education (Chapter 3) completes the profile of the Congolese refugee in Tanzania as one who is assertive, eloquent, and vocal. One illustration of this high level of Congolese assertiveness and sense of initiative comes from a newspaper article describing their reaction to a twofold cut of the food rations (by 40% in July 2000, and by another 20% in November 2000) distributed in the refugee camps in Western Tanzania. Sizable numbers of Congolese refugees want to return to the Democratic Republic of Congo (DRC) because of meager food rations . Early March, 5,000 refugees were found trekking to Kigoma town, claiming they want to go back to the DRC in the face of what they termed as "insufficient food rations."22 The food rations had been cut in all the refugee camps, but neither Burundese or Rwandese refugees ever staged any form of public protest, whereas a large number of Congolese, five thousands persons, left the camp without permission and headed for the provincial capital on their own initiative in an effort to better their situation. The rebellious nature of the Congolese vis--vis any type of state structures comes out equally clearly in one Congolese informant's disdain for Tanzanian civil obedience: You know, these [Tanzanians] are people who don't know their rights . like people who don't have this notion of defending their rights. Everything their government says, it is as if it was God who spoke.

PAGE 331

315 It is in the same spirit that Congolese are significantly more vocal than either Burundese or Rwandese respondents (14 versus 0 and 4% respectively), in stating that their "rights as refugees are violated" in Tanzania (see Table 3-7). One Congolese informant, who was a human rights activist during the Mobutu era, and has lived in Tanzania since 1995, compares the human rights situation in his home country with the one in his country of asylum: The difference is that in my home country from independence till now, the Congolese population continues to live under a dictatorial regime. There is no respect for human rights, a situation that leads to the never ending wars and troubles. But here, there is only an apparent calmness because in reality there are also many human right violations, especially for us refugees, we have become the objects of exploitation by UNHCR as well as by the Tanzanian authorities. Both the Human and Legal Rights Center and the Dar es Salaam branch of Amnesty International (see chapter 2) equally confirm receiving many more Congolese than either Rwandese or Burundese with complaints of "unjust treatment" by formal structures, such as, the UNHCR or the Refugee Department at the Tanzanian Ministry of Home Affairs. The importance that the nurturing of extensive social networks plays in Eastern Congolese survival strategies in Dar es Salaam has emerged on a number of occasions. While one in six Congolese stated avoiding the refugee camps because "it prevents being in contact with others," none of the refugees of the other two nationalities mentioned this as a reason (Chapter 3). We also saw in chapter 3 that significantly fewer Congolese spent any time in one of the refugee camps, preferring instead to head straight for Dar es Salaam with the objective of joining friends and relatives (33% of Congolese versus 20 and 19% of Burundese and Rwandese respectively). Upon arrival in Dar es Salaam, Congolese usually do not hide their efforts to meet compatriots, contrary to either Burundese or Rwandese and they will more often have as their closest friends in Dar es

PAGE 332

316 Salaam other Congolese men and women rather than Tanzanians (Chapter 4). In addition, there appears to be a much higher tendency among Congolese refugees to form associations, albeit on the basis of ethnic affiliations (Chapter 4). All of these findings point to the fact that Congolese refugees in Dar es Salaam are embedded in much larger, stronger and more extensive networks than either Burundese and Rwandese, who seek assistance and support from formal structures in significantly larger numbers than the former (Chapter 5). It is also among the Congolese that the increase in the proportion of women among post-Dar es Salaam social network members is by far the lowest (Table 5-7), suggesting that the situation of forced migration does not affect Congolese refugees' social networking strategies to the same extent as it does those of the other two nationalities. It is the same knack for establishing social relationships despite life in exile, that explains why significantly more Congolese appear as social network members of Burundese and Rwandese refugees than vice versa (Chapter 5). This Congolese aptitude to connect with people is equally reflected in the Tanzanian perceptions, albeit stereotypes, of Congolese as "mixing easily, friendly and generous" (Chapter 2). However, the other side of the coin, describing Congolese as "cheating, conmen and liars" may well be interpreted as a corollary of an inventiveness and industrious creativity gone awry in the desperate economic endeavors of life in exile. The state-led, bureaucratic societies in both Burundi and Rwanda, on the other hand, gave rise to an entirely different national profile, namely one that is less contesting than that of the Congolese, or in the words of the Tanzanian informants quoted at the beginning of this section, less "pushy" or "harsh." At the same time, these Tanzanian informants expressed having very different experiences with Burundese as compared to

PAGE 333

317 Rwandese refugees, despite the fact that both are Hutu, victims of ethnic wars, share the same culture and language, and lived under similar government bureaucracies. The answer may perhaps be found in the different political histories of the two countries and its respective Hutu populations. 6.2.2 Politics While Rwanda, from 1962 onwards, was ruled by a Hutu elite, Burundi continued to be governed by the Tutsi in the post-colonial era (Chapter 2). These political developments lay at the basis of the out-fluxes of Tutsi refugees from Rwanda in the 1960s and 1970s, while hundreds of thousands of Burundese Hutu sought refuge abroad from the 1970s onwards. Often both countries were popularly identified by the ethnic group dominating the government. When I asked Jeanne, for example, if there were differences between the Burundese and the Rwandese, she replied that: In my view, a Burundese and a Rwandese are different. Because you can consider them as the countries of the tribes of the Tutsi and the Hutu [respectively]. As described in the previous section, the respective government bureaucracies were the major economic actor in both countries given that the public sector was the main source of accumulation for lack of significant deposits of strategic natural resources. In order to safeguard their position, both regimes applied a policy of exclusion from power for members of the other ethnic group, for example, through limiting their access to education. In Burundi, where a Tutsi minority felt threatened by the Hutu majority, this policy of exclusion occasionally took the form of selective genocides, e.g., "during the 1972 program, most Hutu with 4 years or more of high school education were either killed or forced to leave the country" (Ngaruko and Nkurunziza 2000: 390). Daniel remembers that:

PAGE 334

318 because with their Tutsi government, it wasn't easy for the [Burundese] Hutu to develop (merger). And many of their big men, I mean persons who have studied, Hutu professors and such, became victims. They [the Tutsi regime] killed many of them [Hutu intellectuals], others fled, while those who stayed weren't given a In fact, it is precisely their membership to the pre-1994 Hutu elite that has rendered them suspect in the eyes of the current Rwandese regime, who accuse them of being part of masterminding the 1994 genocide. chance to continue, that's how it was. Even me, in 1998 when I passed through Bujumbura, I visited their university, l'universit libre du Burundi (ULB), you'll see that the majority of students are Tutsi. Despite the fact that they constitute the minority, there were more Tutsi students. So I think that discouraged the Hutu, others joined the rebels, in an attempt to let off steam maybe . That's why the Burundese Hutu couldn't develop. It is striking, with respect to the refugee populations in Dar es Salaam, that even as Burundese urban refugees enjoy a much higher education level than their compatriots residing in the refugee camps, they still count only half as many secondary and post-secondary graduates when compared to the Rwandese urban refugees (Chapter 3). Yet, being a member of the government elite23 did not just translate into higher education levels, whether in Rwanda or Burundi, it also opened up new possibilities beyond the national borders. Daniel, whose father was a prominent member of the Habyarimana government in Rwanda, recalls that: The president gave us many opportunities to study, for example at the university of Butare, there were many students . we really got this chance to study, to emerge and develop, we were given many opportunities. Twice a year, we could apply for scholarships to go study abroad, and many left for Belgium, France, Canada, and such . Yes, we had many opportunities, we were not isolated (enferm). The larger part of the current Rwandese refugees in Dar es Salaam belonged to Rwanda's governmental Hutu elite until 1994, when the Tutsi-dominated RPF took over.24 Some have been on the run since 1994, others returned to Rwanda in 1996, but later decided to 23 In view of the fact that the Hutu constitute the majority of the Rwandese population (85%), only a minority of them could be members of the government elite. 24

PAGE 335

319 leave their homes again in light of the increase of arbitrary arrests and settlement of scores (Chapters 2 and 4). to a situation whereby "most of the Barundi refugee families living in Tanzania have non-refugee relations in Kigoma region" (1993:20). From the above, the most important difference between the Burundese Hutu and Rwandese Hutu refugees appears to be their societal position of the past decades due to different political structures in the two countries. It not only explains the significant difference in the respective education levels (Table 3-2), but also in coping strategies. As former members of a governmental elite, the Rwandese refugees dispose of financial means to a much larger extent than the two other nationalities. In fact, one in seven Rwandese quoted financial means as their main reason for not residing in the refugee camps (Chapter 3), while they also reported receiving financial support from a significantly higher proportion of their pre-Dar es Salaam alters (Table 5-9). In addition, almost half of all social network members of the Rwandese nationality are in a position to provide the urban refugees with financial support, significantly more than any of the other nationalities (Table 5-11). In comparison to the Congolese and Rwandese, the Burundese refugees' social networks consist of a relatively higher proportion of network members residing in Tanzania (but not Dar es Salaam) and the immediate region (e.g., Zambia, Malawi and South Africa) rather than in the home country, because of their long history as a people in exile in the region.25 Even allowing for the fact that increasing levels of education will induce geographically more expanded networks (Fischer 1978), it appears that the geographical extent of the Rwandese supportive networks exceeds that of both the 25 At the same time, as Daley notes, migrations during the colonial period from Burundi to Tanganyika, led

PAGE 336

320 Burundese and the Congolese refugees' networks (Table 5-9). This is the result of the advantages the Rwandese refugees enjoyed as a political elite which allowed them to 26 Sommers developed the concept of "cultural fear" to indicate the extent to which Burundese refugees are instilled with fear during their daily activities in Dar es Salaam and its historical and cultural roots (1994). develop networks beyond the continent (e.g., Belgium, France, Canada, etc.). Being more cosmopolitan, they also aspire to a significantly larger extent than the Burundese refugees to resettle in a third country (Table 3-7). Being financially less well off and internationally less well connected than their Rwandese brothers and sisters is only half the story of the Burundese, however. The observations of the Tanzanian aid workers confirm the image of a Rwandese who is more self-assured and assertive than the Burundese, but not as pursuing or tenacious as the Congolese. The Burundese, on the other hand, according to one aid worker: if you say "no" [to their request for assistance], they will accept. They will understand you. But, at times, I think they need to make some more effort, try again. This image of the Burundese as a docile, confrontation-evasive person was a recurrent one among informants. In view of the fact that Burundese Hutu have known only suppression and discriminatory measures from their own government during the past four decades, it is not surprising that the Burundese refugees apply a strategy of appearing very submissive to official structures of any kind, be they Tanzanian government offices or humanitarian aid agencies. Having undergone many (usually deadly) military state reprisals in their lifetime, they are understandably more afraid than the other two nationalities of being denounced to state authorities by their Tanzanian neighbors, and of being discovered by the polisi wa upelelezi or Tanzanian immigration officials as Pierre, Christine and many other Burundese informants testified (Chapter 3).26 Keeping a low

PAGE 337

321 profile and not attracting the attention of the omni-present state apparatus (see previous section) was for decades part and parcel of the cultural "tool-kit" of Burundese Hutu, embedded in the political realities and praxis of every-day life. It was an essential survival strategy under a governmental regime, where merely being educated or being politically active sufficed for a Hutu to be regarded as a threat to the existing political order, hence jeopardizing the life of oneself and one's relatives, as illustrated at length by Pierre's testimony (Chapter 3). The above context may also provide us with some clues as to the image of the Burundese as "hardworking" yet "aggressive" circulating among Dar es Salaam's Tanzanian residents (see chapter 2). Even Aristide, who is quite well integrated in Tanzanian society feels that: From a general, cultural point of view, Burundese are not lazy, contrary to many people here [in Dar es Salaam] As persons who were prevented from obtaining higher education levels, most Burundese Hutu, at home, derived an income from agricultural activities or other projects related to manual labor. In addition, we learned that Burundese do not enjoy the same financial means as the Rwandese refugees, nor the extensive, supportive networks of the Congolese. When left with no other option to provide for one's daily needs, the Burundese, in general, will be likely to agree to heavy manual labor, at for Tanzanians, unacceptably low salaries to a much larger extent than the higher educated and well-connected Rwandese and Congolese. Aggression, on the other hand, especially at the domestic level, could possibly be interpreted as a way to vent the frustrations the Burundese urban refugees feel as a result of their continued exploitation, both at home and in exile. However, this issue needs

PAGE 338

322 further research because, apart from the survey among Tanzanian citizens (Chapter 2), I was unable to find any confirmation, much less negation, of this image from any of my through the success of the Zairean music and songs. Until the fall of Mobutu in 1997, it was often considered the language of state agents and the military. informants, regardless of nationality. Having contextualized the image of the Congolese and Burundese refugees using the economic and political national frameworks respectively, there still remains the image of the Rwandese which emerged from the Tanzanian survey as one who "is proud and doesn't mix with others" (Chapter 2), as well as a few unanswered questions from the previous chapters. I believe the answers may be found in the history of the Great Lakes region. 6.2.3 History As mentioned in chapter 3, Bonaventure was instrumental in locating informants, not only Congolese, who are rather easily found and contacted, but Burundese and Rwandese as well. When I asked him to explain how he managed to establish contacts, especially with the Rwandese, he explained that: Us, Kivusiens, we are friends with the Rwandese, we know their language because we received the Rwandese refugees [1994 till 1996] and they taught us their language. In a sense, I didn't appreciate that they did, but these are very proud (orgeilleux) people. When leaving their country, they never wanted to learn the [Swahili] dialect spoken in Congo, they continued to speak their own language, Kinyarwanda . So it is the Congolese who adapted and learned to speak Kinyarwanda, but they, they categorically stayed within their language. That's why you'll find certain Congolese who know Kinyarwanda, because the Rwandese lived with us, but you will rarely find a Rwandese that knows Lingala.27 Swahili, maybe those who have lived here in Tanzania. What is particularly striking in Bonaventure's testimony, is more than the issue of language. The account suggest a very high level of hospitality and flexibility on the part 27 One of the four official languages of the Democratic Republic of Congo, one whose use was popularized

PAGE 339

323 of the Congolese in the Kivu region, to be prepared to learn the language of the Rwandese refugees, in stark contrast to the unwillingness of the latter to express the 28 Swahili, the language of the East African coast was born out of the contact between Arab merchants and the African Bantu populations (Mazrui and Shariff 1994). minimum of courtesies to their hosts, i.e., learning the language. In my view, the "open-ness" of the Congolese Kivu region as well as the "closed-ness" of the Rwandese society find their roots in the precolonial era and, more precisely, at the Tanzanian Swahili Coast. From the late eighteenth century onward an East African commercial empire began to emerge based on the trade in spices, ivory and slaves, and whose nexus was located in Zanzibar, an island off the East African Swahili Coast (Sheriff 1987). The East African coast had been a site of international trade for almost two thousands years,28 but it was the choice of Zanzibar as the capital of the Omani sultanate together with the establishment of a plantation economy on the island that gave an extra impetus to the expanding ivory and slave trade in the mid-nineteenth century (Ibid.: 8-32). Connected to the coast by several long-distance trade routes, the African hinterland was gradually integrated into the international trading system (Ajayi 1990:89). Through four major trade routes, Swahili, as the language of the Arab and African traders from the coast, spread deeply into the continent. It became the lingua franca of a region reaching from modern-day Kenya in the North to Northern Mozambique in the South and Eastern Congo in the West, facilitating contacts with people from outside, as well as among the different ethnic groups in the regions, except for Rwanda and Burundi. Whereas some of the large kingdoms in the Great Lakes region, such as Buganda and Bunyoro in present-day Uganda (see chapter 2), were greatly strengthened by their association with the coastal traders by accumulating firearms and raiding their neighbors,

PAGE 340

324 the kingdoms of Rwanda and Burundi, on the other hand, had limited dealings with the long-distance trade and "were actually unwilling to welcome foreigners in their 29 Catherine (1994) describes that the region was actually under Arab dominion by then, and that it took the Belgians roughly ten years (1885-1894) to take over control of the region from the Arabs. communities" (Ajayi 1990: 91-95). They held out against the foreign trade and successfully withstood the aggression of their neighbors, Buganda and Bunyoro, by now well equipped with firearms from the long distance trade. Rwanda, in addition, placed an "impediment on the free movement of trade into Rwanda" (Sheriff 1987:184). The map by Sheriff (1987:191) of The Hinterland of Zanzibar clearly shows its boundaries excluding Rwanda and Burundi, while large parts of current-day Uganda and most of Eastern Congo, up to the town of Kisangani, were included in this East African commercial empire. On the eve of the colonial era, Swahili had become the lingua franca of Eastern Congo, which opened up the region not only to Arab and Swahili traders,30 but also to missionaries and other foreigners. Burundi and Rwanda, on the other hand, were more or less isolated kingdoms speaking only their own language and purposely not participating in the East African trade routes and its links to the international world. When Ruanda-Urundi, as it was then called, was placed under Belgian administration by the League of Nations after World War I, Bujumbura became its capital (Gahama 2001: 51). The mandate under which Ruanda-Urundi was to be ruled stipulated clearly that no military were to be recruited from its population (Ibid.: 37-45). As such, the Belgian colonial administration had no alternative but to station Congolese troops in Bujumbura, which 29 Even the most famous of all Arab slave traders, Tippu Tip, never succeeded in reaching Rwanda (Verlinden 1995:32). 30

PAGE 341

325 remained under direct orders of the colonial authorities of Congo's Eastern Province. For reasons of convenience,31 the Belgians opted for Swahili as the language of the army, Interestingly, the book by Jean Rumiya (1992) mentions nowhere that the actual capital of Ruanda-Urundi was located in current-day Burundi. administration and colonial labor force. In addition, the Belgian administration encouraged the immigration of Swahili agriculturists in certain parts of Burundi, where they contributed substantially to the cultivation of rice crops destined for the Belgian Congo (Nyambariza 1986:75). As a result of all these colonial arrangements, there was increased contact between the Burundese and the Congolese in Burundi, using Swahili as their medium, whereas Rwanda remained more or less on the fringes of the Belgian colonial system.32 In the early years of independence, many of the Congolese rebels (see chapter 2) after being defeated by Mobutu, sought refuge in Burundi. They integrated rather well, and when the Burundese Hutu revolted against the Tutsi regime in 1972, they could count on military assistance from these Congolese refugees (Reyntjens 1996), because as one Burundese Hutu informant put it: The Congolese like freedom (aiment la libert). So when they saw the oppression, they decided to help their Burundese brothers who had received them so well. In contrast, the contacts between Rwanda and Congo were limited to migrations from Rwanda into Congo, both in the nineteenth as well as the twentieth century (see chapter 2). There were never large movements of Congolese into Rwanda, leaving the country into its somewhat isolated status quo, contrary to the Burundese who throughout the 31 See also Fabian's Language and Power: The appropriation of Swahili in the Former Belgian Congo 1880-1938 (1986). 32Le Rwanda sous le rgime du mandat belge (1916-1931)

PAGE 342

326 colonial and postcolonial period were exposed to an increasing level of contact with their open-minded Congolese neighbors as well as other foreigners. Rwandese society was recurrently described by informants as "secretive" and "dominated by suspicion." One Congolese woman, in a conversation with a Rwandese informant recalls how: in 1994, when the Rwandese refugees came [to Congo], that's when we learned that . for example in a bar, you're having a drink, you can not leave your drink on the table to go to the bathroom or so . So it is because of the refugees that we learned no longer to use glasses . even going to the dance floor, you take your bottle with you . It is said that a stranger won't harm you, so necessarily it has to be someone you know [who may poison your drink]. The Rwandese informant confirms this and says to feel much safer in Tanzania than at home, because: here I can leave my glass [unattended], I go to the bathroom, I come back, I'm not afraid [that someone poisoned my drink]. The general lack of trust of Rwandese may be related to a history of isolation and "closed-ness" of their society, which is exemplified in certain living patterns and socio-cultural customs. For example, instead of living in villages, Rwandese live in high walled compounds surrounded by their fields. One's closest neighbors often live a couple of hundred meters away, even as the country is one of the most densely populated in the world. I also learned from informants that the Rwandese eat in a room in the back of the house, so that when unexpected visitors arrive, they may wait for the family to finish eating instead of being invited to share the meal, as is customary among Congolese, Burundese as well as Tanzanians. From the above, emerges the profile of the Rwandese as a people who prefer to keep their distance from others, and whose lives are ruled by suspicion. For historical reasons, Rwandese were the least exposed to Swahili of all three refugee nationalities and

PAGE 343

327 thus the ones struggling hardest with learning the language as documented by informants' testimonies (Chapter 4). Significantly larger numbers resided in the Tanzanian refugee camps for a number of years before coming to Dar es Salaam (Table 3-5) not only because they are not familiar with Swahili, but also for lack of necessary contacts which the Congolese enjoy through extensive networking and the Burundese because of their long history as exiles in Tanzania. The fact that Rwandese count considerably fewer compatriots among their post-Dar es Salaam alters than either Congolese or Burundese (Table 5-6) already points in the direction of less supportive networks due to the decreased number of homophilous relationships in the situation of forced migration. At the same time, the increase in proportion of female network members among Rwandese refugees' post-Dar es Salaam alters is double that of either Congolese or Burundese (Table 5-7) testifying to the important role (political) suspicions or distrust play in establishing new relationships with men from any nationality in exile (see earlier in this chapter). Both elements undoubtedly contribute to the fact that Rwandese also appear to be the least likely to form supportive associations in Dar es Salaam (Chapter 4). The social network literature suggest that it is the lack of supportive and strong networks that cause significantly more Rwandese to ultimately seek assistance from humanitarian agencies (Table 3-5), regardless of the fact that they appear to dispose of relatively more financial means than either Congolese or Burundese (see previous section). The establishing of new, trusting and supportive relationships during the situation of forced migration is thus hampered not only by an insufficient knowledge of Swahili, but in large part also because Rwandese tend to regard strangers with suspicion, whether

PAGE 344

328 at home or in exile, because it is an attitude based on a world-view that belongs to their cultural "tool-kit" embedded in historical events. It goes without saying that the traumatic events of 1994 have reinforced even further the feelings of suspicion now directed no longer only toward strangers, but toward compatriots as well, whether Hutu or Tutsi. A few remaining findings may now be contextualized in the broader socio-cultural frameworks of the respective nationalities: namely, that more Congolese reported perceiving Tanzanians as hypocrites, more Burundese regard Tanzanians as suspicious, while more Rwandese blamed cultural differences as the roots of the difficulties in establishing relationships with Dar es Salaam's Tanzanian residents (Chapter 3). From the above, it emerges that Rwandese prefer to quote cross-cultural differences as an insurmountable impediment to establishing relationships with the Tanzanian population, creating such a relational distance that it cannot be easily bridged by simply knowing the local language. Burundese refugees, on the other hand, having undergone years of suppression from their own government and national army, are most bothered by the inquisitive nature of Tanzanian nationals, perceived as bent on denouncing them to the authorities and instilling them with fear. The Congolese, on the other hand, are perceived by all informants, regardless of nationality as extremely hospitable and open-minded. Tanzanians, as well, are renowned for their hospitality (see chapter 2 and 3). However, because of a high level of poverty, they are not always able or willing to conform to societal norms. What remains is the formal, spoken invitation to visitors to share a meal or a drink (commonly called karibu chakula), but which is rarely acted upon by actually adding another plate to the table or pulling up a chair. This is something the Congolese

PAGE 345

329 find very hard to accept, and is interpreted by the majority of informants as part of the "Tanzanian hypocritical nature" rather than the unpleasant result of an impoverished state of affairs. The above analyses of the strategies applied to the situation of forced migration by the Congolese, Burundese and Rwandese refugees have been contextualized in their respective national historical, political and economic frameworks through the concept of a cultural repertoire of habits, skills and styles or "tool-kit" (Swidler 1986). The contents of individuals' cultural "tool-kits" are embedded in their ideal as well as material worlds at the macro-level, and the choice they make when constructing certain lines of action in a particular situation at the micro-level is based on the perceived continuity of these strategies' results. It is when familiar strategies no longer generate predictable outcomes that efforts will be made to choose an alternative line of action but "only to the extent that it is sufficiently consistent with the basic qualities of the culture in question as to become part of the shared and unquestioned repertory of practice" (An-Na'im 2002:26). Compatible with Marris' model of coping with loss (see chapter 1), socio-cultural change (or innovation in Marris' terms) will not immediately follow a drastic change in people's lives. Not only did the majority (four out of five) of the urban refugees under study arrive in Dar es Salaam between only one to five years before the time of research (Table 3-5), but even more importantly, only a small minority (less than one in ten) envisages remaining in Tanzania permanently (Table 3-7). Thus the particularities of a situation of forced migration--i.e., most refugees are simply waiting to go home--are likely to form an obstacle to any precipitous socio-cultural change, hence people's

PAGE 346

330 persistent adherence to familiar socio-cultural strategies embedded in the historical, political and economic background of their respective home countries. used to indicate father's sister () or mother's brother () are not combined with the adjective "younger" or "older." 6.3 Age The importance of age as an important social marker in the African context (see chapter 1) is exemplified in a number of African languages, where, not rarely, age terms supersede gender categories. Oyewumi (1997), for example, in her book The Invention of Woman: Making African sense of a Western Gender Discourse, describes how in Yoruba culture (Nigeria) seniority is the primary principle of social categorization. "While most names and all pronouns are ungendered, the third-person pronouns o and won make a distinction between older and younger in social interactions" (40). Kinship terms as well are coded by age relativity, as "there are no single words in Yoruba denoting the English gendered kinship categories of son, daughter, brother, sister" (42). Swahili, as well, has only one, genderless third-person pronoun (yeye), and uses the same word for daughter and son (mwana). There are separate words to indicate brother (kaka) and sister (dada), however they are rarely used without the specification of age: e.g., younger brother (kaka mdogo) of older sister (dada mkubwa). The terms for father's brothers or mother's sisters as well contain indications of age: e.g., mama mdogo is a younger sister of the mother, baba mkubwa an older brother of the father.33 These terms are so engrained in East African cultures that even when speaking French or English, a person will very rarely not include the age-related adjective: e.g., "my little brother" (mon petit frre) or "my big sister" (ma grande soeur). 33 Both terms may also be used to address very close friends of a different age than oneself. Separate terms shangazimjomba

PAGE 347

331 Even as all societies worldwide have age categories and prescribed social relationships based on age, studies on the topic have focused almost exclusively on East As Foner and Kertzer point out: "In most societies the organized age system involves only males, age systems spanning the entire adult lives of females are rare" (1978: 1082 footnote). Africa (Keith and Kertzer 1984), where age "divides individuals and groups in a system that constitutes the general framework of society through which it can be represented in its entirety" (Bonte 1985:57, own translation). These so-called age-system societies are organized around age-sets consisting of "men34 who are initiated in youth during a definite span of time, and as a group share certain constraints and expectations for the remainder of their lives" (Spencer 1976:153). Together, as groups, they go through the different age-grades: e.g., youth, warriors, junior elders and senior elders. Foner and Kertzer's analysis compares twenty one East African age-set societies, half pastoral and half agriculture based, in which the number of defined age-grades may vary from two to eleven (1978:1084). The general pattern arising from these East African age-system societies is one of an age hierarchy that requires respect from the younger ones for the older persons and where age is a prerequisite for gathering increasing political, economic and social power, a system commonly referred to as a gerontocracy. While the "elders are at the center in control of family and settlement affairs," younger men are "on the fringe and associated with the bush" (Spencer 1976:155). Contrary to "our [Western] society [where] most old and young do not have access to the most highly rewarded economic roles" in most age-set societies, "greater rewards are bestowed on members of each successively older age grade" (Foner and Kertzer (1978:1985). 34

PAGE 348

332 A simplified age-set system, roughly based on Spencer (1976), consists of a group of youth undergoing a circumcision rite that allows them into moran-hood.35 Morans are This was not always the case, in some societies young adults were allowed to get married soon after their initiation rites into manhood. expected to demonstrate their independence by leaving their ancestral compound (Llewelyn-Davis 1996:424) and travel around in search for opportunities to prove their manhood, exerting their "wanderlust" (155). At around the age of thirty, they are finally allowed to get married,36 have children and are beginning to acquire economic and political influence in the society, as junior elders. It is only when their own children are adults undergoing the puberty rites themselves, that junior elders acquire the highest status, namely that of senior elders. Transition from one age grade to the next are marked by communal rituals and formalized procedures accompanied by considerable public ceremony. Comparing East African age systems with the importance of age as a social marker in the United States, Foner and Kertzer point out that: In modern Western societies a person's chronological age is but a partial clue to his social location--the roles she/he is likely to fill and the rewards she/he receives. For example, class and ethnic stratification crosscut lines of age stratification; they are a source of variations in age norms and provide other important sources of identity for the individual. In age-set societies, although bases of social inequality other than age are operative, the age system is more definitive. Individuals become part of the formal age-stratification system as members of an age set. Thus to know which age grade an individual occupies . is to have a fairly clear idea of his socially defined roles and tasks, his rights and responsibilities, his relations with age peers and age dissimilars. (1978:1085) It is important at this point to draw attention to the fact that Spencer's, Foner and Kertzer's and many other age system studies were primarily based on data from 35 Moran is the Maasai word for warrior. 36

PAGE 349

333 "preliterate societies" whose ethnographic descriptions date back to the 1940s, 50s and 60s (Foner and Kertzer 1083, footnote). As such, we may expect changes to have 37 Including Eastern Congo, Burundi, Rwanda and Tanzania. occurred over the past half a century, and yet [t]hose communal rituals are [still] common to every African society, and they are certainly continued at the urban centers, even in the climate of a generalized conversion to Islam and Christianity. While in the African villages people of the same ethnic group pass through rituals and become older at each one of them, members of nations and states also become older by receiving education, assuming social roles associated with marriage and procreation, and by going through a variety of national and localized rituals. (Aguilar 1998:17) The disproportionate focus of scholars on East Africa's age set societies has to a certain extent obscured the fact that "age categories and prescribed social relationships based on age exist in the absence of elaborate age settings" as well (Keith and Kertzer 1984:22), and that in any society "people's lives are ordered by societal norms regarding age-appropriate behavior, roles, and status" (29). Given the proliferance of age set societies in East Africa, whose age-systems were based on similar principles and comparable prescribed roles and responsibilities across the region,37 I propose that herein lies (part of) the answer as to why the four age groups consisting of Congolese, Burundese and Rwandese men and women, to a greater or lesser degree, act differently in a situation of forced migration. However, in order to have a better view on how age-related variations are embedded in the larger socio-cultural context, I believe it is advisable to first recall the most important characteristics of their respective coping strategies and social network compositions separately by age category.

PAGE 350

334 6.3.1 Persons younger than 25 years of age The overwhelming majority of persons in this age group is single, not living with a partner and childless. Their education level is the highest of all age categories, except with regard to a post-secondary education as a result of war-related structural problems disrupting the normal course of education. Almost three quarters of respondents in this age group were students at the moment of leaving home, while few were engaged in paid employment or trade. In Dar es Salaam, the youngest respondents do not make a living out of trading activities for lack of capital, rather they are the most likely to be self-employed, i.e., as missiontowns, hairdressers, etceteras (Chapter 3). While the reasons for leaving home do not diverge substantially from those of the other age groups, the motivations to come to Dar es Salaam do. The youngest respondents are much more likely to have decided to come to Dar es Salaam to join relatives or friends than out of concern for the security situation in the refugee camps (Table 3-6). The same lack of political concern or involvement shows from the conditions stipulated for the return home: peace is more likely to be indicated as a sufficient reason to return home than among the other age groups, while significantly fewer of the youngest respondents are concerned about democracy, human rights or ethnic equality (Table 3-7). The worry about finishing their studies is foremost on the younger ones' minds when compared to the other age groups, contrary to the significantly lower level of concern for the lack of work permits or the need to return home (Table 4-1). With regard to their social networks, the youngest respondents have the lowest homophilous rates for age of all age groups (Table 5-6). In other words, they depend to a much larger extent than the other respondents on support from persons older than

PAGE 351

335 themselves (Table 5-7). In addition, they are the least represented age category among the alters of the other age groups. This suggests that younger respondents are dependent on their age seniors, with regard to material as well as immaterial support, to a much larger extent than vice versa. The older generations' economically and politically stronger positions allow them to be providers of support for the youngest refugees. This pattern of dependency is noticeable as well from the fact that relatives make up half of all pre-Dar es Salaam alters among the youngest respondents (Table 5-9). From the above appears the profile of one without dependents and who is very dependent on persons older than him/herself, particularly relatives. This generation is hardly involved politically and lack of economic autonomy or opportunities prevent the youngest refugees to continue their studies or engage in profitable professional activities. They find themselves, as it were, at the bottom of the gerontocratic ladder of life. 6.3.2 Between 26 and 35 years of age In the second age group, half of the respondents are still unmarried. The ones that are married usually, but not always, live with their spouse. Half are childless, of the others the majority have one or two children. Even as they count more post-secondary graduates, more than one third reported being a student at the time of leaving home. Many more are engaged in trade activities compared to the youngest generation, suggesting a higher level of economic means, but being self-employed remains the main source of income (see chapter 3). While miserable camp conditions remain the main reason for coming to Dar es Salaam, this age category is most concerned about the security situation in the camp (Table 3-6). I was told by many male informants that they left the camp to avoid being forcibly recruited by one of the rebel factions. Their major conditions for returning home

PAGE 352

336 are the same as those of the two older generations (peace at home, democracy, ethnic equality etc.) indicating a higher level of political involvement than the youngest age 38 This category was defined for alters who are currently residing on the African continent but not in Dar es Salaam nor ego's home country (see chapter 5). group (Table 3-7). They are twice as likely to complain about the difficulties in obtaining work permits than their younger brothers and sisters, suggesting a higher level of economic autonomy instead of the economic level of dependence characteristic of the youngest age category. They also aspire to leave Tanzania in significantly larger numbers than any other age category (Table 4-1). With regard to social networks, the second age group has twice as many ties with persons of their own age than do their younger peers (Table 5-6). On the other hand, the composition of their networks in terms of the nationality of alters is most inconsistent with those of the other age groups, which are more or less similar. They count significantly more Congolese and fewer Rwandese among their pre-Dar es Salaam alters, while the proportion of Tanzanians among network members met after arriving in Dar es Salaam is substantially lower than for any of the other age groups (Table 5-7). They also have almost twice as many pre-Dar es Salaam alters living elsewhere in Africa,38 and lost contact with a significantly larger proportion of social network members met before arrival in Dar es Salaam than any of the other age groups (Table 5-8). Even as social network members of this age provide somewhat more material help to ego compared to the youngest, among pre-Dar es Salaam alters they provide only half as often financial support when compared to the two older age groups (Table 5-11). The profile emerging from the coping strategies applied by respondents between 26 and 35 years of age is one of increasingly independent persons. Their increased level of

PAGE 353

337 economic self-sufficiency in combination with a still relatively low level of family obligations allows them to be the most mobile of all age groups. This high level of 39 Given that all respondents were requested to list the same number of network members. mobility explains why a significantly higher proportion of their supportive network members reside elsewhere in Africa, i.e., outside of the home country and outside of Dar es Salaam. At the same time, because more Burundese and Rwandese in this age group were refugees in Congo (before relocating to Tanzania), they had the opportunity to establish a significantly higher number of relationships with the supportive Congolese. However, most of whom lost all contact with them afterwards due to the communication problems in conflict zones. At the same time, the high level of mobility feeds their aspirations to leave Tanzania and "try their luck" (kutafuta maisha) in a third country to a larger extent than in any of the other age groups. Although this study did not include social network size as a variable,39 one may logically assume that the social networks of refugees in this age category are larger than those in other age categories, because of the former's increased number of opportunities to meet new people, hence the choice to establish ties with persons with whom one shares more characteristics and/or experiences. As such, it is not surprising that significantly fewer Tanzanians appear among the ten network members listed in the survey, since higher levels of mobility logically provide more opportunities to meet and establish ties with compatriots and fellow refugees in Dar es Salaam. In addition, the persons who become part of the social networks of these mobile respondents are likely to be of their own (mobile) generation, explaining the significantly higher homophilous rate for age

PAGE 354

338 both among social network members met before and after arrival in Dar es Salaam in this age group. compatriots: "Well, we didn't choose to be Rwandese, right? That's God's work. But, me, if I were asked where I wanted to live, really I wouldn't have to think twice. I prefer to live with the Congolese." There remains the fact that respondents in this age group count significantly less Rwandese among their pre-Dar es Salaam alters compared to the other categories. In view of the fact that homophilous rates for nationality are the highest among pre-Dar es Salaam alters, we may assume that it is particularly Rwandese respondents who count significantly fewer compatriots among their social network members met before arrival in Dar es Salaam. The reasons for this phenomenon may lie in the 1994 genocide, and the subsequent resolve of a substantial number of this generation of Rwandese refugees to sever all ties with their home country,40 however more research into this issue is needed in order to provide a well-founded and satisfactory explanation. 6.3.3 Between 36 and 45 years of age In this age category, three quarters of respondents are married and have one to four or more children, while the number of widowed and divorcees is on the increase. Paid employment and trade activities were their main sources of income before leaving home, implying a level of economic and socio-political autonomy exceeding that of the two younger age groups. Yet while trade remains a major source of income after arrival in Dar es Salaam, a substantial proportion engages now in self-employment (Chapter 3). A significantly larger proportion of respondents than in any of the other age categories arrived in Tanzania around ten years ago, at a time when they were between 26 and 35 years of age and belonging to the "mobile" generation (Table 3-5). Probably as a 40 Daniel, for example, would much rather prefer to live among his neighbors the Congolese than his

PAGE 355

339 result of having been in exile in Tanzania for a longer period of time, significantly more have resided in one of the refugee camps and registered with UNHCR than among the 41 However, it should be mentioned that the high number of post-Dar es Salaam alters whose age is unknown (21%) makes it impossible to draw any firm conclusion in this respect. two younger generations (Table 3-5). Having had more first-hand experience, they quote miserable living conditions in the refugee camps as the major motive for coming to Dar es Salaam in significantly larger numbers than any of the other age groups (Table 3-6), and are also significantly more concerned about the lack of work permits (Table 4-1). All of the above findings are logically related to the fact that this generation has the responsibility to provide for the largest numbers of dependents (including more children of a school-going age) compared to all other age categories. Their rate of homophilous relationships for age decreased significantly as a result of the situation of forced migration. Among pre-Dar es Salaam alters, the rate approaches the high level of the second generation, while among post-Dar es Salaam alters, on the other hand, it approaches the low level of the youngest respondents (Table 5-6), possibly indicating a decreasing number of opportunities to establish ties with age mates.41 The higher homophilous rates for age with pre-Dar es Salaam alters combined with their higher economic resources, ensures that significantly more refugees in this age category receive financial support (Table 5-9) from more of their age mates than is the case in any of the other age groups (Table 5-11). The urban refugees in Dar es Salaam of between 36 and 45 years of age carry the heaviest burdens in terms of family obligations. Even as they have been in Tanzania somewhat longer--implying that the decision to leave their homes was taken at a point when they had fewer and smaller children, maybe even were unmarried--they financially

PAGE 356

340 depend in large part on social network members with whom a relationship was established before the flight. Because these comprise a relatively high number of age mates who, following the regional life cycle, dispose of sufficient economic assets and are thus able to support their exiled friends and relatives financially to a significantly larger extent than any of the two younger age groups. 6.3.4 Persons older than 45 years of age. None of the persons in this age group is unmarried, yet almost half are either widowed, divorced or living without their spouse. More than half of the oldest respondents have three or more children, and the majority has, usually adult, children who are not residing with them in Dar es Salaam. Their education level is the lowest of all the age groups, and a quarter was working in agriculture when leaving home indicative of a rural rather than an urban background. The major source of income of this generation are trade activities, at home as well as after arriving in Dar es Salaam, suggesting a relatively high level of financial means. Paid employment was the second most important source of income at home, similar to the third age group, but far less than any of the other generations did the oldest respondents turn to self-employment after arrival in Dar es Salaam (Chapter 3). Compared to the two youngest age groups, twice as many resided in one of the refugee camps for a certain amount of time and registered with UNHCR, a trend similar to the one in the third age group (Table 3-5). Not only did significantly more refugees of this age group quoted having sufficient financial means as a reason to head for Dar es Salaam, they also reported finding accommodation with Tanzanian friends upon arrival in Dar es Salaam to a significantly larger extent than respondents of any of the other age groups (Table 3-7). What is foremost on their minds is the worry for relatives, who

PAGE 357

341 stayed behind or with whom all contact is lost (Table 4-1), and they also complain significantly more about being "minimized" by Tanzanians in general (4-2). The oldest refugees also count significantly more Dar es Salaam residents among their pre-Dar es Salaam alters than any of the others (Table 5-8), and the church or mosque constitutes their most important site for meeting new network members, significantly more so than for refugees of any of the other age categories (Table 5-9). While social network members of this age, particularly among pre-Dar es Salaam alters, are instrumental in providing financial support to the refugees (Table 5-11), the oldest refugees appear significantly less dependent on financial assistance from their pre-Dar es Salaam alters when compared to the third age group (Table 5-9). From the above findings, it appears that while the oldest generation of alters provides most financial support to the urban refugees in general, the oldest refugees themselves are somewhat less likely to depend on financial support from social network members, probably as a result of the fact that they have fewer familial obligations than the third generation. But there is an additional aspect playing an important role in their way of coping with the situation of forced migration. As members of the oldest generation, they feel entitled to unconditional respect from anyone younger in age than themselves. As refugees, however, they do not enjoy the same societal position and status they would have at home, and as such feel particularly offended by certain Tanzanian behaviors vis--vis refugees (see chapter 3). From the perspective of societal status, this generation stands to loose the most when leaving their homes where everyone acknowledges and respects their societal position of elder. I suspect that the reason why the oldest generation is more likely than

PAGE 358

342 any of the others to find accommodation with Tanzanian friends or have more Dar es Salaam residents among their pre-Dar es Salaam alters is one of self-selection. In other words, their pre-flight relationships with trustworthy Tanzanian friends or Dar es Salaam residents ready to provide accommodation and a local structure--indicative of a visitor's status instead of that of a homeless refugee--will play an important role in the decision to come to Dar es Salaam in the first place. In addition, so I was told by informants, the option of self-employment, e.g., being a tailor or hairdresser in someone else's workshop, involves an inferior status and is as such an unacceptable position for someone of a certain age. In summary, we find from the interpretation of the age-related findings of the previous chapters that Dar es Salaam's youngest urban refugees are still very dependent on their relatives, while the second generation is characterized by a high level of mobility enabling them to network with large numbers of age mates covering a large geographical area. Refugees of the third category struggle with the heaviest familial burdens but can count on financial support from pre-flight social network members of the same age, while the oldest generation in exile feels the loss of their societal position at home the most through the perceived lack of respect as a result of their refugee status. Strikingly absent from today's urban situation are well-defined age grades and their prescribed roles as well as initiation rites to indicate promotion from one age grade to the next. The much longer duration of contemporary education plays an obvious role in moving up the age where one may begin to acquire a certain level of independence from puberty to the early twenties. (Aguilar 1998:18). However, as the above analysis demonstrates, even as urbanization largely did away with the formal aspects of traditional

PAGE 359

343 age systems, the roles and expectations related to the different stages of life continue to guide people's behavior, even in the situation of forced migration, and are embedded in the regional socio-cultural context. For reasons of simplicity, gender was purposely left out of the above analysis. Yet, while it is true that the traditional age system was mostly relevant to only men's lives, it is interesting to note that the stages of the traditional life cycle remain discernible from the refugee sample, despite the fact that half of its respondents are female. Whether this indicates that an increasing number of urbanized and educated women currently tend to follow traditionally male patterns of life cycles remains an open question constituting a possible topic of further research. 6.4 Concluding Remarks Being an urban refugee in Dar es Salaam clearly does not involve the same implications and experiences for men versus women, Congolese versus Burundese or Rwandese, nor for young versus old. Embedding the findings regarding the coping and networking strategies of the different social categories in their socio-cultural, historical, economic and political contexts, at either the national or regional level, has de-essentialized the concept of "the" refugee experience, the goal I set out to accomplish in chapter 1. The underlying principle of the above analyses was that people rely on previously learned behaviors, attitudes and strategies when coping with a crisis. It is in the presumed predictability of the outcomes of familiar actions that they hope to find a continuity of meaning of life as they knew it before. Only when the results cease to be predictable will individuals introduce changes in their strategies, on condition that these changes are socio-culturally compatible with their historically acquired perceptual

PAGE 360

344 frameworks. It is in this respect that both refugee men and women will prefer to establish new relationships with women rather than men, because the existing cultural image of women as "apolitical beings" makes them a more prudent option in a situation of forced migration, than men, who are considered to be the political representatives of their national, albeit ethnic groups. Then why does the (average) Congolese, despite his illegal status, continue to walk as if "the street is his" (see chapter 4), and to provoke the humanitarian staff with his "harsh" and "pushy" attitude telling them what "is his right"? One would assume that the vulnerability of their undocumented status in Dar es Salaam and running the risk of being arrested and forthwith deported home by the Tanzanian immigration authorities (see chapter 2) would suffice to induce the Congolese with some behavioral changes. This appears not to be the case and the answer may be found in the Congolese cultural repertoire or world-view. The general disregard for state structures resulting from a rampant system of bribing any public servant or state official, combined with a heavy reliance on social relationships with persons in the right place are the cornerstones of life in Eastern Congo . and of Congolese refugee life in Dar es Salaam. Informants testimonies in chapter 4 relate how the strategies to connect oneself through marriage or adoption to a Tanzanian national are applied mostly by Congolese refugees. The Rwandese, on the other hand, walks while "looking 100 meters ahead, it is someone who is afraid that someone will catch him" (Chapter 4). Even as current events likely reinforced the feelings of suspicion reigning Rwandese lives, we gathered from their historical context that it is an attitude inherited from the past and one pervading Rwandese daily cultural praxis. And yet, this behavior or worldview inhibits the

PAGE 361

345 (re)construction of strong and supportive networks (through an increase of homophilous relationships for gender and nationality rather than the current decrease) thus rendering more difficult the process of coping with the situation of forced migration. Again, past experiences provided the answer in the form of heavy financial reliance on social relationships established long before the flight, together with increased attempts to seek UNHCR assistance (most of Rwandese refugees in Dar es Salaam having been refugees since 1994 and having resided in refugee camps in either Congo or Tanzania or both, for a number of years). This leaves the Burundese, who "walks as someone who is about to fall over . [who is] extremely tired" (Chapter 4) and who accepts official refusals for assistance so easily that it prompted one humanitarian worker to wish that "they should make more effort, try again." The Burundese lack of assertiveness, based on fear of (being denounced to) state structures, and subsequent by missing out on possible humanitarian assistance (according to one aid worker), has made them easy targets for economic exploitation (earning them the image of "hardworking" among Tanzanians or "extremely tired" according to the above refugee informant). At the same time, this fear also inhibits many of the Burundese refugees in Dar es Salaam to establish more strong and supportive relationships with compatriots (see chapter 4) compounding the difficulties and traumas of coping with the situation of forced migration. And yet, the main priority for survival in Dar es Salaam appears to be the avoiding of any contact with official and/or state structures. Having been discriminated against and persecuted by their own state at home, it is this same fear and behavior of avoidance that rules the lives of the Burundese in exile. Disposing of fewer financial means than the Rwandese and fewer strategically

PAGE 362

346 important social connections than the Congolese to deal with arrests and/or harassment from Tanzanian officials, they survive by applying pre-flight strategies of hiding from the authorities. They adhere to keeping a low profile and blending in by posing as Tanzanians (see chapter 4), even if this is to the detriment of establishing new, possibly emotionally rewarding, relationships with compatriots and, of defending their case with UNHCR and other humanitarian agencies. Studies on the effects of the African urban context on traditional socio-cultural patterns of aging remain conspicuously absent from the literature, yet it is certain that in addition to the specificities of urban life, modern systems of education as well have brought "significant changes to the cultural appreciation of age" in African societies (Aguilar 1998:19). Education has affected the age at which young people can begin to acquire economic and political independence and start their journey through life's stages, but it has also, especially in the urban context, brought the patterns of men's and women's life cycles closer together. Higher levels of education not only allow women to marry and start families at a later age than before, it also has provided them with increased access to high profile economic, or even political, positions in society. In addition, increasing numbers of African women migrate independently of men (Tacoli 2001), an aspect of African migration traditionally lacking from the literature (Gugler and Ludwar-Ene 1995). Allowing for the fact that substantial differences remain between the age-related roles and responsibilities of men and women, the above analyses of the Congolese, Burundese and Rwandese refugees' strategies by age category did illustrate their socio-cultural embeddedness in historical (East) African age systems.

PAGE 363

347 Until recently, urban refugee populations were considered to consist predominantly of young, single males who were generally perceived to be largely able to fend for themselves in the absence of humanitarian assistance. Having established that today's urban refugee populations consist of men and women of all ages (see chapter 1), and that being an urban refugee in Dar es Salaam involves experiencing specific needs and problems according to one's gender, nationality and age, it should prove useful to have a look at humanitarian policies vis--vis urban refugees. The final chapter consists of reflections on existing urban refugee policies and methodological pre-occupations as well as theoretical considerations concerning the research project as a whole.

PAGE 364

348 many of the variables that should have been included. For example, sensitive questions related to political, religious, and national affiliations were not included in the questionnaire" (Goitom 1987:136). CHAPTER 7 REFLECTIONS, PREOCCUPATIONS AND CONSIDERATIONS Contrary to a recurrent assumption in the refugee literature that "to become uprooted and removed from a national community is automatically to lose one's identity, traditions and culture" (Malkki 1995b: 508), the findings of this research project have demonstrated that Dar es Salaam's Congolese, Burundese and Rwandese refugee men and women, young and old, apply previously learned behaviors, attitudes and world views to cope with their situation of forced migration. In addition to nationality, both gender and age proved to play a significant role in the variety of coping and networking strategies of forced migrants. While concurring with Daley that "apart from age and gender differences, refugees are differentiated according to class,1 regional origin, party and religious affiliation" (2001:208), I estimated that including all of these elements in one research project would not be the most feasible option, hence the selection of only three--politically less sensitive2--parameters.3 What all refugees share, however, is the experience of stress, albeit trauma, of being forcibly uprooted from their familiar surroundings for reasons that are beyond their 1 The reasons why I did not include class as a sample parameter is explained in the second section of this chapter. 2 Because of the political sensitivity of many of the refugees' situation, I introduced myself as a researcher who was not going to ask any questions related to politics. Many of the informants accepted being interviewed only after this particular condition was agreed upon. Often political conversations did ensue afterwards, but only when a certain level of trust was established. 3 Other urban refugee researchers struggled with the same issue: "[T]he sample survey failed to capture

PAGE 365

349 control. Sometimes this stress is compounded by the fact of being relocated to refugee camps (Beiser et al. 1989) or by violence-related traumas (Bracken et al. 1997), but always "common ongoing sources of severe stress include fears of being repatriated, barriers to work and social services, separation from family, and issues related to the process of pursuing refugee claims" (Sinnerbrink et al. 1997: 463). And yet, while the debate on the efficacy and cultural appropriateness of therapeutic strategies currently applied by relief workers and organizations is still ongoing (Knudsen 1991; Bracken et al. 1997), an increasing appreciation of the level of resilience displayed by the refugees themselves is slowly emerging in the literature. Refugees may well be "perhaps the maximum example of the human capacity to survive despite the greatest losses and assaults on human identity and dignity" (Muecke 1992:520) displaying a high level of resilience confronting adverse circumstances and being empowered by their own strategies of resisting imposed structures (Oliver-Smith 1991), here i.e., a forced relocation to the refugee camps in Western Tanzania. This new emerging image is very different from the one traditionally generated by the International Refugee Regime that presents refugees with "a peculiar kind of speechlessness in the face of the national and international organizations whose object of care and control they are" and whose "accounts are disqualified almost a priori, while the languages of refugee relief, policy science, and "development" claim the production of authoritative narratives about the refugees" (Malkki 1996: 386). It is precisely with these opposing images of "speechless dependency" versus "resilient agency" in mind that I formulated the following reflections on the urban refugee policies vis--vis the Congolese, Burundese and Rwandese refugees in Dar es Salaam as well as the

PAGE 366

350 methodological preoccupations and theoretical considerations in sections two and three of this concluding chapter. In this policy document, all three nationalities, Congolese, Burundese and Rwandese are listed among the fifteen nationalities that "tend to dominate urban refugee caseloads" worldwide (1997c:4). 7.1 "To aid or not to aid?" Conversations with UNHCR officials and other staff members of the UNHCR branch office in Dar es Salaam were all too often replete with elements of a discourse portraying urban refugees as "demanding" and "too expensive." The general perception is that [t]here is a growing tendency whereby refugees prefer to reside in the urban areas than going to camp [sic]. The obvious reason is that life in town, especially when paid for is better and easier than camp life. (UNHCR 2001b:2) This discourse and imaging of urban refugees is not particular to UNHCR's Dar es Salaam office, it is also reflected in its recently issued Comprehensive Policy on Urban Refugees (1997c)4 at the head quarters level in Geneva. This document states that "while constituting less than 2% of UNHCR's refugee caseload" urban refugees "demand a disproportionate amount (estimated at 10-15%) of the organization's human and financial resources" (5) and thus "[c]ontinuing open-ended assistance in an urban setting is not a solution and should be excluded" (14). Hence, the recommendation that in order to avoid unnecessary drains on the organization's ever diminishing financial resources there should be a focus on employment . critical in changing the attitude of international and local communities from one in which refugees are seen as "the problem" to one in which refugees, through their hard work, initiative and skills, are seen as an "asset" and part of the solution to both their own and the host country's economic problems. (17) 4

PAGE 367

351 Throughout the 1997 policy document, however, there is an underlying emphasis on the "irregular status" of urban refugees, on the host government's pre-existing acceptance of refugees residing in urban centers, and on the extending of assistance to urban refugees only "if the government does not object" (19). These particular weaknesses of the 1997 policy document were pointed out in an internal evaluation report (UNHCR 2001c). The report not only criticizes the 1997 policy document for using "terminology which stigmatizes people who are of concern to the organization" but rightly explicates that "[i]f UNHCR staff members regard and treat all urban refugees as potential troublemakers, then they are all the more likely to react in such a manner" (8). In addition, there is an inconsistency between advocating a policy of self-reliance and neglecting the existence of individual government policies to the contrary. A related weakness of the current policy is to be found in its statement that UNHCR should promote self-reliance in ways that "respect the policies of the government." It is unclear how this element of the policy statement is to be interpreted and implemented in countries where government policies actually obstruct UNHCR's efforts to promote self-reliance amongst urban refugee populations by, for example, denying them legal status, identity documents or the right to work. (5) In Tanzania, as the previous chapters have amply illustrated, urban refugees are not allowed to reside outside of the designated areas, i.e., the refugee camps in rural Western Tanzania. The refugees who decide to come to Dar es Salaam despite the official refugee policy, are usually well aware of the fact that material or financial assistance is generally not obtainable from the UNHCR office in Dar es Salaam. However, they do rely on the UNHCR office to provide them with protection as is stipulated in the first paragraph of Chapter 1 of the 1950 Statute of the Office of the United Nations High Commissioner for Refugees, namely that the UNHCR shall first and foremost "assume the function of

PAGE 368

352 providing international protection5 . to refugees" (UNHCR 2000a:41). Providing protection usually (but not exclusively) takes the form of a so-called "Protection Letter" In view of the fact that a refugee can no longer avail upon the legal protection of his own state through one of its embassies. (see also chapter 2), confirming one's status as an asylum seeker or refugee. Its function is to protect the bearer from being arrested by the Tanzanian authorities as an illegal alien and face possible forced repatriation. This letter does not involve the providing of material assistance, for which separate criteria are used, and as such carries no financial or budgetary implications for UNHCR. And yet, as one among many other informants, this Congolese informant, who has been in Tanzania since 1995, bitterly complains about the way the UNHCR Office in Dar es Salaam has been receiving refugees and asylum seekers that come to seek their protection. Going to UNHCR? The reception you get when you go there, they leave you standing outside, and even one time . the police came, they entered into the compound of the UNHCR to arrest people. This was in 97. That event made us even write to the representative at the time, to say that it was not normal that police officers came, entering the [UNHCR] compound and arresting people inside. This shows that we are not really protected here, we have no protection. Most of the informants who had been in contact with the UNHCR office at one point, complained about being left outside to wait for hours on end, on the only day of the week that the office receives the refugees, and being told at the end of the day to come back the following week. Others would camp outside the compound walls of the UNHCR office until being "chased" away by the guards (see also chapter 3). According to UNHCR officials, these complaints concern refugees who have been given definite, negative replies to their requests (e.g., for financial assistance, resettlement, etc.), yet who insist on 5

PAGE 369

353 pursuing their file in the hope of having the decision reversed. Occasionally, "reports of widespread corruption and intimidation by officers of the UNHCR in Tanzania against refugee, no-one would be able convince him to live in a refugee camp because of the appalling living conditions. refugees" surface (The Guardian, March 11th, 2002), but no proof has ever been brought forward in support of any of these allegations. From talking to both refugees and asylum seekers on the one hand, and UNHCR and its implementing partner Umati on the other, I found their interactions to be rife with tensions, misinformations and misunderstandings mostly from a lack of communication. While the refugee agencies are pre-occupied with budgetary arguments based on the conviction that urban refugees first and foremost seek material assistance, the refugees themselves are concerned by the inability (unwillingness?) of the UNHCR office to issue them legal documents and/or residence permits that would protect them from the Tanzanian police and immigration authorities. Both sides complain bitterly of not being listened to and being treated disrespectfully by the other. This is the concrete expression of the communication gap at the root of the clash between the speechless image of the refugee whose voice is appropriated by the international refugee regime, and the uprooted individual men and women who are trying to regain control over their lives by exerting a high level of resilience and resistance against decisions imposed upon them and which they consider unacceptable. For example, the observation of one UNHCR official from the Dar es Salaam office that "[m]uch effort is required to convince them [the urban refugees] to go to the camp6 especially those going for the first time . [and c]ontinuous counseling is the norm" (UNHCR 2001b:2), reveals again the disregard by the refugee 6 One Tanzanian senior official from the Refugee Department at the Ministry of Home Affairs, who had been the Director of one of the refugee camps in Western Tanzania for many years, stated that, if he were a

PAGE 370

354 agency of what the urban refugee might have to say about his/her motivations to reside in Dar es Salaam and resist going the refugee camps. Another instance is the recurrent argument from UNHCR officials that there is insufficient funding to assist urban refugees whose cost per capita exceeds by far that of a refugee in the camp (see chapter 2). However, most informants emphasized that they are not seeking UNHCR's financial assistance, rather they rely on the international organization for documents legalizing their stay in Dar es Salaam and possibly allowing them to exercise their profession (see chapter 4). In fact, if more self-sufficient urban refugees were to be issued with work permits, there would automatically be fewer requests for financial assistance directed at the refugee organizations. Ultimately, the decision to allow refugees to reside in urban centers and engage in gainful employment is dependent on the authorities of the country of asylum. The motivation of the Tanzanian government to deny legal residence and the right to work to refugees in Dar es Salaam is based on the assumption that local integration is a disincentive for repatriation. And yet, The pattern of repatriation to Rwanda after the genocide in 1994 and to some extent that of South Africans in the early 1990s suggests that educated and economically well-off refugees might be the first to repatriate when conditions allow since they are empowered, and have better prospects of playing a key part in the process of reconstruction and rehabilitation in the country of origin. (Rutinwa 2002:23) In addition, there is the proposal by Dr. Musoke from the Center for the Study of Forced Migration to issue a separate class of work permits affordable to refugees (see chapter 2). Given the policy guidelines, as well as academic arguments, in favor of treating the urban refugees in Dar es Salaam as "assets" rather than "the problem" (UNHCR 1997c:17), it is regrettable that UNHCR Dar es Salaam continues to invest so much effort in counseling urban refugees to go to the refugee camps, instead of redirecting their efforts towards

PAGE 371

355 lobbying with the Tanzanian government for a change of the national refugee policy. While it is true that UNHCR at the level of the Branch Office in Dar es Salaam have in the recent past undertaken a few modest attempts to sensitize the Tanzanian Refugee Department to the advantages of issuing work permits to urban refugees,7 I believe that UNHCR, as an international organization could make more high level efforts to advocate their policy with the Tanzanian authorities. In view of the fact that in a number of countries (such as India, Egypt, Kenya, Macedonia, Russia, Syria, Yemen and others (UNHCR 2001c:3)) thousands of refugees are UNHCR registered and, in certain instances, have been allowed to work, it is recommended that expertise gained and arguments collected in the afore-mentioned countries is shared with others,8 so as to allow for an effective lobbying of a more humane approach to urban refugees in general. Advocating at the level of the Executive Committee that employment, educational opportunities and a certain measure of economic and social integration in the country of asylum are important for refugees' well-being, including their psychological and physical health (UNHCR 1994:23) is ineffective, when not accompanied by directives for country offices to advocate among reluctant host countries the right for self-sufficient refugees to legally reside and be gainfully employed in urban centers. Involving the urban refugees in the process, through local refugee committees and elected representatives, would but enhance a worthy and respectful dialogue and exchange of experiences and expectations, and as such, avoid the current, unnecessary misinformation and frustrating gaps of 7 A 1998 letter from the former UNHCR Senior Protection Officer to the Head of the Refugee Department in the Tanzanian Ministry of Home Affairs recommended the issuance of work permits to refugees. 8 Applying the principles of Knowledge Management theory.

PAGE 372

356 communication,9 contributing in the end to the development of a more workable urban refugee policy. See chapter 1 for my personal experiences related to finding informants and gaining access to the respective refugee populations. 7.2 Positionalities Carrying out a research project among non-camp refugee populations, particularly those consisting largely of illegal residents, entails a number of methodological considerations. One is the use of certain sampling methods, i.e., snowball sampling (see more in chapter 1), which cannot guarantee the sample's representativeness according to strict statistical principles of randomness.10 Another issue is the researcher's unavoidable dependence on certain members of the respective refugee populations to introduce him/her into the communities and to assist in carrying out surveys.11 "While using co-ethnics to collect the data [is] necessary for access and language skills, the downside [is] the possibility that it would bias the data collected" (Bloch 1999:377). Occasionally, refugee researchers mention the difficulties they encounter in overcoming the strong feelings of suspicions and high level of mistrust of the refugees resulting in the necessity to omit the registration of certain data, such as for, example, contact addresses (379), real names or other possibly politically sensitive information, such as ethnic affiliations (see chapter 1). Rarely, however, is mention made of the possibility that the data collected are biased because of the interviewees' interpretation of 9 Apparently, there had been regular meetings between the UNHCR office in Dar es Salaam, representatives from the Refugee Department at the Tanzanian Ministry of Home Affairs and local NGOs and elected refugee representatives until 1998. It was decided by the then-UNHCR Country Representative to discontinue them for no apparent reason given, according to refugees and local NGOs alike. 10 As explained in more detail in chapter 1, this was the reason for pre-setting the number of men and women, Congolese, Burundese and Rwandese, and individuals of four different pre-defined age categories to be included in the sample. 11

PAGE 373

357 the interviewer's positionality. I found only few references in the literature pondering whether it is possible that asylum seekers may accentuate their difficulties with the hope that publicity will rally sympathy to their plight. Thus, the possibility that data are being biased [sic] cannot be discounted entirely . (Sinnerbrink et al. 1997:468) Only Knudsen investigated in-depth "how a subject (refugee) [sic] tries to construe a life history in an attempt to reduce the uncertainty of his/her life course and simultaneously secure his/her rights towards the future" (1990:123). Comparable to the way in which the refugees' interpretation of Knudsen's positionality influenced the (re)construction of their life histories (see more chapter 1), I noticed during my research that different groups of refugees presented themselves differently depending on the roles they ascribed to me. A considerable number of Congolese men and women, for example, conspicuously downdressed at our first meeting (i.e., the women covering up their expensive hairstyles with scarves and taking off golden jewelry) in the conviction that I was the representative of an international NGO scouting for a project for "poor refugees" on which to spend money (see also chapter 1). On another occasion, a group of Burundese women, conspicuously professionally attired, presented themselves as the representatives of their respective women's associations, founded to explicitly engage in matters such as conflict resolution and peace reconstruction, expecting me, as an "undercover" official from UNHCR (or UNIFEM) to invite them to international workshops on the topic (see also chapter 4). On most occasions, except for the older Rwandese who considered me to be associated with the International Crimes Tribunal for Rwanda in Arusha and thus avoided to meet with me at all, respondents were appropriating the prevailing discourse on and image of "a refugee" as poor and destitute, and dependent on international assistance, i.e.,

PAGE 374

358 money from a white person. Their strategy was not exclusively reserved for Western researchers, and not always carried out for the same purpose. One informant related that he learned, after an initial mishap, to downdress when going to the Tanzanian Refugee Department in Dar es Salaam lest he be chased away for not being a "real" refugee and risk being refused a temporary residence permit.12 In other instances, informants re-interpreted the offer by some countries to resettle certain categories of refugees as their "right" as refugees to be resettled in a third country (see more chapter 2). When confronted with "UNHCR denying them their right as refugees to be resettled," some refugees took the matter to the Legal and Human Rights Center or the Amnesty International office in Dar es Salaam (and sometimes beyond) so as to "expose the infringement on their human rights as refugees." From these examples emerges not only the agency displayed by the urban refugees in dealing with their situation of forced migration, but equally how "the meanings they make, depend . on the discourses available to them" (Canning 1994:377).13 Discourses such as those on human rights, on third country resettlement programs and on "poor, dependent" refugees are continuously re-interpreted and its constitutive elements used by the refugees as arguments in strategies aimed at regaining control over their lives and securing future ambitions. Of particular relevance to my own research was the strategy of a large number of respondents and informants to present themselves as "alone, poor and needy." These conscious strategies turned out to be an obstacle in collecting certain types of 12 Malkki, as well, doing her research in Western Tanzania, noticed how refugee workers in Tanzania had a "pronounced tendency to try identify and fix the "real" refugee on extralegal grounds . making it possible to claim that given people were not real refugees because they did not look (or conduct themselves) like real refugees" (1996: 384). 13 See also chapter 1.

PAGE 375

359 information, such as, socio-economic background or class, network size, the actual amount of financial support received from social network members, etceteras. One female informant, for example, pretended to be the housegirl in the house she actually owned and had financed by trading internationally in women's cosmetics. Another man asked me for financial assistance to pay the monthly rent of his house while he had just imported several computers into the country intending to sell them locally. On more than one occasion, a respondent started out by stating that "not a single person" had ever helped him/her in any way during the flight from home or the years of residence in Dar es Salaam. Other times, amounts of money sent over from relatives and friends abroad were reportedly smaller than the bank charges involved in the transfer. When instances like these occur, it is important for the researcher to be very aware of the politics involved in the interactions between interviewer and interviewee, as well as to be careful in evaluating the reliability of the stories and data collected and whether to include them in the final analysis. As another corollary of this particular situation, I found that refugees who were socially well integrated in the host community and financially self-sufficient, were more often than not uninterested in participating in the research project. Not only because there was always the risk of exposing themselves as refugees by being seen talking to me, but in addition, they considered that there was nothing useful to gain, in neither social nor economic terms, from spending their time and experiences with a foreign social researcher. It is in the same line that usually only the very needy and really destitute refugees (e.g., female headed households with large families) humbly seek material

PAGE 376

360 assistance from the few humanitarian agencies in Dar es Salaam, thereby reaffirming the circulating stereotypical image of the "poor and dependent refugees." 14 This particular issue is one of the main criticisms of reviewers leveled at Sommers' research on Burundi refugees in Dar es Salaam, and one I completely concur with (Kibreab 2002). Despite the methodological pitfalls of doing research among undocumented refugees in terms of skewed samples, incomplete informations, albeit outright untruths,14 there is still a significant amount of knowledge to be gained from it. The following and final section explores the potential contributions of this research project in the fields of urban refugees, and social networks in forced migration. 7.3 Urban Refugees, Social Networks and Forced Migration Despite (the recent recognition of) ever increasing urban refugee populations on the African continent (Rogge 1986; Kibreab 1996), there remain significant "lacunae in knowledge about African urban refugees" (Kibreab 2002:328) resulting in "a dearth of data on their demographic structure, socio-economic background, treatment and survival strategies" (1996:132). In addition to contributing knowledge to these specific areas, the explicit purpose of this research project was to explore the effects of forced migration on the composition the refugees' personal social networks, as well as the role that these play in the refugees' coping strategies in an urban context. Many of the findings of this research project are compatible with the data generated from the research done on Ethiopian and Eritrean urban refugees in Khartoum, Sudan (Kibreab 1996; Goitom 1987; Karadawi 1987), such as the urban background of the majority of urban refugees; the diversity of reasons to flee the home country; the bypassing of border towns or refugee camps in heading straight for Dar es Salaam (i.c., Khartoum); the on average higher education levels and the proliferation of urban-based,

PAGE 377

361 professional skills; and, the readiness to repatriate when peace at home prevails (Kibreab 1996). One contentious issue remains the composition of the urban refugee population for lack of census data and the impossibility to apply strict statistical criteria in the selection of samples. However, the observation of the above authors (who conducted their research in the 1980s) that urban refugee populations consist of mostly young males has been contradicted by more recent data on from UNHCR urban refugee populations in other countries. My own sample, based on predetermined gender proportions, half men/half women (see chapter 1), can neither confirm nor negate the above observation. Conversations with informants, however, appeared to confirm the trend exposed in recent case studies which have documented an ever increasing proportion of women among African migrants in general. Taking into account furthermore, the growing rates of urbanization in all African countries, I am convinced that urban refugee populations are much more diversified than they were twenty years ago, even as young, males continue to make up the more accessible segment of African refugee populations because of their language proficiency and other, age and gender related, cultural factors. While problems regarding education, employment and the lack of legal papers (Karadawi 1987) were profusely documented, the systematic analysis of empirical data on Dar es Salaam's refugees' social networks allowed the fine-tuning of earlier general observations concerning urban refugees' reliance on remittances from relatives and friends (both from the place of exile and abroad) and refugees' places of social interaction such as, for example, churches (Goitom 1987). Contrary to the refugee-centric approach (Chambers 1986) of the above urban refugee studies, my research illustrated the

PAGE 378

362 relevance and importance of daily contacts between the urban refugees and their Tanzanian hosts through the analysis of the social network data. In addition, the testimonies of the Great Lakes region refugees in Dar es Salaam illustrated a substantial level of appreciation for the hospitality and different types of support provided by the Tanzanian man and woman in the street in spite of a hostile media and circulating negative images. Finally, comparing attributes of social network members with whom a relationship had been established before versus after the flight, allowed the contextualization of the refugees' socio-culturally inspired coping strategies in the respective historical, economic and political structures. This purely anthropological perspective was also largely missing from previous studies on African urban refugees. While the focus of this research project was primarily on the role of social networks in the situation resulting from forced migration, it is a well-established fact that social networks play an equally important role in the very process of (forced) migration, albeit the establishment of transnational communities (Koser 1995; 1997; 2002; Crisp 1999). And yet, refugees seem destined to suffer the same fate at the hands of transnational community studies as they have at the hands of international migrations studies--at best exceptionalism, at worst exclusion (Koser 2002: 138) Koser argues that the perceptions that refugees are expected to return soon to their respective homelands while maintaining few ties with home--differentiating them from "regular" migrants--does not hold true for many of the world's refugee populations. In effect, growing Rwandese, Burundese and Congolese communities worldwide, but particularly in Canada, Belgium and France, have been supporting not only their relatives and friends at home, but also, more recently, increasing numbers of compatriots who found refuge in countries in the region such as, for example, in Tanzania. The social

PAGE 379

363 network data in my study show that one in every twenty supportive network members of the urban refugees in Dar es Salaam reside in a country outside of the African continent. But also network members residing in the refugees' respective home countries, representing between one in five to one in ten supportive ties depending on nationality, are indispensable with regard to providing financial and other assistance. Evidently, the global proliferation of electronic means of communication (e.g., email, mobile phones, bank transfers, etceteras) has been instrumental in this development, just as it has in the establishment of other transnational communities. 7.4 Epilogue Since the time of conducting the research, there have been few changes in the situation of Congolese, Burundese and Rwandese refugees in the region. The withdrawal of the Rwandese and Ugandan armies from Eastern Congo had left a power vacuum, in which competing Congolese rebel groups deployed any means imaginable in order to increase their hold on the local population: e.g., the recurrent massacres of the civilian population in Bunia. It is hoped that the recent arrival of UN led peacekeeping troops may ultimately succeed in stabilizing the situation in Eastern Congo, but in the meantime, Congolese refugees continue to seek refuge in Tanzania. In Burundi, despite the installing of the transitional government in November 2001, the sporadic fighting between rebels and government soldiers continues. The several thousands of Burundese refugees who repatriated to conflict-ridden regions in South Eastern Burundi (IRIN 2003b) during the month of May 2003, cited "reduced food rations, Tanzanian restrictions on their economic activities and their perception of the camps as a prison" (IRIN 2003a) as motivations to return home, rather than a stabilization of the internal situation.

PAGE 380

364 In October 2002, the Tanzanian government decreed that all remaining Rwandese refugees should leave the country by the end of the year, and by December 27, 2002 the "voluntary repatriation" of Rwandese refugees from camps in northwestern Tanzania had been completed (IRIN 2003c). However, during the previous months, there were several reports concerning thousands of Rwandese refugees leaving Tanzania and trying to find asylum in one of its neighboring countries for fear of being repatriated to Rwanda (2002a). Whether to escape the situation in the refugee camps, described by both the UNHCR and the World Food Program as the "worst ever" due to the halving of food rations (IRIN 2003d), or to avoid being "voluntarily" repatriated, Congolese, Burundese and Rwandese refugees are clearly increasingly motivated to not reside in a refugee camp and to join their compatriots and other fellow-refugees in one of Tanzania's urban centers. Curbing the situation not being a workable option, the development and implementation of an appropriate urban refugee policy (guaranteeing the urban refugees' protection under international law) by both the Tanzanian government and the UNHCR, is urgently called for.

PAGE 381

APPENDIX A MAPS Map 1-1: The Great Lakes Region (Courtesy of the United Nations, Department of Public Information). 365

PAGE 382

366 Map 1-2: The Democratic Republic of Congo.

PAGE 383

367 Map 1-3: Rwanda and Burundi.

PAGE 384

368 Map 1-4: Rwandese and Burundese refugee movements, 1994-99 (Courtesy of UNHCR).

PAGE 385

369 Map 1-5: Internally Displaced Persons in Burundi, 27 June 2000 (Courtesy of UNHCR).

PAGE 386

370 Map 1-6: Refugee Camps in Western Tanzania, May 2002 (Courtesy of UNHCR).

PAGE 387

371 Map 1-7: Dar es Salaam, Urban Growth 1945-1998 (Courtesy of University of Glasgow).

PAGE 388

372 APPENDIX B TANZANIAN PERCEPTIONS ON REFUGEES IN DAR ES SALAAM Gender: Ward: Place of Origin: Age: Religion: Level of Education: ________________________________________________________________________ 1.a Do you know any refugees personally (e.g., relatives, friends, neighbours, etc.) that you have regular contact with ? Yes No Don't know 1.b.If yes, how many refugees do you know and have regular contact with in each of the following nationalities ? Congolese : None Less than 5 More than 5 Burundese : None Less than 5 More than 5 Rwandese : None Less than 5 More than 5 2. Those refugees who manage to make an honest living (shughuli hala) here in Dar es Salaam, should they be allowed to exercise their profession and be given a residence permit ? be forced to go to the camp and depend on international aid ? other: . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . don't know 3. What is according to your experience the most common profession among refugees in Dar es Salaam by nationality and gender ? (tick only one per categorie) a. Congolese men Hairdresser Tailor Missiontown Business Musician French teacher Other: . .. .. .. .. .. .. .. .. .. .. Don't know b. Congolese women Hairdresser Tailor Missiontown Business Musician French teacher Other: . .. .. .. .. .. .. .. .. .. .. Don't know c. Burundese men Hairdresser Tailor Missiontown Business Musician French teacher Other: . .. .. .. .. .. .. .. .. .. .. Don't know d. Burundese women Hairdresser Tailor Missiontown Business Musician French teacher Other: . .. .. .. .. .. .. .. .. .. .. Don't know e. Rwandese men Hairdresser Tailor Missiontown Business Musician French teacher Other: . .. .. .. .. .. .. .. .. .. .. Don't know f. Rwandese women Hairdresser Tailor Missiontown Business Musician French teacher Other: . .. .. .. .. .. .. .. .. .. .. Don't know

PAGE 389

373 4. Which of the following nationalities would you say represents the largest group of refugees in Tanzania? Burundese Rwandese Congolese Other: . .. .. .. .. .. Don't know 5. Which of the following nationalities would you say represents the largest group of refugees in Dar es Salaam? Burundese Rwandese Congolese Other: . .. .. .. .. .. Don't know 6.a. How would you feel if your daughter married a refugee man? I would be very happy I would not like it Other . .. .. .. .. . I would accept it I would not care Don't know 6.b. Which nationality would you prefer if your daughter were to marry a refugee man? Burundese Rwandese Congolese Other: . .. .. .. .. .. Don't know 7.a. How would you feel if your son married a refugee woman? I would be very happy I would not like it Other . .. .. .. .. . I would accept it I would not care Don't know 7.b. Which nationality would you prefer if your son were to marry a refugee woman? Burundese Rwandese Congolese Other: . .. .. .. .. .. Don't know 8. Which quality/characteristic do you like most among? Congolese : . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .... .. .. . Don't know Burundese: . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Don't know Rwandese: . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Don't know 9. Which quality/characteristic do you dislike most among? Congolese : . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Don't know Burundese: . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Don't know Rwandese: . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Don't know 10. What do you see as the main (socio-cultural) difference between Tanzanian men and . .? Congolese men: . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . Don't know Burundese men: . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . Don't know Rwandese men: . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Don't know 11. What do you see as the main (socio-cultural) difference between Tanzanian women and . .? Congolese women: . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Don't know Burundese women: . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . Don't know Rwandese women: . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Don't know

PAGE 390

374 Swahili version Biashara Mwanamuziki Mwalimu wa kifaransa Nyingine: . .. .. .. .. .. .. .. .. .. .. Sijui Gender: Ward: Place of Origin: Age: Religion: Level of Education: ______________________________________________________________________________ 1. Je, wewe binafsi unafahamiana na wakimbizi ambao unawasiliana nao mara kwa mara? (Kwa mfano: marafiki, majirani . .) Ndiyo Hapana Sijui Kama ndiyo, ni wakimbizi wangapi ambao unafahamiana wa kuwasiliana nao mara kwa mara? Wakongo : Hamna Chini ya 5 Zaidi ya 5 Warundi : Hamna Chini ya 5 Zaidi ya 5 Wanyarwanda : Hamna Chini ya 5 Zaidi ya 5 2. Je wakimbizi ambao wanaendesha shughuli halali hapa Dar es Salaam, waruhusiwe kuendesha shughuli hizo na wapewe kibali cha kuishi nchini? walazimishwe kwenda kwenye makambi na kutegemea misaada ya kimataifa? Nyingine: . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Sijui 3. Kwa uzoefu wako shughuli wazifanyazo wakimbizi ni zipi? (tick only one per categorie) a. Wanaume wa kikongo Kinyozi Fundi cherehani Missiontown Biashara Mwanamuziki Mwalimu wa kifaransa Nyingine: . .. .. .. .. .. .. .. .. .. .. Sijui b. Wanawake wa kikongo Kinyozi Fundi cherehani Missiontown Biashara Mwanamuziki Mwalimu wa kifaransa Nyingine: . .. .. .. .. .. .. .. .. .. .. Sijui c. Wanaume wa kirundi Kinyozi Fundi cherehani Missiontown Biashara Mwanamuziki Mwalimu wa kifaransa Nyingine: . .. .. .. .. .. .. .. .. .. .. Sijui d. Wanawake wa kirundi Kinyozi Fundi cherehani Missiontown Biashara Mwanamuziki Mwalimu wa kifaransa Nyingine: . .. .. .. .. .. .. .. .. .. .. Sijui e. Wanaume wa kinyarwanda Kinyozi Fundi cherehani Missiontown Biashara Mwanamuziki Mwalimu wa kifaransa Nyingine: . .. .. .. .. .. .. .. .. .. .. Sijui f.Wanawake wa kinyarwanda Kinyozi Fundi cherehani Missiontown

PAGE 391

375 4. Kwa maoni yako unafikiri wakimbizi wengi sana Tanzania wametoka nchi gani? Burundi Rwanda Congo Nyingine: . .. .. .. .. .. .. .. Sijui 5. Kwa maoni yako unafikiri wakimbizi wengi sana hapa Dar es Salaam wametoka nchi gani? Burundi Rwanda Congo Nyingine: . .. .. .. .. .. .. .. Sijui 6. Ungejisikiaje kama binti yako angeolewa na mwanamume mkimbizi? Ningefurahi sana Nisingependelea Nyingine. .. .. .. Ningekubali Nisingejali Sijui b. Je, ungependa binti yako aolewe na mkimbizi toka katika taifa gani? Burundi Rwanda Congo Nyingine: . .. .. .. .. .. .. .. Sijui 7. Ungejisikiaje kama kijana wako wa kiume angeoa na mwanamke mkimbizi? Ningefurahi sana Nisingependelea Nyingine. .. .. .. Ningekubali Nisingejali Sijui b. Je, ungependa kijana wako aoe na mkimbizi toka katika taifa gani? Burundi Rwanda Congo Nyingine: . .. .. .. .. .. .. .. Sijui 8. Ni tabia/mwenendo gani hasa unaoupendelea kwa . .? Wakongo : . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Sijui Warundi : . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Sijui Wanyarwanda : . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Sijui 9. Ni tabia/mwenendo gani hasa usingependelea/furahia kwa . .? Wakongo : . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Sijui Warundi : . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Sijui Wanyarwanda : . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Sijui 10. Ni tabia/mwenendo gani hasa unaotofautisha wanaume wa kitanzania na wale wa . .? Kongo : . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Sijui Burundi : . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Sijui Rwanda : . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Sijui 11. Ni tabia/mwenendo gani hasa unaotofautisha wanawake wa kitanzania na wale wa . .? Kongo : . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. . Sijui Burundi : . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Sijui Rwanda : . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Sijui ______________________________________________________________________________ Date of the Interview: . ../. .../. ... Name of the Interviewer: . .. .. .. .. .. .. .. Place where the Interview took place: . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. Impressions of the Interview : Was the respondent cooperative or reluctant ? Other Comments : . .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. .. ..

PAGE 392

APPENDIX C SOCIAL NETWORKS SURVEY

PAGE 393

_______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 377Name: ________________ Nationality (*): __________________ Registered as Refugee: (Y/N)_______ Age: ______Gender: ________ Town where you lived ?: ________________ Arrival Date Tanzania : ___________ Married (*): __________________ From (year) __________ until (year)_______ Time Stayed at Refugee Camp?_____ Education Level (*): ___________ Job(s) (*) : __________________________ Arrival Date Dar es Salaam: ________ # Children living with you: ______ Ages/Gender: ________________________ Residence of Children not living # Children not living with you: ___ Ages/Gender: ________________________ with you in Dar es Salaam ? _________ How can we contact you ?: ______________________________________ Language used at Interview: _________ Date of the Interview: ______________ 1. Why did you decide to leave your home country ? _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 2. Why did you decide to come to Dar es Salaam instead of the refugee camps _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 3. What are the main difference between your life here and how your life was at home ? _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 4. What are the circumstances under which you would/will return home or do you wish to remain here ? _______________________________________________________________________________________________________

PAGE 394

10 3785. Can you give me the names of the 10 persons (true or false) that have helped you from the moment when you arrived in Dar es Salaam upto now ? They can include any type of help (hali na mali), any nationality, any place of residence, male or female. There is no need to use people's real names. # Name M/F Age Nationality/Ethni-city (*) Mar-ried? (*) Num-ber of Child-ren Resi-dence Job(s) (*) Level of Educa-tion (*) How did you meet, through whom ? (*) Did you know before/ after Arrival DSM ? How long have you known ? (*) # Times you have contact ? (*) How did this person help you exactly ? In what way concretely ? 1 2 3 4 5 6 7 8 9

PAGE 395

3791. Often, before taking an important decision, people like to discuss it first with other people. Over the past 6 months, who have you discussed your decisions with before taking them ? # Name M/F Age Nationality (*) Mar-ried (*) # child-ren Resi-dence Job(s) (*) Level of edu-cation (*) How did you meet ? (*) Did you know before coming to Dar ? How long have you known? (*) Frequency of contacts ? (*) 1 2 2. Suppose you need a large sum of money. Is there anyone you know in Dar whom you could ask to borrow it from ? 1 2 3. Some days things are not going well, and you may feel depressed. When you feel depressed, who do you go to to find consolation ? 1 2 4. In the evenings or in the weekend, when you go out to see a movie, or have some drinks, and such, who are the three persons with whom you most often go out ? 1 2 3

PAGE 396

3805. Do you know of any Congolese/Burundese/Rwandese grassroots organizations here in Dar es Salaam. Are you member of one of these organizations, or do you participate in their meetings ? _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 6. Do you go to church/the mosque ? How often ? Which church do you go to ? Is it a place where you meet a lot of friends ? _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 7. Do you find it more easy or more difficult to establish social relations with people here compared with at home ? Why ? _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 8. In which type of situation will people use the word mkimbizi ("refugee") here ? Does it happen often that you are called mkimbizi ? _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 9. What is your major worry here in Dar es Salaam ? Have you ever been arrested by the Immigration Police ? _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 10. Nakushukuru sana/Thank you very much.

PAGE 397

_______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 381Nom: ______________________ Nationalit (*): _______________________ Enregistr comme rfugi: (O/N)____ Age: ______ Gender: _________ Ville ou vous habitiez ?: ________________ Date d'arrive en Tanzanie :________ Mari (*): ___________________ De (anne) _________jusque (anne)______ Priode vcue au camp ?___________ Niveau d'ducation (*): _________ Job(s) (*) : __________________________ Date d'arrive a Dar es Salaam:______ Nombre d'enfants avec vous: ______ Ages/Gender: ________________________ Domicile des enfants qui ne vivent pas Nombre d'enfants qui ne sont pas avec vous: ___ Ages/Gender: _______________ avec vous Dar es Salaam ? _______ Comment peut-on vous contacter ?___________________________________ Langue de l'interview: ____________ Date de l'interview: ______________ 1. Pour quelle(s) raison(s) avez vous quitt votre pays maternel ? _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 2. Pourquoi est-ce que vous avez dcid de venir Dar es Salaam et non les camps de rfugis ? _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 3. Quelle est la diffrence la plus importante entre votre vie ici et votre vie au pays maternel ? _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 4. Quelles sont les circonstances sous lesquelles vous rentreriez au pays, ou est-ce vous prferez de rester ici ? _______________________________________________________________________________________________________ _______________________________________________________________________________________________________

PAGE 398

3825. Est-ce que vous pouvez me donner les noms de 10 personnes qui vous ont aid lorsque vous quittiez votre pays, lorsque vous arriviez ici Dar es Salaam, et mme ce moment-ci ? Il s'agit de n'importe quel genre d'aide, n'importe quelle nationalit ou domicile, homme ou femme. # Nom H/F Age Nationalit/Ethni-cit (*) Mari (*) Nom-bre d'en-fants Domicile Job(s) (*) Niveau d'du-cation (*) Comment vous avez conn ? Par quel personne? (*) Conn avant ou aprs DSM ? Vous la connaissez depuis? (*) Nombre de contactes ? (*) Dans quel sens exactement cette personne vous a aid ? La situation concrte ? 1 2 3 4 5 6 7 8 9 10

PAGE 399

3 3831. Souvent, avant de prendre des dcisions trs importantes, les gens aiment les discuter d'avance avec d'autres personnes. Pendant les six dernires mois, est-ce que vous avez discut vos dcisions importantes ? Avec qui ? # Nom H/F Age Nationalit (*) Mari (*) # d'en fants Domicile Job(s) (*) Niveau d'du-cation (*) Comment vous avez conn ? (*) Conn avant ou aprs DSM ? Vous la connaissez depuis? (*) Nombre de contactes 1 2 2. Supposons que vous avez besoin d'un grande somme d'argent. A qui est-ce vous pourriez demander de vous l'emprunter ? 1 2 3. Il y a de ces priodes o les choses vont mal, et que vous vous sentez dprim cause des problmes. Avec qui est-ce vous discutez ce genre de problmes 1 2 4. Les soirs, et les weekends, de temps en temps vous sortez allez voir un film, boire un verre, aller danser et vous amuser. Les trois personnes avec qui vous sortez le plus ? 1 2

PAGE 400

3845. Est-ce que vous connaissez des associations communautaires congolaises/burundaises/rwandaises ici Dar es Salaam. Est-ce que vous frquentez une de ces associations ? _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 6. Est-ce que vous allez l'glise ? Souvent ? Quelle glise est-ce que vous frquentez ? Est-ce que c'est un endroit o vous rencontrez beaucoup d'amis ? _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 7. Est-ce que c'est plus facile ou plus difficile d'tablir des rlations amicales avec des gens que vous rencontrer ici qu'au pays ? Pourquoi ? _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 8. Dans quelles situations est-ce qu'on utilise le mot mkimbizi ici ? Est-ce que cela vous arrive souvent d'tre appel mkimbizi ? _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 9. Quel est votre souci le plus grand ou le plus important ici Dar es Salaam ? Avez-vous dj t arrt par la police /immigration ? _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ _______________________________________________________________________________________________________ 10. Nakushukuru sana.

PAGE 401

385 LIST OF REFERENCES Abu-Lughod, Lila 1993 Writing Women's Worlds: Bedouin Stories. Berkeley, CA: University of California Press. Adelman, Howard 2001 From Refugees to Forced Migration: The UNHCR and Human Security. International Migration Review, 35(1):7-28. Adelman, Howard and John Sorenson, eds. 1994 African Refugees: Development Aid and Repatriation. Boulder, CO: Westview Press. African Rights 1995 Rwanda Not So Innocent: When Women Become Killers. London: African Rights. Ager, Alistair, ed. 1999 Refugees: Perspectives on the Experience of Forced Migration. New York: Pinter. Ager, Alistair, with Wendy Ager and Lynellyn Long 1995 The Differential Experience of Mozambican Refugee Women and Men. Journal of Refugee Studies 8(3): 265-287. Agresti, Alan and Barbara Finlay 1997 Statistical Methods of the Social Sciences. Third Edition. Upper Saddle River, NJ: Prentice Hall. Aguilar, Mario I. 1998 Gerontocratic, Aesthetic and Political Models of Age. In The Politics of Age and Gerontocracy in Africa. Mario I. Aguilar, ed. Pp. 3-30. Trenton, NJ: Africa World Press, Inc. Ahearn, Frederick L., ed. 2000 Psychosocial Wellness of Refugees. Studies in Forced Migration Volume 7. New York: Berghahn Books.

PAGE 402

386 Ajayi, J. F. Ade, ed. 1990 UNESCO General History of Africa. Abridged Edition. VII. Africa under Colonial Domination 1880-1935. Berkeley, CA: University of California Press. 1998 UNESCO General History of Africa. Abridged Edition. VI. Africa in the Nineteenth Century until the 1880s. Berkeley, CA: University of California Press. Amadiume, Ifi 1987 Male Daughters, Female Husbands: Gender and Sex in an African Society. London: Zed Books. Amid-Talai, Vered 1997 In Precarious Motion: from Territorial to Transnational Cultures. The Canadian Review of Sociology and Anthropology 34(3):319-333. Amnesty International 1997a Rwanda: Thousands Killed while the World Remains Silent. Electronic document. Http://web.amnesty.org/ai.nsf/print/AFR470371997?OpenDocument, accessed November 15, 2002. 1997b Great Lakes Region: Refugee Crisis Far from Over. Electronic document: http://web.amnesty.org/ai.nsf/index/AFR020091997?OpenDocument&of=THEMES\REFUGEES, accessed October 1, 2002. 1998 Rwanda: "Disappearances" and Killings Continue on a Daily Basis. Electronic document. Http://web.amnesty.org/ai.nsf/print/AFR470261998?OpenDocument, accessed November 15, 2002. 2000a Democratic Republic of Congo: Killing Human Decency. Electronic Document. Http://www.web.amnesty.org/ai.nsf/index/AFR620072000, accessed October 30, 2002. 2000b Great Lakes Region: Refugees Denied Protection. Electronic document: http://web.amnesty.org/ai.nsf/index/AFR020022000?OpenDocument&of=THEMES\REFUGEES, accessed October 1, 2002. 2001 Report 2001: Rwanda. Electronic Document. Http://web.amnesty.org/web/ar2001.nsf/webafrcountries/RWANDA?OpenDocument, accessed November 15, 2002 2002 Report 2002. Oxford, UK: The Alden Press. Anderson, Benedict 1991 Imagined Communities: Reflections on the Origin and Spread of Nationalism. New York: Verso. An-Na'im, Abdullahi and Jeffrey Hammond 2002 Cultural Transformation and Human Rights in African Societies. In Cultural Transformation and Human Rights in Africa. Abdullahi An-Na'im and Jeffrey Hammond, eds. Pp. 13-37. New York: Zed Books.

PAGE 403

387 Anstey, Roger 1970 Belgian Rule in the Congo and the Aspirations of the 'Evolu' class. In Colonialism in Africa 1870-1960. Volume 2. Peter Duignan and L.H. Gann, eds. Pp. 194-225. New York: Cambridge University Press. 1997 Review of Mistrusting Refugees. American Anthropologist 99(3):664. Antonucci, Toni C., and Hiroko Akiyama 1987 Social Network in Adult Life and a Preliminary Examination of the Convoy Model. Journal of Gerontology 42(5):519-527. Bakaji, Mbantu Madila Benyi 1993 The Contemporary Thinking, forms and Practices Concerning Refugees in Africa. A Critique. Master's Dissertation. Dar es Salaam, Tanzania: Department of Sociology, University of Dar es Salaam. Balati, Albert Aidan 2001 Total War in Burundi: a Realist View on the Scramble for Land and Strategic Minerals. Master's thesis. Dar es Salaam, Tanzania: Department of Political Sciences, University of Dar es Salaam. Basok, Tanya 1988 Local Settlement of El Salvadorian Refugees in Costa Rica. PhD thesis. North York, Canada: York University. 1993 Too Much or Too Little: Urban Employment Projects for Salvadorean Refugees in Costa Rica. In Refugee Empowerment and Organizational Change: A Systems Perspective. Peter W. Van Arsdale, ed. Pp. 99-118. Arlington, VA: American Anthropological Association. Beggs, John J. with Valerie A. Haines and Jeanne S. Hurlbert 1996 Situational Contingencies Surrounding the Receipt of Informal Support. Social Forces 75(1):201-222. Beiser, M., with Turner, R.J., and Ganesan, S. 1989 Catastrophic Stress and Factors Affecting Its Consequences Among Southeast Asian Refugees. Social Science and Medicine 28(3): 183-195. Bernardi, Bernardo 1985 Age Class Systems: Social Institutions and Politics Based on Age. New York: Cambridge University Press. Biaya, Tshikala K. 2001 Parallel Society in the Democratic Republic of Congo. In Shifting African Identities. Simon Bekker, Martine Dodds and Meshack Khosa, eds. Pp. 43-60. Pretoria, South Africa: Human Sciences Research Council. Bisharat, George E.

PAGE 404

388 Black, Richard 1998a Putting Refugees in Camps. Forced Migration Review 1(2):4-7. 1998b Refugee Camps not Really Reconsidered: a Reply to Crisp and Jacobson. Forced Migration Review 1(3):31. Networks 21(4):361-373. Bloch, Alice 1999 Carrying Out a Survey of Refugees: Some Methodological Considerations and Guidelines. Journal of Refugee Studies 12(4): 367-383. Bonte, Pierre 1985 Structures d'Age, Organisation Familiale et Systmes de Parent en Afrique de l'Est. In Age, Pouvoir et Socit en Afrique Noire. Abls, Marc and Chantal Collard, eds. Pp. 57-89. Paris: Karthala. Bott, Elizabeth 1957 Family and Social Network. London: Tavistock Publications. Bottinelli, M.C., with Maldonado, I., Troya, E., Herrera, P., and Rodriguez, C. 1990 Psychological Impacts of Exile: Salvadoran and Guatemalan Families in Mexico. Washington, DC: Center for Immigration Policy and Refugee Assistance. Boussevain, Jeremy and J. Clyde Mitchell, eds. 1973 Network Analysis: Studies in Human Interaction. Paris: Mouton & Co. Boyd, Monica 1989 Family and Personal Networks in International Migration: Recent Developments and New Agendas. International Migration Review 23(3):638-670. Bracken, Patrick with Joan E. Giller and Derek Summerfield 1997 Rethinking Mental Health Work with Survivors of Wartime Violence and Refugees. Journal of Refugee Studies 10(4):431-442. Braeckman, Colette 1992 Le Dinosaure: le Zaire de Mobutu. Paris: Librairie Arthme Fayard. 1996 Terreur Africaine.Burundi, Rwanda, Zaire: les Racines de la Violence. Paris: Librairie Arthme Fayard. 1999 L'Enjeu Congolais: L'Afrique Centrale aprs Mobutu. Paris: Librairie Arthme Fayard. Brewer, Devon D. 2000 Forgetting in the Recall-based Elicitation of Personal and Social Networks. Social Networks 22:29-43. Brewer, Devon D. and Cynthia M. Webster 2000 Forgetting of Friends and its Effects on Measuring Friendship Networks. Social

PAGE 405

389 Bureau International du Travail (BIT) and United Nations High Commissioner for Refugees (UNHCR) 1987. L'Intgration Socio-Economique des Rfugis Urbains et Semi-Urbains au Costa-Rica. Genve, Switzerland: UNHCR. Camino, Linda A. and Ruth M. Krulfeld, eds. 1994 Reconstructing Lives, Recapturing Meaning: Refugee Identity, Gender and Culture Change. New York: Gordon and Bleach Publishers. Campbell, Karen E. and Barrett A. Lee 1991 Name Generators in Surveys of Personal Networks. Social Networks 13:203-221. Campbell, Patricia J. and Peter W. Van Arsdale 1994 Preventing the Consequences of Human Rights Abuse: The Case of Refugee Women. In Selected Papers on Refugee Issues: III. Jeffery L. MacDonald and Amy Zaharlick. Pp.31-45. Arlington, VA: American Anthropological Association. Canning, Kathleen 1994 Feminist History after the Linguistic Turn: Historicizing Discourse and Experience. Signs, Winter 1994: 368-404. Castles, Stephen 1998 Globalization and Migration: Some Pressing Contradictions. International Social Science Journal 50(2):179-187. Catherine, Lucas 1994 Manyiema: De enige oorlog die Belgi won. Antwerp, Belgium: Uitgeverij Hadewijch/BRTN/VAR. Center for the Study of Forced Migration (CFSFM) 2001 Review of Refugee Related Policies and Laws in Tanzania. Internal document. Chambers, Robert 1979 Rural Refugees in Africa: What the Eye Does Not See. Disasters 3(4): 381-392. 1986 Hidden Losers? The Impact of Rural Refugees and Refugee Programs on Poorer Hosts. International Migration Review 20(2): 245-263. Chrtien, Jean-Pierre 2000 L'Afrique des Grands Lacs: Deux Mille Ans d'Histoire. Paris: Aubier. Cobham, Rhonda 1992 Misgendering the Nation: African Nationalist Fictions and Nuruddin Farah's Maps. In Nationalism and Sexualities. Andrew Parker, Mary Russo, Doris Sommer and Patricia Yaeger, eds. Pp. 42-59. New York: Routledge.

PAGE 406

390 Codere, Helen 1993 Perception of Caste in Rwanda 1959-1960. In Liber Amicorum Marcel d'Hertefelt: Anthropological Essays. Patrick Wymeersch, ed. Pp. 93-113. Brussels, Belgium: Institut Africain. Collins, Robert O., ed. 1968 Problems in African History. Englewood Cliffs: NJ: Prentice-Hall, Inc. Collinson, Sarah 1999 Globalization and the Dynamics of International Migration: Implications for the Refugee Regime. New Issues in Refugee Research, UNHCR Working Paper No1. Electronic document, http://www.unhcr.ch/refworld/pub/wpapers/wpno1.htm, accessed August 20, 2000. Crisp, Jeff 1999 Policy Challenges of the New Diasporas: Migrant Networks and Their Impact on Asylum Flows and Regimes. New Issues in Refugee Research, Working Paper No 7. Electronic Document. Http://www.unhcr.ch, accessed October 1, 2002. Crisp, Jeff And Karen Jacobson 1998 Refugee Camps Reconsidered. Forced Migration Review 1(3):27-30. Croll, Elisabeth and David Parkin, eds. 1992 Bush Base, Forest Farm: Culture, Environment and Development. New York: Routledge. Cuny, Frederick C. 1977 Refugee Camps and Camp Planning: The State of the Art. Disasters 1(2):125-143. Custers, Raf 2001 Mayi-Mayi: partizanen in Kivu. Electronic Document. Http://www.uitpers.be/archief/nr17/mayi.html, accessed November 11, 2002. Dacyl, Janina W. 1996 Sovereignty versus Human Rights: From Past Discourses to Contemporary Dilemmas. Journal of Refugee Studies 9(2): 136-165. Daley, Patricia 1991 Gender, Displacement and Social Reproduction: Settling Burundi Refugees in Western Tanzania. Journal of Refugee Studies 4(3): 248-266. 1993 From the Kipande to the Kibali: the Incorporation of Refugees and Labour Migrants in Western Tanzania, 1900-87. In Geography and Refugees: Patterns and Processes of Change. Richard Black and Vaughan Robinson, ed. Pp. 17-32. New York: Bellhaven Press.

PAGE 407

391 2001 Population Displacement and the Humanitarian Aid Regime: The Experience of Refugees in East Africa. In Mobile Africa: Changing Patterns of Movement in Africa and Beyond. Mirjam de Bruijn, Rijk van Dijk and Dick Foeken, eds. Pp. 195-211. Boston: Brill. Daniel, E. Valentine and John Chr. Knudsen, eds. 1995 Mistrusting Refugees. Berkeley CA: University of California Press. Davidson, Basil 1992 Black Man's Burden: Africa and the Curse of the Nation-State. Oxford: James Curry. Davis, J. 1992 The Anthropology of Suffering. Journal of Refugee Studies 5(2): 149-161. De Witte, Ludo 1999 De moord op Lumumba. Leuven, Belgium: Uitgeverij Van Halewyck. Des Forges, Alison 1995 The Ideology of Genocide. Issue: A Journal of Opinion 23(2): 44-47. di Leonardo, Micaela, ed. 1991 Gender at the Crossroads of Knowledge: Feminist Anthropology in the Postmodern Era. Berkeley, CA: University of California Press. Drew, Allison 1995 Female Consciousness and Feminism in Africa. Theory and Society 24:1-33. East African School on Refugee and Humanitarian Affairs (EASRHA) n.d. Historical Background and Objectives of EASRHA. Dar es Salaam, Tanzania: University of Dar es Salaam. Eger, Bertil 1979 Colonization and Migration: A Summary of Border-Crossing Movements in Tanzanian before 1967. Research Report no. 52. Uppsala, Sweden: Scandinavian Institute for of African Studies. Ellis, Sue and Sultan Barakat 1996 From Relief to Development: The Long-term Effects of 'Temporary Accomodation on Refugees and Displaced Persons in the Republic of Croatia. Disasters 20(2): 111-124. Escalona, Ana and Richard Black 1995 Refugees in Western Europe: Bibliographic Review and State of the Art. Journal of Refugee Studies 8(4): 365-389.

PAGE 408

392 Express Online 1999 Congo Influx Chokes Refugee Camps. Electronic document. http://www.theexpress.com/express128/news/news2.htm, accessed December 10, 1999. Robert E. Krieger Publishing Company. Fabian, Johannes 1986 Language and Colonial Power: The Appropriation of Swahili in the Former Belgian Congo 1880-1938. New York: Cambridge University Press. Facts on File (FOF) 1997 Facts on Kabila. Electronic Document. Http://www.facts.com/wnd/kabila.htm, accessed October 31, 2002. Farah, Nurrudin 1986 Maps. New York: Pantheon Books. Ferguson, James 1993 De-Moralizing Economies: African Socialism, Scientific Capitalism and the Moral Politics of "Structural Adjustment." In Moralizing States and the Ethnography of the Present. Sally Falk Moore, ed. Pp. 78-92. Arlington, VA: American Anthropological Association. Fischer, Claude S. 1982 To Dwell among Friends: Personal Networks in Town and City. Chicago, IL: University of Chicago Press. Foner Anne and David Kertzer 1978 Transitions over the Life Course: Lessons from Age-Set Societies. American Journal of Sociology 83(5): 1081-1104. Forster, Peter G., with Michael Hitchcock and Francis F.Lyimo 2000 Ethnic Factors in National Integration. In Race and Ethnicity in East Africa. Forster, Peter G., Michael Hitchcock and Francis F.Lyimo, eds. Pp. 94-119. London: McMillan Press ltd. Freeman, Linton C., with Douglas R. White and A. Kimball Romney, eds. 1989 Research Methods in Social Network Analysis. Fairfax, VA: George Mason University Press. Gahama, Joseph 2001 Le Burundi sous Administration Belge. La priode du Mandat 1919-1939. 2nd edition. Paris: Editions Karthala. Gailey, Harry A. Jr. 1981[1970] History of Africa. Volume I: From Earliest Times to 1800. Malabar, FL:

PAGE 409

393 Gasarasi, Charles P. 1987 The Tripartite Approach to the Resettlement and Integration of Rural Refugees in Tanzania. In Refugees: A Third World Dilemma. John R. Rogge, ed. Pp. 99-113. Totowa, NJ: Rowman & Littlefield. 1990 The Mass Naturalization and Further Integration of Rwandese Refugees in Tanzania: Processes, Problems and Prospects. Journal of Refugee Studies 3(2): 88-109. Geertz, Clifford 1973 The Interpretation of Cultures: Selected Essays. New York: Basic Books. Geiger, Susan 1992 What's so Feminist about Doing Women's Oral History? In Expanding the Boundaries of Women's History: Essays on Women in the Third World. Cheryl Johnson-Odim and Margaret Strobel, eds. Pp. 305-318. Bloomington, IN: Indiana University Press. Geiger, Vance 1993 Refugee Cognitive Expectation and Sociocultural Change Theory. In Selected Papers on Refugee Issues: II. MaryCarol Hopkins and Nancy D. Donnelly, eds. Arlington, VA: American Anthropological Association. Geuijen, Karin 1997 Review of Mistrusting Refugees. Journal of the Royal Anthropological Institute 3(3):625-626. Gluck, Sherna Berger and Daphne Patai, eds. 1991 Women's Words: The Feminist Practice of Oral History. New York and London: Routledge. Goitom, Eyob 1987 Systems of Social Interaction of Refugee Adjustment Processes: The Case of Eritrean Refugees in Khartoum, Sudan. In Refugees: A Third World Dilemma. John R. Rogge, ed. Pp. 130-142. Totowa, NJ: Rowman & Littlefield Gold, Steven J. 1992 Refugee Communties: A Comparative Field Study. London: Sage Publications. Gould, W.T.S. 1974 International Migration in Tropical Africa: a Bibliographical Review. International Migration Review 8(3):347-365. Government of Tanzania (GOT) 1966 The Refugees Control Act, 1965. Government Notice N 89 published 11/3/66. Dar es Salaam, Tanzania.

PAGE 410

394 1973 The Immigration Act, 1972. Government Notice N 95 published 25/5/73. Dar es Salaam, Tanzania. 1995 Immigration Act, 1995. Electronic document. Http://www.unhcr.ch, accessed July 7, 2002. 1998 Refugees Act, 1998. Electronic document. Http://www.unhcr.ch, accessed July 7, 2002. Goyvaerts, Didier 2000 About Bantu and Nilotes. In Conflict and Ethnicity in Central Africa. Didier Goyvaerts, ed. Pp. 301-304. Tokyo, Japan: Institute for the Study of Languages and Cultures of Asia and Africa, Tokyo University of Foreign Studies. Gozdziak, Elzbieta M. and Dianna J. Shandy 2002 Editorial Introduction: Religion and Spirituality in Forced Migration. Journal of Refugee Studies 15(2): 129-135. Grosz-Ngat, Maria 1989 Hidden meanings: Explorations into a Bamanan Construction of Gender. Ethnology XXVIII(2):167-183. 1997 Introduction. In Gendered Encounters: Challenging Cultural Boundaries and Social Hierarchies in Africa. Maria Grosz-Ngat and Omari H. Kokole, eds. Pp. 1-21. New York: Routledge. Gugler, Josef and Gudrun Ludwar-Ene 1995 Gender and Migration in Africa South of the Sahara. In The Migration Experience in Africa. Jonathan Baker and Tade Akin Aina, eds. Pp. 257-268. Oslo, Norway: Nordiska Afrikainstitutet. Gupta, Akhil and James Ferguson 1997 Anthropological Locations: Boundaries and Grounds of a Field Science. Berkeley, CA: University of California Press. Hafkin, Nancy and Edna Bay, eds. 1974 Women in Africa: Studies in Social and Economic Change. Palo Alto, CA: Stanford University Press. Haines, Valerie A. with Jeanne S. Hurlbert and John J. Beggs 1996 Exploring the Determinants of Support Provision: Provider Characteristics, Personal Networks, Community Contexts, and Support Following Life Events. Journal of Health and Social Behavior 37:252-264. Hannerz, Ulf 1980. Exploring the City: Inquiries Towards an Urban Anthropology. New York: Columbia University Press.

PAGE 411

395 Hansen, Art 1979 Once the Running Stops: Assimilation of Angolan Refugees into Zambian Border Villages. Disasters 3(4): 369-74. 1981. Refugee Dynamics: Angolans in Zambia 1966-1972. International Migration Review 15(1): 175-94. 1992. "Some Insights on African Refugees," in Selected Papers on Refugee Issues. Edited by Pamela A. DeVoe, Pp. 100-109. Arlington, VA: American Anthropological Association. Hansen, Art and Anthony Oliver-Smith, eds. 1982 Involuntary Migration and Resettlement: The Problems and Responses of Dislocated People. A Westview Special Study. Boulder, CO: Westview Press. Hansen, Karen 1992 Introduction: Domesticity in Africa. In African Encounters with Domesticity. Karen Hansen, ed. Pp. 1-33. New Brunswick, NJ: Rutgers University Press. Harrell-Bond, Barbara. E. 1986 Imposing Aid: Emergency Assistance to Refugees. Oxford: Oxford University Press. 1998 Camps: Literature Review. Forced Migration Review 1(2):22-23. Harrell-Bond, Barbara. and Eftihia Voutira 1992 Anthropology and the Study of Refugees. Anthropology Today 8(4): 6-10. Hathaway, James C. 1991 Reconceiving Refugee Law as Human Rights Protection. Journal of Refugee Studies 4(2): 113-131. 1995 New Directions to Avoid Hard Problems: The Distortion of the Palliative Role of Refugee Protection. Journal of Refugee Studies 8(3): 288-304. Hein, Jeremy 1993 Refugees, Immigrants and the State. Annual Review of Sociology 19:43-60. Hochschild, Adam 1998 King Leopold's Ghost : a Story of Greed, Terror, and Heroism in Colonial Africa. Boston : Houghton Mifflin. Hodgson, Dorothy L. and Sheryl A. McCurdy, eds. 2001 "Wicked" Women and the Reconfiguration of Gender in Africa. Portmouth, NH: Heinemann Holmboe-Ottesen, Gerd and Margareta Wandel 1991 Men's Contribution to the Food and Nutritional Status in the Tanzanian Household. Ecology of Food and Nutrition 26:83-96.

PAGE 412

396 Human Rights Watch (HRW) 2000a Seeking Protection: Addressing Sexual and Domestic Violence in Tanzania's Refugee Camps. New York: Human Rights Watch. 2000b Eastern Congo Ravaged: Killing Civilians and Silencing Protest. Electronic Document. Http://www.hrw.org/reports/2000/drc/Drc005.htm#TopOfPage, accessed October 29, 2002. 2000c The Response of UNHCR. Electronic Document. Http://www.hrw.org/reports/2000/tanzania/Duhweb-10.htm, accessed October 31, 2002. 2000d Arbitrary Detentions, Torture and Ill-Treatment. Electronic Document. Http://www.hrw.org/reports/2000/rwanda/Rwan004-11.htm, accessed November 15, 2002. 2002 Hidden in Plain View: Refugees Living Without Protection in Nairobi and Kampala. New York: Human Rights Watch. Hunt, Nancy Rose 1990 Domesticity and Colonialism: Usumbura's Foyer Social, 1946-1960. Signs 15(3): 447-74. Hyden, Goran 1980 Beyond Ujamaa in Tanzania: Underdevelopment and an Uncaptured Peasantry. Los Angeles: University of California Press. 1983 No Shortcuts to Progress: African Development Management in Perspective. London: Heinemann. Iliffe, John 1979 A Modern History of Tanganyika. New York: Cambridge University Press. Indra, Doreen Marie 1989 Ethnic Human Rights and Feminist Theory: Gender Implications for Refugee Studies and Practice. Journal of Refugee Studies 2(2): 221-242. Indra, Doreen Marie, ed. 1998 Engendering Forced Migration. New York: Berghahn Books. International Rescue Committee 2001 Mortality Study Eastern D.R.Congo (February April 2001). Electronic Document. Http://www.theirc.org/index.cfm, accessed October 2, 2002. Jahan, Rounaq 1995 The Elusive Agenda: Mainstreaming Women in Development. Dhaka, Bangladesh: University Press Limited.

PAGE 413

397 Jayaratine, Toby Epstein and Abigail J. Stewart 1991 Quantitative and Qualitative Methods in the Social Sciences: Current Feminist Issues and Practical Strategies. In Beyond Methodology: Feminist Scholarship as Lived Research. Mary Margaret Fonow and Judith A. Cook, eds. Pp. 85-106. Salaam. Bloomington, IN: Indiana University Press. Jefremovas, Villia 1995 Acts of Human Kindness: Tutsi, Hutu and the Genocide. Issue: A Journal of Opinion 23(2): 28-31. Johnson, Jeffrey C. 1994 Anthropological Contributions to the Study of Social Networks: a Review. In Advances in Social Network Analysis: Research in the Social and Behavioral Sciences. Stanley Wasserman and Joseph Galaskiewicz, eds. Pp. 113-151. London: Sage Publications. Karadawi, Ahmed 1987 The Problem of Urban Refugees in Sudan. In Refugees: A Third World Dilemma. John R. Rogge, ed. Pp. 115-129. Totowa, NJ: Rowman & Littlefield Keith, Jennie and David Kertzer 1984 Introduction. In Age and Anthropological Theory. David I. Kertzer and Jennie Keith, eds. Pp. 19-61. Ithaca, NY: Cornell University Press. Kelley, Kevin J. 2002 New Book Tells of Nyerere's Support for Cuba in Africa. Electronic document. Http://allafrica.com/stories/printable/200205070115.html, accessed November 11, 2002. Kibreab, Gaim 1989 Local Settlement in Africa: A Misconceived Option? Journal of Refugee Studies 2(4): 468-90. 1991 The State of the Art Review of Refugee Studies in Africa. Uppsala Papers in Economic History, Research Report No 26. Uppsala, Sweden: Uppsala University. 1996 Eritrean and Ethiopian Urban Refugees in Khartoum: What the Eye Refuses to See. African Studies Review 39(3): 131-78. 1999 Revisiting the Debate on People, Place, Identity and Displacement. Journal of Refugee Studies 12(4): 384-428. 2002 Review of Fear in Bongoland: Burundi Refugees in Urban Tanzania. Journal of Refugee Studies 15(3):326-328. Kironde, Lusugga 2000 Rapid Urbanisation in Tanzania: the Government's Coping Strategies. In Urbanising Tanzania: Issues, Initiatives and Priorities. Suleiman Ngware and J.M. Lusugga Kironde. Pp. 22-58. Dar es Salaam, Tanzania: University of Dar es

PAGE 414

398 Knudsen, John Christian 1983 Boat People in Transit: Vietnamese in Refugee Camps in the Philippines, Hongkong and Japan. The Migration Project Studies, Occasional Paper No.31. Bergen, Norway: University of Bergen. 1990 Cognitive Models in Life Histories. Anthropological Quarterly 63(3): 122-133. 1991 Therapeutic Strategies and Strategies for Refugee Coping. Journal of Refugee Studies 4(1): 21-38. Koser, Khalid 1995 Information and Refugee Migration: the Case of Mozambicans in Malawi. Tijdschrift voor Economische en Sociale Geografie 87(5):407-418. 1997 Social Networks and the Asylum Cycle: the Case of Iranians in the Netherlands. International Migration Review 31(4): 591-612. 2002 From Refugees to Transnational Communities? In New Approaches to Migration? Transnational Communities and the Transformation of Home. Nadje Al-Ali and Khalid Koser, eds. Pp. 138-152. New York: Routledge. Kpanou, Tossou Bertin 1989 L'Intgration des Rfugis au Bnin: Cas de Cotonou. Master's thesis. Cotonou, Bnin: University of Cotonou. Kuhlman, Tom 1990 Burden or Boon: A Study of Eritrean Refugees in the Sudan. Amsterdam, the Netherlands: VU University Press. 1994 Organized versus Spontaneous Settlement of Refugees in Africa. In African Refugees: Development aid and Repatriation. Howard Adelman and John Sorenson, eds. Pp. 117-42. Boulder, CO: Westview Press. Landau, Loren B. 2001 The Humanitarian Hangover: Transnationalization of Governmental Practice in Tanzania's Refugee-Populated Areas. New Issues in Refugee Research, UNHCR Working Paper No.40. Electronic Document. Http://www.unhcr.ch/cgi-bin/texis/vtx/home?page=publ, accessed September 25, 2002. Lange, Sri 1999 How the National Became Popular in Tanzania. In National Identity and Democracy in Africa. Mia Palmberg, ed. Pp. 41-58. Oslo, Norway: Nordic Africa Institute. Langness, L. L. and Gelya Frank 1985 Lives: An Anthropological Approach to Biography. Novato, CA: Chandler & Sharp Publishers, Inc.

PAGE 415

399 Leach, Melissa 1992. Dealing with Displacement: Refugee-Host Relations, Food and Forest Resources in Sierra Leonean Mende Communities during the Liberian Influx, 1990-1991. Research Report No 22. Brighton, Sussex UK: Institute for Development Studies. Leacock, Eleanor and Mona Etienne, eds. 1980 Women and Colonization: Anthropological Perspectives. New York: Praeger. Lee, Luke T. 1996 Internally Displaced Persons and Refugees: Toward a Legal Synthesis? Journal of Refugee Studies 9(1): 27-42. Lemarchand, Ren 1970 Rwanda and Burundi. London: Pall Mall Press. 1994 Burundi: Ethnocide as Discourse and Practice. New York: Cambridge University Press. 1997 Patterns of State Collapse and Reconstruction in Central Africa: Reflections on the Crisis in the Great Lakes. African Studies Quartely 1(3). Electronic document, http://www.africa.ufl.edu/asq/v1/3/2.htm, accessed September 19, 2002. Lewis, David 2003 Cow Pats Fuel Computers. BBC News. Electronic Document. Http://newsvote.bbc.co.uk/mpapps/pagetools/print/news.bbc.co.uk/2/hi/technology/2957488.stm, accessed May 6, 2003. Light, Deborah 1992 Healing Their Wounds: Guatemalan Refugee Women as Political Activists. In Women and Therapy 13 (3):297-308. Lin, Nan with Mary W.Woelfel and Stephen C. Light 1985 The Buffering Effect of Social Support Subsequent to an Important Life Event. Journal of Health and Social Behavior 26:247-263. Llewelyn-Davis, Melissa 1996 Women, Warriors and Patriarchs. In Anthropological Theory: An Introductory History. R. Jon McGee and Richard L. Warms, eds. Pp. 414-429. Mountain View, CA: Mayfield Publishing Company. Long, Cynthia 1999 Congolese Refugees Overwhelm Tanzanian Camp. Electronic Document. http://204.146.17.49/Disasters/990722tanzania/index_text.html, accessed December 10, 1999.

PAGE 416

400 Longman, Timothy 2001 Church Politics and the Genocide in Rwanda. Journal of Religion in Africa 23(2):164-186. Louch, Hugh 2000 Personal Network Integration: Transitivity and Homophily in Strong Tie Relations. Social Networks 22:45-64. Lugalla, Joe 1995 Crisis, Urbanization, and Urban Poverty in Tanzania: A Study of Urban Poverty and Survival Politics. New York: University Press of America. 1997 Economic Reforms and Health Conditions of the Urban Poor in Tanzania. Electronic Document. Http://www.africa.ufl.edu/asq/v1/2/2.htm, accessed December 23, 1999. MacGaffey, Janet 1986 Fending-for-Yourself: The Organization of the Second Economy in Zaire. In The Crisis in Zaire: Myths and Realities. Nzongola-Ntalaja, ed. Pp. 141-156.Trenton, NJ: Africa World Press. 1992 Initiatives From Below: Zaire's Other Path to Social and Economic Restructuring. In Governance and Politics in Africa. Goran Hyden and Michael Bratton, eds. Pp. 243-261. Boulder, CO: Lynne Riener Publishers. MacGaffey, Janet and Rmy Bazenguissa-Ganga 2000 Congo-Paris: Transnational Traders on the Margins of the Law. Bloomington, IN: Indiana University Press. MacGaffey, Wyatt 1993 Anthropology of History. In Liber Amicorum Marcel d'Hertefelt: Anthropological Essays. Patrick Wymeersch, ed. Pp. 153-161. Brussels, Belgium: Institut Africain. Malkki, Liisa H. 1995a Purity and Exile: Violence, Memory, and National Cosmology among Hutu Refugees in Tanzania. Chicago, IL: University of Chicago Press. 1995b Refugees and Exile: From "Refugee Studies" to the National Order of Things. Annual Review of Anthropology 24: 495-523. 1996 Speechless Emissaries: Refugees, Humanitarianism, and Dehistoricization. Cultural Anthropology 11(3): 377-404. Mamdani, Mahmood 2001 When Victims Become Killers: Colonialism, Nativism and the Genocide in Rwanda. Dar es Salaam, Tanzania: E & D Limited.

PAGE 417

401 Manahl, Christian R. 2000 From Genocide to Regional War: the Breakdown of International Order in Central Africa. African Studies Quarterly 4(1). Electronic document, http://web.africa.ufl.edu/asq/v4/v4i1a2.htm, accessed August 20, 2001. Marris, Peter 1986 [1974] Loss and Change (Revised Edition). London & New York: Routledge. Marsden, Peter V. 1988 Homogeneity in Confiding Relations. Social Networks 10:57-76. Martin, Susan Forbes 1992 Refugee Women. Atlantic Highlands, NJ: Zed Books. Martin, Susan Forbes, and Emily Copeland 1989 Making Ends Meet: Refugee Women and Income Generation. Washington, DC: Refugee Policy Group. Marx, Emanuel 1990. The Social World of Refugees: A Conceptual Framework. Journal of Refugee Studies 3(3): 189-203. Mayer, Tamar 2000 Gender Ironies of Nationalism: Setting the Stage. In Gender Ironies of Nationalism: Sexing the Nation. Tamar Mayer, ed. Pp. 1-22. New York: Routledge. Mazrui, Ali A., ed. 1990 UNESCO General History of Africa. Unabridged Edition. VIII. Africa since 1935. Berkeley, CA: University of California Press. Mazrui, Jamin M. and Ibrahim Noor Shariff 1994 The Swahili: Idiom and Identity of an African People. Trenton, NJ: African World Press. Mbilinyi, Marjorie 1990 "This Is an Unforgettable Business": Colonial State Intervention in Urban Tanzania. In Women and the State in Africa. J.L. Parpart and K.A. Staudt, eds. Boulder, CO: Lynne Riener Publishers. McAllister, Lynne and Claude S. Fischer 1983 A Procedure for Surveying Personal Networks. In Applied Network Analysis: A Methodological Introduction. Ronald S. Burt and Michael J. Minor, eds. Pp.75-88. Beverly Hills: Sage Publications.

PAGE 418

402 McGovern, Michael 1998 Review of Mistrusting Refugees. American Ethnologist 25(1):17-19. Mikell, Gwendolyn, ed. 1997 African Feminism: The Politics of Survival in Sub-Saharan Africa. Philadelphia, PA: University of Pennsylvania Press. Miserez, Diana, ed. 1988 The Trauma of Exile: the Humanitarian Role of Red Cross and Red Crrescent. Boston: Kluwer Academic Publications. Mitchell, J. Clyde 1969 Social Networks in Urban Situations. Manchester, England: Manchester University Press. 1987 Cities, Society, and Social Perception: A Central African Perspective. Oxford: Clarendon Press. Moore, Henrietta 1988 Feminism and Anthropology. Padstow, Cornwall UK: Polity Press. Moore, Sally Falk 1977 The Chagga and Meru of Tanzania. London: International Africa Institute. 1986 Social Facts and Fabrications : "Customary" Law on Kilimanjaro, 1880-1980. New York: Cambridge University Press. Moser, Caroline O.N. and Fiona C. Clark, eds. 2001 Victims, Perpetrators or Actors? Gender, Armed Conflict and Political Violence. New York: Zed Books. Muecke, Marjorie A. 1992 New Paradigms for Refugee Health Problems. Social Science and Medicine 35(4):515-523. Mukagasana, Yolande 1997 La mort ne veut pas de moi. Paris: Editions Fixot. Mwanasali, Musifiky 2000 The View from Below. In Greed & Grievance: Economic Agendas in Civil Wars. Mats Berdal and David M. Malone, eds. Pp.137-153. Boulder: Lynne Rienner Publishers. Mworoha, Emile, ed. 1987 Histoire du Burundi : des Origines la Fin du XIXe Sicle. Paris: Hatier.

PAGE 419

403 Nederveen Pieterse, Jan. 1990 Wit over Zwart: Beelden van Afrika en Zwarten in de Westerse Populaire Cultuur. Amsterdam, The Netherlands: Koninklijk instituut voor de tropen. Nelder, Brian W. 1979 Settlement of Rural Refugees in Africa. Disasters 3(4): 393-402. Newbury, Catharine 1986 Survival Strategies in Rural Zaire: Realities of Coping with Crisis. In The Crisis in Zaire: Myths and Realities. Nzongola-Ntalaja, ed. Pp. 99-112. Trenton, NJ: Africa World Press. 1992 Rwanda: Recent Debates Over Governance and Rural Development. In Governance and Politics in Africa. Goran Hyden and Michael Bratton, eds. Pp. 193-219. Boulder, CO: Lynne Riener Publishers. 1998 Ethnicity and the Politics of History in Rwanda. Africa Today 45(1):7-24. Newbury, David 1986 From "Frontier" to "Boundary": Some Historical Roots of Peasant Strategies of Survival in Zaire. In The Crisis in Zaire: Myths and Realities. Nzongola-Ntalaja, ed. Pp. 87-98. Trenton, NJ: Africa World Press. Ngware, Suleiman and J. M. Lusugga Kironde, eds. 1996 Urbanising Tanzania: Issues, Inititatives and Priorities. Dar es Salaam, Tanzania: University of Dar es Salaam. Ngaruko, Floribert and Janvier D. Nkurunziza 2000 An Economic Interpretation of the Conflict in Burundi. Journal of African Economics 9(3):370-409. Nyambariza, Daniel 1986 L'Occupation des Terres de la Plaine de Bujumbura par les Swahili, 1933-1935. In Socits et Pouvoirs dans l'Afrique des Grands LLLLacs. Dpartment d'histoire, Facult des lettres et sciences humaines, ed. Pp. 53-80. Bujumbura, Burundi: Universit du Burundi. Nzongola-Ntalaja 2001 Ethnic Identification in the Great Lakes Region. In Shifting African Identities. Simon Bekker, Martine Dodds and Meshack Khosa, eds. Pp. 61-80. Pretoria, South Africa: Human Sciences Research Council. Nzongola-Ntalaja, ed. 1986 The Crisis in Zaire: Myths and Realities. Trenton, NJ: Africa World Press.

PAGE 420

404 O'Barr, Jean and Kathryn Firmin-Sellers 1995 African Women in Politics. In African Women South of the Sahara. Margaret, Jean Hay and Sharon Stichter, eds. Pp. 189-212. New York: Longman Scientific and Technical. Okema, Michael 1996 Political culture of Tanzania. Lewiston, NY: E. Mellen Press. Oliver-Smith, Anthony 1986 The Martyred City: Death and Rebirth in the Andes. Albuquerque, NM: University of New Mexico Press. 1991 Involuntary Resettlement, Resistance and Political Empowerment. Journal of Refugee Studies 4(2):132-149. Omidian, Patricia A. 2000 Qualitative Measures and Refugee Research: the Case of Afghan Refugees. In Psychosocial Wellness of Refugees. Frederick L. Ahearn, ed. Pp. 41-66. New York: Berghahn Books. Oppong, Christine 1983 Female and Male in West Africa. London: George Allen & Unwin. Oyewumi, Oyeronke 1997 The Invention of Women: Making an African Sense of Western Gender Discourses. Minneapolis, MN: University of Minnesota Press. Palmer, Catherine 1998 From Theory to Practice: Experiencing the Nation in Everyday Life. Journal of Material Culture 3(2):175-199. Parker, Andrew, with Mary Russo, Doris Sommer and Patricia Yaeger 1992 Introduction. In Nationalism and Sexualities. Andrew Parker, Mary Russo, Doris Sommer and Patricia Yaeger, eds. Pp. 1-18. New York: Routledge. Parpart, J.L. and Staudt, K.A., eds. 1989 Women and the State in Africa. Boulder, CO: Lynne Riener Publishers. Pateman, Carole 1989 The Disorder of Women: Democracy, Feminism and Political Theory. Palo Alto, CA: Stanford University Press. Phillips, James 1996 The Politics of Refugee Identity in Central America. In Selected Papers on Refugee Issues: IV.Ann M. Rynearson and James Phillips, eds. Pp. 222-240. Arlington, VA: American Anthropological Association.

PAGE 421

405 Portuous, J. Douglas and Sandra E. Smith 2001 Domicide: the Global Destruction of Home. Ithaca, NY: McGill-Queen's University Press. Prunier, Grard 2001 The Catholic Church and the Kivu conflict. Journal of Religion in Africa 23(2)139-162. Reed, Holly 1998 The Demography of Forced Migration: Summary of a Workshop. Washington, DC: National Academic Press. Reno, William 1998 Warlord Politics and African States. Boulder, CO: Lynne Rienner Publishers. Reyntjens, Filip 1996 Danse Macabre: Rwanda en Burundi tussen Haat en Hoop. Antwerp, Belgium: Icarus. 1999 La Guerre des Grands Lacs: Alliances Mouvantes et Conflits Extraterritoriaux en Afrique Centrale. Paris: l'Harmattan. 2000 Small States in an Unstable Region Rwanda and Burundi, 1999-2000. Current African Issues No. 23. Uppsala, Sweden: Nordiska Afrikainstitutet. Roberts, A. 1968 Tanzania before 1900. Dar es Salaam, Tanzania: East African Publishing House. Rogers, Alisdair And Steven Vertovec, eds. 1995 The Urban Context: Ethnicity, Social Networks and Situational Analysis. Washington DC: Berg Publishers. Rogge, John R. 1986 Urban Refugees in Africa: Some Changing Dimensions to Africas Refugee Problem, with Special Reference to Sudan. Migration World XIV (4): 7-14. Roy-Campbell, Zaline Makini 2001 Empowerment through Language. The African Experience: Tanzania and Beyond. Trenton, NJ: Africa World Press. Rumiya, Jean 1992 Le Rwanda sous le Rgime du Mandat Belge (1916-1931). Paris: Editions l'Harmattan.

PAGE 422

406 Rutinwa, Bonaventure 1999 The End of Asylum? The Changing Nature of Refugee Policies in Africa. New Issues in Refugee Research, UNHCR Working Paper No5. Electronic Document. Http://www.unhcr.ch/refworld/pub/wpapers/wpno5.htm, accessed August 10, 2000. 2002 Prima Facie Status and Refugee Protection. New Issues in Refugee Research, UNHCR Working Paper No69. Electronic document. Http://www.unhcr.ch accessed March 10, 2003. Sanders, Todd 2001 Save Our Skins: Structural Adjustment, Morality and the Occult in Tanzania. In Magical Interpretations, Material Realities: Modernity, Witchcraft and the Occult in Postcolonial Africa. Henrietta L. Moore and Todd Sanders, eds. Pp.160-183. New York: Routledge. Sanjek, Roger 1990 Urban Anthropology in the 1980s: A World View. Annual Review Anthropology 19:151-86. Schoenbrun, David Lee 1998 A Green Place, A Good Place: Agrarian Change, Gender, and Social Identity in the Great Lakes Region to the 15th Century. Social History of Africa series. Portsmouth, IL: Heinneman Schwartz, Jonathan 1997 Review of Mistrusting Refugees. Social Anthropology 5(3):336-337. Schweizer, Thomas with Michael Schnegg and Suzanne Berzborn 1998 Personal Networks and Social Support in a Multiethnic Community of Southern California. Social Networks 20:1-21. Scott, Joan 1992 Experience. In Feminists Theorize the Political. Judith Butler and Joan W. Scott, eds. Pp. 22-40. New York: Routledge. Scott, John 2000 Social Network Analysis: a Handbook. London: Sage. Sheriff, Abdul 1987 Slaves, Spices & Ivory in Zanzibar. Athens, OH: Ohio University Press. Sinnerbrink, Ingrid with Derrick Silove, Annette Field, Zachary Steel and Vijaya Manicavasagar 1997 Compounding of Premigration Trauma and Postmigration Stress in Asylum Seekers. The Journal of Psychology 131(5):463-470.

PAGE 423

407 Smith, Anthony 1986 The Ethnic Origin of Nations. Oxford: Blackwell. Smith-Lovin, Lynn and J. Miller McPherson 1993 You Are Who You Know: A Network Approach to Gender. In Theory on Gender / Feminism on Theory. Paula England, ed. Pp.223-251. New York: Aldine De Gruyter. Sommers, Marc 1993 Coping with Fear: Burundi Refugees and the Urban Experience in Dar es Salaam, Tanzania. In Selected Papers on Refugee Issues: II. MaryCarol Hopkins and Nancy D. Donnelly, eds. Pp. 13-25. Arlington, VA: American Anthropological Association. 1994 Hiding in Bongoland: Identity F
Permanent Link: http://ufdc.ufl.edu/UFE0002281/00001

Material Information

Title: Embedding the Refugee Experience: Forced Migration and Social Networks in Dar es Salaam, Tanzania
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0002281:00001

Permanent Link: http://ufdc.ufl.edu/UFE0002281/00001

Material Information

Title: Embedding the Refugee Experience: Forced Migration and Social Networks in Dar es Salaam, Tanzania
Physical Description: Mixed Material
Copyright Date: 2008

Record Information

Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
System ID: UFE0002281:00001


This item has the following downloads:


Full Text












EMBEDDING THE REFUGEE EXPERIENCE :
FORCED MIGRATION AND SOCIAL NETWORKS
IN DAR ES SALAAM, TANZANIA


















By

ROOS WILLEMS


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2003

































Copyright 2003

by

Roos Willems

































This document is dedicated to the refugees in Dar es Salaam, Tanzania.















ACKNOWLEDGMENTS

Many persons have contributed towards the success of my doctoral study. I would

first like to thank the members of my advisory committee, namely, Professors Anthony

Oliver-Smith, Paul Magnarella, Goran Hyden and Maria Grosz-Ngate, for their scholarly

advisement. Dr. Oliver-Smith, my advisor, deserves special mention for his guidance and

help from the inception of proposal writing to the production of this dissertation, as does

Dr. Maria Grosz-Ngate. The staff at the Department of Anthropology, Karen Jones,

Patricia King and Salena Robinson, deserve my deep appreciation for their help with

routine administrative activities.

I owe special thanks to the Department of Anthropology and the Center for African

Studies at the University of Florida for the various graduate assistantships that made it

possible for me to complete my doctoral studies, and to the Wenner-Gren Foundation for

Anthropological Research which funded my field research. In Tanzania, grateful

acknowledgements are due to Dr. Masanja of the Department of Sociology and Dr.

Kamanga, Director of the Center for the Study of Forced Migration at the University of

Dar es Salaam, for sharing their knowledge and facilitating numerous local contacts with

institutions and persons alike. Arjun Jain at UNHCR, Dar es Salaam, was indispensable

in providing information and facilitating contacts with potential informants, as was Dr.

Sinje of Umati at Upanga.

I am very grateful to the many forced migrants in Dar es Salaam who were willing

to share their experiences with me and to introduce me further into the refugee









communities, as well as to the staff of government offices and humanitarian agencies in

Dar es Salaam. While apologizing for not mentioning all the names of individuals, I

would like to express my appreciation for my research assistants Denis, Xavier,

Mopetane, Gaspard, Bati and above all, to Justine Buhanda for her untiring and ever

efficient endeavors, without whom this research project would have been so much more

difficult to complete.

Karen, Jasper and Suzan deserve special mention for providing me with

accommodation and very much needed social contacts during my stay in Dar es Salaam, as

do Frieda and Eva for their companionship and encouragements. Special thanks for

friendship and logistical support go to Lilian Osaki, as well as to Rob Duijs, whose frank

comments on earlier writings were very helpful. Also very much appreciated were the

insights and suggestions from the members of our anthropology writing group, namely,

Fatma, Elli, Antoinette, Mercedes and Ermitte, as well as insightful conversations with

numerous other fellow-students. I thank them all.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iv

L IST O F TA B LE S ........ .................... ......... .............................. .................... xi

LIST OF FIGU RE S ............. .............. .............. .................... .. ........... xiii

A B S T R A C T .........x.................................... ....................... ................. xv

CHAPTER

1 CONTEXTUALIZING RESEARCH TOPIC AND CHOICE OF METHODS ..........1

1.1 C ontextualizing the R research Topic............................................ .....................3
1.1.1 Anthropological Concerns...................................................................... 5
1.1.1.1 Anthropology and forced m igration.............................. .............. .6
1.1.1.2 Anthropology and social networks................................ .... ..... 7
1.1.1.3 Forced migration and social networks ............................................ 9
1.1.2 Refugee Studies and Gender Studies.............................................. 12
1.1.2.1 De-essentializing the "refugee-experience" ................................. 13
1.1.2.2 G ender and forced m igration............................... ..................... 17
1.1.2.3 Engendering forced m igration............................... ..................... 19
1.1.3 The Social W orld of Refugees ....................................... ............... 20
1.1.3.1 C hoping w ith loss........................................ .... .... ........ ......21
1.1.3.2 Social netw works ............................ .................... .. ......... 23
1.1.3.3 G ender, nationality and age................................... ............... 26
1.2 Choice of M methods ......... .................................. ............ ................. 31
1.2.1 Collecting Data.................... .. ....................................... 32
1.2.2 Collecting "V oices" ................................... ........................35
1.2.3 Representativeness .................................. .....................................38
1.3 In th e F ield ....................................................................... 4 0
1.3.1 M ethods U sed .................................................. .. .... .. .. ........ .... 41
1.3.2 Selecting R espondents.................................................................... ..... 44
1.3.3 Locating Inform ants ............................................................................47
1.4 Structure of the D issertation ..............................................................................54

2 SITUATING THE DISSERTATION PROJECT AND RESEARCH SITE.............. 58

2.1 B background to the R research Project ............... ..... ................. .....................58









2.1.1 The Refugee Producing Countries.............................................. 59
2.1.1.1 The historical interconnectedness of the Great Lakes countries......60
2.1.1.2 The colonial impact in Congo, Rwanda and Burundi ....................62
2.1.1.3 From independence to the early 1990s...........................................66
2.1.1.4 From the early 1990s onward.................... ...............69
2.1.2 The Refugee Receiving Country ..................................... .................76
2.1.2.1 Nyerere's Tanzania ............................. ................ ........ ....... 77
2.1.2.2 Africa's "open door" policy in Tanzania ....................................... 79
2.1.2.3 The change in clim ate ........................ ......... ............... .... 80
2.1.3 Rwandese, Burundese and Congolese refugees in Tanzania......................83
2.1.3.1 The refugee camps in Western Tanzania .......................................85
2.1.3.2 Urban refugees in Tanzania............. ............................................89
2.2 The Research Site: Dar es Salaam ....................................................... .......... 91
2.2.1 Governmental and Political Structures....................................................94
2.2.1.1 The 1998 Refugee Act and the Refugee Department.....................94
2.2.1.2 The Ten House System (Nyumba Kumi) ........................................96
2.2.1.3 The Center for the Study of Forced Migration............................98
2.2.2 Humanitarian Organizations.................. ..... ... ...............100
2.2.2.1 United Nations High Commissioner for Refugees.........................100
2.2.2.2 Umati ......... ........ .. .. ............. ........ ....... 102
2.2.2.3 O their organizations ............................................. ............... 103
2.2.3 The Social Context ................................................... ... .. .............. ... 106
2.2.3.1 Level of Contact ............... ..... .. ......... ........... ... ............ 110
2.2.3.2 Interm marriage ...................................................... ............... 112
2.2.3.3 R refugee policy............................................... ........ ........ 113
2.2.3.4 Likes, dislikes and differences ...................................................1 14
2.2.3.5 Summarizing findings ..... ................. ..............116
2.3. Conclusion ............. ..... ......... ................. 117

3 HEADING FOR DAR ES SALAAM ........................................... ............... 125

3.1 Localizing V oices .................... ........ .. .......... ............ .... .. .............. 125
3.1.1 Refugees in Post-Independence Dar es Salaam ..................................126
3.1.2 Demographics of the Sample.............................. .............129
3.1.2.1 Religion, marital status, and children ................ ... ............... 130
3.1.2.2 Education level and occupation .............................. ............... 133
3.1.2.3 Arrival in Tanzania, duration of stay in Dar es Salaam and
U N H CR registration ........................................ ......... ............... 137
3.1.3 Introducing the Protagonists ...................................... ............... 142
3.2 L leaving H om e ............................................ ... .... ................. 148
3.2.1 Fleeing W ars .................. .................................... ................. 149
3.2.2 Personal Loss .................. ......................... .... .. .. .. ........ .... 150
3.2.3 Saving O ne's Life .......................................... .............. ............. 151
3.2.4 Political Reasons ..................... ............ .. .......... .... ........ 152
3.3 Avoiding the Refugee Cam ps.................................... .......................... ......... 154
3.3.1 "M iserable" Cam p Life................................ ........................ ......... 156
3.3.2 Security C oncerns............................................. ............................ 157









3.3.3 Joining Relatives or Com patriots .................................. ............... 159
3.4 Arriving in D ar Es Salaam ................... ......................... .............. 160
3.4.1 Locating Relatives or Other Compatriots............................162
3.4.2 Tanzanian Friends ............................................................................. 162
3.4.3 U N H CR ..................................................................................... 163
3.5 Prospects of R turning H om e....... .... .................................. .......... ........ 164
3.5.1 The Return Home ............................... ............... 165
3.5.2 Alternatives to Repatriation...................... ... ......................... 166
3.6 E m erging Profile(s) .................................................. .............................. 168
3 .6 .1 G e n d e r ................................................................................................ 1 6 8
3.6.2 N nationality ................................. ..................................... 169
3 .6 .3 A g e G rou p s............. ........................................................ .. .... .. ... .. 170

4 ADAPTING TO THE NEW LIFE ........................................ ....................... 178

4 .1 L local Interactions ....................................................... .. ........ ........ 180
4.1.1 "Too m any questions" ........................... .............................................183
4 .1.2 L angu age ..................................................... ............... 184
4.1.3 "Lack of respect" ............ .... .... .... ... ........... ... .... .. ..... ..... 188
4.1.4 Religion ........... .............................. 191
4.2 Econom ic Endeavors .............................................. ......... ...............194
4.2.1 Perm its ..................... ............ ....................... ......... 195
4.2.1.1 Applying for a perm it ... ................. ....... .................... ............ 196
4.2.1.2 M carrying a Tanzanian citizen .................................... ............... 196
4.2.1.3 "A doptive" fathers ...................................... .. ............... 198
4.2.1.4 Staying out of prison .............. ....................... .. ............... 198
4.2.2 Kupata riziki ("M making ends meet")............................................. 200
4.2.3.2 Self-Em ploym ent ........................................ ........ ............... 200
4.2.2.2 Trading activities....................................... ......................... 203
4.2.2.3 Paid employment............. ............................ 205
4.3 M ajor Pre-occupations.......................................................................... ....... 208
4.3.1 Education ................................. .......................... .... ...... 210
4.3.2 Sexual Exploitation ............................................................................ 212
4.3.3 A spring to Leave Tanzania................................... ....................... 214
4.3.4 W worrying about Lost Relatives........................... ...... ............... 215
4.4 Social W orlds.................................................... 217
4 .4 .1 M meeting P people ............................................................ .............. 2 18
4.4.1.1 M meeting Tanzanians ............................................ ............... 218
4.4.1.2 M meeting fellow -refugees ..................................... ............... ..220
4.4.1.3 Socializing w ith friends....................................... ............... 223
4 .4 .2 M meeting P laces............ .......................................... .......... .... .......... 22 5
4.4.2.1 The neighborhood ........................................ ....... ............... 226
4.4.2.2 The church/m osque ............................................. ............... 227
4.4.2.3 A associations ......................... ...... ................ ........ .. .......... 228
4.5 C including R em arks ................................................ .............................. 232




viii









5 DE-CONSTRUCTING SOCIAL NETWORKS ............................................... 237

5.1 The Average Personal Social Network.................................... ............... 239
5.1.1 Gender, Nationality and Age .............................. .... ...............241
5.1.2 Other Demographic and Socio-Economic Attributes............. ...............243
5.1.3 Place of Residence, Lenght of Relationship and Frequency of Contact ..245
5.1.4 Ways of Meeting Alters and Types of Support Received.....................248
5.1.5 Sum m arizing Findings........................................ .......................... 250
5 .2 H o m o p h ily .................................................................................................... 2 5 2
5.2.1 Sam e G ender ............................................... ................ ............. 252
5 .2 .2 Sam e N ation ality ........................................................... .....................2 54
5.2.3 Sam e A ge C category ...................... .. .. ............................. ............... 255
5.2.4 Sum m arizing Findings......................................... .......... ............... 255
5.3 D isaggregating E go ................................................................ ....................257
5.3.1 Gender, N ationality and Age Groups .............................................257
5.3.2 Attributes of R relationships .......... .... ........................ ......... ...262
5.3.3 Alters by Ways of Meeting and Types of Support Provided....................266
5.3.4 Sum m arizing F indings................................................... .....................268
5.4 D isaggregating A lters ......... ................... ......... ................................ 270
5.4.1 W ays of M meeting .............................................. ............................ 270
5.4.2 Types of Support Provided .............................. ...............272
5.4.3 Sum m arizing Findings......................................... .......... ............... 274
5 .5 C o n c lu sio n s................................................................................................... 2 7 6

6 EMBEDDING THE REFUGEE EXPERIENCE ...............................................290

6 .1 G e n d e r ................................. ..... ............ ....................................................... 2 9 2
6.1.1 W omen as "Political Objects" ............. ............................................. 294
6.1.2 W omen as "Apolitical Subjects" ................................. ........... ........ 299
6.1.3 "You are who you know." ................ .....................302
6 .2 N atio n ality ..................................................................... 3 0 7
6.2.1 E conom ics ............................................................................ .3 11
6 .2 .2 P o litic s ............................................................... ............................... 3 1 7
6 .2 .3 H isto ry ................................................................................................ 3 2 2
6.3 Age ..................................................... ........... ...............330
6.3.1 Persons younger than 25 years of age ............................................... 334
6.3.2 Between 26 and 35 years of age .... .......... ..................................... 335
6.3.3 Between 36 and 45 years of age ............ .......................................... 338
6.3.4 Persons older than 45 years of age. .................................. ............... 340
6.4 C including R em arks ................................................ .............................. 343

7 REFLECTIONS, PREOCCUPATIONS AND CONSIDERATIONS ...................348

7.1 "T o aid or not to aid?"............. .. ......................... ....... ........ .. 350
7.2 Positionalities.................. ............... ................. 356
7.3 Urban Refugees, Social Networks and Forced Migration...............................3.60
7 .4 E p ilo g u e ...................................... .............................................. 3 6 3












APPENDIX

A M A P S ........................................................................... 3 6 5

B TANZANIAN PERCEPTIONS ON REFUGEES IN DAR ES SALAAM............ 372

C SOCIAL NETWORKS SURVEY............... .............. ......... ............... 376

L IST O F R EFE R EN C E S ........................................................................ ...............385

B IO G R A PH IC A L SK E T C H ........................................... ...........................................413











































x
















LIST OF TABLES

Table pge

2-1. Level of Contact w ith R refugees ................................................. ...... ......... 118

2-2. Tanzanian Perceptions on Prevalence of Refugee Nationalities .........................19

2-3. Tanzanian Perceptions on Refugee Occupations ..........................................119

2-4. Tanzanian Perceptions on Intermarriage with Refugees .................................... 120

2-5. Tanzanian Perceptions on Refugee Policies .................................. ............... 121

2-6. Level of "D on't know R eplies ........................................ ......................... 122

2-7. R efugee Characteristics Liked ........................................ .......................... 122

2-8. Refugee Characteristics D isliked ........................................................................ 123

2-9. Perceived Differences between Tanzanians and Refugees .................................124

3-1. Composition of Sample: Gender, Nationality, and Age Group. ..........................172

3-2. Education Level Compared to Camp Population ............................................172

3-3. Demographic Characteristics of Sample ........... ......................... ............... 173

3-4. Education Level and Professional Occupation ............................................... 174

3-5. Characteristics of Stay in Tanzania and Dar es Salaam .......................................175

3-6. Reasons for Leaving Home and Heading for Dar es Salaam .............................176

3-7. First day in Dar es Salaam and Prospects of Returning Home .............................177

4-1. Com paring Life at H om e and in Exile ........................................ .....................235

4-2. Interactions with and Perceptions of Tanzanian Nationals .................................236

5-1. Ego's alters disaggregated by Gender, Nationality and Age ...............................279

5-2. Social Network Members by Gender, Nationality and Age ...............................280









5-3. Demographic and Other Socio-Economic Attributes ........................................281

5-4. Relationship attributes of Social Network Members ........................................282

5-5. Ways of Meeting Network Members and Types of Support Received .................283

5-6. Homophilous relationships for sex, nationality and age group ...........................284

5-7. Social networks by gender, nationality and age category ....................................285

5-8. Relationship attributes by gender, nationality and age category .........................286

5-9. Ways of meeting and types of support received ......................................... 287

5-10. Alters by ways of meeting gender, nationality and age category ...................288

5-11. Alters by type of support gender, nationality, and age category .........................289
















LIST OF FIGURES

Figure pge

3-1: Religion by gender, nationality and age..... .......... ...................................... 131

3-2: M marital status and living arrange ents ........................................ .....................131

3-3: Number of Children in Dar es Salaam ........................................................132

3-4: Education levels ................................... ............. ........ ..... 133

3-5: Employment activities at home before leaving. .............................................. 134

3-6: Employment activities after arrival in Dar es Salaam ................. ... ..................135

3-7: Y ear of arrival in Tanzania............................................. ............................ 137

3-8: Number of years resided in refugee camp in Tanzania............... ...................138

3-9: Length of Stay in D ar es Salaam .................................................. ..... .......... 139

3-10: Registered with UNH CR ....................................................................... 141

3-11: Reasons for leaving the home country ............ .............................. ...............148

3-12: Motives for choosing Dar es Salaam instead of the refugee camps.......................155

3-13: Accommodation upon arrival in Dar es Salaam .....................................................161

3-14: Prospects on R turning H om e..................................................................... ..... 164

4-1: Comparing life at hom e and in Dares Salaam ...................................................... 179

4-2: Why establishing relationships in Dar es Salaam is difficult..............................181

4-3: M ajor w orries in D ar es Salaam ..................................................... ............ 209

5-1: H om ophilous ties for gender...................................................................... .. .... 253

5-2: Hom ophilous ties for nationality................................ ................... ...... ....... 254

5-3: H om ophilous ties for age category.................................. ....................... .......... 255









5-4: Pre-D ar es Salaam alters by gender.................................. ......................... 258

5-5: Post-Dar es Salaam alters by gender................................................................... 258

5-6: Pre-Dar es Salaam alters by nationality ...................................... ............... 259

5-7: Post-Dar es Salaam alters by nationality............................................................259

5-8: Pre-Dar es Salaam alters by age group........................................ ............... 261

5-9: Post-Dar es Salaam alters by age group ...................................... ............... 261

5-10: Pre- versus post-Dar es Salaam alters ........................................ ............... 262

5-11: Pre-Dar es Salaam alters by place of residence................... ...................................262

5-12: Post-Dar es Salaam alters by place of residence...................................................263

5-13: Pre-Dar es Salaam alters by length of relationship ............................264

5-14: Post-Dar es Salaam alters by length of relationship................ ................264

5-15: Pre-Dar es Salaam alters by frequency of contact.............. .................................265

5-16: Post-Dar es Salaam alter by frequency of contact.............. .................................265

5-17: Pre-Dar es Salaam alters by way of meeting ........... .............. ...............267

5-18: Post-Dar es Salaam alters by way of meeting ...................................................267

5-19: Types of Support received from pre-Dar es Salaam alters...............................268

5-20: Types of Support received from post-Dar es Salaam alters.............. ...............268

5-21: Pre-Dar es Salaam alters by ways of meeting...................................................270

5-22: Post-Dar es Salaam alters by ways of meeting ................................................271

5-23: Pre-Dar es Salaam alters by type of support .................................. ............... 272

5-24: Post-Dar es Salaam alters by type of support...................................................273















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

EMBEDDING THE REFUGEE EXPERIENCE:
FORCED MIGRATION AND SOCIAL NETWORKS IN DAR ES SALAAM,
TANZANIA


By

Roos Willems

December, 2003

Chair: Anthony Oliver-Smith
Major Department: Anthropology

Due to the tragic events in the countries of the Great Lakes region during the

1990s, more than one million refugees have sought a safe haven in Tanzania. Increasing

numbers among these, estimated to be in the tens of thousands, resist going to one of the

refugee camps and head for Dar es Salaam instead, even if it means foregoing any

material assistance from humanitarian agencies and legal residence status from the

national authorities in the country of asylum. Living scattered across the Tanzanian

metropolis, these forced migrants show their resilience under adverse circumstances by

coping with a new language, economic challenges and unfamiliar social environments.

In this process, they rely heavily on their personal social networks, members of

which provide them with financial and material assistance as well as emotional support in

the form of friendship, advice and companionship. This process of rebuilding their

shattered social worlds is analyzed in terms of the gender, nationality and age group of









the refugees. The central thesis is that there is no essential refugee experience, and that

the coping strategies of refugees are embedded in the social, cultural, economic, political

and historical background of the social groups, be they gender, nationality or age groups,

that the individual refugees are member of.

The principal methodological tool applied to the analysis of the data collected

during fieldwork sessions in 2000, 2001 and 2002, is social network analysis. Its

findings are contextualized qualitatively by the accounts, life histories and interviews

with informants of Congolese, Burundese, Rwandese and Tanzanian origin. Social

network analysis is proposed as the tool to situate individual persons' agency and

interactions at the micro-level within the structures and infrastructures (including events

at the regional, national and international level) constituting the macro-framework.














CHAPTER 1
CONTEXTUALIZING RESEARCH TOPIC AND CHOICE OF METHODS

This research project focuses on the self-settlement strategies of forced migrants in

an urban context, and explores how individual social networks are deployed in the

process of coping with displacement. The study is situated in Dar es Salaam, Tanzania,

and concentrates on urban refugees of Congolese, Burundese and Rwandese nationality.

Reconstructing and deploying one's personal social networks becomes an essential part of

survival in a situation where the policy of host governments and international agencies is

to not provide the self-settled urban refugee with material assistance, as is the case in

Tanzania. The variations in the ways in which Congolese, Burundese and Rwandese

refugees, men and women, young and old, adapt to the new life in exile while

reconstructing their social worlds are subsequently contextualized in a larger historical,

socio-political and economic framework of the respective social groups under study.

By April 2002, Tanzania was hosting approximately 512,000 refugees assisted by

the United Nations High Commissioner for Refugees [UNHCR], the overwhelming

majority of whom originated from Burundi (69%), Congo-Kinshasa (25%), and Rwanda

(5%) (UNHCR 2002a).1 Tanzanian refugee policy stipulates that all assistance to

refugees is to be provided only in the "designated areas" in rural Western Tanzania. In




1 The Tanzanian government estimates that the total number of refugees in the country (UNHCR assisted
and self-settled) is as high as one million. This study will concern itself only with refugees of the above
mentioned three nationalities. The remaining 1% of refugees in Tanzania comprise a/o Somali's, Ugandans,
Ethiopians, Sudanese.









the wake of the ever increasing rate of urbanization on the African continent,2 a growing

number of the refugees in Tanzania are urbanites, who prefer to live in the city, even

when this means foregoing the material assistance offered by the international agencies.

Being clandestine residents in Dar es Salaam, no official statistics exist on their number.

Congolese, as well as Burundese and Rwandese, have a long history of finding refuge in

Tanzania due to the crises and political tensions that have reigned in their respective

home countries over the past four decades. This study, contrary to others on the same

topic,3 will focus specifically on "first-generation" refugees, those who have vivid

memories of life in the home country and are still in the first few years after arriving in

the host country. Despite the enormous difficulties encountered in building up a

livelihood in a foreign city, close to all of the refugees interviewed in this study were

categorically resisting the official Tanzanian policy of indiscriminately allocating all

refugees to camps in Western Tanzania.

My argument is that the coping strategies of individual refugees are embedded in

the larger historical, cultural, social, political, economic and national frameworks that

constituted their lives before the flight. Dar es Salaam, at this point in time, provided the

perfect venue for a cross-cultural analysis allowing a comparison between Congolese,

Burundese and Rwandese refugee men and women, young and old. This first chapter

will contextualize the research project both theoretically and practically by situating the


2 According to UNDP, in 1999, 33.5 % of the Sub-Saharan African population lived in urban centers,
compared to 46.5% worldwide. On the basis of a higher growth rate for urban Africa, by 2015 its urban
population will comprise 43.3% of the total, increasingly closer to the global average of 53.2% projected
for 2015 (UNDP 2001).

3 Sommers (2001) in his study of urban refugees in Dar es Salaam, concentrates solely on "young men ...
Burundian nationals who have little or no memory of Burundi," alternatively called "second Burundi
refugee generation" (7) by the author.









research topic within the literature, documenting my phenomenological approach to

fieldwork and methods followed by their subsequent application and consequences

thereof in the field setting.

1.1 Contextualizing the Research Topic

The goal of this research project is a cross-cultural comparison of the ways in

which urban refugee men and women of different nationalities and ages cope with a

situation of forced migration and rebuild their personal social networks. Before

venturing out into any disciplinary concerns, however, I would like to elaborate on the

terms "refugee" and "forced migrant." The legal definition of refugees stipulates them to

be persons having crossed international borders in fear of personal persecution, and as

such entitled to protection from the UNHCR. The 1951 UN Convention defines a refugee

as any person:

... who owing to a well-founded fear of being persecuted for reasons of race,
religion, nationality, membership of a particular social group or political opinion, is
outside the country of his nationality and is unable or, owing to such fear or for
reasons other than personal convenience, is unwilling to avail himself of the
protection of that country. .. (UNHCR 2000a: 6)

In 1969, the Organization of African States [OAU] enlarged that definition (as applied in

African countries) to include persons:

... who owing to external aggression, occupation, foreign domination or events
seriously disturbing public order in either part or the whole of his country of origin
or nationality, is compelled to leave his place of habitual residence in order to seek
refuge in another place outside his country. .. (82)

Over the past decade, a debate has been raging as to whether to enlarge the

international definition of a refugee even further by including "internally displaced

persons" (IDPs) as equally entitled to the protection of the UNHCR (Lee 1996).

However, as this would, according to some observers, challenge the sovereignty of the









nation-state--because IDPs by definition never crossed the borders of their homeland--it

is unlikely that this debate will draw to a close any time soon.

Nevertheless, recognizing that both refugees and IDPs suffer the same experience

of uprootednesss" and of having forcibly left their homes, the term "forced migrant"

emerged to encompass both categories and create a wider focus for both academic and

policy makers' interests (Adelman 2001). As such, "a forced migrant can be defined

roughly as someone who is forced to leave his or her home because of a real or perceived

threat to life or well-being" (Reed 1998:1). It is my understanding that in the field of

Diaspora Studies, the term "forced migrant" is reserved for those who had no say

whatsoever in their "migration," such as for example, the Africans subjected to the

transatlantic slave trade. This is not the guiding principle in the area of refugee studies,

where although named forced migrants, these individuals are recognized to have taken

the decision to leave.4 For the purpose of this dissertation the terms "refugee" and "forced

migrant" will be used interchangeably.5 In this research project, I focus solely on

Congolese, Burundese and Rwandese nationals who sought refuge across international

borders, namely in Tanzania, for reasons ranging from personal political persecution and

situations of war to a motivation based on a lack of economic opportunities due to the

disturbances created in day-to-day life by politically motivated (violent) events. In

recognition of the variety of conditions under which the refugees from the Great Lakes


4... depending on their financial possibilities. This point was brought home to me only when one of the
Congolese informants was recounting how in 1996 when he left, many Congolese were unable to pay the
US$20 boat fare to cross Lake Tanganyika and find refuge and safety in Tanzania. Umutesi's detailed
account of her year long flight across Zaire highlights more than once that the quest for safety is an
expensive one (2000).

5 In the same line, the Center for the Study of Forced Migration at the University of Dar es Salaam in
Tanzania concerns itself almost exclusively with studies of refugees.









region flee their homes and seek asylum abroad, Tanzanian authorities apply different

procedures for different nationalities. As such, Congolese and Burundese asylum seekers

were recognized as primafacie refugees under the 1969 OAU Protocol, while Rwandese

asylum seekers, at the time of research, were expected to undergo an individual interview

before being accorded refugee status, in line with the 1951 UN Convention. Ultimately,

however, there was no difference between the two groups with regard to their rights as

refugees in Tanzania or the challenges that refugee life posed upon them.

1.1.1 Anthropological Concerns

In anthropological terms, refugees are people who have undergone a violent

separation and--unless or until they are "incorporated" as citizens into their host state or

have returned to their state of origin--find themselves in "transition," or in a state of

"liminality" (Harrell-Bond & Voutira 1992). In Turner's words: "Liminal entities are

neither here nor there; they are betwixt and between the positions assigned and arrayed

by law, custom, convention, and ceremonial" (1969:95). Refugees have to deal with

chaos. They have been ripped from their established lives; often their families are

dismembered and their support systems destroyed. Many of them are struggling with a

new and unfamiliar social and physical environment, in which the belief systems or styles

of action they brought into refugee status do not work well. It is in this sense that they

are "betwixt and between" their old and new lives, not knowing what to expect next.

Gold rightly points out that,

while voluntary immigrants have frequently made a conscious decision to accept
the indignities of dealing with an alien culture, refugees [as involuntary migrants]
often cling to their traditional roles and values and may resist adaptation. (Gold
1992:18)









As the following chapters will apply illustrate, the Congolese, Burundese and Rwandese

refugees in Dar es Salaam are not inclined to adapt to the culture of their host country,

except in instances where not adapting could possibly have consequences endangering

their personal safety, such as the use of Swahili as it is spoken in Dar es Salaam (see

chapter 4). In a situation of forced migration, which is ultimately a transient one, given

that the majority of respondents arrived in Dar es Salaam fairly recently and aspire to

repatriate as soon as the conditions at home allow it (see chapter 3), the emphasis will be

on coping rather than adapting.

1.1.1.1 Anthropology and forced migration

Hansen and Oliver-Smith (1982) point out that when people migrate across national

borders as forced migrants, they do not give up their understandings of what various

individual, family, and social behaviors mean. On the contrary, "our responses to change

are always innovative but are set within the parameters specified by our perceptual

frameworks, and these are a function of our past" (Oliver-Smith 1986:16). Refugees will

choose from alternative behaviors those that will allow them to maintain their basic

conceptions of what life is amidst the turmoil and disruptions, a process possibly (but not

necessarily) leading to socio-cultural change through innovative behaviors (Geiger 1993).

It is in this sense that Harrell-Bond and Voutira stated that: "of all the disciplines

involved in the study of human behavior, we contend that anthropology has the most to

contribute to the study of refugees" (Harrell-Bond and Voutira 1992: 6).

Just as anthropology may contribute to the study of refugees--and, on a more

practical note, to the designing of refugee relief efforts as well (Waldron 1988)--the study

of refugees may also contribute to the development of anthropological theory.

Anthropology should not just be "the comfortable anthropology of social organization," it









should include as well "the painful anthropology of disruption and despair" (Davis

1992:149). The particular circumstances of refugee situations allow us to uncover certain

socio-cultural dynamics and realities that could be hard to distinguish or identify in

"normal" life. At the same time, I would like to point out that these socio-cultural

dynamics and realities are not always generally applicable. Rather, members belonging

to one or the other social groups (i.e., men versus women, young versus old, Rwandese

versus Burundese or Congolese) will act following the cultural norms and values

governing the pre-flight societal positions of their respective gender, nationality or age

groups. As such, studying forced migrants through an anthropological lens may

contribute to the further development of social science theory.

1.1.1.2 Anthropology and social networks

Social networks as a methodology emerged in the 1960s out of the Rhodes-

Livingstone Institute (Lusaka) in the wake of the ever increasing urbanization rate in the

Copperbelt area in Zambia.6 While urban anthropology itself, as a "self-labeled body of

research" (Sanjek 1990:151), saw the light in the 1960s, only in the 1980s did it rear its

head in the anthropological literature.7 Rendering all but obsolete the traditional kit of

anthropological research tools designed for small-scale rural communities, the upsurge in

urban anthropology interest induced many researchers to look into the direction of social

network analysis as a more appropriate methodological approach. Indeed, as Hannerz

(1980) pointed out: "urbanites especially ... typically do not draw their sustenance

directly out of the earth, but to a great extent from their dealings with one another" (57).


6 While Bott (1957) carried out research on social networks in the 1950s, she did not as such belong to a
school of thought on social network analysis.

7 See Sanjek 1990 for a literature overview.









And yet, while a 1988 survey of social network researchers found that prior to

1970, 60% of the scholars most influential to the "social network community" were

anthropologists (Johnson 1994: 142), it appeared that by 1988, only 20% of the top

positions within the field were anthropologists, the majority of whom specialized in

cognitive anthropology. The key to the contribution of anthropologists to the

development of the concept of social network appears to have been the general

anthropological interest in social structure in the 1950s and 1960s. As two major

contemporary proponents of social network analysis at the time put it:

... the study of social networks has arisen out of the dissatisfaction of a number of
social anthropologists with the conventional form of structural-functional analysis.
Structural-functional analysis by definition views societies as essentially static [and
thus] is inadequate to cope with the study of rapidly changing complex societies.
(Boussevain and Mitchell 1973: vii)

It has been suggested that when the techniques of social network analysis began to lead a

life of their own, not necessarily linked to a particular social theory (Scott 2000: 7-37),8

more and more anthropologists began to see scientific rigor as a hindrance, impractical

and unnecessary for anthropological research on social networks (Walsh and Simonelli

1986).

Notwithstanding the alleged historical unpopularity of social network analysis

among the majority of anthropological researchers, I believe that this method is

extremely appropriate for studying groups of people faced with tremendous changes,

because it allows for establishing the linkages between micro and macro level events.

One anthropological method, occasionally called situational analysis, is defined as the

"intellectual isolation of a set of events from the wider social context in which they occur


8 A development largely due to the availability of computer systems and software packages able to digest
ever increasing amounts of data.









in order to facilitate a logically coherent analysis of these events" (Mitchell 1987:7). This

methodology was first applied by Gluckman in his account of the opening ceremony for a

bridge in Zululand, showing how elements of the larger social order are at work in

individual situations and need to be incorporated in the final analysis (Rogers and

Vertovec 1995: 5-11; Marx 1990:192). Applying this technique to social networks

allows uncovering the ways in which the composition of an individual's personal network

varies according to the larger world this person belongs to.

The key focus of this anthropological dissertation project is to contextualize the

daily actions (agency) and interactions (social networks) of the Congolese, Burundese

and Rwandese urban refugees (men and women, young and old) in Dar es Salaam by

situating them into the larger framework of their respective historical, socio-cultural,

political and economic backgrounds (structure). Leaving the static principles at the basis

of structural-functionalism far behind it, this approach intends to contribute to the

anthropological understandings of the dynamic complexities of contemporary human

existence.

1.1.1.3 Forced migration and social networks

An important, if not the most important, link in today's world, between forced

migration and social networks is the issue of globalization. The current era of continuing

improvements in transport and communication has seen the migration of ever increasing

numbers of people. Castles estimated that in 1992 over 100 million people worldwide

lived in a country other than that of their birth, and that some 20 million of those, or 20%,

were refugees (1998:179). Academic, and more particularly, social science theorists

could not have been left untouched by this reconfiguration of social life (Amit-Talai

1997), and therefore a growing number of authors are reformulating the classical









anthropological (and sociological) perception of "community as a place" to one that sees

communities as social networks (Wellman 1999: xiv).

At the same time, within the field of international migration, the realist perspective

sees refugees as fundamentally distinct from immigrants, while the nominalist

perspective ponders "whether the category is a social construction masking similarities

with immigrations" (Hein 1993:43). Whereas in earlier research on refugees, the starting

point was a simple dichotomy with immigrants constituting an economic form of

migration and refugees a political form, it has been found more recently that "refugees

share some migratory experiences with immigrants, particularly prior [to] migration and

[concerning their] reliance on social networks" (46, emphasis added).9

Despite theoretical recognition of the close link between forced migration and

social networks, empirical research, in terms of the collection of actual social network

data, seems to have lagged behind. The existence of social networks and their

importance to the survival strategies of forced migrants have been mentioned by only a

few researchers who have done major research projects on urban refugees in Tanzania.

Malkki (1995a), for one, remarks that the Burundi urban refugees in Kigoma "relied on

networks of their own making" (46), however, she did not undertake any systematic

analysis of these networks. Sommers (1994) noticed how the group of Burundi refugees

in Dar es Salaam he studied were able to obtain employment through networks, yet again,

no further analysis was ventured. Outside of Tanzania, there have been only a few small-

scale projects looking into the role of social networks in a refugee situation (e.g., Koser



9 In their last State of the World's Refugees UNHCR concedes that "while the categories are neatly
distinguished on paper, in reality the boundaries between them are far from clear and the connections
between them many" (UNHCR 2000b: 280).









1997; Williams 1993). Williams found, in her small-scale study among Angolan self-

settled refugees in rural Zambia, that the refugees formed social relationships with non-

kin individuals, Zambian or Angolan, to a much larger extent than was previously

assumed. In addition, she discovered that those social networks were the channels par

excellence through which the refugees were able to rebuild their livelihoods in a new and

unfamiliar environment.

At a more general level, Van Duijn et al. (1999) draw attention to the fact that "the

idea that people invest in one another is also one of the basic assumptions in social capital

theory ... It is assumed that people invest in each other to gain future access to different

resources" (191). Likewise, Wellman (1999) argues that in situations of uncertainty, such

as living as clandestine refugees in a foreign country, social networks seem "to be

especially important" (xviii). Notwithstanding, the existing body of research on urban

refugees in Africa (e.g., Kibreab 1996; Kpanou 1989; Weaver 1988; Goitom 1987;

Karadawi 1987; Kuhlman 1990), while recognizing that the main concern of urban

refugees is to find (informal) paid employment, pays little attention to the mechanisms

through which the refugee actually seeks and finds financial and other support. My study

intends to illustrate that the Congolese, Burundese and Rwandese urban refugees in Dar

es Salaam depend on their social networks to either find gainful local employment in the

informal sector as hair-dressers, tailors, daily laborers, etc. or receive financial support

from friends and relatives, whether from the other side of town or from the other side of

the world.

The reliance of these urban refugees on their personal relationships for survival in

the absence of any institutional structures providing assistance may be constructed

through Hyden's concept of the "economy of affection."









It must be said from the outset that the economy of affection has nothing to do with
fond emotions per se. Rather, it denotes a network of support, communications and
interaction among structurally defined groups connected by blood, kin, community
or other affinities ... To be sure, the economy of affection is most prevalent in the
rural community but it is an integral part of the society at large. Its influence
stretches right from the grass-roots to the apex of society. (1983:8-9)

According to Hyden, the economy of affection may be considered an alternative

economy that fulfills an "extremely important welfare function" particularly in African

countries that have experienced a breakdown of the state (10-11), but even more so in

refugee situations. "In no other situation has the economy of affection proved itself more

important as a mechanism for basic survival than where endless flows of refugees have

suddenly overwhelmed a country" (12). Even as Dar es Salaam has not been

overwhelmed by endless flows of Congolese, Burundese and Rwandese refugees, it is

certainly true that in the absence of any assistance from institutional structures (whether

material assistance or residence permits), the economy of affection as a support network

is of the utmost importance to the refugees' survival. The following chapters will

illustrate, however, that Dar es Salaam's urban refugees rely not only on their social

network members from before the flight, but that newly established ties in exile are

equally essential with regard to both material assistance and emotional support.

1.1.2 Refugee Studies and Gender Studies

Very often in the refugee studies literature, the emphasis is on "the" refugee and

"the refugee experience" (Ager 1999), and too little attention goes to the refugee as a

cultural, gendered person with his/her own particular historical, political, economic and

social background. With regard to the gender studies literature, it is not so long ago that

one researcher, while conceding that gender is an organizing principle in the social,

economic, political structures and the historiography of society, underlined that, at the









same time, the cultural construction of gender is embedded in its social, economic,

political and historic contexts (di Leonardo 1991). At the same time, Malkki challenged

the assumption, recurrent in the refugee literature, that "to become uprooted and removed

from a national community is automatically to lose one's identity, traditions and culture"

(1995b: 508).

The term "de-essentializing the refugee-experience" may be interpreted in more

than one way. One interpretation is based--in line with Malkki--on the assumption that

different refugee populations will experience what it is to be a refugee in different ways

because of their respective socio-cultural and historical backgrounds. The second

interpretation of the term relates to the assumption that the experiences that refugees go

through are not exclusive to forced migrants that crossed international borders, but are

shared by persons who were uprooted and displaced from their homes for a variety of

possible reasons (e.g., economic, political or related to development), and who are often

referred to as internally displaced persons [IDPs]. This research project, however,

concentrated exclusively on the first interpretation, namely, the various ways in which

different refugee populations experience life in exile.

1.1.2.1 De-essentializing the "refugee-experience"

In the decades following the second world war, refugees became the focus for the

development of a vast and complex network of institutionalized assistance composed of

host governments, UNHCR and other UN organizations, and also non-governmental

organizations which were assigned or assumed responsibility to deal with their material

needs (Harrell-Bond and Voutira 1992; Zolberg et al. 1989; Adelman and Sorenson

1994). In the wake of this development, a wide range of historically generated Western

conceptions about refugees, rights, nations and international aid has been unconsciously









incorporated into large parts of the current domain of refugee studies (Malkki 1995b).

Indra (1998), rightly, points out that:

Researchers often unproblematically used highly stereotypical social issues and
bureaucratically generated representations of "the" culture, experiences, and goals
of refugees that would never have been acceptable in contemporary ethnographic
treatments... (xii)

Consequently, the refugee was given an as it were "bureaucratic identity" (Zetter 1991)

and a "visual image" accompanied by

a peculiar kind of speechlessness in the face of the national and international
organizations whose object of care and control they are. Their accounts are
disqualified almost a priori, while the languages of refugee relief, policy science,
and "development" claim the production of authoritative narratives about the
refugees. (Malkki 1996: 386)

This passive image of "the" refugee was compounded by the discovery of the so-called

dependency syndrome among refugees due to prolonged residence in camps, where the

bulk of studies on refugees takes place (Harrell-Bond 1986; Van Damme 1995). The

very configuration of the international refugee aid regime in the form of refugee camps--

especially those where access and exit are restricted by the authorities of the host

country--requires the refugees to be not only completely dependent on the camp

management for things such as food and medical assistance, but in addition, local

housing arrangements and configurations are routinely disregarded by international

agencies for the sake of efficiency (Cuny 1977; Ellis and Barakat 1996).

Comparative studies into the living conditions of self-settled refugees outside of

refugee camps, on the other hand, have shown them to be materially less well off than

camp refugees, but much stronger and better healed at the psychological level (Harrell-

Bond 1986: Hansen 1979, 1981). Studies among Mayan Guatemalan refugees in Mexico

in camps (Light 1992) and urban areas (Bottinelli et al.: 1990) showed that, regardless of









gender, the coping process of men and women refugees is accelerated by the

opportunities to regain independent livelihoods, or to feel in charge of their own lives

allowing for a renewed self-esteem and a view on the future. Clearly, the specific

conditions in which refugees are to cope with their situation of forced migration not only

influence the process of adaptation and the extent to which they are able to exercise

agency, but play a determinative role in the multiple refugee experiences.

Even as the majority of research studies among refugees has taken place among

camp refugees, it is estimated that in Africa they make up only 20 to 25% of the total

refugee population (Kuhlman 1994). Refugee camp populations nevertheless continue to

attract the most attention not only from the international aid regime, but also from the

news media and academics (Kuhlman 1994; Malkki 1996; Zetter 1991). Harrell-Bond

(1986) and Hansen (1979, 1981, 1992) were among the first to compare the situation of

camp refugees with rural self-settled refugees (Chambers 1979), but the debate on

whether spontaneous self-settlement or organized camps is the better option for refugees

rages on (Kibreab 1991; Black 1998a, 1998b; Crisp and Jacobson 1998; Harrell-Bond

1998). Whereas the humanitarian perspective has pointed out the psychological

consequences of living in refugee camps for prolonged periods of time (i.e., dependency

syndrome, see earlier), at the political level, there is an increasing tendency among the

authorities of refugee-receiving countries to invoke reasons of national security to

confine the refugees to a restricted or "designated area" (Kibreab 1991). This was also

the case in Tanzania, where until the mid-1990s, the majority of refugees were allocated

plots of land in settlements and encouraged to become self-sufficient while a substantial

minority were issued with work permits, allowing them to reside in one of the country's

urban centers. The influx of one million Rwandese refugees in the wake of the 1994









genocide, together with armed ex-military and other militant groups had a drastic effect

on Tanzania's refugee policy, and from then on, all refugees were required to reside in

"designated areas," a euphemism for the refugee camps in the Western provinces of the

country.

The afore-mentioned studies reported that Africa's rural self-settled refugees

usually join their co-ethnics living just across the border in order to cope with the absence

of assistance from the host government or international agencies (Hansen 1992, 1979;

Nelder 1979; Leach 1992). Increasing urbanization rates across the African continent,

however, have changed the rural/urban ratio of its populations, hence of its refugee

populations. In the mid-1980s, one researcher estimated that probably close to 30% of

Africa's refugee population are urban, and are likely to choose to reside in urban centers

while in exile (Rogge 1986). This research project focuses specifically on African urban

self-settled refugees and their coping strategies in the absence of the possibility of joining

co-ethnics or of receiving humanitarian assistance.

Whereas assisted urban refugee populations have received some attention in the

literature,10 the increasing numbers of self-settled urban refugees have been largely

neglected in academic research (Kibreab 1996). In addition to Kibreab (1996), there is

Sommers' study among young male Burundi refugees in Dar es Salaam done in the early

1990s (1993; 1994; 2001). It should be noted however that his research project

concentrated on only a particular segment of the Dar es Salaam refugee population,

namely second generation Burundi refugees--children from the 1972 Burundi refugee



10 E.g., Ethiopians and Eritreans in Khartoum who received assistance until 1986 (Rogge 1986; Goitom
1987; Karadawi 1987; Weaver 1988), and Salvadorans in Costa Rica (Basok 1988, 1993; BIT and
UNHCR 1987).









influx in Tanzania--who grew up in one of the rural settlement in Western Tanzania and

migrated to an urban setting as young, mostly male, adults. They share with Tanzanian

rural young men, who equally left their villages trying to "find life" (kutafuta maisha) in

the big city, a large part of their experiences.

From the above, it appears that a large part of the refugee literature of the past

decades has been dominated by the stereotyped image of a refugee as a passive recipient

of humanitarian assistance in a rural refugee camp. While some researchers studied self-

settled rural refugees or assisted urban refugee populations, a lacuna exists in the area of

urban, self-settled refugees. And yet this is a population which has known a tremendous

growth in numbers over the past decade due to increasing rates of urbanization11 as well

as the ever restrictive policies of host governments. This increase in numbers of urban

refugee populations worldwide had equally led to a reconfiguration of their composition,

traditionally typified as young, single and male (UNHCR 2001c). Taking into account

variables such as gender, nationality and age will demonstrate the variety of the refugees'

experiences and ways of coping in exile, and de-construct the essentialist definition of

"the" refugee and "the" refugee experience.

1.1.2.2 Gender and forced migration

The anthropology of women, emerging in the 1960s and 1970s, intended to provide

a counterbalance for the male bias, prevalent in social studies at the time, by studying

women (e.g., Amadiume 1987; Hafkin and Bay 1974; Stoler 1989, 1995; Nederveen

1990), an approach which was later termed--somewhat derogatorily--the "add-women-

and-stir-method" (Moore H. 1988:3). Many of these ethnographic studies were based on


11 In 1999, Burundi was reported to have a 8.9% urbanization rate compared to 14.5 % and 39.3% in
Rwanda and the Congo respectively (UNDP 2000).









the concept of the "universal presence of the patriarchal system" prevailing among

feminists and anthropologists at the time. Even as colonialism and capitalist penetration

were not considered solely responsible for inegalitarian gender relations in Africa

(Leacock and Etienne 1980), studies confirmed that male-oriented Christian ideologies

were often reflected in official colonial state policies (e.g., Hunt 1990; Mbilinyi 1990).

The emphasis of the feminist anthropologists of the 1980s was on the sexual division of

labor, the decision-making process within the household, and the deficiencies of the

development strategies in terms of gender (Holmboe-Ottesen and Wandel 1991; Spring

1992; Jahan 1995). They failed, however, to explicitly take into account gender relations,

as well as the experiences of gender by local men and women, and their agency. While

exploring the ways in which economics, kinship and social and political organization are

structured by gender (e.g., Oppong 1983; Amadiume 1987; Leach in Croll and Parkin

1992; Mikell 1997), they stopped short of asking how gender itself is constructed through

culture. The anthropology of gender, emerging from the late 1980s onward, specifically

concentrated on the study of gender identity and its social (West and Zimmerman 1991)

and cultural construction (Moore H. 1988). It explores the relationship between

discursive constructions of women and men and their political economic realities,

whereby both women and men take advantages of the opportunities produced by social

and material changes in their lives and endlessly renegotiate the boundaries of gender

(e.g., Hodgson and McCurdy 2001).

Up until recently, the focus on gender in the field of refugee studies followed the

earlier reorientations within the field of gender studies rather closely. The substantial

increase in research on women refugees focused mainly on a range of conventionally-

defined women's issues, such as health and diet, basic protection issues, women's









employment, education, child care, language training and sexual violence (e.g., Martin

1992; Martin and Copeland 1989; Ager, Ager and Long 1995; Daley 1991). The result

was a growing awareness among refugee agencies of the gender specific problems

women refugees are exposed to. As a result, UNHCR and other agencies started issuing

and distributing explicit guidelines, policy documents and training materials to sensitize

all refugee workers of the necessity to take into account the specific needs and

vulnerabilities of women refugees (UNHCR 1991; 1990).12

At the theoretical level, however, women's issues continue to remain quite

peripheral to the mainstream discourse of academic refugee-studies: "the work is on

refugees who happen to be women rather than on women and men who happen to be

refugees" (Indra 1989:224). Gendering the refugee experience should not involve

additional research on women only, rather it means looking at how both men and women-

-be it in the refugee camps or among self-settled refugees--discursively (re)construct

social realities and renegotiate gender roles and boundaries in the process of coping with

forced migration.

1.1.2.3 Engendering forced migration

The process of coping with loss resulting from forced migration will in the first

instance give rise to an increased importance attached to ethnicity, and traditional

(usually patriarchal) cultural values and familiar rituals. The increased ethnic awareness

of forced migrants--especially in refugee camps where the feeling of alienation is even

more pressing--have been illustrated amply by researchers (e.g., Phillips 1996, Malkki



12 Contrary to earlier contentions, UNHCR currently confirms women refugee and girls to make up 50% of
the refugee population dispelling previously held assumptions that over 80% of most refugee populations is
female.









1995a, Harrell-bond 1986, Camino and Krulfeld 1994). In contrast, Turshen and

Twagiramariya (1998) have argued that war also destroys the traditional patriarchal

structures of society that confine and sometimes degrade women, and that in the very

breakdown of morals, traditions, customs and community, war also opens up and creates

new beginnings. They illustrate their argument referring to specific case studies in Chad,

Rwanda, Sudan and Mozambique where war, genocide and conflicts forced significant

changes in gender roles. As a result of the increasing economic role of women, their

power within the home increased significantly.

The configuration of the personal social networks of refugee men and women

respectively13 should prove to be an appropriate site to explore the agency exercised by

members of both genders in the process of renegotiating socio-cultural boundaries,

especially when compared to the pre-flight situation in the respective countries. In

addition to exploring changing gender dynamics as a result of the situation of forced

migration, I intend to look at the role played by nationality and age as well. The context

for selecting nationality as well as age, in addition to gender, as the main parameters of

the research project is presented in the following section.

1.1.3 The Social World of Refugees

The legal nature of the definition of a refugee and the fact that s/he is considered an

aberration in a world order of nation-states (Ager 1999:90) has given rise to an

impressive body of literature with regard to the legal, political and human rights aspects

of the matter (e.g., Campbell and Van Arsdale 1994; Hathaway 1991, 1995; Lee 1996;

Warner and Hathwayl992). Furthermore, the political, and often also ethnic, nature of


13 The unit of analysis of this study is the individual man or woman, not the household.









the causes that lie at the source of many of today's refugee flows have triggered an

avalanche of studies on how refugees politically and ethnically (re)organize themselves

in exile, again mainly in refugee camps (e.g., Malkki 1996; Phillips 1996). Less attention

has been paid in the literature, however, to the refugee as a social being uprooted from

his or her "home" (Xenos 1993).14 Allowing for the fact that "home" is not just a

geographic notion, but maybe more importantly so, a set of social relations, Marx defines

a refugee as "a person whose social world has been disturbed" (1990:190). However, in

order to better comprehend the process of rebuilding one's social world in exile, it is

necessary first to conceptualize how people deal with loss.

1.1.3.1 Coping with loss

Marris analyzes the process of coping with loss in terms of three elements:

"conservatism, bereavement and innovation" (1986:2). The element of conservatism is

based on the assumption that the impulse to defend the predictability of life is a

fundamental and universal principle of human psychology. Predictability is related to

control, and according to Marris "we cannot control the physical world without

discovering its laws, but we can control our social world, more or less, by imposing laws

upon it" (16). When confronted with a situation of a loss, people will look for a

continuity of the meaning of life as they have known it so far, hence the deep-seated

impulse to defend the validity of what they have learned, because without it they would

feel lost. Because meanings are learned in the context of specific relationships and

circumstances from early childhood onwards, the continuity of this context represents for

an individual his/her identity (5-11). When there is a drastic change in an individual's


14 See also Portuous and Smith's Domicide and the Global Destruction of Home (2001).









relationships or circumstances (e.g., the loss of a close relative, or the event of forced

migration) this will affect the individual's sense of identity, in that the continuity between

past and present is threatened, and life has become unpredictable.

The second element of Marris' process of loss, bereavement, arises not from the

loss of others (e.g., the death of a close relative, or the loss of one's socio-cultural

environment), but from the loss of self, in that the relationships fundamental to one's

identity have been abruptly altered and meaning now needs to be reformulated. As the

meaning of life is defined by the particular experiences of each individual, those

experiences need to be treated with respect, and so where changes imply loss, this loss

must be grieved for.

Grief... is the expression of a profound conflict between contradictory impulses -
to consolidate all that is still valuable and important in the past, and preserve it
from loss; and at the same time, to re-establish a meaningful pattern of
relationships, in which the loss is accepted. (31)

The contradictory impulses within the process of grieving, which seek to reconnect past

and future, ask for a predictable set of "mourning behavior," as formulated in mourning

customs and rituals, in order to protect the mourner from the complexity of the issues

involved (92). The necessity for a mourning ritual applies not only in situations where

one has lost a close relative through death, but also when a drastic change in (living)

circumstances occurs. It is not difficult to see in Marris' example of bereavement in a

situation of slum clearance in Lagos, Nigeria (43-58), the analogy with a situation of

forced migration "as in grieving for the dead, all the purposes and understanding inherent

in those surroundings have to be retrieved and refashioned so that they still make sense of

life elsewhere" (57).









The conservative impulse in itself is not incompatible with innovation, since it

"seeks to consolidate those skills and attachments, whose secure possession provides the

assurance to master something new" (19). Marris says that "we can readily accommodate

to changes, so long as they can first be assimilated to our existing patterns of thought and

attachment" (104). People organize their lives as a function of hopes and goals that are

in line with conventional expectations, and when they realize that these goals have

become unattainable in the conventionally prescribed set of actions, frustration arises.

Only in this situation of frustration, will people innovate and initiate changes which

might affect the predictability of their lives. And so we may expect Congolese,

Burundese and Rwandese refugees in Dar es Salaam to adapt to the new life and rebuild

their social worlds following existing patterns of behavior up to the point where these

behaviors no longer generate the expected outcomes and changes have to be

accommodated to. Having conceptualized the process of coping with loss as a result of

forced migration, we need to de-construct the concept of social worlds and establish a

methodological tool with which to conduct the research.

1.1.3.2 Social networks

The refugees' social worlds may be defined as "the sum of all the [refugees']

relationships and of the forces impinging on them at any moment" and "can be explored

through the social networks [of the refugees] and the changes they undergo" (Marx

1990:189) during the process of forced migration and adaptation to life in exile. The idea

of using social networks as a tool to analyze the dynamic forces that play at the macro-

level is not new. One of the forerunners of social network analysis and member of the

Rhodes-Livingstone Institute (see earlier) stated from the very beginning that the wider

social context is reflected in the composition of the personal social network and that









each personal network, therefore, will be unique though obviously influenced by
such factors as the social position of the persons concerned and the social situation
in which he is placed [and] the stage of the life-cycle he is at. (Mitchell 1969:43)

As such, the composition of social networks will not only be affected by a situation

of forced migration, but will also be influenced by the societal position of the individual

in terms of his/her belonging to various types of social groups, e.g., men or women,

different nationalities and age categories (see more in next section). Marx's "impinging

forces" are structures or macro-elements of a historical, political, economical or socio-

cultural nature related to the functioning, societal role and status of the (members of)

respective social groups. These forces will affect the behavior, albeit the agency,

displayed by individual refugees as members of the various social groups. It is in this

respect, that I intend to disaggregate the social network data collected in order to uncover

multiple patterns of coping with forced migration and rebuilding of social networks,

thereby ultimately de-essentializing the refugee experience and embedding it in several

layers of larger contexts at the macro-level.

Approximately two decades into social network analysis, researchers cautioned that

"it is important that network researchers consciously and clearly specify what they need

to know about networks and what they mean theoretically by relation" (McAllister and

Fischer 1983:87). The relationship (or relation) is the social process that ultimately links

one with his/her social network members. Mitchell defined potential network members

as that category of people who "in terms of the general norms or values of the community

might be expected to provide ego with some type of service or support" (1969:43), and

for the potential relationship to become a link in the personal network "some social

exchange or transaction which converts the possible into an actual social linkage" (Ibid.)

needs to take place. Applying this principle to the topic of this research project, I









postulate that a relationship between an urban refugees in Dar es Salaam and one of

his/her social network members consists of the actual providing of some type of support

in the wake of the situation of forced migration.

Support may include a variety of resources such as "emotional help, personal

service, material assistance, financial aid, social brokerage, and empathetic

understanding" (Wellman 1981:184). The same author found that

While sociability [does] not provide supportive resources directly, [its]
existence affirms self worth and continued network membership. However, we are
also finding non-supportive "structurally embedded" strands to be common. In
such instances, two persons interact with one another only because the larger work
or kinship structure requires it. (Ibid.: 183-4)

In view of the fact that Wellman's research largely took place in stable and well-

structured communities under ordinary circumstances, I propose that "structurally

embedded" strands may not be as common in crisis situations, especially one of

displacement. When one's survival is on the line, it is more likely than not that an

individual will attach more importance to and invest more effort in those ties that he/she

expects to be beneficial in the short term, whether in terms of receiving material or

immaterial support. Thus, for the purpose of this study, social network members are

considered those individuals that have provided the urban refugee in Dar es Salaam with

some type of support from the moment of flight from home onward to the time of the

research. 15

The body of research on support networks of the past two decades--none of which

took place in a situation of displacement or forced migration, and few outside the

Western hemisphere--have uncovered a number of common patterns. A first finding was


15 The name elicitor was formulated conform this definition and will be elaborated on further in the
methods section of this chapter.









that strong ties or ties between individuals with common characteristics (also called

homophilous ties) are more important conduits of social support than weak ties or those

between individuals with dissimilar characteristics (also called heterophilous ties)

(Wellman and Wortley 1990; Lin, Woelfel and Light 1985). Secondly, kin appeared as a

primary source of support while residential proximity proved essential in support

transactions involving material aid (Fischer 1982; Antonucci and Akiyama 1987;

Wellman and Wortley 1990). Lastly, at the network structure level, it seemed that

individuals who are embedded in dense, homogeneous networks receive more social

support in emergency situations than do individuals in wide-ranging networks (Marsden

1988; Smith-Lovin and McPherson 1993; Beggs, Haines and Hurlbert 1996). My

findings on the social networks of Dar es Salaam's urban refugees evaluate whether all of

these patterns hold true in an East-African context of forced migration in an urban

setting.

1.1.3.3 Gender, nationality and age

As mentioned earlier, I decided to focus on three social parameters to disaggregate

the collected data further, aiming to explore whether certain coping patterns, including

social networking, are different among specific social groups: men versus women,

Congolese versus Burundese versus Rwandese, and refugees belonging to different age

groups.

Gender. The study of gender in an African context, as mentioned earlier in this

chapter, has demonstrated the centrality of gender as an organizing principle of society as

it is elsewhere. On the one hand,

accounts of African women as the primary economic providers for their households
in their roles as farmers and traders and descriptions of the complicated gender and
generational rights and responsibilities for household production, consumption, and









distribution have challenged prevailing Euro-American understandings of marriage
and households as partnerships in which men are the main providers and household
resources are pooled for consumption and distribution. (Hodgson and McCurdy
2001:5)

On the other hand, it should also be noted that Christian male-oriented ideologies as well

as the colonial administration had profound effects on urban populations, through, for

example, the organizing of training sessions on "how to become the perfect housewife"

(Hunt 1990). Karen Hansen suggests that it is the visibility of women's economic work

in the fields and markets, together with the assumption that domesticity in Africa

consisted mainly of fragments of Western-style idealizations, that made scholars neglect

"both African reformulations and appropriations of elements of Western-derived

domesticity, and their variously constituted indigenous counterparts" (1992:14). While

fully subscribing to Hansen's contention that the domestic/public divide is not exclusively

organized by gender, and "gender relations are not only or even primarily negotiated

across a politico-jural/domestic divide" (17), I propose that local or popular perceptions

to the contrary may play an influential role in the coping strategies ofDar es Salaam's

refugees. Particularly with respect to the establishing of new ties in the process of

rebuilding one's social networks, my findings suggest that popular perceptions of gender,

of what it implies to be a man or a woman, constitute an essential element of life in exile.

Nationality. The second parameter I choose to define distinct social groups is

nationality. In the literature, African identity has traditionally been closely linked to

ethnicity rather than nationality. Nevertheless, from independence onward, integrationist

policies aimed at overcoming ethnic differences and opposition at the national level,

such as e.g., Mobutu's retour a l'autenticite during the 1970s, generally succeeded at

creating a nationalist identity without excluding the ethnic one. Biaya describes this









nationalist identity "like a garment that can be worn or discarded, depending on the

circumstances" (2001:59).

In a refugee situation, for example, individuals are likely to become more acutely

aware of the important role their respective nationalities may play in their lives as exiles.

The act of applying for asylum in a third country is usually, albeit not always, based

solely on an individual's being a citizen of a particular nation, as is the evaluation of that

application by the authorities of the asylum country. Manipulating one's nationality

hence becomes a strategy of survival. When, for example, the Tanzanian government

decided to forcibly repatriate the hundreds of thousands of Rwandese refugees from its

territory in 1996, more than twelve thousand Rwandese refugees managed to leave the

Rwandese refugee camps and register as Burundese refugees in camps established

exclusively for Burundese refugees (Rutinwa 2002:Internet). They stayed on in the

camps as Burundese even after the repatriation exercise, because from 1997 onward,

individualized refugee status procedures became mandatory for Rwandese, while

Burundese continued to be given primafacie refugee status. Only when it appeared that

over 99% of these applications were successful, did the twelve thousand Rwandese

refugees re-declare themselves Rwandese (Ibid.). As will be described in the following

chapters, the urban self-settled refugees in Dar es Salaam apply the same survival

strategy of "switching" nationalities as camp refugees when the need arises: Rwandese

may pretend to be Congolese, Burundese pose as Tanzanians, etceteras.

Apart from the obvious political implications, I propose that nationality equally

plays a role in the ways in which refugees adapt to the situation of forced migration

socio-culturally, on condition that the individual actually grew up and was socialized in









the context of his or her home country.16 Each country with its specific historical,

economic, political, legal, social and multiculturall conditions constitutes a unique

macro-level framework in which the socialization of its nationals takes place. Without

falling into a determinative, structuralist view of human life or neglecting the role of

agency, certain national patterns of behavior geared towards survival are likely to

emerge. Even as these patterns differentiate to a larger or lesser extent across socio-

economic classes, gender, ethnic or political affiliations or religious congregations, there

exist areas of common norms, values and behaviors often closely linked to perceptions of

one's national identity. Therefore, disaggregating the collected data by nationality should

bring to light existing commonalities among refugees from either Congolese, Burundese

or Rwandese nationality.

Age. In the wake of Evans-Pritchard's research on the Nuer in the 1930s, numerous

ethnographic studies on age in Africa took place in East-African, traditional societies and

focused largely on age-sets17 (e.g., Spencer 1976; Foner and Kertzer 1978; Bonte 1985;

Bernardi 1985). However, age plays a major role as well in African societies not

structured by age class systems. In the introduction to his edited volume The Politics of

Age and Gerontocracy in Africa (1998), Aguilar states that:

the passing of age as a universal, human and biological process can only be fully
understood by looking at its cultural--and therefore localized--processes of
construction, perception, change and adaptation. Perceptions of age do not arise
from a given definition but by sociality, from daily and habitual human activities .
With Africa at large in mind, it is possible to argue that older people are given



16 This argument is operationalized in the selection criteria of respondents (see later in this chapter).

1 Whereas the term age-set is used to refer to a set of persons who were initiated together as youth, and
who will go through the different stages of the age-system together, the term age-grade is used to refer to
the respective stages of the life cycle (Spencer 1976)









some kind of recognition, and that old age seems to be regarded as an attribute
rather than a hindrance. (9)

At the same time, the functionalist perspective of the majority of the studies on age class

systems allowed little, if any, space to incorporate the influence of social change on the

formal structures of age in these societies. The changing perceptions of age among

growing numbers of individuals with different ethnic affiliations in Africa's urban centers

are insufficiently documented in the literature. Increasing education levels, for example,

among the younger (more urbanized) generations, both male and female, are known to

challenge the traditional authority of the older generations in particular situations. And

yet, even

in the 1990s mzee is [still] used in East Africa to show respect for older men
[and women]. Elderhood, however, is not always associated with an active
political role, but sometimes with the non-threatening role of advisor and councilor.
(Aguilar 1998:22)

The daily confrontations and contestations between young and old, whether men or

women, are part and parcel of the ever changing social structures of current day African

societies, particularly, but not exclusively, in the urban context. As with gender, the

social roles that come with age are cultural constructs embedded in larger economic,

political and historical structures. As such, persons of different generations are more

likely than not to display different socio-culturally inspired responses in a particular

situation. It is this logic that made me decide to add age as the third social parameter to

this study of forced migration and social networks.

Arriving in Dar es Salaam, Congolese, Burundese and Rwandese refugees have left

behind a substantial part of their social worlds. Relatives and close friends from home

are not always reachable through telephone or e-mail, even letters may not arrive at their

destination depending on the severity of the situation back home. In addition, they









experience how difficult it is to locate compatriots in a city of three million inhabitants

"where the majority live in unplanned areas [and] where most streets and houses are

not named or numbered" (Kironde 2000:53). To complicate matters even further, there

exists a lot of suspicion and distrust among nationals from the same country usually for

political and/or ethnic reasons. And last but not least, for lack of legal documents, the

urban refugees have to be wary of being arrested by the Tanzanian Immigration Police at

any time. Reconstructing one's social world in these particularly challenging

circumstances is a task tackled in multiple ways as we shall see in the following chapters.

But first, I would like to present the arguments underlying my choice of methods.

1.2 Choice of Methods

Before going from theoretical considerations about the topic in the various

disciplinary fields to the actual geographical research site, important matters concerning

the phenomenological and methodological options of the researcher need to be addressed.

Traditionally, field experience was a conduit sine qua non for a piece of research to be

considered anthropological, based on the perception that only participant observation

allows the researcher to observe and write down in a scientifically objective manner the

"other" culture or society waiting to be discovered (Gupta and Ferguson 1997).

Increasingly, however, the inevitability of the subjectivity or "social, cultural and

political location" (5) of both researcher and those being researched came to be

recognized. Whereas nowadays "participant observation continues to be a major part of

positioned anthropological methodologies, it is ceasing to be fetishized" (17) and is

usually complemented by a vast array of quantitative and qualitative methods.

In addition, traditional anthropological fieldwork usually took place in village

communities where participant observation was considered to be "the" data collecting









tool par excellence. This study, on the other hand, takes place in an urban setting and

among individuals who do not form "a community" in the traditional anthropological

sense of the word. In order to hide their refugee status from the Tanzanian authorities--

occasionally even hiding from their own countrymen for political reasons--the urban

refugees in Dar es Salaam are forced to live a clandestine life.18 For the researcher, this

means that relying mainly on participant observation by talking and living with the

members of the refugee community is not really a viable option, and the necessity of

additional data gathering tools ensues.

1.2.1 Collecting Data

The goal of this research project, linking the adaptation processes, including the

(re)configuration of forced migrants' social networks, to the larger worlds surrounding

them, requires the collection of both quantitative and qualitative data. In view of the

quantitative nature of the social network data, a myriad of qualitative collecting methods

is called for in order to assure the correct interpretation and contextualization of the

quantitative data.

Within the field of social network analysis, there are two basic concepts of what

constitutes a social network, namely socio-centric or "whole" networks, and ego-centric

or personal networks (Johnson 1994: 117). Socio-centric networks represent the ties

between all members, two by two, of a particular group of persons. This set of relations

between a limited number of people is often visualized in a matrix, whereby all the

members of the group are laid out on both the X and Y-axis. The number on the position


18 Notwithstanding, sharing similar experiences with regard to the flight from the home country and carving
out a new life in an unfamiliar environment, they form what Malkki has called an accidental community of
memory consisting of individuals with "accidental sharing of memory and transitory experience" (Malkki
in Gupta and Ferguson 1997: 91).









where a person on the X-axis and another person on the Y-axis meet usually indicates the

intensity or another tie-attribute of the relationship between those two persons. An ego-

centric network, on the other hand, is the set of relations one person--called ego--has with

a certain number of other individuals--called alters. This type of network is often

visualized as a star, whereby the length of the line between ego and one of his/her alters

indicates the intensity or another attribute of a certain type of social relationship between

these two persons.

Whereas the bulk of the research into the statistical analysis of social network

analysis concentrates on socio-centric networks (Freeman et al. 1989; Wasserman and

Faust 1994), ego-centric networks as well received a substantial amount of attention with

regard to the fine-tuning of data-collection and statistical analysis techniques (e.g., Straits

2000; Brewer and Webster 2000; Brewer 2000; Van Duijn et al. 1999; Louch 2000). The

clandestine nature of the life of the Congolese, Burundese and Rwandese refugees not

permitting research on socio-centric networks, this study therefore concentrates solely on

ego-centric social networks. Traditionally, ego-centric or personal networks have proven

to be excellent tools to engage in the study of social support in fields such as clinical and

community psychology, as well as sociology and psychology (Wasserman and Faust

1994: 42; Wellman 1981; Lin, Woelfel and Light 1985, Smith-Lovin and McPherson

1993; Schweizer, Schness and Berzborn 1998; Wellman and Gulia 1999). Applications

in the field of disaster studies have been far fewer (e.g., Beggs, Haines, and Hurlbert

1996; Haines, Hurlbert and Beggs 1996).

However, it is essential at this point to draw attention to the difficulties raised by

many social researchers of doing cross-cultural studies using methodologies that are

based on purely Western concepts and instruments (Ahearn 2000). In an attempt to limit









as much as possible the collection of inaccurate data by using culture-laden concepts, the

questions pertaining to social networks asked in this study focused exclusively on: (1) the

attributes of ego and his/her network members (e.g., gender, age, nationality, educational

level, marital status, profession, etc.); (2) the type of tie (e.g., kinship, friendship,

acquaintance, emotional support, financial support, etc.); and (3) the intensity of the tie

(length of relationship, frequency of contact and place of residence).

After an initial period in the field setting, I decided to reformulate the principal

question eliciting the social network data, also called name generator, from the more

generally used, abstract "Could you tell me who are the most important people in your

life?" to the more practical and concrete: "Could you give me the names often persons

who have helped you from the time you left your home and during your life in Dar es

Salaam? They may include persons from any nationality, men and women, no matter

where they live, and who helped you in any way, be it financially, emotionally,

information-wise, socially, neighbourly, etceteras." In Swahili, the term hali na mali

("situation and riches") was used to indicate that the term "help" was to be interpreted in

both its material as well as immaterial sense. This reformulation of the name generator

was necessary because replying to its first version, respondents would start listing

persons who had been important to them at one point in their life long before the current

situation was forced upon them (e.g., during childhood), and with whom they had since

lost contact. Considering that this is a study on "forced migration and social networks" it

was primordially important to guide people to the current phase of their lives to elicit

information on the persons playing an important role at that particular point in time. It

should be mentioned at this point, that because of the sensitive and precarious situation of

the undocumented urban refugees, I never insisted on collecting the real names of









respondents or their social network members. However, on more than one ocassion, the

informant, after initially introducing him or herself to me under a pseudonym, decided to

reveal his/her real name once a relationship of trust had been established.

"It has been said, rather facetiously, that qualitative research is what you do when

you don't have enough data or don't understand statistics" (Ahearn 2000:15). Clearly,

would it not be more precise to state that instead of focusing on the elusive goal of total

objectivity, qualitative research methods implicitly recognize subjectivity through their

reliance on other people's voices collected through narratives, interviews, life histories,

focus groups and participant observation? In addition, certain types of data cannot be

ethically or methodologically quantified, such as narratives. It is in this sense that

quantitative and qualitative research methods should be considered complementary

instead of each other's opposite. The premise on which qualitative research is based is

"the value of the insider's view: to understand the world from their perspective, rather

than categorizing their experiences out of context or from the outsiders' view" (Omidian

2000: 42).19 The qualitative data or "voices" collected in this study through interviews,

(partial) life histories, media articles, and other published materials represent an avenue

towards uncovering and understanding the multiple interpretations and meanings of

refugee life in Dar es Salaam, and constitute the larger part of the contextualization of the

data collected.

1.2.2 Collecting "Voices"

The "Kuhnian disrobing of'timeless' scientific authority" (di Leonardo 1991: 2) in

the 1960s was one of the major elements of feminist criticisms of traditional social


19 What anthropologists have traditionally called the emic versus the etic view.









science methodology. It was based on a "general rejection of positivism, its claim that

science is value neutral, and [the false conviction] that the scientific method protects

against contamination of findings by subjectivity" (Jayaratine and Stewart 1991: 89).

Feminist researchers at the time blamed quantitative methods for concealing women's

real experience, and subsequently became proponents of qualitative methods as methods

which permit women to express their experience fully and in their own terms.20 The

resulting body of literature on qualitative methodology was generally based on the actual

experiences of the feminist researchers and on the theoretical and methodological

reflections that ensued thereof (Stanley and Wise 1990). Even as it was feminist

scholarship that gave the impetus to the development of this body of literature, its

findings are extremely useful to other social science research as well. Feminism aimed to

give a voice to women who were unheard in a traditionally white, male and Western

world of research, yet at the same time, attention was drawn to the unheard voices of

people--women and men--who were not white, not Western, or simply did not belong to

the privileged layers of society. It is in this sense of giving voice to the unheard ones,

that I consider it worthwhile to have a closer look at the feminist reflections on the use of

qualitative methodologies.

The most self-evident way of giving a voice to those unheard is through personal

narratives, be it life histories, structured or unstructured interviews, etc. Personal

narratives allow the expression of people's experiences and are, as such, a very popular

research method among social scientists. They have produced a wealth of information on


20 Nowadays, there is a growing consensus among feminist researchers that both quantitative and
qualitative research can result in erroneous and misleading results, and that "there can be no single,
prescribed method or set of research methods consistent with feminist values" (Jayaratine and Stewart
1991: 100).









human life and experience previously omitted or overlooked: "what could be truer [sic],

after all, than a subject's own account of what he or she has lived through?" (Scott 1992:

24). Yet, it is exactly this reliance on "experience" as the undisputed evidence that Scott

subsequently puts into question. She argues that experience itself is a subjective

interpretation of events, and that the "project of making experience visible" obscures "the

workings of the ideological system itself' (25). Social scientists therefore need to include

into their analyses the discursive construction of differences and categories that subjects

utilize to describe their experiences.21 Canning (1994), on the other hand, goes a step

further and refuses to accept that discourses are completely determinative of people's

experiences in the sense that "subjects do have some kind of agency, even if the

meanings they make depend on ... the discourses available to them" (377). Moreover,

discourses not merely constitute the domain in which subjectivities emerge, they also

supply the actual arguments for people to challenge their uncontested applicability and to

endeavor changes (395-6).

Applying the above reflections to this particular study of urban refugees in Dar es

Salaam, it becomes clear that the refugees' personal narratives need to be contextualized

through analyzing other texts as well, such as the discourses of refugee agencies, of the

Tanzanian government and local authorities, of the media, and of Dar es Salaam's

Tanzanian residents. Because of the international nature of the research setting and the

politico-historical laden-ness of the refugee situation, time must also be spent on the

political, economic and socio-cultural history of the region. In addition, there will be

ample illustrations of Canning's argument, namely that the refugees in this study exercise


21 This emphasis on representation or discourse, as opposed to the pursuit of a discernible, retrievable
historical "reality" is also known as the "linguistic turn"(di Leonardi 1992:22).









agency through re-interpreting the existing discourses--of internationall agencies as well

as other institutions--to their advantage in the pursuit of individual and personal goals.

1.2.3 Representativeness

It is one thing to contextualize the narratives of the persons interviewed through

discursive analysis (i.e., locating the positionality of the interviewee) however, it is

equally important to take into account the positionality (role and setting) of the

interviewer and the interpretation thereof by the interviewee (Gluck and Patai 1991). This

point is beautifully illustrated in Knudsen's experience as a Canadian researcher among

Vietnamese refugees in South East Asian refugee camps (1991; 1990; 1983). His study

draws attention to the role of the investigator in the construction of the refugees' life

history stories and of the transit refugee camp itself.22 He found that a refugee will

"construe a life history in an attempt to reduce the uncertainty of his/her life course and

simultaneously secure his/her rights towards the future" (1990: 123). These survival

strategies led the refugees to construct life histories that forewent personal past

experiences in order to fulfill the ascribed role of refugees, especially where it concerns

refugees who will resettle permanently, and who thus tend to "perceive the present camp

sojourn in itself as a liminal phase between the past and the future" (125). Eventually, to

some camp dwellers, the story which was first meant to serve as a strategic construct,

started to live its own life, the person no longer able to split the story s/he constructed

from the reality invented. Knudsen's role as a Canadian investigator was perceived by

the refugees as "related to" the Canadian immigration officials who were to decide on the

refugees' individual immigration statuses. Ultimately, in the refugees' life histories past


22 Because the refugees residing in the transit camp were all in the process of being resettled or seeking
resettlement to Western countries.









and present were subconsciously negated in favor of the future due to the specificities of

the refugee transit camp and the nationality of the investigator.

Langness and Frank (1985) have pointed out that this phenomenon is not exclusive

to refugees: "... the very act of asking questions may make interviewees "recreate"

themselves in the sense that, in looking for a coherence in their life histories, they often

tend to reinterpret the past through the present" (107). In addition, the narrating of the

story in itself can be empowering, validating the importance of the speaker's life

experiences (Oliver-Smith 1986; Gluck and Patai 1991). For example, with regard to the

traumatic experiences of the Mayan Indian widows, Zur (1997) found that the women

resisted the official version of events by producing an alternative unofficial discourse:

As they uncover memories, the women rework, relive, make sense of, come to
terms with and integrate the traumatic events which dramatically changed their
lives, thus giving a boundary to their suffering. Together they attempt to construct
a narrative of the past which makes sense to them. (66)

In recounting to each other the memories they had of the events that occurred, the women

were primarily concerned with how this past related to their present lives, not with the

accuracy of the events that actually had transpired. The same coping strategy was

applied by Burundi refugees in the Tanzanian refugee camps when constructing their own

"mythico-histories" about the events that forced them into exile (Malkki 1995a).

Whereas the use of life histories, memories and oral histories will assist the

researcher in being "more sensitive to situationality and variability" (Indra 1998: 19),

informants' re-interpreting of the past in order to give meaning to the present, on the other

hand, does raise the question of the "representativeness" of the narratives collected. The

important questions at this point are: how does one determine what or who is

"representative" and who decides which norm to apply? Or in other words: should









researchers apply themselves "to produce general ethnographic descriptions of people's

beliefs or actions" by "smoothing over contradictions, conflict of interests, doubts, and

arguments, not to mention changing motivations and historical circumstances" (Abu-

Lughod 1993: 9)? Susan Geiger (1992) cautions that a methodology "that features as

major conceptual organizers the positional markers of marginality and representativeness

automatically privileges certain voices and obstructs others through the very framework

imposed" (309). In order to counter that tendency, she advocates the concept of "multiple

truths" to be the leitmotiv for researchers intending to describe people's lives.

Subscribing to Geiger's advice, I intend to give voice to a whole range of individuals,

who are likely to have contradictory experiences and interpretations thereof, despite of all

being termed "urban refugees."

1.3 In the Field

With the aim to take into account all of the above considerations in order to be

attentive to "the different forms of knowledge available from different social and political

locations" (Gupta and Ferguson 1997:37), the methods applied in this study include--in

addition to participant observation and ethnographic interviewing of the refugees--the

following: administering a survey to a number of urban refugees in Dar es Salaam,

administering another survey to a group of Tanzanian residents of Dar es Salaam,

interviews with government officials and staff members of the international and national

organizations concerned with the refugee population in Dar es Salaam, collecting relevant

newspapers articles and analyzing government documents as well as other official and

academic publications with regard to both Tanzania's refugee populations and the

situations in the three major refugee producing countries, i.e., Congo, Rwanda and

Burundi. In addition to determining which methodological tools to apply, the selection









criteria for respondents needed to be thought out as well as strategies to locate informants

and convince them to participate in the survey and share their experiences with an

outsider.

1.3.1 Methods Used

Collecting quantitative and qualitative data as mentioned above was done through a

myriad of methods applied to different groups of people and through different media. In

order to complement participant observation and ethnographic interviewing of

Congolese, Burundese and Rwandese informants, I decided to concentrate on the

following six data collecting activities.

(1) A Refugee Survey was developed in French23 (see copy in appendix c) soliciting

information about the events causing the respondent to flee his/her home country, the

motivations for going to Dar es Salaam instead of heading for one of the refugee camps,

and the first few days after arrival in Dar es Salaam. Included were also questions on the

perceived main differences between his/her life at home and in Dar es Salaam and the

motivation to (eventually) repatriate or remain in exile. Finally, detailed information is

requested about ten social network members who helped the respondent from the time of

the flight onward, and individuals' opinions regarding certain aspects of life in Dar es

Salaam, such as religious activities, establishing relationships with Tanzanians,

encounters with the Immigration police, etc. This survey was administered to 300

refugees in total: 100 of each nationality (Congolese, Burundese and Rwandese), equally

divided between men and women. In addition, 30% of the respondents were between 20


23 The main European language used by nationals from the Democratic Republic of Congo, Rwanda and
Burundi is French. In Congo, French is the language of tuition in schools from the 4th year in primary
school onward. In Rwanda and Burundi French becomes the only language of tuition from the 1st year of
secondary school onwards.









and 25 years of age, 30% between 26 and 35, another 30% between 36 and 45 and 10%

older than 45 years. The survey was administered by a research assistant in the mother

tongue of the respondent. I strived to have a balanced sample as large as possible in

order to increase the statistical significance of my findings. Three hundred was found to

be the maximum numbers of respondents that could be located and surveyed in the time

space of nine months.24

(2) A Tanzanian Survey was developed in English, later translated in Swahili (see

copy in appendix B), and solicited information regarding the Perceptions on Refugees

among the Tanzanian population in Dar es Salaam. It recorded the level of acquaintance

the respondent had with refugees, the respondent's personal opinion on refugee policy

vis-a-vis urban refugees and his/her perception of the professional activities of urban

refugees according to gender and nationality. Equally noted were existing perceptions on

the proportionate numbers of refugees in Tanzania and in Dar es Salaam and how the

respondent would react to the marriage of his/her son or daughter with a refugee.

Ultimately, the questionnaire solicited perceived positive and negative characteristics of

refugees according to gender and nationality, as well as perceived differences between

Tanzanian men and women on the one hand and refugee men and women of the three

respective nationalities on the other hand. This survey was administered by an

experienced Tanzanian social researcher contacted through the University of Dar es

Salaam. A total of 216 respondents were interviewed, equally divided between men and

women. Other parameters applied to have a balanced composition of the sample were the


24 See Bloch 1999 for an interesting overview of the methodological considerations of surveying refugees.









socio-economic status of the place of residence, age, religious affiliation and, the level of

education.

(3) Most government officials interviewed belonged to the Department for Refugee

Affairs at the Tanzanian Ministry of Home Affairs, a few to the Immigration Department

in the same Ministry. The only international agency concerned with urban refugees in

Dar es Salaam is the UNHCR. Many interviews were held with different officials in the

Protection unit, Community Services unit and External Relations unit. Among the

national agencies contacted were Umati, Tanganyika Council for Refugee Services, the

Legal and Human Rights Office, the local branch of Amnesty International and the

Center for the Study of Forced Migration at the University of Dar es Salaam.

(4) National newspapers browsed were: The Guardian, The Daily Mail, and The

African (dailies), and The East African, The Sunday Observer, and The Business Times

(weeklies) over a period of approximately 3 years, from 1999 onward.

(5) Government documents include the 1998 Refugee Act, the 1995 Immigration

Act and others. Among the academic publications are a number of extremely interesting

Master's theses from the University of Dar es Salaam in both Sociology and Political

Science focusing on certain refugee matters relevant to Tanzania. One very important

official publication was the preliminary policy review by five scholars, affiliated with the

Center for the Study of Forced Migration at the University of Dar es Salaam, done at the

request of the Prime Minister's office. Other were the UNHCR Newsletters, and

publications by various other refugee agencies.

Having determined and developed the above listed methods as the ones to be used

in the field, the time has come to establish the selection criteria for respondents and to

locate the informants.









1.3.2 Selecting Respondents

Before going into the ethnography of the refugees of Congolese, Burundese and

Rwandese descent residing in Dar es Salaam, it seems advisable to first have a closer

look at the label "urban refugee." From the literature there appears to exist an increasing

diversity in the composition of urban refugee populations, as well as in their motives to

head for the urban centers instead of the refugee camps. In one of the earliest references

to urban refugees as a separate category of forced migrants, albeit one continuously

increasing in numbers, Gould (1974) defines them as "educated, skilled, often politically

active, and involved with freedom fighters of governments in exile" (416). Subsequent

studies however, the majority of which took place among Eritrean and Ethiopian refugees

in Sudan's capital, Khartoum in the 1980s and 1990s, showed that over time the

proportion of non-urbanites became more important, and that their reasons for preferring

to reside in Khartoum instead of the refugee camps covered a whole range of motives:

search for education, rural to urban migration, ambitions to emigrate, etc. (Weaver 1988;

Goitom 1987; Karadawi 1987; Kibreab 1996). In addition, increasing rates of

urbanization on the African continent have had a determinative influence on the

rural/urban ratio of Africa's refugee populations. Contrary to the situation in the 1970s,

the current profile of the African urban refugee is far more diversified.

The diversity in the composition of the refugee populations of Congolese,

Burundese and Rwandese descent residing in Dar es Salaam, and their motives for going

there in the first place instead of one of the refugee camps will be presented in one of the

following chapters. Given the long history of refugee influxes from the Great Lakes

region in addition to migration flows and inter-marriage patterns among neighbors, I

estimated that defining the selection criteria of respondents deserved thorough









consideration. After studying the existing situation through consultation with Congolese,

Burundese and Rwandese friends "on the ground" as well as going through the existing,

albeit scant, literature, I arrived at the following three general selection criteria.

In view of the fact that this research project focuses on the reconstruction of

individuals' personal social networks in the wake of their forced migration, the first

criteria was that respondents should have a maximum often years, yet a minimum of one

year of residence in Dar es Salaam.25 This selection criterion thus excludes those

Congolese, Burundese and Rwandese who sought refuge in Dar es Salaam in the 1960s

and 1970s, quite a few of whom are now Tanzanian nationals, particularly among the

Rwandese. Secondly, interviewees should have recollections of life at home as an adult,

i.e., older than 15 years of age at the time of leaving home, and these recollections should

be no older than 10 years in order for the individuals to be able to compare social

relations at home and in the country of asylum. This second criteria excludes many

Burundi refugees who spent over 20 years, sometimes 30 years, in exile in (then) Zaire,

and who came to Tanzania only because of the wars in 1996 and 1998 (see chapter 2).

The third criteria pertains to the motive for leaving the home country, and relates to the

principle of forced migration. In addition to persons deciding to leave home for political

reasons or the dangers of war, I decided to also include individuals who came to Dar es

Salaam for lack of economic opportunities at home as a result of the political situation in
26
the Great Lakes region.26 Having established these three general selection criteria for




25 E.g., someone who spent 5 years in a Tanzanian refugee camp, and came to Dar es Salaam only 6 months
ago does not respond to this criteria.

26 E.g., a Rwandese university lecturer lost his job of many years to a "returnee" from Uganda who was
more closely related to the current regime in Kigali than he was.









both respondents and informants, the next step was to further fine-tune the sampling

method with regard to the demographic parameters of the population.

Chambers (1979) estimated that in 1976 only 1.2% of African refugees were urban,

the majority of whom unassisted (382). By 1986 however, Rogge speculates that

"although there are still no specific data on actual numbers of urban refugees for the

continent as a whole, the figure may now be as high as one third the total" (8). If this

holds true, there might be as many as one hundred thousand refugees residing in Dar es

Salaam. Notwithstanding the general consensus on the fact that urban refugee

populations are steadily increasing in size, there remains a lot of conjecture on their

demographic structure. Contrary to widely held assumptions that urban refugee

populations are comprised predominantly of young, single males (Sommers 2001;

Goitom 1987; UNHCR 1997c), recent statistical information shows that "this is not the

case in many cities where substantial numbers of refugees are to be found" (UNHCR

2001c: Internet) as is the case in Dar es Salaam.27 Based on UNHCR's finding that larger

urban refugee populations are more diversified demographically and in order to avoid a

probable bias resulting from free snowball sampling in a clandestine population, I

predetermined the following three demographic parameters.28 The first one was

nationality: the aim was to locate one hundred respondents of each nationality,


27 Given that young males, for reasons of language and culture norms, are more open to contact with
outsiders than women or older men, it is easy to see how this perception came about. The description of
the sampling method used all too often missing in the write-up (Sommers 2001; Goitom 1987), the reader
is left to speculate on the level of its randomness. This holds especially true for clandestine populations
where snowball sampling appears the only indicated method of locating respondents. It is in this context
that I decided that without some pre-established demographic parameters, the likelihood of a skewed
sample (i.e., one that reflects the demographics of the initial contacts) would be all too high.
28 In addition to pre-setting the three demographic parameters, and being dependent on the snowball
method to locate informants, I considered it counterproductive to an unbiased analysis of people's social
networks to survey more than two individuals listed among one respondent's social network members, lest
there should be too much repetition in the alters listed.









Congolese, Burundese and Rwandese, totaling three hundred persons. Secondly, gender

was to be equally represented in each group of one hundred respondents, i.e., fifty adult

men and fifty adult women. Thirdly, four age categories were defined. The principle

underlying the defining of the four age categories was based on marital and child status,

because both marital and child status compliment the role of age in the societal status of

an individual in an African context. Persons younger than 25 years of age are generally

considered to be not yet married and childless, while persons between 26 and 35 years of

age are much more likely to be married as well as have young children. The next

generation, the 36 to 45 years olds, more frequently have larger families consisting of

adolescent children generating pre-occupations of a different nature than the two younger

age categories. The oldest age group in this study consists of persons over the age of 45,

the majority of whom will likely have children of an adult age.29 Having established who

I wanted to survey and interview, the next step was to locate informants.

1.3.3 Locating Informants

One of the most frequently asked questions, when I explain to colleagues and

friends the topic of my research and the fact that the Congolese, Burundese and

Rwandese refugees are illegal residents in Dar es Salaam, is: "Well, if these people are in

hiding, how are you able to contact them and why would they agree to talk to you?" As it

is an extremely pertinent question in this particular research situation, I considered it

worthwhile to elaborate at some length on my experiences in the process of locating

informants and gaining their trust.




29 For the final number of respondents surveyed per nationality, gender and age, refer to Table 3-1 in
chapter 3.









"Trust" is an issue of the utmost relevance to the process of collecting data among

refugee populations, and particularly so with regard to the overwhelming majority of the

urban refugees in Dar es Salaam. The concept of "trust" in Daniel and Knudsen's

Mistrusting Refugees (1995) raised a good deal of dust in the academic world of refugee

researchers (Bisharat 1997; Schwartz 1997; Geuijen 1997; McGovern 1998; Voutira and

Harrell-Bond in Daniel and Knudsen 1995: 207-224). While I agree that assuming

"trust" as an apriori state of humanity does not really have a scientific basis, it is

impossible to deny that becoming a refugee means having to deal with a huge increase in

the lack of predictability of day-to-day life as it was known to them before the flight. Or

as Bisharat puts it: "... trust ... is shattered when refugees are created by systematic

violence and mistrust .. replaces it and stubbornly persists, even when the

conditions that originally fostered the mistrust have abated" (1997: 664).

The high level of "mistrust" existing in refugee camps and settlements, is even

compounded in a situation where refugees opt to self-settle elsewhere, thereby

contravening the explicit official policies of the country of asylum, as is the case for the

refugees in Dar es Salaam. Outside of the officially designated areas in Western

Tanzania, refugees in Tanzania run the permanent risk of being arrested and, as illegal

immigrants, may even be deported to their home country by the Tanzanian Immigration

officials. In such circumstances information becomes a powerful weapon, that, once

released, may be turned against the refugee who lost all control over it. Suspicion runs

high, especially vis-a-vis strangers, such as social researchers who go around asking all

sorts of questions and spreading the word that they are particularly interested in talking to

refugees.









During the first months of predissertation research in the Spring of 2000, I was

fortunate enough to meet with a person closely linked to the University of Florida, who at

the same time happened to be married to a Congolese History professor teaching at the

University ofDar es Salaam. Through her, I met my first Congolese assistant, who was

not a refugee but a legal resident of Tanzania. My assistant and his family were well

integrated and respected in the Congolese community in Dar es Salaam, which consists

not only of refugees, but also of academics, businessmen, diplomats, etc. The ward of

Kinondoni30 was by that time already locally known as the "place where all the

Congolese live,"31 and it was here that I spent many hours during the predissertation

phase of my research project talking mainly to Congolese men and women. Most of

them had come to Dar es Salaam in the 1990s, some registering as refugees with

UNHCR, others accepting being undocumented as an unavoidable fact of refugee life.

Congolese informants were located using the snowball method, and contacts were fairly

easily established due mainly to two reasons: the link to the parents of my assistant, who

were well-established in the community and widely respected, and, secondly, I surmise,

my Belgian nationality. Despite a violent colonial history and the, at times, ambiguous

postcolonial relationship between the two countries, I gathered from contacts with

Congolese informants that at the individual level there still exists a certain level of

affinity towards Belgium and Belgian nationals. Speaking French was an

unquestionable, even indispensable, asset, as many of the educated Congolese refugees

expressed their joy at the opportunity of engaging in conversations and discussions in



30 Centrally situated between Manzese and Mwenge on the Dar es Salaam map 1-7 in appendix a.

31 In 2002, a popular Swahili song by a Tanzanian singer refers to Kinondoni as "Congo."









French. At the same time, few, if any of the Congolese informants appeared at any time

uncomfortable walking around with a white person, thus attracting attention in a popular

neighborhood such as Kinondoni. I was able to meet with people in their homes, at their

place of work, in public places such as local bars, or on the street.

By the time I returned to Dar es Salaam in the Spring of 2001, my research

proposal included refugees of Burundese and Rwandese descent as well for the purpose

of cross-cultural comparisons. A Rwandese research assistant was recruited through

contacts at the University of Dar es Salaam, a Burundese assistant through the French

Cultural Center (Alliance Francaise) in Dar es Salaam. The Rwandese young man's

father had sought refuge in Tanzania in 1959, indicative of his Tutsi background, and had

since acquired Tanzanian nationality. The young man having grown up in Tanzania

spoke only English and Swahili, and never having visited Rwanda, he mastered neither

Kinyarwanda nor French. After numerous fruitless attempts to convince some Rwandese

refugees of the 1990s to participate in the survey and eventually be interviewed by me,

we agreed this line of action was not going to work. At the same time, the Burundese

assistant had little success as well, claiming that the political tensions between Burundi

and Tanzania at the time32 had frightened his Burundese refugee friends to the extent that

they categorically refused to have any part of the research project, let alone meet with the

researcher. Clearly, contacts with the Rwandese and Burundese refugee populations in

Dar es Salaam were not going to develop as smoothly as with the Congolese.




32 In August and September 2001, the press had reported incursions of the Burundi army into Tanzanian
territory claiming Tanzania was supporting the Burundi rebels. The reaction of president Mkapa was to
forthwith announce the involuntary repatriation of all Burundese from Tanzania. This announcement
caused a great deal of unrest and anxiety among Dar es Salaam's Burundese residents.









Over time, continuing the meeting and interviewing of Congolese men and women,

another Congolese research assistant was recruited, a young woman who at one time was

a teacher at the Groupe Interscolaire des Grands Lacs33 through which she had

established contacts with the Burundese and to a lesser extent the Rwandese community

in Dar es Salaam. Having learned by now, that for a variety of reasons that will be

elaborated on in the following chapters, it would be even more difficult to gain access to

the Rwandese refugees than the Burundese, I decided to concentrate on the last ones

first.34 After a few initial conversations and interviews with several Burundi young

men35 through my assistant's contacts from the school, one of them was recruited with the

sole purpose of locating Burundi respondents according to the selection criteria set out

earlier. We devised a routine whereby initial contact was made by the Burundese

assistant inquiring about the willingness of the person to participate in the survey, and if

positive, subsequently setting up an appointment with the Congolese research assistant

who by now was well trained in administering the survey and eliciting the social network

data.








3 This is a school founded by Congolese, Burundese and Rwandese nationals in Dar es Salaam, where
close to 300 children at primary and secondary school level are instructed the French/Belgian school
curriculum in French by around twelve (often professional) teachers, often themselves refugees from the
Great Lakes region.

34 Another possible avenue to the Burundi refugee community in Dar es Salaam through a local NGO
;. ,i ....i. ah I ,,. 't Borders, founded by a Burundi mzee, did not work out either. The several persons I was
referred to by the Director of this NGO, turned out to be political activists and members to one of the
Burundi rebel groups, who wanted to check out whether I was really a social science researcher, but who
were further unwilling (or unable) to introduce me to potential respondents and/or informants.

35 Young men are always the easiest point of entry into any African refugee community in my experience.









The third step was to establish a place and a time, determined by the Burundese

respondent, to meet with me for an interview.36 Usually, both the Burundese and

Congolese research assistants accompanied me on the first meeting. After explaining to

the informant at length (in either French or Swahili depending on his/her preference) who

I was and why I wanted this meeting and ask all sorts of questions, I went over the social

networks data, trying to slowly establish contact and gain trust. With a limited number of

Burundese informants, I made arrangements for further conversations to which neither

assistant would be present. Participant observation, or observation tout court was not

really a viable research method in these circumstances. Most of the Burundese preferred

to meet with me in public places, often in a ward different from the one they resided in,

obviously in an attempt to not attract the attention of their Tanzanian neighbors by taking

advantage of the anonymity a big city such as Dar es Salaam with its three million

inhabitants has to offer. By the time I completed six months of research, in December

2001, I had succeeded in reaching the research goals established for the Congolese and

Burundese urban refugees, but it was still unclear whether I would be able to accomplish

the same among the Rwandese forced migrants.37

In the Summer of 2002, I returned to Dar es Salaam for the final three months of

research focusing this time almost exclusively on the Rwandese refugee population.

Fortunately, during my absence my Congolese assistant and two other Congolese young

men conversant in Kinyarwanda, had persevered in trying to not just establish contacts,


36 In accordance with Spradley's definitions, a respondent is a person who participates in a survey, an
informant agrees to ethnographic interviewing and more frequent personal contacts with the ethnographer.

7 A few months earlier, I had made another attempt at recruiting another Rwandese assistant. This young
woman appeared very enthusiastic, but it turned out that suspicions run so high among the Rwandese--as I
was to learn later--that she refused to give me even her mobile phone number, let alone any other indication
as to how I would be able to contact her, thus impeding any fruitful collaboration.









but more importantly, convince their Rwandesefreres et soeurs ("brothers and sisters")

that participating in the survey or agreeing to talk to me, would not put them in any

danger of being discovered by the Tanzanian authorities.38 In view of the Rwandese

reluctance to divulge any information on paper about their social networks for fear of the

Tanzanian police and the International Crimes Tribunal for Rwanda (ICTR) in Arusha, I

had to change the routine used so far. One of the Congolese middle persons would

arrange for a meeting, at which my main goal was to convince the Rwandese of my "real"

identity by pulling out not only a letter of introduction by the Department of

Anthropology (that I had not had to use even once until then) but also my Gator-1 student

card. The first Rwandese persons I met were, as usual, young men. Most would agree to

meet with me on a second or third occasion for ethnographic interviewing, but in the

absence of any of the Congolese assistants. When a relationship of trust was finally

established, most would also agree to provide the social network data, especially when it

was emphasized that no real names needed to be revealed. This new routine worked well

for young men and women in their twenties and early thirties, but it appeared after a

while that, unfortunately, despite numerous attempts older Rwandese would continue to

categorically refuse to meet with me. They would however, accept to be surveyed on the

social network data and a few other questions by the female Congolese assistant, because

she was notre soeur ("our sister").39 Ultimately, I succeeded in collecting the Rwandese


38 The specificities of the pre-occupations of the Rwandese refugees in Dar es Salaam are complex and
many, and will be elaborated on later. Basically, Rwandese refugees in Dar es Salaam, many of whom
(linked to) former political figures, are afraid of being accused of participating in the master-minding of the
1994 genocide.

39 One was even interviewed and recorded on tape by a Congolesefrere. This Rwandese refugee agreed to
the interviewing and to the taping, but he was too afraid to meet with me personally, as there was the
slightest chance that "as a white person" I should be connected to the UN International Crimes Tribunal for
Rwanda in Arusha.









social network data on Rwandese respondents of all ages, yet had frequent ethnographic

interviews with men and women between 20 and 35 years of age only.

Clearly, not every single Congolese or Burundese refugee contacted agreed to talk

to me, yet, contrary to the Rwandese, the former appeared to base their decision on

individual and personal motives rather than the structural or collective reasons

influencing the decision of the latter. On the basis of feedback from my research

assistants and a number of informants, I learned that whereas the Congolese perceived me

to be the representative of an international NGO, sometimes even of UNHCR, looking

for a project to assist the refugees, the Burundese usually saw me as a possible gateway

to a third country, while the Rwandese would only grudgingly accept that I was not in

any way linked to the International Crimes Tribunal for Rwanda in Arusha or a journalist

bent on publishing a scoop at their expense. The process of locating informants for the

purpose of this research process and the different types of problems that I had to confront,

are in itself an argument in favor of one of the main theoretical stances this research

project is built on, namely the de-essentializing of the refugee-experience.

1.4 Structure of the Dissertation

I would like to conclude the first chapter by a brief outline of the remaining

chapters of the dissertation. Having presented the theoretical and methodological

concerns of this research project in the first chapter, the second chapter aims to

contextualize the research topic and research site. The historical overview of the political

events that took place in the three refugee producing countries under study, Congo,

Burundi and Rwanda and that are at the root of the current refugee out-fluxes, aims at

giving the reader a sense of where the refugees are coming from, and what they have had

to endure before deciding to leave their homes behind. The section on Tanzania situates









the changes in the refugee climate that have taken place in the country since

independence, and that have had direct effects on the daily lives of its refugee

populations, in a more global framework. The second part of chapter 2 describes the

research site, Dar es Salaam, giving particular attention to the political and humanitarian

infrastructure of potential relevance to the urban refugee population. In addition, a

survey on the perceptions of Dar es Salaam's residents on the refugee population paints

an interesting picture of the stereotypes and public images circulating among the

refugees' hosts, as well as their attitudes towards the former.

Having set the stage in chapters one and two, chapter 3 relates to the actual

Congolese, Burundese and Rwandese refugee populations in Dar es Salaam. The

findings of the survey administered to three hundred refugees give an overview of their

demographic and socio-economic composition not only by nationality, but also by gender

and age category. From all the informants interviewed, I selected twelve on the basis of

gender, nationality and age, whose stories and accounts personalize and illustrate the

findings of the survey throughout the following chapters. This chapter also presents the

refugees' motives for fleeing their homes and for opting to go to Dar es Salaam instead of

one of the refugee camps closer to the home country.

Having found a safe haven, the next major challenge to the refugees is to adapt to

the new life in an unfamiliar environment and to make ends meet in the absence of any

material assistance from humanitarian agencies. Chapter 4 relates to difficulties

experienced with regard to the use of language and the confrontations with Tanzanian

residents' prejudices towards refugees. The lack of any official residence or work permit

not only requires a clandestine existence in order to avoid arrest by the Tanzanian

Immigration authorities, but it also complicates the quest for daily survival. And yet,









refugees generally manage somehow to make ends meet, school their children, and at

times, form grass roots associations for a variety of reasons. In the process, they meet

with Tanzanians, fellow refugees and other Dar es Salaam residents, many of whom

become supportive members of their social networks.

The purpose of chapter 5 is to de-construct the refugees' personal social networks

and to explore the extent to which gender, nationality and age play a role in the ways of

meeting network members and the types of support provided. Through the use of

concepts such as homophily, I attempt to uncover the variations that exist between

women's and men's personal social networks, those of Congolese versus Burundese or

Rwandese, and those of the different age groups. Throughout, I pay particular attention

to the effect of the situation of forced migration on the ways refugees establish

relationships with social network members. The analysis of the social network data is

generally quantitative, the contextualization of its findings being the topic of the

following chapter.

Embedding the refugee experiences forms the main goal of chapter 6. The findings

of the previous chapters, namely the variations across gender, nationality and age groups

in the way of coping with forced migration and the reconstruction of personal social

networks, are linked to the relevant macro frameworks. Going back to the premise that

people react to change and are innovative as a function of their previously acquired

behavioral repertoires and cognitive perceptions, a range of political, economic,

historical, social and cultural arguments illustrate the embeddedness of the individual

refugee experience according to the respective social groups the individual belongs to.






57


The final chapter and conclusion delves into a number of policy implications

related to the findings of the above research project, as well as methodological pre-

occupations and theoretical considerations.














CHAPTER 2
SITUATING THE DISSERTATION PROJECT AND RESEARCH SITE

The previous chapter placed the research topic in the existing bodies of literature of

cultural anthropology, refugee studies and gender studies with the aim of investigating

the ways in which it may contribute to these fields of study. In addition, the choice of

methods used was contextualized by perusing the work of a number of feminist scholars

and locating points of similarities with the study of forced migration.

However, before presenting the actual experiences of the urban refugees in Dar es

Salaam, it will prove useful to provide a context to the research project, both in time and

in place. The brief historical overview of the four countries involved, the Democratic

Republic of Congo, Rwanda and Burundi as the refugee producing countries and

Tanzania as the refugee receiving country situates the events at the root of the current

refugee situation in the Great Lakes region and Tanzania. The second part of this chapter

provides a short description of the actual research site, the city of Dar es Salaam, and

presents the existing official, humanitarian and social structures in place that are of

relevance to the urban refugees.

2.1 Background to the Research Project

The structure of this section is unavoidably based on the concepts of the nation-

state and its corollary "the national/citizen" as categorizing tools. One should, however,

particularly in the context of this research project, be aware of the relatively short history









of the nation-state1 and of its currently diminishing independence, whether economically,

socially, culturally or politically, in today's era of globalization.

In this context of closely interacting and localizing processes and forces, it is
obvious that the distinction between domestic and international politics is
meaningless Even to the extent that the nation-state maintains its authority, it
itself is becoming increasingly 'internationalized' or'globalized' (Collinson
1999: 5)

An overview of the current debates on both these points being far beyond the scope of the

dissertation project, it is nevertheless important to mention their relevance with regard to

the current situation in the Great Lakes Region.2 In the following sections, I briefly

present the historical interconnectedness of the three refugee producing countries, and

after a short overview of the colonial and post-colonial period, continue to describe in

brief the actual events that eventually led to "Africa's First Continental War" (Weiss

2000:1). Subsequently, the recent changes in Tanzania's refugee policy are situated in a

more global context, followed by a short history of the refugee flows into the country and

an overview of the current situation.

2.1.1 The Refugee Producing Countries

The dramatic events of 19943 in Rwanda triggered an avalanche of publications

(books, articles, web sites, etc.) on the situation in the region of the Great Lakes, written

by political scientists, historians, journalists, aid workers, politicians and many others, all

trying to make their point known to the international world. Jan Vansina, the most

eminent historian on the Great Lakes region recently cautioned that:



1 Especially in Africa (e.g., Davidson 1992).
2 For a map of the Great Lakes Region of Africa, please refer to map 1-1 in appendix a.

3 The genocide of between five hundred to eight hundred thousand Tutsi and moderate Hutu.









Indeed, most of the accessible sources (especially those generated by the media) are
not reliable, in part because of individual bias or even incompetence, but mainly
because the sources are all part of an intense propaganda war. This campaign is
intended to generate support for one side or the other, whether it be to raise morale,
to facilitate the procurements of arms, or to generate outside diplomatic cover.
Furthermore, atrocities and the violent emotions they generate make it nearly
impossible for any observer to retain a "neutral" stance or to avoid "demonizing"
one side or the other. (1998:37)

The implications for the historiography of the region are clear:

Hence, it is not surprising that all sides in today's Central African crisis reinvent
history. They use portions of the existing historiography written by respected
authors ... to derive novel interpretations of supposedly well established facts.
Their historical arguments range not only over the last thirty years but back into the
colonial period and as far back as the early Iron Age. (1998:38-9)

Vansina's well-formulated considerations concerning the origins of Hutu and Tutsi are

not new to the history of the Great Lakes region, but in the wake of the recent events they

have taken on a whole new dimension. Even before the 1994 events, MacGaffey,

reflecting on (the mythologization of) the history of the ethnic groups of the region,4 had

concluded that "[t]he present, we discover, is mother to the past" (1993:15).

2.1.1.1 The historical interconnectedness of the Great Lakes countries

Leaving the above considerations largely as reflections on the controversies

surrounding the origins, political history and ethnic parameters of the Hutu and Tutsi

populations in Rwanda and Burundi,5 the following overview of the region's history will

concentrate on its historical interconnectedness. The Great Lakes Region consists of

(parts of) five countries: Rwanda, Burundi, Eastern Congo, South Western Uganda and

North Western Tanzania (map 1-1 in appendix a). There are four major lakes in the area:


4 One well-illustrated example of this phenomenon are the mythico-histories of the Burundi Hutu refugees
in the Tanzanian refugee camps (Malkki 1995a).

5 Giving a complete overview of the controversies surrounding the origins of the Hutu and Tutsi
populations in Rwanda and Burundi falls beyond the scope of this dissertation project and the reader is
referred to the publications of experts on the matter: e.g., Lemarchand, Vansina, Reyntjens, Chr6tien, etc.









Lake Victoria, Lake Edward, Lake Kivu and Lake Tanganyika, hence its name. Its

geography is dominated by the Great Rift Valley in which the huge lakes are located and

most parts of the area lie at an altitude of between one and two thousand meters,

combining a pleasant, moderate climate and lush green, fertile hills (Chretien 2000: 11-

12). Linguistically, the myriad of languages in the region can be divided into four or five

families of, often to a lesser or larger extent, mutually intelligible vernaculars (Chretien

2000:34; Mworoha 1987:188). Recent archeological and linguistical research has

established the presence of cattle throughout the region from more than 2,000 years ago6

accompanied by a high level of interaction between its ancient Early Iron Age societies

(Schoenbrun 1998).7 Over time, a variety of political systems saw the light in the region,

but notions of nobility, kingship and courtly life took firm ground by the beginning of the

second millennium only, with "the invention of concepts of hereditary nobility and the

link that was created between nobility and the institutions of instrumental power over

people" (185).8

By the beginning of the 19th century, many of the kingdoms in current-day

Rwanda, Burundi, Uganda and Eastern Congo had developed into states with political

systems based on centralized power. These states were not isolated entities, but


6 Contrary to the assertions of earlier historians that cattle--considered "the" hallmark of the Tutsi
pastoralists--was introduced into the region by the "Hamitic invaders" of the 15th century (see more on this
issue later in this chapter).

7 Linguistic research had shown earlier that the region was populated during the Bantu migrations from
West Africa between 5,000 and 3,000 years ago. More information on Greenberg and Guthrie's linguistic
theories and on how archeological finds eventually helped shape the Bantu migration theories may be
found in e.g., Gailey 1970; Chr6tien 2000; Schoenbrun 1998, Collins 1968.

8 Some oral history sources on the royal dynasties of the Great Lakes region contend to be able to go back
as far as the 15th century. Scientific research however, was able to confirm the reliability of the data
provided up until the beginning of the 18th century only (Chr6tien 2000: 117-172). The names of the kings
of no less than fifteen royal dynasties in the Great Lakes region between 1500 and 2000 are listed in the
appendices of Chr6tien's L 'Afrique des Grands Lacs: Deux mille ans d'histoire (2000).









constituted a communicating network of diplomatic relationships, alliances, dynastic

marriages, commercial exchange and military activities throughout the region,

incorporating not only present-day Rwanda, Burundi and Uganda, but also what is now

Eastern Congo and Western Tanzania (Ajayi 1998:100). In addition, regional migrations

took place as early as the 18th and 19th century, usually from the more populated

kingdoms of Rwanda and Burundi into sparsely populated Western Tanzania (Roberts

1968: xi, 120; Daley 1993:19-20; Egero 1979:30). At the same time, Arab and Swahili

traders from Zanzibar and Bagamoyo on the Tanzanian coast introduced firearms into the

region in exchange for slaves and ivory, thereby ultimately opening the way for European

explorers, as well as Muslim and Christian missionaries. The stage was set for the

military incursions and political and economic impositions of the colonial powers that

were to have a devastating effect on the societal landscape of the Great Lakes region.

2.1.1.2 The colonial impact in Congo, Rwanda and Burundi

Congo was the private property of the Belgian king Leopold II from 18859 until

1919, when it was taken over by the Belgian government under mounting international

outrage over the king's exploitation of the Congo Free State.10 Rwanda and Burundi, on

the other hand, were first colonized by Germany together with Tanganyika, as the

mainland of Tanzania was called. After World War I, Germany was punished and its

colonies redistributed: Tanganyika to Great Britain and Rwanda and Burundi to Belgium.

Whereas in the Congo in general, the Belgians applied a system of direct rule by

removing the legitimate chiefs and appointing the "chiefs of the whites" (Lemarchand


9 This is the year the Berlin Conference took place, where the African continent was divided between the
European powers.

10 See Adam Hochschild's King Leopold's Ghost (1998) for a fascinating account of this time period.









1970: 66), in Ruanda-Urundi (as it was then called) as well as in the Kivu region in

Eastern Congo, they applied the principle of indirect rule as had the Germans before

them. The decision "to follow the traditional lines of authority rather than to create anew

an administrative structure" as they had in the rest of the Congo, was based on the

existence of "micro-kingdoms which tended to hold up rather well against the

administrative, colonial onslaught" in both the Kivu region and Ruanda-Urundi (Prunier

2000:146).

In addition to political impositions, the colonizers introduced a system of ethnic

categorization in all three countries. Making rigid categories out of ethnic groups that

before had somewhat fluid boundaries11 had much less far-reaching effects in the

Congolese ethnic mosaic than it did in the two small kingdoms of Rwanda and Burundi.

These were considered by the colonizers to be inhabited by the same three ethnic groups:

Hutu, Tutsi and Twa. Two of these, the Hutu and Tutsi, were categorized, not by

language, religion or culture (which they shared), but by perceived differences in status

and physical characteristics. The Tutsi were seen as the ruling minority, the Hutu

majority as the agricultural vassals of the rules, and the few Twa as the remaining hunter-

gatherers from the rain forest. Each group was categorized on the basis of certain

physical traits considered markedly different from the others', and scientists of the time

went as far as to put together tables with height, weight, nasal and facial index per group

(Codere 1993: 96). These ethnic labels eventually became the basis of the German and

the Belgian colonial administration and "the listing of ethnic labels on identity cards


11 In his Hutu and Tutsi: Primordial Illusions (1994: 6-16) Lemarchand argues that "status, not ethnic
identity, was the principal determinant of rank and privilege" (10), and that "depending on luck and
individual ability, some Hutu elements could end up with considerable more influence, wealth and social
recognition than the average Tutsi" (11).










issued to all residents served to eliminate the previous flexibility in identity" (Longman

2001:169).12

Another colonial inheritance was the "Hamitic Hypothesis" that explained the

"separate origins of the Tutsi," who were seen as "a civilizing Caucasian influence in

Negro Africa" (Mamdani 2001:47). According to the Hamitic Hypothesis the Tutsi

migrated from the Ethiopian Highlands around the 15th century and "conquered" the

sedentarist Hutu farmers who had inhabited the region since the Early Iron Age Bantu

migrations. The Tutsi further introduced pastoralism to the "uncivilized peasantry" and a

royal state court (Ibid.).13 This myth was the basis for, and at the same time, explanation

of the colonial perception of the Tutsi as "Nilotes" being superior to other African groups

defined as "Bantu."14 The Church was instrumental in co-opting the Tutsi population into

the colonial system of indirect rule by giving Tutsi children preference over Hutu

children in the mission schools, the only avenue for Africans at the time to acquire a

Western education and enter into the colonial administration (Gahama 2001:243-260;

Rumiya 1992:159-168). Up until shortly before independence, the Church "posed as the



12 Umutesi, a Rwandese woman, recalls that in the few years preceding the 1994 genocide some children
"who looked like Tutsi" were chased from school by fellow students and "were not to come back unless
they had the identity papers (livret d'identite) delivered by the Germans to their fathers or grandfathers.
That paper was the only document, according to the students, that could give authentic information as to
the ethnic identity of the Rwandese." (2000: 12). In the same line, other observers remarked that the very
fact that the Interahamwe relied on individuals' identity cards in order to decide whom to kill, proves that
Hutu and Tutsi are physically indistinguishable.

3 For a comprehensive overview of the other theories circulating at one time or another regarding the
origins of Hutu and Tutsi, see Mamdani 2001 (41-75).

14 Whether Hum in Rwanda and Burundi or other ethnic groups in Eastern Congo. It is interesting to note
that contemporary scholarship has (re)classified the Tutsi as a Bantu group (Nzongola-Ntalaja 2001). In
fact, "[t]he terms 'Bantu' and 'Nilotic' refer exclusively to a i,,,, ,,,. distinction and it is only within a
linguistic context that they are scientifically warranted. They cannot possibly be justified if extrapolation is
made from this very context to other than linguistic features such as, for instance, putative historical,
physiological or social attributes" (Goyvaerts 2000:302).









strongest advocate of Tutsi supremacy, largely on grounds of political expediency"

(Lemarchand 1970:73).

In Congo, the Church equally had the monopoly in education through its mission

schools and seminars providing virtually the only route for a Congolese to become an

evolue.15 These educated Congolese over time began to constitute a separate "class ...

different from the untutored mass" (Anstey 1970:202). When their growing political

aspirations remained largely unmet by the Belgians, they eventually turned to revolt. By

that time, however, there were only a handful university graduates and few Congolese

had been trained to take over the administration in the national government offices. This

level of unpreparedness and lack of intellectual capital at the time of independence was to

set the Congolese stage for a long time to come.

In Rwanda and Burundi, when the educated Tutsi began to rally for independence

in the 1950s, the colonial administration suddenly decided to democratize the countries

and give the political power to the Hutu majority. Electoral processes were introduced at

a time when the Hutu educated elite in Rwanda and Burundi constituted a very tiny

minority, and with little fundamental change in the traditional social structure, the stage

was set for a turbulent era of independence (Lemarchand 1970:81-83). Although the two

countries shared many cultural characteristics, ethnic relations in Burundi, for historical

reasons,16 were more fluid and not reducible to a simple Hutu-Tutsi split, hence their


15 An evolue (literal translation: developed or advanced person) was a term given to those Congolese who
appeared to have given up the traditional African lifestyle and assimilated the lifestyle of the "whites": i.e.,
living in a nuclear family, in a house in town, and exercising a paid job requiring a certain level of Western
education and fluent French. They were never to aspire, however, to be completely accepted as a member
of the white community. Representative figures were the clerk, the male nurse, and the teacher (Anstey
1970:195).

16 See Lemarchand (1970) for a detailed analysis and comparison of the historical backgrounds to the
political situations in the respective countries.









"radically different political trajectories, with Rwanda acceding to independence as a

Hutu-dominated republic and Burundi as a constitutional monarchy" (Lemarchand

1994:1).

2.1.1.3 From independence to the early 1990s

In Congo (map 1-2 appendix a), the democratic election of Patrice Lumumba as

the country's first Congolese prime minister raised quite a few eyebrows because of his

socialist ideologies and stem condemnation of the era of Belgian colonization.17 Amidst

the ethnic turmoil of the early 1960s and the secession attempts by some of the richer

Congolese provinces, Lumumba was assassinated18 and a young army sergeant Joseph

Mobutu was considered more trustworthy by the Western powers to rule the Central

African country. Zaire (as Congo was renamed) enjoyed a relative political stability

under Mobutu's thirty three years of rule. However, the lack of skilled and experienced

Congolese cadres to manage the vast country had devastating effects on its economic and

infrastructural development, especially from the mid 1970s onwards after Mobutu's

nationalization exercise (retour a l'autenticite).

The crisis continued to deepen in the 1970s and 1980s (Nzongola-Ntalaja 1986),

and while the state withdrew further and further to Mobutu's immediate entourage, the






17 This section has been intentionally kept to a minimum and is meant only to give the reader unfamiliar
with the region some background information as to the events that took place in the three countries in the
course of the 1990s, ultimately leading to the current situation in the Great Lakes region.

18 De Witte (1999) gives an elaborate account of the time era and is the first to document the direct
involvement of the Belgian military and senior members of the Belgian government in the murder of
Patrice Lumumba. It appears that even the Secretary General of the United Nations at the time publicly
expressed disdain for and failed to recognize the legitimacy of Congo's first democratically elected Prime
Minister.









Zairian people gained an international reputation as ddbrouillards,19 engaging

increasingly in urban "popular" economic activities, i.e., the informal sector (Willame

1997). Biaya (2001) describes the development of a parallel society where "various

associations and local NGOs blossomed and set objectives to resume control of what the

state had abandoned or was unable to control, such as local universities, regional

development councils, and NGOs" (51).

In Eastern Zaire, in particular, there was a proliferation of independent churches as

part of citizens' response to the withdrawal of the state and its infrastructures. These

churches were increasingly involved in development programs based on local initiatives,

at times even setting up "parallel networks of schools, health and agricultural services,

and courts" (Newbury in Nzongola-Ntalaja 1986: 105). They complimented the

"enormously important role of the [Catholic] Church" in "a vacuum from the point of

view of the basic social and administrative services which a normal government is

supposed to provide" (Prunier 2001: 155).

In Rwanda (map 1-3 appendix a), from 1959 onward, increasing Hutu political

violence aimed to dismantle Tutsi power at the local administration level and ultimately

triggered a transfer of governmental power from a Tutsi to a Hutu elite in 1962

(Mamdani 2001: 103-131). The change of regime went hand in hand with multiple

massacres of thousands of Tutsi, producing refugee outfluxes of tens of thousands of

Tutsi to Burundi, Uganda, Tanzania, and Zaire. Rising unemployment among increasing

numbers of educated Hutu and the 1972 events in Burundi (see later) created ethnic



19 A popularjoke of the time was that Article 15 of the Zairean constitution advised Congolese citizens to
Debrouillez-vous. Se ddbrouiller means "to fend for oneself." A ddbrouillard is a resourceful person
coping by him/herself.









tensions that led to the 1973 military coup of Habyarimana (a Hutu) inducing more

Rwandese Tutsi to seek refuge abroad. Habyarimana was re-elected in 1983 as the

president and stayed in power until his assassination in 1994.

In Burundi (map 1-3 appendix a), before independence "caste cleavages were more

numerous and fluid, and less consistent with ethnic differences" than was the case in

Rwanda (Lemarchand 1970: 89). The effect of the 1959-1962 events in Rwanda,

however, made the Tutsi monarchy cling to power yet more strongly--despite several

Hutu victories in democratic elections--and eventually developed into a military

dictatorship. In 1972, a Hutu uprising led to Tutsi massacres in Bujumbura and the

Southern provinces. The response of the Tutsi regime was swift: over a hundred thousand

Hutu were massacred, many of them belonging to the educated elite. The result was an

outflow of hundreds of thousands of Hutu refugees to Rwanda, Tanzania, and Zaire.

After a series of military coups following the 1972 genocide, Buyoya (a Tutsi) seized

power in 1987 in an attempt to restore order. When another Hutu uprising involved more

massacres of Tutsi in the North of the country in 1988, the devastating reply from the

Tutsi dominated army resulted in twenty thousands persons killed and ever increasing

Hutu refugee flows to the neighboring countries (Braeckman 1996:127-201).

A number of researchers have pointed out that the competition for strategic

minerals and land is the main cause of the war in Burundi rather than ethnic hatred

(Balati 2001; Ngaruko and Nkurunziza 2000). Nyerere, the facilitator of the Arusha

Peace Negotiations once said: "The strife in Burundi is cast in ethnic terms. It is not, it is

really about resources" (Daily News June 9, 1997).20


20 More on this specific topic follows in chapter 6.









2.1.1.4 From the early 1990s onward

Lemarchand (1997) rightly proposes that "much of the Hutu-Tutsi confrontations in

Rwanda and Burundi is indeed reducible to the polarization of group identities that has

accompanied the movement of refugee populations from one state to the other"21

(Internet). Mamdani (2001), however, ventures out further than the Rwanda-Burundi or

Hutu-Tutsi dichotomy, locating the very beginning of the Great Lakes region events of

the 1990s in Uganda.

It is estimated that by 1990, the number of Tutsi refugees in Uganda that had fled

Rwanda between 1959 and 1964 numbered two hundred thousand, most of whom were

living an uncomfortable and insecure life in the refugee camps. By the early 1980s,

several thousand of the second generation refugees had joined the National Resistance

Army's (NRA) guerrilla war against Uganda's president Obote, and assisted Museveni in

eventually taking over power in 1986. In appreciation, some of these Rwandese guerrilla

fighters (the most notorious of whom is undoubtedly Kagame22) acquired important posts

in the Ugandan government. In 1990, a proposal from president Museveni to allow non-

Ugandans with a ten year residency in Uganda to apply for Ugandan citizenship, was

voted down in parliament in favor of a citizenship law that required proof of

"indigeneity," i.e., having proof of forefathers residing in Uganda during colonial times.

For the Rwandese Tutsi refugees, "the die had been cast" (182). Having no future

prospects in Uganda, they decided to return home and demand their place in Rwanda.





21 I.e., Rwandese Tutsi refugees to Burundi and Burundese Hutu refugees to Rwanda.

22 Rwanda's current President.









The Rwandese Patriotic Front (RPF) invaded Rwanda in 1990, reportedly with Ugandan

assistance (164-183).

In Burundi, in the meantime, Buyoya was pressurized by the international

community into organizing democratic elections. While higher education systems and

the administrative government offices had slowly been opened to Hutu participation,23

Buyoya remained firm in his refusal to accept Hutu officers in the national army. In June

1993, Melchior Ndadaye was elected as the first Hutu president of Burundi, promising

"in French democracy, in Kirundi, the liberation of a people suppressed for too long"

(Braeckman 1996:152 own translation). The majority of the provincial governors and

local administrative officers were promptly replaced, fueling the feeling of increasing

resentment and feelings of insecurity in Tutsi circles. In the wake of increasing pressure

on the government to integrate former Hutu rebels into the Tutsi dominated national

army, the democratically elected president Ndadaye was assassinated by this very army

in October 1993. Civil war ensued involving the massacre of thousands of Tutsi

civilians. Subsequent repression on the Hutu population by the army led to tens of

thousands being killed, and several hundreds of thousands displaced or in exile.

From 1993 onwards, Burundi has been prey to what observers have called a

"genocide a compte-goutte" (literal translation: genocide drop-by-drop) because of the

systematic usage of terror by Tutsi government forces to intimidate the Hutu population

(Reyntjens 1999:37). By 1997, 5% of the population, or three hundred thousand persons,

were interned in government camps, deemed "necessary" by the Burundese government



23 Upon until that time Burundese Hutu were systematically discriminated against in the educational
system making it sheer impossible for them to participate in national, and to a lesser extent local,
government structures.









"in order to isolate the rebels" (Buyoya in Reyntjens 1997:190). By 2000, close to seven

hundred thousand Burundese, more than 10% of the population, were internally displaced

(map 1-5 in appendix a). While Hutu rebels continued to launch attacks throughout the

country, the Tutsi government forces remained directly and indirectly involved in the

killing of Hutu civilians, all parties truly "caught in the spiral of violence" (USRC 1995).

In Rwanda, in the meantime, the RPF was slowly making its way towards the

capital Kigali, while the international community was putting pressure on the Hutu

government to share power with the Tutsi population. At the same time, Hutu refugee

flows from the RPF controlled areas in the North were pouring into Kigali, and rumors

abounded about atrocities allegedly committed on the Hutu population by the advancing

RPF forces (Umutesi 2000:19-51). The events of October 1993 in Burundi and the

pouring into Rwanda of several hundreds of thousand Burundese Hutu refugees

contributed further to a general climate of fear (Mamdani 2001:191). When on April 6th,

1994, the plane carrying the Rwandese president Habyarimana was shot down by

unknown assailants,24 the Rwandese 1994 genocide started. More than half a million

Tutsi and moderate Hutu were killed by the Interahamwe militia and the Rwandese

Armed Forces (FAR), albeit assisted by parts of the Rwandese civilian Hutu population

later to be known as the genocidaires.252627 By the time the RPF reached Kigali in July


24 It has not been established beyond doubt until now, who was responsible for this action.

25 Many have tried to answer the question: "How was this possible ?" Fear was most likely the primordial
instinct played upon by the extremist Hutu (with the help of Radio Mille Collines) to convince the illiterate
peasant villager to kill his lifelong Tutsi neighbors. "They hear over and over that the Tutsis are out to kill
them, and that is reality. So they act not out of hate as fear. They think they have only the choice to kill or
be killed" (Mamdani 2001:191). Mukagasana's chilling account of the events is consistent with this
analysis in her book La mort ne veutpas de moi (1997) written as a Tutsi woman survivor of the genocide.
See Des Forges' (1995) and Newbury C. (1998) for an analysis of the role of the state in shaping ethnic
consciousness, and Jefremovas (1995) and Longman (2001) for a contextualization of the genocide beyond
the ethnic framework.









1994, more than one million Rwandese were internally displaced and close to three

million crossed the borders seeking refuge in Tanzania and Zaire. Soon, it became

apparent that the Interahamwe and other genocidaires were hiding among these refugee

populations. This complication eventually discredited the UNHCR and other refugee

agencies for not being able to distinguish between civilian refugees and the genocidaires,

thereby jeopardizing the security of the civilian refugee population (UNHCR 1997a).

In 1996, the Tanzanian government, worried about the large numbers of armed

former military and militia members hidden among the civilian refugee population,

decided to forcibly repatriate the one million Rwandese refugees in Tanzania. The

Zairian president Mobutu, on the other hand, who had been a close friend of the

assassinated Rwandese president Habyarimana, was happy to provide shelter for the

Rwandese militia much to the exasperation of the new Tutsi government in Kigali (see

later). After the return of the majority of the 1994 refugees in 1996, the war in Rwanda

was considered over. However, notwithstanding the international goodwill initially

accorded to the new government in view of the genocide, increasingly observers saw a

dictatorship emerging that relied on arbitrary arrests, disappearances, killings of unarmed

civilians and unlawful detentions28 (AI 1997a, 1998, 2001, 2002; Reyntjens 1999:25-37;

HRW 2000b).



26 Both Mamdani and Lemarchand (1997) relay how the Burundi refugees played an active role in the
Rwandese ethnic killings and some of the most gruesome tortures "as if they were settling old scores, even
across the border" (Mamdani 2001:205-6).

27 Jefremovas (1995) estimates that maximum 1 to 2% of the Rwandese Hum population, 75,000 to
150,000 individuals, joined in the killings. As such, the 4.5 million displaced people, 75% of the total
population had unjustly been accused of complicity, even as "many of these millions ..have to answer for
... cowardice" (28).
28 It should be mentioned that increasing tensions between the Tutsi genocide survivors and the Tutsi old
diaspora, occasionally led to open confrontations. The departure into exile of an increasing number of Tutsi









By 1996, the refugee camps in Zaire had been turned into military bases from

which regular cross-border incursions were launched in order to destabilize the new

Rwandese Tutsi government, a situation condoned, albeit supported, by Zairian

government officials because of the alliance that existed between Habyarimana and

Mobutu. At the same time, attacks against ethnic Tutsi in the Kivu provinces2 in Eastern

Zaire, instigated by the Interahamwe militia and the former Rwandese military, also

called the ex-FAR (Forces armies rwandaises), multiplied. They destabilized the region

to such an extent that in October 1996 the provincial authorities demanded that the

Banyamulenge leave the country altogether. This attempt to expel the Banyamulenge

from their homes in Zaire, was "a gift from the heavens" for the Rwandese government,

now in a position to defend its cross-border advances as preventive actions against

another genocidal attack on a Tutsi community (Weiss 2000:3).30 The Rwandese army

invaded the Hutu refugee camps and attacked the Zairian army, causing the flight

westwards of the ex-FAR, the Interahamwe and thousands of Hutu civilians, while the









(e.g., military men, journalists, students and administrators from voluntary organizations) testified to this
mounting tension (Reyntjens 2000: 5-10).

29 Since precolonial times, there had been many migrations into different parts of the Kivu provinces in
Eastern Zaire by Rwandese Tutsi and Hut, commonly called the Banyarwanda. One of these groups in
South Kivu (consisting of mainly Tutsi generally said to have arrived in the region in the 1880s and thus
legally Zairean nationals) for electoral reasons decided to assume a territorial identity (Banya-mulenge
Sons of the mountain Mulenge) rather than an ethnic one (Banya-rwanda Sons of Rwanda). This move was
interpreted by the local Zairean politicians as suspicious and intended to cover up their "real" Rwandan
nationality (Mamdani 2001:234-253).

30 Rwanda had pleaded with the international community time and again, yet to no avail, that the Hutu
military and Interahamwe be separated from the civilian refugees.









vast majority of Rwandese refugees had little choice but to walk back into Rwanda (see

map 1-4 in appendix a).31

In order to portray their actions as something other than an attack against a

sovereign state, Rwanda and Uganda32 were eager to find Zairian allies against Mobutu.

They found one in the person of Kabila. With the help from Rwandese and Ugandan

troops, Kabila reached Kinshasa by May 1997 meeting very little resistance from the

Zairian government soldiers who preferred to flee or join the ranks of the rebels rather

than fighting for a dictator hated by the Zairian people (Manahl 2000). Unfortunately,

Kabila was not able to rally the major political forces of the country around him, nor to

contain the deteriorating security situation in Eastern Zaire (now renamed the Democratic

Republic of Congo). On the Kivu-Rwanda border, the Interahamwe and former Hutu

military continued their operation against the Rwandese government, while further North,

various Ugandan insurrectionist forces continued to use Congo as a base from which to

attack Uganda (Weiss 2000:8). In August 1998, Rwanda, Uganda and several Congolese

army units took control of large segments of Eastern Congo. Before long, Zimbabwe,

Angola and Namibia joined the Congolese side, while Burundi army units were observed

fighting on the Rwanda-Uganda side of the conflict.33 Apart from the Congolese rebel

groups cooperating with or backed by the Ugandan and Rwandese armies, another


31 UNHCR estimated that in the process of breaking up the camps and the flight west ward close to two
hundred thousands persons "remained unaccounted for, with many dying of natural and violent causes"
(UNHCR 1999c: Internet)

32 Ugandajoined the Rwandese invasion because for some time, Zairean territory along the Ugandan border
has been used as a rear base by anti-Museveni forces (Weiss 2000:5).

33 Burundi's Tutsi forces joined the war because Burundi Hutu rebels joined hands with the Rwandese Hutu
in exile and used the DRC as an rear base to launch attacks against the Burundi government. An elaborate
overview of the motivations of any of the other countries to join Africa's First Continental War can be
found in Braeckman's L 'enjeu Congolais: L'Afrique central apres Mobutu (1999).









Congolese rebel group opposing the invaders had been gaining strength in the meantime.

The Mayimayi34 blamed the Banyamulenge for the situation in Eastern Congo, and

targeted all Congolese seen as linked in one way or another with the Rwandese and

Ugandan occupying forces.

In view of the fact that Eastern Congo is extremely rich in mineral deposits, the

assertion posited by certain researchers that "the continuation of seemingly "senseless"

civil wars is sometimes linked to the "rational pursuit of economic goals" by the warring

factions" (Mwanasali 2000: 145) is most applicable to the Congolese current situation. A

2002 United Nations report on the pervasive plundering of Congo's natural resources has

identified three elite networks consisting of "political and military elites and business

persons" (6) many of whom it refers to by name. One of these networks, for example, is

estimated to have "transferred ownership of at least US$ 5 billion of assets from the State

mining sector to private companies under its control in the past three years" (7). In

another example, a Canadian national was recently arrested in Brussels, Belgium,

carrying US$ 500,000 worth of gold bullion bars in his backpack. The arrested man was

reported to have been making such trips bi-weekly, carrying each time the same amount

of gold from Congo to Belgium, the UK, the US and Switzerland for the past four years

(IRIN 2002b: Internet).

The human costs suffered by Eastern Congo's population in Africa's First

Continental War are enormous. A recent study (IRC 2001) estimated that between


34 The Mayimayi have a long history in the province of Kivu in Eastern Congo going back to the 1960s.
Initially siding with Kabila against Mobutu, they switched sides in view of Kabila's perceived ties to
Rwanda. Holding out against the other rebel movements in Eastern Congo--who are either Rwanda or
Uganda backed--the Mayimayi are generally considered the only Congolese militant group that has not sold
out to foreign powers. Their popularity among the population however, suffers greatly from their ruthless,
often cruel and sadistic, modus operandi (Custers 2001).









August 1998 and April 2001, more than 12% of the region's total population of twenty

million died of causes related to the war. With regard to the whole region, USRC

estimated that by the end of 2001, close to 4 million persons from Congo, Rwanda and

Burundi were uprooted. Two thirds of these were internally displaced and, of the other

third, close to 1 million found refuge in Tanzania.35

2.1.2 The Refugee Receiving Country

Contrary to the precolonial centralized states in the Great Lakes region, the territory

of present-day mainland Tanzania knew "a bewildering series of... [migrations] of

Bantu, Nilotic, Nilo-Hamites, and Hamites which resulted in the present highly

fragmented populations and cultures of this area" (Gailey 1981:210). From the late 18th

century onward, the influence of the Omani Arabs--during the previous millennium

limited to the East African coastal towns of Mombassa, Zanzibar and Kilwa--began to

extend to the hinterland through an expansion of the Arab and Swahili36 long-distance

trade. The impetus for the extension of the long-distance trade37 came from increased

contact and trade agreements between the Sultan of Zanzibar and American and

European states. The major social, economic and cultural impact of the long-distance

trade on the societies in the hinterland was the spread of Islam and the Swahili language.


35 Roughly 500,000 refugees receive UNHCR assistance, another 470,000 (mainly Burundi) refugees have
selfsettled in Tanzania's rural Western provinces (USRC 2002:56).

36 The people occupying the narrow coastal strip of East Africa stretching from Mozambique in the South
to Somalia in the North, and its adjacent islands are referred to as the Swahili. For centuries, this region
was a zone of intercontinental trade between the people of the Arab world, Persia, India, China, Eastern and
Central Africa. "In the process, small communities of foreigners settled in the region, intermingled with
and were absorbed into the local population. Consequently, under the influence of foreign settlers the
indigenous population converted to Islam and Africanized the religion into local temper and taste" (Mazrui
and Shariff 1994: 3). Their language, Swahili, is a Bantu language with a Bantu grammatical structure,
although it has incorporated many words from Arab over the past centuries and still continues to do so.

3 Traded goods were slaves, ivory, gum copal and cloves in exchange for sugar, beads, brassware, gums,
gunpowder and cotton cloth (Ajayi 1998:86).









The travels of the abolitionist David Livingstone were a major factor in drawing

European attention to the "inhumane character of the Arab slave trade" in the interior of

East Africa, and the latter half of the 19th century saw a boom in the establishment of

missionary colonies for the freed slaves paving the way for military conquest (Ajayi

1998:99).

After World War I, the Germans, who had colonized Tanganyika (the name for

mainland Tanzania) from the end of the 19th century onwards, were forced to relinquish

authority to the British, who governed the country through indirect rule until 1961.

Colonial policies "accentuated regional differentiation and rivalry," with the aim of

weakening any territorially organized opposition to the colonial government (Mazrui

1990:232-235). One man, however, was able to coordinate and unite the many resistance

initiatives and in 1954, the Tanganyika African National Union (TANU) was born under

the leadership of Julius Nyerere, who was subsequently to become independent

Tanganyika's first president in 1961.38

2.1.2.1 Nyerere's Tanzania

Tanzania is one of the few African countries that have not known civil strife either

before, during or after independence, unlike many of its neighbors. Observers have

attributed the country's stability to the vision of Julius Nyerere but also to the territory's

considerable ethnic complexity of around one hundred and twenty ethnic groups, only

one of which counts more than one million members.39 From the beginning, Nyerere's

goal was national unity, based on a state ideology (the Ujamaa socialism) and one



38 Zanzibar and Tanganyika united in 1964 to form present-day Tanzania.

39 The total population of current-day Tanzania is 30 million.









national African language, Swahili (Forster et al. 2000: 102-105). The Ujamaa

philosophy was based on the idea that the solidarity of Africa's traditional societies could

be brought up to the national level, creating a new national culture.40 With numerous

ethnic groups, each one having its own histories and myths, traditions and heroes, "it was

the political system itself which came to constitute the main base of the national culture"

(Lange 1999: 42). All these factors have contributed to the remarkably low role the

ethnic factor has played in Tanzanian politics over the past four decades, contrary to the

situation in many of the other countries in the region.

One observer points out that:

For Julius Nyerere of Tanzania, socialism was first of all an attitude of mind.
The key opposition, for Nyerere, were not between rival economic systems or
modes of production, but between conflicting moral orientations: selfishness versus
sharing; exploitation versus solidarity; individual acquisitiveness versus communal
mutuality. Socialism, for Nyerere, was the rejection of selfishness; a capitalist, in
contrast was defined as "the man who uses wealth for the purpose of dominating
any of his fellows" (Nyerere 1968). (Ferguson 1993: 83)

Nyerere applied his philosophy not only at the national, but at the international level as

well by extending his country's hospitality to those seeking refuge from armed struggles

against colonialism, racial domination and apartheid. In 1983, he received the Nansen

Award from the Nansen Medal Committee41 chaired by the UNHCR "in recognition for

his outstanding services to refugees" (UNHCR 1999a:10).





40 The overall rate of success of the Ujamaa exercise was put into question by several observers, notably by
Goran Hyden in his Beyond Ujamaa in Tanzania: Underdevelopment and an Uncaptured Peasantry
(1980).
41 The medal is named after the Norwegian explorer Dr. Nansen, pioneer of international humanitarian aid,
who became the first League of Nations High Commissioner for Refugees in 1921. One of his greatest
achievements was the introduction of a travel document for displaced people commonly referred to as the
"Nansen passport" (UNHCR 1999a:11).









2.1.2.2 Africa's "open door" policy in Tanzania

The period from the 1960s onwards until the late 1980s is sometimes described as

the "golden age" of asylum in Africa with many states adopting an "open door policy"

(Rutinwa 1999:5). While in Southern Africa, most independent nations received

thousands upon thousands of refugees fleeing racism, colonialism and apartheid (from

countries such as South Africa, Angola, Mozambique and Namibia), in Central Africa,

Rwandese and Burundese refugees escaping internal political conflicts readily found

refuge in Tanzania, Uganda and Zaire. Later, a number of the refugee receiving countries

in Southern Africa (e.g., Botswana, Lesotho and Swaziland) succumbed to intense

pressure from South Africa and evacuated the South African refugees from their territory.

Other countries, such as Tanzania, Zimbabwe and Zambia, remained committed to the

norm of non-refoulement by accepting the evacuated South Africans (6). During this

area, African host countries, in collaboration with UNHCR and the international

community, ensured not only the refugees' right to protection but attended to their basic

needs as well. Even if refugees were mostly protected in camps, "these camps were

sufficiently large to enable refugees to acquire land and engage in economic activities

which earned them a dignified living" (7).

Tanzania, of all African countries, was one of the most generous countries to

refugees, allowing them full access to social welfare, the education system, as well as the

job market, both in the private and public sectors. They were offered land and even

Tanzanian citizenship (Gasarasi 1990). Many former refugees went on to rise to very

high positions in the civil as well as the diplomatic services. The underlying philosophy

for these policies was Nyerere's strong conviction that Tanzania "could only sustain

genuine freedom if the neighboring countries are peaceful" (UNHCR 1999a:9).









Nyerere's philosophy is best exemplified in his attempt in 1983 to ban the Swahili word

for refugees, wakimbizi and replace it by the term wageni wakazi. The word wakimbizi

comes from the verb kukimbia which means "to run away from" implying negative

connotations such as cowardice and rootlessness. The term wageni wakazi may be

translated as "guest-workers," emphasizing the positive contributions refugees can make

to the national economy of the country of asylum and their sense of self-sufficiency and

economic independence.42

2.1.2.3 The change in climate

The economic crisis of the 1980s and the IMF's and Worldbank's austerity

measures that followed in its wake,43 together with the massive refugee influxes after the

1994 Rwandese genocide, are only some of the reasons behind the drastic changes in

Tanzania's refugee policy from the late 1980s onward. In the aftermath of the

"liberation" of the former Soviet Union and other East European countries, European

countries fearing an immense influx from the East, changed their asylum and migration

policies to become even more restrictive (Escalona and Black 1995) much to the agony of

human rights advocates (Dacyl 1996). This "downward standard-setting led by the

migrant- and refugee-receiving countries of the North" all but accelerated "the erosion of

an international commitment to the refugee regime" (Collinson 1999: 21). In 1995, after

receiving international criticism for closing the borders with Rwanda and Burundi to

prevent further refugee influxes, the then-Tanzanian Minister for Foreign Affairs said that


42 Despite all good intentions, the new Swahili term for refugees did not hold longer than two years because
"it was felt at the annual conference of refugees, that the word was unacceptable because the fact remained
that persons who had fled from their respective countries due to war were [still] refugees" (Kasomangala in
UNHCR 1999a:9).

43 In fact, it was Nyerere's initial refusal to accept IMF conditions for further loans that lay at the basis of
his stepping down as president in 1985, after 24 years of ruling the country.









"it was a double standard to expect weaker countries to live up to their humanitarian

obligations when major powers did not do so" (Rwegasira in Rutinwa 1999: 17).

Following the fall of the Berlin Wall and the end of the Cold War in 1989, African

countries were strongly encouraged to democratize, institute multipartism and organize

general elections. At the same time, the profile of the 1960s and 1970s refugee as a

"freedom fighter" (e.g., from South Africa, Angola, Mozambique) had drastically

changed in the 1980s and 1990s with virtually all refugees coming from independent

countries engaged in internal (often ethnic) conflicts, and thus not conducive to the

formerly existing feelings of sympathy. These negative dispositions were fueled further

by the negative economic and environmental effects on already impoverished host

populations44 of ever larger groups of refugees, ultimately giving rise to xenophobic

feelings. This happened

at a time when most of Africa is democratizing and governments are compelled to
take into account public opinion in formulating various policies. The result has
been the adoption of anti-refugee platforms by political parties which result in anti-
refugee policies and actions by governments. (Rutinwa 1999:1)

Another important reason for the change in refugee perceptions and policies in

Tanzania lies in a heightened sense of insecurity both at the local and the national level.

The media are rife with stories blaming refugees of committing violent crimes and

overcrowding prisons, to retarding national development:

Growing waves of crime and violence perpetrated by refugees in Ngara district
have retarded economic development because Tanzanians are no longer certain of
their security. .. (Ngara District Executive Director Esther Mbigili in The African,
Jan 22, 2000)

The situation reflects how pathetic the security in the region is ... the influx of
refugees in Kigoma fuels crime in the region, leading to the overcrowding of

44 Tanzania is still one of the poorest countries worldwide: 156th out of 174 (UNDP 2000).









prisons ... Incidents of highway robbery using fire arms and home breaking which
were not very common before, now are noticeably increased ... (Kigoma Regional
Police Commander in The Daily News, Feb 26, 2000)

An unfortunately common side effect of refugee influxes (equally observable in other

African countries) is that among the civilian refugees pouring into the country of asylum,

are hiding former military, rebels, or people otherwise engaged in the armed conflicts in

the home countries, who possess weapons. While some will use their weapons to commit

crimes in the country of asylum, others will sell theirs in an attempt to complement the

meager rations provided by international aid agencies. Prices are usually deflated

because of the supply-demand effect. Landau (2001) reports that "AK-47s and other sub-

machine guns are readily available in the refugee-populated areas [in Tanzania] for

between $10 to $20" (23-footnote). Whereas, undoubtedly, there is an increase in crimes

in heavily refugee populated areas, inducing a feeling of heightened insecurity at the local

level, one should not underestimate the extent to which the situation is blown out of

proportion in the media. Refugees are regularly used as scapegoats by local and national

officials for existing infrastructural shortcomings and professional incompetencies.

At the national level, the refugee is seen as an additional threat to Tanzania's state

sovereignty already weakened by the increasingly authoritative power of international

organizations. The IMF and Worldbank dictate to a large extent how the economy

should be run while in the refugee-populated areas, residents have reportedly become

more inclined to turn to international actors than to the state (Landau 2001). Conceding

that refugees are part of more complex migrant diaspora networks45 that crosscut



45 A topic that is currently underresearched, as Van Hoywegen rightly points out. "With respect to Africa,
there is ... an exaggerated focus on crisis related displacement and less so on labor migration. ... It is ...
important to see how refugee and migrant labor networks interact" (2001: 10). The extent of these
interactions became increasingly clear to me while interviewing informants, and discovering that many









international borders often operating on the margins of the law (MacGaffey and

Bazenguissa-Ganga 2000) and bearing in mind the role these trade networks play in the

warlord politics of many of Africa's current conflicts (Reno 1998), one easily understands

why refugeeses have become, according to the authorities, a security concern,

threatening the peace and stability of the Tanzanian nation" (Van Hoyweghen 2001:1).

From the above, it has become clear how local, national, international and global

elements and events contributed to the current refugee climate in Tanzania.

2.1.3 Rwandese, Burundese and Congolese refugees in Tanzania

The history of Rwandese and Burundi refugees in Tanzania, sadly enough, goes

back many years.46 Since independence, Rwandese Tutsi47 and Burundese Hutu have

been coming to Tanzania in multiple, large waves in 1959, 1964, 1973 and 1972, 1988,

1993 respectively. Until 1996, however, there were no major refugee influxes from

Congo (then Zaire), as only a limited number of individuals sought refuge in Tanzania as

political opponents of Mobutu and his regime.48 When war broke out, in 1996 and again

in 1998, large refugee flows poured into Tanzania not only of Congolese, but also of

Rwandese and Burundese Hutu "uprooted" from the refugee camps in Eastern Congo,

where they had fled to from their home countries in 1994 and 1993 respectively. Some


refugees--although genuinely refugees--were part of larger migration networks to other African countries,
but also Europe and Canada.
46 It should be mentioned that refugees from other countries (South Africa, Mozambique, Uganda, Sudan,
Somali to name just a few) have been equally welcome on Tanzania over the past, however their history
falls beyond the scope of this study.

47 In 1994, the year of the genocide and regime change from Hutu to Tutsi, former Tutsi refugees from
countries all over the world returned to Rwanda in large numbers, while at the same time about 1 million
Rwandese Hutu sought refuge in Tanzania and 2 million in Congo (then Zaire).

48 According to Eger6 (1979), of the three thousand Congolese refugees that fled to Tanzania during the
Mulele uprisings of 1964, only several hundred remained by the late 1970s (Mulele was a close ally and
friend of the assassinated Patrice Lumumba).









of these Rwandese camps were actually attacked in 1996 by Rwandese government

troops in their quest to eliminate the Interahamwe militants and other genocidaires.49 Up

until the time of writing, despite several--eventually aborted--repatriation exercises for all

three nationalities, new arrivals keep trickling in due to the unabated conflicts raging in

the Great Lakes region, albeit in another form than the full-scale wars of 1993, 1994,

1996 and 1998.

Rutinwa (1999) mentions the growing trend in the African context of states

refusing entry to asylum seekers or returning refugees to their countries of origin before

the situation is considered safe to return. Tanzania as well, after receiving hundreds of

thousands of refugees from Burundi in 1993 and Rwanda in 1994, closed its borders with

those two countries in 1995, while the Minister for Foreign Affairs openly declared: "We

are saying enough is enough. Let the refugees go home and no more should come" (11).

Not only are refugees refused asylum, thereby contravening international refugee law,

scores of Rwandese, Burundese and Congolese refugees have at times been forcibly

repatriated. In 1996, both Zaire and Tanzania forcibly repatriated 700,000 and 540,000

Rwandese refugees respectively (AI 1997b), and in 2000 Amnesty International reported

that "standards of refugee protection in Tanzania have dropped dramatically since late

1999 as scores of Burundese and Rwandese refugees have been forcibly returned to their

countries" (AI 2000b: Internet). UNHCR as well mentions that during 2001 several




49 The larger part of the 2 million Rwandese Hutu refugees in Congo (then Zaire) were thus forcibly
repatriated (see also earlier in this chapter), yet several hundreds of thousands preferred to flee to Western
Zaire instead. These groups, suspected to be involved in the genocide, were pursued by the Rwandese
government troops (and collaborating Congolese troops) through thousands of miles of dense forests. No
exact numbers exist, but several tens of thousands persons are assumed to have vanquished. Umutesi's Fuir
ou Mourir au Zaire (2000) recounts the personal experiences of the author during this ordeal. See map 1-4
in appendix a.