Transnational childrearing and the preservation of transnational identity in Brazzaville, Congo

Abstract

Drawing on ethnographic research conducted at two ends of an intra- Africa migration flow (Mali and the Republic of Congo), in this article I examine the role of childrearing practices in the maintenance of transnationalism. I consider different approaches to transnational childrearing by migrant parents and their reasons for adopting them, and delineate three common modes. The most widespread and socially validated approach is to send children home from Congo to their parents’ places of origin, where child fostering is widespread, to be raised by relatives for long periods; this approach increases the durability of transnational ties. I use childrearing approaches as an analytical lens to demonstrate the complementarity of multiple forms of domestic organization, mobility and settlement in the intergenerational production and transmission of durable transnational identities. By arguing for greater focus on phenomena such as transnational childrearing, I seek to promote a broader conceptualization of transnationalism.

Some of the most important questions about transnationalism today concern its durability and accessibility: to what extent can migrants pass transnational activities on to their children and grandchildren, and who can take part in them? Researchers in this area have been unable to reach a consensus on these questions, in part because transnationalism is simply too recent in many settings to be analysed as an intergenerational phenomenon. Another problem, however, is that different scholars define transnationalism in different ways. Narrowly associating the concept with regular physical movement across borders, Portes (2001, 2003; Portes et al. 2002) states that transnationalism in the USA is still limited to a small, rather privileged minority of first-generation immigrants. By contrast, those who conceptualize transnationalism in broader terms (for example Itzigsohn et al. 1999; Jackson et al. 2004), taking symbolic and cultural dimensions into account, leave open the possibility of a more diverse group of participants that can include members of the second and subsequent generations.

My goal in this article is to demonstrate the utility of broader definitions of transnationalism encompassing child-rearing activities that extend across national boundaries. I examine household-level transnational connections within a West African migrant population in the Republic of Congo, describe their transmission between generations, and analyse their availability to various groups of people as well as their impact on identity formation. My primary focus is on the ways migrants raise their children in transnational settings. How parents expose a young child to their people, culture, language and/or place of origin has tremendous influence over that child’s socialization, enculturation, language acquisition, and identification. The migrants I study employ different modes of transnational parenting to produce childrearing outcomes deemed desirable in their communities.

Ethnographic research for this article was carried out in Brazzaville, Congo (2005–2006) and in Mali (2002 and 2006), both in the capital Bamako and in the rural community of Togotala (pseudonym) located 160 kilometres north of Bamako. This research built on my three years of previous experience living and working in Mali. With one assistant I conducted in-depth interviews in the French and Bamanan languages with approximately 140 current migrants in Congo and 20 returned migrants in Mali; 98 interviewees were parents of children born in the Congo. Interviewees represent a non-random sample of this migrant population. While we deliberately oversampled parents, we tried to balance our selection of interviewees by gender, ethnicity, economic activity, place of origin and time spent in Congo.

The origins of Congo’s West African population

Although there are few reliable statistics on migration between African countries, evidence indicates that intra-African migration flows involve more migrants than the flows from African to non-African countries. Ba (1996) reckons that 80 per cent of all Africans who leave their home countries remain on the same continent. Even in the Senegal River Valley, renowned for being the origin of labour migrants to France (Condé et al. 1986; Kane and Lericollais 1975; Quiminal 1991), two to three times as many people leave for other African countries as leave for Europe (Ba and Bredeloup 1997). Intra-African migration also has historical depth: unlike many contemporary South–North flows that began in the 1960s or later, migration flows within Africa are often several generations old, and some began well before European colonization of the continent. Intra-African migrants have long formed ‘permanent stranger communities’ (Skinner 1963: 308; see also Shack and Skinner 1979) transmitting ‘Othered’ cultural identities from one generation to the next. Some of these migration flows are circular, like that linking Burkina Faso with Côte d’Ivoire (Cordell et al. 1998), but others fit better with prevailing definitions of transnational migration (for example Glick Schiller et al. 1995: 48) and may not entail definitive return.

West Africans in contemporary Congo constitute an intra-African migrant population with deep historical roots. Shortly after France began bringing colonial troops and orderlies from its West African possessions to Equatorial Africa in the late 1800s, these African personnel created a place for themselves in the newly acquired Congo colony (Balandier 1985; Gondola 1996; Manchuelle 1987). Many stayed on and became traders after their work contracts ended. They brought families to join them from their home communities (located mostly in what are today Senegal, Mali and Guinea). Since the beginning of the twentieth century, Congo’s capital city of Brazzaville has been home to the largest concentration of these migrants in Central Africa. Over generations, they formed a densely-knit community set apart from Congolese society while also inextricably intertwined with it. Despite their diverse nationalities, ethnicities and origins, these West Africans have two important cultural characteristics in common: they are almost exclusively Muslim, and come from an area of the West African Sahel with a mutual cultural heritage issuing from the Mali Empire (c. thirteenth to seventeenth centuries). Most speak languages of the Mandé family (especially Bamanan and Soninke) and are more oriented towards entrepreneurial activities than salaried employment or wage labour. When I refer to West Africans here, I specifically mean the members of this heterogeneous group.¹ While there are no reliable official statistics on the size of this population, reasonable estimates by community leaders put it at 20,000 or more in Brazzaville, with another 10–20,000 in other parts of Congo.

Legal or logistical barriers for foreigners wishing to travel to and reside in Brazzaville are few. Congolese entry visas – unlike those for Western countries – are available in various African capitals to anyone willing to pay the application fee, which is not prohibitively expensive for most applicants. Although Congolese residence permits are costly – about US$ 200 per person per year – enforcement is lax and, in practice, foreigners who wish to stay in the Congo without one often do so. Without severe official constraints on people moving between West Africa and the Congo, migrants in the Congo have developed various transnational connections with their communities of origin and maintained them over the long run. For these reasons, this migrant population constitutes a useful case in the study of transnationalism, offering insight into the durability and accessibility of transnational ties.

A transnational social field and its inhabitants

Brazzaville is only one of many cities to which West African migrants go; migration flows from Mali, Guinea, Senegal and other countries of the western Sahel extend to cities throughout Africa, Europe, North America and increasingly Asia. Several of my informants had previously lived outside their country of birth, particularly in Côte d’Ivoire, prior to migrating to the Congo. Some become successful entrepreneurs, members of the global business class holding multiple passports, owning multiple homes in the Congo, West Africa and elsewhere, and regularly travelling to other continents in the conduct of their affairs.

Other West African migrants, such as so-called aventuriers, are less fortunate. These are young single men often just starting their migratory careers with scarce economic resources. For them, Brazzaville is more a transit point than a destination. Repeated outbreaks of political violence, political instability and a stagnating commercial sector since the early 1990s (Clark 2008; Eaton 2006) have relegated the Congo to a fallback position in the international migration circuit. Aventuriers, usually Malians and Guineans in their twenties, regularly pass through Brazzaville to seek their fortunes in the diamond fields of Angola and Congo-Kinshasa. Many of my younger informants were heading overland to Angola when difficulties forced them to settle in Brazzaville and regroup for another attempt; others were turned back or deported from Angola or Congo-Kinshasa, and a few came to the Congo only after unsuccessful efforts to reach Europe via North Africa. These undocumented migrants’ rights are systematically violated and they are subject to abuse ranging from beatings to confiscation of property by the authorities of the states whose territory they enter.

My analysis of transnational linkages includes aventuriers as well as well-off, well-connected traders. It also extends beyond the most mobile components of this population to more sedentary individuals, including members of the second and third generations. As Hondagneu-Sotelo and Avila (1997) have noted, studies of transnationalism often privilege circulation over settlement, obscuring important aspects of transnational communities from view. Rather than restrict my investigations to mobile elements of a population or even to a single type of mobility, I consider the ‘full house’ of variation (Ferguson 1999: 42) in movement and settlement patterns. My study population includes circular migrants, sojourners and settlers, as well as some who never move at all. The focus of the study of transnationalism is not strictly mobility, but rather how people use transnational practices to stay connected with communities in which they no longer live and from which national boundaries separate them.

A useful concept for envisioning transnational populations in all their diversity is the transnational social field. This exists in migrants’ minds and in the behaviours interlinking them across physical space. Itzigsohn et al. (1999: 317) describe it as ‘a field of social interactions and exchanges that transcend political and geographical boundaries of one nation’ (see also Levitt 2001). Within the transnational social field linking Brazzaville’s West African migrants with their home communities, individuals, goods and ideas flow from one location to another by way of telecommunications, air travel and overland transport. The transnational social field concept encompasses all types of connection, not merely those entailing physical movement.

While much research on transnationalism concentrates on the highly organized upper stratum of hometown associations, political parties and the like (Portes 2001, 2003), my work here fits into the growing literature on transnational families that examines transnational connections operating through households and kin groups. These ties, characterized by informality rather than institutionalization, play a vital but underappreciated role in facilitating the continuity of the transnational social field.

Households, kin groups and children in the transnational social field

In 2002, during fieldwork in the Malian town of Togotala, I interviewed a middle-aged Togotalan woman named Kadi. Her father had gone to Brazzaville in the 1940s, where Kadi was born and stayed for three years before being sent to Togotala to be raised by paternal relatives. She spent the rest of her childhood there without seeing her biological parents. When she was 19 Kadi married a Togotalan man who subsequently emigrated to Brazzaville; three years later he brought her to join him. Only then was she reunited with her mother and father from whom she had been separated since she was a toddler. Kadi told me she had spent 13 years as a wife and mother in Brazzaville, during which time she sent all her own children to Mali between the ages of two and four years so as to shield them from the cultural influences of Congolese society. She only saw these children again when she returned to Mali definitively in her mid-thirties.

Kadi’s case challenges certain received notions of parenthood and family. She and her husband removed their children from Brazzaville’s social environment explicitly to safeguard their cultural identities, just as her parents had done with her. By fostering their offspring out to relatives in Mali, they could ensure that these children grew up speaking their people’s Soninke language, practising their people’s Muslim faith, and experiencing their people’s home community both as a physical and a social space. Parents like Kadi put the social reproduction of their home community before their right to raise their own children. She and many other returned migrants assured me that this practice was common, even the norm for Togotalans living abroad. Beginning fieldwork in Brazzaville a few years later, I wanted to discern just how widespread this pattern of transnational parenting was among West African migrants living there, and what alternatives to it might exist.

Brazzaville’s West African population comprises people at all stages of life, living in multiple types of households. The most visible component of this population is the merchants, who are male and mostly middle-aged, and their younger employees (also male). A number of young West African men also work as ambulant vendors or labourers. Many are single and share rented houses but take their meals with more established West African households where several women cook in rotation. Traders are mostly married, as are some shop assistants and labourers. They tend to marry in West Africa and bring their wives after becoming established in Brazzaville; any children they have before emigrating usually remain behind in West Africa. Other migrant men only get married after going to Brazzaville, receiving brides sent from their home community or finding wives within the migrant West African population. A few marry Congolese women. Finally, some men who marry prior to emigrating opt to leave their wives behind in West Africa, but this arrangement is unpopular: many informants felt that it puts husbands in the position of being tempted to engage in illicit sexual liaisons with local women, thus committing the grave sin of fornication. (There is much less public discussion, however, of the situations of wives left behind.) The most commonly expressed ideal is for husbands and wives to cohabit in the migration destination, especially while the wives are of childbearing age.

Women in this population are represented as dependent migrants: unmarried women move only at the discretion of their fathers, and married women only at the discretion of their husbands. Since women in Sahelian societies are supposed to be provided for economically by their male kin, autonomous female mobility for economic purposes is stigmatized, even to the point of being equated with prostitution (Ba 2002, 2003; Findley 1994). Some women nonetheless do migrate by what amounts to choice: an unmarried woman might arrange to visit her sister abroad and during that visit marry a migrant, thus prolonging her stay for an indefinite period. Such cases are rare, however, and women almost never ascribe their moves to their own agency.² West African women seldom trade in Brazzaville’s markets, although they often trade informally out of their homes.

Polygamy is common in West Africa and among West Africans in Brazzaville. For a migrant husband with multiple wives, several domestic arrangements are possible. Most commonly one wife joins him abroad while the other (or others) remains at home; after a time – at least one year, sometimes more – the wives change position. This arrangement is described as ‘rotation’, in much the same way as the women of an extended family take turns with household cooking, and stems in part from Islamic requirements that a man give his wives equal treatment. Alternately, a husband may live with multiple wives in the migration destination, but this is a more expensive arrangement for West Africans in Brazzaville (where the cost of living is much higher than in their places of origin) and is limited to the wealthiest merchants’ households. These merchants are among the most established members of the migrant community: they may have interests in real estate, customs brokerage, import–export, or other activities, while a few have been involved in the cross-border diamond trade (see Bredeloup 1994, 1995).

In discussing the childrearing practices prevalent in this transnational population, I must be clear about my terminology. I do not refer to these practices as ‘strategies’ because they are not always strictly intentional, but are, as we shall see, mediated by a household’s available means. Nor do I call them ‘types’, which would suggest a typology of discrete behaviours rather than the full spectrum (or ‘full house’) described above. The spectrum, in this case, signifies the wide range of exposure foreign-born children receive to their parents’ home community and country; I use the term ‘modes’ to designate common childrearing approaches clustered around particular points along that spectrum. The three modes I identify below highlight some of the most popular practices my informants and their parents employed, but do not constitute an exhaustive list of available practices.

Maximal mode: maximum exposure to the home community

The life of Kadi, the woman cited above, shows one approach to transnational childrearing: when their offspring are still in early childhood, preferably below schooling age, parents buy them a one-way ticket and send them on a plane with a relative or family friend to accompany them to their parents’ home community, where they usually remain until adulthood. This approach, which I call the ‘maximal mode’, represents the highest degree of exposure to the homeland. Most informants described the maximal mode as the preferred one, and indeed 39 of 99 West African parents interviewed practised it, while another 26 parents stated a firm intention to adopt this mode when their children (mostly infants and toddlers) were old enough. Parents are motivated by the conviction that children cannot grow up ‘properly’ outside their ancestral homeland, and that West African children who do grow up abroad will turn out badly, no matter what pains their parents take over their upbringing and education. This notion constructs the host society as a corrupting influence that inevitably prevents West African children exposed to it from developing their ‘true’ cultural and religious identities. ‘Congolese ways are different from our ways,’ migrants told me time and again during research in Brazzaville, but West Africans living there often saw Congolese values as not only different from but actually inimical to their own. From the perspective of their interpretations of Islam, they equated Congolese ways with such illicit activities as alcohol consumption, fornication and theft. ‘If you let your child stay in Congo and grow up there’, Kadi told me in Togotala, ‘he’ll become Congolese. If he doesn’t become a beer drinker, he’ll become a drug user. That’s why, when our children are a few years old, we send them to grow up here.’

Most migrant parents in Brazzaville employed similar language to describe the importance of sending children back, saying they wanted to provide for their children the same kind of upbringing they themselves had had. This especially means learning the same language (or languages) and religious traditions, and getting to know the same place (the homeland or faso, Bamanan for ‘father house’) and its people. Adults who as children had been fostered to their parents’ home villages emphasized that exposure to their extended families helped them discover where they came from and who ‘their people’ were. In the words of a migrant father in Brazzaville,

We always like our children to grow up in the place where their parents come from, to know their culture. … The place where their grandfathers come from, we call that faso. So that, for us, is very important. Any child we [raise] like that is going to come [abroad again, and] living their life they have no problems. We teach them what is right and what is good, what is not good.

The children’s interests do not solely guide parental concerns about their children’s enculturation: the parents’ own welfare will depend on the degree to which their children acquire the ‘right’ values, especially those pertaining to filial piety. West Africans suspect that a child who ‘becomes Congolese’ will be less likely to honour religious and cultural obligations to provide for parents in their old age. ‘Here, if you bring up a child, it won’t benefit you, that’s why we send them off’, stated a Malian mother in Brazzaville, whose children were both living with grandparents in Mali. The contradistinction migrants make between rebellious and dutiful children is clear in the reply of a father named Draman after I asked how parents could spot the signs of a child turning out ‘wrong’:

OK, we see him smoking cannabis. Or taking pills. Or getting drunk. Or there are children who steal. Right away we send them off, because the upbringings aren’t the same. Congolese children, they depend on their parents for food, right up to 25 years of age. … But for us, already from the age of ten or so, a child says ‘I must feed my father and my mother.’ There’s the big difference. West African children work for their parents. Children here work for themselves. Not only do they work for themselves, but if they don’t have anything it’s their parents who must provide for them, even at the age of 40.

Many migrants see maximal-mode childrearing as the best way to ensure that their children grow up to be part of their parents’ communities (defined in terms of kinship, ethnicity and religion), thereby ensuring some measure of social security for themselves later in life. Sending children back to spend their childhoods in the faso essentially prevents them from becoming ‘second-generation immigrants’ in the place of destination and from assimilating even partially with the host society. The maximal-mode approach of course entails long periods of separation for these parents from their children. The case of Kadi, who for 19 years did not see her biological parents, is an extreme example; today, given cheaper transport and telecommunications, migrant parents are more likely to visit home and stay in regular contact with their children there. But the distance remains great and parents who adopt maximal-mode approaches must give up the kind of face-to-face relationship with their children they could otherwise expect.

Maximal-mode transnational parenting has been the normative approach among West Africans in Brazzaville for generations. Writing of Soninke migrants in the Congo Basin in the 1950s, Manchuelle (1987: 452) asserts that elders ‘had long reacted against what they saw as the “immoral” atmosphere of the Congolese cities by sending their children back “home” to their villages to be educated in the traditional ways, and in the proper Islamic upbringing’. Such responses are widespread among West African migrants today, and not only in Central Africa.³

Parents of maximal mode children come from a variety of socio-economic backgrounds. While the father usually covers the financial expenses entailed in this practice (such as the child’s airfare), the mother may make significant contributions as well, and the kin looking after the child in the home country may be maternal or paternal. Although the social and economic resources maximal-mode parenting requires are not available to all migrant parents, my research suggests that they are within the reach of a majority of West African parents in Brazzaville.

One could argue that maximal-mode parents deploy alternative constructions of parenthood, redefining their parental roles from caregivers to economic providers, as others (Hondagneu-Sotelo and Avila 1997; Silvey 2006) have observed among transnational mothers elsewhere. The reality is more complicated, given the widespread nature of child fostering in West Africa, where two-thirds of households in some areas are home to foster children (Findley and Diallo 1988; see also Bledsoe 1990; Isiugo-Abanihe 1985). Where many children are raised for long periods by their uncles, aunts, grandparents or others, home-fostered children of transnational migrants might not experience the same sense of dislocation, lack of intimacy and loss attributed to children separated from their parents in other transnational contexts. The latter include Chinese ‘parachute kids’ and ‘satellite kids’ sent abroad for their educations (Waters 2005; Zhou 1998), or the children of Filipina migrant workers (Parreñas 2005). This is especially true for children like Kadi who are fostered out at a very young age. I asked Kadi what it was like to be apart from her parents for so long. ‘It wasn’t hard, because I was with my family’, she told me. ‘Plus I wasn’t alone; many kids came from Brazzaville. Whoever was born over there came back [to Togotala] to grow up.’ As in transnational families from the Caribbean (Chamberlain 1999), strong extended families have blunted the impact of parental separation for these West African children; this fact distinguishes these populations from many others where family nucleation is or has become the norm.

There is nothing unusual about sending children ‘home’ across national boundaries. Migrant parents in many settings send their children back to shield them from the cultural influences of the host society. In the USA, for example, this practice has been observed among migrants from Belize (Matthei and Smith 1998), Yemen and Korea (Orellana et al. 2001), Guatemala (Menjívar 2002), and Mexico (Viruell-Fuentes 2006). For most of these parents, however, sending a child back is a form of ‘transnational disciplining’ (Orellana et al. 2001: 588). This is a last resort in response to certain negative outcomes, such as the child’s involvement in criminal activity or social rebellion; many migrant parents invoke the possibility of transnational disciplining to deter misbehaviour by their children. What seems distinctive about transnational parenting among West Africans in Brazzaville is the frequency with which these parents resort to sending children home at an early age, before the host society has a chance to influence the children. While some West Africans (such as Draman quoted above) use home fostering as a corrective measure, most use it preemptively.

Intermediate mode: some exposure to the home community

An alternative approach to transnational childrearing, one I call the ‘intermediate mode’, is for parents to keep children based in the place of destination but send them on occasional visits to the faso. The duration and frequency of these visits vary widely. One approach entails four- to twelve-week visits made by school-age children, timed to coincide with Congolese school vacations. Intermediate-mode parenting has the obvious advantage of enabling mothers and fathers to have more control over their children’s upbringing and to spend more time with them. It also has the disadvantage, however, of stigmatizing the children in the eyes of many West Africans, particularly non-migrants in the parents’ home communities who may feel that individuals raised abroad are culturally impure. I often heard the word tabushi used to designate such individuals. The origins of this word, which exists in a number of languages including Bamanan, Dyula, Soninke and Songhai, are uncertain, its orthography is inconsistent, and its exact meaning is subject to revealing disagreement. One Soninke dictionary defines tabuusi as children of mixed ethnicity (Dantioko 2003), while a Bamanan lexicon defines tabusi as ‘Mande people abroad/ children of Mande immigrants’ (Diaby 2003). The definition of tabushi I adopt comes from West African informants of many different ethnicities who use the term to refer to West African children born and especially raised abroad. This definition carries the pejorative implication that the tabushi is culturally inauthentic and lacks the mores and habits that anyone ‘born and bred’ in their ancestral West African homeland acquires early in life. The term is most frequently applied to children of West African parents born in Côte d’Ivoire, which for several generations has been home to more migrants from the West African Sahel than any other country (see Gary-Tounkara 2008).

Despite some of the above descriptions, tabushi status does not automatically befall children born to West African migrants abroad. Children sent home at a sufficiently early age are seen as no different from their West Africa-born peers; through their parents’ sacrifice, they benefit from the same socialization as other children in their faso. Maximal-mode parenting, therefore, prevents children from acquiring tabushi status. Intermediate-mode parenting, however, does not: children who go to West Africa for relatively short, infrequent visits are only likely to call unwanted attention to their alien upbringing through their unfamiliarity with local landscapes, their ignorance of local customs, and their imperfect command of the local language.

It takes considerable effort and skill for a tabushi in such circumstances to ‘perform’ a purely West African identity successfully. Madu, the son of a wealthy Malian trader in Brazzaville, went to Mali for the first time at the age of 13 and remained there for eight years. (His personal trajectory therefore falls somewhere between maximal and intermediate modes.) ‘I went to learn things, to know this is my faso’, he told me. ‘Now I can know my family, and even when I go away, I know that I come from Mali ka koro’ [literally ‘of old’, or originally]. When he first arrived, he could understand Bamanan (the dominant language in southern and central Mali, as well as in Brazzaville’s Malian population) but could not speak it; within one year, he says, he had fully mastered the language. ‘When I was in Mali, if I didn’t tell you that I was a tabushi, you couldn’t tell. Not from the way I spoke, the way I talked to my parents, even true Malians wouldn’t be able to tell.’ Madu acquired a Malian national ID card and attended school for seven of the eight years he spent in Mali. He returned to Brazzaville to find work at the age of 21.

Informants who adopted intermediate-mode parenting voiced the opinion that the most important determinant of a child’s upbringing is direct parental involvement, not the environment outside the home. ‘For me a child’s education depends on the parents. The child will adopt my own behaviour’, said Baru, a 39-year-old Malian father of four who owned a thriving auto parts business in Brazzaville. He had sent each of his children (apart from the youngest who was still an infant) to visit Mali, and saw these visits as a necessary complement to the upbringing he gave them in the Congo. He was having a house built in his hometown to serve as a kind of country home he and his family could use after relocating to Bamako. He and other parents in a like situation remained wary of allowing their offspring to ‘become Congolese’, but they sought to reduce that risk by setting a proper example in the home rather than by fostering their children to their faso for a lengthy period. The high cost of home visits makes intermediate-mode childrearing off-limits for all but a tiny fraction of parents: only three of the 98 parents in my interview sample used this approach with all their children.

Minimal mode: minimal or no exposure to the home community

Just as parents who adopted intermediate-mode parenting tended to be wealthier than parents who sent their children to be raised in West Africa, those who had not sent their children to their faso at all were often near the bottom of the economic scale. There were 14 of these parents in my sample.⁴ Minimal-mode fathers were generally commercial employees, labourers or itinerant traders who lacked either the financial means or the social relationships necessary to send their children back. A few were relatively well established and could have afforded to send their children to West Africa, but did not have relatives there in a position to look after them. Nearly all parents in this ‘minimal mode’ category expressed the opinion that it is best to send children home, but said they simply were unable to act on that desire; many held out hope that they would be able to send them home eventually. Maryam, a 39-year-old Malian mother, was one of these. She and her husband (a commercial employee) lived with their eight-year-old son in a modest Brazzaville apartment. They could not afford to send him to West Africa, and he was already unable to speak either Bamanan or his parents’ Soninke language, yet Maryam couched their parenting approach in terms of preference rather than necessity:

Now people understand that there’s no reason to send them too young. [It’s best] to have your child with you, so you develop affection for each other. But if you send your child away young, when they grow up you’ve spent all those years apart, and you can’t relate to each other. People have come to understand that it’s not good to send them away too young. So we can leave them even till age ten, once they start school they can take care of themselves. Now we understand that this is better than sending them off very young. … If you send a young child to somebody, you’re giving them a burden.

In fact Maryam’s son was by then in his third year of primary school and was able to look after himself in many ways, but Maryam chose to represent him as still too young and dependent. Like many of the poorest migrant parents, Maryam and her husband had to make do while waiting for their fortunes to change. Though their son attended a Congolese public school, they also sent him in the afternoons to the neighbourhood madrasa, a private Islamic school where young people (mostly the children of West African migrants) learned the Qur’an and the fundamentals of their Muslim faith. Moreover, he spent school breaks with a prosperous uncle in Pointe-Noire whose children made regular visits to Mali.

Brazzaville’s madrasa plays a crucial role in the education and socialization of children in the West African community. It provides an Islamic-oriented learning environment similar to that of madrasas in West Africa, where they have become a widespread alternative (and complement) to secular schooling in recent decades (Brenner 2001); madrasa proponents represent the institution as a bulwark against alien cultural influences. Yet there is a common sentiment among West African migrant parents that even the best madrasa abroad is no substitute for a traditional upbringing in the faso. Salim, a madrasa instructor in Brazzaville, told me that most of his students came from families unable to send them home to West Africa. His own three children (the oldest of whom was ten years old) were in the same position, and though he did everything he could to keep them on the ‘straight path’, he felt that their fate was out of his hands as long as they remained in Brazzaville. They spoke only Lingala (the dominant Congolese language in their neighbourhood), and spent most of their time with Congolese playmates. Like most West African fathers in Brazzaville, Salim did not have enough time to monitor his children’s activities adequately: in addition to teaching in the madrasa he worked in a relative’s shop and taught in a local mosque in the evenings. His wife was busy with domestic duties. Salim worried about his children’s future but saw little alternative to keeping them in Brazzaville, where he felt he could at least provide for them better than he could back home.

During my Brazzaville fieldwork I heard about and occasionally met children of West African migrants who had grown up to ‘become Congolese’ and who were seen as marginal members of the West African community. They represented ‘failed’ outcomes from the perspective of their migrant parents and their community. Some of these tabushis were only nominally Muslim, others had abandoned the faith entirely; many had married Congolese, while others had married similarly marginal second- or third-generation West Africans. Most importantly, they had not learnt to speak Bamanan, Soninke or any other language of the West African population. For them, mainly because of the language barrier, becoming fully-fledged members of their parents’ community was problematic, but they also felt out of place among Congolese. In a society where the problems of everyday life are addressed by mobilizing bonds of kinship, ethnicity and shared regional origin (Milandou 1997), these individuals’ lack of ancestral ties to other Congolese left them without the kind of connections necessary to compete. Intermarriage with ‘real’ Congolese was the only way for them to build up this capital and have children capable of social integration.

Not all minimal-mode children in Brazzaville face such alienation, however: a few were able to integrate successfully into Brazzaville’s West African community despite their lack of homeland exposure. These were the children, or sometimes the grandchildren, of wealthy West African entrepreneurs. One such parent and grand-parent was Makan, a patriarch in his eighties who had spent six decades in the Congo after migrating from Mali as a young man. He was a jatigi, a combination of host, landlord and intermediary who used his connections in Brazzaville to help newly arrived West African migrants as well as transnational traders – particularly diamond merchants – who were passing through. Makan told me he had resorted to minimal-mode parenting because he did not have a relative in Mali who could take his children in. His children had only Congolese citizenship. Nonetheless, every sign indicated that his children, 18 in all, had remained part of the West African community: they were practising Muslims and had married Malians (mostly from his home village). ‘My children haven’t gone to Mali, but they came out fine, their father brought them up well,’ one of Makan’s two wives said. ‘Look at them, they haven’t been to Mali but they speak Bamanan, Soninke, French. … God willing, if my children go back to their homeland, they won’t be ashamed because they were brought up right.’ Many of Makan’s children went to Mali after becoming adults, while some were residing elsewhere in the transnational social field, including France and the USA. Even those of his adult grandchildren who had never left the Congo seemed to be fully a part of Brazzaville’s West African community, and spoke Soninke and Bamanan in addition to Lingala and French. ‘I’ve never been to Mali, but the way they do things, even Malian languages, I’ve been able to learn them little by little. It’s as though I was born and raised there,’ said Fanta, Makan’s 21-year-old granddaughter. She was engaged to marry a fellow Congo-born Malian whose father had come from the same region of Mali as Fanta’s grandfather.

Given the ease with which West African children in Brazzaville can grow apart from their parents’ religious and cultural identities, one may wonder how Makan managed to raise children to identify so strongly with their ancestral homeland without having lived there. Certainly, Makan and his two wives were diligent in their children’s upbringing, took pains to instil them with the customs of their home culture and ensured that they attended madrasa as well as public school. But the same could be observed of many other migrants whose children were nowhere near as integrated into the West African community. To answer this question we must examine the factors shaping ‘successful’ social reproduction in this migrant population.

Determinants of ‘success’ in transnational childrearing

West Africans tend to represent ‘success’ in transnational childrearing as a set of traits, habits, values and skills that the children should acquire while growing up. These include respect for elders, proper Islamic comportment, avoidance of illicit behaviour (such as drinking alcohol), and especially mastery of their parents’ West African language. We have seen that economic and social factors play a role in the adoption of particular childrearing practices. Those migrant parents who employ intermediate-mode approaches are likely to be better off financially than many others: the costs of this mode are high (a round-trip ticket from Brazzaville to Bamako currently costs about US$ 800), and only the wealthy can afford such visits for their children on a regular basis. These fathers often own multiple shops and other businesses. Parents who adopt the maximal mode, which requires only a single one-way ticket, are more likely to be small-scale merchants and commercial employees. Those who adopt the minimal mode are usually the poorest members of the community – petty traders, labourers and the marginally employed. Finally, maximalmode childrearing requires that parents maintain considerable social capital in their community of origin, particularly in the form of kin capable of raising their children there over the long term.

It may be, however, that the determinants of ‘successful’ childrearing outcomes for these migrant parents lie less in the specific mode of transnational parenting adopted than in the parents’ economic and social status within the migrant population. Wealthy transnational entrepreneurs with employees managing their commercial operations can afford to spend more time at home; their wives can also spend more time with their children, since they have domestic servants (either paid Congolese workers or young female relatives brought from West Africa) in their households. They can therefore play a greater role in their children’s upbringing. By contrast, labourers, ordinary shopkeepers and commercial employees like Salim spend most of their time outside the home. Their wives are busy with their cooking rotations and often leave their children to their own devices.

A father’s social position in the migrant population is another decisive factor in his children’s enculturation. Established businessmen like Makan, the jatigi described above whose children were entirely raised in Brazzaville, form the epicentre of the city’s West African community. Because of their stability and knowledge of the host society, they constantly receive new West African visitors and lodge West African guests and tenants, including circular migrants and aventuriers. As vital nodes in migrant networks, they draw in other community members. Their spacious homes are where West African merchants interact with religious and community leaders. They are islands of permanence in a sea of flux. As Agier (1983: 159) writes in his study of migrant traders in Togo, ‘[the] mobility of some is made possible by the sedentariness of others.’ Because of these men’s importance, the rest of the community comes to them. Their children grow up amid constant activity and intercourse in West African languages, which they easily master; they are exposed to constant discussion of their parents’ homeland, thus enabling them to accumulate ‘vicarious memory’ (Chamberlain and Leydesdorff 2004: 235) of a place they have never seen. The fathers pay for their children’s private religious instruction at home. The daughters become highly sought after as brides by other West Africans, guaranteeing their secure membership in the West African community for the rest of their lives.

The picture looks quite different for poorer migrants with less to offer their fellow West Africans. Their homes, cramped one- or two-room apartments literally on the margins of the community, receive few visitors. Their children spend their time with Congolese peers, seldom hear their parents’ language spoken by anyone else, and become more comfortable in Lingala than in any West African language. Without religious tutoring and outside affirmation of their parents’ memories of home, they come to view their ancestral homeland and its traditions as alien. Other migrants are likely to belittle them as tabushis, half-breeds, and are unlikely to see them as desirable marriage partners. Their existence is a liminal one, and their future within the West African population is uncertain. While transnationalism is hardly the purview of only a wealthy segment of this population, as Portes (2003) has argued for Latin American migrants in the USA, it is clear that the social and economic resources that people have available condition the types of transnationalism in which they engage, particularly where transnational childrearing is concerned.

Conclusions

Examining the ‘full house’ of domestic variation among migrants enables us to understand the multivocality of approaches to household organization and child-rearing. It also helps us understand the complex relationships among these different approaches. There is a crucial complementarity between short-term sojourners – usually single migrants with few assets – and settled heads of households with strong ties to the host community. Settlers provide sojourners with material support, knowledge and advice; newcomers and sojourners enable the settled to renew their links with their homeland and provide their children with indirect exposure to it. This complementarity also exists among migrant parents with differing modes of child-rearing and degrees of mobility. Recall Maryam, who could not send her son to Mali, but instead sent him to spend school breaks in another Congolese city with relatives who regularly visited Mali. For these minimal-mode parents, a kin relationship with an intermediate-mode family gave their son the possibility of obliquely acquiring some of the cultural and social assets and ‘vicarious memories’ he could not otherwise obtain in Brazzaville.

Studies of transnational social fields and the people who occupy them should thus be sensitive to the diversity of transnational family organization. There is no uniform development with each member of a transnational population moving in lockstep through the stages of an evolutionary process, progressing from sojourner to settler or from cultural cosmopolitan to assimilated migrant. Transnational lives do not somehow eclipse cultural assimilation or sedentary communities. Rather, given the ‘full house’ of transnational practices accessible to a wide variety of people, what we most need to understand is the dynamics connecting these different modes of living in a transnational social field.

As for the durability of transnational linkages, this research suggests that transnationalism can be passed on from one generation to the next – but only through considerable effort and sacrifice, by parents and children alike. Even for inhabitants of a transnational social field, place still matters a great deal in processes of identity formation. Transnational childrearing requires extremely flexible domestic arrangements and it remains to be seen how technologies facilitating parental communication with far-flung children and their guardians will affect the practice of maximal-mode childrearing. But the example of Makan’s children and grandchildren shows how, given favourable circumstances (notably high social and economic status in the migrant community), even second- and third-generation migrants can be socialized to consider themselves as belonging primarily to their ancestors’ country and/or culture of origin, despite occasional jibes from people born and raised there. With the right combination of economic resources and social networks, it is possible to raise ‘good Malians’ outside of Mali and ‘good Senegalese’ outside Senegal. Such examples may always remain exceptions to the rule among migrants and their descendants, but they show how distinct transnational communities can endure over the long term despite being rather loosely tethered to their geographic point of reference. Through their combination of childrearing practices, West African migrants in Congo may have found a way to alleviate one of the greatest challenges of transnationalism – ensuring social reproduction and the continuity of an ‘Othered’ identity over the long term.

Finally, this case study suggests that research on intra-African connections can contribute to broader discussions of transnationalism. In the absence of strictly enforced immigration laws like those common in wealthy Northern states, migrants who move between and among countries of the global South have a significant capacity to establish and preserve transnational connections over the long term. While rural-to-urban, seasonal and circular forms of mobility have dominated the scholarly literature on South–South migration, it seems clear that South–South transnational migration deserves an equal measure of attention.