In 1936 Chief Philip James Karanja wrote to the colonial government of Kenya regarding the resettlement of a community displaced by European encroachment. He pointed out that some of those with farms that were being appropriated were ‘rich people who have large areas of land in the Reserve.’ In light of this, he proposed that such people should not be awarded more land as part of the resettlement. ‘This will give a chance to the poor to get shambas (farms) and live without trouble in the future.’ ‘It is better to help the poor rather than to help the rich men,’ he concluded.Footnote ¹ This was a context of extreme land alienation and displacement, yet the effects of the colonial policy of white settlement were not equally felt among Gikuyu people. The tensions engendered by this socio-economic polarisation — and its powerful implications for the ability of young Gikuyu men, in particular, to become socially-recognised adults — fed directly into the violence of Mau Mau less than twenty years later.Footnote ² Chief Karanja's charitable view towards the poor was politically astute. But was it surprising?

John Lonsdale's work would suggest that Chief Karanja's sentiments in favour of helping the poor over the wealthy were a reaction to transformations wrought by the colonial economy and its profound inequalities, rather than a continuation from the pre-capitalist era. Lonsdale argued that in the moral economy of Gikuyu society prior to colonisation, ‘the rules of social obligation scorned charity.’ ‘The poor had,’ he continued, ‘no heart, no friends,’ and their ‘hunger kept nobody else awake.’ Given the greater availability of land prior to white settlement, it was ‘only fecklessness’ that ‘excluded the poor from the means of production.’ Lonsdale's survey of Gigikuyu proverbs revealed that while they often praised reciprocity, they did not praise generosity towards the poor.Footnote ³ By the late nineteenth century, at least, ‘the poor were despised,’ even if in part that was because their condition reflected the moral failure of those who were better off to provide for them in times of crisis.Footnote ⁴ The difference between this normative view of the poor and Chief Karanja's outlook points to the value of serious study of poverty and perceptions of it over the long term. Lonsdale's starting point was the late nineteenth century; I want to make a case for a much deeper time perspective.

The question of poverty in precolonial Africa has long been contentious.Footnote ⁵ In large part, this is a problem of definition — poor by what measure? — and of scale — whose poverty? For an earlier generation of scholars who examined the impacts of the trans-Atlantic slave trade and colonialism on poverty at the macro-scale in particular, the definition was framed in relation to economies in other regions.Footnote ⁶ When did Africa become poorer than other world regions and why? The debates have thus often focused on trying to quantify the economic consequences of inter-related processes. First, of the forced removal of millions of people, and the deaths of many more millions still, through the slave trade to the Americas. And second, of colonial conquest, colonial extraction and the continuation of unequal trade patterns.Footnote ⁷ This regional comparison still appeals to some economic historians, although the explanations offered vary.Footnote ⁸ While this research can be valuable in redressing lazy arguments about Africa's inherent economic backwardness, these approaches necessarily flatten the extraordinary socioeconomic diversity of the continent. Recent work, such as that by the economic historian Gareth Austin, has sought to counter this with more dynamic frameworks that allow for historical diversity and contingency.Footnote ⁹

Nonetheless, this macro-level flattening led, in much of the earlier literature, to an idealised conception of an egalitarian approach within precolonial African societies that has retained a broader popularity. A good example of this was when Walter Rodney wrote, that in ‘Africa, before the fifteenth century, the predominant principle of social relations was that of family and kinship associated with communalism.’Footnote ¹⁰ This view echoed the ideologies of, and was in turn echoed by, some of the first generation of leaders of independent African nations. Fifty years ago, for example, Julius Nyerere recognised that poverty had existed in ‘traditional life.’ Still, he argued, it was an egalitarian poverty because ‘all the basic goods were held in common, and shared among all members of the unit… no one could go hungry while others hoarded food and no one could be denied shelter if others had space to spare.’Footnote ¹¹

There was in all of this a desire to romanticise the deeper African past, to argue that while people may not have had much, they shared what they had. In making that case, Bertrand Magubane, Catherine Coquery-Vidrovitch, and others were writing in a very different vein to John Iliffe, who asserted in The African Poor that ‘Africa's splendour lies in its suffering.’Footnote ¹² I encounter arguments for a rosier, more egalitarian past when presenting my research on the history of poverty in eastern Uganda. Some claim, for example, that people in eastern Uganda could not have been materially poor before colonial conquest unless they had no family or kin. While others argue that the philosophy of Ubuntu was ubiquitous across the sub-continent and mitigated any economic disparities. From a historian's perspective, it is hard to reconcile these arguments with the ample evidence for precolonial poverty, for socio-economic inequality and for, at times, highly negative views of the poor in different parts of the continent. Consider ‘The Hunters’ Oath,’ or ‘Manden Kalikan,’ an oral text with origins in the founding of the Mali empire, that referenced both the existence of poverty and a certain ambivalence towards the poor: ‘The evil of poverty ends today in the Manden. / Hunger is a bad thing, / The hungry man knows no shame; / Poverty is a bad thing, / And the poor know no respite.’Footnote ¹³ We would do well to heed Lonsdale's warning against romanticising ‘the moral economy of precapitalist worlds,’ not least because, in his view, ‘the morality of agrarian society was calculating.’Footnote ¹⁴ Perhaps the best way to avoid falling into either trap — of romanticising moral economies or dehistoricising poverty — is by focusing instead on the internal dynamics of African societies and, thereby, foregrounding how people experienced and understood poverty at various points in the past. This has the additional benefit of allowing specific historical contextualisation, rather than abstracting to the level of macro-scale world-systems models. Drawing on diverse methodologies and approaches, such as archaeology, comparative ethnography, conceptual history, historical linguistics, and oral traditions, we can then build comparisons from solid evidence that will help us understand the cross-cultural dynamics of inequality.

Research on the longue durée history of poverty on the continent has highlighted its diversity of forms, but clearly demonstrates the long-standing existence of economic inequality. In 1977, Robin Palmer and Neil Parsons offered a long-term overview of the region's economic history in The Roots of Rural Poverty in Central and Southern Africa.Footnote ¹⁵ The editors, and the authors of several of the chapters in the volume, were primarily occupied with tracing the economic impact of the region's incorporation into global trade networks, with a particular emphasis on the decline in people's standards of living. They noted, ‘in many instances underdevelopment was not merely a matter of increasing economic distortion, dependency, and subordination for the masses of the people; it was also a matter of absolute impoverishment against previous standards of living.’Footnote ¹⁶ Quantitatively measuring the absolute impoverishment that resulted from the slave and other trades remains, even forty years on from the publication of this work, an elusive goal. There is an important divide in studies of poverty, in Africa and elsewhere, between quantitative approaches and qualitative ones. Both Jane Guyer and Morten Jerven highlight, in this forum on poverty, the difficulties with statistical data even for more recent decades, whether because units of analysis hide as much as they reveal or because the quality of the data is too low.Footnote ¹⁷ Yet it is the notion that poverty becomes something that can be measured that underlies, in part at least, its description as ‘modern poverty,’ as Vincent Bonnecase demonstrates also in this forum.Footnote ¹⁸ Quantitative data becomes exponentially sparser as we move back in time. The challenges of using it to meaningfully describe poverty increase as the data becomes more limited. It is, therefore, qualitative studies that often provide the most insight for histories of poverty in the deeper past, not least because they start to address what poverty meant in particular societies and how what it meant changed over time.

Despite the elusiveness of measures for absolute impoverishment, Parsons and Palmer's volume was an important — and still rare — contribution to the history of inequality and poverty in sub-Saharan Africa that sought to provide a deeper context for the developments of the twentieth century. For all the dramatic changes of more recent centuries, they argued that it was possible to ‘begin to distinguish between rich and poor, in a class sense, within the same society on the basis of differential access to and control over the means of production,’ as long ago as the Early Iron Age.Footnote ¹⁹ But it was in the eighteenth and nineteenth centuries, with the expansion of extensive international mercantile networks into the interior, they argued, that ‘distinctions between rich and poor, oppressors and oppressed, took on new dimensions in many areas.’Footnote ²⁰ This raises the question of how forms – and meanings – of poverty changed between the eleventh and eighteenth centuries and how we might uncover those changes in the absence of much of a documentary record.

Writing about West Central Africa, Jan Vansina delved even deeper into the past to explore the necessary conditions for the emergence of notions of wealth and poverty. In How Societies Are Born, he drew on the premise that there are specific material prerequisites that enable the existence of ideas of property and ownership and argued that these in turn make possible notions of poverty and wealth in any given society. Among the foraging communities that lived in the region from around ten thousand years ago, Vansina argued, the necessity of mobility impeded the acquisition of material goods and limited social stratification. Material objects were few and were viewed neither as personal property nor as commodities.Footnote ²¹ Without a concept of personal property, he posited, these communities could not have the concepts of wealth or poverty. Around two thousand years ago, the community speaking proto-Njila led increasingly sedentary lives and produced a wider range of manufactured goods. According to Vansina, this ‘fostered the development of new conceptions of property and the limits of sharing.’ Proto-Njila speakers linked the root *-yéné (‘owner,’ ‘self to the exclusion of others,’ ‘self-same’) to ‘goods, places, or people and stressed that these belonged to this and no other person.’Footnote ²² Further south, where communities began keeping herds of sheep also around two thousand years ago, Vansina made a similar case for a transformation in how people viewed property. If sheep were no longer communally owned, this ‘raised questions as to ownership (community or a leading man?) and inheritance (male or female descendants?).’ When and how this transformation occurred, that is, when and why adult men became the owners, managers, and inheritors of sheep in southern Angola, is a question Vansina side-stepped, writing only that it occurred sometime between two thousand and three hundred years ago.Footnote ²³

Vansina's work prompts further questions: can we develop more granular understandings of when these kinds of fundamental socio-economic changes happened and, if so, how? Can we better understand the meanings attributed to the new statuses that these transformations brought about? There is a strong element of a ‘just-so’ story to the narrative that Vansina offered about the development of notions of personal, alienable, property and men's control over it; his explanation slides towards an endpoint of male economic and political domination. More research on the deeper history of this region would enable us to move beyond a Marxist view of the emergence of classes of rich and poor around two millennia ago and beyond a somewhat evolutionary model of the development of the material conditions necessary for poverty and wealth to emerge. It would also counter the underdevelopment literature that posited ‘African societies merely as victims of asymmetrical exchanges.’Footnote ²⁴ Indeed, recent work by Andrea Felber Seligman on the deep history of the interior trade with the East African coast and Indian Ocean world points to innovation and creativity within African societies in response to trade opportunities.Footnote ²⁵ In particular, we need more contingent histories of economic change, inequality, and poverty that foreground local and regional developments even as we pay attention to broader structural shifts.

Writing this history requires an interdisciplinary approach. Drawing on archaeological and historical evidence, Peter Delius and Stefan Schirmer recently reassessed the Bokoni settlement in Mpumalanga, South Africa, inhabited between the sixteenth and nineteenth centuries. Arguing that it is not possible to understand the economic history of any site outside of its regional context, they also countered simplistic ideas about the relationship between centralised political power and economic development.Footnote ²⁶ The settlement, with its elaborate stone terracing, did not give rise to sharp economic distinctions and the homesteads offer ‘little evidence of hierarchy,’ despite relatively high levels of population density. Instead, recent research at the site highlighted the fact that its inhabitants placed ‘an emphasis on the maximum use of potential agricultural land.’Footnote ²⁷ People in Bokoni were thus able to produce an agricultural surplus that they traded for iron and copper goods, making them a key node in a network of exchange.Footnote ²⁸ Bokoni was, for three centuries, an example of a successful economy that did not result in the impoverishment of some of its members through unequal distribution of wealth, yet was not focused only on subsistence. Predation and violence from neighbours undermined its sustainability and led to Bokoni's eventual abandonment in the early nineteenth century. This predation, argued Delius and Schirmer, introduced ‘a social order built around the logic of extraction and war’ that was premised on socio-economic inequality and political centralisation, in marked contrast to that which had existed in the preceding three centuries.Footnote ²⁹

This new interpretation of the Bokoni settlement shares some features with the model used by archaeologists Susan Keech McIntosh and Roderick McIntosh for Jenné-jeno, Mali, in light of their excavations in the Inland Niger Delta starting in the 1980s. Faced with a very large and complex site, but no evidence of centralised political power, McIntosh and McIntosh developed the concept of heterarchy to explain this ancient West African city. At Jenné-jeno, agricultural or other subsistence intensification was not the response to a rapid population growth between two thousand and twelve hundred years ago. Instead, its inhabitants focused on localised specialisation that enabled them to take best advantage of microclimates, but which was also diverse enough to insure against climatic variation. Their integration into wider trade networks was, again, an essential element in economic stability.Footnote ³⁰ By the ninth century, people living in Jenné-jeno were apparently prosperous and acquiring pottery decorated with ‘geometric white paint on red slip.’ Pottery consumption as a practice was widely shared as ‘these “fugitive-paint” wares are everywhere on the mound, not just concentrated in the homes of the elite.’Footnote ³¹ In contrast to Vansina's argument about West-Central Africa, then, the evidence from the Inland Niger Delta suggests that poverty is not a necessary corollary of personal possessions.

Vansina shed light on the emergence of poverty in a particular context two thousand years ago, and Delius and Schirmer and McIntosh and McIntosh offered counter narratives to that of a continual march towards ever greater economic inequality. David Schoenbrun has instead examined the rise of social stratification, a development that presupposes unequal economic distribution. Rather than pastoralism or specialised metallurgy, Schoenbrun identified surplus production as a key determinant of poverty and social stratification: ‘Surplus allows accumulation and accumulation invites both inequality and followers.’ By the sixth century, people living in the Great Lakes region of East Africa were able to support the work of specialist smelters and smiths through the production of agricultural surplus.Footnote ³² Furthermore, individual wealth appears to have been marked at the time through burial practices.Footnote ³³ Unlike the situation in Bokoni a millennium later, the development of an agricultural surplus led here to the early emergence of social stratification. Poor people, named by those speaking proto-Great Lakes Bantu with the root *-keni, survived by becoming part of wealthier families or households.Footnote ³⁴ We might surmise, therefore, that while inequalities existed, most people could at least survive through relationships of patronage and clientship and through informal networks of fostering.Footnote ³⁵

Schoenbrun argued that in the societies of the Kivu Rift Valley, around the twelfth century, the developments set out above intensified.Footnote ³⁶ This intensification of inequality, in his view, was based on the development of a stable and productive economic base, in this case banana cultivation and cattle pastoralism. Concomitant with this stratification, as well as with a new attachment to cattle herds and to lands especially suited to perennial banana gardens, was the creation of ‘new social categories of lower-status identities.’ In particular, there was a marking of those who were ‘dependents’ in a household, those who had joined because of their poverty, and who retained a lower standing. This ‘heralded the appearance of a new social dynamic of inequality,’ one that was increasingly embedded in multiple ties of patronage and clientship, and which was increasingly difficult, although not impossible, to transcend.Footnote ³⁷ These were also societies that were centralising political power into chiefdoms and monarchies in ways that more often than not served to accentuate economic and social inequalities.Footnote ³⁸ There are, therefore, a number of ways of approaching the study of poverty in Africa over the longue durée.

I approach the history of poverty from a highly qualitative perspective, asking how people understood poverty (and its corollary, wealth) in their own societies. My starting point is explicitly not externally derived definitions, such as those offered by the World Bank or by quantitative studies seeking to use absolute measures. Instead, this research asks how people defined poverty for themselves and how those definitions and conceptualisations changed over time. The area that today straddles the border between Uganda and Kenya has long been a site of linguistic, political, and economic diversity. Over the past thousand years and more, this region has been home to people speaking multiple, distinct languages, both Bantu and Nilotic in origin. They have engaged in a range of subsistence and other economic practices and participated in regional trade networks.Footnote ³⁹ This deep diversity makes the study of how people living here conceived of poverty especially compelling.

Evidence from historical linguistics allows us to see the long-standing diversity in how people in proximate, but different, communities understood poverty and the place of those who were poor in their societies. Three brief examples illustrate this diversity. In particular, they reinforce the importance of not romanticising economic difference in precapitalist contexts. One concept of poverty found in the region is encapsulated by the root *-can- that can be reconstructed to proto-Nilotic around three thousand years ago.Footnote ⁴⁰ Proto-Nilotic speakers used it to mean ‘being poor,’ ‘poverty,’ and ‘poor person.’ But in using this particular root, they emphasised the suffering that went along with poverty and drew direct parallels with other forms of suffering, whether from disease or bereavement. Around a thousand years ago, people speaking proto-Ateker (descended from proto-Eastern Nilotic) used the same root to talk about disturbing and mistreating others. A different conceptualisation of poverty emerged among people speaking proto-Greater Luhyia who settled in the southern part of this area around two thousand years ago.Footnote ⁴¹ They emphasised poverty as lack and, with the root *-tamb- in particular, they drew a conceptual connection between poverty and bereavement, but not with suffering in general. Around fifteen hundred years ago, the community that spoke proto-North Luhyia began to develop highly negative conceptions of the poor as selfish, avaricious, and mean, which they expressed through the root *-tak-. The fact that it is possible to identify radical changes and long-term continuities in the ideas people held about poverty underscores the potential for more qualitative historical studies that reach well beyond the nineteenth century.

This approach to studying the history of poverty requires a deep time-perspective to be able to identify moments of change as languages diverged from each other. The methods of historical linguistics as applied to oral societies necessarily work at the level of centuries not decades, let alone years.Footnote ⁴² But they do enable us to see some of the ways in which people understood socio-economic difference and, crucially, enable us to view that understanding as dynamic and as part of broader processes of historical change. We can, therefore, comprehensively move beyond simple and static generalisations about poverty in precolonial Africa. The kind of history we can write by drawing on historical linguistics and archaeology, alongside comparative ethnography and oral traditions, is different to the history of the African poor in John Iliffe's seminal overview. Furthermore, conceptual histories of poverty drawing on such sources offer a strong counter to Iliffe's pessimistic view that histories of poverty and of the poor in Africa can only be the subject of serious study ‘where written sources survive’.Footnote ⁴³

A willingness to embrace the whole range of methodological approaches and kinds of evidence available to Africanist historians is of course part of this. But of greater interest to me here is that we can write complex, qualitative histories of poverty that focus on dynamics internal to a given society, or even region, while still recognising the effects of external changes and pressures. An example of this is Jeremy Prestholdt's work on nineteenth-century Mombasa that takes a synchronic approach to attitudes towards the poor and towards poverty in the city. To do so, he drew on two particularly rich sources: the works of Muyaka bin Haji, a nineteenth century poet who wrote extensively and critically about social attitudes and norms, and Johann Krapf's dictionary of the Mombasa dialect of Kiswahili published in 1882. Through these, Prestholdt presented an intricate conceptualisation of poverty as ‘a lack of respectability and its signifying objects.’ He noted that for the residents of Mombasa, ‘becoming impoverished was synonymous with decay, dying away, wasting.’ These ideas, according to Prestholdt, could be summarised as ‘to be poor was to want.’Footnote ⁴⁴ And ‘want’ meant, in nineteenth-century Mombasa, both ‘to lack’ and ‘to desire,’ with poverty being the condition of being without and of desiring clothes and possessions that marked a person as respectable.Footnote ⁴⁵ But there was also an understanding of the symbiosis of wealth and poverty: wealth could be used to acquire clients, but the patron had to be careful lest they became one of the ‘many Suahili who were once wealthy people, but who lost all their riches by aspiring after greatness.’Footnote ⁴⁶ In Domesticating the World, we get a detailed and nuanced picture of Mombasans’ notions of poverty and wealth. Historians of twentieth century Mombasa and the Swahili coast more broadly have a rich basis on which to build a diachronic study of poverty as it was understood locally and inflected by regional trends. Because it is a snapshot of a particular moment in time, however, it does not offer us insight into their ideas about economic difference before the nineteenth century or transformations in these ideas.Footnote ⁴⁷