Late one night, less than a year after I had left the field – and not long before Christmas – I had a sudden and unexpected message from Moagi.

‘Hi dear, how are you? We have a problem at home. Kagiso is late, car accident.’

It knocked the breath out of me. I responded in urgent disbelief, asking what had happened, when, where. Moagi did not reply. I tried to reach other members of the family by text, but none of them replied either. Multiple phone calls wouldn’t connect or cut out after a few rings. Eventually, in a state of anxious dread, I got through on the family landline. Lorato answered as if she had been expecting me.

‘Who told you?’ she asked first. I explained I’d had a message from Moagi. She let out a sigh of relief. ‘We’ve been trying to figure out how to get hold of you. We were worried you would hear from someone else first.’

She told me what they knew of the afternoon’s disaster. Kagiso and his fiancée – for whom he had recently concluded negotiations and paid bogadi – had been driving to town in Kagiso’s car. It was a drive Kagiso made every day for work, often multiple times. It was a drive I had made hundreds of times myself. A truck overtaking at high speed hit them head-on. They were both killed instantly.

Everyone except Moagi was already home, and he was expected back from his post, across the country, the following day. Lorato described them to me, sitting scattered around the darkened lelwapa in silence, their faces intermittently lit by their mobile phones as they notified family and friends by text message. ‘Nobody can sleep,’ she said. We sat in silence on the phone for a while ourselves. The last time we’d spoken, we’d been anticipating the second stage of Kagiso’s marriage – a church wedding – and wondering when it might be held and how it should be organised. ‘I don’t need to hear anybody crying,’ she warned, adding, ‘It will be too painful.’ I swallowed and tried to heed the warning; we had each heard the tell-tale catch in the other’s breathing.

Steering ourselves back to safer ground, we started talking through everything that would need sorting out that week: the food to be bought, the programmes to be designed and printed, the tent to be hired, the firewood to be collected. ‘Ija! Ke dikgang hela,’ I said, trying to be light-hearted – nothing but problems! Lorato chuckled. ‘But there’s going to be a serious issue of some sort, isn’t there?’ I added more seriously, with a sudden sense of foreboding. The situation was so difficult already. ‘Gareitse wena, re tla bona,’ Lorato answered, sighing – we don’t know, we’ll see. We stayed on the line for hours, alternately chatting reflectively or sitting in silence – until the sun came up and Mmapula called everyone to begin preparing the yard and house.

Kagiso and his fiancée died early in the week. Funerals were usually held on Saturday, but no one was sure whether the arrangements could be made in time. Kagiso’s fiancée came from a village halfway across the country, and representatives sent by her family – parents, uncles, aunts – had to make their way to Dithaba before preparations could begin. They arrived on the Tuesday; that night, the Legae family hosted a large meeting with their guests to begin the funeral consultations. Unfortunately, my foreboding had been justified: dikgang emerged almost immediately.

‘They’re refusing to let us bury her,’ Lorato explained by phone when I called for an update. ‘When we called them at first they said there would be no problem, we could bury her here with Kagiso. He paid bogadi, right. But now we don’t know what happened, somebody must have changed, because now they’re refusing. Saying the marriage was not finished. They want to take her home.’ The insistence was unexpected and had thrown the meeting into disarray. Both families agreed to meet separately and to reconvene the following day. ‘Haish, wena, ke kgang e tona,’ Lorato commented, dispirited – it’s a big issue.

I called daily for updates, and I received regular text messages from Moagi and my friend Lesedi, who had arrived in Dithaba to help. The debate among the Legaes – including Kagiso’s parents, aunts and uncles, siblings, cousins, and other elder members of the family – was protracted. Some were piqued that the woman’s family could even suggest taking their daughter home for burial when bogadi had already been paid; lengthy exegeses of Tswana law were offered, and it was suggested that bogadi should be claimed back. Others – including Mmapula, Kagiso’s mother – were deeply hurt but could not summon the emotional will to fight, and they thought it best to let the issue go. Alongside these questions of principle ran equally urgent questions of who would meet which of the funeral’s steep costs, who should take on which formal roles (giving speeches, pall-bearing, and so on), how the programme should run, and how the extensive work of preparing for the event itself would be managed – all of which hung on the question of whether or not the woman’s family would contribute.

The two families met together and disbanded again twice more over the next two days, holding separate meetings in between. The woman’s family seemed to be as divided and uncertain as the Legaes were. Some were insistent about taking their daughter home for burial at all costs; others were quietly convinced that her place was with her husband, especially given that they had died together. The same concerns about cost, contribution, and organisation hung over their deliberations. The woman had left behind a young son; he had become close to Kagiso, but Kagiso had not paid the requisite cattle to take the child as his own. The problem of who would take on his care presented yet another thorny decision, entangled with and impinging on the others.

Muffled recriminations began to fly. Some of the Legaes wondered whether the woman’s family wasn’t holding out in order to retain exclusive benefit from the large payment anticipated from the Motor Vehicle Authority (MVA) – a government agency that paid out often significant claims to passengers injured or killed in car accidents. Others suggested that her kin had already stripped the woman’s house of furniture and belongings without a thought for her son’s inheritance. Witchcraft ran as a subtle subtext throughout: a likely explanation for how such a tragedy should befall a young couple, especially given Kagiso’s growing profile in business and the church, but also a risk that hung over each family and between them, should their multiple negotiations go awry and produce intractable ill will. It was a tense and dangerous time, compounded by the deep shock and pain of the two deaths.

Finally, late on Thursday, an agreement was reached. Kagiso and his fiancée would be buried together in Dithaba. The funeral would be held on Saturday morning. The MVA had indeed been consulted and was to provide a substantial sum towards the cost of the funeral. Those who had compromised by allowing the woman to be buried in Dithaba insisted that no expense should be spared. Joint teams, comprising members of each family, were sent to town to locate the best coffins. Modiri contributed no fewer than four cows from the family herd; vast quantities of food were procured by the women; the programmes were unusually large, at A4 size, and printed in full colour – making them exceptional enough to be fought over by those who attended the funeral. And hundreds attended. Most of both families were there, as were neighbours and friends from near and far. Many staff members, volunteers, and clients from the local home-based care centre for people living with HIV – where Kagiso and his fiancée had met and had worked together – were in attendance; they also contributed substantial financial support, and had helped design and print the much vaunted programme. Members of the couple’s church arrived from all over the district. Even the attendance of more remote figures was widely anticipated and rumoured – like the couple who had once run the local orphan care centre where Kagiso and I met, who had long since left the country; and, of course, my own – if ultimately disappointed. The funeral lasted much longer than usual to accommodate not only speakers from both families but also the kgosi, and – in a moving gesture – a spontaneous ceremony conducted by the elderly head of Kagiso’s church, who stood between the couple’s coffins and bound them together in Christian marriage.

***

Once the shock had faded, I was struck by the extent to which the deaths of Kagiso and his wife mapped and condensed the full range of dikgang – issues or conflicts, their negotiations and irresolution – that emerged over the course of my fieldwork. The preparations and the funeral that followed also powerfully demonstrated the ways in which kinship is constituted in crisis and conflict, rather than being destroyed by it. The sudden loss of Kagiso and his wife creates a darkly apt frame in which to draw together the stories of contemporary Tswana kinship I have tried to tell.

The dangers of distance, movement, and moving together – which figure critically in the spatialities of Tswana kinship – are especially pronounced in the case of a car accident. The distance at which Kagiso’s wife was living from her family, the necessity of movement in the couple’s personal and working lives, and the ways in which kin gravitated and were called to the lelwapa on receiving news of the deaths all resonate with the descriptions I have given of kin space. The Tswana gae, or home, is a multiple and scattered place – usually comprised of masimo (farm lands) and moraka (cattle post), with the lelwapa as a lodestone. Its dispersed places are integrated through gendered and generationally differentiated practices of movement, staying, and care work, which seek to strike a careful balance between closeness and distance – and which produce dikgang. Whether in disruptive intimacies and absences, like Dipuo’s during his dalliances with the neighbour’s widow, or in the necessity and risk of sending children like Tefo ‘up and down’ on errands, awkward balances must be struck between keeping family simultaneously together and apart. And it is in the continuous production and negotiation of dikgang that this balance is created.

The need to get away and be away from family, while remaining connected to them, loosely characterises the challenge of personhood as well. As we saw in the construction of Lorato’s house, building is a critical and continuous means of go itirela, or self-making. Building, too, invites a proliferation of problems: in the need to mobilise resources and manage labour, as Lorato did among kin, neighbours, and NGO connections; in highlighting the failures of others – in this case, Lorato’s aunts and uncles – to build; and in reworking one’s actual and potential relationships with relatives and partners, echoed in Lorato’s concerns about marriageability after she had built. Building, in other words, involves an accumulation of dikgang on the part of the builder, and an opportunity to demonstrate one’s ability to manage those issues – making dikgang central to personhood too. Batswana do not build in a vacuum, of course: governmental control of plot allocation, combined with shortened timelines for plot development and the advent of both governmental and non-governmental programmes to which builders have differential access (based on, for example, orphanhood), produces further problems to be negotiated. These dikgang, however, knock builders out of sync, introducing a temporality that interferes with usual tactics of negotiation and frequently frustrates progress. Many noted with dismay that Kagiso had not even managed to build before his sudden death – a fact made all the more bitter because he had helped improve and extend the house at home and had saved an amount substantial enough to build for himself, but had been unable to secure a plot. Even Lorato expressed guilt around this circumstance, having chosen to build for herself rather than giving her plot to Kagiso when he had requested it years earlier.

Similarly intractable dikgang are produced in the spatial practices of social work offices and NGOs working with families in the village. Kagiso was a driver for both the home-based care and orphan care centres at which he worked, and, as we have seen, the ability of NGOs and government to enable the movement of their clients was a key factor in their relevance to families. These agencies presented a surprising parallel to kinship spatiality – being equally scattered, requiring comparable movement, and emplacing the work of care in similar ways. But an inversion was at work: a centrifugal tendency which moved clients away from the lelwapa, and which managed boundaries and access in ways that both competed with and disordered kin spaces. These similarities and divergences demonstrated clear links between kin practice and intervention practice, but also antagonism, displacement, and disruption – a pattern that echoes throughout this book.

A preoccupation with who would contribute what, how, and in what spirit – in terms of things, work, and the sentiments they condensed – saturated negotiations around the funeral much as they saturated home life. For Batswana, kin care is constituted in contributions: specific material things (cattle, food, cash, cars, or clothes, for example); the work of acquiring, producing, or looking after them; and the sentiment that animates making them available to others. As Modiri’s responsibilities for the cattle, Kelebogile’s for food, or Lesego’s for cooking showed, expected and actual contributions define – and are defined by – kin roles, according to gender and age. And contributions, too, are subject to dikgang. The dikgang that emerged around both cattle and food enabled a shifting generational framework, whereby family members may inhabit multiple generational positions at once, creating alternately hierarchical and egalitarian relationships. Thus, Modiri was Kagiso’s brother as well as his father; Kelebogile was Oratile’s sister as well as her mother, and was grandmother to Oratile’s children. And all of these relationships were indexed by responsibilities to contribute. Contributions forgone – as when Lesego stopped cooking – mark a profound threat to these relationships; and the question of who would contribute what at the funeral posed the particularly fraught problem not only of how surviving family members related to their dead, but how they related to one another, within and between the two families.

At the same time, contributions are subject to competing claims. The very things and work that a family expects of a given person are expected by potential partners as well, and they also figure critically in other processes of self-making. Tuelo and Khumo faced these conundrums in trying to acquire things for themselves through others – Tuelo through theft and violence, Khumo through motshelo savings groups – and in the claims to which these acquisitions were subjected. The uncertainty that emerges around what people can and should contribute, what they will contribute, to whom, and for how long means that contributions are a fertile source of serious dikgang. Care, in turn, is routinely subject to crisis. AIDS, of course, is frequently described as presenting a ‘crisis of care’ – a framing that reveals a fundamental misunderstanding of Tswana practices of care. The crisis AIDS presents may differ in degree, but not in kind, from the ordinary crisis of kin care. Interventions in response to AIDS, however – the bulk of which prioritise the provision of the very same goods listed above – do produce crises that differ in kind: by disentangling care things from care work, leaving both subject to competition within families over who might be seen to be ‘contributing’ them, thereby profoundly disrupting the dynamic of contribution itself. These dikgang, like others generated by government and NGOs, evade the family’s capacity to negotiate them. The MVA payout described above posed precisely this threat, but the ultimate choice to contribute it entirely to the costs of the funeral defused its disruptive potential.

While the question of contributions cast a long shadow over the funeral, at the heart of the dikgang that emerged was the drawn-out, highly uncertain process of marriage. I have suggested that intimate relationships become kin relationships through a gradual and carefully managed process of recognition, whereby they become visible, speakable, and known among increasingly wide circles of family. Women achieve recognition most meaningfully through pregnancy (which generally precedes marriage, as we saw with the parallel cases of Lorato and Boipelo); with men, however, it is often first conferred through negotiating marriage. Every shift in recognisability is marked and achieved by dikgang, the negotiation of which progressively expands to include additional relations; and their engagement with these dikgang determines not simply how families might relate to one another, but also the viability of the relationship their recognition shapes. Kagiso’s first attempt to marry a previous partner was ultimately scuppered by her father’s unwillingness to engage in negotiation, which sank the relationship in turn. And the father’s unwillingness was rumoured to stem from unresolved dikgang between himself, the girl’s mother, and her family, demonstrating the intergenerational ripple effects that the characteristic irresolution of dikgang can produce. Although Kagiso and his family had successfully negotiated his marriage on their second attempt, including the payment of bogadi, his wife’s family’s initial refusal to bury her with him underlines the highly tenuous – even reversible – nature of recognition conferred by marriage, especially to the extent that it relies on the indeterminate dimensions of dikgang. The couple’s posthumous wedding was moving in part because it signified a final, irreversible recognition of the sort unavailable in life. It marked the successful negotiation of dikgang between and within the two families; and it settled any outstanding issues sufficiently that the child left behind would not inherit them when it came time for his own marriage.

Although the accumulation and successful negotiation of dikgang promises a stable accumulation of personhood – pregnancy decisively reworks a woman’s position and relationships in her natal family, just as marriage reworks a man’s position in the community, regardless of whether either the child or the marriage survives – building relationships through such dikgang is also risky and prone to failure. In this light, the risk of contracting AIDS becomes one among many risks associated with intimate relationships; if its stakes are higher, they are understood more in terms of potential effects on negotiating recognition than in terms of life and death. Indeed, I have argued that protection against the indeterminacies generated by recognition and the dikgang it generates may be more crucial to Batswana than protection against contracting the disease itself – a possibility that goes some way in explaining stubbornly high rates of HIV infection in Botswana.

The child left behind by the deaths of Kagiso and his wife brings us to the question of children’s mobility, claims of responsibility for their care that are made or rejected, and their potency in asserting the limits of kinship. The son of Kagiso’s wife had been moving between his mother’s house in Dithaba and her family’s house across the country; while he had become used to Kagiso, and we had seen him frequently at home, the decision was taken to return him to his mother’s natal village. Given that he had already been in frequent movement, continued shuttling among kin of the sort we saw with Lesedi’s family was highly likely. I have described this kind of child circulation as an experimental extension of the circulations of kin, the economies of contribution, and the recognisability of relationships. It attracts potential dikgang connected to all three, the management of which serves to articulate which kin might be considered ‘close’ and which ‘distant’. Drawing on Lesedi’s disillusionment with the ‘far kin’ who offered to send her to school but instead treated her like a maid, I argued that child circulation tends to reproduce appropriate distances of relatedness rather than producing new bonds of closeness. It asserts limits, differentiating and distancing kin. Lesedi’s case and that of her cousin Tumi – taken in by a ‘close’ aunt – demonstrated that circulating children among kin already bound by economies of contribution tends to produce irresolvable dikgang managed in much the same way as those linked to contribution would be, leaving relationships unchanged. By comparison, children moving to stay with non-kin – as Bonolo did when he decided to foster himself to the Legae family, in response to ill treatment and a fear of witchcraft at home – does not necessarily involve establishing kin-like relationships, partly because dikgang are suspended and ignored, neither worsened nor addressed. Formal, government-sponsored fostering, in contrast, seeks to form relations of mutual care, responsibility, and love between non-kin, and attempts to provide a permanent solution to dikgang. In this sense, formal fostering seems to collapse appropriate distances among and between families that child circulation would ordinarily reassert, while offering not only an alternative family but an unmarked alternative ideal of kinship in its place – creating a disruption of and direct competition with usual kin practice, reminiscent of that seen in preceding chapters.

In part, Kagiso’s wife’s family sought to address the dikgang generated by her death by turning the funeral into a major event. As we have seen, such events serve to articulate the boundaries of family and to establish its proper relationships to the wider community, while offering opportunities to redefine personhood. The presence of everyone at the funeral, from neighbours to chiefs and churchmates to friends, the provision of food, the programme, and the management of work and space were all reminiscent of the priorities asserted in the family party a year earlier – if with distinct, dark differences. Death itself, and the couple’s posthumous marriage, marked a new configuration of personhood for them both.

I have suggested that the distinctions between family and village relied primarily on the careful management of hiding and sharing – an echo of the recognition dynamic – through which non-kin are drawn into the family’s performance of success, but carefully excluded from the messier realities of dikgang. Attention to dikgang demonstrates ways in which the lelwapa generates and permeates the motse (village), articulating a relationship in which the village is understood to begin in and to be sustained by the home. The deaths of Kagiso and his wife – especially given the hopes people had for their growing prosperity – marked a disastrous sort of inter-familial kgang, making the performance of a successful response all the more critical and complex to manage. This imperative weighed on the negotiations leading up to the funeral, and on the question of how each family could best demonstrate its own ability to respond – forcing the question of which dikgang needed most to be hidden from the other family, which shared, and how. It was partly this consideration, I suggest, that motivated the debate over where Kagiso’s wife should be buried. In the end, the two families seem to have concluded that they were in a much better position to preserve the priority of the relationship between family and community by working together. Jointly, they could draw in the maximum number of people to whom they could demonstrate their encompassing reach (through costly coffins, fancy programmes, and ample food). And, having successfully negotiated the question of the burial between them, they could be relatively confident that they would prove able to contain further dikgang that might arise. Together, they succeeded in drawing the village – or the better part of two villages – into the home, containing any risks involved, and securing the transformation of the couple’s status by doing so.

Finally, the ambivalent influence of government and NGOs around the funeral – at which they are simultaneously absent and present, marginal and critical – echoes the ambivalence both of their influence in the home and of the home’s influence on them. As we saw above, a major financial contribution from a government agency – the MVA – was both a source of suspicious speculation and a means of achieving compromise within and between the two families, in part by alleviating the burden of contributions they faced. Kagiso and his wife were also linked to a variety of AIDS-oriented NGOs in the village through their work and even in their relationship: members of the home-based care project figured strongly among the attendees at the funeral, carried out some of the work involved, and made significant contributions, honouring relations mediated by and through the project on which they had worked together. And, of course, my own connections to Kagiso and the Legaes had been forged through our mutual involvements in the orphan care NGO. At the same time, that NGO had long since closed, and, despite my connection with the family, I was unable to attend the funeral, underscoring an inevitable tenuousness in that connection.

As much as these agencies may succeed in mediating kin relations, ultimately they can neither enter into nor incorporate the family. While the home-based care staff mediated Kagiso’s relationship with his wife and participated meaningfully in the funeral, they were not themselves family and could not have participated in resolving the issue of the wife’s burial. The impacts of NGOs and government agencies on kinship practice are achieved, as it were, by knock-on effect rather than through direct involvement in kin relations as such. Families, for their part, efficiently draw such agencies into the realm of kin practice while carefully excluding them as kin actors.

What these projects do evince, however, is a tendency to deploy kin-like structures and practices – including familiar dynamics of dikgang – in the internal dynamics of and relationships between state and non-governmental agencies at local, national, and transnational levels. Kagiso’s marriage to his co-worker, the fact that his sister worked in the same NGO, and the fact that its founder was a close friend of their mother and took a parental concern in them all, indicate ways in which the home-based care project relied on, mediated, and reproduced kin relations. And those relations, in turn, naturalised, depoliticised, and legitimised its work. But this naturalisation presented a conundrum: NGO and government projects alike draw on multiple, mutually familiar and yet divergent notions of kinship within a discourse and practice that explicitly differentiates them from the family. As we have seen in both social work offices and NGOs, and perhaps most powerfully at the opening ceremony of the NGO campsite, these projects bring an array of Euro-American notions of what kinship should be, and of how it should relate to politics, into jarring juxtaposition with their Tswana counterparts. These contrary notions mirror but also disrupt one another, especially because they are obscured and implicit. This same mirroring and disruption, of course, is evident in intervention programmes’ influence on households. I suggest that the frustration that plagues governmental and non-governmental intervention in families – or, at least, the sharp divergences between their intended aims and actual outcomes – may be traced back to this confusion of kinships, and their tendency to saturate and overwhelm the bureaucratic practice that seeks to contain and instrumentalise them. At the same time, this confusion exacerbates the dynamics of permeation described above: kin logic and practice do not simply escape political projects of containment – they define, motivate, and disrupt them from within.

Everything that we might understand as constitutive of Tswana kinship thus creates dikgang, the negotiation of which produces additional dikgang in turn, in a continuous, fraught, and yet surprisingly innovative and generative cycle. Dikgang, in this sense, form a critical dimension of Tswana kinship. This understanding of dikgang suggests a novel role for crisis and conflict as something more than simply external influences on kinship practice, or as unfortunate but anomalous and fundamentally inconsequential corollaries of being family. I have attempted to make the case that crisis and conflict are, instead, constitutive of kinship. I suggest that crisis is inevitably produced by deep tensions and contradictions in the work to which kinship is put – between, for example, enabling the development of a distinctly individualist personhood while retaining the togetherness and mutual support of family; between creating closeness and maintaining distance, accumulating and sharing (or sharing and separating), recognising and concealing, connecting and dividing, creating ‘publics’ and preserving ‘privacy’; or between multiple ideals of kinship and between its ideals and reality, among other contradictions we have seen. And I suggest further that it is the ongoing negotiation of crisis that enables kin to strike unlikely balances among these opposing imperatives, continuously and creatively. For Batswana, the ongoing negotiation of dikgang both defines and differentiates relationships among kin – by generation and gender – and between kin and non-kin; and, at the same time, it is fundamental to the reproduction of kinship. Self-making, too, emerges as a process of accumulating and managing dikgang, and it waxes and wanes depending on the sort of dikgang that one has undertaken and one’s success in facing and living with them (since their resolution is often suspended indefinitely). The notion that crisis and conflict might reflect not only common experiences but crucial dimensions of kinship helps explain the surprising resilience of kinship in times of major socio-political crisis, such as Botswana’s time of AIDS – and also provides ample opportunity for cross-cultural application and comparison.

The question remains as to how far we can push the idea that crisis or conflict constitutes kinship. Throughout my time in Botswana, so-called passion killings – murder–suicides, usually committed by young men who killed first their girlfriends and then themselves – were rife and were subject to extensive public commentary and concern. Passion killings were often the result of dikgang between partners (and occasionally their families) of the sorts I have described, as well as being a source of serious public dikgang, and they suggest one violent limit on the generative potential of conflict. There is also some question as to whether different socio-political contexts of crisis and conflict work differently on the crisis dynamics of kinship. While I suspect that some of the conclusions drawn here about the AIDS crisis might apply to other public health crises, comparison with different sorts of large-scale crisis or conflict – whether natural disasters, overt political violence, or economic collapse, for example – in different socio-political and cultural contexts might reveal other critical limits to my argument. In the example of Botswana’s experiences of and responses to AIDS, I hope at least to have challenged the prevalent assumption that crisis and conflict simply destroy families, and that the only way of understanding kinship in such circumstances is in terms of breakdown or collapse.

In virtually all of the cases suggested above, of course, there is no single cultural framing of crisis or conflict at work. There are multiple framings. To the extent that each framing invites intervention, and to the extent that those interventions originate in a vast range of different institutional and socio-cultural contexts around the world, to talk about crisis is automatically to make connections and comparisons. Crisis, in other words, is exceptionally well suited to comparative anthropological study. The perspectives I have provided here would undoubtedly benefit from further investigation into the ways in which socio-cultural attitudes towards conflict or crisis, and their implications for families, inform humanitarian intervention ideology and programming originating outside Botswana.

In challenging the assumption that crisis is simply destructive, I have also sought to provide a fresh perspective on the wide array of governmental and non-governmental programmes that take it as their starting point. Part of the motivating concern of my research was to shed light on those factors that consistently frustrate family welfare programming in Botswana, and that frequently produce unintended and highly problematic knock-on effects for the families they seek to assist. As we have seen, most of these factors can be traced back to a fundamental misunderstanding of the elasticity of the Tswana family and of the importance of dikgang in that elasticity. This misunderstanding underpins other problematic assumptions in turn, about everything from the spatial and temporal norms of the Tswana home to the management of resources among kin, from the long-fraught unfolding of kinship reproduction and the life course to relative assessments of risk, from the role and power of children to the relative priority and power of family, as well as – perhaps most crucially – a misapprehension of the ways in which kinship and government or organisational practice ought properly to relate. While many of these programmes have adopted practices reminiscent of Tswana kinship practice, and thereby create an influential resonance or link, their effects have been to disrupt, invert, muddle, overextend, and competitively displace existing kinship practice. Combined with an explicit mandate of alleviating crisis, resolving conflict, and recreating the broken family in an appropriately ‘modern’ shape, these mis-framings create a legacy of disarray that has affected the Tswana family much more deeply than AIDS itself. While the conclusions drawn above suggest a certain inescapability in the dynamics they describe, they also provide a fundamental reframing of the problems facing Tswana families that holds the potential for experimentally rethinking both social work and NGO practice.

Finally, I hope to have provided a case for rethinking the conceptual and experiential relationships between kinship and politics, as we understand them in social sciences research. Michael Lambek argues that kinship is characterised by a ‘surfeit of meaning, relations, and sentiment’ (Reference Lambek, McKinnon and Cannell2013: 242); I have argued that much of the work of kinship for Batswana is to contain, shape, and direct that surfeit and the dangers it presents. The goals of states and transnational organisations working with families might be understood in much the same terms of containment and control (ibid.: 251–5), and of redirecting that surfeit to naturalise and justify institutional exercises of power. Paradoxically, however, agency interventions in family strategies of containment disrupt that work of containment, producing a confused, undifferentiated, and unbounded profusion of meaning and relations. To use Lambek’s terms, state and organisational intervention in kinship exacerbates its ‘immodern’ excesses precisely in the ways in which it seeks to eliminate or modernise them. This disruption and exacerbation is not simply a matter of problematic systems that need to be fixed, however; nor is it simply about the depersonalised and dehumanising effects of bureaucratic systems. It is, I have tried to show, a direct product of the surfeit it seeks to contain: states and transnational organisations fail with families because their work is understood, experienced, and enacted in kinship terms and kin-like practice, and because these terms and practices tap into a multiplex confusion of kinship models. Such analytical possibilities emerge only when we read kinship and politics together, rather than assuming that they are separate and exist in fixed relationships to one another (McKinnon and Cannell Reference McKinnon, Cannell, McKinnon and Cannell2013).

***

A few months after the funeral, I called the Legaes to see how things were going. It had been a difficult time for all of them. Mmapula had not ploughed – the funeral had been held at the beginning of the ploughing season, and she had not been out to the lands since – and so food was in shorter supply than usual. Winter was coming and warm clothes were scarce. The combi-van that Kagiso had run as a school bus to the nearby town had broken down; two of the children who had enrolled in school there were struggling to get back and forth. The younger children had been deeply upset by Kagiso’s death and were inclined to reminisce about their uncle, including by posting photographs of his bogadi negotiations on Facebook. Mmapula had reprimanded them harshly for vocalising their memories more than once, and had taken to making wry comments about how much they ate, as if they hadn’t noticed that their uncle was no longer there to feed them. Meanwhile, she and Dipuo had paid to have the wrecked car towed home, and it remained in the yard behind the house – a fact many friends and neighbours expressed concern about, partly because of its symbolic concentration of grief, and partly because of implicit concerns that it may have been bewitched.

But perhaps most worrying of all, Dipuo had been making more strange and unsettling pronouncements – and they had been taking on increasingly dark overtones. ‘He said something to Khumo about the next one who’s going to go under the ground,’ Lorato said. She wasn’t sure of the context or complaint, but the statement itself was so threatening that it left no room for extenuating circumstance. ‘The old man likes to blame Kagiso’s death on the Bangwato,’ – another Tswana tribe – ‘but these days Modiri is wondering whether it wasn’t him [Dipuo]. It’s like that’s what he’s trying to say. Modiri and the others are planning to call him and tell him that if he doesn’t promise to come back from the lands to stay in the village, they’re out [of his life].’

The call weighed heavily on me long after I had hung up. We had discussed various tacks to be taken on each of the issues in turn. Modiri was already fixing the combi. Khumo was looking into boarding school options. I offered to look into finding good winter clothes coming on sale in the northern hemisphere that I could send. Lorato had agreed to talk to her grandmother about our collective concern over the car and see if she would be willing for us to pay to have it removed. I talked to the children about being considerate towards their elders’ discomfort with speaking about their late son. Oratile agreed to talk to her mother about the way she was speaking to the children. Modiri and Khumo would call their father. It would all take time. None of it suggested decisive solutions – indeed, most of it suggested more problems to come. Being so far away, it felt overwhelming, and I felt impotent.

Over the next few weeks, there was a spate of Facebook activity among the family who used the site. Boipelo created a family Facebook group and posted photographs from the last Christmas I had been in Botswana. Tshepo posted lovelorn status updates; Lesego deftly deflected online suitors; Lorato posted a note to the family group to say that she was moving to the next town for a new job. Moagi wrote to say hello while on a work trip up north to Kasane; Oratile, who had also moved for work, wrote to tell me that she was taking some of the children from home to stay with her for a while. The contrast with the weighty phone conversation was striking: here there was a sense of growth, movement, and possibility.

On reflection, I realised that even the density of dikgang I had heard about over the phone presented possibility, in its own way. Modiri’s insistence on calling his father home opened up the possibility that he and his siblings might successfully assert a new authority. The children’s insistence on vocally reminiscing about their uncle opened up the possibility of reworking their relationships not only with his memory but also with their grandmother. Tshepo, who had been commuting to school, began boarding, which afforded her considerable comparative autonomy for the first time. Khumo, who had been working doubly hard at the lands, was gradually solidifying a claim to continue working them as her own. Each of these possibilities, of course, presented new dikgang in turn; but, taken together, they reminded me that among family dikgang are never completely unmanageable. They are always already in the process of being dispersed, suspended, or transformed into other dikgang, which are also negotiated into new manageability, in a continuous, generative cycle.

I realised that the apparently intractable knot of problems with which I had been presented over the phone had not been given to me for untangling, nor simply to re-entangle me; it was meant to draw me back into the continuous processes of disentangling in which I had a part to play, but that reached well beyond me and involved us all. For all that I had come to understand the dynamics of crisis in the Legae household, my default position was still to frame problems as things that needed solving, possibly by me – an artefact of my time working in both NGOs and Social Services, and of my own personality and upbringing, without doubt. But for the Legaes, including me in dikgang had always been, first and foremost, a way of including me in family. Dikgang were what we shared when we spoke together; they were what brought us together and what kept us together. Moagi would often say, by way of concluding his brief updates on the unfolding dikgang of the funeral, ‘Re mmogo’ – we are together. And in that simple statement, he reminded me that, for all the dikgang we had faced, and for all the directions our lives had taken, we were indeed still together; and that in the face of these new challenges – indeed, because of them – we would remain so.

Figure 11 Fireworks. Celebrating the New Year in the Legae yard.