A Methodological Approach and Conceptual Tools for Studying Migrant Belongings in African Cities: A Case of Zimbabweans in Johannesburg

Abstract: Many scholars view the relaxation of restrictions to enter and reside in South Africa after the end of apartheid in 1994 as a catalyst for a number of changes in regional human migration. In this article, I show that human migration dynamics in post-apartheid South Africa pose theoretical and methodological challenges for studies of belonging among migrants. I argue that the majority of studies of belonging among migrants in South Africa rely on nation-state frames of reference and quantitative approaches. Furthermore, they largely seek to establish general patterns of migration processes. While these studies illuminate macro-level factors which seemingly structure migration experiences, they provide only a partial picture of migration dynamics. Often, they fail to sufficiently attend to context-dependent modes of emplacement among migrants in the urban zones due to the conceptual and methodological tools they use. The purpose of this article, therefore, is to suggest theoretical assumptions, concepts and methodological tools which constitute an alternative design for studying social constructions and practices of belonging among migrants in post-apartheid South Africa. The article draws on my doctoral research of Zimbabweans in Johannesburg for illustrations.

Key words: agency; attachment; belonging; context-dependent dynamic; qualitative content analysis; theory-driven ethnography; semi-structured interviews; social relationships; South Africa; transnationalism; migration

2.1 Migrants and belonging in post-apartheid South Africa: What existing studies show

3. Qualitative Inquiry of Migrant Belongings: Theoretical Premises, Concepts and Methods

Apartheid South Africa preferred temporary, unskilled and oscillatory as opposed to a permanent indigenous work force since the former was docile and incapable of organising itself (PROTHERO, 1974). Based on this logic, foreign African migrants who came mainly from the Southern African region were employed for short-term contracts (usually two to three years long) in South Africa, after which they were made to return to their places of origin to join family members. This trend of regional labour migration was especially pronounced in the mining industry where contract mining has persisted even after the demise of formal apartheid in 1994 (ALEXANDER, 2001; CRUSH & McDONALD, 2000). However, many historical studies have shown that this system of circular labour migration was pervasive in all economic sectors of the country (MLAMBO, 2010). In any case, legal immigration was not possible for African international migrants (PROTHERO, 1974). [1]

Scholarly opinion regarding patterns of migration in post-apartheid South Africa is polarised between those who see continuity with apartheid trends of human migration vis-à-vis those who suggest evidence of change. Some scholars purport that apartheid policies and attitudes continue to influence migration management post 1994 when South Africa gained majority rule (CROUCHER, 1998; LANDAU, 2010a; PEBERDY, 2001). This point of view draws on evidence of strict migration controls which are designed to keep out low-skilled migrants. [2]

Yet using statistics of post-apartheid human migration, others contend that restrictions to enter and reside in South Africa have undoubtedly eased since 1994. Jonathan CRUSH and David A. McDONALD (2000) note that both the volume and diversity of human traffic into South Africa increased after the introduction of majority rule: In the first decade alone legal migration to South Africa increased ten-fold. Furthermore, the population of foreigners in South Africa now includes movers from its traditional migrant source countries in the South Africa Development Community region as well as many other arrivals from the continent (PALMARY & LANDAU, 2009). Also, settlement patterns of migrants have changed visibly since 1994. Whereas before, the majority of foreign migrants lived in the urban peripheries, particularly in compounds on farms and mines, many foreigners now reside in cities, mainly in the inner-city districts (CHERENI & PALMARY, 2010; PALMARY & LANDAU, 2009). Increased visibility of foreigners in the cities has been accompanied by strong and abiding anti-migrant sentiments, prevailing in public discourse (LANDAU & FREEMANTLE, 2010). Furthermore, reports of xenophobic violence are increasingly captured in mainstream media (DANSO & McDONALD, 2001; MAWADZA & CRUSH, 2010; MOSSELSON, 2010). Some researchers speculate that these more recent developments prompt new research questions regarding the social formations of migrants in post-apartheid South Africa (LANDAU, 2010a, 2010b; PALMARY & LANDAU, 2009). Loren LANDAU and Iriann FREEMANTLE (2010), for example, argue that novel social formations are emerging in African cities but we know little about them. These authors maintain that there exists a paucity of a "conceptual vocabulary of belonging reflecting practices of those living in and moving through Africa's cities" (LANDAU, 2010a, p.1). [3]

The need to generate conceptual and methodical tools for studying social formations linked to migration is fully acknowledged internationally (RICHTER, 2012). In South Africa, where migration studies largely employ quantitative approaches in order to illuminate general patterns of migration, context-specific and variable forms of migrant belonging remain underexplored. Moreover, the development of innovative conceptual and methodical approaches in this area remains stunted. [4]

In this article, I suggest an alternative approach to the predominantly quantitative designs adopted in migration studies in post-apartheid South Africa. At the core of the approach described here is the assumption that migration is a context-dependent dynamic and not necessarily a homogeneous experience (SLÁDKOVÁ & BOND, 2011). Based on this assumption, social interactions and relationships as well as situated modes of identification, sustained between migrants, are central to any analysis of migrants' social constructions of belonging. From this point of view, belonging is conceived of as a less individuated affair, and a more situated and contingent experience. [5]

It is important to point out up-front that the approach I present below was developed in the context of my doctoral research with Zimbabwean migrants in Johannesburg (2008 to 2012). The study aimed to explore how Zimbabwean migrants in Johannesburg construct, sustain and negotiate notions of belonging in both destination contexts and places of origin. Additionally, the study examined the implications of migrants' experiences for social policy, especially social protection policies for mobile populations. Study participants were Zimbabwean members of a transnational religious movement which originated in Zimbabwe in the late 1950s. The study used ethnographic methods, mainly participant observation and semi-structured interviews, to generate data. Qualitative content analysis was used in conjunction with data analysis software (ATLAS.ti) to make sense of participants' narratives (FRIESE, 2012). [6]

The article proceeds as follows: The next section provides a background of contemporary human migration linking South Africa and the Southern African region. It reviews existing empirical research on migrant belonging in post-apartheid South Africa, highlighting the theoretical frameworks and methodological approaches that they use. In so doing, the section demonstrates that present-day patterns of human migration between South Africa and neighbouring countries present an anomaly to both theory and methodology, especially those that draw on nation-state frames of reference. This section also shows that South Africa is the migration hub of the region, and it has a foreign born population of slightly above three per cent of the total population (UNITED NATIONS DEPARTMENT OF ECONOMIC AND SOCIAL AFFAIRS, 2009). Section 3 briefly describes the central question which underpinned the doctoral study and introduces the qualitative approach including the theoretical assumptions and concepts used. Section 4 presents a concrete step by step account of the approach using relational data gathered through semi-structured interviews. In the final section, key conceptual and methodological insights of the study are highlighted. It is concluded that conceptualising belonging as a situated and contingent experience helps the researcher to move away from the more state-centred theoretical frames of reference which overlook the quotidian practices of migrants. [7]

Human mobility to South Africa has increased dramatically since 1994. This is largely because South Africa is the preferred destination of African migrants from the Southern African region and beyond. South Africa is host to regular (documented) migrants and irregular (undocumented) migrants, skilled and unskilled migrants, professionals, students, informal traders and asylum seekers all of whom perceive economic opportunities in the country (CRUSH & McDONALD, 2000; WALLER, 2006). [8]

It is believed that undocumented migration intensified after 1994 due to relaxation of restrictions on physical mobility (VIGNESWARAN, 2008; WALLER, 2006). A large number of irregular migrants "border-jump" onto South African territory using undesignated "ports" of entry. And many continue to live in the country irregularly. These migrants are not captured in migration statistics at all. An unknown number of irregular migrants acquire fraudulent permits and other necessary documentation (WALLER, 2006). Whilst a considerable portion of migrants enter the country using legal documentation such as a visitor's permit, they often remain in the country much longer than is legally permissible, swelling the population of irregular migrants (PALMARY & LANDAU, 2009). Therefore, the exact size of the population of foreigners in South Africa is not known. [9]

However, the 2007 Community Survey, a nationally representative survey conducted by Statistics South Africa (STATS SA), estimated that the total number of foreign-born residents is just over 1.2 million or 2.79 per cent of the total population (PALMARY & LANDAU, 2009). The spatial distribution of the foreign-born population suggests that settlement patterns have shifted since the early 1990s. Figure 1 illustrates that prior to 1994, cross-border migrants were concentrated in agricultural and mining areas outside urban zones. After 1994, both internal and cross-border migrants have been concentrated in the country's urban centres, but mainly in Gauteng Province of South Africa (PALMARY & LANDAU, 2009). In 2007 alone, Gauteng hosted 46 per cent of South Africa's foreign born population. About 7.9 per cent of the population in Johannesburg—the biggest city in Gauteng Province and the country—is foreign-born. In some inner city areas like Yeoville, Berea and Hillbrow, more than 50 per cent of the residents are non-citizens (PALMARY & LANDAU, 2009). [10]

The popularity of Gauteng as a destination for both internal and cross-border migrants can be explained by real and perceived economic opportunities in the province, especially in its economic centres such as Johannesburg and Pretoria. Gauteng Province accounts for 34 per cent of gross domestic product of the country. Furthermore, although Gauteng is the smallest of all the nine provinces in that it accounts for only two per cent of the total landmass, it is by far the most urbanized province (PALMARY & LANDAU, 2009).

Figure 1: Distribution of non-nationals in post-apartheid South Africa (PALMARY & LANDAU, 2009, p.23) [11]

2.1 Migrants and belonging in post-apartheid South Africa: What existing studies show

During white minority rule, migration studies sought to examine how larger social and political systems created conditions that influenced trajectories of both internal and international migrants (POTTS, 2010; PROTHERO, 1974). Importantly, these studies demonstrated that structural factors—including urban administration and macro-level processes of capital accumulation—inhibited both physical and social mobility of African migrants in urban areas, thereby generating persistent material insecurity for them (POTTS, 2010). In response to these adverse conditions, non-white migrants adopted coping strategies of which circular migration between the distant city and rustic places was at the centre (MAYER, 1962; POTTS, 2010). Thus, studies of migration conducted during apartheid and colonialism scrutinised the complex ways by which circular migration was institutionalised through law and policy (CRUSH & McDONALD, 2000). [12]

Migration studies conducted in South Africa after 1994 continue to draw on political economy approaches that generally privilege structure over agency in analysing migration phenomena (POTTS, 2010), while relying mainly on quantitative survey methodologies. The "representative survey" remains the eminent methodological approach adopted by most migration researchers. For instance, two prominent migration research institutes in South Africa—namely the Southern African Migration Project (SAMP), and the African Centre for Migration and Society (ACMS) at the University of the Witwatersrand—employed quantitative surveys in the majority of studies conducted since the end of apartheid. Broadly, the studies conducted at SAMP and the ACMS, and by many other researchers, focus on practices of the state and powerful actors in the economy—including institutional arrangements and discourses—that shape immigration and citizenship (CROUCHER, 1998; LANDAU, 2010a, 2010b; MOSSELSON, 2010; PEBERDY, 2001). More often than not, these studies exclusively attend to the mechanisms through which the South African state creates and enforces the criteria for inclusion in its political community, as well as the rights and entitlements framed for migrants vis-à-vis insiders. [13]

Such studies paint an entirely precarious picture for unskilled temporary migrants in South Africa. They clearly demonstrate that, unlike citizens, those foreigners without permanent residence have only partial or no access to de jure socio-economic rights, let alone political rights. Leila PATEL (2005) and others (GRAY, 2006; PEBERDY, 2001) describe in detail government economic and social development policies and programmes through which rights for citizens materialise. What underpins such programmes is a commitment to up-lift the non-white people, those who experienced systematic exclusion during apartheid¹⁾. [14]

One thus finds that post-apartheid migration studies seek to make visible the legal factors and political processes that variably structure the experiences of migrants in South Africa. In particular, these studies explore the significance of international migration for nation-building in post-apartheid South Africa. A common argument advanced by these studies is that state administration of human migration is largely premised on political expediency rather than a rational management approach that maximises economic benefits for the country. Rhetoric designed to paint a spectre of "flooding" and "invasion" of the country by "aliens" from across the Limpopo River²⁾ (MAWADZA & CRUSH, 2010), rather than evidence, seem to influence government stance on migration. For instance, in October 2003, former South African President Thabo MBEKI told Commonwealth Secretary General Don McKINNON that "he [MBEKI] has three million Zimbabweans in South Africa ..." (CRUSH & TEVERA, 2010, p.3). Jonathan CRUSH and Daniel TEVERA (2010) attribute the source of some outrageous estimations of the size of foreign-born population in South Africa to this conversation. Yet, as noted already, STATS SA, South Africa's department responsible for demographic analyses, estimated that the total number of all foreign-born people in South Africa was just over 1.2 million at that time (PALMARY & LANDAU, 2009). Hence, through discursive practices that politicise the presence of migrants in South Africa, state officials effectively position migrants as outsiders in ways that simultaneously construct those who belong to South Africa (MOSSESLSON, 2010; PERBERDY, 2001). [15]

The presence of migrants in South Africa is also politicised because foreigners provide a convenient scapegoat for government failure. Exclusive apartheid policies led to higher levels of poverty, inequality and extreme injustice for non-white groups (PATEL, 2005). Therefore, calls for an inclusive citizenship where all sections of the South African population can effectively exercise their socio-economic rights reflect on government's competences. Consequently, the post-apartheid South African government often finds it convenient to blame migrants for its own failures to deliver on constitutional promises (LANDAU, 2010b; MOSSELSON, 2010). In government rhetoric, migrants from Africa play the villain; they are blamed for taking away jobs from South Africans, expending welfare benefits targeted for deserving citizens, and bringing disease, corruption and crime into the country (PALMARY & LANDAU, 2009). In this way, Africa, north of the Limpopo River is constructed as a threat to the South African constitutional promise of a prosperous country (LANDAU, 2010b). [16]

The wider society similarly exhibits strong anti-migrant sentiments as demonstrated by prevailing derogatory language used to describe foreigners (LANDAU, 2010b). Occasionally, this hatred toward foreigners incites violence aimed at hurting and killing migrants while looting their property (LANDAU, 2010b; MOSSELSON, 2010). [17]

The predominance of political economy frameworks and quantitative methodological approaches has strong implications for the construction of the research subject as well as for the nature of data generated in migration research. Studies which focus on the mechanisms that either enable or impede formal belonging tend to conceptualise migrants through the prism of a citizenship framework, as rights-bearing individuals. Thus, belonging is understood as a more individuated experience. Furthermore, a lack of socio-economic rights and related entitlements for migrants, prevailing exclusive legal frameworks, and coercive state practices create conditions that discourage the fixity of foreigners in the receiving places (see for example, CRUSH & McDONALD, 2000). Circulation migration then, is part of migrants' agency as they respond to insecurities at destination. Except, in so doing, migrants deny themselves a stable locus of belonging in the receiving country. Based on this interpretation, studies of human migration in post-apartheid South Africa continue to frame African international migrants as up-rooted, transient and mobile individuals—sojourners who are always arriving and never settling (POTTS, 2010). [18]

Findings of recent migration research suggest that theoretical frameworks and methodological approaches adopted in studies reviewed in this section are no longer adequate to grasp the dynamics of human mobility in post-apartheid South Africa. Whereas trajectories of many international migrants still conform to the circular movement thesis—as migrants shuttle between places of destination and origin—a significant population ends up living in South Africa's cities much longer than initially planned. Loren B. LANDAU and Iriann FREEMANTLE (2010) argue that, for many African migrants, returning home or moving on to a third country remains a dream for practical reasons such as money, safety and social status. Strictly speaking, cities have become "destinations and not points of transit" (LANDAU, 2010a, p.1). As noted already, these authors (LANDAU & FREEMANTLE, 2010) speculate that this situation gives rise to multiple, place-specific forms of membership for migrants. Because inclusion in such solidarities follows a logic far divorced from formal citizenship and rights frameworks, LANDAU and FREEMANTLE argue that new conceptual tools are needed to account for migration processes. [19]

Approaches that emphasise structure rather than agency are inclined to assume, a priori, a homogeneous experience of inclusion (or exclusion) for migrants. Migrants' exclusion from formal citizenship and the formal labour market, for example, is used to illustrate a lack of inclusion in society as a whole. This, however, seems to be a limited view of belonging among migrants. Numerous qualitative studies of integration of migrants conducted in the Global North demonstrate that exclusion of a person in a subsystem would not necessarily lead to the person's exclusion in another (NOHL, SCHITTENHELM, SCHMIDTKE & WEISS, 2006; ROSENTHAL & KÖTTIG, 2009). Stated otherwise, exclusion of a migrant at one level might not necessarily result in that migrant's exclusion at other levels. In fact, Gabriele ROSENTHAL and Michaela KÖTTIG (2009) argue that experiences of economic and social insecurities, for example, in the labour market and welfare systems, may strengthen internal integration in solidarities of outsiders. Therefore, there is immense value in exploring internal workings of meso- and micro-level solidarities of migrants. [20]

In the following section, I highlight some of the existing qualitative studies of migration which influenced my own approach. [21]

Throughout the world, contemporary human migration presents theoretical and methodological challenges to existing nation-state frames of reference (LEVITT & GLICK SCHILLER, 2004). Hence, researching migration dynamics, including the various aspects of new migrant spaces, requires innovative and flexible approaches which employ new concepts and use qualitative methods in combinations. Anja WEISS's (2006) study of transnational class formation among migrants clearly brings out this challenge. WEISS found that extant notions of class are largely premised on the nation-state frame, and have become inadequate to account for the social position of migrants who live and work in more than one country. To address this shortcoming, she combines a grounded theory approach with semi-structured interviews which are designed to leave room for participants to construct their own narratives. [22]

Numerous other researchers similarly use qualitative designs to explore and theorise various aspects of spaces of migration, including migrant integration and belonging. Marina RICHTER's (2012) study further illustrates the efficacy of a qualitative research in exploring complex transnational spaces of second generation Spanish in Switzerland. In much the same manner, NOHL et al. (2006) used a qualitative research approach to analyse, at various levels, how highly skilled migrants draw on cultural capital during entry into the labour market. Numerous other qualitative studies of contemporary migration dynamics exist (see BORKERT, MARTÍN PÉREZ, SCOTT & DE TONA, 2006). [23]

A number of insights from these and other qualitative studies of migration influenced how I designed my own approach. With regard to the need to avoid methodological nationalism in research, that is, not designing research which is constrained by nation-state frames (LEVITT & GLICK SCHILLER, 2004), the above studies clearly demonstrate that migrants are connected to disparate sites on either side of the border through various quotidian practices (RICHTER, 2012). Border-crossing practices of migrants have two opposing theoretical implications for migrants' social formations. On the one hand, it seems that migrants' social relations—that is, sustained ties or connections between migrants and other persons (KNOKE & YANG, 2008)—are free from spatial limitations such as those imposed by political borders. On the other hand, locality remains important as it is the "place where migrants touch ground" (RICHTER, 2012, §1). This suggests that solidarities of migrants are grounded somewhere, geographically. Hence, networks of migrants are, in essence, place-specific social formations. Therefore, both the nature of membership and the level of internal integration in migrants' solidarities vary from place to place. Also, the point at which belonging to a specific group becomes important varies from migrant to migrant (WEISS, 2006). One cannot, therefore, presume that a stable belonging to we-groups or religious belonging exists (ROSENTHAL & KÖTTIG, 2009). [24]

3. Qualitative Inquiry of Migrant Belongings: Theoretical Premises, Concepts and Methods

This section presents the qualitative approach with which I explored how Zimbabweans in Johannesburg construct and sustain notions of belonging. [25]

From the foregoing, it is clear that migration is largely a fragmented, transitory and place-specific experience (De TONA, 2006). Hence, in order to effectively explore the forms of inclusion in micro-level solidarities of migrants, it is necessary to develop a set of theoretical assumptions about migrants' life worlds, which are no longer in thrall to the nation-state frames of reference. In line with this theoretical orientation, I founded my approach on social practice, a thinking in the social sciences which presumes that social life "involves the working of relationships" (THOMAS, 2005, p.14). In other words, social interaction, not necessarily multi-level structures, is integral to any account of social life (THOMAS, 2005). Therefore, this theoretical posture is more inclined to make sense of the quotidian practices of migrants as well as their distinct relatedness. [26]

By conceptualising migration as a "context-specific dynamic" (SLÁDKOVÁ & BOND, 2011, p.327), this study brought the marginal life worlds and agency of migrants into the purview of analysis. Although one cannot entirely explain away the influence of macro-level structures on migration dynamics, support for this theoretical posture exists in literature. A number of scholars accentuate the role of agency in shaping migrants' experiences. For example, FEDERICO DE LA RÚA (2007) argues that, through their own agency, actors in micro-contexts can potentially distort state structures and systems that constrain their aspirations. The author illustrates that, even when the state develops policies and classificatory schemes to which other actors must conform, often such categories "are negotiated, opposed, adjusted and filled or emptied" by actors within networks of relations at local levels (p.684). Eugenia RAMÍREZ GOICOECHEA (2005) similarly underscores the importance of accounting for agency in migration research. She argues that researchers should attend more to the ways in which social actors, as active builders of their (and others) parameters and conditions of existence, may reconstitute circumstances within which they find themselves. [27]

My conceptualisation of migration as a context-dependent dynamic also profited from literature on transnationalism. As defined by Nina GLICK SCHILLER and Georges E. FOURON (1999), transnationalism represents "a pattern of migration in which persons, although they move across international borders, settle and establish on-going social relations in a new state, maintain on-going social connections within the polity from which they originated" (p.344). It is a unique form of post-migration adaptation whereby migrants create and sustain social relationships in both countries of origin and destination after departure. [28]

This form of post-migration adaptation—also known as simultaneous embeddedness—strengthens the view that political borders and legal frameworks do not necessarily contain the relationships and practices of migrants (LEVITT & GLICK SCHILLER, 2004). Theoretically speaking, quotidian border-crossing practices of migrants—broadly termed everyday transnationalism or transnationalism from below (SMITH & GUARNIZO, 1998)—who are positioned in multi-local social networks can provide "modes of resistance to exploitation and discrimination for migrants" (CRUSH & McDONALD, 2000, p.10). Migrants can use various kinds of support in relational networks and civil society formations from both home and destination countries to resist and challenge inequalities of power which confront them in the global market place (ibid.). Hence, a transnational way of life can potentially enable migrants to resist state practices which constrain their aspirations. [29]

The premise that social life unfolds within relationships between persons (THOMAS, 2005), and that it transcends nation-state boundaries (GLICK SCHILLER & FOURON, 1999), influenced my conceptualisation of belonging. I used belonging not necessarily to represent an individual's membership in a political community³⁾ and the mechanisms by which the political community validates an individual's claim to formal inclusion (YUVAL-DAVIS, 2006). Rather, I conceived of belonging as "an inclination or an inclining from one toward the other, of one by the other, or from one to the other" (ROWE, 2008, p.17). In other terms, belonging is not an individual affair (PROBYN, 1996); it captures "more accurately [migrants'] desire for some sort of attachment, be it to other people, places or modes of being" (p.19). Therefore, in my study, belonging was more relevant to migrants' everyday lives, and not to citizenship frameworks (CHRISTENSEN, 2009). [30]

In my study, the notion of attachment signifies a person's (on-going) need to gain positive evaluation of oneself, which comes from positive evaluation linked to membership of a social group (FEDERICO DE LA RÚA, 2007). Strictly speaking, attachments or identifications manifest in and are part of social relations and practices that are enacted between real persons in concrete settings (ibid.). [31]

I found a qualitative research design, specifically ethnography, appropriate to use in addressing the research question because it brings the researcher closer to participants, and to what is being studied (DENZIN & LINCOLN, 1994). Unlike quantitative approaches, research techniques used in qualitative designs enable the researcher to elicit data in situ (ibid.), and to capture social processes as they unfold (TAVORY & TIMMERMANS, 2009). [32]

Qualitative researchers, especially those who situate their research in the broader ethnographic tradition, face a key methodological challenge regarding how one constructs a unit of analysis from which empirical data are generated. According to Iddo TAVORY and Stefan TIMMERMANS (2009), this methodological challenge is known as casing, meaning the process of "determining the sociological properties and boundaries of the situation at hand" (p.244). This challenge arises in qualitative research since the researcher has to resolve the following question: What is his or her study a case of? The manner in which I conceptualised the ethnographic case in my own study followed the extended case method approach (TAVORY & TIMMERMANS, 2009). This involved first specifying the theoretical guidelines or principles with which to frame the case. In a broad sense, the activities of constructing an ethnographic case were essentially driven by theory. Granted that viewing migrants strictly as rights-bearing individuals failed to account for their everyday practices, framing the subject of research constituted a productive casing activity. The notion that belonging constitutes both a collective and contextual experience (PROBYN, 1996) shaped my view of research participants as connected agents or relational subjects and not necessarily rights-bearing individuals (ROWE, 2005). This implies that research participants were already in relationships that pre-existed fieldwork. In a wider sense, then, field participants were interactants who most likely shared settings, history, and who knew each other, at least, by reputation (WEBER, 2001). [33]

Thus in my doctoral study, theory not only generated a general picture of the life world of migrants, but it also influenced heavily the processes of case-construction. A priori theoretical framing, such as the way in which I conceptualised research participants, implied that the most appropriate case was a collective. In addition, such as collective must be a conventional unit, one that has emerged out of everyday interactions of Zimbabwean migrants with their non-migrant counterparts in South Africa and abroad (RAGIN, 1992). [34]

At the beginning of my fieldwork, the Forward in Faith Mission International (FIFMI)—a transnational religious movement with origins in Zimbabwe (MAXWELL, 1998)—was the most significant case for the study. The FIFMI originated in the 1950s in the townships of colonial Zimbabwe, and its transnational expansion largely relies on international migrant labour networks (MAXWELL, 1998, 2006). As well, during reconnaissance, I had observed that the FIFMI had a large following of Zimbabweans and a small population of non-Zimbabweans. Also, generally speaking, social relationships of migrant members of the FIFMI variably involved non-migrant family members and friends in Zimbabwe. Theoretically, such non-migrants fell inside the purview of inquiry since they participated in various cross-border exchanges with migrants although they remained in the country of origin. To accommodate non-migrants, I deliberately mapped out the boundaries of the case in a less bounded fashion. Therefore, the FIFMI was a relevant case study in so far as it met the theoretical demands discussed in Section 3.1 above. [35]

Ethnography involves studying people in situ, that is, in their natural set-up. In order to collect data about experiences of Zimbabwean migrants, I used participant observation, a qualitative data collection technique whereby the researcher immerses him- or herself in the everyday life of a particular community or situation (McCALL, 2006). In line with participant observation, I immersed myself in the flow and rhythm of two congregations of the FIFMI, namely City Christian Centre and Berea. The two congregations were purposively chosen because they are located in Johannesburg's inner city area, which is home to the largest Zimbabwean born population (MAKINA, 2010). [36]

Participant observation thus enabled me to venture into Zimbabwean migrants' own "turf, a territory they control numerically so that their ways dominate there" (McCALL, 2006, p.4). Between May 2009 and November 2010, I collected observation data regarding interactions and activities of congregants during Sunday services at City Christian Centre and Berea Assemblies. Such activities gave rise to patterned liturgical practices. Aside from the liturgical aspects of the FIFMI, observations also focused on sociability activities of FIFMI congregants such as braai parties⁴⁾, dinner get-togethers, outings and social football events. [37]

At any given point during fieldwork, my orientation assumed anything between "participant as observer"⁵⁾ and "observer as participant"⁶⁾ (KAWULICH, 2005). Throughout fieldwork, I made no effort to conceal the purpose of my research activity. I recorded proceedings of and interactions between congregants during the Sunday service on tape and camera. Additionally, I kept field notes which I reinforced with my own reflections after an observation period. I found it very obtrusive to record congregants' interactions outside the Sunday service on tape. Hence, I relied more on field notes although I captured some events on still photos. [38]

This article draws on interview data generated from selected congregants of FIMI. Hence, there is no need to elaborate activities of participant observation. In the following paragraphs, I describe the manner in which interview data were collected, including the selection of interview participants. [39]

Unlike observation, semi-structured interviews are strategically designed to elicit a personal narrative of a participant's lived experiences and social relations (KVALE, 2007; NOHL et al., 2006). Since semi-structured interviews are typically less rigid, they enable participants to construct their own narratives (WEISS, 2006). In my study, a prolonged period of observation at the two congregations of City Christian Centre and Berea enabled me to interact with potential interviewees and gain an intimate knowledge about them. In addition, this increased the credibility of the data generated (TAVORY & TIMMERMANS, 2009). [40]

Toward the end of my fieldwork, I selected eight interview participants based on pre-set criteria. First, all selected interview participants were regular migrants who had family members in Zimbabwe and South Africa, and who had been living in South Africa for at least two years. Second, selected interviewees represented different categories of migrant documentation including asylum and temporary work permit statuses. Lastly, selected participants came from different family forms such as families where the migrant lived together with his or her spouse in Johannesburg, those families in which one spouse lived in Zimbabwe. Single parent families in which the selected migrant was either divorced or widowed, or had never married were also considered. In addition to the eight migrants whom I interviewed in Johannesburg, I included three spouses in Zimbabwe. However, their narratives are included here since the illustration in Section 4 below uses narratives of two migrants. [41]

In order to maintain consistency, I used an interview schedule comprising themes to be explored. A full record of the interview schedule is included in the Appendix. The theoretical discussion above showed that identifications are inseparable from the social relations that actors create and sustain during interaction, and which in turn, reproduce them (FEDERICO DE LA RÚA, 2007). Therefore, the interview schedule (Sections 3 & 4, see the Appendix) covered aspects of migrants' ties and identifications. Some of the questions in the interview schedule sought to elicit the following information:

Broadly, three aspects of migrants' relationships were explored: configurations, the form and the content of the ties which migrants create and maintain over a defined period of time; the various practices through which migrants and other social actors enacted their attachments and identifications; and attributes of actors (migrants and non-migrants) whose interactions in micro-social contexts produced multiple forms of inclusion and exclusion. [43]

I conducted two separate interviews with each participant in Johannesburg, totalling three hours on average. Interviewees were free to use Shona⁷⁾ or English or a mix of both during the interview. Each interview followed a unique course. Conversations during interviews were not confined to a uniform question-and-answer pattern such as one might find in a questionnaire for a survey. Different tactics of questioning consisting of timely use of open-ended questions, minimal verbal and non-verbal responses, probing, encouragement, commenting and seeking clarification, and reflexive summaries (KVALE, 2007) elicited rich relational data. [44]

Daniel G. OLIVER, Julianne M. SEROVICH and Tina L. MASON (2005) argue that translation and transcription are important stages in data analysis. I initially recorded all interviews on tape. Next, I transcribed them word for word. This was followed by translation of Shona sections of the interview into English to form complete English texts. [45]

In keeping with the theoretically-driven ethnographic design (TAVORY & TIMMERMANS, 2009) adopted in the study, data analysis followed what Hsiu-Fang HSIEH and Sarah E. SHANNON (2005) term directed content analysis or what Deborah FINFGELD-CONNETT (2013) branded a deductive approach to content analysis. Directed content analysis is a variant of qualitative content analysis, a term that describes a group of analytic approaches used to interpret qualitative data (HSIEH & SHANNON, 2005). [46]

When applying this analytic approach in my study, theoretical premises provided the starting point: assumptions and notions about migrants' life worlds and belonging helped to generate the key methodological concepts as well as specifying relationships between them (ibid.). These concepts include attachments, social relations etc. Thus, conceptual considerations, discussed in Section 3.1 above helped develop a coding template (ATTRIDE-STIRLING, 2001). Identified codes were assigned to chunks of text using computer-based software known as ATLAS.ti (FRIESE, 2012). During the process of dissecting text, I learnt that initial codes left out certain aspects of relational data. This led me to alter the coding framework as suggested by other scholars (ATTRIDE-STIRLING, 2001; FINFGELD-CONNETT, 2013; HSIEH & SHANNON, 2005). [47]

Broadly, data about a migrant's ties were not self-evident; they were embedded in that person's narrative. Therefore, after assigning codes to text, the next step involved pulling out the salient themes in the coded data segments (ATTRIDE-STIRLING, 2001). Then, themes expressed in coded text or those that emerged from analysis were organised into larger categories "broad enough to encapsulate a set of ideas contained in numerous text segments" (p.392). Tables 1 and 2 below illustrate the categories which I eventually used to make sense of relational textual data. These include pre-migration tie, post-migration tie, church-based etc. [48]

Additionally, Tables 1 and 2 illustrate that numbers "zero" and "one" represented the existence or non-existence of a pattern of a given tie, e.g., pre-migration tie, respectively. Whilst a study design such as mine unlikely leads to comprehensive statistical analysis such as in social network methodologies, the incidence of codes (as represented by number "one") was used in rank order comparisons of cases. For example, the incidence of the Non-Zimbabwean tie is not the same for Mashumba⁸⁾ (34) and Rose (32) as shown in Tables 1 and 2 below. These differences suggest that although Mashumba and Rose migrated to South Africa at the height of economic meltdown in Zimbabwe (RUTHERFORD, 2008), in 2008 and 2006, respectively, they followed different relational strategies. Although the narratives of these two interviewees are not only typical, they illustrate clearly the complex and variable manner in which migrants enact inclusion in destination settings. [49]

This section demonstrates how narrative data generated from interviews with Mashumba (34) and Rose (32) were analysed and presented. Both migrants lived in Johannesburg and were members of the FIFMI, a transnational religious movement which originated in Zimbabwe in the late 1950s⁹⁾. Furthermore, both migrants were congregants of the FIFMI in Zimbabwe prior to relocation to South Africa. [50]

Mashumba, a father of two girls and a boy is a man with immense familial responsibilities. Yet in 2006, he was forced to close down his valuation company due to the inflation rate which had skyrocketed to unprecedented levels. Short of options to generate income in Zimbabwe, Mashumba had turned his sights to the neighbouring South Africa. After numerous fruitless interviews with Harare-based agencies that were recruiting for the South African job market at that time, fortune smiled again. In 2007, Mashumba got a job opportunity through an agency in Harare to work in the insurance sector in Johannesburg. Under South African immigration law, however, it is illegal to take up employment in the country without appropriate documentation (LANDAU, 2010b). [51]

Hence, Mashumba had one more bridge to cross. In an unexpected twist of events, however, the recruitment agency tasked Mashumba to arrange his own work permit. If passing an interview was a tall order acquiring a permit was equally demanding. After several futile attempts to get a permit at the South African Embassy in Harare that year, Mashumba got even more desperate. And his despair drove him right into the hands of a fake immigration agent at BeitBridge border post¹⁰⁾. There, he was misled into applying for a South African corporate work permit that was not applicable in the insurance industry. Mashumba only learnt that his permit was inappropriate for the insurance job when he reported for work in Johannesburg. He was in a dilemma: Should he choose to remain in Johannesburg and look for another job, he had no money to rent even the cheapest accommodation. Yet, if he returned, he had no means of generating income in Zimbabwe. As I show in the following analysis, the manner in which he inserted himself in relations of belonging helped him to stay on in Johannesburg until he landed a job. [52]

Unlike Mashumba, Rose (32) was single and without children when she moved to South Africa in 2006. In contrast to Mashumba's situation, Rose easily arranged her study permit. Rose's relocation strategy was to study for a Bachelor's degree in South Africa and job-hunt after completion. In contrast to Mashumba, Rose experienced no delays when she applied for a study permit at South African Embassy in Harare. [53]

To recall, part of the purpose of this article was to suggest and illustrate a qualitative theoretical and methodical approach which accounts for the contextual and contingent nature of migrants' belongings. The study on which this article draws tapped into participants' own understandings of their ties as well as the meanings they attribute to them. This was used as an avenue to make sense of migrants' identifications as in the question, "who do you get along with". Mashumba and Rose's accounts of the close ties which they recreated, formed and sustained in Johannesburg are summarised in Tables 1 and 2, respectively. To borrow from literature on network analysis (KNOKE & YANG, 2008), the two are referred to as focal persons¹¹⁾, and the persons with whom they closely related are alters.

Table 1: Mashumba's account of his close ties in Johannesburg (Please click here for the PDF version of Table 1) [54]

When one considers the incidence of different types of close ties (see Table 1 above), one finds that, after arrival in Johannesburg, Mashumba closely identified with co-nationals who were also congregants, friends and members of his extended family. During the first six months or so after arrival in Johannesburg, Mashumba's close ties comprised of Zimbabweans, most of whom he had known prior to migration. These include Mutize, Catherine, Daniel, Lucky, Magwenzi, Tembo and Muchemwa. With the exception of his connection with Muchemwa, all of these relationships were created in Zimbabwe, before Mashumba relocated to South Africa. [55]

Whereas representing a migrant's ties with numbers enables some level of rank order comparison, the migrant's story reveals more insights regarding that person's ties including the context in which they occurred. This is illustrated in the manner in which Mashumba describes his pre-migration connection with Mutize:

"We were like brothers ... I was his wife's youth [leader] at some point ... in Rusape [Zimbabwe]. I came back to live in Warren Park [Zimbabwe] when I got married. That's where I met Mutize and his wife. That's where we developed our relationship as brothers." [56]

This quote clearly illustrates that Mashumba's close relationship with the Mutize's was created and nurtured in the context of the FIFMI Church in Zimbabwe. At some point prior to migration, both Mashumba and Mutize's wife were congregants at the same church in Rusape. Then, Mashumba was a church youth leader. Although Mashumba got to know Mutize through the latter's wife, Mashumba affectionately compares his connection with Mutize to a family tie. Mashumba's relationship with Mutize can be classified as a pre-migration church-based tie as shown in Table 1 above. The same category applies to Mashumba's ties with Tembo and Muchemwa. [57]

In Johannesburg, Mashumba also inserted himself in networks of extended family members. For example, he reconnected with Catherine and Magwenzi, his aunt and uncle, respectively. Similarly, he renewed ties with friends he had known in Zimbabwe. They include Daniel, Lucky and Tembo. Two years on, Mashumba still identified closely with Mutize, Daniel, Lucky, Tembo, Daniel and Magwenzi. In addition, during his stay in Johannesburg, Mashumba created and nurtured new close ties with Tendekai and Victor, fellow Zimbabweans and congregants at City Christian Centre. As shown in Table 1 above, all of Mashumba's close relationships involved co-nationals. [58]

Rose's description of her own close relationships after migration suggests a slightly different mode of emplacement. Although Rose immediately reconnected with Shamiso and Angela, her childhood friends, just after arrival, she included non-Zimbabweans in her inner circle of friends during the first year. These non-Zimbabweans were Marriane and Yolande, two white South African females who were mentors at her college in Johannesburg. Unlike Mashumba, all persons with whom Rose closely identified during the first six or so months in Johannesburg were neither FIFMI congregants nor members of her extended family. Four years on, none of Rose's close friends were extended family members, although a few were members of the FIFMI. As illustrated in Table 2 below, three of the six friends during fieldwork, that is, Christina, Gontse and Viola, were non-Zimbabweans. Christina, Gontse and Viola, were South African women with whom Rose had connected at work. The extent to which Rose valued her close ties with non-Zimbabwean friends is salient in the way she spoke about Christina:

"The other [friend of mine] works at the [name of a newspaper]. She is doing media liaison there ... Christina. At some point she even helped me to find a job ... we do business together. But she [has become] more of a friend because she is more helpful to me ... There was a time ... she was in hospital, I went to see her there." [59]

Although Rose and Christina, a White Afrikaner who worked at the weekly newspaper, connected at work, their friendship had become more personal. I classified this relationship as a post-migration occupation-based tie in Table 2 below.

Table 2: Rose's account of her own close ties (Please click here for the PDF version of Table 2) [60]

Aside from revealing migrants' identifications, an analysis of migrants' close ties has the potential to shed light on the role of economic insecurity in the way in which migrants' insert themselves in local networks. Tables 1 and 2 show that both Mashumba and Rose received instrumental support, that is, help exchanged to address practical problems such as paying monthly rental. On top of that, the two received expressive support, meaning help that nurtures feelings of comfort and leads to an individual belief that one belongs, and is loved, admired and respected. Mashumba's description of his friendship with Tembo below exemplifies instrumental support:

"Mr Tembo helped me quite remarkably. I could [go] around the city ... with him. When I went for my interviews I knew little about Joburg [Johannesburg]. Tembo is into marketing, so he knew a lot of places. So he could go out of his way to drop me off at a couple of places. He also helped me with ZAR¹²⁾ 1000." [61]

Expressive support is illustrated in Rose's description of her friendship with Patricia: "We talk the same language ... She is somebody I can confide in. And she is one person who can really stand up to me and say in my face, 'this is wrong, this is not right'". The importance of close ties as sources of instrumental support for the two migrants in times of need is demonstrated by migrants' perceptions of perceived support. Both Mashumba and Rose indicated that they intended to approach persons with whom they closely related in the event that they failed to pay for their monthly rental. This is illustrated in Tables 1 and 2. Unlike Mashumba, Rose used material exchanges from her close ties with persons positioned outside her national, ethnic and religious boundaries of belonging to address material insecurities. Thus, Rose's account of close ties demonstrates that economic insecurity may influence migrants' identifications with persons outside a migrant's national and ethnic group. [62]

I began this article by demonstrating that both the demise of formal apartheid in South Africa in 1994 and subsequent relaxation of restrictions on immigration provided conditions for the emergence of new migrant spaces in the country. In these new spaces, migrants' quotidian practices spawn numerous but under-researched social formations. The social dynamics connected to new migrants' solidarities constitute an anomaly which requires modification of existing theoretical and quantitative methodological tools for researching migration. Since existing theoretical approaches largely draw on nation-state frames of reference, their value lies in revealing larger structures that shape migration experience rather than illuminating place-specific modes of emplacement among migrants. [63]

A revision of the researcher's assumptions about migrants' life worlds is essential if one is to capture the variable practices and constructions of belonging of migrants. In my doctoral study, I started thinking about migrants' life worlds with the assumption that social life unfolds within relationships between persons (THOMAS, 2005). In the same breath, I imagined migration as essentially a context-specific dynamic (SLÁDKOVÁ & BOND, 2011) and not a homogeneous experience. This tactical move, that is, privileging social relations in the study of experience, directs inquiry away from state-driven citizenship and immigration regimes toward the marginal life worlds of migrants. As illustrated above, this theoretical move prompts a view of belonging as a situated and contingent experience (PROBYN, 1996). [64]

From this optic, the subject of research is no longer constructed as a rights-bearing individual for whom formal rights and entitlement signal fixity and social membership. Rather, the participant, in this case the migrant, is constructed as a relational or connected agent whose social ties are integral to any analysis to belonging. Thus, modes of identification, and not necessarily citizenship frameworks provide migrants with opportunities for constructing a locus of belonging in the destination context. [65]

This theoretical posture, I argued, has key implications for the collection and analysis of data. When migrants are constructed as relational agents the analysis of belonging ought to account for both the participants' relationships and the context within which they unfold. Hence, participants must be in relationships prior to fieldwork (WEBER, 2001). Ethnography, used with participant observation and semi-structured interviews, enabled me to interact with research participants long enough to grasp the "sense in which experience is situated within relationships and between persons" (HORTON, 2008, p.22). Any other designs that enable the ethnographer to interact with research participants for prolonged periods are suitable for collecting relational data that is trustworthy. [66]

The challenge, however, is to delineate the boundaries of a relevant case from which to generate evidence sufficient to make claims about migrant's constructions of belonging. To address this challenge, I adopted a theoretically-driven ethnography, which draws on an extended case method approach (TAVORY & TIMMERMANS, 2009). In the foregoing, I illustrated that a priori theoretical framing, e.g., the conceptualisation of research subjects and their life worlds, is a key exercise in the construction of a relevant case. Thus, a case is relevant to the extent that it meets the theoretical demands of the research question. [67]

Indeed theory plays a central role in the approach described above. Theory not only provides a picture of the migrant way of life, it is useful during the analysis of narrative data. As shown above, it helps the researcher to develop a set of initial codes with which to make sense of relational data embedded in text. Whereas, some might argue that theory may encourage the researcher to "verify the obvious" (FINFGELD-CONNETT, 2013, p.2), it is normal procedure to up-date initial codes in view of further analysis. [68]

This article demonstrates that narrative data generated via semi-structured interviews enable the researcher to analyse social relations of migrants and to "retain and examine the context in which those relationships occur" (McKETHER, GLUESING & RIOPELLE, 2009, p.155). What emerges from the above discussion is that analysing migrants' ties without losing the context in which they unfold holds immense potential for revealing processes of insertion and self-exclusion that underpin migrants' constructions of belonging. This article demonstrated that stories or excerpts of a migrant's narrative reveal more about that person's relationships, perhaps, more than statistical and graphical representations in social network analysis. The stories migrants tell reveal a lot about their aspirations and longings, in ways that statistics cannot. Therefore, stories rather than numbers or arrows have potential to represent a migrant's valuable relationships and their context at once. [69]

Analysis of a migrant's close ties has potential to generate evidence of fixity and embeddedness of that person in relations of belonging at destination, not just in the country of origin. Interrogating migrants' ties also has potential to investigate the role post-migration material and emotional insecurity plays in the way in which migrants embed themselves in relations of belonging at the local level. While not conclusive, the narratives of the two participants considered above suggest that, in the face of economic insecurities, it is possible that migrants may strategically insert themselves in local networks in a manner that transcends national, ethnic and religious belonging. [70]

Zimbabwean economic migrants in Johannesburg: Transnationality and its implications for social policy.

I am a Zimbabwean student studying at the University of Johannesburg. I am studying for a Doctorate in Social Work.

I am conducting research on Zimbabwean migrants in Johannesburg who have close family members in their country of origin. The purpose of my research is to understand the experiences of Zimbabwean migrants living in Johannesburg. I am interested to hear about how the experience of migration affects you, your family and your social and community relations in Johannesburg and at home.

It is my hope that this information may be helpful to all those who would like to improve the lives of migrants and their families such as social service organisations e.g., welfare, religious and governmental organisations.

The issues that we discuss during the interviews will be kept confidential. Your real name and that of your family and friends who speak with me will not be made known to anyone. I would like to speak with you over two interviews.

[If interviewee is a migrant men with a spouse in Zimbabwe: I would also like to interview your spouse at home in Zimbabwe and to learn about their experiences.]

If you feel that you are not comfortable to answer certain questions during the interview, please let me know.

I have also brought a tape recorder along. I would like to record our conversation if you agree to it. The reason is that it will help me to take down your information and your story as accurately as possible.

Would you mind if I use it? [To record notes of participant's consent to recording]

2.2 You have been in Johannesburg for a while now; can we talk about what your life is like now?

3.1 Could we now talk about your relations with people and your involvement with specific organisations in Johannesburg?

3.2 I would also like to know about the friends and relations with people at home.

4.2 How do you cope with living apart from members of your family and extended family?

4.3 Has living apart from members of your family, particularly your children, spouse or elderly parents affected your relationships with them?

Section 5: Family responsibilities, obligations, reciprocity, loyalties and migrant autonomy

5.1 Could we now talk about the ways in which you assist members of your own family and your extended family at home [the extended family includes your siblings, their children, and your aunties, uncles, their children etc., and your spouse's parents, siblings, aunties, uncles etc.). Such assistance includes tasks, duties and roles such as provision of food, clothing school fees, accommodation).

5.2 What role do you play in the provision and care of your children while you are away?

5.3 Has your role as a parent changed since you came to South Africa? If yes or no, please explain how this has changed.

5.4 Which other household activities did you perform before you relocated to South Africa?

5.5 You are living in South Africa, hundreds of miles away from your family members. I would like us to talk about how this has changed how your family makes decisions about important family matters.

5.6 Do you think that your spouse or members of your family have assumed more responsibilities for family matters since you relocated to South Africa?

6.1 I would like us to talk about your visits to your family members in Zimbabwe.

6.2 Let us now talk about members of your family and your extended family who have visited you in South Africa in the last 12 months.

6.3 How many times have you sent money and goods to any member of your family and extended family in the last 24 months?

6.4 How do you maintain contact with members of your family and extended family?

6.5 How often do you communicate with members of your family and extended family?

6.6 Which family ceremonies such as a funerals or weddings in Zimbabwe did you attend since you moved to South Africa?

7.1 Let us talk about the sources of your family income and the needs of your families in South Africa and in Zimbabwe?

7.4 If you have family members with you in Johannesburg, do they have access to public services?

7.6 Does your family access welfare/social services from non-state organisations?

7.7 Does your family access welfare/social services from non-state organisations in Zimbabwe?

7.8.1 Let us talk about disposable assets that your household owns here in South Africa and in Zimbabwe. These are the assets that could be sold at times of stress like residential and industrial properties e.g., stands, plots, houses, cars, livestock, farm implements. We begin with the household you own in South Africa.

7.9 I would like to know if your household participates in any form of rotating credit scheme or burial society here in South Africa and in Zimbabwe. We begin with South Africa.

8. In the event that you fail to work for your family due to invalidity or accident, have you put any arrangements to see your family through?

8.1 Does your household have access to credit provider from banks, or any other loan facilities here in South Africa and in Zimbabwe?

I wish to thank the ZEIT-Stiftung Ebelin und Gerd Bucerius for a generous scholarship awarded during the Ph.D. study on which this article is based. Funding from the Pierre Joubert Fund and the University of Johannesburg Merit Awards is also acknowledged. I would also like to thank two anonymous reviewers for their valuable comments on previous drafts.

1) According to ELLIS (2008), 15.4 million out of a total population of 47.4 million are considered very poor in South Africa. And what is more, over a third of the total population lives below the poverty datum line. In addition, the official unemployment rate is 26.7 per cent, although most estimates put the figure at 40 per cent. <back>

2) The Limpopo River forms the border between Zimbabwe and South Africa. It is often depicted in public discourse as the boundary between South Africa and the rest of the continent to the North. <back>

3) Studies reviewed in Section 2.2 exemplified this dimension of belonging. They all illuminate the mechanisms through which the post-apartheid South African state seeks to enforce criteria for exclusion from its polity and territory. <back>

4) Outdoor parties or social gatherings held outside at which meat is cooked over an open flame. <back>

7) The main native language or vernacular of the Bantu people of Zimbabwe. <back>

9) In Zimbabwe, FIFMI is popularly known as Zimbabwe Assemblies of God Africa (ZAOGA). <back>

Alexander, Peter (2001). Oscillating migrants, "detribalised families" and militancy: Mozambicans on witbank collieries, 1918-1927. Journal of Southern African Studies, 27(3), 505-525.

Attride-Stirling, Jennifer (2001). Thematic networks: An analytic tool for qualitative research. Qualitative Research, 1(3), 385-405.

Borkert, Maren; Martín Pérez, Alberto; Scott, Sam & De Tona, Carla (2006). Introduction: Understanding migration research (across national and academic boundaries) in Europe. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 7(3), Art. 3, http://nbn-resolving.de/urn:nbn:de:0114-fqs060339 [Date of access: June 12, 2013].

Chereni, Admire & Palmary, Ingrid (2010). Social development and migration in South Africa: Directions for further research in the region. The Social Work Practitioner-Researcher, 22(1), 1-16.

Christensen, Anne-Dorte (2009). Belonging and unbelonging from an intersectional perspective. Gender, Technology and Development, 13(1), 21-41.

Croucher, Sheilla (1998). South Africa's illegal aliens: Constructing national boundaries in a post-apartheid state. Ethnic Racial Studies, 21(4), 639-660.

Crush, Jonathan & McDonald, David A. (2000). Introduction. Canadian Journal of African Studies / Revue Canadienne des Études Africaines, 34(1), 1-19.

Crush, Jonathan & Tevera, Daniel (2010). Exiting Zimbabwe. In Jonathan Crush & Daniel Tevera (Eds.), Zimbabwe's exodus: Crisis, migration, survival (pp.1-49). Cape Town: Southern African Migration Programme.

Danso, Ransford & McDonald, David A. (2001). Writing xenophobia: Immigration and the print media in post-apartheid South Africa. Africa Today, 3, 115-137.

De Tona, Carla (2006). "But what is interesting is the story of why and how migration happened". Ronit Lentin and Hassan Bousetta in conversation with Carla De Tona. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 7(3), Art. 13, http://nbn-resolving.de/urn:nbn:de:0114-fqs0603139 [Date of access: April 20, 2013].

Denzin, Norman K. & Lincoln, Yvonna S. (Eds.) (1994). The Sage handbook of qualitative research. Thousand Oaks, CA: Sage.

Ellis, Stephen (2008). Introduction: Migration in post-apartheid South Africa. In Aurelia Wa Kabwe-Segatti & Loren B. Landau (Eds.), Migration in post-apartheid South Africa: Challenges and questions to policy-makers (pp.47-53). Paris: Research Department Agence Française de Développement.

Federico de la Rúa, Ainhoa de (2007). Networks and identifications: A relational approach to social identities. International Sociology, 22(6), 683-699.

Finfgeld-Connett, Deborah (2013). Use of content analysis to conduct knowledge-building and theory-generating qualitative systematic reviews. Qualitative Research, pre-published April 2013, 1-12, http://qrj.sagepub.com/content/early/2013/04/03/1468794113481790.full.pdf [Date of access: May 15, 2013].

Friese, Susanne (2012). Qualitative data analysis with ATLAS.ti. Los Angeles, CA: Sage.

Glick Schiller, Nina & Fouron, Georges E. (1999). Terrains of blood and nation: Haitian transnational social fields. Ethnic & Racial Studies, 22(2), 340-366.

Gray, Mel (2006). The progress of social development in South Africa. International Journal of Social Welfare, 15(Suppl.1), S53-S64.

Horton, Sarah (2008). A mother's heart is weighed down with stones: A phenomenological approach to the experience of transnational motherhood. Cult Med Psychiatry, 33, 21-40.

Hsieh, Hsiu-Fang & Shannon, Sarah E. (2005). Three approaches to qualitative content analysis. Qualitative Health Research, 15(9), 1277-1288.

Kawulich, Barbara B. (2005). Participant observation as a data collection method. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 6(2), Art. 43, http://nbn-resolving.de/urn:nbn:de:0114-fqs0502430 [Date of access: March 17, 2012].

Knoke, David & Yang, Song (2008). Social network analysis (2nd ed.). Los Angeles, CA: Sage.

Landau, Loren B. (2010a). Passage, profit, protection and the challenge of participation building and belonging in African cities. ANU-WIDER, Working paper No. 2010/49, World Institute for Development Economic Research, http://www.wider.unu.edu/publications/working-papers/2010/en_GB/wp2010-49/ [Date of access: December 10, 2012].

Landau, Loren B. (2010b). Loving the alien? Citizenship, law and the future in South Africa's demonic society. African Affairs, 109, 213-230.

Landau, Loren B. & Freemantle, Iriann (2010). Tactical cosmopolitanism and idioms of belonging: Insertion and self-exclusion in Johannesburg. Journal of Ethnic and Migration Studies, 36(3), 375-390.

Levitt, Peggy & Glick Schiller, Nina (2004). Conceptualizing simultaneity: A social field perspective on society. International Migration Review, 2, 1002-1039.

Makina, Daniel (2010). Zimbabwe in Johannesburg. In Jonathan Crush & Daniel Tevera (Eds.), Zimbabwe exodus: Crisis, migration, survival (pp.225-243). Cape Town: Southern African Migration Project.

Mawadza, Aquilina & Crush, Jonathan (2010). Metaphors of migration: Zimbabwean migrants in the South African media. In Jonathan Crush & Daniel Tevera (Eds.), Zimbabwe's exodus: Crisis, migration, survival (pp.364-374). Cape Town: Southern African Migration Project.

Maxwell, David (1998). "Delivered from the spirit of poverty?" Pentecostalism, prosperity and modernity in Zimbabwe. Journal of Religion in Africa, 28(3), 350-373.

Maxwell, David (2006). African gifts of the spirit: Pentecostalism and the rise of a Zimbabwean transnational movement. Oxford: Weaver Press.

Mayer, Philip (1962). Migrancy and the study of Africans in towns. American Anthropologist, 64(3), 576-592.

McCall, George J. (2006). The fieldwork tradition. In Dick Hobbs & Richard Wright (Eds.), The Sage handbook of fieldwork (pp.3-21). London: Sage.

McKether, Willie L.; Gluesing, Julia C. & Riopelle, Kenneth (2009). From interviews to social network analysis: An approach for revealing social networks embedded in narrative data. Field Methods, 21(2), 154-180.

Mlambo, Alois S. (2010). A history of Zimbabwean migration to 1990. In Jonathan Crush & Daniel Tevera (Eds.), Zimbabwe's exodus: Crisis, migration, survival (pp.53-76). Cape Town: Southern African Migration Project.

Mosselson, Aidan (2010). "There is no difference between citizens and non-citizens anymore": Violent xenophobia, citizenship and the politics of belonging in post-apartheid South Africa. Journal of Southern African Studies, 36(3), 641-665.

Nohl, Arnd-Michael; Schittenhelm, Karin; Schmidtke, Oliver & Weiss, Anja (2006). Cultural capital during migration—a multi-level approach for the empirical analysis of the labor market integration of highly skilled migrants. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 7(3), Art. 14, http://nbn-resolving.de/urn:nbn:de:0114-fqs0603143 [Date of access: July 8, 2013].

Oliver, Daniel; Serovich, Julianne M. & Mason, Tina L. (2005). Constraints and opportunities with interview transcription: Towards reflective practice in HIV/AIDS intervention research. Social Forces, 84(2), 1273-1289.

Palmary, Ingrid & Landau, Loren B. (2009). Social protection of migrants in South Africa. Report prepared for the Institute for Development Studies. Johannesburg: African Centre for Migration and Society.

Patel, Leila (2005). Social welfare and development in South Africa. Cape Town: Oxford University Press Southern Africa.

Peberdy, Sally (2001). Imagining immigration: Inclusive identities and exclusive policies in Post-1994 South Africa. Africa Today, 48(3), 15-32.

Potts, Deborah (2010). Circular migration in Zimbabwe and contemporary Sub-Saharan Africa. Suffolk: James Currey.

Prothero, Mansell R. (1974). Foreign migrant labour for South Africa. International Migration Review, 8(3), 383-394.

Ragin, Charles C. (1992). Casing and the process of social inquiry. In Charles C. Ragin & Howard S. Becker (Eds.), What is a case? Exploring the foundations of social inquiry (pp.217- 226). Cambridge: Cambridge University Press.

Ramírez Goicoechea, Eugenia (2005). Immigrants contesting ethnic exclusion. Structures and Practices of Identity, 29(3), 654-669.

Richter, Marina (2012). Researching transnational social spaces: A qualitative study of the Spanish second generation in Switzerland. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 13(3), Art. 8, http://nbn-resolving.de/urn:nbn:de:0114-fqs120385 [Date of access: July 8, 2013].

Rosenthal, Gabriele & Michaela, Köttig (2009). Migration and questions of belonging. Migrants in Germany and Florida. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 10(3), Art. 19, http://nbn-resolving.de/urn:nbn:de:0114-fqs0903191 [Date of access: May 3, 2013].

Rowe, Aimee C. (2005). Be longing: Toward a feminist politics of relation. NWSA Journal, 17, 15-46.

Rutherford, Blair (2008). An unsettled belonging: Zimbabwean farm workers in Limpopo Province, South Africa. Journal of Contemporary African Studies, 26(4), 401-415.

Sládková, Jana & Bond, Meg A. (2011). Migration as a context-dependent dynamic in a world of global inequalities. Psychosocial Intervention, 20(3), 327-332.

Smith, Michael P. & Guarnizo, Eduardo L. (Eds.) (1998). Transnationalism from below, comparative urban and community research. New Brunswick: Transaction Publishers.

Tavory, Iddo & Timmermans, Stefan (2009). Two cases of ethnography: Grounded theory and the extended case method. Ethnography, 10(3), 243-263.

Thomas, Julian (2005). Materiality and the social. In Pedro A. Paulo; Funari Andrés Zarankin & Emily Stovel (Eds.), Global archaeological theory: Contextual voices and contemporary thoughts (pp.11-18). New York: Kluwer Academic / Plenium.

United Nations Department of Economic and Social Affairs (UN DESA) (2009). Trends in international migrant stock: The 2008 revision, http://esa.un.org/migration/index.asp?panel=1 [Date of access: June 12, 2012].

Vigneswaran, Darshan (2008). Undocumented migration: Risks and myths (1998-2005). In Aurellia Wa Kabwe-Segatti & Loren B. Landau (Eds.), Migration in post-apartheid South Africa: Challenges and questions to policy-makers (pp.135-162). Paris: Research Department Agence Française de Développement.

Weber, Florence (2001). Settings, interactions and things: A plea for multi-integrative ethnography. Ethnography, 2(4), 475-499.

Weiss, Anja (2006). Comparative research on highly skilled migrants. Can qualitative interviews be used in order to reconstruct a class position? Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 7(3), Art. 2, http://nbn-resolving.de/urn:nbn:de:0114-fqs060326 [Date of access: March 7, 2013].

Yuval-Davis, Nira (2006). Belonging and the politics of belonging. Patterns of Prejudice, 40(3), 197-214.

Admire CHERENI is a researcher at the Centre for Anthropological Research, University of Johannesburg, South Africa. His Ph.D. thesis, entitled "Zimbabwean Economic Migrants in Johannesburg: Transnationality and its Implications for Social Policy", examined Zimbabwean migrants' notions and practices of belonging and family. It also interrogated informal social protection strategies which migrants employed in order to address risk and vulnerability in a transnational context.

Centre for Anthropological Research, House 10, Research Village
Bunting Road Campus, University of Johannesburg
Box 524, Auckland Park,
2006 Johannesburg, South Africa

Chereni, Admire (2013). A Methodological Approach and Conceptual Tools for Studying Migrant Belongings in African Cities: A Case of Zimbabweans in Johannesburg [70 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 15(1), Art. 6,
http://nbn-resolving.de/urn:nbn:de:0114-fqs140163.

Forum Qualitative Sozialforschung / Forum: Qualitative Social Research (FQS)

ISSN 1438-5627

Creative Commons Attribution 4.0 International License