Lesedi eventually had her own experience of being sent to stay with other relatives, like Tumi had, and like almost everyone else I knew in Botswana. Lesedi’s mother died while she was a teenager, but that did not affect her living arrangements as such; she remained with her grandmother and Tumi’s mother, who continued to look after her, until she finished her public schooling at Form Five. Having failed her exams, she had limited opportunities at home – until she was called by relatives living in the south, in one of the large villages close to the capital.

‘They were far family,’ she explained, ‘on my grandmother’s side – he was my grandmother’s brother’s son.’Footnote ¹ The man’s wife had taken a teaching post in a distant peri-urban village, and they told Lesedi’s grandmother that they wanted to take her so they could help her repeat her Form Five. On the face of it, it looked very much like the sort of help Tumi had been offered years previously, which gave Lesedi hope. ‘But it didn’t work like that,’ she explained, with a look of resentment. ‘When I came to stay there they wanted me to be their maid. They didn’t even take me to the school they promised. They wanted somebody to help them, so they just lied that they’ll take me to school.’

She stayed with them for a year and a half. ‘It was bad … I just had to, to stay there. She couldn’t even give me two pula,’ she added, referring to the wife. At one point, her hosts had even begun passing comments about the cost of feeding her, suggesting that her grandmother should be contributing something for her care. The injustice and disappointment in being expected to contribute to a household when the contributions promised towards her schooling were withheld – along with the opportunities for self-making schooling presented – were still raw in Lesedi’s telling. Lesedi had felt unable to say anything to her host family about the issue or its possible resolution. I asked if she had told her grandmother, through whom the arrangement had originally been made. ‘I didn’t want to stress her,’ she answered. ‘I only told her after I left. Because you know how people are – if you tell, tomorrow it’s like you are trying to destroy people’s families or something. So I just stayed. Also it was hard at home. My brother had just started working, but others were staying with him, plus he was looking after everyone.’

Lesedi had done quite well for herself since then. She had eventually put herself through Form Five exam rewrites and had passed, and she was attending university, which meant that she was receiving a substantial stipend from the government – enough to comfortably cover her expenses, from rent and food to clothing and toiletries for herself and her daughter. The father of her child had a good job and also supported them both financially; he had bought her a car and helped build a house for her in her home village. She was comfortably settled in the capital. Partly as a result of this visible success, and partly because she stayed in the city, close to its amenities and opportunities, she had moved into the role her mother’s older sister, ‘far relatives’, and brother had all played before her: two younger cousins had been sent to stay with her at the time we spoke.

A younger male cousin,Footnote ² who had come to the city to attend agricultural college, was the first to ask to stay at Lesedi’s. She agreed to accommodate him on the condition that he assist with the care of her school-aged daughter. He often cooked, cleaned the house, and played with or babysat the little girl. However, as his comings and goings became more frequent and unpredictable, and as it became clear that he was at risk of being kicked out of school, Lesedi sent for a younger female cousin to come and replace him. The girl had failed her Form Three exams, and Lesedi offered to help her repeat her courses in exchange for help around the house. She prepared meals, cleaned the house and yard, babysat the little girl, and did anything else she was asked. She seldom left, except to attend classes or to make the long, occasional trip back to their home village. Lesedi described these arrangements with some frustration, however, noting the unreliability of both cousins in doing housework and despairing of either making anything of themselves. The parallel between both situations and Lesedi’s own, at a similar age, went unremarked.

While her younger relatives looked after the child and the house, Lesedi had taken on primary responsibility for Tumi’s care after her return from hospital. It was proving onerous. On a recent trip back to their home village for a wedding, at one of the large family meetings that typify such events, Lesedi told me that she had made an explicit move to disengage from any further responsibility for relatives coming and going to the capital: ‘The city is eating us,’ she told them. ‘I don’t want to encourage anyone else to come there. If they do, they should make their own arrangements.’ To a mutual friend, she vowed: ‘From now on, I just want to think about me and my daughter.’Footnote ³ But at the same time, she would continue to need help caring for her daughter, and for Tumi’s infant child, especially while Tumi remained ill. Lesedi may have hoped to escape the cycle of circulating kin, but it seemed unlikely, a matter of needs and obligations beyond her control.

Lesedi’s experience describes many of the ways in which children and young people circulate, are called, sent, and taken in in Botswana – and it charts the trajectory of growing from a circulating child to an adult attempting to manage such circulations, and the perpetuity that characterises those cycles. As a child, Lesedi’s unmarried mother left her ko gae – at home – to be cared for by her maternal grandmother and her mother’s sister. Having a child meant that there was pressure on Lesedi’s mother to work; and work meant being away from the village, in this case in a transnationally mobile manner. After her mother’s death, like many orphaned children, Lesedi stayed where she had been: with her grandmother. As a teenager at a loose end, she was taken to care for the children of distant relatives in conditions that she described as unfair and uncaring, oriented towards labour. And once she had become a mother and had acquired a house herself, Lesedi hosted younger kin going to school in the city, eventually sending for a young cousin from home to assist in the care of her child in exchange for better schooling opportunities – much as her mother’s sister had done for Tumi and Lesedi’s malome had done for her. Perhaps the only sort of circulation she hadn’t (yet) undertaken was of sending a child of her own to relatives for company and help, or for accommodation during schooling or work.

Lesedi’s story is not unusual. Many of the Batswana I knew, girls and boys, men and women alike, had had similar experiences: they were raised predominantly by grandparents, had lived with other kin while working and/or attending school – often in exchange for providing childcare or other forms of help – and, as adults, had taken in the children of relatives for various periods of time. And these practices are not new: Mmapula, the elderly Legae matriarch, had been raised by her own grandmother in the 1950s and had in turn raised her sister’s child (as well as housing several other members of her and her husband’s extended kin for different periods of time). Lesedi’s experience of fraught, unspoken conflict and bad feeling while staying with her ‘far’ relatives – compared with the relative ease of her relationship with her grandmother, or Tumi’s ease with their mothers’ sister – was also typical of others I knew. Hers were, in other words, widely shared experiences of child circulation and of kinship in Botswana.

These diverse situations involve many of the kin-making processes described so far. All cases involve co-residence; free, frequent movement between places of the gae; and care work undertaken in each of those places. They anticipate the contribution of certain resources and labour by the hosting families – especially food, clothing, toiletries, and transport, as well as discretionary funds; but also cooking, guidance, and discipline, or help with schoolwork. And they anticipate the care contributions of circulated children as well – in raising younger children, cleaning and cooking, and mobilising additional resources. There are, however, noteworthy distinctions among the sorts of child circulation described above, which I suggest work to define gradations of relatedness, from ‘close’ to ‘far’ family. Such distinctions are already apparent in the reasons behind children being circulated, which fall into two rough, sometimes overlapping categories: the absence of birth parents (commonly because of work, but also because of illness or death); and the absence of children, specifically children old enough to contribute to the household. And these distinctions vary with the places to which children are circulated, from ko gae (at home) to away. Thus, in the absence of birth parents, ideally children are circulated ko gae, if possible with the absent parent’s mother or older sisters, often in semi-permanent arrangements; whereas in the absence of children who can assist in the work of the household, they tend to be drawn from away, farther from the host’s home both geographically and genealogically, and often for shorter periods of time.

But these distinctions are perhaps most evident in the sorts of conflicts that arise, in the ways in which they are – and aren’t – addressed, and in the people called on to address them. We have seen in the preceding chapters the different ways in which dikgang emerge and are addressed among families living together at home. These same conflicts, and the means of addressing them, are roughly common to situations that arise when children are circulated ko gae. Tumi’s mother’s sister would have addressed any conflicts with her much as her own mother would have; if the issues had been serious enough to involve calling in others to intervene, the same bomalome would have been called in the same ways, by virtue of comparable relationships. In assessing the problems at hand, the quality of relationships among many of the same people would be called up for reflection as if Lesedi had been at the heart of the matter.

When children are circulated away from the gae, similar problems emerge, running the full gamut of dikgang we have explored so far, with the potential to embroil children, husbands and wives, siblings, multiple households, and an extensive range of kin (Alber Reference Alber2018: 144). However, these dikgang are seldom engaged directly, and seldom addressed within the host yard.Footnote ⁴ Instead, they are either carefully avoided (see ibid.: 140), indefinitely postponed, or expressed through – and referred for resolution back to – the family from which the young person was sent in the first place. Lesedi would not have considered raising her concerns directly with her host family; only her grandmother, who sent her, was an appropriate audience, and then not until considerably after the fact. Likewise, her hosts would not have confronted her with their concerns; instead, they would have presented them to her grandmother for resolution. Within the hosting yard, conflicts are actively muffled: fostering adults may pass comment, but only indirectly, and circulated children are expected to hold their tongues respectfully. A grudge-like atmosphere emerges. Expectations and interpretations of the scenario diverge, but they are not voiced, discussed, or reflected on collectively.

The result of this scenario is frequently an impasse. Having not been witness to the causes of conflict, and having no means of hearing the story from both sides without casting aspersions on people who have offered a favour, the family ko gae does not weigh or attempt to establish the comparative truth of each tale, nor reflect on what they may mean for the relationships at hand, nor pronounce judgement. They are, essentially, unable to mediate. Most often they will counsel their child simply to be respectful and do as she or he is told, especially if there are no better immediate solutions available; and if the issue persists and seems impossible to resolve, they will simply summon or allow the child to come home, without further discussion, letting silence and movement resolve the kgang (see Alber Reference Alber2018 for similar strategies, if in quite different contexts).

Lesedi’s comment regarding the risk of telling her grandmother about her poor treatment at the hands of her ‘far relatives’ – for fear she might be accused of ‘destroying someone’s family’ – is telling in understanding this dynamic. The family she risks destroying by speaking ill of their conduct is not her extended family as a whole, nor her natal family, but the family that has taken her in. Like any kin who live together, she is a potential threat; and speech, especially the articulation of discord (or puo), is one of the most potent means of actualising that threat. But, in this case, the threat she poses is greater because it risks drawing kin into conflict who would otherwise carefully avoid it. As such, the threat is best contained by exclusion and distance, silence and grudges, and above all by forgoing active engagement in conflict. By the same logic, the departure of a circulated child will be accepted without remonstration or accusation (see also Coe Reference Coe, Alber, Martin and Notermans2013: 170).

To the extent that the ‘far’ host family in a scenario like Lesedi’s does not engage in inevitable dikgang the way her family ko gae might, they are distanced from her; they do not, and cannot, replace her ‘near’ family (cf. Coe Reference Coe, Alber, Martin and Notermans2013: 157–8). This distancing reflects their distance from other members of her natal family; in this sense, it reproduces the ‘farness’ of their relatedness. By referring the conflict at hand, and its resolution, back to the natal kin, the latter’s unique capacity to engage and resolve conflict is emphasised – reproducing the nearness of their relatedness to the sent-out child.

When I asked her to map out her family however she saw fit, including and excluding whomever she liked, Lesedi did not include the family that hosted her; nor did she include them among the broad range of people who had raised her. Staying with them, caring for their children, and ultimately coming into unresolved conflict with them did not bring her closer to them; it clarified their distance and reasserted their position as ‘far relatives’. In a similar exercise, Tumi listed her mother’s sister who took her to be raised in the city as kin, but did not give her any particular priority – certainly not above her own grandmother and mother. She acknowledged the help she had received from her mmamogolo, but the time spent with her did not change their relationship so much as reaffirm it. Child circulation among Tswana families thus seems not so much to tighten bonds of kinship, nor even to transform those bonds, but to assert appropriate degrees of closeness and distance between kin and to reproduce these differentiations across generations. Circulated children come to know their relatives and apposite ways of relating to them that ensure help in times of need, while containing the danger – suggested in the proverb above – that misreading their likeliness to help might produce. And in the process of doing so, acquiring and demonstrating good judgement in managing dikgang, their circulation contributes to their projects of self-making as well.

But what about child circulation – undertaken either informally or formally – with non-kin? Does it serve to create a sort of replacement or substitute kinship where kin circulation does not? What practices of care, conflict, and resolution does it involve? And how does it compare to kin-circulating practice? In Chapters 11 and 12, I consider these questions with reference to the case of a young man who brought himself to stay with the Legaes during my fieldwork, and the case of Botswana’s first – and only – formal placement of children with a trained foster parent.

Arriving home one twilit evening, trading loud greetings over my shoulder with others in the yard, I walked into the sitting room and had a shock. An unfamiliar young man sat there, alone, glowering up at me from the edge of the couch. I greeted him; he looked away without response. I passed through into the kitchen to put the kettle on, and when I returned I found him unmoved: leaning forward, his elbows on his knees, he clutched a book and stared into a dark corner of the room.

I went out into the lelwapa and asked Modiri surreptitiously who the young man was. He shrugged, took a drink of his tea, and said the boy was waiting for Kagiso.

Later that night, as we sat scattered around the lelwapa after dinner, I noticed Kagiso’s voice in the house. The lights had been turned on in the sitting room, giving it a pale blue glow through the window. The door was closed. I asked Kelebogile what was happening, and she explained that the young man attended church with them and had come to ask help from Kagiso – their sometime preacher – because his family was bewitching him. They were enclosed together in the sitting room praying intently, and they stayed that way until long after I had gone to bed.

The next morning, I was surprised to see the same young man, now in school uniform, drinking his morning tea by the fire.

I didn’t learn the young man’s name for almost two weeks. He and I circled around each other warily, each of us equally confused by the presence of the other. We seldom spoke, unsure how to take one another or what to say. I would sometimes go for days without seeing him, and he seemed to come and go freely, but a great stack of his school papers and books had appeared on the bookshelf in Kagiso’s room, where he slept. I heard from the younger children in the yard that Kagiso had gone to visit the boy’s family to tell them where he was; he had visited the social worker and the school to make similar reports and discuss alternative arrangements – of which, apparently, there were none. There the matter rested.

His name was Bonolo. He had been staying with us for eight months before I asked to sit down with him and hear his whole story. During that time, he had integrated more or less seamlessly with the Legaes. He took on chores of his own almost immediately, including starting the fire in the morning, sweeping, and occasionally doing dishes; he often also went out to the cattle post at weekends to help with the heavy work of finding, herding, and feeding the dispersed herd (see also Archambault and de Laat Reference Archambault, de Laat, Ensor and Gozdziak2010: 196 on chores and integration). He spent many of his weeknights at Kagiso’s shop, helping out and passing the time with the Legae children who worked there. His clothes were mostly hand-me-downs from both Kagiso and Tuelo, and he was served and ate at home with everyone else. He was well liked by the children of the yard and became close to them, spending much of his time at home in their company.

But there were subtle limits to his integration, too. Unlike the other young people of the house, for example, I didn’t feel I could send him for things, or ask for his help. Other adults in the house seldom sent him for anything, although he would often volunteer to go with one of the other boys when they were sent. The chores he had taken on – at home, at the cattle post, at the shop – were all voluntary; I never saw him being asked to undertake any specific tasks, nor scolded for neglecting any, although the men might invite him along on errands. His relationships with the adults in the yard seemed to remain aloof. While he would sometimes seek help with homework or engage in lively debates around various Christian precepts, he did not seek the adults out for advice or attach himself to any of them particularly. And they, in turn, remained aloof from him and avoided inquiring into his background or life. Kagiso – who was running three small businesses, working as a full-time driver, and conducting a clandestine courtship – was seldom home or available. While Bonolo clearly considered him a sort of mentor, their connection did not seem to run much deeper than that.

As companionable as they found him, none of the family members – not even the children – referred to Bonolo using kin terms either. Occasionally family members teasingly referred to Bonolo as ngwana wa ga Kagiso – Kagiso’s child – but these comments were used in humorous banter among the women, seldom made in front of Kagiso, and I never heard Bonolo called that to his face. Kagiso, moreover, was never called Rra go Bonolo (father of Bonolo), even in jest. The closest comparison was with the foundling calf that Modiri had brought back from the cattle post. The women’s commentary seemed to be more about playfully recognising an unexpected potential to provide care in both men, without asserting any real sort of obligation or relatedness. Indeed, the commentary was perhaps more about the fact that neither man had children of their own, while expressing the hope that one day they might.

While there had been occasional meetings between Kagiso and his parents, Mmapula and Dipuo, about Bonolo’s situation, these had never involved the rest of us; we heard of them as if by rumour, long after the fact. (As Bonolo pointed out to me, these meetings never involved him, either.) Barring Tshepo, who was Bonolo’s age-mate and former classmate, none of us had any real idea about Bonolo’s circumstances. We speculated and swapped overheard snippets freely among ourselves, but nobody asked.

Bonolo had a slow, intense, non-committal gaze when he was listening that almost inevitably dissolved into an affable, indiscriminate smile when he spoke – whether he spoke of happy things, or frustrations, or things to which he took exception. So I was uncertain how he actually felt about the notion of being interviewed, or about anything else for that matter. But he was insistent that people should know his story, and even that I should use his real name (which I have done). In fact, he insisted on writing his entire story out, in longhand, before we began talking.

The story, written in English in a confident, broad hand, traced his movements among all the places he had been raised. Having spent time initially in a small town in the south-east, he moved to the northern border of Botswana to begin schooling. He stayed there for several years before moving to Dithaba for a year, and then to the western desert, all by the time he was 12. After a couple of years there, staying with family and in a boarding school, he came back to Dithaba again, and had stayed there ever since. When he moved the first time, at perhaps seven years old, so had his mother – not north with him, but to the far north-eastern corner of the country. By that time, he reflected, she was working and didn’t seem to be ‘into alcohol or any habits unusual … and me also, I saw my photos … it seems like I was well provided [for]’. He took a curiously distant, sceptical perspective on himself. He surmised that they had had to separate and move ‘because of life’. His mother and two of his siblings still lived together in Francistown, and an older sister lived near them with her own children. Another of his sisters ‘lived outside’, as he described his own circumstances, but he could not say where, or with whom, or why.

He was in Form FourFootnote ¹ when his relatives began to abuse him, as he described it, making him ‘do too much household chores and shopping’. The rest of the children in the yard had been too small to help with work around the home, and he had been left with all of it. This complaint, a usual one for people his age, was what he said had finally brought him to our yard. His account made no mention of the witchcraft he had cited on his first appearance. ‘None of them came to hear why I runned [sic],’ he said of his natal family. He lavished praise on the Legaes as his hosts, noting that ‘my mother didn’t contribute any cent, and they didn’t demand nothing [sic]’. He added: ‘I wish the most high to drive me not to forget them … they are my saviours and trusted friends.’

To this narrative, Bonolo had added a family chart. On one page, he drew in his mother’s parents and their descendants, down to his sister’s children. Down the right-hand side of the page, from his mother’s father, he drew an additional, long line to a second grandmother, with a generic dichotomous split line below her, and nothing else. He focused on his mother’s family, telling me about her siblings and their children. As we talked, I realised that he had not been staying with any of them in Dithaba – indeed, none of them were in the village at all. I asked him to tell me more about the second grandmother he had sketched at right angles to his grandfather, and her family.

He explained that she and his grandfather had not been married, and so he had not sketched in that side of the family. He began to do so, with some hesitation. Slowly I realised that one of the women on this branch of the family tree was the one who had taken him to school in Ghanzi; and that one of her brothers was the malome who had followed up Bonolo’s ‘issue’ with us occasionally at home. He then explained that he had been living with this grandmother, two of her daughters, and their children in Dithaba for years – and it was in reaction against them that he had come to stay with us. He described the grandmother and her daughters as people who had raised him, although they were not batsadi (parents). He did not even list his mother among his batsadi. Only his mother’s married parents achieved that status.

Perhaps halfway through the interview, the phone rang, and Bonolo paused to answer it. Uncannily, it was his mother calling. I had heard that she called from time to time to check in on him but had not witnessed a call myself. He smiled and his voice became excited, like a child suddenly, asking about when he could go to visit. As the conversation progressed he became quieter, mumbling assent. Finally, he dropped the receiver with a sigh. She was promising to come to Dithaba to visit his extended family and then take him back with her for the school holidays. He was sceptical. ‘Nna ke blamea mama,’ he said – me, I blame my mother. When he had had his misunderstandings with his family in Dithaba, he explained, she had refused to come. ‘If she had come, they could have known the problem and resolved it,’ he asserted. ‘But she didn’t come at all. Even now she is not going to come.’

We spoke about the future, his plans to study engineering at the university and perhaps go to work for the army or the mines. ‘I want to stay far from my mum, both geographically and emotionally,’ he said, when I asked where he’d like to settle.

As we wrapped up the interview, I mentioned to Bonolo that the government was thinking about launching a formal foster parenting programme, whereby people would be recruited and trained to look after children who were having serious problems living with their families – much as he had. He was categorical in his response: ‘I don’t support that.’ Surprised, I asked him why not. He shook his head. ‘It’s not good to take children from their families; they should know they have responsibility for those children no matter what,’ he explained. I asked what he would tell children in his situation to do. He smiled. ‘I guess they could do what I did. But they should try by all means to solve their problems.’

Most of the Batswana friends to whom I mentioned Bonolo’s presence at home found the situation surprising, even dubious. As common as it is to circulate children among kin, for a child to stay with non-kin is somewhat beyond the pale, and many view it with suspicion. One friend, however, described a very similar situation in his own family. A close friend of his daughter’s had lost her parents in her early teens, and afterwards spent much of her time at their house. When they were making plans to move across the country to the capital a few years later, the girl’s older siblings approached them and asked whether they would consider taking her with them. The siblings explained that she had come to see them as parents, and were concerned she might take their loss doubly hard. And so my friend and his wife agreed. He laughed bitterly as he recalled how difficult it had been to have two teenage girls in the house at once – all the more so because while one was his daughter, the other wasn’t (he used the English phrase ‘foster daughter’ throughout our conversation). He sent her home to her family during the holidays, and he had recently put her into a boarding school nearer to them, retaining responsibility for her fees and upkeep.

There are three telling details in these stories. One is that – contrary to popular assertion – Batswana do indeed take in children from outside their kin networks. The second is that it is often the children themselves who orchestrate these arrangements (Archambault Reference Archambault2010; Leinaweaver Reference Leinaweaver2007a). And the third is that – although they undertake the responsibilities of a family member and are treated in many of the same ways – these children do not necessarily see themselves as, nor are they seen to be, members of their fostering families. They are ‘living outside’ both their natal families and their host families. Like child circulation among kin, then, ‘living outside’ does not extend or replace kinship so much as define and reproduce its limits.

Thus, although Bonolo slept, ate, worked, played, and otherwise stayed with the family in much the same ways as the other boys did, and although he was treated with affection and goodwill, he was not identified – nor did he identify, nor apparently want to identify – as family. No specific claims were made upon him: although he took on chores, he was not sent on errands, he was not scolded, and neither he nor his mother was expected to provide any specific contributions to his upkeep. Nor did he, in his turn, make any specific requests or claims beyond being allowed to stay. He was not taken along to funerals, weddings, or other events, nor was any great fuss made of his presence at home. There was little special effort to get to know him, develop intimacy, or otherwise draw him closer into the family. And Bonolo himself seemed satisfied with this arrangement, preferring to think and speak of his host family as ‘saviours’ and ‘trusted friends’ rather than as surrogate kin.

These limitations become clearest if we include dikgang among the defining characteristics of kinship. Bonolo’s experience with his chosen host family was marked by a surprising lack of conflict – especially considering the frequency of conflict we otherwise experienced at home. Mutual claims, obligations, knowledge, and intimacy were all avoided, I suspect, precisely in order to ensure that there would be few things to fall out about. The Legaes did not get involved with the ongoing disputes in Bonolo’s natal family whatsoever. Although Kagiso visited Bonolo’s family to report his presence with us and hear about the issue at hand (like a mediator might), and although he shared that information with his parents, once it was clear that Bonolo would be staying, Kagiso conscientiously avoided getting involved – or drawing in anyone else. He took the care of Bonolo as a temporary responsibility, but he did not take on the negotiation of the conflicts the situation involved. Neither Kagiso nor anyone else at home was asked to help expedite the issue by Bonolo or anyone in Bonolo’s family. Only Bonolo’s mother was in a position appropriate to engage dikgang with her family; no one sought to replace her.

While, on a superficial level, Bonolo’s experience suggests a kin-making dynamic, closer examination shows that it is anything but – precisely because those situations in which dikgang might emerge are explicitly forgone. Bonolo is not called or sent, nor reprimanded for his movements, and is left to stay as and where he sees fit; neither he nor his family is required to make contributions, nor are their contributions compared with those made by others in his host family; and his pre-existing relationships are neither enquired into, nor discussed, nor made unduly visible. Care, in this scenario, is delinked from dikgang; and, thus delinked, is insufficient to making kin. As Bonolo himself emphasised in parting, responsibility and problem solving are equally critical to kinship. In a context where kinship entails risk, where those who are closest to you are also most dangerous to you, it is perhaps unsurprising to find that a family otherwise willing to provide care would hold the expansion of their kin networks in such careful check.

Figure 8 Tinkering: brothers, mechanics.

Child circulation among Batswana, then, has an unexpected effect: to produce and reproduce nearness and distance in relatedness, whether among kin or between kin and non-kin. Circulation does not extend nor supplement kinship; rather, it defines its terms and limits. And, as a practice, it creates this distinction primarily in terms of differential responses to dikgang. How, then, might government-driven initiatives in formal foster care – where children are removed from environments of perceived abuse or danger and placed with non-kin foster parents trained for the purpose – fare in the Tswana context?

‘They understand informal fostering – that is the practice we are all doing. It’s foreign when we talk about making it legal. That’s what is putting us in trouble. But if there are no relatives, we need law.’

Tumelo and I sat on either side of her wide desk in pools of shadow left by the daylight filtering in through her office windows. It was an unusual moment of quiet. I had visited her previously at the simple concrete block adjoining Water Affairs that served as the Social and Community Development office, hidden from the highway by a string of bars. But on past occasions she had been beset by long lines of caregivers, groups of young people, or the spreadsheet report listing her orphaned clients by name, surname, age, and ward that was to be submitted to Social Services every month. Diminutive and feisty, Tumelo was energetic to the point that I found it difficult to keep up with her; she spoke quickly and changed topics at lightning speed. She was passionate, humble, and quick to laugh, and she had a particularly mischievous, conspiratorial smile.

Tumelo was the social worker who ran the Foster Care Pilot Programme in Dithaba for its duration. When the pilot was launched in 2007, I was responsible for its orchestration at Social Services, in conjunction with a major national NGO. In the programme’s initial phase, we had identified a number of priority districts – including Tumelo’s – and run in-depth training for teams of social workers in each. But, to my knowledge, only Tumelo’s office had gone as far as recruiting parents and placing children.

The idea of formal foster care was still unusual but not altogether new to Botswana when the pilot was undertaken. Social work degrees at the university had long involved a core course in managing foster care, and detailed procedures had been laid out in common law under the Children in Need of Care Guidelines (RoB 2005a). The guidelines provided for the temporary removal of children from their families, by a government social worker, in cases where professional assessment had raised significant concerns of neglect, abuse, or other pressing issues affecting the child’s well-being. Especially in cases where suitable extended family could not be found to take in the children immediately, the guidelines proposed that banks of vetted foster parents from the local community should be trained up for the role, to minimise disruption in the children’s lives. The ultimate aim was to work with families to address their issues and enable the return of children to their original households, or to negotiate their long-term placement with other suitable kin. In the context of the AIDS epidemic and perceived breakdown among extended families, social workers customarily expressed an urgent need for ‘alternative care’ for children, and many were concerned about the overcrowding and inappropriateness of institutional places of safety in this role. But they were equally uncomfortable with the notion of formal foster care. The guidelines had been ten hesitant years in the making, and by 2007 they had seldom been deployed in the removal and placement of children for whom they made provision. The problem was, according to my social work colleagues and my neighbours in Dithaba, that fostering the children of non-kin was fundamentally un-Setswana. Unsurprisingly, then, while the programme was the first of its kind, it had lapsed between my departure from Social Services and my conversation with Tumelo – although the NGO concerned was working diligently with a few remaining government supporters to revive it.

‘I’m not sure how it came to Dithaba,’ Tumelo admitted, as we reflected on the programme’s beginnings. ‘There were so many problems there at the time. Property grabbing was a serious issue.Footnote ¹ Family conflicts.’ I asked her what she meant. ‘Conflicts can be caused by lots of things – maybe jealousy of relatives, fighting over property, or just lack of understanding among siblings. Anybody can report it, though it might not come out clearly that it is conflict, but reading between the lines then one can see.’ I was struck by how mundane the sorts of conflict she was describing were – they were the sorts of everyday dikgang I had experienced living with the Legaes. But Dithaba was often singled out as having been particularly hard hit by AIDS from the start; the subtext of Tumelo’s comment seemed to be that these mundane conflicts were more serious, more numerous, or more frequently referred to social workers as a result.

Tumelo described how she managed the programme as it unfolded, from her two-day training workshop in the capital to the process of briefing the kgotla (customary court), the village development committee, and district councillors on the initiative. ‘They all knew cases’ that they thought appropriate for formal fostering, she noted. Rather than put out a call for volunteers, Tumelo worked in collaboration with these key village representatives to select roughly 20 women who could form a ‘bank’ of potential foster parents. They applied a range of criteria in their deliberations. ‘These were women who knew how to run their families,’ she explained of the candidates, ‘and know how to care. They have a heart for children, and love.’ Their families were stable; many were married, though not all; the number of their children was comparatively few, or the children were already grown up. The women were not necessarily wealthy but managed what they had well. When the women were called to a workshop on the new programme – covering parenting skills, children’s rights, and relevant laws, to which most of them would not have had formal exposure before – all came.

During the pilot, Tumelo had arranged a single removal and placement in the village, for three boys ranging in age from 9 to 13. They had been staying with their grandmother, but there had been fights among the family about food and over who would care for the children. Recounting the case, Tumelo didn’t go into detail – partly out of professional discretion, perhaps, but largely because it was a familiar narrative in the orphan care field and scarcely bore repeating. As we have seen, government provision of food baskets to the caregivers of registered orphans is widely understood as a source of significant conflict and competition among extended families – and as symbolic of their fundamental fractiousness, ruthlessness, and untrustworthiness as care providers for children. Again, the issue struck me as mundane, particularly as a justification for child removal. Tumelo left me to ‘read between the lines’.

In handling the case, Tumelo went to the kgotla first, accompanied by the boys’ grandmother and a letter written and signed by the prospective foster parent, Mma Dineo. ‘It was an emergency situation,’ she explained; she planned to follow the official legislative route, through the Children’s Court in the city, later on – though in the end they never did. But, she pointed out, ‘even if it can go to the courts, it has to go back to the kgotla; whatever is happening should be reported there’. She described the kgotla as a repository of local knowledge in which the movements of children and the promises and obligations of families should be stored – even (and especially) when the children and families themselves had lost track of them. A woman active and outspoken in local child protection initiatives, Mma Dineo had also been insistent about taking the proceedings through the kgotla. ‘She was very cautious,’ Tumelo reflected thoughtfully; ‘I’m not sure what about. Hei! That lady can talk,’ she added, noting with some chagrin Mma Dineo’s frequent visits to the social work office with concerns and complaints about her charges.

The boys had wanted to go to boarding school but instead moved in with Mma Dineo. Everything went smoothly at first – until the food basket and other government resources attached to the boys’ care followed them. Officially, the guidelines on formal foster placements explicitly forbade the provision of material support or remuneration to foster parents, in order to ensure that people did not take children in for ‘the wrong reasons’: exploitation or personal gain. In practice, however – especially given the connection between care and material support in Tswana understanding (Part II) – social workers and trained foster parents all expected that some compromise would be necessary, particularly if children were to be kept in their home villages. Reassigning government provisions to follow the children was the most obvious compromise to hand. The boys’ grandmother became furious with the arrangement and made her disgruntlement clear in public scenes at both the social workers’ office and Mma Dineo’s place. ‘I guess it was just jealousy,’ Tumelo explained, downplaying it, although the public exposure to insults of wrecking a family was no doubt a challenge even to the staunch Mma Dineo. Ultimately, Tumelo stressed, it did not derail the placement.

Shortly afterwards, some unexpected family turned up. One of the father’s younger brothers came looking for the boys, offering to take them. He said his family was angry and they wanted the boys back. The boys seemed to want to go back, too. ‘When we arranged for the boys to be fostered we didn’t know about those relatives,’ Tumelo explained, matter-of-factly. ‘We only found out about them after they came to find the children.’ Knowing that social workers were generally quite thorough in tracing extended families, I asked how they had been overlooked. ‘We didn’t really expect help from them,’ Tumelo explained, ‘and they were difficult to find.’ To reduce confusion, the father’s brother was initially turned away. After the boys were settled, he was called back, had the situation explained to him, and signed off on the placement as well.

A little over two years later, the man returned and offered to transfer the boys to the junior school in his village. ‘The family felt they had completed their punishment,’ Tumelo explained, paraphrasing his rationale. ‘So the boys went. But I just heard on Saturday that they want to come back to Dithaba. They are spoiled. I told Mma Dineo and the family, just accept them, they are children, don’t fight with them.’ Her complaisance seemed strange given the active role she had taken in their removal, placement, and later movements.

‘The placement was a success,’ Tumelo decided, after some reflection. ‘Maybe people feel deeply bothered by children being taken out.’ She shrugged. ‘To have the option of fostering is good.’ She noted that several of her current clients had had to be placed in a local place of safety, which she felt was overwhelmed and often ended up ‘chasing’ children back out to the social workers. ‘I’m not sure what institutions add,’ she mused. ‘Fostering is a way of teaching them it’s very important to have a family.’

Tumelo’s account makes plain the ways in which formal foster care in Botswana differs sharply from its antecedents: the informal circulation of children among kin and between non-kin. Again, these differences revolve primarily around approaches to dikgang. Circulating children among extended kin might be seen in terms of delegating responsibilities of care beyond the usual contribution-oriented economies and their conflict-management strategies, creating perpetually irresolvable dikgang; taking in non-kin as a suspension of dikgang, which neither exacerbates nor addresses them; and formal fostering as a deliberate attempt to decisively resolve dikgang. Where the first two reproduce appropriate distances of relatedness, the last risks conflating and collapsing them, offering not simply a temporary alternative family but an alternative model of kinship in its place.

Tumelo’s description of the dikgang arising among her client families is familiar from the sorts of conflicts we have seen already. While she did not explain how such issues were initially brought to her attention, it is most likely that she would have first come into contact with the families when they registered for the government orphan care programme. She may have been called on to settle intransigent disputes by the family itself, particularly if there were any conflicts over the food basket. Especially intractable problems at home may be handed to government institutions such as the police, clinics, and social workers – generally in the hope that the handing over itself, rather than any solutions that might be engineered, will help preserve the delicate balance of obligations and responsibilities, power and care, within the family. In this sense, families might envision the social worker’s intervention – including the placement of their children in temporary formal fostering situations – as simply a first step in the process of negotiating an ongoing family issue, or as a temporary suspension of that process.

However, in cases like those described by Tumelo, removing a child into formal foster care presents a problematic set of knock-on effects. The child himself, for example, is seldom the singular focus of dikgang, which reflect wider kin dynamics and demand reflection on the trajectories and quality of specific relationships. Battles over property or responsibilities of care and misunderstandings between parents or among their (often co-resident) siblings may all affect a child, but they seldom take the child as their object. A mismatch emerges between the family’s positioning of the social worker as an extrafamilial actor whose involvement might usefully suspend dikgang until the status quo can be re-established, and the social worker’s dual mandate of protecting children and achieving lasting fixes to family crises (whether in specific cases or by promoting alternative models of being kin). And this mismatch is exacerbated by a certain myopia on the part of the state; in spite of social workers’ best efforts in tracing families, the burden of their caseloads makes it virtually impossible for them to recognise the full range of kin affected, how they are affected, and how they intend the social worker to be involved. No wonder, then, that the boys’ father’s brother saw the removal as a punitive gesture rather than as a means of resolving the dikgang with which the social worker was presented in the first place. Critical capacities and responsibilities to contribute care for the boys (and for them to make their contributions in turn) were not only drawn into question but cut off; the ability to resolve dikgang appropriately in ways that involved them was removed, and the repercussions for reciprocal obligations between adults and children rendered deeply uncertain. In other words, the processes critical to forming kinship with, through, and around the boys had been foreclosed.

Worse than this, the family to which the child is removed is drawn into potential dikgang with the child’s natal family. The loss of the child, their work in the home, and any contributions of care they can mobilise is a source of serious bitterness and ill will towards the fostering family, as the grandmother’s fury and public insults demonstrate. In this situation, the social worker is the primary arbiter of conflict, rather than the child’s natal family. As Tumelo’s irritation with Mma Dineo suggests, the position of arbiter is hardly a welcome one for social workers: not only are they overwhelmed with their caseloads, but of necessity they are entirely disengaged from the day-to-day life of their client families, especially their conflicts – which require a great deal of unavailable time and effort to address. Most social workers will therefore hear out an issue, and perhaps offer advice, but will not re-enter the fray. Natal and foster families are thus drawn into kin-like (and kin-affecting) dikgang, without the means of resolution that might build connections between them and contain the risks that conflicts pose. As Erdmute Alber’s informants in Benin reflected of their changing fosterage practices, contemporary ‘fostering brings so many problems into the web of kinship that it is better not to take foster children at all’ (Alber Reference Alber2018: 146).

Beyond these new dimensions of dikgang, the formal foster parenting programme seems to presuppose and decree a certain ideal of closeness or intimacy between the foster family and the fostered child that – as we have seen – may be at odds with the more fraught affect that characterises usual practices of child circulation. The recruitment drive’s emphasis on able parents, who ‘know how to care’, ‘have a heart for children’, and ‘have love’, and the social worker’s willingness to ensure that additional material support is available to women who meet those criteria, are initial signs of this tendency. These attributes were, of course, appropriate to a Tswana mother; Livingston (Reference Livingston, Cole and Durham2007b: 183) glosses them as ‘moral superiority, a patient heart, and kindness’ (see also Ingstad et al. Reference Ingstad, Bruun, Sandberg and Tlou1992). But they are not necessarily the same traits expected of non-kin in looking after a child. Bonolo’s example in Chapter 11 showed us that these characteristics on their own are not necessarily kin-making; indeed, the absence of similar discourse in describing parenting ideals (focused more on ‘raising properly’ or ‘help’) suggests that they are relegated to the background, or at least left implicit. But as fostering families are also drawn into dikgang with their foster child’s family by the placement process, and unable to refer conflicts with the child back to his or her family or the social worker, they are placed in an increasingly isolated, replacement kin position.

Batswana may read formal fostering less as a matter of taking children out of dangerous families to safety than of bringing an entire network of non-kin into a level of partial intimacy and irreconcilable conflict that may make those non-kin themselves especially dangerous. In contrast to informal child circulation, formal fostering seeks to extend, supplement, and replace family; and, in the attempt – which can be only partially successful – it spreads the risk and danger associated with kinship instead of containing them. It presents, in other words, a worst-of-both-worlds scenario. Formal fostering interferes with the processes of differentiating kin that child circulation usually enables, thereby producing ‘a kind of contagion, a moral degeneracy’ (Wagner Reference Wagner1977: 624) that people register when they describe the practice as ‘un-Setswana’.

Legal rights in children were a key focus of early structural-functionalist approaches to child circulation and were used to distinguish adoption – where all legal rights to a child are transferred from natal parents to new parents – from informal fostering, where legal rights remain with the natal parents. In Tumelo’s account, ‘the law’ makes a slightly different distinction: between child circulation of the sort described in earlier chapters and formal fostering – not so much in terms of transferring rights, but in terms of offering protections and clearly structuring the roles, justifications, and processes of fosterage. In both, the crucial distinction the law makes is between arrangements made by, among, and through kin and those made by the state. As Tumelo’s description of her first formal placement shows, ‘the law’ deployed is not simply Roman-Dutch common law, governed by the Children’s Act (1981, 2009) and the Children in Need of Care Guidelines (2005), nor Tswana customary law, but a hybrid of the two.Footnote ² This hybridised notion of law was used to assess the need for children’s removal, to identify appropriate foster parents, and to anticipate any disputes that would arise, in part by identifying those who would mediate them (the social worker, but also potentially the chief). ‘The law’, in other words, takes responsibility for identifying and resolving dikgang among kin, displacing the ethical work usually undertaken by families. And, in turn, it reworks the boundaries between kin and non-kin. It requires and produces a muddling of intra- and inter-familial kin distinctions, of processes by which families manage dikgang, and of Setswana kin ideals. I suggest that it is the power that formal fostering gives ‘the law’ in deciding how families should work that makes many fundamentally uncomfortable with it – not least because it marks a fundamental inversion of what the relationship between kinship ethics, practice, and law should be. The ‘un-Setswana’ character of formal fostering also lies in the law’s attempt to redefine kinship practice and ethics, instead of being modelled on and directed by them.

Historically, much of the literature on fosterage and child circulation understood these practices primarily in terms of prevailing political and economic conditions (Alber Reference Alber2004; Bledsoe Reference Bledsoe1990; Block Reference Block2014; Goody Reference Goody, Alber, Martin and Notermans2013 [1982]: 3; see a similar argument in Alber et al. Reference Alber, Martin and Notermans2013b), often as a response to crisis (Ingstad Reference Hunter2004; see Goody Reference Goody, Alber, Martin and Notermans2013 [1982] on crisis versus non-crisis fostering), and specifically as a reaction to poverty, seeking opportunities for social advancement on the part of the child, the natal parents, or the foster parents (Archambault Reference Archambault2010; Bledsoe Reference Bledsoe1990; Leinaweaver Reference Leinaweaver2007a; Reference Leinaweaver2007b; Stack Reference Stack1974). But it is also an ordinary, widespread kin practice, not simply responsive to socio-political conditions, but also actively engaged in reworking them. Researchers, policymakers, and practitioners alike work on the assumption that child circulation in the context of AIDS can best be understood (and formally deployed) as a response to mass orphanhood and a crisis of care. However, taking cues from Schapera’s (Reference Schapera1940: 246–7) descriptions of practices that are familiar from my own fieldwork over 70 years later, in Part IV I have argued that child circulation forms an integral dimension of the ideals, structures, and practice of Tswana kinship more broadly. Further, I have sought to demonstrate that it serves not simply to extend kin networks or bind them more closely together, but also to differentiate and distance them; that dikgang are critical means and indicators of this differentiation; and, counterintuitively, that such differentiation is critical to the resilience of those networks.

In Part II, I examined ways in which relationships among siblings, between siblings and their children, and between grandparents and grandchildren are all frequently refigured as parent–child relationships – thereby generating a shifting field of generations and intergenerational relatedness. In child circulation, we find processes that work continuously to differentiate these relationships from one another and to prioritise certain parent–child (or generational) configurations over others. Thus, a girl like Lesedi might have an older sister, a biological mother, a mother’s sister, and a grandmother, all of whom are potentially and actually considered mothers to her. But with her biological mother largely absent, which meant that she was raised primarily by her grandmother, her grandmother takes precedence as mother; and Lesedi’s relationships with other members and generations of her family shift to accommodate this precedence. If her older sister lived elsewhere and they seldom saw each other, the sister’s parent role would be diminished in comparison; although, by the same token, if Lesedi went to live with that sister, the latter’s parent role would be gradually prioritised. And these configurations often change over time, depending on the circumstances of the people involved and their enactment of the key processes of kin-making we have explored – above all, their involvement in managing dikgang. While an ‘essential similarity flows between and among’ (Wagner Reference Wagner1977: 623, italics in the original) these relationships, making them each a sort of parenthood, the circulation of children – counterintuitively – works to disrupt that flow.

This differentiation, in turn, enables the multiplicity that is so characteristic of Tswana kin roles and relationships. One can have multiple mothers, be mother to multiple people (siblings, offspring, grandchildren), and be multiple sorts of relative (sister, mother) to a single person, not because these relationships are conflated and interchangeable, but to the extent that they are differentiated and particularised. Multiplicity suggests not simply that anyone or everyone can be someone’s parent, but that several specific people, by dint of their positions in a network of relationships, the responsibilities they undertake (of managing movement, contributing care, and jointly reflecting on and negotiating dikgang), and their explicit differentiation from one another, are one’s parents. Similarly, one can only be sister and mother to someone (potentially or actually) if these are differentiated roles. And it is perhaps this multiplicity above all – and, by extension, the means of differentiation that produce it – that has made families and kin practice so fraught and yet so resilient in the context of the AIDS crisis, and many other crises besides. While there is no question that socio-politico-economic contexts affect kin practice, it is not simply in terms of the stimulus–response effect that is often presupposed; one produces, is implicated in, and adapts to the other. I suggest that it may not be the epidemic itself as much as assumptions about the childcare crisis it has created – and associated policy responses to that apparent crisis – that have begun to introduce new variables into Tswana understandings of child circulation.

In Part V, I move from the creation of appropriate distinctions within and between families to the creation of appropriate distinctions between the family and the village. Taking my cue from concerns about the appropriate relationship between law and kinship noted in the context of formal fostering, I explore the work that goes into ordering interactions between the lelwapa (courtyard, house, or family) and the motse (village or community) or morafe (tribal polity). As in previous chapters, self-making is implicated in – and in turn enables – these processes of ordering and distinguishing. And as in all the scenarios I have explored so far, the management of dikgang plays a critical role as well – this time in terms of its exposure and concealment. These dynamics were thrown into sharpest relief in the frequent public events that characterised village life during my fieldwork – most notably, in a major celebration to honour Mmapula and Dipuo, in the first initiation held in a generation, and in an opening ceremony conducted by a local NGO.

***

INTERLUDE: Tumi’s Note