Abstract
Introduction
In 2010, Salomon
1
decided to establish a new church. Born in the South African province of KwaZulu Natal and later relocating to the township of Khayelitsha on the outskirts of Cape Town, Salomon had been serving as a pastor in a Pentecostal church called the “Apostolic Faith Mission” until he felt he was mature enough to found his own ministry, which he named “Restoration Center.” “First, we didn’t have a place of worship, so we were using somebody’s dining room,” he told me. However, as the number of followers kept growing, they moved on to a nearby garage and from there to a former clinic, which was owned by the South African National Civic Organization (SANCO). They signed a lease with SANCO which allowed them to use the premises for free. Upon moving in, they initially had to refurbish the building, as it had been severely damaged when burglars broke into the house and pillaged it. Much to the frustration of Salomon and his congregation, the
This brief description offers an astonishing contrast to how scholars usually imagine the spatial forms and materializations of religion in contemporary cities. In anthropological and sociological studies, contemporary urban religion is predominantly conceived as materialized and made publicly visible through architecture and aesthetics (Knott et al., 2016), spatially embedded through practices of place-making (Garbin, 2012; Vásquez & Knott, 2014), and globally networked via communication technologies. The idea is that religious life acquires a more or less enduring presence in urban space through architectural registration (Knowles, 2013), which reinforces people’s affective place attachments, and that the religious practices for which places of worship provide material backdrops are platforms for launching claims to urban citizenship. Religious place-making is thus viewed as “the dynamic appropriation of portions of the cityscape” (Garbin, 2013, p. 678) through worship and the building of places of worship.
What the fate of the “Restoration Center” illustrates, instead, is how religious communities’ inability to produce localities by wresting space from surrounding illicit economies can turn the notion of place into something highly elusive. It is not only that it shows the failure to appropriate urban space and produce proprietary locality but also that such failures lead to the creation of specific forms of infrastructural provision as alternative spatializations of religion. It is the conditions under which such forms emerge, as well as their outcomes, that I explore in this article. Significantly, as the insecurities of life in Cape Town’s townships engendered new forms of urban nomadism for religious communities, spatial attachments evaporated and architectural symbolism was virtually eclipsed. While Sunday services filled the hall of the “Restoration Center” with spiritual intensity, effervescent singing, prayers, and other collective socialities, on most other days the hall remained quiet and Cape Town’s notorious winds, howling through the open doors, would produce the only noises to be heard.
In this article, I develop the idea of
This article is based on ethnographic fieldwork carried out in the South African city of Cape Town between 2015 and 2020. Concentrating on the spatial dynamics in the field of charismatic and Pentecostal Christianity, my study focused on the township of Khayelitsha, located about 25 km from the city center and characterized by high levels of unemployment, informality, and social insecurity. My fieldwork involved participation in all aspects of collective religious life, in the everyday life of pastors and their interactions with parishioners and residents, as well as qualitative in-depth interviews with them focused on issues of building, urban space and religion. All of my interlocutors were pastors and members of small, independent congregations with fluctuating memberships between 20 and 100 people. All of the pastors were inspired by the ideal of making a living from their pastoral work but none of them succeeded in that. I begin by briefly reviewing the anthropological debate on urban religion to specify the theoretical novelty of the notion of infrastructuring religion. The rest of the article seeks to ethnographically illuminate the lessons of this approach for theories of space, urban sociology, and the anthropology of religion.
The Spaces of Urban Religion: Place-Making, Architecture and Infrastructure
Inspired by broader social science debates around space and public contestations over places of worship in many parts of the world, there has been a rising scholarly interest in the spatial dynamics of urban religion, its regulation, aesthetic modalities, and affective forms (Berking et al., 2018; Garbin & Strhan, 2017; Becci et al. 2013). Two issues have been central in this context: religious architectures and practices of place-making. Religious architectures imbue places of worship with particular designs and aesthetics and are often seen as affording religious communities visibility in public space (Becci et al., 2017; Verkaaik, 2013). If, as urban sociologist Steets (2016) astutely argued, architecture is a social construction and material objectivation that externalizes elements of the social order, religious buildings become legible as affirmations or contestations of that order, and as ways to articulate cultural hierarchies in urban space. By becoming iconic, religious buildings make present not only the sacred (Knott et al., 2016) but also the symbolic power it accrues and which particular material forms are able to wield.
While much of the debate has centered on places of immigrant religions, especially mosque and temple constructions in the West, African cities have also seen a surge of new religious architectures. Massive Christian prayer camps (Ukah, 2016) and megachurches (Katsaura, in this issue) in West Africa reveal how religious architectures serve as urban signifiers and expressions of power and how their articulations with real estate industries and consumerism reverberate with visions of urban space in neoliberal modernity. However, in the shadows of these iconic projects, an alternative spatialization of religion is manifest in the numerous, physically fragile places of worship that dot the sprawling urban and peri-urban settlements across the continent. If religious architecture and religious buildings are culturally powerful because they are “time-stamped” (Brenneman & Miller, 2016, p. 89), places of worship in South African townships—as I seek to show in this article—must be understood through a different lens, one that accounts for their divergent temporality and their distinct urban embeddedness. While the linking of profane materialities and artifacts into assemblages which enable religious life—which I call “infrastructuring”—occurs in contexts across the world, it has a particular significance in cities in which “urbanity exists beyond its architecture” (De Boeck & Plissart, 2014, p. 233). Importantly, while infrastructures are often associated with large, state-sponsored projects, I focus on material artifacts produced by ordinary citizens in ways others have called “autoconstruction” (Caldeira, 2017)
As cities are, in an important sense, material assemblages made up of heterogeneous artifacts, technologies and practices (Farías & Bender, 2012; Burchardt & Höhne, 2015), research on urban religion has allowed scholars to address vexed questions about the role of materiality in religious life in novel ways and take theoretical ideas from the study of material culture on board (Knott et al., 2016). Significantly, material objects of all sorts are central to the politics of belonging, especially in rapidly urbanizing and ethnically diverse contexts. Conceptual building blocks such as belonging, space, materiality, and visibility have provided the basis for the notion of religious place-making, which Garbin (2012, p. 410) defines as “the appropriation and experiencing of space through various religious activities.”
While surely helpful for addressing the nexus of space and contested religious and social boundaries, this notion seems less adequate for capturing the religious dynamics in contexts where affective attachments to places are of lesser significance than the (often inchoate) cobbling together of artifacts that enable religious assembly. In brief, as modes of the spatialization of religion place-making refers to places as more or less durable, collectively meaningful and identifiable (Tuan, 1977) while architecture points to the aesthetic codes of the built religious environment, which are inherently expressive and affective. Both allow communities to develop place-attachments on the basis of which they imbue localities with religious meaning. Religious infrastructures, by contrast, as I describe them here, are sets or networks of functional artifacts which enable people to create localities for religious life even if the precarious circumstances of township life prevent them from becoming durable, collectively meaningful and identifiable. Importantly, the quality of religious infrastructures is closely linked to the ability of religious entrepreneurs to succeed and grow in South Africa’s highly competitive religious field. While financially well-resourced challengers such as the Universal Church of the Kingdom of God are able to turn elaborated infrastructures
To shed light on these dynamics, I draw on insights from the recent “infrastructural turn” in urban studies (Björkman, 2015; Graham & McFarlane, 2015) and aim to demonstrate the usefulness of infrastructural perspectives for the sociology of urban religion. Comprising the socio-technical apparatuses and material artifacts that structure, enable, and govern circulation—specifically the circulation of energy, information, goods, and capital, but also of people, practices, and images in the urban realm and beyond—infrastructures are a fundamental part of urban social orders. Infrastructures are typically characterized by different degrees of access and therefore mediate both integration and disruption in urban life on the basis of unequal distributions of power. Moreover, infrastructures shape and to some extent produce human subjectivities, habits, and socialities. Thus far, however, questions as to how these infrastructural dynamics affect and shape religious life have largely been left unaddressed. I now turn to these questions, focusing on infrastructuring as a collective process “in the making” rather than a ready-made outcome. My central concern is how infrastructure is turned into a modality of spatialization, one that partially supplants architecture and place-making.
Mobile Infrastructures: Tent Revivals
Practices of infrastructuring religion in Cape Town often blur the boundaries between the permanent and the ephemeral. This is particularly evident in the use of tents as places of worship (Burchardt, 2020). The “tent revival” has been a major element in the repertoire of evangelical Christianity since America’s “Second Great Awakening” in the early-to-middle 19th century and embodies a spatial form that emphasizes “placelessness”—in other words, the notion that people who have heard and responded to the call of the Holy Spirit come together in an improvised, non-permanent place to worship God. Among South African Pentecostals, tents are a central architectural feature of intense public prayer sessions called “tent crusades.” Aiming to spread the word of God, to amplify the power of the Holy Spirit and make it forcefully present in urban space, and to win followers in often highly dramatic rituals of conversion, tent crusades are organized by established enthusiastic pastors who wish to enlarge their congregations and accumulate symbolic capital through the performance of miracles or other acts that testify to their ability to summon the Holy Spirit. More often, such crusades are carried out by younger pastors who do not yet have a church or congregation but aspire to create them. Before a tent revival begins, pastors engage in door-to-door campaigns. As Mandla, one of my informants, told me: “I go from house to house and tell people how I witnessed Jesus and the miracles he did in my life. And I always pray with them. And then they come to the crusade.”
Significantly, to be successful, tent revivals depend on infrastructures and on access to sites, transport, and equipment. While getting authorization from the city council for using urban space for a tent revival is obligatory, crusade organizers also need to talk with residents to win their approval. In Khayelitsha, failure to do so in the past has led several times to residents attacking or looting the tents. One pastor stated bluntly, “If I do the crusade without asking, they will come and destroy the tent.” Whereas many pastors rent such tents, some such as Mandla decide to purchase them. As he explained to me, “Now I have a three-pole tent and I also had to buy the chairs and the instruments. But I had to save money for that.” Yet as the tents are not always in use, there is a need to organize storage room for them as well as for the chairs, usually in shipping containers. In addition, campaigners need to rent trucks to transport their equipment and gain access to electricity for light and sound systems.
In 2008, another pastor I call Dumisani founded his own ministry after, as he told me, his calling by Jesus Christ had been confirmed several times. Together with six others, he held meetings in his family home in a section called Fulani. As the group grew, they decided to look for a bigger place and were allowed to use a nursery on Sundays for their services. By 2010, their numbers had grown to over 70 congregants and they could no longer use this place. Instead, they bought a tent of 9 × 18 m which could accommodate 180 people and pitched it on an empty site after getting permission from the municipality to do so.
However, there were two major problems: First, they could not use the tent during the winter because of the harsh climate. Second and more dramatically, they were unable to keep thieves away from the tent. Several times, thieves broke into the tent and stole parts of it, particularly the poles on which it rested. Dumisani and his friends soon figured out that the thieves had sold them at a local scrapyard, where their object biography would presumably continue with them serving as building materials in another makeshift assemblage of artifacts. Dumisani conjectured that the owners of the scrapyard might even have sent the thieves to get the poles. In the end, he was nonetheless unable to prove that the poles on offer there actually belonged to his tent. In this case, infrastructuring religion turned out to be a fragile practice, as the artifacts that were meant to physically stabilize religious practices themselves began to circulate in illicit township economies.
Securing Property
Infrastructuration is about creating the material basis upon which religious life can begin to take place. Importantly, the concrete material shapes of places of worship are linked to the broader material and physical landscapes of township life. In Khayelitsha, the numerically most significant type of places of worship are small Pentecostal churches. While the ways in which such churches are theologically envisioned and organizationally consolidated have been explored in a number of anthropological studies (e.g., Englund, 2003; Burchardt, 2017), how they are built and thus crystallize materially has yet to be examined. Pastor Melisizwe Maqogi, whom I got to know during my first field research in Khayelitsha in 2006, organized his Sunday services and other religious and social activities in a small community hall on a regular basis. Owned by the municipality, the place was in theory open to all residents of Town Two, the section of the township in which Melisizwe lived, but his occupation of the place was more or less permanent. In fact, when I began to do fieldwork with Melisizwe, he introduced the hall to me as “his church.” However, he was already aiming at the time to purchase a plot of land to build a new and larger place of worship, one in which he would not be a tenant of sorts and which would reflect his visions for the church—namely to grow. As elsewhere in the world, “church growth” is a major theological tenet among South African Pentecostals, one that has been worked out in numerous almost doctrinal statements and theories.
However, Melisizwe faced continuous challenges in finding a plot of land at a price that he and his congregation, including a few members who had some money, would be able to afford. In addition, as a buyer he had to make sure that those who sold the land were in fact its rightful owners. Throughout the 1980s and 1990s, urban expansion in Cape Town had created an urban landscape in which the pattern of ownership was almost illegible, replete as it was with parallel legal titles—claims made in the name of one community or another—which in the end were often enforced by self-appointed community leaders who acted as slumlords. Once the congregation was eventually able to purchase a plot, they began construction work with proper construction materials, as that was central to their vision.
Nevertheless, many of the Pentecostal churches in Khayelitsha, in particular in the informal settlements that continue to sprawl, were constructed out of used leftover materials: pieces of wood, fiber boards, corrugated iron, zinc, wood piles, pieces of plastic, and canvas covers of all kinds (see Figure 1). Such objects were sometimes found by congregation members somewhere in the open veld as leftovers from abandoned construction sites or abandoned buildings. More commonly, these materials were sold by petty rubbish entrepreneurs who made a living out of collecting, trading, and assembling these parts or who established construction companies, knowing that demand always exceeded supply. Interestingly, though perhaps not surprisingly, these were the same kinds of materials newly arriving residents of informal settlements also used to build their first small houses, typically shacks that provided little protection against the cold weather, fierce winds and heavy rainfalls of Cape Town’s winter months. In other words, church constructions strongly resembled residential homes and other buildings in these areas, and from the outside it was usually nearly impossible to tell whether a building was a home or a church.

Church Interior, Photograph Taken the by Author March 1, 2020.
This resemblance, however, was not only a reflection of material scarcity and necessity. It also reflected the multifunctional nature of these places. As with the building Melisizwe used as a church, many buildings served as a place of worship at some point, but also as a social center, a nursery, or a community hall at some other point in time. In a way, material set-ups that allow for multifunctionality are reminiscent of the ways in which certain infrastructures are generic background structures that allow for the implementation of some other, more particular kind of practice (Larkin, 2013).
In general, the changing use of these places also reflects the erratic nature of life in the townships, in which people are forced to take up any of the constantly shifting opportunities for income that may present themselves. What Robins (2002) calls the “hypermobility” and “domestic fluidity” of Cape Town’s poor also contributed to engendering situations in which church congregations would dwindle or even dissolve within a matter of weeks. Melisizwe’s new church building was never finished, as he and his group of sponsors were unable to gather enough money. His congregation shrank over the years as members drifted toward competing communities; his congregants either bought into the promises of the Brazilian Universal Church of the Kingdom of God, 2 which opened a massive new church building very close to his neighborhood to propagate the “prosperity gospel,” or else began to follow one of the evangelists he called “mushroom pastors”—pastors with no theological credentials whatsoever but with the ability to attract followers on the basis of their capacity to do miracles with which the Holy Spirit had presumably blessed them. Some pastors would therefore have to abandon places of worship, having lost their flock, while other pastors would need to accommodate more and more followers.
One other major reason for the ephemerality of these places of worship was that, unlike Melisizwe, many pastors were unable to acquire socially valid claims to ownership or property over their places. Matthew, a close friend of Melisizwe, opened his own ministry in 1986 when he broke away from the Independent Presbyterian Church after moving to Khayelitsha. Having assembled a small but growing community, he held worship sessions in his own dining room. When his house could not accommodate the growing numbers of followers anymore, they moved to a nearby shipping container that belonged to a defunct nongovernmental organization (NGO) and which no one else seemed to claim. However, after they had installed windows and decorated the place, thieves lifted the entire container on a truck and stole it in a nighttime cloak-and-dagger operation. Having lost their place overnight, the community began to rent a hall in a nearby primary school for its Sunday services. When the school administration raised the rents soon after, Matthew decided to construct a church building using the concrete slabs surrounding the area on which the shipping container once stood and adding a zinc roof. When within less than a month’s time the wired fences he had added to secure the plot were stolen and people began dumping rubbish and defecating in front of the building, Matthew grew increasingly frustrated. He therefore reactivated a plan to build a church on another plot about 7 km away from his home, which he had purchased and paid off several years beforehand.
In the meantime, however, members of another church—the Zionist Christian Church, one of South Africa’s biggest churches—had erected a shack church on the unused site. Unable to access his property, Matthew sought the support of the municipality. With the help of the sheriff of the court and in the presence of police officers, municipal employees demolished the Zionists’ shack. The situation quickly escalated as the Zionists rebuilt their shack and threatened to kill Matthew if he continued to assert his claims. With the support of a municipal clerk, they produced their own counterfeit title deed, the written content of which obliged the municipality to secure an alternative plot if the Zionists’ presence on Matthew’s land was ultimately deemed illegal. After lengthy discussions, the Zionists withdrew their claims to the plot and it seemed imperative to quickly create material evidence that Matthew’s community used and owned the land to signal their property claims to local residents. With the help of his congregants, in one night and one day Matthew fenced off the plot and built a church on it, without any architectural plan or plant layout. However, the second removal of the Zionists’ shack and Matthew’s construction of his church happened so quickly that he neglected to notify the municipality. In an ironic turn of events, the sheriff returned to the plot days later and, once again, marked the newly constructed church as ready to be demolished by the municipal caterpillar (see Figure 2). Aghast, Matthew called the sheriff and prevented the destruction. Finally, his title deed had become functional.

Matthew’s Church, Photograph Taken by the Author March 1, 2020.
Many of the improvised church buildings in Cape Town’s townships and informal settlements are not linked to the urban electricity grid and water pipelines. Often, they use extension cables to get electric power from adjacent residents. “You have to ask the neighbors to help you,” one pastor told me. “They charge us 100 Rand per week and we pay that for through the tithes.” In an important sense, electricity is the infrastructural

Worship at Melisizwe’s Church, Photograph Taken by the Author March 8, 2020.
Matthew’s community also got electricity through an extension cable, but unlike others, their plot was linked to the water pipeline. After establishing the church building, they approached the municipal water authority to solicit the connection. A plumber was sent to check the pipes and link the toilet pipes and faucets. It was only afterward that Matthew realized that they had been required to register their land as “church land” prior to the plumber’s visit to enjoy a discount on that service. Faced with a bill of over 26,000 Rand, Matthew’s community was unable to pay and the water authority denied them water provision. As a result, their dream of having their own clean toilets and to thereby enhance their own sense of decency and respectability (Ross, 2009) remained unfulfilled.
Praying in Ruins
While Matthew and his community still clung to their dream of a fully equipped place of worship, articulated both in terms of architecture and infrastructure, other Pentecostal communities were shifting toward a spatial strategy that relied on improvised infrastructure alone, one which suspended place attachments of any kind. Salomon, whom I introduced at the beginning of this article, was initially faced with the choice between using a substandard place for free or renting a site from SANCO, the latter of which appeared slightly superior in terms of building materials and hygiene. Early on, he estimated that a private security guard was needed to keep burglars and looters away from the building, but when SANCO refused to pay for security services he opted for the free site. However, the degree of destruction produced by the first burglaries left him puzzled. When we visited the place together, he said, “Look, they vandalized the whole campus. They vandalized it like hell. I cannot even call it a break-in, that would be a lie. Because if you break into a house you break the locker, take things and go out. But they took everything. I don’t even know what to call it.”
Visiting the place, it indeed appeared as if it had been targeted by a military attack. Criminal gangs had broken into the building and, as mentioned above, took with them not only the sound system, chairs, and all other moveable items, but also toilets, sinks, floor coverings, and some doors. Ripping open wool boards that covered walls and ceilings, they had left behind a swath of destruction (see Figure 4). Walking through the building, Matthew called my attention to the windows where the

Looted Church, Photograph Taken by the Author March 5, 2020.
The level of physical insecurity to which Salomon’s church was subjected had important consequences for the ways they used infrastructure to enable religious life. Repairing the damages did not seem a promising idea to the community because in their view, it would simply attract new burglaries. Instead, they chose to leave the place without doors and forego any sort of investment into permanent infrastructure, for example/such as access to water and electricity. What was to enable their collective worship was an ensemble of mobile equipment, including a sound system, chairs, a portable pulpit, carpets, and even curtains covered in plastic flowers, which they transported between Salomon’s home and the “Restoration Center” on a trailer he had purchased for that purpose.
Conclusion: Religion, Infrastructure, and Urban Nomadism
As we have seen, as a practice of assembling profane materialities, infrastructuring religion is shaped by both the spatial strategies of religious communities and the city’s spatial regimes that operate as conditions of possibility. I wish to pinpoint four central lessons of my analysis for theories of space, urban sociology, and the anthropology of religion. First, while infrastructure is often viewed as pivoting on large-scale, state-driven projects, infrastructuring religion is often premised on the ingenuity of grassroots actors and their ability to improvise when mobilizing both modular materials and human capacities (Simone, 2004). As the practices of the “Restoration Center” illuminated, infrastructuring “from below” reckons with unstable conditions. These actors actually turned infrastructure into a form of logistics in which the necessary was assembled into a mobile toolkit to be used elsewhere if need be. This logistics allowed religious communities to develop new forms of urban nomadism. Second, if religious architecture is chiefly about sacralizing the material environment, infrastructure is about enabling religious assembly where the presence of the divine is not mediated through material portals but through religious technologies, such as prayer. The portability of the Holy Spirit that is seen to enhance the global spread of Charismatic Christianity (Vásquez, 2009) powerfully resonates with this kind of local logistics. Third, practices of infrastructuring religion are borne from the way in which resources such as urban land and building materials are claimed by others. Shifting power relations, criminal economies, and unstable claims to property are all part of a spatial regime that privileges flexible infrastructures over religious place-making.
Finally, I argue that the practices of infrastructuring religion that I explored in this article are not a mere shadow of more formal, permanent, and aesthetically embellished modes of spatializing religion, one of the many “urban shadows” which, according to McFarlane (2008), exist on the edges of urban theory. In fact, I suggest that attention to the infrastructuring of religion articulates and contributes to what Robinson (2002) has called theorizations of ordinary cities. Involving “dynamic economic activities, popular culture, innovations in urban governance and the creative production of diverse forms of urbanism” (Robinson, 2002, p. 540), infrastructural practices around religion are indeed manifestations of ordinary urbanism and in need of further comparative exploration.
