Flags and Shields: Muslim Socialities and Informal Livelihoods in Dar es Salaam

In this article, I explore entanglements of Muslim sociality and informal livelihoods in Dar es Salaam. I propose that Muslim practices of affiliation deserve to be counted among the ordinary strategies that urban majorities employ to navigate the city amidst conditions of uncertainty. Building on Stephan Lanz and Martijn Oosterbaan’s (2016) work on “entrepreneurial religion,” I draw on fifteen months of ethnographic research in Kariakoo, a super-dense market district at the center of the city. The first part of my analysis explores associational life in Dar es Salaam, combining a historical overview with ethnographic description. I explore how urban actors use affiliations to football teams and political parties to generate livelihood opportunities and to secure protection. The second part of my analysis uses emic notions of “flags” (bendera) and “shields” (ngao) to demonstrate that Muslim modes of association operate in strikingly similar ways to those pertaining to football and politics. Engaging with the conceptual vocabu-laries of AbdouMaliq Simone and Erving Goffman, as well as recent debates about “everyday” and “lived Islam,” this article foregrounds the interplay between religious forms and neoliberal processes of informalization as they are unfolding in African cities and beyond.[Tanzania; Dar es Salaam; Urban Religion; Muslim Urbanism; Informal Economy; Associational Life.]


Introduction
T his article contributes to an emerging body of interdisciplinary work on urban religion and religious urbanism (Hancock & Srinivas 2008;Becker et al. 2013;Garbin & Strhan 2017;Berking et al. 2018;Kuppinger 2019). Stephan Lanz and Martijn Oosterbaan (2016) have recently made a significant contribution to this research agenda with their innovative notion of "entrepreneurial religion," which they propose has flourished as part of the "age of neoliberal urbanism" that we currently inhabit. Lanz and Oosterbaan (2016, 494) conceptualize Pentecostal Christianity and "middle-class" post-Islamism as entrepreneurial religions, meaning that they both can be seen to generate "paradigmatic neoliberal forms of organization." 1 More specifically, each of these movements is characterized by a "radically free-market organizational structure," which, the authors argue, maximizes the "entrepreneurial self-government" and security of their adherents (Lanz & Oosterbaan 2016, 497).
While their article marks an important contribution to this emerging literature, the symposium that it introduces restricts the scope of what Lanz and Oosterbaan (2016, 487, 497) call the "metropolitan religious mainstream" to this pair of explicit movements that have "broken away from their traditional cultural environment." This is not to downplay the tremendous influence on urban worlds exercised by Pentecostal Christianity and post-Islamism, as their rapid global expansion continues apace; however, it strikes me that Lanz and Oosterbaan's intervention can also help explain emerging urban-religious phenomena in southern cities that do not fall within the purview of Born Again Christianity or Islamic reform, but that are no less shaped by, or constitutive of, the social and regulatory landscape of neoliberal urbanism.
In this article, I investigate a metropolitan Muslim mainstream in Dar es Salaam that shares little in common with middle-class post-Islamism. This is a very different kind of entrepreneurial religion-one that gives rise to a far more implicit and ephemeral type of religious organization. My analysis draws on fifteen months of ethnographic fieldwork in Dar es Salaam, the entirety of which I spent living in Kariakoo, a super-dense market district at the center of the city ( Figure 1). The data that appears in this article derives from my conversations with individual residents, my use of walking as a participatory research method, and my interactions at street-side gatherings called maskani, where men congregate to discuss current affairs and consume refreshments.
There are two parts to the analysis. In the first, I examine the changing character of associational life in Dar es Salaam, combining a historical account of its transformation from colonial era to present day, with an ethnographic description of some prevailing modes of sociality in Kariakoo. Here I focus especially on the cohesions generated by football 558 associations and political parties, attending to the political and economic shifts that have structured the transformation of these social formations. Bringing these two threads together, I then provide a concrete account of the social and economic productivity of these distinctive modes of sociality in Kariakoo, a neighborhood characterized by a notable concentration of informal livelihood activities. My particular interest in association football (mpira) is not only reflective of its social vitality in Kariakoo, but also the ways that residents constructed me as a mzungu (white person), and more specifically as a white British man. The modalities of whiteness and foreignness that I encountered in Dar es Salaam made it difficult for me to talk directly with residents about religious matters but also they led people to (correctly) infer that I am interested in football. Over time, I came to appreciate how football could serve as an unexpected entry point through which to better understand contemporary dynamics of religious sociality in Dar es Salaam. Accordingly, in the second part of this article, I stage parallels between these non-religious forms of belonging and popular Muslim associational forms. What these comparisons are intended to demonstrate is that religious modes of sociality have undergone a similar process of transformation to those associated with non-religious markers of solidarity. Using two emic terms, namely "flags" (bendera) and "shields" (ngao), I provide a series of concrete examples that illustrate how, when taken seriously as everyday urban practices, Muslim modes of association operate in strikingly similar ways to those pertaining to football and politics. I demonstrate that Muslim practices of devotion and affiliation deserve to be counted among the informal livelihood strategies by which inhabitants "navigate the city"-how residents "cope, work together, deal with threats and develop opportunities, and invest their energies in the making of urban life" (McFarlane & Silver 2017, 458).
In doing so, I draw on three different research trajectories in addition to that of urban religion introduced above. First, I build on the innovative work of AbdouMaliq Simone (2001Simone ( , 2004Simone ( , 2008Simone ( , 2010 on Muslim urbanisms in districts comparable to Kariakoo in Abidjan, Côte d'Ivoire, and Douala, Cameroon. Second, I take inspiration from a recent turn toward the "everyday" and the "lived" in the anthropology of Islam (Deeb 2015;Marsden & Retsikas 2013;Schielke 2010Schielke , 2018Soares & Osella 2010). This research agenda conceptualizes the everyday as a site of contingency where "ethical strivings" informed by normative models of Islam encounter mundane practices, and where Islam rubs up against other "grand schemes" within people's wider life-worlds (Deeb 2015;Marsden & Retsikas 2013;Schielke 2010Schielke , 2018Soares & Osella 2010). Finally, I borrow from the conceptual vocabulary developed by Erving Goffman (1959Goffman ( , 1963Goffman ( , 1971 to explain the microdynamics of everyday encounters among American and British people in the twentieth century. Goffman's manifest influence on recent writing about urban Africa has rarely been made explicit. The resulting analysis challenges existing research on urban religion and informal livelihoods, which has in both cases overlooked the interplay between religious configurations and neoliberal processes of informalization as they are unfolding in African cities and beyond. Flag Neighbourhood E arly on in my time in Dar es Salaam, I bought two pairs of brightly colored Adidas tracksuit bottoms from a street vendor. It quickly became clear that the color combinations that I had chosen-red and white on one hand, yellow and green on the other-are in fact associated with Simba Sports Club (Simba) and New Young Africans Sports Club (Yanga) respectively, the two most popular football clubs in Tanzania who are also "historic rivals" (watani wa jadi). By wearing both pairs of tracksuit bottoms on different days, I was inadvertently displaying my allegiance to both Simba and Yanga. Some of my acquaintances at maskani began to make innuendo-laden jokes about my fluctuating between sides. Eventually, having committed to taking Simba as my team, I felt the need to retire the green-and-yellow pair.
This anecdote is indicative of the tremendous impact that bright primary and secondary colors have on Kariakoo as a visual environment. More than any other neighborhood in Dar es Salaam, the color combinations that I chose when I bought my tracksuit bottoms (and to a lesser extent blue and purple) can be found on every surface imaginable. Banners bearing the names or logos of football clubs hang between buildings; restaurants decorate their walls with posters and framed pictures of players and mascots; stickers adorn shop frames, windows, and vehicles of all sizes. These colors have also come to represent European teams whose profiles have grown enormously in Dar es Salaam in recent decades. For example, red and white are representative not only of Simba but also the English football clubs Manchester United, Arsenal, and Liverpool. At one major intersection of Kariakoo, two brothers own shops facing one another, both of which are painted red: one is decorated with the Manchester United logo and has above its doors the name of their stadium, Old Trafford, with the other featuring the Liverpool crest and the name of their ground, Anfield. In more recent years, a relative of the brothers has opened a store nearby with the words "Man City" written over the door frame.
These bright colors are an integral element of what Peter Lambertz (2015, 42) calls an "infrastructure of distinction." Urban actors use these colors to performatively draw material lines of social difference between humans and non-human objects (Lambertz 2015, 42). The anecdote about my tracksuit faux pas illustrates how color, as a sensory surface, performed difference and distinction on my body. In other words, the act of wearing a color provides a very simple way by which an individual can embody aesthetic difference. My red tracksuit bottoms, in the words of my informants, were my flag (bendera), and so it might be said that urban actors in Kariakoo become flags through such practices. In the months preceding the October 2015 presidential election, even though effectively the same color combinations were in use, they began doing something different: they now performed a material line of distinction along political affiliations. It became perfectly commonplace for a Simba fan to display their support for Chama Cha Mapinduzi (CCM-Tanzania's dominant ruling party) by wearing green-yellow, or indeed for a Yanga fan to display their support for the primary opposition parties, Chadema and the Civic United Front (CUF), by wearing red, white, and blue. While this would occasionally lead to some confusion about precisely what kind of flag a given individual was, it did not deter people from displaying their colors. Where I observed that women are generally less likely to wear football colors than men, this was most certainly not the case with respect to political colors, with many women buying dera gowns emblazoned with party colors and logos from Kariakoo shops. Street vendors experienced a roaring trade in what resembled a conference ID pass containing an image of the wearer's favored presidential candidate. The "pass" was strung on a lanyard that would correspond to the party's colors and would be worn around one's neck.
It was particularly notable at this time that CCM flags could be found on virtually every surface, from shop fronts to vehicles to garments. My research collaborators explained that these flags were not necessarily indicative of voting preference, serving rather as shields (ngao) that were protecting individuals or spaces from unwanted attention or suspicion on the part of the dominant ruling party. The logic here is that law enforcement agents who may target the homes, shops, and stalls of residents are less likely to do so if the inhabitant or proprietor is a supporter of the ruling party. Accordingly, CCM colors were emplaced in a more "fixed" fashion than those of the opposition parties, which were somewhat "itinerant." For example, motorcycle taxi (bodaboda) drivers would parade through the streets flying Chadema flags behind them, sounding their horns as they passed.
Before addressing the social productivity of this infrastructure of distinction as it takes place in Dar es Salaam, I first want to contextualize its emergence with reference to the changing character of associational life in Tanzania over the past century. As the birthplace of both Simba and Yanga football clubs, as well as the Tanganyika National Union (TANU), which later became CCM, I focus especially on the Kariakoo neighborhood.
Associational Life I n the first half of the twentieth century, urban centers across East Africa were undergoing rapid population growth as a result of rural-urban migration. At this time, Dar es Salaam experienced a flourishing of associational life centered around the Kariakoo neighborhood (Callaci 2017, 117). In accordance with racially segregated programmes of colonial planning, Kariakoo was, during these years, the principal residential neighborhood for the city's African population (Brennan & Burton 2007). There emerged what David Anthony (1983, 156) calls a "distinctly urban consciousness," which "conceived the world in terms of dance societies, football associations, trade unions, [and] political groupings"; a list to which could be added Sufi tarika, churches, and "tribal" associations. The British colonial administration tolerated such groups because they were seen as an important means of integrating rural migrants into urban life, of governing so-called wahuni (hooligans), and of delivering social services (Mercer et al. 2008, 112). In addition to being social clubs where diverse people could exchange news and participate in discussions, these organizations typically had a formal structure and management committee: they "collected subscriptions, organized burials, helped members to find jobs, accommodation, and even arranged marriages" (Brennan 2012, 66-67; see also Ranger 1975, 93;Tsuruta 2007, 201). In the decades leading up to independence, associations as diverse as trade unions, dance societies, football clubs, and Sufi tarika were instrumental to the rise of the Tanganyika African Association (TAA) and its successor, the Tanganyika National Union (TANU) (Nimtz 1980;Askew 2002, 95-96, 100;Tsuruta 2007, 203-204). These organizations were used to great effect to mobilize Tanganyikans across ethnic divides, to raise funds, and to nurture new leaders (Glassman 1995, 165;Geiger 1997, 58;Said 1998, 183-185;Askew 2002, 95-96). Associations also provided an effective cover for political meetings at a time when the colonial government was resisting such movements (Said 1998, 184).
From the end of British rule in 1961 and up to the present day, the formal associational life of Dar es Salaam, and Tanzania more broadly, has diminished considerably. There are two principal reasons for this. The first has to do with the onset of state socialism in Tanzania and its legacy. Before the transition to independence, TANU had endeavoured to capture the energies and "moral obligations" of associations, and to redirect them toward the cause of national unity (Bjerk 2015, 49). Following an overwhelming electoral victory in 1960, TANU increasingly sought to restrict the autonomy and activity of associations as a means of consolidating power and monopolizing "economic, political and social space" (Dill 2009, 722). It did this using a variety of techniques, "stripping some groups of their power, excluding others from politics and folding others into the state apparatus" (Paget 2018, 11). In the case of Christian and Muslim organizations, TANU warned against the dangers of mixing religion and politics in order to discourage any forms of collective political participation based on religious identity (Westerlund 1980). The eventual demise of state socialism in Tanzania saw the emergence of a neoliberal phase in the mid-1980s following an internationally administered programme of political and economic liberalization (mageuzi). Yet despite its return to a multi-party political system, Tanzania has not witnessed the kind of resurgence of popular participation in associational activity that many expected to accompany these liberalizing reforms. In other African countries, social actors "endured at the periphery [of partisan politics] and returned with their mobilizational powers intact" to oversee the transition to multi-partyism (Paget 2018, 13). In Tanzania, however, social actors that were not permanently eliminated remained "mobilizationally diminished" or "excluded by norms of political neutrality," leaving TANU to control the transition to multi-partyism (Paget 2018, 13). While Dar es Salaam has in recent decades seen a proliferation of non-governmental organizations (NGOs), community-based organizations (CBOs), and faith-based organizations (FBOs), these have typically failed to build large or representative membership bodies, and the enduring legacy of single-party rule has meant that these have tended to bypass the state entirely in order to solve their own problems or to secure transnational assistance (Dill 2009, 724, 739). Muslim FBOs have been exposed to particularly close scrutiny from the state with respect to their links with transnational organizations (Gilsaa 2015, 53).
The second reason that the formal associational life of Dar es Salaam and Tanzania has diminished in recent decades relates to changing patterns of employment in Dar es Salaam amidst neoliberal programmes of restructuring, and in particular a growing reliance upon informal livelihood activities on the part of residents. As Nelly Babere (2013) demonstrates in her historical overview of the informal economy in Dar es Salaam, while informal livelihood activities first began to see real growth in the 1940s, it was not until the 1980s and the onset of an economic crisis that the proportion of informal operators began to outnumber those in formal employment. By 2006, almost three in five households (57 percent) in Dar es Salaam were dependent on informal livelihood activities as a means of generating income (Babere 2013, 104). In the first decade of mageuzi, local authorities had been more accommodating toward informal economic activity, but from the early 2000s to the present day, traders operating in public spaces have faced increased pressure to be registered and licensed, and have been subject to routine crackdowns and evictions from prime locations (Brown & Lyons 2010, 41-2;Babere 2013;Bahendwa 2013, 155;Msoka & Ackson 2017;Steiler 2018;Malefakis 2019, 19-29). In addition to the largely ineffectual nature of regulatory interventions, the wider governance challenges surrounding the informal economy in Dar es Salaam are further complicated by low levels of participation in operators' associations: the membership bodies of these organizations are poorly representative of the operators most exposed to difficult circumstances, and they have done little to resist the eviction of informal vendors from prime locations by city police (Brown & Lyons 2010, 34-35;Babere 2013, 273).
As in other urban settings across the African continent and beyond, these dynamics of precarity and informalization are fundamentally reordering associational life in Dar es Salaam (Lewinson 2007;Tripp 1997). In response to these pressures, informal networks of traders in Dar es Salaam are increasingly characterized by a certain "fluidity," and even a "near invisibility" (Brown & Lyons, 2010, 42;Malefakis 2019, 5). These emergent groups are very different in form to the associations that were so prevalent in Dar es Salaam during the first half of the twentieth century, and indeed the NGOs and CBOs that have proliferated in recent decades. Rather, they more closely resemble the kind of "loose-knit, ephemeral social formations" that AbdouMaliq Simone (2001, 102) has identified in several cities across the African continent. These highly mobile platforms can serve as "effective vehicles for balancing [the] divergent, yet interconnected needs" of often very different residents (Simone 2001, 113). They permit urban residents to pursue very specific objectives through provisional collaborations, but "without a sense of owing anything [to one another] or having to align agendas" (Simone & Pieterse 2017, 104). In an urban landscape increasingly characterized by profound uncertainty, these dynamic social forms enhance the capacity of residents to come together and improvise responses to unfolding situations while also keeping one another "at arm's length" (Ogawa 2006: 33-34). Because these associations demand far less investment and fewer obligations than the more formal types of organization described above, residents are able to establish a greater number of "belongings" in order to maximize opportunities to connect with strangers. This often entails the creation of a series of "fake dichotomies, oppositions and potential choices" of affiliation that "in fact spring from an extremely limited number of options" (De Boeck & Baloji 2016, 114). Not only is the shifting associational life of Dar es Salaam more ephemeral than it had previously been then, it is also more fragmented: as Kate Meagher (2010) articulates in the context of Nigeria, livelihood networks are increasingly "defined by an individual's portfolio of ties, rather than by collective arrangements," arguably serving to "exacerbate rather than reduce vulnerability" (Meagher 2009, 2;Marks & Stys 2019, 5-6).
Despite the stark disjunctures that exist between these informal networks and the associational life of 1950s Dar es Salaam, some of these "new relational webs" actually redeploy these longstanding cultural forms and social identities in experimental ways (Simone 2004, 7, 24). This is most clearly apparent in the case of football. Where football clubs had previously operated with formal structures and had provided financial assistance to their members in times of need, these associational forms have been retooled in ways that reflect the economic transformation that Dar es Salaam has experienced in recent decades. As such, they have become a far more ephemeral (though no less important) feature of the urban ecology, as I now go on to show.

Recognizing Strangers
I n a conversation about the ongoing struggles between local authorities and informal retailers in Dar es Salaam, one of my friends provided the following account of the more implicit dynamics of these livelihood activities as they play out between street vendors, shop owners, and police in the streets of Kariakoo: The police are asking why [informal business operators] are roaming everywhere; why they are standing on street corners. These people dress well, they eat on time, they have money, even a car. How? What happens is that they work closely with people who own shops and buildings, including renters. They know where is free to rent, who is a thief, who a new person is that might be a danger to the area, where a car parking space might be free.... They try to establish a kind of friendship with the owners of a business, to bring them information, new deals, new sales, to know new buyers, which products are on top. They are also working with the police. This is a perfect illustration of the way that the livelihoods of informal business operators in Dar es Salaam depend on forming ad hoc collaborations with other urban actors. In their efforts to decipher what is taking place in the neighborhood, to acquire "business-critical" information, residents are compelled to make themselves recognizable to one another. The infrastructure of distinction which is characteristic of Kariakoo needs to be understood in the context of this wider politics of legibility. In other words, urban actors "become flags" as a means of acquiring visibility.
As an illustration of the social productivity of this infrastructure of distinction amidst dynamics of precarity and informalization, consider how one of my closest research collaborators, Fundi Chomba, would introduce new people to me. "This is Mr. Wanja, member of CCM and strong supporter of Yanga," he might say, or "This is Mr. Juma [NB religious identity is implicit in his name], he is conservative Simba like me, but has no other ideology." Over time, I realized that by itemizing the associational "ties" of each individual in this manner, Fundi Chomba was providing me with a platform for further exchange insofar as we were able to position one another within "a specific moral or aesthetic universe" (De Boeck & Baloji 2016, 115). Clearly not every urban actor in Kariakoo has the luxury of a third party like Chomba to provide such a concise introduction to the strangers they encounter, but my point here is that color works in precisely this way in Kariakoo: in a materially and sensually congested environment, color operates as a succinct itinerary of one's moral and aesthetic "coordinates." Implicit practices such as these are highly valuable in neighborhoods like Kariakoo because they allow urban actors to generate mutual recognition without being "slowed down by explicit formulation of ideology or verbal meaning" (Elyachar 2011, 96). By becoming flags, strangers become immediately recognizable to one another, giving many visitors to the district the false impression that everyone in Kariakoo knows everyone. On the contrary, it is notable just how impersonal these forms of knowing one another are, presenting a very minimal and generalizable set of social coordinates, and therefore ensuring that residents are able to remain at arm's length from one another and the social obligations entailed by more intimate relationships (Figure 2).
More than granting them a degree of anonymous legibility, by becoming flags, urban actors are also provided with a certain format or "script" for interaction-one that can be used to generate a "temporary co-presence" (De Boeck & Baloji 2016, 121). For instance, on one occasion in Kariakoo, a crowd of young men were directing suspicion at Fundi Chomba and I. Becoming alert to this, Fundi Chomba quickly identified that one of the young men was a Yanga fan and let out a cry of "Yanga oyee!" The man responded with a begrudging nod and a grunt. Though this man was a total stranger to us both, the color of his football shirt allowed him to be recognized without any extended interaction. Despite being an ardent Simba supporter, Chomba used this minimal set of coordinates to stage a form of solidarity with the man. While this was no doubt a rather weak display of solidarity, it was sufficient to overcome the atmosphere of hostility that he had detected because it obliged the young man to reciprocate. On another occasion, Fundi Chomba had accompanied my friend George and I on a walk around Kariakoo as we were taking the photographs that accompany this article. At one point during the day, a man wearing a Yanga shirt asked us sternly why we were taking photos. This time, Chomba batted the man's question right back at him, joking that he "does not deal with" Yanga supporters. This immediately diffused the situation and the man backed off. In other words, by invoking a playful antagonism toward the Yanga supporter, Chomba successfully generated a sense of social distance between him and our group.
In both of these cases, the men wearing the Yanga shirts provided us with a script for interaction, one that offered us a platform with which we could stage similarity or difference. These practices are not only relevant to football but also affiliations to rival brands, rival music artists, rival political parties, rival nationalities, and so on. In a neighbourhood that is constantly steering itself into a web of mutually exclusive belongings, one can be, so to speak, either with or against every individual that one encounters. Accordingly, residents in Kariakoo are, as demonstrated by Fundi Chomba, highly sophisticated at tactically introducing and withdrawing moments of solidarity with people on the basis of their multiple flags.
In what follows, I propose that we attend to how religious identities are being incorporated within the infrastructure of distinction that has been elaborated, providing another layer of horizontal ties. To this end, I demonstrate two ways that Muslim modes of association bear a striking resemblance to those of football with regard to their socioeconomic productivity amidst dynamics of neoliberal urbanism. First, I show how urban actors use Muslim social platforms to open up new avenues of economic collaboration and material support; and second, I consider how they can keep people at arm's length. In doing so, I raise the question of how religious forms and identities are being redeployed by emergent livelihood networks in Dar es Salaam.

Muslim Flags
A s an introduction to some of the ways in which "being Muslim" in Dar es Salaam is bound up with processes of urbanization, consider the following quote from Omary, one of my friends and research collaborators, on the dynamics of rural-urban migration: In Kariakoo, people are town-dwellers. All of those that come to Dar es Salaam adopt this lifestyle. It doesn't matter where you come from, people are the same here; there is a general behavior [...], the behavior of the cities [tabia ya mjini]. Everyone has to find something to be recognized in the city [in order] to cope. They buy certain clothes, a phone, get a new way of talking. There are certain values associated with cities. It's believed that to be in town, you have to be like X, Y, and Z. The good side of city life is that sometimes when we fail, we turn to God. People come to the city and fail in everything, and then they start working with the Bible and Koran. Islam is the easiest way to live: you turn to God, and then your fellow Muslim can give you shelter and money.
Religion here, and particularly Islam, is framed within the terms of a prevailing urban "lifestyle" or "behaviour," which Omary sees as characteristic of Dar es Salaam. Omary very explicitly indicates that being Muslim serves as a channel for rural-urban migrants to access "shelter and money." What I want to stress here is that the operations of Muslim networks of sociality and channels of material support are, no less than football allegiances, predicated on acquiring a certain legibility in the city-in the words of Omary, the necessity of finding "something to be recognized in the city [in order] to cope." Just like being a supporter of Simba or Yanga, being Muslim is an exclusive form of membership that, when expressed through dress practices or "a new way of talking," bestow on an urban actor a particular recognizability in a congested urban environment. As a means of fleshing out these parallels between religious and sporting modes of affiliation in Dar es Salaam, I want to reflect on how Omary's observations about dress practices relate to residents of Kariakoo, where the dress practices of urban actors are more recognizably Muslim than anywhere else in the city. Most academic commentary on Muslim dress and fashion, both in the African continent and beyond, has been focused on dress practices among Muslim women, and especially on practices of veiling (e.g. Moors & Tarlo 2013; cf. Schulz & Diallo 2016). In what follows, I dwell particularly on the dress practices of men in Kariakoo who, as in most prime locations for informal activities in Dar es Salaam, far outnumber women operators (Babere 2013, 282).
The two garments worn by men that are most associated with Islam in Tanzania are the kofia cap and the kanzu tunic. The loose-fitting kanzu is in some respects impractical for those whose work in Kariakoo is physically intensive. However, a number of people told me that in recent years, it has become more common to see young people in Muslim dress not only at Friday prayer or at festivals but also on an everyday basis, with those traveling to Kariakoo from the suburbs making a conscious decision to do so (see also Hirji 2004, 77) (Figure 3). While this reassertion of visible Muslim identity is partly attributable to the shifting terrain of religious politics in Tanzania, it also relates to the productivity that Muslim modes of association have acquired, particularly among informal business operators.
As was seen above in relation to the wearing of football colors, Muslim dress practices facilitate certain kinds of interaction between urban actors in ways that assist them in forming ad hoc collaborations. For example, when my friends were dressed in a kanzu, I found that they were more likely to be greeted with the salutation assalamu alaykum from people in the street. These exchanges would typically be delivered with a degree of playful grandiosity, being peppered with honorific titles such as ustadh and shaykh. As articulated by one man that I interviewed, the kanzu quite literally "attracts greetings from others," affording the wearer a heightened social "exposure" (Goffman 1963, 125). While people already familiar to my friends would go out of their way to greet them in this manner, those who knew them only in passing would also be more likely to do so. These kinds of "passing greetings" are a good example of what Erving Goffman (1971, 73, 76) calls "maintenance rites"-a minimal type of exchange that serves to reinforce a social bond between acquaintances; however, I also observed that Muslim dress generates a platform for heightened social access between total strangers in Kariakoo (Goffman 1971, 79). Muslim salutations can operate as what Goffman (1963,105,129) calls a "relationship wedge," providing not only a "license to approach" a stranger but also a potential scope for social or material claims. As is the case with visual displays of football affiliation, Muslim dress offers as an explicit grammar for interaction, whether this develops into a brief expression of mutual recognition or into a more protracted social encounter.
These cultural practices of greetings do of course have theological underpinnings: as a number of my interlocutors reminded me, the Prophet Muhammad instructed Muslims to greet not only fellow Muslims but "whoever they meet" by saying assalamu alaykum. Fundi Chomba would often quote a hadith, which reports that on the day of Eid, the Prophet Muhammad would return home from prayer by using a different route from which he came. As someone who practices this not only on Eid but on a daily basis, I found it revealing that Chomba's interpretation of this hadith was that this would maximize opportunities for the Prophet to greet people.
On the one hand then, mutual displays of greeting can be understood as "social-recognition rituals," which present a specific way of being together in Kariakoo (Goffman 1971, 75). In the words of one of my friends, Hassan, "When we greet each other, you are putting your mind close to my mind, and it means that I'm 'someone' to you." He added that, "Greeting makes the brotherhood between Muslims stronger." On the other hand, however, the tactical use of greeting is also a crucial competency insofar as it serves as a way of getting to know another's situation (hali), as a public display of social capital, and even as a means of getting assistance. For example, Fundi Chomba advised me that "It is very important to use greetings when seeking something you want or fixing a problem. In Kariakoo, people can rob you and no one will save you, but if you say 'Habari za leo?' [How are you today?] and 'Assalamu alaykum,' people will be pleased with you and will assist." These insights complement AbdouMaliq Simone's (2010, 146, 151) analysis of Muslim practices of sociality in Bangkok where, by sustaining "the minimal identification of themselves as Muslims," Islam serves as an "instrument for a collaborative entrepreneurship"-a "way in which individuals can align themselves and coordinate their lives with each other." Insofar as it attracts greetings, the kofia and the kanzu offer people in Dar es Salaam another means by which they may become a flag. Residents that wear Muslim dress or adopt a recognizably Muslim comportment are participating in a particular act of "being Muslim," through which they acquire a distinct legibility and heightened exposure in the urban environment. This legibility grants them access to a grammar of social interaction that they and others can key into in order to tactically align themselves with diverse people and objects. Indeed, many residents choose to display both religious and sporting forms of membership simultaneously, wearing Muslim garments in combination with football colors and thereby increasing their social exposure (Figure 4).
To be very clear, the ties under discussion do not entail the comparably thick kind of social trust that Neil Carrier (2016, 180-184) describes as characteristic of Somali Muslims that work in the Eastleigh district of Nairobi; they are far more ephemeral, along the lines of the loose social formations already discussed (see also Malefakis 2019, 67-78). No less than other kinds of social identity in Dar es Salaam then, to be Muslim in the manner that I have described is to make use of this identity as what AbdouMaliq Simone (2008, 80) calls a "device" or "node" of "absorption and deflection," one that that serves to steer urban actors in particular directions and into contact with particular people. In what follows, I want to pick up on this notion of deflection and explore how being Muslim can serve as a device through which urban actors may tactically distance themselves from people and objects.
Muslim Shields W hile my research collaborators speak about flags in reference to the aesthetic markers of one's affiliation (be it political, religious, or sporting), the term can also be used in a pejorative sense. More specifically, flags are people that merely follow prevailing norms with an eye to their personal interests at a given moment, but with no real commitment to a viewpoint or ideology. A Swahili proverb used by Fundi Chomba helps to illustrate my point: "A flag follows [the direction of] the wind," (Bendera hufuata upepo). While the proverb can be used to mean many things, Fundi Chomba has invoked it as a political criticism of what he sees as the capriciousness of Tanzanian citizens with respect to their loyalty to a political party or principle-his point being that they are easily led, whether by popular opinion, fear, or the promise of personal gain. The same sentiment may also be applied to Muslim flags, as defined in the foregoing discussion, with the attendant suggestion that those who exhibit their Muslim identity through dress practices are primarily motivated by economic rather than devotional concerns.
Several of my friends expressed to me that they consider the kanzu and kofia to be poor indicators of whether someone was a good Muslim or not, or indeed Muslim at all, with many Christians choosing to wear a kanzu in order to "look distinguished" (see Hirji 2004, 77-78). 2 As one of my friends emphasized: Most who dress in a kanzu and kofia do so to identify as a Muslim leader or scholar, but inside [ndani] they have nothing. Anyone can enter a shop, buy a kanzu, and look like a shaykh when wearing one, but so many of those who do can't answer a single question about Islam.
Along similar lines, Fundi Chomba alluded to this discrepancy between Muslim dress and behavior using another Swahili proverb: "Behavior is not like clothing, it is not easily changed" (Tabia siyo nguo, haibadiliki kwa urahisi). He further elaborated on this in his own words: Islam is not about wearing a kanzu or kofia, it is about walking in a straight line and having good behaviour. [...] If a believer wears Islamic garments, he or she is announcing that they follow the ways of the prophet. They are expected to be more decent and friendly and trustworthy and peaceful; to be good in everything. The kanzu used to be a peaceful garment, an "honest person" [mwaminifu] garment, so if you are people who lie or do forbidden things, why are you wearing one? Why are you speaking loudly and telling the world, "I'm a Muslim"? All people of bad behavior are wearing a high quality kanzu. It is like a shield [ngao] for them so they appear like a nice person.
I want to elaborate here on this notion of shield as Fundi Chomba uses it. This term was earlier encountered in reference to the context of the 2015 presidential election, when political flags were seen to protect individuals or spaces from unwanted attention or suspicion. In Kariakoo, there are clear parallels here between the shield-like operations of political posters on the one hand and kanzu robes on the other.
Several of my friends shared Fundi Chomba's impression of the kanzu as a shield but in two markedly different senses. The first is that the kanzu serves to protect the wearer from himself; it restricts "bad" behavior. One man put it as follows: You can't wear a kanzu and ask a lady for a date. A kanzu brings respect to you. A certain type of person won't do these kinds of things at this time of day wearing this kind of clothing; won't talk to you with a swagger.
A young Muslim interlocutor said something very similar: To wear a cassock or a full Islamic outfit can change you. Even me, I am only Abdallah, but if tomorrow I wear kanzu every day, they will call me Shaykh Abdallah. Those that wear a kanzu prohibit themselves from bad things. The kanzu is supposed to be accompanied by this habit and this other habit is prohibited.
Especially interesting was the fact that Fundi Chomba suggested that to enter a bar or to swear in the street while wearing a kanzu serves to "put down the kanzu and not you." In other words, while "what you have judged on your own" is not generally seen as anyone else's business, the real concern would be that one was denigrating the kanzu itself. These statements betray a conscious sense on the part of some residents of the agency of the kanzu as a moral actor in Kariakoo, as well as the responsibilities owed to the garment by other subjects that participate in the same social ecology.
The second way in which a kanzu can be described as a shield is that it serves to secure the wearer from evil. As Fundi Chomba put it, "To wear a kanzu is like a protection (ulinzi), no matter which type of kanzu." Chomba's intended meaning here concerns moral and spiritual protection. However, my observation is that in precisely the same way, the kanzu can not only heighten the social exposure of a wearer to some types of encounter, but also serve to deflect unwanted interactions by restricting his accessibility among other types of individuals or situations (Goffman 1963, 104, 125). For instance, one man told me that he "likes to wear a white kanzu because [...] with any color, people are easy to connect to you; any sins will be very easy to connect to you." This idea of colors being used to facilitate connections and generate social distance tallies well with the foregoing discussion of football and party politics, shedding particular light on the material properties specific to Muslim garments as "prosthetic" extensions to the body (Gandy 2005, 29). Indeed, beyond merely deflecting other urban users, the kanzu can even serve to shield the wearer from the unwanted attention of law enforcement agents. Earlier it was seen that, since the early 2000s, unlicensed informal traders in Dar es Salaam have been increasingly exposed to the risk of forcible eviction from prime locations or arrest by police on behalf of local authorities. One of my friends explained that "If a police patrol meets a man without a kanzu, they can be called and interrogated, yet if they do wear one the police will say, 'Oh, this is an ustadh [Muslim honorific title] so we won't take him in.'" In other words, men garbed in Muslim costume are engaged in the performance of a social role that obliges them to maintain an impression of moral piety befitting a distinguished Muslim figure (Goffman 1959, 8-9) (Figure 5). Accordingly, police would think twice about targeting them in their crackdowns since such an interaction could be construed as manifestly unwarranted or even discriminatory.
A further reason why the kanzu operates as a shield in this manner relates to the fact that Kariakoo is home to a particularly pronounced number of Friday mosques (Misikiti wa Ijumaa), in the vicinity of which Muslim-coded garments, such as kanzu and buibui, are typically sold. My conversations with residents revealed that mosques have long served as a legitimate alibi for urban actors in Kariakoo who are forced to account for their presence in the city. In other words, when pressed for their reasons for being in the Kariakoo neighborhood, an unlicensed trader who is recognizably Muslim would be able to feasibly claim that they are in the district to perform swala (obligatory daily prayer). Indeed, these urban sites have acquired a further political sensitivity since 1998 when, at a mosque in the nearby Mwembechai neighborhood, two people were shot and killed by state police in riots precipitated by the arrest of a senior Muslim cleric (see Kirby 2017). Because it grants access to this infrastructural assemblage of informal business activities and mosque complexes (see Kirby forthcoming), it might then be argued that being Muslim provides residents with a more effective means of resisting attention from police than belonging to operators' associations.

Conclusion
W hile the historical intersections between Islam and commerce in East Africa and the wider Indian Ocean region are well documented, it is striking that contemporary entanglements between Muslim modes of sociality and informal livelihood activities in the region have received so little critical attention, particularly at a time when these intersections appear to be intensifying. One reason for this is that commentators generally tend to privilege normatively religious settings, activities, and modes of self-presentation when researching religious people, rather than reading these within the wider context of their everyday lives, or indeed taking seriously more marginal or implicit practices that may still be associated with religion (Larkin 2016;Schielke 2010;Soares & Osella 2010). This is particularly true with regard to matters of business and economic activity, which, as Simon Coleman (2013, 206) shows, are not typically considered to have anything to do with religious practice-a supposition that derives from the influence of ascetic forms of Protestant Christianity on European and North American perceptions of what is constituted by "proper religion" (Kirby 2019). Even when research on urban religion does bring these different domains into consideration, religion is all too often framed within terms of "meaning and hope" offered to "impoverished urban residents seeking solace and comfort" (Murray & Myers 2006, 11). In other words, the importance of religion in the associational life of urban Africans is restricted to the moments of redemption and spiritual or social communion that they provide; a mere compensation for an "urban politics of disappointment" (Simone 2013, 160). In this article, I have instead sought to approach religious practices of devotion and affiliation as numbering among the kind of background competencies by which residents come to "see and feel similarity," and to render urban worlds intelligible (Simone 2013;Larkin 2016;Coleman & Vásquez 2017). Joining Julia Elyachar (2011), I have sought to look beyond consciously performed embodied practices or techniques of the self and instead to the "tacit gestures" and "locomotory bodily practices" that co-constitute urban space and communities.
In Dar es Salaam then, as in many other African cities, longstanding modes of sociality are transforming in response to intensified dynamics of informalization and precarity, which have accompanied neoliberal urbanism. The associational life of the city has become more ephemeral and fragmented, being increasingly geared to providing opportunities for establishing loose economic collaborations with fellow residents. Emergent livelihood networks often operate under aesthetic markers which are borrowed from an earlier era of associational life, most notably in the case of football teams. These cultural forms are redeployed as part of a broader infrastructure of distinction through which, using bright colors and specific dress practices, people become flags for their football teams and political parties. By situating these social dynamics in the broader context of informal business, I have shown that these practices are valuable because they allow residents to recognize other people who share the same urban environment in a minimal way, providing a platform for them to generate temporary co-presences with strangers, which do not entail any sustained mutual obligations. Equally, these practices can also be used in order to tactically withdraw moments of solidarity in ways that remain impersonal and widely legible to other residents.
Building on this analysis, I have demonstrated that Muslim socialities are being reordered in strikingly similar ways to other forms of associational life in Dar es Salaam, and particularly those relating to football teams and political parties. First, I considered the flag-like dimensions of what it means to be Muslim in Kariakoo, examining how urban actors increasingly use dress practices to engage in acts of "being Muslim," which open up new avenues of economic collaboration and material support. In other words, being Muslim grants urban actors access to an expanded horizon of possibility with respect to what they can do with fellow residents. Second, I addressed the shield-like qualities of Muslim modes of association, showing how residents can also employ characteristically Muslim dress practices in ways that tactically generate social distance from others, most notably in the case of law enforcement agents who participate in crackdowns on unlicensed traders. I suggest that the enduring plausibility of Muslim affiliation amidst these social transformations is partly attributable to its remarkable dynamism within Kariakoo as an urban milieu, and its capacity to be incorporated among the informal strategies through which residents navigate the city. Pentecostalism and post-Islamism, the forms of entrepreneurial religion in which Stephan Lanz and Martijn Oosterbaan (2016) are particularly interested, are very much present in contemporary Dar es Salaam; however, I propose that this "metropolitan religious mainstream" that I have identified in Kariakoo is no less imbricated with the trajectories of neoliberal urbanism in which these authors are interested. When read alongside a broader turn towards the everyday and the lived within the anthropology of Islam then, Lanz and Oosterbaan's theoretical intervention can also help to explain the emergence of more ephemeral forms of religious association.
More broadly, my analysis raises questions about the ways in which religious forms and identities are being deployed by emergent livelihood networks in cities across the African continent and beyond. Given that the same conditions of precarity, deriving from parallel processes of informalization, are increasingly characteristic of cities across the world, we may find that comparable configurations of religious urbanity are emerging in other urban sites (Mbembe & Nuttall 2008, 4). As such, addressing these questions will contribute to efforts to understand and plan for our shared urban futures.

Notes
Acknowledgements. In the process of writing this article I benefited from stimulating conversations with Nelly Babere, Brian Larkin, Wendell Marsh, Shabani Mwakalinga, and Gana Ndiaye. I am also grateful to Adriaan van Klinken, Erik Meinema, and the anonymous reviewers for their comments on earlier drafts. 1 Following Asef Bayat (2007), post-Islamism here is taken to mean a Muslim movement that can be distinguished from Islamism in its attempt to "fuse" religiosity with democratic and secular values such as rights and freedom, prioritizing "sober work in place of [delivering] a revolutionary blow" (Lanz & Oosterbaan 2016, 495).
2 Shahab Ahmed (2016, 435, 447) convincingly argues that these dress practices should be conceptualized as "Islamic" (rather than "cultural," "Islamicate," or "secular") not only because garments like these are recognizably "Muslim," having been produced by "Muslims acting as Muslims," but also because these practices transform the body of the wearer by bringing it into an "Islamic field of meaning."