This paper examines the intertwined issues of memory and forgetting, focusing particularly on the question of precisely how, and by what social mechanisms, forgetting is accomplished. I discuss how collective and individual forms of forgetting are central to Bourdieu's notion of the habitus, commenting that the habitus is a living paradox, foreclosing (unimagined perhaps because unimaginable) possibilities and opening others only when moments of improvised reflection intervene. Moreover, the systems of the habitus enact a forgetting of the strange, the marginal, the in-between, and even the singular and the autobiographical. I explore these issues through the juxtaposition of formalized, collective Samburu (Kenyan pastoralists) memory forms and the illicit sexual practices that underworld them.

In Zanzibar, many cases in rural Islamic courts involve disputes about whether or not a divorce has taken place outside of court. Zanzibari men have the right to divorce their wives unilaterally through repudiation; and, because many such divorces take place out of the wife's presence, women interpret certain structural events associated with divorce as divorce even when there is no evidence of lawful repudiation. By going to court, women want to legitimize the events of divorce with a receipt of registered divorce. Although the Islamic judge will not validate alleged divorces without proof of repudiation, he does not dismiss the cases as simple misunderstandings. Rather, he stresses the preservation of the marriage through reframing the cases as disputes about marital rights and obligations. Women acknowledge this shift and move to assert their rights to request better maintenance or a court-ordered divorce.

Memory is often considered a monopoly of adults and older children: the younger the child, the less significant the capacity for recollecting. In Côte d'Ivoire, the Beng posit a radically different theory of cognitive development: adults say that the younger the child, the keener the memory. Moreover, such recall is of a specific sort – infants allegedly hold strong memories of a previous existence before birth (wrugbe), where people reportedly live harmoniously and there is never material want. Nevertheless, remembering this space of plenitude can prove agonizing for babies, making their hold on life precarious, and a distressing array of culturally shaped diseases threaten their survival. Protecting against illness requires an elaborate bathing, jewellery and make-up routine twice daily that begins at birth and continues for the first year. All this somatic activity is meant to lure the child fully and definitively into this world, and to counteract the strong call of the afterlife that adults say is created by the infant's own memories. Why is wrugbe, as purportedly remembered by infants, envisaged as a place of plenitude? And why is it located in an historically identified past? In this essay – which is necessarily to some extent speculative given its subject of infant memory – I explore the allegorical implications of the Beng afterlife, suggesting that the attribution of heightened infant memory of an idyllic wrugbe serves as an indirect critique of French colonialism and its aftermath. I conclude by discussing the ways in which memory and forgetting are mutually constructed, with the Beng model offering substantial support for the contention that reproduction in general – and babies in particular – are crucial to this intertwined process.

African modernity surely comprises less pervasive outright violence than is suggested by popular representations ‐ the ‘New Barbarism’, as Paul Richards (1996) calls it. But by now it surely comprises more violent disjunctures than scholars like me, focusing on emergent and continuous change, have addressed. Uncertainty as a recurrent condition of the longue durée in productive and commercial life is rather different to theorize than violent ruptures in the very existence of collectivities or in people's capacity to imagine the future. It is one thing to link conditions to social organization, ecological knowledge and demographic regime (Iliffe 1995; Lesthaeghe 1989), for example to see the connection between the chronic historical vulnerability of the era of the slave trade and the political structures of the time (Ekeh 1990). It is really another to foreground each disruptive event in slow motion, one at a time, and ask how people created particular futures from particular pasts and presents. This shift of focus from chronic conditions to specific intrusions opens up in a new way the space between event history and the longue durée. It even puts the whole idea of the longue durée into temporary abeyance. Are long‐term processes better thought of as created by cumulative discrete and different memories and projections, perceptions and persuasions, rather than by responses straight out of the cultural/institutional repertoire? Or, at the very least, does the relationship of event to process need to be traced out explicitly?