Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Relations of power, networks of water : governing urban waters, spaces, and populations in (post)colonial… Kooy, Michelle Élan 2008

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2008_fall_kooy_michelle.pdf [ 6.79MB ]
JSON: 24-1.0066407.json
JSON-LD: 24-1.0066407-ld.json
RDF/XML (Pretty): 24-1.0066407-rdf.xml
RDF/JSON: 24-1.0066407-rdf.json
Turtle: 24-1.0066407-turtle.txt
N-Triples: 24-1.0066407-rdf-ntriples.txt
Original Record: 24-1.0066407-source.json
Full Text

Full Text

   RELATIONS OF POWER, NETWORKS OF WATER: GOVERNING URBAN WATERS, SPACES, AND POPULATIONS  IN (POST)COLONIAL JAKARTA   by   MICHELLE ÉLAN KOOY  B.Sc., Calvin College, 1998 M.E.S., York University, 2002      A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY   in   THE FACULTY OF GRADUATE STUDIES   (Geography)       THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   May 2008   © Michelle Élan Kooy, 2008 ABSTRACT  This thesis documents the genealogy of the development of Jakarta’s urban water supply infrastructure from 1873 (the inception of the first colonial water supply network) to the present. Using an analytical framework of governmentality, supplemented by insights from postcolonial studies and political ecology, the thesis explains the highly unequal patterns of water access in Jakarta as the product of (post)colonial governmentalities, whose relations of power are expressed not only through discursive categories and socio-economic relations, but also through material infrastructures and urban spaces. The thesis presents material from the colonial archives, Jakarta’s municipal archives, and the publications of international development agencies and engineering consultancy firms. This is combined with primary data derived from interviews, questionnaires, and participant observation of the implementation of current pro-poor water supply projects in Jakarta.  This data is used to document how water supply is implicated in the discursive and material production of the city and its citizens, and to challenge conventional developmentalist and academic analyses of water supply access. Specifically, a conceptual triad of water, space, and populations – produced through, but also productive of government rationalities – is used to explain two apparent paradoxes: (1) the fragmentation of access in Jakarta despite a century of concerted attempts to develop a centralized system; and (2) the preferences of lower-income households for non-networked water supply, despite its higher cost per unit volume. This analysis hinges on an elucidation of the relationships between urban governance and urban infrastructure, which documents the interrelated process of differentiation of types of water supply, water use practices, populations, and urban spaces from the colonial period to the present. This, in turn, is used to explain the barriers being encountered in current pro-poor water supply development projects in Jakarta. The thesis thus makes a contribution to current academic debates over the ‘colonial present’. The contribution is both theoretical – in the emphasis placed upon the materiality of governmentality – and empirical. Finally, the thesis also makes a contribution to the urban and development studies literatures through its reinterpretation of the urban ‘water crisis’.      ii TABLE OF CONTENTS    Abstract ...….………..…………………………………………………………………………….. ii Table of Contents …...……………………..……………………………………………………… iii List of Tables ………………………………….…………….……………………………………. vi List of Figures ……………………………………...………...…………………………………… viii List of Illustrations ...……………………………………………………………………………… ix List of Maps ………………………………….…………………………………………………… x List of Abbreviations & Acronyms ……………………………….………………………………. xi Glossary ……………………………………………………...…………………………………… xv Acknowledgements …………………………………………...……...…………………………… xvi    CHAPTER 1 Relations of Power, Networks of Water: Jakarta, Indonesia ……....... 1      1.1 Introduction: Jakarta’s Paradoxes of Provision ……………………………………..…….. 1      1.2 A New Analytic: Historicizing, Politicizing, and Materializing Inequitable Access …....... 5      1.3 Hydraulic Histories and Research Methods ………….……......…………………..………. 10      1.4 Chapter Outline …………………..………………………………………..………………. 17      1.5 Conclusion: Imagining Jakarta ……………………….…………………….……………… 19    CHAPTER 2 Governmentality: Governing (Post)colonial Natures, Spaces,                         and Subjects ...……………………………………...…………………….  21      2.1 Introduction ...……………………………………………………………….……………... 21      2.2 Governmentality: From Foucault and Beyond ……...……………………………………... 24      2.3 Government of Nature, Space, Subjects in the Postcolonial .…..………………………….. 27           2.3.1 Postcolonial Governmentality .……..………………..……………………………….. 28           2.3.2 Power-Nature-Society & Governmentality …...……….……………………………... 32           2.3.3 Materiality of Government: Layers of Relations of Rule …….….……...…………… 38      2.4 Conclusion: Productive Contradictions - Jakarta’s Urban Water Supply ...……………….. 40       iii CHAPTER 3  Splintered Networks: Urban Water Supply in Jakarta, 1870-1990 … 42      3.1 Introduction ………………………..………………………………………………………. 42      3.2 Artesian Water Supply: 1873-1922 …...…………………………………………………… 45      3.3 Spring Water Network: 1923-1945 ...…………………………………………………….... 55      3.4 Surface Water Treatment Plants: Pejompongan I & II, 1950-1966 ......…………………… 62      3.5 Large Scale and ‘Mini’ Water Treatment Plants: 1965-1990 .……...…..……..….……….. 68      3.6 Conclusion: Colonial & Contemporary Waters …………………………………...………. 76    CHAPTER 4 Invisible Networks: The Failed Development of Jakarta’s                         Water Supply, 1990-2007 …….…………………………………………  78      4.1 Introduction: 50 Years of Failure? ……………………………………………………….... 78      4.2 Invisible Networks I: Hydraulic Modernity and Development Discourse ..………………. 80      4.3 Invisible Networks II: Discounting Diversity and Miscounting the Disconnected …….…. 85      4.4 Failed Modernization I: PAM Jaya System Improvement Project, 1990-1998 .……......…. 90      4.5 Failed Modernization II: Private Sector Participation, 1997-2007 ........…………………... 100      4.6 Conclusion: Exacerbating Inequity ……….……………………………..………….……... 113    CHAPTER 5 The Materiality of Governmentality: Producing Water,                         Constituting Citizens, and Uncooperative Natures …………………… 115      5.1 Introduction: Rendering Technical and Rematerializing Government .….………...……… 115      5.2 The Rise of the ‘Sanitary City’, 1870-1901 ……….……………………………….……… 118      5.3 Hydraulic Modernity, 1901-1945 ……….……………………………….…………...…… 129      5.4 Modernist Monuments, 1949-1965 ……….…………………………….…………….…… 140      5.5 Lubricating Capital, 1966-1990 ……………………………….………………………...… 149      5.6 Conclusion: Uncooperative Natures …………………………………………………….… 159    CHAPTER 6 Preferences of the Poor: Facilitating Fragmentation and Frustrating                         Development ………………………….…………………………………. 161     6.1 Introduction: Problematic and Paradoxical Preferences for Non-Network Waters ………... 161     6.2. Patterns of Water Use and Preferences of the Poor: Rational Responses ……………….… 165           6.2.1 The Thirsty Poor? …………………………………………………..………………… 165  iv           6.2.2 Patterns of Water Use Amongst Low-Income Households in Jakarta ……………..… 167           6.2.3 Rational Responses to Spatialized Access and Service Quality …………..………..… 171           6.2.4 Identity and Access to Public Water Supply Infrastructure: Rational Responses …..... 179      6.3 Frustrating Development and Facilitating Fragmentation …......………………………....... 184           6.3.1 Development Discourse: Water For All the ‘Thirsty Poor’ ….………………………. 184           6.3.2 Output Based Aid: Excising the Political ……………………..……………………… 185           6.3.3 Ambivalent Responses ………………………………………….……………………. 188           6.3.4 Splintering From Below ………………………………………….………………....... 191      6.4 Casting Doubt ………...……………………………………………………………………. 193      6.5 Conclusion: Resistance to Rule, Facilitating Fragmentation ……..……………………….. 195    CHAPTER 7 Conclusion: The Colonial Present, 1873/2007 ………………….....…… 197      7.1 Introduction …………………………………………………………………………...…… 197      7.2 (Post)colonial Natures and Cosmopolitan Urban Theory …………………………….…… 199      7.3 Analytical Gaps and Avenues for Further Research: Scale and Subjectivity …………...… 200      7.4 Possible Futures: Pro-Poor Water Supply ……………………………………………….… 203      7.5 Conclusion: Re-Imagining Jakarta ……………………………………………………….... 205    Bibliography ……………………………………………………………………………………. 206    Appendix 1   Research Timeline, Methodology, and Chronology …………………………….… 250 Appendix 2   List of Research Interviews…………………………………………………........… 251 Appendix 3   Household Questionnaire………………………………………………………...… 256 Appendix 4   Location of Research Sites in Jakarta…………………………………………….… 263 Appendix 5   Ethics Approval Form……………………………………………………………… 265     v List of Tables  Table 3.1  Urban water supply infrastructure development in Jakarta, 1873-1990………….  44 Table 3.2  Artesian wells, pump stations, and reservoirs installed in Batavia, 1873-1921….  47 Table 3.3  Mini-water treatment plants, 1977-1982 ……..……………………………….…  72 Table 4.1  Native development and predicted per capita piped water supply consumption,                            1920-1950 ………………………………………..……………………………...   83 Table 4.2  International development involvement with Jakarta’s centralized water supply                            infrastructure, 1950-1990s ………………………………………………………   85 Table 4.3  (Post)colonial development and per capita piped water supply consumption …...  87 Table 4.4  Network water consumption and connections per house type in Jakarta,                            1994-1995 …………..…………………………………………………………...   89 Table 4.5   Centralized water supply infrastructure development project JUDP-II,                             1990-1998 ………………………………………………………………………   91 Table 4.6  Piped water supply tariff increases throughout the PJSIP ……….………………  93 Table 4.7  PJSIP Targets vs. Project Outcomes ……..………………………………………  93 Table 4.8  Over-estimations of centralized network coverage by the World Bank in 1995 ...  98 Table 4.9  Private sector operator performance targets: Original and revised, 1998-2002 …  102 Table 4.10  Private sector operator performance targets, 2002-2006 ...………………………  102 Table 4.11 Water supply network customers per tariff band, 2003 and 2006 ……………….  107 Table 4.12 Number of low-income households connected by TPJ, 1998-2004 …..…………  108 Table 5.1  Govermentality and water supply infrastructure in (post)colonial Jakarta,                            1600s-1990s ……………………..………………………………………………   116 Table 6.1  Per unit volume costs of water supply: piped, vended, bottled, and groundwater                            in Jakarta ……….………………………………………………………………..   163 Table 6.2  Average L/capita/day consumption, according to socio-economic status ……….  164 Table 6.3  Household expenses for water supply, as a proportion of income ………………  167 Table 6.4  Cost of new household connection to the piped network, according to tariff                            group ………………………………………………………………………….…   172 Table 6.5  Percentage of low-income residents according to Municipality, 2004 …...….......  175   vi  Table 8.1  Research Chronology, Methods, Sources, and Timeline ………………………...  250 Table 8.2  List of Research Interviews, 2004-2007 …………………………………………  251 Table 8.3  Quantitative & Qualitative Field Research Locations: Household survey                            respondents and UPC communities ………………………………………..……   264                                            vii List of Figures  Figure 3.1  Growth of European population and number of household connections in                            Batavia, 1899-1920 ……………...………………………………………………   52 Figure 4.1  Use of public and private water service providers in Indonesia, by income                            quintile ………………………………………….……………………………….   84 Figure 4.2        Deep Wells and Registered water consumption from deep wells in Jakarta,                           1977-1986 ……..…………………………………………………………………   89 Figure 4.3  Decrease in public hydrants throughout PJSIP, 1990-1998 …………………..…  95 Figure 4.4  Unaccounted for Water (as percentage of water produced) in DKI Jakarta,                            1990 – 1997 ……………….………………………………………………….….   96 Figure 4.5  Increase in connections to centralized network, 1990-1997 ……………..………  97 Figure 4.6  Water Charge vs. Average Tariff, 1998-2005 ………...…………………………  104 Figure 4.7  Shortfall for Palyja, 1998-2003 ………………………………………………….  105 Figure 4.8  Shortfall for TPJ, 1998-2003 …………………………………………………….  105 Figure 5.1  Jakarta water supply production versus distribution, 1950-1990 …..……………  153 Figure 6.1  Percentage of residents using piped water, vended water, shallow groundwater,                            and bottled water by socio-economic quintile ……………………………….….   164 Figure 6.2  Expenses on water supply, according to water source accessed …………………  168 Figure 6.3  Type of water source used, according to income ……...…………………………  169 Figure 6.4  Source of water supply, according to regular vs. fluctuating income ……………  170 Figure 6.5  Type of water supply used, according to land tenure and building status ……….  170   viii List of Illustrations  Illustration 1.1  ‘Spaghetti pipes’ distributing groundwater from a deep well to                                          low- income residents in Muara Kamal, North Jakarta ….………….…   2 Illustration 3.1  Artesian well and reservoir at Batoe-Toelis, 1880 …...…………………  48 Illustration 3.2  Water vendors in Batavia, 1918 …………………………………………  53 Illustration 3.3  Water Treatment Plant Pejompongan I ……………………………….…  62 Illustration 3.4  Advertisement for Sunter Paradise housing estate, 1992 ….……………  72 Illustration 4.1  Indonesia is behind in the regional race, 2007 …….……………………  82 Illustration 4.2  Moralizing water theft I: “Stop Curi Air” …...……………………..……  111 Illustration 4.3  Moralizing water theft II …………………………………………...........  112 Illustration 5.1  Microscopic plants and animals from well water in Batavia, 1873 ..……  124 Illustration 5.2  Artesian water hydrant, 1880 ……………………………………………  125 Illustration 5.3  Artesian well and reservoir in the Koningsplein, 1880 ………...……….  126 Illustration 5.4  Artesian well and reservoir at Meester Cornelius, 1880 …………...........  126 Illustration 5.5  Washing, Bathing, and Recreating in the canal …………………………  134 Illustration 5.6              European swimming pool, Batavia ……………………………...………  134 Illustration 5.7  Batavia hotel advertising ‘modern rooms with running water’,1927 .......  137 Illustration 5.8  Advertisement for Eau de Cologne and Toiletzeepen, 1927 ……………  138 Illustration 5.9  Water Treatment Plants Pejompongan I&II …………………………….  141 Illustration 5.10  Fifty three kilometres of colonial cast iron pipes, 1937 ………...………  143 Illustration 5.11  Spring water network pipes, Buitzenzorg to Bogor …………..…………  143 Illustration 5.12  Installation of spring water network main pipeline, 1922 ………………  144 Illustration 5.13  Friendship fountain, channelling international ambitions ………………  148 Illustration 6.1  Private bore well selling water to low-income neighbours in Kamal                                         Muara, North Jakarta …………………………...……………………….   192 Illustration 6.2  Spaghetti pipe network from illegal connection, North Jakarta ……...…  193   ix List of Maps  Map 3.1  Artesian water supply network, 1873-1922 ……………………………….……..  46 Map 3.2  Urban water supply and land use in Batavia, 1897 ………..…………..................  50 Map 3.3  Spring water supply network of Batavia, 1922-1945 ……………………............  56 Map 3.4  Piped water supply network in Jakarta, 1950-1976 …………………...................  63 Map 3.5  Development of Jakarta’s centralized water supply network, 1977-1982 ……….  70 Map 4.1  Jakarta’s private sector water supply concessions, 1998-2023 ………………..…  100 Map 5.1  The outline of the new center: Pejompongan I&II ………………………………  147 Map 5.2  Booster pumps installed in selective areas of Jakarta, 1960-1980s ……………...  155 Map 5.3  Illegal tertiary networks ………………………………………………………….  158 Map 6.1 Density of tertiary pipe network coverage according to socio-economic status,                            distribution cluster PC-049, East Jakarta ………………………………………..   173 Map 6.2 Areas of low water pressure, Eastern half of Jakarta …………………………….  176 Map 6.3  Household connections in low-income settlement: RW 4, Kel.Jembatan Besi .....  189 Map 8.1 Location of Research Sites in Jakarta …...……………………………………….  263   x List of Acronyms and Abbreviations   AcF  Action Contre la Faim  ADB  Asian Development Bank  AKAINDO  Asosiasi Kontraktor Air Indonesia, Propinsi DKI Jakarta (Indonesian Association of Water supply contractors)  Amrta Indonesian Institute for Water Literacy  ATM  Atmospheric Pressure  BAPPEDA Badan Pembangunan Perencanaan (Regional Government Development Planning   Board  BAPPEKO Badan Perencanaan Kotamadya Planning Board (Municipal Development Planning Board)  BAPPENAS Badan Pembangunan Nasional (National Development Planning Agency)  BEST  Institute for Integrated Social and Economic Development  BORDA Bremen Overseas Research and Development Association  BPLHD Badan Pengendalian Lingkungan Hidup Daerah (Regional Environmental Planning Board)  BPS  Badan Pusat Statistik  CIDA  Canadian International Development Agency  DFID  Department of Foreign Aid and International Development  DKI Jakarta   Daerah Khusus Ibukota (Special Capital District of Jakarta)  DPRD  Dewan Perwakilan Rakyat Daerah (Provincial parliament)  FAKTA Forum Warga Kota Jakarta (Jakarta Residents’ Forum)  FORKAMI Forum Komunikasi Pengelolaan Kualitas Air Minum Indonesia (Indonesian Communication Forum on Drinking Water Quality Management)  GTZ  Deutsche Gesellschaft fur Technische Zusammenarbeit (German Development Agency)  HIVOS  Humanist Institute for Cooperation with Developing Countries  HP3/Lestari Healthy Places, Prosperous People/Lingkungan (IDRC Jakarta Focus City Project)  IDRC  International Development Research Center   xi IHS  Institute for Housing and Urban Development Studies  INFID  International NGO Forum on Indonesian Development  INFOG  Indonesian Forum on Globalization  INCIS  Indonesian Civil Society Institute  ITB  Institut Teknologi Bandung (Technical University of Bandung)  IWACO Consultants for Water and Environment (Netherlands)  JBIC  Japan Bank for International Cooperation  JICA  Japan International Cooperation Agency  JMP  Joint Monitoring Program for the Millennium Development Goals  JWSRB Jakarta Water Supply Regulatory Board  KIMPRASWIL Ministry of Settlements and Regional Infrastructure  KIT  Koninklijk Instituut voor de Tropen (Royal Tropical Institute)  KITLV  Koninklijk Instituut voor Taal-, Land- en Volkenkunde (Royal Netherlands Institute of   Southeast Asian and Caribbean Studies)  KKPPI   Committee of Policy for the Acceleration of Infrastructure Development  KOMPARTA Komunitas Pelanggan Air Minum Jakarta (Jakarta Water Consumers Community)  KruHa  Koalisi Rakyat Untuk Hak Atas Air (People’s Coalition for the Rights to Water)  KTP  Kartu Tanda Penduduk (Personal Identification Card)  L Litres  LP3ES Lembaga Penelitian, Pendidikan dan Penerangan Ekonomi dan Sosial(Institute for Social and Economic Research, Education and Information)  LPPSE Lembaga Pengkajian dan Pengembangan Sosial Ekonomi (Institute for Economic Social Studies and Development)  L/s  Litres per second  MDB  Multilateral Development Bank  MDGs  Millennium Development Goals  NEI  Netherlands East Indies  OBA  Output Based Aid  OED  Operations and Evaluation Department (World Bank)  xii  PALYJA PAM Lyonnaise Jaya  PAM Jaya Perusahaan Air Minum Jakarta (Jakarta water supply company)  PBB Pajak Bumi dan Bangunan (Land and Building Tax)  PERPAMSI Persatuan Perusahaan Air Minum di Seluruh Indonesia (Indonesian Association of Water Supply Companies)  PJSIP  PAM Jaya System Improvement Project  PPMA Pusat Pengembangan Masyarakat Agrikarya (Center for Agro Community Development)  PSP  Private Sector Participation  Rp.  Rupiah (Indonesian currency)  RT  Rukun Tenga (neighbourhood sub-district)  RW  Rukun Warga (neighbourhood district)  SB-AB Subsidi BBM untuk Penyediaan Air Bersih (Fuel Subsidy Reduction Compensation Program for Clean Water Supply)  TPJ   Thames PAM Jaya  UFW  Unaccounted for Water  UNESCO-IHE International institute for Infrastructural, Hydraulic and Environmental engineering  UNIKA Universitas Katolik Soegijapranata  UPC  Urban Poor Consortium  UPDATE Urban Poor Data Acquisition and Technical Evaluation project  URDI  Urban Research Development Institute  US-AEP United States Asia Environmental Partnership  USAID  United States Agency for International Development  USAID-ESP United States Agency for International Development – Environmental Services Program  VOC  Vereenigde Oost-Indische Compagnie (Dutch East India Company)  WALHI Indonesian Forum for the Environment (Friends of the Earth International)  Waspola  Indonesia Water Supply and Sanitation Policy Formulation and Action Planning Project   xiii WHO  World Health Organization  WSP  Water and Sanitation Program (World Bank)  WTP  Water Treatment Plant  YLBHI  Yayasan Lembaga Bantuan Hukum Indonesia (Indonesian Legal Aid Foundation)  YLKI  Yayasan Lembaga Konsumen Indonesia (Indonesian Consumers’ Association)                                           xiv Glossary   Air   Water  Air isi ulang ‘Refill bottled water’ – potable water sold by small-scale businesses to customers who bring their own 19 L bottled water containers. Water is sold at between Rp.2,000-3,000 for 19L, in comparison to name brand bottled water distributors, who sell 19L for approximately Rp. 10,000.   Gemeente  Municipality  Indische   Indies  Inlanders  Indigenous (native) subjects of the colonial state  Ingenieur  Engineer  Jakarta Barat  West Jakarta  Jakarta Timur  East Jakarta  Jakarta Pusat  Central Jakarta  Jakarta Selatan South Jakarta  Jakarta Utara  North Jakarta  Kampong ‘village’- in the urban context a description of an informally planned and unserviced low-income settlement  Kecamatan  Municipal district  Kelurahan  Municipal sub-district  Lurah   Head of Kelurahan government office  UFW Unaccounted for Water is the volume of water that is lost during distribution, either through administrative or technical leakage; it is usually indicated as a percentage of total water produced.  Vended water  Water from public hydrant distribution points, sold by ambulatory vendors to neighbouring households. Sold in 40 L containers, vended water costs between Rp.1,500-3,000 per 40L, with cost depending on the distance of the household from the distribution point.   xv Acknowledgements  The guidance, assistance, and encouragement of many people made this dissertation possible. Foremost is Karen Bakker, whose mentorship and counsel were invaluable from the inception to completion of not only the doctoral research, but throughout my years at UBC. Her initial support in the selection of both UBC and Jakarta as places to work was decisive, and her research and politics continue to inspire. I am also grateful for the support – often delivered long distance – of my committee members and other faculty at UBC. Jim Glassman, Juanita Sundberg, and Abidin Kusno provided encouragement and advice both during my time at UBC and in this last year of writing. Michael Leaf in particular provided much collegial advice during each phase of the research and I am very appreciative for his close reading of earlier drafts of the thesis. In Indonesia, I would like to thank Teti Argo, from the Institut Teknologi Bandung. Her own dissertation provided me with valuable information on water supply in Jakarta and fuelled the origins of my own research interests here. In addition, the support she selflessly provided to make my own research possible (securing a research visa and necessary permits) was much appreciated. Institutionally, time spent in Jakarta conducting research, writing, and then working while writing was partly made possible through the assistance of graduate and departmental secretaries from UBC. In particular, Junnie Chung has helped me in many moments ‘at the last minute’ with administrative matters. Financially, I am grateful to the Social Science and Humanities Research Council of Canada for the provision of a doctoral fellowship, and to the Hampton Fund for provision of a small research grant to finance initial archival research in the Netherlands. While in Jakarta, the goodwill, assistance, and curiosity of many people was crucial to the process of both research and writing.  This includes staff from the World Bank, Water and Sanitation Program, Asian Development Bank and bilateral agencies; professionals in the water sector who were passionate to talk about all aspects of urban water supply in Indonesia; and local NGOs who were encouragingly interested in my research and patient with the slow process of producing any results that might be useful for their own campaigns. Everyone I met received me with goodwill, if not quite comprehending what exactly this ‘relations of power’ stuff was all about. In particular, I would like to thank those who gave most unstintingly of their busy time: August Soetisnjo, Foort Bustraan, Jim Woodcock, Marco Kusumawijaya, and Nila Ardhanie. Separate from the above, I am grateful to all of the residents of low-income communities in Jakarta who were willing to talk with me about water, introduced me to and organized their neighbours to talk with me, and those households who answered more formal questionnaires. None of these people are ever likely to read these acknowledgements, or receive any tangible direct benefit from my research, but from them I learned, and gain, so much. In some small way I hope I piqued their own  xvi curiosity, and tried to reciprocate with answering the many questions about life in Canada that they had for me: conditions of urban life, agriculture, and yes – the weather. In addition to the residents themselves, I am also deeply appreciative for the support from the entire staff of the Urban Poor Consortium. Without their willingness to introduce me and then accompany me to talk with the constituents of their communities, none of that part of the research would have been possible. In this regard, I also need to thank Nur Endah Shofiani – my former field research assistant whose energy, humour, and work ethic kept me inspired. I would also like to thank the producers of the popular Indonesian TV show ‘Buleh Masuk Kampung’ (white person goes to the kampong) for providing this phase of my research with a theme song that while ingratiating and tuneless, at least made my field research a hit with children, and gave me not so subtle reminders of my own positionality. Life in Jakarta was made both more productive and pleasurable through the companionship of many people. This includes all of my colleagues at Mercy Corps Indonesia, especially Sandrine Capelle-Manuel and Karen Pesjak who made the time and space for critical dialogue amidst regular scheduled programming; my yoga pupils; and my former roommates Alysson Oakley and Lalan Susanti. A part of the critical academic dialogue that I missed while away from UBC was alleviated through conversation with staff of the Department of Urban Environment, Graduate Studies at the Catholic University of Semarang, as well as Joost Cote, and fellow researcher Michiko Inawami. Friends and family in Canada and the Netherlands have provided years of support and encouragement throughout this process and I thank them all. This includes my parents, Henry and Marg Kooy; my foster parents in Vancouver, Corrie and John Kamphof; my in-laws Jaap and Evelyn Martijn; and my brothers Rodney, Philip and Brendan. There was also, especially, my sisters-in-arms: Deborah Vriend-Van Duinen, Rachel Melzer, and Holly Spaman; their politics, ethics, and humour keep me going. Finally, there is the person who has lived, and worked, with me through all of this - Ernst-Jan Martijn. As my partner in this, and all endeavours, his faith, love, and understanding remained constant. In addition to all of the intangibles, the thesis is also due in no small part to his material contributions – his assistance in translating documents from the Dutch archives, and his technical support in creating the maps, graphics and tables.   xvii Chapter 1  Relations of Power, Networks of Water: Jakarta, Indonesia   1.1 Introduction: Jakarta’s paradoxes of provision   Water is fundamental to life, and an important condition of meeting essential material needs. This material nature of water is reflected in the pursuit of the Millennium Development Goal for equitable access to clean water supply1. However, water is also much more than a biophysical entity. As access to water supply (or lack thereof) may grant a certain quality of life or deny certain ways of living, water also contains and expresses, and is constituted by, relations of social power (Gandy, 2004; Linton, 2006; Swyngedouw, 2004). Chronicling the emergence of the urban water supply network in Jakarta, Indonesia from colonial past to present, this thesis traces how the flows of water in the city have been enrolled within relations of power to alternately grant, deny, enforce, or resist particular ways of living and being. Specifically, identifying how the production of the city’s piped water supply network has been channelled to particular kinds of residents, living within particular kinds of urban spaces, the thesis highlights the resultant co-constitution of waters, spaces, and populations. This conceptual triad, identified as a product of the ways in which relations of power have been worked through networks of urban water supply, is subsequently used to explain the conditions of access within the contemporary city. Undoubtedly, the conditions of access in the contemporary city of Jakarta are inequitable. Of the city’s twelve million residents, less than half are served by the centralized piped water supply network (JWSRB, 2004), which provides lowest per unit volume cost of clean water in the city. Moreover, access to the piped network is fragmented both socially and spatially - dependent upon both where one lives, and who one is. As a result, the vast majority of the city’s residents rely on a complex mix of types of water and service providers: shallow and deep groundwater, piped network water, river water, and bottled water delivered through pipes, pumps, private wells, ambulatory vendors, depots, tanks, and networks (Waspola, 2007). Reflected within the urban landscape, spaces both above and below ground in the city are traversed by a tangled network of public, private, and communal infrastructure for accessing, filtering, and distributing a variety of water sources.   1 The eight United Nations Millennium Development Goals were agreed upon at the UN Millennium Summit (2000). The MDGs set a specific target for water supply: reducing by half the proportion of individuals without sustainable access to adequate quantities of affordable and safe water by 2015. See  1  Illustration 1.1 ‘Spaghetti pipes’ distributing groundwater from a deep well to low-income residents in    Muara Kamal, North Jakarta   Source: Photo by Author (2005)  However, while providing a stark contrast to the western urban norm of universal access through centralized networked infrastructure systems, I argue in this thesis that Jakarta’s urban water supply should not be interpreted as evidence of Jakarta’s lack of development. Rather, I argue that differentiated access is the product of specific practices of development structured by relations of power which are embodied in Jakarta’s water supply infrastructure. This argument is admittedly counter-intuitive, given that decades of urban development by colonial and postcolonial governments, multi-lateral development banks, international aid agencies, and the private sector state, have all – since the 1870s - invested in the consolidation of a centralized urban water supply system. A renewed emphasis on achieving universal coverage through this centralized system emerged in the 1990s with the World Bank’s largest ever investment into Jakarta’s piped water supply (World Bank, 1990b). This was followed in 1997 by the efforts of private sector operators to consolidate the piped water supply system, and the ongoing Millennium Development Goal related pro-poor water supply projects2. As indicated by the existing multiplicity of water networks in Jakarta, despite over a century of development efforts, including a particular emphasis in the last few decades on improving access by the poor, programs have systematically failed to improve inequitable access through the consolidation of a centralized system of universal distribution. While the official (generous) estimate is that 56% of the city’s residents are connected to the centralized network (JWSRB, 2004), reported figures vary  2 Jakarta’s water supply provision was privatized in 1997 resulting in 25 year concession contracts with Suez Lyonnaise des Eaux in the western half of Jakarta, and Thames Water International in the eastern half of Jakarta.  2 significantly, and this figure both includes residents who access the network through public standpipes and mobile water vendors, and excludes the number of households which have a connection but which rely primarily on other sources (e.g. groundwater) due to quality or service concerns (e.g. low pressure). In addition, as is documented in this thesis, the World Bank’s project, the private sector operators, and recent pro-poor initiatives have all been less than successful in increasing access by low- income residents. The limited access, and service, provided by the city’s centralized network despite the decades of urban development projects targeting improvements in Jakarta’s urban water supply system have been interpreted by aid agencies and multilateral development banks as a failure. Subsequently, the perceived inability of Jakarta to develop according western norms and its concomitant perceived lack of progress along this trajectory of urbanization and modernization has been given two different interpretations. First, some of the analyses of international development programs have rendered the problems of persistent exclusion into technical (by which I mean de-politicized, technocratic) interventions, translating political problems of poverty and equity into technical diagnoses amenable to programs of improvement (see Ferguson, 1990; Li, 2007, Mitchell, 2002). As part of this discursive process of ‘rendering technical’ Jakarta’s water supply problems, various agencies and initiatives have identified the lack of access to water supply infrastructure as a problem of corruption (Server, 1996), lack of finance (Akhtar, 2005; World Bank, 1974, 2005), insufficient regulation of both public and private service providers (Braadbaart, 2007; Jensen, 2005), or of rapid rates of urban growth outpacing infrastructural development (Chifos and Suselo, 2000; Hamer et al., 1986). Alternately, academic analyses critiquing these interventions point to the effects of neo-liberal restructuring and its concomitant reshaping of urban infrastructure networks (Bakker, 2003; Graham, 2001; Graham and Marvin, 2001). Specifically, as Graham and Marvin have argued in Splintering Urbanism (2001), the failure of Jakarta’s urban water supply network is explained as part of the ‘collapse of the integrated ideal’ seen in networked utility provision around the world. The fragmentation of access, control, and pricing of network infrastructure leading to ‘splintering’ is attributed to the wider changes in aid and financial flows, technological innovation, social attitudes, and governance – particularly the reconfiguration of citizens’ entitlements in light of newly dominant understandings of the appropriate role of the state in services provision. This thesis proceeds to query both of these perspectives, arguing that they overlook two key characteristics of Jakarta’s water supply which undermine their diagnoses. First, the fragmentation of Jakarta’s water supply network has occurred throughout a century of investment into a centralized piped network system, not merely in the last decade of neo-liberal led development and privatization. In fact, Jakarta’s urban water supply network has been ‘splintered’ since its origins in 1873, and patterns of access remain remarkably similar, with socio-economic class replacing race in the  3 postcolonial period3. Second, as documented in this thesis, there is the fact that the urban poor – who are the most disadvantaged over their lack of access to the piped water network – do not seem to be ‘thirsty’ for piped water connections. Although both paying more for access to clean water and suffering disproportionate health effects from lack of access (see Agatini et al., 2005; Bakker et al., 2006; McGrahanan et al. 2001), low-income residents are increasingly relying on multiple types of water services providers, documented in the growth of small-scale water supply systems and distributors (World Bank, 2004a; Waspola, 2007).  Providing a puzzling contradiction for technocratic experts, the failures of decades of development projects seeking to improve access by the poor seem due in part to what seem as perverse preferences of ungrateful beneficiaries, who continue to combine multiple sources of water, and pay higher per unit cost prices for the waters that they do purchase. Using theory generated in and by northern contexts to (mis)understand the South, Western- based analyses of urban development can only cite these facts as anomalies (see Amin and Graham, 1997; cf Graham and Marvin, 2001; Robinson, 2003). Urban geographies of ‘splintering urbanism’ reliant on the development trajectories of western cities fail to explain the historical dimensions of relations of power that contribute to contemporary global economic processes (cf Coutard, 2002). As a result, this analysis obviates the fact that Jakarta’s centralized urban water supply network has been splintered since its inception rather than since its privatization. In addition, the privileged focus on the recent socio-economic processes structuring inequitable access (privatization) fails to consider how relations of power are also embedded within spatial relations and subjectivities. As a result, the theory of splintering urbanism presented by Graham and Marvin (2001) fails to explain why the poor in Jakarta might still not be thirsty for water from a system to which they are denied access, and have subsequently contributed to additional fragmentation through processes of ‘splintering from below’. On the other hand, developmentalist analyses diagnosing Jakarta’s lack of the urban infrastructural ideal – universal access through centralized network – as the product of failed development or failed modernization exclude any consideration of the ways in which relations of power structure access (Ferguson, 1990; Li, 2007). Moreover, these analyses do not only fail to offer an explanation, but completely ignore some characteristics of Jakarta’s water supply, and its urban poor, in order to maintain their expert credibility. Specifically, the contradictions in Jakarta’s water supply sector – highly splintered patterns of provision occurring throughout 100 years of investment into a centralized system, and a population of non-thirsty yet disadvantaged poor - are problematic for development discourses, as they identify gaps in the ‘expert closures’ of the technical diagnoses of problems and solutions within international development (Li, 2007). Unable to acknowledge these realities, developmentalist analyses and attendant official narratives of urban water supply in Jakarta  3 This is not to be taken to mean that I see categories of class and race as separately constituted, but were rather the categories articulated by the government to differentiate access to water supply, set tariffs, and calculate projections of per capita consumption of piped water. This is discussed in more detail in Chapters Three, Five and Six.  4 have, until recently, simply ignored both the growth of multiple water sources and services providers in Jakarta, and the population of non-thirsty poor4. In response to the inadequacy of the analyses offered by development agencies and western oriented academic theory, this thesis offers a new interpretation of existing patterns of access, inequity, and development in Jakarta.  Using the analytical framework of governmentality, outlined below, I explain current patterns of access as the product of ways in which colonial and postcolonial relations of rule have been worked through water supply, constituting specific relations been urban water, urban space, and urban populations. In this analysis, I argue that it is this imbrication of urban infrastructure, identity, and space which explains how the splintering of the city’s water supply continued parallel to the development of the centralized network – dividing the city along the axes of difference rationalized by government as necessary for rule; primarily, race and socio-economic class5. Similarly, the co- constitution of waters, infrastructures, and subjects explains why the urban poor might not be thirsty for connection to the piped network, as their identity forecloses some options, and makes other more preferable. This alternative explanatory framework is, I argue, both able to acknowledge, and subsequently explain, the reasons why Jakarta has multiple networks of water if there has been over a century of concerted development into a centralized system, and why the poor – who are the most disadvantaged over their lack of access to centralized system – do not seem to be thirsty.   1.2 A new analytic: Historicizing, politicizing, and materializing inequitable access  The conceptual framework deployed in this thesis focuses on the relationship between governmentality, urban infrastructure, urban space, and populations. Using Foucault’s analytic of governmentality (Foucault, 1991) together with critiques from postcolonial studies (Howell, 2004; Legg, 2006, 2007; O’Malley, 1996; Watts, 2003) and analyses of the relations between power-nature- society (Agarwal, 2005; Braun, 2000; Goldman, 2005; Li, 2007), I construct the genealogy of Jakarta’s urban water supply infrastructure development to trace how relations of power mobilized through government have been productive of patterns of provision within contemporary Jakarta.  4 It is only in the last few years that development analyses have begun to acknowledge and examine the multiple forms of provision, and preferences of the urban poor who are not served by the formal, centralized network (see ADB, 2002; Waspola, 2007). I explain in Chapters Four and Six, how these analyses focusing on ‘consumer choice’ and market options derive from a different theoretical framework, and I illustrate how they do not fundamentally address the problems of inequity. 5 The co-constitution of socio-economic class and race in both colonial and postcolonial eras is detailed in Chapter Five, specifically how the constitution of ‘European’ racial identity in the colonial era was informed by middle-class moralities, and the hidden racialization of middle and upper class in the Indonesian New Order era as dominated by Javanese and Chinese-Indonesians.  5 In this analysis the term ‘government’ is not exclusive to apparatuses of the nation state, but rather describes a complex assemblage of discourses and practices, exercised through state and non- state actors both above and below level of the nation, which shapes the conditions in which lives are lived (Li, 2005). The term governmentality refers to a specifically modern set of social relations of power between subjects and the state, in which the central problem for the state is the optimization and/or control of the relationship between population and resources. This relationship is articulated through ‘relations of rule’ in which states and subject interact. Subjects internalize, but also modify (and sometimes resist) the ways in which the state seeks to control their behaviour and social relations. States seek to control, but also modify (and sometimes abandon) specific projects. This last point is extremely relevant in the colonial context, where governmentality operated in modified and sometimes very incomplete ways (Howell, 2004; Legg, 2006; Prakash, 1999). The framework of governmentality has both specific methodological and analytical implications. First, the methodologies through which relations of power are analyzed involve historical, discursive, and political economic dimensions; this is discussed further in the next chapter section. Second, the analysis of relations of power mobilized through government requires a focus on subjects and populations from multiple perspectives, the subversion of binary categories, and attendance to the processes of enacting social relations of power. This means that the construction of subjects, and their abstraction into populations amenable to rule, is explored from multiple perspectives – both the colonial administrator and the ‘native’, for example; or the development expert and the ‘poor’ urban resident, while the binary categories upon which governmentalities rely so heavily, such as modern/traditional or Western/native are questioned, and in some cases, subverted.  A framework of governmentality also requires that we pay close attention to the process of enacting social relations of power; in this thesis, I focus on the process of ‘rendering technical’, which describes the process of depoliticizing, simplifying, and abstracting specific aspects of urban space and urban populations in order to render them amenable to development interventions. These points are explored in different ways in each chapter of the thesis, but particularly Chapter Six, which focuses on the problematic and contested constructions of Jakarta’s ‘poor’ urban residents. In addition, the thesis makes a contribution to the governmentality literature through focusing specifically on what I have termed the ‘materiality of governmentality’. This term is a response to a trenchant critique of much of the governmentality literature, a critique which suggests that this literature has focused too narrowly on discourse and abstract, intangible subjectivities, without acknowledging and incorporating the material aspects and expressions of power (see Harris, 2004; Jacobs, 1996; McEwan, 2003).  Of course, governmentality has always been material; my use of the term ‘materiality of governmentality’ is simply intended to suggest my call for a greater focus on these material dimensions.  In my case, I will focus on the ways in which both colonial and post-colonial governmentalities construct (but are also transformed and ultimately limited by) not only subjectivities,  6 but also urban spaces, infrastructure, and the bodies they seek to control. Subsequently, attending to the ways in which government operates through both discourse and practice, I highlight how relations of power are worked through physical technologies (like urban water supply infrastructure) and actual, tangible, and visceral ‘matter’. This focus on the material in balance with the discursive, and in particular understanding discursive categories as being worked through physical environments - to be both affected and effected by materiality – can contribute to the analysis of conditions of postcoloniality in ways that speak to the current lived geographies of inequity of the present (Yeoh, 2001). As part of this project of illuminating the materiality of governmentality, and using this to explain how conditions of access in the contemporary city remain informed by the colonial past, the thesis illustrates how the ‘layers of relations of rule’ as described by Li (2001) are grounded not only within socio-economic relations and discursive categories – but also within material infrastructure, physical spaces, and ecological processes. As such, this analysis can explain how the moment of origin of Jakarta’s water supply in 1873, and the ways in which it articulated legitimacy of access to piped water, continues to inform patterns of provision and preferences in the contemporary city. I stress that this analysis does not simply superimpose the past upon the present, or suggest a simplistic influence of the colonial upon the contemporary. Rather, in following the examples of Li (2001) and Moore (2001) I highlight how the relations of power from previous systems of rule that are layered both within the discourses and physical environments of contemporary places interact with the present in unpredictable ways. Therefore, while the post colonial era does not obviate effects of previous relations of rule, it is also not prescriptive or derivative for conditions of the present, including conditions of access to urban water supply. In conclusion, by combining three methods of analysis (historical, discursive, political economic) to interrogate the construction of subjects, binary categories, and the process of enacting social relations of power, the thesis thus traces how relations of power mobilized through government have been productive of patterns of water provision within contemporary Jakarta.  Through this genealogy I make particular conceptual claims relating to the analysis of relations between power- nature-society, and the understanding of urban space in the Global South. I posit that these claims, outlined below, are relevant both conceptually for the analytic of governmentality, but are also of direct practical significance for the current ‘programs of improvement’ seeking to improve equity of access to clean water in Jakarta. The following paragraphs identify these claims through an outline of the arguments of the thesis. First, in documenting how Jakarta’s water supply has been splintered since its inception over 100 years ago, I argue that the fragmentation of access observed in Jakarta is not a recent phenomena, as analyses of both failed modernization and splintered urbanism would suggest. On this basis, I claim that Jakarta’s splintered network and persistently inequitable access to clean water is the product of  7 colonial and postcolonial relations of rule that differentiated access to clean water by urban populations first by race, and then by class. The socio-spatial fragmentation evident within the contemporary city is thus a historical product, a pattern of provision which was first inscribed within the city’s first urban water supply infrastructure in the 1870s. In turn, this means that the current splintered nature of piped water in Jakarta is not an example of delayed development, but rather it is the product of relations of rule that - contested, contradictory, and not always intentional in its effects - have never had universal access as a political goal. Second, building on the documentation of socio-spatial fragmentation in Jakarta’s water supply, I identify how relations of power mobilized within government linked the production of urban water supply, urban space, and urban citizens. Highlighting the relationship between urban governance and urban infrastructure that has often been overlooked in analyses of relations of power and production of urban spaces in Jakarta (see Effendy and Kusumawijaya, 2004; Firman, 1998; Kusno, 2000), I identify urban water supply infrastructure as a material artefact of governmentality. I claim that the technologies of production, treatment, and distribution of Jakarta’s physical infrastructure systems are the product of contested and contradictory relations of rule. Moreover, I argue that these particular infrastructure systems and the socio-natures (waters) they produced, were both necessary to, and constitutive of, rule (cf Braun, 2000), as they facilitated the production of particular kinds of (differentiated) urban citizens, and urban spaces. Thus, I agree that infrastructure networks are implicated in the production of urban space and identity, but not necessarily in the ‘splintering urbanism’ thesis way. Third, by tracing the imbrication of infrastructure with identity, I claim that the constitution of subjectivity is related to the material environment – but in a much broader sense than argued by Agrawal (2005). While Agrawal (2005) retains a binary notion of nature/society, and thus limits the contribution of material environments to identity formation by considering only those subjects who consciously articulate and self-regulate actions according to environmentalist practices of conservation and protection of biodiversity as having an ‘environmentality’, I argue that we are all, consciously, or unconsciously ‘environmentalized’. Considering environment as a socio-natural entity, and thus expanding the scope of our interactions with nature to include urban natures, like water, I illustrate how the daily circulation of urban water supply through the bodies of all urban residents has also contributed to the formation of particular identities: illegal vs. legal residents, hygienic vs. contaminated, modern vs. traditional, and moral vs. unethical. Subsequently, as the articulation of particular identities (European vs. native, hygienic vs. contaminated, modern vs. backwards, urban vs. rural, developed vs. in need of development) were made more or less easy in relation to the type of water that one was physically able to access, this gives emphasis to the materiality of government – which I argue constituted citizens not only through discursive categories, but through material practices.  8 Fourth, I argue that governmentality is both a discursive and material project. Relations of power mobilized and contested through government are inscribed in physical space, material objects, and ecological processes as well as socio-economic relations and discursive categories. This avoids the binary of material/discursive or nature/society, replacing it with an understanding of an iterative relationships between discourse and practice, nature/society, or identity and infrastructure. In addition, it highlights the relevance of past relations of rule for the present, as colonial water supply infrastructure (and the spatial relations and subjectivities it alternately enforced and enabled) continues to inform patterns of supply within the contemporary city. Fifth, I claim that the relationship between governmentality and materiality is in fact iterative. Government rationalities and relations of rule were not only constitutive of physical spaces and material natures, but were also constituted in relation to the physical environment (Braun, 2000). In particular, the ‘uncooperative nature’ of water highlights the ways in which objects other than human have made a difference in the ways in which social relations unfolded in Jakarta (see Bakker and Bridge, 2006). Following chapters of the thesis illustrate how the biophysical and kinetic properties of water and the material components of infrastructure systems (electric pumps, aeration systems, iron pipes, PVC networks) either facilitated, or frustrated the ‘conduct of conduct’ that sought the establishment of ‘right relations’ within different urban populations, and between urban residents, waters, and spaces. The remaining arguments of the thesis relate to the ways in which the relations of power involved in the production of Jakarta’s urban water supply have been misunderstood, or ignored by development projects, in particular where they have ‘rendered technical’ the problem of equity of access. Specifically, I claim that these misunderstandings, and subsequent policy decisions and infrastructure projects, have in turn had their own material effects, exacerbating the fragmentation of access. This is illustrated by documenting the effects of technical interventions of multi-lateral development banks on the fragmentation of the centralized piped network, and the failures of pro-poor water supply development projects. The following section of the introduction presents my research methodology and the ways in which research methods and sources have shaped both the conceptual claims and structure of the thesis, but before proceeding further, the term ‘urban poor’ requires clarification. Laden with relations of power, the act of defining, categorizing, and identifying a group of urban residents as ‘the poor’ is acknowledged to be both a technical and political exercise (Escobar, 1995; McClintock, 1992; Mitchell, 1988). In the thesis I do not undertake an attempt to provide a definition of ‘the poor’ of Jakarta, a classification already fiercely debated amongst academics, activists, development banks, and national governments (Amis, 1995; Mitlin, 2005; Mitlin and Satterthwaite, 2002; Satterthwaite, 2001;  9 World Bank, 2006e; Wratten, 1995)6. My use of the term ‘the poor’ refers to populations of low- income and politically marginalized households who are excluded from access to the piped water supply network through failing to meet socio-economic and/or socio-political criteria set by the state. However, while I focus on aspects of legality (I.D. cards and land ownership in particular) and socio- economic status, I do not seek to under-emphasize the multiple aspects to urban poverty, or the many different relations which condition, mitigate, or define urban poverty7. Rather, my emphasis on these two criteria reflects the ways in which they have historically been, and currently are, conditions of access to piped water supply. The focus on legality and socio-economic status, basic criteria for access to urban service such as water, does not however support the popular ‘thirsty poor’ discourse of international development programs equating lack of access to piped water supply as a condition of poverty. In Chapter Six where I document the results of field research on urban water supply in low-income communities, I provide a more detailed description of how communities of ‘disconnected’ low-income households are defined by the state and development agencies – but rarely themselves - as ‘urban poor’. As Chapter Six explains, low-income households negotiate relations of access through various relations of power, and indeed may choose to remain disconnected from the centralized piped network that their legal and socio-economic status precludes.   1.3 Hydraulic Histories and Research Methods  The theoretical framework of governmentality carries with it certain methodological implications, which in turn have informed the limits of the analysis and the structure of the thesis chapters. A discussion of the analytical and structural implications of research methods follow after an outline of research methods and sources used in my research. First, a framework of governmentality requires a genealogical approach, requiring careful attention to the historical dimension of the analysis, and extensive use of archives. Second, it relies heavily on the use of discourse analysis (interpreting discourse as a practice), which requires the analysis not only of official documents but other textual and non-textual sources. Third, it also deploys aspects of political economic analysis, but builds upon this approach with the genealogical and discourse analysis methods mentioned above. In this thesis, the combination of these three forms of  6 Classification of ‘the poor’ in Jakarta by national government according to socio-economic criteria is explained in Chapter Six. 7 In fact, as many studies document, socio-economic and socio-political characteristics are strongly related, with other determinants of urban poverty: lack of social stability, vulnerability to crises, lack of political voice, access to formal and informal social safety nets, and lack of access to urban services (see DFID, 2002; INCIS, 2005).  10 analysis towards an investigation of the conditions of the present through the framework of governmentality entailed the following research methods. First, my research into the ‘history of the present’ for Jakarta’s urban water supply involved examination of over 100 years of the development of Jakarta’s urban water supply infrastructure. Out of practical necessity, an interest in both the past and present requires the use of different research methods, guided by availability of sources and feasibility of data collection. Over the duration of 2003- 2007, the process of research involved different combinations of archival investigation (reliant upon primary and secondary sources), and primary research consisting of in-depth interviews, quantitative surveys, and focus group discussions. The particular amalgam of methodologies used was divided according to four different time periods: • 1870-1950: archival investigation; • 1950-1990: archival investigation supplemented with interviews; • 1990-2004: investigation into World Bank project and privatization relying on primary and secondary documentation, interviews, media analysis; • 2004-2007: ethnographic investigation into issues around urban water supply for the urban poor involved quantitative and qualitative methodologies (surveys, focus group discussions, participant observation, forums, in-depth interviews). In the interests of disclosure, I note that data gathered in 2007 was facilitated by my employment as a water supply project advisor for an international NGO. More information on the research chronology, methods, sources, and timeline is in Appendix 3, a list of interviews conducted over 2004-2007 is provided in Appendix 4.  The research itself was undertaken in a chronological sequence, as I traced the development of urban water supply infrastructure in Jakarta from past to present. Although my interest in Jakarta was initially piqued by the ‘conditions of the present’ - the privatization of Jakarta’s water supply in 1997 and subsequent protests (see Harsono, 2003)-  I began the research project at the origins of Jakarta’s urban condition, the construction of the city’s first urban water supply system in the 1870s. Archival research into the colonial era was conducted in the Netherlands in the summer of 2003, with analysis continuing into 2004. This archival study of the colonial era was then followed by four months of research in Jakarta in 2004 (March; July-October). During these months I used the archives and resource centers of multi-lateral development banks, government departments, aid organizations, national NGOs and Indonesian research institutes, supplemented with interviews, to trace the development of urban water supply infrastructure from 1950-1990. From 2005 until 2007 I conducted the third and fourth phases of the research, tracing events from 1990-present by investigating the World Bank development project of that decade and the subsequent privatization, and the current impacts for  11 and responses of the urban poor. The following paragraphs provide more detail on the specific methods, objectives, sources, locations, and limitations for each of the four research phases. As already stated above, the first phase of research documented urban water supply infrastructure constructed throughout the colonial era, when the city of Jakarta was still the Dutch colonial capital of Batavia8. Archival sources were used to document both the physical development of the piped water supply system (1870s-1940s), as well as the rationalities that initiated, problematized, and enrolled urban water supply as a technology of government. The gathering of empirical evidence and discourse analysis relied to a large extent on engineering documents, many of them published in the official journal of government water engineering (Waterstaats Ingenieur), now housed in the archives of Delft University of Technology. In addition to the feasibility studies, technical plans, and discussions recorded by the engineers in charge of designing the system in Batavia (and similar systems in Surabaya, Bandung, Semarang), the colonial archives of KITLV and KIT provided: records of minutes of the Municipal council meetings in which issues of water supply were discussed; documents on urban planning and municipal regulations published by the Municipality of Batavia (Gemeente Batavia); other primary documentation of the conditions of life in the city, as recorded by residents, local government, “concerned citizens’, and tourists. Dutch language lessons in the spring of 2003, and employment of a native Dutch speaker enabled translation of archival documents from Dutch to English language. All Dutch language documents cited in the thesis are original translations. The second phase of research constructed the chronology of Jakarta’s urban water supply infrastructure development from 1950-1990. This phase of research was also heavily reliant upon archival sources, with both primary and secondary texts recording the history of Jakarta, urban government regulations, and records of life in the city post-independence. Indonesian government records on the operation of the urban water supply prior to the late 1980s were difficult to locate, and where possible interviews with key informants were used to supplement the incomplete archives of national and municipal government departments and agencies (PAM Jaya and Public Works). In-depth interviews were conducted with key personnel in the water sector, government departments, and MDBs who were involved in urban water supply projects during this time period in Jakarta. The entry of multi-lateral development banks and international aid agencies into the water sector in the late 1960s provided empirical evidence relating the construction of the urban water supply systems (from feasibility studies, master plans, and infrastructure projects funded by the Japanese and Dutch governments and the World Bank). Documents were found in offices and resource centers of MDBs, and bi-lateral aid agencies financing the studies and construction of the water supply system, and  8 Established in 1619, the colonial city of Batavia was under the control of the Netherlands East Indies until the invasion of the Japanese army in 1942. From 1945-1949 during Indonesia’s war for independence, the city was nominally under the control of the Dutch. Following Indonesian independence in 1949, the city changed to the present name of Jakarta. See Abeyaskekere (1989) and Cobban (1970) for a complete overview of the city’s historical development.  12 engineering firms and consultants who designed and constructed the systems (Degremont, Nippon Keoi, Amsterdam Water Supply Company).  The third and fourth phases of research were conducted simultaneously, over a three year period in Jakarta (2005-2007). First, following the chronology of Jakarta’s urban water supply infrastructure development, I traced the construction of the network from 1990-2005. This involved two major events, the World Bank’s PJSIP project (1990-1998), and the subsequent privatization of Jakarta’s water supply (1997-present). The physical development of network occurring over this time period, and the effect on access for particular socio-economic and socio-political classes of residents, were investigated through interviews, media analysis, and primary and secondary documentation. Documents were collected from World Bank offices, JICA office, consultancy agencies who did feasibility studies etc (IWACO, NERA), the Jakarta Water Supply Regulatory Board, PAM Jaya, and the two private sector partners at the time (Thames Pam Jaya and PAM Lyonnaise des Eaux)9. Interviews were conducted with key informants in World Bank, JICA, independent consultants and engineers working on the project, Jakarta Water Supply Regulatory Board, PAM Jaya, Palyja, TPJ, local professional water supply associations (PAM Jaya union, Forkami, Akaindo), and local NGOs who mobilized against the privatization (KruHa, INFID, Amrta, Walhi, YLKI). A list of interviews is presented in Appendix 4. The fourth component of my research was an examination of the impact of the World Banks 1990-1998 PAM Jaya System Improvement Project project and subsequent privatization on access to water by low-income urban households. However, although I originally arrived in Jakarta with the expectation that events of privatization would be of negative impact for access to urban water supply for low-income households, engagement with local NGOs working with the urban poor in Jakarta, and dialogue with members of these low-income community members broadened the scope of research into a wider investigation into how preferences and patterns of water use connect in complex ways to urban space and identity. This in turn brought to light the relationships between different kinds of residents (socio-economic class, place of residence and employment, legal status, land tenure status) and various types of urban water supply in Jakarta which were not represented in the official chronology of centralized network infrastructure development. Subsequently, through collaboration with a local NGO, the Urban Poor Consortium (UPC), I conducted qualitative research into perceptions, preferences, and patterns of water access and use in low-income neighbourhoods, particularly in communities where residents were not legal in the eyes of  9 PAM Jaya is the municipal water supply company (Perusahaan Air Minum Jakarta); Thames PAM Jaya (TPJ) is the private sector operator now serving the eastern half of the city, originally owned by Thames Water International but sold in 2006 to Singapore based consortium, Aquatico. PAM Lyonnaise des Eaux (Palyja) is the private sector operator in the western half of the city, owned by French multi-national ONDEO.  13 the state, or had no formal ownership of land and/or housing10. A series of visits over a period of six months, conducted in eight communities across three municipalities in Jakarta11, provided opportunity to conduct focus group discussions and in-depth interviews, particularly with women, who were responsible for the household’s water supply12. A map of the research sites is found in Appendix 5. The selection of these communities was determined primarily by their prior engagement with UPC. The communities were largely characterized by their illegality (although not in all cases), having been previously or currently involved in struggles over evictions, resettlement, and land rights. During the duration of the field research none of the communities were engaging in any specific political action over land, nor were they threatened with immediate eviction, although in late 2007 almost 4,000 households of a UPC community in Kelurahan Penjaringan, North Jakarta were evicted from their settlement under the toll-roads13. To mitigate against too much of my own interpretation being ‘read into’ the responses of community members, information from interviews and focus group discussions was cross-referenced between communities, with other staff of UPC, with a field research assistant, and with primary data collected from other local and international NGOs (PPMA, LP3ES, Mercy Corps, AcF). Language training in bahasa Indonesia and the employment of a field research assistant assisted this period of field research. Unless otherwise indicated, all Indonesian language translations into English are original. In addition to the qualitative research conducted in low income neighbourhoods, a quantitative household survey was conducted. This survey covered 110 households in four of the five municipalities of DKI Jakarta (North, Central, East, West), documenting water use patterns and preferences. The households were selected by random sample from a list of urban poor communities who were the recipients of government funded pro-poor water supply programs. Some of the data from this survey is  10 Illegal residents of Jakarta are identified by the state as not having proper identity document (KTP-Kartu Tanda Penduduk) granting government permission to reside in the city. The total number of illegal residents living in Jakarta is undocumented, as this population is excluded from census and elections. From 2005-2006, the government documented an average of 150,000 unregistered migrants (BPS, cited in The Jakarta Post, 6 October 2007); and according to Jakarta’s Social Welfare and Self-Development Agency, the number of ‘illegal residents’ captured in population raids annually averages at 15,000 (The Jakarta Post, 22 October 2007). Others have estimated that 21% of the city’s residents are not properly registered with the Population and Civil Registration Agency (The Jakarta Post, 31 October 2007). Distinct from the category of illegal residency, illegal settlement in Jakarta is synonymous with the classification of ‘squatters’, residents who live on parcels of land which is either legally prohibited (alongside riverbanks, along railways, under toll-roads, on flood plains), or living on publicly or privately owned empty land for which they do not have legal land tenure. 11 As the largest urban center in the country, the capital city is a special provincial region (DKI-Daerah Khusus Ibukota), consisting of five municipalities (North Jakarta, Central Jakarta, East Jakarta, West Jakarta, South Jakarta). 12 As documented in Dian Desa (1990), and Waspola (2007) there is a strong division of labour between genders. While the small-scale water supply industry of vendors, hydrant operators, bottled water vendors are almost all solely male dominated, women are primary decision makers concerning the selection, treatment, and use of water inside the home. 13 According to data from the Urban Poor Consortium (UPC), in August 2007 4,646 families (18,584 people) were living under the 11-kilometers of elevated roads stretching from Tanjung Priok to Penjaringan, North Jakarta. All were evicted in October 2007.  14 presented in Chapter Six, and published in entirety in Kooy et al (2006); a map of the communities surveyed is in Appendix 5, and the household questionnaire used is in Appendix 6. Although a survey of 110 households in a city with an estimated twelve million residents is of limited scope, the data is cross-referenced and found to be very similar to other data sets generated by NGOs and development agencies (AcF, 2007; Kooy et al., 2007; PPMA, 2007; USAID-ESP, 2007b; Waspola, 2007). Finally, throughout 2007, qualitative and quantitative data was collected through participant observation of the implementation of pro-poor water supply development projects. This period of research also involved participation in urban water supply development forums, and on-going discussion with the private sector water supply providers, MDBs, bilateral aid agencies, international and local NGOs in regards to ‘water for the poor’ in Jakarta. To conclude this section of the Introduction, I draw attention to the ways in which the above outlined research methods, sources, and timeline of research activities all contributed to shaping the ways in which my conceptual claims around the inter-relationships between waters, spaces, and populations have been explored in the subsequent Chapters. In this discussion I note the ways in which the above outlined research methods structured my analysis of co-constitution of waters, spaces, and populations, but I leave a more detailed discussion of the limitations and avenues for future research for the concluding Chapter of the thesis (Chapter Seven). My intention here is to articulate the ways in which this thesis content has been informed by methods and sources, acknowledging the other possible avenues which could have been, but were not, pursued in this particular research project. I first call attention to the ways in which the genealogical analysis implied by governmentality led to a reliance on particular research sources, and how this informed – and limited – my focus on particular sources of water supply (piped), and subsequently privileged my attention upon particular residents of the city. As detailed above, my analysis has derived from a reliance upon source documents that were written down, and written down primarily by a particular sub-set of colonial and postcolonial rulers (the largely European middle class who made up government bureaucracy, and elite Indonesian citizens enrolled within postcolonial bureaucracy). Using documentation derived from rulers, who decided what aspects of urban water supply, government, and urban development were relevant to document, led to a particular bias upon the importance of piped water supply, and largely omitted the voices, agency, and interests of residents other than these elites, including the ways in which they themselves articulated identity in relationship to other types of water sources. Indeed, living only on the margins of archival documents, non-elite subjects often only come to life when abstracted into populations who are defined by government as problematic, and whose identities, behaviours, and relationships with waters (as defined by authorities) required intervention. As a result of the bias in research sources available in archival records, my analysis of the relationships between urban waters, spaces, and populations has been influenced most heavily by the account from ‘above’, rather than ‘below’. In turn, this has resulted in an overt focus on the role of the  15 piped network water supply over time, and the relationship of populations to this source of water supply was given precedence over other possible avenues of research which would have examined in more detail the relationships between production of urban spaces, populations, and non-piped water sources. Following from this focus on the role of networked water supply within my examination of relationship between waters, spaces, and populations, the use of archival sources and documentation by elites also limited my research on the constitution of subjectivity in relationship to the material environment to a focus on the subjectivities of elite populations. This focus derives from the simple fact that certain populations received more detailed documentation in archival sources. In particular, for discussions of urban water supply in the colonial city, archival documents pay significant attention to relations between water supply and the European and native races, but did not distinguish between the different indigenous ‘native’ ethnic categories that today constitute the Indonesian (Javanese, Sundanese, Balinese, Madurese et cetera14), and did not accord proportionate attention to relations between urban water supply and the Chinese or what was termed the ‘Foreign-easterner’ (Indian and Arab).  Therefore, analysis of the co-constitution of subjectivity and urban infrastructure (Chapter Five) is driven by the availability of archival material and follows the priorities of colonial and postcolonial governments who gave attention to particular groups of subjects who held privileged identities as the ‘European’, or the ‘developed’ and ‘modern’ citizens within the (post)colonial city. In addition to the implications of relying upon source documents from the colonial and postcolonial rulers, my focus on the particular sub-set of engineering documents which chronicled the development of the piped water supply resulted in an overt emphasis on the role that technologies and technical aspects of urban water (L/capita/day, numbers of household connections, L/s produced) played in relation to the formation of urban spaces and populations. This emphasis on technical aspects of urban water supply occluded equal focus on the non-technical aspects of urban waters (cultural, social, ecological, and political properties of waters) and the ways in which relations between urban waters, spaces, and populations – and the negotiation of relations around piped water supply -- were also informed by these relations of power. Finally, I note that my reliance upon source documents from colonial and postcolonial authorities, and subsequent emphasis on the role of piped water supply and technical aspects of its development was exacerbated by the particular chronology structuring the timeline of research activities I undertook. As outlined earlier, it was not until the last phase of my research when I conducted field research within low income neighbourhoods in contemporary Jakarta that I ‘discovered’ a population of disadvantaged yet non-thirsty poor. Therefore, by beginning my research by tracing the origin and subsequent development of Jakarta’s piped water supply network from 1870s- 1990s, and largely reliant upon engineering documents for these development details on network  14 See Castles (1967) for a historical account of urban population and ethnicity in Jakarta.  16 expansion, it was not until I began my final phase of research that I became aware of the noticeable gaps in (post)colonial water supply development chronologies (which omitted non-piped water use). This discovery did prompt revisiting of the empirical documentation of the chronology of Jakarta’s urban water supply infrastructure development to consider how and why colonial engineers and government officials, development experts, and Indonesian government documentation all omitted documentation of the multiplicity of urban water supply networks, but it was not possible to revisit the colonial and Indonesian archives to search for different sources of research material. Therefore, although increasingly aware of the discrepancy between ways in which subjects articulated identity in relationship to waters and the ways in which government problematized populations in relationship to waters, this was only able to explored within the present day. A more iterative undertaking of research, shifting from past to present, or beginning in the present city and moving backwards through time instead of beginning in the colonial city and moving forwards, would have guided me towards different sets of research sources and archival materials. In conclusion, although I have delineated here the limits to my analyses of relations of power, water, space, and populations in (post)colonial Jakarta, I stress that I am not creating an apologetic for the following Chapters. The conceptual triad of waters, spaces, and populations as applied to an examination of the conditions of the present piped water network in Jakarta did elucidate the claims made in the above section, and, I argue, provides a particularly productive explanation for the patterns of provisions and conditions of access around piped water in the contemporary city.    1.4 Chapter Outline  Subsequent chapters of the thesis present the genealogy of Jakarta’s urban water supply infrastructure development, a narrative which seeks to both make visible the contradictions within Jakarta’s urban water supply (ignored or obviated by developmentalist and splintering urbanism analyses), and explain their existence as the physical, spatial, and discursive product of relations of power.  The structure, and order, of the chapters themselves follows from the methodological implications of the framework of governmentality; specifically, Chapters Three, Four, and Five each foreground a particular element of the historical, political economic, and discursive analysis. Chapter Six follows this with a focus on the problematic, and contested, construction of Jakarta’s ‘thirsty poor’. The paragraphs below provide an outline of the contents of each of the following six chapters, their relationship to each other, and a rationale for the particular temporal structuring of the material.  In the following chapter (Chapter Two), I present the theoretical framework I use to understand the current patterns, and paradoxes, of urban water supply in contemporary Jakarta. Utilizing  17 Foucault’s analytic of governmentality by bringing it into conversation with postcolonial critiques, and insights from the production of social nature, I articulate how an understanding of the iterative relationship between materiality and governmentality can explain current characteristics of urban water supply in Jakarta, provide insights into conditions of (post)coloniality in cities of the Global South, and contribute to understanding of relations between power-nature-society. I explain how the analytic articulates ways in which relations of power mobilized within government link production of natures, spaces, and populations; and how subjectivity is constituted in relation to material practices and physical environments. Highlighting the inscription of relations of power with physical space and ecological processes, as well as discursive categories and socio-economic relations presents an understanding of how relations of colonial power continue to affect conditions within contemporary cities. Following this conceptual outline, I conduct a historical analysis of Jakarta’s urban water supply through a genealogy of the city’s water supply infrastructure. Chronicling the development of Jakarta’s urban water supply infrastructure from 1870 (the origin of the physical construction of the city’s first piped water supply system) until 1990, Chapter Three illustrates how the piped water supply system was splintered since its inception and demonstrates that the universal provision of clean water through centralized piped network was never a goal of (post)colonial governments. The genealogy of Jakarta’s urban water supply development is continued in Chapter Four, where I continue the chronology of the city’s piped water supply network development from 1990 until present day. Examining the largest international development project to date for Jakarta’s urban water supply (PAM Jaya System Improvement Project) in the 1990s, through privatization, until present day (1990-2007), the chapter documents how and why the two most recent development interventions led by both public and private sector management have exacerbated fragmentation, and failed to improve equitable access. This section of the genealogy of governmentality of urban water in Jakarta combines historical analysis with discourse analysis, and brings a specifically political economic focus on the last two decades of infrastructure development which were rationalized in terms of neo-liberal economic policies. In addition, in this Chapter I foreground the ways in which both public and private sector development failures were premised upon a misunderstanding of the pattern of network development from 1870-1990, in particular the parallel processes of centralization and fragmentation and the resultant growth, over time, of non-piped water supply sources in the city. Following the completion of the chronology of Jakarta’s water supply network development documented in Chapters Three and Four, in Chapter Five I review and extend this time period of 1870- 2007 to conduct a discursive analysis. The discursive analysis in this Chapter highlights the inter- relationships between governmentality and materiality, as I document the ways in which infrastructure was used to configure – and contest - particular natures, spaces, and populations. Extending the timeline followed in Chapter Three, in this Chapter I read the chronology of Jakarta’s water supply  18 infrastructure development alongside the shifts in colonial and postcolonial government which consistently preceded problematizations of urban water supply, and led to particular technical solutions as political rationalities were, literally, ‘rendered technical’ within physical technologies of water supply production, treatment, and distribution. In this Chapter I give particular emphasis to the origin of the governmentality of Jakarta’s water supply, and describe the ways in which relations around waters preceding the city’s first centralized water supply system became problematic to rule over the course of the 19th century. In Chapter Six I complement the colonial focus of Chapter Five by examining the implications of the co-constitution of water, spaces, and populations within the contemporary city; I explore in particular the problematic and contested construction of Jakarta’s ‘thirsty poor’. With a specific focus on how relations of power co-constituting this triad of water, space, and populations are resisted,, I discuss how the patterns and preferences of water supply associated with the urban poor (and ascribed as ‘ignorant’ and ‘undeveloped’) are responses to relations of rule, and identify how they are contributing to the progressive fragmentation of Jakarta’s splintered network, and frustrating current pro-poor water supply development initiatives in Jakarta. Finally, Chapter Seven presents a summary of the arguments of the thesis, identifies limitations of the research and analysis, and highlights directions for future research. Not surprisingly, most of the limitations of the conceptual framework and the acknowledged boundaries of research and analysis suggest interesting avenues for future work. I limit my discussion in particular to issues of scale and subjectivity, concluding with thoughts on how a more detailed exploration of these issues carries implications for further development of the analytical framework of governmentality.   1.5 Conclusion: Imagining Jakarta  The politics of water supply in Jakarta are imminently material: the consumption of contaminated water explained the death of seventeen low-income residents in June 2005 and the death of seven residents from another community in November 2007 (Jakarta Post 27 June 2005; Jakarta Post 29 November 2007). Across Indonesia, nineteen percent of the deaths of children under five are caused by diarrhoea due to consumption of poor quality water (WHO, 2005). With over 80% of the shallow groundwater sources in the city contaminated with e-coli and/or heavy metals, and water borne and water related diseases such as cholera, dengue, polio, leptospirosis, and chikungunya on the rise, a concern for lived geographies of inequity demand an analysis able to engage with “material practices, actual spaces, and real politics” (Yeoh, 2001:457). I argue that the analysis deployed in this thesis is imminently relevant for the current material lived geographies of inequity: highlighting the materiality of governmentality attends to the effects of relations of power within socio-economic, spatial, and  19 discursive processes, and is thus able to make visible the continuing relevance of past relations of power within the contemporary city in ways that speak to current politics of the present (Barnett, 1997; McEwan, 2003; Yeoh, 2001). Part of this politics is making visible what is currently erased from the developmentalist narratives and neo-colonial chronologies of development in Jakarta: disrupting the dominance of the western hydraulic discourse and acknowledging other trajectories of urban development, identifying the failures of development models based on the western hydraulic discourse, highlighting the culpability of the state and international development actors, and recognizing the agency of low income households who make very rational, informed choices opting out of piped water supply. Making visible that which is currently erased or obviated in mainstream discourses of urban development in Jakarta ultimately seeks to contribute to new ‘imaginings’ of a more equitable Jakarta, echoing the visions of the Indonesian architects, planners, artists and academics who were ‘Imagining Jakarta’ in new ways in 2004, illustrating new possibilities for the structure and function of urban space, and suggesting more equitable relations of engagement between residents, infrastructure, public space, and the built environment (Effendy and Kusumawijaya, 2004). However, to make possible the articulation of new kinds of identities in relation to water (the political poor, the rational and strategic low income household), the production of more inclusive urban spaces, and the more equitable flow of urban water supply, it is first necessary to break with the prescribed script of the dominant discourses of developmentalism and failed modernization and suggest new ways of seeing and understanding (Yeoh, 2001). This is the project to which I now turn.   20 Chapter 2  Govermentality: Governing (post)colonial natures, spaces, and subjects   2.1 Introduction  This chapter presents the theoretical framework I deploy in subsequent chapters to analyze the current patterns, and paradoxes, of urban water supply in contemporary Jakarta. The analysis is grounded within Foucault’s theory of governmentality (Foucault, 1978, 1991), using his analysis of the relations between the production of governmental rationalities and the technologies of modern power to examine the specific conditions under which Jakarta’s urban water supply network emerged, developed, and continues to exist. Foucault’s original concept of governmentality is however revised in reference to postcolonial critiques of the contradictions, contestation, and difference in the operation of relations of rule. Insights from analyses of power-nature-society are also read through governmentality to draw attention to ways in which power works through the environment to link the production of nature with space and subjectivity. At the core of my analysis is a conceptual triad - nature, space, and subjectivity – through which I interrogate the mutually constitutive relationship between materiality and governmentality. I argue that an emphasis on the materiality of governmentality is of both of practical and conceptual use: it makes visible the relationships between urban governance and urban infrastructure in Jakarta by identifying and explaining the contradictions within patterns of urban water supply, and it foregrounds the physical environment (and concrete infrastructures) as a field of power through which socio-natures are produced, but also productive of, rationalities of rule.   Building upon the work of scholars who have already initiated conversations between governmentality, postcoloniality and relations of power-nature-society (Agrawal,, 2005; Braun, 2000; Goldman, 2005; Li, 2007; Moore, 2005), in this chapter I explain the co-constitution of the triad of (urban) nature, space, and subjects as the product of shifting colonial and postcolonial governments, who through the ‘conduct of conduct’ effected (consciously and unconsciously) particular physical, spatial, and discursive relations between urban waters, residents, and urban spaces. The core of my analysis seeks to explain how relations of power mobilized through government have been productive of networks of urban water supply that were (and still are) implicated in both the discursive and material production of the city and its citizens; I do this through exploring two particular relationships, that between governmentality and materiality, and between relations of power and resistance. Articulating the iterative relationship between governmentality and materiality, my analysis highlights  21 the role of material objects and uncooperative natures in constituting or resisting discursive categories, and in facilitating or frustrating relations of rule. The materiality of governmentality is also, I argue, a useful contribution to the theoretical and political projects of (re)materializing postcolonial studies (Harris, 2004; Jackson and Jacobs, 1996; Yeoh, 2001), and constructing more ‘cosmopolitan’ conceptual frameworks for understanding urban development in the Global South (Amin and Graham, 1997; Robinson, 2002, 2006). When governmentality is reworked through the analyses of postcolonialism and power-nature-society, it provides a provocative and flexible theoretical framework for exploring the nature(s) of the postcolonial, particularly in urban areas. First, recognition of the ways in which relations of power work through both discourse and material practice to co-constitute identity and infrastructure uncovers some of the ‘hidden ways’ in which the colonial past shapes conditions of the present (Gandy, 2002). Second, when combined with the concept of ‘layers of relations of rule’ – a phrase used by Li (2001) to describe the layering of social, cultural, and economic relations of power that do not simply disappear when government rationalities change, but continue to generate new, and often unpredictable, relations – attention to the materiality of governmentality can highlight the ways in which these layers of relations of power are embedded within physical space and ecological processes as well as social- economic relations and discursive categories. Acknowledging the continuing material effects of past relations of rule respects current lived geographies of inequity, and provides a particularly productive explanation for the contemporary contradictions within Jakarta’s urban water supply. Prior to reviewing the initial development and further applications of governmentality as a field of study, I emphasize that while the theoretical framework I articulate in this Chapter builds upon the foundation of Foucault’s writings on governmentality, I go beyond his particular genealogy of the liberal state. Drawing from Foucault his analytic of governmentality, to examine the emergence, existence, and changes in government, I construct a theoretical framework applicable to different articulations of rule, outside of the West. This is given further explanation in the following section. Accordingly, while scholars from a variety of disciplines have used and revised Foucault’s original writings on governmentality, this chapter is limited to a discussion of the ways in which the analytic of governmentality has been extracted and extrapolated by scholars to analyze conditions of (post)coloniality (Legg, 2007; Li, 2007; Watts, 2003) and articulate relations between power-nature- society (Agrawal, 2005; Braun, 1997; Goldman, 2005). Although many scholars have provided cogent critiques of what Foucault’s analyses of relations of power, and his original outline of governmentality does – and does not – provide adequate explanation for, in particular the social and economic bases of state power and processes of capital accumulation (see Braun, 2000; Li, 2007; MacKinnon, 2000; Rose-Redwood, 2006), I do not address all of these critiques toward a project of advocating the universal utility of governmentality.  22 Instead, following the scholars of postcolonialism and relations of power-nature-society mentioned above, I focus more specifically on what a framework of governmentality does usefully elucidate – that is, when it is revised to include an understanding of the operation of government beyond the state, and through the production of social nature, and when in conversation with postcolonial criticism. Theoretically, I argue that it draws attention to the co-constitutive relations between nature, space, and subjectivity, and highlights the ways in which subjectivity is constituted in relation to the physical environment, and performed through material practices that resist or reify discursive categories. Practically, I argue that the analytic makes visible the relationships between urban governance and urban infrastructure. In Jakarta, this allows us to identify, and explain the contradictions within contemporary patterns of water supply in Jakarta, and offers a different way to understand, and thus to affect, possible futures. The above apologetics for how I understand and employ governmentality does not however meant that I am not attendant to the criticism it has received. In the following sections of this chapter I review the ways in which governmentality has been revised and extended by scholars within postcolonial studies, geographers, and other scholars who analyze relations between power-nature- society. The remainder of the chapter proceeds as follows. I begin by outlining Foucault’s original articulation of governmentality as marking the emergence of a distinctly new form of thinking about and exercising of power that arose in Western Europe over the 16-18th centuries. I then discuss the ways his work has been both contested and extended, reviewing the work of scholars who have extracted the analytic of governmentality and applied it to understand the operation of relations of power within those states, spaces, subjects, and substrates excluded from Foucault’s original analysis. Through this discussion, I follow how the application of governmentality within postcolonial and power-nature-society studies has provided a more complex understanding of relations of power within government - its different modalities, uneven effects, and operation outside of borders of nation-state. In particular, I highlight how these two areas of study (often in combination) have been used to articulate relations of power and resistance, and elucidate the iterative relationship between governmentality and materiality. In the conclusion, I outline the conceptual framework provided by the conversation between governmentality, postcolonial studies, and analysis of relations between power- nature-society, which I use in subsequent chapters to articulate relations between urban water, space and subjectivity within the (post)colonial Jakarta.       23 2.2 Governmentality: From Foucault and Beyond  In this section I outline Foucault’s original conceptualization of governmentality within its origins as a genealogical project of the history of the modern liberal democratic state. Setting aside the (mis)conceptions of governmentality as an evolutionary progression to a superior form of rule (Harris, 2004; Li, 2007), I employ a conceptual framework of governmentality to analyze how different modalities of power are mobilized, and operate, to what end, through projects of rule seeking to shape conduct for definite but shifting ends, with a diverse set of unpredictable consequences, effects and outcomes (Dean, 1999). The concept of governmentality was originally used by Foucault to describe what he identifies as the emergence of a distinctly new modality of rule, that of government, or the ‘conduct of conduct’ (Foucault, 1978). Stemming from Foucault’s interest in the development of disciplinary power, biopolitics, and technologies of subjectification (Foucault, 1983, cited in Rose-Redwood, 2006), and emerging out of an intellectual project that sought to highlight the contingency of the present day liberal democratic state in western Europe, Foucault’s articulation of governmentality resulted from his investigation into the broader mechanisms, techniques and technologies by which power came to be enforced in modern Europe (Legg, 2005). In particular, Foucault sought to demonstrate how a centralized state and its apparatuses (legal, administrative, security) came to make the fostering of life and the care of population a part of a new regime of power that brought “life and its mechanisms into the realm of explicit calculation” (Foucault, 1978:143, cited in Watts, 2003). To this purpose, Foucault elaborates how a new form of modern power emerged gradually over the 16th and 17th centuries in Europe. Following a transformation in ideas about rule and questions over the nature of government that saw the waning of the power of the sovereign, whose right was to ‘take life or let live’, Foucault describes the emergence of government as a new form of modern power that replaced, but did not eradicate, sovereignty, as the ‘right to make live and let die’. Government is thus identified as a significantly different field of institutionalized power than its predecessors of pastoral and disciplinary power. Not concerned with, or secured by, one dogmatic goal, such as the securing of territory, or the granting or taking of life, government takes both the purpose and means of power to be the optimization of life, including non-human life (see Darier, 1998). The purpose of government is therefore to secure a whole host of ‘finalities’ that are identified as necessary for achieving the optimization of all of life; seeking to ‘optimize the conditions of all of life’, government is defined as “any more or less calculated or rational activity, undertaken by a multiplicity of authorities and agencies, employing a variety of techniques and forms of knowledge that seeks to shape our conduct” (Dean, 1999:209). As the means of securing the validity of government and achieving its purposes was arranging the ‘correct disposition between men and things’, or population and territory (Foucault, 1991), Foucault  24 traces the emergence of new ways of thinking systematically about the relations between population and territory as an arena of intervention. This ‘mentality’ of government, how to conduct the ‘conduct of conduct’ was termed ‘governmentality’. Somewhat confusingly, the term is used by Foucault to describe the specific historical emergence of this ‘mentality’ in Western Europe, but is more widely employed as a more general description of the ways of thinking about how to conduct the ‘conduct of conduct’ of life, in a wide variety of geographical and historical contexts (Dean, 1999). Foucault went on to explain how the new purpose of rule as government in turn required new modalities of power. The organization of all of life relies on the operation of more diffuse mechanisms of power, so that the ‘conduct of conduct’ could be achieved for an entire, dispersed population. Distinct from disciplinary power, which reformed individuals through detailed supervision in confined quarters (Foucault, 1977), the concern of government is the well-being of populations at large. Shifting from the level of the individual (disciplinary power/surveillance) to manage the population required distinctive means, as it is not possible to coerce every individual and regulate their actions in minute detail. Rather, government relied on a new field of power which enabled the regulation of the conduct of inhabitants of a territory through a distance, and through diffuse mechanisms (Brown, 2000; Joyce, 2003; Rose, 1999). As Foucault (1991) and other scholars have illustrated, this both relied on and led to the production of new forms of knowledge and calculation (in particular statistics), which brought new domains of population, economy, and territory into the realm of calculation (Braun, 2000; Hacking, 1990; Miller and Rose, 1990; Rose-Redwood, 2006). This new field of power operated through the exercise of tactics and the construction of knowledge, rather than the imposition of law (Foucault, 1991). Thus, power is exercised through the construction of certain truths and their circulation via normalizing and disciplining techniques, methods, discourses and practices that extend beyond the state and stretch across the social body. Operating to “educate desires, configure habits, aspirations and beliefs”, setting the conditions, “artificially so arranging things so that people, following only their own self-interest, will do as they ought” (Scott, 1995:202, cited in Li, 2007:5), power mobilized within government thus acts diffusely to shape the conditions in which lives are lived.   Exercised through complex assemblages of discourse and practice, government is therefore both material and discursive – with a “complex array of apparatuses of knowledge production” leading to physical and discursive technologies of directing conduct (Rose-Redwood, 2006:272). This understanding of the relational, diffuse, and productive natures of power highlights how government enables as it constrains or coerces, working through material practices and discourses that are for the most part, mundane and routine (Li, 2005). In this way, we can understand how the techniques of government do not act or appear as an external imposition; as the natural expression of the everyday interactions of individuals and groups, “they blend seamlessly into common sense as power works to educate desires and configure habits, aspirations and beliefs” (Li, 2007:5). Of course,  25 the ways in which power operates to structure choices and educate these desires is not homogenous or uniform in application: it operates unevenly across differently situated subjects (Butler, 1996; Nightingale, 2006; Stoler, 1995), and operates unevenly across different spaces (Legg, 2006; Raco, 2003), and I argue, with different results upon different socio-natures (Bakker, 2006; Bakker and Bridge, 2006; Braun, 2000). I discuss the uneven application and effects of relations of power in more detail in the following section. Following Foucault’s articulation of governmentality, scholars in the West began to trace the history of changing forms of government authority and practice, focusing in particular on the emergence of neo-liberal rationalities (Barry et al., 1996; Dean, 1999; Dillon, 1995; Gordon, 1991; Lemke, 2001; Rose and Miller, 1992; Rose, 1996). Other scholars, while still focused on the operation of modern power in the West, have traced: how new forms of knowledge contributed to the rationality and tactics of government - specifically in production of insurance and risk (Ewald, 1991; Hacking, 1990), and systems of taxonomy and classification (Halfacree, 2001); the various ways in which government as a field of power operates through self-help programs (Cruickshank, 1996); the bio- political projects of census-taking and statistics (Brown and Boyle, 2000; Elden, 2001, 2003; Hannah 2000, 2001; Legg, 2005; Philo, 2001; Philo and Parr, 2000). The early inter-disciplinary studies of governmentality focused analyses on the operation of state power within the West (see Burchell et al, 1991; Dean, 1999). For postcolonial scholars, this provoked the question of whether or not the concept of governmentality was of any analytical utility outside of the West. Highlighting the relationality between colonial metropoles and territories, some postcolonial scholars even questioned the utility of governmentality as applied to the West (Cooper and Stoler, 1989; Stoler 1995). Since these questions were posed, scholars situated in places with much different histories of rule have taken up the framework of governmentality to examine the articulation and relations of institutionalized power outside of western liberal democratic states (Chakrabarty, 2002; Chatterjee, 2004; Goldman, 2005; Kalpagam, 2000; Li, 2007; Ong, 1999; Prakash, 1999; Valverde, 1996; Watts, 2003). Resuscitating Foucault’s original writings on the co-existence of sovereign and disciplinary power within government, postcolonial scholars in particular have cautioned against earlier misinterpretations of government. As stated by Foucault, “we need to see things not in terms of the replacement of a society of sovereignty by a disciplinary society and the subsequent replacement of a disciplinary society by a society of government…In reality, one has a triangle, sovereignty-discipline- government” (Foucault, 1991:102). Postcolonial scholars have helped to emphasize this fact that disciplinary mechanisms or sovereign power did not cease to play an important role in government, and rather than eradicating disciplinary power, scholars point to how government articulates new relations between modes of power for new modalities of rule (Li, 2007; Ong, 1999, 2006). Therefore, as illustrated by Li (2007) in the military regime of New Order Indonesia, the diffusion of power  26 introduced through government is not to be taken to mean that sovereign and disciplinary power were eradicated, as some have erroneously interpreted (Harris, 2004).  Rather, government introduced a new triad of modalities of power. Sovereignty, discipline and government operated with a new concern for “the population and its optimization –wealth, health, happiness, prosperity, efficiency – and the forms of knowledge and technical means appropriate to it” (Dean, 1999:20). Disciplinary and sovereign power were not therefore eradicated, but different modes of power were applied to differently situated populations, as ‘the forms of knowledge and technical means’ were considered ‘appropriate’ to it. Power does still discipline and control, as exemplified both within colonial regimes (Li, 2007; Mbembe, 2001, 2003; Arnold, 1994; Hussain, 2003; Mehta, 1999), and modern liberal democratic states who reserve this modality of rule for populations judged deficient (Cruickshank, 1999; Dean, 2001; Hindess, 2001; Mehta, 1997; Procacci, 1991; Valverde, 1996). Cautioning against a reading of governmentality that can be interpreted to present an “ethnocentric replay of modernization theory” (Li, 2007:12), scholars working outside of the West, and below and beyond the boundaries of the nation-state, have instead pointed to ways in which government has articulated different modalities of power within the practice of rule (Li, 2007; Valverde, 1996). Contemporary examples of the ways in which sovereign and bio-power are awkwardly combined – such as the oil regimes in Nigeria described by Watts (2003), and the transnational companies supported by western liberal regimes working with the military in Indonesia (Li, 2007) - further reject the concept of governmentality as the evolution towards a superior form of rule. Rather, postcolonial scholars illustrate that the combination of modalities of rule often observed within pre-colonial and colonial governments is not limited to the ‘pre-modern’ era and pre-liberal regimes. Following these scholars, I take up governmentality as a framework within which to analyze how power works (material and discursive), and what it does (production of nature, subjectivity, and space). In the next sections of the Chapter I review ways in which scholars in postcolonial studies and analyses of relations between power-nature-society have both critiqued and extended the analysis of relations of power as understood within the framework of governmentality.   2.3 Government of Nature, Space, Subjects in the Postcolonial  Within this section of the chapter I conduct an iterative dialogue between governmentality, postcolonial studies, and analyses of power-nature-society. I begin with reviewing the ways in which the latter have critiqued and contributed to the analytic of governmentality. Postcolonial scholars examining the practice of government have shown power mobilized within government to be composed of different modalities (disciplinary, self-regulatory); to be uneven in its application across  27 differently situated populations and spaces; to provoke resistance in ways that are productive of new relations of rule; and to contain inherent contradictions in the pursuit of diverse finalities. Following this, I discuss how scholars of the relationship between nature and society have taken up the substrate left unexamined by Foucault to emphasize how rationalities and technologies of government are both articulated in relation to, and worked through, material environments. In the third section of the chapter, I bring together the insights from both of these bodies of work to foreground the materiality of governmentality, suggesting how it contributes to rematerializing the postcolonial and urban geographies to understand ways in which layers of relations of rule inscribed in ecological processes and physical spaces still affect conditions with in contemporary cities of the Global South.   2.3.1. Postcolonial Governmentality  In the earliest dialogue between governmentality and postcolonial studies, scholars criticized conceptions of governmentality as a geographically limited genealogical project (only within the West), which used a totalizing conception of power (Legg, 2005; O’Malley, 1996; O’Malley et al., 1997). Postcolonial scholars sought to correct these notions by paying attention to difference in places, spaces, and subjects, and this section of the chapter outlines three particular contributions which this attention to difference has made to the understanding of governmentality. Specifically, postcolonial scholars have highlighted ways in which relations of rule are resisted or negotiated; the uneven application and effects of government; and the inherently contradictory and incomplete nature of government. First, postcolonial scholars have set limits to government by highlighting ways in which relations of rule are resisted or negotiated (Li, 1999; O’Malley, 1996; Thomas, 1994). Scholars have shown how, in practice, government rationalities were contested by active subject (and subjecting) populations, who always ‘retained the capacity to act otherwise’ (Li, 2007). Repositioning the ‘prickly subjects’ (Li, 2007) and uncooperative natures from passive recipients to active agents within relations of rule, scholars illustrate how the actions and/or responses of both human subjects and non-human objects of government alternately provoke, facilitate, frustrate, and/or rework relations of power in unpredictable ways (Braun, 2000; Li, 1999, 2007; Mitchell, 2002; Thomas, 1994; Watts, 2003). This has in turn highlighted the conceptual difficulty of divorcing the emergence of new problematics of government and the articulation of rationalities of rule from the contestation and contradictions which emerge through the practice of rule (Howell, 2004; Li, 1999, 2007; O’Malley, 1996). Attending to the continual and responsive process of constructing knowledges through which subjects are governed, and which they actively resist and reshape, scholars of the postcolonial emphasize the relationship between resistance and rationalities of rule – or, between power and agency  28 – within the framework of governmentality. Although some argue that as government technologies respond to, and are informed by resistance to technologies of rule they subsume their failures and in fact reinforce and extend the power of ‘experts’ (Ferguson, 1990; Mitchell, 2002; Rose, 1999), other scholars criticize what they identify as an exaggerated closure of resistance within relations of rule (Li, 2007; O’Malley, 1996; O’Malley et al., 1997; Stoler, 1991; Thomas, 1994). However, although those like Li (2007) and Rose (1999) reserve the practice of politics and resistance to rule as a limit to the analytic of governmentality, their work has acknowledged that “new thinking about how to govern arises not only from inspired ideas, but from the pragmatic observation of how things work out in practice.” (Li, 2007:19). Therefore, although differing in their agreement over whether or not the analytic of governmentality can adequately explain the ways in which populations are mobilized to resist and respond to relations of power, postcolonial scholars illustrate the ways in which the contestation, negotiation, and compromise of government as it is practiced to be integral to formation of its rationalities and technologies1. Second, scholars illustrating ways in which modalities of power work unevenly across different populations, spaces, and natures, have brought to the fore the uneven practice of government. Contra Foucault – power does not bleed homogenously across the social body. Paying more attention to differences in the effects and responses to relations of power, postcolonial scholars emphasize that power works differently across the social body. This highlights both the fact that rule is never complete, and is never evenly applied across all populations and spaces (Hindess, 2001; Howell, 2004; O’Malley et al., 1997). By paying more attention to the complicated realities of social location in which rule is practiced – how it was informed by responses of populations and natures - postcolonial scholars have highlighted how relations of power were applied unevenly across different populations, and how in the practice of rule not all bodies and spaces are incorporated into systems of rule in the same ways, to the same effects (Agrawal, 2005; Howell, 2004; Mehta, 1997; Thomas, 1994). From these insights into the partiality and unpredictable effects of the relations of power within government, scholars have shown how the performance of rule and its construction is always the result of a contested engagement (Howell, 2004), often resulting in compromises in order to secure the sense of order (Li, 1999).  The reality of resistance to rule by colonial populations usefully highlights what is true for all forms of government - seeking utopia, government can never be complete. Rather, given its contradictory aims, or what Foucault (1991) terms ‘diverse finalities’ scholars have emphasized government as a continual project being reworked upon and in relation to past relations of rule, at once both necessarily incomplete and inherently contradictory (Li, 2001; Moore, 2001; Thomas, 1994). Government is therefore seen as a perpetual, ongoing project that changes technologies according to  1 Work on relations of contestation and struggle in government have also been explored outside of postcolonial studies, for example in the work of Blomley and Sommers (1999) and Crampton (2003) who look specifically at cartographic technologies of government, and their appropriation by non-state actors such as community groups to suit their own ends.  29 shifts in the ways of thinking about the purpose of government (to discipline/control; to uplift and develop; to self-actualize), and importantly, responses to government by both differently situated subjects, and non-human natures. In addition, as government, as emphasized by Foucault (1991), seeks a ‘whole host of finalities’, postcolonial scholars explore these diverse finalities to illustrate how government is composed of multiple modalities. Tania Li (2007) illustrates how there was no unitary purpose to rule in the Netherlands East Indies, as the diverse concerns of government - to secure orderly rule, entrepreneurial profit, revenues to support the state apparatus, and native improvement – were often contradictory. Therefore, although not minimizing the fact that colonial rule was based on conquest (Mbembe, 2001), this was not the whole story. As other scholars also document, the liberal arts of government were not absent from the colonies but rather emerged in rough parallel to Europe (Cooper and Stoler, 1989; Stoler, 1995). Emphasizing this complexity of government –which combined different rationalities in awkward amalgamations – postcolonial frameworks of governmentality reject the chronologies of colonial government suggested by Harris’ (2004) teleological progression. Instead, analyses of postcolonial governmentality attend to the combination of modalities of rule (sovereign, discipline, government) that came to bear in different times and places. Through acknowledging that colonial government consisted of multiple modalities, and highlighting the multiple projects of government necessary in the aim to ‘optimize all of life’, postcolonial analyses point to how government does, and must, embrace contradictions and make compromises. Some ‘finalities’ of government are in tension with each other, and often productive of perverse results. As illustrated later on in Chapter Six, an example of these ‘perverse results’ of colonial government was the provision of water supply to native areas of the city; initiated as part of the general purpose of government to ‘raise up’ the native population, the program of ‘kampong water supply’ proceeded to set tariffs for water at twice the per unit cost paid for by European residents. Examples like this abound in the colonial context, and resist perceptions of rule as a coherent, complete project, which are presented in some studies of governmentality (see Ferguson, 1990; Scott, 1998). Instead, as illustrated by Cote (2003) and Goldman (2001), there is often competition or disagreement over the means and technologies of government – both between state departments and agencies, and between national and international institutions. Watts (2003) provides a clear example in contemporary Nigeria of how government is not always a coherent state project, and Legg (2005) illustrates how in colonial India resistance was not necessarily only generated in response to state power, but arose from tensions between different institutions of authority over the different rationalities of government. I discuss more in the following section how geographers and other scholars of power- nature-society also argue that government must be conceived of as going beyond the state to include non-state actors and institutions (see Rose-Redwood, 2006), and note that the framework of  30 governmentality should be adjusted to able to account for the tensions arising from diverse finalities negotiated between multiple state and non-state institutions of authority. In summary, work done by scholars within postcolonial spaces has thus set necessary limits to the analytic of power as it is sometimes understood within governmentality. Power is not totalizing or unrolled without resistance, power is not only located within state agencies but operates outside the nation-state boundaries, the state is not necessarily coherent, and government is always a complex process as it works towards ‘diverse finalities’ and thus must embrace contradictions within relations of rule. In addition to correcting the misinterpretations of the operation of power within government by their focus on the material effects of relations of rule (including resistance), postcolonial scholars have also addressed a consistent critique made against colonial studies and its over-emphasis on discursive relations of power. Charged with favouring cultural analysis of discursive categories and the politics of knowledge production over the lived material geographies of inequity, and accused of failing to connect critiques of discourse and representation to the lived experiences of postcoloniality, postcolonial analyses are seen by some as unable to address the material problems of the present day, much less that of the future (see Dirlik, 1994; Harris, 2004; Jacobs, 1996; McEwan, 2003). The discursive analysis of many postcolonial scholars is seen as simply unable to connect to the specific, concrete and local conditions of everyday life, leading to the ‘fantastic optimism’ overlooking the persistent material inequities (Jacobs, 1996; McClintock, 1992). Responding to this criticism, I argue the analysis provided within the framework of governmentality can redress this imbalance between discourse and practice. For, within governmentality, thought itself is conceived of as a material practice, linked to a complex array of apparatuses of knowledge production. Documenting how new ways of seeing and organizing population and territory led to the formation of particular subjects and spaces (Dean, 1999; Gibson, 2001; Hannah, 2000; Mitchell, 2002), scholars illustrate the imbrication of discursive and material relations of power. In addition, while attention to the practice of rule has highlighted the realities of resistance and uneven effects of power, scholars have also noted the ways in which colonial discourses (racialized, pathologized colonial populations) were materialized, and relied on physical practices. As such, scholars have demonstrated how discursive categorization of differences between European and colonial subjects was productive of physical spaces, material environments, and corporeal bodies (Howell, 2004; Legg, 2006), and led to the spatial ordering of imperial and colonial spaces (Kusno, 2000; Rabinow, 1989; Wright, 1991; Yeoh, 1996), and the production of colonial natures (Braun, 1997, 2002; Scott, 1998; Sivaramakrishan, 1999). Related to this, recent work on cities in the south has also emphasized the interrelationship between urban identities and the spatiality of colonial and postcolonial power and discourse (Bunnell 2002, Kothari, 2006; Legg, 2006; Myers, 2006), highlighting the co-constitution of space and  31 subjectivity (Legg, 2006). This work demonstrates the ways in which material things and spatial arrangements make a difference in the constitution of social relations, either reinforcing or contesting discursive categories of subjectivity. Discussion on the iterative relationship between material and discursive relations of power is continued below in the conversation between governmentality and analyses of power-nature-society.  2.3.2 Power-nature-society & Governmentality  As postcolonial scholars refined the concept of governmentality to include populations and non-western spaces notably absent from Foucault’s writing2, so another body of work has reworked governmentality in reference to the environment. Noting that government is defined as the establishment of ‘right relations’ between population and territory, studies of the relations between power-nature-society have shown that territory is not ‘mere substrate’ upon which relations of power unfold, but is itself a field of power both through which government operates, and is itself constituted (Darier, 1999; Braun, 2000; Luke, 1998; Rose-Redwood, 2006). In this section of the chapter, I review the contributions of studies of power-nature-society to studies of governmentality. Following this I suggest how they can be combined with the insights from postcolonial studies to emphasize materiality of governmentality, and then identify the ways in which this materiality re-conceptualizes the conditions of postcoloniality and urbanization in the Global South. First, one intersection between analyses of power-nature-society and governmentality is found within theories of the production of social nature. Attention to the various ways in which relations of power work both upon but also through material natures has emphasized the social production of nature as a socio-natural product of government (Braun, 2002; Rutherford, 2007). While this corrects the notion of a static, un-mutable nature existing outside of social relations as seems implied within Foucault’s conception of territory (Braun, 2000), it also offers analysis of how relations of power work through the physical environment to link the constitution of nature, space, and subjectivity, providing a conceptual tool for deconstructing the binaries of nature/society or material/discourse. Second, by identifying nature itself as an object of government (and thus enrolled within social relations and processes of socio-natural production), scholars have drawn attention to how the production of particular natures was integral to rule (Braun, 2002; Scott, 1998). As particular kind of natures were remade into something useful for the political and economic aims of emerging states (Bakker, 2000, 2002; Gandy, 2002, 2006; Kaika, 2006; Swyngedouw, 1999), certain natures were also necessary for the constitution of particular global, national or local subjects (global eco-citizens, indigenous natives, geological citizens etc), who would act or be in certain ways in relation to the  2 Legg (2006:711) refers to “Foucault’s almost total silence on issues of colonialism”, which he discusses further in Legg (2007).  32 physical environment, the state, institutions of authority, others, and themselves (Braun, 2000, 2002; Kaika, 2004; Scott, 1998). Tracing how nature was made amenable to rule, and the processes by which nature was constructed as a particular domain with specific uses and resultant relations between nature/society, scholars have also shown how ‘nature’ was both inserted with social processes (social nature), and then – importantly – productive of particular spatial relations and social processes. The relationship between the government of territory and population - environment/citizen or nature/subject – and the technologies by which they are both made amenable to rule is an area of study covered by a sub-set of scholars exploring the politics of environmentalism (Darier, 1998; Luke, 1998, 1999; Mitchell, 2003; Rutherford, 2007). Connecting the government of nature (and its socio-natural production through discursive and material practices), to the governance of populations, scholars identify environmentalism as a field of power through which populations are increasingly governed by neo-liberal states, in a ‘green governmentality’ (Luke 1998, 1999; Rutherford, 2007). Originating within western studies of environmental politics, this analysis articulates the ways in which environmentalism has become a new, non-state centered site for the exercise of both disciplinary and self-regulatory power upon both the individual body and global population – enticing certain actions, practices, and attitudes while enforcing against others (logging, hunting) in ways that are often shown to be both race and class specific (Luke, 2003). Describing how new forms of knowledge about the global environment (satellite, statistics) produce particular representations and discourses around natures in need of intervention and management, and contribute to the production of particular environmental discourses (scarcity, biodiversity), scholars document the implications of these ‘truths’ for how they entice or enforce certain knowledges, practices, subject positions, and kinds of morality (the ecological citizen) (Luke, 1997; Mitchell, 2003). Although a significant proportion of work on environmentalism as a field of power stems from analysis of the operations of power within the western neo-liberal state, other scholars working in the nexus of relations between environment and development have demonstrated ways in which environmentalism also operates outside of neo-liberal regimes to manage populations, reconfigure space, and constitute certain subject positions. Circulated through the sustainable development paradigms of multi-lateral development banks and international agencies, certain environmental truths (deforestation, conservation, biodiversity) have been adopted by international organizations and state agencies as a way to manage population in areas of the Global South (Goldman, 2005; Li, 2007). The ways in which these environmental discourses have then enforced particular spatial practices and subject positions reiterates the myriad of ways in which power is worked through the environment. Scholars of environment and development in Southeast Asia have in particular demonstrated how non- western states use international environmental discourses as way to rationalize state control over natural resources, and control recalcitrant populations (hill-tribes, ethnic minorities) by using the environment as a ground for normalizing certain behaviours, and state/society relations (Li, 2007;  33 Goldman, 2005; Vandergeest, 2003). In Indonesia, Tania Li documents how the establishment of community based natural resource management, emerging from discourses of biodiversity and sustainability, has tied ‘rights’ to particular kinds of environmental identities, social organization, spatial practices, and resource management (Li, 2002). This draws attention to how, outside of western states, state/society relations and identity are also being reconfigured – and resisted - through environmental policy. The analysis of relations between power-nature-society within these studies of the politics of the environment in both western environmentalism and environment and development in the Global South has highlighted the ways in which power operates through the environment. Specifically, it illustrates how power operates at multiple sites and scales of governance that involves, but also moves outside, of the nation state. A major contribution of this work is the way in which it de-centers state power, responding to earlier criticisms of governmentality for its singular focus on institutions of the state (Agrawal, 2005; Li, 2001). Acknowledging that “power bleeds across the social body in such a way that governing occurs in multiple sites and through a myriad of techniques” (Rutherford, 2007:294), scholars have extended analysis of government both below and above the nation-state (Blomley and Sommers, 1999). I argue that this can be used to correct earlier analyses of the government of nature (particularly in Global South), which looked primarily at the expansion of state domination and control (see Ferguson, 1990; Scott, 1998), pointing instead to the myriad relations and scales of power which also operate upon and within state institutions. In addition, by conceptualizing government as more than merely state power, it becomes possible to acknowledge states as composed of diverse institutions, not unitary bodies similarly enrolled within rationalities of rule, but rather each negotiating different global/local relations (Agrawal, 2005; Tsing, 2005). Attention to the role of non-state actors in the development of governmental rationalities and technologies is also evident through the work of geographers, who also argue for an expanded analysis of government beyond the state. Rose-Redwood (2006) emphasizes the role of international organizations like the International Red Cross in the work of biopolitics of populations, while also pointing to the work of non-state agencies in what he calls the ‘geo-coding’ of urban spaces in the U.S. during the 18-19th century. Likewise, Blomley and Sommers (1999) and Crampton (2000) both illustrate the use of government technologies (mapping), and the construction of governmental rationalities by non-state actors as cartographic technologies are used as forms of resistance. As cited in Rose-Redwood (2006), Foucault himself insisted that “relations of power, and hence the analyses that must be made of them, necessarily extend beyond the limits of the State” (1980:122). The challenge put to the analytical framework of governmentality is therefore how to account for, and include, ‘governmentality from below’ and from above, and to consider the interplay between state and non- state practices of government (Rose-Redwood, 2006).  34 While providing some key insights into the operation of relations of power mobilized through government, a criticism similar to postcolonial studies has also been made of some analysis of power- nature-society and governmentality, specifically, that attention to discursive relations of power are prioritized over material practices of dispossession and violence (Harris, 2004; Mills, 2003). As attention to the discursive construction of nature, and subsequent formation of rationalities of government and environmental policies looks at how certain places have been reconfigured and re- imagined as sites in need of intervention and management (Luke, 1998), the emphasis indeed seems to be on discursive constructions – representations on maps, satellite imagery, statistics – rather than actual material practices. Critics of this emphasis on discourse argue that the dramatic reconfiguration of landscapes effecting eminently material relations of inequity (forestry, nature conservancies, national parks) can not only be explained through the construction of discursive categories, but needs to attend to enforcement of practices, and material relations of power involved in acts of conquest, dispossession, and exclusion (Harris, 2004). In response to this criticism, I repeat the argument of the previous section: the analytical framework of governmentality, particularly as applied by scholars of power-nature-society in postcolonial settings, enables movement beyond the discourse/materiality binary. First, as postcolonial scholars seek to decolonize western knowledge, and attend to the violence enacted by exclusions enforced by binaries, they have highlighted the production of boundaries between nature and society as a political technology (Braun, 2002). In turn, this provokes rethinking of the relationship between discursive and material relations of power. Second, attention to the practice of government has helped to trace the material processes and practices through which nature is rendered amenable to technical interventions and inserted within political rationalities. Again, this has suggested an iterative relationship between discourse and practice, as power/knowledge was both productive of but also produced through particular practices (mapping). This is illustrated by scholars like Timothy Mitchell, who while acknowledging the state as a discursive category does not take the presence of the ‘state’ for granted but illustrates the practices through which it is physically constituted, practices “through which uncertain yet powerful distinction between state and society is produced” (Mitchell, 1991:78). Similarly, Ferguson and Gupta (2002) illustrate how the discourses productive of state plans are then constituted through the practices of data collection and planning to produce the apparent autonomy and authority of ‘the state’ (Ferguson and Gupta, 2002; Moore, 1998). In addition, complementing the attention of postcolonial scholars to the materiality of governmentality, and the iterative relationship between discursive categories and material practices, other scholars have focused on the relationship between material environments and subjectivity. Illustrating how subjectivity is not only a discursive construct, but is also constituted in relation to material environments, Agrawal (2005) uses the term ‘environmentality’ to describe this imbrication of environment with subjectivity. Originally coined as a term to describe the operation of governmental  35 power through trans-national environmental organizations who exercise disciplinary power through environmental policy (see Darier, 1999; Luke, 1995, 1997), it is used by Agrawal to denote shifts in subjectivities which accompany new forms of regulation of the environment. As Agrawal explains in the context of the Kumaon region of India, environmental regulations were productive of both new ‘natures’ (socio-natures: the scientific forest), and new subjects (unruly, illegal vs. responsible, environmental, conservationist) who were formed not only through these discursive categories, but through their own material practices. Although not without its criticisms, Agrawal’s demonstration of the mutual constitutiveness of (colonial) governmentality, the identities of colonial subjects, and the physical environment, illustrates the government of populations not only through the construction of discursive categories, but through the formation of particular physical spaces and natures. Unfortunately, left unexplored by Agrawal are the ways in which gender, race, and socio-economic class are worked differently through environmental interactions. In addition, as his definition of ‘environmental subjects’ remains limited to those who consciously articulate and self-regulate behaviour according to standardized notions of environmentalism (protecting biodiversity, conservation practices), this neglects that fact that we are all, consciously or unconsciously ‘environmentalized’. This binary use of the concept of ‘nature’ and environment obviates the socio-natural production of the ‘environment’, and of course excludes examination of urban natures. The imbrication of environment with identity that Agrawal does elucidate contributes to other analyses that have explored the formation of subjectivity within the analytic of governmentality (Chakrabarty, 2000; Mitchell, 1989; Prakash, 1999; Scott, 1999). In particular, Agrawal’s contribution is an engagement with the ways in which material practices and environments are part of the constitution of identity. This also provides balance to the emphasis on discursive constructions of subject positions, criticized for failing to seriously engage with the past and present material geographies of inequity. Instead, illustrating how material relations of power work to close off possibilities for the articulation of certain subject positions - what Butler (1993) would term as ‘making more or less easy to articulate’- this analysis can acknowledge the role of spaces and exclusions which are physically enforced. For example, using this analysis in the colony of the Netherlands East Indies, it becomes possible to articulate how the spatial arrangement of access to urban water supply, and subsequent material relations with particular kinds of water that it either enabled or disabled, informed the discursive constitution of colonial subjects. Finally, to conclude this section of the chapter, I return to its opening paragraph, and the second half of the criticism of Foucault’s conception of a static environment existing outside of social processes.  Specifically, as nature is an object of, and produced by government, nature is also productive of government. Moving beyond the material effects and actions of government, this emphasizes the iterative relationship between governmentality and materiality. For, understanding the environment as a field of power means that it must be recognized also as being constitutive of the  36 relations of rule. Hence, while government is productive of certain socio-natures (and attendant spaces and subjectivities), it is itself produced by these material conditions, natures, and practices (Braun, 2000; Ferguson and Gupta, 2002; Mitchell, 1991). Attention to material practices through which government constitutes nature also points to ways in which political rationalities themselves were informed through these material practices. Braun (1997) shows how the process of making nature legible and amenable to government also informed the technologies of government - constituting the state itself as a separate sphere, and the construction of society and national subjectivities (geological citizens). Understanding how material environments are not only a field of power for government, but are in fact productive and constitutive of these relations of power and political rationalities acknowledges natures as actants, “dynamic forces that constantly surprise those who would harness and control them” (Latour, 1993, cited in Li, 2007:18). This in turn suggests that the biophysical properties of nature matter - determining how they are enrolled within government, and acknowledging how they either frustrate, facilitate, or provoke unexpected relations of rule, while they either enable or disable the articulation of different subject positions. Illustrating the former point, Bakker (2006) documents how bio-physical and kinetic properties of water matter in how it is able to be enrolled in neo-liberal projects of privatization. In conclusion, I argue that like postcolonial scholarship, analyses of power-nature-society have also refined the utility and scope governmentality. First, attention to the various ways in which relations of power work both upon but also through material natures has emphasized the social production of nature as a socio-natural or material/discursive product of government. While this corrects the notion of a static, un-mutable nature existing outside of social relations as seems implied within Foucault’s conception of territory, it also offers analysis of how relations of power work through the physical environment to link the constitution of nature, space, subjectivity, providing a conceptual tool for deconstructing the binaries of nature/society or material/discourse. Second, attention to the environment as a field of power through which government operates has highlighted the operation of government at scales beyond the nation state, correcting the state centered analysis of some postcolonial and environment and development scholars which seem to assimilate governmentality with ‘government’ of nation-state, or fail to recognize the production of state/society binary as product of government technologies, or ignore the production of scale through relations of power as a component of governmentality. Third, attending to the materiality of government as it constitutes physical environments considered necessary for rule, has highlighted the ways in which subjectivity is also constituted through material practices. Finally, moving beyond the mere material practices and effects of government, scholars have emphasized an iterative relationship between governmentality and materiality. Understanding the environment as a field of power means that it must be recognized as  37 constitutive of the relations of rule mobilized by government, hence while government is productive of certain socio-natures, it is itself also constituted by material conditions, natures, and practices.   2.3.3. Materiality of government: Layers of relations of rule  Seeing government as being both simultaneously discursive and material allows for the analysis of ways in which relations of power mobilized, appropriated, and resisted within  government are worked through physical technologies (like urban water supply infrastructure), material practices, and actual, tangible, and visceral matter.  Acknowledging this material dimension of government thus attends to the ways in which relations of power are inscribed within ecological processes and physical spaces, as well as in socio-cultural and socio-economic relations. Applying this observation within post colonial urban spaces, to identify how past relations of rule remain layered upon and embedded within both discursive categories, socio-economic processes and the physical environment highlights the relevance of the colonial past for the present, and speaks to the need to rematerialize urban studies within the South (McEwan, 2003; Yeoh, 2001). Using the concept of ‘layers of relations of rule’ from Tania Li (2001) to attend to the physical sediments of previous government rationalities, and noting how they continue to interact with the present as they are “layered over, and articulated with prior formations” (Moore, 2001 cited in Goldman, 2001:513) assists with the project of re-conceptualizing and rematerializing urban studies within the Global South. First, recognizing how relations of power are worked through physical natures, corporeal bodies, and concrete urban spaces, and attending to ways in which political rationalities themselves are articulated in relation to these material environments, there is a new visibility to ways in which new rationalities of rule emerge from and engage with the past. Even if, as highlighted by postcolonial scholars, the effects induced by governmentality are rarely those that are intended, since government is never complete or uncontested, they still “induce a whole series of effects in the real” (Li, 2007:28) which do not disappear when one mentality of government is replaced by another, as for example the shift from colonial rule to independence. Thus, as previous relations of power remain layered both within the discourses and material environments of contemporary places to interact in unpredictable ways to produce new relations of rule, this can add new material dimensions to the analysis of conditions of (post)coloniality. Without being prescriptive or derivative for contemporary conditions, the physical sediments of previous relations of rule do still matter. Attending to the materiality of governmentality within the concept of ‘layers of relations of rule’ is particularly appropriate for analysis of urban water supply in the Global South. As water supply infrastructure is long-lived (often well over one hundred years, as is the case for Jakarta), these circuits and networks simultaneously embody successive relations of rule, through the patterns of water supply  38 infrastructure and water use practices they both enable and disable. As colonial relations of power were concretized within urban infrastructure and embedded within physical geography of the colonial city, the successive eras of government have had to negotiate around and within these colonial relations. Building physically on top of the colonial infrastructure systems, the subsequent layers of Jakarta’s water supply infrastructure were also discursively situated within existing sets of socio-economic, cultural, and political relationships. Therefore, Jakarta’s colonial pipe’s continue to be engaged within new relations of power, and as I explore in subsequent chapters, it is the materiality of governmentality - the ways in which relations of power were inscribed within physical space and ecological processes as well as socio-economic relations - that explains why colonial relations of power continue to inform access and sustain fragmentation. Second, beyond illuminating the continuingly productive relationship between the colonial past and present in the contemporary city, understanding the materiality of governmentality can also aid scholars in understanding processes of urbanization within the Global South. Specifically, the materializing and thus historicizing of relations of power can help to decolonize urban studies, and produce more diverse urban geographies as called for by scholars working in the cities of the Global South (McEwan, 2003; Robinson 2002, 2006). As argued by Robinson (2003), the assumed universalism of theoretical claims that are developed without reference to the range of different cases to which they are assumed to apply, such as cities in the South, need to be challenged. Indeed, recent scholarship on specific cities in the South highlights the importance of Robinson’s critique of Euro- American centered forms of theorizing (see Boland, 2007; Gandy, 2005, 2006; Legg 2006; Loftus 2007; Nijman, 2006; Swyngedouw, 2004). Rejecting the extrapolation of theoretical frameworks developed through research on cities in the North to those in the South, these urban scholars are calling for the creation of alternative, more flexible theories of urbanisation to generate more appropriate explanatory frameworks (Robinson, 2002, 2006). A conceptual framework informed by the materiality of governmentality can, I argue, contribute to this project of generating more ‘cosmopolitan’ analysis of processes of urbanization in general, and the fragmentation of urban water supply networks (and, indeed, many public services and amenities) in particular. An example of the conceptual contribution of the dialogue between governmentality and materiality is illustrated when examining the thesis of ‘splintering urbanism’ (Marvin and Graham, 2000), which, as a theoretical framework developed to explain the ‘collapse of the integrated ideal’ following the fragmentation of access, control, and pricing of network infrastructure in both the North and the South, is still rooted in a Northern context. Although Graham and Marvin acknowledge the differentiation of service provision which characterizes colonial cities (which they characterize as ‘spatial apartheid’), concern with the interrelationship between splintering infrastructure networks and fragmentation of both urban space and social consensus is based on a narrative which assumes the prevalence of the ‘modern networked city’ as a generic phase, or stage of urban development.  39 Therefore, while Graham and Marvin (2001) position Jakarta, with its toll roads and gated communities, alongside other Southern cities as characteristic of a splintered urbanism arising from processes of liberalization, privatization, structural adjustment, and financial speculation, this is an incomplete analysis. Specifically, Jakarta has never ‘achieved’ the modern infrastructural ideal, and the splintering of its urban water supply infrastructure is neither recent, nor is differentiation always oppressive. As Chapter Six later illustrates, resistance to rule has been part of the process of Jakarta’s fragmentation, and splintering can also occur from below. As also noted by other scholars (Coutard, 2002), the ‘collapse’ of an integrated ideal may be attendant upon many particularly situated relations of power, and need not always be negative. I argue that the insight into the both the historical origins, and the variety of relations of power (including resistance), implicated within processes of splintering are visible within the more flexible framework of governmentality. Indeed, as documented in Chapter Three, by tracing the ways in which the colonial rationalities guiding the construction of the original infrastructure have been incorporated into postcolonial government rationalities, the processes of ‘splintering urbanism’ are shown to be much more than an example of the recent, neo-liberal phenomenon.   2.4 Conclusion: Productive Contradictions – Jakarta’s urban water supply  When governmentality is reworked through the analyses of postcolonialism and power-nature- society to reveal the mutually constitutive relationship between materiality and governmentality, it provides the conceptual tools necessary for understanding the production of patterns of water supply in Jakarta. Identifying and explaining the current contradictions within patterns of urban water supply by making visible the relationships between urban governance and urban infrastructure, the analysis foregrounds the physical environment (and concrete infrastructures) as a field of power through which socio-natures are produced, but also productive of, rationalities of rule. First, attending to the ways in which relations of power are inscribed within physical space and ecological processes as well as discursive categories and socio-economic processes gives literal interpretation to the ‘layers of relations of rule’, and suggests how the imbrication of urban infrastructure and identity continues to inform patterns of provision and preferences in the contemporary city. Applying this analytical framework to Jakarta in particular yields an understanding for both the sustained nature of splintering of the city’s water supply network, and attendant persistent patterns of provision. For, unlike the developmentalist analysis of failed modernization, or the thesis of splintering urbanism - analyses which demand deliberate blindness, or merely footnote Jakarta’s realities as anomalies within theory – the sustained splintering of Jakarta’s water supply network and the problematic non-thirsty poor are not only able to be acknowledged, but are explained as the  40 material, discursive, and spatial effects of successive, contested, and contradictory projects of (post)colonial government. As government rationalities have thought and acted in particular ways to articulate the ‘right relations’ between urban waters, spaces, and residents, successive governmentalities have inscribed both physical and discursive relations of rule, which continue to articulate with the present in unpredictable ways. Ironically, by taking what is defined as Jakarta’s deviance from western based urban trajectories – its failure to achieve universal provision of water through centralized distribution system - as a productive point of entry, governmentality provides a framework for exploring the ways in which this fragmentation and subsequent patterns of access are historical products of material and discursive relations of power. Secondly, by identifying how relations of power mobilized within government are materialized within urban water supply infrastructures, the framework of governmentality highlights the often hidden relationship between urban infrastructure and urban governance (Gandy, 2002). Jakarta’s splintered networks are identified as the product of both government strategies attempting to secure the ‘conduct of conduct’ through urban water supply , and resistance to these strategies by uncooperative subjects and unruly natures.  Acknowledging this relationship between urban governance and urban infrastructure is crucial to explaining the splintering of multiple modes of production and distribution throughout the development of a centralized water supply network in Jakarta. For, although the western urban infrastructural ideal - centralized supply – was periodically articulated as a project of successive colonial and postcolonial governments, it was never intended as a universal supply. Rather, the achievement of one standardized universal system of supply was seen as contradictory to government strategies of producing difference, a tactic I argue was necessary to establish and maintain legitimacy of both colonial and postcolonial government. Subsequently, as the following chapters document, patterns of access to the various centralized piped networks were established according to a strategy of urban governance that sought the creation of particular kinds of urban spaces and particular kinds of urban citizens who resided in them (European vs. native; hygienic vs. contaminated; colonial vs. independent; modern vs. traditional; obedient vs. illegal). Providing different qualities and quantities of water supply through different kinds of infrastructure, according to different rationalities of rule over time (modernizing, developing etc), through different modes of power (disciplinary vs. biopower), has been part of the project of producing governable citizens, and reinforcing the authority of various (post)colonial governments. Therefore, rather than failed modernization, or lack of development, Jakarta’s splintered urbanism has been the product of government rationalities, and equally importantly, resistance to these rationalities – by both human subjects and uncooperative natures. The next chapter follows from the conceptual framework articulated here. Conducting a historical analysis of the socio-spatial fragmentation of access to piped water in Jakarta, Chapter Three provides the first of three methods of analysis (historical, political economic, discursive) used to document the governmentality of water in Jakarta.  41 Chapter 3  Splintered Networks: Urban water supply in Jakarta, 1870-1990   3.1 Introduction  In this chapter I document how the progressive development of Jakarta’s centralized piped networked water supply has continually worked to fragment access between different urban spaces and urban populations according to relations of rule articulated by successive colonial and postcolonial governmentalities. Documenting the ways in which colonial rationalities guiding the construction of the city’s first water supply infrastructure were incorporated into postcolonial governmentalities, the chronology challenges the thesis of splintering urbanism as developed by Graham and Marvin (2001). Identifying the socio-spatial fragmentation evident within the contemporary city as a product of a pervasive, persistent rationality of rule governing the production of urban water in Jakarta, rather than a recent, neo-liberal phenomena, the chronology traces how the patterns of splintered access inscribed within the city’s first urban water supply infrastructure in the 1870s was subsequently enrolled to support successive governments who also premised legitimacy upon visible axes of difference between subjects and the urban spaces they occupied: European vs. native, modern vs. primitive, developed vs. undeveloped and so on. Through the analytic of governmentality, the splintered networks of Jakarta’s urban water supply are viewed as the particular and historical product of colonial and postcolonial relations of rule which limited access to the piped networked water supply to particular types of urban citizens. Following from the moment of origin of Jakarta’s water supply in 1873, the chapter also highlights the materiality of governmentality to trace how both the city’s built infrastructure, and the ways in which it articulated legitimacy of access to piped water, continued to inform subsequent physical patterns of access and rationalities of provision. Addressing the criticism of those scholars who argue that analyses of governmentality focus too narrowly on discourse and abstract, intangible subjectivities, without acknowledging and incorporating the material aspects and expressions of power (see Harris, 2004), the chronology in this chapter highlights the material effects of relations of rule. Specifically, identifying the ways in which the water supply infrastructure in the contemporary city has been both physically and discursively layered upon the foundations of the first reticulated networked system built by the colonial government, I highlight how the ‘layers of relations of rule’ described by Li (2001) are grounded not only within socio-economic relations and discursive categories but also within material infrastructure, physical spaces, and ecological processes. Attending to the ways in  42 which both the physical and discursive sediments of previous relations of rule do still matter, I excavate the relevance of colonial pipes for the contemporary city. Deconstructing the narrative of progressive, linear development of the city’s urban water supply system as it is usually presented (see Maronier, 1929; PAM Jaya, 1992b), the chronology of Jakarta’s water supply infrastructure development paradoxically illustrates the continuities between the phases of both colonial and postcolonial progress: each main infrastructure development worked to further fragment the distribution of water supply both socially and spatially throughout the city. With each phase of development into the centralized piped water supply effecting a simultaneous fragmentation of access between urban spaces and populations - circulation of different qualities and quantities of water, different technologies of distribution, and economies of consumption - the colonial patterns of provision are still visible within the contemporary city. The continuity between the past and present in Jakarta’s urban water supply system is typically identified as the result of a lack of progress, and viewed as evidence of Jakarta’s failed modernization (see Robinson, 2002; see examples within Drozdz, 2006; UN-ESCAP, 2007). However, by documenting the original and progressive fragmentation of the piped water supply network, this chapter problematizes both academic analyses of splintered urbanism (Marvin and Graham, 2000) and the developmentalist interpretations of failed modernization. I argue that the failure to provide the western standard of universal provision is not the product of failed modernization as usually interpreted within developmentalist analyses. Rather, fragmentation is the physical and spatial product of colonial and postcolonial governments who never had universal access as a political goal; the production of a centralized urban water supply was only ever a partial project of both colonial and postcolonial government in Jakarta, and always accompanied by attendant development of parallel and partial networks of alternative supplies. The chronology of Jakarta’s urban water supply infrastructure proceeds to follow four key phases of growth in the city’s urban water supply – (1) the colonial artesian water system (1870-1920), (2) the colonial spring water network (1920-1945), (3) the first two large scale surface water treatment plants Pejompongan I&II built after independence (1950-66), and (4) the long decline of the centralized public water supply infrastructure during the 1960-1990 New Order government. These first four phases of urban water supply development are demarcated by the types of infrastructure and water supply technologies that were introduced, but also importantly, through the subsequent fragmentations of waters, spaces, and populations that they each entailed. The (post)colonial rationalities of rule - and resistance to rule - underlying the patterns of deliberately partial provision are explored more thoroughly in Chapters Five and Six, while Chapter Four discusses the dramatic expansion of the city’s urban water supply infrastructure after 1990 and during the World Bank’s Pam Jaya System Improvement Project and subsequent privatization, and the new rationalities of government introduced within these projects.  43 Table 3.1 Urban water supply infrastructure development in Jakarta, 1873-1990                        Year(s)  Infrastructure built and operational  Total Production Capacity (L/s) Number of Household Connections 1873-1889 9 artesian wells 6 reservoirs (660 m3; T = 660 m3) n.a. 1 1890-1922 41 artesian (shallow) wells (T = 50) 6 reservoirs (140 m3; T = 800 m3) 15 pumping stations (121 L/s) 119 Km main pipe network 121 4,090 1923-1930 Spring water supply from Bogor (315  L/s) 1 reservoir (20,000 m3) 154 Km main pipe network (T = 284 Km) 315 18,769 1931-1956 Spring water production capacity increased  by 85 L/s 400 n.a. 1957  Pejompongan I WTP (2,000 L/s)  2,400 n.a. 1966  Pejompongan II WTP (1,000  L/s)  3,400 n.a. 1967 Pejompongan I capacity increased by 1,000  L/s 4,400 n.a. 1973 Pejompongan II capacity increased by 1,000  L/s 5,400 n.a. 1975 Pejompongan II capacity increased by 1,000  L/s 225 Km main pipe network (T = 509 Km) 6,400 n.a. 1977  Mini-WTP Cilandak 200  L/s  6,600 n.a. 1982 Pulogadung WTP 1,000 L/s 6 Mini WTP’s 280 L/s (T = 480  L/s) 7,880 n.a. 1986 Pejompongan II capacity increased by 600  L/s Pulogadung increased by 3,000  L/s 100 Km more main pipe network (T = 609 Km) 11,480 154,900 1988  - 158,620 1990 3,063 Km more pipe network (T = 3,672 Km)  10,400 227,830 n.a. = not available T = Total (reported figures).  Sources: Smitt (1922); Maronier (1929); Eggink (1930), PAM Jaya (1992b); World Bank (1990b, 1998); JICA (1997).    44 3.2  Artesian water supply: 1873-1922  The drilling of individual artesian1 wells begun as a project of colonial government in the 1870s marks the beginning Jakarta’s centralized piped water supply. Accompanying changes in the purpose of colonial government, which would see an influx of new European migrants following new economic policies promoting investment opportunities, the center of colonial administration began to emerge as a proper colonial capital with attendant infrastructure supporting its transformation from ‘Tempo Dulu’ colonial society (see Milone, 1967) to a racialized and spatially segregated urbanized landscape. Supporting a new population of European residents, and contributing to the new racial rationalization of colonial rule, the series of individual wells, supplemented by reservoirs and later fused together within a piped distribution network, was also the beginning of the splintering of city waters between different urban populations and urban spaces. With the centrally regulated and scientifically monitored artesian water strictly limited to the legally defined ‘European’ population, the city’s first water supply infrastructure facilitated the desired divisions between races (European vs. native), and the urban areas they resided (European suburbs and ‘well planned’ residential areas vs. the kampong). Supporting a colonial governmentality that premised legitimacy of rule upon European racial dominance (see Gouda, 1993), the free standing hydrants and reservoirs comprising the city’s first urban water supply system were located in a north-south strip in the central (European) parts of the city, mirroring the spaces of new European concentration. The public to whom the series of artesian wells were made accessible consisted of the city’s 7,000 odd legally designated European residents, less than eight percent of city’s population at that time2. Of course, as noted by scholars of the Netherlands East Indies (Cote, 2003; Stoler, 1992; Taylor, 1983b), the racial category of ‘European’ did not imply homogeneity; as a contested discursive construct, this legal category encompassed an uneasy mix of naturalized Europeans who had been born in the colony, and an increasing number of new middle class arrivals3. This discursive differentiation between different kinds of European residents themselves, and the ways in which this was negotiated through urban water supply infrastructure, is dealt within in detail in Chapter Five; within this Chapter I document how the colonial government attempted to correct the porous boundaries of racial identity by fixing, and physically embedding, the socially constructed categories of citizenship within corporeal bodies and urban spaces. In the rest of the chapter, the terms ‘European’ and ‘native’ as pertaining to access to urban  1 Artesian water is groundwater trapped between layers of impermeable rock, so that when tapped is under pressure and comes out of the ground as like a spring. Groundwater is still used as a water supply in present day Jakarta, but is no longer under pressure, with the water drawn from between layers of impermeable rock now being pumped from depths of 100-200 meters below sea-level. 2 In 1885 Van Raay (1915) records 7,634 legally designated European residents, 25,560 Chinese, and 62,616 indigenous and mixed race residents living in and around area of Batavia. 3 In 1885 the total European population of the Netherlands East Indies was approximately 47 000; 35 000 of whom were born in the colonies, the vast majority of whom were of mixed European and ‘Indische’ descent (Cote, 2003).  45 water supply in the colonial city derive from colonial archival records; they do not differentiate between indigenous ethnicities (Javanese, Sundanese, Balinese, Buginese et cetera), nor between the different cultural competencies and middle class moralities of those who were legally (although perhaps still not culturally) defined as ‘European’.  Map 3.1 Artesian water supply network, 1873-1922  Source: Smitt (1922)  In the 1870s the city’s water supply system consisted of a series of seven artesian wells, with a storage capacity of 660m3, but the system continued to expand in capacity and coverage according to the growth and pattern of development of the European urban population.  By 1920, after 50 years of development of the artesian supply system, there were 28 wells, 12 reservoirs with a storage capacity of 750m3, and 118 km of piped network connecting 3,863 individual households to the piped water supply system (Gemeente Batavia, 1917; Smitt, 1922).  46 Table 3.2 Artesian wells, pump stations, and reservoirs installed in Batavia, 1873-1921  Name & Location  Year Depth (m) Elevation + sea level (m) Year of pump station Pump Capacity (m3/hour) Year of Reservoir Reservoir Capacity (m3) 1. Parapattan I 1873 1.66 6.58 2. Batoetoelis 1873 1.31 3.50   1873 100 3. Glodok I  1873 1.40 3.00 1916  1873 200 4. Schouwburg 1874 1.40 4.25   1874 100 5. Beursplein 1875 1.71 2.00   1875 30 6. Koningsplein 1976 1.89 3.52   1876 180 7. Salemba 1876 1.09 8.38   1876 50 8. Sluisbrug 1878 2.01 5.17 9. Tanah- Abang 1884 2.42 3.32 1917  1916 22 10.  Kemajoran 1892 2.36 3.78 1919 30 11. Mijnwezen 1903 2.87 12. Rijswijk 1904 2.78 3.50 1919 30 13. Djati 1905 2.92 9.77 1917 15 14. Mijnwezen 1905 2.12 2.59 1921 30 15. Tjikini 1907 2.91 7.54 16. Matraman 1908 2.57 10.71 1917 30 1916 22 17. Glodok II 1909 2.51 5.00 18. Parapattan II 1909 2.63 6.58 1919 30 19. Trate 1910 2.00 1.93 1918 30 1918 22 20. TanahTinggi 1911 2.45 5.65 1918 30 1918 44 21. Menteng 1912 2.88 11.30 22. Jacatra 1912 2.63 3.03 1921 30 23. New Gondangida 1915 2.61 4.00 1918 30 1918 50 24. Krekot 1919   1920 30 25. Petodjo-Ilir 1919   1920 30 26. Boulevard 1919   1920 30 27. Sentiong 1919   1921 30 28. Djati-Baroe 1920   1921 30  Source: Data from Smitt (1922)  47 Illustration 3.1 Artesian well and reservoir at Batoe-Toelis, 1880    Source: Afdeeling Waterstaat (1880), Reservoir op den artesischen put Batoe-Toelis (Reservoir at the artesian well in Batoe-Toelis).    Accompanying the racial heterogeneity throughout the 17-18th centuries (see Milone, 1967; Taylor, 1983b), and prior to the colonial government’s construction of the artesian wells, the entire heterogeneous population of Batavia – full-blood European, mixed race Indische, Javanese, Madurese, Malay, Chinese, and other foreign easterners4 -  all relied on different combinations of river water and shallow groundwater. As was described by a military doctor of the Netherlands Indies, residents used various procurement and treatment methods depending on their geographical location within the city, personal inclination, and socio-economic status (Moens, 1873). Although intermittent attempts to solve the perennial problem of poor water quality in Batavia had been undertaken infrequently throughout the 17-18th century, the imprint (and improvements) on the urban landscape prior to the 1870s were negligible, leaving residents to ‘secure for themselves’ a safe water supply (Moens, 1873). Selecting their sources of household water on the basis of cost, geographical proximity, and quality, as perceived through the traditional sensory assessments of colour, clarity, and odour, all residents used a variety of different water sources, treatment technologies, and providers – ranging from chemically treated surface water produced by a Water Fabriek (Water Factory), to the surface water sold by ambulatory ‘toekang air’ (water vendors).  4 Foreign easterners is the classification given by the colonial government to the colony’s Arabs, East Indian populations.  48  Although providing numerous consumer options and allowing for the varying socio-economic status of residents, the multiple waters of Batavia became problematic in the latter half of the 19th century. New scientific discoveries made connections between drinking water consumption and health, and articulated a standardized, biophysical definition of a hygienic water supply (Hamlin, 2000; Maronier, 1929; Moens, 1873). The shift from recognizing ‘many waters’ of various beneficial properties, to only one scientifically defined nature for water with its quality determined by what it did not contain, precipitated the development of centralized systems of water supply throughout the world (Gandy, 2004; Goubert, 1986; Luckin, 2000; Melosi, 2000). Within the Netherlands Indies this discovery accompanied a rationality of rule premising the legitimacy of colonial government upon European cultural superiority (Gouda, 1993), and stimulated the demand for an urban water supply system capable of providing a superior quality to select citizens (see Maronier, 1929; Moens, 1873). Therefore, although financial investments into urban water infrastructure were previously deemed unnecessary, or too costly for serious consideration by the colonial government, the artesian wells and later network were financed completely by the Dutch government and actually delivered water at no cost to users until the advent of individual household connections and a Water Ordonnance in 1910 (Maronier, 1929). Instrumental to supporting the legitimacy of the colonial government, and determining the technical design of the subsequent artesian water supply system, was the specific racialized public to who water was provided. Unlike in the metropole, in the Dutch colony the public provided with this new standardized water supply was composed of a particular, gradually more homogenous group of European residents, who at the end of the 19th century had begun to concentrate their settlement in newly developed suburbs of the colonial capital city5. Building upon the emerging divisions between urban spaces and urban populations, the artesian water supply system mirrored the newly concentrated clusters of European urban settlements, and the hydrants of the artesian water supply limited the circulation of a chemically treated and scientifically monitored clean water supply amongst a specific set of residents. Facilitating the transformation of the city from a collection of de-concentrated riverside garden villas of the ‘Indische’ or ‘Tempo Dulu’ colonial society (see Milone, 1967) to a more spatially segregated European population, the artesian water supplied a new concentration of Dutch government officials, military personnel and private company employees. This made the centralized provision of a uniform quality of water to a homogenous group of consumers both desirable - supporting the transformation of European settlements by providing a water supply alternative to the surface water previously conveniently located alongside estates - and geographically feasible. Meanwhile, the  5 Increasingly segregated colonial satellite towns, ringed by native kampongs, began to emerge around the city’s center (Weltevreden, Tanah Abang, Gondangdia, Meester Cornelis); see Kusumawijaya (1990), Milone (1967), Wertheim (1956).  49 formerly rural areas of native villages swallowed up by the expanding urban area continued to rely upon surface waters for all of their water supply and sanitation needs (Argo, 1999; Wertheim, 1956). Intended only to provide for the European population, who were increasingly encouraged to ‘take care’ for a scientifically defined quality of drinking water (see Moens, 1873), the artesian hydrants were limited to European areas of the city, and not extended to native populations until the last decade of the artesian water supply (1910-20; Eggink, 1930). Recorded as a matter of fact by a colonial engineer, this is the population for whom the provision of artesian water was always intended. “The Europeans were mainly settled in a small strip along ‘den Grooten Postweg’ [the large mail-road] running through Buitenzorg, Meester Cornelius, and Batavia, and so within the neighbourhoods of Salemba, Kramat, Passer Senen, Waterlooplein, Noordwijk and Molenvliet… The Government’s water provision using artesian wells also only reached along that strip. What lay outside that strip was not considered.” (Van Raay, 1915:136).  Map 3.2 Urban water supply and land use in Batavia, 1897     Source: Map created by Ernst-Jan Martijn (2007), based on water supply network map in Smitt (1922), and land use map in Tresling (1897).  50 Building upon the initial segregation inscribed within the artesian water hydrant and reservoir system, later development of the city’s water supply and specifically improvements in distribution technology led to even greater differentiation between the water supplies of the city’s population. From the late 1890s to 1900s an increasing number of European households began to build small piped networks, creating clusters around individual hydrants accompanied by reservoirs. The increasing demand for household connections to pipe water into the home – already the standard in European cities by the 1920s (de la Motte, 2005) – was in large part due to the influx of new types of colonizers from the metropole. From 1890-1920 male migration from Europe to the Netherlands East Indies increased by two hundred percent, and, perhaps more importantly as argued by some scholars, female migration increased by three hundred percent (Gouda, 1993; Stoler, 1995; van Doorne, 1994). With houses located adjacent to the artesian hydrants, European residents built small clusters of pipes, creating ‘spider-web like’ piping systems surrounding the above ground hydrants. Fused together by the Municipality of Batavia after its establishment in 1905, the number of household connections to the artesian network continued to grow, even more rapidly after 1910 when the government formalized its policies on household connections to ensure an adequate quality of service (Maronier, 1929). Of course this progress towards the metropolitan ideal was only geographically feasible for the households who lived within the clusters of formally planned European residential areas where the hydrants had been located. Reinforcing this initial socio-spatial bias, the subsequent investments into increased production (growth in capacity of wells), and increased distribution (growth in network area and addition of pumping stations) followed the expansion of European residential and commercial areas. During this phase of artesian water network expansion, the native spaces in the city were left untouched, and despite representing the majority of the area’s population, native spaces are noticeably absent from the maps of urban infrastructure. Indeed, throughout most of the operation of the artesian water supply system, native residents were spatially excluded from access. Providing the desired contrast between stages of (racialized) development, the majority of the native population living within the urbanized area continued to rely on the numerous surface water channels for all of their water needs. With the colonial government opining that the ‘physical standards of an urban habitat’ were foreign to the ‘traditional society’ of the native population (Tillema, 1913:407), the majority of the city’s residents were left to demonstrate their ‘traditional habits’ of washing, bathing, and defecating along the riverbanks (van Breen, 1919; van Leeuwen, 1920; van Raay, 1915; Wertheim, 1956).          51 Figure 3.1 Growth of European population and number of household connections in Batavia, 1899-1920                 Source: Data from Drost (1917), Maronier (1929, Smitt (1922) and Van Raay (1914).   The efforts to include the native areas of the city within the growing network of water supply occurred much later, following the secure establishment of water supply to European areas, and were decidedly minimal. Despite the introduction of an Ethical Policy in 1901, supposedly guiding colonial governmentality, the mandate of ‘raising up’ the native population was not immediately evident within the ongoing development of Batavia’s urban water supply infrastructure. Instead, the overall policy of the newly established Municipal Council (1905) towards native affairs - and provision of urban water supply - was “Onthouding was het parool, en geen inmenging!” (Abstinence was the motto, no interference in the affairs) (van der Wetering, 1939:308). Until 1918 native residential areas (kampongs) were still considered a legally, if not geographically, separate entity from the European city, and it had its “own life and independent task” for which the European dominated Municipal council was not responsible for (ibid). Therefore, in contrast to the programs of improvement directing government funds to native education, agriculture, and economic development programs6, there was little action taken to ensure that the native population of the capital city had access to safe water supplies (Abeyasekere, 1985). Although  6 From 1901-1913 the colonial budget rose from 160 to 300 million guilders; allocations to the education system for natives occurred in parallel, rising from 1.5% of overall government expenditures to 4.5% by 1913. See Moon (2005) for more on agricultural development programs, Lindblad (1989) for education, and Brooshooft (1977) for Ethical Policy programs.   52 geographically enclosed by an expanding urbanized area, the kampongs were considered an undeveloped and embarrassing aberration to urban life, and not actually part of the modern city (see Abeyasekere, 1985; Cobban, 1988; van der Wetering, 1939). Installing only four public hydrants for the indigenous urban population of over 116,740 residents (Eggink, 1930; van Raay, 1915), the resultant distances between home and water source meant that households had to expend their own time and labour to collect water, or purchase it from an ambulatory vendor. Vendors sold the artesian water at a cost of three cents per twenty litres, a significant amount for native households in the last decades of 19th century, when the average wage was between twenty to sixty cents per day (Abeyasekere, 1987; Booth, 1988).  Illustration 3.3 Water vendors in Batavia, 1918                          Source: Drost (1918), “Water aftapping uit hydrant Passar Baroe Oost. Het water wordt in open petroleumblikken weggepikold” (Water abstraction from hydrant Passar Baroe Oost. The water is vended in open gasoline tanks).   Left to remain reliant upon the most affordable and convenient water sources for their supply while the colonial government facilitated the transformation of European domestic habits and hygiene practices, the continued use of untreated surface water by native populations for drinking, bathing, and washing created both physical and discursive connections between native bodies and disease. “The occurrence of typhus and cholera epidemics were known to be related to the use of unsanitary water versus the use of reliable water…this was especially evident with the difference in hygienic circumstances between European and Inlandische [native] population.”(Maronier, 1929: 225).  53 Perversely, instead of generating a more widespread effort to eliminate the use of unsafe drinking water amongst native populations, the connection between native bodies and disease provided both physical and discursive support for still more segregation between populations. While reinforcing the premise of cultural superiority, the epidemics of cholera and typhus - clearly the product of lack of access to clean water - considered endemic to native quarters (Abeyasekere, 1987) increased the desire for a separation of European and native spaces in the city. Tellingly, the limited action that was taken by the Municipal Council to provide water supply to the kampongs during these years was spurred by this European fear of epidemics spreading from native to European areas of the city. The few public hydrants that were installed in urban kampongs from 1910- 1920 were geographically strategic7. Instead of installing hydrants into the lower northern areas of the city, where flood waters brought the greatest risk of disease and brackish groundwater created the greatest demand for a clean water supply, the European council favoured the native residential areas in the city center (Abeyasekere, 1985; Wertheim, 1956). Proximity to European residential and commercial areas seemed to be the criteria for provision of public infrastructure, and as the death rate in native quarters rose to 60,000/year during this same time period, “one's chances in the stakes for life and death…were increasingly determined by one's position in the racial hierarchy created by colonial rulers” (Abeyasekere, 1987:206)8. In conclusion, as intended by the colonial government, and acknowledged by colonial engineers, Batavia’s first urban water supply system was built for a specifically European population. When the Ethical Policy was introduced at the turn of the century the artesian water supply system gradually began to include native urban residents within its distribution, but government still continued to ensure the demarcation of different waters between different racial groups, and supported the shifting rationality of rule that legitimized European dominance in the discourse of modernity, not race (see Gouda, 1993). European residents were thus provided with water piped into the home and were expected to rely on this single source of water for all household needs (cleaning /washing /drinking /cooking), whilst native residents – if provided at all with artesian water – relied on the corporeal network of water vendors to deliver a much lesser amount of water, at a higher per unit volume cost. Not surprisingly, native households, if using the artesian water supply at all, used only a small amount for consumption needs. After more than forty years of the operation of the artesian water supply system it was simply accepted that  “the Europeans had access to considerable more points of supply” (Heetjans, 1915:305), and hence consumed the majority of the artesian water produced. Providing both the discursive and the physical foundation for the subsequent design of the 1920 spring water network, the pattern of provision embedded  7 Public hydrants were installed in the following native areas of the city: kampongs to the west of Molenvliet (1909), kampong Jaagpad (1910), kampongs Tanah Tinggi and Petodjo (1912) (Eggink, 1930). 8 In 1912 and 1913 the physicians Ouwehand and Van Gorkom published statistics on mortality in Batavia, calculating an annual death rate for the native urban population of over  60 000 (Ouwehand, 1912; Wertheim, 1958); the high rate of mortality was later attributed to polluted water supplies (Abeyasekere, 1989).  54 within the city’s first urban water supply infrastructure set in place the channelling of select waters, into select urban areas – reflecting, and perpetuating, the construction of differences deemed necessary to colonial rule.   3.3 Spring Water Network: 1923-1945  The city’s centralized piped water supply system was already in place throughout the last years of the artesian water system, but with the growth in urban population, the quantity of artesian water supply was increasingly uncertain9. Moreover, the quality of water provided by the artesian system was less and less desirable to a new kind of middle class European immigrant used to conveniences of the Continent (Maronier, 1929). Investigations into rehabilitating and/or replacing the artesian water supply system began in 1906, following the Decentralization Law of 1903 that established Batavia as a municipality with its own local government (Nas, 1990). However, it was only after two full decades of debate over the relative merits and expenses that the city’s spring water supply system was completed. Lauded by engineers and urban planners as a vast improvement over the artesian infrastructure era, the spring water system increased water production capacity to over 350 litres per second, the size of the reservoir grew from 780m3 to 20 000 m3, the city network was extended by over 150 kilometres (Smitt, 1922), and best of all – the water from the pipes could be used ‘straight out of the tap’ without need for aeration or cooling, as was the case for artesian water (Gemeente Batavia, 1937). Delivering pure mountain spring water into the city through fifty three kilometres of iron pipe, the 1920s spring water supply system was designed with a capacity of 350 litres per second to supply ninety percent of all European households with 140 L/capita/day. When the production capacity was later increased to 600 litres per second (Maronier, 1929), the supply was deemed sufficient for a slowly growing capital city that had not yet anticipated the events of coming years, notably World War Two and the transition to independence that followed. The city’s new spring water supply system was represented by colonial engineers as the conclusion of what had been a process of perpetual growth begun in the 1870s (see Maronier, 1929; Smitt, 1922; Van Raay, 1915a; Van Raay, 1915b). Physically, it was laid within the existing small city network of the artesian water supply system, replacing or rehabilitating piped materials in places where necessary but maintaining the spatial pattern. Discursively, the spring water supply system continued to support shifting colonial government rationality – marking the developmental superiority of the European population, whose mastery was no longer premised upon race, but modernity (Gouda, 1993; Mrazek, 2002). Subsequently, build upon the foundations of the artesian water supply infrastructure, the spring  9 See decline in volume of the artesian wells recorded Drost (1918), van Raay (1915a), and Smitt (1922), and debated in Koster (1917) and van Leeuwen (1917).  55 water supply system continued to splinter the city into axes of differential development. Specifically, much different technologies of distribution were used to deliver different volumes of water to different spaces in the city, and the different per unit cost prices associated with each encouraged different patterns of use amongst different types of urban populations (Europeans, Chinese, Arabs, and Natives). Replicating the patterns of provision initiated within the artesian system, the spring water network only exacerbated the divisions between urban spaces and populations: European residents provided with household connections paid half the price for water than did native residents using public hydrants, and subsequently came to use the bulk of the spring water supply. Native households continued to rely on multiple lesser cost sources as an economic strategy and convenience. Although the spring water supply system was the first urban water supply infrastructure intended for eventual universal use, this was seemingly as theoretical as the eventual independence of the colony, as the system continued to differentiate between the developed (European) versus the undeveloped (native) spaces of the city.  Map 3.3 Spring water supply network of Batavia, 1922-1945                     Source: Created by Ernst-Jan Martijn (2007), drawn from maps in Smitt (1922) and Maronier (1929).   56 Illustrating the contradictions inherent within colonial government - and the conflicting rationalities of different levels of national versus municipal authorities - while in principle the spring water system was intended to provide, after thirty years, for the entire urban population, in practice it was designed and constructed in order to serve European residents and commercial interests. Design criteria explicitly stated a production capacity that could serve ninety percent of European households with 140 L/capita/day, sixty five percent of Chinese and Foreign Easterner households at 100 L/capita/day, and thirty three percent of native households at 65 L/capita/day (Maronier, 1929). Colonial engineers, urban planners, and economists who designed the spring water network and set forth these principles of its operation built into the water supply infrastructure the premise that ‘natives had less need’ for networked water supply because they continued to use a combination of different water sources, a practice notably enforced by the pattern of water supply infrastructure development throughout the growth of the artesian systems, and subsequent limited provision of piped water to native areas. Speaking about provision of piped water to native residential areas, a colonial engineer stated that, “in principle the water for human consumption (i.e. drinking or cooking) shall be obtained from the pipes, whilst bathing and washing water will be taken from the rivers or existing wells…So, we don’t at all need to count on all inhabitants immediately being water users; in any case by far not all will obtain all required water from the water delivery system.” (Brandenburg, 1924: p.154)  Accordingly, native households were expected to use, and thereby supplied with, a much smaller volume per capita. Drawing on data from the operation of piped water supply networks in neighbouring cities of Surabaya and Semarang, colonial engineers argued that the target of providing thirty three percent of native households with piped water was even too ambitious. Instead, advocates suggested that eventually only twenty percent of the native population of Batavia would be connected to the spring water network, in comparison with sixty percent of Chinese, and one hundred percent of Europeans10 (van Leeuwen, 1917), and “when one asks about the different categories why widely varied usage figures are assumed, then the reason is partially to be seen in the fact that per lifestyle (mode of living), some have less water needs than others” (Gomperts 1916:12). The fact that for the native population, “a large part of the water used for internal purposes (human consumption) is taken from the water supply network, while bath and wash water is taken from the existing bad wells” meant that although “we know that an average of 50 L per day is not enough for the native, this is used as a starting point for the design of the water system…the assumption is based on the fact that not all the water that they use is taken from the water system” (Gomperts, 1916:13).  10 In Semarang 50% of the Chinese population, and 6% of the native population were connected to the centralized piped water system; in Surabaya, 14% of the Chinese and 25% of the native population were connected (van Leeuwen, 1917).  57 Conveniently, the Council’s initial decision to accept (and encourage) the use of lower quality river or groundwater by native households for all non-potable uses lowered the total costs of the city’s distribution system. As the municipal distribution network was paid for by the city, and not the Central Government, this was important for the fiscally conservative Municipal Council who financed only two of the total nine million guilders needed to build the spring water network11. Distributing spring water through communal hydrants rather than house connections, and thereby lowering predicted consumption to only sixty five litres per capita per day, the extension of the water network into the native areas of the city could therefore be much less extensive, saving the city money. Speaking of plans for the spring water network prior to its construction, a colonial engineer and member of the Municipal Council stated that, “for the kampongs only a few house pipes shall have to be laid down. The usage network shall thus be more simple, and can more or less limit itself to for supply to public washing and bathing places” (Van Breen 1916:9).  The Council’s decision to provide access to the public water supply amongst native population through public hydrants, rather than through household connections, was supposedly justified through the lower per capita usage of piped water amongst natives and the expectation that native households would be unable to afford the initial connection fee of twenty five guilders, and subsequent monthly costs of water and meter rental12. Perversely, both issues – ability to pay and lower consumption – were then reinforced by the pricing policies of the municipal water supply company. Native households that did have access to the spring water supply through public hydrants actually paid more per litre than did the connected European households. Although initially the spring water was distributed at no cost from the public hydrants (like in the artesian system), after the first few years of operation the Municipal water supply company became worried about its economic efficiency and the rising costs of this ‘free’ water supply to native households, and instituted new system of payment (Eggink, 1930; Maronier, 1929). With the system of ‘paid kampong water delivery’ the Municipality formally appointed the already operating native water vendors to distribute the water from the hydrants to native households. However, by charging these vendors the same volumetric costs for water that connected households paid (30 cents/m3), and allowing the vendors to re-sell it at a determined profit margin to take into account their transport labour, native households ended up paying 60 cents/m3! Native households thus paid two times the price paid by those with household connections. The director of the water supply system in the neighbouring city of Bandung recognized the perverse incentives built into this system of ‘paid kampong water supply’ and argued for a more equitable  11 The total cost of the spring water network system was nine million guilders, however the Municipality only had to finance the costs of the city distribution network (two million guilders); the Central Government financed the fifty three kilometres of pipes which brought the spring water from Buitenzorg (Bogor) into the high reservoir at the edge of the city, almost seven million guilders (Maronier, 1929). 12 Meter rental was between 0.5-1.50 guilders/month, depending on size of the meter (Gemeente Batavia, 1917).  58 tariff scheme, whereby those receiving the ‘special luxuries’ of having water piped directly in the home, could, and should pay more for their water than the poor, who should only be charged the cost price (calculated at 7.5-10 cent/m3) (Heetjans 1915, 1922, 1923). Describing the supply of water through public hydrants as ‘very unsatisfactory’, Heetjans pointed out that, “in the cities, where the largest part of the population has to have the water be carried (i.e. use water vendors), it is expensive, even more expensive than water from a house connection.” (1915:248).13 Undoubtedly, although the system of ‘paid water kampong’ water supply secured the financial profits of the municipal water supply company, it reinforced existing patterns of inequitable provision of water between the different populations, and discouraged the use of piped water within kampongs. With surface and groundwater supplies available at more convenient locations, and at much lower cost, the pattern of water use within native households - using a combination of water sources/qualities for different uses – remained as it always had. Colonial hygiene officers often referred to these practices as the result of ‘ingrained habit’; a well known hygiene propagandist stated, “even if properly designed houses with proper water supply and sewerage were available, unhealthy conditions would ultimately continue to be the result of the occupants’ living habits” (Tillema, 1913:79, cited in Cote, 2003). However, as seen below, the reversion to traditional habits was based on simple conditions of affordability and availability as constructed and enforced by the colonial government itself. Supporting this analysis, initially the placement of more public hydrants within the kampongs resulted in increased use of piped water network supplies by the native households. The increasing costs to the Municipality for its ‘free water’ supply budget as a result of increased use by native households was documented from 1920-1924; in 1920 the Municipality spent only 54 000 guilders on ‘free water’, but by 1924, it paid out 120,850 guilders (Eggink, 1930). However, after the regulation of 1926 charged a cost price of 60 cents/m3, native residents returned to their traditional habits, resulting in a corresponding decrease in demand for network water supply in native areas (Maronier, 1929). As recorded by colonial engineers, there was a dramatic reduction in the demand for networked water in the kampongs after the implementation of the ‘paid water delivery system’; by 1927 the budget for ‘free of charge’ water had dropped from 120,850 guilders to only 6,000; by 1930, “concerning the free of charge water provision, so little is being expended that the amount of it is not expressed in the company’s budget anymore.” (Eggink 1930:63). When, perhaps in response to its critics like Ingeniur Heetjans, the Municipal Council addressed the issue of equitable distribution, this also had perverse consequences – further facilitating spatial fragmentation. Initiating a scheme to subsidize house connections in some native neighbourhoods, the  13Engineer Heetjans explains how by paying 1 cent/petro can (approximately 18L), native households were effectively paying 60 cents/m3:  “water vendors pay 30 guilder cents/m3, then charges on average another 30 guilder cents for carrying/delivery of the water, so that the average costs are at least 30 cents/m3 (transport costs will be 30 cents/m3), which means that the water for the buyer has to cost at least 60 cents/m3.  This is considerably more than what is paid for water than by well-off European renters.” (Heetjans, 1923:91).  59 residents who could not afford the cost of either increased land values or higher rents that accompanied the provision of piped water services were effectively displaced (Abeyasekere, 1989; van der Wetering, 1939). The poorer segment of the native population subsequently moved out from the modernized kampongs in the urban center to the outskirts of the city, while those areas with piped water became populated by the new administrative middle class of salaried Indonesians, and Eurasians (Abeyasekere, 1989; van der Kroef, 1954). As the “silent battle for living space”(van der Wetering, 1939:315) continued to displace native communities for the building of new European residential districts and associated services (railways, tramways) (Wertheim, 1956), the spatial separation of two distinct urban societies was also project in part facilitated by the spring water network. Meanwhile, contradictory to the Ethical Policy of the colonial government, the expansion of the piped water network in European areas of the city was in stark contrast to the government’s ambivalent project of kampong water supply. With the extension of the spring water network throughout the European residential neighbourhoods planned for in the initial engineering design, and enforced by a regulation making connection to the network mandatory for households with a rent above twenty five guilders/month14, direct household connections supplied one hundred percent of European households by 1930 (Eggink, 1930)15. Meanwhile, only an estimated five percent of the Eurasian and native population were connected were connected16. In addition, while replicating the pattern of provision set in place by the artesian infrastructure built within European developed areas, the spatial development of the spring water network also continued to follow the pattern of European residential expansion. Looking at the map of the spring water network, the European neighbourhoods in Batavia are clearly indicated through the overlaid grid of water supply pipelines, in contrast to the almost complete invisibility of the kampongs. Thus, the pattern of urban water supply enabled by existing infrastructure of the artesian network, and further encouraged by the design and operation of the spring water network complemented the 20th century production of western enclaves. By the 1930s, after almost a decade of operation of the spring water network, the (racialized) developmental transformation of the colony was considered complete, as “the Indies became an area where Dutchmen, in a Dutch manner, in Dutch environments, led their own lives” (Van Doorne 1983:10). In conclusion, the development of the spring water network only further reinforced the patterns of use and provision visible already in the previous era of artesian water supply. Splintering the centralized water supply system into different technologies of distribution for differently modernized populations,  14 The Housing and Building Ordinance of 1 July 1923 (Bouw en Vonings Ordeening) made it mandatory for all houses of monthly rent over 25 guilders/month to connect to the piped water network (Eggink, 1930). 15 There were 10,392 household connections in European quarters (Eggink, 1930); the total European population of Batavia in 1930 was 37,067 (van der Wetering, 1939), with an average of 3.6 members per household (see Maronier, 1929). 16 There were 6,926 connections outside of the European areas; with an average of five persons per household (Maronier, 1929), the total non-European population served was 25,000, approximately five percent of the total.  60 only served to reinforce the different volumes of water provided to different kinds of residents.  As noted by an engineer commenting on the technical design of the city’s spring water system, “In neighbourhoods with a lot of villas (i.e. higher class), a lot more main pipes are necessary than in the more densely populated small housing areas. Moreover, the water in the larger houses is supplied to more taps, and so, with more luxury, so that the maximum use of the water network in well off neighbourhoods at the peak hours of water consumption rise more (peak is more pronounced, since these houses use more water).” (Heetjans, 1922:248).  Provided with a convenient supply of water piped directly into the home, and delivered at a higher pressure than through hydrants, European households naturally (i.e. as expected and designed) came to use the majority of this water supply. In 1929, the European population comprised only seven percent of the population, but yet consumed seventy eight percent of the residential urban supply, and this was after four years of a program specifically targeting the extension of supply to native households (1926-1929) (Eggink, 1930). With the worldwide depression in the 1930s, and the drop in colonial economy because of falling sugar prices, the campaigns of kampong improvement ended, along with the Ethical nature of colonial government policies (Cobban, 1974, 1988; Cote, 2003)17. By the time the colony’s economy had recovered, and programs of kampong improvement resuscitated in 1938-39, World War Two had begun. Suffice it to say, up until the end of Dutch colonial occupation, access to the city’s public services remained limited to Europeans and a very small section of the native population whose living standards had improved (Booth, 1988)18.            17 From 1927-1931, 1.25 million guilders was spent on improving kampongs in cities throughout Java, with the central government paying 50% of the costs; after 1931 government funding stopped, only to be revived again in 1938 when 500,000 guilders/year was again allocated to be distributed amongst all urban areas in the colony (Cobban, 1974, 1988). 18 Booth (1988) states that although income disparities amongst the native population grew during the last years of the colonial economy, there is no doubt that it was disparities between rather than within ethnic groups that continued to be the most obvious as income gaps even increased during the 1930s. In 1939 a European made 61 times the average Native wage, while a Chinese worker earned 8 times the average Native wage.  61 3.4 Surface Water Treatment Plants: Pejompongan I & II, 1950-1966  Following the tumultuous and protracted transition to an independent Indonesian government (1945-1950) – the events including Japanese invasion, the return of Dutch colonial government, and the subsequent war for independence – the city’s spring water supply system was all but destroyed (Fischer, 1959; van der Kroef, 1954). With a rapid increase in urban population19, and the capital city of Jakarta emerging as the center piece of the new nation of Indonesia, the postcolonial development of Jakarta’s urban water supply system was quick to follow formal independence. Using money from the French government and engineering plans from the French water supply company Degremont, the construction of a large scale surface water treatment plant (Pejompongan I) began in 1952 and was completed by 1957 (Hanna, 1959; PAM Jaya, 1992a). With a second large scale water treatment plant (Pejompongan II) completed by 1966, the volume of water circulating through the city was increased by almost ten-fold, from the colonial era (from 315 L/s to 3,400 L/s), creating the potential for more equitable distribution in the postcolonial city.  Illustration 3.3 Water Treatment Plant Pejompongan I                Source: PAM Jaya (1996). Instalasi Produksi Pejompongan I (Production Installation Pejompongan I). Jakarta, PAM Jaya. © PAM Jaya, 1996, by permission.    19 Population increased in Jakarta, from 823,000 residents in 1948, to 1.8 million in 1952 (Abeyasekere, 1989).   62  Map 3.4 Piped water supply network in Jakarta, 1950-1976                                     Source: Created by Ernst-Jan Martijn (2007).  WTP Pejompongan I was built on the edge of the urbanized area, drawing its raw water supplies from the Ciliwung river, whose waters were routed through the city by the Banjir Canal (Flood canal, built by the Dutch in 1923). Adding 2000 L/s of treated surface water to the city’s distribution network , Pejompongan I raised the potential urban water supply capacity to 2315 L/s. Although the city’s population continued to increase throughout the 1950- 60s, the construction of a second large scale water treatment plant in 1966 (Pejompongan II) added another 1000 L/s.        63 The city’s new water treatment plants were radical in that they changed the bulk of the source of water supply, from the colonial ideal of mountain spring water to treated surface water taken from the city’s flood canal. However, with the circulation of water throughout the city still determined by the colonial infrastructure artefacts (city network), the patterns of provision embedded in the urban landscape continued to guide postcolonial access, and in fact became inscripted within the postcolonial government rationality, which under the Architect President Sukarno entailed building up selective areas of the city to world class standards (see Djakarta, 1962). Becoming a component of President Sukarno’s monumentalist construction of Jakarta (see Kusno, 1997; LeClerc, 1993), the postcolonial development of Jakarta’s urban water infrastructure continued to facilitate fragmentation. The selective provision of networked water supply within particular spaces in the city continued to work both physically and discursively to reinforce desired differences. After independence, the difference rationalized as necessary to rule entailed distinguishing between the ‘modern ideal’ of what all Indonesian’s should aspire to, and the surrounding realities of the ‘sea of kampongs’ (Kusno, 2000). Investment into new water supply infrastructure for the postcolonial city thus continued the colonial pursuit of the centralized ideal to serve the selective strip of modernized spaces in the city, while simultaneously displacing any un-modern elements (bodies, buildings) outside areas of network access. Subsequent investment into Jakarta’s urban water supply went to support the construction of these differences. Building a new vision for the city and the nation within the existing foundations of colonial infrastructure, the city’s increased volume of piped water remained limited primarily to the formerly European areas of the city now being incorporated into the spatial geographies of power of the postcolonial government (Kusno, 2000; Leclerc, 1993; MacDonald, 1995). Therefore, while a minority of the elite residents occupied the residential spaces with privileged access to piped water, and consumed the majority of the public urban water supply, the “the bulk of the population remains without piped water supply and is dependent upon kampong wells; thousands must still resort for laundry, bath, or toilet, or all simultaneously, to the sluggish canals” (Hanna 1961:5). In 1959, fifteen percent of Jakarta’s residents consumed 2000 L/s of Jakarta’s new water supply, approximately between 220-330 L/capita/day, while the majority of the masses continued to rely on untreated river and ground water supplies for all of their needs (Fischer, 1959). While partly the product of a nascent and fragile postcolonial government lacking the technical and financial capacity to finance large scale improvements, the continuation of the colonial patterns of provision – and splintering of the city through infrastructure access – must also be recognized as a product of purposeful postcolonial planning. Throughout the 1950-1960s, piped water remained limited to the central areas of the city covered by the colonial era network because investments that were made into the distribution system remained limited to rehabilitating the existing pipes within the new Indonesian elite residential areas. Concurrently, there was no money invested to improve access for the rest of the population through even basic infrastructure like public hydrants (Abeyasekere, 1989). With the former  64 colonial neighbourhoods occupied by upper class elites seeking housing closer to the center of the city (Abeyesekere, 1989; Argo, 1999), the pipelines built to increase the water supply from Pejompongan I to the reclaimed modern areas of city did not increase the number of residents who had access to the piped water, but merely increased the volume of water that a minority of residents could access – looking suspiciously like the city’s pattern of urban water supply in the spring water era. However, while similar to the colonial pattern in that the postcolonial urban elites consumed the bulk of the residential urban water supply, unlike the colonial era, these residents paid virtually nothing for it. This in fact assured that the government, already in financial crisis by 1957 (see Robison, 1986), would not have the necessary funds to finance network extension or ‘basic needs’ infrastructure for lower-income areas. Water use by the elite households was not even metered until two decades later (1970s), by which time it was stated that the elites living in selectively served spaces of the city were provided with water ‘almost for free’ (PAM Jaya, 1992b). Paying a flat rate of 100 Rp/month  households were consuming on average between 220-330 L/capita/day, this was between three to six times the expected consumption of 60-90 L/capita/day according to which they were being charged (PAM Jaya, 1992b). The progress cited as illustrated by the construction of an ‘adequate water supply’, which provided the ‘urban conveniences embodying the concretization of the revolution’ (van der Kroef, 1954) was therefore only ever experienced by a few. The reality for the majority of the population (80% estimated to live in both the old central city and new peripheral kampongs), remained reflected in the “dirty brown canals and river arms used for any and all needs ranging from washing to excretion." (van der Kroef, 1954:158). In these early years of independence, the continuation of colonial patterns of water provision in postcolonial Jakarta is usually attributed primarily to the lack of technical and financial capacity of the newly independent nation (Chifos and Suselo, 2000; World Bank, 1974). However, while it is tempting to allocate cause to the legacy of colonial government and its spatial apartheid, the splintering of the city’s urban water supply infrastructure was also a product of purposeful postcolonial planning. Notably, the only investment made by the Indonesian government during this period to extend network access targeted elite areas, rather than the un-serviced informally settled low-income areas. Unlike during the colonial period, cost-recovery was not even the leading rational, as these upper class consumers did not even have their use metered. Tellingly, rather than installing public hydrants to address the non-existent services of dense informal settlements growing on the outskirts of the developed areas of the city (Hanna, 1961), in the 1950s the government built new pipelines to deliver water into elite residential areas on the outskirts of the city. Located on the southern edge of the city, the new upper-class suburb of Kebayoran Baru was originally planned by the Dutch post-war government as a satellite town with 100 000 luxury homes for occupation by Europeans and Indonesian elite (Asian Review, 1955; Dorleans, 2000; Roosmalen, 2000). After independence, President Sukarno reclaimed the archipelago of modern urban dwellings as a  65 ‘western showpiece’ intended to help modernize the image of Indonesia (Boddy, 1983). Changing the original plans drawn up by the Dutch colonial government to provide piped water within Kebayoran Baru through water from artesian wells (Kusumawijaya, 1990), upon the completion of Pejompongan I in 1957, pipelines were installed to attach the suburb to the city’s network, selectively channelling water outside of its central city network into the modern urban dwellings (documented in World Bank, 1974). Representing the only area of the city where investment into networked water supply was actually used to increase the percentage of the population with access, the rationality behind those entitled to access the public water supply services could not be made more clear, and the postcolonial pattern of water provision continued to mimic its colonial origins. In a powerful contrast to the “luxury housing built to satisfy the demands of the newly rich or newly powerful” (Hanna 1961:5) were the surrounding kampongs, denied access to the network water supply even through the low-cost provision of public hydrants (Abeyasekere, 1985). The exclusion of the majority of the population from access to the piped water supply was not only that of passive neglect. To be sure, as President Sukarno’s system of Guided Democracy spiralled into economic instability there was a lack of public finances preventing circulation of piped water through urban kampongs (Chifos, 2000), but this was also accompanied by an active displacement of the un- modern elements from the central, networked spaces in the city to the un-serviced periphery. With the decidedly un-modern kampongs and their characteristic poverty considered an embarrassment in their juxtaposition with the ‘world class city’, campaigns to remove the population, and rid the modern city of their  traditional, unsophisticated lifestyles thought to ‘lower the status of the nation’ continued throughout the 1950s and 1960s (Abeyasekere, 1989). The construction of the Asian Games complex and the adjacent inner-city thorough fare Jalan Thamrin and Jalan Sudirman connecting the suburb of Kebayoran Baru to the city center involved the removal of an estimated 47, 000 kampong residents (Abeyasekere, 1989). Following this emptying out of undeveloped land inhabited by the impermanent dwellings of thousands of recent urban migrants, general campaigns to remove the homeless from the city boundaries continued a more targeted approach (Nas and Mallo 2000), and served to depopulate the international spaces of the city that were also, not coincidentally, the areas served by the centralized water supply network.  Relegating much of Jakarta’s undesirable population to the peripheral land far outside of the already limited coverage of the network also served to reduce the perceived demand for piped water supply. By the end of the 1950s, many of the low-income residents of Jakarta had been moved out of the actual serviced area, and there were strong incentives for newly arriving migrants to also settle in these less regulated urban peripheries. Subsequently, this left only twelve percent of the population of the city with access to the piped water supply (Fischer 1959), and it is therefore little surprise the volume of water supply available for urban residents after Pejompongan I was considered more than adequate to meet demand (PAM Jaya, 1992b).  66 Meanwhile, the kampongs that remained within the interstices of the international spaces in the city’s center remained excluded from the provision of public services. While already in the colonial period the water supply through the provision of public hydrants had lagged behind population growth, there was now to be a thirty year period in which population densities would continue to rise while programs of kampong improvement and public hydrants lay dormant (1939-1969). With the absence of piped water supplies in the kampongs remaining the norm until the late 1970s and early 1980s (Argo, 1999), the migrants streaming into Jakarta during the 1950-60s created “vast block interiors that became the sites of the un-serviced urban kampong” (Cowherd, 2002: 173). The consistent exclusion of low-income, informally planned areas from network access left residents with little choice than to continue their historical reliance upon a combination of water supplies (see Argo, 1999). Digging shallow wells (open pits, no covers) and using the surface waters for washing and bathing, low-income residents only purchased drinking water from vendors in areas where groundwater and surface water quality was no longer adequate (Krausse, 1978). Perversely, the visual display resulting from the lack of public water supply and sanitation services in the kampongs, described as a “public strip tease” (Hanna 1961:2) by the international visitors to Jakarta, also served to highlight the kampongs’ undesirable and primitive status, and was used by government authorities to justify exclusion from the modern areas of the city meant to attract, impress, and reflect international ambitions. Subsequently, during the first decades of independence, access to the city’s public water supply infrastructure remained a material and symbolic emblem of a certain class of citizen, not a public right. With the majority of the city’s residents living in spaces without access to the water supply network, the dramatic increase in the volume of water circulating through the pipes only reinforced the discrepancy between the serviced urban spaces and residents, and those who were not. Ironically, although President Sukarno guided the development of the capital city in order to create a “new space intended to be different from both colonial Batavia as well as the surrounding sea of poor urban neighbourhoods” (Kusno, 2000:52), a similar hierarchy of urban citizenship was inscribed into the postcolonial urban landscape. Rather than a radical break with the past, the water treatment plants built within independent Indonesia were constructed upon the physical and discursive infrastructures of the colonial government. This continued the splintered pattern of both access to water supply and subsequent divisions between populations and urban spaces. Although the failure to provide an adequate urban water supply in Jakarta has always been blamed upon the fact that the new water supply infrastructure continued to be circulated through the already then leaky colonial era city network (Fischer, 1959), and the economic crises of the Guided Democracy era are given as a reason for the lack of expansion of the networked area (Berry and Sierra, 1978; Chifos 2000; World Bank, 1974), the significant financial investment made into the city’s  67 first two large scale water treatment plants20, and the extension of the network into certain newly modern areas of the city belies this myth and supports the argument that the distribution of networked water to ‘the masses’ was never part of the political project of first postcolonial government.   3.5 Large scale and ‘mini’ Water Treatment Plants: 1966-1990  When the economic and political instability of Guided Democracy was thrown over in a violent transition (1965), the New Order era government inaugurated a new vision for the nation, and fundamentally different relations between the state and society (see Bourchier and Hadiz, 2003; Pemberton, 1994). However, like its predecessor, this government maintained the same selective pattern of urban water supply within Jakarta, and worked to further splinter access to piped water, albeit along new axes of difference. Brought to power based on their promises of a ‘new order’ for the country, and their promises of development for the nation, the ‘old order of things’ still embedded within the foundations of the city was used to continue the pre-existing pattern of provision – channelling water into politically – and now economically - productive urban spaces, while bypassing the majority of urban residents. From 1965 to the late 1980s, the rehabilitation of existing infrastructure, the construction of two additional large scale water treatment plants, and new investment into mini-treatment plants went to benefit selectively targeted productive and modern areas of the city: industrial and commercial centers, and middle class beneficiaries of the New Order’s program of economic development. Therefore, although water supply production capacity increased three fold in over three decades of New Order rule (1960s-1980s), the provision of piped water supply was extended to less than one-quarter of the city’s population, and covered less than half of the urban area21 (Nihon Suido, 1988; Porter, 1996; World Bank, 1990b). By the late 1980s, the pattern of provision of piped water supply in Jakarta remained consistent with that of previous govermentalities and their physical works.  The shift in government rationality to the New Order ethos directed the production and distribution of Jakarta’s water supply from the 1960s-1980s. Under the New Order, the key to the development of the nation for the benefit of all Indonesians was the mobilization of free market capital (Cowherd, 2002), and the flow of urban water supply was subsequently guided to support the re- spatialization of Jakarta according to the needs of private sector investment. Following the strategic fragmentation of urban space and other urban services under the New Order’s mandate to facilitate economic growth and private sector development (see Cowherd 2002; Effendy, 2004), the  city’s third large scale water treatment plant was built to service a growing industrial area in the Northeast. With  20 WTP Pejompongan I is recorded to have cost Rp. 80 million (PAM Jaya, 1993b); given the fluctuating exchange rate throughout the 1950s this represents anywhere between U.S. $2-7 million dollars. The cost of WTP Pejompongan II is not recorded in Rupiah, but given as U.S. $7 million (PAM Jaya, 1992b). 21 In 1988 the piped network covered 213km2 of DKI Jakarta total area, 649 km2 (Nihon Suido, 1988).  68 Jakarta’s percentage of un-served population on the rise due to continued migration the city instituted a ‘closed city’ policy to curb demand on urban services22, and built Pulogadung WTP to service the Pulogadung Industrial Estate in East Jakarta, an area of the city that Suharto had targeted for the initiation of the New Order strategy of industry led economic growth (Cybriwsky and Ford, 2001; Castles, 1989). For the seventy six factories already located in the P.I.E. (Castles, 1989), 1000 L/s of treated water was already being provided to the area through the construction of a booster pump and large diameter pipeline, channelling 770 L/s of the increased capacity of Pejompongan II (PAM Jaya, 1992b; World Bank, 1974). This was followed in 1982 with the completion of Pulogadung WTP, providing another 1000 L/s to the industrial area (PAM Jaya, 1992e). When in 1987, the capacity of Pulogadung WTP was further increased to 4000 L/s this supply was actually under-utilized, due to the low demand from industry and the lack of distribution networks to channel water to those surrounding residential areas that did need it. With the industries located in P.I.E. largely favouring cheaper sources of water supply there was excess capacity, due to the fact that the planned installation of 550 kilometres of distribution pipes in surrounding neighbourhoods was never completed (JICA, 1997). Reports on the coverage of the water supply network in the 1990s continued to document that networked water supply in eastern Jakarta was still limited to industrial areas (Argo, 1999).    22 From 1966-1976 the population of Jakarta increased from 3.6 to 5.7 million; to curb population growth in the capital in 1970 Jakarta was declared a ‘closed city’ for migrants (Abeyasekere, 1989; Critchfield, 1971).  69 Map 3.5 Development of Jakarta’s centralized water supply network, 1977-1982                                      Source: Created by Ernst-Jan Martijn (2007).  Parallel to the New Order’s program of industrial economic growth, the extension of network water supply into upper class residential areas continued in a similar pattern as the ‘older order’, and its colonial antecedents. Providing a socio-political complement to the program of industrial water supply, residential access to the piped water network was largely limited to the economically mobile and politically obedient citizens of the New Order. Expanding the production capacity of the pre-existing WTPs (Pejompongan I&II), the additional volume of water went to provide the urban elite, who in the 1970s and 1980s were still living in the centrally located elite residential areas, and already integrated  70 within the city’s network since the colonial era (Menteng, Kuningan, Kebayoran Baru)23. Throughout the 1960-70s, the investments into increased production capacity and distribution went to benefit the upper class neighbourhoods, and the modern strip of highrises, luxury hotels, and developments along the city’s center thorough-fare supported by this demographic. Following the rehabilitation of the existing WTPs, there was sufficient volume of water to provide for between forty to sixty one percent of the population, but this public service was still limited to less than fifteen percent of Jakarta’s residents (Abeyasekere, 1985; KIP, 1976), and ninety percent of the kampong population was recorded as being without access to piped water (KIP, 1976; Taylor, 1983a). When from the late 1970s-onwards money went into extending the centralized network to actually include other residential areas, the new areas included within access to water were those that followed the from the New Order urban development trajectory. This New Order trajectory, as noted by other scholars, was one of marked socio-spatial fragmentation (Cowherd, 2002; Kusumawijaya, 2004). With the emergence of an urban property market in Jakarta, its growth due in part as an investment strategy for the New Order’s beneficiaries of the industrial development (Cowherd 2002), the identification of select new spaces worthy of network expansion led to strategic development of water supply services in limited areas of the city to create isolated archipelagos of service. First, in late 1970s, the residential area of Pondok Indah (Beautiful Neighbourhood) was the recipient of the first new distribution network built for residential areas24. With the installation of a primary, large diameter, pipeline channelling water from Pejompongan I to the recently built gates of Pondok Indah, the first gated community in Jakarta became part of the selectively serviced areas of the city. Advertised as a ‘residential enclave’, Pondok Indah was defined as purposefully isolated from its surrounding neighbourhood; while its spatial exclusion was ensured by the construction of a golf course and the re- routing of a river (see Cowherd, 2002), it also appears on city maps as an island of services amidst non- networked surrounding of Southern Jakarta. The development of a new wave of high income residential areas in the early 1980s, supported by New Order planning policies and economic growth, effected an even greater physical fragmentation of the city’s networked water supply. Housing estates built in the north western and north eastern areas of the city were accompanied by the construction of mini-water treatment plants that were physically completely isolated from the centralized supply system. Financed through the partnership of PAM Jaya with the provincial government, industry, and real-estate developers, the production facilities and distribution networks produced small volumes of clean water, and were built only to provide for isolated residential areas and industrial areas. A total of eight mini-water treatment plants built from  23 In 1967 Pejompongan I was upgraded, increasing its capacity from 2000 to 3000 L/s; in 1973 the capacity of Pejompongan II was increased from 1000 to 2000 L/s (JICA, 1997). 24 Pondok Indah was developed by the New Order urban developer Ciputra, a Chinese-Indonesian who was financed by Sukarno crony Salim (Cowherd, 2002). Pondok Indah was one of Ciputra’s first projects, pioneering the ‘gated community’, and while still an upper class elite residential district today in South Jakarta, in the early 1970s it was transformed from 720 ha of rubber plantation (Cybriwsky and Ford, 2001).  71 1977-1982 provided such places as Sunter Paradise, Taman Kota (Garden city), and Citra Gardens with piped water supply. The beneficiaries of the mini-treatment plants and the islands of services they produced area are beautifully illustrated in PAM Jaya’s 1992 publication, where advertisements for real estate properties receiving water from the mini-treatment plants proclaim their ‘24 hour piped water services’ as an attractive amenity to prospective buyers and investors (PAM Jaya, 1992).  Table 3.3 Mini-water treatment plants, 1977-1982  Year Name/Location Production Capacity  (L/s) 1977 Cilandak 200 1980 Pesing 5 Taman Kota 50 Muara Karang 100 Pejaten 5 Sunter 50   1982 Cakung 25 Source: Data from PAM Jaya (1992a), JICA (1997).   Illustration 3.4 Advertisement for Sunter Paradise housing estate, 1992                         Source: PAM Jaya (1992). 70 Tahun PAM JAY permission.  Advertised with the words, ‘Semakin Lengkap, Paradise housing estate profiles the services offe telecommunications, piped water, and other ame    A: 1922-1992. Jakarta, PAM Jaya. © PAM Jaya, 1992, by Semakin Elit’ (Increasingly Complete, Increasingly Elite), Sunter red to potential residents: strategic toll road location, electricity, nities. 72 The splintering of the city’s urban water supply system through the production of physically disconnected production and distribution infrastructure continued into the late 1980s. The construction of the city’s fourth large scale water treatment plant, Cisadane, was actually built physically outside of the DKI urban boundaries, as its production capacity of 3000 L/s was planned in order to provide for the adjacent ‘new town’ areas of Bumai Serpong Damai (BSD). Like Pondok Indah, BSD was built by the prolific New Order urban developer Ciputra and his Suharto-crony financier Salim, and it was again a pioneering project, the first ‘new town’ built from the concept which soon came to dominate urban residential development surrounding Jakarta (Firman, 2004; Goldblum and Wong, 2000; Leisch, 2000, 2002). Bumai Serpong Damai, which was located twenty kilometres southwest of Jakarta in the neighbouring province of Tangerang covers 6000 hectares, and was planned for an eventual population of 600,000 residents (Cybriwsky and Ford, 2001; Struyk et al., 1990; Winarso and Kombaitan, 1997). Also, like Pondok Indah, the elite residential enclave (and its golf course, designed by Jack Nicklaus) was to be provided with piped water from publicly financed water supply infrastructure. The initial economic feasibility study for Cisadane focused on the provision of BSD residents with piped water from the new treatment plant (pers.comm. PT. Ekamitra Engineering SHD.BHD, 10 January 2007); financed by the central government through a loan from the French government, and built through a Build-Operate-Transfer contract with the French water supply company Degremont (World Bank, 1990), the provision of water to higher income residents repeated the pattern of New Order development. However, with the three projected phases of growth of BSD not occurring according to initial targets because of the 1997 economic crisis, only 200 L/s of the capacity is currently used for BSD, and the vast volume of excess water produced by Cisadane WTP is now sold to the private sector water supply partners for distribution to the western half of Jakarta (pers.comm.PT Tirta Cisadane, 15 August 2007). Notably, as will be explained in more detail in Chapter Five, this re-integration could only occur after the World Bank financed the construction of a pipeline connecting Cisadane back into Jakarta’s central distribution network. In stark contrast to the areas of the city designated as productive and modern emblems of the New Order, the kampongs remained areas outside of the New Order’s promise of development. The first program of postcolonial kampong improvement begun by Jakarta governor Ali Sadikin ran from the late 1960s to the mid-1970s, paralleling the period in which water supply infrastructure investments were made to continue to provision to upper class neighbourhoods, and provide new supplies for industrial growth25. However, although marking the first postcolonial program for water supply to low- income neighbourhoods, the kampong improvement programs produced only marginal improvement in the access to piped water (Argo, 1999; Taylor, 1983a). This is because, like in the colonial era  25 KIP began as a local government initiative, with 50% of settlement improvement costs paid for by the Jakarta government, and 50% by the residents. In 1974 KIP was taken up by the World Bank, where upon it was extended as a national development program in Indonesia, and managed by the Central Government up until early 1980s (World Bank, 1995a).  73 programs, only a very small percentage of the budget went towards the provision of piped water supply; from 1969-1974 only five percent of the KIP budget was spent on drinking water facilities (Erni and Bianpoen, 1980). Despite the pressing needs of the kampong residents for basic sanitation and clean water, the projects money was spent on roads and footpaths (Abeyasekere, 1985). The low priority placed on the provision of piped water in the first decade of kampong improvement meant that public hydrants were still a rare occurrence in kampongs in the late 1970s, and drinking water became an increasingly expensive commodity for the poor (see Abeyasekere, 1987; Argo, 1999; Krausse, 1978; DKI Jakarta, 1976). During this time, the poorest families of one low-income neighbourhood in central Jakarta are documented as paying up to five percent of their daily income for a daily allotment of eight litres of water (Jellinek, 1985). Other surveys record low-income households as paying between 10- 25% of daily wage on water supply (Papanek, 1975)26.  After a decade of kampong improvement (1966-77), still only fifteen percent of Jakarta population had access to networked water (Abeyasekere, 1987). Although after the World Bank took up KIP in 1974 more emphasis was put on the provision of piped water into kampongs, this plan was frustrated by the government’s priorities of economic versus social development. The emphasis on economic development during the New Order was reflected in the mandate of the provincial water supply company, which became a publicly owned company in 1973 (PAM Jaya, 1992a).  Mandated “to take part in carrying out the national and regional economic development" (PAM Jaya, 1992b:65), the city was reluctant to extend pipes into low-income areas and as a result the public hydrants that were installed were sometimes not provided with water (World Bank, 1995a). Alternately, if and when public hydrants were actually installed and connected, often the ‘benefits’ deriving from this public infrastructure were appropriated by local government officials who were condoned under the New Order regime to extract their own profits from the public system (Dian Desa, 1990; Server, 1996). Therefore, although the World Bank had envisioned the implementation of an initial ‘minimal cost, basic needs’ water supply infrastructure, and a gradual transition into more complete network coverage through household connections (World Bank, 1974; DKI Jakarta, 1976), the evaluation of the World Bank’s KIP records failure: “Although changes in drinking water source are in the predicted direction [towards more residents are using piped water rather than shallow groundwater], they are surprisingly modest in the aggregate. While it was the intention of KIP to provide public water standpipes throughout the improved areas this component has not been implemented consistently. The primary reason has been the reluctance of the public water supply company to extend its  26 Papanek (1975) records households paying between Rp.20-50 for 20 litres of water, while daily wage is recorded as Rp.200. With the exchange rate of Rp.415 equivalent to $1 U.S., this means households earned approximately $0.50 U.S./day, and paid either $0.05 or $0.12 a day for a water supply of 20 litres.  74 distribution network into so many low-income kampungs because of presumed difficulty in cost recovery in such areas.” (Taylor, 1983a:101).  With the city’s water supply company following the mandate of the New Order, by the end of the World Bank’s program of kampong improvement in Jakarta, the numbers of public hydrants available to low income residents actually declined. In 1987 there were barely 1000 public hydrants spread throughout the city (Porter, 1996). In comparison, the city of Surabaya, which had one quarter of the population of Jakarta, had two times as many public hydrants (Porter, 1996). Understandably, under the New Order ethos, PAM Jaya was ‘reluctant’ to encourage public hydrants because they were simply less profitable (IWACO and WASECO, 1990). PAM Jaya’s (legal) profits came from household connections, and having to meet financial performance objectives, “they continually stymied the construction of public hydrants which made little profit for the company and required significantly more maintenance” (World Bank, 1990b:115). By the end of the 1980s Jakarta’s water supply capacity had increased by more than three times, but still less than one fourth of the city’s residents had access to the piped water supply (Porter 1996). In addition to limited coverage, distribution was decidedly inequitable. Surveys conducted in the early 1990s reported that in low-income neighbourhoods more than fifty percent of the population relied on public hydrants for drinking water supply, and paid up to five times more for their water than residents with household connections; only ten percent of the residents of these low-income areas accessed water through individual connections (Adzan, 2001). Like the patterns of provision from previous eras, the urban water supply infrastructure investments made from 1960s-1980s favoured the select few residents who qualified as citizens. Building upon foundation of the city’s splintered urbanism, the New Order’s transformation of Jakarta’s urban space came to reflect “a deepening social dualism that provides a higher quality environment and more sophisticated infrastructure for the consumer class without having to provide it for all citizens.” (Cowherd, 2002: 271). In conclusion, while the explanation usually given for the lack of service coverage in Jakarta by the Western development agencies is that of a lack of local financing, inappropriate tariff structures and insufficient economic efficiency coupled with a rapid rate of population growth that continually outpaced the investment into public services and urban infrastructure (Chifos and Suselo, 2000; World B