Barbara Heer Cities of Entanglements

**Barbara Heer**, born in 1982, is an anthropologist living in Basel. She holds a PhD from the University of Basel. Her work focuses on urban diversity in Southern Africa and Switzerland. She is engaged in practice and politics.

Barbara Heer Cities of Entanglements

Social Life in Johannesburg and Maputo

Through Ethnographic Comparison

Dissertation at the University of Basel

Published with the support of the Swiss National Science Foundation

### Bibliographic information published by the Deutsche Nationalbibliothek

The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliografie; detailed bibliographic data are available in the Internet at http://dnb.d-nb.de

This work is licensed under the Creative Commons Attribution 4.0 (BY) license, which means that the text may be remixed, transformed and built upon and be copied and redistributed in any medium or format even commercially, provided credit is given to the author. For details go to http://creativecommons.org/licenses/by/4.0/

Creative Commons license terms for re-use do not apply to any content (such as graphs, figures, photos, excerpts, etc.) not original to the Open Access publication and further permission may be required from the rights holder. The obligation to research and clear permission lies solely with the party re-using the material.

### First published in 2019 by transcript Verlag, Bielefeld

#### © Barbara Heer

Cover layout: Maria Arndt, Bielefeld Cover illustration: Barbara Heer, Maputo, 2010 Proofread by Alexa Barnby, Johannesburg Printed by Majuskel Medienproduktion GmbH, Wetzlar Print-ISBN 978-3-8376-4797-6 PDF-ISBN 978-3-8394-4797-0 EPUB-ISBN 978-3-7328-4797-6 https://doi.org/10.14361/9783839447970


### **Acknowledgements**

I would like to express my immense gratitude to everyone – the urban dwellers of Maputo and Johannesburg who let me into their lives, research assistants, colleagues, friends and family, and, last but not least, the University of Basel and funding institutions – without whom this book would not have been possible. The PhD project was funded by an initial grant from the Janggen‐Pöhn Foundation, a three-year scholarship from the Humer Foundation for Academic Talent, a travel grant from the Swiss–South Africa Joint Research Programme, a one-year scholarship from the Forschungsfond (Funding for Excellent Young Researchers at the University of Basel), and it received funding by the Swiss National Science Foundation for open access publication.

Special thanks go to my research assistants Thabo Mopasi, Nnana Tau and Fernando Tivane. As friends, key informants and often co-researchers, they shaped this ethnography distinctively. I am also very grateful to Cristiana Cuamba, Luversan Gerard, Abigail Mujaji, Dona Alcira and many others who offered me their friendship and hosted me in Maputo and Johannesburg. In Johannesburg, Noor Nieftagodien, and in Maputo, Elísio Jossias, acted as local supervisors. I would like to express my gratitude to them and their respective institutions, the University of the Witwatersrand and the Universidade Eduardo Mondlane.

Special thanks go to my PhD supervisors Till Förster, Ulf Hannerz and Elísio Macamo. I had the pleasure of conducting my PhD in an inspiring research environment at the University of Basel, being based at the Institute of Social Anthropology and the Centre for African Studies. Many thanks go to Veit Arlt, Carole Ammann, Kerstin Bauer, Julia Büchele, Sandra Burri, Michelle Engeler, Piet van Eeuwijk, Divine Fuh, Andrea Kaiser-Grolimund, Kathrin Heitz, Andrea Kaufmann, Rita Kesselring, Lucy Koechlin, Peter Lindenmann, Filomena Mazumdar, James Merron, Brigit Obrist, Silke Oldenburg, Sophie Oldfield, Joschka Philipps, Aïdas Sanago, Fiona Siegenthaler, Regina Stalder and Sandra Staudacher. I also would like to express my thanks to colleagues based in other corners of the world, among them Ana Aceska, Alexa Barnby, Antonádia Borges, Claudia Gastrow, Mark Lewis and Annika Teppo. Finally, many thanks go to my family, my husband Martin and my friends who supported and accompanied me in many ways during these fascinating years of doing fieldwork and writing an ethnography.

## **Introduction** Cities of Entanglements

Imagine yourself walking along a narrow sandy path in a neighbourhood of Maputo. Your walk takes you past unplastered brick houses surrounded by yards in which women grind corn and look after children. Thorn bushes, sheets of corrugated iron and walls of crumbling brick demarcate the boundaries of the home spaces, providing both privacy and protection. Children play in the sand and women sell fruit on little stalls (*bancas*). Suddenly, you are confronted with a concrete wall made of unplastered cement bricks and about two metres high. Barbed wire tells you that you should not try to trespass. Across the wall you can see the roofs of colourful villas. If there were no wall, you would see a few security guards and domestic workers sitting in the shade of a tree besides an otherwise deserted street. You might see a resident backing out of a driveway in her luxury SUV (sports utility vehicle). Hidden from sight, you can only imagine the ostentatious mansions with their carefully maintained gardens and swimming pools behind the electrified fences (see the photographs at the end of this chapter).

This particular wall is situated in the northern part of Maputo, and separates an area known as Polana Caniço from another area colloquially referred to as Sommerschield II. The anthropologist Fernando Tivane, a research assistant on this project, showed me this wall at the beginning of my fieldwork in Maputo in 2010. For him, as for many other urban dwellers in Maputo, this and similar walls symbolise the stark urban inequalities and the wish of powerful urban groups not to have to see poverty and thus withdraw themselves from it. Such walls are encountered in many cities across the world, especially in the Global South. In societies with a colonial past and continuing inequality under neoliberal capitalism, walls are often built by those with more power to protect themselves and exclude less powerful groups whom they construct as urban 'others' (Spivak 1985).

Walls are an omnipresent spatial form used across cities and villages to shape the relations of proximity and distance between those who are inside and those who should remain outside. Besides being about materiality, by physically preventing people from crossing the threshold thus created walls are also about the imaginations of urban dwellers about how people in the city live or should live together – or rather apart. By building walls urban dwellers aim to shape the social worlds they live in. Building walls enables them to construct – spatially and in their imagination – urban lifeworlds which seem disconnected from the surrounding city. For many urban elites in African cities and elsewhere, such apparent enclaves – elite neighbourhoods seemingly disconnected from their surroundings – promise the illusion of residential spaces under control and emptied of the 'other', like the urban poor, the criminal, the 'racial' or the national other.<sup>1</sup> Sometimes this involves the desire to build road closures (see chapter 5), or resistance to public housing for the poor in their neighbourhood (see chapter 4).

Walls are not only a structuring device for urban societies but also become their symbol. As symbols, walls send a clear message about the relationship between those inside and those outside, for example that the wealthier possess the power to insulate themselves from the necessity of sharing and interacting with the poorer (Marcuse 1997b: 109). Walls thus "suggest a particular set of relationships between those on the opposite sides of the boundary: Separation, distance, fear, tension, hostility, inequality, and alienation" (ibid: 103). The contradiction between building walls and the ethics of living together can lead to moral outrage and popular uprising (see chapter 5).

In writing about cities shaped by inequality such as Johannesburg and Maputo, walls have come to symbolise the vexed conviviality of urban elites and less aff luent groups. In her seminal work on São Paulo, Caldeira (2000) argues that in what she calls *cities of walls,* differences become organised through the building of walls, separation and the policing of boundaries. She defines *cities of walls* as "cities of fixed boundaries and spaces of restricted and controlled access" (ibid: 304), and as cities where the "experiences of separateness" become dominant (ibid: 355). Murray adapts the framework, *cities of walls,* to Johannesburg and argues, based on a reading of the existing literature, that "spatial strategies of separation, segregation, and isolation operate by marking boundaries and registering differences, imposing partitions and distances, building barriers, multiplying rules of exclusion, designing spaces of avoidance" (Murray 2004: 150). Johannesburg emerges in this narrative as consisting of "disconnected 'micro-worlds' cut off from one another" (ibid: 142).

While the framework, *cities of walls,* can assist us in grasping certain aspects of the way spaces are organised in unequal cities, this book will argue that it constitutes only a partial lens which omits many aspects of urbanity in these cities. This is not least because the focus on walls and the dividing power of segregation ref lect the perspectives of urban elites while largely ignoring the perspectives of less aff luent urban dwellers. Now imagine yourself a domestic worker. You wake up in the morning in Polana Caniço, and every day you walk to Sommerschield II where you are let into one of the colourful mansions – let's take the pink one – by your employer, a property owner. You carry out your daily routine, cleaning the house, washing the dirty laundry and maybe looking after the children. During your lunch break you gossip with acquaintances

<sup>1</sup>  The fact that many texts about contemporary South Africa still use racial terminology of ten puzzles readers not familiar with the South African context. Yet South African society was and continues to be stratified along the troubling category of 'race'. Race is a social construct with meanings and assumptions that have changed over time and dif fer acrosscontext. In this book, when I write about race and use categorieslike 'black' or'white', itis not because I believe in the biologicalrelevance ofthese terms, but because, associalconstructs and politicalcategories,they shaped and continue to shape urban life in Johannesburg and to a lesser degree inMaputo aswell. WheneverIreferin this book to the racialcategories as introduced by the apartheid government or the Portuguese colonialists (as they were used in official discourses, official documents, laws or censuses), I capitalise them (White, Black, Native). When I use the terms uncapitalised (black, white), I refer to the racial and ethnic categories used in everyday life by the informantsthis book is based on. I use them as descriptive terms, just asthe urban dwellers do. Because ofcapitalisation rules, 'Indian' is alwayscapitalised.

who work in the neighbouring houses about family and neighbourhood life, as well as about how you are treated by your bosses, and you share intimate knowledge about their latest family intrigues. You buy food for your own family with the income you get from this job and you know that your employer trusts you and depends on you. The job is hard, sometimes demeaning and receives little recognition from your boss or your own family, yet it will help you to replace the hedge around your simple house in Polana Caniço with solid bricks one day.

The everyday lives of the family in the pink house and the domestic worker from Polana Caniço are deeply entangled with each other. They depend on each other and they interact with each other daily, even if the urban elites tend to pay little attention to this. Besides the daily encounters between domestic workers and their employers in the intimate space of the elite's home, urbanity in Maputo and Johannesburg is also characterised by f leeting interactions at shared shopping facilities, by shared experiences of praying together in church and mosque spaces, and by confrontations around increasingly scarce and contested urban land. Such relationships between people and between spaces, which are thought of and felt to be different from each other, I define as *entanglements*. Entanglements are, based on Nuttall, "a set of relations, some of them conscious but many of them unconscious, which occur between people who most of the time try to define themselves as different" (Nuttall 2009: 12). Such entanglements, based on the shared use of spaces and the interdependencies, exist both inside and outside the walls. They emerge often invisibly in spaces of everyday lives and go unnoticed, not least because their existence often contradicts hegemonic understandings of the city. As entanglements challenge the elites' desire for segregation, there are social processes at work which make them invisible. In this book, it is therefore often the less powerful urban dwellers who point out the existence of entanglements, while for urban elites, as well as academics, they often constitute *blind fields* (see chapter 8).

Colonialism and apartheid forced urban dwellers and scholars to think about cities like Johannesburg and Maputo in terms of abstract, homogenising and simplifying categories of difference like Black and White, European and Natives, *Cimento* and *Caniço*, suburb and township. Such colonial binaries, enforced with state power and inscribed in space through legislation, continue to be powerful in real terms. Despite all the changes that have taken place, these binaries still inf luence the way urban dwellers understand their worlds and scholars analyse cities. The binaries, however, also make us blind to what happens at the boundaries, the intersections, the sites of encounter, transgression and multiplication. By drawing our attention to separation and division instead of connection, these binaries cause us to neglect the sites and moments of entanglements where what is thought of as separate comes together, intersects and becomes altered by the other.

The main argument of this book is that entanglements are constitutive of cities. They are as constitutive of cities as walls. *Cities of Entanglements* presents a comparative urban ethnography of entangled everyday lives in two contemporary African cities which challenges existing approaches that analyse these sites through the lens of segregation. The ethnography roots itself in the way in which people constitute their quotidian lives in adjacent yet socially and spatially segregated neighbourhoods: urban dwellers living in the township of Alexandra and the suburb of Linbro Park in Johannesburg, and residents of the *bairro* Polana Caniço and the elite neighbourhood of Sommerschield II in Maputo. While acknowledging that the fast-changing cities

of Johannesburg and Maputo are shaped by colonial and postcolonial forms of segregation, the book examines the agency and practices of urban dwellers in not only inhabiting divisions, but also overcoming and reformulating them. Despite the many spatial and social boundaries separating the neighbourhoods, they are fundamentally entangled with each other through labour relations, struggles over urban land, and visions of the right way of living together, or rather living apart. In religious spaces, diverging social groups become integrated with each other through faith-based charity. Shopping malls emerge as important spaces of public life in these cities, where urban lives become entangled through chance encounters, competition and fantasies. By exploring such diverse spaces of encounter, the book produces a coherent account of the processes that constitute and transform these urban spaces. Entanglement therefore becomes a means to understand these cities in new ways, highlighting the tentative forming of relationships and linkages across imagined boundaries, even as elites seek to reinstate divisions. This book replaces the framework of *cities of walls* with the framework of *cities of entanglements* and develops a new narrative about social life in what have long been treated as *divided cities.* Adopting a comparative perspective, the book shows how urban entanglements are based on recurring forms of conviviality which take on distinct forms in both cities. Grounded in everyday life in African cities, the book becomes an exploration of how urbanity is changing in unequal cities in a way which resonates beyond the Southern African cases*.* 

### **The Setting**

As mentioned, this book explores sites of encounters where the lives of urban dwellers from two urban areas, comprising four unequal and adjacent neighbourhoods– Polana Caniço and Sommerschield II in Maputo, and Alexandra and Linbro Park in Johannesburg – become entangled. The locality and specificity of these neighbourhoods as well as the diversity of the residents living there stand at the centre of this ethnography, which is why they are brief ly introduced here.

Alexandra and Linbro Park are situated in the north-east of Johannesburg, in what is today Region E of the City of Johannesburg. When these neighbourhoods emerged (1912 for Alexandra and the 1930s for Linbro Park), they constituted the outer fringe of the city. Suburbanisation and urban sprawl have drawn them into the midst of the so-called northern suburbs and they are now surrounded by desegregating neighbourhoods, highways and edge cities<sup>2</sup> with office complexes, malls and gated communities. Alexandra, with a surface area of approximately eight square kilometres and an estimated population size of 340,000 people (Alexandra Renewal Project 2005), experienced decades of unequal investment in public amenities during apartheid. As a result, infrastructure such as electricity, roads, sanitation and schools is still today insufficient in relation to the high population density. Bounded by the Marlboro industrial area to the north and west, London Road to the south, and the Jukskei River to the east, Alexandra is a complex, dynamic and sometimes violent place, with a history of state oppression and political resistance. Because of its size and history, it hosts a more

<sup>2</sup>  The notion of'edge city' has been introduced byGarreau (1991) bywhich he meansthe decentralisation ofshopping and service-sectorwork spaceswhich leadsto new, decentralised urban centres.

multifaceted population than the other three neighbourhoods. Descendants of former property owners comprise an old and inf luential milieu in the township. Many tenant families and hostel dwellers have also been living there for decades. Yet, there is also a large shifting, highly mobile population, many of them with strong connections to their other homes in rural areas or other African countries.

Linbro Park is much smaller in size and population, with a surface area of five square kilometres and only about 1000 inhabitants (own estimation). At the time of the research, it consisted of around 200 large stands between one to two-and-a-half hectares in size, some used for business purposes, but mostly residential, as well as a couple of new warehouses. In the east, the suburb borders on empty veld belonging to Modderfontein where private investors are planning to build a new edge city. To the south, Linbro Park borders on new office parks, and to the west it could almost touch Alexandra were it not for the barrier formed by a landfill site and a national highway (N3).3 Because most of the land is zoned as 'agricultural holdings' the area lacks infrastructure like public sewage and stormwater drainage, which is also why the municipality refers to it as a 'peri-urban' suburb (City of Johannesburg 2010: 5). The properties are largely owned by members of white milieus.<sup>4</sup> Many of them host tenants from racially and ethnically diverse lower and middle-income milieus. Linbro Park is also home to a large number of domestic workers, gardeners and handymen who live and work in Linbro Park.

In the local taxonomy, Alexandra is called either a 'location' or a 'township'. Both terms denote a residential area where during apartheid the non-White population was compelled to live.<sup>5</sup> The idea of 'township' gains its meaning from the relation to White residential areas, the 'suburbs', of which Linbro Park is considered to be one. In everyday language, 'suburb' denotes the formerly Whites-only areas with low residential densities, as well as newly built neighbourhoods. In the local taxonomies of places and people, the binary township–suburb is associated with oppositional stereotypes about their residents, lifestyles, housing types, densities, governance structures, crime levels and neighbour relations. While the notion of suburb in contemporary Johannesburg tends to be associated with free-standing houses and middle-class or aff luent lifestyles, either predominantly white or increasingly mixed, townships continue to be associated with poverty, high density, crime and black milieus. Although Johannesburg has changed dramatically since the end of apartheid, and the types of neighbourhood have multiplied, this binary is still relevant for academic and everyday understandings of the city.

<sup>3</sup>  The suburb used to be larger: to the north, sections of the suburb were transformed into the Linbro Park Business Park and latersome propertieswere expropriated for the construction of the Gautrain.

<sup>4</sup>  Milieus are groups of like-minded people who have similar values, similar ways of life and similar relationshipsto others(Hradil1999: 420). See later in thisintroduction.

<sup>5</sup>  It should be noted that 'township' is also a legal term in South Africa with meanings deviating from the above described binary suburb–township. When agricultural land is meant to be subdivided into smallerstands, it needsto go through an administrative processthroughwhich itislegally proclaimed a 'township' (Mabin 2005b: 19). Many residential areas which the apartheid state created for the non-White population (referred to as'townships' in everyday use)were legally not'townships', butthe land was government-owned, while most of the white suburbs were originally created through township proclamation (ibid:19).

In Maputo, this book focuses on the neighbourhoods of Sommerschield II and Polana Caniço, which are located north of the inner-city neighbourhoods of Coop and Sommerschield, which adjoin the campus of the Eduardo Mondlane University (UEM). In official municipal documents, the two areas do not exist as separate neighbourhoods but both fall under Polana Caniço A, in district 3 of Maputo. Polana Caniço A is one of the biggest and most densely populated administratively defined neighbourhoods in the city, with a population of about 45,000 inhabitants (according to the neighbourhood secretary in 2010) and covering an area of 222 hectares (Malaulane 2005: 17, Araújo 2006: 7). Within Polana Caniço A, fieldwork took place mainly in Sommerschield II, in an area called Casas Brancas as well as an unplanned section of Polana Caniço which adjoins Sommerschield II.6 The level of urban services in the unplanned section is very low, erosion and natural water drainage making the land challenging for permanent settlement. Polana Caniço may appear to outsiders as a homogenous place, yet the population is considerably diverse. There is a milieu of longterm residents who moved to Maputo at the time of the country's independence. Then there are milieus of war refugees from southern Mozambique (Gaza, Inhambane and Maputo province) who f led to Maputo in the 1980s (Costa 2007). A third group consists of Swahili Muslims from the northern provinces (mainly Zambézia) who also came to Maputo during the war or more recently. There is also a group of residents who used to live in the city centre until the early 1990s, and then could not afford to continue living there when housing was privatised. There are also f luctuating mobile groups of students, young couples and migrants on the move from the rural areas to the cities and to South Africa.

The adjacent elite neighbourhood of Sommerschield II consists of about 130 to 150 free-standing houses, laid out in an orderly fashion along planned streets. The neighbourhood takes the shape of a triangle, bordered by Julius Nyerere Avenue, the land reserves of the UEM, and the Rua Tenente General Fernando Matavele, behind which the unplanned sections of Polana Caniço have evolved. The layout was designed in such a way that there is only one street, the *Rua do Cravo*, which leads directly to the adjacent poorer areas of Polana Caniço; the rest of the boundary is constituted by walls. Many of the current homeowners and residents belong to Maputo's Frelimo elite who draw their power from the ruling party. A newer group comprises expatriates working for embassies, NGOs and transnational companies. They tend to rent their houses from local elites who no longer live there themselves. The third group of residents consists of wealthy members of the Indian merchant community of Maputo (mostly Mozambicans of Indian or Pakistani descent), the majority of them Muslims, while a minority are Christians from Goa. This group can again be subdivided into Indian milieus with a long-term presence in Mozambique and close relations to the Frelimo elite, and a

<sup>6</sup>  Maputo's neighbourhoods are characterised by dif ferent degrees of planning,which iswhy the municipality distinguishes between 'unplanned', 'planned' and 'unofficially planned' areas. Completely 'planned' areas mainly include the colonialcity centre. The elite neighbourhood of Sommerschield II is an 'unofficially planned' area: a small-scale subdivision layout was established by municipal officials in the 1990s or early 2000s, yet it was never officially approved by the municipal assembly (Jenkins 2013:184). 'Unplanned' areas, like sections of Polana Caniço, possess no plan known to the municipality. For a more general comment on such distinctions see the footnote on informality later in this chapter.

group of more recent arrivals who moved to the neighbourhood only recently, sometimes as tenants.

In the local taxonomy, Polana Caniço is referred to locally as a *bairro* or *subúrbio* adjoining the city centre. *Bairro* or *subúrbio* and *periurbano* or *periphería,* the last two expressions referring to emerging neighbourhoods on the outskirts of the city, all evoke images of informality, low levels of urban infrastructure and economic hardship in Maputo.<sup>7</sup> These neighbourhoods comprise what during colonial times was called the City of Reeds (*Cidade de Caniço*), the 'informal' slums and shantytowns of the colonial subjects, the Natives. *Caniço,* which still today forms part of the name, Polana Caniço, refers to the non-permanent building materials which characterised the houses in these African neighbourhoods in what was then called Lourenço Marques (see also Bertelsen 2014). City of Cement (*Cidade de Cimento*) referred to the city centre, the spaces of the colonial city which were inhabited by the colonial subjects and *assimilados* (see below), urbanised according to what was seen as European standards, and associated with whiteness. This old colonial dichotomy becomes disrupted or rather expanded by Sommerschield II. For Maputo's urban dwellers, Sommerschield II appears to have been urbanised according to European standards with its tarred roads, street lights and fancy mansions, yet it is a new neighbourhood growing *outside* of the city centre (*Cidade)*. Sommerschield II is therefore 'neither city nor suburb' (*nem cidade nem subúrbio*), as residents would say. Together with other aff luent neighbourhoods like Belo Horizonte in Matola and Bairro Triúnfo along the beach, they are sometimes referred to as elite neighbourhoods (*bairro da elite*), a term which I adopt here. Residents sometimes also describe Sommerschield II as New City (*Nova Cidade*), associating it with a new type of spatiality, if also associated with the lifeworlds of the *City of Cement* (*Cidade de Cimento*) rather than the *Caniço.* Even though the types of neighbourhood have multiplied in Maputo, and the representations thereof as well, the old binary City of Reeds (*Cidade de Caniço*) and City of Cement (*Cidade de Cimento*) still lurks like a shade in the background. The entangled neighbourhoods exemplify and at the same time also transcend and transform colonial dichotomies of space and identity which constitute part of the colonial heritage of Maputo and Johannesburg.

### **Histories of Changing Cities**

Urban entanglements are deeply shaped by the respective city's history. Before focusing on the ethnography of entangled lives unfolding in these spaces, it is therefore vital to step back and look at the urban histories. In order to compare entanglements across cities, it is important to consider at least three dimensions of urban history, namely, shifting forms of urban citizenship, shifting economic inequalities and shifting pat-

<sup>7</sup>  New academic typologies try to do justice to the diversity of the neighbourhoods in dif ferentiating, for example, between the urban, the suburban and peri-urban areas(Araújo 1999:178) or between the urban core, old suburbs, urban fringe and rural fringe (Henriques/Ribeiro 2005: 13–14). Meanings of *bairro* and suburb dif fer across space and time. For example, in the American context, *bairro* refers to neighbourhoods with a high Latin American population. Across the Lusophone world, the term *subúrbio* hasslightly dif ferentmeaningsin dif ferentcountries(Mabin et al. 2013:169).In this book,I avoid the term *subúrbio* in theMaputo context, although itis of ten used as a synonym for *bairro*, in orderto avoid confusionwith the 'suburb' in Johannesburg.

terns of segregation. I understand urban citizenship as based on Lefebvre, who coined the notion 'right to the city', as the right of urban citizens to participate in urban space, urban life, urban politics and urban centrality (Lefebvre 2009 [1968], Meyer 2007: 278). Both in Johannesburg and Maputo, urban life has been characterised by shifting forms of exclusionary urban citizenship, which have left their specific impact on urban encounters and entanglements. These entanglements are also shaped by shifting forms and degrees of economic inequality in the two cities. In Johannesburg, urban dwellers refer to such differences mostly with the term 'class'*,* using expressions like 'the poor' and 'the rich'. In Maputo, urbanites tend to use euphemisms to speak about such differences, for example *sem condiçoes* (without possibilities) and *os que tem* (those who have), while the terms *rico* (the rich) and *pobre* (the poor) are avoided as having negative connotations.<sup>8</sup> Last but not least, forms of entanglement become shaped by shifting patterns of segregation. By segregation I mean all forms of concentration of a population group in space, whether or not this is imposed by the state or based on so-called voluntary choices by urbanites, and independent of whether the population group is marginalised or powerful in the urban hierarchy.<sup>9</sup> Both Maputo and Johannesburg have a history of shifting forms of urban segregation which left their imprint on the contemporary urban form and everyday spatial practices.

For the kind of comparison I propose here, attention to history is vital. The following insights into the urban histories should make the reader aware of the specific contexts shaping the entanglements, as I only refer to some of them during the chapters for reasons of readability. Urban spaces are carriers of the past, shaping the present and the future entanglements, yet they are also changed through urban dwellers' agency. These sections of the urban histories should also enable the reader to draw her or his own comparative conclusions beyond what I have presented. What is most important, though, is that attention to history makes us aware of the diachronic nature of comparative ethnography (see also Fabian 1983). Ethnographic writing, fixing situations through words on paper, stands in sharp contrast to the f luidity of urban life and urban spaces. Since the time of the fieldwork (2010–2012), all four neighbourhoods have undergone significant changes, ref lecting the transience of urban life. Entanglements, spaces and lifeworlds in Maputo and Johannesburg need to be thought of as being always in production, as constantly changing.

### Johannesburg

Johannesburg is located on the high interior plateau of South Africa, also called the Highveld, where gold was found in 1886 on what turned out to be the world's richest goldfields. Unlike Maputo where the need for a transport hub – a port – decided its location, in Johannesburg it was the local geology with its richness in natural resources.

<sup>8</sup>  Therehasbeenimportantdiscussionofthe applicability ofthe category 'class'toAfricansocietiessince they went through distinctly dif ferent economic developments to European societies where the concept emerged (Neubert 2005, Kroeker, O´Kane and Scharrer 2018). Acknowledging these important debates, I use the term 'class' as well as terms like 'elite', 'poor', 'middle class' and 'af fluent' as descriptive terms. As an analytical concept, I use the multidimensional and phenomenological micro-milieu approach (see later in thischapter).

<sup>9</sup>  This understanding issimilartoNightingale's(2012) global approachwhich invitescomparisons across space and time, and broader than Marcuse's (2005) dif ferentiated terminology distinguishing politically imposed segregation from voluntary clustering which rather complicates comparisons.

Today, it is considered one of the world's leading financial centres and is South Africa's economic hub. People who move to Johannesburg usually do so because of work. For the Southern African region and beyond, the city is also a hub for consumer goods that are hard to get in more remote areas of the continent. Today, Johannesburg's wealth derives mostly from mining, manufacturing and banking; wealth from which large parts of the city's population are excluded. The city has about 4.4 million inhabitants (Statistics South Africa 2012a, based on census 2011) and is the largest city in South Africa and the Southern African region. The urban conurbation, including Tshwane (Pretoria) and Ekhuruleni (East Rand) metropolitan areas, is one of the largest metropolitan regions on the continent (Tomlinson et al. 2003: 6). Especially since the end of apartheid Johannesburg's population has increased significantly (by 68.4%, Harrison et al. 2015: 7). While Maputo continues to have a centre-periphery pattern typical of many European cities, Johannesburg tends to resemble an American city with its multi-nodal or polynuclear pattern. Many edge cities have been built from scratch in the last decades. Johannesburg has also merged with surrounding towns that also turned into sub-centres. Despite urban sprawl, it is also a densifying city (ibid: 9), as exemplified by Linbro Park in this book.

The land where Johannesburg was built was not on 'empty' or 'natural' space; the area had been settled for centuries by Khoikoi, San and, later, Bantu-speaking agriculturalists. Johannesburg was founded within the context of a colonial society produced by forceful conquest, slavery, genocidal extinction of whole population groups and a colonial economy based on forced labour. After the abolishment of slavery by the British in 1834, interaction between black labourers and white employers continued to be tightly regulated by the Masters and Ordinance Act of 1841, which legally ensured the submissiveness of the now 'free' workers, for example by punishing servants who disobeyed their masters (Lester, Nel and Binns 2000: 314). The migrant labour system, which brought people from all over Southern Africa to the Johannesburg mines, affected relations of production, family forms and gender relations in rural and urban areas.

The basis for the infamous apartheid geography was already laid in the founding years of the city. Non-Whites were assigned to live in so-called 'locations'. A class geography which distinguished between the eastern, wealthier areas and the western, poorer areas emerged (Beavon 2004: 53-54). Since then the city has continuously grown through suburbanisation. Property speculators bought up farms around Johannesburg and transformed them into suburbs. In 1912, the investor Papenfus created Alexandra and declared it a 'freehold' township, meaning a township where non-Whites were allowed to acquire land (Bonner and Nieftagodien 2008).

Since the city's early years, African workers were seen as temporary sojourners in the city (Dawson 2006: 126). From 1896 on, African workers had to carry passes in Johannesburg, and their right to be in the city was conditional on their employment at a mine. In addition to race and class, urban society was structured around *work membership*. Ferguson defines work membership as the legal and social distinction between urban 'belongers' with waged employment recognised in their pass book and the urban unemployed who were considered illegal 'hangers on' (Ferguson 2013: 229). Africans had to live in barracks or compounds on the mining fields, or in case of domestic workers, on the property of their employers (Beavon 2004: 33). Since its early beginnings, public space in Johannesburg was not accessible to urban dwellers in

equal ways. By 1900, under the Kruger government, Natives and Coloured people were banned from using Johannesburg's pavements (ibid: 40). They had no political rights, received very low wages and even their leisure time was tightly controlled (ibid: 68). In 1911, the Mines and Work Act legally entrenched a colour bar in the mines, meaning that semi-skilled and skilled jobs were reserved for Whites.

During the Great Depression, the gold standard was abandoned, and South Africa could profit from a considerable economic upswing. During the 1930s and 1940s, the central business district (CBD) experienced a building boom, which transformed the city into a little New York (Bremner 2000: 185). Increased incomes for White families, affordable cars, coupled with state subsidies for housing, produced a revolution in suburban space (Mabin 2005b: 11-23). Edwin James Brolin, owner of a plantation between Alexandra and Modderfontein, recognised at that time the economic potential of his land and sold subsections of what became known as Linbro Park (the reversal of the family name Brolin). The enlargement of private space for the White families – from inner-city f lats to a large suburban house – was possible because Black urban dwellers were forced to live in cramped shacks, rooms and compounds. Class and race became inscribed in the spatial order of the city (Crankshaw 2008). The suburbs, which seemed socially, spatially and economically so distinct and separate from the Black areas, were profoundly entangled with them and dependent on them. White families could only afford a large house and domestic workers because apartheid kept black wages low and because it limited competition for urban land in suburbs by excluding Black owners from landownership. Looking at Johannesburg through the lens of entanglements also invites a re-reading of the apartheid past, paying attention to such interdependencies (Nuttall 2009: 2).

In many African societies, colonialism was accompanied by racial segregation (Seekings 2008: 1). Apartheid, however, as implemented after the National Party came to power in 1948, stands out as an extreme, unique case with "systematic depth and breadth" (ibid: 2). All the "powers of a modern state were deployed to order society along 'racial' lines in ways which went far beyond racism and racial discrimination to generalized social engineering around state-sanctioned racial ideology and legislation" (ibid: 2). The 1950 Population Registration Act introduced the rigid classification of every person into a hierarchically organised, caste-like racial system from which social rights and exclusion were derived.10 One's racial group determined one's access to education, work, land, the use of public spaces and sexual relations (see below).

In 1950 the Group Areas Act was introduced, a legal masterpiece for the realisation of the dream of a White city. In the following two decades, all urban zoning schemes were redrawn, adding information to the mutually exclusive occupation and ownership by legally defined racial groups (Christopher 1997: 311, Mabin and Smit 1997: 206). Rezoning provided the legal basis for massive removals. Tens of thousands of people were forcefully relocated from mixed inner-city slums to townships on the

<sup>10</sup>  'White', 'Coloured' and 'Native'were the initialcategories used. 'Indian' (Asian)was added later (Seekings 2008: 2), 'Native' became replaced by 'Bantu' (1960) and 'Black' (1980). Af ter 1970, the category 'Black' was again subdivided into ethno-linguistic groups such as Zulu and Xhosa (Christopher 2002, Seekings 2008: 2). In the post-apartheid phase, 'black' was replaced by 'African' in the census (Christopher 2002: 404-5). 'Black', asin the post-apartheid Black Economic Empowerment policies,refersto black,coloured and Indian today. Racial terminologywas hence notconsistent over time (ibid: 403).

urban fringe and to rural Bantustans. The White and Black spaces were kept separate through empty tracts of land, so-called 'buffer zones' (Harries 2003: 18). Most freehold townships were destroyed; only Alexandra survived. In 1991, when the laws and racial zoning were repealed, 91.4 per cent of the urban population in the country was living in racially designated zones (Christopher 1997: 319).

Apartheid legislation also operated on the micro-level of everyday encounters and lived spaces. From 1948 onwards many laws (so-called 'petty apartheid') were designed to regulate social interaction in public and private spaces and even intimate relations (ibid 2001). The Prohibition of Mixed Marriages Act (1949) and the Immorality Act (1950) forbade marriage and sexual relations across racial boundaries. The Separate Amenities Act (1953) and the Separate Amenities Amendment Act (1960) subjected public facilities to racial segregation. Park benches, public toilets, beaches, graveyards and the like became segregated (Durrheim and Dixon 2005: 2). There were even building laws regulating the distance between the White family's house and the domestic quarters and demanding separate entrances (Christopher 2001: 139). The everyday experience and use of spaces became closely linked to state-imposed racial categories.

From the 1930s to the 1970s, the city transformed from a mining-based industrial capital to a manufacturing-based economy (Crankshaw 2008, Gelb 1991). The African working class milieus were kept in an inferior economic position by the colour bar, while White milieus profited disproportionally from the economic growth. Immigration at that time was easy for Europeans. When, in 1974, the Portuguese population in Mozambique was expelled by the new socialist regime, many moved to Johannesburg. From the 1970s on, deindustrialisation deeply affected the organisation of the city, the spatial structure and urban dwellers' spatial practices. Many jobs for so-called 'unskilled' workers were lost (Beavon 2006: 59). This led to massive structural unemployment among the Black population, who was kept unskilled under Bantu education. At the same time, edge cities started to develop, as shopping and service-sector work spaces became decentralised and fortified gated communities started to emerge (Crankshaw 2008, Marcuse and Van Kempen 2000: 255). Legal provisions were adopted to allow for 'sectional titles' or condominium ownership of property, giving rise to a new form of suburbia, the 'townhouse complex' consisting of row houses or free-standing houses behind walls (Mabin 2005b: 25). Real estate became increasingly used as a financial asset, leading to a commodification of real estate (Beauregard and Haila 2000, Mabin 2005b: 22). Gated communities also served as a model for residents in older suburbs; they increasingly sought to close off their neighbourhoods, to put up road closures and fences around their neighbourhoods (Dawson 2006, Dirsuweit and Wafer 2006, Hansen 2006).

Already in the 1960s, the first shopping malls were built outside the Johannesburg CBD. In 1973, Sandton City mall was opened, and its success encouraged massive growth in large malls in the suburbs (Beavon 2006: 53). By the end of the 1980s, the contemporary pattern of retail decentralisation with its focus on large malls had been established and was strengthened in the 1990s by what Beavon describes as a "virtual shopping explosion" (ibid: 3). In the 1980s, with the weakening of the apartheid state, non-White urban dwellers moved into the city centre, while White residents and capital continued to move from the CBD to the Northern Suburbs. The inner city underwent massive transformation, and became a 'no-go' zone for many suburbanites (Beavon 2004: 204 ff., Bremner 2000, Czeglédy 2004: 27, Morris 1999a).11 Spatial routines characterising suburban life in Johannesburg still today then emerged: many suburbanites commute from their suburban homes to work in an edge city, shop at a nearby mall and seldom venture into the city centre.

During apartheid, upwardly mobile Black milieus were confined to the townships. When the Black middle-class milieus grew in the 1980s, suburb-like sections were built there for them (Mabin 2005b: 18). Yet, as the grip of the apartheid state was weakening, Black middle class milieus also moved to formerly White suburbs, a process which was accelerated by the repeal of the Group Areas Act in 1991 (ibid: 21, 25). The Population Registration Act was repealed between 1991 and 1994 (Christopher 2002: 405). Desegregation emerged as a process occurring largely along class lines (Crankshaw 2008: 1698). Transformation was discursively framed through the multicultural ideology of the 'rainbow nation'. The slogan of the ANC "a home for all" exemplified this ideal of the post-apartheid 'one nation' (Murray and Shepherd 2007: 7). Yet for many whites the unregulated access to amenities like beaches and parks, informal street trading in inner cities, and co-presence of black urban dwellers as neighbours led to feelings of disorder (Ballard 2002).

The white milieus' repertoire for dealing with the rearrangement of space, race and belonging in the changing post-apartheid era entailed emigration to other countries, 'semigration' into privately secured gated communities and assimilation, the strong expectation of established white milieus that the moving-in black middle class groups had to adapt to their lifestyle (Ballard 2002, 2004a, 2004b, 2005, 2010). Living in proximity to urbanites belonging to different racial categories and economic classes went hand in hand with the desire to create walls (Haferburg 2013, Lemanski 2006a, 2006c, Lemanski and Saff 2010, Morris 1999b), reinforcing and creating insider–outsider distinctions based on race, class, citizenship and other (Bekker and Fourchard 2013, Bénit-Gbaffou 2009, Clarno 2013, Clarno and Murray 2013, Vigneswaran 2014). Yet urban conviviality in contemporary Johannesburg is not all negative and gloomy. For example, in shopping malls and religious spaces mediation across the boundaries of race has been observed (Teppo 2011, Houssay-Holzschuch and Teppo 2009, see also Nuttall 2009). *Cities of entanglements* seeks to contribute to this growing literature.

An important aspect of contemporary urban segregation in Johannesburg is violence and fear of crime. Since the 1950s, in townships like Alexandra crime and gangs have severely reduced the quality of life for Black urban dwellers (Bonner and Nieftagodien 2008: 106ff.). From around the 1980s on, crime and fear of crime also increased significantly in the White suburbs, related to the weakening control by the apartheid state and the shift in public spending on security from the former White areas to township areas. From the 1990s up to the 2000s, extremely high crime levels made headlines (Dawson 2006: 131). Recently, crime statistics have decreased, but urban dwellers still have a sense that crime is escalating (James and Collins 2011). Crime has severely affected the way white people relate to, conceive of and perceive black urbanites (Allen 2002) and conviviality in general (Vigneswaran 2014). South African society is also marked by high levels of violence in interpersonal and intimate relations (Collins 2013, Schäfer 2005). Life in many informal settlements and townships is shaped

<sup>11</sup>  Since the 1990s several initiatives have been launched to reinvent the inner city based on the aspiration to 'world city'status(Bremner 2000:187) and many areas are nowsubject to gentrification.

by structural and interpersonal forms of violence (Ross 2010). Different milieus are affected by crime in very different ways, and therefore their perceptions of crime also differ greatly (Statistics South Africa 2012b: 2).

At the end of apartheid, townships like Soweto and Alexandra were in a dire state in terms of public infrastructure, not only because of the unequal spending by the apartheid state but also because, with the end of inf lux control, residential densities increased. Many people from former Bantustans and adjoining countries moved to the city, rented a shack or a room in a township like Alexandra or built their own shack in a newly emerging 'squatter camp', inside or outside existing townships. Although large-scale public housing programmes were put in place, there was rapid growth in urban informal settlements (Beavon 2006: 55, South African Cities Network, Turok 2011, 2013: 169). Johannesburg has a low average residential density, but densities are highly unevenly distributed between township areas, the inner city, the suburbs and peri-urban areas (Turok 2013: 171).

The turn to democracy brought upward social mobility for the Black middle-class milieus, supported by affirmative action policies, but not to the same extent for poor milieus (Modisha 2008, Seekings and Nattrass 2005: 308-313). Income inequality within the Black population is nowadays as high as across the entire population (Seekings 2010: 8-9). Although formal unemployment decreased between 1996 and 2011 (from 29.4 to 25%), it is still very high, especially among the youth (Beall, Crankshaw and Parnell 2000a, Harrison et al. 2015: 5, Seekings 2010: 4-6). The informal economy is key for urban dwellers in places like Alexandra (Beall, Crankshaw and Parnell 2000b, Hull and James 2012, Simone 2004b, 2006b), although the expansion of the welfare state shapes their livelihoods considerably (Ferguson 2013). In metropolitan areas, half of the households receive one or more government grant (Gelb 2008: 81, Seekings 2010: 14). Many urban dwellers are waiting for government housing (Oldfield and Greyling 2015).

Post-apartheid Johannesburg is a multifaceted city, rapidly changing not least because of its many connections across the globe and across the region. Johannesburg is entangled with Maputo through the circulation of people, capital, goods and ideologies (Harries 1994, Helgesson 2008, Miller 2008, Ostanel 2012, Peberdy 2000, Rogerson 2011, Vidal 2010). Many South Africans and Mozambicans living in Johannesburg travel to Maputo and Mozambican beaches on holiday. Maputo also attracts South African capital and companies and is inf luenced by its lifestyles and fashions. Many Mozambicans see *John*, the name that many Mozambican migrants use to refer to Johannesburg, as a place of job and education opportunities, and as a place with broad and affordable consumption options and reliable health services. Acknowledging the importance that Mozambican workers had and continue to have for building Johannesburg's wealth, one may speak of a mutual dependency between the two cities. To avoid overemphasising the boundedness of the comparative cases (Abu-Lughod 1991, see the postscript), such entanglements between the two cities should be kept in mind.

### Maputo

Maputo is located in Delagoa Bay, in the extreme south of Mozambique's extended coastline. It is much closer geographically, and in some ways also socially and economically more connected, to South African cities like Nelspruit, Durban and even Johannesburg than to cities in the extreme north of the country like Nampula or Pemba. The

city serves as the financial, administrative and cultural centre of the country. Maputo holds the position of the primate city in the country, being considerably larger than the other Mozambican cities, although the importance of secondary cities like Beira is growing (Jenkins 2013: 55). At the time of this research (2010–2012), the municipality of Maputo officially had about 1.2 million inhabitants (Conselho Municipal de Maputo 2011: 8). The Greater Maputo conurbation including the neighbouring city of Matola had about two million inhabitants in 2007 (Jenkins 2013: 109). Like many African cities, Maputo experienced rapid growth following independence and, since the civil war in the 1980s, the population has doubled (Ammering and Merklein 2010). The city centre of Maputo is situated on the northern shore of Delagoa Bay. The suburban areas lie in a half circle around Maputo's city centre. They are continuously densifying and growing towards the north and west, merging surrounding places into the city. Along the shoreline of the bay, moving north-east, gated communities, new hotels and apartment blocks have come to replace the precarious fishing settlements in an area called Costa do Sol and Bairro Triúnfo. New urban road infrastructure, among them a ring road and a bridge across the bay to Catembe built after the time of fieldwork, will bring new dynamics to the settlement patterns.

Maputo was founded as a fort by the Dutch in 1781 and remained a small and unimportant trading post until the gold boom on the Witwatersrand demanded a transport hub in Delagoa Bay. This led to the foundation of the city, then called Lourenço Marques, at the end of the 19th century. Workers on the South African mines and cotton farms travelled through Lourenço Marques, and South African gold was shipped to the world through the city's port. In colonial Mozambique, the Portuguese system of forced labour (*chibalo*) and the introduction of the hut tax made rural livelihoods precarious (Cahen 2012, Mamdani 2000, O'Laughlin 2000). Many men, therefore, migrated to work in South Africa's mines (Harries 1994, Helgesson 2008), and many others went to Lourenço Marques, where they came to form working-class milieus, employed as dock, railway and domestic workers (Lachartre 2000: 31, Penvenne 1995). The African population came to refer to the city as Xilunguine, meaning the "place of the stranger(s)" (Jenkins 2006: 125). Similar to Johannesburg, women had more difficulty in establishing themselves in the city than men, as they were seen as having to take care of the rural household while the men were temporarily away for *chibalo* or migrant work. Many women made a living from beer brewing, prostitution and urban agriculture (Morton 2013: 239, Sheldon 2002, 2003).

Colonial citizenship was bifurcated into citizens and subjects (Mamdani 1996) by a set of institutions called *Indigenato*. The Portuguese immigrants and their descendants had full Portuguese citizenship rights and were governed by the colonial state. The African population was considered 'Natives' and fell under so-called customary law. There were, however, also in-between categories, among them the so-called *Assimilados* (Assimilated). A Native could become recognised as *Assimilado* if he could prove to the colonial state that he was 'civilised'.12 The Assimilated, the mixed population (*Mestiços*) and Asians (*Asiáticos*) were ranked higher than the Natives in the colonial hier-

<sup>12</sup>  In order to receive *Asssimilado* status Africans had to subject themselves to an examination of their private lives in which the 'civilisation' level was assessed: They had to be able to read and write, they had to speak Portuguese at home, eat at a table,wearshoes, only have onewife, and not live in a reed hut (Morton 2013: 240).

archy, had identity cards distinct from them and had more rights (O'Laughlin 2000: 13).13 They were exempt from the hut tax and *chibalo*, they were allowed to live in the City of Cement, could move freely and had access to better schools (Lachartre 2000: 47, based on Honwana 1989: 69-70, Morton 2013: 240). Like apartheid, the *Indigenato* needs to be understood as characterised by forms of entanglements, and not just as a form of separation. As O'Laughlin argues: "Just as apartheid was a normatively prescribed separation of worlds that were in reality linked by the exploitation of African labour, so also were the worlds of citizen and the indigenous subject never separate in Mozambique" (O'Laughlin 2000: 9).

The social, political and economic changes brought about by Portuguese colonial consolidation and the *Indigenato* shaped the production of space in Lourenço Marques (Bertelsen 2014: 2756, Mendes 1979). In the first half of the 20th century, differentiation of neighbourhoods according to race and class emerged and soon different residents experienced very different cities (Lachartre 2000: 34, Penvenne 1995: 33). The eastern upper neighbourhoods with the best climate became populated by European milieus (Polana, Ponta Vermelha, Sommerschield), the western neighbourhoods (Alto Maé) were home to working class whites, *Mestiço* and *Assimilado* milieus*,* and the central lower-lying neighbourhoods in the Baixa were mainly home to the Indian population, called Ásiaticos during the *Indigenato* (Bertelsen 2014: 2756, Lachartre 2000).14 The majority of the population, the Natives, had to live outside the boundaries of the European city in what eventually came to be known as the City of Reeds (*Cidade de Caniço*) (Morais 2001, Morais, Lage and Bastos Malheiro 2012). While in Johannesburg urban segregation became enacted by racist zoning laws and massive relocations, in Maputo it was rather enacted by a multiplicity of racist laws as well as racist practices and racial inequalities (Grest 1995: 150).15

<sup>13</sup> *Mestiço* status was related to birth. The *Mestiço* population consisted mainly of children of European and Indian fathers who kept multiple households, usually with an official wife in the City of Cement and with a black lover in the *Caniço* areas (Morton 2013: 239). A similar spatiality of love relations continuestoday, with af fluent men having loversliving in the *bairros* (Groes-Green 2009: 293-5, Hawkins et al. 2009). For gender relationsin contemporary Maputo see Aboim 2009, Manuel 2013.

<sup>14</sup>  Lourenço Marques had a considerable Indian population (6565 Indiansin 1960, Morton 2013: 237),consisting of descendants of merchants who had participated in the Indian Ocean trade since the 18th century and to a lesser degree also Indians immigrating from South Africa where they had worked as indentured labourers on the sugar plantations fields in Natal (Bastos 2009: 44, Bonate 2008: 641- 2, Marx 2004). The majority of Indians who arrived in Mozambique from India, Zanzibar and South Africa in the 19th and at beginning ofthe 20th centurywere British citizens.When India stopped being a British colony (1947),these Indian migrants had to choose between an Indian or a Pakistani passport, while Indians from the Portuguese colonies like Goa received Portuguese citizenship (1961). Due to transnational business practices and the migration offamily branchesto othercountrieslike Portugal and the UK, many Indians nowadays own several passports besides Mozambican citizenship. Nowadays, urban dwellers of ten refer to Indians as *monhés,* which has a derogatory tone, as merchants (*empresários*, *comerciantes*) and, most of ten, simply as Indians. In this book I adopt the local designation 'Indian' to refer to Maputo residents whose ancestors originated from the Indian subcontinent, independent of what passports they hold nowadays.

<sup>15</sup> The *Indigenato* prohibited Africans from owning land in the City of Cement. In addition, landlords discriminated against black tenants (Lachartre 2000: 38, Morton 2013: 242). Around 1900, it was forbidden to construct African type of dwellings (reed huts, later also wood and zinc huts) within the boundaries of the European city (Bertelsen 2014: 2756).

As in many colonial cities, white anxieties around hygiene and health were fuelled and instrumentalised to legitimate a dual city, giving spatial form to the duality of citizen and subject (Bertelsen 2014: 2756, Eckert 1996, Simone 2004a: 14). The City of Cement (*Cidade de Cimento*) became associated with whiteness and European lifestyles. Those classified as Natives were banned from urban amenities like bars, theatres and hotels (O'Laughlin 2000: 15). It was forbidden for Africans to walk through the streets of the City of Cement after nine o'clock at night (Lachartre 2000: 46). As in Johannesburg, the presence of Natives in the city was dependent on *work membership*, meaning that only those with a place of employment registered in their passbook were considered legitimately to be in the city. In the workplaces, an everyday space of encounter, racism and abuse was common.16 The City of Cement was a space of permanent control and risk for the African population (ibid: 48). Many feared and rarely visited the *Cidade* (Bertelsen, Tvedten and Roque 2014: 2756). The City of Reeds became associated with Blackness, tradition, backwardness, African languages and reed huts (Morton 2013: 240). They were seen as temporary spaces, and many shantytowns were destroyed to make space for the growth of the *Cidade de Cimento*. Unlike Johannesburg, where the grip of the state on African urban areas was tight (except for freehold townships like Alexandra), in Maputo the shantytowns developed largely unplanned. In the 1950s and 1960s, the Portuguese colonial municipality admired the way apartheid South Africa engineered its apartheid cities, yet it lacked the necessary state capacity to entrench the same draconic control over Africans lives and spaces (Lachartre 2000: 39). This so-called 'laissez-faire' attitude was typical of the urbanism transferred from Portugal to Lourenço Marques (Jenkins 2006: 110).

From the 1940s to the 1960s Maputo experienced strong industrialisation as the Salazar regime invested much capital to show to other European powers that it was developing its colonies (Lachartre 2000: 43, Morton 2013: 236, Pitcher 2002: 31). Maputo's skyline with modernist high-rise buildings emerged at that time. When Portugal held on longer to its colonies than other European colonisers, it became increasingly criticised for it (Morton 2013: 236). Portugal responded with the rhetoric of 'lusotropicalism', claiming that the Portuguese assimilated the different races and cultures in their colonies in harmonious racial relations (ibid: 237). Yet, at the same time, political culture in Mozambique became strongly shaped by the Portuguese dictatorship with paramilitary, nationalist propaganda and secret police.

After independence, Frelimo (Frente Libertação de Moçambique) continued Portugal's modernist and one-party state attitude, although now coupled with a socialist and nationalist ideology. By the early 1980s, the state dominated every economic sector (Pitcher 2002: 44). The socialist Frelimo regime replaced the *Indigenato*, which the Portuguese had started to dismantle in the 1960s, by an ideology called *Homem Novo*. Through this they aimed to create a new, modern society based on the decolonisation of the mind and national unity replacing ethnic, racial and religious differences (ibid: 53-54). Citizenship became redefined through loyalty to Frelimo and commitment to the party's image of the *Homem Novo,* the modern and rational Mozambican citizen. As in many other socialist contexts, the state did not allow other civic organisations besides the one adhering to Frelimo (Jenkins and Wilkinson 2002: 40). Traditional

<sup>16</sup>  For further research on racial and classrelations during the colonial period,consult Bras 2006, Cahen 2012, Errante 2003,Henriques 2012, Penvenne 1995, 2003, Thomaz 2006, Zamparoni1998, 2000, 2007.

authorities (*regulos*) governing *bairros* became replaced by new neighbourhood structures called dynamising groups (*grupos dinamizadores).* These were led by Frelimo party secretaries, so that the state and party structures became mixed, which continues today in many *bairros*. Residents were instructed to report 'enemies' and 'traitors' to the party structures (Buur 2010: 35, 42, Hall and Young 1997, Harrison 1998, Macamo 2003: 7, Sumich 2013: 100-103).

Independence and the turn to socialism brought massive changes in relation to who lived in the city and where. In 1976, the new president, Samora Machel, proclaimed: "The people will be able to live in their own city and not in the city's backyard" (quoted from Morton 2013: 232). The white Portuguese population f led the country rapidly, their inner city f lats and houses became nationalised, and were given to African privileged milieus like party elites, military veterans and former *Assimilados* (ibid: 233). Frelimo continued the pattern of work membership, culminating in the infamous 'operation production' in the early 1980s, which forcibly removed people considered 'unproductive' from what was now called Maputo and subjected them to a new type of forced labour (Buur 2010: 43).

South Africa and what was then Rhodesia were a key force supporting Renamo (Resistência Nacional Moçambicana) which engaged in a long civil war (1977–1992) with Frelimo. The war destroyed national infrastructure, 1.6 million people f led the country and 3.7 million were internally displaced (Chingono 1996, Geffray 1990, Lubkemann 2008, Pitcher 2002: 104). Many war refugees came to settle in Polana Caniço and other neighbourhoods, also on land that had been reserved for public use or was considered unsuitable for housing. Unplanned occupation became the main form of city expansion (Jenkins 2013: 96). Coupled with natural disasters the civil war led to a deep economic crisis in the 1980s. Frelimo eventually decided to embark on the transition to neoliberal capitalism and adopted structural adjustment programmes with long-term negative effects on state capacity, limiting the municipality's ability to do any form of urban planning (ibid: 97).

Land, however, did not become privatised with the turn to neoliberalism and, still today, continues to belong to the state. The official absence of private landownership distinguishes the Mozambican context, making Maputo unique and distinct for an African city (Jenkins 2001, 2013: 72). Land can officially neither be owned nor sold, people merely receive land use rights (*Direito de Uso e Aproveitemento de Terra,* short DUAT). Only the buildings on the land are private property. In reality, though, an unofficial and illegal land market developed, which is considered socially legitimate and through which the majority of urban dwellers access land (Jenkins 2001, 2009, Jorge and Melo 2014, Sidaway and Power 1995, Unruh 2005).

Privatisation led to the loss of many jobs and gave rise to the growth in the informal economy (Grest 1995, Jenkins 2006, Kamete and Lindell 2010: 902).17 Post-socialist

<sup>17</sup>  'Informal' refers here to the sector of the economy which is not under state control (taxes, labour laws), but this does not mean this sector does not know alternative forms of order (see also Sardan 2009). Already under the socialist (and also the colonial) regime in Maputo, many engaged in informal economic activities, which were seen as illegal and as actions of political dissent (Buur 2010: 27). Small‐scale trade in public spaces has considerably changed the face of the city in the post‐socialist era (Kamete and Lindell 2010: 902, 904, Lindell 2008). Dichotomies of illegal and legal, informal and formal fail to capture the complexities of intermingled and mutually influencing norms and forms in Maputo's production of space. The notion of informality is widespread in the analysis of Africa's

Mozambique experienced several phases with growth measured by macroeconomic indicators, although poverty also increased. Mozambique's Human Development Index is extremely low and will probably continue to be so in the light of the debt crisis which started in 2016. Since the state is heavily dependent on international donors, its accountability with regard to the parliament, civil society and the population is undermined (Macamo 2006). Politics still does not mediate between society and the state, many citizens do not see much relevance in the formal government system and there is a low level of trust in the state (Eskemose 2004, Jenkins and Wilkinson 2002: 40, Lindell 2008, Macamo 2001, Seibert 2007). With the official arrival of political pluralism and freedom of association, urban associations without Frelimo allegiance emerged in Maputo, but they are, apart from religious associations, still few in comparison with associational life in Johannesburg.

Neoliberalism coupled with globalisation brought changes to and differentiation of lifestyles and spaces in the city. During privatisation, many inhabitants of f lats in the *Cidade de Cimento* were able to buy them at low prices (Sidaway and Power 1995: 1480). Later, many started to let their f lats to well-paying foreigners, and moved to the outskirts of the city where they constructed a suburban house with lots of space (see also Ammering 2010, Jenkins 2013, Nielsen 2014). The privilege of being given a f lat was, therefore, key to the social mobility of many of the contemporary middle-class and elite milieus.

The economic and political elites from socialist times could use their position to shape the privatisation process and to secure economic benefits.18 The severe limitations on food and consumption which existed during socialist times came to an end at the end of the 1980s. For urban elites, for whom buying rare consumer goods through the black market was a mark of distinction and a source of danger during the previous dispensation, modern consumer culture became finally accessible. Conspicuous consumption became a sign of modernity and social power among elites (Sumich 2005: 110-111, Veblen 2000 [1899]). In the 1990s and 2000s many bought or rented houses in

economies (Hart 1973) and spaces, although if simplistically applied it implies an overly rigid separation from the formal and an absence of form and order. In reality, even the most informal economies and neighbourhoods are shaped by urbanites'senses of order and norms,which are of ten influenced by the state's norms. With regard to urban spaces, the notion of 'self‐production', of ten used in lusophone contexts (e.g. Jorge and Melo 2014), is a useful alternative to informality. In contrast to the modernistformal planningwhich is dominated by conceived space, guided by conceptions ofspace in architecture and urban planning (Lefebvre 1996 [1974]: 43),the self-production ofspace hasitsrootsin lived space, the everyday practices, needs and meanings of dwelling in and using space. Thisself‐production ofspace orinformality should not be understood as an urban sector, as Roy (2005:148)rightly points out, but as a dif ferent mode of production ofspace.

18 Pitcher distinguishes four groups of national capitalist actors (Pitcher 2002: 153-164). First, there are old, powerfulcompanieswhich are descended from colonialconcessionary companies. Second,there are companies owned by people of non-African descent (Portugal, India, Pakistan, China, etc) which mostly hold Mozambican passports. Both groups depend on clientelist and patronage networks as well as links with the state, some are members of parliament or have joint ventures with politicians (Pitcher 2002: 154-158). Third, there is the group of most recent domestic capitalists, also depending on patronage and corruption,which tend to be former government officials, Frelimo party supporters and former managers ofstate companies(Hanlon 2004). The fourth group consists of domestic actors with no connection to politics: they are mostly African and Indian shop owners, traders and farmers who engage in economic activities at the porous border of formal–informal (Pitcher 2002:165).

areas like Sommerschield II. These new elites emerged as important actors in the gentrification of the city, as real estate investors, as consumers of high-end residential space and as buyers at shopping malls (see also Boetius 2001, Hanlon and Mosse 2010, Sumich 2008a, 2008b). Critique of these conspicuous lifestyles of the elite and allegations of corruption and abuse of state resources have become part of everyday conversations and political debates in the city.

The liberalisation and growth of the private sector also enabled the growth of middle class milieus with university education and stable jobs at an international company, international NGO or the state (Bornschein 2009, Groes-Green 2009, Sumich 2008b, Van de Kamp 2016). Although the middle class milieus may still be quantitatively negligible on the national level, in Maputo they have a significant inf luence on the urban landscape of consumption, housing and leisure. Bars, restaurants, nightclubs, fitness centres as well as increasing numbers of South African chain stores have become the spaces of play and public life of these milieus. For many members of poor milieus living in the city's *bairros* visits to such new urban leisure spaces constitute exceptional events. The many markets, *barracas* (little bars), discos and churches in the *bairros* are the centre of gravity of the public life of these milieus.

In the post-socialist city, there continues to be categories of citizens whose right to belong is sometimes contested, among them the Indians (former *Asiáticos*) (Sumich 2013). During colonial times, the intersections of religion (the majority of them Muslims), class (usually merchants), nationality (many were British citizens till the 1960s) and endogenously oriented culture with close kinship ties and transnational practices gave these Indian milieus an intermediary and highly ambivalent position within the colonial and socialist society, with effects on their social position and on the way they were seen by others which still prevail today. Although they had a higher position than the Natives in the colonial hierarchy, the Portuguese imagination constructed them as dirty, degenerated, effeminate, astute and greedy (Zamparoni 2000: 199-200). For the African population, the *monhés* were important employers, if also demanding and sometimes cruel, as well as givers of credit and providers of consumer goods (Bastos 2009: 49, Zamparoni 2000). During the times of transformation from socialism to neoliberalism, strategic alliances and business partnerships with functionaries and members of the ruling party enabled inf luential Indian merchants to become part of the new economic elite (Carvalho 2008: 114). To this day, colonial stereotypes shape encounters in workplaces and reappear in religious entanglements (see chapter 6).

### **Cities of Entanglements**

Cities bring together differences in close proximity, at unique scales and unique levels of intensity (Fincher and Jacobs 1998: 1). Urban differences and inequality are a major concern in urban development discourses and academic research. In 2010, UN-Habitat claimed that Southern African cities show the steepest degree of socioeconomic inequality, with extreme levels of poverty. African cities in general are counted as being among the most unequal in the world in socioeconomic terms (UN-Habitat 2010: 14). It is important to note, though, that differences, inequalities and emphasised boundaries characterise cities across space and time. Differences need to be recognised as a sustained feature of cities, of urbanity and of urban spaces (Fincher and Jacobs 1998).

Across historical periods and geographical regions, segregation has emerged as a dominant way of inscribing and dealing with urban differences in cities, as well as writing about them. Segregation has characterised many European cities from Medieval times to the present (Nightingale 2012), in colonial cities in Africa and elsewhere (Beinart and Dubow 1995, Nightingale 2012, Njoh 2007), contemporary Latin American cities (Koonings and Kruijt 2007), cities divided by violent conf lict like Mostar (Aceska 2015) or Jerusalem (Bollens 2012), contemporary American cities (Davis 2006 [1990], Marcuse 1995, 1997a, Young 2000) and African cities (Murray 2011). The causes, forms and effects of segregation may vary greatly, but what such cities share is an urban population that needs to find routines and ways of living together in a context of social and spatial inequality. Shaping urban entanglements, social relations across social boundaries, is one of the key ways in which urban dwellers do this.

The notion of entanglement has been explored in various disciplines: in history ('entangled history') there have been systematic discussions of the term, in fields like geography, anthropology and literary studies a more systematic engagement is emerging. In African urban studies, Pieterse (2006: 408) points out how little is yet known about the significance of interactions and relations across class and race divisions. Two urban scholars, the literary scholar Sarah Nuttall (2009) writing on Johannesburg and the sociologist Sanjay Srivastava (2014) writing on Delhi, have more recently suggested the notion as a new register for writing about these metropolises shaped by significant spatial and social inequalities. I build on their work as I believe that the language of entanglements can open us to new understandings of urban life.

The study of entanglements can be seen as forming part of the ethnography of encounters, a transversal field in anthropology which focuses on how "relationships among unequally positioned groups shape cultural processes" (Faier and Rofel 2014: 64). The notion of encounter has, at least in the German translation *Begegnung*, the problem that it appears normative and suggests a specific form of interaction, namely, one through which mutual understanding is enhanced. I understand encounter in a broader sense as a synonym for interactions of all sorts, involving people who consider themselves as different (ibid: 364). Encounter is then a complementary term to entanglements: While encounters refer to *instances* of engagement across difference, entanglements refer to the social relationships constituted through such encounters. Urban entanglements are more than the sum of moments of interaction; they are part of what Simmel called urban sociation (in German *Vergesellschaf tung*, Simmel 2013 [1908]). In entanglements, forms of sociality emerge and by looking at them we can therefore observe urban society in the making.

The notion of entanglement in the way I approach it builds on Nuttall's thesis that in order to understand social life in the midst of changing forms of inequality and diversity, we need to focus our lens on the moments of interaction where the categories spill "out of the routinised confines of the absolute figures" (Nuttall 2009: 10). With its "metaphorical, real and spatialised imaginary" the lens of entanglement entails a "shift in interpretation away from a sense of dualism and frequently normative theorisations towards a more complex and nuanced understanding of the interrelationships" (Houghton 2013: 2793). In the terrains of encounter, a multiplicity of events can take place, from reaffirmations of boundaries to their transgressions or even dissolutions. As the ethnography will show, entanglements are often characterised by ambivalences, contradictions and tensions. Looking at urban life through the lens of entanglement

is a "means by which to draw into our analyses those sites in which what was once thought of as separate – identities, spaces, histories – come together or find points of intersection in unexpected ways" (Nuttall 2009: 11). Entanglement is a deeply suggestive notion, and rather than implying a clear message as to what these specific sets of urban relationships look like, it opens up space for exploring a diversity of encounters with multiple, sometimes contradictory, meanings and effects.

As mentioned, the lens of entanglements also invites a critical re-reading of the histories of Southern African cities, as they point to the fact that apartheid and Portuguese colonialism also entailed intimacy among racially defined groups and other sites of intersection commonly overlooked (ibid: 25). Entanglements, hence, promise to address a problematic which African urban studies are confronted with, namely, to move "away from the tropes of segregation, fragmentation and bifurcation" (Pieterse 2009: 1). Pieterse (2009) warns against simply replacing ideas of polarisation with accounts that solely foreground interconnections and endless networks. This reminds one of Strathern's (1996) critique of the network imagery, arguing that networks can appear to be without limits. The notion of entanglements can be seen as an alternative to the boundedness of spaces in segregation imageries and the limitlessness of network theories: while drawing attention to intersections and social relations, the lens of entanglement also analyses these as being shaped by boundaries, limits, cuts and disruptions of social relations. "In as much as entanglements are about difference and commonality, they are also about limits" (Nuttall 2009: 1). As the book will show in several chapters, entanglements are also characterised by a tension between two fundamental attitudes towards the 'other', which Förster (2013b) calls 'encounter' and 'distanciation'. This tension, which often translates into a tension between proximity and distance, constitutes urbanity (ibid).

Urbanity in cities characterised by inequality differs from urbanity in more equal cities in degree rather than in kind: Finding solutions for living together in close proximity with people considered different is, such is the premise of this book, a defining feature of urban life in general. Based on Simone (2010: 2), I understand *cityness* as the "capacity to provoke relations of all kinds" and cities as "places of thickening connections" (Simone 2004a: 137). *Cities of entanglement*s should not be misunderstood as a specific type of city or an abstract model like 'global cities' (Sassen 1991). This book rather approaches Maputo and Johannesburg as 'ordinary cities' (Robinson 2006a), each with its specific history and future, which a comparative ethnography needs to account for. *Cities of entanglement* is an ethnographic lens which makes visible social processes and practices that exist in multiple forms in diverse cities. With that, this book embarks on the journey of a comparative ethnography of urban entanglements.

There are several shortcomings in urban anthropology and urban studies that the lens of entanglements promises to address. First of all, urban ethnography has long suffered from what I call *mosaic thinking.* By mosaic thinking I refer to urban research that focuses on a single neighbourhood, without paying much attention to the way the neighbourhood is embedded in the wider city and neighbouring areas. Mosaic thinking, often based on an understanding of space as absolute (see below), produces images of cities as made up of distinct patches, each with its colour or identity, and with clearly drawn boundaries between them. Such mosaic thinking can be observed through the history of urban scholarship from the Chicago School's neighbourhood studies (Wirth [1928] 1998) to urban village studies (Gans 1962), to ghetto studies (Hannerz 2004 [1969], Hutchison and Haynes 2012, Wacquant 2008, Welz 1991) and to gated communities studies (Beall 2002, Breetzke, Landman and Cohn 2013, Low and Low 2003). Some of these studies tend to suffer from an isomorphism of space and identity, insufficiently disentangling spatial from social boundaries, without much attention to boundary-crossing connections and sometimes too little appreciation of internal heterogeneity.

*Cities of entanglements,* the approach developed in this book, aims to move beyond such mosaic thinking by understanding the city through the entanglements that undo the prevailing analytics of separation and segregation. As I argue in chapter 2, neighbourhoods have always been entangled with other places; they become constituted through these connections, and hence neighbourhoods need to be understood in relation to other urban areas and the city at large, and not in isolation. Approaching neighbourhoods through entanglements entails an interest in how spatial and social boundaries are created in situations where urban dwellers encounter each other. For urban researchers it is important to follow urban dwellers on their everyday trajectories through the city, as this uncovers the meanings of spatial boundaries for the everyday lives of different milieus. Entanglements show how different urban lifeworlds are not entirely disconnected, and how segregation is never complete. The lens of entanglements invites an empirical investigation of the degree and quality of divisions and connections, and not merely a reproduction of narratives of urban duality and segregation. Attention to entanglements raises awareness of how neighbourhoods that aff luent urban dwellers imagine to be disconnected enclaves are actually also home to domestic workers, who appropriate and use these suburban spaces in multiple, often hidden ways.

A second shortcoming of contemporary urban scholarship, which an approach grounded in entanglements seeks to overcome, relates to the argument made above about networks. In research about African cities, there is a bifurcation between an emphasis of location and place-making and an emphasis of networks, mobility and aspatiality (Bank 2011: 16). While the one strand overstates the limits imposed by spatial structures of inequality, the other strand aggrandises the limitless agency and mobility of urban dwellers. The lens of entanglements hopes to overcome this dichotomy by providing a new language on the way urban life is unfolding in the midst of urban segregation *and* mobile, connected lives.

A third shortcoming is that scholarship of conviviality in urban spaces often overemphasises the significance of interactions in publicly owned public spaces and neglects or even negates the relevance of sociality in other spaces like malls, which tend to be seen as inferior spaces (see chapter 7). Amin and Thrift criticise this shortcoming regarding the invisibility of homes in urban theory:

Strangely the everyday rhythms of domestic life have rarely counted as part of the 'urban', asthough the city stopped atthe doorstep ofthe home. But domesticlife is now woven routinely into the urban public realm ... The rhythms of the home are as much part of city life as, say, the movements of traffic, office life, or interaction in the open spaces of the city. Its rhythms, too, need incorporating into the everyday sociology of the city (Amin and Thrif t 2002:18).

The analysis of entanglements makes us aware that the diverse spaces of urban life such as suburban homes, religious spaces and shopping centres are connected with each other and become constituted in relation to each other. Urbanites move from one space to another in their everyday trajectories, they experience places in comparison to other spaces (see chapter 7), and urban dwellers bring habits of interaction acquired in one place and within one entanglement into the experience of encounters at the other place. In particular, patterns of seeing and dealing with each other acquired in interactions between domestic workers and employers make their reappearance in other settings, for example in stranger interactions at the mall (see chapter 7). Entanglements can, therefore, have a significance far beyond the settings in which they are usually enacted and shape urbanity in general.

A fourth shortcoming that the notion of entanglement promises to address is related to the methodologies of comparisons. The scholarship on urban comparisons which emerged around Robinson (2006) is currently dominated by debates about the politics of urban theorising from the Global South, about epistemologies and typologies of comparisons, but there are as yet few practical answers on how comparison as a qualitative, ethnographic method can be used. Bringing the notion of entanglement into these methodological debates can elucidate that comparison in ethnography should not be an analytical 'last step' in a long research process, but rather a research practice which takes place during every single step in the circular research process. The notion of the *entangled comparer* (see the postscript) draws attention to the epistemological questions related to producing urban knowledge by researching two cities as a situated subject. Entangled comparative ethnography does not bring about logical conclusions on abstract relations between cases and variables, but rather produces ethnographic texts with multiple, ever-shifting cases, and hence narratives about urbanity based on a diversity of urban experiences.

Urban entanglements have a *spatial* dimension, they involve urban dwellers who consider each other as *dif ferent*, and these urban dwellers *shape* urban entanglements. In order to ethnographically grasp and analyse such urban entanglements, there is a need for three further, interrelated concepts: space, milieus and agency.

### Space

Massey points out that the way we formulate the concept of space radically shapes our understanding of the social world (Callard 2011: 299). Harvey introduces a typology of conceptions of space which is useful here. *Relative space* refers to the spatial relationship between objects and the relative nature of a location (Harvey 2006: 121‐122). *Relational space* refers to the space created by relationships between objects or actors. *Absolute space* is a fixed geographical space, a "thing in itself" or "discrete and bounded phenomena" (ibid: 133). In everyday life, as well as in urban studies, there is a tendency to understand cities and neighbourhoods as absolute spaces, as wholes with clear boundaries like walls. Absolute space has also been called the "container model of space": the modern notion of space as a naturally existing *container* which comprises objects and people (Bachmann-Medick 2006: 289). Such a container model of space, which is often present in writing about segregation and what I call *mosaic thinking*, obscures rather than illuminates interdependencies and entanglements across spatial and social boundaries.

*Cities of entanglements* is based on a relational and relative understanding of space, grounded in Lefebvre's spatial theory. The philosopher and sociologist Henry Lefebvre laid the groundwork for the shift from a container model to the social production of space with his theory *The Production of Space*. Lefebvre's understanding of space as always in production (or construction) introduces a diachronic perspective, looking at the becoming and the historicity of space (Lefebvre 1996 [1974]: 37). This is well suited to the study of cities where contemporary social practices evolve within a context that has undergone profound transformations (see above). Following Lefebvre, social space is constructed, negotiated and experienced in the situations of everyday life through the interplay of three dimensions, namely, through (1) the material, (2) the conceived and (3) the lived space.

*Material space* is that dimension of space that can be perceived by the senses, so it can be seen, heard, smelled, touched or tasted. The material comes into existence only through perception, which varies socially and individually (ibid: 113). Maputo Shopping Centre as a material space is perceived differently by a middle-aged member of the Indian elite who has travelled extensively abroad than by a child who has grown up playing on the sandy paths of Polana Caniço. Material space becomes produced through what Lefebvre calls *spatial practice*, namely, through daily routine and infrastructural routes and networks which link spaces (ibid: 38). Key for the constitution of urban material spaces are hence the urban dwellers' trajectories of everyday life which emerge when urban dwellers move through the city to work, socialise, eat and pray (Magnani 1996: 21). In this understanding, a deserted, empty city would cease to exist as a space: no actor would be there to perceive it, no actor would connect it to other objects and spaces through their movements. Hence: "There is no such thing, in a social sense, as an empty space" (Tonkiss 2005: 3). Spatial practices make the city more than simply an aggregate of different urban spaces or a 'mosaic' of separate moral worlds (Park 1915: 608). Rather, spatial practices entail an understanding of space and cities as relational, open and connected (Callard 2011: 302, Massey 2005).

*Conceived space* refers to the conceptual construction of space achieved through mental activity. It is a mental achievement to treat individual elements as a whole (Schmid 2006: 169). When people communicate about such spaces, based on a "system of verbal, therefore intellectually worked out signs", Lefebvre refers to it as a *representation of space* (Lefebvre 1996 [1974]: 39), by which he also means ideology and knowledge about space (ibid: 45). For Lefebvre, this logico-epistemological space is the dominant space in any society and has a substantial role in the production of space (ibid: 38-9, 42). It is intrinsically linked to power. For example, it is through conceived space that governments 'see' their population, shape spatial reality and draw territorial boundaries between insiders and outsiders (Brighenti 2006, Scott 1999). Based on coherence and logic, in conceived space the complexity of life is reduced to abstract categories. In title deeds, zoning and cadastral maps, abstract properties of spaces are measured using a unified metric system (Schmid 2005: 101-102).

Lefebvre formulated conceived space based on the Western experience of "the space of scientists, planners, urbanists, technocratic subdividers and social engineers, as of a certain type of artist with a scientific bent – all of whom identify what is lived and what is perceived with what is conceived" (Lefebvre 1996 [1974]: 38). In African cities, though, as well as in many other cities, there are various non-western, non-scientific forms of knowledge production about space which need to be taken into account. Urban dwellers' representations of space tend to emerge out of their lived experience of spaces, yet they may appropriate and adapt the political discourses, categories and processes imposed by states to represent or legitimate their interests and ideas (see also Scott 1999). They create themselves room to manoeuvre within and around imposed categories (Certeau 1984).

*Lived space* refers to how space is lived and experienced in everyday life, rather than conceptualised (Schmid 2006: 169). This dimension of space, which Lefebvre also calls symbolic space or space of representation, is "space as directly lived through its associated images and symbols, and hence the space of the 'inhabitants' and 'users', but also of some artists, writers and philosophers, who describe and aspire to do no more than describe" (Lefebvre 1996 [1974]: 39). In contrast to representations of space, spaces of representation do not need to follow rules of consistency or cohesiveness, as they are f luid and dynamic (ibid: 41-42). This is also the reason why lived space cannot be fully dominated by conceived space and is therefore also the space of resistance and subversion (ibid: 33). While spatial practices point towards mobility and the network character of urban life, lived space and place making draw awareness to the way urban dwellers develop memories, meanings and attachments to particular places (Bank 2011: 15). The layers of use and meaning of lived spaces can be multiple, complex and diverse for the different groups that inhabit them (see for example Staudacher 2019). These three dimensions, the perceived, conceived and lived space, should not be taken as three distinct types of space, but rather as three dimensions which together come to constitute space (Schmid 2005: 309). With such a model, one can approach the spaces that shape and become shaped by urban entanglements in complex, multifaceted ways.

### Urban Milieus

If entanglements are the encounters and relations between people who define themselves as different, how do we ethnographically assess these differences? In order to bring diversity and difference into the comparative ethnography of the urban, I use the phenomenological concept of urban milieus. I understand social milieus based on Hradil as groups of like-minded people who have similar values, similar ways of life, similar relationships to others and similar mentalities (Hradil 1999: 420). In the shared lifeworld of the city, different milieus come to interact, overlap and come to exist in relation to one other (Dürr 2004: 137), they become entangled. Urban milieus do not come first and then the entanglements between them; entanglements also bring into being the urban milieus.

Phenomenology understands everyday life as consisting of situations in which social actors engage with one another. In these situations, typicality of action and meaning emerge. Based on Förster, I understand milieu as the sphere of the lifeworld in which such typicalities are shared with others and are regarded as normal by everyone (Förster 1997: 158).19 For the ethnographic analysis this means that actors whose typical ways of acting and interpreting the world are not normal to each other belong

<sup>19</sup>  Normality is not the same as familiarity in phenomenology. According to Schütz, a new situation is unfamiliar if the actor does not know how to interpret and act in a situation because she has never experienced it before (and hence possesses no appropriate scheme for interpretation and action) (Schütz and Luckmann 1973: 151). Normality, on the other hand, is when the practices of another person are familiar to the actor (she knows them) as well as normal (she also acts this way). Domestic

to differing milieus. For the constitution of milieus and differences, the habitual side of agency (see below) is crucial, as in Bourdieu's habitus concept (Bourdieu 1977 [1972]). Actors recursively implement structures through the performance of dispositions of acting, perceiving, feeling and judging, which are constituted by past experiences and inscribed in the body (Emirbayer and Mische 1998: 978, Schwingel 1995: 60). Milieu membership shapes the typical interests and goals of actors, as well as typical problems (Förster 1997: 158). For example, residents from milieus living in Polana Caniço and Sommerschield II, and residents from milieus in Linbro Park and Alexandra, may all have certain security problems in their lives, but the types of crime they consider a problem may differ considerably, and this can be a cause of contention in encounters, with relations of power inf luencing whose problem definition becomes socially more relevant.

In contrast to other approaches to social differentiation, the milieu concept assumes that the 'objective' conditions of life (income, living environment etc.) do not totally determine the 'subjective' views, practices and lifestyles of a milieu (Abels 2009: 326, Hradil 1999: 431). Rather, people with very similar education, work and income can develop quite distinct ways of life. The concept is therefore especially promising for the analysis of social differentiation in African societies where unidimensional perspectives focusing, for example, solely on income fail (Kroeker, O'Kane and Scharrer 2018, Neubert 2005a).20 The phenomenological milieu perspective I adopt here focuses on small-scale milieus, and not statistically defined large groups as done in quantitative milieu studies. The ethnographic milieu analysis generalises through ethnography 'from the bottom up': from individual actors to characteristics they share with others.

Paying attention to urban milieus makes us aware that the way people experience the urban, their possibilities of action and their ways of relating to others differs widely across them. This raises a very important, yet often neglected question in comparative urbanism (Gough 2013). *Whose* city are we writing about? *Whose* urban lives we focus on greatly affects the nature of the comparisons (ibid: 877). Understanding urban difference through social milieus is a useful path towards becoming more aware about *whose* city we are writing about.

workers, for example, have intimate knowledge about the everyday routines of the family they work for (familiarity), yet they do notshare the same routine (normality).

<sup>20</sup>  African cities are marked by an increasing inequality and dif ferentiation of lifestyles, but there is a lack of empirical and conceptual work on social dif ferentiation which goes beyond uni-dimensional perspectives. Research of ten focused on socio-cultural dif ferentiation with 'ethnicity' as a key concept or it focused on socioeconomic dif ferentiation (key concepts 'poverty' and 'livelihood') in the context of poverty reduction and development (Neubert 2005b). The concepts of 'milieu' and 'lifestyle' are able to bring these two dimensions together and are hence very promising as analytical frameworks to study social dif ferentiation in Africa (Neubert 2005a). The concept of milieu and its related concept of lifestyle have become prominent in German sociology since about the 1980s (Hradil 1999: 42). They can be regarded as replacements for the classical approaches to social stratification, which analysed the vertical structure of society along economiclines, such asclass theories by Marx and Weber (Abels 2009: 266f f.). For the use of milieu analysisin anthropology see also Bauer (2007) and Kaiser-Grolimund (2017).

### Agency

The aim of urban anthropology should not only be to describe differences between urban milieus, as this would create a static picture of urban lifeworlds. Gupta and Ferguson propose a bigger task for an anthropology in the contemporary world. The task is to explore "*the processes of production of dif ference* in a world of culturally, socially, and economically interconnected and interdependent spaces" (Gupta and Ferguson 1992: 14, italics added). One of the biggest fallacies of research on urban differences would be to take differences as fixed attributes or static forces (Bridge and Watson 2013a: 502). As academics we need to be careful not to exaggerate differences, nor to exoticise or to construct milieus as overly bounded and independent from each other.

Agency is an important concept for drawing attention to the urban dweller's capacity to shape and change urban differences, spaces and urban entanglements. The analytical category agency focuses on the social actor's capabilities of sustaining and altering structural contexts. Agency can then be defined "as the temporarily constructed engagement by actors of different social environments which, through the interplay of habit, imagination, and judgement structure their lifeworld in interactive responses to the problems posed by the historical situation they have to cope with" (Förster and Koechlin 2011: 7, based on Emirbayer and Mische 1998). Actors do not act in situations from a *tabula rasa* but bring with them a stock of knowledge of pre-existing patterns or schemas of actions, of experiences and of meanings (ibid: 971). Through the process of typification, actors recognise sameness or difference in an emerging situation by comparing it with situations in the past. Depending on their agentic orientation toward the past, present or future, they may then repeat, adjust or reconfigure them to the contingencies of the emerging situation (ibid: 971). In everyday encounters between members of different milieus it is through their agency that urban dwellers can shape and change their ways of interrelatedness and, hence, urban society. Because of actors' agency, entanglements are at the same time f luid, shifting and contingent as well as shaped by patterns from the past.

### **Outline of the Book**

How do people live together in cities shaped by inequality? This question guided the fieldwork based on which this book was written. Between 2010 and 2012, I conducted fourteen months of fieldwork in Johannesburg and Maputo. During the seven months of fieldwork in each city, I conducted – with the support of three research assistants – over 140 interviews, and documented over 400 instances of participation, observation and informal conversations in English and Portuguese.21 I explored neighbourhoods, everyday trajectories and sites of encounters in the two cities. My fieldwork was hence guided by three foci which I explored in a circular manner. Firstly, I was interested in how everyday lives evolve in the four neighbourhoods, and I explored them in

<sup>21</sup>  I distinguish participation from observation as two distinct methods, in line with Förster (2001). In participation, many actions take place in an interwoven manner, like talking, doing things, and also seeing. It is not possible, though, to observe and participate at the same time, as they are based on two dif ferent modes of experience. While observation is based on distance, participation is based on direct engagementwith theworld (Förster 2001: 468-9).

ways similar to classical neighbourhood studies as a form of entry into urban life. I mapped spaces, organisations and urban milieus in the neighbourhoods, participated in everyday life and conducted ethnographic interviews. Secondly, I wanted to understand the everyday trajectories of urban dwellers which mostly evolved on paths and routes through the city, linking their neighbourhoods with workplaces, spaces of consumption and of worship somewhere else in the city. I therefore 'followed' (Marcus 1995) urban dwellers through the city, by accompanying them in 'go-alongs' (Kusenbach 2003), by having them draw their spatial routines on a map, and by interviewing them about their trajectories. I also conducted interviews with photo elicitation (Collier 1957, Harper 2002), which means that I showed photographs of specific urban spaces to evoke narratives, and I asked urban dwellers to keep self-administered diaries of everyday routines (Zimmerman and Wieder 1977). Thirdly, I explored spaces of encounters, meaning the sites and themes where the lives of my informants from the adjacent neighbourhoods became entangled. Here I used the same methods, triangulating observation, participation and ethnographic interviews. I approached these spaces of encounter as well as the other two steps with a methodology developed at the Institute of Social Anthropology at Basel University called the *Emic Evaluation Approach*. This approach consists of a triangulation of three different perspectives, namely, the mapping of actors and spaces, social discourse analysis and practice analysis (Förster et al. 2011, Heer 2011, for more details see the postscript).

The ethnography starts in *chapter two* with what seems at first a classical neighbourhood study. The chapter analyses urban conviviality in the densely populated township of Alexandra, stigmatised as a place of poverty and crime. It shows that the township has been constituted in the past and continues to be constituted by struggles around insider and outsider divisions, mostly centred on questions of access to urban land. Exploring the spaces of everyday life like homes, yards, night spaces and spaces of civic life, it becomes clear that the spatial boundaries of the township shape life powerfully, yet urban dwellers nevertheless create connections across milieus and to the rest of the city. The chapter criticises narratives of cities as a mosaic of separate worlds, and shows that Alexandra township, popularly imagined as segregated and uniformly poor, is in reality a deeply differentiated and connected urban setting.

*Chapter three* focuses on the intimate entanglements between domestic workers and their suburban employers in Linbro Park. The domestic worker–employer relationship constitutes a habitual, quotidian and invisible urban entanglement, connecting lifeworlds and spaces in intimate and lasting ways. The chapter shows that in post-apartheid Linbro Park, relations between employers and workers still resemble relations of domination and resistance (Scott 1990). Yet the chapter also shows how affection plays an important role in these relations, pointing to the ambivalences, tensions and contradictions in intimate entanglements. The chapter criticises the bias in urban studies towards public spaces and the neglect of encounters in private spaces.

*Chapter four* looks at the agency of white property owners in Linbro Park who are confronted with municipal plans to construct public housing for Alexandra's poor in their neighbourhood. What is at stake here is a potential redistribution of land in what used to be a deeply segregated urban area. The chapter focuses on the practices of a social group that engages in a politics of loss, opposing change because they may lose a part of their privilege in the changing city. The politics of loss is part of the current boundary re-articulation processes, both real and imagined, between suburb and township. In the chapter, neighbourhood change emerges as the result of multiple actors trying to shape urban space and conviviality according to their own images and visions. By opposing high densities in urban plans, by hoping to sell land to private developers to avoid expropriation and by trying to install a City Improvement District (CID), the aff luent property owners aspire to save their 'country living in the city' lifestyle.

*Chapter five* looks at the agency of elite urbanites living in Sommerschield II and the agency of Polana Caniço's' residents, by exploring the *politics of proximity* emerging at the transforming boundaries of *Cidade* and *Caniço* in Maputo*.* The chapter focuses, on the one hand, on attempts by urban elites to manage the proximity to their poorer neighbours through a road closure which Polana Caniço's residents destroyed overnight. On the other hand, the chapter writes about the larger gentrification processes shaping this urban area, in the course of which many of Polana Caniço's residents decided to sell their land and move to the outskirts of the city. Entangled neighbourhoods exemplify the continuing relevance of colonial dichotomies, yet they are also the sites for their transformation. The chapter ends with a comparative section bringing together some of the insights from the preceding chapters and outlining four axes of urban differences (among them established–newcomer boundaries and distinctions around property ownership), which are important in these neighbourhoods and which should receive more attention in urban research.

*Chapter six* discusses forms of sociality across class and ethnic boundaries which emerge within shared religious spaces in Maputo and Johannesburg. It shows that religion and spirituality are central yet often neglected aspects of urbanity. Religious practices and spaces inf luence the way urban dwellers use and experience the city and how they become entangled with others. The chapter shows that in both Maputo and Johannesburg, religiously embedded charity is an important form of creating entanglements. The paternal bonds created through alms giving and alms receiving are forms of sociation that integrate socially and spatially diverging milieus not despite but because of inequality. The chapter also explores f leeting encounters at a suburban church and a neighbourhood prayer place, analysing the relations between believers emerging through co-presence in religious spaces. While religion provides discourses of equality which urban dwellers utilise to make claims, everyday interactions are still shaped by the memory of colonial and apartheid racial separation which become easily reactivated in moments of competition and conf lict.

*Chapter seven* shows that shopping malls are important spaces of public life in African cities which need to be understood in their own right, rather than being perceived as inferior in comparison to what some see as 'proper' public spaces. The chapters criticise narratives which portray malls as fortified enclaves that exclude the 'African' urbanity. Using Foucault's concept of *heterotopia* (Foucault 1986 [1967]) the ethnography approaches the Maputo Shopping Centre and the Greenstone Shopping Centre comparatively with regard to spatial practices by users, competing representations of the malls, encounters in malls and the relationship between the mall and other urban settings. Malls as entangled spaces of heterotopia are not uniform, bounded spaces, but rather have multiple layer of meanings produced by actors with diverging powers; from everyday users, environmental activists and property investors to mall managers. Surprisingly, shopping activities not only emphasise inequality between aff luent and poor mallgoers, but also bring about instances of togetherness across multiple social

boundaries. The Maputo Shopping Centre emerges as a *heterotopic mirror*, making the urban dwellers more aware of where their place in society is, yet at the same time also inspiring them to imagine a different life. Greenstone Mall, in contrast, works as a *heterotopia of compensation*, a place that compensates for the perceived dangerousness of suburban street life and the hardship of shack living.

The *closing remarks* suggest that we can think of the entanglements in Maputo and Johannesburg as *blind fields* (Lefebvre 2003 [1970]). Urban entanglements contradict the ideology of separateness that the previous epoch – apartheid and colonialism – entrenched. Still today, they may be *blind fields*, things we don't see because they lie outside our perspective, outside what is epistemologically imaginable to us, and outside that which the past epoch *wanted* us to see. The fact that entanglements remain often unseen and invisible is, therefore, not only a question of a *wrong lens* for looking at the urban, but is also about power and ideology. Powerful actors like urban elites may *refuse* to see and recognise entanglements. In this ethnography, it is then especially members of poor milieus, from Alexandra and Polana Caniço, who make claims to recognise the connectedness and mutual responsibility across social and spatial divides, while members of elite milieus tend to turn a blind eye to this. Seeing Maputo and Johannesburg less as *divided cities* or *cities of walls* and more as *cities of entanglements* is, therefore, also a change from not only looking at the two cities from the perspective of urban elites who desire to withdraw themselves from connectedness, towards looking at Maputo and Johannesburg from the perspective of the urban poor for whom connectedness is an indisputable urban reality, and a basic condition for being able to survive socially and economically in the cities marked by inequalities. What we need, hence, is an ethics and politics of interrelation. How we deal with and shape our entanglements is the key ethical and political question of urban futures.

The *postscript* takes a ref lexive stance toward the research process that led to this book and makes methodological comments about comparative ethnography and comparative urbanism. The postscript criticises the fact that comparative urbanism emphasises theory building through comparison, yet lacks an engagement with the hands-on aspects of doing comparative qualitative urban research. The postscript introduces the disciplinary history of comparison in anthropology, outlines the *biography* of the units of comparison and the analytical framework of this book, and draws attention to the importance of the researcher as an *entangled comparer*. Positionality, ref lexivity, the web of relationships in the fields and specificities of the urban context shape the research process and knowledge production through comparative ethnography in important ways, making it not only a way of *thinking cities through elsewhere*  (Robinson 2016b) but also a way of *experiencing cities through elsewhere.* 

By means of comparison, this ethnography aims to move beyond the confines of the scholarship of South African and Lusophone urban studies, which have both long treated the respective cities as exceptional. In this way, we can move beyond provincialism and develop a new form of urban studies which takes into account the diversity of urban experiences (Lancione and McFarlane 2016, Robinson 2006a, Söderström 2014). This book distances itself clearly from an understanding of comparison that reduces entire cities and the complexities of urban practices to fixed variables, or which explains differences and similarities through simple relations of causality. Rather, this research is comparative in the sense that it explores the urban with two cities, four neighbourhoods and manifold urban milieus in mind. In addition, this

ethnography should not be misunderstood as being split into two parts according to geography; rather, the fundamental concepts developed in one chapter feed into the others and vice versa. *Cities of entanglements*, hence, hopes to invite the reader to look at Johannesburg and Maputo, as well as cities in general, in new ways.

©Barbara Heer, Polana Caniço, 2010

©Barbara Heer, Polana Caniço and Sommerschield II, 2010

©Barbara Heer, Sommerschield II, 2010

Google Maps 2012, amended by the author

Google Maps 2012, amended by the author

## **Navigating Belonging?**  Entangled Spaces, Entangled Lives in Alexandra Township

The Chicago scholar, Robert Ezra Park, once argued: "The processes of segregation establish moral distances which make the city a mosaic of little worlds which touch but do not interpenetrate" (Park 1915: 608). This famous, yet problematic quote is symptomatic of what I call *mosaic thinking*, a way of understanding the relation between people and spaces that haunts much research and many theories on neighbourhoods and cities. Mosaic thinking understands the city as made up of distinct patches (neighbourhoods), each with its colour (identity) and clearly drawn boundaries. Mosaic thinking equalises neighbourhoods with specific social categories or identities, it overemphasises the power of spatial boundaries and it pays no attention to the way the different spaces and lifeworlds are connected in everyday life. What I will argue in this chapter, though, is that neighbourhoods need to be understood in relation to other urban areas and the city at large, and not in isolation. Neighbourhoods are always entangled with other places, and they become constituted through these connections. What the chapter also wants to challenge is representations of such neighbourhoods as homogenously poor and disconnected slums.

Neighbourhoods need to be understood as very diverse places. One has to very carefully disentangle neighbourhoods as social spaces from assumptions about their internal social life. In Alexandra, which this chapter focuses on, the notion of community is central to local politics, and dense social networks indeed exist within yards and across yards. But at the same time, it is also a neighbourhood deeply divided by shifting internal boundaries as well as characterised by anonymity and the difficult conviviality of clashing lifestyles. There is great variety in what the neighbourhood as a social space means in the everyday lives of my informants. Defining neighbourhoods as communities has remained popular in geography where they are often defined as having a cohesive sense of identity, a political or social organisation (Konings, Van Dijk and Foeken 2006: 1). Such definitions are highly problematic as neighbourhoods can be anything from administrative boundaries existing only on bureaucrats' maps to anonymous sleeping places to the sources of identity and the horizon of everyday life. Meanings of neighbourhoods and the relevance of spatial boundaries in everyday life have to be researched and not pre-assumed.

Most of the research on Alexandra is in line with a vast strand of research that treats Johannesburg as a 'problem to be solved', looking at segregation, poverty, inequality and social justice (Beall et al. 2002, Bond 2000, Harrison, Huchzermeyer

and Mayekiso 2003).1 Mbembe and Nuttall (2008: 13) criticise the fact that many township studies pay little attention to the imbrications of city and township. Recent ethnographies on townships, for example by Harber (2011) on Diepsloot or by Bank (2011) on East London, have similarities with American community studies (e. g. Gans 1962, Hannerz 2004 [1969]) in the sense that the overall city just lurks in the background. There is little research which looks at suburban and township spaces within the same analytical frame. This bifurcation of research practice contributes to reproducing the image of South Africa as a bifurcated society with dual cities and downplays the deep connections between different social spaces and social milieus. As this chapter will show, Alexandra has always been and still is tightly integrated and interdependent with the surrounding city. It is embedded in complex social, cultural and economic entanglements with the surrounding suburbs, as well as with other cities and far off rural homes in South Africa and other countries. Based on the ethnography of everyday spaces in Alexandra, the chapter seeks to rethink the category of neighbourhood in the relation to the city. The city is not a "collection of independent realms" but a "series of interconnected spaces and processes" (Srivastava 2014: xx). Moving beyond the city as 'a mosaic of little worlds', Johannesburg emerges as a city produced in the links, connections and relationships between worlds within and across neighbourhoods.

### **Alexandra Collectively Imagined: Narratives and Histories**

The history of Alexandra has been documented meticulously by two historians, Noor Nieftagodien and the late Phil Bonner, in *Alexandra: A history* (Bonner and Nieftagodien 2008). History, or rather memory, is very important if not constitutive for many urban milieus in the township. Memories shape the popular imaginations and identities of what 'Alex', which is what the locals call it, is; five of these important narratives (there are many others) are worth mentioning, as they profoundly shape the way Alexandra's residents talk and feel about the township. *Alexandra* becomes constituted simultaneously and contradictorily by the following diverse narratives, meanings and imaginations.

A 70-year-old descendant of what once used to be property owners (see below) and a well-known 'community leader' said to me in a conversation: "We CAN build Alexandra, we CAN change Alexandra (February 2011).2 In this quote, the 'we' refers to an emic notion of 'community' which residents often invoke. The question of who is considered a legitimate member of this 'we' has been a key contentious issue in township politics across the decades. The following five modes of narrating this 'we' and narrating Alex-

<sup>1</sup>  Research on Alexandra coverslocal politicalstruggle history (Bozzoli 2010, Lucas1995), housing (Charlton 2010, Huchzermeyer 2003), participatory democracy and the Alexandra Renewal Project (Khoza 2008, Sinwell 2008, 2010) and xenophobia (Tafira 2009,for xenophobia in generalseeNyamnjoh 2006). Anotherstrand of research focuses on questions of health andHIV (Le Marcis 2004, 2012).

<sup>2</sup>  The majority of quotations by urban dwellers presented in this book stem from ethnographic interviews conducted between 2010 and 2012 in Johannesburg and Maputo. Interviews conducted in English were transcribed by the author. Interviews conducted in Portuguese were transcribed by Fernando Tivane and translated into English by the author. The indicated age of the urban dwellersrefers to the approximate age atthe time ofthe interview. All personal names, exceptthose by the field assistants, are pseudonyms, and minor personal details have been adapted in order to ensure anonymity.

andra – framing it as a home of proud property owners, a struggle moment, a problem to be solved, 'Gomorrah' or as a space of creativity – exemplify the way the township is highly differentiated, and not simply a homogeneous 'slum'. Alexandra becomes constituted by such competing discourses, diverging lifestyles, manifold sociabilities as well as conf licts. Such diversity *within* neighbourhoods tends to go unseen in imaginations of unequal cities as bifurcated into two worlds, be it suburb and township in Johannesburg, or elite neighbourhoods and *bairro* in Maputo. The analysis of *cities of entanglements* thus starts by deconstructing, through the ethnography of Alexandra, the idea of the poor neighbourhood as homogeneous and enclosed. Alexandra emerges as deeply connected to other spaces, while there are also very powerful boundaries characterising its residents' lives.

### Alexandra – African Land in the White City

Alex is the only township that resisted removal and kept freehold title deeds … Many property owners are still battling to get their granted title deeds – a pride of many South Africans who for many years were denied a right to property (Speech by Obed Bapela at Media Launch Alexandra Centenary, printed document in the press kit, 22 May 2012).

In 2012, Alexandra turned 100 years old, and local organisations and politicians celebrated the occasion extensively by pointing to the importance of the township in the larger South African context. Obed Bapela, born in Alexandra and a deputy minister since 2010, emphasised in the speech above how 'Alex' had survived through resistance to removals, a struggle personified by the figure of the property owner (Bapela 2012). This is one of five key narratives about the township, claiming that Alexandra is a freehold township, the only one that survived apartheid, and hence home to proud African landowners. As Eriksen points out, such myths of origin – whether they are recent creations (Hobsbawm and Ranger 1984) or simply particular views of the past – are powerful tools for creating a sense of belonging, for spreading and legitimating a particular view of the present and making political claims (Eriksen 2001a: 273). Similar to what Lentz (cf. Lentz 2013b: 4) experienced in her research on rural land conf licts in West Africa, the descendants of these former landowners in Alexandra were also always eager to recount the history of the neighbourhood and how their ancestors acquired land and were then expropriated.

In 1912, the European investor Papenfus bought the land on which Alexandra stands and declared it a 'freehold' township, where non-Whites were allowed to acquire land.3 Because of the 1913 Natives Urban Land Act, it soon became one of the very few places where Africans could buy urban land. Better-off Africans – commercially successful farmers, or members of the small Black petty bourgeoisie – bought stands in Alexandra and established a social order based on tenant–landlord relationships. In order to pay off their bonds, they rented out rooms and land to poorer Africans seeking a better life in the city. Based on the stable rent income and their business activities they could turn themselves into a Black urban middle class (Heer 2018). As an elderly property owner recalled in 2011: "We were the only Black people to own property then. The gov-

<sup>3</sup>  A more detailed exploration of thistopiccan be found inHeer (2018).

ernment didn't like that at times, that's why they wanted to destroy Alexandra" (Alpoa Youth lecture movement, February 2011).

Landlord–tenant relations became contested, unsettled and more complex when the descendants of the approximately 2000 original property owners were expropriated during apartheid. Tens of thousands of people were removed, yet many disenfranchised property owners could remain in their houses, if only without formal property rights. A complex web of unofficial property rights has emerged since then. Some expropriated property owners managed to continue to exercise exclusive control over their property and extract rents from tenants. Yet there were also many tenants who turned themselves into landlords by subletting rooms and building shacks. In the context of poverty and extreme scarcity of land, extracting rents is till today a key income strategy for many Alexandra residents who manage somehow to gain access to a house, be it by waiting for a government house, by occupying open space or by informally or formally inheriting a house in which parents, relatives or acquaintances have lived.

The descendants of the former property owners have remained a politically powerful, vocal milieu, even though many have lost the material base of their middle-class identity (Heer 2018). Property ownership and expropriation were central to the discourses of the liberation movement (James, Ngonini and Nkadimeng 2005: 827), and still receive considerable attention in ANC political discourse. During the early 1990s, a symbolically charged land reform programme was developed. In the 1990s, 1695 families of former Alexandra property owners made a successful group claim under the Restitution of Land Rights Act (1994). The case was appealed and only in 2016 was an agreement between the landowners and the government signed (Heer 2018), with the topic of land restitution having occupied an important place in the public debates and interventions in Alexandra's urban renewal in the last two decades. A court interdict forbade the Alexandra Renewal Project (ARP), a f lagship urban renewal project in place since 2001, to develop any of the contested properties till the case had been settled. This forced the ARP to construct new housing on the periphery of Alexandra, like at the Far East Bank, and it also increased pressures to find land in nearby suburbs, like Linbro Park (see chapter 4).

As the speech by Obed Bapela at the centenary celebrations in 2012 exemplifies, in official discourse land restitution for former property owners is upheld as an important political goal. In the complexities of local politics, however, exponents of the ANC have an ambivalent relationship with it: they maintain that the property owners are 'self-interested' and that their land claim is blocking development for the rest of Alexandra's residents. They argue that because of the court interdict the ARP cannot progress with the upgrade of the housing stock in Old Alexandra (Thabo Mokgothu, political advisor to the ARP, February 2011; Ron, executive of the ARP, March and June 2012). Like the rural case described by James (2000), the local ANC branch prefers to support the side of the current occupants and tenants than the former property owners. Hence, interest groups like Alpoa (Alexandra Land and Property Owners' Association, see below), which claims to represent the interests of former property owners, engage in many practices to keep the imagery of Alexandra as a historically exceptional place, the only urban neighbourhood with African property ownership during apartheid, alive. In 2011 and 2012, for example, their youth group organised weekly events called 'Lecture movement' where elders from Alpoa recounted their version of Alexandra to young urban dwellers. Activists from Alpoa also actively build relationships – entanglements – with researchers and journalists to ensure that their version of what 'Alex' is enters their writing.

### Alexandra – A Struggle Monument

It's an old population … I think to an extent it's unique. A lot of oursettlementsin South Africa are not old … But they feel very strong and attached to Alex; they are very … very emotional about Alex. And they are very proud of the fact that the government of the day then did not succeed in getting them out. The fact that so many political leaders in South Africa, leaders in business and in politics, came from Alex (Ron, executive employee at ARP, March 2011).

In this conversation, Ron, an executive employee at ARP, refers to a second important narrative about Alexandra, namely 'Alex' as a site of memory for the successful resistance against apartheid, the neighbourhood thus becoming a struggle monument. This narrative emphasises Alexandra's role in the political struggle against apartheid, the many South African leaders who came from the township, and the continuing power in the post-apartheid phase to shape national politics through (violent) protests. This empowering image, which links the political history of the township to the national culture of memory and the master narratives of struggle against apartheid is very common among politicians and bureaucrats, and the middle-aged and elderly residents who experienced the struggle. This narration involves the practice of entangling Alexandra with other South African places, of linking Alexandra's history with the history of the nation, and hence of making Alexandra relevant beyond itself.

Alexandra indeed has a history of self-governance and radical democratic politics. When White suburbs around Alexandra grew in the 1930s, their residents started to demand that what they called a 'Black spot', a black neighbourhood in a White part of the city, be removed. There were four attempts by the state to remove the township, namely in 1940, 1943, 1950 and 1979 (Bonner and Nieftagodien 2008, Curry 2012: 2012). Alexandra residents successfully fought off extinction each time. The threat of removal shaped township politics until 8 May 1979, when the 'Save Alex Campaign' definitively put a stop to it and government abolished the plans.

In the 1980s, during the height of resistance against apartheid, self-governance bodies were put in place in many townships, including Alexandra (so-called 'organs of people's power'). At that time, many young men fought on the streets against the police and the military of the apartheid regime, risking their education and lives by protesting for months and years. One of them was my field assistant Thabo Mopasi, born around 1970 and today father of eight children. He is a descendant of long-term tenants, an ANC (African National Congress) activist and an inf luential personality in the township. He has been involved in many research projects in the township and also supported this study greatly as gatekeeper. For men like him, memories of the struggle against apartheid became inscribed in Alexandra's spaces; when I walked with him past certain street corners, he would tell me about the protests that took place there and the injuries his friends suffered there. These are not happy memories, as they involve violence and the loss of friends to bullets; but remembering his contribution to the overthrow of oppressive structures instils feelings of pride and self-worth in the present. One day I watched the news on television with Thabo and some of his friends;

this was during the Arab Spring and the TV was broadcasting how activists had made the Egyptian president Mubarak step down. The emotions of Thabo and his friends overf lowed, they put their fists in their air and started to sing struggle songs from the 1980s. They were re-enacting what Bank called the "social and cultural style of the comrades", the lifestyle of the youth involved in the struggle against apartheid (Bank 2011). This ethnographic instance exemplifies how the decades of resistance created a distinct and strong political culture and political agency in Alexandra; how the struggle shaped the contemporary "modes of being" (ibid: pos. 77), the ways of thinking, feeling, talking, acting; the ways of relating to others, to the built environment and to oneself. The imagery about Alexandra as a place of struggle is not only a politically instrumentalised discourse of memory but is also inscribed in the township as a lived space.

The memory discourse became formalised and officialised in 2011 by over 100 heritage site signs, a public history and tourism project forming part of a larger 'Alexandra Tourism Development Project' (funded by the ARP and the tourism department). The signs mark places which are considered memorable, like the yard where Nelson Mandela used to live. Although the heritage signs are officially designed as an attraction aimed at tourists, organisations like Alpoa also use them to make sure that knowledge about the past and the related pride and identity are transmitted to the youth.

### A Problem to be Solved

One day in February 2011 I went for a walk through the township with four young men. They were unemployed and created purpose in their lives by engaging in the arts, volunteering at the youth desk of the local Community Police Forum (CPF) and by maintaining an urban food garden project in Far East Bank which was where we were heading. When we walked past a taxi rank, they explained to me that this spot was very dark and dangerous, which was why they patrolled there on weekends with the CPF. When we walked past a cemetery, they explained that the cemeteries were full and residents had to bury their relatives in faraway Midrand, which constituted a spiritual problem. When walking past the Jukskei River they pointed out proudly the many filled rubbish bags on the river bank and explained that they were part of a public work programme engaged in cleaning the river. These men, and many others whom I accompanied on walks through the township, constantly commented on urban issues which they defined as social problems.

Alexandra residents often conceive of living in Alexandra as harsh and there is general agreement on the definition of social problems which affect almost everyone, like HIV, poverty, lack of housing, environmental degradation, teenage pregnancy, crime and drugs. Urbanites who are engages in civic organisations, youth groups, political parties or other community structures, define many of the phenomena they observe as problems, collect information, talk and make political claims about them. The best-informed commentators, researchers, analysts and critics of Alexandra are the township dwellers themselves This everyday construction of Alexandra as a space full of problems present in local politics, everyday political discussions and everyday life is inf luenced by the terminology of social problems used by NGOs and the government. Children become familiarised with these discourses in, for example, in school where they learn about child pregnancy, HIV and drug abuse. It is also ref lected in academic and journalistic work on Alexandra.

A related, very powerful narrative on Alexandra, often upheld by outsiders and newcomers, constructs the township as an extremely dangerous place of crime, a 'nogo' area and hotbed of criminal activities. This outsider view powerfully shapes the social reality of youth lifestyles in the township.

When I have to say "I am from Alexandra", when I am out there, people … change their face,their gesture changes…all atmosphere changes because ofthe area that you come from, because of, you know, the past … crime rate, and stuf f like that. […] They look at me asif I were from a savage area. As you have been around, Barbara, in Alexandra for quite a period oftime, you have seen thatAlexandra people are normal people,they are modern people, it's a civilised community. It'sjustthatthe infrastructure development overshadowsthat; it paints a dark picture to the outsideworld (Tebogo, a 30-something member of Alpoa Youth, February 2011).

In this image of Alexandra as crime-ridden, place becomes a stigma, attached to Tebogo's body, like a 'spatial destiny' and an isomorphism of place and identity (Tonkiss 2005: 45). In particular, young, male and black township dwellers like Tebogo become 'othered', what he articulates as 'being treated like a savage'. Young men from Alexandra especially are confronted with these prejudices when looking for a job. Some resort to giving an address in one of the neighbouring formerly White suburbs on their CV in order to avoid the stigma.

Crime is not only stigmatised by outsiders but is also a social reality. Alexandra is indeed not a safe place to live. The 2005 Benchmark survey conducted by the ARP showed that robbery and assault were the most common crime (Alexandra Renewal Project 2005). Forty-one per cent of the participants said that they did not feel safe at all, while only 20 per cent said they felt very safe (ibid). For women, sexual assault in public spaces and violence by romantic partners are agonising urban realities.

### 'Gomorrah'

There is a further layer to the idea of Alexandra as a problem which deserves mentioning. Young Alexandra dwellers often link the local notion of *ikazi* (home) to the notion of 'ghetto', inspired by globalising countercultures like hip hop emerging from black American ghettos (Castañeda 2012: 177). Rappers and others refer to Alex affectionately as 'Gomorrah'*,* like the biblical city full of sinners eventually destroyed by God. This *ikazi* discourse can be read as a youthful, artistic reinterpretation of Alexandra's stigma as crime-ridden and poverty stricken.

I wish people knew Alex like I do … it's not about shacks that are on little space … It's people coming together from close to far to make a good living … I'm proud to be born here…I'm proud to be aGomorian…(Anonymised Facebook entry, pinboardAlexMews Gomora, January 2012).

There is no place lyk alex A.K.A G-town even celebs wish 2be here every mondy nyt n sundy nyt,hw grate is that huh ...? (anonymised comment on the post above, January 2012).

#### 52 Cities of Entanglements

Growing up close to criminal gangs is portrayed as making township youth especially 'street wise' and 'cool'. The notion of the 'cool' township youth becomes constituted in relation to notions of the 'soft' youth of the suburbs. Letsatsi, the 17-year-old daughter of a middle-class household in River Park, a newer section of Alexandra, who was attending a mixed private school in a nearby suburb, expressed the typical view that the boys from the suburbs are seen as silly and naïve by township youth, not least because they tend to live indoors and knew less about the world than those in the township who were exposed to everything and had to learn how to deal with it. Letsatsi would accompany her suburban friends to what she calls 'classy' clubs in malls in Rosebank or Midrand, but she would actually prefer to go out in Alexandra as she considered herself a 'ghetto girl' who liked the rough style of township clubbing, although she was aware of the risks.

I am a ghetto girl, I prefer Alex. I enjoy there, even though it's not safe. It depends on what kind of people you are around and stuf f. If you are around gangsters, you won't last actually. But if you are around good people in that area, I am telling you, you come outclean (Letsatsi, River Park resident, June 2012).

These examples effectively illustrate the way people subjected to 'othering' do not automatically adopt this position, but employ practices to deal with it, like constructing themselves as superior to the 'soft' suburban youth because of dealing with hardship every day and seeing it as a source of 'coolness'.

### Space of Creativity

A further image, strongly contradicting the narrative of 'ghetto' and spatial destiny, constructs Alexandra as a site of relative freedom and possibilities, which leads to cultural creativity and vibrant lifestyles. Like Sophiatown (Hannerz 1994), Alexandra has seen artistic productions of theatre, music and other forms of art, and is known as a legendary party place to which people from all over Johannesburg, at least from Black milieus, have been going for decades. The relative 'freedom' as a freehold township not completely under the grip of the apartheid state provided a fertile setting for political movements, cultural creativity and leisure spaces. During apartheid, when public life was restricted for the Black domestic workers living on their employers' properties in the White suburbs, many came to Alexandra at the weekend to party. Bonner and Nieftagodien argue that politics and culture "fed off each other" (Bonner and Nieftagodien 2008: 169). Today, Alexandra's nightlife continues to be known all over Johannesburg, not least because of hip hop, pantusla and kwaito artists and deep house DJs, some of whom have become famous nationally (for example Malamesh whose songs were played in the soap opera Zone 14 on SABC1, or Razor from LiquidDeep). There are music production studios, theatre productions, and manifold youth groups engaging in various dance forms, as well as many choirs and artists.

### **Entangled Anxieties in Home Spaces**

With a surface area of about eight square kilometres and an estimated number of inhabitants of 340,000 people (Alexandra Renewal Project 2005), high density living is a powerful everyday reality in the township. During the different phases of the city's history, moving to Alexandra and other relatively 'free' areas was a strategy for urbanites to avoid living in a state-planned township like Soweto or being removed to rural areas. Here, beyond what the state could 'see' (Bollens 2007, Scott 1999), people considered illegitimate urbanites by the state found the space to nevertheless develop an urban existence. However, the attractiveness of Alexandra for people arriving in the city turned housing and urban land into highly scarce, contested resources in the township. The private spaces of houses, shacks and rented rooms in Alexandra are hence also highly political and politicised spaces in the neighbourhood, as they are in many other South African townships. Urban scholarship, however, often overemphasises the significance of struggles around public spaces and neglects what happens in private spaces like homes.

Alexandra is home to a large shifting population of people who arrived recently, newcomers with strong connections to rural areas or other African countries where they aim to return one day. Some of them intend to reside only temporarily in Alexandra before finding a home somewhere else, while others eventually become long-term tenants or even 'owners' of properties they acquired somehow. Nnana, a field assistant living in River Park, and her family, for example, belong to this milieu of newcomers. Nnana's mother moved to Johannesburg in the 1980s, and since then she worked as a domestic worker in Linbro Park, while her children stayed at home in a rural area with relatives.<sup>4</sup> Since they finished school, her children have been moving between Johannesburg and the rural area, depending on where they managed to find work. I became close friends with Nnana and her family during fieldwork and they helped me to get in contact with residents of River Park, the section of Alexandra they were living in. They rented a tiny shack of about sixteen square metres in size in a backyard they shared with other shack dwellers, most of whom also came from neighbouring countries and rural areas. The shack had a large fridge in which the neighbours also stored their food; although they did not consider each other friends, the backyard neighbours shared resources and everyday lives. Nnana's shack was divided into a more public area with a couch where relatives, neighbours and friends often sat during the day, and a more private realm, a large bed hidden behind a curtain, where the five family members (the mother, the two adult daughters and two children) slept. As is common in many Alexandra households, each family member kept her clothes and other personal belongings stored in a bag, neatly separated from the belongings of the others, and squeezed together into one closet.

The second household I gained intimate knowledge of was that of my field assistant, Thabo, who was living with his wife and nine children in a complex amalgam of rooms

<sup>4</sup>  River Park was built in the 1990s, on the eastern border of the township adjacent to the suburb of Lombardy East and the national highway, the N3. It is geographically close to Linbro Park and Greenstone shopping mall, which is why Iconducted many interviewsin 2012 in this area which residentsconsider a section of Alexandra. River Park was built for Alexandra residents who were internally displaced during the civilwar of1991/1992 in Alexandra.Nnana'sfamilywasliving in a rented shack in River Park.

belonging to house in 12th Avenue, where his brother was also living with his family. They had divided the house they had inherited from their parents into two parts, nevertheless Thabo and his brother often fought; such conf licts are typical of the dense living in Alexandra. His wife's parents lived in an adjacent yard on 13th Avenue and owing to a lack of space three of his nine children usually slept at his in-laws. Thabo's wife also spent much time during the day at her mother's house, so that the two houses and yards, although spatially separated, were actually closely interlinked through the everyday practices of their inhabitants. Such interconnected households may even stretch across sub-quarters; daughters living with boyfriends in a new section of Alexandra may spend the day at their mothers' house in 'old' Alex, enjoying the sociability and support networks they still lack at the new place.

When space is scarce, privacy becomes interactionally achieved and highly contested, and is structured by the temporal rhythms of the household. At Thabo's house, there were doors which people could close, for example to take a bath out of the bucket. When a woman was talking a bath, other women and children would walk in and out of the room, but men and teenage boys would stay away, respecting the temporary space of female intimacy. In Nnana's shack where there was no room to retreat to, male visitors would leave the shack whenever someone was about to take a bath, while women would sometimes stay. The person taking a bath would conceal her body behind the curtain separating the living area from the bed. The person would even continue talking to the female visitors while washing behind the curtain. Household members could lie on the bed behind the curtain, napping or watching TV, with little children crawling around on the bed. However, visitors and other family members, sitting not even two metres away on the couch, would not talk to the person behind the curtain, as her gesture of lying down behind the closed curtain symbolised her wish for distanciation, her desire to be left alone.

Besides material boundaries like curtains or doors and social boundaries like silence, time also offers boundaries which enable privacy. When I was staying at Thabo's in-laws' house, I used to share the couch in the living room with his oldest daughter. When everybody else was asleep and the house locked, we chatted about boyfriends, and secretly drank beer which was left over from the weekend. Many teenagers use these quiet nightly hours to escape from the house and to go to nightclubs or meet with a boyfriend in the yard. Privacy in this moment meant being able to evade parental control for a while.

Lack of privacy means being under the constant control of parents, being visibly exposed to others and being forced to interact with others. Attempts to create privacy are often contested in the context of scarcity of private space, which causes anxiety and conf lict. Many Alexandra dwellers experience life as harsh because of unemployment, badly equipped schools, decaying urban infrastructure, interpersonal violence, crime and the like. Many residents hence complain about stress, which, in their view, is significantly enhanced by the anxieties around private space. Many understand the lack of space as pathological and as limiting their agency: "You can't be creative in a space like this", explained Paul, a 60-year-old pensioner in Alexandra (February 2011), referring to the high residential density. Urban dwellers tend to explain social ills like child abuse and domestic violence in terms of the lack of privacy: living on top of each other makes things visible to children and teenagers which they should not see, according to the residents. The head of a primary school explained in an interview that

some brothers abuse their sisters because they imitate the sexual practices of their parents: "The shacks are too small. What they observe, they do" (Head of Carter Primary School, February 2011). Young people are said to be inf luenced by seeing neighbours taking drugs or getting drunk, or by seeing their neighbours who rob houses in nearby suburbs leading comfortable lives.

When I started doing fieldwork in Alexandra, I wanted to explore diverse meanings of public space. But the longer I spent in the township, the more I learnt to recognise that home space and private land are foremost in the minds of Alexandra residents. Home space is the quintessential private space and refers to the spaces within which urban residents 'dwell'.<sup>5</sup> Dwelling here refers to both a place and a process and involves social and spatial practices (Jenkins 2012a: 6). In Maputo and Johannesburg, dwelling and home space are the areas in which I observed major differences in urban life. In Maputo the African *Caniço* urban areas developed largely unplanned and under less control than townships in South Africa. The 'informal' form of urban development on the outskirts of Maputo makes access to urban land for the poor much easier in comparison to Johannesburg. Many of the young people I met in Maputo, from lower to upper income levels, were working towards constructing a house, based on continuous, patient savings. They would buy a plot on the informal land market on the peri-urban fringes of the city, then they would buy bags of cements whenever they had saved some money and they would build a house over the years; usually a simple house, but more or less conforming to what they considered to be a decent house. Such a route to becoming a home owner did not exist for the majority of urbanites I interacted with in Johannesburg. Although many urban dwellers in Johannesburg engage in the self-production of home space, these practices are considered illegal by the state and neither do they conform to the residents' ideas of decent housing. In Alexandra, there are many informal practices of land access, like the occupation of industrial buildings or buildings under construction, the illegal occupation of land in shack settlements, or the renting of an illegally constructed shack at the back of an RDP house.6 But in contrast to Maputo, where constructing an own house on the peri-urban fringe is considered a route to social growth, the illegal and informal forms of land access existing in Alexandra tend to be associated with stigma and poverty. While in absolute terms, poverty is far more rampant in Maputo, relative deprivation in the sense of decent housing seems to be a bigger problem in Johannesburg – at least for many of Alexandra's milieus, who have a vexed, troubled relationship with urban land distinct from what I experienced in Maputo.

Alexandra is only about a kilometre away from aff luent Sandton, the 'Manhattan' of Africa, with its many office towers, fancy shopping malls, international hotels and elegant restaurants. Seeing the skyline of Sandton in the background of the sea of dense housing in Alexandra is experienced and represented by many residents as the epitome of their exclusion and the rampant inequality in the 'world class African city', as the City of Johannesburg likes to brand itself. Alexandra and Sandton, though, should not be seen as separate entities, rather, they are structurally, economically, culturally and socially deeply connected and interdependent on each other.

<sup>5</sup>  I derive the notion from the research programme 'Home Space in African Cities' (Jenkins 2012a).

<sup>6</sup>  The government-built houses are colloquially referred to as'RDP' af ter the Reconstruction and Development Programme introduced in the 1990s.

Sandton City as you see it, it has been built by the people of Alexander. Many people here in Alexander participated in the building of Sandton city (Thabo Mopasi, a 40-something member of the long-term tenant milieu and field assistant, February 2011).

Many Alexandra residents like to emphasise this interdependence. In the quote above, Thabo points out how Alexandra's labour power was used to build the fancy suburb. This point is theoretically extremely important for the way in which we understand Alexandra's relationship with Sandton, namely, as mutually entangled and interdependent. This point builds on a Brazilian strand of favela research, by among others Epstein (1973), who systematically investigated relations between Brasília's' spontaneous, poor areas and the planned city and concluded that these are reciprocal yet asymmetric patron–client relationships, on which both – the elite in the planned city and the poor in the satellite towns – depend. Similarly, Perlman, in her ethnographic study on favelas in Rio de Janeiro, reconceptualised the urban poor not as marginal in the sense of 'outside the system', but as asymmetrically integrated into society (Perlman 1979, see also Aceska, Heer and Kaiser-Grolimund 2019).

Issues around waste and hygiene also materialise as intricate entanglements between Alexandra and Sandton. When 340,000 urbanites share a few square kilometres, the disposal of waste becomes a pressing issue, especially as waste collection by the underfunded municipality is not always reliable. Alexandra is experienced by many of its residents as a very dirty place. The epitome of the dirt is the rats that surface at night.<sup>7</sup> Rats, almost the size of cats, run around the yards and streets. For this reason, residents store their food carefully in buckets. Residents told me with disgust that sleeping children's extremities are sometimes eaten away by the nightly visitors. Many people experience the rat plague as one of the most disturbing and unhealthy aspects of living in the township. In terms of social analysis, the disgust felt and the tabloid media that fuel the hysteria can be read as a preoccupation with social disorder. Social disorder is what dirt is mainly about (Dlamini 2010: 68, Douglas 2005 [1966]: 2).

Muhle, a neighbour of Thabo's in-laws' yard in her forties, explained to me in March 2011 that she usually cleans her handbag before she leaves for work because cockroaches sometimes crawl into it at night. She feared that a cockroach might suddenly crawl out of in her bag in a minibus taxi or at work, which would be very embarrassing for her. Decency and cleanliness considered appropriate for urban places like taxis, malls or the workplace demand everyday practices for keeping the dirt at bay. Muhle and other Alexandra residents were very much aware that the neighbouring suburbs were not as cockroach or rat-invested as Alexandra, not least because waste collection worked better there. The rats are then not only about social disorder but also a reminder of the township's exclusion and marginalisation within the 'world class' city.

Yet even the intimate practices of personal hygiene in the township are connected to the economies of aff luent Sandton. In Alexandra, I asked a couple of friends to keep

<sup>7</sup>  See the remarkable short documentary by the project Hillside Digital/Siyakhona "Rats in Alexandra Township" (www.youtube.com/watch?v=A3oU5MeeoPY). The documentary was broadcasted on national TV (SABC) in 2010. The City of Johannesburg Region E initiated a clean-up campaign on 1 September 2010, presenting it as a response to the documentary. However, rats continue to be a major concern in Alexandra and other townships.

a diary of their everyday routines (Zimmerman and Wieder 1977), based on which we had long conversations. When interviewing Muhle, after she had filled in a self-administered diary, we came across the topic of what she called *toilet rolls:*

Barbara: You wrote here in the diary: "The other lady from the next yard called me to take toilet rolls". What does this mean? Muhle: You know, where she works, they use toilet rolls to wipe their hands af ter washing them. So that lady, she took those toilet rolls from the plastics they throw them into, dried them, and she gave them to us, to use them for the toilet (Muhle, 40-year-old neighbour of Thabo'sin-laws, April 2011).

Women like Muhle's neighbour who work in the malls and the office towers of nearby Sandton collect used paper towels from the toilets they clean. They take them home, dry them, and provide their own household and their neighbours with this paper to be used as toilet paper. Thus, through toilet rolls, Sandton's malls and Alexandra's home spaces become intimately connected.

The inequality for which Johannesburg is so famous not only exists between people who are strangers to each other but also characterises home spaces and families, especially if family members have come to belong to different urban milieus and have moved, for example, to a suburb. The southern boundary of Alexandra is formed by London Road, an east–west route which connects Sandton to Linbro Park and Modderfontein. A couple of warehouses separate the busy road from the previously Whitezoned suburbs of Kew, Bramley View and Lombardy West. What used to be a strict social, legal and physical boundary between the Black township and the White-zoned suburb has nowadays become f luid: Alexandra has – as urban dwellers would say– 'expanded'; squatters have taken over warehouses and constructed shacks around them. However, many Alexandra residents have also become more aff luent and have moved to the neighbouring suburbs. They constitute a new middle class; a middle class that started to emerge slowly in the 1980s and has grown considerably since 1994. In the 1960s, only 15 per cent of middle-class employment was non-White. By 2001 they constituted 50 per cent (Crankshaw 2008: 1695). In Alexandra, such social climbers are hard put to find a house they regard as appropriate to their new social status, so many decide to move out if they can afford it.

Tebogo is about 35 years old, belongs to the long-term tenant milieu and lives with her husband and children in an inherited house in the old sections of Alexandra. Her cousin, Steven, is a pastor and the owner of a successful company. He bought a house in the adjacent Lombardy East, a few kilometres away from where he grew up. The house is an old suburban house with a rough outside but with an expensive interior design. There is expensive leather furniture, shiny tiling and the most recent kitchen technology. The spatial proximity to his relatives and friends in Alexandra allows Steven to keep up his social commitments easily. Many others have moved further away to places like Midrand where they now live in new-built fashionable townhouse complexes. Some of these 'movers out' bring their children daily to their Alexandra homes so that relatives can look after them while they are at work. Some eventually move back to the township because they feel isolated in the townhouse complexes and continue to spend most of their social lives in Alexandra. Moving out of Alexandra also has advantages, however. The geographical distance can provide a social distanciation and a relief from the many demands for money, the conf licts and jealousy from neighbours

and family. Many also move out when they can afford it because they want to escape the crime and violence. Because of these processes, researchers say that desegregation is occurring along class lines: the previously racially defined townships are still almost exclusively black, while the Northern suburbs are becoming desegregated for those who belong to middle- and upper-class milieus (ibid: 1698).

One weekend in March 2011, I accompanied Tebogo and her cousin Steven to celebrate his mother's birthday. His mother was living in Rabie Ridge, a township established during apartheid for 'coloured' people who were forcibly removed from Alexandra. Steven, the wealthy pastor living in Lombardy East, had built a beautiful house for his mother. Tebogo and her cousin's wife were both helping in the kitchen; Tebogo's hands were rough from washing and cleaning the township house of her large family, while the pastor's wife living in the suburb had long gel nails. The pastor's wife was well aware of how a daughter-in-law should behave at such a celebration; she had grown up in a township herself. But as the wife of a wealthy pastor, she had become used to different bodily practices and habits to Tebogo. Both Tebogo and her cousin's wife had brought their babies with them. Tebogo had put four diapers and an extra set of clothing for the baby in a handbag; the baby was wearing a practical dress which had been worn by many children of the neighbours sharing the same yard. Her cousin's wife, however, had brought along a large bag full of diapers, processed baby food in glass jars and countless other bits of baby equipment. Her son was dressed in fashionable baby jeans and sneakers from a famous brand. I noticed the way Tebogo became self-conscious about her milieu-specific child-rearing practices in the presence of her cousin's wife. Tebogo observed that the pastor's wife was constantly cleaning her baby boy's runny nose; she then discovered that her own baby had dried snot under her nose which she had not paid attention to for a while. She cleaned the snot with her figure and wanted to clean her finger on her dress, then looked at the pastor's wife and decided to ask me for a napkin. She thus tried to adhere to ideals of hygiene and cleanliness which she saw as typical of the suburban, better-off lifestyle of her cousin's family. The women then together cooked the food that the pastor had bought for his mother's birthday. Through the familiar, shared practice of cleaning vegetables and cooking together, the social tensions eased a little. On the way home, Tebogo told me that the pastor's wife was very jealous; a comment which showed, I believe, how these entanglements – family bonds criss-crossing township and suburban lifestyles, enmeshing poor and new middle-class urban dwellers – are in a way ordinary yet also contradictory sites of contentions, tensions and conf licts.

### **Solidarity and Power in Yard Life**

Themba lived in the same yard as Thabo's in-laws, where she inhabited a single room, together with her fifteen daughters and grandchildren. As a joke she liked to say that because so many bodies kept the house warm, she never had to use a heater and so she could save money (Themba, March 2011). One advantage of living in close proximity to others is the sociality and mutual help from neighbours who share the same yard.

Here in Alexander, our houses are very close to each other. If you don't have a phone, you go to your neighbour and ask. In Soweto, each and every one has his own yard. So, each and every one is minding his or her business. In Alexandra, it's very nice, it's like we are a family, you don't struggle very much. But: You can't go to a neighbour all the time to ask something to eat. We are 17 here in the house! The neighbour won't able to provide a mealie meal for us(Amahle, resident in Alexandra, March 2011).

In this conversation, Amahle, another neighbour in Thabo's in-laws' yard, points out positive aspects of proximity to neighbours, but also explains the limits of reciprocity in this context of poverty.

Yards are communal spaces shared by different families whose houses, added rooms or shacks are located on the same property stand. Yards constitute a middle ground between the public space of the street and the private space of the home. Yards are gendered spaces, shaped by normative expectations of masculinity and femininity and the gendered division of labour. Once the children have left for school in the mornings, women start their routine of cleaning their houses. Clothes washing they usually do in the yard, and later the women interact with female neighbours who are also busy with their cleaning routine. When the women have finished their chores, they may take a bath and then sit on the fringes of the yard observing the male-dominated life on the street while the children who are too young to go to school play together in the yard. The women only leave the yard once they have finished the daily household chores and have organised someone to look after the children. Many men, often unemployed, gather in less female-dominated spaces like shebeens, or on street corners, or they go around looking for a job.

In summer, women withdraw to the private spaces of the houses, shacks and rooms only at night. Neighbours live close to each other, physically and socially close, and share each other's dramas like a death, a lost job, family fights, as well as their celebrations such as a birthday, a pregnancy or a new job. They give each other emotional support and, if possible, share material resources. Residents of Thabo's in-laws' yard told me that it would be hard for them to eat a nice dinner knowing that the family next door had no food; hence, neighbours often bring each other plates of food. As Simone pointed out, the capacity to treat neighbours, kin and friends well is a central ethos of social networks in African cities where access to material well-being must be channelled through a nexus of social relations (Simone 1994: 17). When a baby is born in a yard, female neighbours bring their old baby clothes to the new mother, and she will pass them on to the next new mother. Sometimes this clothes-sharing circuit remains within the networks of the yard, sometimes people from neighbouring yards participate. The boundary of the yard is therefore a social one rather than a geographical one; the metaphorical boundary of the yard ends where the everyday sharing ends, not where the stand ends.

The households in a yard negotiate a rotation scheme for shared facilities: the women use the shared washing lines in turn, and also clean the shared toilet in turns. Not least because of the different household sizes and different working hours, this often leads to conf lict. In Alexandra, there is a saying that what people fight about most in the yards are the washing lines.

The communal governance of the yard and toilets and the emotional and material exchanges between neighbours hold the potential for tension and conf lict: neighbours may become aff licted by the drug abuse, domestic violence or gang violence that other yard members are involved in. Also the ability to rely on social networks should not be romanticised or idealised: as Marcel Mauss argued so well, every gift demands a counter-gift (Mauss 1990 [1925]). When the balance of reciprocity cannot be reinstated over a prolonged period, conf lict and pressurising expectations, demands and jealousy emerge. According to Bähre "solidarity is not opposed to conf lict, nor does conf lict necessarily take place outside of the realm of solidarity. Instead, rivalry, conf lict, jealousy, and aggression can be at the heart of solidarity networks" (Bähre 2007: 52).

Bähre makes a strong point against romantic notions of solidarity in the way they are sometimes espoused by development research and institutions like the World Bank. Ambivalence is thus key for any analysis of solidarity. My repeated presence in Thabo's in-laws' yard, for example, was a double-edged sword for my host family: on the one hand Thabo's in-laws found it an exciting experience to have a European anthropologist as guest which even brought some prestige, but on the other hand, it increased pressure and demands from their neighbours who thought that I was giving them money.

The yards are spaces of dense and contested conviviality that are also shaped by difference and inequality. One Sunday in February 2011, Thabo and I visited a former fighter in the ANC's military wing (Umkhonto we Sizwe) who was living in the new RDP section in Fast East Bank. As a former police officer, he had a pension and he received health support because of his HIV-positive status. He was a respected elder and had more money than the neighbouring families in the yard. The daughter of a destitute female-headed household living in the same yard would clean and cook for him, and he in turn would assist her family with groceries. During our visit, he became increasingly drunk and started to make explicit jokes about these female neighbours, telling me repeatedly that he was an elder who should be honoured. The symbolic power of his former fighter status, seniority and masculinity, as well as his financial power, placed him in a powerful position over his female neighbours, such that he was able to demand all kinds of services from them.

In the yards, the lives of neighbours who may belong to different generations and lifestyle groups become entangled and may clash. Residents told me that sharing a yard with a criminal gang could be dangerous and annoying, as they commit robberies at night and then party with drugs and loud music throughout the day and night while the families next to them try to lead normal lives. Being around addicts and drunkards and hearing or even seeing violent outbursts in the family life next door is part of what it means to live in the township.

In the past, the social relations embedded in yard life had a strong political significance: they became the basis for 'people's power', a form of self-governance installed in the 1980s as part of the insurrection and struggle against apartheid (Bonner and Nieftagodien 2008: 265, 281, Bozzoli 2010, Lucas 1995). Still today, yard relations and the sharing of resources in yards have an important function, assisting urbanites through the vagaries of economic and other hardships in the city. As Bank (2011) rightly points out, yards are hidden yet very important spaces of cultural life and of social reproduction in townships. Nevertheless, there is still little appreciation of how these yard spaces work socially, and they continue to be under-researched (Bank 2011: 190-191).

### **Street Life and Difference**

The drive from peri-urban Linbro Park with its quiet, green suburban streets (see chapter 3) to the township of Alexandra takes five minutes, but it feels like a drive into a different world. Arriving from Linbro Park one may first drive past River Park and through Far East Bank, new sections of Alexandra built since the 1990s. These were constructed by the state with the intention of providing public housing for Alexandra's poor. The older the planned RDP housing sections are, the more the streets, houses and open spaces show signs of appropriation, adaptation, refurbishing and added constructions. Far East Bank resembles townships all over the country with its typical aesthetics of rows of identical houses. Once one has driven past the middle-class section of East Bank and crossed the Jukskei River, one enters the old sections of Alexandra. Old Alexandra, with its grid layout designed at the beginning of the 20th century, hosts over 85 per cent of Alexandra's households (Alexandra Renewal Project 2005). 'Old Alex', as people call it, has a diverse typology of housing, ranging from old bond houses constructed over 100 years ago to newer houses, as well as many, many shacks. In Old Alex, apart from streets and yards, there is hardly any space unoccupied by housing.

Because of the lack of open spaces and the high density, Alexandra's streets serve many different purposes, which shift with the rhythms of the day and the week. There are informal market stands, street corners where men gamble, residents wash their cars and boys play soccer. While yards are spaces of femininity and children, streets tend to be spaces of masculinity. On specific nights, shebeens may fill up and a whole street may transform into a party venue. When someone dies and the family's yard is not large enough, funeral tents may be put up on the street. In the newer section of Alexandra, streets are even used as spaces for political meetings. Pedestrians usually walk on the street and only move slowly to the sidewalk when a car wants to get through. Car drivers, hence, have to adjust their pace to the many users of the street, which is a social order distinct from Johannesburg's streets outside the township where cars dominate.

Alexandra is anything but a relaxed, quiet place. Rather, it vibrates day and night. It is not a place where urban dwellers let their guard down in public, but people are used to expecting things to happen that impede the f low of everyday life. The eruption of political violence, fights between neighbours, accidents, rape and being mugged on the street are still extraordinary, but people have a sense that this may happen at any time. People who have lived in the township a long time or who have grown up here are affected by this in the same way as newcomers who have moved here recently, but long-term residents have learnt how to move through the township spaces with an air of self-assurance in order to appear 'streetwise' and 'smart'.

In the densely populated streets of Alexandra, the Metro Police, who are responsible for the enforcement of by-laws and traffic rules, are seldom seen and drivers generally prefer not to put on seatbelts. During fieldwork I even was told that I should take my seatbelt off when I drove inside the township, as this would show everyone that I was familiar with Alexandra and hence *streetwise.* "If you are not *streetwise*, you get teased at school and called a 'cheese boy'" (Thabo Mopasi, March 2011). Being called a 'cheese boy' is an insult in the township, meaning a 'weak' man, an expression ref lecting notions of class difference and ideas of hegemonic masculinity. A boy who gets

cheese sandwiches from his mother for lunch at school – an expensive meal which many township families cannot afford – does not conform to expectations of masculinity like toughness, strength and street wisdom. Being *streetwise* is an important urban competence in Alexandra. *Street wisdom* locally means a specific type of urban knowledge and competence that helps urban dwellers to move through the dangerous and confusing streets and footpaths of Alexandra. Having a precise mental map of the neighbourhood as a geographical and social space, that is, being able to recognise potentially dangerous places, people and times of day, assists urban dwellers to move around safely, while also paying careful attention to what is happening around them.<sup>8</sup> I was also told, for example, that I was not *streetwise* when I took time to look for exact coins in my purse when paying at the counter; instead, I should keep my money in my pocket, take out what I needed quickly, and put back the change as quickly as possible. Thabo repeatedly told me to wear a cap so that people could not see in which direction I was looking. A cap would hide my glasses, he explained, which also gave me away as an outsider, as wearing glasses was rare in the working-class township. Although I doubted that such behaviour, symbolising *street wisdom,* would make me, a white anthropologist in Alexandra, appear less of an outsider, Thabo's instructions made me aware of the importance of such bodily practices in Alexandra's public spaces.

Outsiders who come to Alexandra are scared of being recognised as outsiders on the street, which points to the performative, symbolic dimension of *street wisdom*: by moving securely through space, one indicates to potential criminals that one is not an easy victim. There is the saying that Alexandra residents can recognise whether someone is from Alexandra or not, which makes non-Alexandrians even more scared when coming to Alexandra.

There is a dif ference, but I don't know exactly what this dif ference is. Even for people who come from Tembisa or from Soweto, you can see that they are not from here. I don't know how, but you can see that these people are notstaying here. Maybe it'sthe movement or … (Manah, a 40-something neighbour of Thabo'sin-laws, May 2011).

Trying to explain what is difficult to explicate, Manah, a neighbour in Thabo's in-laws' yard, referred to the corporeal habitus, the way of moving through the township. She argued that she could distinguish whether a stranger was used to moving physically through Alexandra's environment simply on the basis of her visual perception of a body.

*Street wisdom* as a specific spatial, corporeal and social urban competence has to be learnt through practice and habit, and it is hence not equally shared among members of different milieus in Alexandra. During fieldwork, I became close friends with Nnana, who was a member of the newcomer milieu, and I often moved in her company through the township. This allowed me to compare the experience with the many times I had been on the streets with Thabo, Thabo's children and many other long-term residents. They greeted someone on every street corner; they were familiar with all the different sections of the township and knew exactly how to get home from all of them. Nnana, on the other hand, who only spent a couple of weeks in the township every year and then returned to the rural area, was only familiar with River Park, the

<sup>8</sup>  Based on hisresearch in a black neighbourhood inWashington,the anthropologistUlfHannerz comes to very similar formulations of the term 'streetwise' (Hannerz 1981: 26)

section where her mother rented a shack. Once, as we drove through the streets of Old Alexandra on our way to the San Kopano community centre, she told me that she had never been to this section of Alexandra before, and stared at the landscape of shacks, cars and people with a combination of fear and fascination. "I never had to go, and I always heard how dirty and dangerous it is" (Nnana, 23-year-old field assistant, resident of River Park, June 2012). As the taxi from River Park could take her directly to the clinic and shops of Alexandra, she had been able to avoid the streets of old Alexandra of which she felt scared. Like many other newcomers she had incorporated outsiders' images of Alexandra as dangerous.

In Lynch's seminal research on urbanites' mental maps, he defined *imageability* as a quality of the built environment which enables urban dwellers to imagine the material city as a mental map and which helps them to move through it with a sense of emotional security (Lynch 1960). Lynch showed that way finding is not a mystic 'instinct', but the result of the consistent use and organisation of sensory cues from the external environment (ibid: 4). What he does not talk about is that there are large differences between milieus in Alexandra between long-term residents and newcomers, in their ability to create mental maps because they use the urban spaces in widely different ways. *Imageability* hence also needs to be defined as the ability of urban dwellers to create a more or less defined mental map of the city as a physical and social space, based on their urban knowledge, experiences and urban competences. The ability to produce a mental map of Alexandra, as well as *street wisdom*, are specifically urban forms of capital which shape the way secure and safe urbanites move through the township, which spaces they go to or avoid, and also the place they hold in the neighbourhood social hierarchy.

Such milieu differences between long-term residents and newcomers become articulated as political categories in Alexandra and are often used to legitimise and delegitimise claims to access to limited urban resources like housing or electricity. The drawing of such political boundaries has shifted historically and situationally, but they tend to revolve around the landlord–tenant relations, distinctions between informal and formal (or legal and illegal), as well as insider–outsider distinctions based on national citizenship and shifting notions of urban citizenship. In the 1950s, Alexandra was declared an area for Natives (Blacks) and in the 1960s, people classified as *Coloured* were forcibly removed (Bonner and Nieftagodien 2008: 178). Not all Black urban dwellers, though, had the right to stay in Alexandra; they were categorised by the state into those considered legitimate urban residents who had legal access to the urban labour and housing market and those considered 'illegal' urban residents whose access to jobs and land was denied (Ferguson 2013: 229). The 1952 Abolition of Passes and Documents Act granted people who were already living in Alexandra rights of permanent residence, while newcomers were excluded from this urban citizenship (Bonner and Nieftagodien 2008: 105). In the 1980s, the Alexandra Master Plan (1980), part of the apartheid regime's carrot-and-stick policy, aimed to upgrade the township and provide housing, but only for 'legal residents', the so-called *bona fides* (ibid: 237). The Burger Alexandra Urban Renewal Plan from 1986 wanted to allocate houses only to those who had arrived before 1986, while those who arrived afterwards were excluded. In the post-apartheid era, when the official racial bifurcation of citizenship rights was abolished, new state categorisations continued the shifting insider–outsider divisions. In 2006, the ARP, on the insistence of the Alexandra Development Forum, shifted housing allocation

from following the housing list, from which mainly old residents would have benefitted, to a block-by-block approach which prioritised shack dwellers (Sinwell 2010). This caused considerable animosity among long-term residents towards shack dwellers who were perceived as 'jumping the queue', meaning receiving an RPD house earlier than others who had been waiting longer. Apart from shack dwellers, foreigners have become the new outsiders in the post-apartheid city (Nieftagodien 2008, Worby, Hassim and Kupe 2008: 26).

Johannesburg's winter days can get quite cold. The use of electric stoves or cooking plates for heating rented rooms and shacks, which are often illegally connected to the city's power network, regularly overburdens the system, leading to power outages. On a cold winter's day in 2012, property owners in River Park called for a public meeting on a neighbourhood street. Electricity had been out the whole day and some property owners blamed landlords who had erected illegal shacks in their properties for rental for overburdening the electricity system with their many tenants. The leader of the River Park Community Police Forum (CPF) had called the meeting by blowing a whistle, an important means of communication in the township in order to call neighbours to collective action. Landlords and tenants came out of their houses and shacks to the street corner; most of them were women and already in their pyjamas and nightgowns.

A group of women, all shack dwellers and tenants, remained on the other side of the road from the gathering, observing the crowd, and only slowly going closer. One of them explained to me: "The people are angry; they are complaining about us, the shack dwellers, and blaming us for the power outage." Those property owners who did not have shacks in their yard suggested angrily that they would cut off electricity forcefully from all the yards with shacks. One of the shack dwellers responded: "We pay rent, we have the same rights as the others." After a lengthy heated debate, the leader of the River Park CPF called for calm and set the rules for the following meeting:

The shack people are threatening the landlords. This decision has to be taken by the landlords, and therefore the people from the shacksshould leave! Everybodywho does not own a house hasto leave (Leader of the River Park CPF, June 2012).

The shack dwellers accepted this angrily and retreated to the other side of the street where they remained observing the gathering. Through networked infrastructure, urbanites "become inevitably bound up with one another" (Kirsch 2005: 208). Regular power outages in areas like River Park in Alexandra are related to the high population density, the many illegal connections and the insufficient investment in electricity infrastructure by the municipality. As the philosopher Iris Young rightly argues, in the distributive paradigm typical of a social welfare state like South Africa, social justice becomes defined as an end-state pattern of distribution, while the unjust institutional conditions which determine the distribution move out of the political debate (Young 1990: 15-37). Instead of directing their anger towards the City of Johannesburg and criticising the institutional context which, even in the post-apartheid context, still produces such unequal distribution of services, the political debate on the neighbourhood street revolved around the allocation of public services among competing social interests. It blamed those who were perceived to have less legitimate access the scarce resources – shack dwellers and foreigners.

By appearing in public something or someone becomes real, Arendt wrote (Arendt 1959 [1958]). In the public many perspectives on the world appear together, "allowing the reality of the world to appear truly" (Madanipour 2003: 168). On this evening, the street corner became temporarily transformed into a space of public debate; temporarily, architectural public space and public sphere came to overlap, very like in the Greek agora (Arendt 1959 [1958]). The notion of the public sphere, the political dimension of the public as a communicative, non-material space in which citizens come together and discuss 'public' issues according to 'rational' standards has been famously coined by Habermas (2002). The liberal public sphere as imagined by Habermas is open to everyone, there are no status hierarchies, and private people come together to discuss 'public' topics which leads to the 'public' opinion (Fraser 2001 [1997]: 112-113). The normative ideals of public spaces and public spheres developed by Arendt and Habermas shape academic discourses in urban studies, yet they need to be criticised. The historical Greek agora was not all-inclusive; it excluded women, slaves and other people who did not have Greek citizenship (Benhabib 1995 [1992]: 124). In 19th century Europe, citizens who did not possess property were excluded from the liberal bourgeois public sphere (Fraser 2001 [1997]: 123-125). These discursive arenas were embedded in the societal inequalities that inf luenced the social positions of actors within them, even though actors may have pretended that status differences were irrelevant (ibid: 123- 125). Also, Habermas' liberal bourgeois public sphere was "structured around significant exclusions" (Crawford 1995: 4, see also Fraser 1994 [1989], 2001 [1997]).

Further, the debate around electricity in River Park was deeply shaped by processes of exclusion, ref lecting inequalities in political power, access to housing, citizenship rights and economic resources in the neighbourhood. Because some of the River Park residents were tenants living in shacks, they were excluded from the decision-making process. Their access to the debate became conditional on property ownership and access to housing. Roy (2003) conceptualises such links between property ownership and political citizenship as 'propertied citizenship'. That is to say, "the right to the city is expressed through home ownership" (ibid: 85). Many property owners in Alexandra have the sense that owning property gives them a special right to inf luence the future of their neighbourhood, at the expense of those who do not have property (for a comparison with propertied citizenship in Linbro Park, see Heer 2018).

A further, in this case non-articulated, implicitly present social category fuelling the exclusion of tenants from this political space is national citizenship; many shack dwellers in the township stem from other African countries and are therefore considered less legitimate residents with less legitimate claims in struggles over scarce resources. While the shack dwellers were observing the meeting from the other side of the road, they felt anger and outrage about their exclusion, but they also silently expressed a sense of threat and fear, based on their past experiences with mob violence in River Park directed against shack dwellers and foreigners. They worried that a mob could indeed cut their electricity cables and turn their anger towards them. The shack dwellers eventually retreated into their yards, electricity returned the next day, and emotions quieted down – at least till the next power outage. This singular event is not simply an anecdote, but an urban situation involving the construction of social boundaries which is typical for River Park, for Alexandra and for politics in post-apartheid South Africa.

### **Civic Spaces and Routes of Becoming**

During the height of the mining industry and racial Fordism, which is how this phase of apartheid has been described (Crankshaw 2008, Gelb 1991), the economy of South Africa and Johannesburg was marked by labour scarcity, which is why migration from rural areas, the larger Southern African region and other continents was key to its growth. In the late apartheid period, however, South Africa radically transformed into a post-Fordist economy with a growing service sector and a demand for skilled labour, which led to a surplus of unskilled labour. In Southern Africa, this economic shift had a radical impact on the region, some argue that it is the second largest change after the macro-historical rupture brought by industrialisation and the rise of capitalism (Ferguson 2013: 230, Seekings and Nattrass 2005).

Getting access to the labour market and constructing a life considered respectable despite unemployment constitute key struggles in the everyday life of most Alexandra residents. A considerable part of the population is permanently locked out of the formal job market and make their living by informal means. A large proportion of Alexandra's households depend on social grants. According to the Benchmark Survey conducted by the ARP in 2005, 80 per cent of Alexandra's households were surviving with less than R3000 a month. The unemployment rate was quoted at 60 per cent in this survey. In the social perception of the residents, the idleness related to unemployment leads to what they see as social ills, like drug abuse, crime and domestic violence.

One way in which Alexandra residents deal with the void and gaps created by a lack of formal employment is through the creation of diverse forms of civic spaces. Civic spaces, how I understand them here, are collective spaces built and existing outside of kin, neighbour and work relations. They offer alternative forms of socialising and subjectification to those offered by the yards, streets, nightlife spaces, shopping malls and churches. The spaces are civic in the sense that their purpose is usually to create a society considered morally good; these spaces are seen as having a social purpose, like assisting the youth to become social beings outside drugs and crime or motivating urban dwellers to become engaged not just for their personal benefit but for the collective good. Hence, these civic spaces contribute substantially to the political, economic and socio-cultural life of the township.

By participating in and co-constituting these spaces, urban dwellers can achieve social respect which may be denied to them in other realms. The resources they build up in these civic spaces – social capital, practical experiences, inf luence and respect – may assist them to become more successful in other fields of life. Amongst many other things, these civic spaces can be seen as a key form of self-help, not only because their purpose is often to 'help' someone or to 'educate', but because they create alternative forms of being and of being together in a city characterised by inequality and exclusion.

Civic spaces in Alexandra differ in their form of institutionalisation, their degree of materialisation and their manifold social meanings and purposes, ranging from institutionalised, materialised spaces like community centres (San Kopano community centre) to imagined spaces of communal action (Tourism Hub, food garden project) to ephemeral youth spaces (kwassa kwassa dance group). Alexandra also has a long history of community self-help. Over the decades, residents have created many welfare projects out of meagre resources in order to assist community members (Bonner and Nieftagodien 2008: 314-320). Although many projects emerged out of Alexandra itself

and many are totally dependent on themselves, larger initiatives in particular depend on entanglements – on financial and other support from outside the township. In the 1980s, South African municipalities and businesses increased investments in community projects in townships as they hoped to undermine the increasing resistance against apartheid.

One lasting institution from that time is the *Alexsan Kopano* (*San Kopano* for short) community centre on 12th Avenue in old Alexandra. The red-brick centre with a large inner courtyard was constructed in the 1980s by a successful collaboration between township activists and aff luent White residents from neighbouring suburbs. They assisted the township dwellers to obtain funding from the development arm of the Protestant churches in *Germany (Evangelischer Entwicklungsdienst, see* Wilson 2003*)*. Nowadays it is administered by a trust; a resident of Linbro Park is also on the board.

The 'multipurpose' centre hosts a library, offices of local NGOs, the local chamber of commerce, the pro-bono branch of a renowned South African law firm, a large community hall and many rooms used for political meetings, church services, functions, training and the like. As an absolute space, by which David Harvey (2006: 133) means the view on space as a fixed geographical space, San Kopano has clearly discernible spatial boundaries: a gate which is locked at night, doors which separate the offices from each other. But, as a relational space, the space created by relationships between objects or actors, it stretches much further. San Kopano is a junction, an intersection of manifold relational spaces. Here residents can access the networks of established township residents who own businesses and who are engaged in ANC or SANCO (South African National Civic Association) networks, which intersect with local NGOs, government branches and other institutions.

San Kopano functions as a hub and springboard for many Alexandra residents, as a space where they can access diverse resources ranging from social networks to books, computers, food parcels, toilets, and information of all sorts. Here young people can attend courses on computer skills or do an internship. Many also come to access computers to compile their CV, to find out how to apply for a university scholarship, or how to raise funds for a project. When visitors at San Kopano meet unknown others, they exchange phone numbers, business cards and information about each other's employment or other connections. Conversations often reminded me of network events like academic conferences.

One day I met a young woman in the bathroom at San Kopano. She was living in the nearby shack settlement and preferred the toilets at San Kopano to the public ablutions. She had just moved to Alexandra from the Eastern Cape to join her sister who had moved there earlier. At the centre she came in contact with the ANC Youth League: "I like to work for community and help people. So I joined the party", she explained to me (Themba, a 24-year-old female shack dweller and ANC youth league volunteer, March 2011). As a 'volunteer' of the ANC Youth League, she was attending political meetings and participating in activities like cleaning up streets and clinics, visiting the elderly and the disabled and organising food donations for them. As a newcomer to the city, participating in the youth league helped her to make friends and get variation into her monotonous life as a young unemployed woman. She included the volunteer experience in her CV, hoping to increase her chances of getting a job.

So the people, organisations and institutions that urban dwellers find at the centre assist them to circulate, namely "to navigate and engage diverse spaces, actors, sensi-

bilities and activities" (Simone 2005b: 519). By sharing all kinds of resources, by creating connections, by generating ideas and projects, the politicians, NGO workers, government officials, volunteers, activists, school children and ordinary urban dwellers who go in and out San Kopano come to be 'people as infrastructure' (Simone 2004b). People as infrastructure do not replace other forms of structures like the state, the market or urban infrastructure, yet they create connections among urban dwellers as well as between urban dwellers and various pools of resources. San Kopano can be understood as an 'assemblage'. Assemblage is a notion that emphasises the processual, undetermined, relational and generative nature of city making, the "eventful, disruptive, atmospheric, and random juxtapositions that characterise urban space" (McFarlane 2011: 651). My field assistant Thabo, for example, regularly assists others to draw up a CV for a job application or get an appointment for a free consultation at the lawyer's office, or he helps journalists to find a good story. Through his and many others' capacity of assembling, connecting, merging and crossing, he is capable of "generating social compositions across a range of singular capacities and needs" (Simone 2004b: 410-11).

One should not forget, however, that networks and assemblages are shaped by inequality and processes of inclusion and exclusion: every network has cuts (Strathern 1996). The assemblage is not equally accessible to everyone: Newcomers, especially those living in more distant sections of the township, may not know about its possibilities and resources. Nnana from River Park, for example, only got to know San Kopano when she accompanied me to visit Thabo.

The most common connections created at San Kopano interlink politically active actors and established milieus in Alexandra and link them 'upwards' with government departments and politicians. There are entanglements across the township boundaries, for example to NGOs and activists from other townships, to companies in Sandton, and to actors from abroad. Nevertheless, they hardly reach, nearby suburbs. There are hardly any relations with civic organisations, politicians and ordinary urban dwellers from the formerly White-zoned and now desegregating suburbs. The apartheid legacy, the spatial heritage of racial segregation and economic inequality, thus still narrows the reach of networks and collaborations at San Kopano.

San Kopano is an example of a civic space which has the form of a firmly institutionalised, architectural space, yet civic spaces in Alexandra can also exist as social spaces in the making, as not-fully-existing, imagined spaces, which nevertheless direct collective action towards a shared goal. In 2010, a group of about twelve to fifteen mostly unemployed men in their twenties created a garden, the Gcwala s'Bahle Garden Project, in a new section of Alexandra, the Far East Bank. They constructed the garden on an empty piece of land which used to serve as a scrapyard. When I met them, they were dreaming about putting up a large tent across the garden, and they hoped to construct a stage where they and other youth groups could hold theatre productions. They imagined that political meetings which were taking place on neighbourhood streets could be moved into the garden. Next to the garden they established a car wash on empty land from which they earned some money. Like many young people in Alex, they were engaged in diverse activities like volunteering for community patrols with the police and participation in a government work programme, as well as theatre, music and dance. Spaces to practice such activities were scarce, and they could hardly

afford to rent a room at a community centre like San Kopano. Hence, they dreamed of turning their garden into their own community centre.

Occupation of spaces which are not occupied by others – like a scrapyard – is hence an important strategy for youth groups in Alexandra to achieve control over space. Justine Lucas and other researchers have shown that political organisations in Alexandra establish political power and legitimacy by seeking to control access to space, especially housing (Lucas 1995). What has been less documented, though, is that seeking control over imagined civic spaces is an important route to achieving legitimacy; a route often sought by young urban dwellers. A similar example is a building called Tourism Hub, which was occupied by a youth group called Alpoa Youth from 2011 to 2013. Alpoa Youth is the youth branch of the organisation Alpoa, which aims to represent the interests of the former property-owning families. The Tourism Hub is a building on the south-eastern border of the Jukskei River containing a few rooms, a half-open hall and a stage surrounded by stone tables and benches. The construction of the Tourism Hub was funded by wealthy individuals and it was intended to serve as a community centre. However, there were ambiguities about which civic or government entity was entitled to administer it, which is why it stood empty for a long time; a 'white elephant' as it was referred to locally. In 2011, the ARP was formally in charge of it, and organised private security personnel to look after the building, while it was still planning how to use it.

In 2011, the then secretary of Alpoa Youth was living next to the Tourism Hub in an area called the 'TB Settlement' because there had once been a tuberculosis clinic there. He obtained the keys from the security personnel and for about two years, Alpoa Youth turned itself into the custodian of the space, claiming it to be a "space for the youth of Alexandra" (Patrick Baloyi, secretary of Alpoa Youth, March 2011). The ARP did not approve of the 'white elephant' becoming the headquarters of the youth group, yet it did not take action against them. Alpoa Youth started to organise music events for the youth (for example the 'land restoration hip hop session') and held all their meetings there. Some of the members spent their days at the Hub, giving them some everyday routine and a reason to leave their cramped homes. Here they socialised with their friends and strategised about activities. The secretary of Alpoa Youth turned one of the rooms into his private fitness studio where he used to lift weights. Alpoa Youth also rented the large facility out to churches and for private parties, and with the money they paid the phone and internet bills and other expenses.

In 2012, a conf lict emerged with the residents' association of the adjoining neighbourhood section, the TB Settlement Committee. The Committee criticised the youth group for enriching itself by renting out the facilities, wanting to gain control of the space itself. In a conversation, one of the exponents of the committee, Kwanele, explained their plans for the place:

We want to apply for government funds, so that we can employ people from the community ascleaners and receptionists.Ican come here every day and look af terthe place, but I need to earn something, I need a salary (Kwanele, a 30-something member of the TB settlementcommittee, June 2012).

This statement makes clear that the conf lict around the Tourism Hub was not least about the possibility of turning it into a source of employment, showing how control over space is an important strategy for economic survival in the township. It also had a political dimension, however; at the same time, Alpoa Youth launched a petition to disband the ARP, strongly criticising it for having failed to develop the township. By appropriating a building that was the responsibility of the urban renewal project, the ARP, Alpoa Youth aimed to challenge the legitimacy of ARP as the development actor in the township, and they hence suspected that the ARP had told the TB settlement committee to challenge theirs. Control over space in Alexandra is always a quest for power.

In different yards in the township, towards five or six o'clock in the evenings, young people set up music boxes and a loud beat starts to fill the early evening air. In these yards, young people study fast, exhaustive dance choreographies which they called *kwassa kwassa* at that time. In 2011, there were seven or eight *kwassa kwassa* groups in Alexandra according to the 27-year-old leader of ABCDE (leader the of *kwassa kwassa*  group, March 2011). In the early evenings, green fields in Jukskei Park and semi-private spaces like yards, sometimes only a few metres square, become transformed into stages where young people train for theatre productions, music performances or dance competitions. These self-organised and mostly underfunded youth groups constitute part of the diverse, large landscape of arts and sports in Alex, compensating partly for the lack of such facilities in the badly equipped public schools. Children and youth between eight and about twenty years old participate in these spaces of sociability and physical exercise.

Many of these youth groups are founded and organised by older youth in their 20s and 30s. A coach of one such *kwassa kwassa* group had succeeded to getting a good job outside Alexandra and wanted to 'give something back', as he explained to me (March 2011). In Alexandra, such youth spaces are usually represented by young people and adults alike as spaces of activity, where the youth are kept busy. Idleness, according to the township residents, seduces the youth to become involved in socially and morally wrong activities like crime and drug abuse. "You have to be creative; you have to do different things to survive here. You have to keep yourself busy" (a 25-year-old member of Gcwala s'Bahle Food Garden Project, February 2011).

Unemployment in South Africa is the highest among the youth. Growing up in Alexandra and living there as a young person is anything but easy. When young people leave school, they enter a social void. There are scholarships available for further education, but one needs the cultural capital of knowing how to apply for them and also high grades, which are difficult to achieve in Alexandra's schools notorious for their low matric pass rates. Simone argues that: "Without any prospect of employment, there is no platform to signal progression from youth to adulthood, like likelihood of viable social reproduction – of family, cultural value, memory" (Simone 2005a: 323, see also Honwana 2012).

Creating and participating in youth spaces of sociability and physical exercise is therefore an important route for young people to avoid idleness, to 'keep busy'. While some youth may spend their free time at the church or in such civic youth spaces, others again party. Alexandra has always had a very vibrant nightlife. The many different places to hangout socially invite one to spend the nights dancing and the days recovering from *babalaza* (hangover). If one lacks money to buy beer in these public places, one can get cheap and easy access to all kinds of drugs and consume them with friends at the home of somebody who has a shack or a room. The boundary between joyful partying, drug abuse and drug addiction, however, is a slippery one. Many young men in their thirties whom I met during fieldwork told me about a phase in their life (mostly

the years after school) in which they became heavily involved in drugs and crime. The desire for consumer goods felt by themselves and their girlfriends, anger at Whites in the context of crumbling apartheid and the search for friendship, parties and fun were among the reasons why young men chose to become involved with criminal gangs. They eventually left these gangs either because they were arrested or because some of the gang members were shot, and very often both.

Civic spaces for the youth like the *kwassa kwassa* dance classes are often talked about in the township within frames like morality and health. In contrast to the stigmatised teenage pregnancy and the 'bad' drug and crime lifestyle, activities like the *kwassa kwassa*, or the Alpoa Youth group or the food garden project were portrayed by the young people as morally good spaces that were supposed to protect the participants from the 'bad' out there. They presented them as places where they were safe from 'bad' inf luences and could learn from the older ones who had survived their difficult life phases. In contrast to the 'rational leisure' described by Burgess (2005), for example boys scout, which is organised from above (the powerful, the adults) in order to educate youth to show discipline and respect, the youth groups in Alexandra are organised bottom up, by the young people themselves or by young adults in their thirties.

Not least because of this perception of civic youth spaces as 'morally good' and educative, the *kwassa kwassa* groups receive high acceptance in the neighbourhood. The yard neighbours where the youth train tend put up with the loud beat of the music and many watch the training sessions. Some neighbours explain that they accept the youth groups in their yards because, for example, "it prevents things", as one neighbour told me (neighbour *kwassa kwassa* training ground, March 2011). In lived reality, though, boundaries between 'good' sports and arts and 'bad' drug lifestyles are not so clear. Sometimes participants of a *kwassa kwassa* group smoke *dagga* (marihuana) in a hidden corner of the yard before they show up for the training session. On Sundays and Mondays, the young men of the Alpoa Youth and the food garden project were hard to find, as they were usually recovering from a *babalaza* (hangover) from the weekend. Nevertheless, the youth spaces which they had created for themselves provided them with some structure to their days and weeks, so that they had a reason to stop partying when the weekend was over. These spaces, which are considered socially legitimate in the township, provided spaces of sociability for the youth as well as routes for these young people to become someone socially.

### **Alex at Night: Of Conspicuous Consumption and Nostalgia**

Alexandra is home to many different milieus and spaces, and it is also marked by different rhythms across the day and the week. Urban dwellers are always on the move, and their daily lives produce urban rhythms (Hahn 2012: 18). In early mornings and at dawn commuters fill the streets of Alexandra, recognisable by their work clothes, either the business attire of those who work in the service industry or, typically, the blue overalls of those who are on their way to a job in industry or a suburban house. During the day and after rush hour, urban dwellers transform the streets from spaces of transport to spaces of multiple practices like hanging around, playing cards, selling things, socialising, observing others or playing soccer. Friday is the busiest and most hectic day, and, as many get their salary on Friday, people say that Friday is the

day of heavy drinking, fighting, domestic violence and assaults on workers returning home with the salary in their pocket. During the day on Sundays, the streets are very quiet except for people on their way to church and the sound of the many church choirs chanting in small houses and garages. Sunday night, though, is the big party night in the township, followed by Monday, the day of *babalaza* (hangover).

Alexandra has a lively and diversified nightlife culture. Differences in age, income and lifestyle preferences create the demand necessary to sustain the diversity of simple shebeens/taverns, *shisanyamas* (places where meat is barbequed) as well as sophisticated bars and nightclubs. Nightlife spaces in Alexandra have a distinct weekly rhythm: many of the drinking places where young people socialise are open every day, but they only fill up on one specific day of the week. On Sunday, young people assemble at Joe's Butchery, on Monday, one goes out at Chicks, on one of the other nights one can go to the Heritage corner and on Friday, if one has enough cash, one can go clubbing at the elitist, luxurious club *Neh!* in nearby Marlboro.

Gender, age and income are important determinants for access to such leisure spaces. Married women, people who lead a religious lifestyle and avoid alcohol, and poor people, who do not have wealthy friends or lovers, seldom frequent such places. While young men are often not subject to much parental control in the township, young women told me that they had to sneak out of their parents' houses at night to go clubbing. Places like Chicks and Joe's Butchery, where the party takes place on the street, are infamous for violent fights and shoot-outs. Also, gang leaders are said to hang out at such places. These places belong to the key spaces in Alexandra where entanglements across spaces and lifeworlds become formed; where otherwise separate lives become enmeshed in one place, confronted with each other.

On Monday nights, the southern section of 15th Avenue is transformed into a night space called, amongst others, 'Chicks'. Youth and young professionals between 20 and 40 years old from diverse townships and suburban areas mingle here, drink beer or cider, chat, observe and enjoy being observed. The heavy smell of tripe fills the night air and the atmosphere is marked by the humming beats of deep house, an electronic music style which at the time of research had largely replaced kwaito as the most popular genre in the township hang out places. The name 'Chicks' stems from the nickname of the owner. The owner of the place started it about 10 years earlier by selling *mogudu* (tripe), a favoured township delicacy, and young people used to sit on the steps to eat it. With time, the social gatherings became larger and by 2012 had turned into one of the most famous party hangouts with a reputation across the entire city.

Professionals and managers working in Sandton offices bump into each other here. Going out clubbing in townships is a fashionable trend among the black middle class in the city. Those unfamiliar with township spaces can be recognised by their lack of *street wisdom*, e.g. women who handle their handbags and cell phones carelessly, or who wear high heels, unsuited to the rough township street. At the beginning of the night, before the street is closed off with fences so that customers cannot bring alcohol from outside, cars make their way through the crowd. Nobody takes 15th Avenue on Monday nights accidentally; rather they do it to show off their cars. The latest German cars, their drivers sometimes holding whisky glasses in their hand, and even large motorbikes are showcased here. People say that sometimes one can even see Lamborghinis and Porsches. Places like Chicks are understood by Alexandra residents as spaces where those who moved out of the township come back in order to show off their wealth.

Many people moved out of Alexandra to the suburbs, but they continue coming here. People who move away miss Alex. They become nostalgic. And they come here to show of f (Thabo Mopasi, a 40-year-old member of long-term tenant milieu and field assistant, February 2011).

Veblen called such practices *conspicuous consumption* (Veblen 2000 [1899]), meaning the consumption of luxury goods and participation in luxury leisure activities in order to display social status and wealth. Chicks is not only about display, though, it is also about the complex politics of redistribution, the creation of patrimonial ties and transactional sex: wealthy men buy drinks for their less aff luent friends, relatives, employees or potential girlfriends, awaiting favours in return.

"Johannesburg is about money", I was repeatedly told during fieldwork, as did a friend working as an assistant at a large South African company in Rosebank (Gaby, my 'host mother', resident of Orange Grove, February 2011). When I was staying at Gaby's house for a couple of months, I was driving a VW golf from the 1980s in a particularly ugly yellowish colour; friends repeatedly told me that while it was wise to drive such a car in a city prone to car hijacking, it was also uncool. "Your car is a statement" Gaby repeatedly told me. Many middle-class people spend their income on branded clothing and the latest German car, mostly on credit. Some people say that Alexandra residents who managed to get a middle-class job continue to live in a shack on purpose, so that they can spend their money on the monthly instalments for their car. "Everything is about being shiny and glittery. People don't buy bags or clothes where you don't see the brand. Everything needs to have a brand" (Gaby, February 2011). For Gaby, this also meant significant social pressure. She had decided against brands and a shiny house as she preferred not to get herself into debt. "But most people don't think like that here, Johannesburg lives on credit." She said repeatedly that she didn't have many friends because of her choices. The social pressure to consume, and the feeling that one becomes socially excluded and does not fully belong if one does not participate in the consumption hysteria is typical of people like Gaby who want to be part of the middle-class suburban lifestyle, but whose salaries of about R20,000 before tax (2012) are too low to get a credit for all the status symbols of a middle-class lifestyle, namely a house, a car, private schooling, brand clothing and going out to fancy places. Debt is then a considerable problem for many people.

The club and restaurant *Neh*! are much more chic than Chicks. Its interior design looks like any nightclub in Johannesburg's northern suburbs. Its location in Marlboro (a former industrial area next to Alexandra) and the stunning views onto Alexandra's skyline, however, ground it in the township. On Sundays, *Neh*! becomes a meeting place for established, aff luent men in their forties and fifties. They grew up in the townships under apartheid and managed to get into very high-level positions thereafter; they belong to a generation who, after the abolishment of racist laws and Bantu education, profited from affirmative action policies (Employment Equity Act, 1998, Modisha 2008, Seekings and Nattrass 2005: 308-313). The term 'black diamonds' was introduced by the marketing industry to describe this black middle class (Southall 2016). These top manager at large South African companies spend their Sunday afternoons on the golf course, and then come to *Neh!* in the evening for a beer, sitting on the veranda overlooking the sea of shacks and cramped houses that is Alexandra. They have replaced their office suits with polo shorts and golf shorts. One of them told me

that as a child he used to walk past the large houses in the suburbs and wonder: "Why do we not have this?" (Jabulani, a 45-year-old former Alexandra resident, March 2011). Now he lived there himself. After saying that to me, he called the waiter and complained that the table cloth was damp.

At club *Neh!* Alexandra becomes the object of a nostalgic gaze by those who left this lifeworld behind. The Swiss doctor Johannes Hoffer defined the term 'nostalgia' in 1688 as meaning the "longing for a home that no longer exists or has never existed" (Dlamini 2010: 15). The nostalgia celebrated at *Neh!* constitutes what the journalist and novelist Jacob Dlamini calls 'ref lective nostalgia'; a nostalgia that is self-aware and combines critical ref lection with affective memory (ibid: 1, based on Boym 2001). What these successful men are nostalgic about are images of the township way of life to which they would not like to return, but which they nevertheless miss nostalgically, and which decisively shaped their sense of belonging.

### **Conclusion: Entangled Spaces, Entangled Lives**

With the transformation of Johannesburg from a Fordist to a post-Fordist city, and the development of edge cities, gated communities and shopping malls, there are striking similarities with other 'postmodern' cities like Los Angeles. It is not surprising, then, that theories developed in the Los Angeles school of postmodern urbanism have been carried over into writings about Johannesburg, so that the city has become framed theoretically as quite distinct from other African cities. Most prominently, Murray (2004) argued that Johannesburg is a 'late-developing', world-class status aspiring city which has become, like São Paulo, incorporated into global circuits of trade and investment. This has led, coupled with local dynamics, to the spatial outcomes typical of postmodern urbanism like socio-spatial fragmentation and polarisation. According to Murray the "spatial dynamics of postmodern urbanism have produced an urban landscape carved into fragments, disconnected 'micro-worlds' cut off from one another" (ibid: 142). While the urban landscape has come to "resemble the glitz and glamour of the 'first world' … the poor and marginalised are pushed aside, allowed to languish in poverty and destitution in impoverished ghettoes in the 'in-between' places of the city" (ibid: 158).

The aim of this chapter was to interrogate and unmake this representation. The ethnography of Alexandra deeply challenges the 'mosaic' vision of community and neighbourhood by highlighting the complexity of one space and its entanglements within and beyond. Indeed, life is harsh in Alexandra, which is also ref lected in what Lefebvre called the conceived dimension of space, namely, how urban dwellers, politicians and academics talk about the township, mostly as a 'problem to be solved'. This representation of the township as a problem is grounded in the everyday experience of lack of space, dirt and other oppressing structural conditions. Looking at Alexandra through this lens, outside observers might regard it as a 'ghetto', a socially and spatially excluded area that is home to poor, marginalised and stigmatised city dwellers (Hutchison and Haynes 2012, Varady 2005, Wacquant 2008). As in many neighbourhoods at the social margins across the globe, young people in Alexandra have developed a specific subculture, appropriating the discourse of marginality and transforming it into a popular counterculture. *Gomorrah* emphasises the agency of Alexandra's

residents; it claims that growing up in the *ikazi* makes the urban dwellers especially strong and streetwise instead of stigmatised.

Nevertheless, this is far from being the only way Alexandra is experienced and imagined. There is also the narrative of Alexandra as a space of urban creativity; a neighbourhood where the lack of state control during apartheid and after made space for innovative forms of sociality, culture and economics. In order to party, people from other townships and suburbs travel to Alexandra's bars and clubs, transforming it into a centre for Johannesburg's night life. As a former freehold township, Alexandra has given rise to an inf luential property owning middle class. Although this milieu was dismantled by the apartheid state, it nevertheless continues to have significant political power in the township. Calling Alexandra an 'in-between' place as Murray does implies that it is somehow 'outside' the proper city. Yet, in political terms in particular, Alexandra has actually been a political centre for the city and even the nation; it is the point of origin of many inf luential political movements, phenomena and actors. The discourse of Alexandra as a struggle monument draws attention to the fact that Alexandra was a central stage for the different phases of political resistance against the apartheid state. With pride, residents list the many Alexandra politicians who came to shape the provincial and even national politics; for example, Obed Bapela who, in 2010, became a member of Zuma's cabinet. Also, less glorious aspects of South African recent urban history are linked to Alexandra. The xenophobic attacks of 2008 started in Alexandra and then turned into region-wide upheaval resulting in tens of thousands of people being internally displaced. Alexandra is thus not only a spatial entity in the city, but also social imagery, constructed as multiple narratives and multiple worlds in a continuous process of social, cultural and political imagination and articulation.

Notions like ghettos and slum imply that the residents of an area are homogenously poor; they tend to be associated with the idea of an isomorphism of place and the social characteristics of the residents. Yet the population and everyday life in the township are highly differentiated, marked by considerable diversity and inequality. The constant inf lux of newcomers into Alexandra over the last century has led to very high residential mobility, high densities and tensions around scarce resources like housing and electricity. In these conf licts, milieu differences become articulated as social categories and are instrumentalised to legitimate and delegitimise claims to access to limited urban resources. The drawing of boundaries has shifted historically and situationally, but they tend to revolve around landlord–tenant relations, distinctions between informal and formal (or legal and illegal), as well as insider–outsider distinctions based on national citizenship as well as shifting notions of urban citizenship. Also, the xenophobic politics need to be understood within the historical context of ever-shifting boundaries of insider and outsider divisions.

Urban studies in Africa recognise a long trend of network approaches, starting with the Manchester scholars and including writers like Simone and the latest assemblage notions (McFarlane 2011). Network studies with their discourse of f luidity and interconnections tend to neglect the power of geographical space like spatial boundaries and questions of geographical accessibility, while privileging relational spaces of networks and connections. It has even been claimed that "African identities display a remarkable capacity not to need fixed places" (Gotz and Simone 2003: 125, quoted in Bank 2011: 16-7). Yet the emotional connection to land and the politics of land are highly constitutive for Alexandra. Everyday life is characterised by opposing tensions, namely, the oppressive structural conditions inscribed in space versus the agency and creativity of urban dwellers in resisting, appropriating and transcending them; the tension between the power of marginalisation and spatial boundaries versus the spatial and social mobility and f lexibility of the residents; the tension between the everyday experience of conf lict, violence, anonymity and social isolation versus cosmopolitanism, dense social networks and mutual help; the everyday experience of hardship and suffering versus the joy of dense sociality, of leisure and cultural production. Bank hence rightfully warns that we need to "be careful not to over-emphasise mobility and movement and the inability of urban Africans to become grounded in the cities and neighbourhoods within which they live" (Bank 2011: 16).

Alexandra is not a disconnected micro-world but is entangled with other places in manifold and complex ways, as the following chapters will develop in more detail. Alexandra has always been and still is tightly integrated and interdependent with the surrounding city. It is embedded in complex social, cultural and economic connections with the surrounding suburbs, as well as with other cities and faraway rural homes in South Africa and other countries (ibid: 14). Although this chapter has focused on spaces within Alexandra, it has shown such connections beyond its geographical boundaries. Economically, the township is deeply entangled with the surrounding suburbs and aff luent areas like Sandton. Important agents of connection are also the aspiring middle-class milieus that have moved out or aspire to move out of the township. In addition, many newcomer milieus connect the township through faraway places like rural homes in South Africa and elsewhere, by circulating between them. Places like the assemblage San Kopano are key spaces where residents from diverse milieus find resources that enable them to circulate across boundaries and where actors from outside like NGOs, academics and tourists become connected to the township. The legacy of the apartheid geography still limits these connections, however, which can be seen in the absence of links to political organisations in nearby suburbs. Entanglements further involve the most intimate spheres of life like hygiene and family relations.

Neighbourhoods hence need to be understood in relation to other urban areas and the city at large, and not in isolation (Castañeda 2012: 160). Alexandra only comes into being through connections, boundaries and interdependencies between residents of different milieus and neighbourhoods. To understand urbanism in cities characterised by segregation, there is therefore a need to study the "complicity, co-operation, boundary-crossing, interpenetration, affiliation and divergence which 'come and go' across the city, its neighbourhoods and its facets" (Simone 2001: 61). The different social worlds in the city are not unconnected, segregation is never complete. We need to empirically investigate the degree and quality of entanglements, instead of reproducing narratives of urban duality and dichotomies. This is what *cities of entanglements* aims to do.

©Barbara Heer, Alexandra, 2011

©Mark Lewis, River Park, 2015

©Barbara Heer, Alexandra, 2011

©Barbara Heer, Alexandra, 2011

## **Intimate Encounters?**  Domestic Work in Home Spaces

Johannesburg has for many years been portrayed as a city dominated by fear of crime (Bremner 1998, Dawson 2006, Dirsuweit 2002). Although crime rates have decreased in the last decade (Dirsuweit 2015), thinking and caring about security is part of everyday routine in the city, almost like eating and sleeping. As "the experience of fear and at times panic lies at the deepest roots of life in the metropolis" (Mbembe and Nuttall 2008: 24), the figure of the criminal takes centre stage in narratives about the city by writers and by ordinary residents, especially in the suburbs. In Linbro Park, a peri-urban suburb adjacent to Alexandra, the residents used to share information about security on an online platform (a google group) between 2010 and 2016. Residents for example reported when they observed a suspicious '*BM*' (black man) on their streets –he was suspicious to them because he was black and a stranger; not a gardener or a tenant.

Many famous writers about modern cities have argued that the stranger is the quintessential urban figure. Cities differ from rural areas because they own a 'public realm' where strangers interact (Lof land 1973). In postmodern cities like Los Angeles, São Paulo and Johannesburg, where deindustrialisation has increased social inequality and spatial segregation despite the end of discriminatory politics, the stranger becomes largely a threat, a criminal, and often also a racial other. The black stranger, usually male, is thought to come from the street, the public spaces, and threatens to intrude into the homes, the private and intimate spaces, which is why they become fortified with walls and electric fences by those who can afford to do so. In Johannesburg, both legal and illegal immigration have increased considerably with the dismantling of apartheid. Here, the criminal stranger has become intertwined with the migrant other, the foreigner.

In writing about African cities, scholars have emphasised the mobility of their residents; rural–urban and cross-border movements are important strategies for the economic survival of many African urbanites. Postmodern African city life demands "a daily reinvention of new pathways for living" (Nyamnjoh and Brudvig 2014: 353). Most famously, it was Simone who, in his poststructuralist writing, established the 'mobile African' as the quintessential urban figure (Simone 2004a). He portrays African cities as cities always in the making through the inventiveness of their citizens, whose assembling and disassembling connections create a constantly changing urban landscape.

Both the criminal stranger in Johannesburg's popular imaginaries and the Mobile African in Simone's tropes are male figures whose lives evolve mainly in public spaces like the streets and the markets. As fear of the criminal has made the privileged retreat behind walls, the male stranger, migrant, criminal and poor converge and make a living hustling on the public streets left to decay.

As the propertied, the privileged, and the powerful have retreated behind the protective shield of gated residential communities, enclosed shopping malls, and other barricaded sites of luxury, the urban poor are of ten lef t to fend for themselves in urban environmentsthat have deteriorated almost beyond repair(Myers andMurray 2006: 5).

This portrayal of Johannesburg and of African cities in general depicts only a particular aspect of urban reality, omitting many other dynamics which are just as relevant for analysing urbanity, like locality and routine, care work, intimacy and home spaces. Approaching African urbanity mainly through male hustlers on the streets and in the markets, like Simone tends to do (e.g. Simone 2006a), ignores the contradictions, ambivalences, interconnections and invisibilities characterising African urbanity; the entanglements between locality and mobility, between strangeness and intimacy, and between public and private.

Economic survival in the African city demands inventiveness and mobility from its inhabitants. Yet it also demands routine: caretakers, usually women, need to dress their children, feed them and send them to school every day. While life is highly unpredictable, these women struggle daily to bring stability and routine into their dependents' life, not least by cooking, washing and cleaning. Mobility and locality are usually intertwined: many domestic workers work in Johannesburg to ensure that their children at home in the rural areas can eat and go to school.

In this chapter, I therefore want to draw attention to an alternative quintessential urban figure – the female domestic worker. The domestic worker–employer relationship constitutes a habitual, quotidian and somehow invisible urban entanglement, connecting classes, races and spaces in intimate and lasting ways. Walls are thought to exert and symbolise exclusion and disconnection from the urban poor; yet in reality these boundaries are crossed daily by members of the 'othered' groups – the poor, the migrant, the black – as domestic workers and gardeners. In contemporary Johannesburg, black women commute daily from township to suburb to engage in cooking, cleaning and caring in the home spaces of the white and black middle classes and elites. Many even live in the suburbs as invisible residents. By listening to the stories and experiences about these deeply asymmetric, yet affective relationships, we learn most about the fundamental social, economic and political processes at work in cities of inequalities.1 Looking closely, these relationships reveal the way in which seemingly of

<sup>1</sup>  In order to protect the domestic workers' and employers' anonymity, I have changed some of the personal details. The domestic workers I interacted with tended to be female and long-term employees hence the analysis reflects less the views of short-term workers which could be dif ferent. The data from the perspective of employers stem mainly from informal conversations or topics brought up in interviewswhen talking about othertopics,while Iconducted systematicinterviewswith the domestic workers abouttheirrelationsto the property owners. Thischapterthen gives more room forthe voices of the domestic workersrather than those of the white Linbro Park residents. Although I aimed to represent the data ascritically as possible, the reader will notice my solidarity with the domestic workers and my sense of responsibility towardstheir repeatedly uttered wish that Ishould make their working conditions visible.

people's different life worlds, areas and social strata are actually intimately entangled, even as the employers, as we shall see in this chapter, often try to make these domestic workers invisible.

When talking about domestic work, we also have to talk about gender in the city. Homes and yards in Alexandra, Linbro Park and elsewhere are highly gendered spaces; deeply related to (European and African) images of femininity and masculinity which regard household chores and child rearing as women's work. In the realm of domestic work in Johannesburg, gender ideologies and practices of European and African origin converge and depend on each other. Aff luent women living in Linbro Park have achieved what could be called, in political terms, emancipation: they are usually employed or even have their own businesses. Feminist movements in the West have left a strong impact on gender relations among the urban, aff luent English-speaking milieus in Johannesburg, at least since the 1960s, according to the historian Rebecca Ginsburg (2011: 150-153). The economic emancipation of White suburban women, however, was intrinsically linked to the institution of domestic work and economic inequality; only with the help of Black domestic workers at home could White suburban women engage in paid employment.

In Xhosa gender ideologies, the association of care work with femininity is institutionalised in the role of the *umtshakazi*, a period during which a newly married wife performs all the household chores for the new mother-in-law, an institution which emphasises the obedience and deference of the wife (Hunter 1933: 264, quoted in Cock 1989 [1980]: 60). Xhosa gender ideology did not envision wage labour (*ukuphangela*) for women. The commodification of care work from unpaid to paid domestic work brought with it a loss of status. According to Mayer, Red and School, the Xhosa perceived domestic paid work as an unclean occupation (Douglas 2005 [1966], Mayer 1961: 245). According to the sociologist Jacklyn Cock, this is one of the reasons why paid domestic work is one of the least prestigious occupations in Southern African societies (Cock 1989 [1980]: 59-60).

Women can afford to outsource care work and household chores to other women when their labour can be bought relatively cheaply (Lutz 2005: 113). Therefore, domestic work is prevalent in highly unequal cities where salaries between the lower- and the middle- and upper- income levels diverge greatly. In 1989 Cock claimed that domestic work was the social space within which Black and White women most frequently encountered each other in apartheid society (Cock 1989 [1980]: 2). This also holds true for the post-apartheid period in Johannesburg, as well as for other contemporary cities. In many cities marked by inequality, domestic work is an important urban institution and a large industry. In Maputo, 6 per cent of the workforce (30,000) were employed in domestic work in 2001 (Jenkins 2000: 213). In Johannesburg, the rate is even higher, with 31 per cent of employed black women engaged in domestic work (census 2001, provided in Peberdy and Dinat 2005: 6).

A set of relations not typically understood as being about the urban – interactions between white home owners and their black domestic workers – are key to understanding conviviality in cities with stark inequality. It is within these routinised interactions that significant entanglements are formed across lifestyles and spaces. In these interactions, dispositions become formed which shape urban dwellers' ways of seeing and acting towards each other, and which become transposed to other urban contexts, like shopping malls and religious spaces, and hence shape urbanity.

It is therefore more than surprising that domestic work rarely enters analysis of the urban. Pieterse points out this gap: "Given the surreptitious presence of service workers in almost every middle-class neighbourhood, why is it that we have so little knowledge and understanding of the effects of these criss-crossing spatialities?" (Pieterse 2006: 408) One reason for this omission may be because these entanglements are formed in intimate, private spaces which are often not considered part of the urban. Bank, based on women's studies researcher Judy Giles (2004: 18), criticises Simone and others for what he calls a 'masculine view of the city' which emphasises mobility, adventure and newness, while a 'feminine view' would be more sensitive to routine, continuity and the importance of fixed places (Bank 2011: 19). The chapter hence takes the analysis of urbanity across the doorstep of the home and draws the everyday entanglements formed in the intimate spaces into the centre of analysis.

### **Linbro Park's Hidden Landscape**

If you walk or drive into the suburb of Linbro Park, you might feel that you have left the traffic-ridden northern suburbs with their large office complexes and malls or the dense Alexandra township with its constantly bustling street life behind. Here in Linbro Park, the roads are narrow and shot through with potholes and sand. There are no sidewalks; only a lawn of grass separating the road from the walls and fences that demarcate the boundaries of the large properties, mostly hidden behind trees. The residential density here is extremely low, even for the spread-out northern suburbs. Residents who own land in Linbro Park and are hence property owners like to stress that the surface area of the suburb is equivalent to the CBD of Johannesburg; a statement which expresses these residents hopes about the land's potential for profit-maximising urban development (see chapter 4).

Only a small section of each property is used for housing. On the rest of the land the property owners have laid out vast grass lawns, they have put up additional buildings for tenants, they keep horses, they have their own businesses, or they may even have a tennis court. Tall trees cast their shade on the properties and the roads. There are some deserted properties, often awaiting new constructions, where high grass and bush have grown over decaying fences and buildings. On other properties trees are regularly cut back by gardeners; yet they are not as tamed as the professionally landscaped environments of gated communities and office parks. Some well-kept properties are surrounded by new walls with electric fences and a security gate; others only have a small, decaying wall. Distances between properties are large, and there are no street lights. The property owners had a say in this: once the municipality wanted to install street lights, residents told me, but they opposed it because they wanted to keep the rural feel. Linbro Park is not a well-known suburb, not least because it is hidden away next to the N3 highway (between Marlboro Drive and London Road) and behind a landfill site. Only in the recent years has it become better known, when the northern fringe of Linbro Park was transformed into an office complex, the Linbro Business Park, and drivers discovered that the roads of the suburb could be used as a short cut to avoid the traffic jams on the highway.

South Africa experienced a sudden mining boom and economic upswing when, during the Great Depression of the 1930s, the gold standard was abandoned (Heer 2018:

189). A revolution in suburban space for the White milieus set off, caused by increased income, increased mobility (from car ownership), and state subsidies for housing (Mabin 2005a: 11-23). Brolin, who was then owner of the land which is today Linbro Park, saw its economic potential, and consequentially subdivided his large farm into smaller agricultural holdings and sold them (Heer 2018: 188). The families who eventually bought these properties were of European origin and generally English-speaking. Some were descendants of farmers whose parents or grandparents moved from Europe to rural Southern Africa as early as the 19th century (ibd. 189). Others came as skilled worker to South Africa during apartheid, and often their relocation was subsidised by the state: The apartheid government was recruiting immigrants from Europe. On the one hand, the aim was to address the lack of skilled workers. But on the other hand, it was also out of fear that the White minority was diminishing (ibd. 189, Segatti 2011: 36).

The colour bar guaranteed that the apartheid-era immigrants who were mostly skilled handymen had good jobs and salaries (ibd. 189). For many White workers, the 1970s was a period of upward mobility into more skilled jobs (Crankshaw 1996: 652). At that time, many Linbro Park residents started their own businesses. Nowadays their businesses are mostly in light manufacturing, agriculture and in the tertiary sector. Many property owners host family members like parents, siblings and grown-up children on their land, usually in additional houses or cottages. Many households in Linbro Park are, like in Alexandra, multi-generational (ibd. 189).

White privilege hence ensured the successful upward social mobility of Linbro Park property owners during apartheid. This is still ref lected in the spatial dimension of their lifestyle In the same decades of apartheid that denied Alexandra's inhabitants their citizenship rights, European immigrants enjoyed easy access to South African passports. They could attend well-funded government schools and universities. Sending their children to good private schools and universities is normal for many of Linbro Park's property owners today, and through that, they make sure that their favourable social position is reproduced. In the many conversations and interviews I conducted in Linbro Park, property owners emphasised that their high social position results from hard work and is hence merited. Such views are typical for groups who enjoyed White privilege (ibd. 189, Gallagher 2003).

Linbro Park has not yet experienced desegregation in terms of its ownership structure. In 2012, there was only one black family who owned one property as well as there were a couple of Indian property owners (ibd. 189). Many property owners rent out cottages to tenants; they are mostly from lower middle-class milieus and from diverse backgrounds (Indian, Afrikaans, white, black). The largest milieu besides the property owners is the milieu of what the Seekings and Nattrass (2005: 248) call 'marginal workers': domestic workers, gardeners and handymen. They work for the property owners and businesses in Linbro Park and nearby suburbs. Many originate from rural areas in the Southern African region (mostly South Africa, Zimbabwe, Mozambique and Malawi); few have grown up in Johannesburg's townships. Many are engaged in multi-local households; living temporarily in Linbro Park is part of their economic strategy to help sustain a home in rural areas or in a township. Until 1984 only 'Europeans' were allowed to legally immigrate (Segatti and Landau 2011: 138). Since 1992, and especially since 2000, the number of documented and probably also undocumented immigrants from other African countries has grown constantly (Segatti 2011: 40). At

the same time property owners left the suburb Linbro Park and transformed their houses into cheap rental accommodation. Both this macro- and micro-trend led to the growing presence of marginal workers in the area: "Long back we used to know each other, we used to know who is staying where. But now, we are so many, and they are mixed up, coming from different countries" (Anna, 45-year old member of the property-owing milieu in Linbro Park, May 2012). Many do not have South African citizenship and little prospects of getting it in the future. Their educational background is very diverse: some went to school only for few years, while others have a university degree, but cannot get a job other than as a gardener. Their children usually attend public schools at their other 'homes' where they get looked after by relatives.

That a suburb is also home to a milieu of workers is nothing new. Since the early days of Johannesburg, White families have hosted their gardeners and housekeepers on their premises, usually in simple rooms in the backyard. While many of the White residents were away from the suburb during the day at work or at school, it was Black domestic workers who populated the White suburban houses and streets (Ginsburg 2011).

While grand apartheid legislation was designed to keep the differently categorised populations apart from each other as much as possible, the desire of White middle-class women to 'outsource' the tiring household chores to cheap Black employees brought the differently categorised urban dwellers together on a daily, habitual basis in the most private spaces of the white population – their suburban homes. In the early days of Johannesburg, domestic work was the domain of Black men and White working-class women. This shifted in the 1910s, when reports of male domestic workers emerged who allegedly assaulted and raped White female employers (Beavon 2004: 96). Domestic work subsequently became the domain of Black women. For men, working on the mines was their main route to employment and building a life in the city. For women, it was domestic work (Cock 1989 [1980]: 20). As women, as black urban dwellers and caretakers engaged in household chores in the private realm of homes, the black domestic worker's role in the urban past and present has remained largely invisible; a *hidden landscape* (Ginsburg 2011).

The institution of domestic work has survived apartheid and still forms part of the suburban lifestyle in Johannesburg. According to Sparks, about 60 per cent of South African households employ a domestic worker (Sparks 2011) and it has ceased to be a privilege of white families. In Linbro Park, the majority of the property owners have full-time or part-time domestic employees, mostly male gardeners and female housekeepers. Many of the full-time domestic workers live on the properties in backyard rooms, others live in nearby township areas like Alexandra (especially River Park) and Tembisa and commute to Linbro Park on foot or by taxi.

Working conditions during apartheid were appalling for domestic workers and were often compared with slavery by workers and scholars alike (Cock 1989 [1980], Ginsburg 2011). Because of strict labour controls and pass laws, many women were living illegally in the city and seldom had access to other jobs (Cock 1989 [1980]: 1). Employers disliked the bureaucracy needed to employ an African according to the laws, and often employed domestics illegally (Ginsburg 2011: 109). During apartheid, there was no legislation regulating hours and salary, so that domestic workers worked up to 80 hours a week. They were an extremely insecure and vulnerable group of workers (Cock 1989 [1980]: 6).

Since the end of apartheid, working conditions for domestic workers have improved. There has been a massive effort by the state to regulate paid domestic work and, since 2004, South Africa has had a progressive employment regulation, not least because of the well-organised domestic worker organisations. 'Servants' were transformed into 'workers' (Ally 2009). In Linbro Park today, working hours have become adapted to office hours. Workers work from around seven or eight o'clock in the morning till four or five in the afternoon. Saturdays and Sundays are usually off and they generally have holidays (see also Ally 2009).

In 2012, some domestic workers I interviewed in Linbro Park were employed without work contracts, not least because they were illegal in the country, and some earned less than the minimum wage. Real working conditions are worse than the legislation prescribes, not least because domestic workers find it difficult to negotiate better conditions because of their illegality and their dependence on these jobs. During apartheid, domestic work provided a possibility for survival in the city for those considered illegal because they were considered Black; in the post-apartheid city it is the 'new' illegals, people who lack South African citizenship, who engage in these jobs. In Marxist terms, domestic workers thus constitute a secondary, marginal labour market which pays too little money for working class 'insiders' to sustain their families which is why they are less ready to do these jobs (Meillassoux 1974). Lebo, a domestic worker from Alexandra who had no identity documents, told me: "Domestic work is too hard, but I have no choice" (Lebo, June 2012).

### **Control and Parallel Social Orders**

John is Zimbabwean and in 2012 was working as a gardener on a property in Linbro Park. He was responsible for maintaining the outdoor areas of the property, like cutting the lawns, pruning the trees and repairing the fences. Musa, a South African woman from a rural township, was taking care of the inside spaces of the houses like cleaning and washing. Domestic work and gardening include a lot of routine and both would know for themselves which tasks would be the most urgent ones the next day. "Early in the morning, I wake up and plan what I want to get done. But then, my boss comes and tells me to do something else. But hey, the boss is the boss," John explained. When the female or male household heads were at home, they would monitor their employees' work closely, would give instructions, would observe how Musa and John were carrying them out, and tell them if they did not meet their expectations. In contrast to the members of the household who move around freely on the property, the domestic workers' movements and actions were tightly controlled and regulated by the employers. One day, a tall truck tore down a large branch in front of the property where John and Musa were working, which blocked the road. The gardener, John, felt that the tree should be removed from the street urgently, but he did not dare to do so as he had orders to work on something else. He therefore called his employer at work to get permission to postpone the other task. "I did not want to get shouted at," he explained to me. Many domestic workers said in the interviews that they resented the close supervision, which they experienced as an insult to their ability to think for themselves and it gave them the sense of being seen as stupid by their employers.

Most domestic workers in Linbro Park are expected to wear uniforms when they are at work. Uniforms mark them physically within the private space of the property and the public streets as being in their 'temporary ritual states' (Goffman 1959: 23) as workers. During apartheid and still today, uniforms help the white suburbanites to make sense of their world (Ginsburg 2011:144). Men usually have to wear a blue overall; women's work attire is a full body apron and a head scarf. The uniform needs to be seen as a visual expression of their position in the social order, namely, to serve (Cock 1989 [1980]: 51). The uniforms materialise and reinforce differences between property owners and their employees. Uniforms make visible the relationship the workers have to the suburban space; a work relationship in which they are subjected to rules and control. When property owners get home from work, they, in contrast, change into informal tracksuits, marking their bodies as being in the sphere of home and relaxation. Clothing are hence visual expressions and reminders of the asymmetric relationship employer and employees have towards the property and towards each other (Ginsburg 2011: 142-144).

Also, in the public spaces of the suburb, the uniforms mark the domestic workers' relationship to the neighbourhood as one of work. In the neighbourhood public spaces, the uniform distinguishes them from other strangers who are not under the control of (white) employers (ibid: 144). Domestic workers can also make use of the trustworthiness which their uniform symbolises. Ginsburg (2011) reports that during apartheid the perception of African women in domestic worker uniform as harmless allowed them to go about illegal activities in suburban spaces without raising suspicion. The same does not hold true for black men today: nowadays in Linbro Park, property owners believe that criminals purposely dress in blue collar overalls in order not to raise suspicion. Because of the many construction sites, many blue-collar workers, who work and sometimes also live in the suburb, are not under the control of property owners as their employers and are often blamed for break-ins.

During apartheid there were many so-called racist 'house rules' which regulated the interactions tightly. House rules were related to popular ideas about hygiene and black germs, which justified practices like separate toilets and separate dishes and cutlery. They ensured that "the fork was not licked by an African tongue" (ibid: 146). Nowadays in Linbro Park, the times of explicit house rules have long past, but there continue to exist unarticulated conventions which regulate the workers' engagement with space.

One day I visited the domestic worker Musa while her employers were out. She was busy cleaning the living room and she invited me to sit down on the couch. Her gesture of inviting me to sit was performed with special emphasis, making clear to both of us that it was a transgression; we were using the space in a way which would not have been appropriate if the household heads had been around: Musa was in a social position to invite someone to sit down and chat. One day I conducted an ethnographic interview with John, and we sat in the dining area of a cottage which was usually sublet. He sat on the edge of his chair as if he felt completely out of place. He knew this room intimately from fixing things, but sitting at a table was clearly an unusual, somehow even transgressive form of engaging with this space. Some of what were explicit house rules during apartheid are hence still present in physical practices and perceptions of what are appropriate and inappropriate uses of spaces for the workers.

In many households, domestic workers move around silently and sometimes their bodies are bowed slightly downwards when they perform their tasks in the presence of their employers. They can enter a room without their employers even taking notice. "It's like they blend into the background and are forgotten", a 35-year-old resident living in Fourways once told me:

They are the underground network of any suburb. They knowall the gossip, everything that is going on in everyone's house and they share who's having an af fair, who's pregnant,who losttheirjob,whose kid is ducking school. People do and say thingsin front of domestics because they don'tthink ofthemas people theway they do oftheirfriends or colleagues(Luversan, resident of Fourways, May 2012).

Goffman calls this role 'non-persons', people who are present during an interaction but are neither performers nor audience (Goffman 1959: 152). The employers assume that if they do not pay attention to the workers, then the workers do not observe them or listen to them (Ginsburg 2011: 150).

Control of domestic workers' activities and uniforms and the implicit and explicit house rules are key devices by which the property owners uphold their privacy despite the intimate entanglements; the parallel social orders thus created ensure that despite continuous co-presence in spatial proximity, social differences do not become blurred. Rather, they are continuously reminded and visualised through uniforms, through different engagements with space and differing rules of interaction. Diverging norms and expectations for domestic workers and family members ensure that within the same room, parallel social orders, the "separate, unequal, but coexisting spheres of white home and African workplace" (ibid: 145), are constituted. These parallel social orders contribute to the fact that the everyday presence of these labour entanglement in the suburban family's lives can somehow fade into the background and be forgotten.

### **Hidden Resistances and Performance of Servility**

The everyday intimate entanglements between suburban property owners and domestic workers are anything but symmetric: differentials in power, class, citizenship and mutual dependency create a deeply unequal relationship. Asymmetry and inequality are not a contradiction to urban entanglements; rather, building entanglements constitutes a key way in which urban dwellers *deal with* asymmetry and inequality. Because the asymmetry and mutual dependence is so powerful in the domestic work entanglements, urban dwellers develop sets of practices, moulding these relations in a way which makes them, somehow, bearable. While the suburban employers employ practices like invisibilisation, social control and separating social orders, domestic workers ease the social tensions and defend their independent agency through practices of hidden resistance.

The anthropologist James Scott (Scott 1990) argued that interactions between employers and servants can best be understood through a theory of domination and hidden resistance. According to him, performative interactions between dominator and dominated are shaped by *public transcripts*, so through hegemonic public conduct. At the frontstage, dependant servants perform subordination; in the backstage, they

question their domination through hidden critique of power, or *hidden transcripts*. Not all asymmetric entanglements correspond to this theory of domination and resistance; yet the structures of domination operating in post-apartheid Linbro Park, which carry the legacy of colonial master–servant relations, can be explicated by this analytical lens.

When the gardener Peter, a Malawian living and working in Linbro Park in the last six years, walks around on the property of his employers in his blue overall, his back is usually bent a little, as if he were trying to make himself smaller or invisible. When talking to his employer, he rarely looks into her face, and constantly nods his head approvingly. He often addresses his boss in a low tone of voice. One day I accompanied Peter to the London Road Church, the charismatic church in the neighbourhood (see chapter 6). When I observed him interacting with friends from his home country, most of them also working in neighbouring suburbs as gardeners, he appeared to be a different person; he showed self-confidence, kept his head straight and used broad gestures when talking, as if he now dared to take up space with his voice and body.

Like many other people employed to work in the homes in Linbro Park, Peter tends to perform deference and servility in front of his employers as well as in front of the other white residents of the suburb. Cock and Ginsburg both described such performances – domestic workers with bowed back, their eyes apparently focused only on their work – for the 1960s and 1970s (Cock 1989 [1980], Ginsburg 2011: 68, 162). The analytical lens offered by James Scott invites us to see these bodily practices within the neighbourhood spaces dominated by the employers as performative dramatisations, which aim at fulfilling the employers' expectations of them as submissive workers. Rather than interpreting these symbolic subjections through the physical posture as expressions of actually felt inferiority, they should be seen as symbolic displays of deference and loyalty, which do not express the worker's self-perception, but are rather a self-portrait of the dominant elites of how they want themselves to be seen (Scott 1990: 18).

Like many other domestic workers I got to know in Linbro Park, Peter felt a certain fear towards his employers, fear of being dismissed and fear of verbal humiliation. During colonialism and partly also during apartheid, punishment of slaves and servants could take the form of physical violence. In Linbro Park today, stories are told among domestic workers that certain property owners allegedly used physical violence against their employees. But if it still takes place, it is extremely rare and exceptional. Yet in interviews, workers often complained about the tone of voice which their employers used when they made a mistake; they reported demeaning comments and getting corrected for petty mistakes. Such verbal acts of correction, especially if they involve a raised voice and demeaning comments, constitute a form of verbal punishment for the domestic workers, which they fear. The domestic worker Buhle talked often to me about this lack of etiquette by her employers.

The way they talk to people, the way they talk to us domestics … But I don't let them, I just turn it back to them. But my colleague, she just keeps quiet. They ... they are God! Maybe they told themselves: 'We are the bosses!'.IfI did somethingwrong, and I amnot aware that it was wrong, you have to come to me nicely and say: "Why did you this and this?" They should notsay: "Come here,come here! Why are you doing this and this, you are so stupid!" (Buhle, domesticworker in Linbro Park, May 2012).

The suspension of self-control and dramaturgical discipline (Goffman 1959), the apparently uncontrolled outburst of emotions is offensive and hurtful for the workers. The workers interpret the shouting as an insult to their human dignity and integrity. Being shouted at creates emotional pain and leads to an atmosphere of fear for the domestic workers. The performance of servility and deference are hence also a measure of self-protection: pretending to be submissive helps to avoid violent words.

In this conversation, Buhle also brought up the topic of reciprocity: she claims that she would defend herself, unlike most of her colleagues who would just keep quiet. The unequal power relation of the employer–worker relationship denies the workers as subordinates the "ordinary luxury of negative reciprocity" (Scott 1990: 23). Many of domestic workers in Linbro Park spoke about a lack of communication. Male domestic employees in particular reported that they did not dare to address their employers verbally.

Maybe one day Iwill bring my family here. Butthese people [the employers] don't allow workers to stay with family. I didn't ask, but I heard these facts from previous workers. The boss didn't tell me openly. So, I don't know if I can ask, I don't know how they will respond (David, gardener in Linbro Park, May 2012).

David assumed that he was not allowed to bring his family to live with him on the property because he had heard it from previous co-workers, but he did not dare to ask about it explicitly. Domestic workers hence often act based on assumed house rules which have never been verbalised and may even follow rules that their employers never laid down for them. Zanele told me that she had a good relationship with her employer, a relationship of hanging out together in the kitchen, of calling each other by personal names and of helping each other beyond the work commitments. Nevertheless, Zanele could not openly articulate critique or concerns in this relationship.

The problem is the bosses. They don't know how to treat people … If you shout at me, I get angry and I cry. I cry until you see that I am angry. If you say sorry, if you realise you made a mistake, I stop crying. I don't like that. Barbara: So, the bosses shout at you sometimes? Zanele: The lady, she has never … ah yes, she did, sometimes (Zanele, domesticworker in Linbro Park, May 2012).

Zanele reported that she retreated into bodily communication like crying as a form of expressing emotions instead of verbally articulating that she was angry. When she wanted to get electricity in her backyard room, she did not ask openly for it, she told me. She complained to her employer about feeling cold in her room. Eventually she wrote him a letter asking for electricity and he then installed it, apologising for not having thought about it earlier.

In a very literal sense, these examples show how within the worker–employer relationship, the workers are not 'speaking subjects' in the sense of Foucault. Their dependant position and their work role delegitimise their words and inhibit their capacity to perform 'speech acts' (Austin 1962, Du Toit 1993: 323). Similar to what Du Toit (1993) describes for farm workers in the Western Cape, domestic workers in Linbro Park have no voice. Sara, a domestic worker, put it like this:

We don't have freedom ofspeech to our bosses, most of us. Barbara: Butwhy do people not speak up? Sara: I think, it's feeling inferior. I don't know how they think, but maybe they think you get cross. They think a lot about you. Even people who went to school, they say nothing, even if it's wrong [what the employer does]. Or they say: "We were born dif ferently, we can't come out with our feelings, we can't say it out loud." It's like someone who is being abused by their husband, but they just can't speak out. It's like that, Ithink,they are ashamed. I don't know(Sara, domesticworkerin Linbro Park,May 2012).

In this quote, Sara included me as white anthropologist in the generic 'you', together with the white domestic employers; it shows how much the property owner–domestic worker entanglement is embedded in local understandings of society as still deeply racialised. In many interviews, domestic workers expressed how they do not dare to ask questions or voice their concerns, but resort to nonverbal communication and acts based on assumptions about what employers expect. Silence and avoidance are hence strategies employed by domestic workers. Avoiding verbal communication is a form of 'impression management' Goffman calls it (1959). Remaining quiet can help domestics to avoid falling out of their performance of submission by, for example, shouting back at the employer. The absence of communication has been noted in literary studies on master–servant relations (Mphahlele 1962, quoted in Cock 1989 [1980]: 83-84). Mphahlele interprets the silence as a weapon by the servant "against the white master who has all the instruments of power on his side. Both of them know this" (ibid: 140,). During apartheid some White employers literally refused their domestic workers the possibility to talk; workers could get fired if they complained (Ginsburg 2011: 157). Sara, in the quote above, presented diverse interpretations for why many domestic workers do not dare to speak to their employers: maybe an inner feeling of inferiority, a fear of being shouted at, a perception of a lack of intercultural understanding, and feeling ashamed about being a victim of verbal violence.

Among domestic workers in Linbro Park there are stories which they tell about their employers to each other; some of them may exist in different versions in other suburbs as well. According to one such story, a close relative of a domestic worker in Linbro Park had died and the worker told her employer that she needed days off to attend the funeral. According to this story, her employer responded: "Why do your kind of people die the whole time?" I was told the story repeatedly, and the storyteller always expressed feelings of moral outrage. Instead of expressing condolences for her loss and acknowledging her grief, the employer talked to the worker as if she was representative of a racial other. Albert Memmi calls this form of speech the "mark of the plural" (Memmi 1967: 85). The employer did not see the domestic worker as an individual but as a representative of a generalised other, the Black poor, the once racially inferior (ibid: 85). It was this depersonalisation and performance of disrespect towards the death and her loss which the workers found to be morally outrageous and insulting.

One way of dealing with these acts of indignation is gossip. When domestic workers visit each other in the evenings, when they talk to other workers on the street, or when they get home to their families at the weekend or at the end of the year, they talk about their employers and special incidents, and about the hardship of domestic work. Many complain about the physical strain, the workload, and about health: many domestic workers suffer from back pain and physical signs of emotional stress like

blacking out. Stories about morally outrageous mistreatment by employers travel and become retold by workers who did not experience it themselves. In gossiping about these experiences and criticising them in a loud, angry voice, domestic workers can express their condemnation of their employers' behaviour and, at least with their co-workers and families, reassert part of their human dignity. Ginsburg (2011: 157) observed similar behaviour in the Johannesburg suburbs during the 1970s and found that such mutual commiseration helped domestic workers to feel better about themselves, even if it did not improve the situation.

Gossip exchanged among domestic workers can be seen as backstage performances (Goffman 1959) in which the workers discuss their roles so that they are able to play their part again in the next frontstage performance with the employer. The reciprocity of speaking back can now be enacted in this back region outside the spaces and interactions dominated by their employers. They constitute *hidden transcripts* (Scott 1990), hidden critique of power uttered in the absence of the powerful. The hidden transcripts are a derivative discourse embodied in "speeches, gestures, and practices that confirm, contradict, or inf lect what appears in the public transcript" (ibid: 4-5). In the private spaces with co-workers and family, domestic workers call their subordination into question, they make public their unhappiness with their social position and, by proclaiming it as morally wrong to be treated as such, they reclaim their social and moral value. The dominant discursive frame within which domestic workers talk about the poor treatment by employers is usually apartheid and racism. Many domestic workers explained situations which they perceived as outrageous as routinised habits which employers had acquired during the apartheid regime: "This is because of the way my employer has grown up" (Peter, gardener in Linbro Park, May 2012).

Hidden transcripts can go beyond verbal acts and can entail forms of retaliation. During apartheid, domestic workers sometimes hid guests in the back rooms and stole from their employers. These were not explicit, coordinated efforts of resistance, but hidden practices of 'infrapolitics' which pushed back against what had offended them (Ginsburg 2011: 22-23). Infrapolitics are low-profile forms of resistance constituted by disguised but nevertheless political acts (Scott 1990: 20). Around properties in the suburbs in Johannesburg there was a hidden net of activities unknown by the employers; practices like hiding lovers, relatives and strangers in their domestic quarters transformed the ideologically White suburbs into lived spaces of hidden cohabitation (Ginsburg 2011: 111ff.). Ginsburg also reports the frequent violation of the spatial and physical etiquette which employers imposed on workers, like drinking from the Whites' glasses and turning on the radio (ibid: 157-9). Cock's study on domestic workers in Eastern Cape in the 1970s showed that 98 per cent of her informants used things that they were not supposed to (Cock 1989 [1980]).

In Linbro Park today, one such practice of covert resistance is the making of unauthorised breaks when the employer is not around. In such situations, when the employer, the audience for their performance as diligent worker, leaves workers take on a backstage style (Goffman 1959: 129). They take possession again of their bodies and reject their employer's control over their work rhythm. Employers tend to interpret such practices as laziness, and hence the workers need to be controlled, employers told me.

In interviews workers suggested that crime can also be seen as practices of deviance, resistance or even retaliation.

At 3rd Road, a lady was cut by the garden boy, she died. When we, the people and the police, tried to ask, the garden boy said: "It's because she didn't want to pay me. All the timeswhenIwantedmymoney,shewouldtellme stories,here andthere. So I got angry. In fact, I didn't want to cut her throat, I just wanted to take all her money. But unfortunately, when I tried taking all the money,she came, and I didn't want to be arrested,so I killed here." So if theywere in good books, therewas nothing like that (Sara, domestic worker in Linbro Park, May 2012).

This story about the killing of a female employer by her gardener because of a wage dispute was told repeatedly to me. The domestic worker, Sara, presented the murder as a direct consequence of the employer's poor treatment. Domestic workers suggest that the best way for employers to protect themselves from 'inside jobs' (when employees assist criminals to get into the house) and crime in general would be to treat their workers well. Workers in Linbro Park did suggest in interviews that inside jobs can be related to a deep-seated hatred generated by yearlong experiences of inequality and attacks on human dignity. Crime hence becomes a tactical form of resistance (Certeau 1984), a form of critique of persistent inequality.

Because they depend on each other in basic ways – the worker needs a salary, the employer a household organised – they do not inhabit separate spheres in the city but come to share what Mbembe calls "the same episteme" (Mbembe 2001). Following Nuttall (2009: 1), being entangled with each other is the "condition of being twisted together or entwined, involved with; it speaks of an intimacy gained, even if it was resisted, or ignored". Domestic work relations as a mode of entanglement are deeply troublesome for both the employers, who often deal with it through invisibilisation and social control, and the domestic workers, who tend to deal with it through forms of hidden resistance and conf lict. There is, however, also a further set of practices through which employers and domestic workers deal with the everyday intimacy in their asymmetric relationships: the moulding of patrimonial ties. This corresponds to a more positive way of experiencing these entanglements, namely, through affection. Although these two sets of practices or attitudes (control/resistance versus patrimonial ties) may appear to be deeply contradictory, they coexist within the same relationship, causing domestic work entanglements to be shaped by the tension between them.

### **Affection and Patrimonial Ties**

In Linbro Park, critique of power is not always relegated to the hidden transcripts; there are also workers, especially women, who feel comfortable enough to raise their voices against their employers.

One day Isaid to thewife: "I am not a slave; Iwork here because I need your money and you need my hands. Don't treat me like I am a slave." Sometimes I think about leaving, finding a job somewhere else, butwhen Ithink about Liz [their adult daughter],she isso good,so sweet (Buhle, domesticworker in Linbro Park, May 2012).

The comparison of domestic work with slavery is quite common among domestic workers. Buhle emphasised the mutual dependency in the relationship, but also talked about her freedom as an economic actor in a free market who could sell her labour to whom she chooses. Then she contradicted the free labour market idea by drawing attention to a key aspect of the intimate entanglements between employers and domestic workers; the affective ties. Despite the depersonalisation, the violent words and the expected performance of servility, domestic work also involves a strong emotional dimension, ref lected in the way that people feel and talk about it. Because she felt affection to the daughter of the household whom she had seen growing up, Buhle was ready to remain with her employer even though she considered the treatment to be humiliating.

Domestic work, as a paid form of care work, is in many ways different from other types of wage labour. In contrast to working in a shop or a factory, there is no clear temporal and spatial separation between workplace and home, especially for domestic workers who live on the property. Another fundamental difference lies in the nature of care work. In households with children, female domestic workers often have childcare responsibilities. Children demand different treatment from inanimate household items. Care work is about relationship building; giving the children a sense of comfort is part of the job (Jochimsen 2003). If care providers are replaced, the quality of the work suffers (Stingelin and Schillinger 2012: 8). A caregiver can feel compelled by societal norms and values to do the job, which makes it difficult for her to quit and leave the persons in need behind. The needs of a dependent child and their fulfilling can create empathy and affection in both the caregiver and the care receiver (Jochimsen 2003).

During apartheid in White middle-class households, the workers were often the most constant presence in young children's life. Many women worked 80 hours a week for a meagre salary, but still felt compassion for the children, the sick and the elderly they took care of (Cock 1989 [1980]: 16, 92). Out of this continuous co-presence and caregiving, a high degree of intimacy could emerge (Ginsburg 2011: 138). Ginsburg finds that the best word to describe this intimacy is love, "the fractured, conf licted, pathological, self-doubting love that often exists among family members of a dysfunctional household, but love nonetheless, including fondness for, knowing of, and dependence upon another person" (ibid: 138).

Gabriel, the son of a property owner in Linbro Park, was brought up by domestic workers who have been with his family for over 20 years.

Barbara: Now who is living here? Gabriel: Just my family and the Zulu staf f … Three of them live here, one of them lives in Tembisa. But they have known me and my brother since I was a little baby. So it's nice, for them, it's nice to see us grow up. "You naughty, Gabri, you naughty!" (Gabriel, 23-year-old son of a property-owning family in Linbro Park, May 2012).

Imitating a variety of English with interference from Nguni languages, which Gabriel saw as typical Zulu intonation, he explained how his domestic worker would call him with a nick name 'Gabri', which indicates the informality of the relationship, the affection, but also the complex asymmetries between the privileged white child who depended on the care given by the black domestic worker. His tone of voice spoke of mutual fondness and he prided himself in the interview for visiting the domestic worker's son in Alexandra regularly whom he called his friend. By marking them with an ethnic (and black) label ('Zulu staff') and by imitating what he perceived to be a typical Zulu accent he made very clear that he assigned them to a different socio-cultural

world, even though their everyday lives were deeply entangled with each other and there was obviously so much affection.

This relationship between domestic worker and the employers' children is often characterised by subtle power negotiations which are played out in a joking manner. The domestic workers are given a limited, surrogate authority by the employer to make sure that children behave properly, an authority often contested by the children. Lebo was working as domestic worker in a household with teenage boys whose lives she knew in detail. She told me that she would shout at them in a joking manner if they made a mess in the kitchen after she had just cleaned it (Lebo, May 2012). Joking and humour assisted in navigating the insecurities about which situation definition applied (Goffman 2008 [1974]: 91). is it a black domestic worker shouting at her white employer, which is a breach of etiquette, or is it a surrogate mother reprimanding her children?

Relationships between domestic workers and their employers can also be personal and affective. Female domestic workers and their female employers in particular build up some degree of intimacy if the relationship lasts for many years. Sometimes work is interrupted, and they sit together in the kitchen, exchanging information and concerns about children, care responsibilities, kin and neighbours.

Barbara: Do you have a good relationshipwith the property owner? Buhle: When she is there, we sometimessit and chat. We like to gossip. We talk about life, about anything. Or we just take the phones. "Come and see this on Facebook!" We sit and we forget [abouttime, aboutwork]. She is a good lady, I don'ttell you lies,she is a good lady (Mosa, domesticworker in Linbro Park, May 2012).

In such moments, employers and workers can develop intersubjectivity, at least with regard to the responsibilities and worries they both share as women taking care of a household and family. Shared gendered duties can become a commonality and a basis for some mutual understanding. Shifting from an attitude emphasising each other's differences to sociable moments of togetherness based on commonalities could be moments where differences are resolved, at least for seconds, where boundaries become transgressed and the fact that one shares an episteme (Mbembe 2001) acknowledged.

Domestic workers' lives are often marked by a succession of social tragedies, like illness in the family, death, abusive love relations and crime. Burdened by these worries, domestic workers may share them with their female employers, seeking emotional and sometimes also financial support. Sharing of intimate worries, though, has limits: domestic violence and personal health concerns are considered too private to be shared. In addition, the information f low tends to be asymmetric. While employers may be well informed about the family problems of their domestic workers, they may try to hide information about marital issues from to them.

Gardeners, I sometimes observed, were treated by female employers in ways reminiscent of mother–son relationships. One day, a property owner, Sophie, showed me her garden and introduced me to her gardener and his wife who was visiting from Mozambique. The gardener experienced Sophie as a warm, caring person who was never harsh and never insulted them, very much in contrast to her husband. Before we continued our walk through the property, she said to the young gardener: "Don't make another baby now that your wife is here. Don't forget, children need to be fed as well" (Sophie, September 2012). Her comment on her worker's personal decision of reproduction and sexuality implied that Sophie assumed she had more knowledge or foresight than this young couple.

Such communicative practices in domestic worker–employer relations can be termed *paternalism*. Paternalism refers to practices in which actors interfere with another person's freedom and autonomy for what the actor believes is for their own good (Bavister-Gould and Matravers 2001: 476). Paternalistic communicative practices entail treating the interaction partner as possessing less knowledge and being in need of moral or other guidance. In the past, paternalistic attitudes formed part of employer–employee relations on farms in South Africa. In these rural contexts paternalism as an economic relation and ideology regarded the employer as father of the workers who took on responsibilities beyond those included in a mere labour contract (Du Toit 1993: 321). Paternalistic relationships were not only typical for rural areas, but also existed in urban settings like the mines (Ferguson 2013: 228) and were an important component of White attitudes towards their domestic workers during apartheid (Cock 1989 [1980]: 68). According to Ginsburg, seeing the domestic worker as infantilised, and in need of guidance helped the employer to mitigate the fears caused by having strangers in the privacy of the home (Ginsburg 2011: 156).

Paternalism leaves its imprint on the way in which white suburbanites sometimes speak to domestic workers in the post-apartheid city: in situations when they give advice like the above, or seek to provide emotional support, they use a specific way of speaking, a high-pitched voice and a dramatised friendliness. When domestic workers gossip about their female employers, they often imitate and ridicule this tone of voice, which comes to stand as a symbolic code for the paternalistic and asymmetric relationship. These learnt patterns of interacting with each other are beyond consciousness and constitute forms in which asymmetry has become inscribed in the body and in the everyday encounters with strangers. At the beginning of my fieldwork in Linbro Park I attended a meeting of domestic workers. They did not know me yet, but because of my white skin and because our encounter took place in a white suburb, they treated me like an employer: Two female domestic workers started to talk to me in an overfriendly way and offered me all kinds of services. I was puzzled, as during my fieldwork in Alexandra no one had ever talked to me this way. With time I realised that these were the typical registers of employer–domestic relationships and black–white encounters in a white suburb.

In Alexandra I sometimes used to go with Thabo to a shebeen; an enjoyable experience, not least because the customers were generally curious about what I, as a white outsider, was doing in a township shebeen. When I visited a shebeen in Linbro Park in the company of a domestic worker, though, I felt intensely out of place, not least because I was observed with suspicion. When we left the shebeen, a young man approached me. He asked whether I was looking for somebody, assuming that I was seeking an employee. And then he said, "The people didn't know if they could talk to you or not" (visit at Pharao's place in Linbro Park, May 2012).

These registers constitute part of the habitual urban knowledge of many urban dwellers in Johannesburg. They are part of the stock of knowledge which urban dwellers activate in specific situations of encounter. They constitute dispositions and modes of seeing each other which are relevant beyond the specific employer–domestic worker

entanglement in the suburban homes; they also become activated in other settings and are, hence, relevant to stranger interactions as well. They inf luence the way urban dwellers interpret encounters in other situations, be it at the shebeen, in neighbourhood spaces, in religious spaces (see chapter 6) or the mall (see chapter 7).

Some domestics have a good relationshipwith their boss, others don't. Some blacks are like this, others are dif ferent. It'sthe same among the whites. Some are nice with their workers, others aren't (Zuzile, domesticworker in Linbro Park, May 2012).

In this conversation, Zuzile, a domestic worker, emphasises the heterogeneity among what she saw as two racial groups. This again shows how strongly employer–worker relations are still interpreted as relations between the racial categories created during apartheid. In that conversation Zuzile also explained that her boss is like a 'friend' or an 'aunty' for her. She called her boss by her nickname, Babs, instead of Barbara. Zuzile who also went by an English name, Susanne, had her boss refer to her by her Zulu name, Zuzile.

It is common for speakers of Nguni languages to carry a so-called ethnic name which indicates their belonging to a linguistic or ethnic group like Zulu or Xhosa, as well as an English name. This is a result of colonisation and the efforts of missionaries who gave a 'church' or 'school' name to children because their previous names were considered complicated, foreign and heathen (Moyo 1996, Neethling 2008: 32). Naming practices in Nguni languages are more than just a label for a particular individual; the names have a meaning which is located in the cultural context. The name may express an aspiration that the parents have for their child, it may express the position within the family, or it may ref lect the particular circumstances which prevailed at the time of birth (Neethling 2008: 33). For Zuzile, not calling each other by their 'official' names, but with a nickname or ethnic name, is a symbolic code for the intimate nature of their relationship. Naming practices are hence not only ways of expressing identity but also ways of expressing one's understanding of the relationship and inf luencing it.

During apartheid, many employers did not know the full name of their domestic workers (Cock 1989 [1980]: 74). They would refer to them by their English name, or even called them by the generic term *sissie* (in the Xhosa dominated Eastern Cape) (ibid: 118). Domestic workers were expected to address their employer as 'madam'; employers would talk about their workers as 'my garden boys' or 'my girls'. These naming practices contributed to the performance of a hierarchical, asymmetric relationship in which the workers were stripped of their personal identity and represented as inferior, childish others. In post-apartheid Linbro Park, parts of these naming practices are continued in habitualised, often unref lected ways. While some employers have adopted the new legal terminology speak of 'workers' instead of 'servants', others habitually speak of 'maid' or 'my girls'. In interview situations and backstage situations with co-workers or family, the domestic workers in Linbro Park would refer to their employers as 'the boss'; when performing servility in front of white employers, many would address them as 'madam'.

A further aspect of a relationship seen normatively as good by the workers is gift exchange, as for example brought up by Zuzile in that conversation. She explained that she whenever she felt hungry, her employer would make her a sandwich or she could make one for herself. Towards month end when money is short many domestic workers do not eat breakfast but are hungry at work till they receive the lunch from their employers. The receiving of food not on a rhythm determined by the employer but rather depending on Zuzile's own biological needs was hence a sign of a 'good' relationship for Zuzile. Zuzile explained that her boss also assisted her financially, for example if she had difficulty paying her children's school fees (Zuzile, May 2012).

In many domestic worker–employer relations the work–salary exchange is complemented by another form of exchange, gift giving. Employers may help their workers financially when they run out of food before the end of the month, they may assist with paying for a funeral, they may pay for the domestic worker to see a doctor or to get medication and they may give them old clothes and food.

Such gift giving by the employer is generally interpreted by employers and employees alike as a symbol of mutual care and a good relationship. Yet, there are also exceptions. Sometimes, employers give the domestic workers food which they did not want to eat themselves. In an interview, Bianca complained how employers would open a pack of chips, let it stand around for days and then give the stale food to the worker.

When my employer gives me stale food, I would like to tell her: "This is not right; this has been open to aweek,why did you not give it to me immediatelywhen you thought you can't finish it? Because I was there when you opened it, am I supposed to eat stale food?"No,we can't talk like that to them,we can't. We need to say: "Oh, you have given me this..." (Imitating a high-pitched voice). Am I your dog? It's wrong (Bianca, domestic worker in Linbro Park, May 2012).

She explained that a worker could not openly criticise the gift of stale food but had to pretend to be thankful. As already mentioned, Mauss makes us aware that gift giving follow rules of reciprocity (Mauss 1990 [1925]). Although one may think of charity and gifts as being given without expectations of receiving anything in return, in reality gift giving is not altruistic at all, but guided by the self-interest of the parties involved (Mauss 1990 [1925], Rössler 2005: 194). The counter-gift which the domestic workers are expected to give is a social attitude, namely gratitude. This normative expectation is so strong that, as Bianca points out so clearly, voicing critique about the donated stale food is literally impossible for the domestic worker. The domestic worker rather performs the expected gestures of gratitude in front of the employer and throws the stale food away on her way home.

Affection, gift giving and sharing some intimacy and informal etiquette are seen in Linbro Park by workers and employers alike as indicators for relationships which they normatively evaluate as good. They often talk about such relationships in nonwork-related terminology, namely, as friendship, kinship or family. Many employers in Linbro Park claim that their domestic workers, especially if they were with them for a long time, constitute part of the family and many refer to them as *Gogo*, a Zulu term for grandmother.

One day, Monika, a female property owner and household head approximately 50 years old, told me that she belonged to a 'new generation' which builds a close rapport with their domestics. She told me that their *Gogo* has worked for them forever, from the birth of her first son until she retired. She was part of the family, she claimed (Monika, April 2012). Performances are often subject to idealisation and actors aim to create an impression of themselves which corresponds to officially accredited values of the

society (Goffman 1959: 39). Monika's reference to 'generation' probably referred to her German forefathers, farmers during the colonial period, or the generation of her parents who were farmers during the height of apartheid. Defining the relationship to her domestic workers as 'family' was probably less a statement about their actual relation and more a claim to being a 'good' employer as one should be in the post-apartheid era.

The same day, Monika told me a story about a former worker. Her company had gone bankrupt a couple of years earlier and they re-employed one of the dismissed company employees as a handyman in their home. He was destitute and they allowed him to sleep at the house, as she explained, and she even gave him many things which they did not need anymore, like kitchen utensils and clothing of her deceased husband, so that the worker could sell them and make money. Then her gardener told her one day that the handyman was stealing from her. The handyman had helped himself to the old material from their bankrupt company like copper cables, which was lying around in the garden, and had sold them. Monika was bitterly disappointed:

The hand which helps them, the hand which gives to them … But the people, they bite the hand again and again. They mess it up themselves and never learn it. That really hurts. I gave him work, I gave him money, I made tea for him and prepared sandwiches. But that is now over! He can't come anymore, because now he is guilty, and he knows it, and he knowsthatwe knowit (Monika, a 50-something property owner, Linbro Park, April 2012)

In this paternalistic narrative, Monika presented herself as the slightly naïve, benevolent employer who did everything in her power to help her suffering employee. Her narrative of how she took care of him when he needed help is reminiscent of a mother taking care of her child (making tea and preparing sandwiches). She felt responsibility, affection and sympathy for him, and she expected gratitude and loyalty in return for her gift. Her disappointment was not so much because of what he stole (the cables were anyway just lying around to be sold one day), but she was disappointed about the fact that he stole and broke the moral 'contract' and did not live up his role as the grateful gift receiver. Among property owners and other white Johannesburg residents I encountered many such stories about disappointing workers whom one treated like family. Ginsberg notes that during apartheid "the most damning accusation for a domestic worker" was not that she was slow or dishonest, but "that she was ungrateful", as ingratitude represented a "rejection of the white madam's role as protector, teacher, and mother figure" (Ginsburg 2011: 156). In the quote above, Monika applied the "mark of the plural" (Memmi 1967: 85), generalising from the individual example to 'black poor' as a general category. Claiming that "they mess it up themselves" she meant that she intended to be non-racial or non-racist, but that 'their' own behaviour made this difficult.

The expectation to be grateful and to serve can be very powerful. One day, armed criminals entered the quarters where the domestic worker Zuzile was living. They held her and her husband at gunpoint. The criminals wanted Zuzile to knock on her employers' door so that they would open, believing that it was Zuzile. Out of fear for her employers' and their children's life, she invented stories about dangerous security devices at her employer's house and successfully scared the criminals away. She encouraged them to burgle the neighbour's property, where only workers were living, which they eventually did. The affective relationship bound her that closely to her employer that she risked her own life for her; and rather put the lives of workers living next door at risk. The reciprocity of this commitment, though, is not always mutual:

Once we don't stand for each other, it means we are going to die. If I tell you that I am sick today, take care of me, because I am taking care of you. If I am working for you, you leave yourchildwith me, I am the half mother, because I love the baby, I love you, I love everyone in the family. So, once I tell you I am sick today, just be fair with me. Some people, they tend to forget this. Every morning, they go to work, they return at six. We remain here, we are their bodyguards, we are their everything. If ever, we were very rough, they could find empty houses. But once I am telling I am sick, or I am stranded, orsomething came up, please help me! They mustcome up for me! Aslong asI am your worker, it means I am part of the family (Zuzile, domestic worker in Linbro Park, May 2012).

The claim "I am part of the family" expresses strong normative expectations on how employers should act towards the workers. Speaking of family or friendship is used to articulate demands on employers, namely, expectations of solidarity, of empathy, and often also of resource distribution.

Domination and resistance, as well as affection and paternalism, characterise the entanglements between workers and their employers in Linbro Park. Some domestic workers brought up all these aspects when talking about the same employer, others related them to different family members. Research on domestic work in South Africa during apartheid also described these oppositional tendencies in domestic work relations (Cock 1989 [1980], Ginsburg 2011). This can be understood as the ambiguous co-presence of two fundamental attitudes in these domestic work entanglements: encounter and distanciation. Encounter is, on the one hand, "an interaction where both actors perceive and recognize the difference of the other, respect it, and try to build on it in their relationship" (Förster 2013b: 242). Distanciation is, on the other hand, "an interaction here two actors adopt a disruptive attitude toward the other, trying to secure an independent agency" (ibid: 242). While affection, moments of drinking tea together and paternalistic ties can be understood as forms of encounter, practices like tight control and gossip can be seen as forms of distanciation, based on disruptive attitudes towards each other. Domestic work entanglements are hence characterised by this deep, troubling tension between affection and mistrust, between moments of togetherness and awareness of deep dividedness, between encounter and distanciation. These deep ambivalence and contradictions are characteristic of urbanity in general (ibid), and of the *cities of entanglements,* Maputo and Johannesburg, more specifically.

### **Invisible Residents of Linbro Park**

Domestics either live in rented rooms or shacks in the suburb or in a nearby township, or they live in their employers' domestic quarters. These rooms are usually found in the back of the properties and generally consist of simple concrete structures with one or

several rooms.<sup>2</sup> As Cock describes for the apartheid times in Eastern Cape, also nowadays in Linbro Park, the standard that the property owners consider to be appropriate for their domestic employees is considerably lower than what they consider as normal for themselves (Cock 1989 [1980]: 35). Nowadays, most of the domestic quarters have working toilets and electricity, but some domestic workers recount that they had to struggle with their employers in order to get these amenities. At the end of the 19th century, male African servants usually had to sleep on the kitchen f loor. When domestic work became black women's work, English sensibilities demanded better accommodation for them. Detached rooms in the yard, called 'back rooms', became the norm for in-house domestic workers. During apartheid there were even building regulations prescribing a certain distance between the dwellings for Whites and domestic quarters (Ginsburg 2011: 10-11). Within such suburban homes, the micro-segregation of apartheid is often continued today; in some properties there are still back entrances for the domestic quarters, which continue to "create separate social universes for the two groups" (Cock 1989 [1980]: 35). This inequality, which was once produced out of feelings of racial superiority, is maintained, consciously or unconsciously, and serves as a sign of distinction between the domestic workers and the property owners.

In the domestic quarters, domestic workers continue to live under the rules of their employers, which many describe as a 'lack of freedom'. Domestic workers are often not allowed to live there with their children, or to receive visitors or to sublet parts of the cottage or room, which is for many township dwellers an important form of income generation. This is one of the reasons why many domestic workers engage in circular migration and maintain a home in the rural areas or another country, where relatives look after their children. In Lebo's (the daughter of a domestic worker in Linbro Park) case, she and her sister grew up by themselves at the rural home because, as a result of family conf licts, their relatives did not look after them.

Me and my sister, we never lived with our mam in one place. She was always working, sincewewere kids. My sisterwas about three and Iwas five. Since thenwe always have to look af ter each other. We lived together in a house without an adult or anyone since I was twelve. It was hard; we had to wait for her to send money home, so that we can buy food … They didn't allowher to livewith her kidsthere, I don't knowwhy. They only allowusto visit. Itwastough. You know,when you are a girl, you need thatthing of your mam, you know, when you met your first boyfriend, you have to tell somebody, your mam. We didn't have that. It was tough. But we managed to behave well, because we know we have to, otherwise ... we going to disappoint her and ourselves(Lebo, 20-yearold daughter of a domesticworker, June 2012).

During apartheid such prohibitions on living with one's children were related to pass laws and the migrant labour system (Cock 1989 [1980], Ginsburg 2011). Nowadays, this is continued through the individual wishes of the employers not to have African family life on their properties. Living in a tiny room while cleaning the employers' ostentatious house forms part of the everyday life of these domestic workers, observing the

<sup>2</sup>  Some property owners deduct rent for from the salary for living on the property, others do notcharge. By law, employers are allowed to deduct 10 percent of the salary for accommodation (Department of Labour 2012).

privileged life of their employers' children who attend private schools while their own offspring are far away and have few prospects of getting a decent education. Such experiences result from the deep entangledness of the different social strata in the city and form part of everyday life for many domestic workers in Linbro Park, which many interpret as racial and economic inequalities. Such aspects of the conviviality in cities shaped by inequality tend to go unseen, unless one draws the entanglements under one's analytical lens.

The property owner is simultaneously employer, landlord and neighbour. Many workers do not dare to question the working and living conditions their employers-landlords offer them. They are in a situation of multiple dependency, as conf lict might not only result in their expulsion from the residential quarters, but also dismissal from work. People without South African citizenship in particular are in position where they lack bargaining power, as they depend exclusively on this salary, cannot access social grants or ask family members to support them, and find it difficult to find accommodation. Yet property owners see the domestic workers who live on their property predominantly in their roles as workers, and not as fellow neighbourhood residents. When talking about neighbour relations or relations with tenants living on the same property in conversations with me, property owners would, for example, never talk about their relationships to domestic workers, which, for them, fell into a totally different category.

Forbidding domestic workers to have family members and sometimes even visitors can be seen as an imposition on the right to family life. It therefore makes sense that Coser (1974) called domestic work a *greedy institution*, an institution which makes total claims on the person, seeks exclusive loyalty and reduces claims to competing roles and status positions (Cock 1989 [1980]: 81-2, Coser 1974). In order to avoid this 'greedy institution' with its strict regulations by employers, many domestic workers live in township areas and prefer to commute. But apart from the 'lack of freedom', which is how domestic workers in the interviews and conversations refer to such rules, living in a white suburb also has advantages. Linbro Park, with its large properties, the abundance of space and tall trees, is experienced by many as a hiding place, a place where they can withdraw. During apartheid, moving into domestic work was an important strategy of accessing residential space for many women who did not have a husband and were therefore excluded from getting a township house (Ginsburg 2011: 132). Nowadays, in the context of the shortage of affordable and safe housing in Johannesburg, having a whole room of one's own in a gated and walled property may not be the worst option for establishing oneself in the city or surviving after a household falls apart.

For Zuzile, living in Linbro Park is a refuge from township life, which she feels is marked by intense sociality and social control. Because of the large properties, one's private life is less exposed to the neighbours. Being a domestic worker almost allows them to live a bit as if they were suburbanites, a lifestyle which many township dwellers aspire to. It is quiet, one lives surrounded by trees and birds, and there is a lot of space in their view.

l like this place, it's not noisy, it's easy going. If Iwant to talk to someone, I go out to talk to someone, if I don't want to talk to anyone, I can hide myself. There is privacy. But in the location [township], you can't hide. I don't mean I hate my place, my location. But the life they are living in the location, I don't like it. It's full of noise. I wish I was rich, I could buy my house (Zuzile, domesticworker in Linbro Park, May 2012).

This is an example of how the entanglements between the domestic workers and employers bring the workers certain advantages, namely, that they can profit, at least a bit, from the amenities of living in a suburb. A major worry for domestic workers who immigrated from other countries is xenophobia and police violence. On 11 May in 2008, young men attacked Alexandra residents whom they categorised as *amakwerekwere* (foreigners). The attacks which the media and many analysts referred to as xenophobic violence spread rapidly to other township areas and, within a few days, led to 62 deaths and tens of thousands of internally displaced people (Worby et al. 2008: 23). These violent events were exceptional in their extent, but the underlying hostility and resentment against foreigners was and still is quite commonplace and present in everyday urban life. In interviews, domestic workers from other countries reported experiencing hostility in daily interactions in public spaces, taxis and shops, as well as with neighbours. The image of South Africa and Johannesburg suffered considerably because of the xenophobia-related violence; going to work in *John*, as Mozambicans call South Africa and Johannesburg, has come to be seen by many Mozambicans and other foreigners as going to live in a violent and hostile place.

This citizen of this country, any time, they can start the xenophobia. So as far as they know that you are a foreigner and that you are staying here [in Alexandra], they just mug you. The problem is that they come at night, which means, no matter what ever, they gonna go and get you inside your place. Staying in Alexandra would be cheaper [than in Linbro Park], but Ican't risk my life (Simon, gardener in Linbro Park, May 2012).

Some domestic workers carefully hide their illegal status from neighbours and friends for fear of becoming a target. Many domestic workers and other marginal workers choose to stay in the white Linbro Park which they perceive to be a safe place, a hideout where they are protected from xenophobia. Simon explained that he did not like to go to the CBD, because he might be stopped by the police and they then demand money. While on the streets of Linbro Park, the police also often check the identity documents of domestic workers, in contrast to the anonymous CBD, these police officers from the Sandringham police station eventually become acquainted with the domestic workers and do not check their papers anymore. Living in Linbro Park thus gives foreign domestic workers a sense of protection from the dangerous anonymity of the city; it is a place where they can carve out a more or less safe existence for themselves within what they experience as a dangerous and violent country.

Living in a suburb also means that domestic workers who usually do not have cars are confronted with the lack of public transport and dominance of automobility typical of the Johannesburg suburbs. Transport in Johannesburg is still marked by the apartheid legacy: the apartheid state tightly controlled the mobility of the Black population and economically delimited their access to automobility (Graham 2007). Private motor transport has primacy in the suburban life of the city (Czeglédy 2004). Property owners drive in and out of the suburb several times a day. Because of the low residential density, it is not profitable for minibus taxis to pass through Linbro Park. Apart from month end when the demand is higher, one usually has to wait for 30 minutes

to two hours for a minibus. Except for some better-off workers who own a *skorokoro* (an old car), most of the domestic workers move around on foot or use minibus taxis. Distances are large and a walk to London Road, which is on the major transport routes, can take up to 40 minutes. For domestic workers who do not live in the same suburb as they work, walking for two hours or more everyday forms part of the daily routine. This reality rarely enters depictions of Johannesburg's northern suburbs.

Cock reports that in the Eastern Cape during apartheid, domestic workers suffered from social isolation because of the inhumane working hours and the lack of free time to attend church or socialise casually with friends and family (Cock 1989 [1980]: 46-49). Nowadays in Linbro Park it is rather the high costs of transport, the danger of walking at night on the streets and the lack of integration into the urban transport networks which makes living in Linbro Park socially isolating. Several domestic workers told me that their friends and relatives who live outside Linbro Park hardly ever come to visit because they would have to walk long distances on the dark deserted roads. "They say: 'Eish, we want to visit you, but your place…'" (Bianca, domestic worker in Linbro Park, May 2012).

Except for running errands, attending church services and visiting friends and relatives, the daily routine and social life of domestic workers takes place within the neighbourhood. The streets and fences, the boundaries between the public streets and private properties, are important spaces of sociability. When walking from one property to another, when taking the dustbin onto the street or working in the proximity of the fence, one would exchange greetings with workers from nearby properties, complain about the hard work and the treatment by the employers and talk about personal relationships.

Many domestic workers spend their time after work doing household chores and grocery shopping, carrying out additional income-generating activities, visiting others and frequenting the local shebeens. In one deserted property there is a *spaza* shop (a small store) and a traditional healer offers his services. Unless employers forbid visitors, workers' quarters are important places for socialising, with a constant coming and going of relatives, co-workers, friends and members of credit and savings associations.

We just know each other, we meet on the streets, we meet in our rooms, we meet anywhere. Some other people go to London Church,they meet each otherthere. Some they go toMorgan [the shebeen]. I go allthese places. They go to Pharaoh [anothershebeen] and meet people. Mostly men. Ja, we just meet, even in the taxis, sometimes we just chat through the phones(Sara, domesticworker in Linbro Park, May 2012).

In 2012 there were two shebeens in Linbro Park, where men and, less often, women would go. Both were former suburban houses, now deserted dwellings, which were transformed into simple places of leisure where people engaged in drinking, listening to music, smoking, discussing and playing pool. Both shebeens were parochial spaces in Lof land's sense, that is, spaces where people know each other from being neighbours, co-workers, friends or acquaintances (Lof land 1989). One would usually sit together in groups of close friends but also interact with others one was acquainted with. These were not places where strangers would venture, as they would not even know these shebeens existed.

And you see the problem with shebeens is, Barbara, a guy from River Park comes in over here, a worker from Linbro Park gets broke, and to get money he tells them what his boss does and how much money the boss has got and where his gun is. And then ofcourse, he knows to come back, to rob the place (Baldwin, property owner in Linbro Park, March 2011).

The white property owners regarded the shebeens with suspicion. They saw them as uncontrolled spaces of trouble and vice, of prostitution and drug trade; spaces from where crime in the suburb was planned and committed and spaces where domestic workers accidentally or on purpose shared private information about their employers with criminals. This shows how many property owners tend to see the private and public lives of their domestic workers as a disturbance in 'their' suburb, thus demanding that this be under their control. Not least because of the fear of crime, property owners have tried to have illegal shebeens closed down several times.

The property owners of Linbro Park have organised themselves into formal neighbourhood organisations, the most important being the Linbro Park Community Association (LPCA).3 The residents run a security initiative which they call a Community Policing Forum (CPF), a subforum of the Sandringham CPF.4 The security initiative organises a neighbourhood watch, called Block Watch. The Block Watch has not always been functional owing to a lack of involvement by residents, but the idea is that every night two residents patrol the whole suburb for about an hour. During the height of apartheid, the suburb had an active associational life and property owners would socialise on the horse grounds, at the tennis clubs and at ballroom dancing. Many property owners claim that they moved to Linbro Park because they wanted to live in a place where they could keep their own horses, their children's favourite sport. Property owners called this lifestyle 'country living in the city' and it still today forms part of the collective and individual identity of Linbro Park property owners. Nowadays, their children have grown up and associational life has been reduced. Neighbours who consider themselves friends now prefer to socialise in restaurants, bars or gyms outside of suburb or meet for a *braai* (barbecue) at their homes. Yet in the wake of the decline of the former spaces of association, new realms of interaction among property owners have subsequently emerged, one being the virtual space of the internet, where property owners exchange news about security, governance issues and stray horses, and the other being the CPF and Block Watch. Many of the property owners who moved to the suburb after the height of associational life have gotten to know each other through Block Watch. Often family members or tenants who live on the same property go on Block Watch together. Neighbours who have only read about each other

<sup>3</sup>  The LPCA is organised as a section 21 company (a non-profit organisation). It is run by a steering committee of property owners which meet once a month. They represent the interests of the neighbourhood vis-à-vis the local government and private developers, and engage in the maintenance of urban infrastructure, either by pressuring the responsible service delivery entities to fix it, or by fixing it themselves. A couple of years ago they collected money for a concrete wall around the suburb, which served as a protection against trespassers from the nearby open veld.

<sup>4</sup>  Community Policing Forum (CPF) is a structure of community involvement in policing and an important element of the community policing approach which is practised by the South African Police Service (SAPS). The CPF aims at liaising between residents and the police (National Crime Prevention Centre 2000, SAPS 2013). The security initiative haslong organised Block Watch and the Domestic Watch.

in the web-based forum meet for the first time physically during Block Watch and use the hour in the car to exchange news and discuss issues around family, business and development of the suburb. One could say that security governance replaced horse riding as a sphere of neighbourly sociability. Fighting criminals together also creates a sense of belonging and togetherness. I was told that some members of the CPF get up in the middle of the night whenever there is a break-in, and chase after f leeing criminals. Such chases have been reported on the web-based forum and photographs of caught criminals posted.

As chapter 4 will show in more detail, a conf lict emerged a couple of years ago over differing visions for the future of the suburb, and since then many property owners report a sense of what they call a 'divided community'. Clara, a 60-year-old property owner in Linbro Park, for example said: "The community is very divided. Everyone has their own criteria of what should happen" (Clara, March 2012). 'Community' refers in this quote solely to the property owners and 'divided' to the split among them; the milieu of domestic workers remains invisible in this conception of community.

In the interviews, I usually asked the property owners about who they knew in the suburb. Mandy, a 45-year-old property owner in Linbro Park, responded: "Ja, people know me, and I know all my neighbours, Andre, John, I know most people that used to live here, Alexis, Phil, I know everyone in Linbro Park" (Mandy, May 2012). Responding to my question, Mandy exclusively referred to other property owners, excluding the many domestic workers on her and neighbouring properties from the 'people'. Both quotes make clear how Linbro Park, as a collective space, is imagined by the property owners as a white suburb, where the black domestic workers are not seen as residents, neighbours, or citizens.

Their intimate entangledness with domestic workers, on whom they depend, is non-existent in this imagination of the world they inhabit. This also means that domestic workers are largely excluded from neighbourhood politics (see chapter 4). The LPCA is an organisation solely for the property owners, domestic workers are neither members nor part of the board, nor do they participate in the web-based discussion forums. This invisibility of the entanglements shaping lives in Linbro Park is characteristic of elites' lives in *cities of entanglements. '*Invisibility', as defined by Honneth (2003), means the absence of recognition of the other as a fellow human being. This social form of invisibility can also have visual aspects, that is, the physical hiddenness of the spaces of domestic workers' public life. Invisibility is also defined by Goffman as the notion of the 'non-person', namely, people who are present during an interaction but are neither performers nor audience (Goffman 1959: 152). This is similar to what Siegenthaler refers to as "social (in) visibility; the lack of mutual visibility (intervisibility) in interaction creates marginalisation and exclusion" (Siegenthaler 2013: 171). For Brighenti, the relationship of looking at each other "constitutes the site of mutual recognition, misrecognition or denial of recognition of the other – in short, the site where we constitute ourselves as 'subjects'" (Brighenti 2010: 27). Turning domestic workers into non-persons means that one does not need to acknowledge their presence or regard their needs as individuals, rather they become reduced to the subject position as workers. Invisibility was also typical for domestic workers during apartheid (Ginsburg 2011). But it is important to note that domestic workers have agency and can resist, negotiate and make use of the invisibility. As described above, living invisibly in a white suburb can be a form of protection from police controls. In the realm of security governance, however, their invisibility becomes especially troublesome, as their own security needs are anything but central.

### **Securitised Entanglements**

Despite their exclusion from official neighbourhood structures, domestic workers become included in security governance in their role as employees. As in many other Johannesburg suburbs, security is a major issue in Linbro Park. Owing to its physical proximity to Alexandra township many think Linbro Park is a prime target for criminals. There are no streetlights in the suburb, and as a result of the untamed growth of bushes and grass on certain properties, many urban dwellers experience the neighbourhood spaces as uncontrolled and as a perfect hideout for criminals. Many people preferred living in a gated community to living in Linbro Park. Following Gabriel, 23-year-old son of a property-owning family in Linbro Park, many "people think that Linbro is crime ridden. My friends look at me and think I am nuts for living here" (Gabriel, May 2012).

A considerable number of the property owners in Linbro Park own guns or rif les, having learnt how to use them during their childhood on a farm or for sport. Besides sport, guns are kept for self-defence in violent encounters with strangers, as such as burglaries or car theft. Most households have a high fence and dogs, many also have electric fences, alarms and a contract with a security company. Domestic workers form part of the neighbourhood's security system: many employers send their workers to the monthly meetings of the Domestic Watch programme.5 The Domestic Watch meetings, which last about two hours should equip "domestic workers with valuable crime-prevention knowledge and skills" (stated on the Domestic Watch timetable for 2013). The programme is implemented in various suburbs in Johannesburg in conjunction with the local residents' associations, the respective police station, and the private security company ADT, which acts as a major sponsor of the programme. The lessons address questions about what security companies do and how domestic workers should interact with them; they contain information about practices used by criminals, such as how they trick domestic workers into disclosing information about their employer's alarm system. Non-crime related topics such as first aid and domestic violence are also sometimes addressed. The aim is to sensitise domestic workers to security issues and to encourage them to report anything suspicious to their employers. As Clarno (2013: 1200) rightly observes, the programme aims to transform domestic workers from potential collaborators with criminals to auxiliary security producers.

Some of the domestic workers in Linbro Park enjoyed the meetings because they were a welcome break from heavy work, while others tried to avoid the meetings because it encroached on the time they needed to get their heavy workload done. Some workers stopped attending because their employers lost interest in sending them or

<sup>5</sup>  The meetings are officially called 'MAD Domestic Watch Meetings', MAD being the abbreviation for 'Make A Dif ference'. At the time of fieldwork, the Domestic Watch programme was running in about twenty neighbourhoodsin the northern suburbs ofJohannesburg. In each suburb,therewere monthly meetingstowhich all the domesticworkers and gardenersin the suburbwere invited.

the workers found it increasingly boring: "There is only so much to say about crime" explained one of them (Ayanda, domestic worker in Linbro Park, April 2012).

At a meeting held in April 2012, a resident from a nearby suburb, Birgit, gave a talk about 'inside jobs', as she had recently completed courses in the field of criminology and was running a small business offering polygraphy (lie detector) services. 'Inside jobs' refers to crime in which domestic workers are involved as collaborators with criminals. At the talk she explained to the approximately thirty domestic workers and gardeners present at the meeting that she could find out with her polygrapher whether a domestic worker had assisted a criminal. In the written version of her presentation which she sent around on the web-based discussion forum after the presentation, she wrote: "Do not get involved. You are putting your life at risk. If I am appointed, I promise you I will catch you out" (Google Group, April 2012). After the talk, the Domestic Watch convenor, Sophie, attempted to reframe Birgit's accusing tone and explained to the workers that Birgit was offering her help and that domestic workers should speak to their employers or to Birgit if they were approached by criminals.

In the round of questions that followed, Helen, one of the domestic workers asked: "What must we do in order for the police officers to understand that we are not the ones involved in the robberies?" (Domestic Watch meeting, April 2012). With this question, Helen attempted to reframe what the actual security problem was – the problem is not that domestic workers collaborated with criminals, but that domestic workers could be falsely accused of collaborating with criminals. As a response, Birgit recommended asking for a lie detector test at the police station and Sophie emphasised that one should create good relationships with employers and not gossip behind their backs, as this could make the domestic worker seem suspicious. This reference to gossip exemplifies how property owners and organisers of Domestic Watch programme saw sociability and communication among domestic workers as a security threat.

Following this, a domestic worker, Sara, made a comment, first shyly, then, when asked to repeat, in a more assured tone: "We are only talking about madams and bosses. Cause, the thing is, it's not only happening to them. We also get raped, we also get killed, we even get killed first" (Domestic Watch meeting, April 2012). Sara was attempting to redefine what security problems actually were; not only did employers need to be seen as potential crime victims but domestic workers' potential victimhood also required acknowledgement. Birgit and Sophie both interrupted Sara immediately and defended themselves. Birgit argued that they were aware of this, but that violence against domestics was simply not the topic of the lesson on this day. The convenor Sophie emphasised understanding:

It's true, our gardeners and domestics are also suf fering. What I have heard in the ten years of doing this work, is that black people can be very cruel to each other … There is more crime, more burglaries in Alex than in white suburbs. We just hear more about it. But I think sometimes, what happens, is … there is the English word 'unwittingly', if I give information out unwittingly, you are not consciously saying it to others (Sophie, Domestic Watch meeting, April 2012).

Trying to act as a mediator between the two clashing world views, the meeting convenor Sophie aimed to recognise the domestic workers' victimhood, yet also emphasised that spilling out information to criminals could happen without bad intentions. During the Domestic Watch meetings Sophie often spoke in a high-pitched voice and lifted her index finger when speaking to the domestic workers. I noted in my fieldnotes: "Her tone of voice somehow reminds me of a kindergarten teacher talking to her pupils" (Domestic Watch meeting, April 2012). In the meeting a month after the talk by Birgit, Sophie apologised to the domestic workers as she had become aware that many of them had it found insulting. At the beginning of the meeting she said: "Birgit has good intentions, but she is not used to talking to my people, because this is what you are, my people … Now please give each other huggies and forgive her in your heart" (Sophie, May 2012). The domestic workers ignored her request to hug each other, as if feeling slightly embarrassed by the paternalistic tone.

People are affected by crime in very different ways, and therefore their perceptions of crime differ greatly (Statistics South Africa 2012b: 2). In a neighbourhood like Linbro Park, not all types of crimes are prosecuted or made socially visible in the same way. Definitions of what constitutes a threat, what are relevant security problems and what are necessary measures against them differ between milieus, so that security governance is a field where competing definitions become entangled, clash, and eventually impose themselves on others. The property owners in Linbro Park become targets of violence largely in relation to crimes aimed at their property. For the property owners, crime is something extraordinary which they try to avoid by investing time and money in block watch, security companies and alarms. For the domestic workers, danger is not something which they can avoid in their everyday life by retreating into walled places, rather, they are constantly exposed to it. For domestic workers, crime aimed at their property is only one threat among many others. A larger threat constitutes different forms of physical violence which are directed at them because of their nationality, gender, position in an intimate relationship or because of their occupation as domestic workers. It includes xenophobic violence, rape in public spaces, violence by intimate partners, and assaults by criminals who want to access their employers' property. By looking at how domestic workers are integrated into security governance, one can unravel their silent fights to be recognised not only as security providers and as security threats, but also as victims of manifold forms of violence.

Because their main mode of transport is walking, domestic workers are especially subject to crime than can occur in public spaces, like mugging and rape. According to domestic workers, criminals are said to know their walking routes and payment rhythms. Deserted, unkempt properties in particular are seen as dangerous, because criminals can hide in them and wait for victims. In 2011, a domestic worker was raped on a deserted property, and domestic workers pressed property owners to cut the high grass on that property. Despite the gravity of the crime, the grass was not cleared, as the members of the neighbourhood's security initiative were unable to reach the absent property owner and did not dare to clear the land without his permission. Instead, they recommended that their domestic workers avoid walking in the suburb after dark. This enraged domestic workers:

They tell us"You mustn't go out at night". I tell them: "Icome home from working somewhere, it's night, and the only person I knowwho can give me sugarto make tea,when I don't have sugarin my house, isstaying at that place. So just because you don'twant to clearthis place, I have to sleepwith hunger." They are notright(Zuzile, domesticworker in Linbro Park, May 2012).

The asymmetry of the patrimonial employer–domestic worker entanglements, coupled with the power inequalities between property owners and their tenants, thus profoundly shape the life of domestic workers in the neighbourhood, beyond the realm of work. The domination of this domestic work entanglement continues in the neighbourhood spaces where domestic workers are denied participation in neighbourhood politics and their leisure spaces are suppressed. It continues even into the realm of security governance where domestic workers need to fight for their security needs to be recognised. This shows how powerful and omnipresent the domestic work entanglement is with regard to sociality in Linbro Park: for both the domestic worker and the employer, it is almost impossible to encounter each other in other ways. The asymmetry characterising this form of entanglement makes it in many ways highly ambivalent. While the property owners recognise the existence of this entanglement to some extent, for example by giving domestic workers an important role in security governance, and by supposedly seeing them as part of the family, they also do not recognise it and thus render domestic workers in many ways invisible. The urban sociality emerging through these entanglements, connecting the suburban middle-class lifeworlds with the everyday life of less aff luent urban dwellers in intimate, asymmetric power relations, is deeply ambivalent and ambiguous.

### **Conclusion: Invisible Entanglements in a White Suburb**

Linbro Park is imagined by the white property owners as a white suburb in which their black domestic workers only live by virtue of their work duties, in order to clean their houses and protect their lives. Through manifold practices – suppression of their private lives, house rules, uniforms – domestic workers have become invisible as residents and citizens of the neighbourhood. Practices of invisibilisation do not make the domestic workers disappear, but in the work entanglements they become reduced from full human beings to workers dependent on their employers. This invisibility also has a few advantages, especially for domestic workers without residence permits, as Linbro Park allows them to hide from the police. It is hence important to acknowledge that domestic workers' agency can resist, negotiate and make use of the invisibility and the patrimonial ties to their employers.

In studying domestic workers' lives in neighbourhoods popularly represented as 'aff luent', 'white' or even 'enclaves' by property owners, politicians and academics provide unexpected insights into the social processes and social relations that constitute these spaces. This shows the multiple realities that exist even within a small area like Linbro Park and the diverse, sometimes also contradictory and competing, ways in which suburban spaces are used and imagined. The focus on the forms of entanglements between domestic workers and employers unravels the many practices white property owners engage in to uphold the fantasy of a white, safe world in a rapidly changing city, a fantasy which constitutively depends on the marginalisation, exclusion and invisibilisation of their most intimate co-dwellers, their domestic workers. For the domestic workers, though, the entanglements with their employers are omnipresent, and they deal with what they experience as scarcely bearable asymmetry with practices of hidden resistance and gossip. What also emerges, however, are moments of intersubjectivity shared when drinking tea together, feelings of affection and patrimonial ties. The domestic work entanglements are accompanied by certain registers of interacting and seeing each other, which urban dwellers reactivate in situations outside the employers' homes, consciously or unconsciously. The domestic work entanglements therefore shape urbanity in the city beyond the suburban homes and spaces, and their significance in *cities of entanglements* cannot be overstated.

Acknowledging the importance of the often invisible domestic work entanglements has implications for the way we think about space theoretically: it shows how in enclave-like suburbs, the conceived space (how the space is thought of and represented) and lived space (how the space is used and its everyday meanings) can diverge greatly, and even contradict each other. While the suburb is constructed mentally by the aff luent residents as 'pure', homogenous, and secluded from rest of the city, in reality its spaces are populated every day by the 'other' as domestic workers. For many domestic workers and their families whom I met, this contradiction gives them a certain sense of superiority and control. They may have no money to live like the wealthy, but they nevertheless know the suburbs very well; they know its dirty laundry, in both its figurative and literal sense. The fact that domestic workers are highly aware of these entanglements, while the aff luent residents tend to ignore them, gives the domestic worker a bit more of power in the unequal city.

Louis Wirth, the Chicago scholar, established a view which has long dominated Western urban studies, namely, that urban encounters are characterised by anonymous, 'rational' relationships and social distance (Wirth 1938: 61-63). Simmel argued that this Western, modern 'blasé attitude' resulted from the continuous exposure of the senses to the stimuli of the large city. Förster has criticised Simmel's basic assumption behind this attitude for being "deeply rooted in Western modernity and its history: the individual's claim to autonomy and to independence from social constraints" (Förster 2013b: 238). By hidden forms of resistance, domestic workers aim to reclaim autonomy in the troubling asymmetric relations, yet their profound dependency limits this. The intimate entanglements around domestic work therefore offer a perspective on urban sociality which could not be more in contrast to Simmel's understanding of urbanity. These entanglements are not about independence but rather about deep asymmetry and conf licted dependency. They are not about anonymity in public spaces but about intimate, everyday encounters in private homes.

Social dependence has long been recognised by the political anthropology of Southern Africa as the foundations of personhood (Ferguson 2013). During the expansion of the Ngoni (Zulu) state in the 1820s, people voluntarily subjugated themselves to the new state. Political power was based on wealth in people; it was a society founded not on the relations of exchange between equal individuals but rather on relations of dependence and hierarchy. Through hierarchical dependency, one could become someone, one could achieve social personhood (ibid: 226). According to Ferguson, apartheid and capitalism disrupted the political system, yet it did not break with this socio-political logic. During apartheid, workers travelled from rural areas in Southern Africa to Johannesburg to subjugate themselves to the oppressive system of the mines and farms. On farms, the logic of paternalism, the socio-political logic of social attachment via hierarchical dependence, emerged between white settlers and black workers (ibid: 229, 239). In the mines as well, there were reports that workers and managers understood their relations in quasi-kinship terms, despite all abuse and racism (ibid). Referring to Cape fruit and wine farms, Du Toit has shown that the supposedly

'pre-modern' forms of paternalistic labour relations were not replaced by the purely commoditised worker–employer relations that idealised models of capitalism imply (Du Toit & Ewert 2002: 91-2). Ferguson argues that in the current phase of massive unemployment in South Africa, the biggest fear for the poor is to become totally 'independent'; without any employer to look after them, without any state to provide them with social grants. "It is not dependence but its absence that is really terrifying – the severing of the thread, and the fall into the social void" (Ferguson 2013: 232). While social inequality means that one is hierarchically dependent on an employer, an abusive boyfriend or the state, 'asocial inequality' means that there is no one on whom one is able to make claims (ibid: 233).

This excursion into dependence, inequality and personhood in Southern African societies is important because it tells us a lot about urbanity in contemporary African cities. The lives of rich and poor, black and white, suburb and township in Johannesburg are not simply segregated and divided, as it is often recounted, but through the realm of domestic work they were and continue to be deeply connected and intertwined through patrimonial ties and asymmetric relations of dependence. The city is, therefore, "not a collection of independent realms … but a series of interconnected spaces and processes" (Srivastava 2014: xx). It is through such entanglements that lives unfolds in cities marked by inequality. Urbanity in *cities of entanglements* emerges in these "unexplored terrains of mutuality, wrought from a common, though often coercive and confrontational, experience" (Nuttall 2009: 11). Through entanglements the lifeworlds of suburb and township become "twisted together or entwined, involved with". These entanglements speak "of an intimacy gained, even if it was resisted, or ignored or uninvited" (ibid: 1).

Domestic work is an important urban labour market in cities across the globe but especially in societies with high income inequality (International Labour Office 2013). This is may be one of the reasons why Western urban studies have yet paid little attention to domestic work; it is a female occupation, taking place unseen in the private sphere, moreover, it is not as established in Western cities as it is in the Global South. Analysing the relevance of domestic work for urbanity is therefore a contribution to Southern urbanism (Parnell and Oldfield 2014). It means putting these deeply colonial and postcolonial urban relationships at the centre of understanding urban sociality. As it is especially the poor urban dwellers and not so much the more aff luent who are highly aware of these entanglements, placing these entanglements at the centre of analysis also means privileging the subaltern view, instead of the view of urban elites who imagine their urban world as disconnected and enclosed.

Domestic work is important for urbanity because it is in these quotidian encounters that habits and modes of seeing each other (dispositions) become learnt. Cock in her seminal study pointed out the importance of these everyday encounters for the everyday reproduction of racial attitudes during apartheid.

Of ten the institution of domestic service is the only significant inter-racial contact whites experience, and they experience the relationship in extremely asymmetrical terms. Many white South African children are socialised into the dominant ideological order and learn the attitudes and styles of racial domination from relationship with servants, especially 'nannies'. The reverse is equally true in that many black children experience the inequality of apartheid and the anger it generatesthrough some experience of domesticservice (Cock 1989 [1980]: 3).

In contemporary Johannesburg, domestic work relations are no longer embedded only in what used to be the apartheid black–white categorisation. With the rise of the non‐ white middle class it ceased to be a privilege of white milieus. Because of the low salaries of domestic workers, urban dwellers from aspiring but economically struggling milieus can afford to employ a domestic worker a day per week. There is currently a research gap on these new domestic work relations.

Domestic work in the post-apartheid city is about more than race, and what race means is rapidly changing, yet these intimate encounters continue to be a key site for the formation of urban knowledge – the stock of knowledge by urbanites that shapes how they understand their urban world and the social relations emerging in it. These dispositions inf luence the way in which urban dwellers interact with strangers and how they interpret encounters in other situations; be it at the mall or in religious spaces. Hence, it is of the utmost importance to recognise private spaces, homes and the intimate relations of property owners and their workers as being part of the urban and as exemplary of the everyday entanglements that link people and places in the face of seeming divisions.

© Mark Lewis, N 3 highway, 2015

©Luversan Gerard, Linbro Park, 2012

© Luversan Gerard, Linbro Park, 2012

## **A Politics of Loss?**  The Threat of Public Housing in a Johannesburg Suburb

Anthropology is about large issues in small places (Eriksen 2001b). In this sense, this chapter is about the agency of a relatively small group of urban dwellers in north-eastern Johannesburg, yet the story is also about the larger struggles of South Africa and its citizens in coming to terms with changing spatial, cultural and economic differences in the post-apartheid city. The property owners in Linbro Park have been engaged in a struggle against a public housing project, so-called RDP housing (referring to the Reconstruction and Development Programme introduced in the 1990s), planned to be built by the municipality in Linbro Park. The area is also rapidly transforming because real estate capitalists have their 'greedy eyes' on the suburb, as property owners express it. Hence, the area is transforming from a residential area with small businesses to a mixed-use area with large office parks and warehouses, which not all the property owners are happy about. Many fear that it is only a matter of time before their 'country living in the city' will come to an end. Linbro Park is thus characterised by competition and even clashes between contrasting visions and interests regarding how the larger urban area comprising the peri-urban suburb and the adjacent township should develop.

The Alexandra Renewal Project (ARP) would like to build RDP housing in Linbro Park, so that township dwellers can move to Linbro Park and become home owners here, as part of the larger spatio-political project to de-densify Alexandra. What is at stake here is thus a potential redistribution of land in what used to be a deeply segregated urban area. What is also at stake is that erecting public housing for Alexandra's residents in Linbro Park could lead to a neighbourhood mixed in ethnic and economic terms. Township dwellers and suburbanites may become neighbours, they may rub shoulders in the neighbourhood streets, and share a social position in terms of owning property.

In an ideal world, one might hope that such a mixed neighbourhood would lead to new forms of being together in the post-apartheid city, that it could turn into a stage where new identities and new relationships become formed, leaving the binary categories of apartheid around race and suburb–township behind. Reality, unfortunately, is more complicated. There are many social processes in place which strongly complicate, hold back, or even obstruct the emergence of something new, be it new forms of sociality, new ways of seeing the world and new spatial redistribution. There is the insufficient capacity of the local government to actually implement the large public housing programme, there is the strong competition by private real estate investors for urban land, and, most importantly, there are the suburban property owners of whom most are opposed to this transformation. Over the last two or three decades they have tried to obstruct, slow down, or at least inf luence the way in which the public housing programme is implemented, ranging from opposing densities in an Urban Design Framework to attempting to push land prices upwards. Instead of envisioning a less divided city, they argue against the public housing project with narratives based on simple binaries of Johannesburg as deeply divided in economic and cultural terms.<sup>1</sup>

The property owners' agency is driven by a fear of loss, that is, losing their properties, their wealth and their lifestyle. Imagining daily encounters with Alexandra residents as neighbours makes the property owners feel threatened as material and social beings. In what I consequently call their *politics of loss*, they struggle, on the one hand, for a Linbro Park as they imagined it to have been in the nostalgic past, namely as a peaceful, cohesive, orderly and implicitly white suburb. They do this by drawing boundaries between themselves and the Alexandra residents, constituting them as 'others'. On the other hand, they fear losing money, as they believe property values may decrease when they have to share their neighbourhood with residents with lower incomes. They are aware, however, that they are losing the battle. Their opposition is a fight of David against Goliath, not least because within the post-apartheid 'rainbow' nation the property owners' stance against public housing for the poor is morally troubling and their inf luence with regard to the larger socio-spatial processes at stake is limited. For this reason, many property owners have already sold their land and left.

Politics aimed at preserving privileges, keeping others out and, therefore, disentangling from them have taken centre stage in national politics across the globe in recent years. In 2016, the UK electorate voted to withdraw from the European Union (Brexit). Donald Trump was voted in as president of the United States with his promise of isolationist politics and the building of a wall on the border with Mexico. There is a rise in ultra-right movements in Europe like the Alternative for Germany (AfD). Last but not least, xenophobia is taking centre stage in South African politics. Conservative forces are on the rise, and hence there is a need to pay more attention to anti-social movements, to politics of loss and politics of disentanglement.

Anti-social movements in African cities have received little attention till now. If they are mentioned, they appear as rather opaque groups: the corrupt elites, the middle-class buyers of gated homes, or the angry xenophobic mobs. This may be related to a sort of overcompensation in African urbanism. Academic and political discourses on African cities have long been dominated by a narrative of doom, African cities tended to be seen as "as examples of all that can go wrong with urbanism" (Myers 2011: 4). Not least to counter this, urbanists have started to draw attention to the endless agentic possibilities that African cities offer and have focused on how urban dwellers make the urban work, based on their agency and creativity (e.g. Simone 2004a). Focusing on urban politics and governance 'from below' in African cities, researchers have made many important contributions, not only to Southern urbanism but to urban studies in

<sup>1</sup>  As in all the chapters in this book, this part fieldwork was also conducted between 2010 and 2012. Linbro Park haschanged considerably since then and many of the interviewees have sold their properties and moved elsewhere.

general (Parnell 2014, Parnell and Oldfield 2014). Yet the overcompensation for the narrative of doom has led to a focus on urban movements that fight for *more* social justice, for example the right to the city movements or service delivery protests in South Africa. Conservative groups whose political agency is driven by the fear of losing something have received less attention.

Linbro Park was once a partially enclosed neighbourhood; the residents paid for a large concrete wall to be built along the eastern boundary in order to stop people from coming into the area from the adjacent open field. Some people wished to have a wall around the whole suburb but struggled to get enough fellow property owners to pay for it. Parts of the wall have since been removed when some properties were repurposed from residential to industrial use. In South Africa and elsewhere, enclosed neighbourhoods and gated communities have sprung up as forms of dwelling to which privileged urban groups retreat (Dirsuweit 2015, Young 2000).

Although Linbro Park does not exactly qualify as an 'enclosed' neighbourhood, as most of the wall has gone, many of the social processes at stake in the construction of such enclosed suburbs and gated communities are also present in Linbro Park. Jennifer Robinson warns us against treating cities as embodiments of abstract types because this limits the reach of comparisons and the diversity of features that we pay attention to. Rather, we should treat them as 'ordinary' and "attend to the diversity and complexity of all cities" (Robinson 2006a: 1). The *ordinary city* approach is also useful for studying changing neighbourhoods. Instead of categorising, ranking and labelling Linbro Park, I suggest it should be seen rather as a diverse, differentiated, contested neighbourhood shaped by a "multiplicity of trajectories of processes" (Massey 2006: 92). Neighbourhood change is the result of multiple actors trying to shape urban space according to their own images and visions. How Linbro Park evolves is thus a "product of political decisions and collective actions in space", and by analysing this we can unravel how agency shapes the effects of larger structural processes (Thompson 2017: 105, 107). In the *politics of loss* the property owners of Linbro Park shape their entanglements in the city by drawing boundaries, yet the state of disentanglement they strive for remains an ideal rather than a reality.

### **Linbro Park – A Changing Suburb**

Once upon a time, Linbro Park was a plantation, owned by the surveyor, farmer and business man Edwin James Brolin, who was of Scandinavian descent (Louw 1981: 14). Probably in the 1920s, he bought the land on what used to be the Modderfontein farm in the north-east of Johannesburg. He planned to supply the mining industry with mine‐props (wooden poles for propping up the mine) but the business idea failed and he decided to subdivide his farms into smaller plots, probably in the 1930s. He gave the area the name 'Linbro Park', as a word play on the family name Brolin. Some of the land he gave to his children who lived there with their families. Other plots he sold (ibid: 14‐15). In 2010, the area had about 228 separate plots (City of Johannesburg 2010). A cemetery, an unkempt plot with scattered graves where the Brolins were buried, reminds residents of the founders of the area.

The property-owning families in Linbro Park have their roots in various European countries, like Germany, Austria, England, Greece and others. They or their ancestors had moved to South Africa from Europe in immigration waves since the 19th century. Many of the contemporary owners grew up on farms in Southern Africa, predominantly in South Africa, Namibia and Zimbabwe. Many relate to such rural roots when explaining their choice to move to Linbro Park.

I grew up in Namibia on a farm … Linbro Park was country, there was nothing here [when we moved here]. This was agricultural, it was the closest that we could get to farming. My husband had a business, and also I wasn't a farmer who wanted to farm with cattle, but I just needed the space, the outdoors (Clara, property owner in Linbro Park, April 2012).

The property owners' ideas of the good life are still shaped by their childhood experiences which they speak about nostalgically. Living in Linbro Park enables them to live a lifestyle close to what they call 'nature' and 'outdoors: hence 'country living in the city'. One resident coined the slogan *'*country living in the city' to describe their way of life, an expression which became widely used by other residents. 'Country living in the city' as an emic notion speaks about Linbro Park as 'lived space', as a space "directly *lived* through its associated images and symbols, and hence the space of the 'inhabitants' and 'users'" (Lefebvre 1996 [1974]: 39, italics in original). The 'in the city' is important, as the property owners do not identify themselves as rural, not least because they all have urban-based occupations and shop in the surrounding malls.

Barbara: What made you come here? Steve: You must understand, Iwas born on a farm, raised on a farm. I have been a farmer's son. To get a little box on top another is not my scene. And I see it, when I go to Germany, I see how there is no space. I come back to Africa and I think 'oh, we are very lucky' (Steve, property owner in Linbro Park, June 2012).

For 60-year-old Steve, a romantic idea of 'African' space associated with freedom and nature is an important reason why he continues to live in South Africa even though many of his friends have emigrated since the end of apartheid. Living in Linbro Park is hence part of a lifestyle that links them to their settler ancestry. The space is crucial to their place of being in and belonging to (South) Africa.

The property owners nostalgically remember the times of apartheid, which for them was the time before the neighbourhood started to change and when there still was cohesion and order. According to them, there was little crime and the fences around the properties were low. The property owners and their children spent their leisure time riding horses and breeding dogs; property owners recount that there was an intense associational life around horse riding. Many of them explain that they bought a property in Linbro Park because it was large enough to keep horses for themselves and their children.

Barbara: Why did you moved to Linbro? Sandra: Because of the horses. My children were into horse riding. And I just loved living here, I love the quiet … Linbro is home to many people who were into horse riding and competing, who could have horses at home, but didn't have to farm. Even the library used to have a hitching point.One could tie the horse there. Istill have a picture of my daughter and her tied up pony in front of the library when she went to get a book for school. It was lovely. But as they say, progress comes along, and all your good living goes down the window (Sandra, property owner in Linbro Park, May 2012).

Looking after a horse, riding out on horseback and participating in competitions also entailed socialising with neighbours. This was to such an extent that horse-riding became an important element of the notion of 'country living in the city', a part of the collective and individual identity of Linbro Park property owners.2 While the youth of Alexandra threw stones and Molotov cocktails against the apartheid regime's military tanks, the youth of Linbro Park rode to the local library on horseback. My interviewees never brought up apartheid in their nostalgic narratives and when I asked how it was to live near Alexandra township, where political violence was rampant during the 1980s and 1990s, answers were usually short. She responded: "It was no hassle at all, no, no, not hassle at all" (Sandra, May 2012). In these nostalgic constructions of the past, the political-ideological context of apartheid is not present and their privileged, safe lives appear normal to them.

The times of the imagined rural idyll have passed, and Linbro Park is now seen as a changing neighbourhood by politicians, by town planners and by residents. The Urban Design Frameworks describes Linbro Park as peri-urban, as "an area in transition from rural to urban" (City of Johannesburg 2010: 5). Thomas, a Democratic Alliance (DA) politician in the ward that Linbro Park belongs to, represented it as an exceptional place:

That's a suburb in transition, and probably one of the only ones in Jo'burg at this time. When you say 'transition', we talk about a changing heartbeat. Between what used to be ten years ago to what's going to be in ten years' time, there's a scary dif ference, two extremes. Thatcan't be very comforting to those who chose their retirement in Linbro Park (Thomas, politician, September 2012).

Also, the property owners of Linbro Park perceive that their neighbourhood has changed considerably from what it used to be. They anticipate that further change will come, largely induced by outside forces that they are unable to inf luence, yet it is it is hard for them to estimate whether this will be in the next five, ten or even 20 years. Dave, a 60-year-old committee member of the LPCA, said "It's a nice place to come home to. I love my house, my home, my neighbourhood, and I am sorry to see it change, but I know the change must come" (Dave, June 2012).

Until the early 1990s, Linbro Park was a largely residential neighbourhood dominated by *lived space*, used by its inhabitants as homes and as a setting for their lifestyle. However, at the end of the 1980s, actors with different interests and intentionalities came onto the scene. Mr Silva, owner of a truck rental company and a European immigrant, started to buy up properties in Linbro Park. The land was cheap, and he saw its

<sup>2</sup>  The roots of horse-riding asleisure in Africa can be traced back historically to Victorian Englandwhere itwas a sport ofthewealthy and the elite, and a sign ofsocial distinction. The Victorian enthusiasm for sports and game like cricket, tennis, golf and rifle shooting wasspread throughout the British empire by British administrators, missionaries, traders and settlers (Stoddart 1988: 654). For colonial culture, sport played an important role in terms of determining and exhibiting the ranking of social groups (Stoddart1988: 657-8).

potential owing to its location close to the highway. One property he used as his home, another one he used for his truck rental business, and others he bought as investments. Over the years, he acquired several properties. Like him, other property owners started to keep businesses on their properties, so that the area transformed from being a purely residential place to a mixed-use suburb. The neighbourhood increasingly came to be viewed by other family-based business entrepreneurs as an economic resource. Gabriel, a 30-year old member of the Silva property development family told me: "For us in Linbro, it's mostly monetary. And of course, we are a business there, so we have this interest in our neighbourhood. But mostly for us, we see it as a business opportunity. We always try to maximise our value" (Gabriel, May 2012).

Hence, a new rationality became important: a relationship to urban land that Lefebvre suggests should be referred to as *abstract space*: a space of instrumental, profit-seeking rationality, a commoditised space that serves the economic interests of the investors. This made Linbro Park more diverse in terms of property use. It was now home to what Aalbers (2006) calls *abstract space-maker*s, who "instrumentalise space for the production of exchange value" (Thompson 2017: 106), living side by side with *social space-makers* whose relationship to urban land is mainly based on lived space and the use value of space.

With the presence of businesses in the area, and especially with the arrival of professional property developers a couple of years later, Linbro Park's spaces as lived, perceived and conceived realities changed. For the long-term residents, the haptic, olfactory, auditory and visual experiences of the neighbourhood were affected. New buildings of a light-industrial and commercial character emerged which residents perceived as visually disturbing among the plots of green spaces and the suburban houses. The sensual habits of the urban dwellers were also disrupted by the noise resulting from the increase in traffic and the noises and smells from small-scale factories. Increasingly, horse riding along the streets became dangerous because there was more traffic, and the land on the northern fringe, which was originally used for horseback riding, was transformed into the Linbro Business Park in the late 1990s. Older property owners were deeply opposed to the transformation into a mixed-use area, and conf lict emerged between the *social* and *abstract space-makers*. The property owner, Baldwin, recounted that that the opposing factions even got into physical fights, damaging each other's fences.

As Linbro Park was zoned 'agricultural', the changes in the use of space from residential to commercial were illegal, as the zoning did not actually permit commercial or light-industrial businesses. This change in land use took place at a time in South Africa when the newly elected post-apartheid government had other worries to the enforcement of urban by-laws. During the 1990s many South African cities and many parts of Johannesburg saw "such a shift to multiple land uses despite the continuation of increasingly irrelevant zoning schemes" (Mabin 1995: 192). It was a time of weak public authority, when the so-called decay of the inner city also took place. Residents lodged complaints with the new local government and approached the Public Protector, arguing that the state was obliged to prevent the illegal use of private property. The local government responded that it could not enforce the by-laws because it lacked the necessary resources (Public Protector 1998: 20-21).

In the nostalgic narratives of older residents, the imagined homogenous idyllic community was destroyed by the question of what the future of Linbro Park should

look like. The conf lict between the faction of the property owners who wanted to keep the neighbourhood as it was and the other faction who wanted the suburb to become more business oriented had disruptive effects on neighbour relations in the suburb. A sense of a *divided community* with a lack of internal cohesion emerged and exists to this day, as the property owner, Clara, explained to me: "The community is very divided. Everyone has their own criteria of what should happen" (Clara, April 2012). Many property owners today see the lack of what they call cohesion as an impediment to their ability to shape the neighbourhood's future. Some even claim that it slowed down development and kept property prices low.

In the 1990s a third rationality, a third *spatial project* claiming space in Linbro Park became relevant: the government, more precisely the ARP, began to plan the construction of low-cost housing for the residents of Alexandra in the neighbourhood.3 Peter, a leader of the communist youth league in Alexandra, expressed it like this: "It's just farmland, you find somebody living in 10 hectares alone, and there is nothing. And you know, within that space you could at least put 50 houses, take 50 families to live there" (Peter, February 2011). Because of the large differences in density – the densely populated township and the agricultural holdings – existing side by side, the idea that the land could be used to alleviate the housing problem in Alexandra is not farfetched. The spatial inequality is seen as a problem by many, be they Alexandra residents like the leader of the communist youth league in the quote above, government officials or even Linbro Park property owners like Andrew, an LPCA committee member: "The realities of life is we have a lot of land that is not being utilised to its fullest potential, and there is other areas that need to de-densify, to try and get other land" (Andrew, March 2011).

Already in the 1990s rumours had emerged that the local government was planning to use land in Linbro Park for public housing. The plans for the construction of low-income housing in Linbro Park became more concrete when the ARP, a larger urban renewal initiative, was started. The R1.3 billion (about 100 million CHF) ARP was a f lagship project with considerable financial, managerial and political support from the African National Congress (ANC) at national, provincial and local level (Sinwell 2010: 30-31).4 The most visible and, from the perspective of the Alexandra inhabitants, the most important mandate of the ARP was and continues to be public housing. The key element of South Africa's housing policy is an income-related capital subsidy used to purchase land, secure tenure, deliver the necessary services and construct a basic house for households that meet the subsidy criteria. The subsidy was introduced in 1994 and has been increased intermittently. There are several forms in which the subsidy has been put in place, but the most well-known and most visible form is the 'RDP house' (Charlton and Kihato 2006: 254), also referred to as 'low-cost housing'. 'Affordable housing', a further category in the housing policy, refers to subsidised bond houses aimed at buyers in the upper end of the low-income market.

<sup>3</sup>  Spatial projects are "coordinated, continuous, collective campaigns to produce and format space according to identifiable logics and strategic goals, pursued by specific actors utilizing particular techniques" (Madden 2014: 480).

<sup>4</sup>  The ARP was long an intergovernmental entity and part of a national Renewal Programme. In 2014/2015 the ARP became mergedwith the JohannesburgDevelopment Agency. Asthis merging took place af terthe time this ethnography refersto (data collection 2012–2012),this book continuesto refer to it asthe ARP.

Over the years, there has been a shift in the way the government, locally represented by the ARP, imagined the future of Linbro Park, a shift ref lecting changes in national housing policy. The first plans were to develop the whole neighbourhood into a low-cost housing settlement. This ref lected the first housing policy in the post-apartheid era that focused on the construction of such mono-functional RDP settlements in order to rapidly meet quantitative delivery goals (ibid: 253). In the early years of the urban renewal project, the ARP constructed 3256 RDP housing units in the township Bram Fischerville (close to Soweto) and 3500 in Diepsloot (a post-apartheid township in the north of the city). Such relocations to distant and monofunctional settlements were increasingly criticised.

We made a mistake in our housing policy since 1994, we should never put housing for the poor, to live on the fringes of the city, on large, cheap land pockets, because that was the motivation, because they said, "We need to accommodate for as many people, where is the cheapest land?" Where is the cheapest land? At the outskirts of the cities. But for the poor person, it's the worst place to be (Ron, executive employee at ARP, March 2011).

At the urban fringes, land was available and cheapest, yet the far-off location of public housing settlements and the lack of transport made life and economic survival harsh (see also Todes 2006: 64). The national housing policy became criticised for contributing to urban sprawl, perpetuating the marginalisation of the poor, and failing to contribute to the integration of the city (Charlton and Kihato 2006: 255). The RDP settlements were also criticised for being reminiscent of the 'matchbox' houses delivered during the apartheid period (Dangor 1998: 359-361).

In 2004, a new housing policy, called Breaking New Ground (BNG), was introduced, which entailed a major paradigm shift in the conceptualisation of the ideal public housing settlement: housing provision should now address poverty more broadly, improve the quality of life, stimulate economic growth, generate assets for the poor and ultimately develop sustainable human settlements (Charlton and Kihato 2006: 257, Department of Housing 2004). In addition, the ARP changed its housing approach: instead of the large-scale relocation of Alexandra residents to faraway areas, much like the policies of the apartheid state, a decision was taken to construct houses in Alexandra and in the vicinity of Alexandra. Accordingly, considerable pockets of land between Alexandra (East Bank) and the highway N3 were transformed by the ARP into RDP settlements, nowadays called Far East Bank. In its first 10 years, the ARP constructed 14,500 residential units (Ron, executive employee at ARP, March 2011). However, it was obvious that for rest of the 21,500 targeted households (numbers from 2008) land outside Alexandra had to be found. With an average density of 125 units per hectare, 90 hectares of land in the vicinity of Alexandra were needed (City of Johannesburg 2008: 17). In the search for land, several pockets of undeveloped land in the vicinity were identified: Frankenveld, Modderfontein, land of the Islamic Trust, as well as Linbro Park. The ARP then conceptualised the future for Linbro Park based on the new BNG housing policy. The main aim would be to develop Linbro Park as a 'sustainable human settlement' in order to accommodate "the housing backlog and future growth of Alexandra" (ibid: 39). Sustainability is a key notion in the BNG policy and the ARP planning documents for Linbro Park, as South African housing policies were

developed in the context of international discourses (Charlton and Kihato 2006: 257, Todes 2006: 64). The idea of sustainable human settlements entered policy discourse at the end of the 1990s and stems from the UN Habitat Agenda. In the BNG policy, 'sustainable human settlements' refer to neighbourhoods where "economic growth and social development are in balance with the carrying capacity of the natural systems" (Department of Housing 2004: 12), thus, understanding sustainability in an economic, social and environmental sense.

In order to address the spatial fragmentation and inequality inherited from the apartheid era, the new housing policy envisages the acquisition, and if necessary, the expropriation of privately owned, well-located land (ibid: 14). This is the broader political context within which the Linbro Park property owners' *politics of loss* emerges.

### **Opposing State Intervention through Tools of Abstraction**

As early as the 1990s, Linbro Park residents heard rumours about the ANC government's plans for a public housing project in the neighbourhood. Only in 2013, however, did the local government acquire its first pieces of land in Linbro Park and it will take another couple of years to complete the project. During fieldwork (2010–2012), apart from a few exceptions, the property owners were fiercely opposed to this state intervention.

States have a particular way of seeing and acting with regard their citizens and spaces. Making the environment and its people legible is a central problem of statecraft (Scott 1999: pos. 117). It is through *conceived space*, the space constituted through mental activity and inscribed in powerful documents, that states 'see' their populations, shape spatial reality and draw boundaries (ibid). For Lefebvre (1996 [1974]: 38), conceived space is "the space of scientists, planners, urbanists, technocratic subdividers and social engineers, as of a certain type of artist with a scientific bent – all of who identify what is lived and what is perceived with what is conceived". In order to 'see' and act, the state depends on 'tools of abstraction' which reduce the complexity of life to abstract categories. These state simplifications like title deeds, zoning and cadastral maps are like abridged maps, which represent only that slice of complex reality that is of interest to the state (Scott 1999: pos. 134). The land zoning in Linbro Park is such a tool of abstraction. It is a state instrument to prescribe and control land use. It is both descriptive and prescriptive: it delivers state information about what type of land use is practised on a certain property and, by changing zoning law, the state can change which type of land use is permitted. Zoning was important for implementing spatial segregation during apartheid: The Group Areas Act (1950), the key legislation designating different areas to different racial groups, was implemented through rezoning.

The discourses, categories and processes imposed by the state can become appropriated by urban dwellers and other actors, and they can instrumentalise them to project their own visions onto space. As zoning is about drawing boundaries per se, urban dwellers can use them to draw boundaries which are in their own interests. As mentioned, the official zoning in Linbro Park prohibited business activities, hence since the 1990s, some business-oriented families and potential investors in Linbro Park have pushed for new town planning documents to be drawn up; the new plans should record, legalise and enable their visions for the future of the suburb. In the 1990s and

early 2000s, different factions paid town planning offices to create precinct plans and reports, none of which were formally approved by the Council and therefore failed to become legally binding (City of Johannesburg 2008: 24-25). In 1998, the 'business' faction achieved an important victory: the first Integrated Development Framework was drawn up for Linbro Park (prepared by Setplan town planners), envisioning mixedland use with a business node, industry, commercial and residential uses. Many residents did not like this framework, as it envisaged an end to their 'country living in the city', yet it was nevertheless adopted by the City. This was for many Linbro Park property owners one of the important battles lost in the *politics of loss.*

In order to represent their interests in these highly technical, power-shaped processes, the Linbro Park property owners acquired specialist knowledge about the legal and political processes, including urban planning instruments. A considerable part of community meetings and LPCA committee meetings is dedicated to the distribution of such knowledge by those who are better informed. Some residents like Amy, a 50-year-old property owner, refrained from participating in these meetings because of the technical knowledge needed: "I am a farmyard lady; I am not into the technical side as such … It is too technical for me, to be honest" (Amy, May 2012). Hierarchies of technical knowledge shape the *politics of loss* and the government–citizen interface in participation processes.

Citizens in post-apartheid South Africa have a constitutionally entrenched right to participate in development planning in local government. The authoritarian apartheid state saw its citizens only as objects of planning; the post-apartheid constitution seeks to put the lives of ordinary people at the provenance of planning (Williams 2006: 200- 201). Historically, the call for participation grew out of the liberation struggles in the politically suppressed townships, but nowadays privileged groups can also make use of it in their politics of loss.

Although the plans by the local government for an RDP settlement in Linbro Park existed have since the 1990s, only in 2008/2009 were they written down in the Regional Spatial Development Framework (SDF, City of Johannesburg 2008/2009). The Linbro Park property owners took part in the public participation process; according to Ron, executive employee at ARP, they achieved a compromise with the residents. Linbro Park should not entirely become an RDP settlement, just a small section of it (5000 instead of originally planned 20,000 housing units). This also ref lected the change in housing policy, however, and the ARP's limited financial capacity (see above).

Finally, the SDF officially earmarked Linbro Park for 'sustainable human settlement' development, envisaging it as 'ripe for development' and found that the area should be 'unlocked' (City of Johannesburg 2008/2009, City of Johannesburg 2008: 38). The vision of the SDF was "to create an integrated, sustainable neighbourhood through infill development on well-located land within Linbro Park" (ibid: 49), envisioning it as a mixed-use and mixed-income development with the idea that the area would "benefit the city as a whole and not just specific groups" (ibid: 47) and include the provision of social and economic infrastructure. This aspect of the BNG policy is often referred to with the notions of integration or inclusion (see Haferburg 2013).

As the next step, the Gauteng Provincial Department of Housing appointed a consortium to develop an Urban Design Framework (UDF), which is a refinement of the SDF. An Urban Design Steering Committee was established, comprising spatial professionals (town planners, engineers), representatives of the ARP, a representative

from the major private developer in the area (Intrapop) and a few property owners who had to be nominated by the residents (Bigen Africa/ADA Urban Design 2009: 7). The steering committee was tasked with developing more precise guidelines for zoning. The Linbro Park property owners could voice their concerns: "At the end of the day, we were quite happy with the framework that was presented. Except: densities" (Dave, LPCA committee member, June 2012).

One key site of contention emerged, the question of residential density. Density is one of the many tools of abstraction used by the state to regulate land use. It prescribes how many dwelling units are permitted per hectare in a certain space. For the local government, high was important for the housing project because of the lack of urban land and the high costs. High density also symbolises the political will to counter urban sprawl and spatial fragmentation. High density is seen as more energy efficient and the provision of urban infrastructure is proportionally less expensive. "Housing density is of central importance to sustainable urban form" (City of Johannesburg 2010: 21). As such, in the UDF the local government pushed for a density of 150 units per hectare.

For the property owners, high density invoked images of squatter settlements and badly maintained high-rise buildings where poor (black) people live. As Andrew, LPCA committee member, explained: "People believe – which is not necessarily correct – the higher the density, the lower the standard of the development" (Andrew, March 2011). For the property owners, low density symbolises their 'country living in the city' lifestyle, and they believe that that the lower the density, the wealthier the residents are. By resisting higher densities, they aimed to change the ARP's plans to provide fully subsidised RDP housing to building subsidised bond houses targeted at the better-off poor, which they considered more tolerable as neighbours (discussed in more detail later in this chapter).

In a community meeting the property owners decided that "the community wants 75 [units per hectare] maximum" (Andrew, March 2011). In order to advocate this position, they appropriated the discourse of the new public housing policy, the BNG, and the notion of sustainability. They argued that it would be 'unsustainable' to have such high densities in the suburb because of the increase in crime and because the public infrastructure would become overloaded. At a meeting about creating a City Improvement District (CID), an exasperated resident explained: 'That was our whole argument, sustainable housing! (meeting about CID, June 2012). However, no agreement between the ARP and the property owners was reached. "Now it's almost like a stalemate. But we will just have to drive it through" (Ron, executive employee at ARP, March 2011). Hence, in 2011, the local government had the plan adopted by the city council, with the densities that the ARP desired against the will of the property owners. In the context of the political pressure on the ARP to deliver high numbers of housing, the property owners' opposition to high densities, part of the *politics of loss*, was unsuccessful.

### **Economic Aspirations – Replacing Lived Space with Abstract Space?**

In Johannesburg, well-located, affordable urban land is a scarce resource. In this sense, a 60-year-old LPCA committee member once said in an interview: "Linbro Park is the best kept secret in Johannesburg" (Dave, June 2012). Strategically located between the Johannesburg CBD, Pretoria and the OR Tambo International Airport, the Gauteng

Spatial Development Framework from 2010 described Linbro Park as an area where a large part of future Gauteng development was expected to take place (City of Johannesburg 2010: 5).

The strategic significance of Linbro Park was only made clearer by the multiple new real estate and infrastructural developments emerging around it. Among them are a train station for the new commuter rail system, the Gautrain, and two examples of 'privatised urbanism' (Herbert and Murray 2015). the construction of Waterfall City, a mixed-use development, with the massive Mall of Africa which was opened in 2017 in nearby Midrand, and the planned megaproject on land which once belonged to the dynamite factory AECI in Modderfontein (Ballard et al. 2017). Since about 2006, professional property developers, though of a smaller scale than in the two cases mentioned above, have also changed Linbro Park. A large property developing company developed a major warehouse and other commercial properties on about twenty plots on the southern fringe of Linbro Park. The company also played an important role for the extension of public infrastructure in the area. Many other developers have bought land around the area and have started building or are still planning to do so.

The planning processes and land sales related to these diverse projects were closely observed by Linbro Park residents and other stakeholders and fuelled fantasies and hopes about the economic potential of Linbro Park land. Many property owners started to accept that they would not be able to defend their rural idyll against the plans of the government and the property developers, and their intentions and priorities started to shift. Instead of seeing their homes mainly in terms of everyday use, social meanings and memories, hence as *lived space*, many started to see their land in terms of its exchange value, hence as *abstract space*. Recognising that their attempts to save their 'country living in the city' was a lost cause, they redirected their *politics of loss* towards the economic dimension of their land. The 50-year-old property owner Amy told me, as did many others: "We want to get the best prices for our properties" (Amy, May 2012). This desire to at least make the highest profit possible out of the sale of the land constitutes a shift in the *politics of loss*, namely, a shift from a relationship to land based on its use value (*social space-makers)* to a relationship based on its exchange value (*abstract space-maker*s) (see below).

Some property owners hoped that because of the demand by property developers, the land in Linbro Park would become so expensive that the local government would not be capable of buying it for public housing. However, investors found it easier to develop areas like Modderfontein or Waterfall, where the land belonged to one property owner. In Linbro Park, as Anthony, a member of a property developing family, explained, ownership was split between multiple people, "all with different motives and interests" (Anthony, May 2012). In 2011, a group of property owners therefore attempted to sell their approximately 20 properties as a packet to big developers. As Andries, a former committee member of the LPCA, explained: "

What we are trying to do is we are trying to get willing people together, under the banner ofthe framework ofthe development plan, and get developersin to fund thewhole development,so thatitis economically not viable for governmentto build (Andries, formercommittee member of the LPCA, March 2011).

Other residents tried to act as mediators between real estate investors and potentially interested sellers in the neighbourhood. As such, they attempted to create competition between the private developers and local government for urban land; a competition they assumed the local government would lose. However, at the time of research, the plan to sell a group of plots to private developers did not work out. The land prices in Linbro Park were low, and demand by private developers to buy land in the neighbourhood barely existed. This affected not only the residents who engaged in this plan of diverting urban development towards private city buildings, but also residents who needed to sell their land for other reasons. Clara, a 60-year-old property owner, had experienced a business bankruptcy, and was suffering because of the difficulty selling her property:

We need to sell, becausewe can't af ford to live here, butwe are notselling, because the buyers are notready.…We had reallyNOinterestin three years and Ireally spent a lot of time on computer, on the telephone, and I got it on my website (Clara, property owner in Linbro Park, April 2012).

To the residents, the property market forces appear to be relatively obscure outside forces not accessible to immediate experience. The residents employed various theories to explain the lack of demand for their land: most importantly, they blamed the ARP, and said that no one wanted to buy land in an area where low cost housing will be built. However, others speculated differently: some argued it was because of the lack of infrastructure like public sewage and because the land was still zoned as agricultural. Others speculated that interest was low because the area was seen as crime ridden and the large properties were expensive and difficult to secure. Others made a link between the low demand and the world recession. Some then also blamed the Linbro Park residents themselves and their internal 'dividedness' and infights which they believed scared developers away. The *politics of loss* hence takes place in a context where larger forces are at play that are not always accessible to experience and hence are subject to many speculations.

In reality, the purchase offers by the ARP which arrived in 2012 were for some property owners the first offers they had received in years. In 2011, the ARP received funds from the national government for the purchase of properties in Linbro Park. In 2012, the Johannesburg Property Company (JPC), the local government branch responsible for land acquisitions, made purchase offers to about twelve property owners on the western boundaries of the neighbourhood (Heer 2018). In these purchase offers, the JPC also made clear that the government would be ready to expropriate if no agreement could be found on the 'willing seller–willing buyer' basis.<sup>5</sup>

<sup>5</sup>  The current South African land redistribution policy favours the 'willing seller, willing buyer' principle based on the market and does not prioritise expropriation. The expropriation of privately owned, well-located land is only envisaged as a finalresort by the newpublic housing policy BNG(Department of Housing 2004: 14). The Constitution recognises the government's right to expropriation if the land should serve "for a public purpose or in the public interest" and if the previous owner is compensated (Ntsebeza 2007:117). The compensation hasto be based on the current use of the property, the history of acquisition, the purpose of expropriation, and on the market value of the property" (ibid:117, Republic of South Africa 1996). In the new Constitution of 1996, the protection of existing property rights, acquired during colonialism and apartheid,was entrenched (the 'property clause', ibid:108).Ntsebeza

This 'expropriation threat' by the ARP, as the residents called it, caused alarm among property owners. The LPCA committee decided to call the affected property owners to a meeting to obtain information from them about what was going on and to empower them to deal in informed ways with the 'threat'. Their biggest fear was that one of the property owners would negotiate with the ARP and sell for a low price because they were scared. They feared that this would lead to a chain reaction and make property prices go down in the whole area (see below).

In May 2012, five property owners who had received purchase offers by the ARP appeared at the meeting. The LPCA committee members first informed the property owners about the legal situation. They explained that when no agreement could be reached on the willing seller–willing buyer basis, the JPC would send an expropriation notice. They recommended that when they received an expropriation notice, they should come to the LPCA and get themselves an attorney. One of them, Dave, said: "If you do not appeal within 60 days, the expropriation becomes valid. Please don't panic, you can fight for ten years" (Dave, meeting on expropriation, May 2012).

The key argument of the LPCA committee was that the property owners should not agree to sell, but they should 'take the expropriation route'.<sup>6</sup> Dave then explained that as soon as the property owner received an expropriation notice along with a compensation offer, all rights to the properties are immediately terminated. Yet the expropriated owners can then start legal proceedings regarding the compensation amount, a process that can drag on for years. The expropriated property owners must prove that market prices are higher than what the government is offering as compensation. In the end, the compensation would be higher than the price the local government offered them on the 'willing buyer–willing seller' basis, not least because the financial loss resulting from expropriation would be compensated. So, Dave explained to the group, "it's financially a better deal to wait for expropriation" (Dave, meeting on expropriation, May 2012). The LPCA committee members promised to assist the affected property owners in this legal fight.

Despite the committee members' explanations, one of the property owners, John, explained that he had want to sell his plot for a long time but had never found a buyer. He therefore entered into negotiation with the JPC, the local government branch responsible for land acquisitions. He had found them very accessible and learnt that they wanted to avoid expropriation because it was complicated and expensive. He concluded: "I want to sell" (John, meeting on expropriation, May 2012). John's statement provoked a heated debate. One of the LPCA committee member's face turned red with anger and he shouted: "If you sell to ARP, everyone else is doomed. Please don't sell to them! You gonna ruin all our chances. Have patience. Expropriation will be better for you than willing buyer-seller" (meeting on expropriation, May 2012).

In this debate, and also in the following weeks and months, property owners in Linbro Park represented John as a traitor, a 'sell out' to the 'community', as they believed that his selling would make property prices drop in the whole neighbour-

argues that this was due to the relatively powerful position of the National Party in the negotiation process. Other reasons may also be that some strands within the ANC and the alliance partners were pro-capitalism (Ntsebeza 2007:116-117). Expropriation law has become a political issue in South Africa in recent years.

<sup>6</sup>  The committee members had read the MA thesis by Breedt (2009).

hood. In the debate that evening, it was suggested that perhaps LPCA members could put their money together and make John a better offer than the ARP. One committee member said: "It's a basic thing, give us a figure and give us a chance to purchase it. If we don't come up with an offer within 14 days or so, you sell." One of the affected property owners did not think this would work: "That's crazy, you don't have so much money, you can't buy up all of them" (meeting on expropriation, May 2012). The idea of making a counter-offer nevertheless calmed heated emotions for the rest of the meeting, yet it was never put in place and John eventually sold his property to the ARP. By 2014, the ARP was able to acquire a handful of properties in Linbro Park on the willing buyer–willing seller basis. Other property owners, however, 'took the expropriation route', engaging in long legal processes.

### **Conceiving a City of Threatening Encounters**

The current expropriation act stems from 1975, from the pre-ANC government. Attempts to introduce a new act that strengthens the government's ability to expropriate have failed (2008, 2013). Land reform has long been debated in South Africa as a rural-agricultural rather than as an urban issue (Beyers 2013, Brown-Luthango 2010: 124). It is housing policy that is expected to address the difficulties with land access for poor urban dwellers. Expropriation, however, has become more politicised in recent years, mainly through the politician Julius Malema and his party, the Economic Freedom Fighters, as well as more radical factions within the ANC, which call into question the right to compensation. In a speech in 2011 at the Setswetla informal settlement in Alexandra, Julius Malema referred to government's difficulties in buying up land owned by whites in the vicinity of the township. He criticised the fact that government had to pay for it: "We have to buy land from whites when they did not even buy land from us" (Molatlhwa 2011).

In the post-apartheid city, being against desegregation and public housing is a morally troubling stance. In the many interviews, conversations and meetings with property owners on the 'ARP threat', I observed that the fear of being judged as racists was hanging like a sword of Damocles over the Linbro Park property owners' heads. For some, this was mostly a problem of public appearance, and in more private spaces they openly talked about negative feelings toward black people whom they did not want to live with. For most, however, it was also about how they wanted to see themselves, namely, as rational people who acted on the basis of objective reasoning and not racist attitudes. This subsection unravels these diverse narratives, entailing different forms of boundary demarcation, which the property owners employ to construct an urban world in which their negative stance towards public housing seems like a logical, justified attitude. The *politics of loss* strongly depends on these stories and theories property owners tell, as they constitute the way in which the property owners conceive of Johannesburg's social and spatial worlds and their own place in it in relation to their Alexandra neighbours. Replacing and moving beyond 'race', their narratives standing at the centre of the *politics of loss,* constructing new boundaries based on stories about class, crime, property ownership and culture. These narratives unravel the *conceived* city of the property owners, what kind of urban world they believe they inhabit and in which way they believe the 'other' are different from them.

One dominant way to interpret the relationship between Linbro Park property owners and their potential future Alexandra neighbours is to see it as a vexed relationship between economic classes, best exemplified by this statement by Kacy, an LPCA committee member, who was also involved in a charitable NGO in Alexandra: "We have no problem with the suburb developing, but we don't want a situation where we have … there is no racial issue, it's not a black–white issue, it's an economic issue" (Kacy, March 2011). Clara, another property owner, understood herself as a person who wanted to 'uplift the poor', and pointed out to me that she had always helped her black employees extensively. Being highly aware that her opposition, the ARP project stands in sharp contrast to the imaginary of the South African nation as a 'rainbow nation' working towards desegregation and racial reconciliation, Clara explained: "The Linbro community does not want them here … It's not about colour, it's about niveau. It's about low-income groups and middle-income groups" (Clara, April 2012). These narratives create a divided world, consisting of two apparently homogenous and opposed social entities. Imagining the 'low-income' Alexandra as opposed to the 'middle-income' Linbro Park, however, overrides the reality that a diversity of milieus live in these neighbourhoods, including, for example, domestic workers in Linbro Park.

The problem of poor and rich people living together in one area is, in their eyes, crime: the 'fact' that, so goes their theory, as soon as poor people become visually exposed to wealth, they turn into criminals. "It doesn't work if you are unemployed, sitting there without work, looking at the houses around you," the former LPCA committee member Linda explained to me (Linda, March 2011). Andries, also a former LPCA committee member, said: "Obviously, exposure. I mean if I walked past here every day and I see a pot of gold inside this fence, ja! Maybe, if I'd never knew about it, I would never make a plan" (Andries, May 2012). By referring to his imagined self as a poor, unemployed person who would also turn to crime if he was in the same situation, he makes clear that this theory of unemployment and exposure to wealth as causes for crime is racially neutral, based on logical inference and not based on prejudice. Everybody can be unemployed and therefore a criminal, it just happens that in contemporary South Africa most unemployed are black, his theory implies.

The everyday theory that when poor people are exposed to wealth they become criminals constructs the relationship between Linbro Park and Alexandra's residents as a relationship between potential criminals and potential victims. "When unemployed people move here, there will be crime," Sarah, a property owner in Linbro Park, explained to me (Sarah, May 2012). In this narrative, she presents the future as a predictable sequence of event, what Derek Edwards (2003) calls a 'script formulation': crime is presented as a typical, almost automatic characteristic of the category 'poor' or 'unemployed'. The course of events is presented as regular, factually robust and knowable in advance without having to wait and see whether it takes place (ibid: 38-41). Amy, another property owner, explained her anxiety about people from Alex potentially moving into the area: "Because of the crime, because of what will happen, I won't be able to protect my family against it. Because these people gonna be unemployed and there is gonna be more crime. There is no doubt about it" (Amy, May 2012).

Crime, therefore, becomes a powerful boundary-demarcation device in the *politics of loss*. The narrative of crime transforms the problematic opposition to public housing into a natural right of threatened citizens to defend their physical security and to protect their loved ones. With that, they try to present themselves in a good light and in ways which confirm socially accepted values (Goffman 1959). Crime hence appears as a logical reason why segregation needs to be continued. This became apparent in the same conversation with Amy:

Barbara: What do you think about the low-cost housing plans? Amy: I would be devastated, to be honest. I know people need places to live, but I see a lot of empty space around, for example in Modderfontein. It would be such a pity. I do look af ter people, I do my bit for society, I do get people, I do feel sorry for people, I do want people to get better, but I also have worked very hard. And I know that I need the money from this place one day. If it was only at the outskirts of the suburb, I wouldn't mind (Amy, property owner in Linbro Park, May 2012).

Scholars have called such narratives and stances as the 'not in my backyard' (NIMBY) phenomenon. The urban dwellers portray themselves as and believe themselves to be being morally good, benevolent, and even charitable people who want to do their 'bit for society', as long as it does not concern their neighbourhood space. Important to this narrative is that it entails a one-dimensional understanding of inequality: the property owners merely problematise the poverty of the Alexandra residents, while they construct their own aff luence as normal and unproblematic. This normalisation of wealth and demonisation of poverty is similar to what has been found in research on gated communities and many white South African residents in cities such as Cape Town and Durban (Ballard 2004b: 56, Lemanski: 2006b: 416). The narratives present the property owners' favourable economic position as earned and something which they worked hard for, and not related to racial privilege during apartheid. Through these narratives, the property owners make the *politics of loss* appear reasonable and rational to themselves.

In some situations, property owners introduce boundaries with the category 'poor' based on the notion of class: they make a categorical differentiation between the 'bad' poor who are unemployed and therefore criminal and the 'better-off' Alexandra residents who have jobs, a decent lifestyle and who hence would be acceptable as neighbours. This distinction is ref lected in the words of the property owner, Baldwin:

There is also a situation ofthe actualsocialclass. I don't mind if he is a middle and lower class, he doesn't have to be a rich man, but I wouldn't be happy if he was an impoverished man or an unemployed man. The problem aboutthe ARP istheywanna move the people from Setswetla squattercamp, I think you know about that, over here. I said to everybody in Linbro Park: "I am going to Setswetla to do an investigation, I'm gonna go there and speak to the people and see what they think about the move." And when I went to Setswetla squattercamp, they allsaid: "When are we going to get our houses given to us in Linbro Park?" Right? Then I went and spoke to people on the East Bank [middle classsection in Alexandra]who own housesthere, and they said: "For God sake, when are you going to take that bloody rubbish into Linbro Park!" Because Setswetla squatterswere robbing the guys on East Bank,who own houses. You follow? And thisis where my problem comes(Baldwin, Linbro Park property owner, March 2011).

By emphasising that he, as a white person, personally dared to go to Alexandra and speak to the different parties, especially house-owning residents in East Bank, he shows that his judgement is based on rational inferences from facts and not on racist prejudice. By means of the boundary drawn between tolerable better-off Alexandra residents and anti-social unemployed crime-prone poor he distinguishes between acceptable and unacceptable types of public housing: property owners like him argue that they would accept 'bonded' housing, aimed at the slightly better-off groups (upper end of the low-income market), for example as teachers and nurses, but not 'RDP housing' intended for the poor. In the eyes of Linbro Park residents, this target group is equivalent to the 'bad' poor with no jobs. People who are able pay off housing bonds are, like the property owners themselves, hard-working citizens who take their destiny into their own hands, they believe. Property ownership hence becomes something shared between themselves and the imagined Alexandra neighbours, a realm promising intersubjectivity and mutual understanding (see also Heer 2018). Owning or not owning property, an important boundary property owners build in their *politics of loss,*  does not explicitly follow racial lines, as there are also non-criminal, better-off Alexandra residents who could maybe afford a bonded house. This, again, helps the property owners to see themselves as non-racist.

Another way in which property owners talk about their stance towards public housing is through property values. As already explained above, when many Linbro Park property owners felt that they could not stop the disappearance of 'country living in the city*',* they no longer saw themselves as being solely interested in preserving their way of life (*social space-makers*), but as rational economic agents interested in maximising profit (*abstract space-makers)*. Amy explained this shift like this:

We want our properties to be worth the money. We want the money. There is lifestyle and there isthe money. I used to be very involved in the lifestyle, and I would not have accepted nothing for my property. Ifthey camewith a 10 million of fer,then [in the past], Iwould have said no.Nowadays, Iwould say: "The money isimportant." YetIwould still be very disappointed to sell (Amy, Linbro Park property owner, May 2012).

Space now became evaluated according to quantifiable criteria, such as size, price per square metre and zoning. Their felt shift in rationality feeds itself off one key theory, namely, the script formulation that public housing will lead to a devaluation of property prices in the area. What Clara, a property owner, said in an interview with me exemplifies this theory: "In the moment when my neighbour puts 100,000 people on his property, my property loses 50 per cent of its values. And that's how people think here and that's why they say, 'keep them away from us'!" (Clara, April 2012).

The property owners argue that when people sell their land to the ARP, this will negatively affect the overall property values in the neighbourhood, as crime will increase and businesses will suffer and they assume that investors will not be interested in putting high quality developments next to the low-cost housing. The assumption that proximity to a township and an increase in a black, poor population leads to a devaluation of property is a common-sense theory widespread among suburbanites and property economists in South Africa (Saff 1998). This white fear of property devaluation has also been observed in many cities around the world in similar situations when members of poorer and racially different categories move into white and aff luent suburbs, for example in the US (Massey 1989). Andrew, an LPCA committee member, also links the Linbro Park situation to other places, namely, with what he knows

about the 1980s and 1990s when inner-city neighbourhoods like Hillbrow and Yeoville experienced rapid devaluation and housing abandonment (Morris 1999a).

The community said [in the public participation process]: "Well if they get in here, it's just going to spread." Because obviously, Ithink that'stheway mostcommunitieswork. If you have one area, and nobody will buy around it because they don't want to be in close proximity to a, let's call it a slum, so eventually, it basically starts to carry outwards. I think Hillbrow and Yeoville are prime example of that. It used to be very good prime, nice locations, but asitstarted sprawling out, peoplewouldn't buy because they were worried about what was happening next door. Then the properties were going there for nextto nothing. And that'swhattheywereworried about here (Andrew, LPCA committee member, March 2011).

Suburbanites as urban citizens worrying about property values have greater moral legitimacy than property-developing enterprises, according to such narratives, as it is not only about maximising their profits but about securing their well-earned retirement. Andries explained:

They all clung on to the property because their investment was going to be their retirement. So one day they planned they would sell it for the five or six million rand a property, or the 10 million rand a property, and that is enough to sustain as a retired person … So the people hold on to it, yet the retirement is diminishing in front of their eyes (Andries, former LPCA committee member, March 2011).

Creating moral legitimacy for their opposition to the ARP, the discourse of wealth earned through hard work constructs them as good, diligent citizens. Linking property values to retirement emphasises that what they are defending in the *politics of loss* is earned and not the result of previous privileges. They also use property values as a boundary demarcation device against the potentially interfering state. As Dave expressed it at the LPCA Annual General Meeting: "The government is attempting to steal land from us for no value" (Dave, LPCA Annual General Meeting, May 2012). For many years, property owners hoped that the government would not be able to obtain the funds needed to buy land in Linbro Park. They expected that the prices offered by the government would be very low. Sarah, a property owner, explained it this way:

I don't see why I must agree to low-cost housing being built here. I really don't. And where are they going to getthe money?Asfar asI know,they have never had the money and that [the rumours] has been going around for about 5 years as well. I am hoping it doesn't ever happen, I really do. Because I mean, I get nothing for my property [from the government], they will give me what the municipality can pay and that's nowhere nearwhat the market value is(Sarah, Linbro Park property owner, May 2012).

This narrative is based on the notion of an 'acceptable price' (Plattner 1989), an imagined market value, which the property owners bring up in contradictory ways. The residents claim that they will not sell to the ARP, as the ARP will offer lower prices than the market, hence, a price which is not acceptable to them. Framing themselves as economic actors aiming to preserve their investment and retirement, they present

this stance so as to be free of moral censure: they are not against housing for the poor, but against selling their properties for below the actual price. This narrative creates the impression that the market prices for the properties are high and their lands very valuable. The reality, however, is different: As mentioned above, at the time of the fieldwork (2010–2012), the free market demand for the Linbro Park properties was low and residents complained that they had not been offered an 'acceptable price' by private buyers either, nor, indeed, had they even received an offer. Some property owners also believe in the conspiracy theory "that they [the local government] were actually trying to keep the price of Linbro Park down", because "they want to buy Linbro Park for Alexandra" (Baldwin, March 2011). Thus, there is a considerable discrepancy between the notion of the acceptable price present in their narratives and the reality of, at the time of research, a very low market demand for their properties.

A further boundary demarcating device property owners used in order to legitimate their *politics of loss* is culture: a further narrative constructs fundamental differences between Alexandra and Linbro Park residents as related to 'lifestyle' and 'cultural' differences. The encounter and living together in the future are imagined as a problem of intercultural communication and disagreements on what constitutes the right way of living and appropriate uses of space. In the following quote, the politician Thomas, from the Democratic Alliance party, theorises more extensively on these differences, linking them to what he defined as 'culture'.

I'm one of the only champions in politics that calls everything a suburb. So whether you're in Alexandra, in Klipfonteinviewor in Linbro Park, they are allsuburbs and that's roughly what it should be. […] Now in Alexandra, you got a township. It is perceived as a township, it has a culture of its own. In Linbro Park, the culture is dif ferent. You bring them together now, and…You get peoplewho say:"right, I am leaving the culture in the township and move to the suburb." So now when you are a stranger in someone else's backyard … and Icall it backyard very loosely, and I don'tthink that's a fair orrightword to use. There is a buzzword called 'ikazi culture'. There's life in the townships. There is 'ubuntu', which means get together and helping each other, so everyone knows each other, you knowyour neighbours. If youwalk around atten o'clock orif youwalk around at six in the morning, there's a buzz: people walk in the street and talk, there's trade happening. The same can't happen in the suburbs. If someone started selling sweets outside my house, there'd be ten complaints from residents trying to get metro police to remove them. These are the kinds ofthingsthat are dif ferent between township and suburb. It's hard to verbalise it with words, you got to feel it (Thomas, politician, September 2012).

Thomas supports the residents' opposition to the ARP but has even more need than the property owners to avoid being judged a racist. As anthropologists and politicians did during apartheid (Cocks 2001: 741-742), he comes to use the notion of culture to justify segregation. He constructs 'township culture' and 'suburban culture' as essentially different lifestyles with different habitualised uses of space, different values regarding what appropriate uses of space and different intensities of neighbour interaction are, which also speaks to different public–private boundaries. Alexandra and Linbro Park thus become framed as two separate, clearly distinct urban worlds, each with its specific, radically different spatiality. Any entanglements would, as follows from this

key narrative in the *politics of loss*, create cultural problems. He comments that the difference is hard to put with words, but something which one needs to experience ("you got to feel it"). This points to the link between this discursive representation of cultural difference to actually lived and perceived urban diversity. Urban dwellers do indeed see, feel and hear that everyday life in the suburb and the township is different from what they are habitually used to in their own milieu. Ballard (2010) has shown that this perceptual aspect of desegregation can be experienced as an attack on the sensual habits by white suburbanites. Linbro Park residents for example fear the erection of shacks and informal trade on the street, uses of space which are confusing for the suburbanites' senses, as they are used to empty, quiet streets bordered by free-standing houses behind representational yards. They interpret these spatialities, which that are out of the ordinary for them, as out of their control, as dirty and dangerous.

Again seeking to legitimate the *politics of loss,* Thomas does not judge the township way of life in the above quote but values it positively as a form of culture. Linking township culture to income and assuming that social mobility would lead to a change in culture, he induces a categorical distinction between people who move voluntarily from the township to the suburb, implicitly referring to the emerging black middle class, who would adapt the 'suburban values', and those who would be relocated by a government housing programme, implicitly referring to very poor township dwellers who will bring with them their 'township culture', causing conf lict among the suburbanites.

When imagining having several cultures sharing one neighbourhood it would appear to be a problem, as in the conversation with Kacy:

You have your Muslims, your Jews, you have your … People like to group together. And how many people are going to be comfortable here?Alot ofthem,they like theirfriends in Alex,they like theirway oflife in Alex,they like doing thingsthatway, you know. Why pull them into another area! Why pull somebody into another area. I mean, I wouldn't be happy sitting in ... (Kacy, LPCA committee member, March 2011).

This narrative constructs differences between the milieus living in Alexandra and Linbro Park as complementary: they are constructed as two culturally distinctive and structurally equivalent entities which would feel uncomfortable if coexisting in close proximity. She reframes segregation from a phenomenon caused by inequality which is evaluated as damaging to a society to a cultural phenomenon resulting from the legitimate desire to form affinity groupings (see also Young 2000). Kacy argues that it is natural and not morally wrong that people feel most comfortable among people who are like them. In this essentialist understanding of culture, entanglements become undesirable.

There is also, however, a more negative interpretation of the conviviality of what property owners construct as two 'cultures': often, the 'ARP coming in' or 'ARP threat' in terms of which Alexandra residents would move to Linbro Park is narrated and experienced as a threat to the property owners' lifestyle. At one meeting, a property owner exclaimed: "These [swearword] across the river want to actually f lush in here" (LPCA meeting on CID, June 2012). This comparison of Alexandra residents with a f lushing river which overruns everything implies that the mixing of the two worlds would not lead to mutual adaptation, coexistence or a new form of neighbourhood;

rather it foresees that the township spatiality would take over, dominate and destroy what the property owners still imagine as the quiet white idyll of 'country living in the city'. The above narratives, by imagining borders of culture and lifestyle, construct future encounters as anticipated conf licts based on an essentialist notion of culture and identity as homogeneous, bounded entities. Imagining a conviviality of several lifestyles in one neighbourhood becomes impossible, and their *politics of loss*, the opposition to the ARP, a logical consequence.

### **Aspirations to Suburban Control**

All of us have been here a long time, very long time. If we could have had it our way, we would have juststayed as we were. But you know, the world is progressing, there is nothingwe can do aboutit. Butja.Wewill be the last people standing (Jess, Linbro Park property owner, May 2012).

During the time of the research, many Linbro Park residents and the LPCA committee members were aware of the limitations of their attempts to counter desegregation. Nevertheless, the property owners liked to portray their stance towards the changes in their suburb as a resistant agency, reminiscent of the narrative of the last standing Gallic village resisting Roman occupation (Goscinny and Uderzo 2004 [1961]). Many residents thought that if they were not able to prevent this new development, they should at least attempt to shape and control it. The creation of a City Improvement District (CID) promised to give them such control, potentially putting in place a new administrative and financial barrier around which property owners could rally.

CIDs are self-taxing, self-help organisations set up by businesses and property owners in order to govern neighbourhoods, especially with regard to public services (Peyroux 2006: 09). They are regulated by provincial legislation and must be authorised by the state so that they can impose levies on the property owners within the defined geographical area (Clarno 2013: 1202, Gauteng City Improvement Districts Act 1997). As such, they constitute a model of privatised, decentralised governance (ibid: 1203). In South African cities, CIDs were first established in CBDs such as Johannesburg, Cape Town and Sandton, and have mainly had economic aims: they were installed to protect and improve values and commercial profits. Since about 2006, there have been calls in Johannesburg for the establishment of residential CIDs in suburbs but the City of Johannesburg has opposed these demands (ibid: 1202).

The idea to create a CID in Linbro Park was pushed by town planners with a stake in Linbro Park. Town planners constitute an important, if also slightly opaque, group of actors in the neighbourhood. By selling their expertise to the interests and power of the state and private capital, this group of professionals inf luences the production of space. Many town planners with a stake in Linbro Park have worked for independent consultants, local government institutions (such as the ARP itself or Johannesburg Development Agency) and property developers. They are well connected among themselves, having worked with each other in a variety of roles in different urban development projects. Certain town planners have been working in Linbro Park for a long time; they have been employed for many development applications and know the suburb intimately (Dave, LPCA committee member, June 2012). One such town planner successfully lobbied for a CID to be mentioned in the Urban Design Framework (UDF) (City of Johannesburg 2010: 49). This idea was also appealing to the members of the LPCA committee.

In the existing literature, the reasons given as to why local actors create a CID are usually the provision of public services, particularly security, and the maintenance of public space. The CID should act as a private government, collecting levies from its members and in return providing or coordinating the delivery of services which the local government provides insufficiently. As such, the spread of CIDs has been linked to the rise of neoliberal forms of governance and 'urban entrepreneurialism' (Peyroux 2006: 09). But in the case of Linbro Park, the attractiveness of the CID lies less in the provision of public services, and more in the promise of achieving control over the development of the neighbourhood. The CID would give the property owners an administrative tool to guard the boundaries of the neighbourhood and ensure order:

If you wanna be in control what's happening in the suburb, you need a CID. It is the only way to control development in your area … Once the ARP comes, it has to participate in the CID. If they don't, it would be mal-intent … The CID sets standards in the area, the CID runs the suburb, and gets a levy from each property. It's part of the title deed. A good example is Illovo Boulevard … You can only police an ARP component with a CID. Do it now, then the ARP has to join (Dave, LPCA meeting on CID, June 2012).

Both town planners and residents alike understood the CID as what one town planner called a "mechanism for assisting the development in the area." (LPCA meeting on CID, June 2012). In other words, they hoped to inf luence development in Linbro Park in such a way that the ARP would not be able to buy the necessary land and, if the lowcost housing project was nevertheless implemented, the residents could retain power over the implementation of land use regulations via the CID body. They expected that this management body could inhibit the erection of shacks and keep away informal hawkers from public spaces. As such, the CID would be an authority with territorial powers through which the previous Linbro Park residents could monitor and enforce compliance with by-laws. The CID would give the property owners significant power over the usage of space by the potential new residents from Alexandra.

The felt need for such a controlling body is related to the boundaries of irreconcilable cultures above described which, in their imagination, separate Alexandra from Linbro Park. The CID is a way of dealing with feelings of dislocation the property owners anticipate experiencing when the new neighbours originating from townships transform the sterile RDP houses into what the property owners see as uncontrollable and chaotic spaces. The call for by-law compliance can therefore be understood as an appropriation of instruments of neoliberal governance by the Linbro Park property owners in order to defend what, for them, is a legible (Scott 1999, Lynch 1960) and safe lived space. Besides the assertion of governmental control over neighbourhood space, the creation of a CID also promises a renewed unification of the 'community' which, they feel, lacks cohesion. The CID should empower them to speak with a united voice, they hoped, which gives their objections to town planning applications or legal steps taken against other actors, such as the local government, more weight and more legitimacy. Improved crime control was also a positive aspect of a CID for the LPCA.

The LPCA made various attempts to establish a CID. Around 2011, the LPCA instructed a consulting company to make a proposal, but it was not followed up. In 2012, a town planning firm with many clients in Linbro Park approached the LPCA with

a proposal to create a CID for them. In June 2012, members of the LPCA committee held a preliminary meeting with a town planner from this company at the tennis club. The town planner explained the legal requirements for creating a CID in the following way: "To establish a CID, you need three things: support of 50 plus one per cent of the residents, a management company set up plus a CID plan" (LPCA meeting on CID, June 2012). The present LPCA members were very much in favour of this proposal, yet they had considerable doubts about their ability to implement it. The residents found it unrealistic to mobilise the necessary 51 per cent of approximate 240 property owners to sign, as community meetings were often badly attended, the community was 'divided' and the LPCA members anticipated that some residents would be opposed to the plans. The LPCA members were also worried about having to pay a considerable sum to the town planning office for drawing up the CID, as the risk of investing all this money and failing to get enough residents to sign up was considered overly high. Two months after this preliminary meeting, the LPCA committee members decided not to follow up with the plan. Failing to put in place a CID was another lost battle in the *politics of loss,* not least because the property owners' sense of 'community', their sense of a 'we' as opposed to the 'other' so strongly constructed in narratives, diverted attention considerably from the lived reality of a neighbourhood in transition, where many property owners had diverging interests and many had already sold and left.

### **Conclusion**

In cities shaped by many inequalities like Johannesburg there are social groups that may lose out when the state intervenes to create more social justice, not least because this changes the power relations between those who benefit from the intervention and those who do not. Hence, there are social groups who engage in a *politics of loss,* opposing change because they may lose a part of their privilege. In current research on urban mobilising, there is a strong focus on progressive social movements that fight for more justice. There is, however, also a need to understand the agency of urban dwellers who want to keep things as they are ('conservative' movements), who aspire to disconnect from others, who see differences as irreconcilable and clashing. They are important forces in cities of inequalities, with considerable inf luence to shape change in the city.

An analysis of the routes for action and the discourses employed by the property owners unravelled their specific visions for the future of living together – or rather living apart. The routes for action chosen are largely within the realm of neoliberal city building and decentralised governance. Property owners appropriate the government's tools of abstraction – like resisting high densities and participating in public participation processes. The urban plans for the suburb that resulted were, according to the ARP, a compromise between residents and the local government, yet in the eyes of the property owners, they failed to convey their interests. Property owners tried to strengthen the presence of property developers in the area. They attempted but failed to make use of the organisational and spatial forms available under neoliberalism like CIDs. Their fantasies of seclusion became contested by the realities of vested economic interests, discordance among the property owners and a rapidly changing city.

The property owners aimed to shape their urban world based on exclusion, the maintenance of difference and the construction of difference as a threat. Their narra-

tives are about a politics of belonging, about defining oneself and the other, of drawing boundaries and narrating differences, of creating and strengthening social and political attitudes based on imagined encounters. The narratives are also about a politics of entanglements, of how they want to live with, or rather apart from, others. The scripted formulations property owners use do not merely represent an 'inner attitude': rather, they construct interests, identities, ways of interpreting the world and acting towards it (Laclau and Mouffe 2001 [1985]: xi). With these discourses, the Linbro Park residents construct and maintain the differences and inequalities. In their powerful narratives they construct the urban world they inhabit, whom they share it with, in what relationship they stand towards them, and they try to shape their surrounding materialities accordingly. By focusing on the narratives, one may unravel the kind of city they construct around themselves – a city full of problematic, even dangerous, entanglements with people whom they believe to be fundamentally, irreconcilably different from them. These become very powerful assumptions and theories beyond which the actors can barely think and with which they legitimate their actions. As such, the discourses have practical, powerful effects and can therefore be understood as ideology (Wetherell 2003: 14, referring to Foucault). Key terms like 'property', 'crime', 'class', 'culture' and 'sustainability' with which the property owners seek to invoke basic assumptions in the listener about universal rights and common sense, come to qualify as what Laclau and Mouffe (2001 [1985]) describe as *empty signifiers*. Such claims to universality are based on the "pure cancellation of all difference", they have to be emptied of particular meanings (Laclau 1996: 38). Such appeals to common sense are typical of discourses that touch questions of race, uttered in a societal context which morally judges racism (Edwards 2003: 40).

When we look at how this urban area is transforming, what emerges is not a teleological narrative of a formerly racially segregated city moving towards becoming a more equal and more just post-apartheid city. Rather, it is a city shaped by a multiplicity of processes and possibilities (Massey 2005) and where the old dualistic categories of black and white, suburb and township stubbornly resist dissolution, and yet become transformed in new dualistic representations of spaces around class and culture, which, like the old racial conceptions of the city, try to impose themselves onto the already existing diversity and entanglements of lived spaces and identities.

Instead of taking Johannesburg as a city consisting of neighbourhoods and social milieus as a starting point, *cities of entanglements* understands the city as always in production and asks about the difference-producing set of relations and processes. The *politics of loss* is part of the current processes at work, re-articulating the boundaries, real and imagined, between what in Johannesburg is conceived as a suburb and as a township. In Maputo, similar yet also very different processes of re-articulation can be observed on the boundary between the *bairro* Polana Caniço and the elite neighbourhood Sommerschield II. I call them the *politics of proximity* in Maputo and this is where this ethnography now takes us.

©Mark Lewis, Linbro Park, 2015

© Luversan Gerard, Linbro Park, 2012

©Mark Lewis, Setswetla, 2015

## **A Politics of Proximity?**  City Building, Enclosure and Expansion in Maputo

Cities are "places of thickening connections" (Simone 2004a: 137). Cities throw people together in spaces and by virtue of proximity, by virtue of existing side by side, "they have something to do with each other" (ibid: 137). Colonial urban governments shaped these relations in African cities like Johannesburg and Maputo profoundly so that "Europeans and Africans … had both everything and nothing to do with each other" (ibid: 137). While in Johannesburg apartheid planning often left ample space between White and Black areas, so that there was enough land to build a highway between Linbro Park and Alexandra, in colonial Maputo, the two typologies of spaces, the *Cidade de Cimento* and the *Cidade de Caniço,* often rubbed shoulders, sometimes only separated by a street (Penvenne 2011: 253). Proximity, and the way urban dwellers deal with it in a transforming city, stands in the centre of this chapter on Maputo.

This chapter focuses on how urban dwellers in two Maputo neighbourhoods, the *bairro* Polana Caniço and the elite neighbourhood of Sommerschield II, engage in what I call the *politics of proximity. Politics of proximity* refers to the manifold practices that urban dwellers, the state and property developers engage in to transform proximity into distance, as well as the imaginations of proximity and the ideas of the urban good that go with is. It is about how urban dwellers and other actors try, through spatial, political and social practices, to shape this peculiar coexistence resulting from an elite and a poor area rubbing shoulders. Like the *politics of loss* (chapter 4), the *politics of proximity* also unravels the way in which neighbourhoods, urban differences and entanglements become reconstituted as material, social and imaginary realities. It expands on what the previous chapter on Linbro Park and Alexandra in Johannesburg started, namely, a comparative ethnography of entanglements through the analysis of the politics of neighbourhood constitution and boundary making in unequal cities.

The *politics of proximity* can probably be observed in many urban areas where wealthy and less aff luent urban groups live segregated, yet in close proximity to each other. This chapter zooms in on a very specific setting: Sommerschield II and Polana Caniço are situated to the north of Maputo city centre, adjoining the old inner-city neighbourhoods of Coop and old Sommerschield, and hence on the boundary between what used to be the *City of Cement* (*Cidade de Cimento*), the centre of the colonial citizens, and the *City of Reeds (Cidade de Caniço*), the 'informal' slums and shantytowns of the Natives, the colonial subjects (Mamdani 1996). Sommerschield II constitutes a new type of elite neighbourhood which only emerged in the 1990s. This chapter sheds lights on the social processes at stake in the building and expansion of this new type of

urban neighbourhood in the post-socialist, neoliberal city. It does so by examining the *politics of proximity* from two ethnographic angles*.* The first angle is an attempted road closure: the ethnography reconstructs how the elite property owners tried to close off Sommerschield II with road closures in 2003 and 2009, and how they failed to do so because their neighbours from Polana Caniço were opposed to the closure. The narratives around these events are part of the oral history of the area, and resulted in a denial of enclosure and denial of superiority.<sup>1</sup> The second angle is the gentrification processes and how residents from Polana Caniço related to them in 2010–2012. The *cidade nova*, a new type of urbanised space that Sommerschield II represents, is currently expanding into Polana Caniço, because it is seen as prestigious owing to its proximity to the city centre, the beach and new malls. In this *politics of proximity*, the poor become offered compensation for relocating to distant places on the urban periphery and living close to the city centre becomes the privilege of the elite in the area.

As Castells said: "Societies are structured around conf licting positions, so the production of space and cities is, too" (Castells 1983: xvi). Polana Caniço A, which is what the larger urban area containing Sommerschield II and Polana Caniço is called in municipal documents, is, like Linbro Park, an area in transition. In this area in Maputo, different visions for the future are at stake, and differing actors use the current openness, the contingency of the space, in order to realise their own visions of what a good city, a good urban life should look like, and how they should be (dis)entangled from others. The *politics of proximity* is based on competition and the interplay between different *spatial projects*: diverse urban dwellers' lifestyles and agentic possibilities, property developers' aspirations and capitalist resources, and the state's visions and its capacities. These spatial projects attempt to produce and change space according to their own logics and goals by applying particular techniques (Madden 2014: 480). Whether these transformations of the urban area will lead to an upgrade of the area from which the current poor residents can profit or whether it will lead to a large-scale expulsion of the poor to the outskirts of the city is not yet clear. Hence, this chapter contributes to the ethnography of cities, spaces and neighbourhoods as "the unfinished products of historical debates and conf licts" (Castells 1983: 318).

The *politics of proximity* and the *politics of loss* (see chapter 5) constitute two different modes of entanglements, emerging through the agency of urban dwellers shaping the boundaries between changing categories of neighbourhoods and people, practices which at the same time constitute these categories in new ways. Both constitute a difference-producing set of relations and processes in the *cities of entanglements.* When analysing them, what emerges is that entanglements are characterised by the ambiguous co-presence of two fundamental attitudes: encounter and distanciation. Encounter is, on the one hand, "an interaction where both actors perceive and recognize the difference of the other, respect it, and try to build on it in their relationship" (Förster

<sup>1</sup>  The first part ofthe chapter aboutthe road closure isless aboutwhat'really' happened and more about how the urban dwellers talk about these events, how they interpret them and what this tells us about how they conceive urban society and about how they negotiate proximity. The following elaborations about the road closure are based on informal conversations and interviews conducted in both neighbourhoodsin 2010 and 2012. The narratives have been interpreted with regard to ethnographic knowledge of everyday life in these two neighbourhoods, which was established through participation and observation. There are no official documents or newspaper articles accessible on the road closure.

2013b: 242). Distanciation is, on the other hand, "an interaction where two actors adopt a disruptive attitude toward the other, trying to secure an independent agency" (ibid). Out of this tension between proximity and distance, between togetherness and dividedness, between encounter and distanciation results a deep ambivalence and even contradiction which is characteristic of urbanity in general (ibid), and for the *cities of entanglements* Maputo and Johannesburg more specifically.

The chapter concludes with a more systematic comparison of empirical aspects concerning the *politics of loss* versus the *politics of proximity* as two different modes of entanglements in Maputo and Johannesburg. The two following ethnographic chapters (chapters 6 and 7) about religious spaces and malls, both providing modes of entanglements which promise what urban dwellers see as positive encounters, weave the material from the two cities together even more. This ethnography, therefore, should not be understood as being split into two parts according to geography, but rather as the fundamental concepts developed in the preceding three chapters feeding into the following ones, and vice versa.

### **Negotiating Proximity through Road Closures**

In 2010, Alfonso was a waiter at Café Sol, an American-owned coffee shop in Sommerschield II. Café Sol is a place of public life where expatriates and members of the local elite mingle, and drink coffee imported from Kenya. Alfonso is originally from Sofala and only recently moved to the capital city of Maputo. At the time, he was living with his wife in a rented room in Polana Caniço, yet he was planning to soon finish his own house in the peripheral neighbourhood of Intaka. At his workplace, Café Sol, he often observed how customers spent 1500 to 2000 metical over a breakfast, the same amount he earned in a whole month, and he used to calculate how much rice he could have bought with the money they spent on one breakfast. "One gets used to these social differences, but it's not easy" he explained to me (Alfonso, October 2010). His rented room in Polana Caniço was at few minutes walking distance from Café Sol, but the spatial rupture, the discontinuity between the spaces of the elite neighbourhood Sommerschield II and the adjacent *bairro* Polana Caniço, made them appear as two distinct, disjunct urban worlds. When my fellow anthropologist and research assistant, Fernando Tivane, and I accompanied Alfonso on his way home from his workplace and crossed from Sommerschield II to Polana Caniço on a street which connects the two areas, Alfonso commented that the residents once attempted to close this street off with a wall, but the residents of Polana Caniço had destroyed the wall the following night. In the following months and years, many urban dwellers, mostly from Polana Caniço, but some also from Sommerschield II, recounted this story when we asked them about the relationship between the two adjacent areas. The story about the wall can hence be considered part of the collective memory or oral history of this rapidly changing area and is symbolic of the politics of proximity that shapes the interactions and spaces.

Sommerschield II consists mainly of two- to three-storey mansions, painted in bright colours from white to yellow and even pink, all equipped with the security aesthetics typical of many upper-class residential areas all over Africa – high walls, electrified fences and security personnel. These security workers, gardeners and housekeep-

ers, are, like Linbro Park, the people most visible on the streets. The proper residents can sometimes be seen arriving or leaving in their luxury cars, typically a black Pajero. The streets are also frequented during the day by residents of Polana Caniço who walk through Sommerschield II on their way to work and back. The aesthetics of the Sommerschield II streets symbolise aff luence in Maputo, and it is hence not surprising that artists sometimes use them as a stage for their music video clips. People say, for example, that the music video of a song by the Angolan singer Yola Semedo was shot there.

Sommerschield II is a young suburb and emerged as a result of processes which have changed Maputo's urban form and society considerably since the 1980s: the shift from socialism to neoliberalism, the partial liberalisation of land and the housing market in 1991 (Jenkins 2009: 102-103, Law 5/91), the arrival of the expatriate personnel of international agencies and companies, and international trends in the private governance of residential space and security (Folio 2010, Morange et al. 2012, Quembo 2010, Paasche and Sidaway 2010). These new geographies of wealth have as yet received little attention in research on Maputo, which rather tends to focus on development-related issues like poverty, health, livelihood strategies and service delivery on the periphery (Oppenheimer and Raposo 2007: 277ff). The land of Sommerschield II, which probably used to belong to the university land reserve (Jenkins 2009: 103), was used as a military camp during the war. With the end of the war, this well-situated land – in close proximity to the city centre, with a nice view on the ocean and not endangered by f looding like the lower lying areas of Costa do Sol – became available for other uses. The neighbourhood was "born out of power", as one of the residents, 70-year-old Senhor Sousa, put it (Senhor Sousa, July 2012).2 Subsequently, officials in the municipality lobbied for an urban plot layout to be designed for the area. In the early 1990s, the municipal planning department drew up an initial plan to urbanise the area without coordination or an overall register, as it ref lected the particular interests of elite groups rather than systematic planning and was never officially approved by the city council (Jenkins 2013: 97, 123-4).

At the time of this research, about three different milieus were living in Sommerschield II: members of the Frelimo elite, members of the Indian merchant elite and expatriates. The current residents of the Frelimo elite heard rumours about the allocation of these plots while they were working in high positions in local or national government or in state-owned companies. They consequently made use of their networks of inf luence to get access to these plots. Residents who gained access to a plot at that time emphasised in the interviews that while they had gained access to the land legally by paying a small registration fee for the necessary paperwork (the usufructure title called DUAT), other neighbours who moved into the area later had bought the land illegally for up to \$40,000 from corrupt government officials.

The members of the Frelimo elite are of Portuguese, Indian and African descent. Graça Machel, politician, widow of Samora Machel and widow of Nelson Mandela owned a house in Sommerschield II, neighbours told us. A former mayor of Maputo

<sup>2</sup>  This elite neighbourhood can be seen as a case of elite capture of previously state-owned property, as happened quite of ten during the privatisation processes in Mozambique (Jenkins 2013: 123-124). The post-1992 privatisation of housing policy led to the emergence of similar neighbourhoods elsewhere, such as Bairro Triúnfo along the coast and BeloHorizonte along the EN4 north of Matola.

was living there as well, about whom residents said that at one time he had been an active member of the residents' association (*commissão dos moradores*). Among our interviewees were a former deputy-minister, a former *vereador* (appointed city councillor) of the City of Maputo, and many other people who had made a career from changing political or bureaucratic positions in the municipal and/or national government. According to one interviewee, the first row of plots was given to *dirigentes* (the rulers), like Graça Machel and CEOs of state-owned companies and banks, and only later the 'smaller' members of the Frelimo elite received plots. The social position of this urban elite, linked through a shared ideology and material interests, based on bonds of mutual loyalty, friendship and even kinship (Sumich 2007: 4-5), is characterised by an intermingling of economic, bureaucratic and political power. Being neighbours in Sommerschield II is thus embedded in what Gluckman called a 'multiplex relationship', meaning an urban relationship which serves many purposes, in which one interacts with each other by way of several distinct roles (Hannerz 1980: 184). It is typical for elites to develop such networks of high density, built by going to similar schools, clubs (ibid: 197) and, in case of Mozambique, working side by side in various roles in the government or state-related companies. So although many members of the Frelimo elite may interact little with each other as neighbours, they do know each other because they belong to related networks.

### Tarring Roads, Claiming Control

The attempted road closure had a longer history in the makings of Sommerschield II as an elite neighbourhood and the production of not only exclusion but also belonging and community through space-making. It was linked to residents' collective efforts to bring into being the neighbourhood that they wanted, beginning with the construction of roads. Doing things together can lead to the emergence of intersubjectivity (Förster 2011: 11). In neighbourhoods, looking together after urban infrastructure, like going together on block watch in Linbro Park (see chapter 4), can foster a sense of belonging among residents. In Sommerschield II, it was the local government that constructed the roads in the 1990s according to the then defined plot layout. However, the roads were not tarred, which increasingly annoyed the residents. As they knew that the municipality had other priorities apart from using its very limited funds to improve infrastructure in what was emerging as an elite area, they decided to form a residents' association (*commissão dos moradores*) which was supposed to take such matters into its own hands. Indeed, at the beginning of the 2000s, the residents' association paid a company to tar the roads and to install street lights. Each property owner was expected to pay a considerable sum (apparently around \$2000) towards this end. Many residents remember this episode in the short history of the neighbourhood fondly as a success. The fact that the residents themselves built the neighbourhood roads (or at least tarred them) is one of the first statements they make when asked about the neighbourhood in interviews and conversations. Such statements express a sense of ownership and responsibility for the neighbourhood. The memory of successfully coordinating action among the many residents has become a form of nostalgia, which serves to constitute feelings of belonging (Maly, Dalmage and Michaels 2013: 758), yet also tendencies to exclude others. Residents also well remember that many neighbours did not actually contribute to the payment of the roads and that the residents' association started to lament the fact that they had no legal instruments to

force them to do so. This issue became increasingly pressing as the roads started to deteriorate and the stormwater drainage became faulty. When many property owners started to move out and rent their houses to expatriates and the Indian merchant elite, it became even more difficult to get money from residents for collective investments. In order to deal with such problems, the residents association came up with the idea of creating what in Maputo is called a *condomínio*. In Maputo, the term *condomínio* refers to enclosed neighbourhoods, security complexes and apartment buildings which are managed under the regulation of condominiums (decree 53/99).3 Condominiums are hence legal bodies of self-governance put in place by property owners in order to collectively govern communal spaces. They are similar to property owners' associations in gated communities with sectional title deeds in Johannesburg and, on a larger scale, City Improvement Districts (CIDs) like the one Linbro Park property owners tried to put in place (see chapter 4). In Maputo, the legal body of a condominium can be put in place if, and only if, two-thirds of the property owners agree to it. Then they can delegate responsibilities to the formally elected residents' association (c*ommissão dos moradores*) to manage and govern the shared spaces (Senhor Dias, executive employee at Department of Construction and Urbanisation, November 2010). The existing residents' association in Sommerschield II did not have such a legal basis nor was it elected by the residents, therefore there were high hopes that installing the administrative entity of a *condomínio* would give them more power to facilitate the collective governance of the area, as Senhor Faruq, a 60-year-old resident of the area, explained in a conversation at his shop:

The residents are not united, they are not interested, and they forget. Everything takes lots of time to get resolved. Because of that we wanted to make a condominium, with the lawof the condominium itwould be so much easier (Senhor Faruq, Sommerschield II resident, October 2010).

In relation to this, residents started to think about the advantages of turning their neighbourhood into a *condom*í*nio*, also in spatial terms by putting up a road closure. If they put up a road closure on the street which connects Sommerschield II and Polana Caniço, the *Rua do Cravo,* they could stop the through traffic which, in their view, put an additional strain on 'their' roads. Pedestrians and car drivers were using the *Rua do Cravo* to get from Polana Caniço to the main road, Avenida Julius Nyerere, and many residents saw this as a problem. Turning Sommerschield II into an enclosed neighbourhood thus seemed to be a promising solution. This was not least inspired by what

<sup>3</sup>  During the socialist period, the high-rise buildings in the city centre were governed by a state institution (APIE). But when the housing stock was privatised and the former tenants became owners of the flats, APIE retreated and problems of maintaining communal areas and elevators were exacerbated. Therefore, in 1999, a regulation about condominiums was put in place (decree 53/99, locally referred to ascondominium law, República de Moçambique 1999) so that owners of flats could organise themselves better. While there is an increasing number of gated communitiesin Maputo (Folio 2007), there is only one area which might be described as an enclosed neighbourhood, namely, the blocks surrounding the presidential residence in the city centre, the so-called Ponta Vermelha (ibid: 249). Up to 2010, only one neighbourhood had become registered legally as a condominium, butthey did notclose of f the streets(Senhor Dias, executive employee at the Department of Construction and Urbanisation, November 2010).

the elites saw in the neighbouring country of South Africa, as the 40-something resident, Senhor Batista, explained: "Many of them exist in South Africa. This [enclosed neighbourhood] was the idea, and not separation of the poor from the wealthy, as many misunderstood it" (Senhor Batista, October 2010). Referring to the criticism and even resistance which emerged among residents from the adjacent Polana Caniço (see below), the president of the residents' association claimed in a conversation with me and Fernando in 2012 that their plan was never to put up a wall. Rather, they wanted to put several blocks on the street so that cars could not pass through anymore but pedestrians could still walk past them. They eventually never went through the formal application process to become a *condomínio* in legal terms, as they found it unrealistic to get two-thirds of the residents to agree. They did, however, try to put the spatial measures in place. In 2003, members of the residents' association heard from municipal officials whom they knew through their elite networks that they would give them political support for a road closure. Subsequently, they erected a wall on the *Rua do Cravo,* which disconnected the two neighbourhoods.

### Popular Rejection of the Wall: *Jà não são mais mais*

We had problemwith thosewho live up there,themagnates,who havemoney.Because they also belong here, we share the same block representative (*chefe do quarteirão)*. But they closed that road, there,where the tar ends, have you seen it? But the people made confusion, they tried to open it. Soon they managed to open the street again, the peoplewent and destroyed it (Senhora Aurora, resident of Casas Brancas, December 2010).

Senhora Aurora, resident of a section of Polana Caniço called Casas Brancas (see below), learnt about the construction of the wall on the very same day. Her block representative called for a meeting with the residents of her section who were, as she recounted, all angry about the road closure. On the same night, the wall was taken down by the population (*o povo*). This is how Senhora Aurora and many other residents of both Sommerschield II and Polana Caniço recount the story as part of the oral history of the area. "The people rose up", is how one 50-year-old resident of Casas Brancas expressed it (Pedro, December 2010). Senhora Aurora emphasised, though, that the neighbourhood representative did not take an official decision that the wall should be removed; instead "it was the people who decided. The people went and destroyed it" (Senhora Aurora, December 2010).

Although details differ, the key message remained the same when these and other residents from Polana Caniço recounted the story about the failed road closure: it is a narrative about how justice won out over an elitist attempt to deal with their proximity to their poor Polana Caniço neighbours by erecting a road closure. In the many interviews and conversations we had with the residents of Polana Caniço and Casas Brancas, the wall was interpreted as an act of keeping the poor out of the neighbourhood of the wealthy, as a statement by the elite of being fundamentally different and leading separate lives apart from the world of the urban poor. Gustavo, who was also waiter at the restaurant Café Sol and lives in Maxaquene, argued, for example:

To me it does not appear just to create awallwhich closes other people out. Ok, it is not easy to have a precarious house next to a house of proper bricks. So the child who lives here and the child who lives there lead very dif ferent lives. The child here drinks milk and the child there does not drink milk nor doesit have toys. But nevertheless,this does not justify a wall, even though these are two dif ferent social classes, which is complicated (Gustavo,worker at Café Sol and resident of Maxaquene, October 2010).

Gustavo felt the need to acknowledge that the lived realities were indeed very different on the two sides of the wall, but he nevertheless considered it ethically and morally wrong to put up such a structure. Referring to notions of citizenship, equality and the equal rights of all people, Senhora Cumbane, the neighbourhood secretary of Polana Caniço, said in a conversation with us: "We are all Mozambicans. There are no reasons to close that road" (Senhora Cumbane, November 2010). She and many others denied in conversations with us that the elite had the right to different treatment and therefore the right to spatial self-segregation. In saying so, many interviewees from Polana Caniço referred to a normative idea of public space, claiming that the streets in Sommerschield II were public and therefore the elite's attempt to restrict access to them lacked legitimacy. Senhora Cumbane, for example, said: "You can't just close somebody's route. It was a public route" (Senhora Cumbane, November 2010). Some drew analogies with other segregationist projects, like the Berlin wall or apartheid. Some also recounted another case of a wall in Maputo, which is generally referred to as the wall of shame (*muro da vergonha*). In preparation for the Africa Union Summit, also in 2003, a wall was built along the road connecting the airport to the city, which hid unplanned settlements from view (Lindell and Kamete 2008). The *muro da vergonha* also led to public condemnation and became part of collective urban memory.

In Maputo, blocks (*quarteirões*) are the smallest unit of local government administration. A block is headed by a block representative (*chefe do quarteirão*) who is supposed to act as intermediary between the residents in her block and the *bairro* structures. Multiple blocks together constitute a neighbourhood (*bairro*), which is again headed by a neighbourhood secretary (*secretária do bairro*). Both the block representative and the neighbourhood secretary are residents who are supposed to be elected by their neighbours for these positions, which they hold on a part-time basis and which are badly remunerated (mostly via fees residents pay for administrative documents). These local structures are usually closely aligned to the Frelimo party and, historically, they stem from the so-called 'dynamising groups', Frelimo's local party sections during the socialist period which were responsible for neighbourhood governance (Seibert 2007: 164-5). The power and the legitimacy of the block representatives and the neighbourhood secretary have decreased considerably in recent years.

In terms of official, administrative neighbourhood boundaries, Sommerschield II does not actually exist as a neighbourhood. Rather, it forms part of the official neighbourhood of Polana Caniço A, and belongs, together with the section Casas Brancas, to the block 49a. If one were to follow these official structures, it would mean that the representative of block 49a would also have certain authority over the elite residents of Sommerschield II as they would have to, for example, come to her house when they needed administrative documents like proof of residence and not go to offices in the inner city. It should be the block representative who is the point of contact and mediator between the residents of Sommerschield II and the municipality, not a residents' association. This is important, as it shows that the constitution of Sommerschield II as a separate neighbourhood resulted not so much from an administrative act by the

city government, but is rather the result of practices by its elite residents which aimed to separate it from its surroundings, and which hence constitutes part of the *politics of proximity*. Contesting this politics in the quote above, Senhora Aurora calls into question the idea that Sommerschield II exists as a neighbourhood separate from Polana Caniço by saying that "they also belong here, we share the same *chefe to quateirão*".

Besides ignoring the official block responsible and forming an independent residents' association, the naming of the area by the elite residents also constitutes a boundary-demarcation practice. The name 'Sommerschield II' does not exist in official terms but is a colloquial name created by the property owners themselves. 'Sommerschield' is the name of the neighbouring, old inner-city suburb, called after the person who held the concession for that land during the early colonial period. In colonial Lourenço Marques, Sommerschield was the home of urban elites and still today many embassies and NGOs are based here. Calling their neighbourhood the second 'Sommerschield' is thus clearly not a modest statement. Close to this new neighbourhood there is a gated community with the same name, which was constructed in 1998 by SOMOCOL (Morange et al. 2012, Vivet 2012: 301). It is not known whether the residents adopted the name from the gated community or whether it was the other way around.

The naming of the area was an important practice in drawing symbolic boundaries between themselves and their poor neighbours. As mentioned, Sommerschield II belongs officially to the administrative unit Polana Caniço A, and most residents know this, as this is written on their DUAT or rental contracts. In 2010, a former resident of Sommerschield II, aged about 70, explained to us that if he said he lived in Polana Caniço, people would look at his fancy car and start laughing (resident of Sommerschield II, October 2010). Polana Caniço includes the word *caniço* which means reed and is therefore the quintessential neighbourhood name for the spatiality associated with the colonial Reed City (*Cidade de Caniço*), hence connoting informality and poverty. The use of the name Polana Caniço to refer to Sommerschield II's large, multistory villas hence appeared strange to several residents from Sommerschield II and Polana Caniço alike. To residents of Polana Caniço and other Maputo urban dwellers, Sommerschield II is seen as the neighbourhood of the wealthy, the elite (*zona da elite*), the magnates (*magnatas*) and big people (*gente grande*), figuratively denoting powerful and inf luential people. The term *mulungu* (white) is also used, not only to refer to the expatriates, but also to the black Mozambicans who have become rich and are therefore like the *mulungu*.

When Fernando and I spoke with representatives of the municipality, like the neighbourhood secretary or the block responsible, they usually rejected the name 'Sommerschield II' and denied that it existed as a distinct neighbourhood. By contesting the name 'Sommerschield II', as well as by complaining that like ordinary Polana Caniço residents, their elite neighbours should also get their documents through the official neighbourhood structures, they claimed their administrative power over their elite neighbours and resisted their neighbours' desire to be different and to construct symbolic and spatial boundaries. The road closure and its physical destruction go hand in hand with the symbolic processes of making and unmaking neighbourhood boundaries. As part of the *politics of proximity*, the road closure was an attempt to divide spatially what residents had already tried to divide imaginatively.

In the narratives about the road closure, the people who rose up (the *povo*, meaning the people, population, but also pejoratively, the plebs) are usually portrayed by the

interviewees from Sommerschield II and Polana Caniço as an angry mob, symbolically representing the sense of justice felt by the residents of Polana Caniço and Casas Brancas. Although many people in the interviews and conversations remember that the road closure had been taken down, nobody claimed to have taken part in it. Senhora Aurora, resident of Casas Brancas, speculated that it was the youth of Casas Brancas who used this road to go to school; the owner of a nearby bar, which attracts people from Polana Caniço and Casas Brancas and allegedly even Sommerschield II, claimed it was his customers. This stands in stark contrast to Alexandra township in Johannesburg where people in conversations with me proudly recounted how they threw stones at the police in the 1980s. In Polana Caniço, civil disobedience was carefully hidden; nobody wanted to admit to having been involved in a case of insurgency against the powerful magnates. Although Polana Caniço's residents considered the *povo*'s actions ethically and morally necessary, none of them claimed authorship of these actions. This is a peculiarity of the Mozambican context and is related to a long history of state repression, secret police and even violence against journalists. It was no coincidence that it was not a formal neighbourhood structure that decided to take down the road closure, but that it was rather left to an anonymous crowd, *o povo*, which the powerful elites and the potentially repressive state could not identify. In a climate of fear of state repression urban dwellers seek ways of expressing direct, active citizenship other than the political spaces provided by the state. Justice was restored outside of the formal administrative procedures. This is quite distinct from the *politics of loss* in Linbro Park where conf licts were often played out in formal settings like public participation processes and association meetings.

In the case of the road closure in Sommerschield II, the media played a considerable role, constituting checks and balances on the power of Sommerschield II's elites. The day after *o povo* had destroyed the road closure in 2003, a national TV channel broadcast a report on the issue in the national news. Senhor Benedito, a 45-year-old resident of Sommerschield II, explained the sudden media attention in the following way:

The people who live in those small [Casas Brancas] houses came from inner city buildings, so they are not just anybody, but there are journalists, engineers, policemen and military commanders. The people on thisside [Sommerschield II] did notconsider this sufficiently. They made the wall in front of journalists, basically (laughs). It is a total disaster (Senhor Benedito, Sommerschield II resident, October 2010).

According to Senhor Benedito, the story entered the national news broadcast because the immediate neighbours in the section Casas Brancas were not as powerless as Sommerschield II's elites assumed; they had networks including journalists and maybe a journalist even lived there. This is not least because the Casas Brancas section of Polana Caniço is a planned settlement where people who lived in the Cardoso building in Maputo's city centre were relocated to under the auspices of a World Bank funded Urban Rehabilitation Project (Pereira 1994). Many of the residents in Casas Brancas work in the formal labour market, for example as teachers, and military personnel also live there. Others have been retrenched from the public or private sector and struggle just like their neighbours in the poorer sections of Polana Caniço to make a living through subsistence agriculture in *machambas* (agricultural fields) on the outskirts of

the city. As outlined in the introduction (chapter 1), many Casas Brancas dwellers were descendants of *Assimilados*, the colonial category of citizenship African Mozambicans could acquire by meeting specific standards of what the Portuguese understood as 'civilised'. The *Assimilado* status opened up certain socioeconomic possibilities and limited citizenship rights (see Penvenne 1995), and many managed to enter what can be called the urban middle classes in the postcolonial period. What distinguishes residents of Casas Brancas from most other residents in Polana Caniço is their relationship to their land: most of them possess the legal documents, the DUAT, confirming their ownership of the houses and their right to occupy their land, something which most families in the adjoining Polana Caniço areas lack (see below). Not least because of this Casas Brancas residents have a considerable sense of entitlement to their area, a sense of propertied citizenship (Roy 2003). Some residents from neighbouring sections emphasise that the people who live in Casas Brancas are from the city centre (*cidade*) and therefore different, while again others negate the existence of such a differentiation. In contrast to Sommerschield II, which tries to constitute itself as an independent neighbourhood, the residents of Casas Brancas do not have such aspirations and consider themselves part of Polana Caniço.

Besides having certain power over their elite neighbourhoods because of their connections to the journalists and the threat of creating a media scandal, residents of Casas Brancas also saw themselves as having certain powers because they were the constituency and voters of the politicians living in Sommerschield II. Senhora Aurora, for example, said: "Who voted for them so that they could get what they have? We are poor. During the election campaigns they came here and asked for our votes (Senhora Aurora, December 2010). Contesting their elite neighbours' practices of drawing a boundary between themselves and Polana Caniço, Senhora Aurora pointed out an important entanglement between them, namely, the one between citizens and democratically elected rulers who depend on the votes of their neighbours. Many Polana Caniço residents pointed out in the interviews that Sommerschield II and Polana Caniço do not constitute separate worlds but that they are entangled in many such ways. Among the political and economic interdependencies and connections, labour relations are the most important. Residents of Polana Caniço work in nearby Sommerschield II as gardeners, security personnel and domestic workers. How would they get to work if there was a wall, the waiter Alfonso asked rhetorically in a conversation with us.

For the residents of Polana Caniço, the failure of the road closure, forming part of the *politics of proximity*, was much more than a failed attempt to keep them from driving and walking through the elite neighbourhood with its beautiful houses (*casas bonitas*). Rather, this moment had multiple symbolic and moral dimensions. It was a successful resistance against the elite's immoral desire to build an enclosed, disentangled world. Senhor Mateus, a block representative in Polana Caniço, explained it to us like this: "The crowd showed them that we are human beings like them" (Senhor Mateus, November 2010). Taking down the road closure meant that the residents of Polana Caniço denied the economic and political elite of Sommerschield II their constitution as essentially different, separate and superior human beings. Senhora Aurora, resident of Casas Brancas, expressed it jokingly in a Portuguese word play: *Já não são mais mais* (Senhora Aurora, August 2012), meaning that they had lost their status as better-offs.

### Contested Normality of Closures

Besides members of the Frelimo elite and expatriates, Sommerschield II is also home to Indian merchants. The majority of them have recently moved to the neighbourhood on the recommendation of friends or family who have already lived there. Those who arrived in the area in the 1990s tend to be home owners, while some of the more recent arrivals are tenants. The Indian merchant elite is a very old milieu in Maputo, and yet a very mobile one, as many have transnational networks with Europe, East Africa and the Indian subcontinent. Indian merchants were present and involved in the Indian Ocean trade as well as regional trade even before the Portuguese (Disney 2009: 354). They played important, if also shifting, economic roles in colonial, socialist and neoliberal Mozambique. Nowadays, many are owners of several family-based companies. Many survived the socialist period economically because of extensive transnational connections, strategic alliances with the ruling elite, as well as by occupying economic niches outside the legal realm regulated by the socialist party-state. In contrast to the Frelimo elite, most of them were not part of the socialist government. Nowadays, many build networks with those in power, not least to sustain and secure the accumulation of wealth in the private sector. Many are today members of the Frelimo party and even members of parliament, have joint ventures with politicians and are said to use political tools for business interests (Pitcher 2002: 154-158).

The Indian merchant elite are said to be especially prone to becoming victims of assaults and robberies because, according to stereotypes, they keep large sums of money from illegal business activities in their homes which they need to get to their business partners and family members in other countries through channels other than the official banking system. Since about 2011, members of the Indian elite have become victims of kidnappings in which the perpetrators demand large sums from the families. When we rang at the residents' doors for interviews in 2012, Fernando often joked with the gardener or domestic worker who opened the door: "Don't worry, we are not kidnappers, we came to talk to your employer." In 2012, many residents of Sommerschield II had replaced their self-employed security guards (*guardas*) with professionally trained security personnel from security companies (*seguranças*). The *guardas* had often worked for the same employer for years and were intimate with the daily routine of their employers as well as the faces of regular pedestrians. The personnel of security companies usually rotate between the houses of customers for whom their security company is providing the service, and therefore possess much less local knowledge. The professionalisation of security in the neighbourhood hence had the contradictory effect of actually diminishing social control, and it had become a more anonymous place.

Many Indian residents of Sommerschield II linked the need for a road closure to arguments about crime and security. Similar to property owners in Linbro Park they felt that crime was a normal outcome resulting from the proximity of poor and wealthy urban dwellers, and that there was therefore a need to protect themselves. Senhor Mostafa, resident of Sommerschield II, explained to us that this is also why the road closure was built: "The intention of the wall was not to have robberies. Anywhere in the world it is like that when there is a small house next to a big house. Nobody wants to have people assaulting their private property" (Senhor Mostafa, October 2010).

In this script formulation, the people who live on the other side of the wall, the residents of Polana Caniço, become constructed as criminals responsible for assaults

in the area. The above quoted Indian merchant assumes that the people living in the 'small houses' in Polana Caniço constitute a security threat to the social order. Another Indian merchant, Senhor Faruq, argued that the people use the street as a passageway and believed that because of this, there were assaults (Senhor Faruq, October 2010). In such narratives, the people from behind the wall become constituted as a dangerous 'other' whose movement through the neighbourhood needs to be limited. Framing road closures as a rational response to crime is also typical for debates around closures elsewhere, such as Johannesburg (Dirsuweit 2007: 54).

The members of the Frelimo elite also considered crime a problem in their area. Most of the residents dealt with it by instituting security measures like electric fences, a sign naming the security firm they had hired and a security guard sitting in front of the house or behind the gate. The more prominent residents even had bodyguards and security personnel from special police forces. Many also had alarm systems and other invisible security devices. Nevertheless, many members of the Frelimo elite contest the security narrative by the Indian merchant elite and question the 'othering' of Polana Caniço's residents. Senhora Constancia, resident of Sommerschield II, regarded this quite critically:

Their idea was that this is a Sommerschield area, which is synonym for power, wealth. But it's right next to Polana Caniço. So they thought that those people from Polana Caniço would come and assault them. [Laughing] But it's not like that (Senhora Constancia, Sommerschield II resident,November 2010).

Rejecting such stereotypes, many members of the Frelimo elite like Senhora Constancia, a high-level government official, emphasised in interviews that the residents of Polana Caniço were hardworking, honest people and not simply criminals. In the interviews, a considerable part of the Frelimo elite judged the attempt to close off the road as morally wrong and rejected the idea that they were in any way involved in it. Senhora Constancia, for example, distanced herself clearly in the interview with us: "It was a very sad situation, because it wouldn't be good to have a wall. I didn't support it, this was not an idea that could come out of my mind" (Senhora Constancia, November 2010). Similarly, Senhor Benedito, a former Frelimo deputy-minister, positioned himself:

It was an action that I personally always found wrong and very risky. It turned into a social revolution,whichwas normal, perfectly normal. Itwas a reaction that I had anticipated, as this neighbourhood was not constructed to be a condominium, the people always used thisstreet (Senhor Benedito, Sommerschield II resident, October 2010).

Apart from the then president of the residents' association, very few of our Frelimo elite interviewees acknowledged having supported the road closure. A key reason for this denial of responsibility and authorship is the negative publicity which the failed road closure received. As mentioned, the destruction of the road closure was broadcast on national TV. The road closure, which up that point had only been discussed inside the neighbourhood and was based on personal relations with municipal officials, was now suddenly revealed to a general public. This was very troubling for these urban elites, as many residents of Sommerschield II were public figures as a result

of their former or current position in the Frelimo party, the government, or public or private companies. The former deputy-minister explained that the road closure hence became very 'risky' for him:

For me, itwas very risky, because Iwasstill in government atthattime. I knew, if a problem happened, it [the culprit] would be me. When it happened, some people thought: "There livessomebody from the government, it must have been him who did it" (Senhor Benedito, Sommerschield II resident, October 2010).

The Frelimo party has its history in the independence struggle against the Portuguese colonisers. It became socialist after independence and now it claims a social democratic orientation if, nevertheless, in a neoliberal context. If Frelimo members publicly claimed their support for a road closure which Polana Caniço residents describe with terms like 'new apartheid' or 'Berlin wall', the public would judge them as hypocrites. The fear of moral censure by the public was especially relevant in this case as the road closure attempt took place shortly before the municipal elections in 2003. Residents say that a major figure in the residents' association was running for mayor at that time.

During the socialist period, there were severe limitations on food and consumption. Even at the presidency, the same food rations as all the citizens received were served at dinners with guests (Sumich 2005: 109). Accessing rare consumer goods was dangerous during the socialist period: one could be denounced by anyone for owning goods bought on the black market (ibid: 110). In the post-socialist period, a modern consumer culture became legally accessible to the elites and conspicuous consumption became an important sign of distinction, of modernity and of social power (ibid: 111). The emergence of Sommerschield II since the 1990s with its ostentatious houses inspired by suburban architecture from South Africa, the Middle East and elsewhere was a result of the elite milieus' desire to display their wealth and participate in global elite lifestyles now that socialism was over.

Together with the conspicuous consumption of the elite, another popular critique of the Mozambican elites emerged. The rampant inequality between the small elites and the vast masses of Mozambicans living in poverty became a common topic of moral and political debate in the many *barracas* (bars) and in family life in front of the TV. Allegations of corruption, abuse of state resources and moral condemnation of the wealthy elite in such a poor country are common topics on social media and increasingly also in newspapers. The anger expressed in the popular uprisings that took place in 2009 and 2010 were not least targeting these lifestyles (Bertelsen 2014).

In 2009, a second attempt at road closure was initiated by a small group of residents consisting of members of the Indian merchant milieu. They took up the idea of creating a condominium and lobbied for all residents to collectively employ a security company to police the area, which would be cheaper and safer than if each property owner contracted its own security company. Relatedly, residents asked the residents' association one more time to try to create a *condomínio* in legal terms but they failed to get the support. The legal process seemed overly complicated and still nobody believed that they would be able to convince two-thirds of the property owners to participate. Despite this negative decision, one member of the Indian merchant elite took things into his own hands: he put up a boom (*cancela*) on the road in front of his house, and he bought the material to construct another boom at the entrance to the neighbourhood to Julius Nyerere Avenue. This time not only were the residents of Casas Brancas enraged, but also the residents' association itself: they informed the municipality and the municipality removed the illegal road closure. Again, the issue was reported in the media and this time the newspaper article associated the case with the name of the then president of the residents' association, Senhor Alim, a well-known CEO of a large bank. In order to avoid further scandal, he stepped down from the residents' association. The association has since then, 2009, ceased to function (Senhor Alim, resident of Sommerschield II, July 2012).

Senhor Alim, himself Muslim and of Indian descent, distanced himself in a conversation with us from 'other' Indians in the neighbourhood. Referring to ideas about citizenship and moral behaviour, he explained that many Indian merchants living in Sommerschield II had come to Mozambique only recently. He and his family, in contrast, had been living here for four generations and were hence 'proper' Mozambicans. He portrayed himself as a trustworthy, sincere businessman who had served in the Ministry of Finance, first under the colonial regime and later during the socialist government, and hence worked hard to attain his position as CEO in a large bank and had earned his wealth. Appropriating popular discourses on elites he constructed the Indian merchant elite as dubious and corrupt: "It is not about origin, but about behaviour. They have behaviour of bribery" (Senhor Alim, resident of Sommerschield II, July 2012).

When looking at the agency of the ordinary residents involved in this *politics of proximity* and their ways of narrating and interpreting what was going on, what became clear is that urban proximity was not only about having access to a space or not, but was also about local understandings of moral values and normativity. In the oral history of the road closure, the population did not take down the road closure just because they wanted to continue using the streets in Sommerschield II. Rather, the road closure and its destruction became a metaphor for the competition between views on how the rich and poor should live together; leading segregated lives, the wealthy retreating into fortified enclaves, or living side by side, seeing and acknowledging each other's differences and sharing certain urban spaces. The residents of Polana Caniço do not so much question the elite's wealth in their narratives, but rather rejected the fact that the elite derived a sense of superiority and desire for exclusion from their wealth. They recalled the deep entangledness of the urban elite and the poor, be it in the form of security guards and domestic employees working in Sommerschield II, or in the form of the votes that the politicians need from Polana Caniço residents.

Being proximate and sharing certain spaces means that the everyday lives of the elite and the poor urban dwellers remain visible to each other. As introduced in chapter 3, Brighenti (2010: 27) argues that the relationship of looking at each other "constitutes the site of mutual recognition, misrecognition or denial of recognition of the other". By taking the road closure down, the residents of Polana Caniço shaped the entanglements between the two lifeworlds in a specific way. They turned the entanglement into a relationship of mutual visibility, entailing the *recognition* of the other as a fellow human being (Honneth 2003). It is this visibility which makes the politics of proximity a deeply moralised engagement. By taking the road closure down, the people, *o povo*, demanded social recognition from their elite neighbours, recognition that they are Mozambican citizens like all others; that they are the same human beings despite all differences. They rejected the elite's aspiration to spatially express with the

wall that they are privileged human beings, whose streets are reserved for them. As in Linbro Park, it is the less powerful urban dwellers who aim to draw entanglements into visibility, while the more powerful urbanites tend to invisibilise them. In the *politics of proximity*, by dreaming of road closure, the urban elites aspire to create their lifeworld as homogeneous and disconnected from the surrounding city, trying to disentangled themselves from Polana Caniço. This points toward what I will argue in more detail in the conclusion of this book, namely, that *cities of entanglement* as a method and analytic of cities tends to highlight visions of the city by the less powerful, while looking at the city in terms of walls and segregation rather reinforces elite visions.

### **Urban Proximity or Suburban Dream: Gentrification in Polana Caniço**

With the end of the socialist regime and the shift to neoliberalism, Maputo was drawn into globalising processes (Morange et al. 2012: 901). Besides the Frelimo elite and the Indian merchant milieu, there is a third milieu living in Sommerschield II, which ref lects these globalisation trends, namely, expatriates from Europe, USA, Brazil as well as diverse African and Asian countries. With the liberalisation of the markets and natural resource boom, many transnational companies have opened offices in the city and brought their employees with them. As the capital city, Maputo is also the home to embassy-related personnel and professionals from the development industry with their families. In Sommerschield II we interviewed, amongst others, several wives of embassy employees from Europe, a Malawian consultant working in the area of accountancy for international organisations, and a family of white Zimbabweans owning businesses in Maputo.

In terms of public life in the neighbourhood, the expatriates are the group which is most visible in public spaces. The neighbourhood has a small commercial centre, built by one of the most inf luential residents. The building of this commercial centre led to considerable discontent in the neighbourhood. In the initial plan for Sommerschield II, the plot in question was reserved for a public space, namely a park. An Indian merchant nevertheless acquired the plot and got the necessary building concession through his networks of inf luence. Some residents went to complain with the municipality, but as the original plan was never officially approved by the Council, their complaint was rejected. It is in this commercial centre that the coffee shop, Café Sol, where Alfonso and Gustavo work, is based. In 2010–2012 it was run by three expatriates, all husbands of female professionals working in the development sector. Café Sol was one of the most important meeting places for the expatriate community in Maputo and the only place of public life in the neighbourhood. The place was also frequented by Mozambican professionals and business people.

Because expatriates are usually tenants and not home owners and because they usually stay for a limited time in the city, they generally do not overtly involve themselves in neighbourhood matters. None was active in the residents' association when it still existed and nor were they involved in the road closure attempt. Yet with their demand for luxury housing in Maputo, they are a major background force in the transformation of these neighbourhood spaces and in the *politics of proximity*. Many of the mansions built in Sommerschield II are actually rented out by the owners to expatriates. Many members of Mozambican elite milieus own several f lats and houses and rent them out, acting as private real estate investors. Some of our interviewees from Sommerschield II told us that they finance their children's tuition at universities in South Africa by renting out their inner-city f lat or other properties. The Mozambican elite and the expatriates are therefore tightly interlinked economically through the rental of housing, constituting an entanglement in the city which tends to be invisible.

### A Changing *Bairro*

In the last few years, the eastern sections of Polana Caniço immediately north and south of Julius Nyerere Avenue have steadily gentrified. Private real estate investors, as well as professional consortiums of several investors, have bought up the houses of Polana Caniço residents who subsequently moved to the outskirts of the city. The previous makeshift, mostly unplastered houses, built and improved over the years by the residents, have been replaced by gated communities, a private hospital and increasingly also free-standing mansions in the style of the elite houses in Sommerschield II.

Slater defines gentrification in a broad, non-normative sense as a process of urban change in which a class of residents who are relatively more aff luent replace the previous residents of a neighbourhood (Slater, Curran and Lees 2004). A higher-income group moves into a lower-income area and the previous residents move out (Lemanski 2014: 4). In Polana Caniço new mansions have replaced much of the old housing stock. In the broad definition by Slater et al. (2004), such new-build developments also qualify as a form of gentrification (Visser and Kotze 2008). Some limited research has been conducted on gentrification in Polana Caniço, mostly focusing on land transactions (Jenkins 2013, Jorge 2016, Jorge and Melo 2014, Vivet 2012). Private developers buy up the land owned by Polana Caniço residents who subsequently move to peripheral neighbourhoods. Apart from this private-led process, a second process at stake involves government plans (which were still vague at the time of research 2010–2012) to relocate the local population and build high-rise buildings instead, into which the former residents should be able to move back. This chapter focuses on the images of the city and the ideals of urban living that guide these two interlinked processes, as well as the agency of Polana Caniço residents with regard to these urban changes.4 These two processes constitute part of the *politics of proximity*: gentrification replaces proximity by distance, pushing Polana Caniço's dwellers to the outskirts of the city. As in Linbro Park, the abstract quality of space – property value – serves as a key explanatory ideology, supressing underlying questions of conviviality and access to the city.

Although Polana Caniço may appear to the outsider as a homogenous place, the population is highly diverse (see introduction). There is a milieu of long-term residents, many of whom arrived in the 1970s. Many of the male household heads used to have formal employment, but many lost it with privatisation. Those who arrived in those early years tend to have larger yards than those who arrived later. There is a large milieu of war refugees from southern Mozambique (Gaza, Inhambane and Maputo province), which f led to Maputo mostly in the 1980s (Costa 2007), many of them being *Tsonga* (*XiRhonga* and *XiShangana*), and which nowadays constitute the majority of the population (in 1998, 40% of the population in Polana Caniço was born in southern provinces outside of Maputo (ibid: 31). Another milieu consists of war refugees and

<sup>4</sup>  For more details on the current planning processesconcerning Polana Caniço A, which are not treated here,see Jorge 2016.

ordinary migrants from the northern provinces, mainly Zambézia. Many of them are Muslim, their culture inf luenced by the Swahili region. In 1998, they constituted five per cent of the population in Polana Caniço, this has probably risen (ibid: 31). Many people also moved from the city centre to Polana Caniço when they were unable to afford to buy their f lat or house in the privatisation of the housing process in the early 1990s. Increasingly, there are also transitory milieus of tenants living in the neighbourhood, who usually live in smaller households than the long-term residents: among them are students studying at the adjacent Universidade Eduardo Mondlane (UEM), young couples in the process of establishing a home, as well as newly arrived migrants, of whom some may plan to stay while others plan to proceed to South Africa. They rent rooms in the backyards of residents or houses from residents who have moved to the outskirts of the city because they wanted a larger yard.

During colonial times, when the *City of Cement* (*Cidade de Cimento*) needed to grow, this usually involved the bulldozing parts of the African *City of Reeds (Cidade de Caniço)*. This is also the reason why in the African settlements solid houses were forbidden in the colonial period: houses made of reed were easier to remove to make space for 'European' urbanisation. In the Aguiar Urban Plan (1952) the land where Polana Caniço is situated today was defined by the colonial municipality as space for the expansion of the City of Cement (Jenkins 2013: 87-90). These plans, though, were never implemented and instead, African urban dwellers settled there. The first residents of Polana Caniço, mostly *XiRhonga* originating in the areas around Maputo, received big stands from the local *regulo* (traditional authority), and many kept *machambas*, fields for subsistence agriculture, which nowadays can only be found on the peri-urban fringe.

During the socialist period, in the Maxaquene neighbourhood upgrading project (1977–1979), parts of contemporary Polana Caniço became a planned settlement for the growing poor population of the city, a section of Polana Caniço that this ethnography does not focus on. During the civil war, many people from the north of the country and neighbouring provinces f led to Maputo and urgently needed housing. The block representatives (*chefes do quarteirão*) hence subdivided the stands of the original residents and gave the land to war refugees who were received as temporary neighbours. Most of the war refugees stayed and *XiRhonga* residents had to get used to their smaller plots, and still today remember the large plots nostalgically.

Polana Caniço has some land subjected to erosion and natural water drainage. Already during socialist times, these lands adjacent to Julius Nyerere Avenue were considered unsuitable for residential use, as the soil was too sandy and unstable. Nevertheless, because of the high population pressures during the civil war and afterwards, refugees and immigrants settled on this land.<sup>5</sup> Because of this, the municipality never granted the residents of the unplanned section of Polana Caniço tenure security for their houses and hence denied their applications for usufructure title (DUAT). During the time of this fieldwork, the residents generally felt that the municipality did not

<sup>5</sup>  Indeed, during the floods of 1998 and 2000, a large crater opened up around this main traffic axis and houses were washed away. Many residents had since been relocated by the state to new peripheral neighbourhoods like Zimpeto and people who still live there are considered endangered (Conselho Municipal de Maputo 2008: 96/59, 183/59). Because of the lack of sanitation and the quality of the soils, Polana Caniço A was the neighbourhood with the highest rates of cholera incidents in Maputo between 1999 and 2004 (Malaulane 2005: 27).

want them to live there. They felt it was hypocritical of the municipality to tolerate gentrification, with new mansions being built on the very same eastern sections of Polana Caniço immediately north and south of Julius Nyerere Avenue where they themselves were not tolerated.

### Plans for the *Nova Cidade*

A central factor in this ongoing gentrification is a redevelopment (*requalificação)* plan for Polana Caniço supported by inf luential municipal officials. The municipality does not aim to formalise or regularise the informal occupation of land in Polana Caniço as it aspires to do for other neighbourhoods that emerged from self-production by urban dwellers. Adriano, the person in charge of the municipal department responsible for these processes, explained that the municipality had plans to redevelop the whole area, including the (re)construction of a road (connecting Julius Nyerere with Vladimir Lenine Avenue), commercial developments as well as new residential spaces and public infrastructure. The PEUMM (Conselho Municipal de Maputo 2008) stated that a development plan would be drawn up for Polana Caniço. In April 2014, a consortium of consultancy companies started to develop the new plans for the municipality. Adriano pointed to the difficulty of implementing such plans: "The big challenge is not to make the plans, but to implement them … We will draw up the plan, how to execute them, we will not know. We will leave this challenge to those who come after us" (Adriano, executive employee at the municipality, August 2012). Adriano found that the implementation of the plans was unrealistic because the municipality had limited financial and technical capacity. It would be difficult to implement the plans, as they were long-term and because after municipal elections many personnel in the planning departments were usually replaced. For Adriano, it was clear that the implementation depended strongly on public–private partnerships and that it could take up to 20 years for the process to be completed.

In November 2010, a well-known architect announced on the national news that there were redevelopment (*requalificação*) plans for Polana Caniço which entailed multi-story apartment buildings to which the current Polana Caniço population would be relocated. In these plans, the current population would stay in the neighbourhood and move to f lats in three- or four-storey high-rise buildings. Similar to the Alexandra Renewal Project in Johannesburg, Maputo's planning officials have come to the conclusion that relocating residents to faraway places should be avoided, as Adriano explained:

We learnt over the years that relocations are not the best solution. Instead of reducing poverty, we were actually creating conditions which would turn the residents even poorer. In consultation with the urban dwellers we came to the conclusion that we should densify (Adriano, executive employee of the municipality, August 2012).

The loss of proximity to the city centre when moving to the peri-urban fringes of the city entailed many costs for the already poor residents. Peripheral neighbourhoods often lack basic services like schools and hospitals, and people would have to pay high transport costs for commuting. One of the main goals of the redevelopment (*requalificação*) of Polana Caniço for the municipality is to densify the city and move the current population into the new high-rise buildings. Adriano, who was in the department in charge of this development, framed this in the nationalist discourses of poverty reduction and development:

The people who will buy the new houses and who will become developed, who will enjoy the newly created infrastructures, are Mozambicans, the citizens of Maputo … If we have the opportunity to save families which are in precarious living situations and bring them to a better level, we will be elevating the living quality and contributing to the fight which constitutes the big challenge of this country, namely poverty (Adriano, executive employee of the municipality, August 2012).

In this discursive frame, the redevelopment plans for Polana Caniço form part of the national fight against poverty. This discourse of the fight against poverty and of development has also been adopted in everyday parlance by many urban dwellers in general and by Polana Caniço's residents specifically. As Sérgio, a Polana Caniço resident in his thirties, explained: "*Requalificação* is development, which we need and cannot negate. Whenever there is development, there are also sacrifices" (Sérgio, August 2012). From this perspective, the gentrification of Polana Caniço becomes constructed as something contributing to a larger, common goal, and the eventual suffering by Polana Caniço residents is a price they will have to pay for the broader development of the nation.

The redevelopment plans are not only about housing for the poor, however. As a matter of fact, the municipal officials regard the land of Polana Caniço as a prime location which should be made better use of than the current slum-like settlement. According to the municipal official, Adriano, he and his colleagues "find that Polana Caniço is one of the most noble zones of the city to live in. If one constructs vertically, one will even have view onto the sea. It is an extremely privileged area" (Adriano, August 2012). Traffic has increased considerably in the last few years in Maputo and many people spend many hours a day sitting in their car or in a *chapa*, the minibus taxis, in traffic. Polana Caniço's proximity to the city centre thus makes it a highly desirable area, besides the wonderful view onto the sea and the fresh air. Investors sometimes approach the municipality looking for land to construct shopping malls, but the municipality has to turn them down because all the municipal land is informally occupied, as Adriano explained in 2012. "So we have to create conditions in order to accelerate the process of investment in the city" (Adriano, August 2012). This enthusiastic support by municipal officials for private investment driven urban development is partly related to the municipality's lack of financial capacity which therefore makes it dependent on public–private partnerships. But it is also because such land transactions are highly lucrative for city officials who give out usufructure titles (DUAT) and demand large 'commissions' from the new users.

Shortly after these plans were announced on national TV, I conducted an interview with Nelinha, a 16-year-old schoolgirl living in Polana Caniço. I showed her a photograph of the Maputo Shopping Centre (see chapter 7), and she started to talk about the gentrification plans that, she assumed, would destroy her neighbourhood as she knew it.

This shop [the mall] is very beautiful … They say they will construct hotels and restaurantsin Polana Caniço and because ofthiswe have to leave, in orderto have more beautiful places like this. It is positive for the nation, but negative for the residents of the neighbourhood. It will become beautiful, but they will put us to a place like Malhazine with terrible living conditions(Nelinha, resident of Polana Caniço,November 2010).

While referring to the nationalistic discourse of development, she uses a notion of beauty to describe the aesthetics of the hotels and restaurants she believes will look similar to the Maputo Shopping Centre or the houses in Sommerschield II. Many residents believe that the municipality aims to transform Polana Caniço into a neighbourhood similar to Sommerschield II. The urban dwellers in Polana Caniço hence call these emerging types of spaces the New City (*Cidade Nova*). City (C*idade)* here refers to the spatiality of the city centre, which used to be the colonial City of Cement, with its high-rise buildings, tarred roads and other urban infrastructure. It is called 'New' (*Nova*) because places like Sommerschield II are emerging outside the former city centre, in areas previously occupied by *bairros* with limited urban infrastructure. The New City (*Cidade Nova*) hence moves beyond and transforms the old City of Cement – City of Reeds duality. As a mode of entanglement, the *politics of proximity* taking place at the boundary between Sommerschield II and Polana Caniço adapts Maputo's old colonial categories to more f luid and multiplied distinctions.

The *politics of proximity* leaves the category *Cidade de Caniço* intact, however, and continues to implement the colonial idea that proper *Cidade,* if now also *Cidade Nova,* should replace *Caniço*. Al-though in 2012 the redevelopment plans had not yet even been put down on paper, private property developers had already begun to speculate in the anticipated upgrading of the area. Private investors started to buy up land from Polana Caniço's residents and built luxury mansions, already starting to build what they thought the municipality was planning for the future of Polana Caniço. The new urban plan did not yet exist formally, yet through their networks of inf luence real estate investors nevertheless acquired usufructure right (DUAT) and building permission. The unwritten plans and ideas by the municipality provided the context in which the real estate investors seeking profitable investments already felt safe enough to invest in Polana Caniço. Adriano, the municipal official in charge, told us in 2012 that he had warned private investors to be cautious with new constructions, as in the yet to be written urban plan, a road could potentially be drawn to go through their property. However, destroying a new development by a private investor (*Nova Cidade)* would not make much sense, he explained, as the aim of the municipality anyway is to have private investors modernise and urbanise Polana Caniço (*Caniço)*.

If we come to the conclusion that there is a significant development on a piece of land, but that the road would actually cut through this land, we could potentially make a small detour. So we have to find a way of adjusting the reality without creating situations which don't make much sense (Adriano, executive employee of the municipality, August 2012).

The gentrification driven by private investors is hence welcomed by the municipality. Even though the building of *Cidade Nova* by private investors may be technically illegal, it replaces *Caniço* with *Cidade,* and this is the overall aim of the municipality and investors alike. Many residents internalise this in a discourse of beauty ('more beautiful places', as Nelinha said above), coming to lament but also accept this as an indisputable

part of progress (see also Harms 2012). In the *politics of proximity,* the high property value of this land in a prime location as the *abstract* qualities of space, appears as a legitimate and even a just explanation for this social and spatial transformation:

We will not inhibit that the private sector is aggressive and of fering more to the urban dwellers than what the state will of fer them. That is the law of the market. We don't have problems with the people who make those transactions. Overall, we want the *requalificação* to take place. If the people receive what they demand in these transactions and decide to leave, that is alright, they are contributing to the *requalificação* (Adriano, executive employee at the municipality, August 2012).

Instead of seeing Polana Caniço as *lived space*, where residents have built up their houses and live conveniently close to the city centre, he attributes to the residents an intentionality based on *abstract space,* the monetary value of their land. In this way of legitimising the transformation, moving to the outskirts of the city becomes a purely economic and not a social issue.

### Fearing the State

The social position and agency of Polana Caniço's urban dwellers within the *politics of proximity* should be seen within the larger Mozambican context of state–citizen relations. While they regard gentrification pushed by what they see as a demagogic state as a threat, the possibility of selling to private investors seems like salvation. Their social position is shaped by tenure insecurity. Until the early 1990s, the houses in Polana Caniço were still mostly made of reed, even though the colonial prescription to build in reed had been abolished. The choice of material was due to financial difficulties, but also because the war refugees were expected to build reed huts, as they were long considered temporary residents of the neighbourhood who should return home one day. With time, the residents started to invest their savings in the construction of brick houses and, over the years, the reed houses and the spinose hedges were replaced with more durable materials. Sérgio, a 30-something resident of Polana Caniço, explained to us:

Polana Caniço used to be allreed huts,fromthe university tillmy house.Only nowthere has been evolution, people are purposely constructing houses ofcement, even though they know that they are not secure, that the land does not belong to them. There are areas where people really should not live, but they nevertheless build there. Why? Because if they become relocated, they will receive more compensation. When the time has come, they tell them to validate their house and will demand corresponding compensation. And they know that if the house is of reed, reed can be removed easily (Sérgio, resident of Polana Caniço, August 2012).

As Sérgio explains, building with more durable materials is also a means of engaging in economic speculation regarding future compensation, as residents receive more money from a potential buyer, be it the state or private investors, for a brick house than for a reed building. Replacing the reed house with a house of cement brick, though, goes beyond economic interests. By transforming their house from a reed hut, associated with the stigma of rurality, backwardness and poverty, to a house of bricks they

materially express their urbanity, local understandings of decency and their right to the city. It is also a built form of resistance against the view held by the municipality that the houses in this area are illegitimate and temporary, which is why Polana Caniço residents do not have tenure security for their land (see also Gastrow 2017).

When in 2010 it was announced on national news that Polana Caniço would be redeveloped and that multi-story apartment buildings would be constructed, the topic dominated everyday conversations among the residents of the neighbourhood. In 2012, when Fernando and I revisited the interviewees, many immediately brought up the insecurities related to what they then called rumours and gossip (*fofoca*) about them being relocated by the state so that high-rise buildings could be built. The tenure insecurity felt as a result of these rumours was on a different scale to the ordinary lack of legal documents (DUAT) which is common for many Maputo dwellers living in unplanned or unofficially planned *bairros*. Being relocated by the state was not seen as a promising future by the residents we talked to; they feared that the compensation paid by the state would be much lower than private investors could pay. The state only gives compensation based on the evaluation of the house, residents would say, while private investors would also pay for the land which officially cannot be sold.

The residents of Polana Caniço have a low level of trust in the state and the majority conceive themselves as subjects of a demagogic state rather than as citizens with rights they can claim. When Fernando and I asked Zeferino, a 35-year-old household head in Polana Caniço, for his opinion on the relocation plans, he responded: "Who are we to say 'no' to the municipality? We can agree or disagree, but it is them who decide if they will arrange new land for us and if we have to leave from here" (Zeferino, resident of Polana Caniço, July 2012). Because the residents were aware that the state defines their land occupation as illegal and inappropriate, they did not feel that they had rights which they could claim. In addition, the state is the owner of the land, and not the residents. Senhor Justino, a 47-year-old resident of Polana Caniço, saw it like this:

The government is God, if he decidesto remove you, you will be removed, with or without the use of force. With which document will you show that you were allocated this land legitimately? Where will you go to complain? (Senhor Justino, resident of Polana Caniço, August 2012).

The expression "where will you go to complain?" (*você vai queixar aonde*?) is often used by urban dwellers when talking about the state. If an individual commits a crime against you, you can technically go to the police or take the person to the court. But if the state itself is corrupt and treats its citizens badly, there is no higher political order to which one can complain. Senhora Paula's house is located right next to the wall separating Polana Caniço from Sommerschield II and she and her neighbours were approached by private investors (*empresários*) in 2012 to negotiate a land sale. When asking whether she preferred to be relocated by the municipality or by private investors, 70-year-old Senhora Paula responded in the same way as many other Polana Caniço dwellers had told us before:

The municipality takes you away from here and then leaves you alone somewhere in the bush (*mato*), in aworse house than you have now,with only a living room and a bedroom. The private investor, though, will take you away and put you into another house,

and you can negotiatewith him to receive the samewhich you had … The state, though, does what he intends to do. The state does not discuss with you, but it gives orders. If you deny, the state removes you with force (Senhora Paula, resident of Polana Caniço, July 2012).

So the majority of the informants did not see the possibility of taking action against the state, which they regarded as potentially violent and demagogic. In order to avoid ending up somewhere in the *mato*, relocated by the state to a place which they did not chose and with a meagre house provided by the state, many of the residents had started to prepare themselves long ago by buying a piece of land somewhere on the outskirts of the city:

I have a field (*machamba*) in Katembe and a plot where I am constructing. On the day the governmentwill need thisland, I already have my land there. Many others are planning to go to Marracuene. Almost everybody is prepared here, almost everybody has something outside (SenhorMateus, a 55-year-old block representative in Polana Caniço, November 2010).

This means that the residents in Polana Caniço are not so much passive victims, as active agents in the *politics of proximity.* By already investing in land and a house on the outskirts of the city, many residents had created a sense of security in this context of uncertainties, as they would have a place to go to whenever the state started relocating. Land is then a major economic asset for households in Polana Caniço, whose livelihoods are usually based on a combination of the formal working-class income of a few household members (security personnel, domestic workers) and informal economic activities.

Many Polana Caniço residents do not think of themselves as having the potential to inf luence the state and are highly critical of the state's plans and the gentrification process. These include, for example, the residents of Casas Brancas; because they were relocated to Polana Caniço through the World Bank urban rehabilitation project, most of them possess usufructure rights (DUAT) for their houses and therefore also feel better equipped to defend themselves against the state: "The government gave us title to our properties, so the houses belong us. We can sell, if we want, but the government can't tell us to leave" (Senhora Aurora, December 2010). Private investors have also approached the owners of houses in Casas Brancas, telling them that they had better sell, or the state would remove them. Some residents of Casas Brancas therefore believe that the plans for the relocation by the state are rumours constructed intentionally so that the residents would be willing to sell.

In addition, many residents who do not possess the DUAT have a strong sense of ownership of their land. They see themselves as city makers and urban pioneers, turning the 'bush' into something 'beautiful':

Before everything herewas bush, itwasthe populationwho made it become beautiful. And now the investors want this space because it is indeed beautiful. It's the same out there, for example in Marracuene, it will become beautiful as well, and then they will arrive,tell usthattheywill pay us because they liked the place. Thenwewill move even further away. And once we made it beautiful there again, they will come again, tell us

they like it and of fer us money to leave. And the nation is growing (laughs) (Simão, resident of Polana Caniço, July 2012).

In the eyes of the Polana Caniço residents, it is not private investors and the state who make the city at first but them. With 'further away' Simão refers to the unworked, rural land without housing or infrastructure, which people call 'bush' (*mato).* Bush becomes beautiful (*bonito*) when the urban dwellers clear the land, build their houses and transform that land into a *bairro* belonging to the city, and hence into an urban settlement. In its early years, Polana Caniço was also *mato,* and it was the residents who cleared the land, built the houses, and only much later did the municipality install (very limited) urban services. In this same social process, the city is now growing at the perimeter, in the so-called peri-urban areas, with urban dwellers buying up rural land and transforming it into city, making it 'beautiful'. Because of this incremental logic of production of space in Maputo, driven by the agency of urban dwellers, many of Polana Caniço's residents anticipate that in the future, once they have settled in their new homes on the outskirts of the city, they will have to relocate again, because, again, the municipality and private investors will want to buy up their land, profiting from the fact that the urban dwellers, as urban pioneers and city makers, have made the rural land urban, valuable and 'beautiful'.

Because of this emic understanding of how the city develops, the majority of Polana Caniço's urban dwellers do not believe that f lats for Polana Caniço's residents which they can afford will ever materialise. For the youth whose public life is oriented to the city centre, moving into such a f lat would be ideal, as they aspire to such lifestyles and they could profit from the proximity to the activities in the city centre. Many older residents, however, especially those who have migrated from rural areas to the city, can hardly imagine themselves living in f lats; women could not pound maize and peanuts in a f lat and they could not keep chickens or open a small informal stall (*barraca*) to supplement their incomes. Besides these lifestyle-related values and aspirations, many Maputo residents have also observed that such housing projects aimed at 'ordinary' income groups usually do not work out the way they were planned. The people from the Cardoso building were initially meant to be only temporarily relocated to Casas Brancas, with the plan being that they would return to their city centre f lats after they had been rehabilitated. But hardly anyone could afford the new rents, so they stayed. In recent developments such as Intaka (Zimpeto), people had been relocated a few years previously because new houses had been constructed on their land. They had been offered new houses in the new development but would have had to pay the difference between the value of their old house and the new house, which nobody could afford.

Many residents believed that the state did not have the capacity to build these f lats for the residents. They also believed that gentrification was not about improving their lives, but about removing the slum-like conditions. One of the residents said in a conversation with us: "They will only remove the rubbish, not the beautiful houses" (resident of Polana Caniço, December 2010). As mentioned, many see the current processes and the state's behaviour as hypocritical in these gentrification processes. They have observed the way in which the municipality removed residents from the areas around Julius Nyerere Avenue considered unsuitable for housing and yet *Cidade Nova* was built

there, as Senhor Mateus, a 55-year-old block representative in Polana Caniço explained to us:

Thepeoplewithmoneywanttocomehere andorderpeople toleave.Downthere [south of Julius Nyerere Avenue] they claimed that when it rains, the houses get full of water, therefore people had to leave. The people lef t and nowthere are constructions(laughs). There are good houses now there where a Mateus had to leave (laughs). And it doesn't rain, does it? (Laughs) … it means they prepared the soil so that they could construct there. But they could not prepare the soils for us, but they threw the people to the outskirts(laughs)(SenhorMateus, block representative in Polana Caniço,November 2010).

Senhor Mateus criticised the discourse of inhabitability and insecurity employed by the municipality which served to legitimate the removal of the residents, while the new, wealthy occupants immediately received what they had been refused, namely, usufructure title (DUAT). Some residents then anticipated that there would be resistance to the removals: many would not want to leave and would give the municipality a headache (*dar dor de cabeça*). As some residents in Polana Caniço work for the government, they would feel entitled to defend themselves, some urban dwellers claimed in conversations with us in December 2010.

In 2012, a single small house standing on the road connecting Julius Nyerere Avenue to Marginal Avenue symbolised this potential for resistance by the urban dwellers in the *politics of proximity*. The small house was surrounded by empty land, waiting for new houses to be built and, as the previous *Caniço* houses had been destroyed, probably to make space for a new gated community or free-standing mansions. As residents in Polana Caniço explained to us, the owner of this house had agreed to sell his land to a private investor but was not content with the quality of the house – without electricity and water – that the buyers had built for him on the outskirts of the city. He therefore refused to leave before the private investors had provided him with a decent house.

### Suburban Dreams and High Hopes

In the ongoing gentrification, Sommerschield II's elite and Polana Caniço's residents rub shoulders, living in proximity to each other, and become slowly but continuously transformed in a relationship of distance, with many of Polana Caniço's residents literally expelled to the city's outskirts. The municipality, investors and urban dwellers do not conceptualise this as a process of elite residents moving into Polana Caniço, which could lead to a mixing of Sommerschield II and Polana Caniço, but in the local conceptions it is imagined as a new type of spatiality, *Cidade Nova*, which replaces the *Caniço.* In this process, the private investors profit from the fact that Polana Caniço's residents see the potential removals by the municipality as a threat, because it makes the residents more willing to sell to them.

Over the last few years individuals representing real estate investors have been approaching residents of Polana Caniço and asking them to sell their houses. According to interviewees, many of these *empresários* are Indians, but others are also Portuguese and African Mozambicans. In these negotiations, the residents agree to sell their houses if the private offer is good enough: they usually want compensation for the house or a new house, access to an alternative piece of land in a peripheral neighbourhood and some additional money to start or expand their informal economic activities.

During the process of negotiation or at the latest when they come to an agreement, the residents usually call someone from the neighbourhood structures, the block representative (*chefe do quarteirão*), to witness the process. The block representative or the neighbourhood secretary (*secretária do bairro*) then gives them a document identifying the plot and the name of the new user (Jorge and Melo 2014: 27). Urban dwellers consider these land transactions socially legitimate (see also Jenkins 2013: 76). This social legitimacy is enhanced and expressed by the presence of the local authorities in these transactions. Strictly speaking, though, these land transactions takes place outside the legal, formal and regulated realm. As explained in the introduction, from a legal point of view, all land in Mozambique belongs to the state and is in this sense 'public': there are no owners of land, but users get the usufructure title (Direito de Uso e Aproveitemento de Terra, DUAT) from the state (Jenkins 2009). Land cannot be owned and therefore cannot be sold, only the properties built on the land can be legally traded. Hence, the residents in Polana Caniço speak specifically about 'selling the house' instead of 'selling the land'. It is nevertheless clear to everyone that it is actually a transaction of land and not of the house. The buyers are interested in its favourable location, its proximity to the inner city, and they replace the small house of the former owner with a colourful elite mansion. Because these transactions constitute a grey area, the process and especially the prices paid are surrounded by secrecy. The neighbourhood secretary, for example, complained to us in August 2012 that she was never told how much the sellers received.<sup>6</sup>

In 2012, many interviewees told us that they or their neighbours had been approached by private investors who had initiated negotiations with them to sell their land. Senhora Paula is a 70-year old former domestic worker. Her former employer, a minister, had constructed a house for her in Polana Caniço where she was living with her children and grandchildren. Because her plot is adjacent to Sommerschield II, the land is especially attractive for investors and about three or four groups of buyers had already come to talk to her. She was only willing to sell, she told us, if the private investors would provide her with a nice piece of land, an equivalent house, some money to start informal economic activities plus transportation money for the children to go to school. Although this sounded like a wish list, it was probably not unrealistic that she would eventually find a buyer who would agree to her conditions. While in relation to the state, the Polana Caniço residents saw themselves as relatively powerless subjects, with regard to private investors, they felt they had more agentic possibilities. They could negotiate with investors because they had something to offer which the wealthy *empresários* wanted. As Senhora Jacinata expressed it:

When we came here, it was bush, there were snakes and monkeys. We fought to have what we have today. If now someone arrives and wants us to leave, we don't leave unless we are satisfied. And satisfaction is easy, they just have to pay enough (Senhora Jacinta, block representative in Polana Caniço, July 2012).

With regard to private investors, the residents conceive of themselves as economic actors in a field shaped by market forces. In this field of the market they are more powerful than in relation to the state. In their claims they refer to the investment in

<sup>6</sup>  For more in-depth research on the sale processes and pricing see Jorge (2016).

time, effort and money which they made in the past to develop the neighbourhood and their houses, and which now legitimises their demands for financial compensation and access to a new plot. This is also how the neighbourhood secretary, Senhora Cumbane, sees the ongoing land transactions: "The residents move away voluntarily after they have come to an agreement with the buyers. The buyers have to negotiate with the families and pay compensation. Nobody was forced to leave" (Senhora Cumbane, neighbourhood secretary, November 2010). Like municipal officials, the neighbourhood secretary also used a monetary rationale to remove the process from moral censure; she portrayed it as a free choice bringing economic profit, instead of coercion.

If also better than *mato,* the current spatiality of Polana Caniço with its high density in comparison to peri-urban neighbourhoods, its twisted sandy paths and organically grown, makeshift boundaries between irregular plots does not conform to the normative ideas of order and urban aesthetics among Polana Caniço's residents. The name *Polana Caniço* carries stigma, as the term *Caniço* makes it the quintessential *bairro* of the colonial City of Reeds with its urbanity conceived as deviating and inferior. A reed hut could be compared to what in South Africa constitutes a shack; a shelter, but not a proper house, and associated with poverty and backwardness. As mentioned, in the local understandings of the gentrification, Polana Caniço does not actually become merged with Sommerschield II or *Cidade Nova* but becomes replaced by it:

This neighbourhood is in transformation. In a couple of years, it won't have the official name Polana Caniço anymore. You can already see there how they are constructing, how they are buying property af ter property and construct those houses of Sommerschield. So Sommerschield is growing also here in some time. Everything will become Sommerschield. If I construct a multi-story house here, it's already Sommerschield, it's not Polana Caniço anymore (laughs) (Emílio, resident of Polana Caniço, December 2010).

Outsiders see Polana Caniço as a place of crime and alcoholism, and home to people with little education. Many residents of Polana Caniço have internalised or share these stigmatising images and only a few expressed a sense of ownership and identification in the interviews. In contrast to Alexandra township where many of the long-term residents identified strongly with the neighbourhood and proudly recounted its history, Polana Caniço's residents rarely expressed such pride. Some even jokingly described Polana Caniço as a rubbish dump (*lixo*, *lixeira*). Just a few informants talked about Polana Caniço in positive terms as a space of sociality and community; many spoke about suspicion among neighbours, witchcraft, malicious gossip and negatively connoted social control.

Their normative frame of reference, though, refers not so much to the ostentatious villas of neighbouring Sommerschield II but to the peri-urban areas which have developed in the last few years, like Kongolote, Intaka or Marracuene. Here urban dwellers can get larger plots than they have in Polana Caniço, they can keep chickens and grow food in large gardens. Many urban dwellers aspire to move to such areas, as there is "more space and more comfort" and it is "quieter and more relaxed" (Nelinha, 16-year-old resident of Polana Caniço, December 2010). In these peri-urban neighbourhoods, the local neighbourhood secretary and block representative often parcel up the area themselves, imitating what they consider to be the standards demanded by

the state (Jorge and Melo 2014). The new peri-urban neighbourhoods emerging today have a much higher aesthetic regularity than the unplanned, 'inhabitable' sections of Polana Caniço; urban dwellers mimic their images of state-defined urban standards, what Nielsen calls 'inverse governmentality' (Nielsen 2009, 2010a, 2010b) and Jenkins describes as "desire for a sense of physical order" (Jenkins 2012b, 2013: 177).

For this aspiration to a large, orderly, quiet plot, many urban dwellers are ready to give up the advantages that Polana Caniço offers: the proximity to the city centre with all its amenities like schools, hospitals, public life and other urban services. The price, though, is high: because of the lack of public transport the daily commute from the outskirts to school or work in the city centre can easily take two to three hours during rush hour, often at the back of an uncomfortable lorry instead of a much safer minibus taxi.

Here you have the advantage that everything isclose by. Out there, there are not even hospitals yet. But this will emerge. When the population in those areas grows, the government will think of putting a hospital there. It always depends on the movement of the people. Also here in Polana Caniço it was first bush (*mato*) and acacia trees. And nowit'sfull of houses and they are even constructing condominiums(SenhorMateus, a 57-year-old block representative in Polana Caniço, July 2012).

These areas generally have no electricity and water when the first urban residents decide to move there; they anticipate, as explained above, that the state or private entities will soon provide better services. They anticipate that they will maybe suffer in the short term because of the long distances and the lack of urban amenities on the outskirts, but in the long run, they anticipate, they will come closer to realising their dream of suburban living with a larger plot in a more organised, quiet neighbourhood. Some residents also appropriate the discourse of national development and the fight against poverty in order to legitimate these economic processes.

Barbara: So people who have money buy out the population here. Is this a kind of an expulsion by people who have money? Célito: They are not expelling us from Polana Caniço,they are rather helping us. Therewhere I go, Iwillconstruct a newhome. When another one arrives and will buy me out again, it means that the neighbourhood is developing, and the nation is growing. It's a good thing ... And for me, I have already won against absolute poverty (Célito, resident of Polana Caniço, July 2012).

In this conversation, Célito claimed that the gentrification enabled the poor residents to win the "fight against absolute poverty" and did therefore not constitute a morally problematic expulsion of the residents. Because of their suburban dream, some Polana Caniço residents are extremely interested in selling to private investors. What limits the residents' agency on the land market, though, is that not all plots are as equally sought after as Senhora Paula's. Many wish to receive visits from wealthy *empresários*, but they lack the networks for actively seeking buyers. Others wait for them to come back after their first visit, hoping that their high demands did not scare the investors off.

Many journalists and academics criticise the ongoing gentrification in Polana Caniço from a moral point of view, arguing that it entails an expulsion of the poor to

the outskirts of the city (Jenkins 2013, Jorge 2016, Jorge and Melo 2014, Lachartre 2000, V. R. 2014, Ventures Africa 2014, Vivet 2012). But this critique seems justified; what tends to go unseen is that for some of Polana Caniço's residents, it is also an opportunity to come closer to what they normatively see as ideal urban living: selling their land enabled them to get a larger, more orderly plot on the fringes of the city. The increase in the value of the land they lived on, resulting from the investors' interests in making the Cidade Nova grow even more, enabled them to realise their suburban dream. For others, though, the *politics of proximity* also brought losses and entailed social risks. Young Polana Caniço residents in particular saw it in negative terms.

### Social Risks

Many of the older family heads in Polana Caniço have experienced dramatic changes in the course of their lives, be it the shift from colonialism to independence, becoming soldiers or refugees during the civil war, becoming victims of natural disasters or becoming retrenched from a privatised company where they worked for decades. Although the social networks with neighbours in the neighbourhood are important for their economic survival, many of the older household heads are not especially concerned about disruptions to their social relations once they are relocated:

Barbara: Will you miss the neighbourhood if you have to leave? Senhor Mateus: If they remove me, I may missit. But also if you have a son, the son can die. If the son dies, you become sad, but he died. With time, you will forget, and a new life starts … I will make new friends there (laughs) (Senhor Mateus, block representative in Polana Caniço, November 2010).

It is then especially the young people who are worried about the emotional and social dimension of moving away. Nelinha, a 16-year-old schoolgirl, had lived all her life in Polana Caniço and had grown up in a relatively stable political context, if also shaped by economic hardship.

Nelinha: Polana Caniço is a great neighbourhood. Barbara: Why? Nelinha: I think it is because my roots are in this land ('terra'). They are saying that they will remove everyone from Polana Caniço to construct new houses. It will be very difficult for me to get used to a newplace. Whatwill hurt me mostisthe separation from the neighbours,the friends because they will be in dif ferent neighbourhoods. To start a new life out there will be difficult (Nelinha, resident of Polana Caniço,November 2010).

While for the older generation moving around and building up new social capital constituted their normality, for the youth the prospect of losing their stability and familiarity is unsettling. They anticipated that the relocation would disruption their social relationships and break the emotional connection they felt for Polana Caniço as a lived space, despite its stigmatised image. This f lexibility of the older generation and the fearfulness felt by the youth is remarkable, as it contradicts hegemonic images of elderly people as inf lexible and the youth as more dynamic.

Home ownership is an important value for young people in Maputo, who aspire to having a family. Constructing a house is considered to be part of the process of becoming an adult and a basis for a secure livelihood. For many people, home creation and

home building have higher priority than the social recognition of marriage (Jenkins 2013: 186). But young people like Nelinha, who do not yet have plans to start a family, have different preferences: their youthful lifestyle is rather oriented towards the city centre and their quality of life depends on urban proximity. Mariana, 24-year-old basketballer player, preferred to live close to the city centre; her life happened there, be it her basketball training, university studies or the places of public life like the restaurants and discos she goes to (Mariana, resident of Polana Caniço, December 2010). Other young people, who aim to go to university, aspire to finding a job in the formal labour market and participating in the middle-class public life taking place in the city centre, do not share the suburban dreams of their parents. For them, moving away from Polana Caniço and losing the proximity would mean an increase in living costs due to high transport costs and an exclusion from urban public life in the city centre. This raises a further aspect of what the *politics of proximity* in Maputo is about: quality of life in the city also depends on access to urban amenities like jobs, university and places of recreation. With the increasing land values and gentrification pressures this proximity becomes a privilege of people who can afford it, while less aff luent urban dwellers have to fight for a place in the *chapa* and travel long distances.

In 2010 and 2012 we also encountered residents like Simão who were sceptical about the promise of the suburban dream. Simão also dreamt of a suburban house, but his agency was less oriented towards an idealised, distant future and based rather on a critical evaluation of the immediate future. Simão and others rejected the view that Polana Caniço's residents are individualist economic actors, instead constructing themselves as social beings embedded in many not only supporting but also constraining social relations. They would come to the fore if a home owner received monetary compensation for the house, which would be a high social risk:

Some people may go drinkingwith the money. The person may be rich in that moment, but hewill be rich only for a month, and then the moneywill have evaporated, because money is nothing. Money is nothing. If you don't hold on well to the money, everything will go bad (Simão, about 36-year-old resident of Polana Caniço, July 2012).

For Simão, the large amounts of cash which the residents may receive as compensation are dangerous and a source of conf lict. He anticipates that criminals might hear about them and rob the urban dwellers of their fortune. Husbands might run away with the money, leaving their wives and children without a roof to sleep under. Another resident told us that in such a situation, neighbours and relatives suddenly have emergencies and exert social pressure on those with new fortunes to help them.

Investing surplus income in property, as members from both poor and wealthy milieus do in Maputo, is more rational than leaving it in the form of money in a bank account, where it is subject to inf lation. In contrast to a house, money is not a durable asset, it can be lost, stolen, used up for immediate consumption without long-term benefits, and it can be given away easily to relatives in need. Because of this scepticism towards money, many urban dwellers in Polana Caniço do not just ask for monetary compensation from the private investors but also want a new piece of land and often a house. Trading their house and plot for another one is then a safer form of transaction than transforming it into money which could become subject to normative expectations of sharing and helping out. More sceptical urban dwellers also think that moving

to the outskirts comes with long-term losses. Many urban dwellers in Polana Caniço rent out extra rooms on their properties to university students and other members of the transitory milieu. This additional income source would be lost. Many also consider the living costs on the outskirts as higher, not least because of the high transport costs. They fear that the immediate benefits may come with long-term losses.

Residents in neighbourhoods like Polana Caniço are often represented in the media and Afro-pessimistic discourses as poor, powerless actors living in slum-like conditions. One could narrate the story of gentrification as a discourse of neoliberalism pushing the poor out of central neighbourhoods to the outskirts. Media reports speak of 'bulldozing' and rising rental prices which push the poor out of the city centre (Ventures Africa 2014, V. R. 2014). Further research would be needed in order to establish systematically what variables (e.g. income, age or other milieu and lifestyle aspects) shape the actor's outlook on the gentrification processes and how the relocation to the outskirts affects their livelihood, public life and imaginaries. What the ethnography could unravel, though, is that in 2010–2012, the residents of Polana Caniço evaluated the current transformation processes in diverse ways. Uniformly, all of them perceived the potential relocation state as a threat, but many found that selling to private investors was an opportunity to improve their lives and work towards realising their dream of a suburban home. Many of Polana Caniço's residents are, if struggling to survive economically, also enterprising property investors and have already acquired plots in peri-urban areas or diligently negotiated with wealthy real estate investors to trade their plots for new ones. For some of the residents, the gentrification is a chance to realise the dream of suburban living rather than a misfortune. Others again were more pessimistic with regard to the realisation of these dreams.

What this section wanted to show is that in the *politics of proximity* evolving at the boundary between Sommerschield II and Polana Caniço, Polana Caniço's urban dwellers do not appear as a homogeneous group. While some of them see the gentrification as an opportunity to realise their suburban dream, freeing them from having to live in a stigmatised *Caniço* area, others see it as an exclusionary urban process which expels them to the distant outskirts of the city, taking away their privilege of living relatively close to the many amenities of the city centre, and entailing many social risks.

Entanglements are characterised by the ambiguous co-presence of two fundamental attitudes: encounter and distanciation. Encounter is, on the one hand, "an interaction where both actors perceive and recognize the difference of the other, respect it, and try to build on it in their relationship" (Förster 2013b: 242). Distanciation, on the other hand, is "an interaction where two actors adopt a disruptive attitude toward the other, trying to secure an independent agency" (ibid: 242). From this tension between proximity and distance, between togetherness and dividedness, between encounter and distanciation, deep ambivalences and even contradictions result, characterising the *cities of entanglements.* The *politics of proximity*, which is currently leading to the replacement of Polana Caniço with *Cidade Nova*, transforms the situation of proximity, with residents of Polana Caniço and Sommerschield II basically rubbing shoulders in a situation of distanciation: elite urbanites increasingly occupy the valuable land in the proximity of the city centre, while the less aff luent urbanites move to its outskirts. Instead of leading to a mixing or a dissolution of the colonial binaries *City of Cement*  versus *City of Reeds*, these processes lead to a multiplication of such spatialities.

By focusing on the *politics of proximity* in the urban area of Maputo comprising Polana Caniço and Sommerschield II, this chapter unravelled the complexities of the ambivalence between encounter and distanciation, between proximity and distance, in the contemporary, postcolonial capital of Mozambique. It expanded on what the *politics of loss*, the previous chapter on Linbro Park and Alexandra in Johannesburg, already started, namely to develop a comparative ethnography of entanglements through the analysis of the politics of neighbourhood constitution and boundary making in unequal cities. Both chapters focused on the boundary processes between neighbourhoods which exemplify some of the deep divisions characteristic for each city: in Johannesburg a formerly Black township and a formerly White, peri-urban suburb in transition, and in Maputo a *bairro* and a new elite area. Thinking about cities through their entanglements invites us to focus our lens on such moments and sites in which "what was once thought of as separate – identities, spaces, histories – come together or find points of intersection in unexpected ways" (Nuttall 2009: 11). The *politics of proximity* and the *politics of loss* both unravelled the way in which these points of intersection between the urban differences of two neighbourhoods become reconstituted as material, social and imaginary realities.

### **Maputo and Johannesburg: Comparing Entangled Neighbourhoods**

Entangled neighbourhoods exemplify and at the same time also transform colonial dichotomies of space and identity which constitute part of the colonial heritage of Maputo and Johannesburg. In the unmaking and remaking of boundaries between the neighbourhoods, types of neighbourhood emerge that are new but also need to be analysed in the context of the urban histories. Researching entangled neighbourhoods comparatively is a useful lens for unravelling the way colonialism still shapes African cities and how the legacies of colonialism are changing into something new such as the *Cidade Nova* and the like owing to the agency of urban dwellers, macroeconomic processes and actors like private property investors and the state. Researching entangled neighbourhoods takes us to the spatial legacies of the colonial city models, and the contestations around urban land unravel the ongoing powerplay between the different 'spatial projects*'* aimed at transforming the spaces in terms of their own visions. The transformations and contestations around urban space taking place at the boundaries between Sommerschield II and Polana Caniço and between Alexandra and Linbro Park are, therefore, not only about changes in these African cities as material space, but are also about the emergence and competition between new imaginaries and aspirations for the way urban spaces, urban lifestyles and urbanity in Maputo and Johannesburg should look in the future.

The ethnography has unravelled two key ways in which the neighbourhoods under study are entangled. Firstly, the neighbourhoods and lifeworlds are economically interdependent, which expresses itself in the labour entanglements, the everyday encounters between the residents of Linbro Park and Sommerschield II as employers and the residents of Alexandra and Polana Caniço as domestic workers, gardeners and security personal. The employers depend on their workers because they literally need to get their dirty laundry washed and the workers, in turn, depend on their employers for the salary. The second way in which the respective neighbourhoods are entangled

is through the politics of urban land and housing which potentially redistributes and reuses space in the larger urban area. In Johannesburg, this comes down to a *politics of loss* for Linbro Park property owners who struggle with the disappearance of their 'country living in the city' lifestyle. In Maputo, it comes down to a *politics of proximity* for Polana Caniço residents, many of whom trade their valuable land close to the city centre for a plot on the urban periphery. Sommerschield II's elite residents failed to erect a road closure to distanciate themselves from those they consider to be 'others'.

These two types of entanglements, domestic work and the politics of urban land and neighbourhood boundary making, are about deeply structural aspects of everyday life and, hence, about basic needs like having a place to live, earning some money, eating, washing clothes, rearing children, feeling safe and so on. This may distinguish them from the forms of entanglements the rest of the book is concerned with, the entanglements forming in and around religious spaces (see chapter 6) and encounters at the mall (see chapter 7), which are more about leisure and lifestyle and both entail a promise of symmetry and positive encounters, which the reality does not always live up to. While at the church and at the mall new forms of sociability may emerge (see chapters 6 and 7), in the realm of domestic work and the politics of urban land, structural inequalities tend to become reinforced in sociabilities characterised by spatial or social distance.

In the labour entanglements and the entanglements around the politics of proximity and the politics of loss, the invisibilisation of entanglements and the 'other', as well as the normalisation of structural inequalities, plays a considerable role in the practices of elite urban dwellers. The detailed analysis of domestic workers living in Linbro Park pointed to the many practices which the property owners use to deny their workers a 'right to the neighbourhood', a right to being recognised as equal residents. The obligation to wear uniforms, prohibitions against hosting family members in their quarters, and ignorance of the domestic workers' security needs on the badly lit roads are among the examples of invisibilisation practices. By normalisation I mean the many narratives elite (and partly also the less aff luent) urban dweller employs in order to legitimise what is going on. Scripted narratives in which Alexandra and Polana Caniço residents become 'othered' as criminals therefore play an important role. A further aspect of normalisation is that when Linbro Park and Sommerschield II residents spoke to us in interviews, they generally constructed the poverty of their adjacent neighbours as a problem while regarding their own wealth as normal. Linbro Park property owners emphasised how they had earned their land and houses through hard work, ignoring the years of racial privilege they had profited from during apartheid. Sommerschield II's Frelimo elite emphasised in interviews how they, as political and bureaucratic leaders, had gained access to the land legally and did not pay large bribes for it, normalising their privilege through their political merits.

Although the boundaries between the neighbourhoods are shifting and ample plans and interests exist for a more mixed future, from the perspective of the urban dwellers the neighbourhoods and the everyday life of their residents still constitute clearly separate worlds and a conviviality in proximity is barely imaginable, especially not for the aff luent residents who tend to 'other' the residents from the adjacent neighbourhood as prone to crime. This sense of separate worlds is not so much a lived reality as an aspiration, which reveals itself in the fact that in both Maputo and Johannesburg, the ideal form of living for the aff luent residents would be to have an enclosed neigh-

bourhood with walls, road closures and an appointed self-governing body granting them power and control. Yet in both the *politics of loss* and the *politics of proximity*, the aff luent residents fail to achieve this. They fail not least because they fail to mobilise enough neighbours for the collective actions needed, showing that there are diverging interests and weak social cohesion among the urban elites. It is the residents of Alexandra and Polana Caniço who manage to exert political pressure collectively, like *o povo* who tore down the road closure in Maputo and the many civil society organisations and political parties active in Alexandra.

As chapters 4 and 5 have shown, entangled neighbourhoods are spaces of competition where diverse urban actors struggle for power over territory and inf luence in the area's future. In these struggles over space and power, differences become politicised and instrumentalised in order to grant privileged access to resources for certain groups and legitimate the exclusion from these resources of other groups. 'Rich' and 'poor' (class), 'black' and 'white' (race), 'African' or 'Indian' (ethnicity) are significant differences that residents mobilise in these entangled neighbourhoods, and they correspond to the way in which the two cities are often discussed. The comparative ethnography, however, also unravels the importance of four further urban differences that go beyond these more commonly discussed categories, and which tend to receive too little attention in urban studies. Among them are (1) boundary making between the *established* and *newcomers*, (2) privilege related to *property ownership*, (3) competition between *abstract space-makers* and *social space-makers*, and (4) differentiation into *densifiers* and *de-densifiers*.

Firstly, in terms of intra-neighbourhood differentiation, all four neighbourhoods are characterised by ample polarisation between *established residents* who come to be regarded as insiders and *newcomers* who are considered urban *outsiders* (Elias and Scotson 1994 [1965], Wimmer 2013). This newcomer-established boundary can even by more powerful than ethnicity or class. In Alexandra, residents who recently moved to the township often have less knowledge about the township, have fewer social networks they can rely on, and their voices and views are less present in township politics than those of long-term residents. Who is considered a newcomer and who an established inhabitant has shifted over time, but what has remained constant is that these boundaries become politicised for exclusionary claims around access to employment and housing. In Linbro Park, the property owners, ironically descended from European settlers, have become the 'insiders', anticipating Alexandra residents becoming newcomers in 'their' suburb (for a similar case see Ballard 2005). Among the property owners, this creates anxiety and fear, emotions which are continuously reconstructed in the scripted narratives property owners tell each other about the irreconcilable differences. The dream of establishing a City Improvement District (CID) can be interpreted as an expression of the wish by established residents to impose control over newcomers. In Sommerschield II, insider and outsider distinctions are a very important boundary separating the Frelimo elite and the Indian merchant families who have a more transnational outlook and who moved to the neighbourhood more recently. The more established residents tended to construct the newly arrived Indians as 'others' in the interviews, as people who are corrupt, as people who bought their land illegally instead of getting it through networks of political inf luence, and as the people who initiated the second road closure attempt, although the other residents did not agree. There is also a considerable closure of networks between the two milieus which

also constitute themselves through differences in religion and occupation. In Polana Caniço, an insider–outsider boundary exists with, among others, Muslims from the northern provinces who tend to be treated as outsiders, many of whom moved to Maputo during the civil war and after.

Secondly, another urban difference that emerges from the comparative analysis of entangled neighbourhoods as very important is *property ownership*. On the one hand, owning property or not creates differentials in the residents' sense of entitlement to have a say in their neighbourhood, while, on the other hand, it creates important intra-neighbourhood relationships, namely between landlords and their tenants. I use the notion of property ownership here in a social rather than in a strict juridical sense: A narrow understanding of property ownership in Mozambique is problematic, as legally people can only possess the DUAT (usufructure title) and cannot own land. In Alexandra, it is equally complicated, as during apartheid the former property owners were expropriated, and now there is a multiplicity of socially more or less accepted claims to exclusive rights to land and houses which one could refer to as 'informal' ownership. Hence, I define property ownership broadly as including these diverse forms of socially accepted, exclusive land rights, ranging from usufructure title to expropriated property owners, acknowledging that the formalisation of these rights is often a very important political project in struggles around property ownership.

What emerges is that in all four neighbourhoods "the right to the city is expressed through home ownership" (Roy 2009: 85). In River Park, landlords and other property owners excluded tenants when they were having a neighbourhood meeting about the lack of electricity. The access to the debate was thus conditional on property ownership and access to housing. Roy (2003) conceptualises such links between property ownership and political citizenship as *propertied citizenship*. Many property owners in Alexandra have the sense that owning property gives them a special right to inf luence the future of their neighbourhood, even at the expense of those who do not have property (for more on this point see Heer 2018). In Linbro Park, the property owners have organised themselves into a residents' association called the Linbro Park Community Association (LPCA), believing that they as property owners alone constitute the 'community'*.* In Polana Caniço, one may observe that the milieus living in Casas Brancas, who own the DUATs for their land and houses, position themselves more self-confidently in neighbourhood politics. They feel less threatened by the rumours about potential removals by the state as they have legalised access to their land, and they criticise more openly the elites' failed attempts to erect a road closure. In Sommerschield II, residents complained that the more property owners rent out their houses to expatriates the more difficult it is to collectively mobilise, because contact with property owners is more difficult when they no longer live there themselves and because short-term tenants are less interested in doing something together to improve the area. Property owners and tenants may also have significantly different interests with regard to the transformation of an area, which makes it a very important urban difference in neighbourhood politics. In Alexandra, expropriated property owners have been blamed for 'blocking' urban renewal, as a court interdict related to their case prevents the urban renewal project from upgrading the 'Old Alex' sections of the township (see Heer 2018). A further aspect is that in Linbro Park, domestic workers were often not only employees but also tenants of their employers, which complicated their relationships and made them even more dependent on them. This linkage between domestic work as a

job and the workers' private lives as tenants makes domestic work a 'greedy institution'. Property ownership and landlord–tenant relations are very important differentiations in neighbourhoods which up to now have still not received the attention they deserve. They may also be conceptualised as a mode of entanglement, shaping conviviality in these cities.

The third urban difference which emerges in these entangled neighbourhoods in transition is the tension between urban dwellers and property investors whose relationship to urban land is primarily an abstract one based on its economic value (*abstract space-makers*), and urban dwellers whose interests are focused on the lived dimensions of spaces, meaning urbanites for whom the neighbourhood is primarily a place they call home and a place where they can live the lifestyle they choose (*social space-makers*). Over time, urban dwellers may shift from one orientation to another, like the property owners in Linbro Park. The Linbro Park property owners also felt that their neighbourhood was 'divided' and ridden by internal conf licts because of this: *abstract space-makers* were pushing for a mixed-use suburb, a vision in which they saw the promise of lucrative land deals, while *social space-makers* were opposed to this because they wanted to save their 'country living in the city'. Both in Linbro Park and Alexandra, residents had a strong sense of belonging and many constructed a good part of their identity based on their neighbourhood, which is also why the politics of urban land and housing is a very emotional issue for many. Sommerschield II, in contrast, is a very young suburb and many people we interviewed do not have such a strong sense of belonging, and it would make little difference to them if their mansion were in a similar neighbourhood like Bairro Triúnfo along the coast or Belo Horizonte in Matola. For older residents of Polana Caniço, potentially moving away from Polana Caniço was simply another change in their lives, similar to the many they had experienced before. For some, it could even improve their lives, as living in Polana Caniço was not something they felt proud of or nostalgic about. It was especially the young people who felt their way of being to be threatened by the potential relocation, doubting whether they would be able to lead a similar life with access to urban amenities and their friends elsewhere.

A fourth urban difference, closely related to *social* versus *abstract space-makers*, is the tension between *supporters of densification*, urban actors who aim to increase population density and the intensity of spatial use, and *opponents to densification*, urban actors whose ideals of urban living entail living on large spaces in a place with low residential density. In Maputo, many Polana Caniço dwellers found their dense settlement inferior to other neighbourhoods and they aspired to peri-urban living with a large stand and quiet on the outskirts of Maputo. This is actually very similar to the ideal of 'country living in the city' upheld by many Linbro Park's property owners who do not want to give up their lifestyle with their horses and large stands. The transformation taking place in both Maputo and Johannesburg has a similar impact on these opponents of densification: they eventually decide to sell their properties and move to the outskirts of the city – in Maputo the peripheral neighbourhoods, in Johannesburg peri-urban areas further out of the city. At the end of the fieldwork, it was not yet clear whether the announced high-rise f lats to be built for the population of Polana Caniço would become a reality. Nor it was clear when and how RDP housing would be built in Linbro Park, and whether the residents moving over from Alexandra would find the necessary urban infrastructure like public transport and schools there. So whether

ordinary urban dwellers who are *supporters of densification,* who are in favour of turning these neighbourhoods into more intensely used places, will actually profit from the ongoing transformation is yet unclear, and unfortunately, there are reasons to be sceptical. What is clear, though, is that private property investors, the municipality and property owners who are *supporters of densification* and are pushing for densification are already making profits.

What the comparative ethnography also points to is the importance of larger normative frameworks within which the transformation of the urban areas in Maputo and Johannesburg is taking place and which shapes morals and values regarding segregationist practices by the urban elites. In both Maputo and Johannesburg, aff luent urban dwellers tried to employ similar tools of spatial segregation, like road closures and appointing self-governance bodies like CIDs and *condominio*. In both cases, these segregationist practices were subject to moral judgements. The societal normative frameworks, shaped by the city-specific history of segregationist policies, inf luence the way different urban milieus talk and feel about the elites' segregationist practices. What emerges in the comparison is that in Johannesburg, elites' segregationist practices tended to be more accepted and more normalised than in Maputo. In Maputo, milieu differences were noted, with segregation being more normal for the Indian merchant elite than for the milieu of the Frelimo elite of whom many, at least in the interviews with us, criticised the road closure. What is normal and what is not also differentiates urban milieus (Förster 1997), and this is also shaped by the colonial past. What might play a role here is that many members of the Frelimo elite had distant family or acquaintances who lived in neighbourhoods similar to Polana Caniço. The lifeworlds in Polana Caniço may not be normal to them but they are at least familiar. The degree of segregation, the degree of invisibility is still today less pronounced in Maputo's elite lifeworlds than in those of Johannesburg.

What also needs to be mentioned are the significant differences in crime rates and, equally important, as Hannerz (1981) argues, the perception of danger by urban dwellers. Crime levels in Maputo have risen in the last few years and the recent kidnappings of wealthy Indians significantly diminished the sense of security, especially among the Indian business elite. However, crime levels are still significantly below the rates experienced in Johannesburg; I met far more informants in Linbro Park who had been a victim of crime or lost a family member in a murder than in Sommerschield II. This also contributes to the fact that 'othering' Alexandra residents as criminals is more normal and morally more acceptable for Linbro Park property owners than it is for residents of Sommerschield II.

The many elite residents of Sommerschield II belonging to the Frelimo party once fought against colonialism, supported a socialist regime and nowadays claims to engage in the fight against poverty. These social values stand in strong contrast to the desire to privatise streets and erects road closures. The public sphere, public moral outrage and public opinion hence play an important role in this Frelimo elite's agency, and when the story about the attempted road closure was broadcast on national news, they withdrew their support. In a country like Mozambique where there is the public perception that politicians are corrupt and unaccountable to their constituency, it is surprising to see that on the localised level of neighbourhood politics, public opinion indeed plays an important role in checking and balancing the actions of urban elites.

In Johannesburg, there have been debates on the ethics of road closures (Dirsuweit 2007), yet at the time of the research in 2010–2012 there was little public debate on what was going on in Linbro Park. Land restitution and expropriation were politicised topics in the national public sphere, yet there was only one instance when Malema, on visiting Alexandra, claimed that whites should be expropriated without compensation, referring to the white neighbourhoods surrounding Alexandra (Molatlhwa 2011). Although Johannesburg has quantitatively more communication channels such as national, urban and even neighbourhood media than Maputo, there is almost no public sphere linking Linbro Park and Alexandra. Residents and political organisations of both neighbourhoods hardly know what is going on beyond the N3 highway. Local politics is extremely insular, which also means that the problems and perspectives of residents of other urban areas are seldom considered (Young 2000: 214), with the exception of fellow property owners. As the political scientist, Iris Young, pointed out, one of the detrimental effects of segregation is that it impedes the formation of inclusive political arenas (ibid: 209).

One point of contrast in the respective urban histories of Johannesburg and Maputo needs to be pointed out to better explain such differences in the contemporary normative frameworks and also differences in the way in which urban dwellers talk about segregation. The Portuguese colonial government claimed that it upheld a policy of interracial mixing in the colonies, an ideology also called 'Lusotropicalism'. *Assimilados* and *Mestiços* were pointed out as proof of the apparent policy of mixing. The rhetoric of mixing was contradicted, however, by the reality of segregation that the Portuguese colonisers put in place (Nunes Silva 2016: 11-12), which left a lasting impact on the way urban dwellers talk about segregation. In Johannesburg, on the other hand, categorising people and spaces as 'black', 'white' or 'Indian' is done unref lectedly and is considered normal by the majority of urban dwellers. In Maputo, people also use such categories but actually consider them to be impolite and inappropriate. Linking spaces to racial or ethnic groups and economic classes is done with much more hesitation in Maputo than in Johannesburg, at least in frontstage situations. Johannesburg, in contrast, has been a draconian example of rigid urban segregation, and urban dwellers have been used to an apartheid rhetoric of racial segregation, and interpreting and talking about their urban world in terms of race continues to be normal. This may explain why the opposition by the white property owners to public housing does not create moral outrage among the neighbouring Alexandra residents: such attitudes and political practices are still very normal and widespread in the city.

A normative framework which shapes both cases distinctly is neoliberalism which has left its imprint on both countries since the beginning of the 1990s in terms of an ideology of the market, shaping urban governance and the economy of urban land. The municipalities' funds are limited, hence they depend on public–private partnerships and provide the legal contexts for self-governance bodies like CIDs and *condomínios*, if also imposing conditions which the elite struggle to fulfil. Real estate investors are in both cases a major force in the urban production of space, aiming to produce *abstract* commoditised space that serves the interests of capital and produces profits for the investors. In both cases, property owners prefer to sell their land to private developers rather than the state, because they expect to get a higher price from them. In contrast to negotiations with the state which can resort to coercion, both groups find that in negotiations with private developers they have a better bargaining position, being market agents with something valuable to offer. Urban dwellers who shift from *social space-makers* to *abstract space-makers* emerge as important actors, giving real estate investors an advantage in their competition with the state over the valuable and increasingly scarce urban land.

To sum up, when comparing entangled neighbourhoods in Maputo and Johannesburg, two modes of entanglement around land and neighbourhood boundary drawing emerged: the *politics of proximity* in Maputo and the *politics of loss* in Johannesburg. This comparative section sought to point the reader to some of the similarities between the two cities: among them the importance of boundary making between the *established residents* and *newcomers*, privilege related to *property ownership*, competition between *abstract space-makers* and *social space-makers*, and differentiation into *densifiers* and *de-densifiers*. It also aimed at highlighting differences in the degrees of moral judgements resulting from different degrees of segregation and the related ways of talking about race and class, while there are also similarities, resulting from the shared embedding in neoliberal economics. The aim of comparing Maputo and Johannesburg as *cities of entanglements* is, therefore, not so much to establish explanations of causality the way other comparative methods would do, but rather to deepen our understanding of the social processes at play in these cities.

©Barbara Heer, Sommerschield II, 2011

©Martin Dürrenberger, Maputo, 2015

©Barbara Heer, Intaka, 2010

## **Building Communities?**  Paternalistic Bonds and Religious Spaces

The comparative ethnography of entangled neighbourhoods has revealed that in the localised realm of neighbourhood boundary politics there are strong aspirations by urban elites to reinforce urban differences by defending old and imposing new forms of segregation. Although the visions upheld by municipalities tend towards mixed neighbourhoods, it is as yet unclear whether these will materialise. Elite neighbourhoods hence tend to be emptied of encounters with 'others', and if they do take place it is within the hierarchical and socially invisibilised work relations between employers and their domestic workers or security employees. What the chapters spoke little about is the venturing of the elite into the poor areas; mainly because it is quite rare. Only sometimes do employers visit their domestic workers in the township; they may go to express condolences to a distant relative in the *bairro* or they may make use of a cheap service offered by an informal business. In *cities of entanglements*, it is rather groups at the lower end of the social hierarchy that cross social boundaries into spaces dominated by those at the upper end of the social hierarchy, rather than the other way around. The same holds true in the following two chapters.

Chapters 5 and 6 will focus on two sets of urban spaces which entail the promise of something new; a promise of more equal spaces, of symmetrical relations and of the emergence of sociality and community across boundaries. On the one hand, this is the urban realm of religion, where, as the scriptures say, everyone should be the same in front of God, and it is the spaces of the shopping malls, where, apparently, it does not matter where you come from, as long as you know how to 'mall'. The two chapters examine the forms of encounters, the patterns of social relations and maybe even belonging (and exclusions) that emerge in these places, and the stories of hope, conf lict and disappointments urban dwellers tell. The potential of religious spaces and malls to build urban communities is, as the two chapters will show, vexed and ambivalent.

### **Religion in Entangled Neighbourhoods**

The significance of urban religion – as everyday routine, as ways of understanding and imagining the (urban) world, as a space of belonging and exclusion – has long been underestimated.<sup>1</sup> This is not least because many urban scholars stem from secular, Western backgrounds and have inhibitions or even a defensive attitude towards anything religious. Yet religious routines constitute a considerable part of the everyday life of many urban dwellers in Maputo and Johannesburg, as well as in many other cities across the world. Religious practices and experiences shape the imagination and the perception of cities, urban life and urban spaces. Orsi (1999: cover), for instance, claims that "[p]eople work on city spaces and realities in their religious practice, as the city works on them". In Africa, the anthropologist Filip De Boeck and the photographer Marie-Françoise Plissart have famously shown that Kinshasa "contains many cities in one," among them an invisible, spiritual city (Boeck and Plissart 2004: 16-17). As the editors to *Routes and Rites to the City* (Wilhelm-Solomon et al. 2017: 7) rightly point out, religion in Johannesburg has been little acknowledged and, if it was, then as a functionalist space where the poor could retreat to from what Murray calls a "heartless world" (Murray 2011: 171). In studies of Maputo there is also a lack of understanding of how much religion shapes urban lifestyles and social relations (an exception to this is Van de Kamp's research on urban women and Brazilian Pentecostalism, e.g. Van de Kamp 2016). Many strands of Islam and Pentecostalism understand religious practice and values not simply as a separate sphere of everyday life, but as all encompassing. Analyses of the urban with no attention to its invisible, spiritual dimensions miss this.

In Maputo and Johannesburg, Christianity is the majority religion (71% in Maputo, 76% in Johannesburg). Muslims constitute a minority, but Islam has become more publicly visible in the last few years in both cities (1.7% in Johannesburg, 5% in Maputo, Statistics South Africa 2004: 24, from census 2001; Conselho Municipal de Maputo 2010: 29). South Africa is less diverse in religious terms than Mozambique where some provinces in the north have a Muslim majority and where Muslims constitute between 20 and 30 per cent of the national population (Morier-Genoud 2007: 240). In both cities, spiritual practices related to ancestry, witchcraft and local forms of medicine and healing are also constitutive of the invisible urban realm. For some, such practices, which are locally coined as 'tradition', are loathed in the version of the Abrahamic faith they follow, while for others they form a normal part of everyday life.

Racial segregation under apartheid left its imprint on the religious landscape of South Africa. During apartheid, there were no segregationist laws governing religious spaces, yet the Group Areas Act indirectly broke up integrated churches as it violently destroyed racially diverse areas. The segregated living and large travel distances produced racially segregated religious spaces (Christopher 2001: 145). Even burial grounds became segregated. The politically dominant former Dutch Reformed Church fragmented into racially separate entities. As many churches legitimised segregation in their discourses, many Africans rejected European forms of Christianity, which evoked the mushrooming of African Independent Churches (Anderson 2005: 70). Christianity

<sup>1</sup>  Parts of thischapter have been published in the journal *Anthropology Southern Africa* (Heer 2015b, ©Taylor & Francis) as well as in a German version (Heer 2013). I thank the publishers and editors for the permission to reprint them.

also profoundly shaped resistance to the racist regime and played a role in the democratisation of South Africa (see also Comaroff and Comaroff 1986).

Among the new additions to the religious landscape in Johannesburg are Charismatic or new Pentecostal churches which grew in line with the charismatic movement in the US. The development of these churches first occurred along racial lines in South Africa. They emerged first in White areas (1980s) and then in Black communities (1990s), with both the township and suburb version appealing largely to upwardly mobile milieus (Anderson 2005: 75-80, Balcomb 2004: 18). While the White Charismatics had strong connections to Western religious trends, the new Black churches developed quite distinct characteristics as a result of the struggle against apartheid (Anderson 2005: 88). The gap between Black and White Charismatics was and continues to be also a gap between rich and poor; the white, middle-class churches based in formerly White suburbia tend to be much larger and wealthier (ibid).

As Christianity is so prominent in all milieus in South Africa, there were many hopes about the integrating potential of churches after the end of apartheid (Ganiel 2006: 13). With the growth of black middle-class milieus, immigration from other African countries, desegregation and the formation of new lifestyles in Johannesburg, the landscape of charismatic churches is also changing. Nowadays, many of the formerly White suburban churches are positioning themselves as non-racial or multiracial and aim to attract both black and white believers, for example, the Jubilee Community Church in Cape Town (Ganiel 2006, 2007, 2008), the Rhema charismatic mega-church in Johannesburg (Balcomb 2004) and the Rivers Church in Sandton. This chapter will focus on the LRC Church in Linbro Park (at the time of fieldwork called London Road Church, then renamed Love Reaching Communities Church*)*, also a formerly White Charismatic church which claims to be multiracial. Every Sunday, believers from the township Alexandra and the surrounding suburbs congregate at this Charismatic church to sing and pray together. It thus constitutes a space of encounter, a contact zone between diverse urban lifeworlds in the city, carrying the promise of symmetric encounters in a city shaped by inequality.

In Lourenço Marques, which is what Maputo was called during Portuguese colonialism, Catholicism was the official religion. During the socialist period religious organisations were first forbidden and then tightly controlled by the state.2 The change from a socialist regime to neoliberalism and the ideal of multiparty democracy in the 1990s also involved the liberalisation of the religious market. Since the 1980s religious activities have been on the rise in Mozambique (Morier-Genoud 2007: 238). Liberalisa-

<sup>2</sup>  Socialist Frelimo first banned vernacular spiritual practices as well as religions like Islam and Christianity. Christmas became *dia da família* (family day). The early post-independence Frelimo government pursued an ideology which emphasised that in the new nation there would be "no place for racial, ethnic, 'tribal', regional or religious dif ferences" (Bonate 2008: 643). The socialist Frelimo intended to erase 'obscurantist elements' which they perceived as 'backward' and contradicting the modernist norms ofrevolutionary 'scientific'socialism (Bonate 2007a: 57). Frelimo banned most practicesrelated to the invisible, spiritual world, like initiation rites, and traditional healing and ancestor-related rituals, aswell asreligionslike Islam and Christianity. Full religiousfreedom was only introducedwith the democraticconstitution of 1990. But already at the beginning of the 1980s, Frelimo reduced its anti-religious policies as they were generating negative international publicity (Morier-Genoud 2007: 234, Van de Kamp 2016). Frelimo also feared that the Renamo guerrilla movement might receive support from Arab or Muslim countriesif they continuedwith religiousrepression (Morier-Genoud 2007: 242).

tion has also left its material imprint on Maputo's cityscape; large Brazilian Pentecostal churches and new ostentatious mosques have been built. Since the 1990s evangelical churches have proliferated, especially Brazilian ones forming part of the spread of transnational Pentecostalism (Van de Kamp 2016).

Islam plays an increasingly important role in the Maputo. Islam made inroads into the northern Mozambican coast in the 8th century and became associated with the Swahili ruling elites (Bonate 2007a: 56). Over the centuries, a locally and regionally rooted conception of Islam developed (Bonate 2006, 2008: 638). After 1900, Islam expanded, caused by the arrivals of Sufi orders at the end of the 19th century, the development of markets and infrastructure under the increased Portuguese presence, and the social and economic insecurities caused by the colonial domination (Bonate 2007b: 138, Morier-Genoud 2007: 235). Nowadays, the northern provinces are the stronghold of Islam in Mozambique. In the academic literature, this form of Islam is often referred to as African Islam or Swahili Islam, belonging to the Shafi'i Sunni school of Islam, with Sufi inf luences (Bonate 2007b).

In the course of urbanisation, which accelerated after independence in 1975 and during the civil war, many Muslims migrated from rural to urban areas, from small towns to cities, and from the northern provinces to Maputo in the south. Nowadays, there is a considerable Muslim population in Polana Caniço and other Maputo neighbourhoods. Muslims came to Polana Caniço as war refugees and economic migrants from the northern provinces, especially from Zambézia. In 1998, northerners constituted five per cent of the population in Polana Caniço (Costa 2007: 31); in the meantime, this has probably risen. Many settled on the inferior land in Polana Caniço subject to erosion, and hence some were relocated to the outskirts of the city in the course of the rehabilitation of Julius Nyerere Avenue (see also Vivet 2012).

Maputo is also home to an Indian Islam, upheld by (descendants from) immigrants from the Indian subcontinent. The Mozambican Indians are religiously diverse: Maputo is home to Catholics from Goa and a minority of Hindus. The majority, though, is Muslim, a minority belonging to an *Ismaili* orientation of Islam, and the majority belonging to the *Hanafi* Islamic school. Like the African Muslim milieus, the Indian Muslim orientations are also shaped by different movements, among them also Sufi tendencies (Pires 2008: 19). As in other countries, the inf luence of the Saudi Arabian schools of Islam has increased in Mozambique in the last few years, among them the so-called Wahhabism, creating new differentiations and increasing competition and conf licts within and between Indian and African Muslims.3

According to Bonate (2008: 652), most of the differences among Mozambican Muslims arose during pre-colonial and colonial periods, but those related to ethnicity and race in particular declined after independence and were replaced by a main ideological divide between Sufis and the so-called Wahhabis. At the local level of the neighbourhood and in the everyday life of Muslims in Maputo, however, denominational differences play an insignificant role, with local conceptions of race and culture as well as economic inequality being perceived as the main lines of division. The Muslims living

<sup>3</sup>  In 2017, violence emerged in the northern province of Cabo Delgado, perpetrated by groups which appeared to be inspired by violentIslamic extremistideologies. Thisis having a considerable impact on the politics around Islam in Mozambique. The ethnography of this chapter is based on data collected between 2010 and 2012 and doestherefore not include such newdevelopments.

in Polana Caniço and Sommerschield II seldom used labels to refer to different Islamic movements. Because the denominational differences play a small role in their everyday life, knowledge about the complex denominational landscape constitutes specialist knowledge which they did not possess.

Until today, many members of Indian milieus live in the *Baixa*, the downtown commercial city centre of Maputo. In recent years, suburban living with a large house in a quiet neighbourhood has become fashionable among wealthy Indian milieus. Many moved to elite areas like Sommerschield II which nowadays has a considerable Indian Muslim population. In 2011, residents opened a Muslim prayer place, a *musallah* in a half-finished building in Sommerschield II. During its short existence (2011–2013) this Muslim prayer place was also frequented by some Muslims living in the adjacent Polana Caniço. This chapter provides insight into the religious relations between Muslim residents from Polana Caniço and Sommerschield II, based on interviews about visits to the *musallah* with its promise of entangling the different lifeworlds in less hierarchical ways.

Many religious practices follow a rhythm, and these rhythms again shape the rhythms of the city (Lefebvre 2010 [1992]), the spatial practices of urban dwellers, as well as public spaces and public life. Faiths bring together believers from different milieus on a weekly or even daily basis in the religious spaces where they collectively worship. On Sundays, the otherwise bustling city centre of Maputo calms down; the shops and offices are closed, there is little traffic and the passage ways also are empty as the informal traders stay at home, just like the many office employees, workers and shoppers whom they usually sell their goods to. The only people on the street are groups of men and women in formal clothes of Portuguese fashion (men in pants and shirts, women in skirts) on their way to one of the many churches. In a *bairro* like Polana Caniço, churches claim aural space through their singing; singing can be heard on the neighbourhood streets on Sundays, coming from the believers congregating in small buildings constructed of corrugated iron with a thatched roof and no electricity. Likewise, in Alexandra, the bustling street life calms down on Sundays and the distant sound of singing changes the township soundscape. While during the week the main traffic routes are used by working-class commuters wearing blue-collar overalls, on the weekends, churchgoers wearing differently coloured uniforms, indicating their membership of a specific congregation, take over the public spaces.

In downtown Maputo Islam is increasingly visible. Men and women in clothing marking their Muslim identity – among them white, long robes for men and diverse styles of veiling for women – drive and walk on the streets, stand in front of mosques and work in the many Indian-owned shops in the Baixa. Especially during Ramadan and on Fridays, men in prayer attire may be seen on their way to one of the five daily prayers at the mosque. In the evenings, the call of muezzins mingles with the noise of the traffic. More than twenty mosques, some of them large, new and prominently situated at major transport axes like the Masjid Taqwa on Eduardo Mondlane Avenue, claim the space for Muslims architecturally in the city. On Sundays, the Maputo Shopping Centre and a popular park, the Jardim dos Namorados, become transformed into spaces of Indian (Muslim) public life. During certain times of the year, like Ramadan or around Christmas, religious activities increase significantly, and religious rhythms may temporarily become more dominant in everyday life.

Christian church services on Sundays can take the whole day. For many women, Sunday services and prayer meetings during the week are the main social activities they engage in. Volunteer church groups have spiritual purposes like bible study, prayer and evangelisation, yet they are also support networks for church members: fellow church members visit sick neighbours, organise food for the destitute, provide moral and emotional support for each other in family conf licts or personal dramas. Besides the spiritual functions and social services, the churches are also safe spaces for sociality like the fostering of friendship and love relations, in a city perceived by many to be anonymous, shaped by self-interest and ill will. Young people who belong to church groups find friends there, and in Maputo especially, they may go out together to public parks, beaches or even a shopping mall. Many young people attend church. Here they find romantic partners with whom they share values and lifestyle. They may not, for example, drink alcohol, go out to bars or be gold diggers (*pessoas interesseiras*) who abuse friendship or love to gain access to resources. As Cecílio, a 22-years old resident of Polana Caniço, explained:

Nowadays, you can't trust anybody … My friend can for example go af ter my girlfriend … The friendships I want are not the friendships you can have nowadays. Friendships today are about personal interest, and if not aboutthat, it's about drinking. I don't drink and I don't like *pessoas interesseiras* (Cecílio, resident of Polana Caniço, August 2012).

The narrowing of networks along the lines of religious denominations, the preference for socialising and falling in love with fellow churchgoers of the same age are practices of protection from social ills in African cities which are seen by religious urbanites as places full of dangers and vices. Yet the security is not all encompassing; a boyfriend met at the church may be found to be cheating, an apparently religious girlfriend turns out to be a gold digger after all. Christian interviewees told me that some people who use witchcraft pretend to be highly religious in order to deceive people about their 'dark' practices. In the elite areas of Sommerschield II and Linbro Park, religion plays less of a role in everyday life than in Polana Caniço and Alexandra. Property owners in Linbro Park and members of the Frelimo elite have generally had a Christian upbringing, but nowadays they attend a church service only on special occasions like Christmas or a wedding. The Indian Muslims in Sommerschield II are the most religious milieu among the elites studied in this ethnography.

Habit and routine are, besides imagination and judgement, a key element of agency (Emirbayer and Mische 1998). Yet much research on African cities emphasises improvisation, f luidity and creativity and overlooks the routine aspects of African urban life. Urban religion and its related repetitive spatial movements, ritual practices and regular interactions contribute to the urban social order and consistency (Katsaura 2017: 12). Such religious routines become shaped by specificities of the religion as well as the urban form. Maputo is still characterised by a centre–periphery spatial form with central functions like formal jobs, services and commercial centres allocated to the city centre. Many workers living in Polana Caniço therefore frequent mosques in the *Baixa* during the day, where their workplace is situated, together with Muslims from other milieus and neighbourhoods, and in the evenings, they frequent the African Sufi mosque in their own neighbourhood that is less anonymous and less mixed. There are a few Muslims living in Polana Caniço who cross the neighbourhood boundaries and go and pray in the *musallah* in Sommerschield II. Indian Muslims living in Sommerschield II sometimes drive to the new mosque in the Bairro Triúnfo, a growing urban area along the coast, with gated communities, villas, shopping malls, hotels, offices spaces and now also a large mosque. The emerging polynuclear tendencies thus become ref lected in religious routinised trajectories of aff luent urban dwellers, impacting again on the forms of urban entanglements.

Many Muslim believers frequent a range of mosques and prayer places at different times of the day, depending on where their worldly responsibilities like work have taken them. People who are Christians tend to attend the same church over a long period, which is also the case in Johannesburg. For township dwellers transport costs can be high, so that practical concerns like the proximity of a church can be important. Churches on the boundary between suburb and township like the LRC Church are therefore attractive to township dwellers who have limited means yet are eager to circulate across boundaries and venture into spaces where they can meet new people. For aff luent milieus in Johannesburg the location of a church is relatively less important, as they lead car-based lives and can better afford to drive considerable distances across the spread-out, polynuclear city. There are also township dwellers who are ready to take a long minibus taxi drive to get to church, but then it is usually a church they joined when they were living elsewhere. So there are multiple ways in which religion shapes urban life in these neighbourhoods in Maputo and Johannesburg: on the one hand, religious rhythms constitute urban spaces and urban times in ways which ref lect and shape the spatial order of the cities. On the other hand, religion shapes urban sociality, be it social relationships between family members, friends or neighbours, entanglements across social boundaries within religious communities, and chance encounters between urban strangers. While bringing people together in new ways, the religious realm is at the same time also characterised by subtle and overt forms of division.

### **High Hopes and Deep Disappointments in Religious Encounters**

All human beings are the same, be it across the sexes, the races, there is no dif ference, be it between black andwhite. Important about being aMuslim isto be good to the others,to be honest,to serve as an example for others, and notto hurtthe next one (Senhor Jaul, local administrator of the Masjid Ioonusin Polana Caniço, July 2012).

God's kingdom is not, you know, for whites or blacks, it'sjust God's kingdom, we are all children of God (Terry, elder of LRC Church in Linbro Park, May 2012).

These statements by religious leaders in neighbourhoods in Johannesburg and Maputo claim that religion is an urban sphere where everyday differences – class, race or ethnic origin, gender – are dissolved.<sup>4</sup> In front of God, the quotes assert, urban dwell-

<sup>4</sup>  The ethnography of Muslim religious relations in Sommerschield II and Polana Caniço is based on interviews with residents as well as visits to prayer places and mosques collected between 2010 and 2012. As both my field assistant Fernando Tivane and I were not Muslim, and me being a woman, our possibilities to participate and observe in these spaces was limited. The data for the LRC church case

ers whose differences usually shape their interactions in everyday life meet as equals, united by their faith. Christianity and Islam, both Abrahamic religions, ideologically promise equality to their believers (e.g. Renard 2011: 162-164). Both the notion of the Islamic *ummah* and the Christian church express the idea of an imagined religious community constituted by equals, coming together in religious encounters, here understood as interactions between co-believers. Religious spaces are sites of everyday urban rituals where spaces become routinely shared, where habitualised encounters with difference take place, and new connections may emerge (Katsaura 2017: 3, 10). The promises of equality in the *ummah* or the church by urban religions raises hopes for the transformational potential of such habitualised encounters, yet there are also risks of deep disappointments.

### Uncovering Inequality in the Post-apartheid Rainbow Church

According to the leadership of the LRC Church, the roots of the church lie in the late 1980s when Rigby Wallace, a white South African who would later become the lead elder, founded a multiracial pre-school in Lombardy West, a suburb in the vicinity of Linbro Park and Alexandra. The fact that the school was open to all races is still considered worth mentioning on the website of the school, which nowadays exists as a private Christian school adjacent to the LRC Church, and leaders highlight this fact when telling the history of the church. In 2012, about 400 people were regularly attending the Sunday service according to Terry, who was one of the church elders in 2012. The church had about eight to ten ministries including children's and youth groups, a coffee shop run by the church, group-based bible study, and a social ministry called *Malukhanye* in which food was donated to the poor (see below). In 2012, the elders were planning to open a church branch in Alexandra; this was put in place in 2014. The church leaders call the church's denomination Charismatic. The church was aligned to the New Covenant Ministries International (NCMI), a group of allegedly loosely linked churches (New Covenant Ministries International 2005). NCMI belongs to the new Pentecostal churches which emerged in South Africa in the 1970s and 1980s, in response to the charismatic movement in the US.

The LRC Church is situated on the southern fringes of Linbro Park. Churchgoers from nearby Alexandra walk to the church, which is situated close to the entrance of the suburb, along a main road, London Road. More aff luent believers, usually coming from their homes in the suburbs, take the London Road off-ramp from the N3 highway to reach the church by car, often driving past fellow churchgoers from Alexandra who are on foot. The LRC Church was not always based in Linbro Park. The church moved several times in the northern suburbs until affordable land was found in Linbro Park in 2000. The church leadership saw the opportunity of being close to Alexandra as a chance to expand their church and tap into new religious markets.

entailed more participation and observation; the ethnography is based on many visits to Sunday services and the *Malukhanye* feeding scheme, as well as interviews and informal conversations with churchgoers, volunteers and the church leader Terry. Both the LRC church and the Muslim prayer place in Sommerschield II have changed rapidly and considerably since then. The LRC church opened a branch in Alexandra and united with a church in Boksburg. The *musallah* in Sommerschield II only existed between 2011 and 2013.

If you look over that way, it's incredible how close we are to Alex. And if you look there [towards Sandton] it's a crazy thing. I suppose it's not unique in the world at all, but there you have the af fluence and riches of Sandton – some of the best properties and prices for properties are there in Sandton – and right below that is Alex. We always felt that God has notjust placed us here for nothing. We really feltwe must have some footprintinAlex, because that'swhywe are here.We could have been anywhere, butwe felt thatthisiswhere Godwasleading us. So that'sthe kind ofchurchwe are (Terry, elder of LRC Church, May 2012).

The proximity of wealth and poverty which characterises this part of Johannesburg became interpreted by the church leadership as their religious calling and their location in Linbro Park an expression of God's will that they should evangelise in Alexandra.

It is noteworthy, though, that on the LRC Church website, the proximity to Alexandra is not mentioned in the description of the location, although there is ample reference to neighbouring aff luent suburbs (London Road Church 2013, Love Reaching Communities Church 2015). This omission is an expression of the ambiguity that characterises the church leadership's stance towards their township constituency; while they do have the goal to include black township milieus, they are also afraid of alienating aff luent suburbanites (for more details see Heer 2015b). This is also the reason for several changes to their name. First named Waverly New Covenant Church in reference to the suburb in which it was then located, the leadership renamed it with the Zulu name *Duduza* ('to comfort'), because they wanted to attract residents of the township where African languages dominate everyday life. As the name seemed to imply a black school and a township church, they eventually became afraid of losing suburban churchgoers and changed the name to London Road Church, referring to its new location on London Road. However, the London Road offramp from the N3 highway, which is on the border between Linbro Park and Alexandra, suffers from a reputation as a crime hotspot. Because the proximity to Alexandra comes with negative connotations, the leadership removed the reference to the locale and changed the name after my fieldwork in 2012 to Love Reaching Communities Church.

In 2012, the majority of churchgoers at the LRC Church were from middle class and elite milieus living in formerly White suburbs like Linbro Park or Lombardy West, as well as the new gated communities and townhouse complexes in north-eastern Johannesburg. While the majority of aff luent churchgoers were white, there was also increasing numbers of Indian and black middle-class families, as well as a number of racially mixed couples attending the church. This ref lects the growing black middle classes in Johannesburg and the increasing desegregation and diversification of Johannesburg since the 1990s (Crankshaw 2008, Southall 2016). In our interview, church elder Terry remarked, "We love the fact that we are 35 per cent black, and the rest white, roughly, because that's what it should be (Terry, May 2012). The LRC Church did not keep official registers of attendees during the period of my research, so these numbers are estimations. My informal conversations and interviews showed that black churchgoers tended to belong to the suburban middle classes, were domestic workers living in the surrounding suburbs and/or were township dwellers from nearby Alexandra. Many churchgoers from Alexandra started to frequent the church because they were recipients of food donations and prayers in the Malukhanye scheme. Many were unemployed and dependent on social grants and piece jobs. Some Alexandra residents

were better off but could not afford to move out of the township and considered the church a space where they could expand their social horizons. This is quite typical of the new Pentecostalism: many charismatic churches have a strong appeal to upwardly mobile and aspirational milieus (Anderson 2005: 67).

In its organisational culture, the LRC Church displays the three dimensions typical of the globalisation of Charismatic Christianity (Coleman 2000). Among them are the use of mass media like the internet and video clips to communicate its ideas, a social organisation with strong transnational links and management ideas derived from corporations, plus a global Charismatic meta-culture (Anderson 2005: 66-7). On the website, they portray the LRC Church as a 'bunch of ordinary people'. This ref lects the discourses employed by the NCMI, the group of churches they were aligned to in 2012. The mission statement (New Covenant Ministries International 2005) with the telling title *Who are these guys?* claimed that their churches know no internal hierarchies, but rather worked based on 'teams', 'partnerships' and 'relations'. This discourse of egalitarianism and informality, which downplays hierarchy, is encountered throughout statements by the leadership and in texts on their website. Like many other charismatic churches, their theology is shaped by the prosperity gospel (Coleman 2000), both rooted in and ref lecting the middle-class lifestyles of urban Johannesburg. Status symbols like iPads, ostentatiously used by the preachers for reading their sermon, and sporty, yet pricy, brand clothes play an important role in the religious performances. This may also be the reason why aspirational township dwellers attend the church; the sermons had little to do with the realities of township living, as the church leaders constructed for them the ideal, suburban consumer world to which they were aspiring and hoped one day to belong. At least on Sundays, they could be part of this world and build bonds with people whose lifestyles they aspired to.

Brian, a 30-something musician, grew up in Alexandra. In 2011, when he attended the LRC Church, he was living in Tsutsunami, a new section of Alexandra just across the highway from Linbro Park. Although his identity and everyday life were rooted in Alexandra, his lifeworld as a successful musician extended well beyond the township and he had professional and personal networks in the television and music industry. As a volunteer in various organisations, he knew many government employees and politicians. In 2012, his livelihood consisted of rental income from a f lat that he had acquired through illegal occupation, irregular music gigs and by working for a Community Work Programme, a government scheme targeted at under- and unemployed people. His steady rental income as informal landlord and his broad social networks put him in a better social position than, for example, rural migrants working as domestic workers who had to pay rent and who did not know anybody whom they could ask for help. Brian's mother was a devout Christian. Unlike many other young people in Alexandra who started drinking early, Brian did not drink alcohol for a long time, not least because of his mother's strong Christian orientation. But when his stepfather started to throw him out of the house at night when he was about 18, he had nowhere else to go but nightclubs and thus started drinking. After a turbulent time of drug abuse and crime, Brian started to go to the LRC Church in his early twenties, not only because he was living nearby, but also because he was curious to experience a different kind of church.

Bertha, a 30-year-old resident living in the River Park section of Alexandra at the time, was unmarried and taking care of a foster child and her sister's children whom

she was living with. Bertha and her sister owned a house and supplemented their income by renting out shacks in their backyard. They had received the house from the government after they had been internally displaced due to township violence between hostel dwellers belonging to rival political parties in the early 1990s (Bonner and Nieftagodien 2008: 359). Although Bertha had no other income, the household itself was financially sound due to the monthly rents, her sister's salary from an office job and social grants. Worrying that she had not yet found a husband, Bertha has been going from church to church seeking spiritual support. Bertha can be described as a church shopper (Brunn, Jetton and Palmquist 2015), trying out different spiritual remedies. When LRC volunteers came to evangelise in River Park in the 2000s, she decided to attend a church service: "[I went there] because I love to observe too much, I love to learn too much, so I went there. And it was incredible" (Bertha, June 2012).

For both Brian and Bertha, interacting with people from diverse milieus at the LRC Church was something desirable and exciting: a key motivation to join the church was their desire to cross the boundaries of their everyday lives. They perceived the church in a very positive light as a space enabling them to become entangled with other urban lifeworlds to which they were otherwise little connected.

I went to the church because I just wanted to experience a dif ferent environment. I grew up in a church, you know, but now, I just wanted to see how white people praise. Do they accept black people or they pretend like they do? Because you know, we are from the apartheid era, some of us still don't exactly accept the rainbow nation, you know(Brian, resident of Alexandra, June 2012).

For Brian and Bertha, the promise of multiracialism at the church, the coming together of white and black urban dwellers, raised the hope and expectation that the church would finally be a space where new forms of sociability (what the Anglican theologian Desmond Tutu called the 'Rainbow Nation') would be possible.

While for white suburbanites visiting the LRC was moving through familiar territory, churchgoers from the township like Brian and Bertha had to cross everyday social boundaries to get there. I call such practices of crossing everyday boundaries *circulating*, drawing on Simone (2005b: 519), who defines circulation as "practices that enable residents to navigate and engage diverse spaces, actors, sensibilities and activities across the city in its entirety, or at least across domains larger than the quarters where residents work and/or live". For many economic and social reasons, urban dwellers circulate beyond their everyday world to actively seek encounters at the LRC Church with residents from other social backgrounds. Some do this in, amongst others, the hope of creating new connections which could help them to find a job. Circulation is, therefore, a key urban practice for the creation of new urban entanglements.

For Bertha and Brian, the LRC Church was initially unknown territory and their first visits entailed some degree of uncertainty. Both had interacted with white people before, but not as congregants of the same religious space. As Bertha recalled, "I was there and I was a *black* person in front of *white* people. So I didn't know how to react and I didn't know what to do" (Bertha, June 2012). Yet, when I asked each of them about their first visit to the church, both Brian and Bertha were nostalgic. Bertha declared that "It was wonderful. I loved it," whilst Brian proclaimed, "I felt great, you know, because it was for the first time that I was among all the South African races" (Brian, June 2012). They associated their co-presence with white churchgoers with the ideal of the post-apartheid society, what Brian referred to in the quote above as 'rainbow nation'.

With time, they learnt to move securely through this new social environment, which they both evaluated as a personal achievement. But the initial fascination and enthusiasm soon diminished when they became better acquainted with the church and its believers and began to feel that they were being dealt with differently to the white churchgoers. Hoping to experience encounters shaped by symmetry, they were disappointed when they felt that the broader historical and contemporary inequalities nevertheless made their inroads into the entanglements emerging at the church. Positive experiences therefore became juxtaposed with feelings of exclusion whenever they felt that they were being treated as 'others'.

One type of situation sometimes causing disappointment was the manner in which they were greeted when arriving at the church. Before Sunday morning church services, elders and volunteers usually stood at the entrance of the church to greet the churchgoers, performing friendship and familiarity in an informal, casual manner. Brian and Bertha sensed that the people welcoming them to the church felt restrained towards them, in a similar manner to how they had felt when they visited the church for the first time. Both found that the greetings between black and white churchgoers were shorter than between whites. Brian noted that white churchgoers only looked brief ly into his eyes and then quickly looked away. Bertha had observed the practice of complimenting young mothers on their beautiful babies and emulated it. She noticed, however, that when she made such a compliment, the young white mothers tended to move on from her relatively quickly, while the same compliment would lead to a much longer conversation if both churchgoers were white. Bertha and Brian both concluded that their white interlocutors did not feel 'comfortable' in their presence and possibly even felt scared. As Bertha remarked, "When we hug them or talk to them, you can *see*  that ... white is white. Like they are afraid" (Bertha, June 2012).

According to Goffman (1959: 9), social actors create common ground in a shared situation by coming to a temporary agreement on their social reality and on the roles and identities that each will assume (see also Crossley 2011: 29). The situational definition created in the LRC greeting ritual entailed a definition of equal partners in a religious relationship marked by informality and familiarity. Situational definitions also have a moral character (Goffman 1959: 13) which, in the LRC Church, created a context in which Brian and Bertha expected to be treated in the same manner as white churchgoers. Goffman (1959: 53-55) points out that each participant's belief in the authenticity of how the other participants presented themselves could easily be jeopardised. Thus, the insecurity and discomfort which Brian and Bertha sensed in the white churchgoers (whether true or not) highlighted for them a contradiction between the performance of equality and friendliness and their experience of insincerity. They came to interpret the interaction as an inauthentic show which Brian described cynically in the following exchange:

Brian: If you are there for the first time, they make sure that they cover [up] all the negativitiesthat you may see. But when you go there continuously, that's when these scandals or illnessesstart to now reveal themselves. Barbara: Can you give me an example?

Brian: Racism …When times goes by, you see that, no, [these] people,they are notreally comfortable amongst black people (Brian, resident of Alexandra, June 2012).

In terms of urban entanglements, these greeting rituals constitute on one level of experience, namely the situational definition, a transgression of the spatial and racial boundaries shaping life in the city, yet on another level of experience, the perception of discomfort felt by the 'other' leads to a reaffirmation of the very same boundaries. What emerges are, therefore, entanglements with contradictory meanings, shaped by a tension between the new and the old, between being together and remaining apart.

Bertha and Brian also drew on other situations in the church in their critical evaluation of the performances of white churchgoers and elders. Brian, for example, noticed that the same white churchgoers who would greet him so warmly at the church entrance had driven past him as he was walking to church without offering him a lift. In his view, this would be a moral obligation if they were indeed equal partners in a religious community:

No one stops and gives you a lif t. If it has been raining, and if the road has got potholes, youwill getsprinkled by thewater. They just pass by. Some even hoot orlif ttheir hands to show you that they can see you. Then atchurch, when you enter, now they startsmiling at you (Brian, resident of Alexandra, June 2012).

On the terrace outside the church building, the church leadership put up a table where churchgoers could help themselves to free coffee and tea after the church service had ended. Inside the building, the church had opened a coffee bar with comfortable couches which sold expensive cappuccinos or latte macchiatos. Here people who visited the service for the first time were invited to a free cup of coffee as well as cake to ensure that they would stay for a while and chat with church volunteers. Yet, the creation of two social spaces shaped the distribution of believers across the church building and thus the manner in which they interacted. The apparently well-intended act of providing free drinks in a site that was spatially separated from the coffee bar that offered drinks for sale meant that churchgoers of the suburban middle class congregated inside the church while those who availed themselves of the free drinks, usually poorer township congregants, would socialise on the terrace outside. Brian recounted his memories of his first visit to the church:

If you are coming for the second time, on the second week, what they will do is, if you can't af ford the cappuccino and the espresso,then youwill have to go outside the building and just watch and make yourself the cheap cof fee (Brian, resident of Alexandra, June 2012).

This informal micro-segregation created a form of social mingling after the church service which reminded one of apartheid racial segregation: the outside crowd tended to be black, the inside crowd white. Brian believed that the church leadership had purposefully made this arrangement to protect white churchgoers from the discomfort he felt they were clearly experiencing in the co-presence of black believers. The church leadership, aware of the micro-segregation, also interpreted it like Brian, in terms of racial categories. This quote by the elder, Terry, for instance, shows that in his interpretation of the micro-segregation was about black and white churchgoers who do not mingle because of their race:

You may see on a Sunday that many of the black folk are sitting here, in the sun, talking to each other. And then you look in there [at the cof fee bar] and it looks like there are very few black guys af terchurch, because they are all mingling here. We thought: "No, this is wrong, they must mingle." And then we realised ... we can't force things! So if that'stheway people feelcomfortable ...We love everybody in the church,we see them as brothers. But if they are meeting here, and that'stheway theywanna meet on a Sunday, and the white guys wanna sit there, then so be it! (Terry, elder of LRC Church, May 2012).

Like Brian, Elder Terry regarded the micro-segregation as a personal issue that resulted from churchgoers' desire to create comfort zones around them, spatially separated from those belonging to what were perceived as different racial groups. Since the church leadership framed it as voluntary self-segregation, they did not regard it as their responsibility to interfere. In practice, though, they could have easily altered the spatial arrangements of the two spaces and lowered the prices of the coffee sold to remove the economic exclusivity of the coffee bar. However, as church structures were dominated by middle-class white males, it was their norms and values, rooted in a suburban middle class, which were taken as given and implemented. It was not an absence of sensitivity by the church leadership to race, but rather a lack of attention to class difference which produced this micro-segregation. Both Bertha and Brian stopped attending this church, not least for these reasons. Their high hopes of new forms of entanglements at the church turned into deep disappointment.

### Colonial Images in the Postcolonial Mosque

Senhor Justino originates from the province Zambézia in northern Mozambique, and was living in a simple house in Polana Caniço, relatively close to Casas Brancas and Sommerschield II. During the civil war, when he was studying to become a teacher, he was conscripted to the army, and he was never able to return to finish school. Deprived of the possibility of social mobility through education, he nevertheless decided against the harsh rural life of the field (*machamba*) in Zambézia, and when he was released from the army in 1992, he set off for Johannesburg to work on the mines. However, like many others on their journey to Johannesburg, he was stranded in Maputo and eventually decided to stay there. In 1996 he was able to buy a little house in the unplanned section of Polana Caniço A. Since then, he had developed strong networks in the neighbourhood and was well known, yet he retained a rural orientation and was planning to return to Zambézia when he is old. His wife returned to Zambézia in 2009 and since then he had been living in Maputo without her, with three of his children and other relatives. His house had become the entry point and initial safe haven for family and kin from Zambézia who wanted to start a life in Maputo, where economic prospects are greater than in the north.

His everyday life in 2010 and 2012 was shaped by the rhythms of the household, the rhythms of the world of commerce in the Baixa and the Islamic rhythm of prayer routines. He was working as a stockman at a trading company owned by a Muslim family of Indian descent in the city centre. Because life was increasingly expensive, he

preferred to walk instead of taking public transport. During the lunch break, he would sometimes go to pray at one of the mosques in the Baixa. Senhor Justino usually did his grocery shopping at the local market in Polana Caniço, the Mercado Compone. He did not frequent bars or restaurants but, like many men of his age (he was around 47), he preferred to drink alcohol in the private space of the home, where his drinking could not be seen by others; not least because as a Muslim he was not supposed to drink. On the weekends he sometimes took a minibus taxi, a *chapa,* to visit friends in other *bairros*, friends who typically were fellow Zambezians, co-workers or former neighbours.

He did not always conduct the obligatory five prayers per day, one of the five pillars of Islam, but especially on Ramadan, the rhythm of praying, fasting and breaking the fast would dominate his routines. Since about 2007, he had been going to a neighbour's backyard where a group of Muslims assembled to pray together. Recently, the group leader had managed to receive a generous donation by a wealthy Muslim woman from South Africa, and he constructed a mosque in Polana Caniço, with the name *Masjid Ioonus*. Since 2012, Polana Caniço has had a new, bright green painted mosque with arches and decorations alluding of small minarets, architecture which expresses the mix of Arab and African inf luences so typical of the northern parts of Mozambique. It is probably one of the most representative buildings in the whole *bairro,* where most of the residents cannot usually afford to invest in plastering or painting their simple houses. In comparison to the new inner-city mosques, it is, nevertheless, still a very modest construction and blends neatly into the neighbourhood. Here Senhor Justino would come in the evenings during Ramadan for a simple *If tar*, the communal meal shared at the mosque when the fast is broken after sunset. From 2011 till 2013, Senhor Justino also frequented the prayer place (*musallah*) in nearby Sommerschield II, a couple of hundred metres from his house, just across the *Rua do Cravo*. Because of this proximity he would sometimes cross the social and spatial boundaries between Polana Caniço and Sommerschield II, and pray there, side-by-side with residents from this adjacent elite area.

One of these residents is Senhor Ismail. He is a Muslim of Indian descent and a business owner. He bought a house in Sommerschield II in 2009 and moved there after a Muslim Indian architect had refurbished the house to make it suitable for hosting an extended Muslim Indian family. In 2010 and 2012, he was living there with his wife and children and regularly hosted his parents and siblings for prolonged periods. At 60 years old, this household head was suffering from high blood pressure, which is why he had to disconnect himself from the stressful business rhythm of the city. Instead, his everyday life and everyday trajectories through the city in 2010 and 2012 were shaped by the needs of his body, the rhythm of praying as well as the rhythms of meetings with the various religious and political organisations he was involved in.

Senhor Ismail aspired to go to the mosque daily and to pray five times a day. He often used to get up early in the morning for the first prayer and then went back to sleep. For the second prayer around midday, he would take a drive to the beach to pray at the new impressive mosque *Abdurrahman Bin Auf* in the Bairro Triúnfo. For the third prayer in the later afternoon he would attend a mosque wherever he was at the time, for example the *musallah* in Sommerschield II, a mosque in the Baixa or a mosque close to his office in the inner-city neighbourhood Alto-Maé. The two evening prayers he often did at home. He would meet many friends and acquaintances at the mosques

and chat before and after the prayer. Yet there were also many people, especially at the inner-city mosques, praying side by side with him who were strangers to him.

Senhor Mattar, the 55-year-old administrative head of the Muslim prayer place, explained that Muslim families in Sommerschield II had developed the need for a prayer place in the neighbourhood because friends and relatives had become victims of kidnappings, and driving around came to be considered a risk. With a prayer place in the neighbourhood, it was also be easier to perform the five prayers:

We sawthatfrom here [the neighbourhood]itis a bitfar away [from the mosquesin the city centre],which makes you lazy to go and pray, butwe have to pray. So that's howthe idea came up (Senhor Mattar, resident in Sommerschield II, August 2012).

An Indian Muslim property investor who was building several new houses in Sommerschield II offered the Muslim families a room in a half-finished house which they could use as a prayer place until it was sold. Senhor Mattar explained that the owner of the building would pay for electricity. When smaller expenses needed to be made, Senhor Mattar would ask the thirty-odd families involved to contribute. The *musallah* was located in the rear section of the neighbourhood, basically on a construction site, so only informed insiders knew that it existed. In spatial terms, the *musallah* constituted part of the invisible city, only visible to those who knew about its existence. Men would come here to pray, and in the afternoon the room would be used as *madrasa*, as a place for the religious teaching of children. According to Senhor Mattar, residents from Polana Caniço prayed there, but only a few, "as they have their own mosque there" (Senhor Mattar, resident in Sommerschield II, August 2012).

While the LRC Church in Johannesburg becomes a site for entanglements between the suburb and the township, the *mussalah* in Sommerschield II does so less for reasons related to the specificities of Islamic religious practices and more as a result of the spatial form of the city. Maputo continues to be characterised by a centre–periphery model with central functions focused on the city centre. This monocentric urban form shapes religious spatial practices and the spatial location of encounters: neighbourhood mosques tend not to attract urban dwellers who live elsewhere, as they either attend a mosque close to where they live or in the city centre. Senhor Justino worked in the city centre where he prayed during the day, and only sometimes in the evenings did he attend the *mussalah* in Sommerschield II, as his house in Polana Caniço was very close to the boundary between the two. The majority prefer to visit Polana Caniço's own mosque because of its spatial and social proximity. In general, the Muslim interlocutors in Maputo were less bounded to one specific religious space than the churchgoers in Johannesburg: believers attend various mosques and Muslim prayer places at different times of the day, depending on where their worldly responsibilities like work take them to.

### *Religious Ideals and Urban Reality*

In Islam, everybody is equal, be it between the sexes and the races. There are no dif ferences, be it black orwhite. Central for being Muslim isto do good to other people,to be honest, to be a good example, and not hurt others. Islam means peace. Unfortunately,

some people don't follow these rules (Senhor Jaul, local administrator of the Masjid Ioonusin Polana Caniço, July 2012).

A striking feature of the conversations Fernando Tivane and I held with leaders of local mosques and prayer places was their pointedly positive portrayal of their religion and of Islamic values. They were eager to explain to us the importance of normative prescriptions on how Muslims should live their everyday lives, like the five pillars of Islam. Many would emphasise that their religion is benevolent and based on values like humility, modesty, equality, non-racialism and tolerance. The emphasis of benevolence and non-violence in the conversations with us as non-Muslims may have been a defensive strategy against stereotypes which equate Islam with terrorism. Also, the Muslim urban dwellers we interviewed sometimes referred to Muslims who they saw as very conservative and potentially under Saudi Arabian inf luence as people with long beards *(homens de barba grande*) or, jokingly, as *Al-Qaeda.* 

Mozambican Muslims tend to be characterised as deeply divided, with different factions competing with regard to proper Islamic doctrine, the Islamic calendar, the right to represent the Islamic community, political inf luence and access to resources, both of the state and of transnational Islamic organisations (Alpers 1999, Bonate 2007a, 2007b, 2008, 2010, Morier-Genoud 2000, 2007). This stands in contrast to the extent to which Muslim leaders and ordinary Muslims downplay differences among them and emphasise their equality, despite cultural and other differences. Islam, we were told by various religious leaders on the neighbourhood level, does not know the separation of races, classes or cultures, but for prayer, everybody would come together. The emphasis of values like equality by Muslim leaders also shares similarities with and is maybe also inf luenced by the Frelimo discourses of socialist times when regional, racial or ethnic differences were not supposed to be acknowledged, and when wealth should neither be accumulated nor shown conspicuously.

The Muslims from Polana Caniço referred to themselves as 'black', 'African' or 'from this land' (*da terra*), while they categorised the Muslims living in Sommerschield II as 'white', 'Indians' and/or 'merchants' (*comerciantes*). Unlike in Johannesburg, racial terms (*negro, branco*) were used only in a few informal situations in Maputo, as people feel they have a colonial baggage and negative connotations. In the following exchange with Senhor Mattar, who was responsible for the prayer room in Sommerschield II, I brought up my observation about the differences in the local forms in which people articulated notions around race in Maputo and Johannesburg.

Barbara: There is a huge dif ference in how people talk about race here. In South Africa people are very used to speak of black and white, while here people don't really use these terms so of ten. Senhor Mattar: Here we don't really have this [segregation] a lot. Herewe laugh together,we are alwaystogether. We for example sitin the mosque next to our employeewho is Muslim,who is black and prayswith us, eatswith us at the time for eating during Ramadan, and is there with us. It makes no dif ference in colour (Senhor Mattar, resident in Sommerschield II, August 2012).

In this conversation, Senhor Mattar brought up the example of praying together in an inner-city mosque to illustrate that Mozambican society is less racially segregated than South Africa used to be. He portrayed the co-presence of 'black' and 'white' Mus-

lims in the moment of praying as something especially noble and valuable about Islam. When we asked Senhor Adil, a 25-year-old religious teacher at the mosque in the *bairro* Costa do Sol, a neighbourhood adjoining Polana Caniço, about what categories of believers attend his mosque, he explained that Islam is actually a world religion which does not belong to a specific race, but: "I don't want to diminish nobody's status, but I can say that the people who come here are all of the black race" (Senhor Adil, July 2012).

Feelings of cultural belonging and cultural distance play an important role here. Locally, these differences are articulated as different 'origins' or 'upbringing', and sometimes also as 'race'. In the following exchange, Fernando and I asked Senhor Justino, a Muslim resident of Polana Caniço, to compare his perception of the *musallah* in Sommerschield II with the Masjid Ioonus in Polana Caniço. Senhor Justino, who sometimes went to this specific *musallah,* responded:

For example, if I die, a white man can come here [for mourning] but he will not spend thewhole night. Itwill be my…let'ssay,the people of my racewhichwill accompany my family. They [the Indians]will assistthem, help them, buttheywon't drink tea.One also feels more comfortable side by side of the brothers. In the Qur'an, apartheid does not exist, butwell … (Senhor Justino, resident of Polana Caniço, August 2012).

In this conversation, Senhor Justino expresses the physical and emotional experience of difference using the notion of 'comfort'. Like other Muslim interviewees, he compares social distance at the mosque to apartheid South Africa and not to Portuguese colonialism. This is perhaps because apartheid existed 20 years longer than Portuguese colonialism, was much stricter in terms of the segregation of races, and because it was more directly accessible to the experience of the interviewees who had travelled before to South Africa.

As a visitor to the *musallah*, the Muslim prayer place in Sommerschield II, Senhor Justino became acquainted with some of the Muslim residents from the elite neighbourhood. These encounters with co-believers were, however, somewhat f leeting, restricted to being present in the same space, praying next to each other and exchanging greetings. The experience of praying together can nevertheless be understood as a liminal phase in which ordinary social hierarchies and differences are dissolved in a temporary *communitas* (Turner 1991 [1969]) of praying. The temporary anti-structure and the spiritual liminal experience of equality, though, is limited to these moments and do not change social structures and social boundaries.

At the time of the interview in 2012, Senhor Justino had never engaged with his Muslim neighbours from Sommerschield II, except from praying together at the *musallah*. African Muslim interviewees hence tend to report closure of Indian social networks; Indian parents are said to oppose to intercultural marriages of their children, business partners are usually kin and the spaces of public life like the Maputo Shopping Centre are seen as being 'dominated' by Indians, rather than allowing for co-presence (see chapter 7). So the social networks which are strengthened through praying together usually follow ethnic or cultural boundaries: Indian and African Muslims socialise separately from each other, also at the mosque. The entanglements which emerge are hence tenuous, ephemeral, and more related to a sense of shared belonging to a religious *ummah* than to actual moments of togetherness.

The reason for keeping up the rather rigid boundaries between 'Indian' and 'African' Muslims, as seen locally, is not so much an 'origin', 'race' or denominational orientation, but rather extreme economic inequality. When Fernando asked about the differences between African and Indian Muslims, Senhor Justino responded: "Not religious ones. They may exist for economic and financial reasons" (Senhor Justino, interview conducted by Fernando Tivane, March 2013). Using an array of strategies, many Indian merchants managed to expand their businesses even during the socialist regime and entered the economic elite of the country (Bastos 2005, Carvalho 2008). Many Indian Muslim residents own printing, clothing or household supply businesses and are considerably well off. Muslims in Polana Caniço, in contrast, are mostly engaged in a daily struggle for survival, like the majority of Maputo's population. They either have low paid jobs or try to make a living in the informal sector.

The workplace thus turns out to be a key space for everyday interactions and encounters between these diverse lifeworlds where specific forms of entanglements and mutual attitudes become shaped. The Indian business owners depend on cheap labour and offer jobs which the African Muslims need in order to sustain their families. As already mentioned, Senhor Justino, our key informant from Polana Caniço on these matters, was employed as stockman in a trading company owned by a Muslim Indian family in the Baixa. During the interview in 2010, when asked about his employers, Senhor Justino inclined his head in the direction of the villas of the neighbouring Sommerschield II: his employer did not live in Sommerschield II, but Senhor Justino assigned them to the same social category, namely wealthy Indian merchants (*comerciantes*).

These decade-old asymmetric entanglements between African Muslims and Indian Muslims at the workplace, lasting from colonial to socialist to neoliberal times, produced mutual stereotypes and prejudices which again feed into religious entanglements. Indian Muslim interviewees refer to their African workers as lazy, uneducated, unreliable and lacking responsibility, which is why, as they explained us, they need close supervision at work. Indian Muslims, on the other hand, have the reputation of being abusive employers who pay higher salaries to their employed kin than to African workers. These stereotypes observed today are clearly still shaped by the colonial work relations between the *Asiáticos* who occupied an intermediary position in the colonial hierarchy and the Natives, the colonial subjects. "They are big religious people, but also big tribalists" (Senhor Justino, interview conducted by Fernando Tivane, March 2013). Inside the mosque, Indian Muslims pretend to be brothers and good people, Senhor Justino claimed, but in the economic realm they only look after themselves and their kin.

The economic inequality between African and Indian Muslims is also related to politics. Indian and other Muslim elites have increased in political and economic power and risen in civil society (Morier-Genoud 2007). Urban dwellers often talk about the political inf luence of wealthy Indians in terms of corruption; referring to stereotypes about *comerciantes* and the hidden and illegitimate inf luence which they allegedly have.

Our employers pay us af ter two months only and even call us monkeys. This is all against Islam. If they had come here because of Islam, they would have changed these behaviours. But they are showing of f the money they have. And they are showing this country how weak the governments are (Jorge, visitor to the Masjid Ioonus in Polana Caniço, July 2012).

Jorge questions the legitimacy of Indian Muslims' claims to religious leadership in the Mozambican *ummah*, maintaining that they came to Mozambique with economic intentions, and not to practise Islam. African Muslim milieus in Maputo, in contrast, have a marginal political position. The northern Mozambican Muslim leadership has not been able to play a significant role in the official Islamic public sphere (Bonate 2007a: 57). In Maputo, the African Muslim milieus constitute a regional and religious minority. They are seen as people stemming from the 'north' and are called *Xingondo* by many, which is a derogatory designation for an ethnic group from the north, the *Macondo*. Common stereotypes about people from the northern provinces or *Xingondos* entail lack of education and, more importantly, sympathy for the former rebel movement and contemporary opposition party, Renamo.

Islam presents hegemonic ideologies of equality, providing values and norms with regard to the equality of human beings before godly powers, which African Muslims take up in order to make normative claims about how they expect to be treated by Indian Muslims. Yet at the same time, they point out the large discrepancies between ideals and reality. Like the churchgoers from the township, the leaders of the mosques and the Muslim residents of poor neighbourhoods do not hesitate to point out the hypocrisy inherent in discourses of egalitarianism and discriminating practices by the more powerful members in their religious community. Interactions between co-believers as equals inside the religious space, enacted before, during and after the actual worshipping, lead to troubled entanglements, characterised by a hope for symmetry, and the disappointment caused by the reaffirmation of unequal structural positions.

### **Faith-based Charity and Paternalistic Bonds**

Muslims living in *bairros* and the New City (*Nova Cidade*), despite all their differences, share a religious rhythm of five daily prayers, which may bring them together in routinised, yet largely impersonal, encounters in mosques, especially in those in the inner city. Similarly, in Johannesburg, suburban life and township life come to intersect at Sunday services, in friendly interactions between acquaintances. Yet, besides bringing people together for praying and worshipping, religions also work on urban relations by creating specific religiously embedded forms of economic relations: more precisely, faith-based charity practices lead to specific forms of religious entanglements between less entitled and aff luent urban dwellers. Bringing together an apparently benevolent person who gives and an apparently thankful person who receives, this asymmetric and hierarchical form of entanglement creates paternalistic bonds across economic and other divides. These relations are accompanied by specific behaviours and expectations that express themselves in these encounters.

### Practices of Zakat in Maputo

There is a time of the year, Ramadan, inwhich they [the Indians] are generous. They distribute food to allthe mosques, groceriesforthe people to eat atsunset. I don't have the money for this, I have to ask for it, write letters to ask the *direções* [city centre mosque administrative bodies] to send groceries. So, they send them here for the community. At the end of the month, five days before Ramadan ends, theywillsend more groceries to distribute to the people as *zakat*. *Zakat* is a tribute which you have to pay every year (Senhor Jaul, local administrator of the Masjid Ioonusin Polana Caniço, July 2012).

During Ramadan in particular, small mosques like the Masjid Ioonus in Polana Caniço A and the mosque in the neighbouring *bairro* Costa do Sol receive food donations from Indian Muslims for the *If tar*, the festive, communal meal eaten together after the sunset. The *If tar* served in mosques in the city centre also attracts destitute Muslims to break the fasting there. In the early evenings during Ramadan, Muslims, whose ragged clothes indicate their economic hardship, walk through the gleaming hallways of the Maputo Shopping Centre on their way to the Muslim prayer place on one of the upper f loors where food will be served. The prayer place at the mall is funded by the mall owner, Bachir, a controversial and wealthy Muslim businessman. Muslim interviewees report that public figures like him sometimes pay for the *hajj,* the pilgrimage to Mecca and the fifth pillar of Islam, for Muslims who cannot afford it themselves. Offering *If tar* to those in need and paying for the *hajj* are considered morally highly valued acts of charity.

Islamic charity practices even shape the rhythms of the city and the spatial practices of destitute urbanites. After the most important prayer of the week, the prayer at lunch time on Fridays, Muslims are expected to give voluntary alms (*sadaqah*) to the poor. Hence, on Fridays, elderly people with no support system, orphans, the disabled and other destitute people go to the Baixa, the commercial area of the city centre. They seat themselves strategically in front of the mosques or next to Muslim-owned businesses, hoping to be the beneficiaries of this religiously grounded acts of charity. Senhora Maria, an old woman living in a reed building in Polana Caniço which a church built for her, walks the long distance to the Baixa on Fridays, hoping to get alms. Strategically capitalising on the Muslim obligation to help the needy, people depending on alms thus adapt their weekly routine to the Muslims' religious rhythm (see also Capurchande 2004).

Senhor Ismail, the Muslim Indian business man living in Sommerschield II, supports an Islamic NGO which funds the construction of housing for people in need in the northern provinces. Charitable funding, social projects and emergency relief programmes are a recent institutionalised form of *zakat* charity, which give the money transfers a national and very often a transnational dimension (Kochuyt 2009: 103). Charity is a central value in Islam (ibid: 99), as *zakat* is the third of the five pillars of Islam, which should be upheld in order to be a good Muslim. *Zakat* is a commandment to care for the poor (ibid: 99). As Senhor Jaul, the religious teacher from the mosque in the neighbouring *bairro,* Costa do Sol, explained, Muslims are obliged to pay a certain percentage not of their income but of their assets. The poor (*os que não tem condições*) are exempt from this obligation, however. Other religions like Christianity also endorse charitable practices in a similar way, but Islam, according to Kochuyt, is quite unique

in its codification and elaboration of charity as the mandatory *zakat*, which even has similarities to a tax system (ibid). Practices of *zakat* vary greatly across contexts and milieus, based on differing interpretations of the Qur'an, different patterns of local institutionalisation and differences in how Muslims live it.

Humility, modesty and charity (*caridade*) are Islamic values which Muslim interviewees repeatedly brought up in the conversations with us, ref lecting Islamic discourses about almsgiving for the fulfilment of religious norms and duties (Weiss 2007: 5).

Notto have a good carteaches you to be humble. If you have and the other doesn't have, it doesn't mean that you can stop being humble. I give you an example. We annually pay a tax of 2.5 percent. If all the Muslimsin the world followed this general rule, there would be no poverty, there would be no hunger. But we as human beings are subject to failure andweaknesses(Senhor Mattar, resident of Sommerschield II, August 2012).

Senhor Mattar mentions the ambivalence between Islamic values and lived reality. The obligation to pay *zakat* has transformative and redistributive potential, yet the implementation fails due to 'human nature'. *Zakat* is hence related to Islamic normative ideas about wealth and appropriate attitudes towards inequality. In the *Qur'an*, wealth is described as being given by God (not earned as in protestant ethics); as Allah is generous to the wealthy, they also have to be generous to others, Senhor Mattar explained. Ramadan is seen by many of the Muslim urbanites in Maputo as a time in which the wealthy, through the practice of fasting, experience hunger, which is an everyday, physical and emotional experience of a large part of Maputo's urban dwellers. By sharing this experience, fasting is supposed to make people humble and empathetic to the lot of poor fellow Muslims.

These same values also entail prescriptions on how Muslims should display their wealth: "People can enjoy these riches as long as they do not become a source of pride, envy or greed" (Kochuyt 2009: 100). In Maputo, young men, mostly the children of wealthy Indian merchants, own Japanese racing cars and hold car races and driving skill competitions at night on public roads. When asked about these *corridas*, many of the (African and Indian) Muslim urbanites we talked to expressed their disapproval of these practices. Senhor Jaul liked to watch these races himself but, he emphasised, only the legal ones taking place on the race track of the local automobile club, the ATC. Such races are not against Islamic doctrine, he explained, as long as they are not done on public roads. Not so much because it is against the law and dangerous, but because it is a form of conspicuous consumption.

Senhor Jaul: On publicroads, you are showing of f. Barbara: What do you mean by that? Senhor Jaul: To show 'I have'. For us, this is not a good quality. One should limit one's expenses; one should not exaggerate while others are suf fering. One hasto remember the have-nots. Because of that, fasting is good as it makes the haves learn how it is to suf fer, they feel it, they take on the place of the poor and he feels how it feels not to eat nor drink, and he startsto be more sensible (Senhor Jaul, local administrator of the Masjid Ioonusin Polana Caniço, July 2012).

It is common knowledge that the cars, the motor tuning, the paintwork and the tyres which get burnt at each race are extremely expensive. Racing them on public roads is hence criticised as a form of 'showing off'.

Fasting and practices of *zakat* are thought to contribute to the construction of the Muslim *ummah* despite the economic divides. The general rules of *zakat* prescribe that receivers may not be members of the nuclear family. This means that, similar to the incest taboo (Lévis-Strauss 1968), the mandatory *zakat* obliges Muslim believers to create social relations outside the nuclear family. It obliges the better-offs to create some sort of connections to those who are poor and marginalised, people whom they would not normally encounter in their own social networks. *Zakat* therefore gives Islam a transcending drive and is an important practice for the creation of a sense of belonging to the *ummah* despite the social, economic and cultural cleavages in other spheres of life. "Through the zakat the centrifugal forces of economic inequality are checked and balanced so that they do not lead to a disintegration of the Muslim community" (Kochuyt 2009: 105). It is also prescribed that *zakat* may only be given to Muslims; hence, *zakat* reinforces the boundary with non-believers. This type of entanglement, which pulls the diversity of Muslims together, is nevertheless based on an insider–outsider distinction, namely, between those who deserve to receive and those who don't.

*Zakat* can be conceptualised as a form of gift exchange which follows the rules of reciprocity (Mauss 1990 [1925]). Gift giving is not as altruistic as it might appear, as it is also guided by the self-interest of the parties involved (Rössler 2005: 194). Applying Mauss's approach, one may distinguish three phases, namely, giving the gift, accepting the gift and giving a counter-gift. Through acceptance of the gift the receiver agrees to enter a relationship with the donor in which he now owes a counter-gift. If the counter-gift is not of a similar quality and quantity to the initial gift, the balance of power shifts between giver and receiver, resulting in an asymmetrical relationship. In charity like *zakat*, the counter-gift is hidden, as charity has more the character of a *Geschenk* than a *Gabe*. What the receiver is expected to return is not of material quality, but a social attitude towards the relationship with the giver: an attitude of gratitude and acceptance of the own inferior position in the relationship. In practice, the receivers do not take on such a passive role as the normative expectations and the description of the ideal type imply; those in need of resources make active use of this norm of charity for their own purposes, like the alms seeker who go to the Baixa each Friday. In order to find funds for the construction of the Masjid Ioonus, Senhor Jaul, its leader at the time, wrote a project proposal and sent it to wealthy individuals who were known for their charitable practices and to Islamic organisations. Through these networks their proposal came to the attention of a Muslim Indian widow from South Africa, who was looking for a project to which to donate part of the inheritance of her deceased husband.

Happily, there was a woman of your race [addressing the white anthropologist], she is Muslim. She lost her husband and inherited that money. But she wanted to do somethingwith the money thatthe deceasedwould also benefitfrom …There isthis promise from the Creator oftheWorld inwhich he saysthatfor everyonewho does good actions on earth, there will be a place called paradise. So that's how our mosque was created, for the benefits of the Muslims of the neighbourhood and, notwithstanding, for the

benefit of the deceased husband. This is the spiritual field (Senhor Jaul, local administrator of the Masjid Ioonusin Polana Caniço, July 2012).

The deceased donor of the money for the construction of Masjid Ioonus receives a divine recompense: he becomes considered *halaal*, purified of sins, and guaranteed a space in paradise (Kochuyt 2009: 108). So, the donation by the widow was beneficial for everyone: the Zambezian Muslims in Polana Caniço received a mosque and the good deed would, on a spiritual level, ensure her deceased husband a good afterlife in heaven.

*Zakat*, as Kochuyt pointedly argues, is thought to transform potential social revolutionaries or criminals, which the poor could turn into, into obedient and loyal members of the *ummah* who accept their inferior position because they receive (ibid: 106-7). Faith-based charity and the entanglements it creates does not strive for real redistribution leading to the transformation of the social hierarchy. Rather, in the entanglements emerging from faith-based charity, the status quo is maintained while creating feelings of solidarity and gratitude between rich and poor (ibid: 106), which makes it easier for both sides to cope with these ambiguous entanglements embedded in structural inequality.

### Food Donation and Prayers in Johannesburg

The ethnography of the LRC Church in Johannesburg shows the same, important characteristic of paternalistic entanglements based on exchange. While faith-based charity pulls people closer together, it also draws a boundary between them and the outsiders who do not deserve to take part. The LRC Church runs many different programmes for its believers in addition to the weekly Sunday service. One of them is a programme called *Malukhanye*, instituted to make an "impact on poor communities" (Love Reaching Communities Church 2015). This 'ministry to the poor', as they call it, consisted of a weekly event aimed at the 'needy' in the church (Terry, elder of LRC Church, May 2012). In 2012, it took place on Tuesday mornings in the church building. The Malukhanye ministry started with bible study and prayer, and then church volunteers distributed plastic bags filled with bread and vegetables to the attendees. Church volunteers collected these food donations from church members and from nearby supermarkets. Such food programmes and other charitable activities are widespread among Christian churches in Africa and, specifically, among progressive Pentecostal congregations (Freeman 2012, Heuser 2013, Swart 2012).

The Malukhanye events I attended between April and June 2012 were led by two middle-aged white church volunteers. They led the morning's proceeding, standing, like the elders at the Sunday services, in front of the audience who consisted of about 50 to 80 people. No technology, not even a microphone, was used, which was in stark contrast to the Sunday services. With enthusiastic friendliness the two greeted everyone, aiming to foster a jolly, energised atmosphere. Their voices tended to be highpitched and shrill, and sometimes they cracked. At the same time, they were observing everything restlessly with suspicious eyes, as if they were worried that something could go wrong at any minute.

A couple of grey-haired, elderly white women, one of whom had initiated the food programme years previously, were usually seated in the first row. On my first visit, one of these older women called me to sit in front with them and not with the recipients of

the food. She did not explain her request, but she assumed it was obvious to me, too, that the beneficiaries and the representatives of the donors (which she had categorised me as based on my white skin) had to sit separately. What resulted was a separation of black and white churchgoers in the hall, not actually based on race but based on the different positions they had within this setting. On the one hand there were the white, aff luent givers and on the other hand the black, poor receivers, connected by the paternalistic bonds being constructed and performed at this Malukhanye event, which were in many ways similar to the relationships witnessed in other settings in Linbro Park. These older ladies did not take an active part, unlike the many younger volunteers hovering around. Yet they were visible to everyone in the first row and acted as role models. They enthusiastically proclaimed 'amen' during the prayers and diligently consulted their bibles, thus showing the township dwellers how they were supposed to act. These older women also policed the audience with scrutinising looks: sometimes they threw chastising gazes at the mothers of children who were too noisy. If the gazes were not enough to keep order in place, they sometimes indicated with a slight nod of their head to one of the volunteers that they should approach the mother in person and tell her to keep her children in check. The interactions within this setting were extremely troublesome for everyone, shaped by mutual distrust and tension.

The receivers of the food donations during the Malukhanye events I attended consisted of two groups: there were the regulars, mainly pensioners and a few young people who attended the LRC Church on a regular basis, not just the Malukhanye event, and who were familiar with the setting and its rules and the volunteers. Many of them usually sat close to the front and had a bible with them, which they paid visible attention to. They moved around comfortably, knowing the premises and the volunteers. When I arrived for my second visit, many of them greeted me in a friendly way. The second group, though, seemed to be outsiders in this space. This group consisted mainly of young women, somewhere between 18 and 30 years old, many of them accompanied by children too young to be at school. Many of them were wearing very old, ragged, long skirts and worn-out shoes. Some were very thin and had unhealthy glassy eyes, most did not have extensions in their hair. With their ragged clothes they appeared to belong to the poorest of the township dwellers, a style of dressing very different from the stylish, yet inexpensive clothes many of my female friends from Alexandra liked to wear. They usually sat in the back rows where they chatted to each other and looked after their children instead of listening; anyway, because the sound system was not on, one could hardly hear the volunteer's preaching among the children's voices, and many of these women had not mastered English well enough to understand the sermon. They obviously did not feel felt at ease, and they were constantly looking around, as if they were not sure about what was going on and what was expected of them.

Apart from two volunteers organising the event and the elderly women who were from white, middle-class milieus, the majority of volunteers at this event themselves came from Alexandra. These volunteers organised the tea break and the food distribution in a laid-back manner and did not show the same overemphasised friendliness as the volunteers in charge did in their interactions with the recipients. Nevertheless, these volunteers were constantly and critically observing the food recipients and explained the rules to them, if necessary. This emphasis on order and rules was especially remarkable, as in the same church space on Sundays, church leaders and volunteers distinctly aimed to create an atmosphere of informality, belonging and lack of

hierarchy in the way they greeted churchgoers, and also in the sermons. The Malukhanye events, by contrast, were marked by mutual distrust and discomfort.

This became especially apparent after the coffee break at one Tuesday event. The volunteers took a moment to explain the rules of the event, specifically for the people who they felt were not behaving appropriately. People who disturbed the sermon by arriving late would not be given food, they explained. Women with little children should sit close to the exit and they should leave the church quietly if the children were crying. These rules were more important than the previous bible reading for the functioning of Malukhanye: while the volunteers in charge had read the bible in English, the volunteers from Alexandra talked to the food recipient in Sotho or Zulu, translating what the volunteers in charge had told them. The volunteers from Alexandra hence took on the role of 'go-between' (Goffman 1959: 151), explaining what role the food receivers were expected to play in this performance. The instructions were mainly aimed at the young mothers, and not the regular visitors who seemed to have learnt the rules of the game and who were usually there without children.

Besides noise and being on time, a further worry of the Malukhanye organisers was about the respectability and legitimacy of the food recipients, whether they were actually 'poor'. The category 'poor' was not institutionally defined and access to the event was not controlled: anybody could participate, as long as they abided by the rules of the event and as long as the volunteers could agree that they belonged to the 'poor'. The participants had to disclose their identity by writing down their name and phone number in a register. According to Scott, such collection of information about the subjugated is an important instrument of control for institutions that exercise power (Scott 1999). The volunteers sometimes doubted the authenticity of the poverty, and suspected that some of the people present had a 'dark secret' (Goffman 1959: 141) incompatible with their performance, namely, that they were not as poor as they appeared to be. On my third visit, volunteers told me that they had heard that some of the young women had actually sold the food donations once back in the township instead of using it for their families. For the volunteers, this was an act of treachery and it confirmed their suspicion that some of the donation receivers were not so poor after all. By selling the food the young women exerted agency and decided for themselves what they wanted. With the money made from selling the food, they could buy different groceries, clothes or school books. For the volunteers, however, selling the food donation was incompatible with their own expectations of how the good 'poor' should act; they should be passive, thankful and humble about what they received, and not actively take charge of what they did with the charity they benefitted from.

A further worry of the volunteers was that the attendees only came to get the food donations and were not interested in the religious aspects. Hence, the attendees were obliged to be present at nine o'clock for the bible reading and prayer, and they were not given food if they arrived late and missed the prayers. This highlights a problematic contradiction and ambivalence characterising certain faith-based charity projects as they link power over resources to evangelisation. On the one hand, the church claims that they feed the 'poor' but on the other hand the volunteers demand that the receivers participate in the bible teaching and even expect the receivers to be intrinsically committed to the praying.

The encounters between the destitute township dwellers and the church volunteers were hence characterised by mutual distrust, suppressed and sometimes open con-

f lict caused by misunderstandings, disappointed expectations and diverging interests. The interactions and relations enacted at the Malukhanye event stood in sharp contrast to the Sunday church services where the working consensus was that the co-present churchgoers stood in an equal, friendly and informal relation to each other. Malukhanye was a performance of charitable gift giving enacted by the benevolent middle-class white donors, facilitated by diligent black volunteers, and aimed at poor Alexandra residents who were expected to thankfully and passively receive both material and spiritual support. The volunteers with white, middle-class backgrounds still attempted to show equality through their dramatised friendliness in the interactions, which may be typical for such paternalistic relations. Yet they also expressed a clear expectation of obedience and subjugation. Many of the attendees were thus extremely careful and even fearful in their interactions with the volunteers. When I tried to chat to some of the mothers who did not know me during the coffee breaks, they looked away and responded shyly and in low, scared voices, assuming that I was a volunteer. On one such Tuesday, I left the event before it ended, and I saw a group of these mothers sitting outside the church on the grass. They were chatting freely to each other, laughing and watching their children. The interactions inside the church hence had a frontstage character for these young women where they aimed to foster the impression of being submissive, poor young mothers. At this moment, the churchyard was a backstage where they could shed their roles as beneficiaries. The tension that seemed to stiffen their bodies inside the church seemed to be gone.

In contrast to the paternalistic relationships typical of the labour entanglements around domestic work in Linbro Park, which were characterised by intimacy, the entanglements around charity at the Malukhanye event were somewhat impersonal and, as some of the food recipients only came irregularly, ephemeral. In the relationship between a domestic worker and an employer, the routine of proximity and feelings of affection can ease some of the existing tension because of the hierarchy structuring their relationship. In the more anonymous religious setting of Malukhanye, the volunteers as powerful actors resorted to rules and control as a reaction to insecurity and their need to draw boundaries. This shows that paternalism, hence relations of giving and receiving, is a repeating pattern in urban entanglements, even though it can take diverse forms.

### **Conclusion**

Some researchers have observed a religious turn in urban studies (Wilhelm-Solomon et al. 2017) and the intersection between religion and urban spaces has received more attention recently (Becker 2011, Cimino, Mian and Huang 2012, Desplat 2012, Garnett and Harris 2013, Pinxten and Dikomitis 2009). The religious dimensions of urban life are nevertheless still marginal in the mainstream of urban studies; urban religion being largely absent from recent paradigmatic handbooks (Bridge and Watson 2013b, Parnell and Oldfield 2014). In times of terrorism threats and 'religious' wars, religion becomes portrayed in public debates, especially in Europe, largely as a problem and as a source of conf lict. Religious leaders, as exemplified in this chapter, tend to paint a rosy picture of religion as a source of peace, tolerance and equality. In this context of increasingly polarised debates about religion, the importance of academic, empirical

research which studies religious practices from a non-normative stance, unravelling what believers make out of religion through their agency, becomes even more apparent.

Religious practices and religious spaces impact on the way urban dwellers use and experience the city and how they become entangled with others. In Maputo and Johannesburg, religion provides discourses of equality, which urban dwellers utilise to make normative claims about religious socialities, even though the resulting everyday encounters are still shaped by the memory of colonial and apartheid racial separation, which become reactivated especially in moments of competition and conf lict. Religious spaces are increasingly important spaces of encounter in entangled cities, yet the unequal power distribution within these spaces complicates and limits their potential for creating new forms of conviviality and can lead to open or hidden conf lict. Within religious spaces, this is partly dealt with through religiously embedded charity. Alms giving and alms receiving create social ties, paternalistic bonds, across class and ethnic boundaries. Practices of *zakat* among Muslims in Maputo and schemes like food distribution in Johannesburg integrate socially and spatially diverging milieus not despite but because of the inequality between them.

The LRC Church and the Muslim spaces in Maputo are deeply urban spaces. They bring people together in repetitive, rhythmic ways, yet many remain, to a large degree, strangers to each other. In both the LRC Church and the *musallah* in Polana Caniço, there is coexistence and intersection of diverse socialities. These religious spaces become crystallisation points for dense social networks consisting of close relatives, neighbours, friends and co-workers, as well as the acquaintances with whom one socialises on other occasions and whom one may have come to know there. But parallel to this, there are also many strangers in these spaces with whom one has only one thing in common, namely, that one shares a faith and happens to pray or worship together in the same religious space from time to time. The stories by the interviewees about encounters in religious spaces then usually refer to interactions with acquaintances and the experiences, hopes and disappointments they bring about, be it the disappointment of not being offered a ride in their car, the sense of a large social divide expressed by specific looks from a distance, or the feelings of discomfort experienced in a space shared with those whom they see as culturally distinct. Nevertheless, the instances of praying together are meaningful, powerful symbols for these urban dwellers, encounters full of promises for a more equal urban world. Against all odds, these moments of togetherness create feelings of belonging, even though they become easily ruptured in situations of conf lict and disappointment where urban dwellers refer to the old models for imagining society – the ghosts of racial apartheid and colonialism – to explain what is going on.

The paternalistic entanglements across the economic divides created through faith-based charity are of a more mediated, more distant quality than the paternalistic relations between domestic workers and their employers, which are deeply personalised and intimate. In these religious spaces the paternalism tends to be impersonal, more like between the state and its citizens, between a programme and its recipients, and less like that within a (post)colonial family. Givers and receivers may not know each other well, yet the social effects of paternalism are similar to those in domestic work: it creates social ties across economic divides not despite but because of them, turning the social positions into complementary, interdependent interests. Paternalistic entanglements create positive feelings like thankfulness and the sense of doing

something good, yet they are also imbued with potentially disappointed expectations, mistrust and hidden conf lict.

It is important to emphasise that philanthropic practices in South Africa and elsewhere are not at all restricted to the wealthy and powerful who give to the 'poor'; giving practices are commonplace in poor communities like Alexandra as well (Everatt et al. 2005). There is, however, a qualitative difference in between charity relations entangling people across economic divides and relations constructed through self-help initiatives in which the givers conceive of themselves as belonging to the same lifeworld with similar living conditions to the receiver (which are by definition not entanglements). Charity relations crossing economic divides often involve the construction of the poor as 'other' (Spivak 1985), which is less prominent in the self-help initiatives as encountered, for example, in Alexandra township. This is also why entanglement is not a synonym for relationship*,* but a specific type, namely, a relationship in which the actors see themselves as different. *Cities of entanglements* claims that these specific sets of relations crossing social and spatial boundaries are key to understanding urbanity in cities characterised by inequality and, hence, key for urban theory.

©Fernando Tivane, Maputo, 2012

## **Spaces of Freedom?**  Shopping Together, Living Apart

In 2007, two shopping malls opened their doors in Maputo and Johannesburg, the Greenstone Shopping Centre in north-east Johannesburg and the Maputo Shopping Centre in downtown Maputo.<sup>1</sup> In Johannesburg, this new mall was one amongst many others in the differentiated shopping centre landscape. Yet in Maputo, the gleaming golden mall constituted a new and unique type of urban space, although it has received competition from new malls since then. In many cities around the globe, shopping malls have become important spaces where urban dwellers of all walks of life – not only the middle classes and the elite – engage in manifold practices, ranging from grocery shopping to socialising with neighbours, playing arcade games and people watching. Every society and every mode of production produces its own space (Lefebvre 1996 [1974]). In this sense, the shopping mall is one of the quintessential spaces of globalising, neoliberal economies. Shopping malls emerged in the second half of the 20th century, first in the USA, then in Europe and South Africa, and since the 1990s, increasingly in many other cities of the Global South. One of the reasons for their mushrooming is that shopping centres are considered relatively secure investments (Goss 1993: 21). Not least because of the volatility of the financial markets, property has become a favourite asset. Shopping centres economies of scale and the all-encompassing planning make them attractive to real estate investors in the Global South and elsewhere, especially where states struggles to provide urban infrastructure and public spaces which are considered attractive by the growing aspirational classes. As we have seen in the previous chapters, such capitalist interests in property have also transformed Maputo and Johannesburg neighbourhoods, being a key force in the politics of proximity (see chapter 5) and the politics of loss (see chapter 4).

In large African cities, especially in Southern Africa, shopping malls are bourgeoning. In South Africa, where most of the existing research on African malls has been conducted, the construction of shopping malls has profoundly inf luenced the urban consumption landscape and public life since as early as the 1960s. In other cities like Maputo, shopping malls are a new phenomenon and related to the deepening of economic liberalisation. In the Southern African region and beyond, retail investors from

<sup>1</sup>  Sections of this chapter have been published in the journal *Anthropological Forum* in an article written together with Aceska (Aceska and Heer 2019, © Taylor & Francis). A previous version of parts of the chapter has been published in an edited book (Heer 2017). I thank the respective publishers, editors and co-author for the permission for reprint.

South Africa play a major role in the shaping of mall landscapes (Miller 2008, Miller et al. 2008). It is likely that in the coming decades more malls along the lines of the South African model will be built across Africa.2

The chapter unravels the way in which shopping malls function as important spaces of public life in African cities, and how they stand in specific relations to broader society and to other urban spaces. Rather than being perceived as inferior spaces in comparison to so-called *proper* public spaces or as fortified enclaves which exclude the African urbanities, malls should be viewed as relational and connected spaces serving as spaces of encounters in African cities where entanglements, hence forms of sociability crossing social boundaries, emerge. Using Foucault's notion of heterotopia, the ethnography will illuminate that malls are not uniform, abstract spaces, even if mall builders design them as such, but rather gain multiple layers of meanings, emerging in diverse forms of practical appropriation, in competition with diverging local conceptions of what the malls are and in manifold entanglements with other urban spaces.

### **Beyond the Enclave: Malls as Entangled Spaces of Heterotopia**

Shopping malls are privately owned spaces of public life. Containing multiple shops and diverse leisure establishments such as restaurants, movie theatres, and sometimes even game arcades, they are planned, developed, managed and owned by a nonstate entity (Falk 1998: 15). Historically, they can be seen as descendants of the European 'Passage', an architectural ancestor of the mall invented in European cities at the turn to the 19th century (Watson 2009: 55). Shopping malls can also be seen as a contemporary form of urban market (Wehrheim 2007). They differ from other markets in African cities by being strongly regulated spaces with differentiated mechanisms of security and control, with rules set up and enforced by the private owners and their security personnel. Often, they have very clear spatial boundaries setting them apart from the surrounding state-managed urban spaces.

Although malls are private spaces with regard to their ownership structure, from the point of view of users, they are spaces of urban public life. Lof land defines public life as social life characterised by the co-presence of "personally unknown or only categorically known" urban dwellers (Lof land 1989: 453). As Sennett (1983 [1974]) points out, public life is about visibility and accessibility. Accessibility to shopping malls is, however, subject to the many limitations and rules imposed and sanctioned by the mall management. Helten (2007) even suggests approaching shopping malls as total institutions (Goffman 1961). Many malls exercise a high degree of control and regulation of people's behaviour. The architectural designs of many shopping malls even resemble the panopticon and prisons (Helten 2007: 245). Yet power and control tend to be "exercised through a seductive spatial arrangement, where the experience of being in the space is itself the expression of power" (Allen 2006: 454). The landscape artists and interior designers aim to create a sense of place which attracts urban dwellers to

<sup>2</sup> The shopping landscape in Maputo and Johannesburg is changing rapidly. The following ethnography refers to 2010–2012, and the shopping circuits thus described may have changed since then (in Johannesburg with the opening of the Alexandra Mall and maybe the Mall of Africa, and in Maputo with the opening of Baia Mall and others).

the mall, often in unconscious ways (Goss 1993: 22). The shopping mall planners therefore distinctly aim at shaping lived space (Lefebvre 1996 [1974]), meaning the way urban dwellers feel and act in space and what they associate with it.

One may distinguish two opposing academic discourses on shopping malls: a negative discourse which sees malls as enclaves and a more positive discourse which acknowledges malls as new spaces of urban sociability. The more positive discourse emphasises the agency of urban dwellers and the mall's potential for providing space for new forms of publics and public life (Heer 2017, Houssay-Holzschuch and Teppo 2009, Leeuwen 2011, Murray 1997, Nuttall 2004, Perez, Salcedo and Caceres 2012, Underhill 2004, Van Eeden 2006, Vries 2008), whilst also acknowledging the multiple negative aspects like forms of exclusion. This ethnography seeks to contribute to the literature focusing on malls' capacities to both open up opportunities and reproduce systems of exclusion. In the negative discourse, shopping malls have become symbols or epitomes for the loss of public space within the context of the increasing privatisation and commercialisation of cities (e.g. Sorkin 1992). Many urban scholars claim that malls are a phenomenon of the globalised economy, diminishing locality and human agency (Olivier 2008, Salcedo 2003). They tend to be seen as spaces of hyper consumption and social control (Dirsuweit 2009, Fu and Murray 2014, Marks and Bezolli 2001, Mbembe 2004, Miles 2010, Miller et al. 2008, Salcedo 2003). Malls have been said to limit political and democratic practices like demonstrations (Németh 2009). By excluding the poor and other marginalised groups they are seen as constituting fortress-like, exclusionary and elitist spaces sanitised of urban poverty and decay (Dirsuweit 2009, Dirsuweit and Schattauer 2004). Because of the mobility of South African capital in Africa, the spread of malls regionalises or globalises class and race divisions in African cities (Miller et al. 2008).

Critical voices claim that the emergence of malls has consequences for conviviality in the city as such. In generally reinforcing tendencies toward urban sprawl they contradict the often stated goals of urban integration (Beavon 1998, Turok 2001). Contributing to the fragmentation of the urban landscape, they are drivers of postmodern forms of segregation (Caldeira 2000, Davis 2006 [1990], Murray 2008, 2011, 2013, Paasche and Sidaway 2010). Caldeira most prominently proposed the notion of a 'fortified enclave', under which she subsumes gated communities, offices complexes and shopping malls. These enclaves are typical of cities like São Paulo, Los Angeles and Johannesburg, and "negate the main characteristics of the modern democratic ideal of urban public space" (Caldeira 2000: 304). This view has come to dominate academic discourses on malls in African cities. Myers and Murray (2006) argue accordingly:

With the increasing privatisation of basic social services and the overall neglect of physical infrastructure, the privileged middle class with global connections have increasingly cloistered themselvesinto the fortified enclaves of Western modernity that have materialized in cities all over Africa. Asthe propertied,the privileged, and the powerful have retreated behind the protective shield of gated residentialcommunities, enclosed shopping malls, and other barricaded sites of luxury, the urban poor are of ten lef t to fend for themselves in urban environments that have deteriorated almost beyond repair (Myers and Murray 2006: 5).

According to the *Oxford Dictionary*, enclave in the strict sense of the term means "a small territory surrounded by a larger territory whose inhabitants are of a different culture or nationality" (Waite 2013: 294). In urban studies, most prominently, Marcuse defined enclave as a "area of spatial concentration in which members of a particular population group, self‐defined by ethnicity or religion or otherwise, congregate as a means of protecting and enhancing their economic, social, political, and/or cultural development" (Marcuse 2005: 18). Understanding shopping malls as enclaves as Caldeira (2000) and Myers and Murray (2006) do has, however, several shortcomings.

The first fallacy of approaching malls as enclaves is that it implies that the users of the mall constitute a homogeneous group, clearly delineated from people outside the mall, who, by implication, do not frequent the mall (Aceska and Heer 2019: 49). As this chapter will show, well-off urban milieus may dominate malls, yet there is also ample space for their appropriation by less aff luent milieus. In addition, in both Maputo and Johannesburg, they are spaces of religious, ethnic and even economic heterogeneity rather than marked by homogeneity. The metaphor of enclave thus reproduces the *mosaic thinking* of the Chicago School, which understood the city as comprising patches of social groups, neatly separated and distributed across urban spaces.

The second fallacy is that the metaphor of the enclave does not distinguish between the users of malls and the builders of malls. It assumes that the urban elites desire to retreat from the city, and hence mall builders respond to this need. In reality, this is a bit more complex. Although the property developers, mall management and shop owners clearly depend on the users' demand, it is a fairly oversimplified understanding of consumption to assume that the interests and ways of seeing malls of mall builders and mallgoers simply coincide. Often there are multiple and even contradictory representations of shopping malls purposed by diverse actors, as this chapter will show. In addition, mall builders and mall users have a distinctly different relationship to these spaces. For mall builders, they are abstract spaces (Lefebvre 1996 [1974]: 49). Abstract space serves the purpose of creating profit, it is based on exchange relations and it is quantified and assessed based on its exchange value. Its use value is only relevant as long as it can be subjugated to quantifying measures such as price, size and the costs determining its exchange value. Based on such abstract calculations, property owners and property developers assess, plan, sell and manage malls, they are *abstract space-makers* (see chapter 5, Aalbers 2006, Thompson 2017). As the discourses by mall managers and mall architects show in this chapter, their thinking and acting is based on the idea that by shaping space, they can determine who uses malls and how they use them. This biased and old conception of space as a container (Bachmann-Medick 2006: 289) assumes that space determines what takes place within it (Aceska and Heer 2019: 49).

The production of space (Lefebvre 1996 [1974]), however, is an ongoing process and is not finished when the cranes and diggers have left. Despite the tight surveillance and restrictions on accessibility as constraining structural conditions, urban dwellers can to a certain degree *appropriate* these abstract spaces (Carrier 1996) and turn them into spaces of sociality, self-stylisation (Nuttall 2004) and leisure. By acting in and on space, urban dwellers imbue malls with meanings, memories and dreams; as *social space-makers,* they constitute malls as lived space (Lefebvre 1996 [1974]). Consumption practices may be silent and barely visible, but they may even entail hidden forms of resistance (Certeau 1984). These two rationalities – the builders and users, *abstract* and *social space-makers* – need to be disentangled in order to fully understand the social functioning of malls.

The third fallacy of the metaphor of enclave is that it overemphasises the spatial and social rupture between malls and the surrounding city, whilst removing attention from the many connections between them (Aceska and Heer 2019: 49). Overstating separation and segregation, the metaphor renders invisible the many entanglements and relations to the 'outside' which are as constitutive of the malls as the ruptures. Understanding malls as enclaves is a form of what Amin calls 'telescopic urbanism', namely, a sense that the city is "a collection of settlements with varied geographies of affiliation rather than as the sum of its parts" and not a "field of shared life and common rights" (Amin 2013: 477). But actually, only through the entanglements with other spaces do malls become constituted.

A fourth fallacy of the rather negative appraisal of malls is that it is caught up in a modernist public–private dichotomy, lamenting the economisation of what is nostalgically imagined as a civic space open to everyone (the former public space). Yet feminist research and others have amply shown that each *agora* (Arendt 1959 [1958]) and public sphere (Habermas 2002) is characterised by processes of exclusion, and that the idea of unitary modern public space is a naïve ideal, especially for societies characterised by diversity and inequality (Benhabib 1995 [1992], Fraser 2001 [1997]). Again, the public–private dichotomy overstates the importance of closure as a defining feature of malls, whilst neglecting other aspects.

As an alternative to conceptualising malls as privatised public spaces and enclaves, this chapter suggests approaching shopping malls as *entangled spaces of heterotopia*. Michel Foucault coined the notion of heterotopia in a lecture in 1967 (for more details see Foucault 2009: 13-14). By the term, he meant various spaces and institutions that interrupt the normality and continuity of everyday life and ordinary spaces and are hence *other spaces* (Dehaene and Cauter 2009: 4). While utopias represent society in a perfected, utopian form, yet are unreal spaces, heterotopias are really existing spaces:

There are also, probably in every culture, in every civilization, real places – places that do exist and that are formed in the very founding ofsociety – which are something like counter-sites, a kind of ef fectively enacted utopia in which the real sites, all the other real sites that can be found within the culture, are simultaneously represented, contested, and inverted (Foucault1986 [1967]: 24).

What makes shopping malls heterotopic spaces is that, amongst other things, they have the capacity of "juxtaposing in a single real place several spaces, several sites that are in themselves incompatible" (Foucault 1986 [1967]: 25). In a shopping mall, there may be a gym next to a fast-food restaurant, or a marketing discourse alluding to nature, while for the construction of the mall, nature was actually destroyed. In a movie theatre at the mall there is the projection of a three-dimensional space onto a two-dimensional screen, hence two spaces which are actually incompatible. Heterotopias are spaces entailing multiple logics, and they become constituted by the play between binaries such as public–private, normal–extraordinary, imagined–real, homogeneous–heterogeneous, and are thus fundamentally ambiguous and contradictory.

A further key characteristic is that heterotopias "presuppose a system of opening and closing that both isolates them and makes them penetrable" (ibid: 26). Malls, prisons or hamams are not freely accessible like public spaces, but there are certain systems in place that determine entry, like purification rituals, a juridical sentence, or clothing standards. In addition, heterotopias are often linked to '*slices in time*', they thus constitute a break with the time of ordinary life; museums enable us to travel in time, cemeteries are places where the dead are put to rest for the rest of eternity, and, as we will see, shopping malls enable temporary refuge in urban utopias, a rejuvenating pause from bustling city life (ibid).

The most important aspect of heterotopias for the analysis of malls, though, is their heterotopic relation to other urban spaces: heterotopias "have a function in relation to all the space that remains" (ibid: 27). They are "spaces that have the curious property of being in relation with all the other sites, but in such a way as to suspect, neutralize, or invent the set of relations that they happen to designate, mirror, or ref lect" (ibid: 24). Foucault distinguishes six types of heterotopia: heterotopias of crisis, of deviance, of illusion, of compensation, of festivity and of permanence. As this comparative ethnography reveals, the Maputo Shopping Centre can be seen as a *heterotopia of illusion*, which exposes "every real space … as still more illusory" (ibid: 27). The mall makes visible what is usually not acknowledged, for example that which is often seen as an urban reality, namely, that urban elites can live enclaved lives disconnected from others in the city, which becomes exposed as an illusion at the mall. Greenstone Mall, in contrast, emerges as a *heterotopia of compensation*, whose "role is to create a space that is other, another real space, as perfect, as meticulous, as well arranged as ours is messy, ill constructed, and jumbled" (ibid). The notion of heterotopia hence proves useful for comparative ethnography.

Heterotopias cannot be understood only in terms of themselves; their meanings derive from the entanglements, the relationship to other urban spaces and to the larger urban society. The concept of heterotopia brings therefore into focus for the urban dweller the relationship in which the mall stands with the rest of the urban world and the place it takes within their urban lives. This allows the analysis to move beyond the problematic idea of malls as enclaves, separated and disconnected, deterministic and absolute spaces. Malls are rather relational spaces, their social roles and meanings only emerging in relationship to other urban spaces. This also means that malls always need to be analysed in relation to the specificities of their urban contexts. Although malls may be financed by foreign money and built according to transnational architecture, they are nevertheless lived and imagined locally, and hence only come into being in relation to the local setting. This invites comparative research on malls across diverse settings, in ways that do not presuppose what malls ought to be and how they ought to be used, but in ways which are empirically open to all the variety of uses and meanings malls take on in diverse cities and for diverse groups.

This analysis of shopping malls as entangled spaces of heterotopia, as I develop it here, focuses on four analytical aspects, namely, (1) spatial practices by users, (2) representations of the malls, (3) encounters in malls, and (4) the relationship between the mall and other urban settings. Shopping has spatial dimensions (Gregson, Crewe and Brooks 2002). Shopping-related movements between home, work and consumption spaces form part of the spatial practices (1) through which urban dwellers link different urban spaces, embedded within the rhythms of everyday life, and therefore contribute to the production of the space of the city (Lefebvre 1996 [1974]). Urban dwellers weave together particular shopping spaces which results in shopping circuits, recognised in

their totality only by their users (Magnani 1996: 23). The Maputo Shopping Centre and the Greenstone Shopping Centre are just one space among many that urban dwellers weave into their individual shopping circuits, which differ from others depending on milieu background, the transport possibilities urban dwellers have, where they live, how much they earn; however, gender, lifestyle preferences, and whether they have to look after dependants or not may play a role. Any analysis of malls needs to ask which role this specific mall plays in urban dwellers' shopping circuits.

Yet malls also need to be analysed with regard to the multiple, contradictory and competing representations and discourses (2) upheld by diverse actors about the mall. This second aspect grasps that malls are not only material but also conceived spaces, and that different actors imagine and represent the malls differently. The relation between these multiple representations is shaped by power; there is often a competition over how the mall should be seen, and the importance of these representations not least derives from the fact that they can inf luence social practices, for example whether or not people shop at a mall.

The third analytical aspect in the empirical analysis of entangled spaces of heterotopia refers to the sociabilities and urbanisms through which encounters at malls may emerge (3). These encounters bring about co-presence and the mutual visibility of urban dwellers who live apart, and whose lifeworlds become entangled in new ways at malls. These encounters – often ephemeral, sometimes more intense – across economic and ethnic boundaries cause urban dwellers to imagine other realities and question their place in the broader society. Urban dwellers link, juxtapose and compare these mall encounters, based on habitualised experiences and urban knowledge, with encounters in other settings. The meaning of entanglements created at malls can only be understood in relation to other entanglements. This leads to the fourth analytical aspect: malls need to be analysed with regard to their heterotopic relationships to other urban spaces or urban realms (4), like urban religion, domestic work or street life. Malls may turn out to be heterotopias of compensation as in the Johannesburg case, compensating for inequalities in encounters outside of the mall, or heterotopias of illusion in the Maputo case, as it exposes otherwise hidden aspects of urban reality. This unravels the way malls work on people's imagination of their social place in the city as a material and social setting.

### **Maputo Shopping Centre – The Mirror of Mozambique**

The Maputo Shopping Centre (locally called Maputo Shopping) is located in the city centre of Maputo, more precisely in the Baixa section of the inner city, and is easily accessible on foot. Pedestrians walk through the mall on their way home and informal traders sell their goods at the entrances. This is quite unusual for a mall. Malls in the US and in South African cities are commonly built on empty land in suburban areas, only to be reached by car, as is the case with Greenstone Shopping Centre (locally called Greenstone). This central location of Maputo Shopping ref lects and reinforces the monocentric form of the city, which is, however, gradually shifting towards a more polynuclear form, as exemplified by malls emerging in areas of expansion such as along the coast line. Maputo Shopping consists of two main buildings, a large courtyard and an outside parking area usually jammed with cars. The mall attracts its customers with a shiny façade consisting of grey aluminium panels and window fronts made of blue mirrored glass. This is a contrast to Greenstone Mall and many malls based at transport axes on the urban outskirts. These malls tend to have an inward-oriented 'siege architecture' (Murray 2004: 142) which pays little attention to the outer façade. The interior design of the main building of Maputo Shopping is strongly inf luenced by Middle Eastern aesthetics and mall architecture. The f loor is covered with elaborate mosaic patterns made of shiny marble tiles. Many young people from Polana Caniço experienced this building as exceptionally beautiful; they describe it with the term *bonito*, which they equally apply to the luxury mansions in the elite area of Sommerschield II. In contrast, members of the urban elite of African and Portuguese descent, who had travelled extensively to malls in South Africa and elsewhere, tend to perceive it as ugly and overloaded. They read the decoration as distinctly 'Indian' and associate the mall with the Indian ethnic group. Raquel, a 40-year-old resident of Sommerschield II, explained it like this:

Raquel: The shopping centre as a building is a bit overly Indian (laughs). Its decoration … but I understand, as the owner is Indian, right. Barbara: But in which way is it overly Indian? Raquel: It's overly golden,the decoration is … it's not a simple thing, it's not modern. This is when I compare it with malls in South Africa where I normally go (Raquel, resident of Sommerschield II, January 2011).

Maputo Shopping belongs to the company MBS, which is also the colloquially used abbreviation for Mohamad Bachir Suleman, the owner of the mall, who is indeed a well-known, aff luent Indian merchant. He apparently invested 32 million dollars in the 52,000 square metre mall (Canalmoz 2007). The initial plan of the mall was drawn up by an architect, but later the owner changed it according his own taste and had it built by a Chinese construction company (Sebastião, manager of the Maputo Shopping Centre, November 2010). Since then, the mall has been a work in progress rather than a completed endeavour, with facilities added or rebuilt over time. The mall is owner-built, as well as owner-managed, which is also quite rare for a mall. The owner himself, making himself an actor in the transnational f low of ideas, travelled to Brazil and headhunted an experienced shopping mall manager. In 2012, the mall owner Bachir did not have a contract with a private security company (so-called *seguranças*) but employed the security personnel himself (so-called *guardas*). In 2012, the enforcement of the official rules (according to a sign at the entrance it is forbidden to smoke, to drink alcohol in the hallways, to carry weapons, to run, to take photographs or film, or to leave children unattended) was not particularly strict, however. In the afternoons and at the weekend in particular, groups of children roamed the centre without their parents, and were seen running up and down the escalators. Sebastião, the mall manager, claimed in an interview that informal traders were not allowed at the mall entrances, but in reality, they were seen standing there quite often, although security personnel sent them away from time to time. This relatively relaxed implementation of the rules stands in contrast to Greenstone Mall where I could neither observe trespassing by users nor tolerance by the security employees. The sometimes contested boundary between inside and outside and disjunctions between official rules versus lived reality question the enclave character and total control which the existing literature often describes for shopping malls.

Maputo Shopping contains a supermarket and many small boutiques which specialise in clothing, electronics, books and gifts, as well as a movie theatre, a gym and a games arcade. There are two food courts: one is on the fourth f loor and this is where the non-Muslim mall users who can afford to buy food at a restaurant usually mingle. The other is an open court between the two buildings called Guebuza Square. Many of the restaurants here are Indian-owned and focus on Muslim customers; they offer halal food and do not sell alcohol. In the middle of the square there are chairs and tables where people are allowed to sit without being customers of a restaurant: here mall visitors, usually in groups of two to ten, sit down, socialise and people watch. They often include mall visitors with little means who have bought food or drink at the supermarket.

Like an onion, Maputo Shopping contains several layers ranging from more public to more private spaces, with increasing access restrictions. Besides the mall spaces open to the customers, there is a Muslim prayer room on one of the top f loors, only accessible to people who share the faith of the owner. Then there is a back region, as Goffman (1959) would call it, to which only the workers have access, including service rooms and the offices of the mall management. The most private and most secluded spaces are the mall owner's private rooms, probably situated on the topmost f loor. The existence of these private rooms is publicly announced by a golden elevator which is reserved solely for Bachir, a symbol for his wealth and inaccessibility. In November 2014, the owner was nevertheless kidnapped by armed perpetrators in this very mall and he was only released a month later. Such incidents undo the perception of malls as highly secured and secure spaces. Equally, in Johannesburg, shopping malls are not safe from attack (Mtakati 2014).

In 2010, the centre was attracting about 300 to 400 people per day and around 12,000 people a month. The centre management was content with these numbers but wanted to increase them by around 20 per cent (Sebastião, mall manager, November 2010). Not least because of this, the centre management opened new facilities; in 2012, a mini-theme park, the Maputo Game Centre, was installed on the fifth f loor, mainly to make the mall more attractive to families. This venue, which includes coin-operated entertainment devices (games arcade), simulates a fantasy underwater world. On Sunday afternoons when Maputo Shopping is at its busiest, the games arcade – the air filled with the sounds of video games and carousel music – becomes inhabited by families and groups of children and teenagers from diverse social milieus. One of the salespersons confirmed this: "It attracts all kind of people, all the different classes, humble people and people who own more" (Salesperson at Game Centre, July 2012). In such a world of fantasy, spaces open up for ephemeral entanglements across lifeworlds.

### Global Modernity or Money Laundromat?

When the mall was inaugurated on 8 May 2007, the then president of Mozambique, Armando Guebuza, gave a speech in which he portrayed the mall as a driver of economic development and modernisation in the commercial sector, as it increased competition and diversified the shopping possibilities (Canalmoz 2007). Referring to the regional and global competition between cities, Guebuza argued that the mall constituted a new architectural contribution to the cityscape which would improve the image of the city as a whole. The owner, Bachir, claimed at the opening that from then on it would be possible to buy in Maputo the best goods in the world (ibid). He represented Maputo Shopping as the only place in Mozambique where urban dwellers could indulge in all the achievements of modernity and the global consumer culture. These inaugural speeches, based on tropes of progress and development, represented the mall as the entry point to the global consumption-scape of brands and shopping experiences. They portrayed the mall as a global, cosmopolitan island in the midst of a yet developing city and nation.

Such tropes of modernity and development need to be contextualised to the Mozambican history of consumption. During the socialist phase and the civil war, consumption possibilities in Maputo were very limited (Pitcher 2002: 90). At this time of nationalised industry, the often sparsely stocked shops and wholesalers were owned by the government and by Indian merchants (Miller 2006: 154). The acute shortage of imported goods led to considerable illicit border trade and a black market (Gastrow 2003: 11). During the transformation from a socialist to a market economy, the liberalisation of prices and free choice of suppliers enabled the expansion of the retail sector. The growth of elite and the middle-class milieus, plus the presence of well-paid expatriates, led to a considerable demand for consumer goods in the city. In 2008, the municipality considered the retail sector to be one of the most important and most dynamic economic sectors of the city (Município de Maputo 2008: 118-9). The retail landscape in Maputo is nowadays highly diversified (Dawson, Findlay and Paddison 1990).

The end of apartheid in neighbouring South Africa ended the sanctions and opened the way for the investment of over-accumulated South African capital into other African countries. Foreign investment started to come back to Mozambique after 1994, when investors perceived the country as politically stabilising and the government as more legitimate (Pitcher 2002: 151). In the last few years, South African retailers have entered the Mozambican market, which, in turn, has become more and more competitive (Lucas 2007: 1). In 2008, the municipality of Maputo believed that this trend would continue in the future (Município de Maputo 2008: 119). Nevertheless, tariffs, transportation costs and rents captured by traders make goods relatively expensive in the city, which is why many Mozambicans like to shop in South Africa and informal cross-border trade is a lucrative business.

In 2012, movie theatre was inaugurated at the mall. At the ceremony, Armando Artur, the then minister of culture, congratulated the Portuguese cinema distributor Lusomundo for bringing new 3D technology to Maputo. He claimed that this was a "big contribution to the growth of the country in terms of its development" (Artur quoted in Canalmoz 2012). A representative of Lusomundo, Luís Moto, called the new 3D technology 'revolutionary' and claimed that Maputo 'deserves' the best technology in the world (Canalmoz 2012). Following the fight for independence from colonialism, the socialist 'revolution' and the turn to democracy, the revolution of consumption is now taking place, guided by the Maputo Shopping Centre, it seems. This discourse of modernity and newness, however, silences local history. Cinema has a long history in the city but the introduction of Structural Adjustment Programmes in the 1990s led to a decline in national cinema production (Power 2004: 261).3

<sup>3</sup>  The first cinematic exhibition in Maputo took place in 1890 (Power 2004: 264). During colonial times, the local population was not allowed to enter the cinemas; under the Portuguese they were exclusive, racialised spaces.Nevertheless, moviesreached the local population in the form of mobile cinemas. In

At the opening of the 3D theatre, the mall management emphasised that the new movies would be screened at the same time as in Europe and South Africa (Sebastião, mall manager, August 2012; Canalmoz 2012). This statement referred to the fact that until 2012, movies were screened in Maputo only a couple of months after they had been screened elsewhere and after many people had already watched them at home using cheap DVD copies bought on the black market. With the immediate screening of new movies, Maputo had become synchronised with the rest of the world – the mall allowed this African city to connect to global time.

The Brazilian mall manager, Sebastião, was an important promulgator of the discourse of modernity and a significant actor in the f low of ideas from elsewhere to Maputo. His frame of reference was not the local shopping landscape of markets, informal stalls and *mercearias* (small shops); but compared Maputo Shopping to malls in Brazil and South Africa whose standards of quality and professionalism he believed they should aspire to. He saw his job as going far beyond managing a facility. Rather, acting as a facilitator of what he understood as modernisation, his task was to inf luence local practices of going to the mall and shopping. The urban population had to be taught how and when to go to the mall.

Shopping malls are a novelty in Mozambique … People only now started to develop the habit of going to the mall. Before that, they shopped in the city centre. Now we succeeded that they associate shopping with the shopping mall (Sebastião, manager Maputo Shopping Centre,November 2010).

He explained that through their marketing, customers in Maputo would develop the habit of frequenting the mall in order to buy a gift for commercialised holidays like Mother's Day or Valentine's Day. His goal was to make 'malling', meaning going to the mall for shopping, leisure and entertainment, an important habit in Maputo public life. In this sense the management saw itself as pioneer; they were creating new urban practices by offering a new space to the urban population.

This discourse of economic development and modernity was not, however, positively received by all users of the mall. In an interview with Nelinha, the 16-year-old school girl living in Polana Caniço, I showed her a photograph of the mall and, as mentioned in chapter 5, she immediately associated it with the planned gentrification of her neighbourhood and the threat of removal.

<sup>1975,</sup> Frelimo inherited 42 cinemasfrom the Portuguese colonialstate (ibid: 263-4). Cinema has always occupied an important position in the making ofidentitiesinMozambique (ibid: 278).Atthe beginning of the 1990s, the movie theatres were in state of neglect and many were taken over by the Universal Church, an influential Brazilian Pentecostalchurch (ibid: 276). According to Van de Kamp (2016:13)this was not only a change of the use of these buildings, but also a shif t in public life. As movies were no longer screened, going to church was welcome entertainment. The national cinema production and screening companies never recovered. Mozambican investors were taken over by foreign private sector involvement (in this case by the Lusomundo media group, the biggest communications group in Portugal). Most of the screened movies are nowadays Hollywood productions. Prices that were once quite low have become inaccessible for the majority of the population. Movie theatresin Maputo also have a hard time because of pirate DVDswhich can be bought on any streetcorner.

They say they will construct hotels and restaurants [similar to this mall] in Polana Caniço and because of this we have to leave … And imagine one day I would be walking past where my old house was and I would see such a shop [like the mall]. I would stop to look and think about the house which was there originally. I would cry, point to the inside where my bed used to be, where my room used to be, where the kitchen used to be. So the *saudade*(longing, yearning)will become very strong. Butthen Iwould have to leave the city which used to be my home in order to reach the outskirts where there is no civilisation, in the complete bush. So thatis difficult. Butitwould be good for the country. For the country it would be very good (Nelinha, resident of Polana Caniço, November 2010).

While the shiny façade of the mall promises the illusion of modernisation for everyone, Nelinha sees through it, seeing her old house being replaced by a similar development in Polana Caniço. This excerpt from the interview exemplifies how many urban dwellers experience the current economic growth symbolised by the mall and the luxurious mansions constructed in Sommerschield II. Many people like Nelinha see it is a development which is not for them, a development from which they are excluded. Ordinary salaries stay the same while living costs rise constantly. Within the nationalist discourse of modernisation employed by Frelimo, which Nelinha picks up on in the quote above, the plight of the individual becomes subordinated to the higher goal, the economic growth of the nation. For this young woman, the 'beautiful' mall becomes a symbol for this ambivalent process of modernisation which Maputo is undergoing, and which threatens her neighbourhood and her lifeworld. The mall as heterotopia hence emerges through the entanglement of Nelinha's home and biography with the larger processes of gentrification, embedded in ambivalent understandings of modernity.

The discourse of the mall as Maputo's entry point into global modernity is quite powerful, not least as the management can reproduce this image in promotional texts on their website, in newspaper articles and in speeches by politicians and the owner at opening ceremonies. There is an alternative representation, however, which is in competition with this discourse of global modernity. It is a counter-image upheld by the urban dwellers, which portrays the mall as a money laundromat, as a place that assists its owner, Bachir, to launder illicitly earned money. This is a very powerful and highly political image, with the power to alienate many customers from frequenting the mall, exemplifying the entangledness of malls and politics.

The end of the civil war and the first democratic elections in 1994 not only brought political stability and economic growth to the country, but it also led to an increase in organised crime. Because of the long Mozambican coastline and the weak state institutions in the 1990s, the country became one of the main transit routes for drug trafficking, trafficking in human organs and money laundering.4 Knowledge about this illegal and dangerous economy consists of rumours, scandals uncovered by the newspapers, and a few research reports such as the one by Gastrow and Mosse, not least because it is dangerous to research such networks (Gastrow and Mosse 2003, Irish-Qhoboshe-

<sup>4</sup>  Drug trafficking is an old trade in Mozambique, though. Apparently, Swahili towns in northern Mozambique have been involved in this trade for centuries, even before Portuguese colonialisation (Gastrowand Mosse 2003: 50).

ane 2007: 3). One of these transnational drug networks probably consists of Pakistani and Mozambican citizens belonging to the Indian merchant milieu (Gastrow and Mosse 2003: 48). Many of them are involved in the import/export business and it is a commonly held believe that certain Indian business men smuggle drugs hidden in the fridges, chickens and other goods they import (ibid: 51). These illicit activities are said to become facilitated by close connections to inf luential politicians and there is the popular perception that with enough money, criminal actors can buy themselves immunity (ibid: 51-52). In everyday conversations, in TV debates, newspaper editorials and readers' letters the view that the economy, politics and the illegal are highly intertwined is regularly raised.

Drug trafficking is accompanied by the demand for money laundering. Some of the money is spent on objects of conspicuous consumption like cars, and some of it is said to be channelled into real estate investments in Maputo and elsewhere (Gastrow and Mosse 2003: 52, Hanlon 2010). There are rumours that a considerable portion of the new high-rise buildings that have been added to the Maputo skyline in the last few years were financed by capital from such sources. The illegal economy seems to be one of the factors pushing new urban developments in Maputo. These rumours become crystallised in the Maputo Shopping Centre:

Barbara: Do you sometimes go to the Maputo Shopping Centre? Senhor Sousa: I don't enter *lavandarias* (laundry shops). I only enter the laundry of my grandmother. When I have driven past Maputo Shopping, I have to disinfect myself at home. (Laughing). I am joking, but I have indeed never entered (Senhor Sousa, resident of Sommerschield II, July 2012).

For Senhor Sousa, a 70-year-old Christian Indian living in Sommerschield II, as for many other urban dwellers, Maputo Shopping is anything but a neutral space; rather, they perceive it as the embodiment of the owner's economic power. The metaphor *lavandaria* (laundry) alludes to the rumour that the mall is an instrument for channelling the profits from the illegal economy into real estate investment.

As already mentioned, MBS, the way in which urbanites colloquially refer to the owner Mohamed Bachir Suleman, is a public figure, and has the reputation of being one of the wealthiest people in the city. The mall users have heard many rumours about the apparently dubious origins of his wealth. These rumours reached an even wider audience and garnered more legitimacy when the United States blacklisted the owner of the mall on the basis of being a drug baron in June 2010 (Hanlon 2010, US Department of the Treasury 2010).5 Many mall users observe that a considerable number of

<sup>5</sup>  In June 2010, Bachir became named a drug baron by theUS president BarackObama underthe Foreign Narcotics Kingpin Designation Act (US Department of the Treasury 2010). This explicitly designated the Maputo Shopping Centre and two of Bachir's other businesses as being part of his drug trafficking network. This designation made it illegal for any American businesses to conduct business with Bachir'scompanies and American citizenswere forbidden to enter them. The American embassy even called on their American and Mozambican employees not to shop at the mall (Hanlon 2010). Bachir himself, the Frelimo government and media close to Frelimo have since denied the allegations. As a result of the scandal all the banks at the mallclosed their branches within a few weeksin order to protect their reputations(Club of Mozambique 2010). At least until 2012 therewas no ATM at the mall and customers could only pay in cash.

the luxurious boutiques at the mall hardly ever have customers, which makes them think that these shops are really just a front for money laundering. According to the centre management, the mall recovered quickly from the decline in customer turnover after the scandal, but research showed that many members of the urban elite avoid the mall, as they do not want to be associated with the rumours.

What also fuels this image of the mall is the fact that the outside food court has been named Guebuza Square. The name is in English ('square' instead of *praça*), which alludes to Nelson Mandela Square, the European-styled piazza in Johannesburg's most famous mall, Sandton City, which is named after the South African icon Nelson Mandela. Guebuza Square is well known across the entire country, as it is often used as a stage for marketing events by brands advertised on TV. The mall management named this hypermodern space Guebuza Square in order to honour the then president, Armando Guebuza (2005–2015), and in order to underline the commitment of the Indian business man to the Mozambican nation. However, mall users interpret the name quite differently: Bachir and Armando Guebuza are said to be close friends, and Bachir is known to have funded Frelimo campaigns with considerable sums of money. The naming of the square after the president is for the urban dwellers the material representation of the interlacing of political and the economic power, and also of the connection of the president himself to criminal networks. In informal conversations, urban dwellers would say the name Guebuza Square with an ironic or sarcastic tone, humorous sarcasm being a common way in which urban dwellers in Maputo articulate their view that their country and their government is malfunctioning.

The mall as entangled heterotopia thus makes visible for its users what is otherwise supposed to be invisible, namely, the entanglements between the realm of politics and economic power. As heterotopia it exposes the democratic ideal, that is, that a president should represent the interests of the population and not be captured by powerful economic players, as an illusion. In addition, malls as conceived spaces come into being through multiple, contradictory and competing representations and discourses upheld by the mall management, politicians, users and others.

### Maputo's Shopping Lifeworlds

Urban dwellers transform malls into far more than mere spaces of consumption. As in markets (Watson 2009), at malls multiple practices and forms of sociality become enacted. In Maputo, urban dwellers themselves distinguish between two ways of using the mall: on the one hand, there is *fazer compras*, which refers to buying things in its literal sense. On the other hand, there is *passear*, which means to go for a walk or to stroll around and refers here to a leisure activity involving moving through space, mostly in company of others, observing others and enjoying the space simply because it is there. In actual shopping trips, the two usages, leisure and shopping, are often intertwined. The distinction, though, is empirically and analytically important, and draws attention to the differing position of the mall within users' shopping circuits and emic perceptions. Users who come to the mall mainly for shopping (*fazer compras*) situate it within their shopping circuits and compare it to other consumption spaces, while users who frequent the mall for *passear* locate it within their leisure circuits and compare it to other spaces of leisure and public life. Different milieus use the mall differently: members of Muslim-Indian elite milieus frequent the mall for socialising and leisure, while urban elites of Portuguese or African descent tend to frequent it only

for shopping. Among residents from Polana Caniço, who frequent the mall less often than more aff luent urban dwellers, both uses exist. There seems to be a tendency for young people from Polana Caniço to explore the mall as a leisure space, while adults buy things there they cannot find elsewhere. So for different users, the mall takes on different meanings in everyday life. Yet these lifeworlds of people who live apart in the city – for example in the entangled neighbourhoods of Sommerschield II and in Polana Caniço – come to intersect at the mall.

### *From the Beach to the Mall: Indian Public Life*

The Maputo Shopping Centre has an important place in the leisure circuits of Indian milieus, both Muslim and non-Muslim, as well as Muslims from other regions. Non-Indian urbanites tend to refer to them as 'Indian', an imprecise term conf lating people from the Pakistani and Indian diaspora in Mozambique, the majority of them Muslims, with Muslims from the Middle East and other Muslims wearing the Islamic clothing fashions inf luenced by Saudi Arabia. I have adopted this local taxonomy, but I use it in the plural (Indian milieus) to avoid erasing the diversity.

As the shops of Indian merchants are closed on Sundays, this is an important day for the fostering of social relations and enjoying sociability. On Sunday afternoons many come to the mall with their extended families where they bump into relatives, friends and acquaintances. Members of Indian milieus populate the tables in the restaurants, sit in the open Guebuza Square and stroll through (*passear*) the hallways of the mall. So, on Sundays, the mall is transformed into a space of experience, leisure and sociability. The building itself with its elevators and its ornate decorations, as well as the strolling visitors, become attractions for the Indian crowd who enjoy, always in groups of friends or family, the co-presence of intimates, relatives, acquaintances and strangers who somehow belong to their Indian-Muslim social world.

Many members of the Indian elite living in Sommerschield II, like Senhor Mattar, the 55-year-old head of the prayer place in the neighbourhood, frequent Maputo Shopping regularly.

Barbara: So do you plan to meet there with friends on Sundays or how does it work? Senhor Mattar: No, the friends come if they are around, maybe some may be travelling or tired, butsometimesit happensthat we all meet there. Then the table becomes very large. We go therewith children, in a group of maybe eight people (Senhor Mattar, administrative head of the Muslim prayer place in Sommerschield II, August 2012).

The Indian or Muslim identity ascribed to the mall, especially on Sundays, is not least because of the women's Muslim dress, increasingly inf luenced by fashions from the Middle East, such as the black, f loor length hijabs. Yet such dress practices are anything but homogenous: there are a few fully veiled women, then there are women wearing a hijab-like robe but no headscarf, others again combine the headscarf with Western jeans, and many women also wear Western clothing without a headscarf and even short skirts. The majority of the Muslim men wear jeans and t-shirts, yet there is also a minority with long beards and prayer robes. This diversity can even be found within a group of people spending time at the mall together, for example a fully veiled women walking around with her cousin in a mini-skirt. But for everyone, going to the mall entails dressing up: mallgoers put on their good jeans, their new sneakers or their

new hijab. Women in *capulana* (African cloth) and destitute people (*pé descalço*, figuratively and literally poor people without shoes), both associated with the rural or the peri-urban, are a rare sight, especially on Sundays.

The main activity of the Muslim-Indian visitors at the mall is not shopping (*fazer compras*); shopping bags and pushing trolleys are rarely seen. Instead, many of these mallgoers walk up and down the mall in groups of two to eight people (*passear*). They tend to be multigenerational groups consisting of family members, or they are in certain age groups which are either gender segregated (e.g. ten women in their early twenties) or consist of several friends or married couples of a similar age. People tend to stroll around slowly on Sunday afternoons, adults and children walking up and down the elevators, strolling through the hallways, crossing Guebuza Square outside and then returning back inside. Groups of children play games together in the games arcade, run around the hallways glancing into shop windows, and then return for a quick break to the restaurants where their parents are sitting with relatives and friends. The children eat chicken and fries, and then continue their games. Parents seldom tell their children not to run around or to be quiet. They enjoy socialising in the midst of strangers while their children are playing in an environment which they experience as child friendly and safe. The composition of the groups in the restaurant constantly changes, with friends or relatives arriving and leaving. Within minutes, the waiters rearrange tables and chairs in order to make space for these shifting group compositions. This constant adaptation of chair–table arrangements to the forms of sociability is typical of restaurants in Maputo and, hence, turns the mall into a typical Mozambican social space.

On Sundays, many families of Indian milieus congregate with relatives at home, and more private celebrations like weddings, birthdays and funerals take place. Yet, when there is no family gathering planned, many enjoy going to spaces of public life (*passear*) on Sunday afternoons. Before the mall opened in 2007, the beach front along Marginal Avenue and 10 de Novembro Avenue served as a meeting place for Indian milieus on Sundays, as 60-year-old Senhor Ismail and others explained:

SenhorIsmail: In the past people used to sit attheMarginal, butsince theMaputo Shopping opened, people go there as it is very convenient … People go there for eating, for shopping. They bump into each other and meet each there. Barbara: They don't go to theMarginal anymore? SenhorIsmail:Only very few. In the pastit used to fill up, driving pasttherewas almostimpossible (laughing)(SenhorIsmail,resident of Sommerschield II, October 2010).

Some interviewees referred to these social gatherings as the 'marriage market', where parents look for potential husbands and wives for their children, and where young people themselves come to see and be seen.<sup>6</sup> So when the mall opened in 2007, this 'marriage market' moved from the beach to the mall, the mall thus being appropriated and integrated into already existing forms of sociability and patterns of public life. According to Sebastião, the centre manager, there was no plan for the mall to become

<sup>6</sup>  In the past, parents had a great influence on their children's choice of marriage partner in Indian milieus but people told us that this influence is decreasing. Indian milieus still have a reputation of opposing exogamous marriages, however, i.e.with Mozambicans of African descent.

a space of public life for the Indian milieus, it simply happened. Customers who frequent the mall for *passear* (leisure) instead of *fazer compras* (shopping) are welcomed by the management as they observe that the families nevertheless spend money at the shops and restaurants.

### *Not Good Enough: Urban Elites' Expectations*

Aff luent and middle-income urban dwellers who do not feel that they belong to the Indian milieus use the mall quite differently. For them, the mall is a place for shopping and they evaluate it accordingly. Raquel, a resident of Sommerschield II of Portuguese descent, explained to me: "I sometimes go to the mall to buy specific things, but I don't go there to hang out (*passear*)" (Raquel, January 2011). They go to the mall because goods are sold there that they cannot find elsewhere or because the prices are lower. Many urbanites do not buy all their groceries at one shop but tend to buy specific products at specific supermarkets or grocery stores because prices, freshness and quality of the products can vary considerably. Many members of the urban elite buys groceries at one of the many small, often Indian-owned, grocery stores in the city centre (*armazens*, *mercearias*), or they buy them at the Mercado Central, the central market, or at informal stands on inner city streets. They increasingly also frequent the growing number of supermarkets, like the HyperMaputo at the Maputo Shopping, and the South African owned Shoprite, Woolworths, Game and Mica. None of the residents of elite Sommerschield II who I interviewed shops at the many markets in *bairros* for groceries, like the nearby market Compone; these are clearly outside their shopping circuits.

Some of the new supermarkets are located outside the city centre and shopping there enables elite urban dwellers to stay away from the inner city which many elite interviewees experienced as noisy and clogged with cars. Besides the Maputo Shopping, members of the elite frequent the Polana Shopping Centre, situated in the upper part of the city centre, and the Shopping Centre Super Marés, which opened its doors in 2012 on Marginal Avenue at Costa do Sol beach, close to the new upmarket gated communities and villas. Super Marés is owned by a local business man, but the architecture shows clear inf luences of contemporary mall architecture in South Africa. This shopping facility is situated some distance from the city centre and forms part of an emerging edge city development. In an online newsletter aimed at the expatriate community, the new shopping mall is advertised as a place where one can park without stress, alluding to the lack of parking in the inner city (Club of Mozambique 2011). There are several more mall projects being planned for Maputo, many involving South African investors. Urban elites' everyday trajectories become increasingly transversal in the sense of driving from suburb to suburb rather than from suburb to the centre and back.

Members of the urban elite, as well as less aff luent urban dwellers, feel that the shopping opportunities in Maputo are still limited and of inferior quality in comparison to nearby South Africa or Europe. The shopping circuits of Maputo's elite are hence deeply transnational: many elite families drive several times a year to Nelspruit, a South African town about three hours away by car from Maputo. Besides enjoying sleeping at a four-star hotel, going to the movies and eating at South African restaurants, the families go to buy clothes in Nelspruit, where they are cheaper and of better quality than in Maputo where much of the clothing is imported from China. For many, the clothes shopping circuits even expand to other continents: business trips or family visits to Europe or the Middle East are used as an opportunity for clothes shopping and the purchase of consumer electronics. Until few years ago, many even bought groceries in South Africa. With the depreciation of the Mozambican metical in relation to the South African rand since 2000, however, and the later trade liberalisation in the SADC region, price differences have been reduced.

Hélder, a 55-year-old member of the Frelimo elite living close to the fish market, in a neighbourhood similar to Sommerschield II, claims to not have bought a single piece of clothing in Maputo since 1995 (Hélder, resident of the fish market area, September 2010). Like many others, he finds the prices of clothes and furniture at the Maputo Shopping overpriced and the quality poor in comparison to abroad. Jorge, a 30-something, well-paid lawyer living in the inner-city neighbourhood Coop, likes to buy branded clothing, but he suspects, again like many others, that the branded products sold at the Maputo Shopping may be counterfeits imported from China. Raquel does not like the Maputo Shopping because the presentation of the goods does not conform to her standards and disconcerts her senses. The supermarket smells of meat and there are f lies, she explained to me (Raquel, a 40-year-old resident of Sommerschield II, January 2011). The products on the shelves at HyperMaputo are indeed sometimes full of dust and damaged, and not always of impeccable quality (Club of Mozambique 2016).

Like many others, Senhor Benedito, a former deputy-minister living in Sommerschield II, does not go to the mall for leisure purposes: "I go to one of the shops to buy something and then I go home. I don't go to stroll around (*passear*)" (Senhor Benedito, 47-year-old resident of Sommerschield II, August 2012). One important reason why many members of the elite emphasise that they do not *passear* at the mall is that they fear that by spending time at the mall, they become publicly associated with hidden networks of power and corruption (see above). Some claimed it was also about ethics; they have hesitations about spending their money in a mall allegedly constructed with money from the drug trade.

### *Too Expensive, but there are Escalators: Mallgoers from Bairros*

The meanings of heterotopias derive from the entanglements, their relationships to other urban spaces and to the larger urban society. In order to understand Maputo Shopping as a social space, it is necessary to look at the place the mall takes within the larger everyday shopping routines of its diverse users. For the majority of the residents from Polana Caniço, Maputo Shopping does not form part of their everyday shopping circuits, mostly because they have the perception that the prices there are too high. Their shopping circuits are rather centred on the markets, informal stalls (*bancas*) and small grocery stores (*contentores, barracas*) in the *bairros*, although they tend to buy electrical appliances and plastic household utensils in the Baixa section of the city centre.<sup>7</sup> For clothing and consumer electronics, there are parallel legal and illegal markets based on transnational connections, as these goods are too expensive

<sup>7</sup>  Informal traders buy household items in the city centre and sell it in the 'bairros' at higher prices, so that it is cheaper to buy them in the city centre. For fruit and vegetables people say it is the other way around, as they either originate from the informal traders' own fields or traders buy them in bulk at the wholesale market on the outskirts of the city; buyers in the city centre pay a higher price due to transport costs.

for Polana Caniço residents at normal shops. Many residents buy cell phones and similar items on the black market *Estrela* or through underhand dealings with acquaintances. When my cell phone was stolen in 2010 my friends told me half serious, half joking to look for it the following day at the Estrela black market. Young men travel to Johannesburg to buy stolen goods, usually through criminal networks constituted by co-nationals, and sell them back in Maputo, as Leandro, a 30-something former stolen goods trader, explained to me (Leandro, August 2012). Many *bairro* residents buy second-hand clothing from Europe (locally known as *roupa da calamidade*) on one of the many markets (especially Xipamanine), as they are of better quality and cheaper than the Chinese clothes sold in the city centre stores. Many residents of Polana Caniço also make a living based on these trades, for example by selling the second-hand clothing they buy from importers, who are usually Indian merchants (Brooks 2012: 288). As urban dwellers living in Polana Caniço can generally not afford to go on shopping sprees to nearby South Africa, entrepreneurs have made a business out of cross-border trade. The transnational networks of Indian merchants, criminal networks with South Africa and cross-border trade enable urban dwellers in Maputo to access consumer goods they could otherwise not afford.

Olívia, who was about 30 years old in 2012, was such a second-hand clothing seller on the market Compone, yet even though her husband was minibus driver it was still hard for her to feed her four children. She was a neighbour of the family I was staying with from time to time in Polana Caniço, and we became something like friends, although I did not know how to speak XiShangana and she did not speak much Portuguese. When she found out about my interest in the Maputo Shopping Centre, she proposed that we go there together. Because she believed that she needed money to enter the mall, she had never gone there: "What will I do in there?" she asked herself. Accompanying me seemed a good enough reason (Olívia, August 2012).

We eventually went to the mall together one week day. Olívia wanted to go into the supermarket to check the prices of powdered milk. She found to her surprise that it was actually cheaper to buy it there than at the *contentor* in her neighbourhood. When walking through the supermarket, Olívia put a shopping basket on her arm, not because she was used to it but, as she explained, because she had always wanted to shop this way. At a *contentor* in the *bairro*, she had to stand in a queue till the sales clerk, working behind a counter, gave her the goods she ordered from the shelves. Touching packed groceries and studying the prices was therefore something special for her, which she only did once a year when they did bulk shopping at Shoprite for the festive season.

Earlier that day, before we had left her house to go to the mall, she asked in a wave of insecurity: "Will I not embarrass you?" The question deeply troubled me but made me aware of how much going to the mall meant venturing into a 'foreign country' for her; how much it meant crossing social boundaries. Before we left her house, she had replaced her *capulana* with a white skirt and a colourful blouse and she put on her best shoes, her wedding sandals, as she explained to me. Yet there were things she could not change: a couple of her front teeth were missing, for example. She had braided her natural hair into neat rows, but in contrast to the shiny hair extensions of other women at the mall, her hairstyle was very simple, not least because of the tiny bits of maize stuck in it from pounding maize in the morning. For people living in *bairros*, being clean and neat, which is what they think urbanites should look like, demands work and

even tricks. Many office workers change their shoes before they enter their workplace, as their other shoes are full of sand from the *bairro* paths they had to use to get to work. Like the office, the mall is an urban space for which one's body needs to be prepared to go, and where urban dwellers can feel deeply out of place if they struggle to keep up to what they see as normative standards. Going to the mall even entails acquiring certain skills and bodily knowledge. Like many others who have never been to the mall, Olívia knew that the building had escalators but she had never been on one. When we went on the escalators, Olívia grasped my arm firmly and felt deeply anxious, preferring to take the stairs on the way down. Her experience of the mall stands in contrast to her everyday physical practices, and hence constitutes a space of exception, a break with the ordinary.

Many residents of Polana Caniço we talked to, even 47-year-old Senhora Aurora from the middle-class milieu in Casas Brancas, thought that "one cannot go to Maputo Shopping without money" (Senhora Aurora, August 2012). Gustavo, a resident of Maxaquene (a neighbourhood next to Polana Caniço) and a worker at Café Sol in Sommerschield II, went to the mall for the first time out of curiosity because he had never been on an escalator before. He told me that he was surprised that the prices were similar to other places, and as a consequence he started to go to the mall regularly to buy specific products at HyperMaputo which were cheaper there than elsewhere. When he went there for the first time, he bought something small to take home which he did not actually need, almost like a souvenir: "I had to come home with something from the Maputo Shopping Centre" (Gustavo, resident of Maxaquene, October 2010). Many inhabitants of the *bairros* never go to the mall, yet for those who do, the mall is something extraordinary, almost like a tourist attraction. "I felt well and as if I was not in Mozambique" (Senhora Aurora, August 2012). Some, like Gustavo, start including the mall in their shopping circuits for specific bargains. Yet only a few adults from Polana Caniço go to the mall for leisure, and if they do, it tends to be people with a formal job or who have had a lucky break, or to celebrate a birthday or a graduation one of the restaurants, for example at the popular restaurant Mimmos in the upstairs food court, rather than at one of the *halaal* restaurants downstairs.

Young people from Polana Caniço were less hesitant than the adults to appropriate the mall for their own purposes. As the mall is situated in the city centre, children and youth can easily sneak into the mall after school on their way home. Many young people from *bairros* venture regularly to the mall as a leisure space (for *passear*), which for them is full of interesting experiences they do not need much money for. Riding up and down the escalators, staring at the shop windows and other mall users, and sharing cheap food like biscuits with each other is not expensive. In contrast to the members of the Indian milieus who tend to come as families, these children and young people rarely come here with parents, but rather with groups of friends from school, from church or from the *bairro*.

On the trip to the mall with Olívia, the second-hand saleswoman from Polana Caniço, we were roaming around the games arcade on the fifth f loor when Olívia's attention was caught by an Indian family whose children were sitting on a carrousel, each with a pack of juice in their hands, and the mother holding a pizza carton. Imagining herself in this mother's position, Olívia commented that she would like to do that, too – come here with her family and eat together. Living in a *bairro* like Polana Caniço where many of her neighbours and probably also herself have experienced hunger, trying to make a living in times of hardship, makes the social practice of eating luxury foods like pizza at the mall extraordinary.

Eating at the mall is also important for the youth. Children of all ages, some in school uniform and others not, sometimes gather in the afternoons on Guebuza Square in large groups and arrange the tables so that they can sit in a circle and share cheap biscuits and soft drinks bought at the HyperMaputo. For many of the interviewees from Polana Caniço, eating something whilst sitting with others at Guebuza Square, be it pizza or chicken ordered at a restaurant, be it *bolacha* (biscuits) or ice cream bought at the supermarket, was an important aspect of trips to the mall, which they remembered fondly in conversations about the mall. Sérgio, a 30-year-old university student living with his family in Polana Caniço, told me that he did not like going to the mall with his girlfriend, as she would want him to buy chocolate and other things which he considered unnecessary. While she was dreaming about eating out, he explained to me, he would be worrying about the empty bag of rice at his mother's house. At the mall, he experienced conf lict between the demands of his girlfriend and his need to give money to his family for basic necessities like rice (Sérgio, August 2012). Unlike in romantic relationships, where the man in the breadwinner role is often expected to pay, in outings in groups, be it a group of friends from the church, from school or from the neighbourhood, everyone is generally expected to contribute to the meal, depending on how much money they could organise. In the context of the economic hardship typical of urban life in Maputo, eating such extraordinary food like ice-cream, pizza or chicken at the mall took on a ritual dimension. Doing something extraordinary together, although knowing that the money could be spent more wisely, creates a kind of 'communitas' (Turner 1991 [1969]) among people whose commitment to the group became expressed by their financial contribution. Social relations then become constructed through the practice of shared consumption, strengthening some social relations while weakening relations with absent others (Miller 1995, 1998).

### Heterotopia of Illusion: The Mall as Mirror

A security guard of the Maputo Shopping Centre once told me proudly: "Maputo shopping mall is the mirror of Mozambique" (security guard, October 2010). When I asked what he meant by this metaphor, he explained that the mall is unique in the country, that it is beautiful, and that it is the first thing in Maputo tourists see when they arrive in a cruise ship at the nearby harbour. When ref lecting on the security guard's mirror metaphor, I came across Foucault's formulation of heterotopia in which the mirror also plays a pivotal role, and which turned out to illuminate the ethnography of shopping malls as social spaces. According to Foucault, heterotopias have a representational, mirroring role with regard to society and other urban spaces.

In the mirror, Isee myselftherewhere I am not, in an unreal, virtualspace that opens up behind the surface; I am over there, there where I am not, a sort of shadow that gives my own visibility to myself,that enables me to see myselftherewhere I am absent:such is the utopia of the mirror. But it is also a heterotopia in so far as the mirror does exist in reality,where it exerts a sort ofcounteraction on the position that I occupy. From the standpoint of the mirror I discover my absence from the place where I am since I see myself over there. Starting from this gaze that is, as it were, directed toward me, from the ground ofthis virtualspace thatis on the otherside ofthe glass, Icome back toward myself; I begin again to direct my eyes toward myself and to reconstitute myself there where I am. The mirror functions as a heterotopia in this respect: it makes this place that I occupy at the moment when I look at myself in the glass at once absolutely real, connected with all the space that surrounds it, and absolutely unreal, since in order to be perceived it hasto passthrough this virtual point which is over there (Foucault 1986 [1967]: 24).

The mirror is an actual existing object, yet when I look at myself in the mirror, I see myself in a place that actually doesn't exist, namely in a mirrored image of the reality. This tricky relation between the real and the unreal is for Foucault characteristic of all heterotopias, making them places of powerful illusions and fantasies. Linking my reading of Foucault with the ethnography of the mall, this means that, first of all, like the mirror, which is an actual existing object, the mall is a materially existing urban place that people go to. Urban dwellers integrate the Maputo Shopping into diverse shopping circuits, giving the mall a distinct, yet milieu-dependent, role in their everyday spatial practices. Secondly, Foucault says that the mirror has the power to turn the actors' own existence more real, because from the mirror, they can look at themselves, and see how they or their world 'really' is. This is also the case for the mall: the mall as mirror renders the urbanites' own reality more visible to them, they can see themselves from a new perspective (through the mirror), and hence they become more aware of their reality. As the guard explained above, the mall as a mirror ref lects and makes visible to Mozambicans and tourists the achievements of the country, like the recovery from civil war and socialist times with empty shelves, being now a nation where at least the better-offs are able to participate in global consumer culture. Hence, looking at the mall is like looking into a mirror; it makes you aware of where the nation is actually standing, as if it were the embodiment of the mall managers' and politicians' discourses of modernisation and national progress. Yet it also renders visible, it materialises a further aspect of Mozambican reality: in the eyes of many users, the building and Guebuza Square is concrete proof that the rumours of corruption and the links between the legal and the illegal, between Guebuza and Bachir, are real. The mall hence makes socially visible the otherwise hidden social relations ruling the economy and politics, and at the same time unmasks what seems to be visible reality as an illusion. Not least, because of this mirror-like capacity the mall becomes formed in relation to all these other realms of urban life.

The third aspect of the mall as mirror is that there is something fundamentally unreal about the mirror, as it projects the actors' real selves onto a virtual space "on the other side of the glass", which is an "unreal, virtual space that opens up behind the surface" (ibid: 17), a place where the actor are actually absent, but the mirror makes them believe that they are there, in this other world 'inside' the mirror. Translating this to the mall, it means that the mall enables urban dwellers to project themselves into a virtual, unreal world, which mirrors the real world but, like a utopia, can also suspend and overturn in. Because of this capacity, mirrors, shopping malls and other heterotopias have one fundamental effect on human beings: they spark their imagination. Heterotopias invite actors to ref lect on their place in the world, make them more aware of what their reality is like and at the same time inviting them to imagine alternative ways of being in a utopic world.

Nelinha, the 16-year old school girl from Polana Caniço, walked past the large mansions hidden behind high walls in Sommerschield II on her way to school every day, and sometimes she saw the residents driving past her, sitting behind the darkened windows of their luxury cars. On some days after school, she used to go to Maputo Shopping, either with school friends, or alone to stroll around (*passear*), or to visit relatives working at a shop in the mall. She and her friends would walk around, stare at the windows and also stare at the Indian elite. At the mall, for once the elite are not hidden behind walls or car windows but share the same public space and can hence be observed, studied and judged.

Barbara: How did you feel when you were there? Nelinha: Uneasy. Because there were only (pause) big people (*gente grande*). I felt very small. I justwalked pastthem, my eyes fixed on the floor … I just watched, from the outside [of the shops], but Icouldn't enter to buy anything. See, my parents(pause) they are poor. So there is alwayslack ofsomething (Nelinha, resident of Polana Caniço,November 2010).

Being in this space together with *gente grande* (euphemism for wealthy and powerful people) felt uncomfortable to her, not least because the high prices and her empty pockets made her feel excluded. This provoked her ref lections about her own reality, making her aware of her own social position, the poverty of her parents and the many unsatisfied desires. The mall has an extraordinary relationship to reality, according to Nelinha. It is a fantastic place which renders visible her inner dreams, it is a paradise of abundance where everything she desires is available, the only thing lacking is money:

It is something from a dif ferent world and exaggerated. It seems as if you were in a world of dreams, as you see everything, everything that you want. You think: "If I had money, Icould simply take this, and that and that."When you get out ofthe mall, itfeels asif you were having a shock. Asif inside it was … I don't know, but you always walk out dif ferently from the way you entered. You enter and something changes there inside, andwhen you get out, you breathe, and you notice howyou returned to the normal life (Nelinha, resident of Polana Caniço, August 2012).

For Nelinha, being at the mall was a deeply transformative, almost religious experience, which she found extraordinary and, like an experience of liminality (Turner 1991 [1969]), made her feel different once she left the building and returned to the busy streets and to real life. When I asked her whether she felt better or worse after the visit at the mall, Nelinha explained:

There is always the feeling of deception, but nevertheless, I usually feel much better, because I think … I can still dream, I still have the capacity to achieve what I want. This is the spirit you feel when you get out: "I see things which I cannot buy and neither my parents can buy it for me." So I think I will study now with even more ef fort to be able to give to my children what I was not able to have (Nelinha, resident of Polana Caniço, August 2012).

Observing the wealthy provokes ref lection on Nelinha's own place in society; it triggers her imagination about her own future, her plans to study in order to move out of

poverty. Because she is young, she reasons, she can still dream about her future. The visit to the mall and walking through the world of the wealthy did not so much leave her feeling deceived about life, but rather motivated her to study harder. "One day I will come back, one day I will shop here" she thought (Nelinha, November 2010).

### *Where Minorities become Majorities*

In the neoliberal, post-socialist context of Mozambique, the Indian milieus occupy an ambivalent, tension-laden position. After the abolishment of the socialist *homen novo* ideology, citizenship became relatively undefined and a 'f loating signifier', a symbol with no fixed meaning. Many Indians fear becoming the new 'outsiders' (Sumich 2013: 114). At Maputo Shopping, though, they distinctly claim their space in the city and in urban society. In contrast to other spaces where they are a small minority, here they dominate the place, especially on Sundays, which does not always lead to an easy conviviality. For members of the Indian milieus, going to the mall is mainly about being together with members of their own milieu, while paying little attention to others.

On a Sunday in August 2012, I observed an Indian family, probably a mother, her husband and their daughter, sitting at a table in the middle of Guebuza Square for about 40 minutes, and they were, just like me, the anthropologist, observing the crowd. They performed 'civil inattention' (Goffman 1963) towards members of African milieus and barely looked at them, while they openly gazed at mall users from their own religious and ethnic background. They commented to each other about them, apparently without worrying that the observed could see that they were being gossiped about. They greeted and were greeted by other Indians and Muslims. In interviews, Indian users of the mall told me that they have a heightened awareness of being observed and being gossiped about at the mall by members of their own milieus. Senhor Ismail, 60-year old Indian resident of Sommerschield II, claimed that: "I don't' like the ambience very much … It's full of people, and there is a lot of gossip, a lot of talking (laughing)" (Senhor Ismail, resident of Sommerschield II, October 2010). For him, Maputo Shopping is not an anonymous place; here he sees and is seen by others who belong to his own social world, encounters which contribute to constituting this very same social world. This exposure to the Muslim-Indian public and the related gossip is experienced by him and others as social pressure to conform to certain morals and values, which is why he presented the mall as a place that he did not like.

The social closure enacted through looks and civil inattention provokes critique from urban majorities who have become the minority at the mall. Many non-Indian interviewees told me that they feel uncomfortable at the mall on Sundays afternoon and prefer to frequent it on other days when the Indian presence is less.

Diogo: On theweekendsthe mall becomesfull of Indians. The majority of black people therefore go on weekdays when it is emptier. Barbara: And why do they prefer to go during the week? Diogo: I don't know, because it's less full. And some also say: "It's full of those *monhés*, I don't like it there" (Diogo, 21-year-old resident of Sommerschield II, August 2012).

Historically, the term *monhé* was applied to Swahili and African Muslim chiefs in the 19th century. In the 20th century, the Portuguese used the term to refer to Muslim Indians and people of Indian descent (Bonate 2007b: 139). Nowadays in Maputo, *monhé*

is used by urban dwellers to refer to members of Indian milieus of diverse origins. It has, however, a clearly derogatory tone, and it invokes stereotypes about the allegedly racist attitudes of Indian milieus which held an intermediary position during colonialism. In such moments of competition over urban spaces, when the Indian milieus, which otherwise constitute a minority, become a dominant majority in the Sunday mall, urban dwellers reactivate such colonial stereotypes about the allegedly racist attitudes of these milieus. They critique the Indian milieus' dominance with reference to the past system of racial hierarchy.

Twenty-year old Jacinta, a student living in the Casas Brancas section of Polana Caniço, often travels to Nelspruit to buy clothes which she sells in Maputo. In an interview about her experiences of South African cities, she told me that she sometimes felt unwelcome in restaurants with predominantly white customers in South Africa They would attend her because she was a client, but she felt that they did not actually want her there. When asked about whether spaces like this also exist in Maputo, she compared it to an experience at the Maputo Shopping Centre.

I remember that once at the Maputo Shopping, I was entering the mall with friends. Some *monhés* came and blocked our passage because they wanted to enter first. This created confusion, we almost fought. We felt as if they wanted to enter first because they are *monhés* andwe are black. But this doesn't exist here. If they are not happy, they should return to India …Here you have to learn to live togetherwith blacks because it's a country of blacks. But it does not exist a lot here, I don't feel it of ten (Samira, about 20-year-old resident of Casas Brancas, August 2012).

In Samira's narrative, Maputo Shopping becomes constructed as a place where racial hierarchies from colonialism can, especially on Sunday afternoon, return to the city. The sociality at the mall provides glimpses into the problematic past, a past which instances of encounters can bring back to reality in this heterotopic space. Like a history book, the mall as heterotopia makes her aware of the urban past, it renders the colonial heritage of the city more visible to her, while also increasing her awareness that these times are over, that she is now also a full citizen, and that she can claim her right to be in this space by literally entering the door first.

### *Heterotopic Relation between Malls and Domestic Work*

Approaching malls as heterotopias unravels how malls are entangled with other spheres of urban life, one of them being the labour entanglements existing in the realm of domestic work. When Olívia and I were strolling around the games arcade, Olívia drew my attention to an Indian-Muslim family, consisting of a woman in *hijab* and head scarf, her husband and a couple of their children, who were accompanied by a woman in a uniform with an apron. The woman in the apron was a domestic worker looking after the youngest child, a toddler. Olívia became angry at this sight and explained to me that it was despicable of this employer to make her domestic worker wear a uniform at the mall. She exclaimed with moral outrage: "They want her to use these clothes in order to mark difference, in order to show the difference to the public" (Oívia, August 2012). I asked how she believed the domestic worker felt in this situation. The domestic worker must be unhappy, she explained, as aprons are only supposed to be worn inside the house and were inappropriate for the mall. Olívia felt

negatively affected by the situation and disapproved strongly of the behaviour of the *patrão* (employer). For Olívia, denying the domestic worker the right to present herself as an urban citizen with decent clothes in this space of public life was an act of discrimination by the Indian employer. Later on the same day, on our way to the supermarket, we walked past another Indian family accompanied by a domestic worker. Olívia exclaimed: "Look, again, they force the domestic worker to wear work clothes in public to mark difference" (Olívia, August 2012). Only then she explained to me that she had looked for a job as a domestic worker recently herself. She was hence not only feeling pity for the domestic worker, but she imagined herself in her role, hence seeing herself where she was not (Foucault 1986 [1967]: 24).

At Maputo Shopping, the mall as a mirror makes urban dwellers more aware of the reality of urban society, its hierarchies and their social position in it. For Olívia, the mall makes visible the way employers treat their domestic workers, that they oblige them to wear uniform to mark social difference in public. This visibility of relations, which are usually lived in the intimate spaces of the home, renders the urban elites subject to critique and they become objects of moral judgement. The optical illusion of the mirror makes Olívia see herself in the role of the domestic worker, exposing her to the harsh reality which seems even harsher in the public of the mall. Maputo Shopping hence becomes a *heterotopia of illusion*, which exposes otherwise hidden aspects of urban reality, even exaggerates them, and also exposes "every real space … as still more illusory" (ibid: 27). The mall makes visible what is usually not acknowledged, namely, that the elite's lifestyles are deeply entangled with the lifeworlds of poorer milieus in the city, and domestic workers are needed to make their lives work, even on a family outing. What is often seen as an urban reality, namely, that urban elites can live enclaved lives disconnected from others in the city, is exposed as an illusion at the mall.

### **Greenstone Shopping Centre**

The Greenstone Shopping Centre is a regional mall situated on a hill in north-eastern Johannesburg, next to Edenvale and Modderfontein. Until a few years ago, this land formed part of a belt of green veld which became transformed into the spatial expressions of privately driven city building. The new area referred to as Greenstone is made up of secured business parks, warehouse complexes, shopping malls, townhouse complexes and gated residential communities with free-standing houses. The Greenstone Shopping Centre itself constitutes the centre, so to speak, of this new edge city officially called 'Greenstone Hill'. In contrast to the Maputo Shopping Centre, which is spatially integrated into the city centre, Greenstone Mall corresponds more to the type of suburban malls, as it is located in an area of urban sprawl, it is difficult to reach on foot, and it has inward-oriented architecture.

The inside of the mall consists of one long alley surrounded by shops and restaurants on two storeys. The mall is divided into two parts by a large main square with restaurants, movie theatre, and a large glass front leading to an artificial green area outside. This area has an artificial grass f loor where users can sit down, with playground equipment for children and a fountain. One can see the blue sky if looking upwards, but the square is surrounded by multi-storey car parks in grey cement. With its 85,000 square metres Greenstone Mall is a third larger than Maputo Shopping. It

has the facilities typical of a South African regional mall, with national retailers (Game, Edgars etc.), of which some seek to attract better-paid suburbanites (Pick n Pay, Woolworths), and some cheaper shops like Mr Price and PQ Clothing cater for the youth and people in search of lower prices. There are also a few boutique stores which do not belong to a chain. They moved here from the town centre of Edenvale which was the commercial centre for the larger area till the mall opened. The mall has leisure facilities like a movie theatre, a couple of restaurants, an indoor playground for children and even a licensed gambling venue. The shops and facilities cater largely for families with mid-range incomes; it is neither as fancy as the famous Sandton City mall nor is it oriented towards low-budget customers like some malls in the townships.

From the hill one has a nice view onto the high-rise buildings of the lower-lying Johannesburg CBD and the surrounding suburbs. John, a design architect who was engaged in many mall projects in Johannesburg and elsewhere explained to me that the builders probably did not take advantage of this view onto the cityscape because the inward-oriented architecture draws mallgoers attention to the appearance of and sociability with other mallgoers, which is more enjoyable than looking at the cityscape: "I believe that people enjoy seeing people. They interact with people in the way they pass each other, in the way they see each other dressed. There is a far stronger attraction to that than looking at a nice view" (John, October 2012).

Unlike the Maputo Shopping Centre where the boundaries between inside and outside are gradual and where there is a certain tolerance by the security personnel of informal traders at the entrance, the Greenstone Shopping Centre is tightly controlled and there is no room for economic activities apart from the formal. In contrast to the Maputo Shopping Centre where Guebuza Square is used by many pedestrians as a shortcut through the city, Greenstone Mall is not located on any existing pedestrian routes and is therefore only entered by people who specifically want to come to the mall. In contrast to Maputo Shopping where one can observe occasional beggars in Guebuza Square and poorly dressed children running up and down the escalators, at Greenstone Mall, the presence of visibly destitute people is exceptionally rare. The entry to the parking area is guarded by security booms and often security personnel stand at the boom and control the cars. The boom signals clearly that one is entering a space controlled by private ownership and private security companies. Heterotopias are contingent on such mechanisms of opening and closing, not least because they distinguish them clearly as different from spaces of ordinary life (Foucault 1986 [1967]: 26).

Greenstone Shopping Centre is owned by the Sasol Pension Fund, Sasol being a large South African company in the mining/energy sector. Sasol hence owns several malls in the country, not least because "large regional shopping centres perform better than any other segment of the property market in South Africa" (John, design architect, October 2012). During apartheid there were, however, also political reason for investing in malls: shopping malls in South Africa became a target for real estate investors because sanctions prohibited them from investing abroad (Marais 2001: 123). Sasol, nowadays a transnational company, was founded as a public corporation tightly linked to the apartheid modernist project and was privatised in 1979. The company has grown to what it is today as a result of subsidies and protection by the apartheid state and the cheap cost of African labour during apartheid. The profits generated at the mall will be used to pay the pensions of Sasol's workforce, a workforce which is still made up mostly of male, often Afrikaner, employees. Many mallgoers do not know

#### 248 Cities of Entanglements

who or what the owner of the mall is, and hence the history and politics behind Sasol's surplus capital and the complex ways in which the money spent by the consumer in the mall f lows back to the pension pay-outs of Sasol employees remain uncommented on in the city, a reality hidden behind the shiny façade. This stands in contrast to Maputo Shopping where the economic powers behind the mall are constantly commented on by urban dwellers.

### Abundance, Pollution and Exclusion

In the representations of Greenstone Mall, at least three different discourses are involved: there is (1) the narrative created by the marketing branch of the management in the promotional material which represents the mall as a space full of abundance, even nature and social relations; (2) there is a counter-discourse by environmental activists which blame the property developers for destroying nature, and (3) there is the discourse employed by the architects and planners of malls which portrays them as totally planned and calculated institutions with invisible walls which keep the poor away.

"Simply sensational shopping", was the slogan of the mall management on the Greenstone Mall website in 2013, and they promised that the mall offers "something for everyone" (Greenstone Shopping Centre 2013).8 Here you can buy "everything you need" for school at the stationery retailer CNA and customers can "eat as much as you like" for R75 at the Breakfast Buffet at the Cape Town Fish Market restaurant. In these website texts, marked by the extensive use of adjectives and superlatives (like 'trendy' fashions, 'exquisite' gifts, 'fabulous' home furnishing, 'fantastic' food outlets), the mall management portrayed the mall as a space of endlessly available goods. Yet the website promised more than mere consumption: it claimed that the mall is "a delightful oasis to those seeking a respite from it all", portraying Greenstone Mall as a secure oasis of abundance, to which urban dwellers can retreat from busy, stressful and chaotic city life. The introductory text on the website in 2013 announced that the mall was opened in 2007 "much to the delight of communities from Edenvale and surrounding areas". The reference to 'communities' indicates that the management want the mall to be seen not as an anonymous commodified space, but as a 'warm' and 'welcoming' place where the local 'communities' meet. With that they emphasise the rootedness of the mall, the connection to the locality and meaningful social relations. The logo of the mall consists of the name 'Greenstone' against a backdrop of two lines which allude to hills, probably representing Greenstone Hill on which the mall is located.

In the book *The Urban Revolution,* Henri Lefebvre pointed out that nature becomes one of the key problems in an urbanising world. What has once been abundant – space, water, air – becomes scarce, yet in advertisement, nature becomes fetishized. While nature is actually shrinking in the city, signs of nature and the natural are "multiplying, replacing and supplanting real 'nature'. These signs are mass-produced and sold. A tree, a f lower, a branch, a scent, or a word can become signs of absence: of an illusory and fictive presence" (Lefebvre 2003 [1970]: 27).

The fetishization of 'nature' in the naming of the mall as 'Greenstone' as well as its logo becomes particularly clear when juxtaposing it to the second representation

<sup>8</sup>  The following analysis is based on texts on the website as it was in April 2013 (Greenstone Shopping Centre 2013) aswell as promotional pamphletscollected during fieldwork in 2012.

of the mall upheld by environmental activists. Peter (ca. 65), is a resident of nearby Modderfontein whose parents moved from England to South Africa many decades ago in order to work for the dynamite company AECI. The company AECI was given about 4000 hectares of land by the Kruger government, land which then served as a safety buffer zone between the dynamite factory and the residential areas of Johannesburg. AECI constructed a workers' town, Modderfontein, for the European artisans (for more details see Behrens 2005). What used to be workers' houses separated according to their country of origin ('ethnic villages') has been transformed in the last few decades into aff luent gated communities and retirement villages where many of the former AECI employees live. In the early 2000s, AECI's property arm Heartland Properties, began to transform the former buffer zone, which was no longer needed because technologies had changed, into highly profitable urban land by pushing for its rezoning, putting in urban infrastructure like sewage in place, and selling it. The first large piece of land, 320 hectares, is where Greenstone Shopping Centre, as well as many strip malls and gated communities, were built.

The rapid transformation of green veld into the Greenstone area was not welcomed by some long-term residents of Modderfontein like Peter, not least because environmental pollution has become a major problem. The Modderfontein area contains a system of dams and a nature conservation area (Modderfontein Nature Reserve) with wetlands and a lake zoned as open land. These are areas which are not supposed to be built on. Since the Greenstone edge city has been built, the dams have suffered from severe pollution. Political activists, residents and politicians blame Heartland for this; apparently the bulk sewage infrastructure put in by Heartland before selling the Greenstone land for development was not large enough to carry the high density living.

Peter hence founded a small organisation lobbying to protect Modderfontein's nature. He remembers nostalgically his childhood days spent playing on the dams:

Before allthis happened,whenwewere kids,we played and learntto swim here,safely, and we fished and it was also a place where frogs and much more lived (Newsletter by Peter,November 2013,sent through Linbro Park Community Google Group).

Peter blames AECI and Heartland for selfish, profit-oriented practices which led to the pollution of rivers and dams, now "filled with sewerage, mountains of ash, trash, paint, oil and millions of golf balls" (Newsletter by Peter, November 2013, sent through Linbro Park Community Google Group). In 2013, AECI sold the second large piece of land, 1600 hectares, to the Chinese developer Shanghai Zendai for 1.06 billion rand. The Chinese company was planning to build a new Sandton-like city (Scott 2015, Ballard et al. 2017). The transaction was one of South Africa's largest single foreign direct investments. The sale was received very critically be the local activists as they fear that the Chinese will not be interested in ecological rehabilitation of the waterways.

Greenstone Mall and its representations as a place of abundance situated on Greenstone Hill upheld in its advertisements, or as a place where nature has been destroyed, a representation upheld by environmental activists, would appear to be an embodiment of Lefebvre's nature–urban relation. At the same time as nature becomes scare, which the environmental activists criticise, the "ideological naturalization becomes obsessive" (Lefebvre 2003 [1970]: 27). In contrast to Maputo where the counter-discourse that Maputo Shopping is a place of money laundering is very popular and even

inf luences shopping practices, in Johannesburg the counter-discourse of environmental destruction is little known beyond residents of Modderfontein who read the local newspapers.

In academia, malls are often portrayed as 'enclaves' and this also corresponds closely to the ways in which practitioners like John, a mall designer, represent malls, namely, as totally planned and calculated institutions with invisible walls which keep the poor away. John had been involved in many retail projects over his career. When developing a mall, he explained to me, "you want to create a machine which is going to draw as much money out of the surrounding population as possible" (John, design architect, October 2012). Shopping mall experts like him have developed a whole science of shopping including calculating walking distances, choosing certain colours and materials and avoiding others, choosing the lighting and the arrangement of shops and goods in such a way as to create an environment that seduces the urban dweller to consume maximally and to feel at ease (Chebat and Morrin 2007, Helten 2007: 247).

While urban dwellers appropriate spaces and imbue them with meanings, mall designers regard malls in abstract numbers, such as the internal rate of return (IRR) or foot traffic statistics. Despite this appraisal of the mall in abstract terms, John, like other mall planners, is also highly aware of the social relevance of malls. "It becomes a social heart of the city, for those who can afford" (John, design architect, October 2012). Yet he emphasised that the first motive for mall designers is to create profits, and the social characteristics are only of secondary importance. Mall investors and mall designers see urban dwellers primarily as consumers with spending power.

In the 1980s many malls had 'themes', yet this architecture of fantasy and escapism became widely criticised by architects in South Africa and elsewhere and was eventually replaced by what John calls 'international modernism', which is almost "Zen-like where less is more", a functional architecture with f lat and monochromatic materials, usually working with rectilinear forms avoiding curves (John, design architect, October 2012).9 This architecture of disconnection from cultural references makes sense in the post-apartheid city, according to John:

It has become this non-architecture, an architecture that is a statement of disconnection from the realities of people's traditional heritage. Which is just as bad and disconnected asthe theme architecture. The reason for that is, I think, what they call the rainbow nation. There are many dif ferent ethnic groups that make up South Africa. There is the fear that by selecting one ethnic group you are going to drive away others (John, design architect, October 2012).

<sup>9</sup>  Some of the financially most successful malls in the country, according to John, are themed malls, as their interior design does not demand adjustment to changing fashions. Other shopping malls tend to have a very short lifespan, as ways of lighting, types of shop front and fashion in design materials change quickly, and hence af ter a mere seven to ten years the finish of the mall is outdated. Investors either have to re-invest in order to keep up with newer malls or the mall will enter a downward spiral inwhich the high-end stores andwealthy customerswill leave and value storeslike factory outletswill enter (John, design architect, October 2012). All over Johannesburg there are indeed half-empty malls which have losttheir appeal and struggle to survive.Designed ascommercialspaces, appropriation for other usesis difficult, apart from fitnesscentres and outletstores(Parlette and Cowen 2011).

In contrast to apartheid where many shopping centres were reserved for white people, nowadays many different 'ethnic groups' interact at the mall. In terms of class, however, mall developers do not want to be as inclusive, but mall planners often purposely design barriers for poor people:

That is a wonderful transition that has happened. The only barrier in the design of shopping centres is the barrier to poor people. You try to keep them out in many respects. And that's done in a subtle way. If you ask any of the developers, they will not be as honest asI. They willsay: "No, the shopping centres are completely open to all levels of society." Whilstthatistrue in the ef fectthatthere is no physical barrier,the centres are designed to suit those that have money (John, design architect, October 2012).

According to John, the upper classes feel uncomfortable and unsafe in the co-presence of the very poor "because they perhaps are not as refined in manners and their methodology of dealing with each other. They are loud, they are maybe aggressive, they may be dirty", as John explained. The lower classes will also have feelings of insecurity "because they do not have as much as those who have more." In order to keep a mall profitable, a 'line' against the very poor needs to be drawn. Poor people driving away the rich at malls is the "biggest social reason for failure in South Africa" (John, design architect, October 2012).

In township malls the expenditure per capita is lower and customers who arrive by public transport (minibus taxi) stay longer in a mall than people arriving by car. Therefore, the malls and hallways in township malls are designed to be wider in order to accommodate greater numbers of people. At a mall like Greenstone, John explained, designers avoid wide hallways, as large open spaces can appear empty and anonymous. Designers rather try to create pedestrian congestion, especially in front of shops, in order to force people to enter them. Details like the number of toilets, the standard of the fittings in the toilets and the walking distance to the toilets can be used to inf luence consumer behaviour, such as how much time customers spend at the mall. Pedestrians are not preferred costumers as they have very low incomes, but taxi ranks – for people who can at least afford to pay the minibus taxi – are often integrated into the parking areas of South African malls. The taxi ranks, however, need to be positioned wisely, John explained, as car owners get uncomfortable if crowds of people with big packets walk past their newly leased cars which could get scratched.

For the selection of stores the notion of 'comparative shopping' is important for mall designers and marketing specialists: the various clothing stores are often positioned relatively close to each other so that customers have more options. The choice of the main supermarket ('anchor store') of the right size is crucial, as different supermarket chains appealed to different customers. The desire to keep the poor out is also the reason why many shopping malls do not provide many benches or other free seating areas apart from restaurants where mallgoers are obliged to order something: "As soon as you provide too much of those facilities, you allow the poor to occupy your spaces, which drives away the rich" (John, design architect, October 2012). Representations of malls by their designers often do not correspond to the way users actually appropriate malls, however. Approaching malls as entangled spaces of heterotopia entails multiple logics and is constituted by irresolvable tensions and contradictions, like the one between the advertisement of the mall as 'Greenstone' and its destruction

of nature, and the contradiction between the fantasy of exclusion of the poor and the messier reality.

### Shopping Routines in Johannesburg

The notion of enclave emphasises the disconnection and closure of a space. Foucault's notion of heterotopia, however, draws our attention to their entanglements with other spaces, and to systems of opening and closing which not only isolate heterotopias, but also "make them penetrable" (Foucault 1986 [1967]: 26). Shopping malls only come into being as spaces with social relevance by virtue of their connections and exchanges between inside and outside. Malls, like Greenstone Mall, can only create profit on the condition that urban dwellers integrate them into their shopping and leisure circuits and prove willing to cross the boundaries between home and mall, between public space and the privately owned space of public life, and often also between social worlds. Any ethnography of malls hence needs to explore how urban dwellers integrate the mall into their everyday urban trajectories.

### *Suburban Mall Routines*

HERE'S one thing that no other city in the country can of fer like Johannesburg does: shopping … shopping for Africa. From the upmarket Sandton Shopping Centre and Nelson Mandela Square to the fun Oriental Plaza and not-to-be-missed Rosebank Flea Market, Joburg is a shopper's paradise (Official website of the City of Johannesburg 2013).

Johannesburg has an elaborate and diversified shopping landscape. There are large, regional malls like Greenstone, strip malls without an entertainment character, old, deserted malls which lost customers because newer, fancier malls were constructed close by (e.g. Rosettenville Junction); new malls which cater for specific lifestyles like the middle-class youth (The Zone@Rosebank) or the commercial design scene (Design Quarter). Shopping malls constitute the central spaces in the shopping and leisure circuits of aff luent urban dwellers like the property owners in the suburb of Linbro Park. It was not always like this. Before the construction of the Greenstone edge city, the aff luent Linbro Park residents used to shop in Edenvale, a separate town on the East Rand belonging to the neighbouring municipality Ekurhuleni which has become drawn into the polynuclear conurbation of the Johannesburg city region.

Nowthewhole area hasjust filled [with newsuburbs and malls]. I have been around 20 yearsin Edenvale, Ican see the changes, I am becoming one of those 'remember when' (laughs) (Laura, 45-year-old resident of Edenvale,working in Linbro Park, May 2012).

With the construction of the new neighbourhood Greenstone Hill, with its malls and townhouse complexes, shopping circuits shifted away from Edenvale's town centre. Only a few local businesses, often family-run stores owned by ethnic minorities like Indians, Greeks, Portuguese, Italians or Jews, were able to survive (see also Hyslop 2005: 8).

Till the 1970s, suburbs mostly had a residential function and suburbanites commuted by car to the Johannesburg city centre for work, leisure and shopping. With the

edge city developments, these spatial practices started to change. For the majority of middle-class shoppers from the northern suburbs the CBD is no longer of importance, as on their way to the city centre they pass several malls with the same items found in the centre of any first-world city (Beavon 2000: 3). Also, members of the Linbro Park property owning milieu seldom venture into the CBD, and some even get lost on the streets of the CBD if they have to pick someone up from the long-distance bus station or show visiting family members around.

In order to grasp cityscapes and the way urban dwellers live in and constitute cities ethnographically, there is a need to replace conceptions of space as absolute, bounded territories with relational perspectives (Harvey 2006: 121, 133). In absolute terms, the mall is located at a distance of about five kilometres from the suburb Linbro Park, which takes about ten minutes by car. In relative terms, Greenstone Mall is 'just around the corner' for property owners in Linbro Park, in terms of how they describe it. Many of them visit the mall daily or at least a couple of times a week in order to buy groceries, go to the bank, visit the gym or eat out. For the property owners and many other milieus in Johannesburg, shopping malls have become the preferred site for leisure and consumption activities (see also Van Eeden 2006: 39). Today, urbanites even venture to shopping malls to visit movie theatres, bars, nightclubs and children's play parks.

Johannesburg's post-Fordist, postmodern shopping-scape depends on automobility (Czeglédy 2004). Since crime increased in Linbro Park in the 1990s, residents no longer even walk within the suburb, but drive to visit neighbours or attend community meetings. The drive from Linbro Park to Greenstone takes ten minutes if there is no traffic; in Johannesburg, where suburbanites are used to commuting long distances, these five kilometres constitute a short drive. Some of the tenants living in Linbro Park take much longer drives of up to 35 kilometres to areas to the east, like to Benoni and Boksburg, where groceries are cheaper. "The closer you get to Johannesburg's central areas, the more expensive the shops get" (Liz, 55-year-old resident of Linbro Park, June 2012).

Anna, tenant in Linbro Park, used to live with her brother who was a property owner and worked at a bank in Greenstone Mall. She usually took the backroads to the mall, as the highway was often packed with traffic. Every day she drove through the same entrance and parked her car in the same section which she chose carefully because of its proximity to her workplace as well as the supermarket (Anna, April 2012). In order to find her car again in the large, concrete jungle of the car park, she remembered landmarks like the name of a bar written on the concrete wall, after which she knew she had to turn left to return to her car. Because the many parking areas and the several entrances all look alike, users of the mall depend on way-finding devices (Lynch 1960: 3), mainly the fronts of restaurants and stores, in order to avoid getting lost.

Property owners in Linbro Park explain that they go to Greenstone because it is practical, and not because they seek out a specific public. "It's convenient and quiet" Claudia, about 30 years old, living in nearby Edenvale and working in Linbro Park, explained to me (Claudia, March 2012). Rachel, resident of Linbro Park, liked Greenstone because "it is big, open, there is everything you want, it has banks, and it is in the area" (Rachel, May 2012). Sixty-five-year-old Sarah, also a property owner in Linbro Park, works as manager of a shop at a nearby mall. She describes Greenstone Mall as a 'working class' mall, by which she does not refer to blue-collar workers, but to

the work-related rhythm of quietness and business which characterise the mall. The mall becomes busy at lunch time when employees working nearby meet friends or colleagues at the mall, it gets busy after work in the evenings when employees come to buy milk and bread, and on the weekends when families and couples come here for leisure and their weekly shopping. Unlike Sandton City and other high-end malls where the elite who do not depend on office hours mingle on weekdays, she pointed out to me, Greenstone Mall is quiet during the week.

Elderly people from surrounding suburbs frequent the mall. Nearby old age homes organise transport for their pensioners who spend some hours in the morning at the mall, shopping, strolling around and sitting at restaurants, till the transport fetches them again. Laura, a 45-year-old resident of Edenvale working in Linbro Park, explained to me: "My mother stays at a retirement home in the area. Monday, they go to Greenstone, Wednesday, they go somewhere else, Friday I think they go to Riebeek mall" (Laura, May 2012). In the mornings, mothers come with their toddlers to the mall to meet their friends after they have dropped the older children off at school. In the afternoons, teenagers come on romantic dates, drinking milkshakes together until they are fetched by their parents. For the children of Linbro Park property owners who are now in their twenties, however, Greenstone Mall does not form part of their leisure circuits, as they associate it with 'family' and prefer to go to the movies, restaurants and bars in more fashionable malls in Bedfordview, Sandton or Rosebank. Unlike the Maputo Shopping Centre where users tend to socialise at the mall in larger groups of up to ten people, the frequency of Greenstone is more individualised. Customers walk around alone, in groups of two or as nuclear families, ref lecting the more individualised social structure of the city. The walking speed at Greenstone Mall generally tends to be higher than at Maputo Shopping. During the week visitors tend to walk at a fast pace with a clear aim. The Mozambican strolling around (*passear*) at a slow pace for the sake of enjoying the public and the architecture is rather rare. On the weekends many Greenstone users' walking speed is more relaxed, but their attention tends to be on the shops and the goods on display and not so much on other mallgoers. The shopping mall is hence integrated in the diverse everyday practices and spatial routines of urbanites living in surrounding suburbs like Linbro Park. Malls hence emerge as deeply contextual and relational spaces, constituted by the way Johannesburg residents use the mall within the rhythms of their everyday life.

### *Township Shopping Circuits*

Greenstone Shopping Centre are just one space among many that urban dwellers weave into their shopping circuits. The shopping circuits Alexandra residents engage in in their everyday life look quite different from the shopping routines of the Linbro Park suburbanites. The retail facilities as well as physical access to them still ref lect apartheid inequalities, yet there is also considerable change. During apartheid, shopping malls were predominantly frequented by white milieus. The Reservation of Separate Amenities Act (1953) gave the local state as well as private enterprises the right to reserve their publicly accessible facilities for specific racial groups. Petty apartheid separated everyday lives and spatial practices on a micro-geographical level (Houssay-Holzschuch and Teppo 2009: 355). Poverty and lack of transport also restricted the possibilities of participating in the mall landscape, which started to emerge from the 1970s on. Township dwellers' shopping circuits continued to take them to the CBD

and the informal shops in the township (Tomlinson and Larsen 2003: 48). Larger retail developments were forbidden in townships during apartheid and up to 1994 townships only had small, informal retail businesses (*spaza* shops) (Ligthelm 2012: 37, Tomlinson and Larsen 2003: 49). Still today township dwellers buy small quantities of the groceries they require daily like milk and bread at these shops. The Pan Africa commercial area, based in the industrial area of Wynberg at the entrance to Alexandra, has served Alexandra as a retail and transport hub for decades. The small supermarkets and butcheries have the reputation of being very cheap, so that domestic workers living in the nearby suburbs like Linbro Park come to shop here. Since the 1990s, the spending power of township residents has increased significantly. National retailers have thus become interested in expanding into these untapped markets, not least because suburban areas have become saturated (Ligthelm 2012: 37). Like many other townships in the last few years Alexandra has also received its own shopping malls. In 2009, Pan Africa Shopping Mall was opened, in relation to the urban renewal initiative, the Alexandra Renewal Project (ARP). The Pan Africa Mall is seen by many as a sign of the development of the township and a symbol for the growing inclusion into the aff luent consumption world of Johannesburg. In comparison to Greenstone Mall, however, many see Pan Africa Mall as quite ordinary. Bertha, resident of River Park, explained her comparative experience of the two places like this: "The first time I entered Greenstone, I was like 'wow'. And when I enter Pan Africa I am like "ok, this is simply a mall" (Bertha, June 2012).

The shopping landscape has become diversified for Alexandra's residents. Many Alexandra residents continue to go to the CBD where small, often immigrant-owned shops offer low prices. Many of the women interviewed compare prizes across the different supermarkets available in the Pan Africa area and buy goods where they are cheapest. For township shopping rhythms, time is an important factor. Township dwellers who get a monthly salary buy small items like bread at a nearby *spaza* shop, and then buy in bulk once they get their salary. Shopping rhythms by households with multiple income sources, though, depend on more complex rhythms. Bertha, 30 years old and living with her sister in River Park, used go shopping on the first, the tenth and the twentieth of the month. On the first day of the month, she receives child grants, on the tenth she receives the rent from their tenants (she and her siblings owns property in River Park), and on the twentieth her sister gets her salary. As many households can only afford to shop once they receive the grants, the rhythms of the malls they frequent are entangled with the rhythms of grant pay-outs.

Based on the interviews and observations I could distinguish at least four different types of mallgoer from Alexandra at Greenstone Mall. The first type are members of aspiring middle-class milieus who own a car and have adopted similar shopping practices to suburban milieus, except that they continue to live in the township, often for financial reasons. They drive to Greenstone or another mall almost daily to buy groceries, and malls have become a firm part of their leisure and shopping circuits. The second type of mallgoer from Alexandra is people who shop for 'specials', so people who only go to the mall because of a specific sale. These members of poorer milieus hear about sales and specials on the TV, in newspapers or pamphlets, and go to buy in large quantities. The third type of visitor is the young township dwellers who go to Greenstone for what they call 'window shopping' (see below). with or without money in their pockets they go to the mall as a leisure activity, to look at goods or people, similar to

the Mozambican *passear* (strolling around). On Saturdays in particular I used to bump into acquaintances from Alexandra at Greenstone Mall who were there for 'window shopping', in the company of their romantic partners or friends. The fourth type of township mallgoer from Alexandra is the workers who are employed at the Greenstone Mall or at nearby malls and spend their lunch breaks there.

Greenstone is also about five kilometres away from Alexandra, yet for township dwellers, it is more complicated to get there than for suburbanites who own a car. Alexandra's convenient location in the midst of the northern suburbs puts it in the proximity of many malls in the surrounding areas, like Balfour Park, Norwood and Greenstone, but it is complicated to get there by public transport and relatively expensive. In order to save transport money, many River Park residents, the part of Alexandra situated closest to the N3, prefer to walk to Greenstone. There are no sidewalks, which means that pedestrians walk along dangerous roads. Apart from speeding cars, criminals are also major obstacles on the walk from River Park to Greenstone. Muggers waiting on the pedestrian bridge across the N3 highway are well aware of when people receive their social grants or salaries and mug workers and mallgoers on their way. Some workers, therefore, avoid the bridge and directly walk or run across the N3 highway, which is obviously very dangerous. In addition, trails have emerged along the highway where pedestrians often walk. The concrete jungle, where usually only cars speed by, has become appropriated by urban dwellers, making the city the product of many builders (Lynch 1960: 1). On the one hand, this shows that Greenstone Mall is indeed characterised by many barriers, limiting its accessibility for people with no car and little income. On the other hand, it shows the importance of acknowledging that despite these barriers, at least some township dwellers regularly venture to Greenstone Mall, weaving it into their shopping and leisure circuits. Because of urbanites' agency, exemplified by the trails along the highway, a mall sitting on a distant hill like Greenstone becomes entangled with township spaces.

### Heterotopia of Compensation

Shopping malls as heterotopias have, in relation to the rest of the spaces, a function (Foucault 1986 [1967]: 27). Maputo Shopping Centre has the function of illusion: for its users, the mall dismantles reality as illusory, it renders visible things which are otherwise invisible, and it sparks imagination. The Greenstone Shopping Centre, in contrast, has the function of compensation. As an entangled heterotopia, the mall compensates for what urban dwellers find is lacking in other urban spaces. Such heterotopias of compensation create "a space that is other, another real space, as perfect, as meticulous, as well arranged as ours [the rest of the spaces] is messy, ill constructed and jumbled" (ibid: 27, 21). Heterotopias as counter-spaces are real places "situated outside all other spaces destined to efface, to neutralize, to compensate or purify the spaces they oppose" (Boyer 2009: 54). Greenstone Shopping Centre is such a heterotopia of compensation in relation to the troubled, cramped shack living, the danger of chance encounters with strangers in public streets and the deep inequality felt in domestic work relations.

### *Safe Chance Encounters*

Many of the Linbro Park property owners have become regular customers at stores in Greenstone Mall and the surrounding strip malls. Interactions between customers and shop employees are in general informal and friendly; after few visits they acknowledge that they know each other, and the relationship becomes more personal. Greeting rituals (Hello, how are you?") often transform into brief conversations about the weather and many soon ask for each other's names and remember them. When I accompanied Regula, a 30-year old Swiss tenant living in Linbro Park, to the gym on Greenstone Hill, I noticed that she greeted the gym employees by name and vice versa. Gabriel, the 23-year old son of a property-owning family in Linbro Park, works out every day at the fancy Virgin Active gym at Stoneridge (one of the strip malls at Greenstone) and often eats there at the Fruit & Veg City supermarket with his best friend.

The managers are all Portuguese and Greek boys, they know us. And they always put extra pasta for our bowls. They are the best(Gabriel, 23-year-old son of a property-owning family in Linbro Park, May 2012).

Becoming a regular customer at a bar, restaurant, shop or gym meant for them that urban anonymity became transformed into acquaintanceship. Although the supermarket chain Fruit & Veg City has multiple stores across the city of Johannesburg, the branch at Stoneridge takes on a different meaning for Gabriel because of his casual social relations to the managers. Many young people drive across town to a specific franchised branch of News Café, a bar popular among middle-class youth and young professionals. Although the food and drink served at News Café in Sandton are identical to the News Café branch in Bedfordview, customers may have become friends with the manager and other customers, social relations which give the otherwise anonymous franchised bars a distinct identity. Through the construction of social relations as well as memories attached to spaces, apparently identical franchised branches become lived spaces with distinct place identities. Such forms of appropriation are easily overlooked, as they do not leave traces nor are they visible to outsiders.

Linbro Park residents go to Greenstone mostly with the intention to shop and less for leisure, but nevertheless social relations with other customers become enacted at the mall in ephemeral situations of interaction. As explained in chapter 4, with the end of apartheid and conf licts about competing future visions for Linbro Park, the active, leisure-based associational life in the suburb has diminished since the 1980s. Victorian sports have ceased to be a key realm for the construction of neighbour relations and nowadays it is through the internet and in activities related to security governance that residents build up at least some neighbour sociability. The mall is one of the very few urban places where neighbours from Linbro Park can bump into each other by chance. Neighbours may have very little to do with each other in everyday life, but serendipitous encounters at the mall help to keep up the relations, and to remind each other of the neighbourliness, so that these relations can more easily be reactivated in case one needs help from a neighbour, for example because a horse has strayed or a burglary has taken place.

Laura was living with her family in Edenvale and working at the King's School which belongs to the charismatic church LRC in Linbro Park. For her, the mall has a bit of the character of a town centre where: "You often meet somebody that you know. From church, from school, your neighbour. You don't even get to see your neighbour usually, but when you see them it's in the mall (laughs) or driving past" (Laura, May 2012). So not only neighbour but also church relations are fostered through chance encounters at the mall. This social relevance of Greenstone Shopping Centre for the strengthening of religious ties is even mentioned in sermons at the LRC Church in Linbro Park. At one church service, a volunteer said:

Yesterday Iwentto Greenstone Shopping Centre,to go to a shop, justto buy something. But then I only came out two hours later. I bumped into people from the church, we wentto have cof fee. You can't go to the mallwithout meeting anybody from the church (volunteer,church service LRC Church, September 2012).

The church not only provides spiritual guidance but also creates a warm, sociable environment, which stretches beyond the church into other urban spaces like the mall and which is strengthened through chance encounters. Apart from the mall, the possibilities for chance encounters have become very limited in Johannesburg where suburbanites rarely walk on public streets, moving around largely in their cars. Chance encounters with strangers in public settings like streets are mainly seen as threatening by suburbanites: they suspect the stranger may be a criminal aiming to hijack their car or to grab their handbag. In relation to these other urban spaces, where chance encounters are dangerous and avoided, the tightly secured and orderly spaces of shopping malls with their franchised gyms and restaurants become heterotopias of compensation. Here suburbanites can enjoy the serendipity of urban public life, mingle safely among strangers, build friendly acquaintanceships without any responsibilities, and keep up the feeling of belonging to certain 'communities', like the church or the neighbourhood, without actively investing much time in them.

### *Escape from the Shack*

Surprisingly, this enjoying of safe encounters, compensating for experiences in other urban realms, also shows up in narratives by mallgoers from Alexandra, the big difference being, however, that Alexandra residents need to cross social and spatial boundaries to get to the mall. While the mall in Alexandra, the Pan Africa Mall, is in their view a space belonging to the township lifeworlds, Greenstone is experienced as a social space determined by different public orders, populated by other milieus than they ones they share everyday life with, and with a different sense of place. Bertha, 30-year-old resident of River Park, considers Greenstone to be tidier, quieter and less tiring than the bustling Pan Africa commercial area with its many informal traders on the street.

[Pan Africa] It'scheap and it's less, but it looks untidy. When you get out from the mall, you can't even move yourtrolley from there to there, because everyone is making noise, everyone is selling. When I come back, I have this huge headache (Bertha, resident of River Park, June 2012).

At Greenstone Mall, Bertha and others believe, customers are expected not to shout or to run. At the end of the month when workers receive their salaries the supermarkets at Pan Africa commercial area become very busy, a time which many customers experience as stressful, not least because they have to remain aware of pickpockets. There are cameras in shops, but they cannot capture who steals from whom in the pushing and shoving. Greenstone, on the other hand, Bertha explains, "is open", and hence you can "do your shopping peacefully" and "feel free" (Bertha, June 2012). Being 'free' here does not refer to the idea that one is able to purchase whatever one likes, but rather to the safe sense of place, to being able to let one's guard down. Because of the tight security at the mall, mallgoers are protected from encountering "people that you didn't even want to meet" (Letsatsi, resident of River Park, June 2012), like pickpockets. At Greenstone, you could lose your purse, and because of the mutual visibility in the open hallways, there is more social control, and someone would give it back to you, Bertha explained.

You can throw something, maybe a purse,someone willsay: "Sorry mam, you lost your things." You can learn and understand. A person can see a person. But in Alexander, someone would pick it up and take it. I would also do that (Bertha, resident of River Park, June 2012).

Nnana10, field assistant from River Park, was living a life in limbo, sometimes working part-time in a coffee shop or as a domestic worker in Linbro Park, sometimes living at 'home' in a rural area where life was cheaper and where the shack was larger than the one her mother rented in River Park. In 2012 she considered her life to be difficult and felt that she lacked perspectives for her future. The lack of private space in the small shack she shared with her family was often difficult for her, especially when she was angry or stressed. In order to avoid fighting at home, she sometimes escaped to the mall, where she could calm down. "And when I come back, I am better again" (Nnana, June 2012). For Nnana, being at the mall was like entering a dream world, which compensated for the world outside and enabled her to forget about the intricacies of everyday life. Window shopping was for her a form of escapism and a ritualistic cleansing experience which she even compared to going to church. Entering the mall without money could be 'painful' for Nnana, as she explained to me, but this would not hinder her from enjoying being there.

Nnana and many other township dwellers live in conditions of high density with a lack of private space. This situation can cause anxiety and stress, which many township dwellers experience as pathological (see chapter 2). Privacy and private space are constantly contested privileges in Alexandra. For these township women, escaping to Greenstone compensates for the continuous invasion of their personal space taking place in dense township living. As a heterotopia of compensation, Greenstone is a more orderly, quieter and safer space than the bustling commercial centre of Pan Africa, allowing township dwellers to let their guard down and feel 'free'. For these moments, the mall as heterotopia neutralises the hardship of everyday life, it provides an escape into an illusionary world where everything is inverted: density turns into openness, lack turns into abundance, sorrow turns into carefreeness, at least for a couple of moments. Also managing to look good and to receive compliments, albeit with simple clothes (Aceska and Heer 2019), can give a sense of victory and achievement,

<sup>10</sup>  InHeer (2017) I referred toNnana as Sbongile for reasons of anonymisation. Because she desired so, I use in this publication her actual name.

providing an instantaneous compensation for the harsh reality of urban inequality where you can only lose because you were born in the wrong neighbourhood, went to the wrong schools, and you have the wrong passport, or none at all.

### Domestic Workers: Glimpses of Freedom

Rita was a domestic worker living and working in Linbro Park in 2012, and she sometimes frequented Greenstone Mall to which she drove in her *skorokoro* (Sesotho word used to refer to dilapidated old cars). She used to buy her groceries at the cheaper supermarkets in Alexandra, in the Pan Africa commercial area, but sometimes she ventured to Greenstone just for leisure and window shopping. She would enjoy having a coffee at one of the restaurants and observing the other visitors who had more money than her, she told me. While mixing with these people from all races she would feel 'free', but also a bit isolated, especially as she was generally short of money. For Rita, this 'freedom' was not possible in Linbro Park. She and many other workers living in domestic quarters in Linbro Park experience a 'lack of freedom' in the suburb, not least because their landlords, who are at the same time their employers, impose house rules like forbidding visitors or having their children live with them (see chapter 3). Many also experience a 'lack of freedom of speech', by which domestic workers meant in interviews that in the asymmetric relation of dependence it is very hard to speak up to their employers. So for Rita, the freedom of mixing with aff luent suburbanites at the mall was linked to her experience of a 'lack of freedom' whilst living in a white neighbourhood, entangled in a dependent, paternalistic relationship with the property owners. Hence, for her and others, the mall offered alternative, new forms of encounters between township dwellers and suburbanites beyond what she was used to within the realm of domestic work. Here, she enjoyed moments of togetherness, of solidarity, even hints of feelings of equality with white women, compensating for the structural inequality shaping her interactions with her employers in Linbro Park.

Nnana's mother worked for a property-owning family in Linbro Park for decades, while she and her sister grew up with relatives at 'home' in a rural area. Sometimes they came to stay with her mother for a couple of weeks, and they played with the children of the property owners.

I started meeting white people when I used to visit my mam, when she used to live in Linbro Park. Iwouldfind other kidsthere,thewhite kids, andwe played together. That's howI have metwhitesfor the first time. But now, that I am older, Ican go to a mall, you know, and find white people. They are not so bad, like we have been told, you know (laughing) (Nnana, resident in River Park, June 2012).

At the time of the research, Nnana sometimes stayed for a couple of weeks in her mother's shack in River Park, and for the rest of the year she lived in a township in her rural 'home'. Her encounters with whites had become reduced to the workplace – she was sometimes employed as a day-worker for property owners and family businesses in Linbro Park – and to the Greenstone shopping mall. While her experiences at the workplace were often rather negative, as she found the work overly taxing, badly paid and she often felt poorly treated, her narratives about encounters at the mall were more positive. When I accompanied Nnana to the affordable clothing store PQ at Greenstone, she repeatedly drew my attention to white customers, as she found their pres-

ence quite remarkable. Other interviewees from River Park mentioned that they were positively surprised to see white people in the low-cost shops at the mall, because they shared the stereotype that all whites are wealthy. Looking for bargains side by side at PQ and knowing that 'the other' was also concerned about affordable prices, produced positive feelings, and what I call *shopping solidarity.* In situations of shopping solidarity, social boundaries become crossed and ephemeral feelings of equality and togetherness based on these similar shopping concerns, namely finding bargains, emerge.

When I enter a shop today in Alexandra mall, I find a woman, when I go to Greenstone mall, I find a mother doing the shopping. I am like "Madam, this is so expensive, how food costs too much." And then we talk about it (Bertha, resident of River Park, June 2012).

Bertha does not refer to a racial category in this narrative, but from the fact that she addresses the stranger with 'Madam', a term used in domestic work relations, it can be deduced that she has a woman from an aff luent suburban milieu in mind. Apart from sharing concerns about money, their similar roles and concerns as mothers may lead to shopping solidarity, and hence provoke an instantaneous sense of fellow feeling among female strangers who share concerns regarding the care of their dependants and households. The women would make use of those moments of standing side by side at a shelf checking prices to start a friendly conversation, commenting on the goods that the other was scrutinising, complaining together about the rising grocery prices or recommending that the other buy a certain piece of clothing. Sometimes men would start talking to Nnana in shops and ask for her opinion about a gift they were buying for their girlfriend or wife. These short interactions constituted an key part of the pleasurable experience at the mall: "You get to meet different people from different places, you know, you get to know different people, how they operate" (Nnana, June 2012). She would first compliment the person with a simple 'hello' and from the tone of voice in the response, whether it was friendly or disinterested, she would decide whether to start a conversation or to walk away. Rejected encounters often became recounted in negative terms in the interviews, interpreted as a re-enactment of racial boundaries. Basani, a 26-year-old woman living with her boyfriend in the same yard as Nnana in River Park, remembers one negative experience with a white man at the mall. She was standing in the line to draw money from an ATM and the white man in front of her asked her to move backwards. She asked why and he responded: "You are too close, I need air." She then talked with her friend in Xitsonga and the white man got angry, complaining that they were insulting him. In the interview she speculated about his intentions, wondering whether he was thinking that she as a black woman wanted to steal his money and was therefore standing so close to him in the queue. But she also recounted positive experiences across racial boundaries at the mall, for example a white man who wanted her advice when he was buying clothes for his baby (Basani, resident of River Park, June 2012).

When I asked Bertha whether it made a difference if these mall users she would brief ly chat to at the mall were black or white, she responded that only a minority of them would still have 'empty knowledge' (a euphemism for racism).

Forsome people, they still have that thing of 'we arewhite and they are blacks'. But normally, people change. Some of them change because of the religion [LRC Church]; they have learned that we need to love one another. They have learned that, now, they are white, but they still need domestic workers, and they are black. So they need to learn to love, because those people, they are keeping their children, and those people, they are cooking for them,so they need to learn. And look at the garden boy, it's a black guy (Bertha, resident of River Park, June 2012).

In this quote Bertha links her experiences with white people at the mall to the discourse of the LRC charismatic church (see chapter 6), which she had frequented for a couple of years. She links the need for whites to become less racist to the everyday realities of interactions in the workplace and the often-unacknowledged dependency of white women on their domestic workers and gardeners for child rearing, cooking and security. Living apart yet shopping together at the mall thus enables these women to create sociabilities and ways of belonging which move beyond the old racial categories, even though these become easily reactivated in encounters experienced as negative. The mall becomes a stage where chance encounters across social boundaries can take place, chance encounters imbued with considerable symbolic meaning derived from their heterotopic relationships with other urban realms like domestic work, urban religion and street life.

### **Conclusion**

Whether the mall constitutes a normal place or not, a part of everyday routine or rather of rupture, varies across urban milieus and lifestyles and is, hence, less determined by the architecture of the space than by the place urban dwellers accord the mall in their everyday lives. For the residents of aff luent areas, both in Maputo and Johannesburg, the respective mall has a firm place in their daily routines, they go there regularly to shop and to stroll around; it constitutes a familiar ground where they feel comfortable and know how to act. For the residents of poorer areas, again in both cities, the malls hold an extraordinary place in their urban lives, the visits are special occasions, the urbanites need to cross social and spatial boundaries to get to the mall where they constitute a minority. Going to the mall is something which they have to learn; the mall may become familiar, more or less known to them, but not entirely normal.

At the beginning of my research I told the DA politician, Thomas, about my plans to do research on encounters between suburbanites and township dwellers at the Greenstone Shopping Mall, and he reacted with surprise, because, as he pointed out, people from Alexandra do not shop at Greenstone. When I talked to a domestic worker in Linbro Park about the politician's reaction, she responded: "They [people from Alex] do shop at Greenstone! Not all of them, some go to Pan, but they do!" She wanted to know whether the person who had told me the contrary was black or white. When I said that he was white, she said. "Ah, some white people think we can't go shopping where they go because they think we can't afford. But even with a little money we can go there and shop" (Agnes, domestic worker in Linbro Park and resident of River Park, April 2012).

I remembered this conversation when the ethnographic analysis of Maputo Shopping revealed the extent to which Indian mallgoers more or less politely ignore the

presence of African customers, while observing people from their own milieu. While interviewees from the poor neighbourhood of Polana Caniço in Maputo and Alexandra in Johannesburg told many stories about encounters with urban elites at the mall, in interviews with the urban elites few such stories came up. They paid little attention to mall users who did not belong to their own social world; for them it was rather encounters with members of their own milieus that created the sociability they talked about in interviews. When I asked Sarah, a resident of Linbro Park who managed a shop at Greenstone, whether residents from Alexandra frequented the mall, she responded: "I think they do, because (laughs) I know they have an awful lot of problems there with security" (Sarah, property owner in Linbro Park, May 2012). Referring to Alexandra's residents as security threats at the mall, she denied them recognition (Honneth 2003) as fellow consumers, making them socially invisible. Such social invisibility of marginalised mallgoers in the eyes of the dominating group has also been observed elsewhere. Van Leeuwen, in her ethnography of middle-class lifeworlds and malls in Jakarta, describes how the middle classes were 'blind' to the presence of beggars at the mall:

The ability to overlook people unless they complied visually with middle-class standards of dress and appearance seemed an innate quality of the peoplewho did comply with this dress code. I never saw, for instance, children take notice of beggars, street children orscavengers, except when forced to, that is, when approached whilst in a car. Itseemed asif they simply did not exist, asif they did not form part of the socialconfiguration of the moment (Leeuwen 2011:168).

Mall builders and mall architects like John portray malls as exclusionary spaces where the poor are made to feel uncomfortable on purpose, so that the wealthy feel safe and consume maximally. Mall designers believe that they can use architecture to attract or repel groups of customers and seduce them to consume. This conception of malls as abstract spaces, as containers that determine agency, is intentional. It is not an innocent representation but an ideology serving their capitalist strategy: they want potential investors to trust them that malls are perfectly designed money-making machines, and hence invest capital in their projects.

This ideology of malls as tightly controlled, secluded spaces of the wealthy has impacted on academics' views on shopping malls. Portraying malls as enclaves of the elites, disconnected from urban realities and hence like a foreign body in the organically grown urban tapestry, ref lects the ideology of mall builders, of how they want their malls to be seen by potential investors. The lived reality of malls, however, is much more entangled, multiple and contradictory. The representations of the malls by their marketing specialists are inf luential: so, the portrayal of Maputo Shopping as entry point to global modernity, where the city becomes connected to the synchronicity of global time, and the narrative that Greenstone Mall is a safe place for rejuvenation and abundance where customers can f lee the bustling city life. These hegemonic images of the malls, however, become contested by counter-representations of these spaces, upheld by mallgoers as critical citizens: in Maputo, there is the image that Maputo Shopping is a place of money laundering, where the hidden relations between the political and the economic become manifested in Guebuza Square. In Johannesburg, environmental activists criticise the destruction of nature and pollution the Greenstone Mall and surrounding gated communities have brought to this urban area. Malls are hence anything but uniform, uncontested urban spaces; rather there are multiple and competing ways of conceiving malls. Although the mall management and mall builders have the most resources and the power to spread their definitions of what malls are – through marketing activities and professional knowledge – they do not have total control over who uses their malls, how urbanites use them and how they conceive them.

Seeing malls as enclaves is a limiting, partial perspective on these urban spaces. Rather than emphasising disconnection and disembedding from the local context, the ethnography of the two malls drew attention to the way urban dwellers integrate the malls into their shopping circuits and everyday routines, in ways which ref lect and constitute their milieu membership. Urban dwellers embed the malls through their ways of appropriating and using them within the local context, giving them a local 'f lair', like the Mozambican way of rearranging tables and chairs constantly in line with the changing group composition of families and friends spending time together at Guebuza Square. Malls need to be understood as relational spaces, constituted by the spatial practices of their users, and receiving their meanings only from their relationship to other urban spaces. Understanding malls as heterotopias places these multiple relations to other urban spaces at the centre of the ethnographic analysis; malls as social spaces cannot be understood only in terms of themselves.

Maputo Shopping, on the one hand, can be seen as a *heterotopic mirror,* making the urban dwellers more aware of what urban reality looks like and of their place in society, yet at the same time it also inspires them to imagine a different life, not least because here the lives of urban elites, otherwise hidden behind walls, become visible to them. As a mirror, the mall renders the invisible visible, and makes reality intelligible, like the hidden relations between political and economic power, between the legal and the illegal. The mirror enables the mall users to look at the city from a mirrored angle and makes intelligible where the city is heading in the future: an urban development driven by the interests of capital in which gentrification will expel Nelinha and other mallgoers from Polana Caniço to peri-urban areas.

Greenstone Mall, on the other hand, emerges as a *heterotopia of compensation*. Suburbanites, and less often township dwellers, venture here to replace the danger of stranger interactions in public spaces with the safe sociability of shopping side by side in the orderly, quiet, spacious mall. The mall is a utopian world, yet also a real place, where township dwellers can pretend to be aff luent, where they can f lee to from their small shacks and enjoy the glitz and glamour, where interactions with white women evoke hopes and dreams about a transforming world. For suburbanites it compensates for the lack of urban joys in their other spaces of everyday life, like the tightly secured, walled homes, the lonely cars, the streets, where strangers are always a threat and where chance encounters are a problem. The mall promises safe mingling in anonymous crowds, with ephemeral encounters with acquaintances, neighbours and distantly known fellow churchgoers, which is a sociability distinct from the denser social ties and even social control enacted by the Indian elite at Maputo Shopping.

These distinct meanings of the two shopping malls under scrutiny – *heterotopic mirror* versus *heterotopia of compensation* – are also ref lected in the relationship between the mall and domestic work in the two cities. In Maputo, the mall turns the social relations between Indian elites and their domestic workers visible, and hence invites

critique and moral outrage from mallgoers at this essentially colonial relationship perpetuated in the postcolonial city. In Johannesburg, the mall offers an alternative, new form of interaction between township dwellers and suburbanites beyond the domestic work relationship. Here, domestic workers are freed from the asymmetric, paternalistic relations and can enjoy moments of togetherness, of solidarity and maybe even feelings of equality with white women, instances which compensate for the structural inequality shaping their domestic worker relationships outside the mall. In both cities, in both malls, the stock of knowledge and memories of encounters in domestic work relations in suburban homes distinctly shape the way in which encounters at the mall become interpreted and experienced. Because of such heterotopic relations between the mall and domestic work and between other diverse spaces of urban life, be they places of private or public life, our understanding of cities can only be enhanced if we focus on the entanglements of spaces and people.

©Barbara Heer, Maputo Shopping, 2011

©Barbara Heer, Greenstone, 2012

©Barbara Heer, N3 highway, 2011

©Barbara Heer, River Park, 2015

## **Closing Remarks**  Urban Entanglements as Blind Fields

Maputo, and even more to the point, Johannesburg, have long been and continue to be treated in the academic literature as *divided cities*, characterised by divisions resulting from a legacy of segregation inherited from colonialism and apartheid, as well as experiencing new forms of segregation invoked by neoliberal city building and the anxieties of urban elites around crime. That these two forms of segregation exist and shape Maputo and Johannesburg is uncontested, and the ethnography thus presented reaffirms this. The chapters have even focused on extreme cases of spatial segregation, namely, an urban area divided into an aff luent suburb and a poor township in Johannesburg, and an urban area divided into a wealthy elite neighbourhood and a poor *bairro* in Maputo. What the book fundamentally questions, however, is the idea that these spatial segregations are accompanied by absence or at least irrelevance of the social relations crossing these spatial boundaries, like the notions of *divided city*  and *cities of walls* seem to imply.

Although the urbanites lead different lives in these adjacent yet separate neighbourhoods, their lives are fundamentally entangled in many ways: through the politics of land in the changing urban areas, through the mutual dependency in the sphere of domestic work, through relations formed by praying together and charity in religious spaces, and through the new forms of sociality emerging in shopping malls. Obviously, there are many more ways in which these urbanites are connected, some involving direct interactions, others related to larger social processes like globalisation and digitalisation. It is impossible to "recognise, let alone to take up and respond to, all those threads by which any individual, or any place, is connected to the world" (Massey 2006: 93). *Cities of entanglements* hence emerges as an analytical perspective, rather than as a type of city.

The shift from *divided cities* to *cities of entanglements* also entails a shift in the theory of space, namely, moving from understanding space as a container to a relational and processual approach. The way how we think about space is "of fundamental importance" (ibid: 90). Spaces in Maputo and Johannesburg only come into being as real urban spaces when urban dwellers integrate them in their everyday routines, when they link them through their spatial practices that take them across the city. It is also impossible to grasp the power of spatial and social boundaries without following urban dwellers who try to cross them, without analysing what happens at the intersections. If we understand neighbourhoods, malls and churches not only in terms of themselves but also in relation to other spaces, we can grasp what makes them urban, namely, their meanings, functions and connections to each other and to the broader urban society.

*Cities of entanglement* understands space as always in the making, never finished and never *whole*, and, based on Lefebvre (1996 [1974]), it assumes that space becomes constituted through the interplay of three dimensions, namely *material, conceived* and *lived space*. In terms of *material spaces*, the possibilities for urban encounters are deeply shaped by the allocation of places across the urban geography, the transport possibilities urbanites have, walls and access restrictions, but also practical aspects of urban life, like household care, food, consumption and work. Space thus raises one of the key social, political and ethical questions, namely: How we are going to live together? It is through space that we are confronted with the existence of others (Massey 2006: 92). Because of the increasing scarcity of well-located urban land residents from adjacent neighbourhoods become entangled. By driving wedges into material space, urban elites aim to disentangle themselves. The politics of neighbourhood space is therefore a key realm where urbanites attempt to shape their city according to their own visions. Arranged around belonging and exclusion, spatial politics becomes a locus of power struggles between entangled lives. In the *politics of loss* in Johannesburg and the *politics of proximity* in Maputo, urban dwellers aim to inf luence these material conditions for living together and living apart by trying to keep the *others* out. This manifests in Johannesburg in the opposition to public housing by property owners, and in Maputo in the elite residents' attempts to erect a road closure. The ethnography, though, also shows that wealthy suburbanites fail to realise their fantasies of secluded lives in the fast-changing cities: cities which are drawing them more into multiplex relatedness than into accepting the defence of their walls.

In terms of *lived space*, urban entanglements have a fundamental impact on subjectivities, on the constitution of urban milieus, and on the way people see their own social position in urban society. Urban milieus do not precede the entanglements between them; it is rather through entanglements that urban milieus become constituted. The book shows that in these cities recovering from segregation and confronted with neoliberalism, public spaces are not the most central in organising urban life. It is in spaces of encounters – homes, places of prayer and malls – where entanglements across spatial divides and social distance become shaped, that urban society in the making can be observed. In suburban homes, patron–client relations between employers and domestic workers create a precarious balance between proximity and distance, exploitation and mutual support. In religious formations, urbanites deal with inequality by forming paternalistic ties, a form of entanglement which allows for positive feelings, if at the cost of hidden tensions. Paternalistic ties lie at the foundation of sociality, emerging in these spaces characterised by inequality. While religious spaces like charismatic churches or local mosques promise new, prophetic communities of equal believers, they are urban spaces shaped by the same structural inequalities as the rest of the city; hopes for encounters and change are met with disappointment. Shopping malls in Maputo and Johannesburg are not merely spaces of meaningless consumption and sites of exclusion and control; they are appropriated by urban dwellers in diverse, often subtle ways and transformed into spaces of urban public life where lives become entangled through chance encounters, competition and fantasies.

Space is always shaped by a contemporaneous multiplicity of processes (Massey 2006: 92), which reveals itself in the ethnography by the co-presence of multiple actors

with diverse and diverging interests, visions (hence *conceived space*) and strategies for the future of neighbourhoods. The interplay between these multiple actors is fundamentally shaped by power differentials between them, be they aff luent and poor urban dwellers, property developers, mall designers, politicians, activists, academics and the like. Looking back on colonial and apartheid urban planning, both Maputo and Johannesburg have become shaped by the desire of powerful actors to separate, to keep apart. Nevertheless, both societies were based on the exploitation of African labour by colonial enterprises and by colonial households, and hence always entailed fundamental entanglements and interdependencies between the colonial citizens and subjects (Mamdani 1996). Domestic work in Maputo and Johannesburg today is probably the urban sphere where there has been most continuity since the colonial past, because in these intimate interactions colonial patterns become relatively easily repeated. Despite the omnipresence of domestic employees working and, in the case of Johannesburg's suburbs, living in aff luent neighbourhoods, this urban reality is seldom recognised and remains invisible. Entanglements can hence be seen as what Lefebvre calls *blind fields*: we tend to see urban spaces incompletely, we tend to have a "a blind spot on the retina", we are blind and do not even know it (Lefebvre 2003 [1970]: 29). According to Lefebvre, blind fields emerge when one epoch (in the *Urban Revolution* he talks about the replacement of the industrial epoch by the urban epoch) becomes replaced with another, and one tries to understand the new epoch with the ways of seeing developed in the old epoch:

What does our blindnesslook like?We focus attentively on the newfield,the urban, but we see it with eyes, with concepts, that were shaped by the practices and theories of industrialization, with a fragmentary analytictool that was designed during the industrial period and is therefore *reductive* of the emerging reality. We no longer see that reality; we resist it, turn away from it,struggle against it, prevent its birth and development. The urban (urban space, urban landscape)remains unseen. *We* still don'tsee it. Is itsimply that our eye has been shaped (misshaped) by the earlierlandscape so itcan no longersee a newspace?(ibid: 29, italicsin original)

The metaphors and typologies used to speak about segregated cities, like divided city, suburb–township*, bairro–cidade,* originate from a different period; in Southern African cities the periods of colonialism and apartheid. Looking at these cities today through these terms creates *blind fields*, things we do not see because they lie outside our perspective, of what is epistemologically imaginable to us, and what the past epoch did not want to see. The fact that entanglements remain often unseen and invisible is not only about a lack of knowledge or education, not only a question of a *wrong lens* for looking at the urban. It is also about refusing to see, about pretending not to see (ibid: 31) and, hence, about power and ideology.

It is no coincidence that claims to recognise the fundamental connectedness of urban worlds have been raised by urban dwellers living in both Alexandra and Polana Caniço, urbanites who usually stand on the less powerful side of the entanglements. There was for example Thabo Mopasi, field assistant from Alexandra, pointing out how Sandton City mall was built by workers living in the township. Senhora Aurora from Casas Brancas in Polana Caniço raised the issue that the elites living in Sommerschield II who wanted to close off the road are those who come to ask for their votes during

election campaigns. There was the waiter Alfonso working at Café Sol in Sommerschield II, who asked rhetorically how he and others would get to work if there was a road closure. Especially blind towards entanglements are the more powerful urban dwellers. The aff luent majorities at the malls, be they the Indian elite in Maputo or the suburbanites in Greenstone, exclude the present poorer minorities from their social formation by never looking at them and not becoming aware that they are there, too. The property owners engage in manifold practices that render their domestic workers socially invisible, for example by prohibiting them from hosting family in their rooms or by closing down and controlling the shebeens, their spaces of leisure and public life. There was the property owner Mandy, who explained that she knew most people in Linbro Park but merely listed the names of other white property owners to illustrate this. The property owners' emic notion of 'divided community' referred to a conf lict between more business-oriented property owners and those who wanted to keep the area purely residential, and not, as one might think, to a division between property owners and tenants, or employers and domestic workers. Being blind to or not seeing entanglements is not a state but a process, which demands many social practices of invisibilisation. Urban elites like to conceive of their lifeworld as homogeneous and disconnected from the surrounding city; many of their practices aim at disentangling them from others or at least allowing them to believe that they are disentangled. Their segregationist practices are nevertheless confronted with resistance expressed in diverse forms, ranging from violent acts to subtle forms of appropriation. In Maputo, residents from nearby Polana Caniço resisted the road closure and rose up to take the newly constructed barrier down. Domestic workers gossip or take unauthorised breaks as hidden forms of resistance. At the mall, poor urban dwellers appropriate urban spaces, playfully participating in the world of consumption despite having little or no money, although mall planners like to sell malls to investors as enclaves designed for users to consume maximally. The change of perspective from *divided cities* or *cities of walls* to *cities of entanglement* is thus also a change from not only looking at the city from the perspective of urban elites who desire to withdraw themselves from connectedness to also looking at the city from the perspective of the urban poor for whom connectedness is an indisputable urban reality, and also a way of surviving socially and economically in a city marked by inequality.

(In)visibility and recognition or denial of entangledness has a fundamentally moral and ethical dimension. When explaining that the Sommerschield II elite come to ask for their votes, Senhora Aurora made a political claim; not only are the politicians accountable to them, but they also stand in a dependency relationship with each other, and they are hence responsible towards each other. In Linbro Park, when domestic workers refer to their employers in familial terms like 'aunty', they make claims that the employers have a responsibility towards them, which goes beyond the mere payment of a salary. In both cases, this invocation of responsibility by less powerful urbanites succeeded. The elite in Sommerschield II gave up their desire for a road closure. In Linbro Park, some property owners have sold up and moved away but still continue to support their former domestic workers, for example by employing them in the households of their children who remained in the city, or by building a house for their employee in her rural home. When urban elites, politicians or other powerful actors try to keep entanglements invisible, this is also deeply political. It serves to ignore the less powerful urbanites' claims to the right to the city or to the neigh-

bourhood. It allows the elite to live in aff luence in a poor city with little guilt, and it helps to suppress feelings of responsibility for others beyond gift-giving. Because this is what the recognition of entangledness is about: it is a recognition that we stand in relationships with others, that our actions have an impact on others, and that we therefore have a responsibility towards them. This entails a very different ethics from the ethics of neoliberalism, which claims that actors are only responsible for their own well-being and maximising their own profits, even at the expense of others and at the expense of nature. It also demands a different ethics to the one invoked by the Linbro Park property owner Kacy, who said that "people like to group together", claiming that the desire to be with like-minded people is a legitimate reason for segregation. Instead, what we need is an ethics and "politics of interrelation; a politics which, rather than claiming rights for a rapidly multiplying set of identities, concerns itself more with challenging, and taking responsibility for, the form of the relationships through which those identities are constructed, in which we are individually and collectively positioned and through which society more broadly is constituted" (Massey 2000: 246). Taking responsibility for others, even though I may not know them, may have never met them and may not have an immediate link to them, is the ethics we need in order to create an urban world where it is possible to live together in difference and where urban dwellers work together to reduce structural inequalities. How we deal with and shape our entangled positions in the contemporary world is the key ethical and political question of our times.

## **Postscript: Entangled Comparers** Experiencing Cities through Comparative Ethnography

In urban studies, comparisons are *en vogue*. In her seminal book *Ordinary Cities* (2006a), the geographer Jennifer Robinson prepared the ground for comparative urbanism by claiming that cities should be compared beyond the North–South divide, and that comparisons should be key tools for postcolonial urban theorising from the South. This should enable urban studies to move beyond being a discipline largely rooted in the Northern experience and deprovincialising urban theory (Huffschmid and Wildner 2013, Lemanski 2012, 2014, McFarlane 2010, Nijman 2007, Parnell and Robinson 2012, Robinson 2006b, 2013). In scholarly debates on South African cities and on Lusophone cities, repeated calls for comparisons have been raised. Urbanists working on South Africa have argued that comparing South African cities with others is crucial in order to move beyond framing them as exceptions and special cases (Mbembe and Nuttall 2008, Parnell 1997). As regards Mozambique, researchers observe a 'Lusophone exceptionalism' which has been criticised for inhibiting comparative gestures between Lusophone and other African countries (Pitcher 2002: 9‐10). Comparisons hence promise to disrupt boundaries of knowledge and research, to *think* Johannesburg and Maputo through elsewhere (Robinson 2016b) and to generate new understandings of the urban based on the diversity of cities and urban milieus. This book situates itself in this field of comparative urbanism.

Besides reclaiming space for comparisons, Robinson significantly advanced the debate on the politics of comparative theory building. By developing new typologies of comparisons based on expansive literature reviews, she pushed for new and experimental ways of doing comparisons and engaging with complex questions of epistemology around comparison (Robinson 2011, 2016a, 2016b). She proposes to

… reimagine comparisons as involving the broad practice of thinking cities/the urban through elsewhere (another case, a wider context, existing theoretical imaginations derived from other contexts, connections to other places), in order to better understand outcomes and to contribute to broader conceptualizations and conversations about (aspects of) the urban (Robinson 2016b: 3).

While the literature on comparative urbanism is constantly growing, making it literally a new turn in urban studies, the epistemological and methodological logistics of doing comparisons continue to be challenging and hamper the application of these new ideas (Lees 2012, Lemanski 2014: 2945). This is not least because discussions focus on the pol-

itics of theory building through comparison but also on comparative empirical research itself (Gough 2013), by which I mean the process of data collection and analysis through which urban comparisons come into being. This is a serious lacuna in the debate for at least two reasons: Firstly, focusing the debates on comparison in theory building and the abstract work of comparative arguments rather than on how comparative data become constituted, removes comparative urbanism from the urban dwellers' and researchers' lived experience of cities into the realm of philosophy. Secondly, the lack of concise, intelligible formulations of the way such new ways of doing comparisons may look in practical terms inhibits students from entering the field of comparative urbanism, in ways which go beyond solely making references to the debate as an expression of one's commitment to a more global urban study. This understanding of urban comparisons as a political and theoretical orientation, rather than as a way of undertaking empirical research, finds its culmination in Robinson's article with the telling title 'Thinking cities through elsewhere' (Robinson 2016b). Here, she would appear to claim that comparative urbanism does not necessarily need to be based on actual empirical comparative research but that, in her view, "perhaps the most useful comparative tactic in urban studies is the case study, brought into creative conversation with a wider literature" (ibid: 18). While writing their single case studies, comparative urbanists should read across contexts, improving their own analysis and theorising based on other people's written cases. I will call such an understanding of comparison *thinking cities through elsewhere,* as comparison here refers to intellectual endeavour, enacted while sitting at a desk. In this postscript I make a case for a much broader understanding of comparison, namely, as *experiencing cities through elsewhere.* Through comparative ethnography the researcher becomes involved, hence entangled, with at least two places and strives to develop an analytical framework and a form of description which speak about both cases. As this postscript will show, comparative ethnography hence entails not only *thinking* but also *experiencing* cities through elsewhere, as through comparative ethnography the ethnographer becomes deeply involved with the spaces and the people she researches. In comparative ethnography, the actor who conducts the comparison, the comparer, is not a detached analyser but an involved person, shaping and shaped by the experiences she has in diverse urban contexts. Instead of understanding comparison as a form of analysis conducted at home, maybe even sitting in an armchair, I approach comparative ethnography as a circular process in which the mind, and even the body of the ethnographer, is involved in constant comparisons along the way. In this postscript I make transparent to the reader the process of fieldwork and analysis on which this book was based. It focuses on the processes before the text was written up and introduces some of the intricacies of conducting fieldwork in two places as a single ethnographer. From the point of view of how knowledge is formed, I contend that there is a substantial difference between *thinking* and *experiencing* cities through elsewhere.

### **Plurality of Comparisons**

In order to approach ethnography comparatively, it is important to start thinking about comparisons as being always in the plural and not in the singular. Multiple comparisons take place in the field, in the data analysis and in the writing. The plurality also refers to the written-up text, as out of a single comparative research project diverse

forms of written-up comparisons can result. Thinking about comparisons in the plural, not in the singular, also makes us aware of the multiple ways in which comparison is understood, be it in different disciplines, different schools of thought, or even in everyday life. In conversations about malls in Maputo, urban dwellers repeatedly told me things like "You can't compare the Maputo Shopping Centre to Mandela Square in Sandton". What they meant by this was not that it is impossible to describe differences and similarities between the two malls but rather that the two are very different from each other, and that they regarded Mandela Square as superior in terms of architecture, the way goods are displayed, the range of brands sold and suchlike. In everyday use, to compare means to claim that two things are similar (Handler 2009: 627) and claiming that they are not 'comparable' also means that as shopping malls they are in a different league. This points to two things: first, that what people mean when they speak about comparison can differ greatly, and second, that there is a diversity of comparative practices even in everyday life and within research projects.

The table below presents an overview of some of the comparative practices as they may be encountered during an ethnographic research project. *Comparison as social practice* means that comparison is first and foremost a social and cognitive everyday practice, not an academic method. Humans always compare, whether intentionally or not (Strauss and Quinn 1997). According to phenomenology, actors possess a stock of knowledge from past experiences, which they relate (compare) to a current situation, and which thus shapes their actions (Emirbayer and Mische 1998: 971, Schütz and Luckmann 1973). *Comparison as data analysis* refers to the fact that academics compare constantly during data analysis, even in non-comparative research projects. The scientific development of concepts and codes is based on comparison (Kant 1969 [1980], Strauss and Corbin 1990). In the grounded theory approach, for example, the coding of data moves constantly between pieces of data and the searching for codes which grasp several pieces of data (ibid).


Table: Overview ofcomparative practices during an ethnographicresearch project

*Implicit or invisible comparisons* refers to the fact that ethnographers often make comparisons between field and home during the research process, and often also between reality and personal normative ideals. In such invisible comparisons, one's home culture is "the constant hidden references in relation to which the unknown culture can be described as different" (Caldeira 2000: 7). Because such invisible and implicit comparisons can have a great impact on our analysis, as well as on ethnographies based in one context, ref lection about such comparative practices is relevant for the whole discipline and not just for comparative ethnographers.

By *armchair comparison* I refer to the 19th century armchair anthropologists like Edward Tylor who, sitting in an armchair, constructed comparisons based on data collected by others (see below). Today, this refers to comparative practices where comparison is understood as distinct and separate from ethnography, as a form of analytical anthropology which brings together data or cases developed and written by others so as to construct comparative conclusions. *Life project comparisons* refers to anthropologists like Marilyn Strathern, who conduct comparisons based on data which they themselves have collected in different places over the course of their career. Armchair comparison and life project comparisons both happen *af ter* the fieldwork was conducted. This differs from *team comparisons*, larger research projects in which several ethnographers work on a similar topic in multiple places at the same time, in which comparisons emerge through mutual visits, workshops and co-authorship. In comparative team projects as well as in *single-researcher comparative ethnography*, in which one researcher conducts fieldwork in two or more places and writes about the data herself, the comparative perspective is present from the beginning when writing the proposal. In terms of the way comparative knowledge is formed these are relevant differences. In order to expand our grasp of the diversity of comparative practices, it is important to look at the history of anthropology, as the use of comparison as a method has changed considerably over time.

### Positivist Roots

Comparison in the humanities and the social sciences has historically been imported from the natural sciences and was therefore grounded in positivism (Schriewer 2003: 14). By positivism, I refer to the scientific paradigm (Kuhn 2012 [1963]) that emerged from the natural sciences, which assumes that there is an objective reality and context-independent data. Positivist approaches to comparison in anthropology go back to Tylor. His cultural evolutionism encompassed a theory about universal laws developed on the basis of the systematic comparison of cultural forms (Tylor 1889). At that time, comparison was seen as central to the expansion of knowledge that the new sciences and imperialism were thought to bring (Melas 2007: 20-22). With his comparisons Tylor aimed to show that there were many similarities among the different 'civilisations' and 'cultures' and that one therefore had to recognise the 'psychic unity of mankind' (Tylor 1889: 44). Analogous to comparative urbanism, which aims to undo hierarchies between cities of the North and South, Tylor also aimed to undo a hierarchy of thinking. With his comparisons he criticised the then prominent distinction between inferior and superior races, arguing that differences exist because of culture, not because of race. Each culture that he drew into his comparisons was at a different stage of cultural evolution, he theorised. Through comparison, he hence introduced a new hierarchy of thinking, namely, between 'primitive cultures' and 'civilisations'. Tylor's comparisons and the theory of evolutionism have received ample critique since then, among other reasons because he totally ignored the context of the data he analysed in his armchair, and because these temporalising comparisons deny the 'primitive cultures' coevalness (Fabian 1983, Melas 2007).

There have been many other comparative anthropologists, for example the diffusionist Friedrich Ratzel and the cultural morphologist Leo Frobenius (1933).1 It is important to note for the current ref lection on comparison that Frobenius' approach was to a certain degree inductive. During his twelve voyages to Africa (1904–1935), Frobenius studied material and immaterial cultural forms across Africa and distinguished them into *Kulturkreise* (culture areas) and *Stile* (styles) (Straube 1990, Streck 2001), which he then compared to each other. So the units of comparison, in this case the different *Kulturkreise* and *Stile*, did not exist at the beginning of his research but rather constituted the key results of the study. This differs from a hypothesis-driven comparison which defines units of comparison from the onset. However, rather than being interested in the specificities of African cultural forms, his aim was to contribute to the grand theories which were in vogue at that time.

Another milestone in the history of comparison in anthropology, which should be seen in as critical terms as the one already mentioned, are the Human Relations Area Files (HRAF) by Murdock and his colleagues at Yale. This compendium of world cultures was meant to provide data for cross-cultural comparisons for anyone who wanted to use it. Like Tylor, the HRAF was a negative example in the history of comparison, as there were massive problems with the empirical foundations of the data, not to speak of the decontextualising, mathematical techniques used to identify 'correlations' between cultures and cultural 'universals' (Moore 1993, Yengoyan 2006: 139).

The anthropological critiques of such comparisons are as old as the discipline itself. The cultural relativist, Franz Boas (1896), was one of the first to formulate key concerns about Tylor's comparative method. Boas made an important claim which is still relevant today, namely, that cultures have to be understood first in their own specificity, and only thereafter should comparisons be drawn (Boas 2004, Bohannan and Glazer 1988, Dürr, Kasten and Renner 1992). Although cultural relativists strongly criticised speculative comparisons in the style of Tylor, comparison remained important for the discipline (e.g. Benedict 1946 [1934]). Mead's famous ethnography of Samoa had comparative aspects, as she compared growing up in Samoa with the troubled teenage phase in American society in the 1920s (Mead 1928). Comparing the 'field' and 'home' in explicit terms can contribute to addressing social problems in the anthropologists' own societies, something which Mead and Benedict saw as anthropology's public responsibility (Fox and Gingrich 2002). This public responsibility was appealed to again in discussion regarding the crisis of representation (Marcus and Fischer 1986). Through comparison, anthropology can render the 'home' culture of the anthropologist visible as culturally specific and not as human nature (Handler 2009: 635). Com-

<sup>1</sup> Frobenius collected material and immaterial cultural forms across Africa and ordered them in space with the help of cartography. His *Kulturkreise* are a spatially and temporally ordered typology of culturalforms,which he explained through Ratzel'stheory of dif fusion. Linking the competing theories of dif fusion and evolutionism, Frobenius argued that every culture has an own *paideuma* characterised by an evolutionary process explaining the dif ferences, yet similarities come into being because of dif fusion and contact (Frobenius1933,Haller 2005: 41).

parison therefore has the potential to serve as a tool to expose assumptions and ideologies of the 'home' societies.

More careful, more contextualised and so-called 'controlled' approaches to comparison were developed between the 1940s and 1960s by British structural functionalists, as they moved from comparing 'cultures' to comparing 'societies' (Brettell 2009: 652, Fortes and Evans-Pritchard 1987 [1940], Yengoyan 2006: 140). One way of moving beyond the decontextualising HRAF style of comparison was to make 'regional' comparisons, as anthropologists believed that it was easier to compare cultures that were similar, and that by making comparisons on a geographically or culturally limited scale, they could control the number of 'variables' shaping differences and similarities (Eggan 1954, Holý 1987b: 3). The idea of 'controlling variables' is misleading, however, as complexity (and hence the number of 'variables') is related to the detailed nature of research rather than 'objective' similarities between the fields (see also Strathern 1992). In addition, the important factors that shape the topic of study will only be clear at the end of the comparative process and should not assumed at the outset. Last, but not least, what should also be mentioned is that the history of comparative methods also entailed approaches which paid a lot of attention to context, for example the scholars of the Manchester School under Gluckman from whom Robinson draws her inspiration for comparative urbanism, as well as anthropologists working on Melanesia like Sahlins, Strathern and Godelier. There has always been a heterogeneity of comparative methodologies (Fox and Gingrich 2002: 5, Strathern 2002: xiii).

### Crisis of Comparison

Since the 1960s, not least because of the anthropologist Clifford Geertz, anthropology has moved away from generalisation towards description and meanings (Geertz 1983, Yengoyan 2006: 141). From the 1970s onwards, key categories like culture became questioned (Abu-Lughod 1991, Lentz 2011, Lentz 2013a). With the shift to theories of practice and agency, culture became deconstructed as a problematic, imprecise category to explain social action (Ortner 2006, Yengoyan 2006: 143). As the units of analysis in cross-cultural analysis were 'cultures', this had severe consequences for comparative anthropology: what were anthropologists actually comparing, if not 'cultures'? The consequences of these shifts have not yet been sufficiently discussed.

In the crisis of representation comparison per se was rarely discussed. The debate tended to focus on ethnographic authority and the critique of the apolitical and ahistorical nature of anthropology (Clifford and Marcus 1986, Marcus and Fischer 1986). The anthropologist John Hutnyk (1990) was one of the few who linked the debate to systematic ref lection on comparison. Hutnyk (1990: 83) pointed out that the crisis of representation made anthropologists acutely aware of the complexities of the cultural realities and the subjective nature of interpretation, so that the comparison of two such interpretations came to be seen as troubling or even impossible. Although he also believed that comparison "thrives on simplicity" (ibid: 94), he did not call for comparison to be abandoned; not least because he argued that every ethnographic description is always comparative even if this is seldom acknowledged (ibid: 82).

Few people shared Hutnyk's claim to continue with comparisons, despite these new challenges. In sum, the crisis of representation made anthropologists turn away from comparison (Hannerz 2010: 547). Many believed that the move towards the use of local concepts inhibited comparison (Yengoyan 2006: 142-143). As the writing cul-

ture debate questioned cultural translation, suggesting that it was embedded in power relations and necessarily imperfect (Asad 1986), so comparison also came to be seen as a colonial, distorting act. Still today, many scholars see comparison as based on an 'imperial ideology' (Zanker and Newbery 2013: 110). A further important critique was that universalist, objectivist comparisons were related to the construction of grand theories, and had therefore to be abolished, together with these totalising theories. Ideas like causality were replaced by meaning, multivocality and relativism (Yengoyan 2006: 142). In retrospect, comparison became seen as a huge fault in rather than a great achievement of the discipline (Gingrich and Thelen 2012: 395).

There were anthropologists who continued to practise comparative methods, but they were those who did not engage with the crisis and largely remained within the positivist framework (e.g. Mace and Pagal 1994). Questions like 'controlling variables' related to the so-called 'Galton's problem' (the apparent 'problem' that cultures are never fully independent from each other) dominated their ref lections. Even in a recent edition of a seminal textbook by Bernard (2015) on methods in cultural anthropology, the cross-cultural method introduced is firmly grounded in positivism and aims to 'test hypotheses' (Ember and Ember 2015). Thinking about comparison in positivist terms with positivist terminology such as testing hypotheses, variables, comparability and causality hence still haunts anthropology, and there is an urgent need to free the discipline from this baggage. The paradigm shift from positivism to interpretivism and constructivism is, however, slowly leading to a new body of comparative anthropology which considers the interpretive turn and the crisis of representation.

### Interpretive, Post-Crisis Comparative Approaches

It is in thematic, often interdisciplinary, fields that calls for comparison have been raised anew in the last few years. Among others, calls for comparisons emerged in the field of transnationalism and in the political anthropology of citizenship, in the debate on the ontological turn and related fields like multi-species anthropology, and in the already mentioned interdisciplinary, geography-dominated field of postcolonial comparative urbanism. Since the crisis of representation, only three anthropological collections (Gingrich and Fox 2002, Holý 1987a, Scheffer and Niewöhner 2010) have systematically explored new styles of comparison which depart from the positivistic comparative methods.

There are at least *four* particularities which appear across these bodies of literature and which can be seen as the shared basis of an emerging field, which I call interpretive, *post-crisis* comparisons, emerging after the postmodernist *crisis* of representation. Firstly, these scholars share the idea that comparisons do not receive the attention they should. Secondly, they argue that comparison is a key tool for deprovincialising and questioning established knowledge. Thirdly, they claim convincingly that new methodologies should not let themselves be limited by ideas like incommensurability, and fourthly, they believe that practices and processes need to be central instead of fixed units.

First of all, across these literatures, there is agreement that anthropology cannot do without comparison, despite its difficulties and shortfalls (Gingrich and Thelen 2012: 398, Strathern 2002). Even extreme relativists are engaged in cultural translation and therefore compare (Fox and Gingrich 2002: 20). If we fail to ref lect on comparison, we fall into the trap of making problematic, implicit and unnoticed comparisons

of self and other, of the exotic and the known (Gingrich and Thelen 2012: 398, Gupta and Ferguson 1997). As the anthropologist Sian Lazar (2012: 353) says so pointedly, with the crisis of representation, the baby (comparison as method) was thrown out with the bathwater (positivism and objectivity). These scholars find this problematic, as the production of anthropological knowledge is based on all sorts of comparisons. The question is less about whether we compare and rather about *"what kind of recognition* [italics in original] scholars give to this basic human activity" (Fox and Gingrich 2002: 20). If anthropology wants to be a self-ref lexive science, comparison needs to be included in methodological and epistemological debates. Thinking about comparison is more than merely ref lecting on a certain method; it also entails ref lecting on anthropological knowledge in general.

Besides these epistemological arguments there is also ample recognition that globalisation and the increasing entangledness of spaces and places across the globe demand comparisons: "These global connections and the heterogeneous local responses to them legitimate a renewed comparative agenda for anthropology and related fields" (ibid: 7). Because of increased global connectedness, it is not only researchers but also practitioners who engage in comparisons, for example there are urban planners working in municipalities and politicians who readily adopt urban policies developed in other cities (Ward 2010). As people across the globe become subjected to similar processes and models, it is necessary to compare how actors engage differently or similarly to them. A very practical argument for renewed comparison is research funding. In countries such as the United Kingdom and Switzerland, research funding institutions increasingly demand interdisciplinarity and collaboration (Lazar 2012: 353).

The second particularity shared across the interpretive, post-crisis approaches is that they find that comparison, even in its positivist version, has a subversive potential and can call existing knowledge and frameworks radically into question. By comparing 'civilised' and 'primitive' cultures, by explaining variations as cultural difference and denying the explanatory value of race and biology, Tylor made a strong political statement against Darwinism and racism (Yengoyan 2006: 140). Through comparison, Strathern (1997) critically investigated concepts like gender and dismantled them as not being culturally neutral but as emanating from the West. Goody (2006) also used comparison to show how concepts about society and history usually take Western societies as starting points. By comparing European civilisation with other histories, he dismantles the claim of the uniqueness of European civilisation and unsettles our understanding of European history.

It is therefore no coincidence that fields which aim to break established boundaries call for comparison: the ontological turn in anthropology and multi-species anthropology calls into question the centrality of humans as the only actors in a world also constituted by non-humans. In this debate, comparison, based on Strathern's writings, has become reassessed as a central tool for thinking about different ontologies. Tsing (2014), for example, develops a comparison in the form of a cultural analogy between mushrooms and human actors. In the field of postcolonial, critical urban studies, Robinson calls for comparison because it should contribute to the deprovincialisation of urban theory. Already the Manchester anthropologists researching the Copperbelt had critically scrutinised the theories of the Chicago school by researching cities comparatively but contemporary urban theory framing cities in the North and South as distinctly different forgot about these debates (Robinson 2006a: 5-7). By comparing

widely different cities we can decentre Eurocentric and America-centric notions of urbanity (Robinson 2006a, 2011).

The third particularity of this emerging field relates to the *how* of comparison. These scholars find that comparisons should be more experimental, more diverse and they should not let themselves be limited by outdated methodological ideas like incommensurability (Robinson 2011). In order to understand this methodological critique, one needs to scrutinise in more detail how positivist comparisons work. Positivist comparers firmly believe that comparisons are something that can fail. Failure happens, for example, because researchers make so-called 'category mistakes': they set off to study apples but find out that one of the apples is actually a pear. This is a serious problem for positivists because of their deductive approach aimed at testing theory. In positivist comparison, the process of data collection is the execution of a plan drafted at the beginning of the study, a theoretical framework is drawn up and is then 'filled' with data. For the positivist comparer, the initial design of the comparison is absolutely crucial; this is where she, informed by theories, develops hypotheses which she then tests. As this initial framework is like a fixed shell, it is possible to make mistakes in the construction. The data may not fit (category mistakes), or the data may be overly different (incommensurable).

Because positivist analytical frameworks are built to test theory and not thought to be adapted to empirical reality, they cannot easily replace the predefined category 'apples' with a broader category 'fruits'. Neither can they use the contradiction between reality and their category to ref lect on what their initial misreading of the pear as an apple tells us about apples and fruits and our conceptualisations of them. This is why positivists are greatly concerned about sampling and comparability in the planning of the research.

Interpretive, post-crisis comparative approaches depart significantly from that. Category mistakes and apparently incommensurable difference are not seen as a failure of comparison but as a useful tool for thinking about our categories. All the different approaches included in this review, like the critical urban studies' call for comparison (Robinson 2006a, 2011), studies of citizenship (Handler 2009, Lazar 2012), comparisons inspired by Strathern (Holbraad and Pedersen 2009, Tsing 2014); and Detienne (2008 [2000]), claim that we should expand our horizon to compare things which have previously been perceived as incommensurable. Abolishing the idea that the things compared must somehow be similar is an important aspect of unsettling the canon of positivist comparison (Nader 1994: 87). Robinson (2011) criticises the fact that assumptions of fundamental incommensurability of different kinds of cities have limited comparative research on urbanity (ibid: 2). By calling for cities to be treated as 'ordinary', she proposes experimentation with comparisons across widely different contexts (Robinson 2006a, 2011).

Lazar (2012) coins the useful notion of *disjunctive comparison*. In disjunctive comparison, two quite different 'things' are placed next to another (Strathern 2002: xvi) so that the data thus placed can talk to each other (Lazar 2012: 351). With the notion of disjunction she signals that the conviction of post-crisis comparers like Detienne (2008 [2000]) and Handler (2009) that it is possible and interpretively productive to compare 'things' which have very little in common and which would have been considered incommensurable by positivist comparers. Strathern is "perhaps the mistress of disjunctive comparison" (Lazar 2012: 351) as she compares, for example, contemporary

gender relations in 'Euro-America' today with gender in Melanesia in the 1970s (Strathern 1997).

The fourth particularity of interpretive approaches to comparison is their call to move away from comparing fixed units towards comparing practices and processes. What do we actually compare? How are the 'things', or the 'units of comparison' constituted? Anthropological cross-cultural comparison was long thought to work across collectives, meaning "social groups conceptualised, roughly, as species are conceptualised in the natural sciences", like nation-states, cultures, societies, tribes or races (Handler 2009: 628-9, see also Moore 1993). These apparently naturally existing things with clear boundaries and internal coherence were thought of as units of analysis. The positivistic comparers took their 'objects' of study as things which are simply given and could be compared. With the epistemological turn, however, social facts were no longer regarded as things but as constructions (Holý 1987b: 15). Abu-Lughod famously argued that the notion of culture in anthropology is an "essential tool for making other" (Abu-Lughod 1991: 470). Cross-cultural comparative methods became regarded with high suspicion as they compared 'cultures', assumed to be stable, highly integrated and self-contained (Fox and Gingrich 2002: 2). So comparers who take the crisis of representation and the deconstruction of notions like cultures as essentialist seriously are confronted with the need to reformulate and rethink what they actually compare. It is important to acknowledge that objects of comparison are socially constructed (Scheffer 2008: 283). Fox and Gingrich argue that units of comparison should not be "accepted as discrete, homogenous and stable entities at all", but they should be understood as "differentiated, changing results of wider developments, within their fuzzy boundaries" (Fox and Gingrich 2002: 19). Many claim that we should study processes instead of outcomes (Moore 2005, Strathern 1981, Robinson 2011). Building on these insights on what a renewed anthropological approach to comparison may look like, in the next section I will outline the way in which the comparative fieldwork for this book evolved, advocating for a style of comparison which is circular and takes into account the deeply subjective side of comparative ethnography which calls for ref lexivity.

### **The Biographies of Units of Comparison**

The philosopher Ralph Weber (2014) draws attention to an often neglected aspect of debates around comparison by pointing out that comparisons have a temporal dimension. At the beginning of a comparative research project, in the 'pre-comparative moment', the researcher develops a *pre-comparative third*, a heuristic idea of what the cases should be cases *of*; she also develops *comparanda*, namely the things she aims to compare, and plans data collection accordingly. During the comparative moment, the researcher collects data, engages in data analysis, produces generalisations and compares them. The post-comparative moment refers to the end products of the comparative project, for example the written-up published article or written-up chapters of a theses. Here, the final *comparata* and the final *post-comparative third* are represented through the writing or the presentation of tables.

Understanding comparison as a process and not solely as a moment is very useful for comparative ethnography. Yet Weber's linear understanding is inadequate for a methodology grounded in induction where the relationship between concepts and

data is one of circularity. Comparative ethnography needs to be understood as a circular process, with the mind and body of the ethnographer involved in constant comparisons along the way – comparisons taking place in the field, in the re-reading of the data, in the process of writing up. The 'thirds' (what it is a case *of*, the overarching themes or concepts) and the 'cases' (units of analysis, *comparata*, things) change constantly over time, involving "constant critical ref lection by the researcher as well as a delicate balance of both immersion into and distance from social reality" (Förster 2011: 13). Thirds and cases may even be different in each written-up article, each chapter or even each subchapter, emerging from the ethnographic research process. Instead of thinking about comparison as moments with a pre- and post-phase, I suggest the notion of *biography* be mobilised in order to speak about the temporality of comparative projects. This is inspired by Scheffer and Niewöhner (2008: 281), who argue that interpretive comparison demands a ref lective stance by the comparer towards the *biographies of comparables*, namely, how things and thirds were "produced, defined, traced, employed and dismissed". They call for *thick comparisons*, hence grounded in thick description (Geertz 1983), which I also understand as a comparative attitude that entails critical self-ref lection on the transformation of analytical frameworks, cases and thirds (*biography of comparables*) during the course of the project. In this section, I will therefore describe the biography of the units of comparison in this book. Thick comparison should also, though, as I argue in the following section, include critical self-ref lection on one's involvement with the field, about how specificities of the field and the positionality of the people involved shaped comparative fieldwork.

Between 2010 and 2012 I spent 14 months doing fieldwork in Johannesburg and Maputo. This was divided into two months of preliminary study (January 2010 in Johannesburg, February 2010 in Maputo), a core field phase (September 2010 to January 2011 in Maputo, February 2011 to April 2011 in Johannesburg), and a follow-up study (April 2012 to June 2012 in Johannesburg, July and August 2012 in Maputo). Based on ethnographic methods like participation and observation (Förster 2001) and diverse forms of interviewing (Meuser and Nagel 2002, Spradley 1979, Wetherell 2003), I worked with the Emic Evaluation Approach, consisting of a triangulation of three different methodologies, namely, the mapping of actors and spaces, social discourse analysis and practice analysis (Förster et al. 2011, Heer 2011). Moving between Switzerland, South Africa and Mozambique several times meant constantly moving between immersion, literature review, writing and adaptation of fieldwork focus.

My initial interest in urban spaces was sparked by a debate in my home city of Basel in the summer of 2009 when I was writing my research proposal. The head of the urban development department of the City of Basel stated in an interview in the local press that the youth culture of barbecuing sausages on the Rhine riverbank harmed the city's image and that the 'cultural niveau' needed to be raised (Loser 2009). This polemic sparked a debate in the city on how and by whom the public spaces along the Rhine should be used, how urban society evaluated certain lifestyles, and who had the power and resources to turn their image of the good city into a social reality. Power, social diversity, morality and politics seemed to culminate in public spaces. When I started reading on public spaces, I became aware of ongoing debates about the privatisation and commodification of public spaces in the 'postmodern' era in cities across Europe (Selle 2002), the US (Davis 2006 [1990], Low and Smith 2006), Brazil (Caldeira 2000) and Africa (Murray 2004). I was initially interested in cities and urban spaces, and the selection of the cities as research settings only came after that. I decided to do fieldwork in Maputo because I wanted to get to know Lusophone Africa and make use of my Portuguese skills which I had acquired a long time ago in an exchange year in Brazil. One of my PhD supervisors, Till Förster, inspired by Robinson's (2006a) call for comparative urbanism, suggested a comparison with Johannesburg, not least because questions of the privatisation of public space were especially relevant there (Bremner 2006, Dirsuweit 2007, Parnell 1997, Peyroux 2006). I was familiar with South Africa, as I had spent an exchange semester at Rhodes University.

In 2010, during a preliminary study of two months, my intention was to get to know as many parts and spaces of the city as possible, using methods like 'go-alongs'<sup>2</sup> (Kusenbach 2003) and exploratory walks and drives.3 My aim was also to get an overview of diverse, more or less public spaces in the two cities, and in some of these spaces I experimented with systematic observation, a non-participatory form of observation (Beer 2003). In addition, I simply spent lots of time in shopping malls and bars, and in Maputo also in public squares and parks. At the time I was inspired by what I call the *public space approach*, a research methodology used by anthropologists like Setha Low (2000), Kathrin Wildner (2003) and others which takes the material, architectural public space as a starting point for the ethnography. In their studies, both Low and Wildner describe a specific public space and the social practices and specific events emerging in them, and they interpret their case within the broad context of the city, of the society and of the nation at large. Both Wildner and Low chose central places with high symbolic meaning for the city: the Zócalo in Mexico (Wildner 2003) and two plazas in Costa Rica (Low 2000). When I was back in Basel after this preliminary study, I wondered which of the many public spaces I had mapped in both cities I should focus on and include in the 'sample' for my comparison. This proved to be a complicated question.

In Johannesburg, I had learnt that for many urban dwellers from townships and suburbs, the spaces in the inner city do not form part of everyday routes. For aff luent milieus in particular, the inner city had become a 'no-go' zone and they preferred to spend their leisure time in shopping malls in the suburbs. In Maputo, however, the downtown area (Baixa) had retained its function as a centre for the majority of urban dwellers, despite being experienced by better-off milieus as chaotic and exhausting. If I were to compare an inner-city public space in Johannesburg with one in Maputo, I would compare two architectural spaces which, from a social point of view, could have totally different social meanings for totally different urban milieus. I felt that the public spaces I thought I needed to select at this early stage of research, without yet understanding much of what was going on these spaces, would greatly inf luence *whose* public spaces and *whose* city I would write about.

<sup>2</sup>  Go-alongs are 'naturally' occurring situations rooted in everyday routines, which are influenced but not determined by the presence of the anthropologist (unlike in exploratory walks). They basically entail participation and observation on the move (Kusenbach 2003).

<sup>3</sup>  In exploratory walks, the anthropologist moves through the city similarly to the literary figure of the flaneur (Benjamin and Tiedemann 1982), familiarises herself with the complex urban environments by perceiving as much as possible with heightened senses. By walking, the anthropologist creates an encounter between herself, the materiality of the city and the rhythms, atmospheres, orders, noises, smells, actors and other aspects of everyday life on the streets, mediated through hersenses(Magnani 1996:16-17, Paasche and Sidaway 2010:1556, Wildner 2003: 7).

In addition, being a newcomer to both cities, I felt that I was not able to interpret the data I had collected by observing and participating in the life in these public spaces. Looking through my fieldnotes back home in Basel, I realised that by merely hanging around in public space I would not be able to understand what role these spaces played in people's everyday lives. I had no idea how the people I talked to at the malls, in the bars and the squares lived, what their everyday routines looked or how the space where I met them inserted itself into their routines. When re-reading Wildner (2003) and Low (2000) I became aware that in their ethnographies, the everyday urban lives of the users of these spaces were lacking, which meant that my difficulties were not so much related to the fact that I was at an early stage of my PhD but that it was a problem of the methodology. A complicating factor was also that the management of Maputo Shopping gave me research permission to conduct research *in* the mall, but they forbade me from re-visiting interviewees at their homes or somewhere else *outside of* the mall. This meant that I could not accompany mallgoers back to their homes and everyday lives. Accordingly, the *public space approach* seemed more and more inadequate to me, as I completely lacked the contextual knowledge to interpret what I observed in these spaces.

In both Johannesburg and Maputo, I was immediately drawn in the lifeworlds of urbanites of my age who had university degrees and were working in the public sector or for large companies. Making friends with them was easy, as we shared a similar background and without much effort I got to know their places of leisure, their homes and ways of moving through the city. The snowball approach, meeting new people through my existing friends, rapidly created a bias towards the urban experience of middle-class milieus. As I was interested in how urban dwellers deal with urban difference and inequality, I did not like the idea of restricting myself to one urban milieu and essentially to one perspective on the urban.

After a lot of ref lection, I decided to temporarily move away from the public spaces: I wanted, instead, to look at urbanites' daily lives and hear and observe when and where they actually meet other people because I thought this was what my interest in public space was all about. I wanted to switch from being a sole interviewer and outside observer to becoming a participant, somebody who accompanied urban dwellers in their lives and on their visits to public spaces. With that, the biography of my thirds took an important turn. My preliminary concept changed from 'public space' to 'public life', by which I understood the part of life that happens in the 'public', defined as the sphere of urban life where one meets people with whom one is not linked through kinship or other kinds of close personal relations (Lof land 1973, Sennett 1983 [1974]: 16). I also decided to do what many urban ethnographers have done before me, namely, to use neighbourhoods as an entry point, as spaces of immersion in everyday life, where I hoped that it would be easier to participate in everyday life and establish familiarity than in the anonymous public spaces.

Neighbourhoods are one of the preferred units of analysis for urban ethnographers. The first version of a neighbourhood approach was developed by the ancestors of urban anthropology, the Chicago School of Sociology. A neighbourhood approach basically means that one draws the boundary of the study, the limits of the field, according to the administratively or otherwise defined boundaries of a neighbourhood. In both cities, Johannesburg and Maputo, I chose two neighbourhoods as a starting point for my research, neighbourhoods which on the one hand exemplified the large urban divisions shaping these cities, and on the other hand were situated right next to each other. On the one hand, this was for practical reasons, so that I could save travel time and visit informants in both neighbourhoods on the same day. And on the other hand, the constellation of spatial proximity and large social differences seemed like an ideal setting for finding answers to my key interest, namely, how urban dwellers deal with the large differences and inequalities in everyday life.

Assisted by research assistants (see below), I explored the neighbourhoods on foot or by car and I asked people to take me to different places which were important in their daily lives. This eventually led to a general mapping of the neighbourhood and some of its places with importance for public life. The other important starting points were qualitative interviews with different residents about their everyday life, their spatial trajectories through the city, their perceptions of their own and the other neighbourhood, and so on. In each neighbourhood, I interviewed about ten to fifteen residents with a qualitative question guide, which I continuously reworked and adjusted to include new topics and new spaces that previous interviewees had brought up. The interviews covered basic data about biography, livelihood, engagement in neighbourhood organisations, neighbour relations, their daily routine, modes of transport and many other topics. I also asked specifically about the frequency, use and perception of places that previous interviewees had mentioned, like shopping malls, parks, the inner city, bars and religious spaces. Sometimes I asked them to show on a city map where their everyday trajectories took them to and often, I showed them photographs of places, which inspired interviewees to tell stories. In Alexandra, I asked a couple of friends to keep a diary of their everyday routines (Zimmerman and Wieder 1977), based on which we had long conversations. These diaries gave me new, unexpected insights into their everyday life, like the fact that some households in Alexandra re-use paper towels from Sandton's toilets as toilet paper (see chapter 2). My lengthy and repeated presence in the neighbourhoods had the advantage of slowly building up rapport with various urban dwellers, the advantage of being able to visit them repeatedly, of hanging out in places where the regulars started to know us. Restricting my attention for a couple of months to one neighbourhood allowed me to develop a more in-depth picture of a section of these metropolises.

With time, my attention was drawn to topics and spaces where the everyday lives of residents of the two neighbourhoods intersected. In Maputo, I was fascinated by the stories people told us about the road closures which members of the elite had apparently built and residents of Polana Caniço had destroyed. The Maputo Shopping Centre crystallised as a key place of public life for residents from Sommerschield II, and I learnt that residents of Polana Caniço also had interesting stories about their mall visits to share. In Johannesburg I found out that Linbro Park residents were worrying about their futures, about when public housing for Alexandra residents would be built. I started to accompany people as much as possible to activities related to these points of intersections, so I went with friends to the mall, attended meetings of the Linbro Park Community Association (LPCA) and hung out at the LRC Church.

Back in Basel after this lengthy second field trip, the biography of the thirds and case studies took another turn. When analysing data on the themes and the spaces where the trajectories of the urban dwellers from both neighbourhoods and milieus intersected, I came to understand them as *spaces of encounter*, hence spaces where urban dwellers come to engage with each other. During the last field visit, I focused my research activities even more on such spaces of encounters. For example, I returned to shopping malls for systematic observation but now, in contrast to the beginning of the fieldwork, I had the necessary contextual knowledge to actually understand what was going on in them socially. I had been to the mall many times in the company of mall users, they had shared their stories about mall visits with me, and I knew their lives outside of the mall.

In this circular process of adapting concepts and focus to the emerging data, a threefold focus emerged which I call the *multi-sited neighbourhood approach*: starting from unequal yet adjacent neighbourhoods, following (Marcus 1995) urbanites across the city, and then focusing on spaces of encounter, the sites and themes where the lives of my informants from the adjacent neighbourhoods became entangled. This threefold focus allowed me to zoom in on several cases and themes which emerged from the fieldwork and which seemed relevant to the specificity of the localities yet also enabled comparisons.

Unlike in classical community studies where neighbourhood boundaries are used to delimit the case studies, in this comparative ethnography neighbourhoods rather served as heuristic starting points and as a "window to complexity" (Candea 2009: 37). Initially, I believed that the neighbourhoods would have the role of *arbitrary locations* (ibid), arbitrarily chosen entry points to study urban complexity. It was for me an empirically open question as to whether they were really existing social formations or merely urban quarters drawn on a map by urban planners with little significance for the everyday lives of the urban dwellers. It was, hence, never my intention to actually compare neighbourhoods. Over time, however, I became aware of how, for many of my informants, their neighbourhoods were important places of belonging and many of their activities that I documented, especially concerning neighbourhood governance and politics, actually contributed to creating this sense of belonging. Besides being geographical places and arbitrary locations for me as researcher, the neighbourhoods also turned out to be what Förster (2013a: 8) calls 'intentional objects', namely, shared images which "orient actors toward specific content". As such intentional objects it would be interesting to compare the four neighbourhoods.

When writing the PhD manuscript for submission to the university, I developed a framework called *spaces of encounter* (Heer 2015a) in which the comparisons did not so much involve the cities but the case studies of the spaces of encounter. Only when rewriting the manuscript for publication did I come across the notion of entanglements in Srivastava's (2014) work, and the geographer Sophie Oldfield pointed me towards Nuttall's book (2009), which allowed me to formulate *cities of entanglements*  as a more overarching framework in which the cities as such also came into view. Up to now, this has been the last step in the biography of the units of comparison and thirds. However, if I were to continue working on the material, I have no doubt that new theoretical frameworks, more sophisticated thirds and quite different units of comparisons would emerge. This is very typical of a circular and interpretive research process. An interpretive comparer starts off with tentative sensitising concepts, which are then continuously adapted. What the research is a *case of* is often unclear until the very end (Ragin 1994: 121). *What* I was comparing and how I was analytically framing it has hence been constantly changing, and this f luidity and circularity of data collection and analysis is a specificity of comparative ethnography. The comparisons written up, the comparative arguments made, the similarities and differences implicitly or

explicitly pointed out in this book are, thus, just a few among the many comparisons comparative fieldwork enables us to do. Like any other form of research, comparative ethnography is never finished and always imperfect.

### **The Entangled Comparer**

An extremely important epistemological difference between *thinking cities through elsewhere*, which I understand here as comparing one's case study by reading other people's cases or by comparing data collected by others, and *experiencing cities through elsewhere*, that is, comparative ethnography as fieldwork conducted by a single researcher in at least two places, lies in the role of the comparer. The comparer, that is, the person who conducts the comparison, has received little attention in methodological ref lections on comparisons, which is a problem, as the comparer is the key locus, if not to say the embodiment of the comparison. Actually, it is through the comparer – her body, her experience, her analysis, her writing – that the two or more fields become fundamentally entangled. The comparer is herself an actor who renders the two places more connected by moving between the two places, by bringing people from both places in contact with each other, and by being present in at least two academic fields. Focusing attention on the role of the comparer reveals that knowledge production in comparative ethnography becomes deeply shaped by the positionality of the comparer and her research assistants, the manifold relationships that emerge during fieldwork, and the many specificities of the urban contexts.

### Positionality and Politics

Since the crisis of representation in the 1980s and the epistemological and political shifts since the 1960s, ref lexivity has increasingly become a marker of good quality ethnography. During thick description, the ethnographer should give an account about her personal and culturally moulded attitudes, perceptions and conceptions. The ethnographer needs to be aware of herself as a historical subject (Förster 1997: 39). This should also be so for comparative ethnography: comparers have to ref lect on their positionality (Melas 2007: 3) and give an account of the diverse relationships that shaped the research process and, hence, knowledge production. For this, I again mobilise the notion of entanglement, drawing on its use by feminist researchers from the field of political ecology. Entanglements draw attention to the way in which researchers themselves are "situated in and often beneficiaries of the very politico-economic systems under consideration in our research" (Sundberg 2015: 117). Rather than seeing the comparer as an actor standing at a distance from the data, producing objective conclusions, the comparer should be seen as standing in multiple and mostly asymmetric relationships with informants, relationships deeply shaped by the manifold specificities of the urban context and, in the case of researchers from the North and from privileged backgrounds, often benefiting from the same hierarchies they describe and criticise in their work.

Comparison involves power relations and responsibility. Following Hobart, "relations of similarity and difference are not given in the empirical phenomena themselves but are generated by the people who act on them and decide, using criteria of their own choosing, to which class, category or concept they conform" (Hobart 1987, Holý 1987b: 16). Representation and comparison of other cultures entails intellectual and academic hierarchies, as the critique of comparison in the 1980s pointed out, and post-crisis comparison therefore has to ref lect on the power relations and public responsibility of comparers (Fox and Gingrich 2002: 9). Nader famously argued that anthropology needs a comparative consciousness, meaning that anthropologists need to be more conscious and responsible about the comparisons they conduct (Nader 1994: 89). But new forms of comparison, she claims, have to cope with questions of process and hegemony, which is not easy. One has to acknowledge that comparison involves the negotiation of unequal power relations "between and among the networks and processes of social actors under study, the author(s), and the audience of readership" (Fox and Gingrich 2002: 19). One's position in social hierarchies at home and in the field, the social milieu in which one grew up and in which one writes, the scientific training one has received, the university where one is based all shape the anthropological gaze.

I come from a family of teachers and I grew up in rural Switzerland. Finding this Swiss valley too narrow, I went abroad, first for a school exchange year in Brazil, later for an internship in India, and during my studies in anthropology and gender studies I spent a semester at Rhodes University in South Africa. In Brazil, India and South Africa I was confronted with what I experienced as extreme, violent inequality between social classes, culturally and racially defined groups, while I had easy access to a privileged social position in the social structure. How people deal with such inequality has hence been a question that troubles me, ethically, politically and empirically. This question did not arise because I did not know inequality from Switzerland. Actually, as the child of secondary school teachers, who constituted part of the local elite in this rural part of Switzerland, I grew up in a house on a hill looking down on the high-rise buildings where the worker families lived. Living a life which I felt was privileged was as much part of my upbringing as the social and feminist values that my parents lived in their work as teachers and as active members and leaders of civil society organisations.

While my milieu and biography have certainly shaped my outlook on the world, my research interests and hence my comparisons, the comparative research process, in turn, has had lasting effects on me. When I started my PhD, I joined the feminist section of the social democratic party in Basel-Stadt, the city that I moved to in order to study and where I now live. In 2018 I became a member of parliament in the City of Basel, so that I am now (also) a politician. My political outlook in Basel is deeply coloured by my experiences in Maputo and Johannesburg. I agree with Sundberg when she calls for an *ethics of entanglement*, by which she means that researchers should "be involved in the struggle for a just world from and in our own sites of entanglement and engagement" (Sundberg 2015: 123). What I do not agree with, however, is when academics believe that they can change the world through knowledge production alone. As I argued in the conclusion, the fact that entanglements are blind fields, unrecognised aspects of urban and human reality, is not just a question of a lack of knowledge or research gaps but is about the denial of responsibility towards others by more powerful groups, it is about not *wanting* to see. Filling research gaps and presenting our results to broader audiences is hence not enough for an engaged anthropology, I contend. Directing the futures of cities is about power struggles between different future visions, and if anthropologists want to be involved in the struggle for a just world, they have to become part and parcel of these struggles, bringing their analysis

into these struggles, and not just sit as apparently neutral experts on the side lines (Scheper-Hughes 1993). There are many routes to how academics working at universities can do this, for example by raising their voices in ongoing public debates, or by working with or being active in NGOs that do advocacy work.

### Fieldwork in Webs of Relationships

Neely and Nguse (2015) propose the notion of entanglements to "think through how researchers' and research subjects' relational positionalities shape knowledge" (142), and how "research is entangled in a web of relationships" (141), ranging from relationships with the informants and research assistants, friendships we make during fieldwork, supervisors, colleagues and so on. Research and positionality emerge from interactions and both the researchers' and their interlocutors' positionalities shape knowledge production. For comparative ethnography, it is important that the comparer makes transparent the entanglements under which the data were collected because this renders the process of knowledge production open to scrutiny by the reader (see also Ammann, Kaiser-Grolimund and Staudacher 2016). One key set of entanglements is the relationship between researcher and research assistants which profoundly shapes the development of rapport with other urban dwellers. Throughout the fieldwork I worked with three research assistants with different personalities, different resources and, most importantly, different positions within the local landscapes. In Maputo, I had the pleasure of working with Fernando Tivane, himself an anthropologist. He had just finished his licentiate (*licenciatura*) at Eduardo Mondlane University (Tivane 2010) and was working as a lecturer in anthropology and doing his master's thesis. He eventually became a colleague, key informant, co-researcher, co-analyser and close friend. Almost all the research activities in the neighbourhoods of Maputo we did together and when I was back home, he transcribed the interviews.

Being a XiShangana speaker who had moved from Gaza province to Maputo as a child, he could more easily establish rapport with residents from Polana Caniço than I could. He was, however, not an insider to the neighbourhood, firstly because he was living in a different neighbourhood, Laulane, and secondly, as an anthropologist working at the university he belonged to a different milieu. Appearing in people's yards and the neighbourhood streets always together, as a European and an African anthropologist, was not only well received but many residents of Polana Caniço saw this as a colonial relation complicated by the fact that I was a white woman employing a black man. Hence, Fernando and I rather downplayed my origins from a European university and the hierarchy in our relation. We showed our research credentials from Eduardo Mondlane University, signed by the neighbourhood secretary, and we presented ourselves as a team of researchers, not as an anthropologist and her research assistant. Some people suspected that we were lovers, as it was apparently difficult to imagine a different relationship between a man and a woman, something which caused some complications for Fernando who was getting married at that time.

The lives of the elite milieus living in Sommerschield II were just as unknown to Fernando as they were to me. In the conversations with the aff luent residents it was sometimes helpful to emphasise my connection to a European university, as the residents associated everything European with prestige. In addition, by emphasising that the information they gave us was destined for a PhD written abroad, we could instil the trust in these inf luential politicians, public servants or company managers that we were not journalists or spies who would use the knowledge to harm them or expose their private lives in the local public sphere. Some had studied in Europe themselves and nostalgically remembered these times in their conversations with us. So, in contrast to Polana Caniço, where my European origin was rather a hinderance, in Sommerschield II we could use it as a door opener. In Polana Caniço, Fernando could more easily establish rapport, in Sommerschield II both of us were strangers.

In Alexandra, I worked with Thabo Mopasi, a 40-year-old Southern Sotho-speaking Alexandra resident and member of the long-term tenant milieu, who had been involved in many projects on the township and often introduced outsiders like me, mostly researchers and journalists, to Alexandra. Thabo was a great gatekeeper to local community leaders and other well-connected residents, and he had a thorough knowledge of and involvement in the township's history and politics. Alexandra is a place with a lot of everyday violence, from the private realm of the household (domestic violence, sexual abuse) to public spaces (violent muggings, rape and murder, car hijackings). Unlike Polana Caniço with its narrow paths, Alexandra has mostly tarred roads and orientation is easy due to the grid pattern (except from the shack settlements). Although I soon felt comfortable walking around on my own, Thabo and others fiercely insisted and without tolerating exception that I should not do that. Unlike in Maputo where Fernando and I spent lots of time on the neighbourhood streets, sitting at bars observing neighbourhood life, in Alexandra, I rather hung out 'inside', so at Thabo's office at San Kopano community centre, at Thabo's in-laws' yard, or at other friends' houses. As a well-known and engaged personality, and a fervent ANC activist, Thabo had his own agenda regarding what my research should be about, namely township politics, while I also wanted to get to know less high-profile aspects of township life. With time, I was able to build up relationships with people beyond Thabo's social world, and I became more independent by driving around in the township on my own, which, in contrast to walking around alone, my friends considered safe for me.

During my third stay in Johannesburg I got to know Nnana, the daughter of a domestic worker employed in Linbro Park. She was living with her mother in River Park, a section of Alexandra constructed in the 1990s, situated on the eastern border of the township and in geographical proximity to Linbro Park as well as the Greenstone area. Nnana grew up with her sister in a rural area while her mother had moved to Johannesburg in the 1980s to work as a domestic worker in Linbro Park. The life of her family has for many years been tightly entangled with the life of her employers. In contrast to Thabo, an insider with dense social networks and knowledge of Alexandra, Nnana and her family were outsiders, shack dwellers who constantly moved between their rural 'home' and Johannesburg, depending on where life is cheaper and where they find employment. By accompanying Nnana in her everyday life, I came to see Alexandra through the eyes of a newcomer with few social networks, with comparatively little knowledge of the township, and who was scared and insecure about moving through the township. Without Nnana, I believe, my take on township life would have been considerably different.

Linbro Park was the only neighbourhood where I went around on my own, without a research assistant, because I could easily establish rapport with the suburban residents of whom many had migrated from Europe or at least had family members there. In the interviews, most Linbro Park property owners were very friendly and informal, and they signalled that, although I was not a South African, they nevertheless saw me

as a fellow 'white'. Many aff luent residents were nevertheless also suspicious that as an anthropologist, I would write about them in critical or negative terms, and they were worried that I might judge their way of life. Irving Goffman's writings about performances are very helpful for understanding such interview situations in which actions or attitudes which are inconsistent with broader societal ideals, for example sensitive issues like social differences, inequality and cultural, racial or ethnic others, are discussed (Goffman 1959, Wetherell 2003). Interviews should not be analysed only in terms of what is said, but as social situations in which actors do things with words (Keesing and Strathern 1998: 40). Interviews do not produce 'objective', raw data, but are interactions in which interviewee and interviewer jointly engage in the construction of social reality (Sarangi 2003: 65-67). Both the researcher and the interviewee engage in self-presentation (Goffman 1959). In my conversations and interviews with property owners in Linbro Park I avoided talking extensively about my involvement in the adjacent township because it would have confirmed their suspicions that I would write about them critically and because being "seen as a member of the ingroup or outgroup can easily inf luence what is said and how something is said" (Sarangi 2003: 67).

In conversations with domestic workers in Linbro Park, sharing stories about my fieldwork in Alexandra helped to build trust. Building relationships with them demanded patience, as they initially placed me within their social landscape as an employer or they believed I was the daughter of an employer. Something similar happened sometimes when I met female township dwellers in Alexandra who worked in the suburbs as domestics. Some of them activated the registers typical of domestic worker–employer interactions. They talked to me if I were a potential employer, using a specific, for me artificial sounding, tone of voice, signalling obedience and friendliness. In order not to be seen as a white suburbanite and potential employer, it again helped to emphasise that I was an anthropologist from Europe. Unlike in Polana Caniço, where being a European researcher raised suspicion, in Alexandra residents valued it as something positive to have a foreign researcher writing about the township, not least because many proud township dwellers aspire for 'Alex' to become better known to the wider word, like Soweto already is. Being white, however, also meant being seen as wealthy. Thabo and his family received social pressure to share the money that their neighbours thought I was giving them.

One of the most striking differences between fieldwork in Johannesburg and in Maputo is that while in Johannesburg I was often seen as a somehow exotic stranger, who was potentially interesting as a tool to make one's experiences and views known to a wider world, in Maputo relating to me as a stranger was somewhat troubling for many of the people we met, as if it entailed many social risks for them. Would I make money from the photographs I took of Polana Caniço's streets? Would I do damaging things with the information they were giving me? Many people were initially worried that I was a spy or a journalist. While in Johannesburg, I felt that people were relatively open toward strangers, in Maputo, I experienced social closure and low levels of social trust. It took significantly more time to get access to people's networks in Maputo. Fernando and I were rarely invited to political or social events in Polana Caniço or Sommerschield II, and even my very close friends took their time before they invited me to family gatherings. To give an example, in Johannesburg phone numbers are almost public knowledge, lists of residents' phone numbers are put up on websites or sent around via e-mail, and people gave me the phone numbers of friends or neighbours

without hesitation. In Maputo, phone numbers are something very private which people believe they are only allowed to pass on to others if the owner of the number has given permission. Sharing one's contacts with others is dangerous, as one could pass on the social risk embodied by the stranger to a member of one's network, which could damage the relationship. In many such instances in Johannesburg, I felt that people shared their social capital with me in order to make their networks grow, whilst in Maputo I experienced that social capital was rather seen as something which could decrease if they shared it with me.

Distinct communication styles also went along with these city-specific ways of relating to strangers. In Maputo, I observed that politeness was extremely important in stranger interactions, and urban dwellers' initial interactions with me tended to be reserved and cautious. In terms of body language, it often felt as if the person would incline their upper body away from me in order to protect their personal space. I learnt that I had to be patient while building relationships. Silence, rather than asking many curious questions, could ease the tension in a first encounter. In Johannesburg, in contrast, informality and friendliness characterised the first interactions and being verbally present was important, as by speaking people establish rapport and hierarchy. If people were suspicious or had fears, they would hide them behind talking a lot and performing friendliness. In conversations people often asked me many questions first and started to tell me things they thought I should be interested in without necessarily waiting for my questions. Such city-specific ways of relating to strangers significantly shaped the research process and I had to adapt my ways of interacting with people depending on where I interacted with them.

While communication styles and levels of social trust differed across the two cities, other local aspects which shaped the research differed across neighbourhoods and milieus. Sommerschield II and Linbro Park, where aff luent milieus were living, were both characterised by a strict temporal and spatial separation of private and public spheres exemplified by the residents' everyday rhythms of leaving for work in the morning and coming back only at night to retreat into their walled homes. Their core social relations were not so much centred on the neighbourhood but extended to other aff luent areas in the city, which they often visited by car. In these two aff luent neighbourhoods, fieldwork was largely anthropology by appointment, we rang people's door bells or called them to make an appointment, and there was little public life in the neighbourhood we could participate in. In contrast, in Polana Caniço and Alexandra, many urbanites make a living from informal economic activities, and many economic and social activities took place within the neighbourhood public spaces during the day, giving me more opportunities to observe and participate. Many residents in these neighbourhoods depended considerably on their neighbours, ranging from sharing food in times of crisis to looking after each other's children. I could appear at people's houses without appointment and I could more easily participate in their everyday life. Another reason why access and participation were in general easier in the *bairro* and the township than in the suburb and the elite neighbourhood may also be related to the power relations present in 'studying down' and 'studying up'. While the elite milieus felt entitled to say no when I asked for an interview, members of poorer milieus may not always have felt empowered to defend their privacy against what some may have seen as an intrusion by an anthropologist. In addition, people with fewer resources

may have been more interested in building a relationship with me than elites, as some were initially hoping to get access to my money or networks.

### Experiencing Cities through Elsewhere

The comparer plays a distinct role in the constitution of the knowledge through comparative ethnography. In the analytical comparison, working with data produced by others, hence *thinking cities through elsewhere*, the comparer is a potentially detached actor who, metaphorically sitting in an armchair, compares data accessible to her as text, be it raw data or written-up research produced by others. In the case of comparative ethnography conducted by a single person, engaged in *experiencing cities through elsewhere*, the comparer also has access to non-text data; the embodied and non-predicative experiences of fieldwork. Not all the data collected can be transformed into written fieldnotes. Many remain 'headnotes' (Ottenberg 1990), non-written memories of events, as well as incorporated knowledge, emotions and memories of smells and sensations, which have become inscribed in the comparer's body and which accompany her when she moves from one context to the other. These embodied aspects of research, non-written and pre-predicative memories and newly learnt habits of seeing and feeling are as much part of the data corpus and inf luence data analysis and writing, although in a different modality from the explicated data (Ottenberg 1990, Sanjek 2001: 266).

What distinguishes ethnography from other qualitative research methodologies is its focus on practice, instead of solely approaching everyday life through interviewing. Practice analysis aims to understand acts of ordinary life which are deeply embedded in habitual attitudes that actors are seldom aware of. The ethnographer can only grasp such non-predicative aspects of urban life by living there, by participating in it (Förster et al. 2011). Hence in comparative ethnography, the data and the comparer cannot neatly be separated, as the comparer-cum-fieldworker is also part of the data: fieldwork is an encounter and dialogue between two parties and ethnographic writing is a construction (Clifford and Marcus 1986, Emerson, Fretz and Shaw 2011 [1995]: 11). This also means that analysis does not start once one is back home but is a continuous process starting in the field (Strauss and Corbin 1990).

The ethnographer as a social, academic and embodied subject hence becomes shaped by the fieldwork experience in different contexts. I usually became conscious of such embodied aspects of the comparison shortly after I had travelled from one place to another. This excerpt from a field book entry was written when I had just moved to the Johannesburg suburb of Orange Grove after having lived in inner city of Maputo for five months.

I notice that all the anecdotes or examples which I bring up in small talk have to do with Maputo. "In Maputo, many people don't have hot water" (which Gaby, my host mothercould hardly believe); "in comparison toMaputo,trafficin Johannesburg is very organised", in Maputo this and that. The memories which I have in my mind and with which I compare Johannesburg in my everyday life are all related to Maputo, and not anymore to Switzerland. SometimesI try not to talk about Maputo in order not to bore people. Now living in a Johannesburg suburb, I really do miss the busyness of Maputo'sinnercity where I wasliving before, the modernist high-rise buildings, the *kizomba*  music, and even the polite reservation by strangers. Here, on the other hand, I feel less observed as a white woman on the suburban streets, people may look at me, but

they will rarely talk to me. When I go jogging, my nose doesn't get irritated by a sharp smell of pee and I don't have to look down on the concrete to avoid stumbling over the cracked cement (from my field diary, translated from the German, January 2011, living in the formerly white suburb Orange Grove, first days af ter arrival from five months' fieldwork in Maputo).

The excerpt shows how my stock of knowledge, both cognitive and embodied, had become shaped by the experience of living in Maputo and had become the lens through which I encountered Johannesburg. Instead of relating the physical experience of living in the suburb Orange Grove to living in my everyday life in my hometown of Basel I was comparing it to Maputo. My own body became part of the comparison.

Living arrangements during fieldwork in the city considerably inf luence where and with whom one spends everyday life, what kind of routines one develops and gets to know, and how one experiences the city more generally. Yet it is not always possible to choose one's living arrangements entirely in a way which one thinks would be best for research. In Maputo, I wanted first to live in Polana Caniço, yet when I finally found a room with a family I only managed to stay there for two weeks, as the hygienic conditions were difficult. I also struggled to travel from the *bairro* to the inner city, together with all the other commuters, as the minibus taxis were full and I regularly lost in the competition for a seat. So, I a rented a small place (a *depend*ê*ncia*) in the city centre, and later I lived with a middle-class, *mestiço* family in an inner-city neighbourhood where I could move around freely in the evenings as well (which I could not do in the *bairro* where there were no street lights). In Johannesburg, I lived with middle-class migrants, a Zimbabwean and a Swazi family, in the suburb of Orange Grove, and on the last field visit I rented a room at a bed and breakfast in Linbro Park.

Once in Maputo I was mugged by a homeless person armed with a rusty knife at dawn in the inner city when I went for a walk with a friend. After that event, for quite some time, my body released adrenaline when I walked past a stranger on a Maputo street when the light was fading. In Linbro Park in April 2012, there was a spike in armed burglaries, which was widely shared on the neighbourhood Google group that I was following intensely at that time. Every day I read about the previous nights' successful or attempted break-ins, which also involved shoot-outs. The bed and breakfast in which I was staying was surrounded by an electric fence and guarded by ten dogs. Despite these measures, I had some restless nights, with adrenaline again pumping in my veins. Coming from safe Switzerland, I had never been a victim of crime before even though I had travelled extensively and lived abroad, and I have no doubts that my restless nights in Linbro Park were related to my body's memories of being mugged in Maputo. This 'extreme participation' (Heer 2011) changed my view on crime and security considerably, as it enhanced my ability to understand what it means to live in cities with higher crime rates than I was used to in Switzerland.

There were, nevertheless, considerable differences across the two cities in terms of security. Official crime rates and, equally important, as Hannerz (1981) points out, the perception of danger by urban dwellers is considerably lower in Maputo than in Johannesburg. In Maputo, I did not really believe the warnings by friends that living in Polana Caniço would be too dangerous for me, while in Alexandra, it was me who did not dare to live there, even though my field assistant, Thabo, wanted me to. But I did stay at Thabo's in-laws' house for two weeks, which helped me greatly to have a glimpse

of what it meant to live in the dense living conditions typical of the township. Staying with Thabo's family at least for this short time created a sense familiarity and intimacy with them and their neighbours in the yard, so that it became a long-term setting for *'*deep hanging out' (Geertz 1998), for immersing myself in everyday life. I usually drove out of the township at dawn for security concerns, which limited my ability to participate in evening activities in the township.

My fieldwork involved many such switches from one social world to another, not only between the cities but also within the cities. In Maputo, there were days in which I woke up at the family home in Polana Caniço and in the evening I attended a fancy function in an upmarket bar. In Johannesburg, I drove daily from the quiet suburb to the bustling Alexandra township and back. Advocates of a purist form of ethnographic fieldwork would criticise that such multi-sitedness hinders deep immersion in the lifeworlds of urban dwellers from the four neighbourhoods, and this is also a doubt I had myself during fieldwork. Urban fieldworkers have long lamented the lack of a sense of a knowable social whole (Ferguson 1999: 18). With time, however, I realised that this daily switching between the different social worlds was not just an exotic practice of an anthropologist leading an unusual daily life in these cities. Indeed, I observed that this formed part of the everyday life of many of the urban dwellers as well. In Maputo, I learnt how many young women from the peripheral *bairros* dress up on the weekends and become part of the urban youth scene in fancy nightclubs. Some of my middle-class friends with university diplomas in their pockets set off to work in rural areas in the north of the country, worried about living conditions there. One of my hosts in Johannesburg was very happy to activate her networks for us to go clubbing in Soweto and Alexandra, as she was curious herself to expand her horizon in the city where she has been living for decades. Circulation (Simone 2005b), crossing everyday boundaries, is a quintessential urban practice, speaking about aspects of everyday urban life that research focusing on one public space, on one neighbourhood or on one milieu can seldom grasp.

When I moved back from staying at Thabo's in laws' house in Alexandra to the suburb of Orange Grove, I experienced something which my informants in the township had repeatedly told me in interviews and conversations, namely that they find the suburbs quiet. Back in my rented room in Orange Grove, I suddenly understood this, as I missed the noise of playing children and distant music, the smells of the neighbour's dinner terribly, and I found it absurd to have a large room all to myself, in a 150 square metre house inhabited by about four people including me. It was especially in such moments of change, of switching from one context to another, before my senses had had time to adapt to the new context, that my self was a comparative subject, experiencing Maputo against the backdrop of Johannesburg, or the suburb against the backdrop of the township. The longer I was in one place, the less my fieldnotes contained comparative remarks, and the more I dived into the realities of local living. Most ethnographers are familiar with this, as they may experience something similar during the first days after arriving in the field, when their most recent memories still concern their home country and when their attention is drawn to things because they are different from home. Yet I believe that for knowledge production, it does make a difference whether our apprehension of something as *dif ferent* results from an implicit comparison with our home country or whether it stems from an implicit comparison with another research site. This touches not only the question of the production of anthropological knowledge, but also how objects of the anthropological gaze become formed.

### **Moving Comparative Ethnography Forward**

Although anthropology was once founded on the application of comparative methodologies, since the crisis of representation and the paradigmatic shift from positivism to constructivism it has largely criticised comparison as an explicit method. Because of the resulting lack of interest in comparison, the positivist comparative methodologies have not yet been replaced by a new paradigm of interpretive comparative approaches in anthropology. There is a considerable degree of insecurity among anthropologists about what type of methodological requirements comparisons should fulfil nowadays. The revival of comparative methodologies induced by geographers like Robinson (2006) and, in the meantime, many others, should hence be embraced as an opportunity by anthropologists to dig deeper into the analytical, logistical, epistemological and methodological challenges which comparisons raise. This is because, on the one hand, anthropology is a thoroughly comparative endeavour yet has grossly neglected to consider what comparison means today, and on the other hand, because anthropology can, with its commitment to interpretivism and ref lexivity, contribute important insights to the ongoing debate on comparative urbanism.

Up to now, Simone (2004a, 2010) has been recognised as the main contributor to comparative urbanism through ethnography (Robinson 2016b). Simone, however, has engaged little with questions of the *how* of comparative ethnography, by which I mean the methodological processes of data collection, analysis and writing. The long-past yet still important debates on representation in anthropology (Marcus and Fischer 1986) drew attention to the fact that the written-up research, the ethnography as a book, is a construction of other people's construction (Geertz 1973), a complex literary and academic genre, which can be quite distinct from the ethnographic research process that preceded it and within which the knowledge written down in the book was formed. This (ethnographic) research process of comparisons has not yet received enough scholarly attention by comparative urbanists, which is why this postscript aims to raise aspects that should contribute to filling this gap.

I have argued that reducing comparative urbanism to reading across contexts ('thinking cities through elsewhere') is a disappointing turn in the development of the debate because it neglects the potential of knowledge production through *experiencing* urban life in two cities. This postscript, therefore, focuses on the methodology of comparative ethnography, not so much in terms of the written-up analysis but more in terms of comparative fieldwork. It focuses on the circular process before the written-up text, namely, on how the comparisons come into being and the hands-on practical work of conducting a comparative research project as a single ethnographer. Thinking about comparisons in terms of the entangled comparer draws attention to the importance of positionality, ref lexivity, the web of relationships and specificities of the urban context for the methodology, for the research process and knowledge production. Moving between Johannesburg and Maputo, between the four neighbourhoods, between social worlds and urban spaces, shaped the fieldwork in intricate ways, which dismantles expectations that some researchers schooled in other epistemologies may have for a solid comparative method, like the 'comparability' or 'replicability' of data. Fieldwork access, my relationships to research assistants and interviewees, my practices of immersion were deeply affected by, for example, neighbourhood-specific private–public boundaries and everyday temporal rhythms, so that not only the data but also the methodology itself became deeply shaped by the specificities of places and people. Comparative ethnography means that the entangled comparer has to constantly adapt herself to the emerging data and webs of relationships.

There are five strategies which I would like to suggest at the end of this book to deal with the dangers of comparison. One danger is the exaggeration of difference, the exoticising of other societies and the construction of units of comparisons as overly bounded, inherently consistent units with little interaction between them (Lazar 2012: 351). When arguing that "X is like this, while Y is like that" there is the inherent risk of creating dichotomies and overemphasising differences or similarities (Nader 1994: 92). In order to avoid essentialism and the construction of bounded units, anthropologists need to be conscious and responsible about comparison (Nader 1994). I suggest the following strategies: firstly, one should focus comparisons on social practices, discourses and social constructions instead of comparing 'wholes' like geographically defined places in order to avoid the trap of essentialism and boundedness (Abu-Lughod 1991). The second strategy of writing against essentialism is thick comparison. As I have argued before, comparers need to take account of the *biographies* of their units and themes which emerge in a circular rather than a linear process. Moments of disjunction, of apparent incommensurability, should be used as moments to push one's conceptualisation of what is going on further. The third strategy of avoiding the illusion of bounded units is by focusing on entanglements: Abu-Lughod (1991), Nader (1994), Robinson (2011) and others have argued for the inclusion of connections in the comparison, be they historical and contemporary, be they between the field sites to be compared, between the informants and the anthropologist, or between informants and the audience of the ethnography (Abu-Lughod 1991). The fourth strategy is to focus on differences within. As Strathern (1991) argues, differences are not only to be found between things; they are also constitutive of things and reside in them. Things are always composed of further things (Holbraad and Pedersen 2009: 374-5). Comparative ethnography hence does not always need to consists of a cross-cultural or cross-city comparison: describing similarities and differences within what we set off to compare is very important in order not to fall into the trap of bounding the cases too much. Last, but not least, as a fifth strategy, I contend that the main aim of comparative ethnography should not be to make comparative statements like "Maputo is like this and Johannesburg like that because of Z", but rather to develop descriptions which relate the case studies to each other in ways that also leave room for other interpretations, for not-yet-made comparisons and for the apprehension of the entanglements of manifold processes and complex causalities. The goal is to find a way of framing, of writing about Maputo and Johannesburg, which leaves room for the specificity of each city yet also speaks about cities in general. Rather than searching for data to fill in a pre-existing framework, the goal is to develop a framework which tries to do at least a little justice to the diversity and complexity of everyday urban life in two cities. *Cities of entanglements* hopes to have done that.

### **Bibliography**


Amin, A. and Thrift, N. (2002). *Cities: Reimagining the Urban*. Cambridge: Polity Press.


Ammering, U. and Merklein, A. (2010). "Städtische Armut in Mosambik: Einführung zur Fallstudie Maputo." In Coy, M. (ed.), *Städtische Armut im südlichen Afrika: Durban und Maputo*. Innsbruck: Geographie Innsbruck, pp. 21–34.

Anderson, A. (2005). "New African Initiated Pentecostalism and Charismatics in South Africa." *Journal of Religion in Africa* 35 (1), pp. 66–92. doi:10.1163/1570066052995843


Certeau, M. (1984). *The Practice of Everyday Life*. Berkeley: University of California Press.


Chingono, M. (1996). *The State, Violence and Development: The Political Economy of War in Mozambique, 1975-1992*. Aldershot: Avebury.

Christopher, A. (1997). "Racial Land Zoning in Urban South Africa." *Land Use Policy* 14 (4), pp. 311–323. doi:10.1016/S0264-8377(97)00025-2

Christopher, A. (2001). "Personal Apartheid." In Christopher, A. (ed.), *The Atlas of Changing South Africa.* 2nd ed. London: Routledge, pp. 139–157.


Harrison, P. (eds.), *Changing Space, Changing City: Johannesburg Af ter Apartheid*. Johannesburg: Witwatersrand University Press, pp. 2–39.


Organization. Available online at www.ilo.org/wcmsp5/groups/public/---dgrepo rts/---dcomm/---publ/documents/publication/wcms\_173363.pdf, accessed on 2 Apr. 2017.


Lemanski, C. (2012). "Everyday Human (In)Security: Rescaling for the Southern City." *Security Dialogue* 43 (1), pp. 61–78. doi:10.1177/0967010611430435


Schäfer, R. (2005). *Im Schatten der Apartheid: Frauen-Rechtsorganisationen und geschlechtsspezifische Gewalt in Südafrika*. Münster: LIT Verlag.


Schmid, C. (2005). *Stadt, Raum und Gesellschaf t: Henri Lefebvre und die Theorie der Produktion des Raumes*. Stuttgart: Steiner.


Scott, J. (1999). *Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed*. New Haven: Yale University Press.

Scott, K. (2015). *Plans Underway to Build Gauteng's R84 Billion Smart City in Modderfontein.* Available online at www.thesouthafrican.com/plans-underway-to-buildgautengs-r84-billion-smart-city-in-modderfontein/, accessed on 18 Feb. 2015.


*to the City: Mobility, Diversity and Religious Space in Johannesburg*. London: Palgrave Macmillan UK (Global Diversities), pp. 1–30.


# Social Sciences

Carlo Bordoni Interregnum Beyond Liquid Modernity

2016, 136 p., pb. 19,99 € (DE), 978-3-8376-3515-7 E-Book PDF: 17,99 € (DE), ISBN 978-3-8394-3515-1 EPUB: 17,99 € (DE), SBN 978-3-7328-3515-7

#### Iain MacKenzie

### Resistance and the Politics of Truth Foucault, Deleuze, Badiou

March 2018, 148 p., pb. 29,99 € (DE), 978-3-8376-3907-0 E-Book PDF: 26,99 € (DE), ISBN 978-3-8394-3907-4 EPUB: 26,99 € (DE), ISBN 978-3-7328-3907-0

 Franziska Meister Racism and Resistance How the Black Panthers Challenged White Supremacy

2017, 242 p., pb. 19,99 € (DE), 978-3-8376-3857-8 E-Book: 17,99 € (DE), ISBN 978-3-8394-3857-2

# Social Sciences

 Cameron Harrington, Clifford Shearing Security in the Anthropocene Reflections on Safety and Care

2017, 196 p., hardcover 79,99 € (DE), 978-3-8376-3337-5 E-Book: 79,99 € (DE), ISBN 978-3-8394-3337-9

 Benjamin Heim Shepard, Mark J. Noonan Brooklyn Tides The Fall and Rise of a Global Borough

February 2018, 284 p., pb. 29,99 € (DE), 978-3-8376-3867-7 E-Book: 26,99 € (DE), ISBN 978-3-8394-3867-1

 Ilker Ataç, Gerda Heck, Sabine Hess, Zeynep Kasli, Philipp Ratfisch, Cavidan Soykan, Bediz Yilmaz (eds.) movements. Journal for Critical Migration and Border Regime Studies Vol. 3, Issue 2/2017: Turkey's Changing Migration Regime

2017, 230 p., pb. 24,99 € (DE), 978-3-8376-3719-9

and its Global and Regional Dynamics

All print, e-book and open access versions of the titles in our list are available in our online shop www.transcript-verlag.de/en!