Forests, Redd+ Environmental Rule and Everyday Practices in Dr Congo

FROM GEOCODED TO ENTANGLED LANDSCAPE: FORESTS, REDD+ ENVIRONMENTAL RULE AND EVERYDAY PRACTICES IN DR CONGO

SUBMITTED IN FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR IN DEVELOPMENT STUDIES AT THE UNIVERSITY OF ANTWERP

CATHERINE WINDEY

SUPERVISORS:

PROF. DR. GERT VAN HECKEN, INSTITUTE OF DEVELOPMENT POLICY (IOB) – UNIVERSITY OF ANTWERP PROF. DR. JOHAN BASTIAENSEN, INSTITUTE OF DEVELOPMENT POLICY (IOB) – UNIVERSITY OF ANTWERP

EXAMINATION COMMITTEE :

DR. ADENIYI ASIYANBI, DEPARTMENT OF GEOGRAPHY – UNIVERSITY OF CALGARY PROF. DR. ESTEVE CORBERA, INSTITUTE OF ENVIRONMENTAL SCIENCE AND TECHNOLOGY – UNIVERSITAT AUTÒNOMA DE BARCELONA PROF. DR. TOM DE HERDT, INSTITUTE OF DEVELOPMENT POLICY (IOB) – UNIVERSITY OF ANTWERP PROF. DR. SARA GEENEN, INSTITUTE OF DEVELOPMENT POLICY (IOB) – UNIVERSITY OF ANTWERP PROF. DR. MARJA SPIERENBURG, AFRICAN STUDY CENTRE LEIDEN – LEIDEN UNIVERSITY

Windey Catherine Doctoral dissertation

ISBN: 9789057286667 Depot number: D/2020/12.293/20

iii

ACKNOWLEDGMENTS

More than six years ago, I decided to leave my life in Paris where I was working as a management consultant, to start a PhD in Antwerp. A few months before, Gert and Johan – who became my supervisors – had kindly accepted to discuss my (very thin) research proposal and trusted me enough to advise me to apply for an academic assistant position at the IOB. I remember my first day at the IOB on October 1st 2014. After a short introduction to the institute given by Vicky, I was sitting alone at my empty desk in front of my computer with no other task, at the time, than reading and thinking about my PhD project. It was scary and exciting at the same time. Spoiler alert: the PhD process never stopped to feel like that but, what a hell of ride! Little did I know then how life-changing the next following years would be, professionally and even more so, personally. A thesis does not say much about the relationships and experiences that have shaped it, so I would like, in what follows, to thank the people who have been part of this journey in some way or another. Apologies to those of you whom I may have omitted by accident but whom I am grateful to nevertheless.

One of my biggest debts is to the communities of Kisangani hinterland, with whom I have worked since 2015. As I mention in this manuscript, I have often felt and still feel so powerless faced with the extreme disparities between me and these people. I will probably never be able to give back all they have given me, except for a strong, continuous commitment to report and support the struggles of these communities. This thesis, the papers I have published and the talks I have given at conferences would simply not have existed if these people had not shared part of their stories with me. While doing fieldwork in DR Congo was challenging at times, I truly enjoyed the exchanges with these people and, every year, I was looking forward to seeing them again. Thank you so much, to all of you, for your welcome and kindness. Merci beaucoup à vous tous, pour votre accueil et pour votre gentillesse. I joyfully remember many faces, places and moments but I would like to say a special thanks to Georges from KP23, Jean- Faustin from KP13, Josué from the Tembo cocoa plantation and Gabi Agalua from Alibuku with whom I have talked so often over the years, and who have helped me to better understand the lives of their communities.

I also want to thank the many people at state institutions, government and REDD+ implementing agencies, civil society organizations and NGOs in Kinshasa and Kisangani, who have accepted to answer my questions. They are also an important part of what is written in this PhD thesis, and I enjoyed talking to them despite our (often) diverging views.

During the research, I worked with several research assistants and transcribers, and I would like to acknowledge their work here. First and foremost, Dieubeni who has accompanied me ‘in the field’ throughout (almost all) the years of my research in Kisangani hinterland. Merci Dieubeni pour ta

sympathie, ton sérieux dans le travail et ton soutien dans ma recherche. On a bien travaillé ensemble mais aussi beaucoup ri. Thanks also to Héritier and Gervais who assisted me during my first round of data collection. Thanks to the students who transcribed and translated my interviews: Cédric, Delphin and Isaac. I am grateful as well to the researchers and professors in Kisangani who helped me more indirectly with my research. Papy Bambu Liena, professor in sociology at the University of Kisangani, with whom we have launched a VLIR-UOS action research project since 2019. Alphonse Maindo and Justin Kyale from Tropenbos-DRC, Jean de Dieu Malongola from OCEAN NGO who know so much about the topics I work on and with whom I had many discussions. Merci à tous.

Thanks to everyone who made me feel welcome where I stayed during my field visits. During my stays in Kisangani, I could count on the logistical support of R&SD, a consulting company who is responsible for the implementation of a EU/CIFOR project. I want to thank Quentin, Gloria, Belly, Raoul, Jugel and the rest of the team for providing me with official invitations, welcoming me at the case de passage, allowing me to rent a motorbike and, generally, making my stay easier. My fieldwork would have been much more complicated without the help of R&SD as I had little or no practical institutional support when I started. Thanks also to my temporary housemates at the case de passage, over the years: Vincent, Mélissa, Nils, Eugène, Dino, Christian, Jean-Yves, et la team de Yangambi: Martin, Mathilde and Jules. We had fun. In Isangi, to the welcoming environment at the mission Les Filles de La Sagesse, in particular Soeur Mado. It was such a peaceful place to stay at. In Yangambi, chez les Prêtres. In both places, I have had many enriching conversations with my hosts and the other guests.

There are also many people who made my stays in Kinshasa la Belle (some say Kinshasa la poubelle but I cannot help looking at the incredible energy of this city) so enjoyable. Tamiley and Deker who drove me safely around the city. They were always so friendly, available and helpful. Thanks also for the great conversations with the priests at the Frères Capucins in Limete. I also want to thank Camille who welcomed me in Kinshasa during my first stay in 2015, and then in 2016, and with whom I exchanged a lot on the REDD+ process as she was working on the same topic in another area of the country. We also had some nice nights listening to live music and going out in this vibrant city. Finally, special thanks to my friend Will. We met at random in 2016 on the tarmac of Kisangani’s decrepit and chaotic airport, and quickly became friends after that. When in Kinshasa, I have spent a lot of time in his apartment overlooking the Congo river, and his British humour made the evenings out memorable.

Such extensive fieldwork would not have been possible without the financial support I received along the years from the IOB Research Fund, the FWO Research Foundation – Flanders, and the Flemish Interuniversity Council VLIR‐UOS.

Back to Belgium now, where it all started. I want to start by thanking my supervisors, Gert and Johan. From the bottom of my heart, thank you so much, both of you for your guidance, your constructive

comments and trust in my work over the last six years. I honestly do not think I could have had better supervisors. You have always encouraged me in my project and given me enormous freedom to pursue my own research interests, and follow my academic path. I have also much enjoyed writing with both of you for several papers. A PhD and academia can feel lonely but we have an amazing team which I am proud to be part of. I am happy to be able to continue working with you for the next year at least, and let’s hope for many others. Gert, of course, I also want to thank you for being much more than a supervisor, and having become such a good friend. You are a sparkling person and excellent researcher and professor. Thank you for sharing the ups and downs of my (PhD) life and so many hilarious moments at conferences and evenings out in Antwerp and abroad.

Thank you also to Paige West for mentoring me during my research stay at the Department of Anthropology at Columbia University. I admire her work and it has greatly influenced what is written in this thesis. I am also grateful to have met and worked with a number of researchers and friends from research institutes around the world: Elizabeth Shapiro-Garza, Jennifer Casolo, Esther Marijnen, Diana Vela Almeida, Aaron Vansintjan, Pamela McElwee and, last but not least, Vijay Kolinjivadi who has now joined our research team at IOB.

Here I also take the opportunity to thank the members of my jury: Adeniyi Asiyanbi, Esteve Corbera, Tom De Herdt, Sara Geenen and Marja Spierenburg who have carefully read my manuscript and provided insightful comments that have helped me to sharpen the message I convey, and (hopefully) to improve the quality of the thesis you have now in your hands.

At IOB, there are many other people I want to thank for being amazing, brilliant colleagues and friends and making this place so homely. In particular, I would like to thank some of my fellow AAPs/PhD students who have become much more than colleagues. Miet, my previous office neighbour, with whom I have spent (and still spend) an incredible amount of time talking about our lives. I truly admire your strength. Mathias, Lisa, Mollie, Hanne and Sarah for your continuous support and friendship. We have shared this PhD experience together, through laughs and tears, and have become good friends. Thanks Mathias for your kindness and rational mind; Lisa for being my AAP buddy since day 1 and for being so caring; Mollie for your sharp sense of humour and our almost daily Messenger chats; Hanne for your lightness and free/party spirit; and Sarah for sharing an office with me and accepting my sometimes chaotic ways with such sweetness. I am happy I met all of you. Nanneke, my previous office mate and friend: I am grateful we have remained in touch over the years and I am impressed by your perseverance to pursue your academic career while being a great mom to Areini. Being surrounded by so many bright people is a blessing.

Special thanks also goes to Frédéric, for giving many very good advices in the final years of my PhD. Vicky, for giving the best pep talks and being one of the pillars of the IOB. An, for dealing with the financial

vii

aspects of doing research with such care and enthusiasm, and for being our secret Sint-Niklaas (we all know). Joëlle and Katleen for your administrative help throughout my six years at IOB. Joëlle again, for introducing me to yoga. Kristof and Fergus, for our work together on the VLIR-UOS project. Sara G, for sharing your passion for DRC and its people, and for giving classes together. Marijke, for the trip in Bukavu in 2017. I also want to mention some of the many (ex-)colleagues with whom I shared some tasks at a point or another, happy hours or small talks: Antea, Cassandra, Danny, Dimitri, Eliane, Els, Eugenia, Filip D.M., Gabi, Greet, Hans, Ivan, Janus, Juan, Katrien, Kelbesa, Klara, Loresel, Marjan, Mark, Micha, Michael, Mitte, Nadia, Nathalie, Neil, Nicole, Nik, Reginas, René, Sahawal, Sara D., Sarah T., Simon, Stef, Tobias and all those who I have (undoubtedly) forgotten.

A sincere thank you also goes to my friends in Belgium and abroad who make my life lighter and funnier. Thank you for listening to my complaints about PhD life and for reminding me that there is much more to life than academia and its pitiless competition.

Almost last but certainly not least, to my dear parents, Arielle and Peter, and my sister, Sarah, whom I love dearly. I am so grateful to have such a close bond with my family. Your constant support and trust in me give me endless reserves of strength. You have created the best environment to grow up in and have continuously motivated me to seize every opportunity. You encouraged me to do the things I really liked, and to follow my dreams even if that meant moving to another country or travelling alone in DR Congo and elsewhere. Your kindness and unconditional love mean so much to me. Sarah, I am so proud of what you have accomplished over the last years, and so happy you brought Maïa, my goddaughter, to this world. She is such a source of joy and you are a great mom. A vous quatre, nous sommes ensemble, pour toujours.

I also would like to thank my sweet grandmother, Gelinda, and late grandfather, Hugo, who have taken such good care of my sister and me every Wednesdays (and more) of our childhood, and whose passion for DR Congo where they lived for six years certainly played a role in my deep interest for this country.

Finally, to the most important man in my life, Nick, my love. I do not believe in God nor am I an overly romantic person but there is a little bit of destiny in our story. Your stay at IOB as an AAP was probably the shortest in IOB history but just enough to get to know each other and start our relationship. The PhD and you were not a match, but thankfully we were. You truly are a beautiful person who makes me want to be a better person through your very being. Thank you for your love, for your patient and unconditional support, for letting me cry on your shoulder when I am down, for being so optimistic towards everything, and for reminding me what really matters in life. There is not one day that goes by when I don’t think of how happy I am to be with you.

viii

CONTENTS

ACKNOWLEDGMENTS v

LIST OF FIGURES AND TABLES xi

LIST OF ABBREVIATIONS xii ABSTRACT xv

INTRODUCTION 1

EPISTEMOLOGICAL AND METHODOLOGICAL APPROACH 13 1. Engaging with multiplicity in times of ‘post-truth’: a performative relational onto-epistemology 13 2. Methodological reflections for studying messy complexities 17 3. Doing embodied research: positionality and reflexivity 20 4. Overview of methods and field sites 29 References 35

PART I : Abstracting Congolese forest: mappings, representational narratives, and the production of the plantation space under REDD+ 39 Introduction 41 Chapter 1. Interrogating maps and mapping: from hegemonic representation to performativity 45 1.1. Contextualizing: modern mapping and (colonial) power 45 1.2. Unfolding the concepts: performativity and relational mapping, tools for (re)thinking about maps in environmental governance 51 Chapter 2. Representing forests and population: a contingent assemblage 53 2.1. Creating consensus on contested data 53 2.2. Threatened carbon stocks: a seemingly neutral and disconnected representation 56 2.3. What (non-)forests? Which people? 65 2.4. A well-oiled discourse 68 2.5. A convenient reality in a political ‘messiness’ 71 2.6. Logoization 74 Chapter 3. Spatializing the eco-functional plantations 77 3.1. Geospatial planning as governance 77 3.2. (Re)creating the enviropreneurial commodity labourer 81 3.3. Advancing the (speculative) socially responsible and green company 88 Conclusions 95 References 98

PART 2 : Entangled (after)lives in Kisangani’s forested hinterland: rubble of imperial rationalization and the unruly multiplicity of socio-spatial-temporal relations 107 Introduction 109 Chapter 4. Living in connectivity: landscape as a socio-natural and spatial-temporal phenomenon 113 4.1. Consideration 1: Dwelling 113 4.2. Consideration 2: Trans-scale multiplicity and power-geometry 115 4.3. Consideration 3: Becoming – time in the landscape 117 4.4. Landscape: a conceptualization 118 Chapter 5. Articulating tenure, identities and forests: untidy relations 121 5.1. Anatomy of a conflict 122 5.2. Alibuku landscape: a story-so-far 129 5.2.1. Alliances, fluid appropriation and forests of belonging 132 5.2.2. Bounding forests and identities 137 5.2.3. Replenishing forests and identities 144 5.2.4. The extractive commodity turn 147 5.2.5. Two weights, two measures 148 5.3. Ruderal socio-political forests 159 Chapter 6. Sangisa: plantation shadows and multispecies peri-urban socio-ecologies 163 6.1. Making fields, producing relations, shaping forests: a socio-ecology of farming 166 6.1.1. Cutting out a field: a more than technical-agricultural meaning 167 6.1.2. Choosing crops 176 6.1.3. Seasonality 183 6.1.4. Organizing production: between mutuality and individuality 187 6.1.5. Entanglements 192 6.2. Forest changes as intensification and social-moral transformations 194 Conclusions 203 References 207

GENERAL CONCLUSION: FROM GEOCODED TO ENTANGLED LANDSCAPE 213 References 223

Appendix 1: Coding system to anonymize interviews 227

LIST OF FIGURES AND TABLES

Figure 1. Nasa maps of fires around the world, August 2019 2 Figure 2. Geological map of the Belgian Congo (1949) 49 Figure 3. Map of main types of mineral, plant and animal production in the Belgian Congo 50 Figure 4. Graphic presentation of the importance of direct deforestation drivers at provincial level 57 Figure 5. Synthetic representation of main direct and underlying deforestation drivers and important explanatory variables 58 Figure 6. Updated synthetic representation of main direct and underlying deforestation drivers and important explanatory variables 58 Figure 7. DRC’s forests under threat 61 Figure 8. Priority zones of the CAFI Programme Intégré Oriental 62 Figure 9. Participatory map carried out with communities inhabiting the perimeter of a logging concession in the hinterland of Kisangani 64 Figure 10. Democratic Republic of Congo: a threatened carbon stock 66 Figure 11. Preliminary land-use plan and key REDD+ program activities in the Maï Ndombe province 81 Figure 12. Participatory tenure map of the intervention zone of the PPRGII Pilot Project in Isangi 84 Figure 13. Basic administrative divisions for rural areas. 124 Figure 14. Excerpt 1 from the Forest Management Plan of ‘Alibuku 018/11 concession’ 125 Figure 15. Excerpt 2 from the Forest Management Plan of ‘Alibuku 018/11 concession’ 126 Figure 16. Landsat 8 satellite map of a part of ‘Alibuku 018/11 concession’, from 10/05/2017 128 Figure 17. Excerpt from an ‘official’ map of Kisangani and its direct hinterland 131 Figure 18. Participatory map conducted with the Bawi ethnic subgroup (03/05/2018).. 143 Figure 19. Participatory map of the land division between villages and clans on the ‘Buta road’ between kilometre points 9 and 18 (11/03/2016) 156 Figure 20. Participatory map of Pumuzika village on the former Amex-Bois road (14/05/2018). 165 Figure 21. Distribution of fire alerts in the hinterland of Kisangani 185

Table 1. Key stakeholders involved in the REDD+ process in DRC 55

LIST OF ABBREVIATIONS

ADIKIS Action pour le Dévelopment Intégral de Kisangani (Action for the Integral Development of Kisangani) AGEDUFOR Appui à la gestion durable des forêts en République démocratique du Congo (Support to the sustainable management of forests in the Democratic Republic of Congo) AFD Agence Française de Développement (French Development Agency) AFOLU Agriculture, forestry and other land use ANT Actor-Network Theory APILAF Association pour les initiatives locales en Afrique forestière (Association for the Promotion of Local Initiatives in the Forested Areas of Africa) BTC-CTB Belgische Technische Coöperatie – Coopération Technique Belge (Belgian Development Agency – now Enabel) CAFI Central African Forest Initiative CCBA Climate, Community & Biodiversity Standards CDF Congolese Franc (DRC’s currency) CIFOR Center for International Forestry Research CODELT Conseil pour Conseil pour la Défense Environnementale par la Légalité et la Traçabilité (Council for Environmental Defense through Legality and Traceability) CSO Civil Society Organization DRC The Democratic Republic of Congo ENGO Environmental Non-Governmental Organization ESCO Edmond Schlüter & Compagnie EU European Union FAQ Frequently Asked Questions FCPF Forest Carbon Partnership Facility FEC Fédération des Entreprises du Congo (Federation of Companies of Congo) FIB Fédération des Industriels du Bois (Industrial Loggers’ Federation in DRC) FIRMS Fire Information for Resource Management FLEGT Forest Law Enforcement, Governance and Trade FMP Forest Management Plan FONAREDD Fonds National REDD+ (DRC’s REDD+ National Fund) FPIC Free, Prior and Informed Consent (UN-REDD Safeguard) FSC Forest Stewardship Council GFW Global Forest Watch GHG Greenhouse gas GTCRR Groupe de Travail Climat REDD+ Rénové (Reformed REDD+ Climate Working Group) IFAD International Fund for Agricultural Development

xii

IFCO Compagnie Industrie Forestière du Congo (Forestry Industry of Congo Company) IGC Institut Géographique du Congo (Geographic Institute of Congo) INEAC Institut National pour l'Etude Agronomique du Congo Belge (National Institute for Agronomic Study in the Belgian Congo) INERA Institut National des Etudes et de Recherches Agronomiques (National Institute for Agronomic Study and Research) JICA Japan International Cooperation Agency KP Kilometre Point MECNT DRC’s Ministry of Environment and Sustainable Development NERICA New Rice for Africa NGO Non-Governmental Organization NRDS National Rice Development Strategy OCEAN ONG Organisation Concertée des Ecologistes et Amis de la Nature (‘Joint Organization of the Environmentalists and Friends of the Nature’ NGO) PGDF Programme de Gestion Durable des Forêts (REDD+ Sustainable Forest Management Program) PRAPO Programme de réhabilitation de l’agriculture en Province Orientale (Agricultural Rehabilitation Program in Oriental Province, IFAD) PRODAT Programme de Développement Agricole dans le District de la Tshopo (Agricultural Development Program in the Tshopo District, BTC-CTB) PPRGII Projet Pilote REDD+ Géograhiquement Intégré d’Isangi (Geographically Integrated REDD+ Pilot Project in Isangi) RCD Rally for Congolese Democracy RDZ Rural Development Zone REDD+ Reducing Emissions from Deforestation and Forest Degradation RFUK Rainforest Foundation UK RGC Réseau Géodésique Commun (Common Geodesic Network) RN4 Route National N°4 (National Road nr. 4) STS Science and Technology Studies UNDP United Nations Development Programme UNFCCC United Nations Framework Convention on Climate Change UN-REDD United Nations Programme on Reducing Emissions from Deforestation and Forest Degradation VCS Verified Carbon Standard VLIR-UOS Vlaamse Interuniversitaire Raad – Universitaire Ontwikkelingssamenwerking WWF World Wildlife Fund WRI World Resource Institute

xiii

xiv

ABSTRACT

This PhD thesis explores Congolese forested landscapes as constituted both by (renewed) planning attempts to reorganize forests and livelihoods into a ‘green plantation economy’ and by everyday lived practices that draw humans and nonhumans together. The basic premise of this thesis is that how we understand and conceptualize landscapes and their environmental changes play an active role in shaping them. Integrating approaches in political ecology, critical cartography and anthropology, and using qualitative data from multi-sited fieldwork, the thesis engages with two ways of framing and attending forests and deforestation in the Democratic Republic of Congo.

The first empirical part examines the representational politics of landscapes embedded in the Reducing Emissions from Deforestation and Forest Degradation (REDD+) policy scheme adopted by the DRC to end the loss of its natural forests and become an emerging economy by 2030. It focuses in particular on the use of maps and geospatial analysis as instruments of environmental rule for diagnosing problems and designing transformation of landscapes into spatially-demarcated zones of production and conservation, i.e. the ‘geocoded landscape’. The analysis deconstructs the entrenched, linear explanations of deforestation as caused by inefficient smallholder swidden farmers’ practices. It also demonstrates how socially weightless (yet political) framings of (de)forested landscapes further increase inequalities and reestablish the rationalizing and homogenizing logics that characterize identities, human labour and ecological complexity in the plantation model.

The thesis’ second empirical part moves away from REDD+ abstract landscapes and takes the reader on a journey in Kisangani’s forested hinterland, which has been identified as a deforestation hotspot. The case study builds on a processual, relational conceptualization of landscape to reconstruct complex, socially grounded understandings of human-environment relations shaping the area’s changing forests. The natural forests REDD+ is concerned with, are in fact naturalcultural spaces that are continuously constituted, over time and across scales, through multiple (violent) practical encounters between people and between people and things (such as land, trees, crops or weeds), i.e. the ‘entangled landscape’. More particularly, this second part focuses on the relations between land tenure, identities and forests, on the socio-ecology of swidden farming and on local understandings of forest change. The data show how various group identities are dynamically articulated with national citizenship in relation to the evolving environment and various land uses, defying containerized logics of ethnic-based environmental governance. They also highlight the diversity among swidden farming practices that are shaped by various ecologies of forested land, historical and structural inequalities and the materiality of crops themselves. Finally the results suggest that processes of land alienation and intensification, rather than purely uncontrolled expansion for subsistence farming, are at the root of ‘anti-social’ forest cover change.

Throughout the thesis, the analysis demonstrates that prevailing expert representations of Congolese forests and their dwellers carry embedded notions of value and productivity that not only render illegitimate a plurality of human-environment entanglements and practices, but also push forward the very same (colonial) mechanisms that contribute to deforestation and uneven development in the first place. The conclusion argues that to address the various facets of the so-called ‘Anthropocene’, we need to engage with multiple dimensions of justice as much as shift focus from human-in-nature to human- with-nature. The thesis, in this sense, also provides a theoretical contribution to ‘more-than- representational’ approaches to landscapes by integrating them with the type of political economic analyses at the heart of the political ecology tradition.

xvi

INTRODUCTION

Over the last decade, environmental change in its many forms –climate change, ozone depletion, deforestation, biodiversity loss among others– has received growing attention in policy circles and in the media, with increasingly visible global civil society movements urging governments to take more radical actions, still with only mitigated success. I am writing this introduction in the middle of another (yet related) crisis: the Covid-19 pandemic which has confined billions of people to their homes across the world resulting in a radical but temporary reduction in greenhouse gas emissions (GHG) (Ambrose, 2020; Le Quéré et al., 2020). Agriculture, forestry and other land use (AFOLU) contributes to nearly 25% of global GHG which accounts for deforestation, fire, wood harvesting, and agricultural emissions including croplands, paddy rice, and livestock (Romijn et al., 2018). For years, some experts have shown the links between increases in infectious diseases and “big farms” (Wallace, 2016), i.e. the industrial monocultural model of food production, of livestock in particular, and with the large-scale land use changes, such as deforestation, that they trigger (e.g. Afelt, Frutos, & Devaux, 2018; Allen et al., 2017; Loh et al., 2015). Such analysis is almost absent in mainstream media as much as the fact that the burden of environmentally mediated infections is, unsurprisingly, disproportionately affecting the rural poor (e.g. Garchitonera et al.; 2017; Wallace, 2016).

Not so long before the current health crisis, in the summer of 2019, deforestation precisely, and in particular the massive related blazes in Brazil captured the world’s attention by means of mass media. Very quickly, supported by ‘evidence’ of alarming NASA satellite images (Figure 1), the blazes were put in parallel with fires across Central Africa, in particular in Angola and the Democratic Republic of Congo (DRC) where the research informing this manuscript took place. The screenshot of NASA’s Fire Information for Resource Management (FIRMS)’s platform went viral in the news and on social media platforms, with French President Emmanuel Macron tweeting “our house is burning” (Macron, 2019a) and announcing aid initiatives to tackle the problem in both regions (Macron, 2019b). While some articles (e.g. RFI & Hird, 2019; Willingham, 2019) had rightly cautioned against comparing fires in these two regions too closely – leading to erroneously equate permanent industry-driven deforestation in the Amazon, with much more environmentally sustainable cyclical land management in Central Africa – in most cases the image was circulated without comment or much analysis, equating the two situations in terms of their environmental impacts (see BBC News (2019) for a sample of Twitter posts). What penetrated people’s imaginaries was the dramatic loss of two of the most important carbon stocks that regulate the global climate and of “sanctuaries for hundreds of endangered species” as framed in an article in The Sun on the topic (Lockett, 2019), rather than the loss of human habitats and sources of livelihoods.

The mass diffusion of such depictions and the urgency they confer, like a (corona)virus which is seemingly affecting people around the world in an undifferentiated, equal way, increases awareness about environmental harms and risks, yet “reduces the mosaic of human activity in the web of life to an abstract, homogenous humanity” (Moore, 2016, p. 82) that is equally harmful to a pristine, wild and carbon storing commonly-shared nature, and that is equally impacted by its destruction. I begin with such a powerful image because, paradoxically, it captures the ways forests and their changes are separated from the specificity of place, i.e. away from their social production, discounting the variety of lived experiences (Debord, 1995 [1967]) and of socio-environmental struggles as much as removing “inequality, commodification, imperialism, patriarchy, and much more from the problem of humanity-in- nature” (Moore, 2006, p.82). It invites us to refigure locality, history and justice within these planetary debates, which is what this PhD thesis attempts to do by scratching below the surface of some of the ‘spectacular’ red dots on the DRC map. The thesis simultaneously deconstructs the how, why and by whom the country’s forests are constructed as an object of urgent intervention and it reconstructs these spaces as ‘more-than-abstract-nature’ under threat.

Figure 1. Nasa maps of fires around the world, August 2019 (BBC News, 2019). This image which was extracted from Nasa’s Fire Information for Resource Management (FIRMS) was largely disseminated on social media and in various newspapers from BBC and RFI to Forbes.

It is precisely the spotlight given to Congolese forested landscapes for their role in climate change mitigation that brought me to DRC, and more particularly to Kisangani’s hinterland where intentional fires for cultivation seasonally pop up alarmingly on satellite monitoring systems. Since 2009, the portrayal of DRC as an underdeveloped environment in crisis has driven, millions of dollars of

international funding for the country’s participation in the international policy framework Reducing Emissions from Deforestation and Forest Degradation (REDD+). The programme intends to end the loss of “natural forests” by 2030, by “transforming the Democratic Republic of the Congo’s development path towards a green economy” (CAFI-DRC, 2015, pp. 2-4) which combines forest protection, climate change mitigation, economic growth and human development in a fast and sustainable manner (CAFI-DRC, 2015). To achieve this goal, the DRC has adopted what is labelled as a REDD+ ‘landscape’, ‘jurisdictional’ or ‘ecosystem’ approach1 in the international environmental and development community. Despite its broad and fuzzy definition, this approach aims to integrate all stakeholders, sectors and multiple land- use types such as intensive agriculture and logging with a larger range of environmental and developmental goals such as food security and biodiversity conservation at the landscape-scale (CGIAR, 2013; Freeman, Duguma, & Minang, 2015; McCall, 2016; Nielsen, 2016; Sayer et al., 2013; Turnhout et al., 2017).

The landscape approach has gained momentum as a new conceptualization of REDD+ at the United Nations Framework Convention on Climate Change (UNFCCC) as from the mid-2010s (Nielsen, 2016). Initially, REDD+ was conceived as an incentive-based mechanism to conserve forests in developing countries, in which payments were conditional on the delivery of the ecosystem service that forests provide through carbon sequestration and storage (Corbera, 2012). From its outset, in 2005, REDD+ was framed as a new ‘all-win’ conservation paradigm which could reduce carbon emissions in a cost-effective way by encouraging ‘green’ land use while raising poor land users’ income (Wunder, 2006). Over time, following years of under-achieving projects and relentless opposition at national and international levels (Asiyanbi & Lund, 2020), REDD+ has progressively included more actors and expanded its scope and focus from a carbon- and forest-centred approach to a single management process that addresses broader dynamics of multiple land use in forested landscapes (den Besten, Arts, and Verkooijen, 2014; McCall, 2016; Turnhout et al. 2017).

In recent years critical scholars, political ecologists in particular, have underlined how the constant evolution and promises of change in REDD+ discourses are in fact characterized by much policy persistence, resembling longstanding dynamics of the conservation and development industry (Asiyanbi & Lund, 2020; Fletcher, Dressler, Büscher, & Anderson, 2016; Lund, Sungusia, Mabele, & Scheba, 2017). Building on Foucault’s (1991, 2008) concept of ‘governmentality’, forests-cum-development paradigms

1 As shown by Sayer et al. (2013), there is no universal definition of the ‘landscape approach’. In the REDD+ literature, some publications use the term ‘landscape approach’, others ‘ecosystem approach’ or ‘jurisdictional approach’ and still some others use the two terminologies interchangeably in the same text depending on whether they particularly emphasize an economic, ecological or administrative conceptualization of the ‘landscape’ (Weatherley-Singh and Gupta, 2017). However, these different publications generally refer to the same environmental governance model that aligns multiple land-use types (the so-called ‘landscape’) with administrative jurisdictions and to coordinate multiple goals, initiatives and stakeholders. For the sake of clarity, in this thesis I refer only to the ‘landscape approach’ and consider it as equivalent to the ‘ecosystem approach’ or ‘jurisdictional approach’ in the context of REDD+ architecture. 3

such as REDD+ have been conceptualized successively as forms of ‘green governmentality’ (Luke, 1999; Rutherford, 2007), ‘environmentality (Agrawal, 2005), ‘neoliberal environmentality’ (Fletcher, 2010) or ‘eco-governmentality’ (Valdivia, 2015), emphasizing their overarching structure of stability. Overall, these concepts highlight a particular type of power to govern human-environmental relations through a set of calculated, rational means and strategies used by a heterogenous assemblage of actors and institutions to shape and order human behaviours to manage environmental crisis.

Li (2007) has analysed how neoliberal governmental rationality in Indonesia has always been shaped, from the colonial period to present by three main goals: 1) the improvement of the population by reconstituting them as orderly villagers; 2) the improvement of the productivity of the land as a source of revenue and prosperity; 3) the improvement of the landscape by protecting forests from use and abuse. McElwee (2016) similarly argues that interventions to manage and protect the environment in Vietnam have in fact always been about seeing, planning and improving population through intervening in disparate areas such as labour availability, rural settlement or markets; a strategy that she terms ‘environmental rule’. Similarly REDD+ is concerned with the establishment of new subjectivities – which involves certain forms of identity, behaviours and practices – and the restriction of use and access to forests to protect carbon stocks (Astuti & McGregor, 2015a). This form of ordering society and nature involves two interrelated practices: first, the identification of deficiencies that need to be rectified, i.e. the ‘problematization’ practice; second, ‘rendering technical’, i.e. a “whole set of practices concerned with representing the domain to be governed as an intelligible field with specifiable limits and particular characteristics” (Li, 2007, p.7). Typically, REDD+ heavily relies on complex algorithms for the economic valuation of carbon, computer modelling, satellite imageries and maps to produce ‘truth’ about nature and people, and “prescribe appropriate conduct for population to act upon” (Astuti & McGregor, 2015a, p. 23; Robertson, 2012).

Various scholars have emphasized that such a ‘knowledge economy’ in which certain kinds of meanings and experts are privileged (Büscher, 2014) leads to a social disembeddedness of ‘nature’ (McAfee & Shapiro, 2010). It downplays local users’ knowledge and various values attributed to ‘nature’ and hence the connections between social and environmental domains, on the one hand, and power relations that shape environmental resource management, on the other (Kull et al., 2015; Swyngedouw, 2011; Van Hecken et al., 2015). REDD+ thus reproduces the dichotomy between Nature and Culture characteristic of Western modernity, that contributes to the further disenfranchisement of local forest dwellers from their lands and livelihoods in postcolonial settings thus perpetuating social and environmental injustice (Latour, 1993; Sikor, 2014; Sullivan, 2014; Tsing, 2005). Throughout this scholarship, indeed, one of the important question has been why are the people dispossessed from their lands, labour, and resources

through processes of commodification and dispossession, also often the most blamed for ecological ravages caused by these processes (Blaikie, 1999; Bryant, 1998; West, 2012).

A mounting number of recent case studies has also shown how REDD+ environmental rule is variously translated into national REDD+ strategies and how it interacts with everyday economic, socio-cultural, institutional and political practices on the ground, and with existing inequalities, vulnerabilities and injustices (e.g. Asiyanbi, Ogar & Akintoye (2019), Astuti & McGregor (2015b, 2017); Benjaminsen (2014); Chomba, Kariuki, Lund, & Sinclair, 2016; Leach & Scoones (2015); Milne et al. (2019)). Few studies, however, have empirically looked from such a political economic perspective at the ‘new’ conceptualization of landscape in REDD+ conservation and its operationalization (but see Clay, 2016; McCall, 2016), and at the frictions with embodied, practiced and lived landscapes of everyday life. To my knowledge, critical analyses of the discourses that underlie REDD+ strategies in DRC are scarce (but see Ehrenstein, 2013, 2014) –in contrast with the more abundant literature for other countries like Indonesia– and there is, more generally, only limited academic empirical material available on DRC’s REDD+ implementation (but see e.g. Aquino & Guay, 2013; Reyniers, Karsenty , & Vermeulen, 2016; Samndong, Bush, Vatn, & Chapman, 2018; Samndong & Kjosavik, 2017).

Theoretically, the poststructuralist turn in political ecology has been extremely productive in deconstructing the politics of environmental knowledge and interventions such as REDD+, as much as in opening up the category of the environment itself by adopting a critical orientation toward nature-society relations. Such approaches remain key to challenging unjust environmental policies. Yet, quite paradoxically, in analysing the landscapes in which these policies unfold, the focus has generally remained on human-in-nature and has paid too little attention to the ways nonhuman and material entities contribute to shape sociality –i.e. socio-political relations (Castree, 2003; Rose & Wilson, 2019 ; White, Rudy, & Gareau, 2016). White et al. (2016, p.87) assert that political ecological analyses tend to keep an “excessive social view of nature”. To address the multiple facets of the ecological crisis, it is necessary to shift our focus to human-with-nature. In this way, posthumanist and more-than-human theories that seek to move beyond the nature-culture divide think of “humans as always in composition with nonhumanity, never outside of a sticky web of connections or an ecology [of matter]” (Bennett in Whatmore, 2006, p. 603, original emphasis). The use of posthumanist approaches in political ecology (e.g. Hinchcliffe, 2007; Robbins, 2012; Sundberg, 2011) has, however, been criticized, by some, for being apolitical (e.g. Lave, 2015; Malm, 2018) by giving too much agency to the nonhuman (Castree, 2003). The challenge remains to engage with such relational ontologies and epistemologies that trace human and nonhuman entanglements while also maintaining the deep commitment to the political project of political ecology. A project “that learns from historically and geographically specific (albeit always relationally defined) political ecological circumstances […] and that refuses an appeal to an abstract

‘people’ or Anthropos, [and to abstract forests] while also learning from ordinary people and their situated knowledges and practices” (Loftus, 2019, p.7).

In light of these debates, the central question that guides the research in this thesis can thus be formulated as follows:

How are forested landscapes constituted through human-nonhuman practical encounters in the current politics of REDD+ and in the geographies of everyday life?

To answer this main question, the thesis is divided into two main empirical parts. The first part explores the politics of the REDD+ landscape approach in DRC. The main sub-questions guiding the first empirical part are:

How is the REDD+ landscape approach constructed and translated in DRC? What are the discourses and knowledge practices informing the national strategy? What kind of socio-ecologies and subjectivities are enacted? And who are the ‘winners’ and ‘losers’ in this process?

The second empirical part explores the geographies of everyday life. The main sub-questions guiding this second part are:

How do some of the everyday power-laden practices drawing human and nonhuman together constitute lived landscapes through time and space? How are identities and forests’ multiple meanings co-constituted by land tenure and land use practices? How can forest change be understood in relation to social life?

Outline of the thesis

To answer these questions, and as mentioned before, the thesis is divided into two main empirical parts. Before digging into those, the thesis exposes the epistemological stance and methodology that have informed this research. More particularly, I develop an approach to engage politically with human- nonhuman relations. After that, each of the empirical parts contains an introduction, a theoretical framework that guides the interpretation of the empirical data, two empirical chapters, and a conclusion.

Part I is concerned with the making of a ‘geocoded landscape’ (Rose-Redwood, 2006) under REDD+. It analyses the construction and use of seemingly neutral cartographic representations of Congolese forests and of geospatial governance for producing green economic landscapes under REDD+ process. Chapter 1 starts by exploring the entangled history of modern mapping with political economy. It then outlines a relational understanding of mapping that emphasize the recursive interplay between maps and the socio- natural world. Chapter 2 examines the role of remote-sensing maps and geovisualizations within DRC’s

‘messy’ REDD+ agencement of forest experts, REDD+ consultants, state actors and CSOs, and of other material, discursive and historical elements, to produce knowledge about forests, people and deforestation. It pays attention to the ways geographical knowledge allows us not only to plan certain actions of socio-ecological modernization but also to solve other relational problems between stakeholders with different interests. Chapter 3 then turns to the REDD+ strategy developed in DRC, how it is embedded in particular historical discourses of modernization and the role of geospatial governance and planning in organizing forests and people. I emphasize the power inequalities between actors in the production of REDD+ integrated landscapes.

Part II is concerned with ‘lived landscapes’ (Lefebvre, 1991). It explores the relationality of Kisangani hinterland’s landscape, in that it is constituted to porous networks of socio-ecological relations and everyday practices. It aims to bring to light the entanglement of multiple temporal layers and socio- environmental legacies of (pre-/post-)colonial, capitalist, and present ruly and unruly modes of being that are co-constitutive of forest landscapes and socio-spatial-temporal identities. Chapter 4 outlines a conceptualization of landscape processual understanding that articulates the four dimensions of (multispecies) sociality, materiality, spatiality and temporality. Landscape, I argue, is simultaneously historical and emergent, social and natural, material and imagined. It is produced out of a constantly negotiated, conflictual process of multiple activities and identities. Chapter 5 first addresses the articulations of forests, tenure and identities in the constitution of a contested landscape. I am attentive to the dynamism of tenure and identities in relation to changing forests’ values and changing circumstances, both in the social and in the physical environments. Chapter 6 then explores the socio-ecology of farming. Farming is the is the very life blood that sustains most inhabitants of the Kisangani hinterland from peasant farmers to many urban elites. I look at the dynamism of swidden agriculture and the ways it is shaped by human cultivators (their agency), socio-economic relations across time and space scales, seasonality and the materiality of resources (forests, land and crops). The chapter also looks at local understandings of socio-ecological change. The concluding section of Part II lays out other ways to think of deforestation that differ from REDD+ framings that I deconstruct in Part I. Changes in land use practices and their ensuing ecological changes touch deeply on identities and social relations. I also show that the environmental crisis is first and foremost a social crisis that arises from inequalities.

The general conclusion then attempts to relate both perspectives on landscape that I have adopted, and the different understandings of socio-ecological relations and change, and the ways they can be negotiated. Such insights highlight how technoscience, diverse knowledges and crucial issues of environmental justice are and should be woven together.

References

Afelt, A., Frutos, R., & Devaux, C. (2018). Bats, Coronaviruses, and Deforestation: Toward the Emergence of Novel Infectious Diseases? Frontiers in microbiology, 9, 702-702. doi:10.3389/fmicb.2018.00702

Agrawal, A. (2005). Environmentality: Technologies of Government and the Making of Subjects. Durham: Duke University Press.

Allen, T., Murray, K. A., Zambrana-Torrelio, C., Morse, S. S., Rondinini, C., Di Marco, M., . . . Daszak, P. (2017). Global hotspots and correlates of emerging zoonotic diseases. Nature communications, 8(1), 1124-1124. doi:10.1038/s41467-017-00923-8

Ambrose, J. (2020, 12 April). Carbon emissions from fossil fuels could fall by 2.5bn tonnes in 2020. The Guardian. Retrieved from https://www.theguardian.com/environment/2020/apr/12/global- carbon-emisions-could-fall-by-record-25bn-tonnes-in-2020

Aquino, A., & Guay, B. (2013). Implementing REDD+ in the Democratic Republic of Congo: An analysis of the emerging national REDD+ governance structure. Forest Policy and Economics, 36(0), 71-79. doi:http://dx.doi.org/10.1016/j.forpol.2013.04.003

Asiyanbi, A. P., Ogar, E., & Akintoye, O. A. (2019). Complexities and surprises in local resistance to neoliberal conservation: Multiple environmentalities, technologies of the self and the poststructural geography of local engagement with REDD+. Political Geography, 69, 128-138. doi:https://doi.org/10.1016/j.polgeo.2018.12.008

Asiyanbi, A., & Lund, J. F. (2020). Policy persistence: REDD+ between stabilization and contestation. Journal of Political Ecology, 27(1), 378-400.

Astuti, R., & McGregor, A. (2015a). Governing carbon, transforming forest politics: A case study of Indonesia's REDD+ Task Force. Asia Pacific Viewpoint, 56(1), 21-36. doi:10.1111/apv.12087

Astuti, R., & McGregor, A. (2015b). Responding to the green economy: how REDD+ and the One Map Initiative are transforming forest governance in Indonesia. Third World Quarterly, 36(12), 2273- 2293. doi:10.1080/01436597.2015.1082422

Astuti, R., & McGregor, A. (2017). Indigenous land claims or green grabs? Inclusions and exclusions within forest carbon politics in Indonesia. The Journal of Peasant Studies, 44(2), 445-466.

BBC News. (2019, 26 August). Amazon fires: Angola and DR Congo 'have more blazes'. BBC News. Retrieved from https://www.bbc.com/news/world-africa-49471644

Benjaminsen, G. (2014). Between Resistance and Consent: Project–Village Relationships When Introducing REDD+ in Zanzibar. Forum for Development Studies, 41, 377-398.

Blaikie, P. (1999). A Review of Political Ecology: Issues, Epistemology and Analytical Narratives. Zeitschrift für Wirtschaftsgeographie, 3-4, 131-147.

Bryant, R. L. (1998). Power, knowledge and political ecology in the third world: a review. Progress in Physical Geography, 22(1), 79-94.

Büscher, B. (2014). Selling Success: Constructing Value in Conservation and Development. World Development, 57(0), 79-90. doi:http://dx.doi.org/10.1016/j.worlddev.2013.11.014

Castree, N. (2003). Environmental issues: relational ontologies and hybrid politics. Progress in Human Geography, 27(2), 203-211. doi:10.1191/0309132503ph422pr

CAFI-DRC. (2015). Plan d’investissement REDD+ (2015-2020). Retrieved from http://www.cafi.org/content/dam/cafi/docs/drc- documents/DRC_2015_Plan%20Investissement%20National%20REDD%20RDC%2024%20Nove mbre%202015.pdf

CGIAR. (2013). Going from Red to Green: the landscape approach. CIFOR Forests News Blog. Retrieved from http://www.cgiar.org/consortium-news/going-from-red-to-green-the-landscape- approach/

Chomba, S., Kariuki, J., Lund, J. F., & Sinclair, F. (2016). Roots of inequity: How the implementation of REDD+ reinforces past injustices. Land Use Policy, 50, 202-213. doi:https://doi.org/10.1016/j.landusepol.2015.09.021

Clay, N. (2016). Producing hybrid forests in the Congo Basin: A political ecology of the landscape approach to conservation. Geoforum, 76(Supplement C), 130-141. doi:https://doi.org/10.1016/j.geoforum.2016.09.008

Corbera, E. (2012). Problematizing REDD+ as an experiment in payments for ecosystem services. Current Opinion in Environmental Sustainability, 4(6), 612-619. doi:http://dx.doi.org/10.1016/j.cosust.2012.09.010

Debord, G. (1995 [1967]). The Society of the Spectacle. New York: Zone Books. den Besten, J. W., Arts, B., & Verkooijen, P. (2014). The evolution of REDD+: An analysis of discursive- institutional dynamics. Environmental Science & Policy, 35, 40-48. doi:http://dx.doi.org/10.1016/j.envsci.2013.03.009

Ehrenstein, V. (2013). Les professionnels de la préparation. Aider la République Démocratique du Congo à réduire sa déforestation : Programme redd+. Sociologies Pratiques(27), 91-104. doi: 10.3917/sopr.027.0091

Ehrenstein, V. (2014). Géopolitique du carbone : L’action internationale pour le climat aux prises avec la déforestation tropicale. (PhD Thesis), Ecole Nationale Supérieure des Mines de Paris, Paris.

Fletcher, R. (2010). Neoliberal Environmentality: Towards a Poststructuralist Political Ecology of the Conservation Debate. Conservation and Society, 8(3), 171-181.

Fletcher, R., Dressler, W., Büscher, B., & Anderson, Z. R. (2016). Questioning REDD+ and the future of market-based conservation. Conservation Biology, 30(3), 673-675. doi:10.1111/cobi.12680

Foucault, M. (1991). Governmentality. In Burchell, G., Colin, G. and Miller, P. (Eds.), The Foucault Effect: Studies in Governmentality (pp. 87-104). Chicago: University of Chicago Press.

Foucault, M. (2008). The Birth of Biopolitics: Lectures at the Collège de France, 1978-1979. New York: Palgrave Macmillan.

Freeman, O. E., Duguma, L. A., & Minang, P. A. (2015). Operationalizing the integrated landscape approach in practice. Ecology and Society, 20(1). doi:10.5751/ES-07175-200124

Garchitorena, A., Sokolow, S. H., Roche, B., Ngonghala, C. N., Jocque, M., Lund, A., . . . De Leo, G. A. (2017). Disease ecology, health and the environment: a framework to account for ecological and socio- economic drivers in the control of neglected tropical diseases. Philosophical Transactions of the Royal Society B: Biological Sciences, 372(1722), 20160128. doi:10.1098/rstb.2016.0128

Hinchliffe, S. (2007). Geographies of Nature: Societies, Environments, Ecologies. London: Sage Publications.

Kull, C. A., Arnauld de Sartre, X. & Castro-Larrañaga, M. (2015). The political ecology of ecosystem services. Geoforum, 61, 122-134.

Lave, R. (2015). Reassembling the Structural: Political ecology and Actor-Network Theory. In Perreault, T., Bridge, G. and, McCarthy, J. (Eds.), The Routledge Handbook of Political Ecology (pp. 213-223). London & New York: Routledge.

Leach, M., & Scoones, I. (Eds.). (2015). Carbon Conflicts and Forest Landscapes in Africa. Oxon: Routledge.

Le Quéré, C., Jackson, R. B., Jones, M. W., Smith, A. J. P., Abernethy, S., Andrew, R. M., . . . Peters, G. P. (2020). Temporary reduction in daily global CO2 emissions during the COVID-19 forced confinement. Nature Climate Change, 10(7), 647-653. doi:10.1038/s41558-020-0797-x

Li, T. M. (2007). The Will to Improve. Governmentality, Development, and the Practice of Politics. Durham & London: Duke University Press.

Lockett, J. (2019, 27 August). RED HEAT Shocking satellite images show there are MORE wildfires raging in Africa than in the Amazon as blazes devastate the world’s ‘second green lung’. The Sun. Retrieved from https://www.thesun.co.uk/news/9803695/satellite-image-congo-africa-fires- amazon/

Loftus, A. (2019). Political ecology III: Who are ‘the people’? Progress in Human Geography, 0309132519884632. doi:10.1177/0309132519884632

Loh, E. H., Zambrana-Torrelio, C., Olival, K. J., Bogich, T. L., Johnson, C. K., Mazet, J. A. K., . . . Daszak, P. (2015). Targeting Transmission Pathways for Emerging Zoonotic Disease Surveillance and Control. Vector borne and zoonotic diseases (Larchmont, N.Y.), 15(7), 432-437. doi:10.1089/vbz.2013.1563

Luke, T. W. (1999). Environmentality as green governmentality. In Darier, E. (Ed.), Discourses of the environment (pp. 121-151). Oxford: Blackwell.

Lund, J. F., Sungusia, E., Mabele, M. B., & Scheba, A. (2017). Promising Change, Delivering Continuity: REDD+ as Conservation Fad. World Development, 89, 124-139. doi:http://dx.doi.org/10.1016/j.worlddev.2016.08.005

Macron, E. (2019a, 22 August). Our house is burning. Literally. The Amazon rain forest - the lungs which produces 20% of our planet’s oxygen - is on fire. It is an international crisis. Members of the G7 Summit, let's discuss this emergency first order in two days! #ActForTheAmazon [Tweet]. Retrieved from https://twitter.com/EmmanuelMacron/status/1164617008962527232

Macron, E. (2019b, 26 August). La forêt brûle également en Afrique subsaharienne. Nous sommes en train d'examiner la possibilité d'y lancer une initiative similaire à celle que nous venons d'annoncer pour l'Amazonie. #G7Biarritz. Retrieved from https://twitter.com/EmmanuelMacron/status/1165966950641610753

Malm, A. (2018). The Progress of This Storm: Nature and Society in a Warming World. London: Verso.

McAfee, K. & Shapiro, E. N. (2010). Payments for Ecosystem Services in Mexico: Nature, Neoliberalism, Social Movements, and the State. Annals of the Association of American Geographers, 100, 579- 599.

McCall, M. K. (2016). Beyond “Landscape” in REDD+: The Imperative for “Territory”. World Development, 85(Supplement C), 58-72. doi:https://doi.org/10.1016/j.worlddev.2016.05.001

McElwee, P. (2016). Forests Are Gold: Trees, People, and Environmental Rule in Vietnam. Seattle: University of Washington Press.

Milne, S., Mahanty, S., To, P., Dressler, W., Kanowski, P., & Thavat, M. (2019). Learning From ‘Actually Existing’ REDD+ A Synthesis of Ethnographic Findings. Conservation & Society, 17(1), 84-95. doi:10.2307/26554473

Moore, J. W. (2016). The Rise of Cheap Nature. In Moore, J. W. (Ed.), Anthropocene or Capitalocene? Nature, History, and the Crisis of Capitalism (pp. 78-115). Oakland: PM Press.

Nielsen, T. D. (2016). From REDD+ forests to green landscapes? Analyzing the emerging integrated landscape approach discourse in the UNFCCC. Forest Policy and Economics, 73(Supplement C), 177-184. doi:https://doi.org/10.1016/j.forpol.2016.09.006

Reyniers, C., Karsenty , A., & Vermeulen, C. (2016). Les paysans sans terre et REDD+ en RDC: Les logiques locales face aux interventions internationales. Cahiers africains (Conjonctures Congolaises) 87, 199-225.

RFI, & Hird, A. (2019, 27 August). Congo Basin vs the Amazon: a tale of two forest fires. rfi. Retrieved from http://www.rfi.fr/en/africa/20190827-africa-forest-fire-amazon-deforestation

Robbins, P. (2012.) Lawn People: How Grasses, Weeds, and Chemicals Make Us Who We Are. Philadelphia: Temple University Press.

Robertson, M. (2012). Measurement and alienation: making a world of ecosystem services. Transactions of the Institute of British Geographers, 37(3), 386-401. doi:10.1111/j.1475-5661.2011.00476.x

Romijn, E., De Sy, V., Herold, M., Böttcher, H., Roman-Cuesta, R. M., Fritz, S., . . . Martius, C. (2018). Independent data for transparent monitoring of greenhouse gas emissions from the land use sector – What do stakeholders think and need? Environmental Science & Policy, 85, 101-112. doi:https://doi.org/10.1016/j.envsci.2018.03.016

Rose, J., & Wilson, J. (2019). Assembling Homelessness: A Posthumanist Political Ecology Approach to Urban Nature, Wildlife, and Actor-Networks. Leisure Sciences, 41(5), 402-422.

Rose-Redwood, R. S. (2006). Governmentality, Geography, and the Geo-Coded World. Progress in Human Geography, 30(4), 469-486. doi:10.1191/0309132506ph619oa

Rutherford, S. (2007). Green governmentality: insights and opportunities in the study of nature's rule. Progress in Human Geography, 31(3), 291-307. doi:10.1177/0309132507077080

Samndong, R. A., & Kjosavik, D. J. (2017). Gendered forests: exploring gender dimensions in forest governance and REDD+ in Équateur Province, Democratic Republic of Congo (DRC). Ecology and Society, 22(4). doi:10.5751/ES-09753-220434

Samndong, R. A., Bush, G., Vatn, A., & Chapman, M. (2018). Institutional analysis of causes of deforestation in REDD+ pilot sites in the Equateur province: Implication for REDD+ in the Democratic Republic of Congo. Land Use Policy, 76, 664-674. doi:https://doi.org/10.1016/j.landusepol.2018.02.048

Sayer, J., Sunderland, T., Ghazoul, J., Pfund, J.-L., Sheil, D., Meijaard, E., . . . Buck, L. E. (2013). Ten principles for a landscape approach to reconciling agriculture, conservation, and other competing land uses. Proceedings of the National Academy of Sciences of the United States of America, 110(21), 8349-8356. doi:10.1073/pnas.1210595110

Sikor, T. (2014). REDD+: Justice effects of technical design. In T. Sikor (Ed.), The Justices and Injustices of Ecosystem Services (pp. 46-68). Abingdon, UK: Earthscan/Routledge.

Sundberg, J. (2011). Diabolic Caminos in the Desert and Cat Fights on the Río: A Posthumanist Political Ecology of Boundary Enforcement in the United States–Mexico Borderlands. Annals of the Association of American Geographers, 101(2), 318-336. doi:10.1080/00045608.2010.538323

Sullivan, S. (2014). The natural capital myth; or will accounting save the world? Preliminary thoughts on nature, finance and values, LCSV Working Paper No. 3, Leverhulme Centre for the Study of Value. School of Environment, Education and Development. The University of Manchester.

Swyngedouw, E. 2011. Whose environment? The end of nature, climate change and the process of post- politicization. Ambiente & Sociedade, 14, 69-87.

Tsing, A. L. (2005). Friction: An Ethnography of Global Connection. Princeton and Oxford: Princeton University Press.

Turnhout, E., Gupta, A., Weatherley-Singh, J., Vijge, M. J., de Koning, J., Visseren-Hamakers, I. J., . . . Lederer, M. (2017). Envisioning REDD+ in a post-Paris era: between evolving expectations and current practice. Wiley Interdisciplinary Reviews: Climate Change, 8(1), e425-n/a. doi:10.1002/wcc.425

Valdivia, G. (2015). Eco-Governmentality. In Perreault, T., Bridge, G. and, McCarthy, J. (Eds.), The Routledge Handbook of Political Ecology (pp. 467-480). London and New York: Routledge.

Van Hecken, G., Bastiaensen, J., & Windey, C. (2015). Towards a power-sensitive and socially-informed analysis of payments for ecosystem services (PES): Addressing the gaps in the current debate. Ecological Economics, 120, 117-125. doi:http://dx.doi.org/10.1016/j.ecolecon.2015.10.012

Wallace, R. (2016). Big Farms Make Big Flu: Dispatches on Infectious Disease, Agribusiness, and the Nature of Science. New York: Monthly Review Press.

Whatmore, S. (2006). Materialist returns: practising cultural geography in and for a more-than-human world. cultural geographies, 13(4), 600-609. doi:10.1191/1474474006cgj377oa

West, P. (2012). From Modern Production to Imagined Primitive: The Social World of Coffee from Papua New Guinea. Durham: Duke University Press.

White, D., Rudy, A., & Gareau, B. (2016). Environments, natures and social theory. New York, NY: Palgrave.

Willingham, A. J. (2019, 27 August). Yes, more fires are burning in sub-Saharan Africa than in Brazil. But context is important. CNN. Retrieved from https://edition.cnn.com/2019/08/27/world/africa- wildfires-amazon-trnd/index.html

Wunder, S. (2006). Are Direct Payments for Environmental Services Spelling Doom for Sustainable Forest Management in the Tropics? Ecology and Society, 11(2).

EPISTEMOLOGICAL AND METHODOLOGICAL APPROACH

1. Engaging with multiplicity in times of ‘post-truth’: a performative relational onto- epistemology

The starting assumption of this thesis is that how we understand and conceptualize landscapes and their environmental changes play an active role in shaping them. Feminist and post-colonial/decolonial epistemologies have emphasized how Western capitalism, colonialism and patriarchy run in parallel with Eurocentric modern technical-scientific knowledge, and in turn legitimize themselves by discrediting and trivializing other knowledges and practices (e.g. Gibson-Graham, 2006, 2008, 2014; Haraway, 1991; Mignolo, 2009; Santos, 2014). This process is what Haraway (1991, p. 189) has called the “god-trick of seeing everything from nowhere”– which she links to visualizing/mapping technologies that I analyse in this thesis; or Santos (2018)’s ‘Cognitive Empire’; the kind of epistemic violence that Fanon (1968) already denounced fifty years earlier. All argue that Enlightened rationality under the guise of neutrality, hides a specific position (white, male) that creates an idea of universal knowledge and humanism. The ontological order characteristic of modernity is based on dualisms such as culture/nature, subject/object, human/nonhuman, mind/body that simultaneously creates hierarchies of class, sex/gender, and race (Escobar, 1999; Haraway, 2003). Western-centric ontologies and epistemologies thus have ethical- political effects as they marginalize and disqualify a plurality of ways of being, meanings and knowledges by rendering them invalid, subjective or non-existent. They served to dehumanize the ‘primitive’, ‘emotional’ colonized during the colonial period and still mainly “relegate the colonized spaces to the periphery and to the past […] (even if that ‘past’ paradoxically exists in the ‘present’)” (Alcoff, 2007, p. 87). The corollary assumption is that, through our research practices, it is possible to enact multiple modes of being and knowing that were actively produced as ‘non-existent’ (Santos, 2014).

Quite similarly to Santos’ (2014) ecologies of difference, political ecology has engaged with deconstructing dominant explanatory narratives of environmental change that ignore certain types of knowledge and reconstructing other explanations that can address socio-ecological issues and the needs of poor communities in a more effective and just way (Blaiki, 1999; Forsyth, 2001, 2008). The core argument of these critics of dominant knowledge is that issues of social and environmental justice are rooted in epistemic and ontological injustice. Yet, the premise of my approach is that all knowledge, i.e. both ‘expert’ and ‘local’, is incomplete, partial, cultural, social and political, i.e. ‘situated’ and historically contingent (Haraway, 1991). As Robbins (2003, p. 248) asserts, “[p]olitics is prior to the conflict between the two environmental knowledges; it underlies the epistemologies themselves”. Yet, as Price (2013, pp. 18-19) argues, it is only problematic that our epistemological choices are about some groups gaining power and achieving political agendas “if one does not distinguish between: power1 – which is the

transformative capacity of the agent; and power2 – which are relations of domination, exploitation, subjugation and control, which can also be described as generalized master-slave(-type) relations”.

Unveiling the epistemological violence of “certain hegemonic types of truth making” in environmental policy (Neimark et al., 2019, p.2), poses the challenge of affirming a material reality to global environmental problems – that may go beyond direct human experience and exist uniformly across space – such as carbon emissions with which REDD+ is concerned, while highlighting some of the multiple, localized understandings of environmental degradation and paths to tackle it (Carolan, 2004; Forsyth, 2001). This is an increasing challenge in a political time of ‘post-truth’ in which “terms such as alternative facts and fake news [are] becoming part of an everyday vocabulary” and in which environmental policy- making institutions “are now under attack from populist authoritarian discourse and policies” (Neimark et al., 2019, p. 2). I believe that a naked science (Price 2013), allows us to speak power to post-truth (Neimark et al., 2019). Such naked science means making my ontological and epistemological positions clear – how do I conceive of knowledge-making and its role s–, and making the processes of science clear – what is my agenda and positionality, as I expose in the next few paragraphs.

I first envisage the nature and construction of knowledge in social sciences as being performative, that is, it “contribute[s] towards enacting the realities that [it] describe[s]” (Callon, 2006, p. 7). There is thus no such thing as a ‘world out there’ independent of the way we represent it, rather there is a constant interaction between realities and the procedures for representing and reporting on them (Law, 2004; Miller, 2011). Such a stance is eminently political, in that different ways of knowing, which are always partial and situated, will produce different effects in the world. In a performative epistemology, practices and beings have a material reality but they do not speak conceptually for themselves. It is the ordering, the assembling, classifying processes, the ways in which we put things in relation and interaction that is performative.

Callon (2006, p. 13), “refer[s] to [scientific] statements and their worlds as socio-technical agencements”, a concept which is rooted in Actor-Network Theory (ANT)2 (e.g. Callon, 1999; Latour, 2005; Law, 1986; Law & Mol, 1995; Mol, 2002). Agencements combine heterogeneous human and nonhuman elements, material and technical devices, that have the “capacity of acting in different ways depending on their configuration” (Callon, 2006, p. 13). The term agencements is very close to the more common concepts of ‘arrangements’ and ‘assemblages’. However, contrary to these concepts which suggest a divide between human agents and things, agencements has the same root as ‘agency’ and thus stresses the idea that networks of elements come together and act as a whole (Callon, 2006). It means that processes of meaning-making about the world are constructed and performed not only by scientists but in

2 In this section, I use ANT for its epistemological and ontological underpinnings to construct my “meta” philosophical stance, rather than as a theory for the analysis of my empirical results. Yet in part I, I will also use ANT as a theoretical lens. 14

knowledge practices that include a “whole range of participants that extend far beyond people” (Law, 2004, p. 19; Miller, 2011), e.g. satellite imageries and maps as I analyse in Part I. Yet, ANT only attributes a weaker agency to nonhuman actants than to humans. Such nonhuman actants do not have agency outside the networks of relations they are embedded in. Moreover, as Callon (2006) insists, we should not understand this stance as an opposition between performativity and constativity/representation, or as if there was no material reality. Rather, the concept of performativity emphasizes that knowledge- making practices are not ‘outside’, representing social phenomena from another place, but that they are embedded in their objects (Mitchell, 2008). Knowledge statements and their objects are in a process of co-evolution (Callon, 2006).

Such a relational stance is present among a wide range of disciplines including anthropology and critical geography, and notably resonates with Haraway (1991)’s feminist material-semiotics, Barad (2003)’s agential realism, ‘Deleuzo-Guattarist’ rhizomatic relationality (Deleuze & Guattari, 1987) or Ingold’s (2000) phenomenological approach. Generally, these perspectives along the socio-materiality of actor- network theory are ‘onto-epistemologies’ (a term coined by Barad (2007)) that neither see everything as socially constructed nor everything as pre-existent. Put differently, it means that practices are constituted by both meaning and materialities; discourse shapes, inscribes the world but does not materially produce it (Barad, 2003). There is a dialectical relationship between discourse and materiality, i.e. they are “in tension with one another […] connected in a continual interplay […] and co-emergence” (Putnam, 2015, p. 707). Reality is both physical and meaningful; meaning cannot exist without material existence (Haraway, 1991). In this sense, “[p]erformativity focuses attention on the ongoing, dynamic, relational enactment of the world”, on how “[it] is constantly being made and reconfigured in material-discursive practices”, without giving primacy to discourse or materiality (Orlikowski & Scott, 2015, p. 700; Putnam, 2015). To paraphrase Callon (2006, 2009) these processes of explanation maintain both the existence of a reality that resists – i.e. they do not ‘create’ it – and the idea of a reality that is multiple, constructed, instituted – i.e. they ‘make it happen’. Because knowledge and perception are constituted relationally, they can only be partial and situated (Haraway, 1991; Ingold, 2000). In other words, and as I develop in the methodological section, the research process, through method agencement, necessarily brings specific features of reality into view while obscuring others.

In the critical landscape literature, similar onto-epistemologies challenge us to become ‘more-than- representational’ (Lorimer, 2005) that is a “shift in thinking towards conceiving of the world in practical and processual terms, or, in other words, as something that [i]s in a perpetual state of becoming” (Waterton, 2018, p.92) and in engaging with practices, affects and (nonhuman) things that intertwine with the production of meaning (Müller, 2015). Practices as the title of this thesis indicates, are central to such an approach. I am interested in how landscapes emerge in an array of practical encounters that

draw together humans and all kinds of nonhuman things. Discourse-knowledge itself is a specific kind of practice (Whatmore, 2006). The style of thinking I adopt in this thesis thus conceives of a messier and complex world and inevitably brings together heterogeneous elements and conflicting perspectives rather than a linear narrative of forested landscapes and their changes. Yet, it is not an apolitical, “uncritical pinboard of complexities [and] multiplicities” (Saldanha, 2003, p. 428) as some critics have accused scholars drawing on ANT and more-than-representational theories of doing (Lave, 2015; Merriman et al., 2008). Instead, throughout my work, which is committed to decolonial and feminist epistemologies and political ecology’s concerns, I draw explicit links to various historical structures of domination and inequalities such as (neo)colonialism, imperialism, capitalism or racism in which the co- production of the human and nonhuman is embedded (Lave, 2015), with the defined agenda of transforming an unjust world.

The perspective I adopt thus combines epistemological, ontological, political and ethical dimensions. It strives to simultaneously provide an account of situatedness for all knowledge claims and knowing subjects, a critical practice for recognizing the situatedness of our own interpretations, “and a no- nonsense commitment to faithful accounts of a ‘real’ world” (Haraway, 1991, p. 187). The issue then is no longer to seek disentanglement from the world but how to engage with it. There is no choice to be made “between the good of truth[s] and such other goods as politics, aesthetics, justice, romance, the spiritual, inspirational and the personal” (Law, 2004, p. 15) – there was, in fact, never such a separation, it was just concealed. Truth remains a good. As Law (2004, p. 148) decisively puts it:

“method [agencement] does not work on the basis of whim or volition. It needs to resonate in and through an extended and materially heterogeneous set of patterned relations if it is to manifest a reality and a presence that relates to that reality. […] Method [agencements] that do not produce presences that have to do with truths may be attractive, there may be other reasons for generating them, but whatever they are, they are not about the out-therenesses of possible realities.”

The conceptual frameworks I build in Part I and Part II both draw from this political (or performative) relational onto-epistemology to engage with ‘landscape’. Part I uses a relational (sometimes referred to as processual) approach to satellite imageries and maps to analyse their roles within REDD+ socio- material agencements in conceiving forested landscapes and the performative effects of representations in terms of nature-society relations and socio-environmental justice. Part II offers an alternative understanding of these same forested landscapes through a relational conceptualization of landscape as an emergent set of pastpresent heterogeneous, unfolding and intertwined human, nonhuman and material (unequal) relations and everyday, embodied practices.

As I explore in the next section, the set of methodological tools for this agenda does not remain limited to a definite number of methodological and data collection rules and processes or to a determinate socio- political phenomena (e.g. sex, class, gender) but engages with combinations, hybridizations to meet the struggles against domination while recognizing uncertainty and partiality – i.e. we cannot know everything deemed important (Law, 2004; Santos, 2018).

2. Methodological reflections for studying messy complexities

“Method is not, I have argued, a more or less successful set of procedures for reporting on a given reality. Rather it is performative. It helps to produce realities. It does not do so freely and at whim. There is a hinterland of realities, of manifest absences and Othernesses, resonances and patterns of one kind or another, already being enacted, and it cannot ignore these. At the same time, however, it is also creative. It re-works and re-bundles these and as it does so re-crafts realities and creates new versions of the world” (Law, 2004, p. 143).

I want to build on John Law (2004)’s After Method: Mess in social science research again to briefly look at how we might think about social research methods in line with a performative relational onto- epistemology. Grasping some of the workings of environmental rule and multiple socio-spatial practices that constitute landscapes inevitably requires us to deal with ‘mess’ and ‘complexity’. We need to recognize that “events and processes are not simply complex in the sense that they are technically difficult to grasp […]. Rather, they are also complex because they necessarily exceed our capacity to know them” (Law, 2004, p. 6). What does this imply?

First, this means that we should challenge the adoption of methods that produce simplicities about an inherently ‘messy’ world; there are no single nor simple answers (Law, 2004; Van Hecken, Bastiaensen, & Windey, 2015). Law and Mol (2002) exhorts us to practise a non-classificatory science that refuses to make a single simple order that expels complexity and openly recognizes its incompleteness. Rather, it is at the place where different simplifications and various modes of ordering (styles, logics) meet comfortably or in tension that complexity is created. This PhD thesis in this sense is made of ‘partial connections’ (Strathern, 1991) between disparate socio-political, economic and cultural configurations. My analytical and writing strategy articulates the disciplinary traditions of critical geography which, put simply, highlights geographies of domination or how the social production of space (re)produce power and inequalities between people and place, and anthropology which studies specific conjunctures and culturally endowed everyday material practices that shape landscapes in which the researcher becomes embedded.

Empirically, I have found the ethnographic-type of fieldwork approach, the open flexible research design (Silverman, 2014), the multi-method strategy (Mason, 2000), the focus on surprises (Tsing, 2005) and

attention to mundane details that all characterize it, useful to attend the messiness of practices. Fieldwork, as Mason (2000, p.87) states, involves the researcher in “observing, participating, interrogating, listening, communicating, as well as a range of other forms of being, doing and thinking”. While I have largely used ‘conventional’ qualitative and participatory research methods to collect data (see 3.2.), daily practices and affects that tend not to be listed in academic texts’ methodological sections have certainly shaped the research processes informing this thesis. I am thinking (non-exhaustively) of: forging some closer relationships with research participants and nonparticipants; despising some people for their views or behaviours, such as the owner of a logging concession who continuously complained, to the ‘white’ regular customers at his hotel-restaurant, about the hardships of doing business in the ‘Congolese chaos’; covering thousands of kilometres on dirt roads and trails and feeling exhausted (everyday); walking in forests and having the feeling of suffocating because of the heat and vegetation density; visiting colonial buildings and talking with people inhabiting them; sharing meals with participants in participatory exercises; getting frustrated (often) by things I perceived as cultural and/or power differences and doing some introspection to analyse my reactions; becoming aware (more than ever) of my whiteness. I come back to some of these embodied experiences in the next section on positionality/reflexivity. These are certainly not methods; that is not my point. My point is that field research is also an embodied practice from which our understanding of socio-ecological phenomena cannot be neatly disconnected. As Coffey (1999, p. 68) rightly says:

“[O]ur engagement with the field is both intellectual and physical. We cannot divorce our scholarly endeavours from the bodily reality of being in the field.”

Second, as I have mentioned before, in contrast with certain assumptions and (often hidden) normativities that are attached to research methods, we need to get rid of the idea that if we use the ‘right’ methods we can achieve certainty, that we can arrive once and for all at stable conclusions about the way things ‘really’ are, as well as of expectations for generality. In short, “we need to unmake our desire and expectation for security” (Law, 2004, p. 9). This goes in line with my ontological and epistemological stance outlined earlier: scientific knowledge can only provide a partial knowledge of the constituents of the socio-ecological phenomena. This PhD thesis should thus be seen as an open-ended account of the landscape of Kisangani’s hinterland. This leads to the third statement: since most methodological tools are engaged in simplifying and describing social realities they do not produce a neutral understanding but rather actively perform realities. Methods are always social, political, and are about what kind of social science we want to practice but also about what part of reality we prefer to represent and constitute. Methods necessarily produce presences and absences as well as arrangements

with political implications; they are creative; “they re-craft realities and [enact]3 new versions of the world” (Law, 2004, p. 143). In short, “methods are fully of the social world that they research; […] they are fully imbued with theoretical renderings of the social world” (Law, Ruppert, & Savage, 2011, p. 4). Thus, this means that I recognize that my own research and its methods, themselves enact, perform certain aspects of the social world while not performing others (Callon, 2006).

Furthermore, it also implies that myself, as a researcher choosing concepts and methods, and as part of a particular epistemic community –I will return to this in my positionality/reflexivity section– I contribute to the ‘performation’ of reality (Callon, 2006). This PhD thesis shows not only how different methods – and epistemologies – to understand the same landscape and its changes produce different understandings but also the socio-material geographies these understandings (may) produce when mobilized in environmental-development projects in specific areas.

This unavoidably leads us to questions of reflexivity where the researcher is himself/herself the object of reflection. Being reflexive is to recognize that our research and methods craft and re-craft realities and reflect on how and why we do this. I will come back to such important questions in next section.

Following this logic, if we assume that science and its methods are no longer about representing a definite social world, but rather that they should “manifest realities that are indefinite and that as part of this, it is important to appreciate […] that the indefinite [is] not necessarily a [sign] of methodological failure” (Law, 2004, p. 154), what does this imply in terms of the rigor and validity of such a research agenda? I argue that while these questions should remain an important concern for researchers, it also clearly appears that the findings of such research cannot be integrated into traditional formalistic validity criteria because they hinder creative and critical reflection (Jones & Murphy, 2011). I build on Patton (in Sumner & Tribe, 2008, p. 112)’s alternative, non-exhaustive lists of quality criteria for social constructivist research and for critical change as well on Chambers (2015)’s thoughts on rigour for complexity and Bradbury and Reason (2001)’s Broadening the Width of Validity, to suggest the following criteria that could be applied to the research agenda I engage with:

• subjectivity acknowledged and embraced; • reflexivity and praxis, i.e. understanding my own background and how I act in the world; • trustworthiness and authenticity, i.e. fairness and coverage of others’ perspectives; • eclectic methodological pluralism and improvisation, i.e. methods sensitive to context and engagement with extended ways of knowing

3 Law (2004) here uses the world “create”. I prefer to stick to the term “enact” – that Law actually uses interchangeably with “create” – which is more in line with the assumptions and terminologies of the performativity thesis and as I think the word “create” might give the wrong impression that I adopt a strongly constructivist ontological stance. Rather, as lengthy explained in the previous section, I believe that knowledge (or meaning-making) and reality are in a process of co-evolution. 19

• significance, i.e. inquiry well-grounded in everyday concerns and struggles of people • adaptive iteration, i.e. methods evolve throughout an inquiry • triangulation and plural perspectives, i.e. for crosschecking, comparing information but also for capturing multiple perspectives by different people and different institutions (between and within) • contributions to dialogue, i.e. encouraging multiple perspectives and in particular the perspectives of the less powerful; identify potential change • particularity, i.e. doing justice to unique cases but a case study can illuminate other situations; • conceptual-theoretical integrity, i.e. bringing more order to complex phenomena but anchoring it in people’s experience; reasonableness of the interpretation is approved by peers

In the next two sections, I first reflect on the role of my positionality and praxis in studying pastpresent socio-ecological relations in DR Congo and then discuss the methods that have informed this research. As we have understood by now, the approach I adopt does not prescribe any specific set of methods but rather great flexibility that articulates different data types derived from different methods.

3. Doing embodied research: positionality and reflexivity

Something… must be wrong somewhere, if the only way to understand our own creative involvement in the world is by first taking ourselves out of it (Ingold, 1995, p. 58)

In the previous sections, I have emphasized the situatedness of our knowledge (Haraway, 1991) which means that we need to take responsibility for the positionality (social and political location) that we actively articulate to construct meaning, make claims and interpretations. At the same time, feminist geographers have stressed that such a position is non-essentialized – it is not the locus of an already determined set of values – and emerges continuously from historical relational experiences and practices such as fieldwork – it is not one single fixed ordering point (Alcoff, 1988; Haraway, 1991). In this way, in reflecting on positionality, it is necessary to go beyond simple identifiers/identities but rather through an ensemble of particular life experiences and worldviews that have influenced the way I have conducted this research. Scrutinizing how our potential social positions may have influenced relationships with informants in the field and may narrow our view, “is a fine way to start a dialogue with readers” (Lichterman, 2017, p. 38, original emphasis). Yet, if we avoid a simple realist epistemology in our research practices, we should be equally modest about assuming that we know which part of our identity really influenced our interactions, what our informants really think about us, if and how a researcher impact their lifeways (Reyes, 2018; Robertson, 2002). A social position is not “a reality ‘out there’ that works mechanically on the [researcher’s] claimsmaking process” (Lichterman, 2017, p. 38). With these preliminary thoughts in mind, in this section I present a more personal account on some of the ways I

believe my positionality “emerge[d] and shifte[d] in the contiguous processes of doing and writing about” my research (Robertson, 2002, p.791).

I have already made my epistemological position clear, as well as my research focus on glocal human- nature connections that constitute Kisangani’s hinterland landscapes, and my commitment to continuously interrogate and question power relations in such constitution, from an explicit decolonial agenda for socio-environmental justice. This stance connects me to an epistemic community of similarly ‘engaged critical’ scholars (Shapiro-Garza, Kolinjivadi, Van Hecken, Windey, & Casolo, accepted for publication), the research participants in historically socially, economically and/or politically marginalized communities where I conducted fieldwork, and a global cosmopolitan community of actors who oppose imperialism, capitalism, racism and patriarchy in their practices (Katz, 1994).

Thinking outside my ‘academic self’, the choice of my research focus and field sites is undoubtedly tied to my personal background. I am a white, relatively young Belgian woman from an upper middle-class family with working-class origins. My mother was born in Lusanga (formerly Leverville) in the then Congo Belge* where my grandfather worked en brousse* for several Unilever oil palm concessions – a characteristic example of colonial agro-capitalism – during the last six years of colonial rule. The years he spent in DRC left a profound mark on him and my grandmother, and the stories of their experience filled the Wednesday afternoons I spent with them during my childhood, and the ‘Congolese memoirs’ he wrote during the last ten years of his life. My grandparents’ stories, and their ambivalent feeling about the impact of colonization – violent but the bearer of ‘development’ – have stayed with me and introduced me, albeit not in academic terms, to some of the logics of colonialism and contradictions of the workings of modernization that I expose throughout my analysis in this thesis. They also played a role in developing my profound interest in Central Africa and social injustices which was cultivated by my involvement as a representative for Oxfam and Amnesty International at my high school, and drove my academic orientation. I was first trained as an Anthropologist (Master) which initiated me in postmodernism/poststructuralism and fieldwork in Cameroon where I worked on community-based forestry. I have always largely identified as a (Far-)Leftist in European political standards and was engaged in alter-globalization movements for some years during my studies. Yet, my wanderings and own dissonance in combination with an anthropological interest for what (my fellow students and) I saw as the ‘enemy’ discipline and with a fear of not finding a job with my initial training led me to study a complementary Master in Business Economics and to work for four years as a Management Consultant which felt so often at odds with my personal values. How did the interweaving of my personal trajectory, disciplinary belonging, accountability towards research participants (i.e. let them speak for themselves), the expected research output (a PhD dissertation in Development Studies) play out in the field research process and relations, and in my interpretations?

Before I started my field research in DRC, despite what I would summarize as my decolonial awareness, I do not think I had ever fully – i.e. bodily, emotionally and mentally, rather than conceptually and intellectually only – realized the “abiding force of whiteness” (Hendriks, 2017, p.685) and privilege in “the world order” (Ferguson, 2006). During some volunteer work I did in the Eastern part of Uganda in 2007 and two months fieldwork in Southeast Cameroon in 2008 for my MA in Anthropology, I had obviously experienced my whiteness-as-otherness as I was constantly reminded of it by people calling out to me as mzungu or la Blanche and by just being noticed everywhere I would go. In the hinterland of Kisangani, my mere presence in a village would gather dozens of pairs of eyes, mostly children who had never seen a white woman but also many adults, fixedly staring at me while I was interviewing a villager. This often generated a feeling of oppression in me as I was regarded as the ‘different one’; something that as a white Belgian living in Belgium I never experience. Yet, in its association with the nationality of the former colonizers, my whiteness appeared more than ever before as otherness-as-a-hierarchical-position-in-the- world that postcolonial theorists have long described (e.g. Fanon, 1952/2008).

One of the first times I recall this explicitly struck me was during an informal focus group discussion I had with young people in a village in the Isangi Territory (some 150km from Kisangani), at the very beginning of my field research in 2016. We were discussing the difficulties these young people were facing in terms of land access and what they perceived as economic decline compared to their parents and grandparents’ generations. They then proceeded to make comparisons with Europe’s ‘development’, interpreting to a large extent the differences between DR Congo and Europe through a framework of ‘blackness’ versus ‘whiteness’ to make sense of their respective qualities. “God has blessed the white skin”, they said, “black skin is darkness”.4 Europe’s wealth in their discourses was also explained through a different kind of witchcraft than the one used by black Congolese people: a productive, compassionate ‘white magic’ as opposed to a destructive, individualistic ‘black magic’. My mere presence in their villages was a sign that development would come. One young person in the group said:

“I think one day Congo will be like Belgium, because the lady [me] came here, and since my birth I had never seen a white woman. But today we see whites coming to us, so things will evolve.”

I asked why would they think that whites only could bring ‘development’. They answered that they, Congolese, could also bring development through community-based systems of investments such as tontines, but that the kind of big development “that could allow [them] to buy a plane” (read: ‘modernity’) could only be brought by the whites. Inequality in such discourse is somehow naturalized.

Rubbers (2009), in his study of the relation between racial identity and post-colonial memory in Haut- Katanga, DRC, has noted similar stories. He classifies them as the story of the biblical curse on Africans

4 7-IFG-YALI-20160326. 22

that were spread by Catholic missionaries and underlined colonial propaganda, of the mastery over occult forces achieved by whites, and of the abandonment by Belgium which goes together with demands for Belgium to support them in present day. In this way, Congolese often refer to the Belgians as their “uncles”, a metaphorical kinship, that emphasizes a common trajectory. These are the kind of discourses I encountered many times in my conversations both with rural people and with Kisangani’s city dwellers, and very often arose spontaneously outside the formal interview or setting, for instance in casual conversations with my research assistant or with Congolese researchers I exchanged with. My interview guide was also not specifically focused on post-colonial memory except for my last fieldwork during which I conducted interviews on geographies of nostalgia. However, such stories and constant references to the colonial past spontaneously came up in interviews and with time, influenced my interests, follow up questions and analysis as I explain below.

I have recounted other anecdotes such as someone telling me in a village that it is God who sent me to them; another one asserting that “Congolese it’s the absolute mediocrity”5 as compared to whites; an academic assistant at the University of Kisangani arguing that the Belgians should not have left Congo so fast and that the South African model “where the whites stayed” would have been better for the country; or my research assistant telling me that he is held in higher esteem by his circle of acquaintances now that he is working with a white. In the same conversation these stories were, yet, sometimes also articulated with the violent abuses of colonialism and the current evilness of the West in constantly destabilizing the DRC. This fourth story articulated around the White-Belgian/Black-Congolese binary is described by Rubbers (2009, p. 276) as a “theory that recognizes the invisible hand of imperialism behind the hardship of life” emerged more often with more educated people and people in a more powerful position, such a politico-administrative function. In Kinshasa, in particular, the story came up in some conversations with some government actors. Here, my presence triggered some of them to emphasize that Belgians and whites, more generally, are encroaching on DRC’s sovereignty – one told me that I should tell my fellow countrymen at Enabel and other organizations that they should get out of here – and are looting Congolese natural wealth at the expenses of its people.

My aim here is not to provide an analysis and detailed explanations of such stories as this could probably be the topic of a thesis in itself. It is rather to concentrate on the ways the ‘white-Belgian’/’Black- Congolese’ binary and my whiteness within this binary affected my relations ‘in the field’ and the research process and analysis overall.

5 95-I-BEN-20190514. 23

Navigating relations and emotions

The imaginaries associated with whiteness and the colonially-entrenched inferiority complex (Fanon, 1952/2008) that was coming out of these stories made me feel disconcerted and uncomfortable as I was confronted with the fact that I was entering deeply unequal relations. I was not merely in a position of difference but in a position of hierarchy; the very one I went to the field being critical of – and continued to be throughout. With my informants and other Congolese acquaintances who had explicitly talked about the power of whiteness, I usually shared my visions of colonialization, how I felt ashamed about it and I tried to deconstruct the white/black binary construction with them. Often, this generated some laughter or embarrassed giggles as, my research assistant explained to me, this was not the position a white is expected to take or usually takes in their brief encounters with villagers. I also usually confessed in these conversations that my mother was born in the then Congo Belge* during the time my grandparents spent there. It was, according to my interpretation, always seen positively and as if we somehow shared, as individuals, some common history. Yet, it happened more than once that villagers told me I was coming back to find the treasure (usually diamonds) left by my grandparents in Congo even when I did not mention my grandparents’ presence in Congo. Such rumours about White Belgians are apparently common, even for those whose families have no direct link with the colonial project. It is thus very possible that informants might have hidden or downplay some information, for example about the presence of diamond quarries.

I often felt that my whiteness – probably together with my female gender – facilitated access to the field partly because of the above power relations but also because of the curiosity it aroused. I generally felt sincerely welcome in people’s yards and houses, and people were by and large very willing to talk openly with me. The eagerness for many villagers to come and greet me, to offer me their best chair to sit on and just to share some of their time with me, showed a sincere generosity. It made me feel very grateful but at the same time uncomfortable as it made me very conscious of the deference that people showed sometimes towards me. I often refused the best chair, especially when it meant that someone else had to stand up or sit somewhere else less comfortable. As Sultana (2007, p. 379) has similarly argued, I believe that “[s]uch little actions, however mundane, are not insignificant […], and speak to the embodied situatedness of me as the researcher that I had to constantly keep in mind”. On the other hand, I feel that despite being a woman, my whiteness did not facilitate my relations with rural women. In DRC’s rural world, women experience multiple discriminations, one of them being their exclusion from the political sphere. Therefore, many of them felt uncomfortable talking to me as they considered such an exchange with a white researcher to be outside their domain. I did have fruitful exchanges with some women but many others declined an interview with me, or if they agreed to talk to me, they would feel noticeably uncomfortable and provide particularly short answers. There are of course other reasons why

it remains more challenging to conduct interviews with women in Congo’s rural areas. One, for instance, is that during the day, women are carrying out various livelihood activities (fields, preparing food, etc.) and are thus rarely available for an interview. Getting women to talk to me appeared sometimes even more complicated as my assistant was a Congolese male, and the fear of provoking a husband’s jealousy was real –both from the interviewee’s and my assistant’s sides.

With some interviewees, my ‘white’ presence was explicitly associated with the expectations that I could lobby for the village for the recognition of their land, to be included in one or the other rural development project (e.g. REDD+), or directly offer them some money. In the latter case, it was usually not my interviewees who asked for money or compensation but rather other villagers who were part of the small audience that generally gathered when I would conduct an interview in their village. I suspect that the stories I mentioned above need to be partly understood in this context – I have also explained above that they should not be understood only in relation to my white researcher’s position though. The expectations associated with my presence in the field, to a certain extent, show a researcher-researched relation underlined by “a norm of asymmetric reciprocity, which demands that whites support the poor, and that Congolese people adopt an attitude of submission and circumspection” (Rubbers, 2009, p. 272). Yet, my Congolese research assistant who regularly carries out field research for various organizations, told me that such expectations were quite similar for better-off Congolese city dwellers going to rural areas to conduct research. I have always tried to manage these expectations by introducing myself as a PhD student and being very explicit at the beginning of the interview about my identity, research goals, and that I had little to offer except for a faithful account of their stories and struggles that, I hoped, could contribute to make their voices heard in forest policies in the future. I always gave my contact number in Belgium and DR Congo to my interlocutors who asked to keep in touch.

Sometimes, some people became very insistent in getting money. In these situations, in spite of myself, I felt frustrated and annoyed to be perceived as a ‘living wallet’. I regularly had this feeling outside the direct context of my research, for instance, when I was randomly arrested by policemen to pay for a supposedly missing driving document for the motorbike I rented or, at one or other roadblock, by a civil servant collecting cash, usually for a higher-ranked official. At other times though, when I was for instance less exhausted by the field, I would take this exact same situation with a smile and engage in a playful conversation with the ‘racketeer’.

Navigating the range of direct emotions I felt during my interactions in the field definitely felt like an entire part of the research process. In the moment, I caught myself reproducing classical culturalist explanations of such behaviours, that I am so critical of in my work. With distance from the direct interaction though, my epistemic and theoretical worldview largely oriented my interpretations of these

encounters through the lens of the logics of violent (post)colonial hierarchies and deep inequality in the world order (Ferguson, 2006) that directly influenced my analysis – see below.

While I have no illusion that relations with my informants were ever fully equal, I also do not believe that my relationships were only defined by unequal power play (Robertson, 2002). The way I interacted with rural people with an informal attitude, empathy and careful listening, also letting them ask me more personal questions or questions on ‘what does Belgium look like’ allowed me to form some relations of trust. I really enjoyed very much conducting interviews and exchanges with rural people. Although I had created a quite extensive interview guide, in order to create a good interaction, I was very flexible in the way I conducted interviews rather than following the guide strictly (see point 4). I was always open to topics that would come up during the interviews, even if I would not always follow up on these topics if they were unrelated to my research theme. I believe all these elements improved the quality of the interview, the power balance between me and the interviewee but also opened some real space for a co- construction of knowledge, in which my informants were (albeit mostly implicitly) empowered to influence the course of my research and my analysis. In the introduction of Part II for instance, I explain clearly that at first I had not specifically planned to explore a land conflict among various communities and with a logging company, nor the zone (the ‘Alibuku landscape) where this conflict was unfolding. In the same way, it is because informants were denouncing discrepancies between colonial maps of their territories and more recent maps – albeit with many contradictions in their discourses (see chapter 5) – that I started to be interested in maps as instruments of environmental rule and the relational power of maps (Part I).

I also created some more sustained relationships with regular informants whom I regularly visited, not for interviews but just to keep in touch. These informants talked French fluently, unlike most of my rural interviewees, and so it greatly facilitated more direct interaction and connection. I often stopped by to say hello on my way to other villages where I had planned interviews. Conversations with them became much more informal with time and, I felt these relations were a bit more equal and based on a mutual interest in talking together. I also often discussed my interpretations with these regular informants – as much as with my research assistant and other Congolese researchers, see point 4 – and sought clarification if some things that came up during my interviews were unclear or fuzzy. Generally, as an outsider I offered people a space to vent and talk about themselves, their communities and their everyday practices. Many told me that it was so rare to have (what they perceived as) more empowered people like me, truly listening to their struggles. In my analysis, although partial, I was always committed to remaining faithful to the relations I forged in the field and to the stories my informants shared. I am also very conscious that the huge differences between our ‘worlds’ undoubtedly left some gaps in my capacity to fully grasp the lived experiences of my rural informants. I believe, however, that assiduous

interviews, carefully collected data and the triangulation of data I strived to achieve allowed me to generate “resonant, thick descriptions and subtly evocative interpretations of people's lives in [some of] their messy complexities” (Robertson, 2002, p. 788). Yet, it is clear that in contrast with my privileged position, I often felt very powerless and conscious about my lack of direct ‘giving back’. Ansoms (2013) reminds us that when we leave the field, the communities we researched have had an instrumental impact on our lives (i.e. obtaining a PhD thesis that can lead to a future academic career) while they are left in the same social conditions and daily struggles we met them in. Through a VLIR-UOS funded action- research project with the University of Kisangani and the NGO Tropenbos-DRC that I co-promote, I hope to ‘give back’ in some ways by presenting my research to environmental NGOs and civil society groups advocating for local communities’ inclusion in forest policies, and engaging in more long-term engagement with the people and sites I have conducted research with.

My field research was not only limited to my interactions with rural people in the hinterland of Kisangani. For my ‘expert’ and ‘institutional’ interviews, I found it generally easy to contact people and organize a meeting. I do not recall any interview request being declined. People were extremely generous with their time during the interview process itself. Whether through a phone call or by email, I presented myself as a Belgian PhD student working on the interface between REDD+ and (g)local human-environment relations. During the exchange itself, I was careful not to share very openly my critical views of REDD+ that were growing with time. I wanted indeed to keep the access open. To what extent did my positionality as a white female student play a role in giving me access to experts, programme managers and government officials? As I said before, it can be difficult to assess. I could clearly see a difference between Kinshasa and Kisangani in that endeavour. In Kisangani, there are not so many NGOs and expatriates as compared to Kinshasa or Goma, and even less so white women. Congolese Provincial government representatives (e.g. Provincial Minister of Environment) or REDD+ Coordinators as much as the whites living there were usually intrigued by my presence, as a ‘lonely’ researcher (read: not attached to any project) in what is considered to be a remote place. I could turn up in person at their office to request an interview and they would generally meet me immediately.

Kisangani is small and I moved around the city in moto-taxis, which made me very noticeable. In the city, I quickly got to know most of the ‘whites’ living here, from EU/CIFOR programme’s implementers to the owners of logging concessions. There are about three restaurants where whites go to have dinner. Informal conversations with them clearly informed my research. When I was doing research in the villages in the Lubuya-Sector – thus closer to Kisangani than Isangi and Yangambi (see below), I was in fact staying in Kisangani at the case de passage* of the EU/CIFOR FORESTS project where I interacted a lot with the people working for the project. Although we are not in regular contact when back in Belgium, I spent many evenings talking with some of the project’s members – if they were in Kisangani; I was also often

alone. Entering that house – although it was decrepit and frankly dirty with rats running around in the kitchen at night – often felt like a relief after some tiring days of fieldwork in the villages. I could take off my researcher’s hat and finally become unnoticed. I did not plan to stay at the case de passage at first. When I started my fieldwork I was staying at a hotel and was navigating my way in Kisangani without any institutional ties. One night, I met by coincidence the manager of the EU/CIFOR programme. He too was astonished to see a young white woman going around the hinterland on a motorbike. So, he suggested that during my next stay in Kisangani, I stay at the shared house for a small amount of money per night. I ended up always staying there during my many stays in Kisangani. Navigating between my researcher’s hat and my non-researcher’s hat was sometimes a challenge and posed some ethical questions. I often felt that I was doing covert ‘participant observation’ among the programme’s members although that was never my aim. Yet, I could not simply ignore important contextual information that I was getting on the functioning of such developmental-environmental programmes. I was, however, more openly critical of some of the programme’s underlying narratives than during formal interviews, precisely because of the informality of our interactions. In the thesis, I have only used one quote coming out of an informal conversation during that ‘participant observation’ although the latter has certainly had a much more indirect impact in my analysis.

Interpreting the data

I have already given clues in the previous paragraphs about the ways my positionality and relations in the field impacted the research process. At the beginning of this section, I mentioned that I was initiated in critical and postmodern thinking probably during through masters in Anthropology. I also emphasized my commitment to feminist and postcolonial epistemologies of enacting multiple ways of being and knowing. This has stayed with me but my fieldwork made it evolve. The repetition of ‘stories in Black and White’, nostalgic references to colonial modernity, as much as the ways racialized hierarchical relations took shape in the field despite my hope to avoid such reproduction, made the blurring of binaries, of the ‘us’ and ‘them’ in my analysis difficult. This made me pay particular attention to the sticky rubble left by colonialism and imperial modernization on landscapes, minds and imaginaries that I show throughout this thesis.

However, how should I respond when I encounter the seeming reproduction of a monocultural perspective, instead of the alternatives and resistance to modernization discourses I expected? As a white Belgian scholar, how should I engage with the aspirations of local collaborators which reproduce ideas of Western economic and ecological modernization while also striving to co-produce alternative ways of thinking and being, attempting to undo entrenched relationships of violent colonial appropriation of land and life? The narratives that people repeated to me expressed a real sense of exclusion from the promised modernity, perceived as an ideal to attain, and an appeal to membership in global society (see

also Ferguson 2006). They represented material, emotional and bodily experiences and needs with consequences on resource creation and use, even if they contradicted postcolonial critiques of (eco)modernization and alternative worlds. In encountering these views, “engagement” meant needing to embrace the “disconcertment” (Law and Lin 2010) I felt as my critical assumptions were shaken, and to adopt a “hyper self-reflexive practice” regarding my personal and institutional positioning. My encounters and collaborations in DRC have pushed me to reflect on how socio-economic privilege might work to permit and enable a “strong” critique of (eco)modernization, as well as on my own complicity within a system I critique (Kapoor 2004). Engaging through praxis with divergent, or even contradictory understandings, rather than sticking to the tidiness of one theoretical frame, requires what Mignolo (2009) refers to as “epistemic disobedience.” In this thesis, this entails engaging with thick description in order to avoid essentializing the struggles of my informants and providing a better account of how such struggles are articulated in terms of distribution and inclusion in the system, not only in terms of alternatives or resistance to it. In Part II of this thesis, I use story-telling as a political method to challenge hegemonic eco-ethnogovernmentality by multiplying the connections between different periods of time, place, relations and practices.

Finally, I also see this PhD thesis as my journey through the reflections I have formed over time when ‘in the field’. The very format of academic writing does not truly allow me to show such a process of reflections, hesitations, constant questionings and dissonance sometimes. In many cases I have found postmodern and postructuralist theories – and some of their inevitable binaries – useful to my analysis but also limited. I engage as much with posthumanist thinkers that opened new ways of thinking about my empirical data. Such evolutions and navigations between different schools of thought are not always apparent in the final product – the thesis – of a research process. There are certainly other theories that could have been as useful to interpret my results, and which might highlight other aspects of the data I gathered in DRC. Again, these are choices I have made, both consciously and unconsciously. I thus insist that I do not consider this thesis to be the end of the research; as in any research, I myself produce ‘absences’. Questions and uncertainties are still lingering, as I highlight in my general conclusions. I hope that this manuscript will convey to the reader the research journey I went through.

4. Overview of methods and field sites

For this PhD project, I carried out a total of nearly 7 months of field research between May 2015 and June 2019 in Kisangani’s hinterland and in Kinshasa (DRC), combined with a desk study of REDD+ policy documents and relevant colonial archives. The research was organized in 5 different field trips. Each time, I usually spent a week in Kinshasa –which used to be a necessary stop to go from Brussels to Kisangani before Ethiopian Airlines started to organize flights through Addis Ababa– to carry out ‘institutional

interviews’. The rest of each field trip –the majority, thus– was dedicated to interviews with various actors, and in particular peasant farmers, in the peri-urban area of Kisangani and in the city itself, as I detail below.

Expert/institutional interviews and desk study

In order to understand the knowledge agencements that make sense of deforestation and forest degradation in the DRC and that inform REDD+ environmental policies and strategies in DRC as well as forest governance, I conducted a total of 85 semi-structured expert/institutional interviews and conversations which mostly took place in Kinshasa – where most of REDD+ is in fact happening at the moment (see below) – and in Kisangani; and seven of them through Skype. The interviewees included international and national forest experts, technical advisers and consultants from REDD+ and conservation projects’ lead implementing organizations, with government agencies and state actors as well as with representatives of international and national environmental NGOs (ENGOS), Civil Society Organizations (CSOs) plus private actors – some of the latter engaged in REDD+ discussions. I was particularly attentive in getting a range of interviewees that encompasses different, sometimes diverging perspectives across DRC’s REDD+ policy domain. This range represents 34 different organizations directly connected to the REDD+ process itself (at different degrees), showing how the latter is governed in a typical ‘neoliberal governmentality’ fashion (Foucault, 1991, 2008). As I became increasingly interested in the roles of maps in producing particular discourses on deforestation, I held interviews specifically with DRC’s mapping agency’s directors at national and provincial levels, and with GIS specialists working on DRC’s forest monitoring, and land and country planning. In addition, I also participated as a participant observer, in a three-day round table of civil society organizations on forestry reforms and REDD+ in DRC, in Kinshasa. Such participant observation allowed me to directly delve into the conversations, debates and contestations around these processes. Other organizations represented in these interviews are all stakeholders of land and forest governance, at a more general level than purely REDD+, which again allowed the diversification of perspectives.

I identified people to interview by checking REDD+ policy documents as well as through the snowball method, with interviewees identifying additional people that could be interesting for me to meet. In Kisangani, my research assistant (see below) also helped me to identify institutions that were interesting in regard to the topics coming out of the research with local communities (see below). As I mentioned earlier, generally, access to ‘elites’ was fairly easy – both with international and national ‘experts’ and with State organizations – and I always felt welcomed. I presented myself as a PhD student from Belgium who is looking at the implementation of the REDD+ process and at land and forest use practices in Kisangani’s hinterland. These extensive exchanges with practitioners have certainly influenced the focus of my research by giving me clues of some of the important issues related to REDD+. Maps, for instance, 30

became an important topic of my work not only because they were fuelling conflicts in the areas I worked in but also because geospatial planning and the need for accurate maps came up so many times in interviews with REDD+ ‘experts’.

It is important to note here that at the beginning, I did not plan to conduct so many interviews in Kinshasa. Yet, as I note in Part I – chapter 2, it quickly appeared that the planification and ‘strategical thinking’ was happening in the capital, in a very centralized manner. Many of the REDD+ practitioners I met in Kinshasa, in particular the international ones, had never been to Kisangani – or just for three-day trips – although they were talking about programmes taking place in the city’s hinterland. In some cases, practitioners somehow became the interviewer and asked me many questions about what was ‘really’ happening in the field, with project X, or e.g. in a protected area targeted by a programme. In the context of some arguments I will make later in this thesis, it is interesting to note that when I announced to some interviewees that I was doing fieldwork in Kisangani’s hinterland, they were often surprised, as Kisangani even if it is a city, is largely seen as l’intérieur* – literally ‘the inside’ which connotes remoteness and backwardness in many Kinshasans’ minds (Petit & Trefon, 2006). DRC’s political economy is dominated by Kinshasa even if its wealth arises ‘from the bush’.

For the desk study, reports and accounts on REDD+ (from international organizations, government agencies, NGOs, consultancies, advocacy groups, news reports) were either directly accessed on the internet or obtained from organizations’ representatives when documents were less accessible or unshared/private. As in some interviews references to agricultural practices and policies during the Belgian colonial period were made, I also consulted some colonial archives dealing with (indigenous) agriculture and rural economy in the Belgian Congo to compare it with present day REDD+ discourses on deforestation and agriculture. Such archives were also particularly useful in understanding the ways in which colonial ‘indigenous policies’ still shape everyday land tenure and use practices. The documents were collected at the Library of the Africa Museum in Tervuren (Belgium) and on online archives databases that include digitalized historical documents (e.g. https://archives.africamuseum.be/, http://arch.arch.be).

Fieldwork in Kisangani’s hinterland

I also undertook some qualitative fieldwork in the hinterland of Kisangani, i.e. within a radius of 150km around the city that is the city’s supply area. Kisangani is located at the heart of the Congolese rainforest, unlike other large cities such as Kinshasa or Goma that are at the forest edge. Kisangani is also unique for its location on the Congo River which makes it an important commercial and transport hub. What brought me first to the hinterland of Kisangani was its inclusion in several REDD+ programmes. Following studies on deforestation drivers (Part I), the supply area of the city was identified as a deforestation ‘hotspot’

mainly induced, as REDD+ studies’ synthesis states, by shifting cultivation, ‘uncontrolled’ artisanal activities and population growth. It hence was chosen as the intervention zone of three different REDD+ programs financed by multilateral aid and one private project, all situated in Kisangani’s hinterland. At the time of the study, the REDD+ geographically integrated pilot project in Isangi (PPRGII) was ending (2016) and another private REDD+ initiative (JADORA) was and is still ongoing. The two other REDD+ integrated projects were (slowly) starting although activities not yet concretely implemented. While I spent several weeks interviewing inhabitants of the villages targeted by PPRGII and JADORA projects, I have explored many different places in the hinterland, some close to the city, some much further away (up to 150km away). I was always guided by the topics that came up in interviews. As I explain in the introduction of Part II, I spent a lot of time researching what I call the ‘Alibuku landscape’ in the perimeter of a logging concession whose entrance is situated about 40 km from Kisangani. I became particularly interested in this area as it was the site of a fierce, long-lasting multidimensional conflict between several groups, generated by resource development and fuelled by maps, and by historical socio-spatial relations. In fact, my interest in Kisangani’s hinterland never stopped growing over the years. Over time, I became more and more interested in the different temporal layers that have shaped this landscape and everyday practices. I visited many former colonial plantations and buildings in which heterogeneous unruly forms of lives and human-environment relations have emerged spontaneously. Kisangani hinterland’s landscape in general offers a microcosm to explore landscape as a relational-spatial-temporal process that defies conceptions of abstract, containerized space that are embedded in REDD+ approaches.

The field research includes a total of 111 semi-structured (mainly), unstructured, in-depth and informal interviews, and 10 focus group discussions, all aimed at understanding local processes and meanings of land tenure and resource use, socio-spatial identities, and local perceptions of environmental change. In villages targeted by PPRGII and JADORA projects, interviews also dug into the perceptions of REDD+. Conscious that my position as a Western researcher might influence the respondents’ answers to these sets of issues – e.g. respondents’ desire to provide the ‘right answer’ on environmental change – in the targeted village, I interviewed communities and households included in REDD+ projects as much as those who were not. Similarly, I triangulated my methods by not relying on formal semi-structured interviews alone –see above and below. I have also tried as much as possible to obtain plural perspectives by interviewing people from different social groups (‘autochthones’, ‘newcomers’, women, young, elderly people, notables*, city dwellers, etc.). Yet, it was particularly difficult to conduct interviews with women for multiple reasons, as I already explained in the previous section.

In order to deepen my understanding of socio-spatial relations, I carried out 6 low-tech participatory mapping exercises combined with focus group discussions. Far from using maps from a positivist/realist epistemological stance that emphasizes the ontological security of maps, I envisaged these mappings as

a contingent, relational discursive practice that can produce different ontologies. Maps in this view are “merely a reflection of land use [and tenure] at a particular time under a particular set of circumstances” (Roth, 2009, p. 122) and a product of specific negotiations. In fact, when presented with a topographic map at 1: 200 000 scale produced in 2011 by the Catholic University of Louvain, that includes rivers, main infrastructures, protected areas, cities and some village’ names, communities asserted that it contained omissions and mistakes; namely that lots of smaller rivers and roads were missing and that the position of some villages was wrong as were some village’ and river’ names. During participatory mapping sessions, positions of things were the object of lively debates among communities. These spaces of exchange were incredibly rich for understanding the struggles around land and the multiplicity of tenure relations that characterize the hinterland’s landscape, as well as complex relations to maps and the colonial past. Such participatory exercises/focus groups always lasted for 4-5 hours, and after the group would progressively dissolve, I would often have separate informal conversations with some of the participants.

In addition to the participatory mappings, I asked 3 land users to draw individual sketch maps – hand- drawn cognitive maps – of their land use spaces. Although (counter-)maps were seen by communities as a powerful way to claim land rights, both low-tech mapping exercises revealed their limitations in understanding a complex and dynamic tempo-spatiality. Indeed, participants seemed to not always feel comfortable with the act of mapping partly because it did not correspond to their ways of knowing and dwelling in space. Inspired by Roth (2009), I thus complemented the counter-maps with 4 transect walks through different land-use areas, allowing me to highlight movement and flexibility. Through the bodily experience of fields, fallows, and forests, I came to understand the logics of a space that is not separated between the wild and the non-wild, nor in zones of use. In my interviews, and through individual sketch maps, I had tried to get clear-cut answers on where each type of activities is taking place or on how land is divided between families, it is through walks in village’s land use areas that I got to better understand their spatiality.

For my field visits to villages, I worked with the help of a Kisangani-based research assistant who mainly acted as a translator from Lingala/Swahili (and sometimes local languages) to French – i.e. I conducted all the interviews myself – but who also helped me to decode some socio-cultural practices and meanings. Everyday conversations with him provided invaluable contextual information.

It is also important to note that during my first fieldwork in 2015, I was accompanied for 4 days by one of the PPRGII REDD+ project managers (OCEAN NGO). I was very interested in collaborating with a local organization. Yet, I quickly felt monitored and realized that it influenced the answers given by my respondents. The results of the project were in fact nowhere to be seen despite a success story told in the official evaluation reports. I thus decided to work independently for the rest of my field research that

year and over the next four ones, and I specifically chose an assistant who had no personal links with, nor research experience in the villages I investigated. I have, however, always remained in close contact with the PPRGII managers and would always meet them formally and informally every time I was in Kisangani. In fact, I discussed many times about the issues I was investigating with representatives from OCEAN NGO and Tropenbos-DRC, the two main environmental NGOs in Kisangani. These discussions gave me great insights into their work with communities, their relationships with international donors, and generally over the main issues I researched. Since 2019, I have in fact been collaborating with Tropenbos-DRC along with Université de Kisangani on an action-research project on the political ecology of forest resource management.

As in any field research, (participant) observation, informal conversations and daily life in Kisangani have all contributed to enrich my data analysis. Over the years, I have built many relationships with Congolese and foreigners working for a range of different organizations. I have spent much time just sitting and talking in villages, with priests and nuns at the Christian missions I often stayed at.

In the manuscript, quotes from respondents are anonymized as much as possible through a coding system available in Appendix 1.

References

Alcoff, L. (1988). Cultural Feminism versus Post-Structuralism: The Identity Crisis in Feminist Theory. Signs: Journal of Women in Culture and Society, 13(3), 405-436.

Alcoff, L. M. (2007). Mignolo’s Epistemology of Colonialit. CR: The New Centennial Review, 7(3), 79-101.

Ansoms, A. (2013). Dislodging power structures in rural Rwanda: from 'disaster tourist' to 'transfer gate'. In Thomson, S., Ansoms,A. and Murison, J. (Eds.), Emotional and ethical challenges for field research in Africa: the story behind the findings (pp. 42-56). Basingstoke: Plagrave Macmillan.

Barad, K. (2003). Posthumanist Performativity: Toward an Understanding of How Matter Comes to Matter. Signs: Journal of Women in Culture and Society, 28(3), 801-831. doi:10.1086/345321

Barad, K. (2007). Meeting the universe halfway: Quantum physics and the entanglement of matter and meaning. Durham, N.C.: Duke University Press.

Bradbury, H., & Reason, P. (2001). Conclusion: Broadening the Bandwidth of Validity: Issues and Choice- points for Improving the Quality of Action Research. In Reason, P. and Bradbury, H. (Eds.), The handbook of action research: Participative inquiry and practice. London: SAGE Publications.

Callon, M. (1999). Actor-Network Theory – The Market Test. In Hassard, J. and Law, J. (Eds.), Actor- Network Theory and After. Oxford: Blackwell Publishers.

Callon, M. (2006). What does it mean to say that economics is performative? In CSI Working Paper Series (Vol. 005). Paris: Centre de Sociologie de l’Innovation. Retrieved from https://halshs.archives- ouvertes.fr/halshs-00091596/document

Callon, M. (2009). Elaborating the notion of performativity. Le Libellio d’Aegis, 5(1), 18-29.

Carolan, M. S. (2004). Ontological Politics: Mapping a Complex Environmental Problem. Environmental Values, 13, 497-522.

Chambers, R. (2015). Inclusive rigour for complexity. Journal of Development Effectiveness, 7(3), 327-335. doi:10.1080/19439342.2015.1068356

Coffey, A. (1999). The Ethnographic Self: Fieldwork and the Representation of Identity London: SAGE Publications.

Deleuze, G., & Guattari, F. (1987). A Thousand Plateaus: Capitalism and Schizophrenia. Minneapolis: University of Minnesota Press.

Escobar, A. (1999). Steps to an Antiessentialist Political Ecology After Nature. Current Anthropology, 40(1), 1-30. doi:10.1086/515799

Fanon, F. (1968). The Wretched of the Earth. New York: Grove Press.

Fanon, F. (2008). Black skin, white masks. London: Pluto Press. (Originally published 1952)

Ferguson, J. (2006). Global Shadows: Africa in the Neoliberal World Order. Durham and London: Duke University Press.

Forsyth, T. (2001). Critical realism and political ecology. In A. a. L. Stainer, G. (Ed.), After postmodernism: critical realism? (pp. 146-154). London: Athlone Press.

Forsyth, T., & Walker, A. (2008). Forest Guardians, Forest Destroyers: The Politics of Environmental Knowledge in Nother Thailand. Seattle and London: University of Washington Press.

Foucault, M. (1991). Governmentality. In G. Burchell, Gordon, Colin and Miller, Peter (Ed.), The Foucault Effect: Studies in Governmentality (pp. 87-104). Chicago: University of Chicago Press.

Foucault, M. (2008). The Birth of Biopolitics: Lectures at the Collège de France, 1978-1979. New York: Palgrave Macmillan.

Gibson-Graham, J. K. (2006). A Postcapitalist Politics. Minneapolis: University of Minnesota Press.

Gibson-Graham, J. K. (2008). Diverse economies: performative practices for `other worlds'. Progress in Human Geography, 32(5), 613-632.

Gibson-Graham, J. K. (2014). Rethinking the Economy with Thick Description and Weak Theory. Current Anthropology, 55(S9), S147-S153. doi:10.1086/676646

Haraway, D. (1991). Simians, Cyborgs, and Women: The Reinvention of Nature. New York: Routledge.

Haraway, D. (2003). The Companion Species Manifesto: Dogs, People, and Significant Otherness (Vol. 1). Chicago: Prickly Paradigm Press.

Hendriks, T. (2017). A darker shade of white: expat self-making in a Congolese rainforest enclave. Africa, 87(4), 683-701. doi:10.1017/S0001972017000316

Ingold, T. (1995). Building, dwelling, living. How animals and people make themselves at home in the world. In M. Strathern (Ed.), Shifting contexts. Transformations in anthropological knowledge (pp. 57-80). London: Routledge.

Ingold, T. (2000). The Perception of the Environment: Essays on livelihood, dwelling and skill. New York: Routledge.

Katz, C. (1994). Playing the Field: Questions of Fieldwork in Geography. The Professional Geographer, 46(1), 67-72. doi:10.1111/j.0033-0124.1994.00067.

Latour, B. (1993). We Have Never Been Modern. Cambridge, MA: Harvard University Press.

Latour, B. (2005). Reassembling the Social. An Introduction to Actor-Network-Theory. Oxford and New York: Oxford University Press.

Law, J. (Ed.) (1986). Power, Action, and Belief: A New Sociology of Knowledge? London: Routledge and Kegan Paul.

Law, J. (2004). After Method. Mess in social science research. Abingdon: Routledge.

Law, J., & Mol, A. (1995). Notes on Materiality and Sociality. The Sociological Review, 43(2), 274-294. doi:10.1111/j.1467-954X.1995.tb00604.x

Law, J., & Mol, A. (2002). Complexities: An Introduction. In J. Law & A. Mol (Eds.), Complexities: Social Studies of Knowledge Practices (pp. 1-22). Durham, NC: Duke University Press.

Leach, M., & Scoones, I. (Eds.). (2015). Carbon Conflicts and Forest Landscapes in Africa. Oxon: Routledge.

Lichterman, P. (2017). Interpretive reflexivity in ethnography. Ethnography, 18(1), 35-45. doi:10.1177/1466138115592418

Lorimer, H. (2005). Cultural geography: the busyness of being `more-than-representational'. Progress in Human Geography, 29(1), 83-94. doi:10.1191/0309132505ph531pr

Mason, J. (2000). Qualitative Researching London: SAGE Publications.

Merriman, P., Revill, G., Cresswell, T., Lorimer, H., Matless, D., Rose, G., & Wylie, J. (2008). Landscape, mobility, practice. Social & Cultural Geography, 9(2), 191-212. doi:10.1080/14649360701856136

Mignolo, W. D. (2009). Epistemic Disobedience, Independent Thought and Decolonial Freedom. Theory, Culture & Society, 26(7-8), 159-181.

Mitchell, T. (2008). Rethinking economy. Geoforum, 39(3), 1116-1121. doi:http://dx.doi.org/10.1016/j.geoforum.2006.11.022

Mol, A. (2002). The Body Multiple: Ontology in Medical Practice. Durham, N.C.: Duke University Press.

Müller, M. (2015). More-than-representational political geographies. In Agnew, J., Mamadouh, V., Secor, A.J., and Sharp, J. (Eds.), The Wiley Blackwell Companion to Political Geography (pp. 409-423). Chichester: John Wiley and Son.

Neimark, B., Childs, J., Nightingale, A. J., Cavanagh, C. J., Sullivan, S., Benjaminsen, T. A., . . . Harcourt, W. (2019). Speaking Power to “Post-Truth”: Critical Political Ecology and the New Authoritarianism. Annals of the American Association of Geographers, 1-11. doi:10.1080/24694452.2018.1547567

Orlikowski, W. J., & Scott, S. V. (2015). Exploring Material-Discursive Practices. Journal of Management Studies, 52(5), 697-705. doi:10.1111/joms.12114

Petit, P., & Trefon, T. (2006). Introduction. Expériences de recherche en RDC :méthodes et contextes. Civilisations, 54, 9-24.

Price, L. (2013). Think Piece. Naked Science: Avoiding Methodolatry in an Environmental Education Context. Southern African Journal of Environmental Education, 29(2012/2013), 10-24.

Putnam, L. L. (2015). Unpacking the Dialectic: Alternative Views on the Discourse–Materiality Relationship. Journal of Management Studies, 52(5), 706-716. doi:10.1111/joms.12115

Reyes, V. (2018). Ethnographic toolkit: Strategic positionality and researchers’ visible and invisible tools in field research. Ethnography, 21(2), 220-240. doi:10.1177/1466138118805121

Robertson, J. (2002). Reflexivity Redux: A Pithy Polemic on "Positionality". Anthropological Quarterly, 75(4), 785-792.

Robbins, P. (2003). Beyond Ground Truth: GIS and the Environmental Knowledge of Herders, Professional Foresters, and Other Traditional Communities. Human Ecology, 31(2), 233-253. doi:10.1023/A:1023932829887

Roth, R. (2009). The challenges of mapping complex indigenous spatiality: from abstract space to dwelling space. cultural geographies, 16(2), 207-227. doi:10.1177/1474474008101517

Rubbers, B. (2009). The story of a tragedy: how people in Haut-Katanga interpret the post-colonial history of Congo. The Journal of Modern African Studies, 47(2), 267-289. doi:10.1017/s0022278x09003838

Saldanha, A. (2003). Review Essay: Actor-Network Theory and Critical Sociology. Critical Sociology, 29(3), 419-432. doi:10.1163/156916303322591130

Santos, B. d. S. (2014). Epistemologies of the South: Justice against epistemicide. Boulder, CO: Paradigm Publishers.

Santos, B. d. S. (2018). The End of the Cognitive Empire: The Coming of Age of Epistemologies of the South. Durham, N.C.: Duke University Press.

Shapiro-Garza, E., Kolinjivadi, V., Van Hecken, G., Windey, C., & Casolo, J. J. (accepted for publication). Praxis in Resource Geography: Tensions between Engagement and Critique in the (Un)Making of Ecosystem Services. In M. Himley, E. Havice, & G. Valdivia (Eds.), Routledge Handbook of Critical Resource Geography. London & New York: Routledge

Silverman, D. (2014). Interpreting Qualitative Data (5th Edition ed.). London: SAGE Publications.

Strathern, M. (1991). Partial Connections. Savage, Maryland: Rowman & Littlefield.

Sultana, F. (2007). Reflexivity, Positionality and Participatory Ethics: Negotiating Fieldwork Dilemmas in International Researc. ACME: An International Journal for Critical Geographies, 6(3), 374-385.

Sumner, A., & Tribe, M. A. (2008). International Development Studies: Theories and Methods in Research and Practice. London: SAGE Publications.

Tsing, A. L. (2005). Friction: An Ethnography of Global Connection. Princeton & Oxford : Princeton University Press.

Waterton, E. (2018). More-than-representational landscapes. In Howard, P., Thompson, I., Waterton, E., and Atha, M. (Eds.), The Routledge Companion to Landscape Studies (2nd Edition ed., pp. 91- 101). Oxon & New York: Routledge.

PART I

Abstracting Congolese forest: mappings, representational narratives, and the production of the plantation space under REDD+

Introduction

Now when I was a little chap I had a passion for maps. I would look for hours at South America, or Africa, or Australia, and lose myself in all the glories of exploration. At that time there were many blank spaces on the earth […]. But there was one yet—the biggest, the most blank, so to speak— that I had a hankering after. […] True, by this time it was not a blank space any more. It had got filled since my boyhood with rivers and lakes and names. It had ceased to be a blank space of delightful mystery— a white patch for a boy to dream gloriously over. It had become a place of darkness. But there was in it one river especially, a mighty big river, that you could see on the map, resembling an immense snake uncoiled, with its head in the sea, its body at rest curving afar over a vast country, and its tail lost in the depths of the land.

Joseph Conrad, Heart of Darkness (1899, Ch.1 p.8)

If one looks at Google Maps and Satellite views, the Democratic Republic of Congo (DRC) still appears as the “unmapped within the map” (Hiatt, 2002, p. 223), a huge ‘blank’ or, rather, green space with few depicted rivers, villages and roads even as one zooms in. Once depicted as impenetrable, a terra incognita for scientists, the dense Congolese rainforest – the second largest contiguous expanse of tropical forest in the world – has rarely been more scrutinized, visualized and mapped than over the last decade, since the DRC engaged in the Reducing Emissions from Deforestation and Forest Degradation (REDD+) process in 2009. Very limited data on the country’s forest dynamics, annual gross deforestation rate and its drivers was available at the time and hence a “comprehensive forest management and zoning process” was nearly impossible (DRC-MECNT, 2008; Thompson, 2011, p. para. 6). Geo-coding spaces, to paraphrase Pickles (2004, pp. 4-5), has indeed been historically a key instrument in Western modernity projects for delimiting, ordering and controlling territories. One of the first major investments in the REDD+ readiness (preparation) phase was to produce such scientific geo-coded data – in particular geospatial analysis of satellite imagery – and maps of Congolese forests. On this basis, a ‘national consensus’ on an annual gross deforestation rate of 0.23%6 and on the main drivers of deforestation was established and served as an input for building the country’s REDD+ strategy, for prioritizing actions and for monitoring forest landscapes.

On the basis of this consensus, the REDD+ strategy in DRC has largely evolved towards what is labelled as a jurisdictional or integrated landscape approach7 in policy discourses (McCall, 2016). The landscape

6 As this part shows later, this figure is not definitive and still highly debated because various studies using different methods, different types of imagery and different definitions of forests/deforestation yield different estimates. 7 As I explained in the introduction, in REDD+ literature, both terms are used and refer to the same model for aligning multiple land-use types with administrative jurisdictions and coordinate multiple goals, initiatives and stakeholders. In the remainder of part I, I only refer to ‘landscape approach’.

approach broadens the initial REDD+ emphasis on climate, forest, carbon and performance towards the integration of multiple stakeholders and multiple productive land-use types – such as agriculture and mining – with a larger range of environmental and developmental goals on a wider landscape scale. This new discourse purports to be a bottom-up management approach that can emerge from ongoing negotiation and reconciliation between the competing land uses and objectives of various stakeholders – bureaucrats and experts within policy arenas, private actors, individual farmers and local communities in the Global South (Arora-Jonsson, Westholm, Temu, & Petitt, 2016; Nielsen, 2016; Sayer et al., 2013).

Emerging critics of the landscape approach have, however, questioned this all-inclusive discourse. McCall (2016, p. 59) argues that the ‘landscape’ conceptualization “focuses on the application of ‘landscape’ as a science-driven tool for analysing [and managing] ecosystem and inter-sectoral relationships” as if all stakeholders were equivalently empowered and so their knowledge, representations and values about (desirable) landscapes and land uses, too. In his case study of a landscape conservation approach in Southeast Cameroon, Clay (2016) similarly notes that the project was largely premised on expert knowledge and authority, such as advanced statistics to calculate species richness or Geography Information Systems to analyse landcover. The resulting overly spatialized interpretation reproduces uneven power relationships by promoting a particular vision of conservation coexisting with productive and extractive activities through the use of zone-based, static land use models that disregard local communities’ society-environment relationships and marginalize their livelihoods. The continuous emergence of such new models seems to be inherently linked to strategies of ‘rendering technical’, depoliticization and building consensus to distil complex realities and reinforce authoritative narratives that legitimize interventions while leaving the challenges of previous conservation and development approaches largely untouched (Lund, Sungusia, Mabele, & Scheba, 2017). The crucial issues arising from this emerging critical scholarship on ‘new’ landscape approaches are thus related to what their dominant narratives are, what and whose knowledge counts, how their framing is legitimized and implicitly or explicitly prioritizes competing stakeholders’ interests in these seemingly ‘all-inclusive’, hybrid landscapes.

Political ecological scholarship on forest governance (e.g. Blaikie, 1999; Fairhead & Leach, 1996; Forsyth & Walker, 2008; McElwee, 2016; Robbins, 2001) has provided valuable insights into the ways in which Western techno-scientific knowledge often leads to simplistic and largely ‘misread’ narratives of deforestation that help constrain the livelihoods of local communities while creating space for more powerful capital interests. These narratives are based on long established representations of nature and some trace their origins back to the colonial period, as early as the mid 19th century when issues such as deforestation and desertification entered policy debates for the Colonies (Grove, 1995). Complex society- environment relationships are distilled into seemingly apolitical abstract categories and reconfigured

within a global environmental space and a unitary political-economic rationale. Igoe (2013) specifically emphasizes how scientific and aesthetical visual articulations of nature have contributed to its abstraction, unitarization and formation as a monolithic eco-functional object of intervention, optimization and commodification, while concealing and marginalizing alternatives and opposition to this vision. Similarly, Forsyth and Sikor (2013) warn that the use of remote-sensing data for defining and measuring global REDD+ objectives might encourage both large-scale reductions in deforestation rates and industrial selective logging rather than shifting cultivation and improvement in forest quality/biodiversity that could be more beneficial locally. The argument here is not that scientific explanations of deforestation are invalid but that the selectiveness of measurement methods, their take on these issues, what problems they highlight or obscure “are shaped by the imprint of dominant narratives from which they dr[a]w their intellectual inspiration and legitimacy” and by powerful international and national environmental policy-making institutions (Blaikie, 1999, p. 133).

This first part builds on these sets of critiques and insights from Science and Technology Studies and critical cartography to empirically analyse the controlling geospatial science-driven narratives and seemingly neutral cartographic representations of Congolese forested areas informing the genesis of a REDD+ landscape approach in DRC. I adopt a relational understanding of mapping (Kitchin & Dodge, 2007) to understand how maps within DRC’s REDD+ assemblage of human actors, interests, (historical) discourses, narratives and material phenomena contribute to reifying and shaping particular socio-spatial relations and identities, and in whose interests. My analysis is about the politics of environmental knowledge and the perpetuation of discursive and epistemic dispossession that allow a transnational network of hybrid actors – acting as “interpretive communities” (Mosse, 2004, p. 646) – to impose their meanings of the landscape, and that underpin various accumulation strategies and the (re)production of inequalities (Igoe & Brockington, 2007; West, 2016). These “complex acts of dispossession” (West, 2016, p. 1), I show, start with the (almost) exclusive use of satellite data, mapping and visualizations at the cost of other epistemologies and with the misrecognition and subordination of local forest users who are cast as unproductive and environmentally unfriendly, and are denied any control over the definition of their struggles, identities, social worlds and needs (Fraser, 2000; West, 2016). To paraphrase Fraser (2000), issues of epistemic (mis)recognition are fundamentally intertwined here with social and distributive (in)justice, and material (dis)possession. Supposedly objective cartographic, semiotic, verbal and visual productions of Congolese forests and their users in fact legitimize old discourses of ecological modernization and spatial control, and help to produce desirable ‘integrated’ green economic landscapes eligible to more powerful external actors. These discourses, I argue, are rooted in what some critical scholars (Haraway, 2015; Haraway et al., 2016; Haraway & Tsing, 2019) have recently termed the ‘Plantationocene’, that is a homogenization of ecologies into the system of extractive and enclosed plantations relying on forms of waged but alienated labour, that starts with the elimination of other

agricultural and forest living practices. Epistemic injustice, I argue, thus contributes to the restructuring of not only space and socio-ecological interactions but also of socio-economic relations and individual subjectivities for the needs of the regeneration of capital; all are inextricably imbricated. More broadly, my critique is concerned with “unveiling some of the epistemic silences of Western epistemology […] to allow the silences to build arguments” (Mignolo, 2009, p. 162) and challenging hegemonic “rhetorics of representation” at the roots of all uneven development (West, 2016). This analysis is particularly important as, to my knowledge, few studies have taken a critical geography and political ecology perspective to analyse the discourses that underlie REDD+ strategies in DRC (but see Ehrenstein, 2013, 2014) – while the literature is more abundant for other countries such as Indonesia. It also generally adds to the limited academic empirical material available on DRC’s REDD+ implementation (but see e.g. Aquino & Guay, 2013; Reyniers, Karsenty , & Vermeulen, 2016; Samndong, Bush, Vatn, & Chapman, 2018; Samndong & Kjosavik, 2017). In a country that is usually mentioned for the militarization of its conservation (e.g. Virunga National Park), its spectacular poverty, failing institutions and violent wars, I believe it is essential to reveal a less spectacular but enduring “slow violence” (Nixon, 2011) of environmental rule. In Nixon (2011, p. 2)’s terms, slow violence is “a violence that occurs gradually and out of sight, a violence of delayed destruction that is dispersed across time and space, an attritional violence that is typically not viewed as violence at all”.

The remainder of Part I is divided as follows. In chapter 1, I outline my theoretical and epistemological underpinnings – critical cartography, power/knowledge, assemblage and performativity. In the next two chapters I develop my arguments by looking at the case of REDD+ in DRC. Chapter 2 shows how diverse geospatial imageries, maps, visual and discursive elements are drawn together to give authority to claims blaming local people’s livelihood practices for deforestation. It emphasizes the contingency of DRC’s REDD+ maps and narratives by highlighting their roots in colonial discourses and their function as a convenient reality in a socio-political context characterized by constant (re)negotiations. Chapter 3 analyses how this rhetoric of representation has led, in DRC’s strategy and investment plan, to a monoculture of bounded rationalized space managed through privately-held plantation concessions to the detriment of communities. I further show how it recasts forest users’ subjectivities into standardized categories of the ‘socially responsible green company’ and the ‘enviropreneurial commodity petty producer/labourer’ integrated in international markets while completely neglecting power inequalities between actors and complex issues of sovereignty over land and resources, their management and conservation. The conclusions summarize this first part’s findings and emphasizes the ongoing slow violence that treats both people and natural ecosystems as disposable and malleable to powerful economic interests.

1. Interrogating maps and mapping: from hegemonic representation to performativity

Anyway, we can say that, in the current era in which technical means allow fast realizations, even more so than in the past, it is economical and rational to start the equipment of new countries with the establishment of the map. […] In a developing country, the map appears as a unifying factor, in the action of all those who are concerned with prospection and planning.

Albert Gilliard, Cartographie congolaise (1953, p.4)

Existing maps have been drawn in function of economic interests. Everyone had specific concerns and everyone has developed their own reference point. […] Maps produced by NGOs today, that is not maps’ production because they work out their maps for their own interests. That is why it is relative.8

Director of the Geographic Institute of Congo (IGC)9

1.1. Contextualizing: modern mapping and (colonial) power

The fact that maps and cartographic visualizations10 are imbued with and have power has long been shown by critical cartography (e.g. Crampton, 2010; Crampton & Krygier, 2005; Harley, 1988; Lefebvre, 1991; Pickles, 2004; Sletto, 2002; Winichakul, 1994; Wood, 1992). They are an entire part of the political economy of government (Crampton, 2010, p. 63). Historically, and in particular since the early 19th century, mapping has been entangled with political decision making, administration (metropolitan and colonial) and monitoring of territories, resources and populations – that is ‘geographic governance’ (Crampton, 2004) or ‘cartographical reason’ (Olsson, 2007). Building on Foucault’s concepts of governmentality and biopolitics – i.e. a type of political rationality concerned with the well-being of populations and used by different types of authorities and agencies to shape human conduct – Crampton (2004, 2007), Pickles (2004) and Rose-Redwood (2006) among others argue that the geo-coding of the world through mapping is in fact a fundamental requisite to this form of governance. As maps along with other statistical and census tools allowed the spatial tracking of social, economic and environmental phenomena and other spatial patterns and processes, they also became a vital instrument for states and new types of governing institutions to identify and address all types of societal problems (Kitchin, Dodge, & Perkins, 2011). Similarly, maps have played a key role in (geo)surveillance and probably even more so

8 Both quotes are own translations from French. 9 From 2 interviews 31-CSI-KIN-20160407, 48-CSI-KIN-20170531. 10 Though I will provide a more complex understanding of ‘maps’ and ‘mapping’ later in this section, for the sake of clarity, please note that I use the term “map” as an all-encompassing term for various conventionalized spatial visualizations that provide a bird’s eye view – looking straight down from above – of a place, a landscape, a phenomenon or a process with “a consistently applied reduction in scale” (Dodge, Kitchin, & Perkins, 2011, p. x). They can be static, paper based or dynamic, interactive. They non-exhaustively include: conventional maps, Geographic Information Systems (GIS) maps, satellite image maps, 3D maps, etc. While I only look at ‘expert’ mapmaking in Part I, maps are also increasingly open-source and collaborative. 45

over the last few decades with the rapid advances in mapping and algorithmic technologies that can track and profile individuals and groups/populations as a whole or in the case I am interested in, forest use (Crampton, 2010). Geospatial technologies and mapping can hence inform governance interventions, expose wrongdoings and “[(re)order] lived lives into markets [or] potential profits” (Kitchin et al., 2011, p. 391). In this way, ‘power through map’ is maybe more than ever, ubiquitous.

One essential aspect that makes geospatial mapping and representations so powerful and central to governance is the way they dematerialize and produce a ‘radical abstraction’ of space in points, lines, squares and surfaces (Lefebvre, 1991; Massey, 2005; Pickles, 2004). Maps, by de-socializing the landscape and rendering places, people, human and nonhuman things and complex lives and processes as mere dots or pixels, makes power from a distance easier because it allows bureaucrats, developers or planners to take action on “bodies of unique places” without witnessing the human and social consequences (Harley, 1989, p. 14; Kitchin et al., 2011). Thus, in addition to a familiar sense of power in cartography, somehow ‘external’ to maps –or, power through maps – maps also have an ‘internal power’ (Harley, 1989). In other words, power is inherently constituted in the very design of maps, in the selectivity, inscriptions and silences they produce and in the very type of meanings and values they embody (Kitchin et al., 2011; Pickles, 2004). Even something as ‘benign’ as a mapped visualization of forests and population, as I explore in this first empirical part, has a politics that expresses an implicitly embedded social vision (Harley, 1989) that typifies and structures certain ways of knowing, thinking, conceiving and being in space.

In two landmark works exploring the relationship between maps and nation states, Winichakul (1994) and Anderson (1983/2006) have shown the key role modern geography has played in the arbitrary and artificial construction of nationhood and the correlated European imperial and capitalist expansion; the specific purpose of modern nation-states being to control and acquire territory. In their argument, nationhood, since it is a socio-spatial reality that “can never be experienced in its totality” needs the modern map as a mediator “in perceiving and conceptualizing such macrospace in its totality” (Winichakul, 1994, p. 55). The latitude-longitude grid and the bird’s-eye views convention of modern maps anticipated boundary lines and territorial units within a larger space on the Earth’s surface. It became the model for thinking, imagining and legitimizing a desired, projected socio-political realm and for inscribing identities. Nations are thus imagined communities (Anderson, 1983/2006) whose origins, territories and indispensable boundaries, values and practices, i.e. their geo-body (Winichakul, 1994), are an effect of maps and cartographic technologies and representations.

Early modern maps, Winichakul (1994, p. 114) further argues, also contained an embedded desire of expansion as the latitude-longitude matrix covered the entire globe, “full of blank squares waiting to be filled in”; blank in terms of hegemonic Western representational science. Concurrently, colonial

expansion contributed to raising the issue of unknown, blank spaces on the maps “to the forefront of geographical concerns” (Hiatt, 2002: 239) and enabled the filling in of these spaces with the help of colonizers, surveyors and military forces (Anderson, 1983/2006; Winichakul, 1994). The blanks on the maps were of course analogous to the idea that these regions were empty, negating any forms of inhabitation and social formations, and simultaneously testifying to a lack of possession: “[they] signif[y] land, rather than territory, earth rather than ownership” (Hiatt, 2002, p. 248). Consequently, these blank spaces on the map ‘invited’ exploration and legitimized political expansionism and acts of demarcation supported by a European cartographic system of knowledge and perceptual framework (Anderson, 1983/2006; Hiatt, 2002). In other words, maps by making far-flung places knowable allowed European states to project and assert their political and commercial power from a distance and thus ultimately contributed to both the brutal violence of colonialism and capitalist expansion (Kitchin et al., 2011). In this way, as early as the 15th century but more intensively from the 18th and 19th centuries, the development of and investment in foreign mapping coincided with the territorial and commercial interests of European countries (Crampton, 2010; Hiatt, 2002).

For instance, Vandermotten (2008) shows that the founding of two Belgian Geography Societies was boosted by Leopold II and the maps they produced became key instruments of colonial propaganda for legitimizing and advertising the colonial project to Belgian public opinion and business investors. As in other colonies, Belgians also exported this specific system of representation to arbitrarily trace and fix the borders of its Congolese colony so as to include Katanga’s rich mineral reserves, and to repel the British and Germans (Turner, 2007). From 1925-1930, topographical, soil and geological mapping (Figure 2) was also initiated as it became increasingly useful to mining operations and industrial agriculture, the two main economic sectors at the time (Baert, Ranst, Ngongo, & Verdoodt, 2012; Gilliard, 1953). It hence facilitated large-scale accumulation strategies (Peluso, 1995). The transformation of Congolese nature and land into extractive and agricultural resources was also soon made visible on maps; conveying a powerful idea of a re-ordered space into uniform areas of production (Figure 3). For governance, ethno- linguistic maps which were supposedly based on historical territorial occupation were also largely used by the Belgian colonial administration to govern a very diverse population of hundreds of different groups with long-term consequences for local inter-ethnic and tenure dynamics (Huggins, 2010).

While the power of maps is used to (re)produce specific power relations for states and dominant institutions their inherent power has also offered new ways to resist for oppressed and marginalized groups, or for social causes. Anderson (1983/2006, p. 175) shows that paradoxically, in European colonies, maps were at the heart of anticolonial nationalisms as the “logo-map penetrated deep into popular imagination” to create a feeling of membership and to unify and organize hundreds of different ethnic groups – some of them not even aware of the other’s existence before. The interests expressed in

colonial maps could in this way be made to work for the dominated and marginalized whose resistance somehow emerged from maps (Crampton, 2010). The movement of counter-mapping has generally allowed a myriad of groups to claim and delineate territories and resources they have long managed, in response to state or private land grabs and to industrial exploitation (Peluso, 1995). Maps’ sovereignty is also increasingly challenged by participatory mapping and the emergence of a new ‘populist cartography’ in which non-expert mapmakers have gained (some) access to public geospatial data and the production of maps (Crampton, 2010; Crampton & Krygier, 2005). These developments, while they could be seen as weakening the power of the map, in fact “take up maps and politics in an explicit manner to provide alternative mappings of space not represented by official state agencies” and make other equally powerful claims (Crampton & Krygier, 2005, p. 25). The Environmental Justice Atlas which maps ecological conflicts and sites of resistance around the world, or the Land Matrix initiative that maps land grabs around the world are examples of this. In this way, despite first appearances, maps have never become post-political (Crampton, 2010). A power that is not only exerted on the map by the institutions within which it is produced and disseminated but that also comes from the map. A power that is not only political but also symbolic, ideological and epistemological.

Figure 2. Geological map of the Belgian Congo. 1949. (Cahen & Lepersonne, 1949)

Figure 3. Map of main types of mineral, plant and animal production in the Belgian Congo. 1937-1938. (Laude, 1938)

1.2. Unfolding the concepts: performativity and relational mapping, tools for (re)thinking about maps in environmental governance

As such the map does not represent the world or make the world: it is a co-constitutive production between inscription, individual and world; a production that is constantly in motion, always seeking to appear ontologically secure. Kitchin, Perkins, and Dodge (2009, p. 21)

What emerges from the detour in the entangled history of modern mapping with political economy I outlined above, is the fact that maps are propositions about the world and ‘ontogenetic’ in nature (Kitchin, Gleeson, & Dodge, 2013). In other words, there is a two-way traffic between material realities and the practices for representing and understanding them. From this post-representational, actor- network theory perspective, the scientific method is not a set of purely technical and innocent procedures for reporting on reality (Law, 2004). Rather, it is an assemblage of contingent, provisional unity between heterogeneous human and nonhuman components that include material, discursive and social artifacts (Anderson & McFarlane, 2011). It is the assembling processes of these components, in specific time and place, that are performative, i.e. they contribute to enacting pre-existing social and material contexts (Callon, 2006; Law, 2004; Yeow & Faraj, 2014). Reassembling can hence produce different ontologies (cfr. counter-maps) and often follow unintended routes as they are contingent on the fostering of new relations and the entrance of new elements (Anderson and McFarlane, 2011, p. 126).

Western modern mapping, however, tends to disentangle itself from such multiplicity, indefiniteness and complexity. Especially when they are associated with planning, modern maps appear as the default comprehensive view of the disposition of things in space and of a static “world ready-made for life to occupy”, from “a mind that is situated above and beyond it” (Ingold, 2000, p. 235 & 241). Satellite-based maps, even more so, convey an idea of accuracy and total vision and total knowledge of the planet through a nonhuman, transcendent bird’s-eye view (Kwan, 2007). However, with an alternative more- than-representational understanding, maps should be understood as a set of unfolding practices enacted to solve relational problems (e.g. spatial distribution of deforestation drivers). They are constantly in the making, contingent, relational and ontologically unstable; they are always mappings, coming together in particular times and places (Kitchin et al., 2013). The issue here then becomes one of looking at the interplay between the map and the social world. It means connecting the material product of mapping, i.e. the map, backed up with wider assemblages of power/knowledge (e.g. REDD+) and “discursive and material practices shaped by personal, social, embodied, political and economic relations”; and “how they perform as actants in the world shaping knowledge and actions” (Kitchin et al., 2013, p. 15). The power of maps is relational rather than fixed and complete.

Moreover, if mapping is about crafting and bundling, and thus only makes some constituents of ‘the world’ present, it necessarily and simultaneously produces manifest absence and Otherness (Law, 2004). The other leading question of this first part is hence: what silences do the maps within the REDD+ assemblage produce and how do these silences and otherness continuously emerge? My entry point will be the cartographic products themselves. As Wood and Fels (2008) argue, the key to this inquiry is the articulation of the map and the paramap, i.e. the verbal and nonverbal discourses that surround the map to position it and through which argument and authority are conveyed. The paramap, in turn, consists of the perimap and the epimap. The former refers to the physical map: title, legend, colours, graphs, etc. The latter refers to the discourse surrounding the map, that shapes the reading and perception of the map. Our adopted approach thus refocuses (critical) cartographic research as “sciences of practices, not [(un)truthful] representations” (Kitchin & Dodge, 2007, p. 342). Such conceptual insights allow us to deconstruct the apparent objectivity and neutrality of maps to highlight their sociality, i.e. how mappings differently affect various actors and social relations such that technoscience and crucial issues of environmental justice are and should be woven together. The question in this first empirical part thus becomes what kind of space, socio-spatial relationships and subjectivities are activated (or not) through mappings’ genesis and unfolding in the wider DRC REDD+ assemblage of people, (historical) discourses, narratives and material things? And who are the ‘winners’ and the ‘losers’ in this process?

2. Representing forests and population: a contingent assemblage

The ones who destroy the forest are the peasants. Industrial logging does not destroy it. Peasants finish the job. The other big problem is that they burn [trees] to make charcoal.11 One international REDD+ Technical Advisor12

The DRC’s REDD+ assemblage has given primacy to remote-sensing satellite mapping – largely seen by policy makers in this context from a positivist perspective as detached global knowledge par excellence – as an entryway to solving the problem of deforestation. In DRC, three interrelated, American-based initiatives producing geospatial data based on Landsat satellite imagery have been highly reinvested by REDD+ and forestry practitioners for forest and policy monitoring. Among these initiatives ‘Global Forest Watch’ (GFW), a ‘near-real time’ online forest monitoring system initiated by the World Resource Institute (WRI) with 40+ partners, is praised as ‘the’ technical feat finally offering reliable, objective forest data. The assemblage of GFW partners – whether users or funders is never made clear by WRI – comprises conservation non-profits, major environmental organizations supporting and employing ecosystem/landscape approaches, and multilateral cooperation agencies funding REDD+ programs as well as blue-chip private corporations including two of the world’s largest agribusiness industries/palm oil buyers. As WRI’s CEO asserted, thanks to GFW ‘[f]rom now on, the bad guys cannot hide and the good guys will be recognized for their stewardship’ (WRI, 2014, p. para.2). However, GFW does not mention who the ‘bad’ and ‘good’ guys are; this is left to each user, as GFW presents itself as a neutral actor, supporting sustainable management of human-environment relationships.

In this chapter I provide a detailed analysis of the relational crafting and use of remote-sensing mapping and geovisualizations within the DRC’s REDD+ assemblage of forest experts, REDD+ consultants, state actors and CSOs, and of other material, discursive and historical elements. I show how this assemblage has contributed to enact a single abstract story of community-induced deforestation that says very little about socio-political and power relations that shape forest use and change, and that virtually ignores local knowledge, thinking and living models.

2.1. Creating consensus on contested data

In 2008, when DRC became one of the first countries to apply for the World Bank and UN REDD+ readiness funds, the country’s proposal indicated an annual 0.2% to 0.25% gross deforestation rate for the 1990- 2000 period (DRC-MECNT, 2008). However, there was no definitive estimate as very few satellite remote- sensing-based data were available. Between 2008 and 2014, at least 10 remote-sensing studies were

11 Quotes from interviews have been translated from French as literally as possible by the author. 12 44-RCS-KIN-20170529. 53

carried out (Ickowitz, Slayback, Asanzi, & Nasi, 2015). Yet there is still no definitive estimate – neither in terms of forest cover or deforestation – largely because these studies use different methods, periods of observation and definitions of forest and deforestation (Ickowitz et al., 2015). Depending on the study, gross deforestation rates vary from 0.15% to 0.25% for the 1990-2000. Despite these data discrepancies, the overall picture is of some forest loss, especially of secondary forest and particularly concentrated in peri-urban areas and the densely populated Eastern part of the country (Ickowitz et al., 2015). As for estimations, most official REDD+ programs’ documents, working papers and reports produced by international organizations (e.g. Aquino & Guay, 2013; DRC-MECNT et al., 2015) use the annual gross deforestation rate of 0.22% or 0.23% for the 2000-2010 period, and on this basis, a projected rate of 0.41% for the 2010-2035 period. This rate is substantially lower (fifth to half) than the rates of Malaysia, Indonesia or Brazil but the highest and fastest growing within the Congo Basin, which justified DRC’s inclusion within the REDD+ international program (Aquino & Guay, 2013; DRC-MECNT, 2008; Ehrenstein, 2013).

In the same way, no recent specific study on deforestation drivers in DRC was then available – only a few cross-national, regional and local studies existed (DRC-MECNT, 2008; Ickowitz et al., 2015). One of the first major goals of the REDD+ readiness phase was to gather scientific data to inform a ‘national consensus’ – between international development partners, state actors, the private sector and civil society all engaged in the REDD+ process (Table 1) – on the main drivers of deforestation. The expert- based ‘national consensus’ was grounded almost exclusively in the use of Landsat remote sensing satellite maps and geospatial expertise. Although the studies’ synthesis acknowledges variabilities between provinces, the ‘national consensus’ posits shifting cultivation and population growth as the main direct and underlying drivers of forest loss, with industrial logging and mining activities having only limited impacts, though they are recognized as potential (but inevitable) future threats to forests (DRC-MECNT et al., 2015; UN-REDD & DRC-MECNT, 2012). In this way, right from the first pages of the 142-page long strategy, after announcing that DRC is a “threatened forest giant” – by rural communities’ activities as we learn a couple of pages later – the text goes on to describe the country as a “potential minerals, agriculture and energy giant” with “a largely untapped potential” (DRC-MECNT et al., 2015, p. 12, own translation). “Rural communities’ activities [are] at the heart of current processes of deforestation and degradation” while, for instance, “large-scale agriculture, whose expansion is necessary to the country’s development, will certainly become the major deforestation driver in DRC in the next few years” (DRC- MECNT et al., 2015, pp. 25-26, own translation).

Type Role Stakeholders

National REDD+ Fund (FONAREDD), National Coordination REDD+ (CN–REDD), Ministry of Environment and Sustainable Development Government Overall coordination of REDD+ (MECNT), Direction of Inventories and Forest Entities process and programs Improvement (DIAF), Ministry of Finance, Committee for Technical Reform, Province of Maï Ndombe–Ministry of environment.

Central African Forest Initiative facilitates (CAFI), World Bank (Forest Carbon Partnership Facility), UN-REDD program of the United Nations (FAO, PNUD, PNUE), African Development Bank (Congo Donor and Provide technical and financial Basin Forest Fund / Forest Investment Program). Technical support for REDD+ activities and Governments of Norway (NORAD) –biggest donor Partners projects country–USA (USAID/CARPE), Japan (JICA), France (AFD), UK (DFID), Germany (GIZ), European Union.

CAFI acts as the coordinator for aligning bilateral and multilateral assistance.

Invest and implement REDD+, Federation of Wood Industries (FIB), Forest Private actors agriculture, logging, Resources Management (FRM), Wildlife Works conservation and other projects Carbon (WWC), NOVACEL13, SOGENAC14

Climate and REDD+ Working Group (GTCR) –a coalition of Congolese environmental CSOs, World Civil Society Implement projects, provide Wild Fund for Nature (WWF), Wildlife Organizations and technical support and ensure Conservation Society (WCS), Conservation International civil society participation in International (CI), World Resource Institute (WRI), NGOs REDD+ process CIRAD, National university institutions, Forest Governance Observatory (OGF), Satellite Observatory of Central Africa Forests (OSFAC)

Table 1. Key stakeholders involved in the REDD+ process in DRC. Adapted from Fobissie, Alemagi, and Minang (2014) and Mpoyi, Nyamwoga, Kabamba, and Assembe-Mvondo (2013)

A range of critiques on the quality, validity and reliability of the data yielded in four REDD+-sponsored studies has been formulated by some scholars (e.g. Ickowitz et al., 2015; Moonen et al., 2016). Similarly, the absence of voices from among the local populations – interviews were carried out almost exclusively with experts and forest administration officials (UN-REDD & DRC-MECNT, 2012) – and of on-the-ground research has been denounced by several international activist NGOs15 (e.g. Rainforest Foundation UK16 –

13 Congolese company that owns large plots of land in the Northwest of DRC. It used to be a commercial cattle ranch but has now moved into the carbon credit business, and planted 4200ha of acacia trees. 14 Congolese cattle and meat company present in the Maï Ndombe Province where the first REDD+ jurisdictional program has been implemented. 15 International environmental NGOs (like Greenpeace, Rainforest Foundation UK (RFUK) or Global Witness) much more than Congolese environmental CSOs are contesting these studies and the framing of the REDD+ program in DRC. While a detailed analysis of the heterogeneity of opinions among NGOs/CSOs would exceed the scope of Part I, it is important to note that their views on the REDD+ process, forestry reforms and industrial logging are sometimes divergent. 16 50-INGO-SKY-20180423. 55

RFUK, Global Witness, Greenpeace). While the pilot projects during the REDD+ readiness phase were supposed to confirm the studies’ results through assessments on the ground so as to better inform the National Strategy, most projects were delayed and just started when the REDD+ framework strategy was first validated in 2012 (Kipalu & Mukungu, 2013; Reyniers et al., 2016). In legitimizing this narrative, an assemblage of selective geospatial data, maps, cartographic, graphic and discursive (expert) representations of land cover change has hence become the unique “neutral language of science”, seemingly extracted from the messiness of the socio-political world (Li, 2007, p. 7). Present in REDD+ official documents and communication mediums, they simplify perceptive judgements and act as reference objects to assign blame and spatially plan REDD+ actions (Ehrenstein, 2014; Wood & Fels, 2008).

2.2. Threatened carbon stocks: a seemingly neutral and disconnected representation

In the official synthesis of the four studies on deforestation and degradation drivers used by REDD+ policy makers in DRC, what first attracts the reader’s eye is the geospatial illustration of the “consensual opinion of experts” (UN-REDD & DRC-MECNT, 2012, p. 13, own translation) on deforestation drivers on a map of DRC’s ten provinces (Figure 4). It is the only map in the synthesis. It lacks any geographical precision and immediately gives an overwhelming impression that the story is extremely simple. Catching the eye of the reader, the dark grey colour – red in the paper version of the report – renders particularly visible and dramatizes deforestation driven by swidden slash and burn agriculture and fuelwood collection/harvesting overall the country. On the other hand, the map omits and silences the considerable variations within smaller scale levels. In fact, as a GFW expert told me17, because current Landsat satellite resolution does not allow for the differentiation between different type of land use activities in the absence of land use maps, quantitative geospatial models in DRC are only based on explanatory variables identified by experts and meta-analyses of drivers of tropical deforestation.

Besides this map, in the paramap, five tables and nine charts order the results from the different studies in terms of direct and underlying deforestation drivers. From one table/charts to another, the weight of the various drivers in terms of their impact on deforestation varies. Yet, the end of the synthesis proposes a “quick reading of the main conclusions of the synthesis’ work” (UN-REDD & DRC-MECNT, 2012, pp. 28, own translation) through two last concluding figures (Figure 5 & Figure 6) and a short verbal explanation to confirm the threat. As it would be difficult ignore variations in studies completely, these two last charts appear however rather obscure. Figure 5 – which is taken up extensively in REDD+ official documents – lists a series of direct and underlying deforestation drivers as well as ‘important variables’ without giving any weighted shares or estimations. The repetition of certain drivers under more than one column, with

17 85-ITSP-SKY-20191011. 56

slightly different labels, obscures the understanding. The short explanation below the chart indicates however that it clearly shows that rural population’s activities are the main drivers of deforestation. Figure 6 takes up the drivers presented in Figure 5 and gives some weight to each of them. This chart suggests (causal?) correlations between drivers based on experts’ observations. However, no clarification of which factors are correlated to each other, no causal direction nor further verbal explanations are provided. While the two charts might bear witness to the complexity of deforestation drivers in DRC, their conclusion yet goes back to the initial story provided in the introductory map (Figure 4): the activities of the rural population, and in particular slash-and-burn (shifting) agriculture and fuelwood exploitation, as well as population growth are the main drivers of deforestation.

Figure 4. Graphic presentation of the relative importance of expert-perceived direct deforestation drivers at provincial level (UN-REDD & DRC-MECNT, 2012). It immediately establishes the initial focus and viewpoint: rural population’s activities are threatening DRC’s forests.

Figure 5. “Synthetic representation of main direct and underlying deforestation drivers and important explanatory variables” (DRC-MECNT, FCPF, & UN-REDD, 2015; UN-REDD & DRC-MECNT, 2012, p. 28). It includes neither figures nor estimations.

Figure 6. “Updated synthetic representation of main direct and underlying deforestation drivers and important explanatory variables” (UN-REDD & DRC-MECNT, 2012, p. 29). The more the driver is perceived as important, the bigger the rectangle’s size.

I now turn to the website of the main REDD+ investment program CAFI-DRC18 (CAFI-DRC, 2017b), which serves as the main communication medium for the latest information on REDD+ progress in Congo Basin countries. The website contains an extensive section with 29 Frequently Asked Questions (FAQ) about “CAFI and the forestry sector in DRC” (CAFI-DRC, 2017a). A more recent map-paramap assemblage appears on all of the three FAQs’ subpages that deal with deforestation drivers, and points back again to the same direction (Figure 7), acting as a reference object that gives an authority to the CAFI REDD+ strategy (Wood & Fels, 2008). One question/subpage claims that it seeks to provide ‘proven’, evidence- based facts on the main drivers of deforestation and land degradation. The subpage starts with a short and straightforward text – the epimap – stating and emphasizing in bold script that the expansion of subsistence activities (slash-and-burn agriculture, and fuelwood collection and harvesting) is the main cause of deforestation and forest degradation and is hence closely correlated to the spatial distribution of population. The text also highlights – as the other 28 FAQs uniformly do – that “contrary to popular belief, [industrial] forest exploitation is not systematically a driver of deforestation and degradation” (CAFI, 2017b, para. 2, own translation). As such, the FAQs seem to largely defend and legitimize the highly contested19 Sustainable Forest Management Program (PGDF) that supports logging companies in developing their forest management plans. Yet, as explained below, the extent to which shifting agriculture represents permanent deforestation and the role of industrial logging in deforestation are both very much debated issues, and often involve much more complex socio-ecological dynamics.

Then comes the map and its perimap (Figure 7). The map displays four carefully chosen human and nonhuman elements: forest cover loss, population distribution, roads and ‘forest’ concessions. It draws on three spatial data sources: GFW (forests), WorldPop (population) and DRC’s Common Geographic Reference System (roads and administrative limits). The map highlights forest cover loss for the period 2000-2014 in an eye-catchingly bright red colour and superimposes it on population density explicitly suggesting a direct causal link. In fact, ‘population’ and ‘forest cover loss’ are completely intertwined on the map, making it difficult to distinguish one from the other. Logging concessions – framed as ‘forest concessions’ – are represented in a non-threatening green, the traditional colour of trees and forests, suggesting the absence of deforestation in those spaces. The (para)map design, with its particular use of map colours and bold characters in the text, distracts the reader from the forest fragmentation that appears within these concessions. It also conceals which concessions are actually active and which ones are not. Since the 2002 logging moratorium, only 10 logging companies are authorized to officially harvest 12 million hectares – a tenth of DRC’s forests (Karsenty & Ferron, 2017; Lawson, 2014). The official volume

18 CAFI (Central African Forest Initiative) and its related fund FONAREDD (Fonds National REDD+) is a partnership between the DRC, and multilateral and bilateral donors, for the implementation of the REDD+ investment phase. It acts as a coordinating body of the REDD+ process in DRC. Its website has become the main reference regarding the REDD+ overall strategy and programs in DRC. 19 PGDF was and still is contested by some international activist ENGOs and some Congolese CSOs as I describe later in this first empirical part. 59

of wood produced has historically never exceeded 400 000m3 per year since 1990, 5 times less than Cameroon’s 2.2 million m3/year – where the impact share of industrial logging on deforestation is much higher than in DRC – and only 4% of Central African countries’ total production (FCPF, 2016; Karsenty, 2016a). A large volume of actual industrial logging production however stays under the radar as lots of artisanal logging permits “have been issued illegally to industrial logging companies and used for industrial-scale logging” and as some other companies sub-contract part of their exploitation to smaller ‘artisanal’ loggers who then resell their volume to the concession owner (Karsenty, 2018; Lawson, 2014, p. 6).

In the same way, the map in Figure 8 justifies priority zones of the CAFI REDD+ Programme Intégré Oriental by framing areas of community activities as deforestation “hotspots” , a.k.a. “zones with high population density” (FONAREDD RDC & UNDP, n.d., p. 13). The legend points out that the hotspots correspond to livelihood activities, i.e. slash and burn agriculture (in red circles), fuelwood production, artisanal logging and mining. Industrial activities in the area are not identified, e.g. in the Tshopo province, 16 logging concessions (occupying 21.5% of the province’s territory) and 2 oil-palm plantations; neither are protected areas (4% of the Tshopo Province; 31% of Bas-Uélé Province). Similarly, smaller private plantations and former colonial plantations are not mapped as cadastral data is practically non-existent20. In fact, except for Figure 7 in which logging concessions’ presence is carefully rendered insignificant as I detailed above, both maps totally omit to show the importance of this territorial organization and fragmentation. While these logging and agricultural concessions are not all active or exploited at the moment, their bounded perimeter still engenders particular socio-ecological dynamics and forces an increasing number of peasants to encroach on other forests for their livelihood activities21. Also smaller logging roads, whether they are still used or not by the company, are invisible on the map (Figure 7). Yet they explain both logging exploitation patterns and the encroachment of what is seen through a satellite bird’s eye view as a road infrastructure vacuum, showing much more complex – partly indirect – links between industrial forest exploitation and deforestation.

20 29-CSI-KIN-20160407. 21 15-CCSO-KIS-20160229, 17-CSI-KIS-20160301. 60

Figure 7. DRC’s forests under threat. Main map on the CAFI REDD+ investment’s program website (CAFI-DRC, 2017b). In this static map, the causal link between population and forest cover loss is made explicit through its dramatic representation in bright red while logging concessions (identified as forest concessions) are presented in a non-threatening green.

Figure 8. Priority zones of the CAFI Programme Intégré Oriental, determined on the basis of “deforestation hotspots” framed as community-induced (FONAREDD RDC & UNDP, n.d.).

In 2018, I carried out a participatory historical mapping exercise (Figure 9) with communities inhabiting the perimeter of a logging concession – a perimeter that is in fact often pointed out by REDD+ practitioners as a strong example of forests’ destruction by shifting cultivators – whose main entrance is situated at about 40km from Kisangani. It tells a particular story about the evolution of the company’s exploitation activities in space and time and how it has attracted all sorts of local livelihood and commercial activities. In fact, land and forest use patterns are very diverse. Before the arrival of the logging concession only a few villages and artisanal small-scale diamond mines (see orange diamonds on the map) were present in the zone. The map represents main logging roads in red and skid trails – bretelles d’exploitation* – in yellow. Zones of industrial exploitation are identified in yellow dots around skid trails and periods of exploitation are in black.22 Exploitation, according to participants, started in 1994 near Alibuku, which has since become a secondary commercial centre for charcoal – makala* in Lingala. The largest part of makala* production is controlled by urban elites from Kisangani who employ local labour. The zone around Alibuku is very much decimated of trees, mainly due to charcoal production, and there is particularly easy access as the main logging road (in red) is still maintained and used by the company’s trucks; trucks that are also being used by some producers for the transport of charcoal to Kisangani. Over time, and with the progression of logging exploitation and of main and secondary skid trails, ‘migrant’ families attracted by easier market access have settled there; a few houses progressively growing into villages. Their main activity is agriculture, which is facilitated by the previous clearance of some areas for opening skid trails. After the company has (temporarily) left the area, these families and villages usually remain, sporadically maintaining roads and skid trails so as to access markets with a (motor)bike. The closest one is to Alibuku and thus to Kisangani, the more one finds private concessions acquired by urban elites to develop commercial cocoa plantations along the skid trails and sometimes as far as 5 or 7km from the main logging road. Artisanal loggers (see dark brown spikes on the map), who are often powerful political and military elites also take advantage of the company’s logging roads to carry out their own activities. In fact, as several informants in the area told us, some of these artisanal loggers sell their timber back to the industrial logging company; information that is corroborated by reports from Global Witness (Global Witness, 2012). Roads, villages and multiple types of land use by different kinds of actors have thus followed the extension of the industrial companies’ activities inside the dense forest. Yet, REDD+ maps render these processes invisible and rather identify tree cover loss in these types of zones as ‘primary forest perforation’ by shifting cultivators pushed away by population growth.

22 Note that the name of the company has shifted from Amex-Bois to Trans-M, then Cotrefor and now IFCO due to numerous accusations of illegality and links with Hezbollah. It seems, however, that the very secretive owner has always remained the same despite assertions to the contrary. A study of these issues would be beyond the scope of this thesis but Global Witness and The Sentry have both written reports on the topic (Global Witness, 2019; The Sentry, 2018). 63

Figure 9. Participatory map carried out with communities inhabiting the perimeter of a logging concession in the hinterland of Kisangani (11/05/2018). It highlights the progression of the company’s logging activities across space and time, and the diverse land and forest use patterns it generates.

2.3. What (non-)forests? Which people?

The static maps used directly for policy purposes give us in fact much less information on what is really happening with the forest. They omit and silence the forest itself. None actually mention forest cover or the exact period and type of forest loss that happened. The various (para)maps only refer to undefined forest cover loss and silence different categories of forests and deforestation. It also does not refer to forest degradation – an explicit REDD+ concern regarding forest quality – nor to biodiversity or any alternative local definitions of forest (use). Yet, if we focus on cover loss, some GIS scholars (e.g. Ickowitz, 2006; Ickowitz et al., 2015; Molinario, Hansen, & Potapov, 2015; Samndong et al., 2018) – among whom some contribute to GFW – have shown that shifting cultivation has varying impacts on DRC’s forest ecosystem. Some of them in fact argue that “the majority of tree cover loss […] is accounted for by shifting cultivation onto previously farmed lands rather than new deforestation” (de Araujo Barbosa, Maschler, Bonfils, & Molinario, 2018, p. para. 5), hence showing different interpretations of land use change. Therefore, they specifically refer to the impact of shifting agriculture on forests as “forest disturbance” as its impact is not always permanent. In the same way, the generic definition of forest in DRC’s REDD+ official documents is an adaption from the definition provided by FAO and the Clean Development Mechanism, that is a minimum area of land of 0.5ha with a tree crown cover of minimum 30% with trees attaining a minimum height of 3m or trees with the potential to reach this height at maturity in situ. It can include (industrial) tree crop plantations like acacia, cocoa, rubber tree plantations (GDRC, 2018; UN- REDD & DRC-MECNT, 2012); the differentiation between these types of ‘forests’ is in fact not possible with Landsat remote-sensing technologies used for DRC’s forest monitoring23 (GFW, n.d.). GFW geospatial experts told me24 that they are therefore very careful with terminologies and use “tree cover loss” – instead of forests or deforestation – precisely because GFW/Landsat geospatial technologies and models – which are used by DRC’s REDD+ practitioners – are not able to differentiate between plantations and other types of forest areas. Interestingly, when CAFI stands for industrial logging and its low impact on forest landscapes, it is strangely much more careful with terminologies and highlights that “the loss of intact forest landscapes should be minimized for the reasons cited in the article but this should not be confused with deforestation” (CAFI-DRC, 2017c, p. para.1). Local understandings are, unsurprisingly, totally absent in any document.

Despite various understandings of (non-)forests, in DRC’s REDD+ discourse, forest is understood as a unified category of carbon stock, as is also made clear by another map (Figure 10) – the only one in the REDD+ national investment plan. In Figure 10, remote-sensing maps are transformed into a rough

23 However, there are increasing investments in the analysis of high-resolution satellite imagery to better detect tree plantations from natural forests. However, these efforts do not include DRC at the moment and have been concentrated on Brazil, Indonesia, Malaysia, Cambodia, Peru, Colombia, and Liberia (GFW, n.d.). 24 83-ISTP-SKY-20190806, 85-ITSP-SKY-20191011. 65

cartographic representation that performs the forest as scarce carbon stocks, i.e. simultaneously as a threatened and a possessable resource (Wood & Fels, 2008), while rendering all other elements invisible.

Figure 10. “Democratic Republic of Congo: a threatened carbon stock” (CAFI-DRC, 2015). Remote-sensed maps were transformed into a rough cartographic representation to render ‘carbon stocks’ and ‘threatened carbon stocks’ visible while making all other elements and nuances invisible.

Population is not defined either. Returning to Figure 7, the (para)map refers both to ‘population density’ and ‘spatial distribution of population’, (unsurprisingly) telling us that human presence and forest cover loss are correlated. Focused on Tshopo province, Figure 8’s epimap refers to “high population density” (FONAREDD RDC & UNDP, n.d., p. 13). Yet, as the document recognizes, densities largely differ from one zone to another: from more than 400 people/km2 in Kisangani to 3 people/km2 in more remote regions (FONAREDD RDC & UNDP, n.d.). The underlying narrative here is that shifting cultivation is in itself sustainable but became unsustainable because of a shortening of fallow periods in zones with strong demographic pressure. As in all REDD+ documents, no figures comparing past and current fallow lengths, nor ideal length are ever provided; a fact that Ickowitz (2006) similarly demonstrates in her critical reflections on shifting cultivation and deforestation in Tropical Africa, in which she demonstrates the little evidence of a dramatic and linear process of shrinking forests due to shifting cultivation. Then, adding to the confusion some other parts of the CAFI website – on which Figure 7 appears – as well as the REDD+ studies on deforestation drivers and strategy documents use the term ‘population growth’. These concepts have very different meanings and implications. The REDD+ strategy makes things clearer: the lack of family planning and education leads to uncontrolled population growth that “poses huge problems in terms of environmental and natural resources preservation, for forestry ecosystems among others” (DRC-MECNT et al., 2015, pp. 77, own translation). The strategy does not provide a further explanation for this causality. No political-economic or historical contextualization is provided regarding the resettlement of population around roads during colonial times, or conflicts forcing relocation, or about unequal access to resources, distribution of wealth or market access. Heterogeneous dynamics of resource use among a seemingly homogeneous ‘population’ remain hidden. Yet, as I show throughout this manuscript, such information could lead to very different understandings of deforestation drivers. However, the REDD+ synthesis of deforestation drivers in DRC supports the idea that the identification of the underlying drivers is very complicated and direct observation of these underlying drivers impossible (UN-REDD & DRC-MECNT, 2012, p. 25); meaning in the document’s vocabulary, that it is impossible to observe/represent them with geospatial technology, from an office in Kinshasa or from elsewhere in the world. ‘Population’ is the exception: it appears to be easily represented (geo)spatially and is largely depicted as a technically manageable matter of family planning.

So, REDD+ (para)maps through data aggregation simultaneously co-produce homogeneous and radically abstract categories of ‘forests’ and ‘population’. A pixeled forest carbon stock threatened by a growing population is all that remains. The ‘blank’ Congolese forest spaces are filled in with particular meanings and simultaneously erase both local understandings of forests and the complex spatiality of everyday life, i.e. daily life practices embedded in particular places that define specific resource use patterns, forcing a new sense of what forests could become through REDD+. Of course, these maps’ and visual representations’ symbolic power is the conjugation of several discourses. Historical and socio-political

affairs, as I show in the following section, permeate the representation of distant forest places. The cartographer cannot be (solely) held accountable that the maps are reshaped and reused far beyond his/her control.

2.4. A well-oiled discourse

The (para)maps’ propositional logic of a community-induced deforestation was continuously restated – especially at central level in Kinshasa – during our interviews with national and international REDD+ consultants in DRC and with various other state and non-state actors involved in or targeted by the program. In analogy to the maps and their distant gaze, the REDD+ process is very centralized in Kinshasa – “REDD+ is only Kinshasa” as an expert for the CAFI Programme Intégré Oriental told me25. However, the narrative on deforestation has already trickled down. Shifting agriculture is seen as irrational both for economic growth and for forest protection; as being previously ecologically sustainable and an anchored tradition as opposed to modernity26. Its continuation is very much painted as the result of economic insecurity, a lack of alternatives and a lack of know-how to do better, coupled with increasing demography – “3.03% every year!”, one REDD+ Provincial Coordinator insisted27. The idea which often came up in my interviews was that without economic security it is not possible to stop slash-and-burn agriculture, resonating very much with WWF’s report on the Maï Ndombe REDD+ ‘green development’ project (WWF Forest and Climate Programme, 2016, p. 8):

“Communities in Mai Ndombe face profound poverty, insecure land tenure, and lack of economic opportunity, all of which encourages the rapid exploitation of resources and discourages sustainable land use practices.”

On numerous occasions during (informal) conversations with international (white) consultants, the discourse of local peasants as victims of their lack of economic opportunity easily switched to a blatant blaming discourse that portrays ‘Congolese’ as only having a utilitarian vision of the environment in which forests have no intrinsic value and are only used to fulfil human needs. The communication/awareness manager of a REDD+-like conservation project – a white man/woman born in Belgium and raised in DRC – who organizes environmental education programs told me28: “Children still have innocence and curiosity, and they see the intrinsic value of the environment but adults… they just have a materialistic vision”. In this narrative, local communities’ dependence on forests for their livelihoods is tainted with patronizing moral and racial claims representing a historically dominant mode of thinking about conservation in which local communities simply do not understand the importance of nature and need

25 56-RCS-KIS-20180515. 26 14-RCS-KIN-20160225, 17-CSI-KIS-20160301, 28-PA-KIN-20160406, 63-ITSP-KIN-20180529. 27 72-CSI-KIS-20190520. 28 82-ITSP-YAN-20190530. 68

to be educated about it (e.g. increasingly through targeted environmental awareness plans – see below) or kept away from it (e.g. protected areas).

Such a stereotypical image of conservative and lazy peasants shone through in the discourses of some Congolese state actors and REDD+ programs’ coordinators. They frame the persistence of shifting cultivation as a matter of “people [...] just being conservative” and simply reproducing what their ancestors did as they have “mentalities that are very hard to change”.29 Or in a similar fashion: “it’s just easier to burn than to stay in one place”30, and using primary forests also allows peasants to avoid the difficult task of weeding in fallows and of constructing fences to protect fields closer to the village from free rambling small farm animals.31 Villagers, in this way, just “do not understand that it is for their own good to do intensive monocultural agriculture” as asserted by a REDD+ project assistant.32 In the villages around Kisangani already targeted by REDD+ activities, this well-oiled discourse of communities’ “behavioural problems” was also repeated by local opinion and civil society leaders who are selected and trained within REDD+ project activities, and by some villagers informed through REDD+ communication and awareness plans, including radio announcements (Kamanda Mangamfu, Mumponga, & Lisingo, 2013, p. 18). The cited culprits were systematically “the fields, makala and wood sticks33 collection”34 or population growth – although exact figures on shortening fallow durations were not once provided35 – as well as their lack of environmental awareness before REDD+ arrived. One of this REDD+ village opinion leader stated36:

“there will be consequences if we cultivate large plots of lands. We will finish the forest and be like in the desert. And then we will feel the consequences of climate change in our region […]. They say we need to save the forest, they gave us an earbashing: if you want to do a field in the primary forest, don’t go over one hectare.”

In a village targeted by the Isangi REDD+ Pilot project, another one said in the same accusatory line37: “what pushes us to destroy the forest is that we do not have the right agricultural method. […] We did not pay attention to the forest. It is for our children that we are now interested in REDD”. While before “they lacked wisdom”, REDD+, some of these communities’ members told me38, came to enlighten them on the importance of forests, how they put them at risk with their daily activities and how they should

29 16-CSI-KIS-20160301. 30 71-CSI-KIS-20190518. 31 21-PA-YAF-20160328. 32 8-CTSP-KIS-20150516. 33 Wood sticks are used mainly for house construction. 34 Interview 20-CCSO-ISA-20160321. 35 As an anecdote, a Provincial REDD+ Focal Point mentioned that at a time “one would leave the fallow for a very long time, 50 years even” [17-CSI-KIS-20160301]; a fallow duration that seems very unlikely. Some other people mentioned historical fallow durations of 10 years, without clarifying which exact period they refer to. 36 47-I-YAF-20160329. 37 2-IFG-YAE-20150520. 38 29-I-YALO-20160322. 69

behave in the future. “We heard on the radio that the forest is disappearing because of population growth”39. The particular carbon value of forests emphasized by REDD+ also started to permeate people’s understandings: forests are sources of “oxygen […] of good air”40 and converting a forest area into a field “will release like an atmosphere”41. On the contrary, the impact on deforestation of industrial companies in the zone of Isangi REDD+ project (PPRGII) was rarely mentioned except when the president of the local Agricultural Administration Council asserted that “even if an agricultural plantation deforests, it stays in a limited space while [local] population, they, are always on the move”42. Yet, a 8874ha oil palm plantation – of which only a part is currently harvested – and two logging concessions, one of which stopped its activities in 2009, were established in the project’s intervention zone. Villagers who were not “opinion leaders” for REDD+ were more likely to mention the destruction of their forests by industrial logging companies, comparing the large amount of trees these companies log to the limited amount they are able to cut due to a lack of the means to do so and poor access to the timber market. These sets of discourses, I believe, emphasize REDD+ “propaganda work” on deforestation more than the in-depth forest values people actually hold. In fact, as I detail in Part II of this thesis, understandings I collected on forest-related issues with communities outside REDD+ projects’ particular intervention zones were different and much more diverse, as some people had simply never heard of REDD+, the terms ‘climate change’ or ‘deforestation’ in certain cases.

A detour through colonial archives shows that this discourse was already developed under the Belgian colonial administration. As early as 1908, the colonial administration was indeed preoccupied with modifying indigenous farming practices that were seen as unproductive and destructive of the environment, including as from the late 1930s a particular concern for shortening fallow periods43 and its consequences on soil fertility due to an increasing population density around urban centres and increasing export agriculture (e.g. Clement, 1952; Malengreau, 1949). Colonial agricultural policies44 shifted from resettlement schemes and a system of compulsory cash crop cultivations to the late-colonial Paysannat Indigène* – usually translated in English as ‘Indigenous Peasantry’45 – supported by the Fonds du Bien-Être Indigene*46 that was concerned with the “moral and material raising of the natives” (De Wildeman, 1940; Malengreau, 1949; Moeller, 1936, p. 235). “Active propaganda”, notably through community leaders (notables), was recommended to convince peasants of the positive effects of

39 10-I-RBU-20160307. 40 31-I-YALO-20160323. 41 41-I-BU-20160326. 42 20-CCSO-ISA-20160321. 43 Again, in the colonial archives I collected, the historical average fallow period is never mentioned. However, ideal fallow period in the colonial Paysannat Indigène* system varied from 15 to 21 years (e.g. Clement, 1952; Malengreau, 1949). 44 See Clement (2014) for a summary of the evolution of agricultural policies in the Belgian Congo from 1885 to 1960. 45 In the remainder of this thesis, I keep to the French term Paysannat Indigène* to mark a clear distinction between this particular colonial system and the general ‘indigenous peasantry’. 46 Funds for Indigenous Well-Being. 70

changing their traditional agricultural modes (Clement, 1952, p. 142). This excerpt from a publication of the Belgian Ministry of Colonies that looks at the “problem of indigenous rural economy” (Dubois, 1952, p. 29, own translation) is a striking illustration of the similarity between REDD+ claims and colonial ones:

“Whoever knows the customary agricultural methods applied by indigenous peoples for many years knows that those ones, while they were of much use to ensure the food subsistence of their producers, or even of those who live in industrial and extra-customary centres, no longer meet the ceaselessly increasing present needs and that besides, they lead to a threatening destruction of natural resources. […] The protection of soil capital as well as improvement in productivity have appeared as an inescapable necessity.”

In chapter 3, I come back to some historical underpinnings of REDD+ approach to the control of local agricultural practices.

2.5. A convenient reality in a political ‘messiness’

The geospatial narratives of the forests threatened by peasants are not only historically embedded but should also be understood contingently in the current political context of DRC. These narratives appear as a convenient, apolitical and stable point of consensus for all actors in a political context characterized by constant reshuffle, contestation and (re)negotiations that often hinders policy implementation (Englebert & Tull, 2013). While bad/weak governance, corruption, lack of government accountability and institutional issues are often raised as drivers of deforestation – and often as justification for non-state control/hybrid governance, see chapter 3 – these are first and foremost political matters which are by definition much less consensual. In this way, there is a strong lack of intersectoral and interministerial collaboration and coordination that, for instance, leads to the overlapping of different types of land-use concessions, e.g. mining or logging attributed in protected areas. However, this incoherence and uncoordinated situation guarantees “the power and the sources of income of individual ministries” (Wehkamp, Aquino, Fuss, & Reed, 2015).

In a typical instance, in 2017, GFW alerted that an area of 300ha had been cleared in the concession of a foreign logging company located close to Kisangani. It appeared that an influential Congolese businessman from Kinshasa and apparently close ally to the then president Joseph Kabila, was developing a large-scale industrial oil palm plantation in an area whose access has been facilitated by logging roads47. When I visited the agricultural concession in 2019 – with the authorization of one of the owners48 – a well-informed manager in the concession told me49 it had been enlarged to 1000ha and was aimed to be enlarged by another 3000ha by the end of 2020. The concession’s development, he continued,

47 59-PA-KIS-20180521, 63-ITSP-KIN-20180529. 48 70-PA-KIS-20190518. 49 74-PA-RIT-20190524. 71

benefitted from foreign investment and more particularly, from the secretive American Blattner family’s Congolese business empire.50 The owner I met admitted his agricultural plantation is inside the perimeter of a logging concession but said that his own company would be much more beneficial to communities. This is a well-known case among REDD+ stakeholders in Kinshasa who have clearly expressed their frustration. However, no legal action was and can be taken – which the manager of the logging company confirmed during an informal conversation in May 2019. Yet it does not inflect REDD+ mainstream discourse on deforestation or the appropriateness of REDD+ strategies and collaborations in DRC. Rather, this case as well as others is often used to explain how difficult it is for foreign logging companies to work in such a political environment and therefore to legitimize REDD+ investments in these logging companies and in increased spatial (and in fact political) control (see chapter 3).

Similarly, there have been important tensions between international donor and technical stakeholders, some CSOs and the DRC government regarding logging concessions and the lifting of the 2002 forest moratorium, which prevents the attribution of new concessions. International stakeholders and some CSOs51 oppose the immediate and unconditional lifting of the moratorium that the government is pushing for, arguing that the necessary conditions of good governance have not yet been reached. REDD+ and its investments in Sustainable Forest Management for operating logging companies – I talk about this program in section 3.3 – are meant to lead to the lifting of the moratorium. REDD+ has indeed been strongly tied to the vast reform of the forestry sector in DRC that has been underway since 2002, with the drawing up and promulgation of a national Forest Code. This reform has largely been pushed and financed by the World Bank who counted on the formalization of the forestry sector to reflate the country's (green) economy (Mpoyi et al., 2013; Trefon, 2008). As in other Congo Basin countries, the reform already laid out a decentralized and participative forest management and required companies “to submit management plans that specify socially and environmentally sustainable practices” (Clay, 2016, p. 134). While the application of the Forestry Code has been largely non-existent for a long time, REDD+ is a way to reactivate it and further revive the sector and its rational management. In this way, “the main private sector interest in REDD+ discussion has been from two industrial corporations in the wood sector: the Fédération des Industriels du Bois (FIB) and the Fédération des Entreprises du Congo (FEC)” (Mpoyi et al., 2013, p. 43), who according to RFUK52 were also interviewed as ‘experts’ for the REDD+-sponsored

50 It is difficult to assess the real extent of the Blattner’s family business and control of the Congolese economy. They own at least two groups: Groupe Blattner Elwyn and Safricas. Groupe Blattner Elwyn’s website mentions activities in the agroindustry, logging sector, logistics (transport) and telecommunication. However, a quick search on Google rapidly shows that Elwyn Blattner also owns or used to own (majority) shares in other companies such as the former BIAC – a bank that went bankrupt in 2017, for which Elwyn Blattner briefly went to prison – and the mining company Sodexmines. Elwyn Blattner’s brother and his group Safricas-Congo S.A. is similarly involved in various sectors of the economy (construction, transport, agriculture) and owns, among other businesses, the airline company CAA. Some of the information I display in this Part I was, however, obtained through my interviews. Due to this fuzziness, I refer to the “Blattner Conglomerate” in the remainder of this first part. 51 Among ‘the’ Congolese civil society there is, however, a specific coalition of CSOs supporting the immediate lifting of the moratorium (Coalition de la Société Civile pour la Levée du Moratoire*). 52 50-INGO-SKY-20180423. 72

studies on deforestation drivers. Similarly, the Ministry in charge of forestry is also in charge of REDD+ along the Ministry of Finance.

While the discussions regarding the moratorium have been tense, the point all parties at the central level agree on and legitimize with the maps and geospatial visualizations reproduced above, is thus that industrial logging does not lead to deforestation and if and when it encourages conversion to agricultural land, “it is true only locally”53. For ministers at the central level, assigning logging concessions can represent much larger financial interests (rent-seeking behaviours) than supporting local livelihoods and artisanal logging, as a high-ranked official at the Tshopo Provincial Ministry of Environment confessed54. Conversely, at Provincial level, artisanal logging can be much more profitable as cutting permits are delivered by Provincial Governors and as legal and illegal taxes can be directly collected at this level. In such a context, while Congolese CSOs constantly challenge state institutions’ legitimacy to make decisions regarding forest policies and REDD+, the large majority of them never question the simple mainstream narrative of deforestation precisely because it appears to be depoliticized. Focusing on communities’ practices seems an easy way to move REDD+ forward while insisting problems of governance are tackled as a priority, both at government and corporate levels, would be a much more tedious and longer path. Most major national CSOs strongly support the REDD+ process partly because their emergence as much as their survival depends on it: the REDD+ coalition of Congolese CSOs (GTCRR) has been directly created and financed through REDD+ programs.55 Conversely, the participation of Congolese CSOs – as ‘experts’ who, typically, are urban elites – in one of the REDD+ studies on deforestation drivers is often brought up in conversations by both Congolese forest state administration and REDD+ practitioners to legitimize the orientation of the programs. The involvement of civil society and resource (expert) informants in the production of the REDD+ study and strategy serves as a powerful justification to refute alternative understandings of deforestation and to implement programs. Geospatial expert-based studies and satellite maps were thus almost systematically called upon as a neutrality label, normatively superior to the socio-political discussions I just described. Greenpeace DRC and other international NGOs like Global Witness or RFUK who, as mentioned earlier, are the only ones contesting the narrative of community- induced deforestation and strongly oppose the lifting of the moratorium are conversely perceived as “blocking the country’s [REDD+] progress”.56

53 28-PA-KIN-20160406. 54 53-CSI-KIS-20180430. 55 For an account of the emergence and role of Congolese CSOs see Ehrenstein (2014). 56 61-CCSO-KIN-20180528. 73

2.6. Logoization

Local communities are in this way simultaneously portrayed as easy apolitical culprits and victims that are desperately in need of a REDD+ green development intervention while past, present and future industrial exploitation and real governance are largely let off the hook. This narrative, I have shown, has emerged from a mutual co-constitution of remote sensing data, maps and a whole regime of (expert) claims tied to a particular socio-political and historical genealogy. What I aimed to show here is not whether the representations contained in REDD+ policy maps are fundamentally true or untrue – in my view they activate one of multiple realities. Rather, I emphasized how maps and REDD+ discourses mutually and contingently unfold in context to make a single coherent reality of deforestation emerge, which not only enables the planning of certain actions of population control and modernization but also the resolution of other relational problems between stakeholders with different interests. In actor- network theoretical terms, this fact is highly networked and therefore real and authoritative, i.e. successful, while effectively refuting other alternative interpretations (Callon, 2006; Latour, 1987; Law, 2004). REDD+ assemblage of visual representations and maps impose a homogeneity or “unified spatial ontology (this is there)” (Kitchin et al., 2009, p. 13) of people and forests in a political economic context characterized by contestation, to facilitate governance and control, and (unconsciously) reinforce a discourse born and propagated during the colonial period. As Wood and Fels (2008, p. 7) assert, “insisting that something is there is a uniquely powerful way of insisting that something is”, in the same way that silences and omissions of maps represent a deliberate and political choice rather than an innocent one. In this sense, REDD+ maps of Congolese forests emerge as quasi “logos” – “map-as-logo” – (Anderson, 1983/2006, p. 175) that secure an appearance of purity with all socio-political and geographic features and heterogeneity removed from it and, that are disseminated in different forms through various mediums, e.g. policy documents, online platforms or awareness-raising programs. As I have demonstrated, it guides how people see, understand and act in the world (West, 2016).

In this set of both conscious and unconscious, ideologically grounded ‘rhetorics of representation’ (West, 2016) and totalizing logics, it appears very clearly that local communities’ understandings of forest use and change, and of socio-political space have been completely ignored. As I discuss in the next section, geographical knowledge and planning is the mediator (Winichakul, 1994) behind every stage of conceiving and implementing DRC’s REDD+ strategy, of thinking, projecting and concretizing particular conceptions of a green development space that reshape local socio-spatial and socio-ecological identities, political and material sovereignty. As the logoization of forest socio-political space in REDD+ policy maps suggests emptiness and incompleteness – lived experiences and (more complex) socio-ecological identities are not acknowledged, people and forests are detached from their local geographical contexts – these ‘blanks’ become the material for refilling the map, that is re-charting and reorganizing the

territory from a distance (Hiatt, 2002). In a country where industrial exploitation of resources is still largely limited to extractive mining and logging – the latter remains a small sector in comparison with other tropical countries – and therefore more prospective than actual, the filling of these ‘blank spaces’ takes on an even greater importance. The REDD+ unidimensional simplified story of deforestation, I show in the next section, pushes a monoculture of eco-modernization, homogeneity, and control rooted in the (colonial) system of plantations – what Haraway (2015) has referred to as Plantationocene – to the detriment of communities’ authority over land and their particular socio-ecological relationships but to the advancement of the economic benefits of powerful actors.

3. Spatializing the eco-functional plantations

“[…] the development of various productive sectors of the economy (agriculture and cattle breeding, extractive industries, logging, etc.) in DRC is capital. The challenge consists of reaching economic and social objectives while controlling forest loss. In a context of acute pressure tied to the expansion needs of these various sectors and to population growth, spatial planning is a key tool. It will allow the definition of a process of optimal, rational and coherent use of space and resources, in particular forest resources” FONAREDD RDC and UNDP (2016, p. 8, own translation)

As repeatedly asserted in the DRC’s National REDD+ Strategy and Investment Plan, and in my interviews the ‘national consensus’ on deforestation drivers served as the basis to establish the country’s REDD+ overall direction. Direct and underlying deforestation drivers (see Figure 5, blue and red columns) have been plainly translated into eight pillars for interventions or “outcomes […] aimed at generating emissions reduction and development co-benefits” (CAFI-DRC, 2015, p. 26, own translation). There are four sectoral pillars supported by four transversal ‘enabling’ ones. The sectoral pillars include: 1) sustainable agriculture, i.e. settle and intensify agriculture; 2) sustainable wood energy production and development of alternative sources of energy; 3) sustainable forest management, i.e. formalization and development of sustainable production forests for a green socioeconomic development and; 4) mines and hydrocarbons exploitation. The transversal pillars include: 1) improved governance; control of population growth; 2) land use planning and; 3) and land tenure securitization (CAFI-DRC, 2015; DRC-MECNT et al., 2015). The rationale is to facilitate DRC’s forest transition by combining the implementation of major structural reforms in terms of spatial land-use planning and land tenure and, the development of multi- sectorial, industrial and economic activities and the promotions of public and private investments (CAFI- DRC, 2015; DRC-MECNT et al., 2015; FONAREDD RDC & UNDP, 2016).

3.1. Geospatial planning as governance

DRC’s national REDD+ strategy thus goes beyond conserving, capturing and enhancing carbon stocks in trees and rather represents an evolution towards a landscape approach that presents itself as a tool for spatial planning necessary for piloting a green development combining “Congolese forest preservation, economic growth and the development of the Congolese people” (DRC-MECNT et al., 2015, p. 36). The CAFI program, as one of its stakeholders told me57, is “really a rational approach, to reinforce control”, it is a “big zoning and land use plan for development”, a vision that was often repeated in my interactions with other stakeholders. While in Latin American and Southeast Asia – where the expansion of commodity-driven intensive/industrial agriculture, cattle ranching and logging are the most significant

57 49-PA-SKY-20180419. 77

deforestation drivers – the landscape approach to REDD+ has been developed out of a need to incentivize agribusiness and logging companies to reduce their impacts on forests, in DRC the rationale appears to be the reverse (Turnhout et al., 2017; Weatherley-Singh & Gupta, 2017). Here, since the main identified deforestation driver is shifting subsistence agriculture, the goal is to supervise and incentivize local rural communities to settle their activities and maintain forest cover in order for DRC to be able to respond to the increasing private and international demand for commercial and industrial agricultural land, timber and ores (DRC-MECNT et al., 2015). The particular focus is on forestry and agriculture, i.e. sectors that are for now far less developed than the extraction of mineral resources (which, in DRC’s REDD+ documents, is barely mentioned and investigated for its impact on deforestation). My interviews with REDD+ stakeholders in Kinshasa and my analysis of REDD+ documents certainly revealed their awareness of the impact large-scale exploitation, palm oil in particular, will have on deforestation in the future. However, they also frame these activities as the only rational use of resources that can enable the country’s green development, contrasted with Congolese people’s “anarchic use of space and resources” (FONAREDD RDC & UNDP, n.d., p. 28) or “irrational exploitation of natural resources” in one CSO’s representative’s terms.58

The references to irrationality and the need for geographic planning, in a more implicit way, also point to Congolese bad state governance. Spatial planning and the creation of bounded spaces would in theory prevent the overlapping of different types of land-use concessions granted by different ministries (see section 2.4) and allow for better monitoring. The outsourcing of the planning and monitoring functions of the state to non-governmental agencies such as WRI and private actors, e.g. logging concessions, in cooperation with decentralized entities whose interests are often divergent from the ones of the central state, appears to be a way of circumventing the state.59 Efforts to establish top-down but non-state-led micro-zoning plans at landscape levels (in REDD+ integrated programs) in fact shifts to a certain extent the responsibility of monitoring and maintaining carbon stocks from the state to decentralized entities, local communities and private actors – see below. In a country where, in 2018, the state ‘illegally’ reinstated three logging concessions (650 000 ha) to Chinese companies and is likely to lift the moratorium in the next couple of years despite the non-compliance with sustainable forest governance agreements signed with international donors, this strategy seems to aim at reaching (speculative) carbon emissions reduction targets without relying on improvements in state governance. My interviews with several high-ranked officials at the Ministry of Environment showed that lifting the moratorium was a priority and was strongly tied to both a political will to assert the State’s sovereignty over international actors such as CAFI, and rent-seeking behaviours – as one of them said, “a man has got to seek a living”.60

58 1-PO-CCSO-KIN-20180423. 59 1-PO-CCSO-KIN-20180423. 60 66-CSI-KIN-20180529. 78

The REDD+ process in DRC is in fact particularly donor-driven while state and national ownership is very weak (also see Fobissie et al. (2014)). Accusations of ‘bad governance’ typically justify the hybrid governance framework (see Table 1) developed by REDD+ donors and partners, and the top-down socio- spatial engineering of this landscape approach to REDD+. While REDD+ programs somehow function at the margin of the Congolese state, they allow international demand – more specifically, Euro-American demand – for Congolese carbon, certified timber, ‘green’ cash crops to be satisfied in the future through ad hoc spatial control in specific landscapes. The involvement of the private sector has indeed increased gradually as the perspectives of REDD+ in DRC entering its investment phase became clearer (DRC-MECNT et al., 2015). Certainly, DRC’s REDD+ strategy is also about investing and developing rural communities but it fundamentally sees their socio-spatial activities and identities as a hindrance that needs to be reconfigured so as to become merely enablers – rather than producers – of green economic development, through their waged labour or as ‘enviropreneurial citizens’ (Baldwin & Meltzer, 2012) conserving forests and carbon for the sake of the global planet and humanity.

Controlling categorizations

In the creation of a bounded economic space, the constant use of essentializing discursive binaries such as artisanal versus industrial, artisanal versus formal, subsistence versus productive, slash-and-burn versus intensive, collective versus individual, revealed in my analysis of interviews with REDD+ implementers and of programs’ documents and legitimized by the mainstream narrative of DRC’s deforestation, seems particularly key. “We have never seen that the collective can manage anything correctly”, one (international) REDD+ main Technical Advisors said.61 In an attempt to explain the reliance on traditional subsistence agriculture rather than on cash crop/productive agriculture in the hinterland of Kisangani, a Provincial Coordinator of REDD+ programs told me62: “here people are passive […] they are less dynamic, entrepreneurial than in other regions”. The program document of the Programme Intégré Oriental implemented, among other regions, in the Tshopo Province similarly calls “the province’s economic model, just like the one of the rest of the country […] an economy of extractive rent rather than productive” that leads to significant forest loss (FONAREDD RDC & UNDP, n.d., p. 28). Such discourse is not restricted to peasant agriculture; Geenen (2012) has for instance analyzed a similar “formalization canon” (p. 323) in policies related to artisanal and small-scale mining in DRC.

In this way, in REDD+ integrated schemes, the creation of carbon value functions through land use zoning plans in which the fixed localization of perennial cash crop plantations and planned industrial logging is combined with the reduction and intensification of subsistence shifting agriculture and agroforestry carbon sinks (CAFI-DRC, 2015; FONAREDD RDC & UNDP, n.d.). Geographic planning at landscape and local

61 14-RCS-KIN-20160225. 62 16-CSI-KIS-20160301. 79

levels is thus at the forefront of integrated programs. The demarcation of land-use zones (e.g. permanent production forests, conservation zones, subsistence agriculture zone, agroforestry plantations, large- scale agricultural land), advocated by the REDD+ spatial-based approach to landscape operates with the same practices of extractive industries, for delimitating territory and resources and negotiating the circumstances of extraction, production and social benefits’ distribution (Clay, 2016). Looking at the preliminary land-use plan of the ‘flagship’ Maï Ndombe REDD+ integrated landscape program – which is currently by far the most advanced one in DRC – illustrates particularly well the eco-functional spatial model REDD+ has adopted (Figure 11). As in the colonial maps in Figure 2 and Figure 3, spaces are abstracted into categories of land cover – note, again, the dark green colour to represent logging concessions – and organized into uniform areas of production and homogeneous groups.

In what is in fact a very complex plan that tackles 7 of the 8 pillars of REDD+ strategy, private conservation concessions operated by both a conservation company (Wildlife Works Carbon) and three logging companies operating in the area, are ‘integrated’ with logging exploitation, industrial agroforestry plantations operated by a private actor and, small-scale agro-systems managed by individual small farmers and in which perennial cash crops (oil palm, coffee, cocoa, rubber) are developed. Such mapping of an integrated landscape neglects power inequalities between actors and does not make any distinctions between what place belongs to whom and how it was accessed which makes lands grabbed for green or extractive purposes justifiable, as part of a unified landscape (Anderson, 1983/2006). In the next two sections I unpack this socio-political space. I first focus on communities and the reconfiguration of their socio-spatial relations into ‘enviropreneurial’ (Baldwin & Meltzer, 2012) commodity producers/labourers, and then on private actors – in particular logging companies – and their representation as green social agents.

Figure 11. Preliminary land-use plan and key REDD+ program activities in the Maï Ndombe province (FCPF, 2016).

3.2. (Re)creating the enviropreneurial commodity labourer

The focus of REDD+ integrated programs on perennial cash crops, and in particular oil palm, cocoa, coffee and rubber tree (hevea) stems from the claim that “their development is in practice inescapable in the face of the continuously growing world demand” (CAFI-DRC, 2015, p. 131). Typically, the rationale is twofold. First, it claims that perennial cash crop agriculture requires 5 to 8 times smaller land areas compared to short cycle shifting agriculture – again, no figures with actual and ideal fallow lengths are ever provided – while also absorbing part of the workforce which would normally turn to subsistence shifting agriculture. Second, for equal labour, a household which cultivates one hectare of perennial crops on its land would earn 1500 to 2000 dollars a year compared to 400 to 700 dollars a year with subsistence crops (CAFI-DRC, 2015). Cash crop plantations are to be associated with agroforestry systems that combine intensified subsistence agriculture –rendered possible by appropriate rotations of monoculture and organic and chemical agricultural inputs – with fast growing acacia trees (used for charcoal production), as well as with afforestation/reforestation areas of fast growing tree species, in particular

acacias (DRC-MECNT et al., 2015). This agricultural system differs from common agricultural methods practiced by local peasants in the hinterland of Kisangani as they plant pêle-mêle*, that is several types of crops on the same plot of land, without organizing them in rows or blocs. They partly do so because of remaining tree trunks that render the organization in straight lines quite complicated. According to REDD+ logic, a monocultural, intensified rotational system would be better suited to providing income to farmers selling their surplus on local and urban markets and would thus help convince farmers to stabilize their practices in addition to the small incentives they receive by planting acacia and respecting a land zoning plan. In other projects (e.g. REDD+ Novacel Sud Kwamouth) the aim is for farmers to sell their production to a (usually monopolistic) industrial agricultural actor in the area, the latter hence being able to increase its production capacity without much investment while ensuring income for local farmers (DRC-MECNT et al., 2015; FCPF, 2016). Except for the result-based payments typical of REDD+, this scheme blatantly resembles the Paysannat Indigène* system implemented by the Belgian colonial administration in the 1940s and 1950s, after decades of compulsory cultivation schemes63 – cultures obligatoires* (Clement, 2014). In fact, several state actors and REDD+ stakeholders referred to Paysannat Indigène*.64

Paysannat Indigène* was an integrated rural development scheme that grouped farmers together along delimited (uniform) cultivation blocs, and monitored production and fallow periods according to scientifically pre-defined agricultural cycles. It was a blueprint model characterized by homogeneity, rationality and uniformity. Subsistence crops were associated with cash crops. As I alluded to in section 2.4, this system had various related goals. From an economic perspective, it aimed to increase productivity and channel agricultural production, not only for peasants’ wellbeing but also to provide food supply for cities and for the paid workforce working in extractive and agricultural industries, as well as to ensure the export production volume (coffee, cotton etc.) (Clement, 1952). From an imbricated socio-environmental perspective, this model of (semi-)settled monocultural agriculture was seen as the solution “to ward off the dreadful consequences of nomadism – translated into the destruction of forest resources reserves and soil erosion – and to gradually and as quickly as possible shift from an ill-defined collective property to an individual property protected by the State” (De Wildeman, 1940, p. 4). As Clement (2014) shows in his analysis of the implementation of the Paysannat Indigène* system in the Equateur Province in the 1950s, while participation in the scheme was in theory voluntary, it was still imposed on villages from above. Farmers did not participate in its design, and in practice it became a more or less coercive system despite its branding as “‘education’ in colonial new-speak” (Clement, 2014, p. 271). Ultimately, economic incentives for Congolese producers were also limited as they sold their

63 In reality, compulsory cultivation was never totally abolished and was upheld until the very end of colonialism, with just a mere reduction of days devoted to compulsory cultivation from 60 to 45 per year (P. Clement, 2014). 64 14-RCS-KIN-20160225, 44-RCS-KIN-20170529, 53-CSI-KIS-20180430, 71-CSI-KIS-20190518. 82

production “to monopolistic traders and European agricultural companies at prices that were fixed between the state and these companies” and that were largely to the disadvantage of the farmers (Clement, 2014, p. 276). Finally, Paysannat Indigène* interfered with customary land tenure rights through the imposed demarcation of cultivation blocs and allocation of fields to individual farmers. In this sense, it accentuated a process already long set in motion, of forcing farmer communities to remain in a bounded space, while many of them used to be itinerant with land tenure rights that were not static.65 The lack of knowledge and willingness to engage with the complexity of customary tenure fuelled an increase in land conflicts.66

As in the colonial Paysannat Indigène*, REDD+ integrated programs are pragmatic calls for shifting cultivators to catch up with Western agricultural standards and expert-driven alternative agroforestry that claims to be based on traditional agricultural practices. Both schemes invoke and perform the figure of the entrepreneurial commodity smallholder tied to his land, free from any imposed scheme while simultaneously using a discourse of participation and collective benefits. The two systems also seem to bear the same fundamental issues, stripping off the social and political content of agricultural systems. Typically, in participatory micro-zoning plans, zones are bounded into categories of use in spaces that were characterized by periodicity, extension and movement, and communities are categorized into strikingly homogenous tenure groups (e.g. Figure 12) ignoring both heterogeneity within so-called ‘communities’ and fluidity and complexity of ownership and authority in such spaces (see Windey and Van Hecken (2019) and part II). The distribution of REDD+ investments and result-based payments are supposedly based on the respect of these micro land use plans and rules (FCPF, 2016; FONAREDD RDC & UNDP, n.d.). On the one hand, such community boundaries might allow for the securitization of their land rights against (further) land grabs; a point that is often invoked by REDD+ stakeholders to justify their interventions. Participatory maps have in fact already been used by communities in DRC, supported by international environmental NGOs, to reclaim their rights against logging companies, for instance in the Safbois concession in Isangi territory where part of my fieldwork took place. On the other hand, local peasants lose the freedom to cultivate what and where they want. Moreover, they trigger conflict and marginalize and render invisible the numerous ‘migrants’ – some of whom have created villages in spaces that were considered vacant – who negotiate their access in time and space, often in an ad hoc manner (again, see Windey and Van Hecken (2019) for an example).

65 Interestingly, today’s deforestation hotspots targeted by REDD+ are often along the roads around which communities were forcibly resettled. 66 As my research revealed, the impact of Paysannat Indigène* on land conflicts and a deregulation of customary tenure systems can still be observed today in numerous land conflicts. 83

Figure 12. Participatory tenure map of the intervention zone of the PPRGII Pilot Project in Isangi, Tshopo Province (OCEAN ONG, n.d.).

However, except for such maps, REDD+ programs in the hinterland of Kisangani have yet to prove that they (will) work with whole community groups. In the case of the PPRGII pilot project, only a few households were selected to develop monocultural plantations. Informants repeatedly asserted that individual household heads participating in the agricultural program – local elites, i.e. notables* – had been chosen simply because of their close relationships with the implementing NGO’s program managers – themselves elites from Kisangani – or due to their affiliation with the dominant clan, which created disappointment and conflict in several communities.67 One of the dangers here lies in the contradiction between theory-propaganda and practice. Under the cover of participation and inclusion, REDD+ projects as in the case of PPRGII usually create participatory land tenure maps (e.g. Figure 12) that raise expectations that REDD+ will be implemented for the whole community and benefits be shared equally. Participatory land use maps are also supposedly created to manage communities around agreed-on land use goals. However, these maps rather appear as ‘maps-as-logo’ (Anderson, 1983/2006) in the context of information/awareness-raising campaigns (see section 2.4) and of the UN-REDD safeguard of Free, Prior and Informed Consent (FPIC) or, as Ehrenstein (2014) similarly notices, as a rhetorical exercise to meet the request of some international environmental NGOs for the systematic use of participatory tenure maps. In practice, projects advance a system of private property rights that aligns with the settlement and intensification of agriculture, cash crop and carbon production pushed by REDD+. It also facilitates control and accumulation, and strengthens social differentiation.

Both the CAFI and FIP PIREDD integrated programs in Tshopo have similarly focused their primary efforts on identifying former cash crop plantations that can be restored as well as ‘private’ smallholder farmers who already planted cocoa plants, coffee or oil palm trees or are interested to do so68, so as to support them technically and financially to develop green business models and transform them into a “professional class of small farmers” (FCPF, 2016, p. 54). These actors already have formal – rather than customary – individual land titles and the programs claim they will support those who do not, to access such individual land property (FONAREDD RDC & UNDP, n.d.). In the absence of land registers and a marked lack of clarity regarding any information on land ownership, identifying plantations and private landowners turns out to be complicated as it likewise was when I asked the programs’ managers to provide the list of private landowners that will be supported by the CAFI and FIP integrated programs. However, my on-the-ground research shows that most actors with land titles are urban (military, political and intellectual) elites from Kisangani and Kinshasa, and the Nande, an ethnic group from North Kivu whose members have escaped conflicts but have a major commercial power in Kisangani – they own most shops and businesses. These elites easily gain access to land by offering villages’ chiefs very low prices or compensation such as some chickens or a goat. Chiefs often agree without consulting the larger

67 E.g. 30-I-YALO-20160323, 31-I-YALO-20160323, 33-I-YALO-20160324. 68 38-CSI-KIS-20170505, 54-CSI-KIS-20180430, 72-CSI-KIS-20190520. 85

clan or village as the customary tenure law would normally require. With the agreement of the local chief, titles are then attributed by the cadastral administration.69 In some other cases, powerful individuals from the military or politics take over pieces of land by force.

Typically, these actors acquire land to develop cash crop agriculture, to produce charcoal and/or, in some cases, for artisanal logging; often combining all these activities. The size of land can vary from ten hectares to several hundreds of hectares, e.g. 700ha for the concession of a general in the Congolese army. In lots of cases, these lands are not yet exploited and are acquired for speculation, awaiting more financing to develop a plantation or to sell them later, e.g. to agribusinesses, when prices increase as a result of the city’s and infrastructure’s expansion.70 However, for the last couple of years, Nande people have been heavily investing in large cocoa plantations and, to a lesser extent, oil palm plantations, with the ambition of controlling these markets in the Province as they easily sell their product at a very low price to the only cocoa trader in the area, ESCO-Kivu71. ESCO-Kivu (Edmond Schlüter & Compagnie), a Swiss-registered agribusiness company – the quasi monopolistic cocoa trader in Eastern DRC – has boosted the revival of cocoa in North Kivu and Ituri by directing farmers towards its production, and now became the official partner of REDD+ programs in Tshopo Province (ASSECCAF, 2016; FONAREDD RDC & UNDP, n.d.). What is not mentioned in any documents though is that the coffee supplier Schlüter SA, of which ESCO-Kivu is a (discrete) subsidiary, was acquired in 2016 by OLAM – the giant Singaporean agri- commodity group. OLAM, through the established networks of Schlüter, has ambitions to develop its market share of speciality coffee from East Africa in both Europe and the United States (Terazono, 2016).

For private urban elites converted into plantation owners and for ESCO-Kivu/OLAM, REDD+ could hence become an opportunity to start or enhance intensive cash crop production on currently unexploited lands for supposedly helping protect DRC’s carbon sinks. While this could be positive for Congolese national GDP –assuming the tax system is functioning – and for creating revenues and employment, this contrasts with CAFI and FONAREDD discourse of community involvement and participation and tenure securitization, and in fact also with the discourse of forests’ protection. REDD+ programs do not engage with the complexity and fluidity of local land tenure systems. In practice, local peasant communities – shifting farmers – do not have legal rights to land and their access is now subordinated to the interests of the global economy, in terms of its demand for cocoa, palm oil and carbon. Lots of local peasants in the hinterland of Kisangani indeed ‘rent’ their land from the community that is deemed to be the first occupant and have no prospect of any securitization. Finally, because of food insecurity and limited means to pay labourers to help maintain the plantation, local peasants are largely prevented from

69 Therefore, conflicts over land are extremely frequent in DRC, both because customary law and positive law are often at odds and because of communities’ internal power struggles. 70 56-RCS-KIS-20180515. 71 88-I-RAB-20180516, 99-I-KIS-20190517. 86

investing in cocoa plantations, lots of them told me. This set of issues was raised in two focus groups I conducted in two villages in which a small number of peasants planted cocoa trees with the support of a bigger producer who owns a large plantation in their landscape – see below.72 Moreover, local purchase prices for cocoa produced in the region of Kisangani remain volatile, from $0.90 to $1.50 per kilo depending of the harvest season, as three bigger producers told me.73 It thus adds a layer of risk for local farmers and communities to invest in cocoa. In this way, at the moment, the main cocoa producers in the hinterland of Kisangani are rarely the same shifting farmers REDD+ programs are supposedly supporting. While REDD+ programs claim to aim to strengthen existing channels and value chains, their partner ESCO- Kivu/OLAM is the same (quasi-) monopolistic company that buys all cocoa production in the targeted areas in the Tshopo Province. The fact that some more powerful operators with large plantations already provide cocoa at current market prices set by ESCO-Kivu/OLAM might complicate this objective and in practice prevent local peasants from entering the market, or just force them to sell at low prices to bigger actors. For instance, in the area of Alibuku, 36km from Kisangani, one important Nande producer is already supporting some local peasants to engage in very small-scale cocoa production and sell their production back to him both to ensure social peace, i.e. “avoid local peasants to steal the production”, and to increase his market share.74 Alternatively, in zones of dense cash crop exploitation, local peasants – in particular ‘migrants’ who have to negotiate their land access with local chiefs as is the case now – might just have to become (occasional) labourers in private plantations for maximum $1.50 to $2.00 a day.75

Regarding the risk cash crop plantations present for forests, the impact of the plantations I visited are already visible on the ground and on the GFW tool, in the “hotspots” identified by REDD+. Some large areas of secondary forests have been cleared for these monocultural plantations installed by more powerful urban elites, the Nande and others. In fact, exploitation surface areas tend to increase with people’s growing financial means and easy access to individual property rights. Some plantations are growing fast – over 200ha for one I visited – and are planned to increase as land is still easily (and cheaply) granted by local communities. The number of Nande starting new concessions is also constantly increasing as in North Kivu and Ituri, competition for land is stark and prices are high.76 In correlation with this movement and as I detail in Part II – chapter 6, poorer local cultivators often seek new cultivation areas further away in the forests as they sell their lands closer to road infrastructure.

REDD+ rationale with its essentialising images of ‘population’ and ‘shifting agriculture’ thus not only purposely overlooks the diversity of the Congolese socio-political landscape but also performs particular

72 15-IFG-ARBU-20190522, 16-IFG-RAB-20190522. 73 88-I-RAB-20180516, 92-I-ARBU-20180519, 99-I-KIS-20190517. 74 88-I-RAB-20180516. 75 88-I-RAB-20180516, 92-I-ARBU-20180519. 76 99-I-KIS-20190517. 87

landscapes of commodity production that affect socio-spatial relations and identities. Local shifting peasants are now supposed to provide the world with commodities (carbon and cash crops) through the reconfiguration of their socio-spatial relationships within the dynamics of globalization and its economic relations (Arora-Jonsson et al., 2016). In fact, they are mainly excluded from entering such markets and dispossessed from their lands. My point is of course not to dismiss the importance of thinking of various strategies that can reduce rural poverty and deforestation – I will come back to this in my general conclusion. Rather, it is to point out how ‘geographic programmation’ and rationalisation as it is conceived now discards the full complexity of social identities and of people’s lives which ironically serves as a vehicle for their misrecognition, and dismisses crucial power differences between actors and between different access to resources in the landscape (Fraser, 2000). In REDD+ projects, the entangled push for the ‘stabilization’ and rationalization of agriculture and for international market integration perpetuates colonial logics of concentration and exclusion through status and social subordination; with a class of restructured and formalized private urban landowners and giant commodity traders (e.g. ESCO- Kivu/OLAM, Blattner Conglomerate) controlling the production, and another class of smaller stigmatized ‘shifting cultivators’ becoming dependent casual labour reserves or petty commodity producers at the margin of (industrial) plantations (Fraser, 2000; Mosse, 2010). It thus creates new political arenas around new resources, further increasing inequalities, social conflicts, and rent-seeking behaviours. Perhaps even more worrisome than the growing inequalities between these groups, is how REDD+ reinforces processes of land/green grabbing and accumulation to the advantage of bigger, international industrial companies in these new green economic landscapes.

3.3. Advancing the (speculative) socially responsible and green company

Nowhere else in the world could I have been able to do what I've done, at 33 years of age. That's why I call this the wild west. If you're willing to put yourself on the line, it's wide open out here. Elwyn Blattner, 20 December 1989, in an interview for The New York Times article “Getting Rich in Zaire: An American, 33, Tells How” (Noble, 1989)

In DRC’s REDD+ strategy industrial actors’ role in ‘carbon sinks’ management is put at the forefront, raising real concerns about social and environmental justice as local communities become only peripheral to the dominant industrial economy in integrated landscapes. The example of OLAM/ESCO-Kivu I mentioned in the previous section is particularly representative of how DRC’s REDD+ investment plan supports powerful agri-commodity companies to gain market shares through organic/responsible commodities (e.g. cocoa) labels that consumers in Europe and the US are increasingly fond of. In the same way, and as I have shown earlier, the logging sector has also been particularly involved in REDD+ discussions and has become a privileged actor for investment and result-based payments. Industrial

forestry and REDD+ are in fact so entangled that there are serious conflicts of interest for forest engineering consulting firms who provide technical assistance to both industrial logging companies and international forest governance programs like REDD+, as well as to the Congolese forest administration. A French firm I interviewed on several occasions, which provides services to logging companies to develop their forest management plans has openly lobbied for the integration of logging concessions in the REDD+ program so as to allow logging companies to financially benefit from REDD+ investments for their forest management plans and their ‘conservation zones’.77 This consulting firm has been involved in the design of DRC’s REDD+ strategy and investment plan and has now become one of the executive agencies for the Maï Ndombe REDD+ integrated/jurisdictional program in the Western part of the country.

In this context, a US$18 million Sustainable Forest Management Program (PGDF) of the French Agency for Development (AFD) has been approved within the REDD+ investment plan. The overall objective of the program is to develop a legal, formal and sustainable logging sector by backing different actors (industrial, artisanal and community forests) in adopting the Sustainable Management Method and supporting the elaboration and application of a forestry policy (FONAREDD RDC, 2016). PGDF is in fact the continuation and extension of the AGEDUFOR program that was already financed by the AFD for five years to support forest governance and the formalization of the industrial logging sector. Both programs can be understood in the larger framework of the Forest Law Enforcement, Governance and Trade (FLEGT) action plan of the European Union that aims to halt trade in illegal timber and thus reinforce control, and of other environmental labels such as the Forest Stewardship Council (FSC). Due to frequent campaigning action, the industrial logging sector in the Congo Basin suffers from a negative image that has induced a drop in European/US demand for African timber, and in particular DRC (Karsenty, 2016a). Cleaning up DRC’s industrial logging sector is a necessity. However, it raises issues in terms of fairness of competition, and social and environmental justice. Both the obtaining of labels and the adoption of the FLEGT scheme by DRC supported by REDD+ would make it easier for industrial companies operating in DRC to win or extend their market share on the EU market.78 But the overall reinforcement of forest governance in DRC would also allow international logging industries to increase their market share on the large domestic market. As the PGDF program’s document asserts, at the moment, industrial logging companies are subject to both formal and informal taxes that hinder their competitivity against artisanal logging which is only subject to informal taxes (FONAREDD RDC, 2017). With the formalization of the artisanal logging sector, the latter would also become subject to formal taxes and thus give a competitive advantage to the industrial sector. Finally, at the level of concessions, REDD+ offers the possibility for logging companies already advanced in their forest management plan (FMP) to be compensated for their

77 28-PA-KIN-20160406, 30-PA-KIN-20160407, 46-PA-KIN-20170529. 78 Most of DRC’s industrial export of logs is currently destined for China, while Europe remains for now the main market for processed products (Karsenty, 2016a). Asian markets for DRC timber exports will likely exceed the European market in a couple of years. 89

effort in (supposedly) reduced impact logging or the extension of their conservation areas (FCPF, 2016). The establishment of FMPs spatially organizes and rationalizes production through inventories, the geolocalization of high-value timber and optimization of harvest schedules. It also organizes different bounded land use zones for regulating communities’ access and resource use, and for ensuring the delimitation of conservation zones. It also allows for the definition of levels of socio-economic compensation for communities. Production zones are, in this way, theoretically ‘protected’ from community and other artisanal activity encroachments. Geospatial planning, again, is at the forefront of getting rid of wider spatial and socio-economic dynamics.

As denounced by a coalition of international NGOs – in particular Greenpeace and Rainforest Foundation UK – and to a certain extent Congolese CSOs (see below), the ecological, economic and social impact of an expansion of industrial logging is highly debatable (Bouessel du Bourg, 2017; Caramel, 2017a, 2017b; Greenpeace et al., 2017). By law, logging concessions have to draw up social responsibility contracts with communities that include socio-economic compensation for the communities located within the perimeter of the concession; i.e. a corporate-centred model that replaces the state in providing social services. Typically, the PGDF celebrates this model and legitimizes logging companies’ activities by emphasizing its role in community development, as opposed to failed state governance. In line with some arguments of this coalition of NGOs, my inquiry field research in one logging concession (see also Part II) showed that small healthcare centres or schools had been built but that because the operational costs such as teachers, health workers or maintenance are not taken care of, these buildings often remain empty or are abandoned after a few months or years. The few tens of thousands of dollars offered by the logging company in compensation cannot in fact replace proper state investment in social services. It also obscures wider structures of inequalities. In some cases, direct arrangements between the company and the local chief and elites have not benefited the rest of the community. Wages and living conditions for the few company’s workers remain also particularly low. Finally, in terms of its ecological impact, as already touched upon in chapter 2, industrial logging is very controversial in DRC, and it is recognized that it largely operates illegally. Yet CAFI still supports the program by advancing the argument that studies on deforestation drivers have shown the small role of industrial logging concessions on deforestation – especially if they have forest management plans as REDD+ promotes79 – in contrast with communities’ land use practices and ‘illegal’ artisanal logging (CAFI-DRC, 2017a). The basis of the program is thus not, and even cannot be, put into question and is seen as absolutely necessary for the country’s economic growth.80 In fact, my interviews with some of the Congolese civil society’s actors that are the most active in REDD+ negotiations, e.g. the program officer of the REDD+ Climate Working Group (GTCRR)

79 There is, however, an ongoing debate about the impact forest management plans have on deforestation, see (Brandt, Nolte, & Agrawal, 2016) The response: (Karsenty, 2016b) And counter response: (Brandt, Nolte, & Agrawal, 2018), as well as: (Cerutti et al., 2017) 8051-CSI-KIN-20180424, 60-RCS-SKY-20180523, 61-CCSO-KIN-20180528. 90

or the director of the Council for Environmental Defense through Legality and Traceability (CODELT)– and my participant observation at the round table of the civil society on forest policies in April 201881, show that over time most Congolese CSOs have also validated the general objectives of the program on the same basis.82 In these discussions, spatial planning and monitoring, which participants referred to as programmation géographique* in French, was raised as the sine qua non for the lifting of the moratorium, and the CAFI PGDF program “really touches on these aspects”, said the director of CODELT.83 CAFI’s support to the logging sector appears here again for CSOs, as I explained in section 2.5, a leverage to counter the state’s bad governance. The director of CODELT continued, in a quote that is illustrative of the arguments CSOs have developed:

“Planning decisions at provincial level reduces the omnipotence of the national government. When the management plan is [legally] approved, it binds the authorities. […] These are strategic and vital projects for the country. We, civil society, have very well understood that we need to continue with CAFI.”

The PGDF and forest management plans are not the only means through which some powerful industrial (logging) actors benefit from REDD+ strategies. For instance, in ‘integrated’ landscapes, logging concessions and industrial agricultural concessions are adjacent to conservation concessions – see Figure 11. While to an outsider they might appear unconnected, most of them are in fact owned and managed by the same actors as logging companies have simply transformed some of their production concessions into conservation concessions. In Maï Ndombe, for instance, the country’s largest operating timber company SODEFOR84, a subsidiary of the giant NordSud Timber mainly held by Swiss and Portuguese capital and based in Lichtenstein, has 11 concessions that represent more than 2 million hectares of forests (Global Witness, 2018; WRI & DRC-MECNT, 2010). Under the REDD+ program, 4 of these concessions are to be converted into conservation concessions for which SODEFOR can be financially rewarded for its carbon performance, counting as avoided deforestation. In the same way, one of the major commercial and industrial actors in DRC, the Blattner Conglomerate, that has been running businesses in diverse sectors including timber exploitation, agriculture – it is the biggest palm oil producer in DRC – and mining since just after independence, has increasingly shown its interest in REDD+. Since 2009, it has converted its 348 000ha logging concession, Safbois, into a conservation concession, Jadora – of which 187 000ha are primary forest – in the Isangi territory where (one of) its 8874ha palm oil tree

81 1-PO-CCSO-KIN-20180423. 82 As I showed in chapter 2, Congolese CSOs’ position in the REDD+ process is generally not one of resistance and opposition. 83 67-CCSO-KIN-20180529. 84 The forestry sector in DRC is an oligopolistic sector and largely controlled by foreign capital. Two companies – SODEFOR and one owned by the Blattner family conglomerate – are responsible for more than half of the total declared production and export volumes (Lawson, 2014). 91

plantations, Busira Lomami, is also established.85 The conservation concession is now a private REDD+ project co-owned by Jadora LLC, an American, privately held company selling carbon credits. While it is difficult to precisely define what logging actors’ exact interests are in converting their concessions besides speculation – at the moment logging seems more profitable than carbon selling – there are some potential lines of explanation. Going into WRI’s Forest Atlas of the DRC and one Global Witness’ report shows that some of SODEFOR’s concessions had no declared production or showed no sign of activity (through satellite imagery) for several years before their conversion (Global Witness, 2018; WRI & DRC- MECNT, 2010). In the case of Jadora/Safbois, the conversion from an exploitation to a conservation concession was strongly criticized by civil society and several NGOs. They contested the legality of the conversion and suspected it had been done to avoid the increasing scrutiny of and revolt against its illegal logging activities and the non-compliance with the social responsibility contract Safbois had with local communities (see e.g. Parenti (2007) who shows the extent of the discontent just a couple of years before the conversion).86 Some other critics claim that Safbois had been converted just because the company realized there were significantly fewer prime high value timber species than expected or that it had already harvested all of them, hence strongly decreasing the company’s revenues or even forcing its activities to stop.

In the meantime the Jadora/Safbois REDD+ project has obtained the Verified Carbon Standard (VCS) and Climate, Community & Biodiversity Standards (CCBA) certifications, and has been advertising its environmental and socio-economic accomplishments on the web for a few years.87 Yet, in the name of conservation and protection, local people have been further evicted from shaping their own landscapes and benefitting from any potential wealth carbon sinks could produce. In villages situated around the Jadora/Safbois concession people told me that while they have received some brief training sessions on the objectives of the conservation concession to sequester carbon and on alternative agricultural practices (rotational monoculture), they have barely received any compensation in the first 7 years of the Jadora concession’s existence despite being counted as beneficiaries in official evaluation reports.88 When I visited the area, there were indeed some demo fields and demo fish farming but they were used by Jadora’s two permanent workers rather by the supposed beneficiaries of the project. Communities’ access to the concession are nevertheless officially further restricted through the conversion from a

85 Blattner Conglomerate is also involved in the establishment of a new oil palm plantations in the CFT logging concession, as developed in section 4.1.2. It has also, in a quite opaque manner, sold five of its logging concessions (previously SIFORCO) to the Chinese company Booming Green. 86 In fact, the accusations of non-compliance with CSR contracts are rife in DRC (other concessions owned by SODEFOR and Blattner Conglomerate have been accused). The logging sector is also largely operating illegally, as regularly denounced by Global Witness and Greenpeace, and as I mentioned earlier. 87 E.g. https://www.jadora.co/what-we-do, http://www.coderedd.org/members-partners/jadora/ or https://standfortrees.org/en/protect-a-forest/isangi-congo-rainforest-conservation-project 88 42-I-YAF-20160328, 45-I-YAF-20160329, 46-I-YAF-20160329. 92

logging to a conservation concession, e.g. hunting or cultivating.89 No socio-economic compensation contract or other form of agreement seems to have been signed between Jadora and the local communities while they existed when it was still a logging concession.90 Communities often brought up the opacity of Jadora’s statues and activities that is related to carbon’s invisibility, as exemplified by this quote91:

“Safbois, we saw how they cut down our trees and took it on their boats to sell it in Kinshasa or abroad. That we could clearly see. But for the carbon gas, it is not visible. […] Why are some Jadora’s workers paid, on which basis? They do not produce anything, but they are regularly paid at the end of each month and they sell the gas at the level of the World Bank; that is where they get their benefits from. But here, we do not see it. […] So we cannot react because we do not know what they take away from us. If carbon gas was clearly visible or so we could protest, but we do not see it. How can we claim it?”

The only thing they saw, as other villagers told me, is that “when white people came, they entered the forest to look at the game animals, to look at the water, to look at the trees […] and put some devices there”92; “they put some red signs [on the trees] that say we cannot touch and if we touch America will get angry”.93 “Is Jadora a logger, is Jadora an NGO, so who is Jadora?”94 asked another one. Yes, as they rightly point out, industrial actors can now cash in on REDD+ to carry on and even bolster their business- as-usual extractive activities, e.g. through PGDF’s supported forest management plans, and engage in speculative (invisible) carbon trading through the conversion of concessions that were anyway not economically viable, simultaneously appearing as economic and environmental saviours. And yes, in DRC’s overly spatial approach, ‘landscapes of [capitalist] production’, ‘landscapes for speculation’ (the carbon stocks valorized by REDD+) and ‘landscapes for contemplation’ (e.g. protected areas) can coexist while leaving local communities who dwell in them at the margin (Igoe, 2013).

89 43-I-YAF-20160328. 90 42-I-YAF-20160328. 91 43-I-YAF-20160328. 92 47-I-YAF-20160329. 93 45-I-YAF-20160329. 94 46-I-YAF-20160329. 93

Conclusions

By adopting a relational understanding of mapping and geospatial planning that reconnects them to wider systems of power/knowledge and the social-material world, this Part I has demonstrated how historically hegemonic ‘rhetorics of representation’ (West, 2016) continuously reproduce a monoculture of productivity (de Sousa Santos, 2014) and bounded space materialized in privately-held plantations, through seemingly new and more inclusive environmental programs. I have argued that the idea of Congolese rainforests as ‘blank spaces’, i.e. not scientifically mapped (Hiatt, 2002), and the corresponding heavy REDD+ investment in and diffusion of expert geospatial analyses and representations has pushed western scientific interpretations while excluding local knowledge, struggles around resources and people’s definition of their own identities and social reproduction which I explore next, in Part II of this thesis. In multiple ways, cutting edge geospatial data, technologies and planning and the eviction of other epistemologies have become the first mediator to think, legitimize and re-order landscapes and lived lives into global ‘green’ commodity markets (e.g. carbon, cash crops), and for inscribing people’s subjectivities as (underpaid) plantation labourers or petty commodity producers at the margin of the dominant economy ruled by more powerful actors (Haraway, 2015; Lefebvre, 1991; Mosse, 2010; Winichakul, 1994). Landscapes are rewritten in a regulatory and extraction-driven manner.

Empirically, I have shown that satellite-based maps perform as neutral reference actants – secure representations within an epistemic regime of claims and meanings – to control people and space from a distance, seemingly out of the ‘messy’ Congolese socio-political world (Anderson, 1983/2006; Harley, 1989; Kitchin et al., 2013). Through a careful crafting that creates presences and absences, dominant mappings and representations in REDD+ assemblage contribute to the production of a unified spatial ontology (this is there) (Wood & Fels, 2008) of ‘deforestation’ that blames the livelihood practices of a seemingly homogeneous category of ‘local population’. Local communities’ subjectivities are uniformly framed as the poor but harmful shifting cultivators, a first step to justify control over their socio-spatial practices. Simultaneously the complexity of land use and power struggles over resources are obscured, and past, present and future industrial exploitation are mostly let off the hook. These knowledge claims, I have demonstrated, are embedded contingently within colonial discourses that stigmatized indigenous peasants in the same ways, and within the political context of DRC and of REDD+ negotiations. The figure of the harmful shifting cultivator and its neutrality label – it is proven with geospatial data – helps resolve relational problems between stakeholders engaged in REDD+ discussions with different interests. The contingent maps, in this context, emerge as logos, from which socio-political heterogeneity and inequality is removed, and that permeate people’s imaginaries of forests and of what is possible, plausible and desirable (Anderson, 1983/2006; Igoe, 2013).

Fundamentally, as the second chapter of this first part has shown, this monoculture of knowledge is intertwined with a monoculture of socio-tempo-spatiality (de Sousa Santos, 2014) that privileges a linear, sequenced and bounded time-space associated with the rationality of formal industrial and commercial activities. It corresponds to what is seen as a productive time-space that is legitimized in REDD+ discourses through its economic significance for the country’s green growth and its insertion into global economic relations. This uniform discourse simultaneously creates the non-existence of periodic, discontinuous and unbounded time-space associated with the irrationality of informal artisanal and subsistence activities. It is perceived as an incidental time-space in a double sense: as unproductive, backward and secondary for national growth, and as unplanned, disordered and uncontained in relation to the natural environment. On this basis, rights to land and forest resources and legitimacy to deforest (or not) are distributed. Clearing trees is acceptable as long as it is planned and bounded in space and time, like continuous and linearly growing industrial exploitation, something one international REDD+ Technical Advisor referred to as “accompanying deforestation”95, that is, a regulated or planned deforestation as opposed to communities’ unplanned deforestation (FCPF, 2016). As he further added, “cash crop plantations like oil palm or cocoa destroy the forest; but they do it once and for all while subsistence agriculture, that one, is insatiable”. Thus, it is also about the kind of crops that really count, which is intimately tied to the valued tempo-spatiality and socio-economic relations of rational modernity and the global economy. Current communities’ activities are seen as problematic because of their ever-expanding and disordered movements in an uncontained space while the intensive, continuous but linearly growing exploitation in a bounded space – that can be as big as a 250 000ha logging concession or a 9000 ha oil palm plantation – is legitimate. In fact, the technicity of REDD+ for measuring and monitoring forest areas, carbon stocks and emissions, which heavily relies on satellite remote-sensing technologies, inventory data and management plans, seem to directly privilege private gridded concessions with their homogeneous and planned activities. Surveillance of dynamic socio-ecologies through satellite imageries is much more difficult. Through these processes, the REDD+ landscape approach, despite its purported holistic way of looking at land use with special attention given to local, native and marginalized groups, continues to push Western techno-scientific knowledge to arbitrarily create green economic and ‘rational’ socio-spatial identities and practices.

By rendering local socio-ecological practices insignificant, blank in terms of their economic importance, places can be filled with other models. The particular valued productive, rational time-space promoted by REDD+ also produces standardized subjectivities and relationships. We have seen two figures emerging: the socially responsible green company and the enviropreneurial commodity producer/labourer. In DRC’s REDD+ imaginary, logging or agribusiness companies emerge as agents that

95 14-RCS-KIN-20160225. 96

are more capable than the irrational Congolese state and local peasants to protect carbon stocks and provide socio-economic wellbeing. This justifies investment in management plans and conservation concessions that allow companies to win or extend into new markets while restricting ‘the harmful shifting cultivators’ access to land and resources. Simultaneously, a class of private landowners and commodity traders is supported and further integrated in international markets (e.g. cocoa) while cultivators with now stabilized farming practices become labour reserves or petty producers of both cash crops and carbon. As such, REDD+ presents opportunities to rethink agricultural systems and for some to escape from poverty traps. But past and ongoing injustices are left totally unaddressed, as are the risks of expanding agricultural and extractive activities.

I believe that this analysis is important, as it sheds light on the various processes of “slow violence” of environmental rule, which causes delayed destruction and is dispersed across times and space (Nixon, 2011). It first manifests in the systematic absence of other (complex) modes of knowing, being and inhabiting the world. Local forest users never speak of and for themselves and are hence denied their ‘intellectual sovereignty’ and ‘representational sovereignty’ that is “the ability for a group of people to depict [and making meaning of and by] themselves with their own ambitions at heart” (Lewis in Raheja, 2010, p. 30; West, 2016). These forms of epistemic dispossession are closely imbricated with further material dispossession by elites or corporate forces and to “the subjection to capitalist law of the value of common goods and resources” (de Sousa Santos, 2014, p. 45). As people are treated as disposable and malleable, so are natural ecosystems. They are homogenized, rationalized and reorganized to enhance efficiency (e.g. of carbon production, of timber extraction etc.), with long-term socioecological consequences. Highlighting such slow violence is of the utmost importance in a country like DRC. First because its (almost uniform) portrayal as a place of failure and dysfunction (Ferguson, 2006) legitimizes external top-down green economic interventions and control. Second because, while there is a growing attention to commodity-driven deforestation in the Amazon, risks of displacement of commodity production in these (for now) more discrete landscapes are real.

Finally, to be clear, and as I have argued elsewhere (Windey & Van Hecken, 2019), I am not advocating against any use of geospatial knowledge in REDD+ environmental governance. I rather argue that if we understand the power of maps as relational rather than inevitable, cartography can start to bring in socio- environmental justice concerns to enact what was rendered invisible or marginalised by dominant representations of the world. To take up Mignolo (2009, p. 177)’s call, if knowledge-making is really concerned with people and environmental well-being rather than control for imperial interests, it should start “from local experiences and needs rather than from local imperial experiences and needs projected to the globe”.

References

Anderson, B. (1983/2006). Imagined Communities: Reflections on the Origin and Spread of Nationalism. London & New York: Verso Books.

Anderson, B., & McFarlane, C. (2011). Assemblage and geography. Area, 43(2), 124-127. doi:10.1111/j.1475-4762.2011.01004.x

Arora-Jonsson, S., Westholm, L., Temu, B. J., & Petitt, A. (2016). Carbon and Cash in Climate Assemblages: The Making of a New Global Citizenship. Antipode, 48(1), 74-96. doi:10.1111/anti.12170

ASSECCAF. (2016, October). Exportation Cacao RDC (secteur Beni), année civile 2015. Retrieved from http://asseccaf-cacao.blogspot.com/2016/10/exportation-cacao-rdc-secteur-beni.html

Baert, G., Ranst, E. V., Ngongo, M., & Verdoodt, A. (2012). Soil Survey in DR Congo – from 1935 until today. Paper presented at the Meeting of the Section of Natural and Medical Sciences, Gent.

Baldwin, A., & Meltzer, J. (2012). Environmental Citizenship and Climate Security: Contextualizing Violence and Citizenship in Amazonian Peru. In Environment and Citizenship in Latin America (1 ed., pp. 23-38): Berghahn Books.

Blaikie, P. (1999). A Review of Political Ecology: Issues, Epistemology and Analytical Narratives. Zeitschrift für Wirtschaftsgeographie, 3-4, 131-147.

Bouessel du Bourg, C. (2017, July 18). Un programme de l’AFD accusé par des ONG de menacer les forêts d’Afrique centrale, Press. Jeune Afrique. Retrieved from http://www.jeuneafrique.com/457159/societe/programme-de-lafd-accuse-ong-de-menacer- forets-de-rdc/

Brandt, J. S., Nolte, C., & Agrawal, A. (2016). Deforestation and timber production in Congo after implementation of sustainable forest management policy. Land Use Policy, 52, 15-22. doi:https://doi.org/10.1016/j.landusepol.2015.11.028

Brandt, J. S., Nolte, C., & Agrawal, A. (2018). Deforestation and timber production in Congo after implementation of sustainable management policy: A response to Karsenty et al. (2017). Land Use Policy, 77, 375-378. doi:https://doi.org/10.1016/j.landusepol.2018.04.056

CAFI-DRC. (2017a, September). CAFI et le secteur forestier en RDC: Foire aux questions. Retrieved from http://www.cafi.org/content/cafi/fr/home/events/drc-forests---frequently-asked- questions.html

CAFI-DRC. (2017b, September). CAFI et le secteur forestier en RDC: Foire aux Questions. FAQ 03: What contributes most to deforestation and forest degradation in DRC? Retrieved from http://www.cafi.org/content/cafi/fr/home/all-news/drc-forests---frequently-asked- questions/what-contributes-most-to-deforestation-and-forest-degradation-in.html

CAFI-DRC. (2017c, September). CAFI et le secteur forestier en RDC: Foire aux Questions. FAQ 06. What is the link between the loss of intact forest landscapes and logging concessions? Retrieved from http://www.cafi.org/content/cafi/fr/home/all-news/drc-forests---frequently-asked- questions/what-contributes-most-to-deforestation-and-forest-degradation-in.html

Cahen, L., & Lepersonne, J. (Cartographer). (1949). Carte Géologique [Geological map]. Retrieved from https://esdac.jrc.ec.europa.eu/ESDB_Archive/EuDASM/Africa/maps/afr_zr1001_ge.htm

Caramel, L. (2017a, July 24). Forêts du Congo: des scientifiques dénoncent à leur tour le projet de l’AFD. Le Monde Afrique. Retrieved from http://www.lemonde.fr/afrique/article/2017/07/24/forets- du-congo-des-scientifiques-denoncent-a-leur-tour-le-projet-de-l-agence-francaise-de- developpement_5164263_3212.html

Caramel, L. (2017b, June 26). L’Agence française de développement se défend de livrer les forêts du Congo aux grandes concessions. Le Monde Afrique. Retrieved from http://www.lemonde.fr/afrique/article/2017/06/26/l-agence-francaise-de-developpement-se- defend-de-livrer-les-forets-du-congo-aux-grandes-concessions_5151102_3212.html

Cerutti, P. O., Suryadarma, D., Nasi, R., Forni, E., Medjibe, V., Delion, S., & Bastin, D. (2017). The impact of forest management plans on trees and carbon: Modeling a decade of harvesting data in Cameroon. Journal of Forest Economics, 27, 1-9. doi:https://doi.org/10.1016/j.jfe.2017.01.004

Clement, J. (1952). Etude relative au paysannat indigène. In Royaume de Belgique - Ministère des Colonies (Ed.), Contribution à l'étude du problème de l'économie rurale indigène au Congo belge. Communications présentées par la Délégation Belge à la Conférence africaine de l'Economie rurale indigène (Jos, Nigérie - 17-24 novembre 1949) (Vol. XLIII, pp. 135-157). Bruxelles: Service de l'agriculture du Ministère des Colonies et du Gouvernement Général du Congo belge.

Clement, P. (2014). Rural development in the Belgian Congo: the late-colonial "indigenous peasantry" programme and its implementation in the Equateur District (1950s). Bulletin Des Séances, 60(2), 251-286.

Conrad, J. (1899). Heart of Darkness (Deluxe Edition ed.). New York: Penguin Books.

Crampton, J. W. (2004). GIS and Geographic Governance: Reconstructing the Choropleth Map. Cartographica: The International Journal for Geographic Information and Geovisualization, 39(1), 41-53. doi:10.3138/H066-3346-R941-6382

Crampton, J. W. (2007). Maps, Race and Foucault: Eugenics and Territorialization Following World War I. In.Crampton, J.W. & Elden, S. (Eds.), Space, Knowledge and Power: Foucault and Geography (pp. 223-244). Burlington: Ashgate Publishing Limited.

Crampton, J. W. (2010). Mapping: A Critical Introduction to Cartography and GIS. Oxford and New York: Wiley-Blackwell.

Crampton, J. W., & Krygier, J. (2005). An Introduction to Critical Cartography. ACME: An International Journal for Critical Geographies, 4(1), 11-33.

de Araujo Barbosa, C., Maschler, T., Bonfils, D., & Molinario, G. (2018, January 16). New map helps distinguish between cyclical farming and deforestation in the Congo Basin [Blog post]. Retrieved from https://blog.globalforestwatch.org/data/new-map-helps-distinguish-between-cyclical- farming-and-deforestation-in-the-congo-basin de Sousa Santos, B. (2014). Epistemologies of the South: Justice against epistemicide. Boulder, Colorado: Paradigm Publishers.

De Wildeman, E. (1940). Paysannat Indigène [Archive]. Congrès Colonial Belge, Ve session, Nr. 25, 1-46.

Dodge, M., Kitchin, R., & Perkins, C. (2011). Introducting The Map Reader. In M. Dodge, R. Kitchin, & C. Perkins (Eds.), The Map Reader: Theories of Mapping Practice and Cartographic Representation (pp. x-xxiii). Oxford: Wiley-Blackwell.

DRC-MECNT. (2008). Transmission R-PIN de la République Démocratique du Congo. Kinshasa Retrieved from https://www.forestcarbonpartnership.org/sites/fcp/files/DRC_R-Pin_Fr_Finale.pdf.

DRC-MECNT, FCPF, & UN-REDD. (2015). Stratégie-cadre nationale REDD de la République Démocratique du Congo. Version 3. Retrieved from Kinshasa: https://www.forestcarbonpartnership.org/sites/fcp/files/2015/March/Strategie- cadre%20nationale%20REDD%20de%20la%20RDC.pdf

Dubois, G. L. D. (1952). Politique agricole en milieux ruraux au Congo belge. In Royaume de Belgique - Ministère des Colonies (Ed.), Contribution à l'étude du problème de l'économie rurale indigène au Congo belge. Communications présentées par la Délégation Belge à la Conférence africaine de l'Economie rurale indigène (Jos, Nigérie - 17-24 novembre 1949) (Vol. XLIII, pp. 29-37). Bruxelles: Service de l'agriculture du Ministère des Colonies et du Gouvernement Général du Congo belge.

Englebert, P., & Tull, D. (2013). Contestation, négociation et résistance: L'État congolais au quotidien. Politique Africaine, 129(1), 5-22. doi:doi:10.3917/polaf.129.0005

Fairhead, J., & Leach, M. (1996). Misreading the African Landscape. Society and Eoclogy in a Forest Savannah Mosaic. Cambridge: Cambridge University Press.

FCPF. (2016). Emission Reductions Program Document (ER-PD): Mai-Ndombe Emission Reductions Program, Democratic Republic of Congo. Retrieved from: https://www.forestcarbonpartnership.org/system/files/documents/ERPD%20DRC%20final%20 draft%20Jan%202016.pdf

Ferguson, J. (2006). Global Shadows: Africa in the Neoliberal World Order. Durham & London: Duke University Press.

Fobissie, K., Alemagi, D., & Minang, A. P. (2014). REDD+ Policy Approaches in the Congo Basin: A Comparative Analysis of Cameroon and the Democratic Republic of Congo (DRC). Forests, 5(10). doi:10.3390/f5102400

FONAREDD RDC. (2016). Note d’Idée de programme1 au Fonds National REDD+. AMI n°4 – Programme de Gestion durable des forêts. Retrieved from http://www.fonaredd-rdc.org/wp- content/uploads/2017/06/NIF-FONAREDD-GESTION-FORET-AFD-05072017.pdf

100

FONAREDD RDC. (2017). Document de programme au Fonds National REDD+. Lot n°4 – Programme de Gestion durable des forêts. Retrieved from http://www.fonaredd-rdc.org/wp- content/uploads/2017/06/20170630_AFD_GESTION-DURABLE-DES-FORETS.pdf

FONAREDD RDC, & UNDP. (2016). Programme d’Appui à la réforme de l’Aménagement des Territoires. Retrieved from: https://www.cafi.org/content/cafi/fr/home/partner-countries/democratic- republic-of-the-congo/drc-fonaredd-programmes.html

FONAREDD RDC, & UNDP. (n.d.). Programme Intégré Oriental. Retrieved from: http://www.cafi.org/content/dam/cafi/docs/drc-documents/DRC- Approved%20Programmes/DRC-PI-Orientale-UNDP/DRC-UNDP-PIREDD-Oriental-Prodoc.pdf

Forsyth, T., & Sikor, T. (2013). Forests, development and the globalisation of justice. The Geographical Journal, 179(2), 114-121. doi:10.1111/geoj.12006

Forsyth, T., & Walker, A. (2008). Forest Guardians, Forest Destroyers: The Politics of Environmental Knowledge in Nother Thailand. Seattle & London: University of Washington Press.

Fraser, N. (2000). Rethinking Recognition. New Left Review, 3, 107-120.

GDRC. (2018). Niveau d'émissions de référence des forêts pour la réduction des émissions dues à la déforestation en République Démocratique du Congo. Retrieved from: https://redd.unfccc.int/files/2018_frel_submission_drc.pdf

Geenen, S. (2012). A dangerous bet: The challenges of formalizing artisanal mining in the Democratic Republic of Congo. Resources Policy, 37(3), 322-330. doi:https://doi.org/10.1016/j.resourpol.2012.02.004

GFW. (n.d.). Does GFW distinguish tree plantations from natural forests? Retrieved from https://www.globalforestwatch.org/howto/faqs/faq-does-gfw-distinguish-tree-plantations- from-natural-forests.html

Gilliard, A. A. (1953). Cartographie congolaise. Bruxelles: Institut Royal Colonial Belge.

Global Witness. (2012). L'art de l'exploitation industrielle au Congo: comment les compagnies forestières abusent des permis artisanaux pour piller les forêts de la République Démocratique du Congo. Retrieved from https://cdn.globalwitness.org/archive/files/lart%20de%20lexploitation%20industrielle%20en% 20rdc_lr.pdf

Global Witness. (2018). Total Systems Failure: exposing the global secrecy destroying forests in the Democratic Republic of Congo. Retrieved from https://www.globalwitness.org/en/campaigns/forests/total-systems-failure/

Global Witness. (2019). Buyers Beware: How European companies buying timber from Industrie Forestière du Congo risk falling foul of EU laws Retrieved from: https://www.globalwitness.org/pt/campaigns/forests/buyers-beware/

Greenpeace, Witness, G., Norway, R. F., UK, R. F., Naturelles, R. R., & GTCR-R. (2017). La Norvège et la France menacent les forêts de la RDC. Retrieved from https://www.greenpeace.org/africa/fr/publications-fr/665/la-norvege-et-la-france- menacent-les-forets-de-la-rdc/

Grove, R. H. (1995). Green imperialism: Colonial expansion, tropical island Edens and the origins of environmentalism, 1600-1860. Cambridge: Cambrige University Press.

101

Haraway, D. (2015). Anthropocene, Capitalocene, Plantationocene, Chthulucene: Making Kin. Environmental Humanities, 6(1), 159-165. doi:10.1215/22011919-3615934

Haraway, D., Ishikawa, N., Gilbert, S. F., Olwig, K., Tsing, A. L., & Bubandt, N. (2016). Anthropologists Are Talking – About the Anthropocene. Ethnos, 81(3), 535-564. doi:10.1080/00141844.2015.1105838

Haraway, D., & Tsing, A. L. (2019) Reflections on the Plantationocene: a conversation with Donna Haraway & Anna Tsing moderated by Gregg Mitman/Interviewer: G. Mitman. Edge Effects Magazin. Madison: Nelson Institute at the University of Wisconsin-Madison.

Harley, J. B. (1988). Maps, Knowledge, and Power. In Cosgrove, D. & Daniels, S. (Eds.), The Iconography of Landscape: Essays on the Symbolic Representation, Design and Use of Past Environments (pp. 277–312). Cambridge: Cambridge University Press.

Harley, J. B. (1989). Deconstructing the Map. Cartographica: The International Journal for Geographic Information and Geovisualization, 26(2), 1-20. doi:10.3138/E635-7827-1757-9T53

Hiatt, A. (2002). Blank Spaces on the Earth. The Yale Journal of Criticism: Interpretation in the Humanities, 15(2), 223-250. doi:doi:10.1353/yale.2002.0019

Huggins, C. (2010). Land, Power and Identity: Roots of violent conflict in Eastern DRC. Retrieved from https://www.international-alert.org/publications/land-power-and-identity

Ickowitz, A. (2006). Shifting Cultivation and Deforestation in Tropical Africa: Critical Reflections. Development and Change, 37(3), 599-626. doi:10.1111/j.0012-155X.2006.00492.x

Ickowitz, A., Slayback, D., Asanzi, P., & Nasi, R. (2015). Agriculture and deforestation in the Democratic Republic of the Congo. A synthesis of the current state of knowledge. Retrieved from http://www.cifor.org/publications/pdf_files/OccPapers/OP-119.pdf

Igoe, J. (2013). Nature on the Move II: Contemplation Becomes Speculation. New Proposals: Journal of Marxism and Interdisciplinary Inquiry, 6(1-2), 37-49.

Igoe, J., & Brockington, D. (2007). Neoliberal Conservation: A Brief Introduction. Conservation and Society, 5(4), 432-449.

Ingold, T. (2000). The Perception of the Environment: Essays on livelihood, dwelling and skill. New York: Routledge.

Kamanda Mangamfu, R., Mumponga, E., & Lisingo, J. (2013). Plan local de communication pour la promotion du projet pilote REDD géographiquement intégré d'Isangi. Kisangani: OCEAN ONG.

Karsenty, A. (2016a). The contemporary forest concessions in West and Central Africa: chronicle of a foretold decline? Retrieved from http://www.fao.org/forestry/45021- 04023cd52f4619cd28fe747b7e42c167f.pdf

Karsenty, A. (2018). La crise de la filière européenne du bois tropical en Afrique centrale. Retrieved from https://www.willagri.com/wp-content/uploads/2018/06/Dossier-Crise-Filiere-Europeenne- Bois-Afrique-Willagri.pdf

Karsenty, A., Cerutti, P., Doucet, J.L., Putz, F. E., Romero, C., Bernard, C., Eba’a Atyi, R. , Douard, P., Claeys, F. , Desbureaux, S., Ezzine De Blas, D., Fayolle, A., Fomete, T.,Forni, E., Gond, V., Gourlet-Fleury, S., Kleinschroth, F., Mortier, F., Nasi, R., De Wasseige, C. . (2016b). Do Forest Management Plans in Congo Lead to Greater Deforestation? Forets D’Afrique Centrale, Web.

102

Karsenty, A., & Ferron, C. (2017). Recent evolutions of forest concessions status and dynamics in Central Africa. International Forestry Review, 19(4), 10-26. doi:10.1505/146554817822295957

Kipalu, P., & Mukungu, J. (2013). Statut du processus REDD+ en République Démocratique du Congo. Retrieved from http://www.forestpeoples.org/sites/fpp/files/publication/2013/05/redd-drc- frenchmay2013.pdf

Kitchin, R., & Dodge, M. (2007). Rethinking maps. Progress in Human Geography, 31(3), 331-344.

Kitchin, R., Dodge, M., & Perkins, C. (2011). Introductory Essay: Power and Politics of Mapping. In Dodge, M., Kitchin, R. & Perkins, C. (Eds.), The Map Reader: Theories of Mapping Practice and Cartographic Representation (pp. 388-394). Oxford: Wiley-Blackwell.

Kitchin, R., Gleeson, J., & Dodge, M. (2013). Unfolding mapping practices: a new epistemology for cartography. Transactions of the Institute of British Geographers, 38(3), 480-496. doi:10.1111/j.1475-5661.2012.00540.x

Kitchin, R., Perkins, C., & Dodge, M. (2009). Thinking about maps. In Dodge, M., Kitchin, R. & Perkins, C. (Eds.), Rethinking Maps: New frontiers in cartographic theory (pp. 1-25). London & New York: Routledge.

Kwan, M.-P. (2007). Affecting Geospatial Technologies: Toward a Feminist Politics of Emotion. The Professional Geographer, 59(1), 22-34. doi:10.1111/j.1467-9272.2007.00588.x

Latour, B. (1987). Science in Action: How to Follow Scientists and Engineers through Society. Milton Keynes: Open University Press.

Laude, N. (Cartographer). (1937-1938). Congo Belge. Principales productions minérales, végétales et animales. Voies de communication [Map]. Retrieved from http://www.expocongo.be/content.php?m=6&r=3&sr=5&doc=159&l=fr

Law, J. (2004). After Method. Mess in social science research. Abingdon: Routledge.

Lawson, S. (2014). Illegal Logging in the Democratic Republic of the Congo. Retrieved from https://www.chathamhouse.org/publications/papers/view/198642

Lefebvre, H. (1991). The Production of Space. Oxford & Cambridge: Blackwell.

Li, T. M. (2007). The Will to Improve. Governmentality, Development, and the Practice of Politics. Durham & London: Duke University Press.

Malengreau, G. (1949). Vers un paysannat indigène. Les lotissements agricoles au Congo belge. Bruxelles: Institut Royal Colonial Belge.

Massey, D. (2005). for space. London & New York: Routledge.

McCall, M. K. (2016). Beyond “Landscape” in REDD+: The Imperative for “Territory”. World Development, 85(Supplement C), 58-72. doi:https://doi.org/10.1016/j.worlddev.2016.05.001

McElwee, P. (2016). Forests Are Gold: Trees, People, and Environmental Rule in Vietnam. Seattle: University of Washington Press.

103

Mignolo, W. D. (2009). Epistemic Disobedience, Independent Thought and Decolonial Freedom. Theory, Culture & Society, 26(7-8), 159-181.

Moeller, A. (1936). La Politique Indigene de la Belgique au Congo. Journal of the Royal African Society, 35(140), 233-240.

Molinario, G., Hansen, M. C., & Potapov, P. V. (2015). Forest cover dynamics of shifting cultivation in the Democratic Republic of Congo: a remote sensing-based assessment for 2000–2010. Environmental Research Letters, 10(9), 094009.

Moonen, P. C. J., Verbist, B., Schaepherders, J., Bwama Meyi, M., Van Rompaey, A., & Muys, B. (2016). Actor-based identification of deforestation drivers paves the road to effective REDD+ in DR Congo. Land Use Policy, 58, 123-132. doi:https://doi.org/10.1016/j.landusepol.2016.07.019

Mosse, D. (2004). Is good policy unimplementable? Reflections on the ethnography of aid policy and practice. Development and Change, 35(4), 639-671.

Mosse, D. (2010). A Relational Approach to Durable Poverty, Inequality and Power. The Journal of Development Studies, 46(7), 1156-1178. doi:10.1080/00220388.2010.487095

Mpoyi, A. M., Nyamwoga, F. B., Kabamba, F. M., & Assembe-Mvondo, S. (2013). Le contexte de la REDD+ en République Démocratique du Congo. Causes, agents et institutions. Retrieved from https://www.cifor.org/library/4094/

Nixon, R. (2011). Slow Violence and the Environmentalism of the Poor. Cambridge, MA: Harvard University Press.

Noble, K. B. (1989, December 20). Kinshasa Journal; Getting Rich in Zaire: An American, 33, Tells How. The New York Times, p. 4.

OCEAN ONG (Cartographer). (n.d.). Projet Pilote REDD Géographiquement Integré d'Isangi (PPRGII). Cartographie participative et Micro zonage: Espace traditionel des droits de la communauté de Yalosuna (Chefferie de Kombe) [Map]. Kisangani : OCEAN ONG.

Olsson, G. (2007). Abysmal: A critique of cartographic reason. Chicago: University of Chicago Press.

Parenti, C. (2007, October 22). The Fight to Save Congo’s Forests. The Nation. Retrieved from https://www.thenation.com/article/fight-save-congos-forests/

Peluso, N. L. (1995). Whose Woods are These? Counter-Mapping Forest Territories in Kalimantan, Indonesia. Antipode, 27(4), 383-406. doi:10.1111/j.1467-8330.1995.tb00286.x

Pickles, J. (2004). A History of Spaces: Cartographic reason, mapping and the geo-coded world. London & New York: Routledge.

Raheja, M. H. (2010). Reservation Reelism: Redfacing, Visual Sovereignty, and Representations of Native Americans in Film. Lincoln: University of Nebraska Press.

104

Robbins, P. (2001). Fixed Categories in a Portable Landscape: The Causes and Consequences of Land- Cover Categorization. Environment and Planning A: Economy and Space, 33(1), 161-179. doi:10.1068/a3379

Rose-Redwood, R. S. (2006). Governmentality, Geography, and the Geo-Coded World. Progress in Human Geography, 30(4), 469-486. doi:10.1191/0309132506ph619oa

Roth, R. (2009). The challenges of mapping complex indigenous spatiality: from abstract space to dwelling space. cultural geographies, 16(2), 207-227. doi:10.1177/1474474008101517

Sayer, J., Sunderland, T., Ghazoul, J., Pfund, J.-L., Sheil, D., Meijaard, E.,. . . Buck, L. E. (2013). Ten principles for a landscape approach to reconciling agriculture, conservation, and other competing land uses. Proceedings of the National Academy of Sciences of the United States of America, 110(21), 8349- 8356. doi:10.1073/pnas.1210595110

Sletto, B. (2002). Producing space(s), representing landscapes: maps and resource conflicts in Trinidad. cultural geographies, 9(4), 389-420. doi:10.1191/1474474002eu256oa

Terazono, E. (2016, 24 October). Olam snaps up East African premium coffee specialist. Financial Times. Retrieved from https://www.ft.com/content/71702eef-915e-320e-acb4-51bad40c477e

The Sentry. (2018). Bank Scandal in Congo: Urgent Action Needed to Hold BGFI Bank DRC to Account for New “Terrorist Treasury” Revelations, Hezbollah Financier Links Retrieved from https://thesentry.org/reports/bank-scandal-in-congo/

Thompson, S. (2011, February 2). New Interactive Atlas Will Help Democratic Republic of Congo Manage Forests. Retrieved from https://www.wri.org/blog/2011/02/new-interactive-atlas-will-help- democratic-republic-congo-manage-forests

Trefon, T. (2008). La réforme du secteur forestier en République démocratique du Congo: défis sociaux et faiblesses institutionnelles. Afrique contemporaine, 2008/3(227), 81-93. doi:10.3917/afco.227.0081

Turner, T. (2007). The Congo Wars: Conflict, Myth and Reality. New York: Zed Books.

UN-REDD, & DRC-MECNT. (2012). Synthèse des études sur les causes de la déforestation et de la dégradation des forêts en République Démocratique du Congo. Retrieved from http://www.forestpeoples.org/sites/fpp/files/publication/2013/05/synthese-etudes-causes- ddfinalaout2012.pdf

Vandermotten, C. (2008). L’histoire de la géographie belge à travers les biographies nationales. Belgeo, 1, 105-122. doi:10.4000/belgeo.10209

105

Weatherley-Singh, J., & Gupta, A. (2017). An ecological landscape approach to REDD+ in Madagascar: Promise and limitations? Forest Policy and Economics, 85, 1-9. doi:https://doi.org/10.1016/j.forpol.2017.08.008

Wehkamp, J., Aquino, A., Fuss, S., & Reed, E. W. (2015). Analyzing the perception of deforestation drivers by African policy makers in light of possible REDD+ policy responses. Forest Policy and Economics, 59, 7-18. doi:https://doi.org/10.1016/j.forpol.2015.05.005

West, P. (2016). Dispossession and the Environment Rhetoric and Inequality in Papua New Guinea. New York: Columbia University Press.

Windey, C., & Van Hecken, G. (2019). Contested mappings in a dynamic space: emerging socio-spatial relationships in the context of REDD+. A case from the Democratic Republic of Congo. Landscape Research. doi:10.1080/01426397.2019.1691983

Winichakul, T. (1994). Siam Mapped: A History of the Geo-Body of a Nation. Honolulu: University of Hawaiii Press.

Wood, D. (1992). The Power of Maps. New York: Guilford Press.

Wood, D., & Fels, J. (2008). The natures of maps. Chicago IL: Chicago University Press.

WRI. (2014). RELEASE: Global Forest Watch: Dynamic New Platform to Protect Forests Worldwide [Press release]. Retrieved from https://www.wri.org/news/2014/02/release-global-forest-watch- dynamic-new-platform-protect-forests-worldwide

WRI, & DRC-MECNT. (2010). Atlas forestier interactif de la République Démocratique du Congo – version 1.0, Document de synthèse. Washington, D.C.: World Resources Institute.

WWF Forest and Climate Programme. (2016). Significant stories: designing green development in the Congo Basin. Retrieved from https://wwf.panda.org/wwf_news/?261832/Significant-Stories- Designing-Green-Development-in-the-Congo-Basin

Yeow, A., & Faraj, S. (2014). Technology and Sociomaterial Performation. Paper presented at the Information Systems and Global Assemblages. (Re)Configuring Actors, Artefacts, Organizations, Berlin, Heidelberg.

106

PART 2

Entangled (after)lives in Kisangani’s forested hinterland: rubble of imperial rationalization and the unruly multiplicity of socio-spatial-temporal relations

107

108

Introduction

On a typical day of fieldwork, sitting in Merci Babili’s yard on one of those iconic blue plastic chairs usually offered to visitors, I was carefully looking at an undated map and a decree dating back to the Belgian colonial era; the first among many other colonial documents I would be presented with over the course of my field research. Merci Babili was passionately talking about the non-recognition of his ethnic subgroup’s land and forests by a foreign logging company. The logging company, he told me, is destroying their relations both with forests by anarchically cutting trees, pushing animals away and limiting access to forests; and with other ethnic subgroups by imposing new land boundaries on which socio-economic compensation for certain communities are defined, fostering division through inclusion and exclusion. The revelation of this fierce, long-lasting multidimensional conflict generated by resource development and fuelled by maps, quickly complicated simple narratives of disrupted landscapes, local96 people’s relations to their forests, indigeneity and dispossession. During the previous couple of weeks of interviews, such land conflicts had become prominent topics for my informants despite my then naïve attempts to direct their focus solely on forest use and environmental change themes – this was what I needed to know. In fact this conflict appeared as a powerful entry point for grasping processes of socio- ecological change (Bierschenk & Olivier de Sardan, 1997). I thus proceeded to investigate this conflict, which so vividly animated the communities I visited, not for its own sake but for what it could reveal about the complex socio-ecological relationships and the ways environmental regulation, tenure systems, kinship connections, spiritual beliefs, material socioeconomic practices and different modes of being all intertwine to produce forest landscapes in the hinterland of Kisangani (Jacka, 2015). Forest change in this context is not just an environmental and spatial issue caused by shifting agriculture and population growth alone. It is also a social and moral problem. The conflict became a first opening to understand the relationality of the landscape I was starting to explore – a landscape that can never be completely captured through maps or governed through planning, but which is apprehended and continuously built through “the business of dwelling” in it and through “porous networks” of socio- ecological relations and everyday practices (Ingold, 2000, p. 192; Massey, 1994, p. 121; 2005; Roth, 2009).

The conflict as a point of entry to the social embeddedness of the landscape was not the only thing that struck me during my conversation with Merci Babili. Looking at what he insisted were colonial-era documents he had collected at the provincial representation of the Congolese Geographic Institute and which he was carefully conserving in his hut, I began to think of the meaning and the continuing presence of the past in this landscape. As many other people I encountered later would similarly do, Merci Babili

96 The italic font aims to emphasize the ambiguity of the term “local” which I purposefully do not associate with “indigenous” and which cannot be reduced either to the people physically living in a particular place. As others have long shown in different ways (e.g. Igoe & Brockington, 2007; Massey, 1993a), this second part in fact demonstrates the multiplicity of “the local” and how it is shaped by forces well beyond its notional spatial and temporal boundaries. 109

was appealing to the objectivity of Belgian colonial scientific and legal accounts (maps, orders and decrees) to prove his ethnic subgroup’s customary rights to this contested landscape, as primary occupants. The literature on legal pluralism in post-colonial African nations has long shown the ways formal law made during colonial administration co-exists, interacts and competes with evolving (informal) customary orders (e.g. Griffiths, 1986; von Benda-Beckmannn & von Benda-Beckmannn, 2006). That such a hybrid system existed or that Merci Babili’s map did not in fact show much, was of secondary importance at this stage; although, as this second part of the thesis demonstrates later, it helps to explain why environmental planners with their frequent lack of understanding of (and/or willingness to engage with) complex and multiple land use dynamics, are so often left with unexpected challenges and outcomes. At that moment, it is rather the temporality of the landscape, to borrow Ingold (1993)’s expression, that veritably arose. It was surrounding me. In Kisangani, formerly Stanleyville, a colonial station established by Stanley in 1883 which was also at the time a centre of the Zanzibari-Arab slave trade, and in almost every site I visited in the city’s hinterland, traces of the (pre-)colonial past were present materially, discursively, cognitively and symbolically. As Ingold might put it, the landscape bears witness to the passing of time; it contains a living memory of all who have lived in it, humans and nonhumans (Abelman, n.d.). Merci Babili’s documents were the very metaphor of the temporality of the landscape as much as the conflict in this logging concession’s area was a striking example of the continuous emergence of past and present elements in shaping landscape, socio-spatial and socio- ecological relations. The colonial maps and decrees, the infrastructural ruins, the remainders of colonial houses and buildings which people continue to occupy, the former colonial plantations, a certain sense of longing, the constant references to the Belgian oncles*97, or the hybrid modes of governing and being, all became pointers to the ways in which “the ruinous effects of [racial imperialism,] capitalism and nation-making become embodied in human-environment relations”, incorporated in the landscape, and bear on people’s minds, in often unpredictable ways (Stoetzer, 2018, p. 300; Stoler, 2013). I would quickly discover that what stood out to my Western eyes as ruins or rubble from the past are not passive environments but rather are constitutive of heterogenous forms of (after)lives (Stoetzer, 2018). This colonial rubble is in fact part of a confusion of traces from multiple eras, e.g. ancient tradition, the Zanzibari-Arab ivory and slave trade, and the present one, that are constitutive of the landscape of the Kisangani hinterland (Gordillo, 2014). My aim is thus to bring to light the entanglement of multiple temporal layers and socio-environmental legacies of (pre-/post-)colonial, capitalist, and present ruly and unruly modes of being and governing that are co-constitutive of forest landscapes and socio-spatial- temporal identities.

97 When they address Belgians, Congolese often say “les Belges, ce sont nos oncles” – “the Belgians are our uncles”– to emphasize the common history they once shared and still share. 110

I start this part with some conceptual reflexions that engage with complexity thinking or “worldview thinking”, teasing out connections between social and spatial theory, political economy, ecology, history, and multispecies interactions to examine a landscape in motion. Chapter 5 then looks at the articulation of forests, tenure and identities in the constitution of a contested landscape. I argue that ways of belonging and appropriating (forested) land are dynamically articulated in connectivity with multiple encounters across time and space and through multiple everyday forest use practices and meanings attached to various natural resources (uses). I show how successive waves of boundary- and resource- making processes shape the multiplicity of tenure and of identities in which a collective appropriation of forests coexists with an individualization of citizenship and land tenure, in often ‘unruly’ ways. Such articulation, I show, is far from static and is continuously negotiated in relation to new practices and changing circumstances in the natural and social environment. In many cases in fact, I will talk about land use practices in the chapter that deals with tenure, and about access where I talk about farming and land use. Both are so intimately intertwined that separating them for the sake of clear academic writing is not always an easy task. In chapter 5, I explore the heterogeneity and dynamism of farming, and the ways imperial exchanges have created ‘hybrid socio-ecologies’. The agricultural system of the hinterland blurs the boundaries between ‘subsistence’ and ‘cash crops’, ‘traditional’ and ‘modern’. I argue that farming and socio-economic relations are co-constitutive processes, and that various species (the soil, trees, crops, weeds, etc.) are co-productive agents of these processes. Farming, I show, is not an equal practice for all and (lack of) access to land and labour highly influences its economic and socio-ecological outcomes. Such differentiated dynamics inform local understandings of environmental change, which are widely interpreted in moral terms, as I show in the last section of chapter 6. In the conclusion, I bring these findings together and emphasize the entanglements between the past and present, invisible and visible human and nonhuman agents that interact to produce various socio-spatial-temporal relations and constitute the ‘landscape’.

111

Picture 1. The classic blue chair

Picture 2. The (undated) map of Bafwasende territory that Picture 3. The 1906 decree establishing the Merci Babili showed to me. Badombi chiefdom to which Merci Babili belongs.

112

4. Living in connectivity: landscape as a socio-natural and spatial-temporal phenomenon

Congolese everyday governance and political life at all levels have been characterized, by some analysts, in many respects by constant contestation, negotiation and hybridity (e.g. Englebert & Tull, 2013; Lund, 2006; Titeca & De Herdt, 2011; Vlassenroot & Delaleeuwe, 2008). Conflicts inside and outside the parliament in Kinshasa, between central and provincial levels, between the National Assembly and a particular ministry, as much as between the State and civil society or between the government and ‘the international community’ are commonplace and fill Congolese newspapers on a near-daily basis. Closely related to contestation, Englebert and Tull (2013) argue that negotiation plays a central role in social and political relations and seems to be the main practice of – rather than instrumental to – the exercise of politics in the Democratic Republic of Congo (DRC). The aim of such practice, whether within or between the government, the army and multiple non-state actors, appears to be the inclusion and participation of everyone in the state system, mainly to access its economic opportunities (e.g. negotiation of contractual terms with resource exploitation companies). As a result, Congolese political governance, as much as society, is in a permanent state of flux and uncertainty (of norms, laws, formal institutions, etc.) in which power is exercised by a multitude of actors who, somehow paradoxically contribute to reproduce the State and public authority (Englebert & Tull, 2013; Titeca & De Herdt, 2011).

I want to build on these considerations on Congolese politics to reflect on the centrality of relationality, negotiation, flux and emergence of the landscape which, as this second empirical part attempts to demonstrate, all characterize the Kisangani hinterland. Everyday Congolese socio-political practices and the formation of forest landscapes echo each other and are, by definition, mutually constitutive. Below, I attempt to provide a conceptualization of landscape that think of space and place in terms of such relationality, i.e. relations/connections and practices precede identity. I start by borrowing Ingold (1993, 2000)’s use of the term landscape in his dwelling perspective as a building block to develop my thoughts on connectivity, multiplicity and temporality. I additionally draw on spatial theorists, in particular Doreen Massey and her relational approach to space and place, and (environmental) anthropologists who have been interested in the spatio-temporal aspects of human-environment relationships. I conclude by providing a conceptualization of landscape stimulated by these discussions.

4.1. Consideration 1: Dwelling

Building on his research among hunters-gatherers and their worldviews, Ingold (1993, 2000) emphasizes the concept of landscape as running against the classic divide between culture (human) and nature (nonhuman) that dominates Western thought. Contrary to the notion of space which Ingold associates with the abstract (human-formed) representation of earth as a surface to be occupied or inscribed,

113

landscape is not ready-made. That is to say, landscape is not a container inscribed by forms and meanings, but rather “meanings are gathered from it” (Ingold, 1993, p. 155; original emphasis). In this way, landscape is defined as “the world as it is known to those who dwell therein, who inhabit its places and journey along the paths connecting them” (Ingold, 1993, p. 156). Landscape is thus a verb rather than a noun; it is a mutual constitution of being and place through movement/action. Ingold’s dwelling perspective is both ontological and epistemological. Ontologically, landscape’s features and subjectivities mutually come into being through the engagement of people in the world, through a process of incorporation of a pattern of everyday life activities and an array of relations between and among human and nonhuman components. Relations in this understanding are always simultaneously ecological and social, and the forms of the landscape arise alongside, the forms of the activities. Epistemologically, knowledge of the social and physical world is acquired through movement and immediate experience in it; it is not a matter “of making a view of the world but of taking up a view in it” (Ingold, 2000, p. 42; original emphasis). So to understand the landscape requires the privileging of the understandings of those who are involved in it through their everyday movement and practices. In some of his more recent work, Ingold (2011, p. 129) offers a convenient summary of his approach:

“Landscapes take on meanings and appearances in relation to people, and people develop skills, knowledge and identities in relation to the landscapes in which they find themselves.”

Although not from such a phenomenological approach, Tsing (2017, p. 7) has similarly stressed the dynamic aspect of landscapes as “gatherings of ways of being in the making” and the heterogeneous (multispecies) entanglements that constitute them. Such understandings of landscapes strongly emphasize the inseparability of society and environment. As noted earlier, such shift to connectivity (Rose, 2011, 2017) radically challenges gendered and racialized culture/nature “hyperseparation” that ecofeminist philosopher Val Plumwood (1993) considers to be a central intellectual move in Western thought. This hyperseparation creates hierarchies of superiority and control between genealogies of dualistic structures: culture over nature, mind over matter, reason over emotion, men over women, master over colonized (Rose, 2011). Connectivity, in contrast, refers to intra- and cross-species exchange pathways through which “the wellbeing of one is enmeshed in the well-being of others”; one’s interests are entangled in the interests of others (Rose, 2011, p. 27). It erases demarcations between ‘the wild’ and ‘the cultivated’. In the case at issue in this Part II, these reflections help to understand how forests and trees – like the matsutake mushrooms in Tsing (2015)’s book – manage to entice and connect local forest dwellers, urban elites, foreign loggers, environmental project promoters and wildlife; they live with and through forests, which all shape them in turn. Forests and their multiple resources are the basis of material, social and moral/spiritual life; they are instrumental (not incidental) in coproducing the lifeways and identities of forest users and dwellers (Tsing, 2015). In other words, material, nonhuman things – in

114

this second part, I am interested in crops and weeds – are thus physical constituents of sociocultural and economic life, as much as the meanings attached to it (Bakker & Bridge, 2006).

The notion of connectivity, for Rose (2011), also entails a moral, ethical dimension that is of particular interest in my analysis. As mentioned in the introduction and as I explore later in this second empirical part, forest use and change appear not only as environmental issues but mainly as relational, moral and spiritual ones, i.e. they result from social and moral transformations of society, and from actions that, intentionally or not, break “connectivity, mutuality and the flourishing of beings and relationships” and a shared moral domain (Jacka, 2015; Rose, 2011, p. 62). In such a context, when invoking the significance, value and meaning of forests or a particular species (e.g. tree species, insect, etc.), people usually invoke their entanglement in complex cultural and socio-economic relationships rather than an abstract value; they “read the landscape for its social as well as its natural stories” (Tsing, 2005, p. xi).

4.2. Consideration 2: Trans-scale multiplicity and power-geometry

As Roth (2009) has similarly put it in her use of Ingold’s ‘dwelling’ concept, the arguments outlined above are particularly useful to overcome the limitations of abstract space conveyed by maps and to make sense of dynamic socio-spatialities and socio-ecological relations such as the peri-urban landscape of Kisangani. I concur with her that Ingold’s (2000) dwelling approach however carries an issue of scale, as does in my view – to a lesser extent though – Rose (2017)’s connectivity thinking, which we need to overcome if we wish to grasp the full relationality of the landscape. The dwelling/connectivity perspective indeed seems too limited to “the local”, which is somehow conceived as bounded, with too little attention to larger processes and networks of connections that exceed the geography and immediate character of an area but that all contribute to shape the landscape (Massey, 2005; Roth, 2009). Moreover, it does not sufficiently look at the everyday politics of the landscape, that is, the ways in which a landscape is always a contested field of practices, constructed through contestation, negotiation and dispute. The risk of such shortcomings is to fall into ‘parochialism’ (Massey, 1993b) or some sort of ‘identity politics’ (Fraser, 2000) through which the study of the landscape would mostly focus on identifying and reifying a unique and drastically simplified group or place identity. The overall effect is to misunderstand the wider spatialities of social and ecological relations and to deny the complexity of people’s lives, the multiplicity of their identifications, affiliations and relations to place. This kind of approach is not only often tinged with romanticism and nostalgia, it erases intra-group variations and requires the drawing of boundaries between an inside and an outside. It might even encourage exclusivist localism, ethnicism, purely ethnic- based territoriality and sovereignty, or displace claims of social and distributive justice (Fraser, 2000; Massey, 1993a, 1993b). Epistemologically, it might also ignore other understandings of landscapes (e.g. maps, satellite visualizations) that might also contribute to shape ‘local’ relations.

115

Massey (2005)’s proposition to theorize space and place as heterogeneous, lively and contested, in this way productively complements the landscape dwelling/connectivity perspective and greatly helps us to further integrate geographical and spatial thinking with the understanding of socio-ecological systems in a ‘trans-scale’ and temporal perspective. Massey’s first radically anti-essentialist proposition conceptualizes space and place as the product and modifier of interrelations, both human and nonhuman. These interrelations, however, are not limited to the ones happening in the local place but arise “from the immensity of the global to the intimately tiny” (Massey, 2005, p. 9). She thus rejects traditional binaries between place and space. The specificity of place, its uniqueness, Massey (1993a, p. 66) tells us:

“is constructed out of a particular constellation of relations [and movements], articulated together at a particular locus […] but where a large proportion of those relations, experiences and understandings are actually constructed on a far larger scale than what we happen to define for that moment as the place itself, whether that to be a street, a region or even a continent.”

Massey (2005, p. 9)’s second proposition, which comes from her deep interest in processes of globalization, positions space/place as the sphere of multiplicity, “in the sense of contemporaneous plurality; as the sphere in which distinct trajectories co-exist” and overlap in many and unpredictable ways. Such a view of space/place can be usefully linked up with feminist geographers Gibson-Graham (2002, 2014)’s deconstruction of the ‘global-local’ binary and understanding of economic identity as a coexisting multiplicity of socio-economic practices and transactions. For Massey (2005), in fact, space and place, the socio-political, and identities/ entities are co-constituted through a constantly negotiated, conflictual process of multiple activities, practices and movements through and across space. Her proposition here strongly echoes the contested and negotiated character of the Congolese State that political scientists (see above) have underlined. It also resonates with anthropologist Stephen Gudeman (2008)’s conceptualization of socio-economies as built on a tension between making calculations and maintaining relationships with others, between self-interest and shared concerns, and that both ‘sides’ are intermingled.

The conflictual, negotiated dimension of space/place highlights that groups and individuals are differently positioned in relation to these flows and interconnections. These power relations both shape and reproduce spatiality and mobility, e.g. through the definition of rules or the construction of borders and boundaries, framing differential geographies between groups; a process that is captured in Massey’s (1993a) notion of “power-geometry”. Ferguson’s (2006) ‘Global Shadows’ exemplifies such uneven relationality arguing that while globalization has ‘reached’ Africa, the vast majority of Africans are excluded from or ‘shadowed by’ it, despite a claim to ‘membership’ of global society (and access to its

116

goods and forms). Globalization, rather than meaning total interconnection and the removal of borders, generates discontinuous transnational networks of social relations, power and capital that selectively and hierarchically constitute “global society” and reconfigures the continent’s socio-cultural, political and economic relations. In this Part II, I show how such connectedness to former Belgian imperialism and current global political-economic order contribute to shape socio-ecological relations in the Kisangani hinterland’s landscape.

4.3. Consideration 3: Becoming – time in the landscape

At the center of Ingold (1993, 2000)’s dwelling perspective, Rose (2011, 2017)’s connectivity thinking or Massey (2005)’s relational approach to space lies an explicit reintroduction of the temporal (temporality/time) into the spatial, through their common concept of ‘becoming’. Space/place, in their view, must be acknowledged to be always in the process of becoming, that is space/place is always under construction, never a closed system, with conditions of life never set once and for all (Rose, 2011). It is ontogenetic in nature precisely because it is the product of “relations which are necessarily embedded material practices which have to be carried out” (Massey, 2005, p. 9). For Massey, space and time are thus fundamentally interwoven: “[t]here is no choice between flow (time) and a flat [static] surface of instantaneous relations (space)” (Massey, 1992, p. 81). This means we need to think in terms of “space- time”.

Although Massey (2005) provides a critique of Bergson’s characterization of space, such understandings of the spatio-temporal/temporal-spatial very much resonate, to me, with Bergson’s concept of ‘duration’, which sees time as a constant flow and as lived, experienced time. Duration is a “virtual whole that encompasses an infinite number of actual fluxes” (Posman, 2012, p. 111). Unlike mechanical clock time or a linear, chronological homogeneous timescale, duration also emphasizes a continuous flow of the past into the present but also a ‘maze’ of multiple interconnected spatial-temporal trajectories (Schatzki, 2010). Similarly, Massey (2005, p. 24)’s space-time is conceived as a “simultaneity of stories-so-far” while Ingold (1993, p. 152) talks of the landscape as “an enduring record of – and testimony to – the lives and works of past generations who have dwelt within it, and in so doing, have left there something of themselves”. Because the landscape is imprinted with the activities of predecessors, bodily orientations, skills, sensibilities and dispositions of humans and nonhumans are conditioned by the work of predecessors and will in turn contribute to fashioning the landscape of their progenies, through the very process of their dwelling (Ingold, 2000).

Social relations of the past thus do not just disappear but continuously emerge and are intertwined in the present becoming of the landscape, as I alluded to in my introductory reflection on colonial disruptions and ‘ruination’ characterizing the hinterland of Kisangani. The entanglement of time and space/place and

117

the notion of ‘becoming’ indeed invites me to additionally draw on recent scholarship on ruins (Stoler, 2013; Tsing, 2015), rubble (Gordillo, 2014) and ruderal ecologies (Stoetzer, 2018) in particular. Such scholarship posits ruins or rubble as alive rather than dead, as both tangible and invisible, and engage with imperial ‘ruination’ as an active process that shapes the psychic, social and material space in which people are left and live, and that weighs on the future. Gordillo (2014) mobilizes ‘rubble’ from successive waves of destructions from Spanish colonialism to the more recent expansion of industrial agribusiness in the Argentinian Chaco, as an analytical figure to explore the ongoing effects of violence and memory that interpenetrate space, social (after)lives and subjectivities. His notion of ‘rubble’ is important as it defestishizes elite valuations of ruins as monuments and includes non-material forms of damage. Building on this approach, Stoetzer (2018)’s ruderal98 analytical lens broadens our attention from destruction to heterogeneous and unexpected socio-ecological lives that emerge amid rubble – located, in her study, in Berlin’s “inhospitable environments created by war and exclusion” (Stoetzer, 2018, p. 308). It can be related to the ecological concept of “edge effect” that describes the increasing biodiversity occurring at the boundary of two ecosystems and the “juxtaposition of contrasting environments in one ecosystem” (Stoetzer, 2018, p. 297). Such conceptualizations shift our understanding of the ruins or rubble as passive elements of the past – as separated from the present (Gordillo, 2014), as behind, finished in a linear conception of time – to ruins/rubble as persistent, power-laden and active trajectories, as a “verb that unite apparently disparate moments [and] places” (Stoler, 2013, p. 7). In this Part II, I am thus interested in visible and invisible imperial, capitalist rubble as indicators of the interplay of entangled histories, encounters and heterogeneous human-environment relations shaping landscape and identities, and in their unintended and unruly forms of re-appropriation through everyday ‘dwelling’, that witness to the multiplicity and never-ending ‘becoming’ of the landscape.

4.4. Landscape: a conceptualization

In the above theoretical discussion, I have consecutively and sometimes simultaneously used the concepts of landscape, space and place following the terms used and conceptualized by the authors I discussed. In this second part of the thesis, however, I operationalize the concept of landscape for which I adopt, based on the above insights, a relational, processual understanding that incorporates the four dimensions of (multi-species) sociality, materiality, spatiality and temporality. The landscape is a provisional expression of a set of heterogeneous, unfolding and intertwined human, nonhuman and material (unequal) relations and practices ‘so-far’ acting across scales and whose future is negotiated. Landscape is simultaneously historical and emergent, social and natural, material and imagined. It is

98 As Stoetzer (2018, p. 297, original emphasis) writes, “[t]he term ruderal comes from rudus, the Latin term for rubble. A common term in urban ecology, it refers to communities that emerge spontaneously in disturbed environments usually considered hostile to life: the cracks of sidewalks,” the spaces alongside train tracks and roads, industrial sites, waste disposal areas, or rubble fields.” 118

continuously ‘becoming’ in the connectivity of past, present and yet-to-come human and nonhuman dwelling practices, produced in relation to social, political, cultural and economic processes, and is thus multiple and dynamic. Such a conceptualization is thus a radical move away from the abstract and static landscape of most environmental and land use plans, maps, or even paintings or photographs. In the two next empirical chapters, I read the Kisangani hinterland landscape through this conceptualization.

The landscape I describe is a peri-urban, tropical forested area in a radius of 150km around Kisangani, DR Congo’s third largest city. This area is often referred to as the rural hinterland or supply area of the city, and in classic typologies of spatial development and planning, the boundaries between the city and this rural hinterland are assumed (or desired) to be socially, spatially and temporally fixed. It is, we will see, a first misconception and I rather understand the rural forest environment around Kisangani and the city itself as part of the same continuum, in terms of socio-ecological, economic and cultural relations. They are co-constructed in relationality. This is an area where imperial rubble, ‘global’ and ‘local’ capital, an extractive foreign company, global environmental policy (e.g. REDD+), heterogeneous forest dwellers and users, and multiple forest resources interact to ‘make’ and negotiate the landscape. The Kisangani hinterland’s landscape thus offers a microcosm to explore landscape as a relational-spatial-temporal process that defies conceptions of abstract space, space as a container, and to examine processes of ‘slow violence’ (Nixon, 2011) in which socio-environmental injustice, histories of racial colonialism and extractive and socio-ecological modernization “work through human-nonhuman relations in sticky and persistent ways” (Stoetzer, 2018, p. 315). In the messiness of this conflict, the intertwined simultaneities of past and present complex trajectories of making a livelihood and of social, economic and cultural reproduction that shape the landscape, have become progressively apparent. They also offer another lens to understand environmental change, which engages with social and environmental justice.

119

120

5. Articulating tenure, identities and forests: untidy relations

In March 2016, during my second research stay, as I was listening to Kisangani hinterland forest dwellers’ views on socio-ecological relations and change, an increasing number of people mentioned a fierce conflict that was opposing several groupements*, i.e. ethnic subgroups – I will come back later to this terminology. The conflict was taking place in what I refer to, in this manuscript, as the ‘Alibuku landscape’. Since 1994 this landscape has been home to a 263,632-hectare industrial logging concession (COTREFOR, 2016) whose entrance road is located about 24km north of Kisangani, and leads to the small town of Alibuku, about 40km from Kisangani. I became particularly interested in this landscape not so much for the extractive activities of the logging company for their own sake but as an exemplary case where the instrumental abstraction and rationalisation of ‘forests’ and ‘population’ misrecognizes the full complexity and socio-historical and ecological embeddedness of human access to forests, and fosters processes of exclusion through renewed ‘ethnic-based’ governance. This approach, widespread in environmental policies, is based on two assumptions that are particularly prevalent in discussions about indigeneity and ‘traditional’ knowledge. First, it assumes that communities are homogenous and that identity/ethnicity is concrete and unchanging. Second, it assumes that such fixed identity/ethnicity affects patterns of resource use, i.e. the former is prior to the latter. In this chapter, I challenge these assumptions. Collective identity here is not considered fixed but rather as a process emerging relationally through everyday forest use practices and regulatory regimes (Lau & Scales, 2016). This chapter should be seen in the light of the larger project of this thesis, which sees landscapes as constantly brought into being through new series of power-laden socio-material practices.

I will explore such complexity and fluidity by using the conflict that arose in the Alibuku landscape as an entry point. Section 1 thus starts by describing the anatomy of the resource conflict triggered by COTREFOR/IFCO’s forest management. In section 2, I then dig into the ‘story-so-far’ of the Alibuku landscape as it was recounted to me and complemented by secondary data, highlighting multiple historical layers and disruptions. This section emphasizes the significance of the past and provide a historical contextualisation of the region that lays the foundation of a form of ‘ethnic-based’ forest governance embraced by ‘Sustainable Forest Management’ techniques promoted by REDD+ policies and used by COTREFOR/IFCO Alibuku concession, and of the conflict that ensued. I pay particular attention to pre-colonial socio-ecological relationships that were fundamentally unsettled by colonial rule but that still play a fundamental role in today’s socio-ecological relations although they are largely ignored by environmental regulators. Section 3 provides an analysis of such a story and of today’s tenure dynamics that shape forest landscapes in the Tshopo Province. Based on the story of the Alibuku landscape and the wider historical contextualisation, I argue that access and use of resources in the Kisangani hinterland landscape is shaped by entangled encounters across time and space and by simultaneous but

121

heterogeneous social, political and economic relations. I show that forest access and use are always negotiated in connectivity with meanings attached to the natural resource at play, emphasizing how wider relations of capital and transformations of forest products into commodities shape such access. Overall, my analysis demonstrates the adaptability and dynamism of modes of access and appropriation, and identities in response to a changing socio-ecological environment.

5.1. Anatomy of a conflict

In January 2016, Cotrefor/IFCO99, a Lebanese-owned logging company announced its Forest Management Plan (FMP) for its ‘Alibuku 018/11 concession’ had been approved, after four years of producing forest inventories, drawing maps and management scenarios, and conducting socio-economic assessments of the neighbouring communities. Since the promulgation of the Forestry Code in 2002, logging companies have indeed been required to submit FMPs for each of their concessions at the risk of losing their exploitation titles. While the code and its application has been (very) loosely enforced, REDD+ ambitions to reactivate it and support logging companies in the drawing up of FMPs through programs such as the US$18 million Sustainable Forest Management Program (PGDF) of the French Agency for Development (AFD) – see Part I, chapter 3. FMPs are contractual planning tools for long-term forestry management (25 years in DRC) between the logging company and state institutions at multiple levels. They are the quintessence of scientific forestry as they spatially organize and rationalize production through inventories, geolocalization of high-value timber and optimization of harvest schedules (division into homogeneous 5-year harvest blocks and annual felling areas). It is assumed that they facilitate the estimation of sustainable levels of timber harvesting and monitoring and control of illegal exploitation (e.g. harvesting outside the defined annual cutting area). Such plans also organize different bounded land use zones for regulating communities’ access and resource use, and for ensuring the delimitation of protection and conservation zones. On the basis of such plans, socio-economic compensation levels for communities are also defined. Production zones are, in this way, theoretically ‘protected’ from encroachments from communities and other artisanal activities.

However, the approval of Cotrefor/IFCO’s FMP did not appear to be received with the same enthusiasm by the communities who inhabit the concession’s perimeter. I became interested in two elements of the FMP that directly impacted communities’ landscapes and that elicited tensions with people’s livelihoods and socio-political organisation and fuelled outright conflict. One concerns the mapping of boundaries of the groupements’*, forested lands on which a concession operates so as to distribute the benefits of the social responsibility contract (Figure 15). Before carrying on with the story-so-far of this landscape, I

99 As mentioned in part I, the name of the company has shifted from Amex-Bois to Trans-M, then Cotrefor and now IFCO due, notably, to numerous accusations of illegality and links with Hezbollah. It seems, however, that the very secretive owner has always remained the same despite assertions to the contrary. 122

believe it is important here to quickly clarify the meaning of some key administrative entities – which are a colonial legacy – in their relation to customary organization that will help to understand the conflict. Because of the complexity of the political-administrative decentralization in DRC, I will only focus on the subdivisions that are important to understand the conflict. A groupement*100 is a traditional, customary group of people (who claim to be) united by an actual or perceived common ancestor and set up as an administrative entity that is under the authority of a customary chief usually chosen for his affiliation/kinship to the original ruling family. In customary ‘ethnic’ terms, a groupement* is a subdivision of an ethnic group which in administrative terms (roughly) corresponds to a secteur or chefferie*. The groupement* is a subdivision of a secteur/chefferie* which are attached to a territoire* , and each groupement* is sub-divided into villages* that are in turn sub-divided into clans*. Secteurs/chefferies*, groupements*, villages* and clans* are traditional and are organized according to customary laws. Groupements* and villages* are simple administrative divisions but territoires* are political- administrative divisions on which – and this is important for my story – political representation is defined, i.e. territoires* are electoral constituencies for general elections (Yetilo, 2010). FMPs are signed with groupements* although customary rules of access to land are traditionally managed at villages* and clans* level. In practice, customary ownership and tenure rights are much more complex and dynamic, and the colonial political-administrative and FMP’s divisions do not always make sense on the ground. However, because of the benefits associated with being organized as groupements*, villages have an interest in organizing themselves as such, as we will see. In this way, the director of the Congolese Geographic Institute (IGC) asserted that while there used to be about four thousand groupements* after independence, there are now more than six thousand.101

It is well known that such clear hierarchical ethnic-administrative divisions were imposed by the colonial administration and although these groups had some social and linguistic homogeneity, these clear-cut boundaries ignored the territorial fluidity and historical alliances that characterised the region’s socio- ecological and political organisation before the colonial period (see e.g. Turner, 2007; Verweijen & Van Bockhaven, 2020). The consequences of such administration are still strongly felt, as exemplified by the story below. To facilitate the understanding of the reader not initiated in such ethnic-administrative complexity, I will mainly refer to ‘territories’, ‘ethnic groups’ (= secteurs/chefferies*) and ‘ethnic subgroups’ (=groupements or villages) which is sufficient to grasp the essence of the argument (see Figure 13 for a schematic representation of these divisions). As we will see, the number of groups living in the Alibuku forests I mention in this chapter shows how such static categories are not able to explain the shaping of this socio-ecological landscape.

100 In the remainder of Part II, I keep to the French terms groupement* used by my informants for the sake of clarity as there is no exact English translation available for some entities, and to mark a distinction between different types of entities. 101 Interview 48-CSI-KIN-20170531. 123

Figure 13. Basic administrative divisions for rural areas. Assembled by the author based on Yetilo (2010)

The social responsibility contract establishes specific socio-economic investments by the company for the ethnic subgroup whose customary ancestral lands are determined and acknowledged to be within the perimeter of the logging concession. The amount of annual compensation for each ethnic subgroup is calculated based on the surface area of their customary forests included in that year’s felling area.102 The salient quality for being recognized as customary owners concerns a timeless presence in the exploited forests and the concomitant status of ‘first occupant/comer’. In the absence of recent administrative maps, maps of forests’ boundaries of the different ethnic subgroups are drawn ad hoc and financed by the logging company. Boundaries are delineated on the basis of colonial-era administrative decrees and ethno-linguistic maps complemented with GPS coordinates collected by a team composed of a company’s representative, ethnic subgroup representatives (chiefs and mainly elderly people), and a representative of the territorial administration. The maps resulting from the combination of these sources need then to be officially approved by the ethnic subgroups’ chiefs and the territorial and provincial state administration.103 Four ethnic subgroups were finally recognized by Cotrefor/IFCO’s forest management plan as customary owners: Abata, Baangba, Bevenzeke and Boumbua. The Bevenzeke ethnic subgroup, however, has been accused by several other ethnic subgroups dwelling in the logging concession’s perimeter (and who have not be recognized in the FMP) (see Figure 14 & Figure 15) of being a “pirate groupement”104 that has been created just to extend the constituency of Bafwasende Territory, which the Bevenzeke claim to be part of. The Badombi, in particular, asserted a large part of their forests were unjustly attributed to the Bevenzeke in the FMP.

102 Interview 30-PA-KIN-20160407. 103 Interview 30-PA-KIN-20160407. 104 Interview 54-I-RAB-20170509. 124

The second element of the FMP that raised my interest concerns the spatial organization and delimitation of the concession into four broad categories of use: industrial timber production area (by far the largest), the rural development zone (where communities are allowed to convert forest into agricultural land or any other use), conservation zones (where any human activity is strictly forbidden) and protection zones (along water streams and fragile zones where logging activities and forest conversion are forbidden) (AGEDUFOR, 2015). The demarcation of these land-use zones is mainly premised on three socio- ecological categories that enable the estimation of community forest conversion rates: current area accounting for shifting cultivation, density of forest cover and existing road networks, all identified by satellite imagery. The model estimates the Rural Development Zone (RDZ) for a 25-year period, taking into account (potential) interventions to improve agricultural productivity and settle practices through projects such as REDD+ (AGEDUFOR, 2015). Once established, the zoning plan needs to be formally approved and signed by the ethnic subgroup chiefs and provincial administrative authorities. However, the chiefs and their communities have little to say about the suggested ZDR’s boundaries as they are defined on geospatial models.

Figure 14. Excerpt from the Forest Management Plan of COTREFOR/IFCO ‘Alibuku 018/11 concession’ (COTREFOR, 2016). The division of the concession between five groupements* (Abata, Bangba, Bevenzeke, Boumbua, Lubuya b). Lubuya b, however, was not allowed to sign the management plan or any contractual agreement with COTREFOR/IFCO because the Bevenzeke claim (and gained) rights on this area of land.

125

Figure 15. Excerpt from the Forest Management Plan of COTREFOR/IFCO ‘Alibuku 018/11 concession’ (COTREFOR, 2016). The concession’s Rural Development Zone in 2012 is represented in grey. In red, the planned extensions of the zone towards 2039, at the end of the plan’s validity.

In 2017, about two years after the plan’s approval, the RDZ’s boundaries were already reached and activities involving tree clearing (agriculture, charcoal production or artisanal logging) were taking place even within the limits of the production zone, i.e. outside of the RDZ. This completely ‘remapped’ the zone, with unpredictable patches of forest cleared for livelihood activities, defying the mathematical models that assumed homogeneous patterns of shifting cultivation and of occupation (Figure 16). As two consultants working for the AGEDUFOR program (see 3.3.) involved in the establishment of the FMP told me with consternation, “everything is moth-eaten in this concession” despite the communities having approved the limits of the RDZ in a participatory way.105 During the same period, the 10-year conflict that pitted the Bevenzeke ethnic subgroup against other ethnic subgroups living in the area, together with the resentment against the company for what was considered to be the unfair distribution of socio-economic benefits were also at their heights, with the main logging road blocked several times by the various communities.

The two AGEDUFOR consultants were particularly aware of the fact that excluding non-autochthonous users from the FMP was particularly inefficient in terms of planning. In fact, they were interested in

105 Interview 63-ITSP-KIN-20180529. 126

reading my work about the area as the limited time they have to go on field missions does not allow them to grasp the full complexity of the issue. On the other hand, two forestry experts at a French forest engineering consulting firm, involved in the FMP’s formulation pointed out to me106 that communities only created problems when they realized there were no more felling areas in their customary forests – and they thus could no longer receive compensation. Ethnic identity and first occupancy are in fact invaluable concepts for local communities through which they can articulate their struggles with the transnational extraction business, as I come back to in section 5.2.

In the colonial administrative documents and the few postcolonial maps I have consulted, the limit between the Lubuya-Bera sector – which the Badombi claim to belong to – and the Bafwasende sector – which the Bevenzeke claim to be part of – is clear and shows that the portion of forests the two groups are fighting for are in the Lubuya-Bera sector. I could have stopped my inquiry there. After all, the ubiquity of maps in these communities’ discourse – I come back to this below – might show their legitimacy as they have been re-appropriated and have helped some of them to confront an exclusionary colonial past and claim their rights against extractive companies. Of course, as we have understood by now, this would not illuminate the complex socio-spatiality of this landscape.

In that spring of 2016, I proceeded to get to the story of the Alibuku landscape, home to the COTREFOR/IFCO concession and of the communities dwelling to better understand the root of a conflict that they had repeatedly experienced over the previous 10 years.

106 Interview 36-ITSP-KIS-20170505. 127

Figure 16. Landsat 8 satellite map of a part of COTREFOR/IFCO Alibuku concession, from 10/05/2017. The mapped RDZ was totally reshaped as it extended outside the boundaries that were expected to be reached in 2039. It also shows that the zone extended in unpredicted areas further away from the homogeneous planned RDZ. I highlighted some of the RDZ’s unforeseen extensions with a red circle. The written indications on the map were added by one informant who provided this map anonymously.

128

5.2. Alibuku landscape: a story-so-far

As Vansina (1990) had already long ago noted, there is a tendency to amputate forest societies of the Congo Basin from their pre-colonial past. In this way, there is a disproportionate amount of studies on the Pygmies of Central Africa, framed as the ‘real’ indigenous people of the region, to the near total exclusion of the Bantu forest dwellers, for instance, who constitute the largest group in the region. Western Bantu societies are too often portrayed as one cultural and political monolith, surviving in and determined by a hostile environment, organized in small communities ruled by hierarchical, even authoritarian institutions of chiefs and kings and where little innovation or change happened before the colonial era. Although the equatorial region of Congo shared a common ‘tradition’, Vansina (1990) argues that there were lots of sub-regional variations that were later ignored by the colonial ruler, and institutions continuously evolved to accommodate socio-environmental challenges, triggered by ‘local’ phenomena as much as by external actors; it was a changing continuity.

Many people, in the contested Alibuku landscape, have recounted the following story of the vast forests of Alibuku. It starts with the encounter between the ancestral family of what is today called the Bevenzeke ethnic subgroup, which is part of the Bali ethnic group, and with the Badombi ethnic subgroup, which is an entity composed mainly of members from both the Kumu and Manga ethnic groups107. It continues with forced resettlement, a rebellion and commodity resource extraction. The story as I tell it here comes from different interviews in which my narrators often engaged passionately. I use the names Badombi and Bevenzeke, and of other groups like Bawi or Topoke because that is how people named their belonging groups during the interviews I conducted with them. However, it is possible that these groups came to identify themselves as a unity under the Belgian colonial administration which, for administrative and resettlement purposes often assembled smaller, non-related communities in chefferies*, sous- chefferies*, i.e. two sorts of administrative units; or inversely, divided bigger communities, for instance when they were considered as troublesome. As I mentioned earlier and as the story will show, group identities are also created or reactivated when political and economic interests are at stake (see Turner (2007) for an example of similar dynamics).

To better understand the history of the Alibuku landscape (Figure 17) and better ground the analysis, I provide important elements of contextualisation that characterize the groups of the larger hinterland of Kisangani. These elements are based on both secondary material and interviews I conducted in the larger area of study, i.e. not limited to the Alibuku landscape. Regarding precolonial history, I back up some of my data gathered during interviews with with Vansina’s (1990) seminal book Paths in the Rainforest in

107 It is unclear which administrative entity the Badombi group corresponds to. On maps, Badombi is indicated as a village, localité*. However, in reality, the Badombis are spread in different locations and therefore often identify themselves as a ‘community’. 129

which he reconstructs the social history of Central African forest societies. In the absence of written documents and little archaeological sources, Vansina has performed such reconstruction through the “words and things technique” (p.11-16) that associates large amounts of linguistic and ethnographic data to trace the key institutions, ideas, technologies and relationships of these societies. It is important to note that Vansina’s approach, while widely praised, has also drawn controversy and contains some weaknesses (see e.g. Macgaffey, 2013; Newbury, 2007) – as any method in fact. Yet it remains one of the most detailed historical account of pre-colonial life in the Congolese rainforest, and many of its elements corroborate the data I collected. In line with some of the thoughts that have extensively explained in my epistemological-methodological chapter, I certainly do not claim that the reconstruction of the ‘story-so- far’ of the Alibuku landscape is complete. As Van Reybrouck (2012) would say, it is a story ‘reassembled’ on the basis of multiple local voices from different groups inhabiting a landscape where new tensions and conflicts have been generated by resource development.

A final note: while the subtitles seemingly follow a chronological history, in telling this ‘story-so-far’, I in fact continuously integrate elements from the past and from the present to highlight the ‘duration’ of the landscape and the entanglement of multiple periods in shaping relations to land and forests. Grasping this ‘story-so-far’ is not only key to understanding the impact of overly spatial REDD+ landscape approaches materialized in the formulation of COTREFOR/IFCO’s Forest Management Plan, and the conflict that it fuelled,to understand how landscapes are built in connectivity between human and nonhuman actants.

130

Figure 17. Excerpt from an ‘official’ map of Kisangani and its direct hinterland (UCL-Géomatique, IGC, RGC, & AfricaMuseum, 2011). The black shape corresponds to what I refer to the ‘Alibuku lanscape’ (the conflictual zone) I have investigated and has be drawn by myself.

131

5.2.1. Alliances, fluid appropriation and forests of belonging

The story begins before the Zanzibari-Arab colonial conquests. At that time, two sages*108 told me109, people and villages moved frequently as it was part of their culture, tradition and system of food production (shifting agriculture and hunting). Such movements, they told me, apparently became more frequent to escape the Zanzibari-Arab slave raids around 1880-1890.110 “Our story starts with elephants, with the hunting of elephants”, Albert, who identifies himself as a descendant of the Badombi, narrated.111 “When our ancestors were chasing the elephant, they found a virgin forest”, Albert continued, “and when they killed the elephant in this virgin forest, they preferred not to go back where they came from”. The Badombi ancestral family, as the story tells, quickly encountered the ancestral family of the Bevenzeke ethnic subgroup, Agokologo, who was an “arrow launcher” hunting chimpanzees.112 They recognized each other’s language because the Kumu and Bali ethnic group from which they are descended from, had signed a blood oath after many years of ‘tribal’ wars. As the chief of the Bevenzekes told me113, “Kumu and Bali, they are uncles, we cannot spill blood”. It emphasizes the importance of their ties as the term ‘uncle’ – Noko* in Lingala – seems to have appeared around the 12th or 13th century in the inner Congo basin to stress the socio-political alliance and equality between two founding groups (Vansina, 1990).114 The two families thus started to live side by side in these sparsely populated forests rich in game. A Badombi man then married a Bevenzeke woman. Ties between the two families became even stronger. As part of the traditional extended family relationships, the brothers of the Badombi man or of the Bevenzeke woman – this varied according to which group I was talking to, as this is a key part in the claim for primary occupancy because it is a patrilineal system in which land is passed down generationally through men only – came to settle in the territories of their in-laws. The brothers were given a large portion of forest and could carry on their livelihood activities as they pleased. All families lived together in peace.

As other scholars contend (e.g. Ndaywel è Nziem, 2012; Vansina, 1990), before the colonial period, the socio-political organization of the forest regions in this part of DRC was indeed characterized by alliances and absorptions formed between different families and groups through marriages, wars and trade. Wars and competition for resources could also cause the emigration of some groups rather than their absorption, and thus resulted in the fragmentation in plurality of villages. Alliances and the resulting

108 Traditional sages are older people who are responsible for passing on oral history and traditions to young people of the village. They are also usually consulted when important decisions have to be made in the village. 109 59-I-RAB-20170511. 110 This is not part of the specific story I am telling here but it provides an insight into some of the destructions these socio- ecological landscapes have gone through. 111 54-I-RAB-20170509. 112 59-I-RAB-20170511. 113 58-I-RAB-20170510. 114 The term ‘uncle’ is still of major importance to describe socio-political relationships between some groups. My research assistant uses it to describe particular groups to which his own kinship group is tied. Congolese also generally refer to Belgians as their ‘uncles’ because of what they see as a common political history, despite it being a domination rather than a tie. 132

socio-political organizations were dynamic and flexible by definition (Vansina, 1990). The communities created through alliances were very variable in terms of size, level of hierarchization and political organization. However, they were generally characterized by decentralization – with the village as the main level of power – but cooperation between village settlements – in some cases larger chiefdoms and kingdoms were created although it did not mean they were regrouped ‘physically’. The Eso kingdom or Kombe chefferie, which today is referred to as Topoke ethnic group, in the Isangi territory seems to have been such a larger kingdom according to some of my informants, although political power was only exerted at local/village level.115 Leaders of villages or larger entities were generally chosen for their conciliation skills and their capacity to protect others notably through physical strength and supernatural powers.116 These villages elaborated common ideologies of shared kinship from a common ancestor and of first occupancy to encourage cohesion, as illustrated by the story of the Alibuku landscape, as well as common ritual institutions. For instance, as Vansina (1990) asserts and as several of my informants117 recounted during interviews, initiation and circumcision of boys would happen at one time in the sacred forests. This cohesion was thus partly virtual and cognitive, rather than physical (Vansina, 1990). These webs of social relations were also key in diversifying livelihood strategies (Ndaywel è Nziem, 2012). For instance, in the Isangi territory, there have always been strong alliances and commercial exchanges between the Lokele “riverside residents”118 (riverains*) who are fishermen and traders – they are “on the side of the water”119, and the Topoke people of the land (terriens*) who are farmers – they are “on the side of the forest”120. Fish could thus be swapped for manioc.

The forest environment and its related production modes both played an important role in shaping this type of relatively small, decentralized socio-political organization. Hunting, as the story of the elephant shows, and the decreasing presence of game animals who would avoid the vicinity of villages after a few years, often led a village to resettle so as to limit the distance between the settlement and the hunting area. Similarly, shifting agriculture which became the dominant mode of production in these tropical forests, encouraged movement. Each year, households would open a new principal field in the forest to ensure high fertility. With the passing of years, the distance from the initial settlement to the main field would increase, causing a whole village to start a new settlement closer to the new field. Movements could also be triggered by overpopulation although low densities of population were generally the norm in this region of Congo (Vansina, 1990). These movements should not be understood as long-distance migrations but rather a dynamic use and appropriation of land. The landscape was thus shaped through the interrelated movement of humans and nonhumans, i.e. the movement of animals, the availability of

115 28-I-YALO-20160322, 29-I-YALO-20160322,40-I-YALI-20160326. 116 51-I-YAF-20160330. 117 E.g. 8-I-RBU-20160307, 28-I-YALO-20160322, 34-I-YALO-20160324, 59-I-RAB-20170511. 118 2-IFG-YAE-20150520. 119 29-I-YALO-20160322. 120 Idem. 133

non-timber forest products or the cycles of soil fertility were completely intertwined with human’s dwelling and social activities. Their paths in life were interdependent and co-constructed.

With such a level of decentralization in the political organization and movements, clans became important units to regulate territorial access to ancestral forests. The clan was and is a spatial unit within a village that is composed of kin related nuclear families descending from a founding ‘father’ – a ‘big man’ as Vansina (1990) calls him – although in ancient times it could also include friends, clients and/or various dependents. Decisions were made between relatively local-level leaders. Moreover, even if villages were physically impermanent because of these movements, cognitively the former village was perennial, i.e. “there was the continuation of the former village and its territorial rights” (Vansina, 1990, p. 78). In this semi-nomadic system, land appropriation largely happened through the inhumation of the common ancestors that then gave rise to sacred forests and rivers (Ndaywel è Nziem, 2012). Through the stories that were recounted to me, forests as hunting and gathering domains, as sites of worship and traditional rites (e.g. boys’ circumcision) and as symbols of power and prestige – the more forests a village/district had, the greater its power – in fact appear to have played an important role in the past, in creating cohesion between the villages and families despite this semi-nomadic way of living. Forests were communal to the network of groups created through marriage alliances. A sense of permanence, although false in terms of physical reality because of the fragmentation of the ancestral lands, was thus essential to social and ritual life. The relationship to land was social and spiritual.

As informants told me, cemeteries, the sacred and ancestral names given to these natural elements and transmitted through oral histories and dwelling in the landscape, as much as the rests of villages or campements* – ‘camps’ in English, which refers to the temporary settlements used during fishing, hunting, gathering but also harvesting seasons – all became signs that the land was in use and thus appropriated by a group.121 A dweller of the Alibuku landscape on the former Buta road said in that regard: “the story is that our grandparents built camps to go trap [hunt]. Then they would leave the camps. They also made traditional ponds. The camps are what marks the boundary between the Bamanga and the Bali [two ethnic groups].”122 Because many different dialects characterize ethnic (sub)groups, naming seems to have played a significant role in marking territory too as Georges, a Bawi (a subgroup part of the Abata groupement*) tried to explain to me during a conversation in which we discussed the arbitrary nature of names. It is worth quoting us at length as I believe it is very insightful to understand the inseparability of the socio-cultural and natural in local worldviews:

Me: you said there were signs in the forest, that were left by your ancestors?

121 E.g. 57-I-RAB-20170510, 63-I-ARBU-20170515, 74-I-RBU-20170522. 122 63-I-ARBU-20170515. 134

Georges: Yes there were signs. You went there. You saw. We have historical sites there, where our forefathers lived. Me: but you did not cite any historical site, you just cited rivers. Georges: Rivers are traces for us. There are traces that justify our presence… Traces that we lived there. Me: the traces? Natural elements? Georges: Yes, natural elements. Me: but… everybody can know the names of rivers, it does not prove that it belongs to a group. […] A same river can be named in different languages. Georges: No, it is the first naming of the river… it’s with the naming of the rivers that we make sense [of our territory]. Me: so you justify this [your customary space] through the way rivers are named? Georges: yes… […]. If you make a discovery, who is going to attribute a name? Who is allowed to attribute this name? Me: the one who discovers. Georges: in the same way, we, who were over there, our ancestors who lived over there, they attributed the names to our rivers. That’s the reason. Me: they transmitted this information through oral history? Georges: yes, exactly. Me: how does it work this transmission? Georges: you know, we had grandfathers who told [the story] from generation to generation. For instance, now, we have grandchildren, we tell them that, over there, is where our ancestors lived. We tell them the names of rivers: Todi, Ngadi, for instance. Children have to know these names, over there.123

Socio-natural elements thus played and still play an important role in creating and maintaining a sense of group belonging. To paraphrase Tsing (2005, p.198), every time people look at the vegetation, people recall nodes of sociality and their social history; and use these stories to negotiate current access and other social responsibilities. The memory of ancestral lands that expresses a social identity and belonging to a group is very much present among Congolese people even today, and each person is attached to an ancestral land where his/her group’s ancestors are inhumated; they are ‘shadow communities’ that exist in the evidence of the forest (Tsing, 2005). Historical campements*, as I show in chapter 6 more extensively, still play an important role in the land use practices of the Kisangani hinterland’s inhabitants. The collective memory of land and group’s history, that is the knowledge of both the social and physical worlds, is transmitted through oral tradition but also through movement and dwelling in forests for livelihood activities. Men bring their male children early on during their hunting and gathering activities to show the natural elements such as rivers, trees or hills that have historical and social significance to the group. “A child grows up with this knowledge [of our forest] […] It is like that”, the chief of the

123 74-I-RBU-20170522. 135

Bevenzeke groupement* told me.124 “If we do not go with children over there, where our forebears lived, we will lose the history”, Georges the Bawi said.125 In this patrilineal system in which women have to move from their birth homes to their husband’s households and cannot inherit, women are largely excluded from such knowledge transmission.126 Hunting is reserved to men; “here, a woman is not allowed to hunt”, I was told during a participatory mapping exercise127– Women are generally more tied to the house, fields and near secondary forests for gathering, and hence tend to have less extensive land use patterns and thus less access to the knowledge of the group’s forests acquired though the experience of them.

Forests in this collective imaginary are intimately linked to social and material reproduction. Forests represent a link between generations as they are the place where the memory of previous generations and traditions is kept as much as a material asset that provides livelihoods for the present and future generations.128 Forests are also important symbols of power as the abundance of primary forests a group holds reflects the greatness of ancestors – e.g. their ability to form alliances, their conquest of new territories through wars or through occult powers. Nestor, a Topoke in Isangi told me to explain the Topoke’s vast forests: “If you were weak, you didn’t have forest. Thus, our ancestors who were warriors robbed forests to the other ones who were weak. […]”. 129

While social identities are to this day very much related to ancestral forests, the Alibuku landscape and the Kisangani hinterland’s landscape more generally were thus characterized by a high level of territorial fluidity and socio-political structures that were not very hierarchical but mostly based on kinship alliances and autonomy. Under such conditions, which were common to much of tropical Sub-Saharan Africa, customary territoriality consisted in ‘using’ and ‘dwelling’ in the land for production, social, ritual and spiritual purposes rather than ‘holding’ it as a commodity (Pottier, 2005). Competition for resources and power definitely happened, e.g. wars, but access to forests was continuously negotiated in time and space to adapt to changing environmental and socio-material conditions, e.g. through new alliances or, the opposite: fragmentation, rather than being permanent and static. Autonomy, competition and negotiation were all sides of the same coin. The absence of defined land rights convinced the anthropologist Paul Bohannan that the term ‘tenure’ was a Western term that cannot be applied to many regions of Africa, except for some densely populated areas or “areas of high agricultural fertility in East and Central Africa” (Pottier, 2005, p. 58). In the same way, generalizing the definition of ethnic identities attached to bounded territories and spatial distribution, as the colonial administration would later do,

124 58-I-RAB-20170510. 125 74-I-RBU-20170522. 126 39-I-YALI-20160326. 127 13-PM-ARAB-20180514. 128 I will come back to such forests’ values many times again in this second empirical part. 129 28-I-YALO-20160322. 136

was not possible. In fact, the diversity in socio-political organizations in the Congo basin renders any kind of generalization difficult. The next part of the story explores how such fluidity was containerized under the colonial rule. The next part of the story explores how such fluidity was containerized under the colonial rule.

5.2.2. Bounding forests and identities

The Belgian colonialists arrived. Willy, a wonderful storyteller from Azunu, a village in the forest landscape of Alibuku shared the story with great details on the basis of what his father told him.130 At the time – the dates are not clear but Willy mentions a period between 1928 and 1935 – the chief of the nearest colonial station and his team were navigating the Lindi and Tshopo rivers and tracing forest paths, with their big boots, to identify forest dwellers. They carried binoculars which helped them spot the dwellers in these dense forests. When they arrived in an inhabited place, the villagers would prepare a camp for the colonials who would stay for a few days. They requisitioned literate people – les lettrés* – to circle the neighbouring forests. Willy tells me that the Belgians said to their aids: “go find these people and ask them who they are, ask them the name of rivers and streams, and note them down”. This information would help to draw the maps my informants showed me so many times. During these expeditions, as many people told me, the Belgians found the Badombi and the Bevenzeke. However, as the archives show, the Bevenzeke were registered as a groupement X* of about 7 000 people (Moeller, 1936). Their name, as Willy told me, just disappeared. As the Belgians were constructing a road – today known as the ‘former Buta road’ because it was replaced by another larger one, later on – they resettled part of the Bevenzeke along it, next to the village known as Alibuku – which is today the entrance to the Cotrefor/IFCO concession – where they were grouped with other communities including some Badombi. Willy again explains: “we had our forest here but we were ordered to stop cultivating and eat all the remaining food. ‘You cannot live here anymore’, the Belgian said. We did not understand how the Belgian could take us where we were all bunched up like this […]”.131 Ayola I, Albert’s (whom we met earlier) grandfather (a Badombi), became the chief of this heterogeneous chefferie* in Alibuku.132 Access to forests was restricted and only the actual villages and surrounding farmland were recognized and registered as indigenous land. Decrees (see Picture 2) established the new limits of each chefferie* and sous-chefferie*. The vast forests the Bevenzeke and Badombi had dwelled in for centuries and that they considered to be their common heritage, were grabbed by the colonial state through their registration as vacant lands (see also Peemans (2014) on this topic). The ancestral alliances were untied.

130 61-I-RAB-20170513. 131 Idem. 132 54-I-RAB-20170509. 137

The Belgians thus dissolved the Bevenzeke and dispersed them to Alibuku and other areas as they were identified as a threat to public order. The Bevenzeke, several of them told me, were part of the Anyota* or ‘leopard-men’ phenomenon. Anyota* was a secret society that originated in the Bali ethnic group as a movement of opposition to the colonial power and in particular, to its occupation of ancestral lands and to the colonized people who were too keen to cooperate with the colonials. The Anyotas* ritually killed chiefs from rival communities and to a lesser extent white colonials, leaving traces that gave the impression they were killed by leopards.133 The Bevenzeke and more generally the Bali ethnic group have, to this day, the reputation among other groups of being powerful due to their use of witchcraft and grigris*. The chief of the group, with whom I often had discussions, often wears a leopard-skin hat in reference to the Anyotas* he told me. The leopard has and continues to be a major emblem of political and spiritual/occult power since a much longer time (see e.g. Vansina, 1990). The Bevenzeke were thus scattered among multiple villages by the Belgians and many died.

By the beginning of the 1950s, a new road – the present ‘Buta road’ (RN4) or ‘Buta road’ going up North from Kisangani – had opened. It instigated a second wave of resettlement melded with the forced mobilization and modernization of indigenous peasantry. Some of the groups who had been put on the ‘former Buta road’ and who did not have sufficient space to farm – including some Badombi clans – were once again displaced and resettled in paysannats* along this new road (see Figure 17). As many informants described134, villages were rationally organized by the colonials, in corridors or agricultural blocks alongside each other that were then divided into clans – a kin group who claim to descend from a common paternal ancestor, thus a group of related nuclear families, and households, 50m2 per household (see an example in Picture 4). The limits of each paysannat* were drawn on maps. As I will show later, the legacy of such individualization of land rights to this day defines the division of primary and secondary forests135 and cultivated land between villages, clans and households, in most communities I visited along historical (colonial) public roads, i.e. to be distinguished from more recent private roads like logging roads. The ancestral communal forests that were deemed ‘vacant lands’ decades before remained in the Bevenzeke’s and Badombi’s collective memory.

While the Badombi were removed from the ‘former Buta road’, the Belgian administration installed Congolese World War II veterans along this former Buta road on an area of about 4000 hectares. The veterans, who were part of the same military group but from different ‘ethnic (sub)groups’ were offered their own ‘customary’ land and had their own villages and chiefs.136

133 Anyotas* have been represented in colonial ethnography and official reports as barbaric and cannibals. In the stories I heard, some people referred to a group eating a white Belgian. It is, however, difficult to distinguish whether they are real, or symbolic stories told to reinforce the powerful image of a group, or internalization of colonial imagery. 134 3-I-RBU-20160304, 7-I-RBU-20160305, 9-I-RBU-20160307, 10-I-RBU-20160307, 51-I-YAF-20160330. 135 I come back to these categories in section 5.2.4 and chapter 6. 136 10-IFG-RBA-20170519. 138

The extended and fluid territoriality and the collective social and spiritual belonging to the land, characteristic of the pre-colonial period, were thus dramatically shaken by the colonial administration. The Belgian colonial rulers imposed a form of ‘ethnic governance’ or ‘ethnogovernmentality’ (Hoffmann, 2019). The latter worked through a new regime of customary law that encompassed the territorial and social regrouping of people into chiefdoms ruled by traditional chiefs – who had to be recognized by and had to act in compliance with colonial rule – and strictly subdivided and bounded spatially (Verweijen & Van Bockhaven, 2020). As one of my older informants (born in 1943) from Yafunga in Isangi territory recalls from his father’s stories, “because in Yafunga we were all dispersed, the [colonial] State reunited us”. People were in fact forbidden to settle in remote areas of the chefferies* until the 1950s (Likaka, 1994). “When the white came and [saw] there was a multitude of notables*137, he saw fit to establish one person as chief”, he continued.138 Only in some cases were some groups deliberately kept divided. As Verweijen and Van Bockhaven (2020, p. 5) assert “[t]his tendency was informed by [the Belgians’] own society’s orientation towards centralized, hierarchical, territorialized forms of political organization sanctioned by tradition”. This process imposed an abstract, geocoded and “containerised” (Vlassenroot & Huggins, 2005, p. 122) view of territory and identity, and succession and inheritance were standardized with primogeniture imposed (Vansina, 1990). At the same time, through forced and voluntary migration for industrial and commercial agriculture purposes in the colonial period, thousands of people were also “moved permanently out of their indigenous rural sphere” (Vlassenroot & Huggins, 2005, p. 120). As from 1910, Congolese people needed a passport to move around the territory. They could only obtain this passport with the authorization of the village or chefferie* chief. Over the next forty-five years, the colonial administration tried to limit rural exodus by restricting people’s mobility and reinforcing a sense of belonging to the indigenous community (De Clerck, 2004).

The land that was not recognized as customary lands – terres indigènes* – because it did not show obvious traces of inhabitation, cultivation and exploitation use became so-called ‘vacant land’ – terres vacantes* – and customary ‘ownership’ was not recorded. Notwithstanding the extensive forms of land use – hunting, gathering, regular resettlement – and the spiritual and social importance of ancestral forests, these vacant lands became state capital that could then be sold to white settlers or commercial corporations, under a ‘modern’, as opposed to customary, system of property rights (Huggins, 2010; Pottier, 2005; Vlassenroot & Huggins, 2005). Under Leopold II’s Congo Free State, this law also provided the basis for forcing Congolese villagers to collect forest products, such as ivory, wild rubber and copal, for the benefit of the state or the concessionaries who had acquired ‘vacant land’ (Clement, 2014). Later, compulsory cultivation schemes that forced peasants to grow cash crops for the benefit of the Belgian

137 Notables* were and are traditional leaders in a community that can be chosen for their affiliation to the original ruling family but mostly for what are considered to be social, intellectual and spiritual qualities (e.g. people who went to high school or who have a formal job). 138 51-I-YAF-20160330. 139

colonial State and European agribusiness companies, and the seemingly voluntarily system of paysannats indigènes* – see Part I – would further constrain people’s livelihoods and movements with consequences to this day.

In this way, the Bawi living on the ‘Buta road’, where the Belgians relocated them, “cannot go further than 1-2km inside the forests […] because they are so many” and there are blocked by the limits of the then paysannat*, Dieu-Merci told me. Another dweller who preferred to remain anonymous told me that on the maps “you will see the villages along the Buta road”, but what they consider to be “their forests have remained where they came from”. 139 When the Belgians placed the Bawi group on the big road, other groups had already been resettled and the Bawi were “wedged” between these groups on lands that were not theirs.140 To this day, the Bawi consider themselves to be venants*, newcomers141 (albeit with full user rights) on the land they occupy on the ‘Buta road’ as their ancestral forests are further away as shown on the participatory map I conducted with this group, in Figure 18.142 In pre-colonial times, the semi-nomadic clans of a village would split to create new aggregates of families when their population grew too large. Alternatively, entire villages would move to occupy new lands or fallows they had left for decades when the distance between settlement and principal fields became too great. Colonization violently ignored the flexible structure and mobility of clans.

Land was also often allocated unevenly by the Belgians. As it was explained to me during another participatory mapping exercise on the ‘Buta road’, although “the whites divided in function of the number of people [in a clan and village]”, villages and clans who were resettled first and last were the ones who received more land.143 Yet, the politico-administrative position of some clans’ leaders in the colonial administration also explains why some groups were allocated more land than others.144 Batiamaduka (KP13-18), for instance, received quite a big area of land compared to others because it was one of the first villages to be relocated on the ‘Buta road’, its inhabitants told me. They also revealed that the village also produced the first Chief of the Lubuya sector –in which these villages are situated. He could thus influence the repartition of land. “The chief of the collectivity-sector is from Batiamaduka; it’s him who walked with the whites to distribute lands”, said Pastor Mutambaï from Batiamaduka.145 The same would happen at the level of the clan. Gaspard, an inhabitant of Batiamaduka, for instance, explained that his clan inherited large areas of land (about 20 hectares according to Gaspard) because his grandfather was

139 63-I-ARBU-20170515. 140 64-I-ARBU-20170515. 141 I explain these categories in next section 5.2.3. 142 Note here that the participatory map also includes a multitude of villages, for example on the ‘former Route Buta’, that are not recorded on official maps (see e.g. Figure 17 for a comparison). 143 4-PM-RBA-20160311. 144 See also Vinez’s (2017) study which notes similar trends. 145 15-RBU-20160309. 140

a kapita* under the colonial administration, i.e. a village chief designated by the colonials rather than by tradition and who was in charge of having the orders of the whites carried out.146

The Christianization conducted by the ‘White Fathers’ and other missionaries during the entire colonial period also contributed to the erosion of the spiritual and social significance of forests. As traditional animist beliefs and rituals, ancestor worship and witchcraft practices were repressed, progressively merged with or simply erased – although never totally – by Christianity, ways of relating to forests and to their multiple nonhuman and more-than-human dwellers profoundly changed. When I asked questions about sacred forests, many insisted these were part of a past tradition that had been abandoned because they had “followed the Word of the Lord”, “everybody starts to pray”.147 When asked about the importance of forests, one said: “There are two ways of answering. For people who do not believe in God, they know their importance […] but for us who pray, such as me who prays to God, I know forests are sacred for tradition but I do not see their importance”.148 Yet, most so-called Christian religious movements remain largely syncretic and it is difficult to know to what extent the spiritual essence of these sacred forests has been eroded as my informants might have wanted to hide such information to a non-initiated white. In any case, the importance of sacred forests is often (strategically) called upon when claiming primary occupancy and a group’s belonging to particular forests –see section 5.2.4.

To summarize, while the Belgian colonials claimed to respect existing customs, in practice, the large diversity in socio-political organizations, the fluidity of access to forests and the extended dwelling spatiality were all ignored and annihilated, and Congolese people’s land and labour were alienated. It would have profound, ongoing consequences on socio-ecological systems and people’s ability to adapt to environmental change, as I show in the following sections. Many scholars who research the very conflictual Eastern part of DRC have shown how the colonial institutionalisation of the link between a bounded ethnic identity and land access are a sustaining factor of violent conflicts (e.g. Huggins, 2010; Pottier, 2005; Turner, 2007; Verweijen & Van Bockhaven, 2020; Verweijen & Vlassenroot, 2015; Vlassenroot & Huggins, 2005)149. As Vansina (1990, p. 239) tragically puts it “[c]ustomary law was the headstone on [the] grave” of forests societies on which the colonial’s own cognitive view had already been imposed on every aspect of their daily life. But let me go back to the story which continues in the post-colonial period.

146 11-RBU-2016030, 51-I-YAF-20160330.. 147 E.g. 40-I-YALI-20160326, 46-I-YAF-20160329, 11-PM-RBU-20180503. 148 57-I-RAB-20170510. 149 Such studies, to my knowledge, do not exist for the central forest regions, such as the Tshopo Province although conflicts around land fill up civil courts. 141

Picture 4. Aerial picture of the paysannat* agricultural blocks in the Belgian Congo (Malengreau, 1949)

142

Figure 18. Participatory map conducted with the Bawi ethnic subgroup (03/05/2018). It emphasizes the distance between the ‘official’ village Bawi on the Buta road (here identified as Nouvelle R. Buta) where the group was displaced during the colonial period and that is taken up on official state maps, and their ancestral forests.

143

5.2.3. Replenishing forests and identities

During the long decades of colonial rule, the vast forests of the current Alibuku concession thus became unoccupied, emptied of its inhabitants through several waves of displacement. The Bevenzeke allegedly started to come back to the area of Alibuku in 1961, two sages* told me.150 They were, however, no longer identified as a group since it had been dissolved by the colonial administration. The Badombi who had stayed, welcomed the Bevenzeke and the two groups “peacefully cohabited”.151 In 1964, the eastern Simba152 rebellion in opposition to Kasavubu and Mobutu, and the counter-attacks supported by Belgium and the United States to free Stanleyville, which became Kisangani in 1966, killed thousands of people and imposed a climate of terror that pushed people – and probably rebels, too – to take refuge in the dense forests. These movements from the city back to the forests continued over the next few decades with successive “Liberation Wars”153 that had some of their worst fighting in Kisangani and with the long decline of the Congolese economy. People who escaped violent conflicts or who lost their jobs settled in the rural areas around Kisangani and returned to agriculture as they could not sustain their livelihoods in the city. Many did not return to their ancestral lands, which were sometimes hundreds of kilometres away as they had been forced to migrate under the colonial administration or decided to migrate during Stanleyville’s peak urbanization and industrialization period (see e.g. Barou, 2000). They thus settled on the nearest and most accessible lands and forests that were deemed fertile and rich in forest products. But, my informants insisted, they knew that these forests must belong to a group even if they seemed empty.154 In practice people told me, when land was occupied, it was easy to gain access to land by just negotiating with the customary land chief. However, these ‘newcomers’ were (and are) considered as venants* or locataires*, i.e. ‘tenants’ with no customary ownership of the land they use – I will come back to this.

During our conversations, a lot of these venants* – best translated as ‘newcomers’ – insisted on the Congolese constitution that asserts that any Congolese can freely go and settle anywhere on the State’s territory. “Same father, same mother, same country, you can settle where you want” said one of them.155 A Bevenzeke similarly said to justify their return to their ancestral lands: “the Congolese constitution says that everybody is free to come back to where he came from. That’s why we are here”.156 Such mobility

150 59-I-RAB-20170511. 151 61-I-RAB-20170513. 152 Simba means lion in Swahili. The rebels believed that, through shamanism, they could transform into lions during battles and that their adversaries’ bullets would be transformed into water. This ideology can be found today among the Maï Maï rebels in Eastern DRC. 153 “Liberation Wars” refer to 1) the rebellion mounted by Laurent Désiré Kabila and his Alliance of Democratic Forces for the Liberation of Congo (AFDL) to push Mobutu out, in the period of 1996-1997 and 2) the rebellion mounted by the Rally for Congolese Democracy (RCD) and Movement for the Liberation of the Congo (MLC) to remove Laurent Désiré Kabila from office (1998-2003). Kisangani was the stage of violent confrontations between opposing factions. 154 57-I-RAB-20170510. 155 62-I-RAB-20170513. 156 59-I-RAB-20170511. 144

of Congolese people away from or back to their ancestral lands was encouraged by the Zairianisation state ideology – sometimes called Authenticité* – and ‘modern’ law that Mobutu introduced as from 1973. As an autochthonous inhabitant said: “after independence, President Mobutu said we need to be united; it is in that period that newcomers started to arrive”.157 Zairianisation was an effort to get rid of the vestiges of colonialism, its ethnic-based administration and the strong restrictions on people’s mobility, in order to establish a singular national identity, notably by limiting the influence of traditional chieferies and declaring the territory as belonging to all Congolese. However, new networks of land control based on alliances between new rural market actors, politicians, soldiers, administrators and representatives of the customary system developed, and further eroded the customary social organization and economic structure (Vlassenroot & Huggins, 2005). The double system of land access/tenure introduced during the colonial era was in fact maintained and used “opportunistically by those able to do so” (Vlassenroot & Huggins, 2005, p. 123), as is still the case now. While poor ‘newcomer’ peasants were considered as ‘tenants’ because they could not afford buying property rights, a privileged class of elites close to the Mobutu regime acquired lands and took possession of former colonial plantations, factories and other commercial and productive assets, often leaving them under-utilised. For instance, the oil palm plantation ‘Belgika’, neighbouring the Alibuku logging concession, was given to a general in the beginning of the 1980s, as informants living in and around the plantation told me.158 After some years of continuous decline in the palm oil production, this general started to rent parcels of the plantation to individuals of the neighbouring communities – both ‘autochthones’ and ‘allochthones’ who came to work in the plantation during colonial times – who also had to pay to access processing facilities by giving part of their production to the general. As in the Alibuku landscape, these communities were and are made up of ‘autochthones’, i.e. registered by the colonial administration as customary chiefs of the villages and lands bordering the plantation, and ‘allochthones’ from other territories who were workers on the plantation during the colonial era and continue(d) inhabiting the workers’ camp, which has become a fully-fledged village. When the general died, the plantation was inherited by his brother who carried on the system until around 2010 when the autochthones rebelled and reclaimed the plantation. Since then, allochthones have to pay plantation’s access rights to the autochthons – e.g. for a parcel of about 2ha, the payment is two 20-liter cans of palm oil per month (which amounts to about CDF26 000 or $19, at market price).159

Elite predation of land during Zairianisation can thus be seen as a continuation of land expropriation during the colonial era. It further contributed to social fragmentation not only between a class of rural politico-economic elites and the rest of the rural population but also within this rural population. While

157 24-I-RBU-20160315. 158 95-I-BEN-20190514, 96-I-BEN-20190515, 97-I-BEN-20190515. 159 97-I-BEN-20190515. 145

the colonial ethnic-based land governance and the restricted mobility it imposed had forged bounded lands, forests and identities, the Zairianisation process pushed national integration, the end of customary land law and mobility. ‘Native’ rights were converted into rights of use, not ownership (Huggins, 2010). In practice, ethnic belonging which was closely tied to claims to land and forests became articulated with Congolese citizenship. It contributed to the emergence of a dichotomy which is of great significance today to understand land and forest access in most of the DRC: ‘indigenous’ or ‘entitled’ – ayant-droits* – to a particular area on the one hand, and those who are seen as ‘migrants’ or ‘newcomers’ – venants* – on the other. ‘Newcomers’ are however simultaneously ‘indigenous’/’entitled’ on their ancestral lands even if they have left them for a long time and do not use them. In this way, Didier, a Topoke inhabiting the Alibuku landscape asserted to me that “na congo, mopaya aza te” which means literally “in Congo, there are no foreigners”, whilst insisting that other groups installed on the Topoke’s ancestral lands are “venants*”, newcomers.160 Such rhetoric implies that if Didier were to go back to his ancestral lands, he would be entitled to get free access to these lands even if it means displacing ‘newcomers’ that are currently using it. Joseph, another ‘newcomer’ farmer and inhabitant in the Alibuku landscape, who comes from the Equateur Province explained this process in these terms:

“For example me, I live here but I know that my forest is there [in Equateur province]. If I leave here, I go back there and other people are settled… First I will approach them, we will discuss […] so that I can get my space back. There are allochthones, so they cannot get [my lands] and me, the owner, not use them. It’s the same for us who are at the Bali. If a Bali comes and he wants to settle here, I don’t have the capacity to ban it because I know that it is his space. […] If I leave [my lands] and there is someone settled on my land, I have the right to reclaim [it], that he gives my land back so that I cultivate it.

We always know that forests belong to other people. How do we know… […] If I have a child, I will tell my son: ‘you have to know that I am from Equateur but we live here on other people’s land, even if we use [resources], even if we carry out this or this practice, it is not our forest… we will leave one day because here we paid for these forests.’ […] If I go back home, I will not pay, because there are the forests of our ancestors, and so I will use [land] without anybody to disturb me. […] I have ancestors, the father of my father has told my father, my father has told his children, these children have told the histories to their own children, and so on. I have children who will tell their own children…”161

160 65-I-RAB-20170516. 161 57-I-RAB-20170510. 146

With such increased mobility and migration due mainly to economic decline and conflict, the vast forests of Alibuku slowly started to be inhabited again but the biggest wave of ‘newcomers’ was still to come, with commodity resource extraction.

5.2.4. The extractive commodity turn

With the liberalization and boom of the diamond trade in the mid-1980s, thousands of poor and unemployed people, and of subsistence farmers who saw an opportunity to improve their livelihoods, entered deep in the Congolese forests to try their luck. The forests of Alibuku landscape were soon scattered with diamond quarries. Some families from the Mbole ethnic group coming from Opala Territory, located on the other side of the Congo river about 120 kilometers away from their current location, were among them. Nestor, the chief of a Mbole village – Azunu – located 40km down the logging road from Alibuku, narrates.162 The first Mbole had already started to arrive in this forest during the Simba rebellion. They found a virgin forest and could settle as they wish although they vaguely heard that the Badombi were the customary owners of this forest. As jobs were scarce in Kisangani and as a Mbole “he could use the arrows” for hunting, he joined his extended Mbole family. A few years later, he found diamonds and opened a quarry for which he received an official permit. He called other Mbole and they “became numerous”. People from other groups as well: “the Mongo, the Tetela, the Ngando were all in this forest to mine”. Families became villages and each village opened its quarry. “If we read the documents of mining zones, there are like maybe about seventy [quarries] that belong to the Mbole, Topoke, Mongo…”, Nestor continued. By then, the Badombi had made themselves known to these ‘newcomer’ communities of miners as the customary primary occupants of this land and they were thus acting as primary occupants and thus as chiefs of the land. Access to land was easy and they could freely use forests and their products. Quarry owners – i.e. those who had legal permits – had to give a small percentage of the benefits from diamond trade to the Badombi. But the latter were just considered as “visitors” who would come and collect their percentage such as a tax collector would do, rather than as main occupants of the land.163 All other livelihood activities were in fact uncontrolled. The type of resource at play is indeed a fundamental aspect of access rights. For traditional subsistence land use activities such as shifting cultivation, hunting and gathering access is largely non-monetized because these activities are conceived of as local and non-commodified – the motive is subsistence rather than profit and they are carried on by poor, similar communities who sell their surplus at local markets. The products of these activities are also considered renewable and cyclical. In contrast, access to land for commodity production for external markets and profit-making activities – as opposed to subsistence – is

162 62-I-RAB-20170513. 163 62-I-RAB-20170513, 68-I-RAB-20170516, 85-I-ARAB-20180509. 147

monetized. In such activities, the resource is considered to be permanently extracted from its native environment. Dieu-Merci, a Bawi villager explains such differences:

“Using the forest for agricultural activities, it is not a problem. [For agriculture], it means for example to plant manioc, eat it without taking away our forest. What is problematic is for example with diamond extraction. In this case, follow up must be done because it is something that brings in money. […] There is a difference between the money from diamond and the one from manioc. With diamond, one becomes rich in a few minutes.”164

In many conversations I had with farmers, such forest resource use is often framed as a modicum of economic development associated with white colonialism. One farmer of about 50 years old, who lives in the former workers’ camp of the colonial oil palm plantation ‘Belgika’ – see above – told me during a walk around the former plantation’s facilities: “White people, they can transform the soil”. I asked, “what do you mean?” Grabbing some soil in his hand, he answered: “you know, they change the soil into money”.165 Talking about his former diamond quarry, Nestor said: “It was like a business, like a company… It was like when the white came”.166 This modicum of economic development is thus logically accompanied by a mode of access and tenure, a sort of proprietary ownership, that was initiated by the colonial government and further asserted after independence.

5.2.5. Two weights, two measures

In the 1990s, however, diamonds “disappeared” and all these communities of miners had to turn to agriculture.167 At that time, there was still no road in this area, there were only “narrow forest paths used by the Mbole and the Mongo to place traps […] They set up their camps in the forest”.168 The road arrived with industrial logging around the mid-1990s. AMEX-BOIS was the first company to obtain a logging permit to develop its activities in the area of Alibuku. Communities asserted that at that point conflicts had not yet burst out. Albert, a self-identified Badombi from Alibuku said169: “[w]hen the company arrived, all of us, we were considered as one community […] we were held as one person”. This, except for the ‘newcomers’ like the Mbole or Mongo who would never be able to claim primary occupancy, e.g. they are Lingala speakers unlike the claimed customary owners who are Swahili speakers, and they recognize their ancestral lands to be in other territories. “AMEX-BOIS would give [corrugated iron] panels, [food] rations and other goods to the Badombi from the Lubuya-Bera sector, to the Bamanga [from Buta sector] and to the Bali170 [from Bafwasende territory]”, Albert added. The conflict started around 2004

164 64-I-ARBU-20170515. 165 Informal conversation, not recorded. 166 62-I-RAB-20170513. 167 Idem. 168 56-RBU-20170509. 169 54-I-RAB-20170509. 170 The Babali he is referring to are the ones who would later organize as the Bevenzeke groupement* a subgroup of the Babali tribu*. 148

when the concession was inherited by another company, TRANS-M-BOIS171. As Léonard and Ferdinand said, “it is as from the promulgation of the law that specifications172 appeared […] Before that, there were verbal arrangements with communities. During that time, nothing was imposed on forest companies. [The company] cut trees without really giving anything in return, only little arrangements (rice, machetes, etc.)”.173 As the Congolese Forestry Code pushed by the World Bank to formalize the forestry sector to develop the country’s green economy, the FMP created new economic opportunities, e.g. different forms of compensations, in the local arena that pushed communities to more fiercely claim indigeneity and primary occupancy. When the specifications, i.e. the social responsibility contract part of the FMP that includes financial compensations for the communities174, had to be signed by communities in 2004, the group of Bali people started the process of (re)establishing themselves as the Bevenzeke groupement*, to claim primary occupancy and thus customary ownership of the forests TRANS-M-BOIS was harvesting. A conflict burst out although it seems that the contract’s specifications still gave compensation to the three groups (Badombi, Bamanga and Babali). As Dieu-Merci, who earlier emphasized ‘money’ as driving the dynamics of customary land rights and access, provides a similar analysis to the conflict: “People remain calm as long as the money of the logger has not yet arrived”.175 In the same way, the provincial director of the Geographic Institute of Congo (IGC) once informally told me during a conversation in 2016 that before the increase in logging exploitation in that area, they received only a few complaints; when there are lots of complaints, “it is an indication that there is business going on”.

Marking appropriation on paper

In 2011, the concession was transferred to yet another company, COTREFOR.176 Yet a new contract had thus to be agreed on with communities. This time, a whole perimeter of forests that the Badombi ethnic subgroup had always claimed to be theirs, was ascribed to the Bevenzeke. The Badombi thus stopped receiving socio-economic compensation. By then, the Bevenzeke had indeed been officially re-established as a groupement* with the help of a Bali political elite running for the deputy seat of the Bafwasende Territory at the National Assembly – Bafwasende being the territory of the Bali ethnic group of which the Bevenzeke are a subgroup. Many people in villages, representatives of local NGOs or in the administration told me that the deputy’s strategy was to enlarge the Bafwasende constituency to include more people who would cast their vote for him.177 By establishing the Bevenzeke as a groupement*, both parties’

171 TRANS-M-BOIS is one of the companies of the CONGOFUTUR group mostly owned by Lebanese capital. 172 ‘Specifications’ which is a translation from the French term cahier des charges refer to the social responsibility contracts logging companies have to sign with local communities inhabiting the concession perimeter. See point 4.1. 173 53-I-RAB-20170508. 174 This financial compensation is managed by a local management committee which decides which project will be funded with the allocated money (e.g. building a school, pay for children’s school fees, etc.), with the approval of the company. In practice, there is a lot of embezzlement and the company often does not bother to monitor how the funds are used. 175 64-I-ARBU-20170515. 176 As I have mentioned before (part I), despite the name change, it seems, however, that the very secretive owner has always remained the same despite assertions to the contrary. 177 E.g. 19-CCSO-KIS-20160317, 25-CSI-KIS-20160404, 37-CCSO-KIS-2017050. 149

interests were served. Nestor, the chief of the Mbole village Azunu, remembers: “because they [Bevenzeke] were a minority [in the area], they could not become a groupement* […] but with the many newcomers in these forests they could become a majority. It is thanks to the newcomers that they had the necessary demography to be recognized by the administration”.178 The Bevenzeke claim that this is finally a fair recognition and distribution after decades of a misrecognition that was instrumentalized by politicians – and they have made their presence symbolically clear as illustrated in Picture 5. During interviews, pre-colonial ancestral ties to sacred forests, authenticity, tradition and their role of “forest guardians”179 were heavily called upon by the Bevenzeke to justify their rights.180 While they recognize their marriage alliances with the Badombi, they are careful to frame them as a landless disparate group: “The Badombi do not have a place. We gave them this place because they are our brothers-in-law […] These Badombi they do not have a fixed place, they do not have a tribe. They speak all the local languages from here, sometimes Bamanga, sometimes Bakumu […] They are like the waste of a tribe”.181,182 In such claims, the rubble of ethno-administrative maps following waves of displacement and the official maps that were created on their basis since independence, play an ambiguous role. On the one hand, each group’s chief and notables* preciously keep and call upon specific maps and their accompanying decrees from different periods (e.g. 1906, 1928, 1933, 1935 or 1938) that are supposedly showing the right limits. Colonial maps, despite them saying very little about the delimitation of forests that were then considered as ‘vacant’, are framed in these various groups’ discourses as a sort of “white’s man magic” (Turner, 2007, p. 62) with an apparent ontological security that allow groups to claim more formal rights to land in the face of new economic stakes. This quote from an interview I conducted with the chief of the Bevenzeke probably best illustrates such common discourse:

“I am very happy to learn that you are a student at the university in Belgium. It is good, because the Belgians colonized us. It is they who know the territorial limits of the Democratic Republic of Congo, and then of each province and of each territory. We have the maps from 1938 and so on.”183

On the other hand, the many conversations I had with villagers about maps and participatory mapping exercises showed that this apparent ontological security was quickly shaken when I presented them with maps from different years and, more generally, when maps were not tied to a specific political-economic claim. During an interesting participatory mapping exercise with a group of Bawi on the Buta road (see

178 62-I-RAB-20170513. 179 58-I-RAB-20170510. 180 58-I-RAB-20170510, 59-I-RAB-20170511, 10-IFG-RBU-20170519. 181 10-IFG-RBU-20170519. 182 This quote comes from a conversation I had with a group of self-identified Bevenzeke who do not live on the Alibuku logging road but on the former Buta road. At another moment of the conversation, they asserted that the Bakumu are their uncles and that it is the Bakumu who gave them this land along the former Buta road because their own forests (in the contested perimeter of the COTREFOR/IFCO concession) were not accessible at the time. This illustrates not only long alliances but also the fact that ‘newcomers’ and ‘entitled’ are two sides of the same coin. Identities are dynamically articulated. 183 58-I-RAB-20170510. 150

Figure 18), this kind of ambiguous relationship appeared clear. A few days before the participatory mapping process, the Bawi’s chief had presented me with a colonial map to justify their claims against the Bevenzeke. I wanted to further explore their relations to maps. As I mentioned in the methodological section, I conducted such participatory exercises with the idea of uncovering the ontological insecurity of maps and show that they are a reflection of tenure at a particular time under a particular set of circumstances and a product of specific negotiations (see also Roth, 2009). The instructions were minimal: I just asked them to draw the location of their lands and forests, and neighbouring villages. During the exercise itself, I first encouraged them to draw their map on a tracing paper under which we placed a recent map of the area – which is based on colonial maps too (see Figure 17) – so as to have an accurate scale. This map does not include any specific limits between the ethnic subgroups, so as to avoid creating a conflict – once again, I was naïve. They quickly took it away as they argued that lots of villages were wrongly positioned or simply absent on the map, and that the names of several rivers are wrong because different groups name rivers differently in their local languages. The Bawi threw away their draft and started another, this time choosing to use two different maps from the colonial era (including the one they had shown to me a few days earlier) which, they said, represent the ‘real’ limits. During the three-hour process, there were lots of discussions and disagreements on the rivers presented on the colonial maps and their actual location, as well as on the limits of their forests. As one of them asserted, communal forests are delineated by landmarks that users have left over generations. In fact, the exercise revealed the difficulty of delimiting fixed boundaries and categories of occupation and that forest territories are known through dwelling and landmarks of occupation rather than through abstract representation. Yet it also confirmed the authority of maps as the Bawi never really let go of the colonial maps during the exercise although they clearly appeared to be limited by them.

151

Picture 5. The health centre built by COTREFOR/IFCO for the Bevenzeke group. It has never been in use, but it is a strong symbol for the Bevenzeke to display their customary ownership of these lands.

Practices of access in the Alibuku landscape

Maps, while they have allowed the Bevenzeke to make formal claims to land in the specific context of COTREFOR/IFCOs management plan, in fact say little about the everyday practices of access. The newcomers I met now largely recognize the Bevenzeke’s customary ownership because the official authorities from the Province and the concession’s maps have established it this way and for the easy forest access they grant. In theory such customary ownership should give the right for the Bevenzeke to control access to their land and forest. In fact, forest and land access for newcomers in search of cultivable land is very “liberalized”, as a Mongo inhabitant of the area asserted.184 The chief of the Bevenzeke sometimes comes by to ask for a small compensation, especially if people are cultivating some cash crops, but in general they let the newcomer subsistence farmers freely carry out their agricultural and other livelihood activities. In fact, because the Bevenzeke are a minority, they can only exercise little control over what they claim to be their ethnic subgroup’s lands and forests. The further one goes away from the busy village of Alibuku, which has become a commercial centre for charcoal and remains under ‘control’ of the Badombi and Bevenzeke, and from the main Amex-Bois logging road, the easier access to land for peasant newcomers is, as competition for land decreases. In villages like Azunu on the Amex- Bois road or Pumuzika on the former Amex-Bois road, peasant households who wish to build a house in the village have to be granted access by the Mbole chiefs who opened diamond quarries in the 1980s.185 In such remote villages, for farming, newcomers who open their fields in primary forests, i.e. in their understanding, forests that have not yet been opened by others for agricultural activities or that at least

184 57-I-RAB-20170510. 185 12-PM-RAB-20180511, 13-PM-ARAB-20180514, 8-IFG-RAB-20170516, 67-I-RAB-20170516, 68-I-RAB-20170516. 152

no longer bear clear traces of agricultural activities, do not have to ask to any customary owner group for access. As the road continues to be built further into the forest to allow logging activities, newcomer families continuously settle and gradually become villages, without any control from the Bevenzeke or the Badombi that still claim ‘ownership’ of these forests. Indeed, in areas further from Kisangani where densities and demand for land become much lower, primary occupant/owner communities often grant access to primary forests quite easily, and often for free. The dwellers of Azunu explained this during a participatory mapping exercise/focus group:

“Everybody can go into the primal forest. When someone wants to cultivate in the primal forest, when he cuts down the trees for his field, this space becomes his. In traditional villages, where there is a majority of autochthones, where people have their ancestral lands, it’s different, it’s more complicated. In those villages, everybody cannot access so easily and freely. But here, since there are only newcomers, allochthones, everybody can enter the primal forest and make their own field.”186

If a family wants to establish its fields in fallows or secondary forests, i.e. in their understanding, older fallows that have already been used previously by another family for its agricultural activities and bear clear traces, they will have to request access from this primary occupant household directly. Indeed each household has its ‘individual’ plots of cultivated fields, young and old fallows. As everywhere in the hinterland, once a patch of primary forest has been cleared and used for agriculture, it ‘belongs’ to the household – in fact, to the male head of the household – that has cultivated it, even when it is transformed into fallow and later into secondary forest. Farming and tenure, as I further develop in chapter 6, are thus fundamentally intertwined. To distinguish the specificity of such patches of softly appropriated secondary forest biotope from shorter fallows, that are more easily distinguished from primary forest for the lay person, I refer to them as ‘post-agricultural secondary forests’. Individual appropriation is thus based on an unwritten ‘right of axe’, that is the principle of primacy of first clearing. The planting of fruit trees, usually avocado, safou or oil palm trees187, and the installation of small animal traps often mark such soft appropriation – to be differentiated from private property or commodity (see 4.2.1). It signals to the passer-by that the space is occupied. In the more remote villages of the Alibuku landscape, the ‘right of axe’ can be exerted by ‘newcomers’ – i.e. non-Badombi and non-Bevenzeke – which is not the case in more densely populated areas. I also come back to this in chapter 6. Access to already ‘appropriated’ fallows and post-agricultural secondary forests for farming activities, as everywhere in the hinterland, is negotiated directly between the claimant and the primary occupant/owner. The former receives temporary use rights – he/she becomes locataire*, i.e. tenant of the plot – sometimes for free where pressure on land is really low but mostly in exchange for a small

186 12-PM-RAB-20180511. 187 8-I-YAF-20160330. 153

payment in kind like a chicken, a goat, a pig, a machete, or increasingly in money where pressure on land is high. Post-agricultural secondary forests, however, remain accessible to the whole clan or village in which, many inhabitants are ‘newcomers’/non-native and do not descend from the same ancestor – for activities that do not transform forest vegetation such as hunting, gathering and collecting. In these cases, ‘visitors’ just have to ask for simple permission to access the owner’s post-agricultural secondary forests.188 Old post-agricultural secondary forests in which original forest vegetation has clearly regrown returns to the ethnic subgroup’s communal domain. When a patch of post-agricultural secondary forests returns to the group’s communal domain, dwellers of the hinterland usually refer to it again as ‘primary’ forest – ngonda* in Lingala or forêt primaire* in French. Such return to the communal domain can, however, be subject to contention, in particular in more densely populated areas.

The organization and practices in more remote areas of the Alibuku landscape starkly contrast with more densely populated areas, for instance villages relocated along the ‘Buta road’ during the colonial era where ethnic subgroups’ land boundaries were rigidly defined and who are faced with increased land scarcity precisely because of their relocation along the road. Indeed, in places with higher densities and competition for land, and where older post-agricultural secondary forests/primary forests are very scarce, ‘autochthones’ assess their power over land more strongly in the face of newcomers. Land is more strictly subdivided between villages and clans, as inherited from the Paysannats Indigènes* system and its boundaries are constantly contested, sometimes leading to fierce conflicts. This is illustrated by the participatory map of land division between several villages on the Buta road, in Figure 19. The little remaining primary forests and swamp forests (which are generally not used for farming purposes), in such areas, are communally held by village communities but increasingly under pressure of alienation by some clans and households who want to claim their rights to them, sell them to newcomers, private organizations and elites. Increasingly, some households also try to ‘privatize’ the land they have claimed through customary practice (i.e. through opening a field) by obtaining some documentation signed by local chiefs and administration. Such documents do not have legal force but offer security locally (see also Vinez (2017) who has observed similary dynamics). Newcomer farmers in need of agricultural land have to “negotiate financially”189 – about $35 per season for 100m2 –190 and on an agricultural seasonal basis; when all products have been harvested, the tenant has to leave or renegotiate a new rental agreement – in lower density areas, rent duration is more flexible. Although access to land is rarely denied, permanent and extended use rights in these areas are never transferred to newcomers – unlike the most remote villages such as Pumuzika. They can, theoretically, obtain formal property rights under the statutory system but in practice, only wealthier (local/village) elites acquainted with the system could

188 57-I-RAB-20170510, 74-I-RBU-20170522. 189 3-I-RBU-20160304. 190 7-I-RBU-20160305. 154

afford to do so and it simply hardly ever happens as it also bears an important social cost – i.e. too many inequalities among kinship members. Poorer ‘landless’ peasant farmers hence prefer to settle in less ‘controlled’ areas further away from Alibuku village, deeper in the COTREFOR/IFCO concession than on the ‘Buta road’. As tenure is so intimately tied to land use, I return to such aspects in chapter 6.

155

Figure 19. Participatory map of the land division between villages and clans on the ‘Buta road’ between kilometre points 9 and 18 (11/03/2016). Although not all subdivisions are visible (e.g. forests of village Batiamaduka), the red lines indicate the division between villages while the red circles indicate the lands of each clan such as they had been rationally organized by the Paysannats Indigènes* system (on the map, it is particularly visible at the bottom centre, in the area between KP13 and KP14).

156

Everywhere in the Alibuku landscape (and in fact, in the hinterland), access differs for people looking to secure private concessions with registered legal titles, mainly urban elites. Since the building of the Amex- Bois road and its progressive extension as logging activities expand, politico-military elites from Kisangani and Kinshasa who are attracted by the area’s easy access, have also acquired large plots of land – sometimes by force – within the perimeter of the concession to carry out diverse land use activities (see chapter 2). Such acquisitions have pushed away long-time farmers from the area. Joseph, a farmer in Likoko in the Alibuku landscape explains:

“You see, when people come with money, they buy concessions inside the entity where we are cultivating. […] Someone has robbed my fields of a value of 8 hectares, where I had planted my palm trees, my avocado trees… I can show you, someone has robbed the products that were inside, I cannot go there anymore. […] If I brush against [this area] with my feet, if I only cross the limits to cut my palm nuts, I go to the Centrale prison. Where people buy concessions in the area, where we are, we suffer a lot, we suffer. We cannot talk, we are suffering.”191

Acquiring land has been, since the Mobutu era, a way of consolidating political and economic power, status and prestige (see also Peemans, 2014). Along the main logging road, as everywhere around Kisangani, signs indicating private property of military generals (Picture 6) are very common although many of these concessions seem to under-utilized as their owners are mostly in Kinshasa. The number of private concessions in the Alibuku landscape is, more generally, constantly on the rise and acquired by urban dwellers to carry on some for ‘large-scale’ charcoal production, and/or artisanal logging in combination with agricultural activities. More recently, many concessions have been acquired by Nande people – an ethnic group from North Kivu who are the commercial elite of Kisangani – who are attracted by the land’s low prices and are developing cocoa plantations. As for the Alibuku landscape, many informants, both concessions’ owners and local farmers inhabiting the area, told me that the Badombi villages sell these large plots of forested land to elites, despite the Bevenzeke having been established as primary occupants and customary owners of this land area in COTREFOR/IFCO social responsibility contract.192 Others asserted that both Badombi and Bevenzeke (sometimes together) grant and sell land to these urban elites.

For such acquisitions, cash and payment in kind are required. As an example, the local agronomist of a 100-ha private cocoa plantation told me the owner paid $50 per hectare, 1 goat, alcohol and rice to the village chief193; a payment that is, in theory, redistributed to the whole village. Whereas a temporary rental happens mainly on fallow land and is managed by individual households, the permanent sale of a

191 57-I-RAB-20170510. 192 E.g. 19-I-RAB-20160312, 94-I-RBU-20180522, 101-I-RBU-20190523. 193 88-I-RAB-20180516. 157

piece of land has to be approved by the whole clan or village, depending on whether the purchased land is situated on the clan’s or village’s communal land. As Liena, Maindo Mwanya, and Maindo (2014) have also shown in their case study of land acquisitions by Kisangani’s urban elites, prices and compensation largely differ from one area to another, according to multiple factors such as distance to the city, ease of access, the presence of primary forests – which is more valued – or simply communities’ ignorance of land value. Then, “at the land register, one pays a lot for the title, it costs much more than to get the customary lands”, the same managing agronomist added.194 Knowing how much it costs, however, is an obstacle course that I was not able to complete despite long hours spent at the land registry with different civil servants and officials. Concession owners are similarly reluctant to disclose information about (mainly unofficial) payments to administrations and officials. As one Congolese expert, a native of Kisangani, for the CAFI REDD+ integrated program who owns a private land concession himself, told me: “everything is so opaque with land estate”.195 Liena et al. (2014) make a similar observation in their study, and assert that land buyers have to make many payments at every step of the process, for which they do not obtain a receipt – no figures are provided though, precisely because of this opacity.

Picture 6. One in many signs indicating the private concession of a military general.

194 Idem. 195 56-RCS-KIS-20180515. 158

5.3. Ruderal socio-political forests

As expected, this story which was reconstituted though the many conversations I had with the heterogeneous dwellers of the Alibuku landscape and of the larger Kisangani hinterland never provided an answer to the primary occupancy question, nor did the historical – and the updated – maps of the region. The participatory nature of the mapping of community land tenure and use boundaries was intended to legitimize COTREFOR/IFCO’s forest management plan (FMP). As we have seen, through the participatory process the FMP assigned bounded areas of the concession to four ethnic subgroups (Abata, Baangba, Bevenzeke and Boumbua), to the exclusion of multiple other groups (e.g. the Badombi, the Bawi) with the idea that as customary land owners they would regulate access to forests and have customary power and control over the rural development zone (RDZ). Reproducing colonial relations of power and control, the FMP imposes a framing that people living together in one space are a homogeneous group and are governed by one common chief, thereby ignoring the heterogeneity within so-called ‘communities’ and different interests at local level but also the symbolic, cultural and temporal aspects of the landscape. It compels communities to operate in a reductive model and understanding of land appropriation and use based on prescriptive and fixed categories of ‘indigenous land’, customary tenure and property rights, that undermines various groups’ access to and use of land. Such misrecognition has not only triggered a conflict but has also led the limits of the RDZ to be reached much sooner than expected (see section 5.1).

What makes the case of this landscape particularly compelling is that it allows us to understand how forest access, land tenure and identities are dynamic and constructed out of a constellation of socio- ecological and political-economic relations and everyday practices across time and space, in a constant state of ‘becoming’ rather than linear and ‘ready to be occupied’. It highlights the multiple layers of the past that have seemingly vanished but, in fact, never utterly (Gordillo, 2014) do so. I have shown that amid the rubble of successive waves of forced resettlements and land expropriation, of increased migration due to rebellions and economic downturn, of nationalization, and of capitalist expansion, the sense of belonging and social identities through collective appropriation of forests have become articulated with an individualization of citizenship and land tenure, in often ‘unruly’ forms. The landscape thus consists of privately held concessions, household (more or less) appropriated fallows, post- agricultural secondary forests and rented land set inside a matrix of ethnic subgroup’s common primary forests imprinted by traces from previous generations and locally named water streams. Such articulation is far from static and defies a containerized socio-temporal-spatiality. Rather, it is continuously negotiated and (re)shaped in relation to everyday land use practices (e.g. agriculture versus mining) the evolving environment in which different groups live; that is, in relation to changing circumstances such as increased densities and competition for resources, the evolution of the natural and built environment,

159

e.g. primary forests, fallows, road construction, and the emergence of new sets of (local, regional and transnational) political and economic relations. The ‘story-so-far’ of the Alibuku landscape is a story about socio-political change as much as it is about “environmental changes and the transformations that structure people’s interactions with the land” (Jacka, 2015, p. 17).

I have shown in this way that bounded and mapped socio-spatial identities from the colonial era are reactivated in today’s claims for recognition in the face of an increasing commodification of land and resources while also overlapping with a fluid appropriation, porous boundaries and looser tenure for traditional activities such as shifting agriculture, hunting and gathering of forests products that was characteristic of the pre-colonial period. I have argued that the type of land use practice, the framing of a resource (commodity or not) and the type of user is key here. Groups assert their ‘entitlement’ to land or primary occupancy more strongly in relation to resource uses that are framed as permanent ‘modern extraction’ – e.g. diamond, logging, cash crop plantations that yield cash – and carried out by users external to the entitled group. Land and forests are then treated the most like private property. Lund and Rachman (2018) have similarly shown in a case study in Indonesia that claims for property rights and citizenship often intertwine and are often articulated with claims of ‘indigeneity’.

Natural resources and their meanings are thus agents that shape human social relations and inversely; their pathways are enmeshed through everyday practices (Jacka, 2015; Rose, 2011; Tsing, 2015). Similarly, user rights for land use activities considered ‘traditional’ also become stricter when the availability of communal primary forests196 – to the ethnic subgroup (groupement*), village or clan – declines or pressure on land resources increases as a result of various interconnected dynamics such as urban exodus, road construction or land alienation by urban elites. For a large part, these dynamics happen at a far larger scale than the Alibuku landscape itself (Massey, 2005), such as the boom of the diamond trade in the 1980s or the push for cacao production by ESCO-Kivu/OLAM (see 3.2) which has prompted some Nande urban elites to buy land in the Alibuku landscape. Identities are a co-constitutive product of these constellation of socio-ecological relations, relationally shifting from an affirmation of national citizenship to ancestral intergroup alliances, to a reaffirmation of ethnic subgroup belonging, or even clan and household membership. Pieces of forest and land have thus multiple connections to these categories of belonging.

The reshaped COTREFOR/IFCO concession landscape-map (Figure 16) thus came into being through the incorporation of this array of relations to forest and land. As I show in the next chapter, the containerized idea of land use it imposes through land use zones – also advocated by the REDD+ spatial-based approach to landscape – similarly misrecognizes the multiplicity of socio-ecological relationships in the Kisangani

196 Again here, I use the term primary forest in its social understanding. These forests might not be ‘primary’ in a biological sense, but they are named ‘primary’ by their inhabitants. 160

hinterland’s landscape. As Souviens, a young man from the Bawi village on the former Buta road, told me during a transect walk we did after having naïvely tried to make him draw a detailed sketch map of his household’s land use area: “you see, that is why you need to walk with us in the forest, you need to see, so that I can explain the forest to you, and what we do in it”.197 Knowing through taking a view in the world; knowing through being (Ingold, 2000).

197 81-I-ARBU-20180502. 161

162

6. Sangisa: plantation shadows and multispecies peri-urban socio-ecologies

In May 2018, I did a transect walk in Pumuzika village’s land and forest use areas with my research assistant and two villagers.198 Pumuzika is a remote village located about 86km from Kisangani on the ‘former Amex-Bois logging road’ that is only accessible with a motorbike or mostly by bicycle as very few villagers can afford a motorbike. To reach the village from the city, you first need to take the quite densely populated first 24km of the Buta dirt road, which was renovated in 2012, along which villages succeed one another in the way the Belgian colonial administration had organized them in paysannats indigènes*. It is an area that is part of what remote-sensing experts call the “rural complex”, i.e. “a distinctive agricultural land cover mosaic surrounding the network of inhabited areas found along rivers and roads in the forest block of the DRC” (Molinario et al., 2017, p. 1). At kilometre 24, you turn East on the current Amex-Bois road towards Alibuku where the base of the first logging concession, Amex-Bois, that exploited the area used to be located. Around Alibuku which has become a centre for charcoal production, the forest has mainly disappeared giving way to a vast area of cleared land. The forest progressively reappears in alternation with a mosaic of private concessions owned by urban elites, villages, fields and fallows as you continues your way towards Azunu, the village at the roundabout between the former Amex-Bois road to the South and the COTREFOR/IFCO logging road, which is still in use, to the North. Turning to the South, the road becomes increasingly difficult to ride as it becomes narrower, degraded and highly weed-infested due to lack of maintenance. The forest is denser here with only a few settlements until you finally reach Pumuzika.

Pumuzika is a former diamond quarry mainly located in the COTREFOR/IFCO concession and inhabited by the Mbole cultivators, as they identify themselves. We already met them in the previous chapter. The village represents a perfect example of the crossroads between commodity-driven and subsistence land use. At first, I was curious to see the high value timber trunks that TRANS-M-BOIS – an industrial logging company – had apparently abandoned in the village’s forests probably more than half a decade before. Villagers had mentioned the trunks’ presence during previous conversations. On the way, however, as I was interacting materially and bodily with this “landscape of dwelling” (Ingold, 1993), I quickly became more interested in the many variations of the physical environment’s features and the unpredictable location of land use activities carried out in this landscape. Leaving the village of Pumuzika behind us, we were following a former skid trail that had last been used by the industrial timber company, TRANS-M- BOIS, about eight years before, my two guides told me. Just like in other villages I had visited, some farm animals (chicken, pigs, goats) were roaming in the direct vicinity of the village, between and behind the houses. A bit further on our way, small fields started to alternate with fallows that became thicker as we moved further away from the village. Farmers and Pumuzika inhabitants, busy with their activities,

198 20-TW-ARAB-20180507. 163

greeted us on the way. The path became narrower. Artisanal loggers used to come here when the skid trail was still maintained by the company. We had walked for about an hour and a half towards the ‘primary’ forests and a former exploitation area of TRANS-M-BOIS where timber trunks had been abandoned, when we came across a campement*, i.e. a small temporary camp in the forest that is regularly used for several weeks to carry out diverse agricultural and forest activities, next to a new field in preparation. It was in the middle of what I considered to be primal forest. It was a very old post- agricultural secondary forest, I would quickly learn; local inhabitants have an amazing capacity to recognize the marks of previous agricultural use. The field had been opened to plant makemba*, i.e. plantains in Lingala. Makemba*, one of my guides told me, thrive in (primary and secondary) forested areas much more than in younger fallows. A few minutes further away, in what was an even denser forest, we arrived at the former site of the village when it used to be a quarry, guides told me. I would not have guessed there had been a village here. They left this site when the (now former) Amex-Bois logging road was built, as it gave them better market access to sell their agricultural products. Yet they have never really left this space as it remains central to their everyday livelihood practices of hunting and gathering. We proceeded to cross the Ambada river on foot, where villagers come to fish, to finally reach what my guides called, for the first time during our walk, primal forest. We followed the narrow trails they regularly use in their nightly hunting expeditions – game animals do not come out during the day, they said – and discovered many abandoned timber trunks. They knew where each trunk was exactly located. Feeling totally disoriented and breathless by the density of the forest and this unusual environment to me, one of the guides told me he, too, is scared when he enters a forest he does not know. But these are “their” forests; they know every tree and liana, and have shaped them through their everyday trails, blurring the boundaries between experiencing the landscape, understanding and making it.

164

Figure 20. Participatory map of Pumuzika village on the former Amex-Bois road (14/05/2018). During the transect walk, we followed the former skid trail that leads to the ‘primary forest’ (FOR VIERGE), at the bottom right of the map. Note the map is not at scale and the distance to primary forests is longer than the map might make appear.

Picture 7. A makemba* field in preparation in the middle of a dense secondary forest.

165

I began this chapter with this short story of my journey to and in Pumuzika’s forests because it captures the variety of ecologies and heterogeneity of the Kisangani hinterland’s landscape. Such a socio-ecology could once again not be grasped from satellite imagery or simplistic cartographic understandings I had seen and tried to reproduce myself through participatory and sketch maps. The disorientation I experienced during that walk has been something I often felt as my Western philosophical system that separates the human from the nonhuman, the wild and the cultivated, the subsistence from the commercial, the abandoned and the ‘in-use’, the ‘traditional’ from the ‘modern’ was challenged.

Swidden agriculture farmers use the words sangisa* in Lingala, which means mixing, blending or gathering together or alternatively mobulu* – disorder or indiscipline in English, or pêle-mêle* in French199, i.e. jumble, hotchpotch – to describe their polycultural field.200 Sangisa* speaks to the dynamics and heterogeneity of ‘peasant’ agriculture in the Kisangani hinterland, which I explored in the first part of this chapter, and the ways they shape and are shaped by an evolving socio-natural environment. As the previous chapter hinted, relations to forests, land access, the built and natural environment, and power differential between groups and individuals explain some of the farming land use patterns in that what, where, and how people cultivate depends in part on these aspects. Sangisa* thus points to the ways the socio-cultural mingles with the ecological in that the forest-scape in the hinterland of Kisangani is profoundly anthropogenic or naturalcultural Haraway, 2003), fashioned in cycles by human activity over time. What might appear to a satellite bird-eye’s view as primal ‘untouched’ forests are often places where previous generations have dwelled and that remain attached to collective socio-cultural memory and shape today’s land use activities. Sangisa*, finally, is the entanglement of local understandings of forest change with social and moral meanings, as I show in the second part of this chapter.

6.1. Making fields, producing relations, shaping forests: a socio-ecology of farming

Elanga* in Lingala which means simultaneously ‘field’, ‘garden’ or ‘what is planted’, is the very life blood that sustains most inhabitants of the Kisangani hinterland from peasant farmers to many urban elites. It is a central expression of the culture of this landscape, in that it does enable people to make a living but also contributes to social reproduction, group’s cohesion and identities. As Rösler (1997) has similarly observed about agriculture in the Ituri region, the shifting agricultural system in Kisangani hinterland is a syncretism of both pre-colonial (‘traditional’) and colonial (‘modern’) elements. It is constantly adapting to changing socio-ecological conditions. Features of today’s version of shifting cultivation, a millennial practice are to a certain extent adapted from the cultures obligatoires* and Paysannat Indigènes* policies

199 French has a large influence on Lingala. In a Lingala sentence there are always several words borrowed from French, as in this case with the word “pêle-mêle”. 200 E.g. 18-SM-TW-RBU-20180501, 31-I-YALO-20160323, 40-I-YALI-20160326. 166

which, as I have already mentioned before, have durably transformed the rural socio-ecological world rather than faded away. Moreover, as much as practices of access can hardly be generalized, farming in the hinterland is not a homogeneous practice but rather matches the variety of its users. Farming, I will show over the next few sections, is shaped by a four-way relationship between human cultivators (their agency), socio-economic relations across time and space, seasonality and the materiality of resources (forests, land and crops).

6.1.1. Cutting out a field: a more than technical-agricultural meaning

In the area’s shifting cultivation system, a farmer has one or two – sometimes more – principal field(s) every year. The process, which is adapted to the soil, flora and climatic characteristics of the rainforest, starts with clearing the undergrowth, cutting trees and burning the dried out vegetal material, as a farmer explains:

“This is the habit. [We] need to plant in a space that is really open, and not under the trees in the primary forests. You cannot plant under the trees, because the seeds won’t grow really well. It is thus necessary to cut trees […] and then burn to make this space really fertile. That is how we can obtain a good yield.” 201

The field is then used for one or more years. As weeds become uncontrollable and fertility declines, the field is left in fallow for several years and drives the farmer to move to another fallow, post-agricultural or primary forest to cut out a new field. However, as it will become clearer, farmers do not ‘just go anywhere’ in the forest, as some discourses on ‘harmful irrational shifting cultivators’ might portray them.

When choosing a field site, a number of ecological, social and cultural considerations are taken into account. Ecologically, the main driver for farmers to clear a ‘primary’ forest or a post-agricultural secondary forest is to increase crop yield and avoid the weeds that grow more rapidly on short-fallow land. In a primary or secondary forest, the dense vegetal material made up of tree trunks, leaves and undergrowth, rots and gives nutrients to the soil which ensures a good yield and less weeding maintenance –done by hand with the help of a machete or hoe – in comparison with younger fallows. Weeding is considered by farmers to be their major constraint on production. While the PPRGII and JADORA REDD+ projects in Isangi push people to work in their existing fallows for several years – i.e. to sedentarize the agricultural system, without weedkillers and fertilizers farmers are forced to engage in hard (and expensive) weeding labour as Claude, a pastor and farmer from Yafunga explained:

“If we use the fallows time and time again, weeds grow. […] We have fallows but we do not have tractors to cultivate. They say we have to repeat [cultivation] on the same place two or three times, or five years, but we do

201 6-I-RBU-20160305. 167

not have chemical fertilizers, we do not have the means to cultivate the same place several years. In the primary forest, you can cultivate and get to the harvest [period] without even having to hoe because the weeds start to grow later. […] Despite the hard work [to open] in the primary forest, you will have some rest before the harvest. After clearing, burning and sowing the field, you will be able to rest. […] But in the fallows, you have to weed out and hoe at least three times before the harvest. […] If you don’t have financial means [to pay for the weeding work], your harvest won’t be good.”202

Héritier and his brother, two farmers of Yalosuna, whose clan has access to quite a large primary forest – see below – similarly said: “[in the primary forest] the soil is really fertile. Even if you do not push the cuttings in the ground… just throwing is enough and it will grow; even if you throw rice on the ground, it will grow without any problem”.203 “We can leave our fallows for six years because we have large areas of forests; that’s why we let some parts grow”, they added. Moreover, a field that has been freshly cleared in the primary forest can be used for two to three times before being left to fallow. The fallow durations reported by my informants, however, vary greatly from one farmer to another. Some cultivate the same plots of land “more than three times”204, others leave their land fallow for two or three years, while still others are able to leave it for five to six years which is, according to the ones who do so, sufficient for the soil to regain its fertility.205 Fallow durations are rarely fixed for the same farmer either, depending on his/her circumstances.

Cutting out a field in the middle of a primary forest or mature post-agricultural secondary forest, i.e. that is not ‘individually’ appropriated by a family, is also a way to combine farming with other forest livelihood activities, such as hunting, gathering or fishing in the numerous small streams that run in the forests. Forest products, game animals and fish are more abundant as one gets further away from the village. Moreover, a farmer never fells all the trees in a plot, as many are essential to other activities. Trees along water streams are particularly important to fishing. As Cobra, a young farmer from Yaliobaga explained, “if I cut all the big trees, how would I eat fish? If I cut all the trees, fish in the stream will die. […] During the rainy season when the water comes up, the fish that were hidden under the tree roots will come out”.206 Feral fruit trees, in particular oil palm, safou and avocado trees are kept for future gathering as well as ‘caterpillar trees’ – arbres à chenilles* – which are an important seasonal source of proteins. Trunks, lianas and the branches of some trees are used as diverse construction materials for the house.

Other tree species are highly valuable for charcoal (makala*) and artisanal timber that are in increasing demand in cities. In areas closer to Kisangani or to evacuation routes (especially rivers), such trees can

202 45-I-YAF-20160329. 203 34-I-YALO-20160324. 204 17-I-RBU-20160312. 20534-I-YALO-20160324 35-I-BU-20160324 206 7-IFG-YALI-20160326. 168

provide another source of income from the primary forest’s field. Gaston, a dweller from Likoko on the Amex-Bois road where loggers and charcoal producers are present, said:

“[in the primary forest] the work is really hard, but we choose to do it because you need to go there to have good [quality] plantains, and fell mogoya/afrormosia, liboyo [two high value trees for artisanal loggers], Limbalu [one of the favorite tree species for charcoal] […] Then you can hope that this one month that year, with God’s help, is gonna be good.”207

Farmers are not engaged in artisanal logging as it theoretically requires a permit – yet many loggers do not have it – but especially technical skills, tools (a chain saw) and relatively high investment and labour costs (see e.g. Abdala and Lokoka (2009); Lescuyer et al. (2014) for studies on artisanal logging in Kisangani’s hinterland). They rather sell the tree pit, in their newly cut primary forest’s field to artisanal loggers. Selling prices greatly vary from one species to another, its quality and volume, and depending on the distance from the road/river and from Kisangani – e.g. $20 to $40 per pit of 5 m3 for Afrormosia to $10-$30 per pit of 5 m3 for Sapelli or Sipo (Abdala & Lokoka, 2009). Regarding charcoal production, some farmers produce it themselves while others just sell the tree pits to ‘specialized’ charcoal producers, who are mainly urban dwellers from Kisangani or newcomer landless allochthones. Charcoal production can be an important subsidiary source of income for some households. Looking at a stack of wood ready to be transformed in his fields, Souviens and Jean, two young farmers from the Bawi village on the former Buta road, told me, “Makala, it’s the fast intervention because in a few days [about 15] it’s sold […] one tree can produce 10-12 bags, each sold locally at 8000CDF”.208 In this way, farming in the primary forest cannot be isolated from other land and forest use activities, and is connected to the city, the needs and activities of its inhabitants.

The availability of accessible, arable forests is a major element that influences farmers’ field location and capacities to have sufficient fallow duration. In Yalosuna or Yaliaboga, two villages I visited in the Isangi territory, the hydromorphic soil that characterizes the main part of their forests – they call them forêts marécageuses* – is one of the main constraints farmers told me they are facing. Such soil is much less suited for cultivation and farmers generally prefer to avoid using it although those who do not have sufficient land are forced to cut out fields in such swamp forests. Yet, as I show in section 6.1.2., many farmers have started to cultivate rice and crop seeds that tolerate such conditions – seeds that have been disseminated by various agricultural development projects. Access to markets, by road or river, is another main factor in explaining a household’s decision making process for field location, as Frédéric, the farmer from Yafunga simply explained: “we respect [calculate] the distance from the house and the fields, you won’t go too far in the forest to cultivate there… how will you come back with the agricultural products

207 55-I-RAB-20170509. 208 81-I-ARBU-20180502. 169

to the house?”.209 Another farmer, who settled on the ‘Buta road’ similarly said: “in order to give a high economic value to my activity [farming], I need to be on the road”.210 Indeed, with lower transport costs and closer proximity to markets, farmers are able to get a higher return on their products.211 Farmers whose fields are further away from important road/river access and in more remote villages that render access to market difficult, such as Pumuzika or even Banjwade much further on the ‘Buta road’, reportedly have smaller fields, and thus lower revenues, sometimes only engaging in subsistence farming.212 I will return to these differences.

Lands that are closer to the village and roads, and even more so in areas closer to big markets and the city, are generally the ones under more pressure and thus can have an important impact on the availability of arable forests and rotations. Etienne, a farmer from Yalosuna whose lands are easily accessible from the main road says: “the ones who have primary forests further away from the village are fine, but on my lands, fertility has declined.”213 As I further develop below, on the main historical commercial roads where villages have been resettled and wedged together during the colonial era (see chapter 5), and especially as one is closer to Kisangani, cutting out a field in a primary forest is simply not possible anymore. Georges, a Bawi farmer in the busy KP23 village on the ‘Buta road’ that was renovated in 2012, which further increased pressure, said: “we cannot continue with the system [of cutting out new forests]; here we only return to fallows”.214 This situation, as I explain below, pushes ‘landless’ farmers to look for new land away from their village of residence. Roads that have been opened more recently such as logging roads and skid trails are particularly attractive to these farmers as they allow them to access primary/older secondary forests for their agricultural activities while still giving them a fair access to markets. In Yafunga, people reported that when the SAFBOIS logging concession, – now converted into the REDD+ conservation concession, JADORA – was still operating, they used to be able to open fields anywhere in the concession as long as “they would let [SAFBOIS] know if there were [high value] trees” 215 in the field. Similarly, in COTREFOR/IFCO concession, these dynamics can be observed “in real time”; the movement of an ever-growing number of farmers is intimately connected to the felling of trees, as loggers are progressing further into the primary forest to harvest previous exotic wood species (see chapter 5 and chapter 2, section 2.2 where I describe such dynamics in logging concessions). With the sporadic maintenance of these roads and trails after exploitation, the accessibility of such areas is, however, not guaranteed. As I explain below, more marginalized ‘landless’ families who do not have the means to rent land elsewhere do not always have an alternative to cultivating ‘deeper’ in forests, outside

209 46-I-YAF-20160329. 210 63-I-ARBU-20170515. 211 60-I-RBU-20170511, 81-I-ARBU-20180502. 212 E.g. 13-PM-ARAB-20180514, 14-PM-BEN-20190516, 55-I-RAB-20170509. 213 32-I-YALO-20160323. 214 74-I-RBU-20170522. 215 43-I-YAF-20160328. 170

the rural complex where access to land can be easier but away from their kin and social relations, and markets. The physical and built environment are thus certainly not the only factors to take into account in our understanding of land use patterns.

Collective and individual socio-cultural and economic factors present both important opportunities and constraints for the field’s location, rotation and size, which inversely impact households’ livelihoods. The location of fields is eminently social in that the nature of tenure arrangements and social relations shape the farming/arable land. I have shown in chapter 5 that villages and clans216 act as spatial units and land owning groups within the larger customary space of a groupement*, but that such categories fluctuate, adapting to changing circumstances and that depending on the areas and resources at stake strict delimitations and access rules are more or less enforced. In the traditional system, where autochthones assert their rights, each village is composed of a ngonda* – a forest with high trees that is also often referred to as primary forest or zamba*, which means forest in general – that is partly communal to the whole village and partly divided between clans.217 In fact ngonda* is rather a mature secondary forest that was cultivated a long time ago by a clan’s ancestor. It is thus ‘marked’ and transmitted within a clan from generation to generation but ‘cultivated’ is no longer the discerning label that characterizes its appropriation, unlike elanga* (field), masokolo* (fallow) and zamba ya sokolo* (literally, “fallow forests” which refers to older fallows). Part of the ngonda* also remains communal to the whole village as no traces of the clan’s ancestors seem to remain, and is considered to be the village’s “reserve”.218 Fields, fallow and post-agricultural forests, unlike the ngonda*, belongs to ‘autochthonous’ families – members of the clan. A family can extend its agricultural land and forest by asking permission from the clan’s chief to cut out new fields in the clan’s ngonda* or from the village’s chief to go into the communal ngonda* – although many reported that in practice, some farmers do not ask for permission which creates intra- community conflicts.

Clearing forest is thus strongly regulated by social organization. The structure of clans is, by definition, always changing as children are born and get married – women come to live on the lands of their husbands’ clans. Such is the structure of land distribution and appropriation as vegetation evolves in various stages over time. Elanga*, the garden-field, in this way cannot be thought of as outside other categories of forest in the village’s extended socio-ecological space; they are part of the same continuum and cohabit through successions, cycle and change (Robbins, 2001).

As mentioned before, unlike other land use practices such as hunting and gathering which mark ancestral communal appropriation, farming is the only way of marking ‘individual’ land appropriation (usufruct

216 As a reminder, a clan is a group of nuclear families who have family ties with a common ancestor. 217 4-PM-RBA-20160311. 218 12-RBU-20160308. 171

right of axe). “We have a collective system, land belongs to everyone [but] the individual strength has to be used to get a share of these lands”, Georges a Bawi farmer on the ‘Buta road’ summarized.219 In this system of father-to-son inheritance, to ensure that a household’s male children will each have land to cultivate within the clan’s lands, the household (male) head has to keep sufficient land in fallow or open enough forests to pass down to his children. “Here in villages, each father has his part of land and each child follow in the footsteps of his father”.220 Failing to cut out new fields, puts descendants at risk as Elizée, a widowed female farmer from Yafunga who can use some of her brother’s lands said: “if you don’t use the primary forests, you will struggle to leave anything to your children”.221 Tenure and farming are, in this way, built into each other. Cutting out a field is thus not only an ecological decision to ensure high yield but also allows a household’s and clan’s social and material reproduction, as Elizée continued:

“When a dad brings children into the world, he has to enter the primary forest and if he has four or five children, when these children grow, they will ensure succession to their father because it is in the primary forest that there are good harvests: plantain bananas, manioc, rice is of good quality. And there [in the primary forest], the father is going to cultivate four or five times and once his children are grown up, they ensure succession. It is like for the whites: you buy land for your children. We do the same with the primary forest.”222

As this quote hints at, to prevent older fallows returning to the clan’s or village’s “common pool” of land, farmers are naturally encouraged to rotate between their plots of land so as to clearly mark appropriation. This ‘rule’ thus plays a role in limiting expansion. In areas where competition for land is high, the moment when a fallow returns to the clan’s collective land can be a bone of contention and lead to conflict.223 Suzanne, a married female inhabitant of the busy KP13 on the ‘Buta road’ told me that this used to happen at her parents’ time already:

“Every parent burnt the bush, the forest, and it became their own part. But it happened that when someone had cultivated their field, after the harvest, they neglected to plough their bush for a while. When the owner stopped like that, other ones came to move the limits for their own benefit.”224

Roger, a notable* from Yafunga explains how he managed to keep his (grand)father’s land while he had moved away from the village to work for an oil palm plantation company in Imbolo, in the Isangi territory:

219 74-I-RBU-20170522. 220 50-I-YAF-20160330. 221 48-I-YAF-20160330. 222 Idem. 223 Conflicts of limits between families and clans are very common in the hinterland. I will not describe such land conflicts further, as I have already dedicated chapter 5 to contested land. 224 21-I-RBU-20160314. 172

“I have a big responsibility. I have two wives and twelve children. This is why, although I worked for a company [elsewhere], [I kept] the known land of our grandfather, of our parents. I was working for the company and there, they gave us oil palm seedlings and I would come back to plant them here [in Yafunga]. I knew that one day I would leave the company and come back to the village. I had a field here that I sowed when I was still a worker at the company, and now I have about 8 hectares. And when I came back here, I started to clear the primary forest to leave some fallows to my brothers.”225

As a former worker for a multinational private company and as a notable*, Roger is however in a better- off position than most households and in an area where ngonda* is more available than in denser areas.226 His autochthonous status also allowed him to plant perennial oil palm trees, one of the most enduring marks of appropriation. In fact, cutting out a field in a ‘primary’ forest is not an equal practice for all.

Chapter 5 amply discussed boundary-making processes, forced sedentary resettlement and the uneven distribution of land between villages and clans under the Belgian Colony. In a shifting agriculture system, they have important impacts on households’ farming opportunities and decision-making processes, and engender important inequalities. Semi-nomadic pre-colonial settlement, I have shown, was built into the integrated farming-hunting-gathering system, and villages’ forced sedentarization has profoundly shaken this socio-ecological equilibrium. In the sites I have researched, this appeared particularly acute in areas on the ‘Buta road’ where Paysannat Indigènes* were enforced and which, due to their proximity to Kisangani, are now under increasing land pressure. Overall, villages that were part of paysannats indigènes* programs which, as a reminder, pushed towards sedentary monocultural agriculture and a shift from collective to individual property, have much less communal/collective forests than other areas. Clans “have remained in the divisions established by the Belgian colonialists”227 – they also make continuous reference to the paysannat* maps – even as population grew for multiple reasons (see 5.2.). In most villages all clan communal land has already been divided between families, which increases conflicts. Such dynamics have dramatically obstructed farmers’ capacities to adapt to changing ecological conditions, e.g. in particular the constant presence of weeds in younger fallows, by using extensive rotations, as they were able to do in pre-colonial times. In such a space, the forest shortage is rather due to past land use changes – older deforestation – than recent ones (see also e.g. Moonen et al. (2016) for a quantitative analysis of deforestation on three road axes around Kisangani).

Where paysannat indigene* was imposed, land was often unequally divided between villages, clans and families, creating many inequalities. “Each child has its plot within its clan, whether its plot is large or

225 50-I-YAF-20160330. 226 Yafunga is located in the former SAFBOIS logging concession, which has been converted into a REDD+ conservation concession (JADORA). Such concessions are precisely chosen because of the availability of primary forests although local peasant farmers continue to use these spaces for their agricultural activities. 227 3-I-RBU-20160304. 173

small, that’s where he needs to work”, Victor a dweller of KP13 reminded me.228 In this way, for instance, the Batiamaduka and Bafamba villages between KP9 and KP18 were attributed large plots of lands unlike some other neighbouring villages (see 5.2.2. and Figure 19). By definition, these inequalities also trickle down to households and their farming practices in the present. Clans and families who were attributed (or attributed themselves) large pieces of land have engaged in selling and renting land. Some inhabitants, Peter, a teacher and secretary of the local development committee from KP12 told me, “don’t even cultivate; they just rent their lands out”.229 In the Ambese clan of the Bafamba village, Mundu an old man who is the clan chief and former village chief, is a typical example.230 Because of his status and membership of the dominant clan, he still ‘owns’ more than 5 hectares – which is a lot in this part of the hinterland, and it does not even include the plots already transmitted to his children – of which he always rents a portion in exchange for money. Mundu’s clan also sold several plots of land to the brewery company Bralima and urban elites. In these ‘privileged’ clans, as I show below, some are able to cultivate 3 to 4 hectares a year in fields that are close to water or road market access. It allows them to ensure higher income than most while simultaneously preventing these lands returning to the clan’s or village’s communal lands thus ensuring their family’s own social and material reproduction. With their revenues, men from privileged clans are sometimes able to purchase some more land, such as Peter has done.231 On top of his family’s land in KP 12 on the ‘Buta road’, he has purchased a plot of land from a clan in KP11 for $500 and a goat, as well as a plot in Kisangani for his children to study there. He does not have a registered legal title for it but rather a document signed by the local chiefs. Such growing practices are, by many, interpreted as land grabbing and, as I show in section 6.2., one of the major local understandings of socio-ecological change.

In contrast, ‘children’ from more marginal clans face much more insecurity. Some younger people, in particular, simply do not have (sufficient) access to fertile land and are forced to rent small plots and/or use short fallows, or still, engage in alternative activities such as charcoal production. In Yaliaboga a village surrounded by swamp forests in the Isangi territory (see above), a group of young people told me:

“Where we are allowed to cultivate, it’s not really fertile anymore. That’s our reality. Those who have a little money rent some lands to sow and, after the harvest, you give back the forest to the owner. That’s how we do.”232

In other cases, such as the small Bawi village at KP23 on the ‘Buta road’, the lack of available fertile land pushes some households to go back to their ancestral lands where road access is particularly difficult. As I mentioned in chapter 4, the memory of ethnic subgroups’ and clans’ ancestral lands, and/or village’s

228 24-I-RBU-20160315. 229 4-I-RBU-20160305. 230 7-I-RBU-20160305. 231 4-I-RBU-20160305. 232 7-IFG-YALI-2016032. 174

former (not so) abandoned site, has remained present and plays a central role in understanding today’s land use patterns that satellite imageries capture but interpret as ‘new perforations’. People’s dwelling space, as my exploration of Pumuzika’s land use area revealed, is much more extended than what the main residential village location and more intensive land use area might suggest at first. Using campements* in the forest is a major practice everywhere in the Kisangani hinterland. Campements* are typically mature secondary forests – so old that villagers refer to them as primary forests or ngonda* in Lingala – that are used for various livelihood activities, mainly for farming and hunting/gathering but also for mining. They are usually located several hours away by foot from the village, at the edge of the core land use area that is made up of fields, fallows and secondary forests in a first circle around the village. In terms of tenure, they are usually considered to be the village’s communal forests. Some farmer households who lack or do not have access to forests closer to their village and the road “renew these campements*”233 for their agricultural activities. “One has to farm. Always. Even if it’s really far, you will just build a small hut and stay there”, one said.234 The farmers making use of campements* stay in the forest for one or several weeks, as Frédéric, the farmer from Yafunga explained:

“It’s in our blood… it’s in our blood. Like me here, I have burnt a new field [in the ngonda*]; if I have to go every morning, it’s gonna be difficult… back and forth, back and forth every day. So I just take my wife and my children with me and we go stay over there. When I am finished with sowing, we just come back home and wait for the harvest.”235

Campements* are in fact used by the villagers for multiple activities such as hunting and gathering or fishing, but also the collection of sticks that serves as construction material. Because they are much further away from inhabited areas, game animals are more abundant than in the core land use area of the village. In villages such as Pumuzika or Banjwade, some campements* are the remains of mining sites where some creusers* –‘diggers’ – continue to go in the hope of finding a gold nugget or a small diamond as they are farming. For other communities, however, campements* are linked to a much more ancient use and an attachment to ancestral land that was deemed ‘vacant’ under the colonial administration. “It’s in our blood”, said Frédéric.236 While they allow some dwellers of the hinterland to cope with shortages of forest in the village, they are also ways to “guard the graves of the ancestors”237 and the forest territory of a group through other land use activities I just mentioned. Safou trees (especially) and other ruderal plants show that “the campements* of our parents were here”, said Frédéric.238 In some cases, campements* have progressively become small permanent villages again, from which their

233 46-I-YAF-20160329. 234 27-I-YALO-20160322. 235 46-I-YAF-20160329. 236 Idem. 237 59-I-RAB-20170511. 238 46-I-YAF-20160329.. 175

inhabitants come back “only to go to the market or if there is a funeral here, in the [main] village”, added Frédéric.

Hidden campements* are part of a history of resistance that has stayed in the collective imaginary. During the 19th century, villages broke up and scattered in the forests to escape Zanzibari-Arab slave traders, two Bevenzeke sages* told me.239 Forests’ campements* continued to be shelters from the violence of Belgian colonialists. Likaka (1994) has shown that during the colonial era, Mbole peasants used hidden settlements in the forests to avoid colonial control – “coercion-free villages” (p. 611) – and attack tax collectors, crops supervisors, and policemen. Campements* allowed them to informally keep control over communal lands that had been grabbed by the colonial administration. On several occasions, informants or acquaintances told me in a sort of secretive tone, that there were peasants hidden in the forests, preparing to resist a conservation project in Yangambi for example, or another time, the State.240 Albeit in a less ‘heroic’ fashion, for some households, campements* represent a sort of ‘runaway’ (Likaka, 1994) from unequal socio-political structures that prevent them from accessing land in the village. They can sustain their livelihoods without depending on a ‘landlord’. In this way, today’s ubiquity of the use of campements* in the hinterland, and their historicity, suggests that they are more than new strategies to cope with recent forest and land degradation for allochthones in search of land or natural resources to exploit (see e.g. Koy, Ngonga, & Wardell (2019) or Reyniers, Karsenty , & Vermeulen (2016) who tend to make this argument). Dwellers in villages which do not report such degradation make use of campements*, and they are also used by autochthones who are not specifically in search of land. Such small-scale population and village movements have, since ancient times, been a profound characteristic of the region’s tradition (Vansina, 1990). Campements* are thus as much a cultural expression as an ecological response.

Ecological determinants, access and socio-historical relations to forested lands thus all influence farming land use patterns. Such entanglements similarly shape a farmer’s elanga*, their fields-gardens.

6.1.2. Choosing crops

Typically, the association of a very mixed range of food crops is a major feature of swidden agriculture in the hinterland of Kisangani. Far from the straight rows with single crops typical of Western agricultural systems, the field in a swidden system is a wonderful jumble of variety, not only of different crops but also within each crop type as farmers usually grow different varieties of the same crop. Manioc, rice, maize and plantain are the main root and cereal crops in the hinterland, with rice and maize mainly for sale. Other roots and leguminous plants such as yams/sweet potatoes, groundnut and Vigna cowpea –

239 59-I-RAB-20170511. 240 E.g. 75-I-MAS-20170522. 176

often referred to as niébé* in French or bean – are also common and can represent a good source of income for some farmers. In general, these types of crops occupy most of a field, but they are complemented by other vegetable crops such as chives, tomatoes, eggplants, squash, peppers and amaranth that are usually of secondary importance to households. All grow together in the same field, here and there interrupted by fallen trunks and stumps. Some fruit trees – avocado, passion fruit, safou – might be planted intentionally to complement the diet or mark down land’s appropriation but most grow naturally or have become feral, i.e. they grow out of thrown away seeds or have ‘escaped’ from the field, and are harvested from the wild. The fruit trees will grow among other herbs and trees, as the field is abandoned in fallow. Much later, it will be the only sign remaining of past occupation. Fruit trees “fall somewhere in the gap between cultivation and the wild” (Tsing, 2005, p. 178).

The unruly variety of crops and plants in the garden has many advantages for peasant farmers. Ecologically for instance, the various crops support each other’s growth. As Moïse, a farmer from Yafunga in Isangi territory explained, “if you only plant rice and the wind comes, it will all be knocked down […] Therefore I put manioc seedling here and there to protect rice from the wind; that is the system”.241 Planting crops and seed varieties that have different growing durations also ensures soil fertility with a slow transition to fallow land but also resilience in the face of irregular rainfalls. Although in the Tshopo Province, manioc is by far the main staple, crops are chosen so that they complement each other to ensure the household’s food intake – and variety in tastes – and income over a season or a year. For instance, if the cultivation of one type of crop fails due to unforeseen reasons such as pests, weather or soil conditions, the household is still able to survive thanks to the other crops. It is the anticipation of uncertainty that, in this sense, defines a large part of farming decisions. As Pierre, a farmer in Yafunga in Isangi territory asserted: “this untidy way of doing is what allows us to win the season […] Monoculture is not able to support the family”.242 “It is not possible to do a field in monoculture because our basis here is manioc. If you do not cultivate manioc and you only have a field with maize, your family will not live well”, another said.243 The different maturity periods of the multiple varieties of each crop also allow peasant farmers to spread out the sale of their products and, for instance, sell at a better price outside the main harvest period when the supply is low.244 Indeed, some crops are mainly intended for subsistence – with manioc as the essential one – while others are intended for sale – with rice and maize as the main cash food crops in the hinterland (see section 6.2). As Moïse from Yafunga further explained, “we have some crops that give us money and some others are there for us to eat […] Rice, groundnut or cowpea are to put for sale, manioc and plantains will not give the same yield [in terms of cash]”.245

241 43-I-YAF-20160328. 242 40-I-YALI-20160326. 243 35-I-BU-20160324. 244 13-PM-ARAB-20180514, 35-I-BU-20160324. 245 43-I-YAF-20160328. 177

Diversifying crops in their elanga* is part of households’ livelihood strategies. Yet, the elanga* should not be thought of as a static and isolated system. Which crop(s) a family privileges has to do both with the physical and built environment and with broader historical, economic-political conditions. Manioc and bananas (plantains and others) have historically been and remain a major part of the diet in the hinterland. As Vansina (1990), explains, although it is not known when and how it reached the region, its impact was major in the expansion of the Western Bantu cultivators into the rainforests of the Congo basin. Bananas are ideally adapted to the evergreen rainforests. They thrive even in the absence of a dry season, they only require the field to be cleared from about two-thirds of the trees before being planted; it thus saves labor and “helps the forest to regenerate faster on the fallow field” (Vansina, 1990, p. 61). Manioc was introduced by the Portuguese slave traders in the mid-sixteenth century, along with maize, groundnuts and lima beans. The yield of manioc equalled that of bananas although it required more work in the field, which was essentially carried out by women or slaves. It became the first grown crop, in particular along the major trade routes as it was “the most important to the slave trade itself because it could be processed into ‘bread’ [referred to as chikwangue* in Lingala, today], which would preserve well for several months”; an ideal food for slaves and travelers (Vansina, 1990, p. 214). Exactly like bananas, manioc can be grown and harvested the whole year round (it is not seasonal); its roots rarely rot. It was the ideal crop. Maize or groundnuts, at that time, did not become so important in the rainforest, because their required more open surfaces and thus more clearing, and had better yield in moist savannas kinds of soils. Niébé* beans which were already part of the diet in ancient times – but usually grown as secondary crops – continued to be grown and mixed with European bean varieties. They were also very filling for slaves. The cultivation of rice was vigorously introduced by the Zanzibar-Arab slave and ivory traders in the then Stanleyville-Ponthierville246 region (Leplae, 1932), and was greatly intensified during the colonial period when it became an export cash crop along others. Overall, the introduction of these new crops and their forced cultivation was accompanied by agricultural intensification, especially in labour. At home and on the other side of the world, Congolese bodies and ecosystems were bent and subjugated by the plantation economy.

In the first twenty years of the 20th century, “indigenous” farmers in the area were forced to increase their rice production to resupply workers of the Compagnie des Chemins de Fer du Congo Supérieur aux Grands Lacs Africains and later military troops during the World War I period (Ibanez de Ibero, 1913; Leplae, 1932; Likaka, 1994; Ministère des Colonies, 1910; Van Reybrouck, 2012). Rice, continued to be one of the major food crops cultivated in the region under the system of compulsory cultivation – cultures obligatoires* – scheme introduced as of 1917 and then paysannat indigene* – see section 1.1, Figure 3. Along with maize and groundnuts which were for long cultivated as adjunct food crops, rice became an

246 Ponthierville is now called Ubundu. 178

even more important export in agricultural products of the Belgian colony around 1935 thanks to the depreciation of the Congolese franc and favourable maritime transport rates (Banque Nationale de Belgique, 1948; Jewsiewicki, 1979; Vanderlinden & Stengers, 1994). Such ‘food crops’ became “purely indigenous crop cultivation” in that they were not grown in European plantations (Delsupehe, 1944, p. 58). The National Institute for Agronomic Study in the Belgian Congo (INEAC)247 – today renamed the National Institute for Agronomic Study and Research (INERA)– in Yangambi was in charge of developing improved seeds to rationalize peasant agriculture which were then distributed and multiplied in the paysannats* (Mission C.E.E. - Congo, 1963). As I explain further in section 6.1.3., export crops always came first in the imposed crop rotation calendar. It is an empirical challenge, however, to know exactly which crops were privileged in each paysannat*. From my interviews with older inhabitants of the hinterland, rice, groundnuts and cotton were the major cash crops their parents had to cultivate during the colonial era. The role of maize is less clear.248 We know, from archives, that as from 1944, the paysannat Turumbu* located in the Isangi territory was specialized in rice production and the multiplication of improved seeds that were grown at INEAC (Mission C.E.E. - Congo, 1963). Seeds produced and multiplied by the Turumbu were disseminated across the territory and the province. Jean, the chief of the Batiafeke village on the Buta road and his father remember the story of their elders:

“When our grandparents started with rice, we weren’t there […] but the whites were distributing rice [seeds], they gave one bag per two people to cultivate. After the harvest, each one would bring half a bag of rice so as to fill in one full bag [per two people] to give back to the white. The rest of the harvest, which was 5 or 6 bags, would belong to you. […] The whites would only control whether you had given back one bag.”249

While local rice production progressively decreased after independence notably due to the dumping of cheap imports, since three decades, the domestic demand for rice has continuously grown with increased urbanization –in DRC, rice is an important staple food crop in cities much more than in rural areas –with the exception of Eastern Kasaï where it is an important part of rural population’s diet (MADR-RDC, 2013). At the same time, Asia’s biggest rice producing lands are increasingly under pressure due to the expansion of Asian megalopolises, and climate change-related rising sea levels and tropical cyclones that submerge rice-growing deltas and seriously deteriorate rice production (Redfern, Azzu, & Binamira, 2012). Following these trends, the pan-African rice research organization AfricaRice with the support of bilateral and multilateral funding – in particular the Japan International Cooperation Agency (JICA) – has pushed rice production in African countries, including DRC, which adopted a National Rice Development Strategy (NRDS) in the beginning of the 2010s. As from 2012-2013, improved rice seeds (from the NERICA variety)

247 The system of paysannat indigene* was in fact strongly promoted by the INEAC. 248 E.g. 13-I-RBU-20160308, 51-I-YAF-20160330. 249 10-I-RBU-20160307. 179

have been distributed to farmers in several regions of the country and has resulted in an increased production in several regions including the Tshopo Province and, for instance, particularly the Isangi territory as I explain below (MADR-RDC, 2013; Radio Okapi, 2013a, 2013b). On the demand side, city dwellers from both Kisangani and Kinshasa – where production is also sent – and the Bralima brewery, owned by Heineken International, in Kisangani, whose beers are made from rice among other ingredients, are good commercial outlets for local production.250 Moïse, still the same farmer from Yafunga referred to such dynamics in our conversation:

“Here, our parents used to grow rice but not in large quantities like now. […] But for the moment what is really purchased is rice, and so everybody must grow rice. […] Here in our economy, we take into account the purchase trends in Kisangani. During the period we called RCD251, it was the period when people bought a lot of fufu*. At that time, nobody was busy with rice, we were only busy with manioc. Rice has become necessary now. Our first buyer is Bralima which comes from Kisangani to buy [our rice] on our small markets like Yafira or Lisaliko. […] There was also a period during which workshops for growing niébé* were organized, and people were buying niébé, so everybody started to grow niébé*. That was in 2006 when presidential and local elections started.”252

Yet, the importance of each cash food crop can vary from an area of the hinterland to another. In comparing analyzing the extracts of interviews in my theme “agricultural practices, I observed that rice, for instance, appear to be more prevalent in more distant sectors from Kisangani, e.g. Isangi territory or in Pumuzika, than in the immediate peri-urban area of Kisangani, e.g. the first 30km of the ‘Buta road’. There are several related explanations for these differences. In the case of Isangi, one can be related to the particular environmental features of part of this territory. Contrarily to the Alibuku landscape for instance, hydromorphic swamp forests make up a substantial part of the territory’s forests, e.g. in the villages I visited like Yalosuna, Bula or Yaliaboga (see map UCL-Géomatique, IGC, RGC, and AfricaMuseum (2014)). Several local farmers in these villages mentioned that the variety of rice they cultivate is particularly adapted to these types of forests (shallow rice).253 With an increasing pressure on land, certain varieties of (shallow) rice allows them to valorize these large swamp forest areas which are less adapted to other types of crops and which are generally not used for cultivation by dwellers of the hinterland – they are used for fishing.254 In the beginning of the 2010s, Isangi has largely benefitted from the Agricultural Rehabilitation Program in Orientale Province (PRAPO) financed by the International Fund for Agricultural Development (IFAD) and the Agricultural Development Program in the Tshopo District

250 43-I-YAF-20160328. See also https://bralima.net/fondation-bralima/. 251 Moïse refers here to the Liberation Wars at the end of the 1990s – see section 5.2.3. 252 43-I-YAF-20160328. 253 1-I-YALO-20150518, 2-I-ISA-20150519, 34-I-YALO-20160324. 254 7-I-RBU-20160305, 8-IFG-RAB-20170516, 36-I-YALI-20160325. 180

(PRODAT) of the Belgian Development Agency (BTC-CTB at the time, now Enabel). These programs have pushed the production of rice in these areas and have supplied farmers with seeds and the area with small warehouses, rice huller machines and even motorbikes – for transport of production. The rice huller machines in theory had to be managed by farmers’ cooperatives or local development committees but have in fact been re-appropriated by private actors who make users pay for access.255 The remote village of Pumuzika in the Alibuku landscape has received similar support from the Food and Agriculture Organization (FAO) in 2011.256 The local transformation of rice increases the product’s quality and attractiveness, and in combination with local stocking thus allows producers to obtain better prices. The low perishability of rice – which is increased with good storage – is particularly important in more remote areas where the degradation of road and transport infrastructure renders the access to markets more difficult and thus less frequent.257 In comparison, plantains, for instance “do not sell well here [in Isangi], they do not yield much money”, Pierre from Yafunga asserted258 while “maize can be sold and bring money only if you process it [into flour]”259 another explained. In the absence of local flour processing equipment – which are much less common in the area as these crops have not been so much the target of projects – plantains or maize are more perishable and have less added value than transformed rice. Moreover, the latter has also a better volume-weight to price ratio than other crops (the unit is the bag rather than kilos) and thus decreases the cost of transport. Moïse, who has stopped growing maize in favor of rice explained this to me: “we used to do maize when SAFBOIS [a logging company that operated in Yafunga] was still here because then we could sell it at the boat but now the sale of maize does not work well”.260 Indeed, where logging companies operate, transportation of smallholder’s agricultural products is facilitated through the opening and maintenance of logging roads and as companies’ trucks and boats are used by farmers as means to bring their products to the markets in Kisangani or Kinshasa. Elizee, a widowed woman farmer for Yafunga, told me:

“If we want to sell in Kinshasa, it is because the rates, especially for maize, are much higher. You can sell a bag at 70 000 or 80 000FC while here it is 5000 to 8000FC, rarely 10 000FC. We took advantage of the boat of SAFBOIS because we did not have to pay anything, we just had to show our authorization. It was a right [we had] because the company settled on our land and so the boat took, each time, five people for free”.261

In areas that are closer to Kisangani, the fact that farmers are able to sell off perishable products such as manioc roots and leaves, plantains and garden produce more easily and at a better price on the larger

255 1-I-YALO-20150518, 13-CSI-KIS-20150522, 23-ITSP-KIS-20160401. See also (Alongo, Osombause, Bolakonga, & Lokinda, 2012). 256 13-PM-ARAB-20180514. 257 See also Bolakonga Ilye, Disonama, Michel, and Lebailly (2012) who have made similar observations. 258 40-I-YALI-20160326. 259 35-I-BU-20160324. 260 43-I-YAF-20160328. 261 48-I-YAF-20160330. 181

urban market of Kisangani partly explains their lower reliance on rice. Unlike the more remote villages of Isangi territory, some intermediaries/retailers come from Kisangani to buy products directly at the production location, and farmers can also rely on local workforce specialized in the transport of agricultural production to Kisangani by bicycle or motorbike at a lower cost. One particular socio- ecological condition was also mentioned by farmers on the first part of the Buta road (KP9-KP24) to explain the giving up on rice crops. In this area made of fallows and no primary or secondary forests, birds seem to destroy rice fields. Emmanuel, the chief of Batiamaduka village at kilometer 15 on the Buta road explained this in this way: “Rice is eaten by small birds. In the past, birds did not come because there were big forests. They could not get through. Therefore, we do not do rice cultivation anymore”.262 Moreover, as I showed earlier, on this part of the ‘Buta road’, villages often report a shortage of ngonda* and secondary forests, which seems to have pushed some farmers to focus on other profitable cash crops such as niébé and groundnuts that grow well in shorter fallows.263 As in Isangi for rice, the production of groundnuts by some better-off households on the ‘Buta road’ has also been boosted by two projects (APILAF and ADIKIS) which had distributed seeds to selected farmers – those who had their own easily accessible fields on the road. In response to the ecological particularities of the area, other local development NGO projects have also recently tried – with poor success until now – to introduce these villages to shallow rice varieties that they can plant in the swampy soils of the small remaining primary forests. The diversification of land use activities – in particular charcoal production – which can be more profitable, and of actors is also higher in areas closer to Kisangani and can also explain their lower reliance on rice as a main source of income than in more distant geographical areas.

While it was pushed during the colonial era, perennial cash crop agriculture such as oil palm, cocoa, coffee or cotton is not very widespread in the hinterland, except for the recent plantations developed by Nande elites from Kisangani – see Part I chapter 3. During the Zairianisation period, as I mentioned in chapter 5, colonial plantations and factories were redistributed to Mobutu’s privileged class of elites and went through a long decline. Cocoa and coffee plants were quickly decimated by fungal diseases. As I explained in Part I – chapter 3, despite external pushes in this direction, farmers today are particularly reluctant to engage in cocoa or coffee for several reasons. First, it poses a threat to their food and income security. Cocoa and coffee are highly sensitive to pests and diseases and both require more maintenance and labour input than any other crops (in particular manioc and plantains). And, they only start to produce pods from about three years after planting. As Moïse, a farmer from Yafunga who had tried to plant cocoa without success, said: “cocoa is a monocrop, so we cannot associate with other crops”264; “it is impossible to abandon food crop cultivation”, said another farmer during a focus group

262 22-I-RBU-20160315. 263 As I show in next sections, because these crops are labor-intensive, only some better-off farmers can focus on such crops, and manioc and maize are the favorite for poorer households. 264 43-I-YAF-20160328. 182

in Kazombo on the ‘Amex-Bois road’.265 Second, these are purely export crops whose purchase price in the hinterland remains volatile. Third, as we have seen, many farmers in the hinterland do not have any land security, and planting perennial crops on rented land would be a lost investment in most cases.

Oil palm had another trajectory after independence. Oil palm is native to tropical Africa and was used (as a wild tree) or cultivated since ancient times (Vansina, 1990). Oil palm is, historically, of central importance to the diet and social life of the rural world. Its nuts are used for oil, wine and soap, while the tree provide larvae for food, timber for furniture, and fronds for craftwork. Its production is usually a subsidiary activity to the elanga* and remains very small-scale and local. Oil or wine is extracted by hand in local artisanal processing units. A few farmers I met who have some sort of land security have small plantations of one to four or five hectares, like Victor a dweller of KP13 on the ‘Buta road’ who inherited a small plantation in another village that his father, who lived in Kisangani, started in 1972 on a land he had purchased.266 Others, like B in ‘Belgika’, 62 kilometres from Kisangani on the ‘Buta road’, use the oil palm from former colonial plantations despite the deterioration of the trees’ yield. Yet in many cases, oil palm are just planted behind the residential areas or in the fallows rather than as ‘plantations’, or are harvested ‘from the wild’ in the villages’ and clans’ forests. In most cases, the extra production is sold locally.

The popularity of some crops as cash food crops over others is thus dynamic and often pushed by various (f)actors. The farmer’s elanga*, field-garden, is connected to the wider landscape. It is created and shaped in connectivity with diverse socio-ecological and economic relations across space and time: from the Atlantic trade and the increasing Belgian colony’s agricultural export revenues to developmentalist programs and the needs of a multinational brewing company; from growing urbanization in DRC and around the globe to the impacts of climate change on Asian agricultural lands. The materiality (or ecology) and value of the crops also plays a fundamental role in coproducing land use practices, as I further show in the following subsections.

6.1.3. Seasonality

The cycle of seasons, of rain and drought and the ecology of multiple crops and weeds still largely shape the patterns of agricultural activities. While in this region of DRC, which is situated directly on the equator, it rains almost the whole year round and drought is relative, the seasonal phenomenon is real, with important variations in rainfall. These fluctuations give rhythm to the agricultural calendar. Generally, farmers in the areas I have investigated report that they organize their food crop production over a year, divided between season A in which they prepare a principal field, and season B in which some of them

265 16-IFG-RAB-20190522. 266 24-I-RBU-20160315. 183

have a subsidiary field.267 However, as I explain below, in the areas of the hinterland closer to Kisangani, like the first 25 kilometers of the ‘Buta road’ that I have investigated, people report that they do three to four fields a year. In this subsection though, for the sake of clarity, I will focus on the most common agricultural cycle in season A and B, before developing on factors that explain variety between different areas and households’ strategies in the following subsections.

While the dry period marks the beginning of the agricultural cycle, the exact beginning of the activities might fluctuate with weather conditions – seasons have become disrupted according to many farmers 268– but also depends on the chosen area for cultivation. In season A for instance, if a farmer decides to use a plot in the primary forest or in a denser post-agricultural secondary forest in which there are large trees, felling can start as soon as September and clearing is carried on until February. With younger fallows and less dense secondary forests, however, clearing starts later and is done between the end of November and February, during the most distinct dry period of the year so as to allow for the vegetation to dry thoroughly before burning at the end of February. In season B, clearing happens between May and June, and burning in July during the short dry season. When exactly the farmer starts clearing thus generally depends on the age of the fallow and its density but the arrival of the dry period marks the beginning of an agricultural cycle. However, most farmers use younger fallows in season B because the dry period is particularly short and thus does not allow the vegetation to sufficiently dry before the heavy and longer rainy season that lasts from August to November. The longer and often heavier rains of the long rainy season can also put the crops at risk which leads farmers to privilege season A for their principal field.269 Moreover, because clearing for season B overlaps with harvesting of the first crops of season A – see below – farmers face a labour bottleneck. In some cases, as illustrated in Picture 7 taken in May 2018, some peasants open areas for planting plantain bananas that are much less sensitive to weather variations and require less care after planting (Vansina, 1990). Yet for many, and in most parts of the hinterland for the reasons I just mentioned, the second field of the year in season B is subsidiary to the main one and some cultivate on the same plot of land for the two seasons, in particular if they just cut out a field in the primary forests. As I explain below, only better-off households can in fact afford to cut out a second – or more – important field in denser fallows or forests in season B. Such differences between seasons A and B are also visible in Figure 21, which shows a comparison of the distribution of fire alerts captured by NASA’s Active Fires satellite system in the Kisangani hinterland between December 2019 and April 2020 – which corresponds to an extended season A felling/clearing-burning period to account for possible seasonal climate variations – and between May and September 2019 – which

267 Please note that in the Opala territory which is located South-West of Kisangani and which I have not investigated, farmers grow rice in season B. This seems however to be an exception in the Province. 268 E.g. 6-I-RBU-20160305, 15-RBU-20160309, 17-I-RBU-20160312, 20-I-RBU-20160314, 43-I-YAF-20160328. 269 12-RBU-20160308. 184

corresponds to last year’s extended season B felling/clearing-burning period.270 In the whole ‘Orientale’ region that includes the Tshopo Province, more than fifteen times fewer fires are recorded in season B compared to season A. Note that there is a major exception in the area South-West of Kisangani. This region, which I have not investigated, corresponds to the Opala territory which presents a particular agricultural cycle as rice is primarily grown in season B, unlike other areas of the hinterland. Note also that the areas closer to Kisangani, as I have already hinted at and will develop in the following subsections, also present a substantial number of fires, indicating the opening of more than one important field a year.

Figure 21. Distribution of fire alerts in the hinterland of Kisangani, the period December 2019-April 2020 (season A) on the left, and between May and September 2019 on the right (season B of last year). In the period Dec 2019-Apr 2020 a total of 196 804 fire alerts were recorded in the whole ‘Orientale’ region, while only 12 099 were recorded in the period May-Sep 2019. Source: (GFW Fires, 2020a, 2020b)

Seedling roughly takes place between March and May in season A and August-September in season B, at the beginning of each rainy season to ensure the crops receive ample water. In one season, the chosen crops will depend on the type of fallow/forest a farmer is using for his field, as crops are more or less adapted to different biotopes. If a farmer has just cut out his field in the primary forest, plantains and manioc are generally the favoured crops because they tolerate some shade (Alexandre, 1989; Vansina, 1990) if the clearing and cutting of trees has not been totally finished271. Rice, maize, niébé* cowpea or groundnuts usually come into play in the second or third cycle and are productive in fallows and secondary forests. I often heard that the best manioc and plantains come out of primary forests, although manioc is always a must and is cultivated in every single field, every single season. The association of crops is thus not the same from one year/season to the next but neither is it from a farmer household to

270 Please note that data more than one year old cannot be accessed on the Global Forest Watch Fires online interactive platform. 271 As I explain in section 5.3., farmers usually keep some tree species e.g. for charcoal production or artisanal logging. 185

another because it is highly dependent on other important socio-economic variables that I develop in the following subsections.

Harvesting is extended over more than a year as crops have different growth durations. In one season, rice and/or maize, niébé* and groundnuts are usually planted a few weeks before the non-seasonal perennial manioc and are harvested first (July-August) as they reach maturity after three to six months depending on the varieties. These crops provide a quick source of income. As one farmer in Isangi said, “it brings money fast”272, which allows households to further invest in their fields, pay for their children’s school fees and cover other household expenses. Such crops are, for this reason, mainly grown in season A rather than in season B. After the harvest of rice, maize, groundnut or niébé* farmers uproot and weed out the remaining rice/maize weeds to allow a better yield for the other remaining crops. Manioc, depending on the variety, usually takes six to twelve months to reach maturity and is harvested progressively, over the course of (often) more than a year.273 It provides a steady and nutritious source of food for households as manioc roots and leaves are the most popular food in the area, e.g. in the form of chikwangue*, fufu* and pondu*.274 As two farmers in Likoko on the Amex-Bois road told me: “[regarding] the harvest of manioc and plantains… to leave the field totally empty it can take up to two years, even three”.275 Plantains are in fact a long term investment as they can be planted and harvested throughout the year, and produce for several years – two to three years – while other cultivations progressively disappear and the field slowly becomes fallow again. Stands of plantains are also often found as a secondary crop in the small ‘permanent’ gardens – a kind of kitchen garden – right behind a household’s residential area – which is usually composed of several small huts and a separate kitchen area– along with a few other fruit trees, like oil palm trees, for the family’s own everyday consumption.

Although weather seasonality has historically determined the rhythm of activities (see e.g. Vansina, 1990), such a seemingly uniform agricultural calendar divided in seasons A and B is an inheritance from the colonial system of cultures obligatoires* and paysannat indigènes*, as many informants asserted.276 The latter organized peasant farmers in rectangular cultivation corridors and rigidly structured their lives according to a crop rotation calendar inherited from the cultures obligatoires* system and adapted to local conditions, under the supervision of a moniteur agricole*. In the region of the then Stanleyville, as I showed in the previous subsection, rice and maize became the main cash crops and determined the cultivation cycle; a practice that has thus largely remained. Farmer households, my informants said, had to individually cultivate two fields of about 0.5ha to 1ha in rotation over a period of a year – the size of

272 2-I-ISA-20150519. 273 E.g.10-I-RBU-20160307, 35-I-BU-20160324, 55-I-RAB-20170509, 81-I-ARBU-20180502, 14-PM-BEN-20190516. 274 Chikwangue: stick of fully-cooked cassava wrapped in a (usually banana) leaf. Fufu: ball of boiled cassava flour (sometimes corn flour or a mix of both). Pondu: cassava leaves; when cooked they look like spinach. 275 55-I-RAB-20170509. 276 E.g. 6-I-RBU-20160305, 30-I-YALO-20160323, 51-I-YAF-20160330. 186

the plot still varied from one household to the other. At the beginning of the rotation, in season A, peasants had to grow rice or maize alone to maximize output. Staple crops such as manioc, plantains or groundnuts came in subsequent cycles. After a few cycles –some informants said one rotation, but it is difficult to assess – the plot had to be left in fallow. In the system of cultures obligatoires*, fallows had been greatly shortened compared to pre-colonial practices but were set (in theory) for ten to fifteen years in the paysannats indigènes* system. It is unclear whether such fallow durations were implemented in practice (Clement, 2014; Malengreau, 1949). The legacies of the colonial peasant agriculture policies appear clear in the widespread adoption of rice and maize as main cash crops –see also 6.1.2 – and the way cash crops generally still structure the agricultural cycle. The contiguity of cultivated lands and crops was aimed at “introducing monoculture sooner or later” (Malengreau, 1949, p. 14). Today’s typical elanga*’s multicropping system shows that this never truly happened and fields were replenished with multiple species. In fact, as Likaka (1997) and Ross (2017) have shown, ‘indigenous’ peasants would engage in forbidden intercropping practices as hidden resistance strategies to undermine the production cycle. Yet, more recently, the more a farmer has land security and economic power, the more he has tended towards a cash crop plantation mode of production, using paid workers, and focusing on extending the size of his fields and on higher-value crops, as I show below. Similarly, although the resettlement for paysannats indigènes* still has important socio-ecological impacts, the pre-set field- fallow cycles in a corridor did not remain in place nor did its strict labour division. Rather, various ecological, socio-economic and cultural factors shape differentiated farming practices and land use as I develop in the two following subsections.

6.1.4. Organizing production: between mutuality and individuality

The family household – which is usually composed of an autonomous male adult (head), his wife/wives, and non-married children – functions as the primary production and consumption unit. As I have mentioned before, each household has its own fields, fallows and post-agricultural secondary forests that are inside a matrix of communal clan/village and groupement* forests. Gender is a significant variable in the organization of farming activities and the shaping of land use patterns. For each annual or seasonal principal field(s), men, almost exclusively, are in charge of preparing the field, that is cutting trees (in primary and secondary forests), clearing fields and burning, while women are in charge of planting, hoeing, weeding and harvesting although men participate in these activities when the need is great and especially at the main harvest period for rice and maize, i.e. for cash food crops.277 However, while the domestic unit functions as the basis, farming ties each household to a wider social network in the village through the pooling of the group’s workforce. While the paysannats indigènes* system had imposed on farmer households the requirement to work individually in a corridor, next to their fellow clan members,

277 48-I-YAF-20160330, 13-PM-ARAB-20180514, 14-PM-BEN-20190516. 187

today’s production model is rather built on a tension between reciprocity/mutuality – maintaining relationships with others – and privatization.

In the whole hinterland, farming is mostly organized through a system of work tontines called likelemba* in Lingala or ristournes* in Lingala/French which is the most common usage. Ristournes* are relatively stable, informal mutuality groups of men or women – they are never mixed – which function as integrated systems of credit-saving and community work. Ristournes* usually group ten to fifteen members but can (more rarely) be larger, e.g. twenty-five members in the case of a ristourne* organized by young men.278 A ristourne* can be organized on several bases: age, church affiliation, “tribe” or clan affiliation –l ike in Pumuzika where Mongo, Mbole, Tetela all have separate ristournes*– but financial means remain the main determiner, i.e. each member needs to have the same financial capacity to contribute to the credit- saving system.279 It requires a high level of trust between its members and therefore households who recently arrived in a village, e.g. for renting land, or are isolated geographically like some households who just settled on new logging roads in the COTREFOR/IFCO concession, are much less likely to be part of a ristourne*. In contrast, people with a wider social network are often part of several ristournes* in their community.

Every week or every month, each member puts a certain amount of money in the common moneybox; and every week or every month, the total sum is transferred to one of the group’s members by turns – this order can be changed if a member faces adverse events, e.g. a funeral. Didier, who is a member of a ristourne* grouping of Topoke dwellers in Azunu, explained:

“According to capacities, we do ristournes*. In mine, we are ten [people], we each give 20 000280 [Congolese francs] every 10 days. We give to one person, after 10 days, 200 000. It works in rotation […] you do it only with honest [reliable] people. […] This amount corresponds to the [financial] means of high-class people. But almost everywhere, the mamas do something like 1000CDF, 500CDF, 3000CDF according to their capacities.”281

The money needs to be used for concrete activities, investments or events, such as buying corrugated iron panels for the house or a bicycle to transport agricultural products, or investing in an oil palm plantation, rather than for personal saving or everyday life such as buying food. These ristournes* are also tied to agricultural activities. Members work in rotation on each other’s land. Forms of extended kinfolk’s collective labour can be traced to pre-colonial times, for the tasks of felling and clearing (Vansina, 1990). Just like for savings, work ristournes* implies that the cultivation area is about the same for each member, and rarely exceeds 1-1,5 hectare. Such collective work happens periodically, following the

278 5-IFG-RBA-20160314. 279 13-PM-ARAB-20180514. 280 20 000CDF is about 14-15$. 1$ = 1400CDF. 281 65-I-RAB-20170516. 188

schedule of activities of the agricultural calendar: felling/clearing, sowing, (less commonly) weeding, and main harvests of short cycle cash crops (rice, maize). Men ristournes* are in charge of felling/clearing, while women ristournes* are in charge of sowing/weeding. Only some days of the week are dedicated to mutual work, while for the rest, members work individually in their own fields. A simple meal is prepared by the host of the day to feed his/her assistants in the field. Ristournes* are thus a collective, reciprocal mode of production that allows everyone to have relatively bigger fields and maximize their output while solidifying community bonds and maintaining equality between the members.

The collective labour force of ristournes* is also increasingly used as a wage labour force to increase cultivation areas, production and profit. Ristournes*, usually those made up of young men, provide paid ‘farming services’ on the land of other farmers in exchange of cash, food (rice, beans and oil) and drinks (palm wine). When ristournes* engage in wage labour, they are generally referred to as associations* and typically function as cooperatives.282 “Those who have a bit of financial means use associations that gather 20, 30, 40 people; you discuss the price and they start to plough and plant”, Suzanne a dweller from KP 13 on the ‘Buta road’, where paid labour is popular, explained to me.283 Some of these better- off farmers use the ristournes* system for part of their fields and associations* for another. They are particularly common closer to Kisangani, where money circulates and where someone in the household is more likely to have a small source of cash income in the city, such as being a moto-taxi driver or having a stall at the market. This phenomenon, many said, is quite recent. Before, “we would negotiate with a few young people to cultivate the fields. It was just a help […] We would not pay them. We would prepare some food… bushmeat and manioc. When they had finished working, [they would eat]”, Suzanne who is not older than 40, said. Likaka (1997), in his study of colonial cotton cultivation, traces the reliance on paid collective work to colonial forced cotton cultivation, during which some cultivators had to call upon their kin to help them work in their plot, which would otherwise not be ready when the controllers came. With the changing economic context, additional cash compensation became increasingly required, just like it is now – see also 6.2.

Only better-off farmers who have access to larger plots of land and have some financial means can afford such services. They allow them to extend their cultivation areas and invest in shorter-cycle crops which require a ‘fast’ harvest – unlike local varieties of manioc that can remain in the soil for long without rotting. Such associations* of collective labour mainly engage in typically male tasks, i.e. felling and clearing, while weeding is left to women – chiefs, notables* and local elites in the rural world usually have more than one wife – and small children or paid daily laborers as I show below. The price of services varies but are about $70 to $115 per hectare for felling and clearing, depending of the density of the

282 I understood that here are some associations* which are independent from the ristournes* system but the members of associations* I talked with all explained there were also organized in a system of mutuality among them. 283 21-I-RBU-20160314. 189

vegetation and the area in the hinterland. Nestor, a notable* and teacher from Yalosuna who uses his teacher allowance to pay associations* told me:

“My small salary does not allow me to make ends meet. So I use this small salary to pay some young men who will clear for us. When you give something to these young people, you will have a large field, much larger than those who relied solely on their own efforts. […] I had rented some extra land in Bombola [for 30 000CDF and a goat], a [primary] forest quite far from our forests. […] The young cleared two hectares. The yield was really high. It helped us for three years.”284

The “Association pour le Développement des Jeunes de Batiafeke” created by young men from KP13 on the ‘Buta road’ works every Saturday for other farmers – each association* has its own schedule. Every member has to be present, and the “irregulars” have to pay 1000CDF every time they are absent on working day. The money the association* earns is put in a common cash box which is managed by a trustworthy treasurer who is part of a management committee. Wages are thus not directly distributed to members but are pooled and then used to finance goods, events or members in need:

“If a member of our association faces a problem like sickness, death of a relative, etc., we give him a sum of money. […] We buy televisions [for the group] to follow Canal+, plastic chairs, plates… Note also that the money we earn is also used to buy tools for the association. If you are twenty members in an association, we sometimes buy twenty plastic chairs and each member receives a chair. […] If we have worked for someone who appealed to our workforce and, for example, he pays use 20 000CDF… First all members will see this money before it is put in the till. Or, if we manage to earn an extra 100 000CDF, then we can share it between members. […] If nobody had problems [and there is remaining money], we organize a party for the association every year or two to thank God who protected us.”285

Associations* represent an economic alternative for some young people who are faced with land shortage and need some source of income while reinforcing ties with their generational community. The productive power of their collective labour thus serves simultaneously the reproduction of community and individual maximization, in the sense that reciprocity remains at the core of its functioning but it is also used by better-off individuals in their self-interest, i.e. to extend the size of their fields (generally to 3-4ha) and increase their yield. Yet, the production of wealth by these better-off individuals is generally considered to be constructive for the community, precisely because the money circulates within it (Gudeman, 2001, 2008). The money young people of the associations* make is not individually appropriated, it serves the young people, as much as the parents they help, while also producing value for particular farmers in the community (I further develop such understandings in section 6.2.). In Gibson-

284 28-I-YALO-20160322. 285 5-IFG-RBA-20160314. 190

Graham (2014)’s diverse economies framework, this kind of organization of labour could be categorized as ‘alternative paid’, ‘alternative capitalist’.

There is yet a third form of farming labour that one finds in the hinterland: purely privatized labour. While better-off farmers use associations* in the periods of major (male) agricultural works (clearing/felling, big harvests of cash crops), some of them also use extra individual daily labourers for the regular field maintenance and harvests. Daily labourers are paid by task, around $1 to $1.50 for an area of 10m to 30m. Most of these daily labourers are recently arrived ‘allochthones’ who are less integrated into community structures and have more difficult access to land, or autochthonous ‘landless’ young men (sometimes women) who are unable to generate an income from agriculture. For ‘autochthones’ who have access to land and forest resources, the salary is often considered too low compared to the income they can get out of their own fields and related activities, e.g. charcoal production. Daily labourers are called upon by the most well-off local farmers who grow more labour-intensive crops, e.g. groundnuts, maize, rice, on large plots of land or by the few farmers who have a small plot of perennial cash crops such as oil palm or cocoa. Urban elites who have acquired private concessions for staple and perennial cash crop agriculture, rely both on permanent paid labourers to take care of their fields, and on daily labourers during the major agricultural works. I observed that paid permanent labourers in private concessions are very often ‘imported allochthones’ who do not come from or live in the villages directly neighbouring the concession; they have a hut within the perimeter of the concession and are allowed to go back to their families on Sundays. Josué, the agronomist manager of a Nande cocoa plantation in the Alibuku landscape, told me that employing people who do not come from the area “ensures they are not distracted”, i.e. by their daily lives and social relations.286 Unlike ristournes* and associations* this individual form of labour is less ‘pro-social’, in the sense that it does not directly contribute to produce community social relations, in particular because many of the workers and users are not members of the local communities where the production takes place. Yet, some mechanisms of redistribution are generally developed by the users to avoid local accusations of illegitimate appropriation (Ferguson, 2006). One Nande cocoa producer, as I mentioned in Part I – chapter 3, has started to provide some local peasants with technical advice and cocoa seedlings to support them in very small-scale cocoa production which they sell back to him. As the Nande producer told me, this ensures social peace, i.e. “avoid local peasants stealing the production”.287

Agricultural labour is thus, more or less, productive of social relations. Yet, the access to different types of labour is constrained by access to land, social and economic capital, and in turn has an impact on agricultural output and income, and on farming land use patterns.

286 88-I-RAB-20180516. 287 88-I-RAB-20180516. 191

6.1.5. Entanglements

Over the last few sections, I have shown how historical environmental conditions (both physical and built), the materiality and socio-ecology of crops and weeds, and historical and present social relations of land access and labour use all coproduce farming practices and their socio-ecological outcomes. All aspects are intimately tied to and affect each other. Let me conclude this section by taking two examples to make clearer how such entanglements function.

Pastor Mutambaï is an autochthonous notable* from Batiamaduka, a village which was allocated large plots of land during colonial times, and pastor of the village’s largest church.288 He has managed to customarily secure large plots of land that are located fairly close to the ‘Buta road’, 14km from Kisangani, which enables him to easily sell off his agricultural production. Due to this close location to the city and a main commercial road, there is a general shortage of ‘primary’ forests, but Pastor Mutambaï is from a leading clan and can count on a large labour force to work in his shorter fallows. Indeed, thanks to his activity as a Pastor, which is paid through donations, he pays associations* and, occasionally, daily labourers to take care of his fields. Because he is so well connected in the community, he also receives the help of its congregation. In season A, he starts with maize in association with Vigna cowpea (niébé*), groundnuts, eggplants and manioc. He does not grow rice because of the birds that destroy it but if he lived in Yafunga, for instance, where the environment is more adapted to rice, he would certainly grow it too. Thanks to the paid labour Pastor Mutambaï uses, he is able to cultivate large areas of several cash crops with short maturity periods (3-4 months) and which require substantial maintenance (ploughing, weeding) and harvest work: they easily rot if they are left for too long in the soil after maturation so they need to be harvested in a few days. The harvest of niébe* and groundnuts is particularly labour- intensive.289 For this reason, only better-off households who are able to afford paid labour use niébe* and groundnuts as main cash crops. After the harvest of his short-cycle crops, Pastor Mutambaï sells them to buyers from Kisangani at a good price thanks to his proximity to the city and to the low supply of local niébe* and groundnuts in the hinterland. “Groundnuts provide a lot of money”, he said. The Pastor can then start a new rotation in season B, getting an even better price as the supply of staple crops is much lower outside season A. With such short-cycle crops and the good income they provide, he is actually able to do three rounds a year, sometimes on the same field, but more often rotating from one short fallow/secondary forests to another – he ‘owns’ a large plot of land. To extend his land further or access more fertile primary forests, Pastor Mutambaï will be able to rent or buy some other plots with the income he generates, such as Nestor from Yalosuna or Peter from KP12, whom I mentioned earlier, have done. And in the future, with such securitization of land, Pastor Mutambaï might even be able to plant some cocoa or oil palm trees as some privileged households have started to do, in the hope of

288 15-RBU-20160309. 289 See also 43-I-YAF-20160328. 192

further increasing his income. In local understandings of socio-ecological changes, he definitely contributes to ‘desertification’ but his position as Pastor means that the production of his wealth is seen as morally and socially legitimate.

At the opposite end of the social scale, Francisca, a poor single mother of two in Yaliaboga, a village in Isangi, illustrates the insecure conditions and cycle of precariousness poorer households and individuals find themselves trapped in, and a different set of socio-ecological pattern.290 Francisca explained to me that for some time she had been able to use, in rotation, a plot of land of about 2 hectares that her grandfather, a non-native, had received in exchange for the help he had provided to a family of Yaliaboga. However, despite this arrangement, she is now required to pay a one-off amount of CDF 75 000 – about $53-54 – and game to the family to be able to continue using this land. With such a small plot of land, she is only able to cultivate 0.5ha per year so as to make sure for the soil to regenerate during a 2-3-year fallow period. Francisca has planted a few oil palm trees that allow her to prepare and sell palm wine locally. “With this small activity, if I earn a bit of money, I can pay people to help me with the field”, she said. She has indeed to count on paid labour to help her fell and clear her plot of land. She is part of a ristourne* with some women who help her with sowing, but she mainly works by herself and with her children. Although she cultivates manioc in association with some rice or maize and other secondary food crops, the latter are only for her and her two children’s consumption. Manioc, on the other hand, allows her to generate a bit of cash by selling chikwangue* and pondu* locally as she cannot afford transportation to larger markets; retailers do not come so far and paying for transportation is of little interest for a product as cheap as manioc.291 Manioc does not need any processing, except for cooking chikwangue*, unlike rice that requires transformation and thus transportation and access costs to a processing machine. In the field, manioc is less sensitive to seasonal variations, requires less maintenance – in contrast to rice or maize for instance, which are much more sensitive to weeds (Alexandre, 1989; Vansina, 1990) – always provides a good yield and can be harvested six months to a couple of years after planting. It thus ensures food security (and a small income) for the most vulnerable households, like Francisca’s, who cannot afford to pay maintenance laborers and who cannot absorb shocks as much as better-off households do. Francisca had to face such a shock in 2016 when I interviewed her in March of that year:

“As at the moment I do not have any means [because of illness], I will not have a field this year. […] If I get 5000FC, I will be able to buy some sugarcane, which will allow me to ferment some 25 litres of palm wine and with this money I will be able to live and get treatment for my disease. […] There is nobody from the village who helps me, but I have a small field from last year

290 39-I-YALI-20160326. 291 Manioc crops are generally under women’s control. They are often in charge of selling manioc products, whereas rice or maize which are considered as cash crops are under men’s control. 193

from which I can [still] harvest some manioc little by little. This year it will not be possible to sell any manioc292. Next year, I will be able to sell manioc.”293

Isolated newcomer households who have just settled in a village to rent land or young men whose families have not provided them a plot of land to cultivate294 can be trapped in similar situations to Francisca’s. With limited access to land, financial means and community integration, they often have no choice other than to cultivate only one small plot of land per year, either in younger fallows which forces them to mainly rely on manioc and/or have a much lower yield as they cannot afford paid labour. They are unable to grow at the fast pace Pastor Mutambaï does. Alternatively, poorer/marginalized households leave their communities and encroach on primary forests much further away from Kisangani, which often results in a diminution of their profit on the small production they do manage to sell. These households just all mainly engage in ‘purely’ shifting subsistence agriculture. Sometimes, they can engage themselves in paid labour in the form of associations* or as daily laborers (or both), in their village or beyond. In any case, poorer and marginalized household farmers are usually left at the margin of a type of capitalist agriculture that is slowly spreading in Kisangani’s hinterland.

Pasteur Mutambaï and Francisca’s stories thus show how multiple ‘shifting’ farming practices are shaped in relation to various ecologies of forested land, historical and structural inequalities of gender, access to land and/or socio-economic capital, and what it means for material and social reproduction as much as for forests. The materiality and ecology of the crop and weeds themselves are not passive but important co-productive agents in producing particular socio-ecological relations of capital accumulation and dispossession, and thus in generating inequalities between farmers of the hinterland. These differentiated dynamics produce various socio-ecological outcomes and inform local understandings of environmental change, as I develop in the next section.

6.2. Forest changes as intensification and social-moral transformations

Forests – ‘the pristine wild’ in Western imaginary – as I have shown, cannot be thought of as separate from the cultivated, nor from the socio-cultural; they are part of the same continuum. When peasant farmers talk about forests, they always simultaneously talk about their life activities and their social stories. Thus, when they evoke ecological perturbations, they not only evoke climatic variations, forest and soil degradation but also speak explicitly or implicitly about societal transformations, accumulation and dispossession. In part I-chapter 2, I described the ‘REDD+ work of propaganda’ and the ways simplistic

292 As the remaining cassava can only cover her own consumption. 293 39-I-YALI-20160326. 294 As a reminder, in the patrilineal system characteristic of this region, land can only be passed down generationally through men, and women have to move from their birth homes to their husband’s households. Note, however, that all women are not equal when it comes to their limited access to land. While the ‘rule’ is that land can only be transmitted to men, with the evolving customs, some better-off fathers and local chiefs are able to transmit land to their daughters. 194

narratives on deforestation and harmful shifting cultivation have partly penetrated local discourses on environmental change in certain areas targeted by communication and awareness plans, including radio announcements. In this section, I am interested in endogenous stories and understandings of forest and soil changes and in the ways they connect to the main activity of the inhabitants I researched, i.e. farming.

During my first three periods of fieldwork in the Kisangani hinterland, I systematically asked people about the changes they had seen over the years in their natural environment, trying to avoid technical words such as deforestation or degradation. It quickly became clear that their answers followed the patterns of deforestation that are visible on satellite imageries. Villagers deeper in the Alibuku landscape (Azunu, Pumuzika, Malaïka), outside the ‘rural complex’ were often surprised by my question, looked around at the natural landscape and told me there are, obviously, plenty of forests. Installation along the road built by COTREFOR/IFCO is much more recent than for villages on ‘historical roads’. People dwelling in villages along historical roads, where they were resettled during the colonial era, report increasing distance from the village to primary forests. Roger, an old notable* from Yafunga, told me:

“if you try to compare, at the time when I grew up, we had the virgin forest just next to the house, but today the primal forest has gone away from the village. […] Now you need to go beyond one or two kilometres from the road [to find primary forests].”295

Unsurprisingly, distances from the road to primary forest are greater in the ‘rural complex’ on the Buta road close to the city. Inhabitants of these villages often explain that “here it used to be a large forest” before the Belgians opened the road in the 1950s.296 Yet, they simultaneously refer to their large ancestral primary forests or to other groups’ forests, located away from the road. As Chef Assumani, a clan’s chief on the ‘Buta road’ said, “the Manga, Ngelema, Bali all still have their forests. Even the Kumu who are in the forests297, they still have their forests. All have forests with high-size trees. [This is because] the Belgians just gave us 5km2 of forests.”298 Primary forest shortage in their understanding is relative rather than absolute, in that it is seen as an issue of local scarcity due to high population density, land alienation by some groups and intensification in specific places, that is in turn the legacy of past and recent resettlement, urbanization, development and roads. Igo’s quote, a farmer on the KP13, summarizes most of the discourses I heard on the Buta road. It is worth quoting him at length:

“If someone [a villager] needs money […] and that someone else from Kisangani comes to ask for a plot of land to rent with money, then the [villager] will rent out fields that he had left in fallow. Thus now when money

295 50-I-YAF-20160330. 296 E.g. 3-I-RBU-20160304. 297 The Kumu are considered to be one of the native ethnic groups of Kisangani and its hinterland. It is the dominant ethnic group in the Lubuya-Bera Collectivity-Sector, and thus on the first 18 kilometers of the Buta road where I worked, but they are spread in all forests around the city. As a reminder, the Badombi are considered to be a subgroup of the Kumu. 298 13-I-RBU-20160308. 195

comes, fallows are cultivated. And when these people [from Kisangani] rent land in this way, they usually use the [same] land for long periods. Let’s take the example of the Amex-Bois road. It used to be a primary forest. That was our Lubuya collectivity’s forest reserve. But there are people who come from very far [e.g. Nande from North Kivu, urban elites from Kinshasa] to buy large areas of land and today, this area […] has also become like a desert. Now, big companies that harvest timber destroy the forest. They fell several big trees per day. Even the trees that would produce caterpillars are cut. Today we are forced to buy caterpillars to eat because it is difficult to find them around here. The one who wasn’t a cultivator, has become one today. The one who wasn’t producing charcoal, does it today. But even this charcoal, we, local dwellers, we used to use one tree trunk for several months. But those who come with their chainsaws, they feel ten to twenty tree trunks per day. If they manage to exploit for a whole month or two, the whole forest disappears.”299

There are several key elements in this quote on which I will build. First, local inhabitants emphasize an important link between industrial logging and the expansion of other land-use change activities in concessions. I already developed such aspects in Part I – chapter 2, so I will not come back to this here. Second, it is the intensification of logging, charcoal production and farming – I will focus on the latter in particular – by certain classes of actors, which is rendered possible by various dynamics of land grabbing, that is causing forests to turn into ‘deserts’. This idea of desertification, which was very often mentioned to me in this area of the hinterland, refers to what is considered to be an overuse of resources (e.g. over- farming, over-hunting, ‘over-‘charcoaling’) on a limited, fixed (often customarily or legally ‘privatized’) spatial area, that does not allow sufficient forest regrowth to occur. “They say the forest is disappearing… but the forest has not disappeared, because we still have space to cultivate. But we rather noticed that […] there are only small-size trees and tall grass [in the immediate vicinity]”, said Tine, a woman dweller of KP13 who farms with her son. 300 Local farmers particularly refer to one type of weed that their fallows are covered with and that they name, quite poetically, ‘mokili mobimba*’ which means “all over the world”. It echoes the loss of biodiversity, in terms of plants, game animals and non-timber forest products, which they often mentioned. Similarly, another inhabitant a bit further along the Buta road clarified the concept of ‘desert’:

“I am referring to desert because in the same area, people cultivate it more than three times and they exterminate all the big trees there. It is in this way that there remains only desert, but it is not in the real meaning of desert because we still produce in it.”301

299 12-RBU-20160308. 300 5-I-RBU-20160305. 301 17-I-RBU-20160312. 196

In other words, there are indeed new areas of expansion that go with the construction of road infrastructure; but the heart of the problem for peasant farmers of the Buta road and Amex-Bois road around Alibuku, lies in the intensification and lack of land rotation that infrastructure development brings. For many, this also leads to a reduction in fertility. This understanding of deforestation- desertification, in this sense, reflects the philosophy of shifting agriculture. The ngonda*, ‘primary’ forest, is not a pristine place independent from human activity; it is built in cycles by humans and nonhuman forms of life alike. As Pastor Mutambaï whom I presented just above said: “today we have issues because we stay in the same place while at the time, our ancestors would move barrier-free […] they considered that the whole forest belonged to them”.302 Thus, the issue for them, and as expected from the analysis I have developed over the few last chapters, lies in the perturbations of an extended and cyclical spatial- temporality that characterized traditional land use and tenure practices.

Third, the above long quote from Igo indicates why and how such intensification is possible, by whom it is carried out and the inequalities it generates. I will only be focusing on agriculture but many of the arguments I am making can be applied to charcoal production and artisanal logging as similar processes appear in these sectors. The alienation of land and labour is, in many ways, what underlies local discourses on deforestation-desertification. The allotment of large areas to dwellers from Kisangani is often mentioned. Two sages* from Alibuku noted: “our Badombi neighbours are selling concessions from twenty to a hundred hectares to the Nande; someone who does one hundred hectares, that is a desert foretold”.303 Yet, peasant farmers, refer even more often to intra-community land alienation by some better-off “individuals who can, like that, cultivate two, three, four fields”.304 Forest shortages that follow from the reduction of fallow years (intensification) due to an expansion of annual crop land in the family/clan’s land area, is consistently linked by my informants to the rise of paid associations* labour and daily labourers. Paid labour is simultaneously understood as a cause and an effect of socio-ecological change. They are, for many, a response to family forest and land scarcity, unequal access to land and land alienation while allowing further intensification and expansion of annual crop land, for certain groups. We have seen how this two-way relationship also forces some poorer and marginalized households to encroach on forests and grow small fields where land access is easier (e.g. deeper in the Alibuku landscape).

In such explanations social and moral transformations, seen in both a positive and negative light, systematically come up in the conversation. The monetization of the economy and the access to material goods, i.e. ‘development’ in their words, are drivers of such changes. “Since the abundance of money and the arrival of development, people do large fields to make money”, said Loku, a middle-aged inhabitant

302 15-RBU-20160309. 303 59-I-RAB-20170511. 304 11-RBU-20160307. 197

of Yalosuna.305 While their grandparents and parents were engaged in a bartering system, and thus just needed to do small fields for their household’s subsistence only, Boni, an old notable* of the same village asserted that: “the white showed us how to do the market [economy], he opened roads […] but he was hiding the money from us; but after independence, we opened our eyes and we could make money and change came”.306,307 The ‘notion of lack’ powerfully started to ‘colonize’ the minds (Fanon, 1968). Markets and native currencies existed in pre-colonial times, but the Western monetized economy spread with colonialism. It is under the system of cultures obligatoires* that agriculture started to become an (export) cash-making activity and that agricultural wage labour emerged, and would continue, as I have shown, to be embodied in human-environment relations and incorporated in the landscape (Stoetzer, 2018). Peri-urban dwellers, said my informants, have developed increasing needs since then, and with “mundialization”308. “Today, goods have become visible everywhere and people do everything to get them; there was the 4x4 bicycle and then another brand and so on”, said Pastor Augustin from Yaliaboga in Isangi territory.309 Corrugated iron roofs, motorbikes, televisions, etc. are all signs of modernity as for most peri-urban population throughout the region (see also Trefon (2011)). This money is also what allows them to pay for school fees, and for the lucky ones to send their children to school in Kisangani. Emmanuel, the chief of Batiamaduka, 15 km from Kisangani city centre added: “people want to have a better life, like people in the city. Just like our African brothers (Central African Republic, Sudan, etc.) have changed, we want to change too”.310 It is the urban-rural binary, i.e. culture-nature distinction, exported by colonialists, that my informants aim to break (Comaroff & Comaroff, 1997). And, in the hinterland, “the biggest company is the forest”, said Chef Assumani; without any other economic alternative, large fields have become for better-off rural households and for some dwellers of Kisangani the only pathway to ‘development’.

Socio-economic and ecological change, in many peasant farmers’ discourses, has gone hand in hand with moral, spiritual change. Belief systems are, as I have mentioned earlier, integrated with socio-ecological systems, not a separate dogma. The replacement of traditional beliefs – ‘la coutume*’ – by monotheism (mainly Christian religion but also other modern religious syncretic movements), many asserted, has helped make forests lose their sacredness and thus their inaccessibility – although such sacredness is sometimes strategically used to justify control over land (see also 5.2.2 and 5.2.5). This loss of sacredness, they say, is in turn what pushes people to overuse forest resources. Forests are no longer a source of

305 30-I-YALO-20160323. 306 29-I-YALO-20160322. 307 Just like anybody who has done research in DRC has probably observed, the relation to the colonial era is particularly ambiguous. My informants often referred to the colonial period as the ‘golden era’, often playing down its violence and the alienation of Congolese people’s land, labour and money. I have become increasingly interested in such collective memory but the topic would go beyond the scope of this manuscript. 308 22-I-RBU-20160315. 309 36-I-YALI-20160325. 310 68-I-RAB-20170516. 198

ritual focus as I have already noted. Believing in God rather than in spirits, charms or secret societies, leads to development. “Religion says that we should follow the life of the Jews… and the Jews lived better than us”, said Boni, the old notable* from Yalosuna.311 This assertion echoes Comaroff and Comaroff (1997)’s analysis of Christian missionaries’ attempts to develop a sense of materialistic individualism among South Africans, and recast their personhood and notions of value. “Darkness hunt is what benefited Europe’s development”, said the director of the INERA herbarium in Yangambi, during a conversation over the evolution of Congolese agriculture and deforestation, seeming to forget that the ‘development’ they praise was built by the colonized people’s land, labour and money.312 But as Vansina (1990, p. 245) has shown, back then already the missionaries who forcefully campaigned against traditions found followers among young people, “to evaluate the core of their own tradition by the criteria of a foreign biblical tradition”. The colonial imagery of ‘darkness’, the corollary of the ‘lack’, has deeply and enduringly colonized Congolese minds and partly shapes understandings of socio-ecological change. Previous assertions, as much as what follows, have to be understood in this context. Yet as I will soon return to, traditional beliefs, e.g. witchcraft, remain central in making sense of today’s socio- ecological transformations, creating a particularly ambiguous feeling towards the loss of traditional practices as both a driver of detrimental ecological change and of a positive economic change, of social conflict and of community development.

For the majority of peasant farmers in the Kisangani hinterland, such ‘modernization’ remains an “economy of desired goods that are known, that may sometimes be seen, that one wants to enjoy, but to which one will never have material access” (Mbembe, 2002, p. 271). Socially, the peasant farmers’ society of the hinterland has bifurcated between those who manage to cultivate large fields, to have good yield, and to access (more or less) ‘modernization’ but contributing to ‘desertification’, and those who do not. As in many domains, witchcraft appears as a dynamic response to such socio-ecological changes. Fonoli*, a form of witchcraft came up many times in conversations about farming, forests and deforestation. Fonoli*, people from different groups told me, is a mystical, occult force that ‘magically’ captures people, makes them amnesiac and enslaves them for agricultural work in the witch’s fields.313 Accusations of fonoli* are aimed at individuals who have bigger and/or more productive (high-yielding) fields than others. The reality of its existence and use is not questioned in my informants’ assertions. One inhabitant on the Buta road explained its functioning:

“Fonoli is a system in which someone is killed through witchcraft so that this person who is supposedly dead goes working, at night, in the fields of the one who killed him. This person is not dead in a real way, he stays in the

311 29-I-YALO-20160322. 312 81-CTSP-YAN-20190530. 313 E.g. 10-I-RBU-20160307, 25-I-RBU-20160316, 26-I-YALO-20160322, 28-I-YALO-20160322, 30-I-YALO-20160323, 34-I-YALO- 20160324. 199

forest so that he can be used for working in the fields. We notice in the morning that this person’s field [of the witch] is always clean [weeded]. We ask ourselves: ‘how can his fields remain so clean every morning?’ And then we understand that someone works in it at night.”314

Nestor, a notable* from Yalosuna asserted that fonoli* practice appeared during the colonial period, to mobilize a workforce in the fields.315 Fonoli*’s imaginary indeed echoes forced labour under the cultures obligatoires* system, and Rösler (1997) notes that in Ituri witchcraft accusations were legion if crops failed on some fields while others were spared. The occult power of fonoli* nowadays still accounts for agricultural intensification and expansion – and its consequences on forests, and individual good fortune. Informants who shared their knowledge about fonoli* – always with a lot of caution because “it happens in the world of darkness”316 and people fear accusations – usually mentioned, a few minutes before in the conversation, that the competition between villagers and conflicts over land had increased with the spread of ‘modernization’ and its impact on the natural environment. Joseph, the well-off chief of Yalosuna who participated in the PPRGII REDD+ project awareness-raising workshop, formulated it this way:

“Deforestation yes… because when you [a neighbour] cultivate a lot, it bothers me. We are in the same family, really, but you, you are busy cultivating a lot and I am not strong to cultivate a lot. This creates conflicts. It depends on everyone’s force or on financial means. If you, you have sufficient means and me I don’t have any, I will be angry against the one who has sufficient means. Because me, I do not find anything. You, you benefit a lot, you have got televisions, water in your house… The forest is for us, so I won’t be happy. Then I will create problems for you. The forest has to be distributed. […] ”317

In this quote, I translated the word ‘force’ in French literally to ‘force’ in English to emphasize its link with occult power, as it is meant by Joseph, the author of this quote. ‘Force’ in many of my informants’ discourses has intertwined connotations in the context of farming and land tenure: as economic capital, physical strength of labour and occult power. The three forces are all framed (often indistinctively) as human and nonhuman agents of socio-ecological changes that the inhabitants of the hinterland are experiencing. They are in fact quite impossible to disentangle from each other in peasant farmers’ hidden interpretative transcripts although they are carefully distinguished in their public transcripts shaped by Christian worldviews (Scott, 1992). The strength of paid labour, especially in its collective form, is the

314 25-I-RBU-20160316. 315 28-I-YALO-20160322. 316 26-I-YALO-20160322. 317 26-I-YALO-20160322. 200

public positive equivalent of the dark negative fonoli* force, as this quote from Nestor, a notable* and teacher from Yalosuna illustrates:

“Fonoli, to maintain the fields, does not contribute [to the community], it is only individual. It’s an anti-development system. We want the development of our community. If someone just pursues his own development at the expense of others, it is thus an anti-development. He has an abundant labour force thanks to fonoli. That’s not good. We, to diminish this system, we pay young people to clear our forests and maintain our crop fields. […] If fonoli was visible, it would be better. Thus this business is clandestine. It’s a really bad force.”318

At the same time, farmers who employ young labour can also be accused of fonoli*. I had a long conversation about fonoli* with Victor from KP 13 on the ‘Buta road’ where this occult power, he asserted, is widespread –it is close to markets and thus the size of fields is higher than in more remote areas. When I asked him whether those who use fonoli* are not using paid labour, he answered: “they pay people from the associations so that they do not get exposed but they do not use them often”.319 Clearly, fonoli* is tied to the tensions generated by capitalist socio-ecological relations and the increasing social differentiation between farming households in the hinterland (see also Geschiere (1997) for such analysis). Fonoli* stories are in this context best understood as stories of anti-social accumulation and dispossession of land and human resources.

‘Environmental’ issues are thus widely interpreted in fundamentally social and moral terms (see also Ferguson (2006) on African ‘moralized’ economies). In these discourses, forest shortage and desertification driven by the intensification of production and land alienation are simultaneously pushed by socio-economic and moral transformations. Such changes are productive of new types of social relations, e.g. in the form of labour associations*. Alternatively, they also break mutuality as the non- regeneration of forests mirrors the non-regeneration of social relations in the kin group and across generations. In this way, local understandings of ecological change invariably include issues of social and environmental (in)justice.

318 28-I-YALO-20160322. 319 25-I-RBU-20160316. 201

202

Conclusions

The main idea that I have tried to articulate in this exploration of Kisangani’s hinterland is that landscapes are sites of provisional encounters between past and present, invisible and visible human and nonhuman agents that interact to produce various socio-spatial-temporal relations (Massey, 2005; Tsing, 2017). Such a conception takes materiality, (multispecies) sociality and multiplicity seriously, and allows us to think across a variety of time and space. I have argued that land tenure and land use practices are co-produced dynamically in relation with the socio-materiality of resources, the physical and built environment, and an array of coexisting socio-cultural and economic forces. It is the temporary expression of these practices and relations ‘so far’ that constitute the landscape. Kisangani’s hinterland is made by a history of movement and connection (Tsing, 2005).

In chapter 5, through the story-so-far of the contested Alibuku landscape, I showed how successive, ongoing waves of resource and boundary-(un)making practices shape the multiplicity of tenure and of identities. The rubble of violent colonial ethno-administrative and capitalist environmental rule, of postcolonial land grabs and the construction of Congolese authenticity, of wars and of commodity resource extraction all interplay with ‘traditional’ ways of belonging and forest appropriation. Tenure in Kisangani’s hinterland cannot be understood through simple (post-)colonial, static categories of ‘customary’ and ‘legal’ nor can identity be reduced to ‘indigeneity’, citizenship or ethnic (sub)group affiliation; categorizations that are reified by management plans for capitalist timber extraction. Rather, I have demonstrated that land tenure practices and identities are continuously contested and (re)negotiated in relation to changing circumstances such as population density, forest cover classes and availability, the type of land use, the meaning of the resource and the status of the user. Entanglements between human and nonhuman elements thus create the patches of this landscape (Rose, 2011; Tsing, 2005). In this way, fluid appropriation and looser tenure, and the use of natural and ruderal landmarks, are legion in more remote areas, for ‘traditional’ land use activities where primary forests are abundant and densities are lower. Bounded ethnic or clan identities, and entitlement to land as collective ‘primary occupants’320 are asserted more strongly in the face of higher levels of land or natural resource commodification and of competition. Yet, such group identity is also articulated in tension with national citizenship to claim individual property land rights – independent of the group – for instance in the face of higher competition for resources within the group itself. In both cases, I have argued, it is when resource uses are framed as permanent modern extraction that land and forests are then treated the most like private property. The apparent ontological security, the static aspect of the abstract space of maps that is associated, in local framings, with Western ‘modernity’ is then called upon to record a group’s permanent appropriation or

320 In many other contexts, the term ‘indigenous people’ would in fact be used but in DRC, the only people truly recognized as ‘indigenous’ are the Batwa and Baka pygmies. 203

the boundaries of a private concession. Such processes of (sometimes relative) privatization, I have shown, are growing in the hinterland, with urban elites continuously acquiring land for charcoal production, artisanal logging and agricultural activities. Tenure, identities and land uses are thus intertwined in often unexpected, complex articulations, shaped by histories (of disturbance), culture, and power (Li, 2000; Stoetzer, 2018).

The evolving multiplicity of land tenure matches that of agricultural practices and their associated socio- ecological changes that I explored in chapter 6. My aim, in this chapter, has been to show how forests and fields are part of the same continuum and can be read as social spaces (Tsing, 2005). I started with my journey from the city to the primary forests of Pumuzika village to emphasize the ways the wild and the cultivated, ‘the commercial’ and ‘the subsistence’, and the traditional and the modern all mingle like the crops in the field-garden of peasant farmers. I have demonstrated that, the local system of shifting agriculture is characterized by a four-way relationship between human cultivators, socio-economic relations across time and space, environmental conditions and the materiality and ecology of resources (forests, land and crops). The field, elanga*, of a rural farmer can in this way not be thought of in isolation from other larger processes. I have first shown that in ‘cutting out a field’, land tenure institutions (access to land), availability of forests and their products, road infrastructure and market access, are important considerations. The very act of ‘cutting out a field’ and of farming, I have argued, does not only allow people to sustain livelihoods but it also allows for social reproduction, as agricultural fallows and secondary forests are passed on from one generation to another. Moreover, farming in the primary forests allows people to engage in other livelihood activities such as hunting, fishing and producing a bit of charcoal. I have shown, however, that the extended spatiality characteristic of ‘traditional’ shifting cultivation has been highly impeded by colonial resettlements along roads and in paysannats indigènes*, and has produced very unequal outcomes for different groups and households. In areas closer to the city particularly, these changes along the continuous introduction of new short-cycle cash crops, and increasing population density, urbanization and integration into markets of goods, have produced new sets of socio-economic relations and ecological effects. In practice, different classes of farmers and two more or less integrated farming models are prevalent in the hinterland, mirroring the cultural syncretism that characterizes it. In this way, a ‘traditional’ (more or less) extensive subsistence farming model based on systems of collective, mutual labour on softly appropriated lands coexists in tension with a more intensive cash-crop agricultural model that requires high input of wage labour on more privatized lands, leading to what inhabitants qualify as ‘desertification’ and land alienation. I have insisted that in such processes, the crops and weeds are co-productive agents (Tsing, 2017). The fast pace of farming and regular weeding maintenance required by most short-cycle cash food crops – in comparison to the slow pace farming that defines the cultivation of more traditional crops such as manioc and bananas – implies that only better-off farmers are able to adapt, such as households within clans endowed with more land

204

and with (relatively) good market access, or urban elites. Land intensification and cash food crops allows this more ‘privileged’ class of farmers to further expand their lands, usually away from the village. Poorer and more marginalized households, I have shown, have often no other choice than renting small plots of land to more endowed families – thus contributing to the overall land use intensification – or encroach on forests further away from Kisangani, sometimes ‘escaping’ to campements* – where access to land is easier but access to markets more difficult.

I have stressed in this sense that the interrelations happening in one place and at a particular time generate effects in geographically distant places, and across time. From the Zanzibari-Arab, Portuguese and Belgian imperialisms to development-environmental programs, from growing urbanization in DRC and around the globe to the impacts of climate change in Asia that have all contributed to create new hybrid socio-ecologies, the shaping of the Kisangani hinterland’s landscape is a spatio-temporally dispersed process, always in becoming.

In essence, I argue in fact that such processes of socio-ecological change need to be simultaneously understood in fundamentally spatial and temporal terms. The socio-spatial-temporal nexus is impossible to disentangle (Massey, 2005). The cyclical worldview that characterizes the integrated shifting farming- tenure system is profoundly shaken by the accelerated linear permanency of intensified and privatized agricultural systems. In the former, the cyclical sequence from field to fallow to forest to field and the slow growing, slowly harvested crops eaten locally allows the simultaneous reproduction of the living and non-living kin and of the nonhuman world. The very act of farming, in the extended spatial-temporal timeframe of the cyclical system, is an act of connection, of nourishing social relations with the ancestors who lived where crops are now planted, with the extended family through reciprocal labour systems (ristournes*) and with next generations. Forests are recurring as much as social relations. In the system of land and labour commodification and intensification, on the other hand, the short growth of cash food crops is matched by the short fallow rotations and labour contracts based on short-term wage relations rather than on long-term kin relations. The crops themselves permanently leave the community to be sold on markets rather than consumed locally. The linear expansion of cultivated land, often accompanying intensification, under the control of a smaller number of people contrasts with the cyclical nature of the ‘purely’ shifting agriculture system. Farming then is not as reproductive of the group as cyclical processes of land rewilding and labour reciprocity over a long timeframe fade to the profit of permanent land rights and individualized wage labour. I have demonstrated how such spatial and temporal rationalization and the related idea of permanency – the resource is extracted and does not come back to the community – shape local understandings of ‘anti-social’ forest cover changes. Forests are disappearing simultaneously with community relations; relations are indeed always simultaneously ecological and social, and the landscape takes the forms of activities.

205

Changes in land use practices and their ensuing ecological changes thus touch deeply on identities and social relations. Within rural farmer communities of the hinterland, understandings of such changes are articulated in social and moral terms underpinned by Christianity and expectations of modernity and progress –formulated as ‘development’ (Ferguson, 2006; Jacka, 2015). Wage labour and the growing of cash food crops are seen as pathways to bring transformations in people’s material lives but also as (potential) breakers of mutual and reciprocal ties normally sustained by the cyclical nature of socio- ecological relations. The destructive forces of capitalism, such as the social and environmental impacts of resource use intensification, find fertile ground in local philosophies in the hinterland, with adapted versions of witchcraft stories (fonoli*), seeking new answers to this new crisis (Jacka, 2015; Vansina, 1990). As much as syncretism defines spiritual beliefs in the hinterland and in the whole DRC, contradictions and articulations are inherent to what it means to live in a hybrid peri-urban area where the rubble of imperialism, Christianization, different systems of values and virtue, global and local capital, extractive timber companies adapting their practices to REDD+ global environmental policy, heterogeneous forest dwellers and users, and multiple forest resources interact to ‘make’ and negotiate the landscape. Yet, without romanticizing sorcery in any way – its effects can be very destructive – I believe that the stories farmers told me also remind us that the environmental crisis is first and foremost a social crisis that arises from inequalities; that the exhaustion of the ‘natural’ is the exhaustion of social life.

206

References

Abdala, B., & Lokoka, R. (2009). Etude de cas sur l'exploitation artisanale de bois à Kisangani et ses environs. Retrieved from http://ocean-rdc.org/data/documents/2rap11.pdf

Abelman, J. (n.d.). The Dwelling Perspective. Retrieved from http://www.groundcondition.com/essays/the-dwelling-perspective/

AGEDUFOR. (2015). AGEDUFOR, vers une gestion durable des forêts de la RDC. Retrieved from http://www.oreade-breche.fr/documents/index.html

Alexandre, D. Y. (1989). L'arbre et le maintien des potentialités agricoles en zone intertropicale humide In M. Eldin & P. Milleville (Eds.), Le risque en agriculture (pp. 115-129). Paris: Orstom.

Alongo, S., Osombause, J., Bolakonga, B., & Lokinda, F. (2012). Etude de filières agropastorales et agricoles dans les villages cibles (Territoire d'Isangi, R.D. Congo). Yangambi: Institut Facultaire des Sciences Agronomiques.

Bakker, K., & Bridge, G. (2006). Material worlds? Resource geographies and the `matter of nature'. Progress in Human Geography, 30(1), 5-27. doi:10.1191/0309132506ph588oa

Banque Nationale de Belgique. (1948). Bulletin d'Information et de Documentation - L'évolution de la situation économique du Congo belge de 1939 à 1947. Bruxelles: Dewarichet.

Barou, J. (2000). Migrations et travaux forcés en Afrique subsaharienne à l'époque coloniale. Hommes et Migrations(1228), 51-61.

Bierschenk, T., & Olivier de Sardan, J.-P. (1997). ECRIS: Rapid Collective Inquiry for the Identification of Conflicts and Strategic Groups. Human Organization, 56(2), 238-244.

Bolakonga Ilye, A., Disonama, M., Michel, B., & Lebailly, P. (2012). Influence de l'enclavement sur l'économie rurale en République Démocratique du Congo : Cas du Territoire d'Opala. Retrieved from https://orbi.uliege.be/bitstream/2268/137818/1/Article%20Orl%C3%A9ans.pdf

Clement, P. (2014). Rural development in the Belgian Congo : the late-colonial "indigenous peasantry" programme and its implementation in the Equateur District (1950s). Bulletin Des Séances, 60(2), 251-286.

Comaroff, J. L., & Comaroff, J. (1997). Of Revelation and Revolution: The Dialectics of Modernity on a South African Frontier (Vol. 2). Chicago: University of Chicago Press.

COTREFOR. (2016). Plan d'aménagement de la concession Alibuku. Retrieved from http://www.cotrefor.com/fr/concession-alibuku/

De Clerck, L. (2004). L'organisation politique et administrative. In E. Lamy & L. De Clerck (Eds.), L'ordre juridique colonial belge en Afrique Centrale: Eléments d'histoire (pp. 121-195). Bruxelles: Académie Royale des Sciences d'Outre-Mer.

Delsupehe, G. (1944). Le Paysannat Indigène au Congo belge. (Licencié), Université Catholique de Louvain, Louvain.

Englebert, P., & Tull, D. (2013). Contestation, négociation et résistance : L'État congolais au quotidien. Politique Africaine, 129(1), 5-22. doi:doi:10.3917/polaf.129.0005

Fanon, F. (1968). The Wretched of the Earth. New York: Grove Press.

207

Ferguson, J. (2006). Global Shadows: Africa in the Neoliberal World Order. Durham & London: Duke University Press.

Fraser, N. (2000). Rethinking Recognition. New Left Review, 3, 107-120.

Geschiere, P. (1997). The Modernity of Witchcraft : Politics and the Occult in Postcolonial Africa. Charlottesville, VA: University of Virginia Press.

GFW Fires. (2020a). Distribution of Fire Alerts 1 DEC 2019 - 14 APR 2020. Esri Fire Report for Orientale (DR Congo). Online: Global Forest Watch Fires.

GFW Fires. (2020b). Distribution of Fire Alerts 15 MAY 2019 - 15 SEP 2019. Esri Fire Report for Orientale (DR Congo). Online: Global Forest Watch Fires.

Gibson-Graham, J. K. (2002). Beyond Global Vs. Local: Economic Politics Outside the Binary Frame. In Herod, A. & Wright, M. (Eds.), Geographies of Power: Placing Scale (pp.25-60). Oxford: Blackwell Publishers.

Gibson-Graham, J. K. (2014). Rethinking the Economy with Thick Description and Weak Theory. Current Anthropology, 55(S9), S147-S153. doi:10.1086/676646

Gordillo, G. R. (2014). Rubble: The Afterlife of Destruction. Durham, NC: Duke University Press.

Griffiths, J. (1986). What is Legal Pluralism? The Journal of Legal Pluralism and Unofficial Law, 18(24), 1- 55. doi:10.1080/07329113.1986.10756387

Gudeman, S. (2001). The Anthropology of Economy: Community, Market and Culture. Oxford: Wiley- Blackwell.

Gudeman, S. (2008). Economy's Tension: The Dialectics of Community and Market. Oxford, UK: Berghahn Books.

Haraway, D. (2003). The Companion Species Manifesto: Dogs, People, and Significant Otherness (Vol. 1). Chicago: Prickly Paradigm Press.

Hoffmann, K. (2019). Ethnogovernmentality: The making of ethnic territories and subjects in Eastern DR Congo. Geoforum. doi:https://doi.org/10.1016/j.geoforum.2019.10.002

Huggins, C. (2010). Land, Power and Identity: Roots of violent conflict in Eastern DRC. Retrieved from https://www.international-alert.org/publications/land-power-and-identity

Ibanez de Ibero, C. (1913). La mise en valeur du Congo belge : étude de géographie coloniale. Paris: Librairie de la Société du Receuil Sirey.

Igoe, J., & Brockington, D. (2007). Neoliberal Conservation: A Brief Introduction. Conservation and Society, 5(4), 432-449.

Ingold, T. (1993). The Temporality of the Landscape. World Archaeology, 25(2), 152-174.

Ingold, T. (2000). The Perception of the Environment: Essays on livelihood, dwelling and skill. New York: Routledge.

Ingold, T. (2011). Being Alive: Essays on Movement, Knowledge and Description. London: Routledge.

Jacka, J. K. (2015). Alchemy in the Rain Forest: Politics, Ecology, and Resilience in a New Guinea Mining Area. Durham, NC: Duke University Press.

208

Jewsiewicki, B. (1979). Le Colonat Agricole Européen au Congo-Belge, 1910-1960: Questions Politiques et Economiques. The Journal of African History, 20(4), 559-571.

Koy, J. K., Ngonga, A. M. M., & Wardell, D. A. (2019). Gestion de la Réserve de Biosphère de Yangambi en République Démocratique du Congo à l’épreuve des dynamiques d’installation des villages et campements (1939–2015). Anthropos, 114(2), 399-416. doi:10.5771/0257-9774-2019-2-399

Lau, J. D., & Scales, I. R. (2016). Identity, subjectivity and natural resource use: How ethnicity, gender and class intersect to influence mangrove oyster harvesting in The Gambia. Geoforum, 69, 136-146. doi:https://doi.org/10.1016/j.geoforum.2016.01.002

Laude, N. (Cartographer). (1938). Congo Belge. Principales productions minérales, végétales et animales. Voies de communication. 1937-1938. Retrieved from http://www.expocongo.be/content.php?m=6&r=3&sr=5&doc=159&l=fr

Leplae, E. (1932). La crise agricole coloniale et les phases du développement de l'agriculture dans le Congo central. Bruxelles: Institut Royal Colonial Belge.

Lescuyer, G., Cerutti, P. O., Tshimpanga, P., Biloko, F., Adebu-Abdala, B., Tsanga, R., . . . Essiane-Mendoula, E. (2014). Le marché domestique du sciage artisanal en République démocratique du Congo État des lieux, opportunités, défis. Retrieved from www.jstor.org/stable/resrep02230

Li, T. M. (2000). Articulating Indigenous Identity in Indonesia: Resource Politics and the Tribal Slot. Comparative Studies In Society and History, 42(1), 149-179. doi:10.1017/S0010417500002632

Liena, P. B., Maindo Mwanya, F., & Maindo, A. (2014). Acquisition de terres par les populations urbaines auprès des communautés rurales de Kisangani. Retrieved from http://www.tropenbos.org/publications/land+acquisition+by+urban+dwellers+from+rural+com munities+around+kisangani+(drc)

Likaka, O. (1994). Rural Protest: The Mbole against Belgian Rule, 1897-1959. The International Journal of African Historical Studies, 27(3), 589-617. doi:10.2307/220760

Likaka, O. (1997). Rural Society and Cotton in Colonial Zaire. Madison: University of Wisconsin Press.

Lund, C. (2006). Twilight Institutions: Public Authority and Local Politics in Africa. Development and Change, 37(4), 685-705. doi:10.1111/j.1467-7660.2006.00497.x

Lund, C., & Rachman, N. F. (2018). Indirect Recognition. Frontiers and Territorialization around Mount Halimun-Salak National Park, Indonesia. World Development, 101, 417-428. doi:https://doi.org/10.1016/j.worlddev.2017.04.003

Macgaffey, W. (2013). A note on Vansina’s invention of matrilinearity. The Journal of African History, 54(2), 269-280.

MADR-RDC. (2013). Stratégie Nationale de Développement de la Riziculture. Retrieved from https://riceforafrica.net/images/pdf/NRDS_drc_fr-min.pdf

Malengreau, G. (1949). Vers un paysannat indigène. Les lotissements agricoles au Congo belge. Bruxelles: Institut Royal Colonial Belge:

Massey, D. (1992). Politics and Space/Time. New Left Review, 196, 65-84.

Massey, D. (1993a). Power geometry and a progressive sense of place. In J. Bird, B. Curtis, T. Putnam, G. Robertson, & L. Tickner (Eds.), Mapping the futures (pp. 59-69). London: Routledge.

Massey, D. (1993b). Questions of Locality. Geography, 78(2), 142-149.

209

Massey, D. (1994). Space, Place, and Gender. Cambridge: Polity Press.

Massey, D. (2005). for space. London & New York: Routledge.

Mbembe, A. (2002). African Modes of Self-Writing. Public Culture, 14(1), 239-273.

Ministère des Colonies. (1910). Bulletin Agricole du Congo Belge. Bruxelles: Direction Générale de l'Agriculture.

Mission C.E.E. - Congo. (1963). Annexes Fascicule II: Secteur Primaire - Agriculture. Retrieved from http://aei.pitt.edu/66603/1/AGRICULTURE_SECTEUR_PRIMAIRE.pdf

Moeller, A. (1936). Les Grandes Lignes des migrations des Bantous de la province orientale du Congo belge. Bruxelles: Institut Royal Colonial Belge.

Molinario, G., Hansen, M. C., Potapov, P. V., Tyukavina, A., Stehman, S., Barker, B., & Humber, M. (2017). Quantification of land cover and land use within the rural complex of the Democratic Republic of Congo. Environmental Research Letters, 12(10), 104001. doi:10.1088/1748-9326/aa8680

Ndaywel è Nziem, I. (2012). Histoire du Congo: Des origines à nos jours. Bruxelles: Le Cri.

Newbury, D. (2007). Contradictions at the Heart of the Canon: Jan Vansina and the Debate over Oral Historiography in Africa, 1960–1985. History In Africa, 34, 213-254. doi:10.1017/S0361541300009797

Nixon, R. (2011). Slow Violence and the Environmentalism of the Poor. Cambridge, MA: Harvard University Press.

Peemans, J.-P. (2014). Land grabbing and development history: the Congolese (RDC) experience. In Ansoms, A. & Hilhorst, T. (Eds.), Losing your Land: Dispossession in the Great Lakes (pp. 11-35). Woodbridge, Suffolk: James Currey.

Plumwood, V. (1993). Feminism and the Mastery of Nature. London: Routledge.

Posman, S. (2012). Time as a Simple/Multiple Melody in Henri Bergson's "Duration and Simultaneity" and Gertrude Stein's Landscape Writing. Mosaic: An Interdisciplinary Critical Journal, 45(1), 105-120.

Pottier, J. (2005). ʻCustomary Land Tenureʼ in Sub-Saharan Africa Today: Meanings and contexts. In C. Huggins & J. Clover (Eds.), Land Rights, Conflict and Peace in Sub-Saharan Africa (pp. 55-75). Pretoria, South Africa: Institute for Security Studies.

Radio Okapi. (2013a). Isangi : la production du riz a augmenté de 25 % en 2013. Retrieved from https://www.radiookapi.net/economie/2013/10/14/isangi-la-production-du-riz-augmente-de- 25-en-2013

Radio Okapi. (2013b). Province Orientale: poursuite de la multiplication des semences améliorées de riz dans la Tshopo. Retrieved from https://www.radiookapi.net/actualite/2013/02/03/province- orientale-poursuite-de-la-multiplication-des-semences-ameliorees-de-riz-dans-la-tshopo

Redfern, S. K., Azzu, N., & Binamira, J. S. (2012). Rice in Southeast Asia: facing risks and vulnerabilities to respond to climate change. Retrieved from http://www.fao.org/3/i3084e/i3084e18.pdf.

210

Rose, D. B. (2011). Wild Dog Dreaming: Love and Extinction. Charlottesville: University of Virginia Press.

Rose, D. B. (2017). Connectivity Thinking, Animism, and the Pursuit of Liveliness. Educational Theory, 67(4), 491-508. doi:10.1111/edth.12260

Rösler, M. (1997). Shifting cultivation in the Ituri Forest [Haut-Zaïre]. Civilisations, 44 | 1997, 44-61. doi:https://doi.org/10.4000/civilisations.1608

Ross, C. (2017). Ecology and Power in the Age of Empire: Europe and the Transformation of the Tropical World. Oxford: Oxford University Press.

Roth, R. (2009). The challenges of mapping complex indigenous spatiality: from abstract space to dwelling space. cultural geographies, 16(2), 207-227. doi:10.1177/1474474008101517

Schatzki, T. R. (2010). The Timespace of Human Activity: On Performance, Society, and History as Indeterminate Teleological Events. Plymouth: Lexington Books.

Scott, J. C. (1992). Domination and the Arts of Resistance: Hidden Transcripts. New Haven & London: Yale University Press.

Stoetzer, B. (2018). RUDERAL ECOLOGIES: Rethinking Nature, Migration, and the Urban Landscape in Berlin. Cultural Anthropology, 33(2), 295-323. doi:10.14506/ca33.2.09

Stoler, A. L. (2013). ‘The Rot Remains’: From Ruins to Ruination. In Stoler, A.L. (Ed.), Imperial Debris. On Ruins and Ruination (pp. 1-35). Durham, NC: Duke University Press.

Titeca, K., & De Herdt, T. (2011). Real Governance Beyond the “Failed State” : Negotiating the Education Sector in the Democratic Republic of Congo (DRC). African Affairs, 110(439), 213-231.

Trefon, T. (2011). Urban-Rural Straddling Conceptualizing the Peri-urban in Central Africa. Journal of Developing Societies, 27(3-4), 421-443. doi:10.1177/0169796X1102700408

Tsing, A. L. (2005). Friction: An Ethnography of Global Connection. Princeton & Oxford: Princeton University Press.

Tsing, A. L. (2015). The Mushroom at the End of the World: On the Possibility of Life in Capitalist Ruins. Princeton, NJ: Princeton University Press.

Tsing, A. L. (2017). The Buck, the Bull, and the Dream of the Stag: Some unexpected weeds of the Anthropocene. Suomen Antropologi: Journal of the Finnish Anthropological Society, 42(1).

Turner, T. (2007). The Congo Wars: Conflict, Myth and Reality. New York: Zed Books.

UCL-Géomatique, IGC, RGC, & AfricaMuseum (Cartographer). (2011). Degré Carré de Kisangani. Planche N.0/25

UCL-Géomatique, IGC, RGC, & AfricaMuseum (Cartographer). (2014). Degré Carré de Yangambi. Planche N.0/24

Van Reybrouck, D. (2012). Congo, une histoire. Paris: Actes Sud.

211

Vanderlinden, J., & Stengers, J. (1994). Pierre Ryckmans 1891-1959 : coloniser dans l'honneur. Bruxelles: De Boeck Université.

Vansina, J. (1990). Paths in the Rainforests. Toward a History of Political Tradition in Equatorial A. Madison: The University of Wisconsin Press.

Verweijen, J., & Van Bockhaven, V. (2020). Revisiting colonial legacies in knowledge production on customary authority in Central and East Africa. Journal of Eastern African Studies, 1-23. doi:10.1080/17531055.2019.1710366

Verweijen, J., & Vlassenroot, K. (2015). Armed mobilisation and the nexus of territory, identity, and authority: the contested territorial aspirations of the Banyamulenge in eastern DR Congo. Journal of Contemporary African Studies, 33(2), 191-212. doi:10.1080/02589001.2015.1066080

Vinez, M. (2017). Division of the Commons and Access to Land on The Frontier : Lessons from The Colonial Legacy in The Democratic Republic of Congo. Online. Retrieved from http://pubdocs.worldbank.org/en/748201495654708183/B2-history-abca.pdf

Vlassenroot, K., & Delaleeuwe, N. (2008). Négocier et contester l'ordre public dans l'Est de la République démocratique du Congo. Politique Africaine, 111(3), 44-68. doi:10.3917/polaf.111.0044

Vlassenroot, K., & Huggins, C. (2005). Land, Migration and Conflict in Eastern DRC. In Huggins, C. & Clover, J. (Eds.), From the ground up : land rights, conflict and peace in Sub-Saharan Africa (pp. 115-194). Pretoria, South Africa: Institute for Security Studies. von Benda-Beckmannn, F., & von Benda-Beckmannn, K. (2006). The Dynamics of Change and Continuity in Plural Legal Orders. The Journal of Legal Pluralism and Unofficial Law, 38(53-54), 1-44. doi:10.1080/07329113.2006.10756597

Yetilo, J. K. (2010). La sous-administration territoriale en République démocratique du Congo. Etat des lieux et perspectives. Pyramides(19 | 2010), 105-128.

212

GENERAL CONCLUSION: FROM GEOCODED TO ENTANGLED LANDSCAPE

In May 2019, a few days before the end of my last fieldwork, I had one of the most memorable and unsettling conversations of my journey in Kisangani’s hinterland, with the director of the INERA herbarium in Yangambi, Elasi. The small town of Yangambi, which I have mentioned before, is situated about ninety kilometres West of Kisangani, on the North bank of the Congo river. It is a fascinating place for anyone interested in precarious life among colonial capitalist ruins where stories of decay and of renewal are interwoven (Tsing, 2015). It used to be a wealthy colonial cité* with opulent houses, a cinema, a private club with its tennis courts and swimming pools, and was home to the National Institute for Agronomic Study in the Belgian Congo (INEAC), one of the world’s most important tropical agriculture and forest research centres at the time. It was the ‘quintessence’ of colonial scientific capitalism: research served the productivity of export plantations, ranging from large-scale industrial and family plantations owned by European settlers to Congolese smallholder farms organized in paysannats indigènes*. Next to the research station, the 225 000-hectare Yangambi Biosphere Reserve was created in 1939 with conservation and research of biodiversity as the main objectives. Today, for any first time-visitor, arriving in Yangambi might appear as a trip back in time. Its decrepit colonial buildings and furniture, its factories at standstill, its former oil palm, cocoa, coffee, or rubber tree plantations have been progressively and spontaneously taken over by heterogeneous plants, trees and weeds, and at a first sight seem unexploited. I have, however, shown throughout this thesis that such ‘ruins’ are home to multiple forms of lives. The town has been described as a “mythic place” (UE en RDC & Nasi, 2019) that is now “chaos” or a “apocalyptic vision” (Choukran, 2019). On its Twitter account, the Center for International Forestry Research (CIFOR) claims to be “bringing the community and the research center back to life” (CIFOR, 2019a) through its EU-funded REDD+-like €27-million programme, which advertises environmental awareness programmes, the renovation of a wood biology laboratory and herbarium, the plantation of 300 000 fast-growing trees for “biomass production” to increase carbon storage and respond to local energy needs so as to preserve the Biosphere Reserve, and the creation of 600 direct employment through these plantations (CIFOR, 2019b, 2020). The programme also aims at opening a museum that gathers the golden age of Yangambi, the responsible for environmental awareness raising told me.

Elasi, who is about 65 years old and whose herbarium he is managing was renovated through the CIFOR/EU programme, was sharing his views with me about this new attempt at conservation and modernization, what it means to live in Yangambi and spirituality. It was a long conversation more than an interview. He told me how people in Yangambi “have regressed” since independence, “they have taken the forest back for their fields”.321 The lack of maintenance of Yangambi’s research and agriculture

321 81-CTSP-YAN-20190530. 213

infrastructures deeply upsets him. He said: “here I have everything. All the forest areas that I need; but we do not have any scientific history. We are a country that comes out of the forest.” The discourse he share with me was one about the total ‘failure’ of Congolese people to continue what the Belgians had started. “We have a lot of money in Yangambi, we have 500 hectares of oil palm, 500 hectares of cocoa and so on; we have all the best products. But all that we do is picking”, he said. In his opinion, the CIFOR/EU programme would not be sustainable after its end. “What you do not want to accept, you Europeans, is that there is a lot of darkness in our minds. It is this darkness that causes one to have a 500- hectare plantation and not exploit it”. Older people, the ones who learnt with the Belgians to plant in lines, “just stopped doing it like that”. For Elasi, the darkness that Congolese people are holding to is responsible for the destruction of all the good things brought by development programmes. “Europeans, you think that you can skip the phase of spiritual change and that is why after you leave, everything gets destroyed”, he concluded.

As I mentioned earlier in this thesis, I have heard Elasi’s colonial Manichean discourse (Fanon, 1968) many times and in many places in Kisangani and its rural hinterland, though usually not always in such extreme wording. In this postcolonial and post-conflict context where industries and infrastructures are quasi absent, that the end of rational colonial management of soil and forest resources has marked the end of progress, that traditional beliefs and ways of being and doing are ‘anti-development’, are common narratives reinforced by the multiple developmental-environmental discourses that peasants (in particular) are on the receiving end of. At the same time, life in the rubble of colonial capitalism has continued and Kisangani hinterland’s inhabitants have developed hybrid and resilient subsistence and market livelihood practices in interdependence with their physical environment, that defy simple categorizations of environmental planners and agronomists. Similarly, social relations and identities have dynamically evolved with land and forest resource use and change. Complex and interdependent trends of socio-environmental transformation such as increased demographic pressure and urbanization, intensified levels of land concentration, the growing importance of cash food crops, the massive arrival of cheap Chinese material goods, and a (still relative) agricultural and resource use intensification redefine peri-urban ruralities. Rather than a static, stuck-in-the-past society, my aim has been to show its permanent evolution and adaptation to socio-ecological change, as much as its (re)production through multiple practices and social relations through and across space and time (Ingold, 2000; Massey, 2005).

Revisiting forests-people relations

In the preceding chapters, I looked at some of the frictions between the ways DRC’s forested landscapes and their changes are constituted by REDD+ practices and discourses, and in the lived landscapes of everyday practices. In doing so, I have emphasized two ways of conceptualizing a landscape that enact different socioecological realities, narratives of deforestation, environmental imaginaries and possible

214

responses: as a representational, geocoded landscape (Part I); and as a more-than-representational, entangled landscape (Part II).

I empirically showed the persistence of colonial representations of small-scale peasantry and attempts at agrarian transition through privatization, intensification and monoculture in the present DRC’s REDD+ narratives and in everyday landscapes. Apolitical and ahistorical narratives on deforestation drivers fail to recognize the slow socioenvironmental violence of such attempts and grossly mischaracterize peasant agriculture as a monolithic, static bloc. I have relied on empirical evidence to help identify how and why such narratives emerge. Cartographic representations based on geospatial satellite data, I have argued, have been key in problematizing deforestation, but only as part of a larger socio-political and epistemic agencement from which local forest dwellers are excluded. It is through processes of negotiation and contestation between the multiple actors involved in DRC’s REDD+ strategy-making that maps of Congolese forests acquire their meaning and have effects in the world. As such, the geocoding of space is part of the larger practices of ‘rendering technical’ and simultaneous depoliticization that Li (2007) or Ferguson (1994) have analysed in Indonesia and Lesotho respectively.

While REDD+ maps appear as “natural arbiters” for defining landscapes, in practice they reflect an “already settled dispute about the nature of […] landscape[s]” (Robbins, 2001, p. 175). The satellite and conservationists’ gazes focus on a complete mosaic of trees (including industrial tree plantations) and easily quantifiable carbon sinks, leaving little room for the fluidity, diversity and cyclical nature of forest patches in shifting agricultural systems. I have shown that, as a result, landscapes are produced as grids of bounded units of land use and tenure groups. The geocoded landscape demarcates fixed spheres of humans and nonhumans, production and conservation. It privileges a regular, linear socio-spatial- temporality that assumes that a “good livelihood requires permanent fields [and titles], just as a healthy forest requires permanent reserves” (Tsing, 2005, p. 175) and just as a good forest guardian requires the assignation of a permanent, delineated indigenous identity effecting the “ethnic spatial fix” (Moore in Li, 2010, p. 391). Geospatial governance thus consolidates particular conceptions of nature-society relations.

I have stressed that today’s forests and the geocoded landscape bears the imprint of colonial boundary- making processes that are rendered invisible by purely geospatial satellite analysis and that yet need to be taken into account to understand today’s socio-ecological relations and forest cover change. One of the essential points that I have sought to make is that present understandings of DRC’s forests fail to recognize that what they see as encroached primeval forests were always social spaces until the Belgian administration denied their long-term occupation through multiple livelihood activities. The resettlement of people along roads for a semi-sedentarization of cash crop agriculture, away from protected or productive areas, created some of the conditions of today’s deforestation and forest degradation

215

dynamics. More particularly, I have demonstrated three important interrelated dynamics. First, the loss of large tracks of land and the grouping of scattered villages, coupled with population increase in areas closer to the city prevents forest regrowth – to a certain extent – in what GIS experts define as the ‘rural complex’. Second, the unequal distribution of land between groups – which has an impact on their descendants – forces some ‘landless’ peasants to find arable land further away in the forest outside of the rural complex, i.e. a phenomenon that is interpreted by the satellite gaze as forest fragmentation. In some cases, ‘landless’ peasants are going back to their ancestral lands that were deemed ‘vacant’ in the colonial era. In others, they simply follow logging roads and skid trails. These processes are reinforced by land grabs through agricultural intensification and expansion by some wealthier households. Third, the cash and subsistence crop rotation systems imposed through cultures obligatoires* and paysannats indigènes* schemes have indeed introduced intensification. In areas with a short distance to markets, closer to the city where money circulates, and where some groups have alienated larger plots of land, intensification often comes with expansion as the former increases return to land and incentivize expansion of crop land. I have shown the human-nonhuman entanglements between land, labour practices (and markets), crops and weeds in driving these processes.

Ironically, while REDD+ strategies promote such intensification to reduce forest degradation and deforestation, in local understandings of socio-ecological change, it is precisely the combined effect of intensification and land alienation that drives what local people refer to as ‘desertification’, i.e. the non- renewal of forest that the extensive shifting agriculture system allowed. The co-dependency between forests, farming practices and social reproduction through land is apparent in these local understandings. Rather than a geo-coded worldview that separates forests from arable land, it is the very practice of cutting a field out of the forest for farming and the forest natural regeneration processes that allow the social reproduction of households and groups over generations. Here again, the nonhuman plas an active role in the fabric of social life. Deforestation-desertification is in this way understood mainly in social terms rather than in environmental terms only. My findings also help to understand the encroachment of forests outside the rural complex that are not due to an ‘irrational’ expansion of shifting agriculture but rather complex dynamics of intensification, land alienation, construction of new road infrastructure and ancestral relations to forests.

This leads me to a second important argument I have made about the shortcomings of the REDD+ geocoded landscapes, i.e. the reification of simplistic understandings of homogeneous, ‘local’ communities that misrecognizes the full complexity of social identities, and dismisses crucial power differences between actors and between different access to resources in the landscape. In this peri-urban space, large plots of land legally owned by urban elites coexist with many different types of land access by ‘autochthones’ and ‘newcomers’. The REDD+ geocoded landscape promotes a double discourse on

216

land rights. On the one hand, the ones who are portrayed as ‘traditional’ shifting cultivators – i.e. without land titles and not growing perennial cash crops– would in theory be able to get recognition only if they conform to a prescriptive model of landholding that is collective and inalienable (Li, 2010). This is not yet happening in practice (see chapter 3). I have shown, with the example of a logging concession’s forest management plan, i.e. an ‘integrated’ landscape on a smaller scale, supported by REDD+ programmes, that the promotion of imaginary homogeneous primary occupant communities allows some groups to claim rights and compensation while excluding others. In such claims, the geo-coded understanding of landscape (e.g. with colonial maps) is strategically mobilized. In this way, as Astuti and McGregor (2017) have argued, forms of land grabs and land claims can align and overlap.

Moreover, as much as categories of land and forests cannot be strictly bounded on a map, individual and collective appropriation coexist in a complementary fashion and both are contingent on the types of resource use and larger social relations. Such more individualized land rights are of a different kind than the legal property rights mostly held by urban elites and promoted by REDD+ strategies. Within communities, my findings show that many seek to sell, rent or buy land – even if such processes are a source of social conflicts– while maintaining a collective land tenure, e.g. for hunting and gathering, and identity. As Li (2010, p. 399) aptly summarizes, “[t]his cannot be seen simply as a matter of choice. When there is a choice, it is shaped by a desire and expectation that capitalist [I would rather use ‘commercial’] agriculture will produce good returns, money to educate children, and other benefits.” However, renting out and selling land are for many rural people a matter of economic survival rather than capitalist accumulation. On the other hand, the ‘productive agricultural entrepreneurs’ who have started to develop perennial cash crops plantations (cocoa) can be granted individual land rights. As the capacity to grow perennial cash crops is very much dependent on economic capital, such a binary view of collective versus individual landholding risks further reinforcing class formation and inequalities.

Revisiting socio-economic subjectivities

The binaries between ‘darkness’ and ‘light’ in Elasi’s discourse thus echoes the many binaries between subsistence and commercial, shifting and intensive, collective and individual, nature and culture, forests and fields that guide REDD+ environmental rule for the ‘protection’ and ‘improvement’ of forests’ and people’s lives (McElwee, 2016). My argument has been that the portrayal of shifting cultivators as impoverished, economically unproductive and harmful to the global forest commons, justifies accumulation by more powerful actors in the landscape, such as industrial logging companies, private conservation organizations but also ‘more local’ actors – generally elites with close link to the city – who have formal property rights on their lands. All are deemed more efficient and responsible economic and environmental caretakers. This process of land grabbing is far from new but is rather a continuation of colonial and postcolonial policies. The Kisangani hinterland has not yet suffered the worst of large-scale

217

land grabbing compared to other regions of DRC (e.g. the Kivus) or many other parts of the world (Peemans, 2014). However, as I have shown, it is a growing trend as city dwellers are increasingly acquiring agricultural space in the hinterland and as development-environmental interventions, such as REDD+, continue to push for various forms of land privatization. This systematic deconstruction of the peasant economy and identity also precipitates a particular kind of subjectivity – well underway during the colonial period – of a private agricultural entrepreneur growing green perennial cash crops – cocoa is the crop pushed by REDD+ programmes in the hinterland– and replanting trees on his lands. The expected result in the end might be fewer people with their own land as private landowners are expanding their cultivated land but more jobs and income for the new ‘landless’, and a higher productivity for export cash crops. I have shown how such a process develops (see also West (2012)).

In the introduction to this thesis I referred to the mass diffusion of images of forests on fire. I mentioned how such images become autonomous in that they are spread without any sophisticated arguments and different situations get networked together, imposing unity where there was none. On a global scale, the harmful African ‘slash-and-burn’ peasants – often opposed to the primitive Pygmy forest guardians – become equated with the Amazon’s commodity-driven deforestation. Both are equally seen as invading pristine forests, without paying attention to what kind of deforestation is happening (e.g. permanent or temporary). For the fomer it has always been their very habitats, and nothing is said about how they have historically conserved the forest through the very nature of their socio-ecological relations. This is where certified commodities, such as carbon credits, timber or smallholder produced cocoa – yet transformed by giant commodity groups as in the case of OLAM/ESCO-Kivu – come into play. West (2012), in her multi- sited study of certified coffee produced in Papua New Guinea, convincingly argues that the story brought by marketers to the wealthier ‘conscious buyers’ of Europe and North America bundles images of poverty, fair labour and good environmental standards as scarce. By doing so, it automatically adds value to the coffee which can be sold at a higher price. At the same time, she says, since coffee producers are always portrayed as poor and primitive in fair-marketing campaigns, consumers who are meant to feel connected with the “economy have-nots of the world” have the impression that any money the producers make compared to “bad exploitative capitalism” is a great improvement; without actually knowing to what extent these narratives depart from the real lives of coffee producers (West, 2012, p. 247).

In the same way, a triple movement is operating as Congolese peasants are portrayed as poor people who have no other choice than burning forests down for their meagre living – “it is the cheapest way to clear land” says BBC News (2019) in its article about last summer blazes – and as their activities are grossly compared with large-scale farming in other places of the world. First, it becomes acceptable to buy conservation in the form of carbon credits or certified cocoa at a low price to the poor (Karsenty, 2007).

218

As these kinds of images circulate with ideas of pristine forests, it also becomes acceptable to evict them from their lands. I have insisted many times that forested landscapes are simultaneously social and natural. Second, it creates two classes of peasants: those who have the means to engage in sustainable perennial cash crop production and those who do not. I have highlighted such unequal capacities in the Kisangani hinterland, which are related to food and income security, and high labour input it demands. Such considerations actually largely shape a household’s choice of any particular crop (not only cocoa). Those who are not able to engage in such production thus “get written off as underdeserving of political attention” (West, 2012, p. 250). One of my arguments has been that the unified discourse of the impoverished shifting cultivators precisely obscures diverse dynamics and inequalities among this seemingly homogeneous population, and simply disregards the causes of structural inequality in the peri- urban landscapes in which REDD+ intervenes. In light of my fieldwork in the hinterland, it also dismisses the widely spread mutual forms of labour-savings – based on a relative equality between their members – that allow many rural dwellers to ensure sufficient production and to deal with uncertainties. These different types of erasure lead to uneven development. Third, large private companies engaged in REDD+ and certification processes in DRC also gain market shares through responsible commodities and are seen as helping the former jobless poor peasants through the provision of cheap wage labour. Here again, attention is taken away from the fact that such large concessions are granted at the central level without taking into account the specific values of these lands for their inhabitants, and that in many cases their presence strongly increase social conflict – as in the case of the Alibuku landscape in the COTREFOR/IFCO concession – often leading to very unequal outcomes.

My argument is not that such subjectivities of the wage labourer or individual entrepreneurial smallholder do not exist in the local peri-urban arena. Rather, it is that REDD+ strategies reify them as the unique subjectivities and modes of being, i.e. desired pathways to development. In the intertwined CIFOR/EU REDD+-like programme’s and Elasi’s discourses with which I started this conclusion, the idea of “bringing communities back to life” is linked with the establishment of tree plantations that create employment. Yet, there are diverse, coexisting and enmeshed forms of socio-economic relations at the ‘local’ level that have evolved with changing socio-ecological conditions. Purely mutual, reciprocal forms of production-savings solidify community bonds and equality, enable them to collectively face uncertainties and risks while ensuring household’s food sovereignty. They coexist with associations*, which are at the intersection of mutuality and individuality and have appeared more recently, especially in areas closer to Kisangani where smallholder commercial and intensified agriculture is more present and where many young people are facing land insecurity and are thus in need of wage labour. Then, purely privatized labour has appeared with the increasing sizes of fields of labour-intensive cash food crops or perennial crops. These different types of socio-economic relations are both a continuity of

219

century-old collective ways of managing land and forests, and adaptations to the increased monetization of the economy and access to material goods that are typical of peri-urban areas.

From geocoded to entangled landscape: towards more just socio-environmental politics

Deforestation is a high-stakes, highly politicized issue in times of global climate change, as it relates to people’s everyday lives and livelihoods. It is simultaneously a reality, an agenda, a problem, a context, a narrative and a discourse (Leyshon, Geoghegan, & Harvey-Scholes, 2018). Throughout this thesis, I have articulated a political relational onto-epistemology to show the complex interactions between epistemic, material dispossession and uneven development while identifying, analyzing, and deliberating some aspects of socio-economic and ecological change in relation to various human-nonhuman entanglements across space and time. My aim has not been to deny the real forest cover change that is happening in the hinterland of Kisangani but rather offering more complex place-based explanations on the making of forested landscapes and their changes, and to link them to socio-environmental justice issues. The relationship between deforestation/degradation and endogenous population growth and ensuing land scarcity is certainly not linear and cannot be simply understood in neo-Malthusian terms. Degradation is relative rather than absolute, in that it is an issue of local scarcity constituted socially and historically through land tenure decisions and various land use practices ranging from land grabbing by some groups and resource use intensification in specific places, to past and recent resettlement, urbanization, development and road infrastructure building. In this sense, the analysis I have developed in this thesis should not be seen as a ‘complete’ causal explanation of socioforest change nor as collection of empirical fragments, but rather as “an effort to germinate connections and openings” (Whatmore, 2002, p. 6) that complicate simplistic understandings of ecological change.

We need to move beyond abstract notions of forests and beyond rigid stereotypes of a homogenous, holistic community of ‘shifting cultivators’ with archaic attributes who are either collectively destroying the environment and have flawed relationships with both nature and the market or, in contrast, who are managing resources only with considerations of social and intrinsic values (Agrawal & Gibson, 1999; Karsenty, 2008; Leach, Mearns, & Scoones, 1999). Farming in the hinterland of Kisangani has multiple affective, cultural, social and economic aspects that remain hidden in most discourse on DRC peasantry (Peemans, 2014). Recognizing complexity and diversity, revealing various community aspirations, worldviews and forms of human-nonhuman relations can lead to more socially and environmentally just outcomes (Asiyanbi, Ogar, & Akintoye, 2019). As Gibson-Graham (2008, p. 624) states:

“The technique of reading for difference has a number of effects. It produces recognition of the always already diverse economic landscape in all geographical regions. It clarifies the choices we have in the policy realm to

220

support and proliferate diversity, to destroy or allow it to deteriorate, or indeed to promote uniformity.”

REDD+ has adopted a ‘new’ landscape paradigm that reestablishes the logics of rationalizing and homogenizing identities, human labour and ecological complexity typical of the plantation model, that engenders ‘slow’ social and environmental violence (Nixon, 2011). Yet, landscape, I believe, is a good concept to start from to advance towards more progressive, just and sustainable REDD+ socio- environmental policies. It requires, however, a radical move away from representational understandings of landscape as a “passive backdrop for human activity” (Tsing in Aisher & Damodaran, 2016, p. 296) and planning. This implies a shift from human-in-nature to human-with-nature perspectives. Landscape’s retheorization for such an ethical agenda needs to unsettle the entrenched binaries of modernity between nature and culture; object and subject; and to be understood through a more-than- representational perspective that pays attention to its visible and non-visible dimensions. Thinking of landscape relationally has several implications. First, looking at the role of nonhuman actants (e.g. forests, crops, maps) in shaping social life offers pathways to enact certain socio-spatial and temporal relations over others. For instance in my case study, short-cycle crops like rice – in comparison with manioc – engenders an acceleration of production time, more intensive labor and land use intensification – or expansion if fertility decreases – that shape the landscape in particular ways. Thinking through these natureculture entanglements can help us to inform us about different socio-ecological pathways we can adopt.

Second, thinking of the landscape through a relational and more-than-representational approach allows different ways of knowing, through practices, bodies and affects to play a more legitimate part in decisions about landscapes. Such a destabilization of representational regimes restitutes epistemic justice and allows dwellers concerned by REDD+ environmental policies to articulate their own struggles (e.g. regarding property rights) without having to “adopt dominant ways of knowing nature and living with nature in order to improve distributional equity and fairness” (Vermeylen, 2019, p.89). It thus opens ways to think outside REDD+ prescriptive approach to justice towards multiple dimensions of justice that critical scholars working on REDD+ are increasingly urging policy-makers to adopt (e.g. Satyal, Corbera, Dawson, Dhungana, & Maskey, 2020; Suiseeya, 2016).

Looking back at geospatial knowledge, a more-than-representational epistemological approach to landscape encourages maps’ re-socialisation, i.e. accounting for their performativity and their partiality, as well as their provisional and emerging character (Windey & Van Hecken, 2019). It is about imagining other ways of approaching mapping that pay much more attention to how and why decisions are made, to the diverse context-dependent ways maps can emerge, and how maps and the world are always co- constituted. This type of mapping rather than enframing is much more open to negotiation and debate,

221

and to performative interaction with more elusive dimensions of existence. More importantly – as it is impossible to disentangle the world from procedures to report on it, as the power of maps is relational rather than inevitable – such a cartography can bring in socio-environmental justice concerns. It indeed opens up to enacting what was rendered invisible or marginalised by dominant representations of the world, and to a willingness to account for what ‘surprises’ modernist science such as different socio- spatial identities and ways of inhabiting the world. As Roth (2009, p.223) the digital environment allows us to increasingly move beyond the abstract and “restructure the data into myriad of maps-as- contingent-products” showing seasonality, intensity of activity, and socially differentiated patterns.

As I have said before, and as I have argued with others elsewhere (Shapiro-Garza et al., accepted for publication), the reading for complexity and difference, yet needs to be embedded with questions of global inequality and its consequences (Ferguson, 2006). The narratives I encountered, such as the latest one from Elasi, seemed to express a real sense of exclusion from the promised modernity, perceived as an ideal to attain, and an appeal to membership of global society (Ferguson 2006) even if they seem to contradict postcolonial critiques of (eco)modernization and alternative worlds. By using stories as a writing strategy, I have tried in this thesis to highlight such tensions and to provide a better account of hybrid uneven geographies and multidimensional peasant struggles. Yet, I found it difficult sometimes not to fall into universalizing ambitions and binaries as emphasizing the ‘totality’ of racial capitalism or imperialism can also help to articulate struggles for a more socially and ecologically just world. It is, however, “at the places where different simplifications meet [that] complexity is created” (Law & Mol 2002, p.11) and that we can start enacting other possible, plausible and desirable worlds.

222

References

Aisher, A., & Damodaran, V. (2016). Introduction: Human-nature Interactions through a Multispecies Lens. Conservation and Society, 14(4), 293-304. doi:10.4103/0972-4923.197612

Agrawal, A., & Gibson, C. C. (1999). Enchantment and Disenchantment: The Role of Community in Natural Resource Conservation. World Development, 27(4), 629-649. doi:http://dx.doi.org/10.1016/S0305-750X(98)00161-2

Astuti, R., & McGregor, A. (2017). Indigenous land claims or green grabs? Inclusions and exclusions within forest carbon politics in Indonesia. The Journal of Peasant Studies, 44(2), 445-466. doi:10.1080/03066150.2016.1197908

BBC News. (2019, 26 August). Amazon fires: Angola and DR Congo 'have more blazes'. BBC News. Retrieved from https://www.bbc.com/news/world-africa-49471644

Choukran, L. (2019, December 16). Yangambi n'est pas mort! Politico. Retrieved from https://www.politico.cd/la-rdc-a-la-une/2019/12/16/yangambi-nest-pas-mort.html/50789/

CIFOR. (2019a, December 6). “We are pleased to see how these plantations are bringing the community and the research center back to life..." A new initiative in #DRC promotes forest landscape restoration for bioenergy production | http://ow.ly/k0Nq50xqp9c #AFR100 [Tweet]. Retrieved from https://twitter.com/CIFOR/status/1202777239592538118?s=20

CIFOR. (2019b, 10 December). Yangambi fait peau neuve grâce au projet FORETS. Retrieved from https://www2.cifor.org/forets/corporate-news/yangambi-fait-peau-neuve-grace-au-projet- forets/

CIFOR (Producer). (2020). Yangambi pôle de Développement. [Video file] Retrieved from https://www2.cifor.org/forets/en/video/yangambi-pole-de-developpement/

Fanon, F. (1968). The Wretched of the Earth. New York: Grove Press.

Ferguson, J. (1994). The Anti-Politics Machine: development, depolitization and bureaucratic power in Lesotho. Minneapolis: University of Minnesota Press.

Ferguson, J. (2006). Global Shadows: Africa in the Neoliberal World Order. Durham and London: Duke University Press.

Gibson-Graham, J. K. (2008). Diverse economies: performative practices for `other worlds'. Progress in Human Geography, 32(5), 613-632.

Ingold, T. (2000). The Perception of the Environment: Essays on livelihood, dwelling and skill. New York: Routledge.

Karsenty, A. (2007). Questioning rent for development swaps: new marketbased instruments for biodiversity acquisition and the land-use issue in tropical countries. International Forestry Review, 9(1), 503-513.

223

Karsenty, A. (2008). Des « communautés locales » problématiques. In Méral, P., Castellanet, C. and Lapeyre, R. (Eds.), La gestion concertée des ressources naturelles. L’épreuve du temps (pp. 277- 288). Paris: Gret-Karthala.

Law, J., & Mol, A. (2002). Complexities: An Introduction. In J. Law & A. Mol (Eds.), Complexities: Social Studies of Knowledge Practices (pp. 1-22). Durham, NC: Duke University Press.

Leach, M., Mearns, R., & Scoones, I. (1999). Environmental Entitlements: Dynamics and Institutions in Community-Based Natural Resource Management. World Development, 27(2), 225-247. doi:http://dx.doi.org/10.1016/S0305-750X(98)00141-7

Leyshon, C., Geoghegan, H., & Harvey-Scholes, C. (2018). Landscape and climate change. In Howard, P., Thompson, I., Waterton, E. and Atha, M. (Eds.), The Routledge Companion to Landscape Studies (pp. 453-463). London, UK: Routledge.

Li, T. M. (2007). The Will to Improve. Governmentality, Development, and the Practice of Politics. London: Duke University Press.

Li, T. M. (2010). Indigeneity, Capitalism, and the Management of Dispossession. Current Anthropology, 51(3), 385-414. doi:10.1086/651942

Massey, D. (2005). for space. London & New York: Routledge.

McElwee, P. (2016). Forests Are Gold: Trees, People, and Environmental Rule in Vietnam. Seattle: University of Washington Press.

Peemans, J.-P. (2014). Land grabbing and development history: the Congolese (RDC) experience. In Ansoms, A. and Hilhorst, T. (Eds.), Losing your Land: Dispossession in the Great Lakes (pp. 11-35). Woodbridge, Suffolk: James Currey.

Roth, R. (2009). The challenges of mapping complex indigenous spatiality: from abstract space to dwelling space. cultural geographies, 16(2), 207-227. doi:10.1177/1474474008101517

Satyal, P., Corbera, E., Dawson, N., Dhungana, H., & Maskey, G. (2020). Justice-related impacts and social differentiation dynamics in Nepal's REDD+ projects. Forest Policy and Economics, 117, 102203. doi:https://doi.org/10.1016/j.forpol.2020.102203

Suiseeya, K. R. (2016). Transforming Justice in REDD+ through a Politics of Difference Approach. Forests, 7(12). doi:10.3390/f7120300

Tsing, A. L. (2005). Friction: An Ethnography of Global Connection. Princeton & Oxford: Princeton University Press.

Tsing, A. L. (2013). More-than-human sociality: a call for critical description. In Hastrup, K. (Ed.), Anthropology and nature (pp. 27-43). London: Routledge.

Tsing, A. L. (2015). The Mushroom at the End of the World: On the Possibility of Life in Capitalist Ruins. Princeton, NJ: Princeton University Press.

UE en RDC, & Nasi, R. (2019, December 12). Merci à nos partenaires du projet @CIFOR_forets à l'#UNIKIS, Université de Kisangani et à #Yangambi [Tweet]. Retrieved from https://twitter.com/UEenRDC/status/1205083406993969152?s=20

224

Vermeylen, S. (2019). Special issue: environmental justice and epistemic violence. Local Environment, 24(2), 89-93. doi:10.1080/13549839.2018.1561658

West, P. (2012). From Modern Production to Imagined Primitive: The Social World of Coffee from Papua New Guinea. Durham, NC: Duke University Press.

Whatmore, S. (2002). Hybrid Geographies: Natures Cultures Spaces. London: Sage Publications.

Windey, C., and Van Hecken, G. (2019). Contested mappings in a dynamic space: emerging socio-spatial relationships in the context of REDD+. A case from the Democratic Republic of Congo. Landscape Research, online first. DOI:10.1080/01426397.2019.1691983.

225

226

Appendix 1: Coding system to anonymize interviews

A) Expert interviews

E.g.: 1-RCS-KIN-20150508

1) Interview number

2) Category of actors

CSI Congolese State Institutions CCSO Congolese Civil Society Organizations (some of them manage REDD+ projects, e.g. OCEAN) CTSP Congolese Technical & Scientific Partners INGO International Non-Governmental Organizations ITSP International Technical & Scientific Partners PA Private Actors RCS REDD+ Coordination Structures (special government entities including international consultants from multilateral funding organizations) 3) Location of interview

ISA Isangi KIN Kinshasa KIS Kisangani RAB Route Amex-Bois RIT Route Ituri YAF Yafunga YAN Yangambi SKY Skype 4) Date of Interview (yearmonthday)

B) Interviews, informal focus groups and participatory mapping in villages

E.g. : 45-I-YAF-20160329

1) Interview, informal focus group or participatory mapping number

2) Type of method

I Interview IFG Informal focus group PM Participatory mapping 3) Location of interview

ARAB Ancienne Route Amex-Bois ARBU Ancienne Route Buta BU Bula ISA Isangi KIS Kisangani MAS Masako RAB Route Amex-Bois

227

RBU Route Buta RIT Route Ituri YAF Yafunga YALI Yaliaboga YALO Yalosuna YAN Yangambi 4) Date of Interview (yearmonthday)

228

229