In African history, emancipation from slavery and migration are usually seen as closely related processes that together stimulated the growth of a free, mobile labour market within an expanding colonial economy.Footnote 1 In precolonial times, slaves either had no freedom of movement or such movement depended on the will of their masters.Footnote 2 With the introduction of colonial anti-slavery legislation at the end of the nineteenth and the beginning of the twentieth century, slavery gradually ceased to be a legal institution, and slaves gained firmer control over their mobility. Many men and women ran away from their former masters; some returned home, whereas others headed to more attractive rural and urban destinations to start a new life.Footnote 3 Still others never fled, but because they did not own land and were now responsible for all fiscal and marriage expenses in their newly established households, they were hard-pressed to earn cash and to improve their situation.Footnote 4 They too began to migrate to cities and plantations.
Notwithstanding a general rise in labour mobility, however, freedom and opportunity of movement should not be mistaken for actual movement: the possibility of moving is one thing, the ability to do so is another. As Igor Kopytoff and Suzanne Miers warned long ago, a liberal view of slavery, centred on freedom, does not allow us to see it as being part of a more complex social system.Footnote 5 As they argued, if slavery is an institutionalisation of marginality, then acquiring a capacity to move was more complex than the mere attainment of a legal right to move. In many parts of West Africa, former slaves have continued to suffer from overt as well as subtle forms of discrimination after the formal end of slavery.Footnote 6 There are reasons to believe, therefore, that the less visible effects of slavery on migration dynamics have not ceased with its legal abolition, and that the transition from slavery to free labour mobility was, in fact, more problematic than previous scholars have suggested.
Focusing on post-emancipation male emigration in Sabi, a Soninke or Serahuli village of the Upper River region of the Gambia, this article shows that people of slave descent have suffered from relative immobility compared to the freeborn. To be more precise, the article presents a case of ‘involuntary immobility’, whereby migration is desirable, but inhibited by various factors.Footnote 7 Economic emigration, a predominantly male domain, had been relatively high in Sabi throughout the early twentieth century. In the 1950s, the volume and intensity of migration increased, with villagers leaving for a number of new destinations in West and Central Africa. This was immediately after the period of emancipation from slavery – formally ended by the British in 1930, but effectively taking place in the two subsequent decades. By then, Sabi freeborn had lost control over the mobility of slaves, and the latter migrated alongside the former. However, quantitative and qualitative evidence suggests that men of slave descent have emigrated less than their freeborn counterparts.
My argument is that the persistence of social liabilities linked to slave descent after emancipation has partially prevented slave descendants from accumulating the resources needed to out-migrate. Firstly, when international migration intensified in the 1950–60s, households of slave descent had only recently acquired autonomy. These households could devote time to farming their own plots of land and thus sponsor the migration of their younger men, but they often remained stuck at the subsistence level and/or could not afford to relinquish an indispensable agricultural workforce. A second and related point is that although slavery was formally over, being identified as a person of slave descent meant being relegated to the margins of society. After abolition, slaves were included as an endogamous status group into the existing tripartite social hierarchy of the village – hooro (freeborn nobles), nyaxamalo (freeborn casted artisans), and slave descendants.Footnote 8 The last group are still anachronistically referred to as ‘slaves’ (komo, sing. kome). Prevented from weaving kinship relations with the freeborn by the rule of group endogamy, the komo could only create weak ties with the long-standing migrant networks of freeborn groups, and thus took time to develop networks of their own.
In other words, although migration in Sabi has on the whole increased since emancipation, it has been spread unevenly across the social structure. This underscores François Manchuelle's thesis of a ‘downward social spread of migration’ among the Soninke of the Haut-Sénégal (Upper Senegal) during the twentieth century.Footnote 9 Jean Schmitz has described an analogous phenomenon in Futa Tooro (in the Middle Senegal valley), where enduring dependency on former masters delayed former slaves’ entry into the migratory circuits that developed in the area from the 1930s onwards, first toward urban centres, and then abroad.Footnote 10 In passing, Schmitz calls this delay in migration a ‘rank effect’, an expression which he unfortunately does not qualify further, though it is one I nonetheless find evocative and analytically suited to describing migration dynamics in Sabi.
In what follows, I use ‘rank’ as a synonym for ‘social status’ in a hierarchical order, but also especially as a transitive verb meaning specifically, ‘to classify with others of a similar type’.Footnote 11 As both Josiah Heyman and Michael Kearny have argued, with respect to a different object of study, classification involves two interrelated processes: assigning categorical statuses to subjects and positioning them within a socioeconomic field.Footnote 12 Heyman and Kearney are interested in the implications of political-legal classification for economic class formation in the US-Mexico borderland. Given that the social structure of Sabi has primarily been based on status rather than on class, to avoid confusion the term ‘ranking’ is preferred to that of ‘classification’. In addition, I am not concerned with categories employed by state officials but with societal practices of identification and inclusion/exclusion.Footnote 13 I thus suggest that the relative immobility of slave descendants in Sabi can be described as an effect of rank insofar as the subaltern integration of people identified as komo has, on average, slowed down their entry into circuits of long-distance migration.Footnote 14
The notion of rank effect places emphasis on the process of becoming a person of slave origins during the inter- and post-emancipation period, rather than on static structures and fixed status identities. In Sabi, a number of people became komo without being enslaved. From the late nineteenth century through to the first half of the twentieth, a large number of immigrants settled in Gambian Soninke villages. Some of the newcomers were only allowed to marry slave descendants because the freeborn thought of them as strangers who, like slaves, lacked certified social origins and attributes associated with freeborn status. Through marriage, such strangers became members of the group of slave descendants. Their children inherited slave descent from their mothers and with it social obligations to their former masters. What is interesting about these immigrants and their descendants is that neither group was ever without freedom of movement; rather, for many, though by no means all, being ranked as komo affected their socioeconomic position and in the long run caught them in varying degrees of immobility vis-à-vis international emigration.
By linking immigration to emigration across generations of travellers, this article aims to contribute to the broader debate on mobility and emancipation in Senegambia and Western Sudan. As a core region in the peanut-growing basin, the Upper River region of the Gambia was a major migrant destination for slaves who gained freedom during the period of colonial abolition. Manchuelle has argued that emancipation in the Upper Senegal (1905) turned many slaves into seasonal migrant farmers to the Gambia and Senegal, who the British called ‘strange farmers’.Footnote 15 Marie Rodet has added that departures from the Upper Senegal were often permanent in the early twentieth century, and that they included women and entire families, too.Footnote 16 Finally, Kenneth Swindell and Alieu Jeng have traced similar movements in the Gambia valley, where many slaves became internal agricultural migrants after abolition.Footnote 17 While this shows that former slaves took advantage of new opportunities to migrate, I argue that many of these mobile ex-slaves were precisely the immigrants who Sabi elites ranked as komo, and whose descendants suffered from immobility at a later stage. The bitter irony of such biographies constitutes a further and compelling reason for revising the teleology and temporality of the emancipation-mobility nexus still prevailing in African history.Footnote 18
This article is based on archival and oral sources collected during 14 months (2006–8) of field research in the Gambia. After 1930, slaves, and to some extent emigrants of all statuses, make only a cursory appearance in the colonial records.Footnote 19 Moreover, in contrast to the Soninke of the Upper Senegal, very few studies have been published on the Soninke of the Gambia. Oral sources are therefore a necessary complement to written documents for the kind of micro-scale, trans-generational analysis attempted here. Whereas such evidence may pose limits on generalising my arguments, I hope to show that this kind of analysis affords insights into histories that were seldom recorded and might otherwise soon slip into oblivion.
The relative immobility of slave descendants in sabi (1902–2008)
Sabi is squarely located within the history of migration of the Upper Gambia and Senegal river valleys, two historically interconnected areas mainly populated by Fulani as well as by various speakers of Mande languages, most notably the Soninke. This has been a space of multiple human, commercial, and cultural mobilities propelled by diverse forces and practised differently by people of both genders and different sociocultural backgrounds. I restrict my focus to patterns of long-distance economic emigration among Soninke men because of the prominence these have had in the region in numerical and socioeconomic terms (see below). In the eyes of most Sabinko, male emigration is the main avenue to individual and, most importantly, collective prosperity. For the same reason, male emigration has been of utmost importance to slave descendants’ strategies of emancipation.
The history of migration in Sabi can be divided into three phases.Footnote 20 The first phase ranges from the late-nineteenth to the mid-twentieth century and is characterised by immigration and by some regional trade and labour migration. Sabi itself was a product of immigration in the Gambia valley. It was founded in 1902 by some families who came from Sabi Xase (literally Old Sabi), a village in the Jeriko river valley in the precolonial Kingdom of Bundu (Eastern Senegal).Footnote 21 Philip Curtin described Sabi Xase and other Soninke villages in Bundu as ‘clerical-commercial communities’ situated along the main trading route connecting the Senegal and Gambia valleys, and linked to well-known Muslim merchant networks (the juula).Footnote 22 In the second half of the nineteenth century, the Sabinko fled political turmoil and settled on the north bank of the Gambia valley. They then crossed the river in search of cultivable land, and founded Sabi. In their new location, the Sabinko recreated their commercial-clerical community under Fulani rule, continuing to trade along regional routes to Mali (gold, cloth), Futa Jallon (cloth, cattle), and Guinea-Bissau (gunpowder, tobacco). They also eagerly invested in the rapidly expanding commercial cultivation of peanuts.
Immigration continued to prevail in the first half of the twentieth century. In 1932, Sabi's population was estimated at 908 inhabitants, making it the third largest village of the Fuladu East district.Footnote 23 Since Sabi had been founded a generation earlier by only a handful of families, this impressive growth was clearly due to new arrivals. Except for the kinsmen and fellow villagers coming from the east, the new settlers were a heterogeneous group, including war refugees, mercenaries, traded slaves as well as slave traders, French subjects crossing the border to Sabi to avoid conscription during World War One, skilled labourers (typically casted artisans) as well as Islamic scholars, and above all seasonal migrant farmers.Footnote 24 During the first two decades of the twentieth century, the Upper River attracted a sizeable portion of the people who left the Upper Senegal and the Western Sudan in search of better opportunities.Footnote 25
After the Second World War, emigration rather than immigration gathered momentum in Sabi. In this second phase, Soninke trading networks rapidly expanded and diversified. By this time, the Soninke-speaking communities of the Gambia had become autonomous from the eastern ones. Nevertheless, there remained linkages and similarities between these two areas. In the mid-1950s, many young men and established traders from the Upper Gambia and the Upper Senegal rushed to Sierra Leone to search for diamonds.Footnote 26 As the commissioner of the Upper River noted at the time, many such ‘prodigals’ returned after the 1956 ‘stranger drive’ in the mining areas.Footnote 27 But many found their way to Sierra Leone again, or went to other West and Central African countries. The British did not attempt to stop emigrants; in this period, they were more preoccupied with bolstering agriculture. Indeed, after the Second World War, and throughout the diamond rush (except for 1954), agricultural production remained buoyant thanks to regular rainfall, favourable marketing conditions, and crucial technological improvements.Footnote 28 This also indicates that emigration was driven by opportunity, not by economic crisis.
There were also differences between the Upper River Gambia and the Upper Senegal. Manchuelle has shown that Soninke migrants from communities along the Senegal river came to dominate naval occupations in the late nineteenth century.Footnote 29 This led them to live and work in coastal cities, where they were hired by commercial ships going to France, and eventually paved the way for the well-known emigration of Soninke to France, starting in the 1950s–60s.Footnote 30 In contrast to the Upper Senegal Soninke, among the Gambian Soninke labour migration linked to riverine and coastal cities remained weak. Throughout the first and second phase of migration, some young men left as seasonal migrant farmers for Senegal and other parts of the Gambia, but short- and medium-distance trade migration linked to the peanut economy remained the most popular form of mobility prior, or in parallel, to undertaking long-distance emigration to West and Central Africa.
The severe drought of 1968 can be taken as the beginning of the third and ongoing phase of migration. In this phase, which concerns us only marginally here, international migration continued to intensify and diversify, while simultaneously becoming costly and selective. Soninke regional trading networks shrank, and migrant diasporas developed in other commercial hubs in West and Central Africa. In the 1970s, trade migrants began to leave for Europe and North America, opening the route for later generations of (labour) migrants. Since the 1980s, Euro-American destinations have polarised migratory flows, as in the rest of Senegambia. In parallel, the rural economy has progressively declined under the combined effect of environmental degradation, cessation of state subsidies, and shortages of labour.
Migration is not evenly distributed in Sabi. Many villagers, including some komo, hold that slave descendants have been less active and successful as migrants than their freeborn counterparts. This widespread opinion is seemingly corroborated by statistical data. In 2006–7, I conducted a basic survey across Sabi households in order to estimate the socioeconomic incidence of migration. In a sample of twenty households (about 10·9 per cent of the village total), ten were komo.Footnote 31 Although the sample is limited and the figures should be approached with caution, the survey gives us an indication of slave descendents’ relative immobility.Footnote 32 Approximately 15 per cent of the male active population (approximately aged 16–55) in komo households were abroad, as opposed to 45 per cent of freeborn (hoore and nyaxamala) ones.Footnote 33 A study of Koina, a Gambian Soninke village 50 kms east of Sabi, carried out in the 1970s, further suggests that the relative immobility of the slave descendants, and possibly that of other marginal classes, is far from being new. In Koina, 65 per cent of the hooro generated income from migrant occupations as opposed to 53 per cent of the nyaxamalo and komo. Footnote 34
It is regrettable that although many studies of Soninke and regional migration exist, post-emancipation emigration of slave descendants is rarely discussed in detail. Where this is done, it is apparent that, besides becoming on the whole more mobile, slave descendants have taken to migrating later and in smaller numbers compared to the freeborn. In Gidimaxa (the Mauritanian side of the Upper Senegal), there were pitched battles between Soninke slaves and masters, and slaves were not allowed to migrate until after 1960.Footnote 35 In other parts where freedom of movement was achieved much earlier, many slave descendants nevertheless had difficulties in accessing routes and networks. The Soninke elites were long acquainted with long-distance trade migration and thus monopolised the aforementioned labour markets emerging along the Senegal river.Footnote 36 This probably gave them a head-start over slave descendants when the time came to leap into the international circuits of labour mobility. Although migrants from all status groups began to flock to France in the 1960s, evidence to be discussed in a later section shows that in 1990s France, komo migrants from the Upper Senegal were still fewer than freeborn ones.
Albeit fragmentary, the evidence from other regional locations suggests that the relative immobility of slave descendants in Sabi is neither an isolated nor a recent phenomenon. A host of other social factors might have affected the emigration of slave descendants in Sabi, producing the quantitative and qualitative data that I gathered in 2007–8. At the same time, since there have been regional variations in dynamics and temporalities of emancipation and migration, the reasons for the relative immobility of slave descendants require a closer and historically grounded investigation of the ways in which slave descendants have approached migration in Sabi.
free to move: the end of slavery
In contrast to places like Gidimaxa, Sabi slave descendants faced few formal constraints on their emigration during and after abolition. The village elders have generally little to say about any resistance to the emigration of former slaves. A historical overview of emancipation will further corroborate their statements: while the elites remained in control of the social order, by the mid-twentieth century formal constraints on mobility had been lifted.
In the Gambia, abolition began in 1894 with an ordinance (later reiterated in 1906) that banned the internal slave trade and provided domestic slaves with the possibility of buying their freedom. As Alice Bellagamba has shown, the British feared disruptions to the production of peanuts, which relied to a large extent on the exploitation of slaves. Policy implementation was thus delayed until after the First World War, progressively leading to the 1930 Ordinance, which outlawed all forms of slavery.Footnote 37 Bellagamba has also argued that in the relatively remote Upper River (annexed to the British protectorate only in 1901), the ruling elites were able to mitigate the destabilising effects of legal emancipation, and to exert pressure and coercion on former slaves to remain within their sphere of influence.Footnote 38 In the 1930s, there were still a number of people in bondage in this region, even though many slaves had already achieved considerable autonomy.Footnote 39
Economic change catalysed social emancipation. Peter Weil has shown that in the kingdom of Wuli, on the north bank of the Upper River, labour obligations on slaves were being reduced at the end of the nineteenth century.Footnote 40 Seasonal migrant farmers provided commercial producers with a flexible workforce, which, unlike slaves, could be deployed during labour bottlenecks and released during slack periods.Footnote 41 Like second-generation slaves, these migrants worked four to five mornings per week (then reduced to three, or even two), in exchange for food, shelter, and land. Slaves were similar, though not equal, to free labourers who worked for landowners/masters in order to access land to cultivate.Footnote 42 Swindell and Jeng have thus concluded that the transition from slavery to free labour was key to both peanut cultivation and emancipation.Footnote 43
Sabi was no exception to these regional trends. While abolition prompted some flight by first generation slaves, who were unmarried and knew how to return to their home places, Sabi received slaves from other villagers who sought better social conditions.Footnote 44 These included a group of komo who migrated from Gambisara, and are still known as Gambisara komo. Occasional unrest and permanent departures were also recorded. However, rather than openly contesting their status or fleeing, most slaves aimed at improving their situation: they acquired land to farm and established independent family compounds.Footnote 45 Their workload was progressively reduced to one or two days a week, or to some days during the agricultural season. In the 1950s, in Sabi some komo still worked in the fields of their former masters on Saturday and Sunday mornings, though it is difficult to estimate the extent of this practice.Footnote 46 I suspect that only wealthy patrons were still able to exert influence on their former slaves. In the meantime, Sabi's wealthy households also relied on the labour of seasonal migrant farmers.Footnote 47 Inflows of such farmers fluctuated according to the price of peanuts, the price of rice (used to feed guest labourers), environmental, and political conditions. On the whole, in Fuladu East, the estimated number of ‘strange farmers’ sharply increased in the first two decades of the century, peaking at 4,278 in 1916, and settling between 2,800 and 3,300 in the 1920s. Despite declining prices and marketing conditions in the 1930s, production continued at high levels and the inflow of seasonal migrant farmers remained roughly the same. There were, however, dramatic variations in this period, and arrivals fell sharply during the Second World War.Footnote 48
The overall result of abolition was not outright emancipation but ‘manumission’, according to Claude Meillassoux's definition: the near eradication of the master's control over production and reproduction, but with the maintenance of the state of kome and some ties of dependence vis-à-vis the former master.Footnote 49 As discussed earlier, komo became an endogamous status group placed at the bottom or margin of the social hierarchy.Footnote 50 Nevertheless, by the 1950s, most slaves were free persons, and former masters replaced slaves with seasonal migrant farmers. Consequently, legal constraints on the mobility of former slaves were few. The struggle over mobility sometimes took place on other grounds, especially on financial ones. In Jafunu (Mali), Soninke slave migrants were expected to remit part of their gains to their masters, and in some cases this practice continued until the 1960s.Footnote 51 In Sabi, some former slaves shared their earnings with their masters, but this practice was soon discontinued except as an initiative of the successful kome migrant wishing to settle the socio-moral debt of bondage with his former master: occasionally komo still pay their former masters – usually a symbolic payment consisting mainly of kola nuts – to free themselves from any social and moral debt.
The immigrants integrated as slave descendants
The absence of jural-political barriers to mobility becomes especially evident when the group of slave descendants is disaggregated. Sabi's slave descendants do not share a common history. Some komo arrived with their masters when the village was founded, or soon afterwards. Others arrived as traded slaves. A large portion of Sabi's komo originate, however, from the immigrants who arrived during the period of abolition. Although, as we saw, these immigrants helped to accelerate processes of emancipation in the first half of the twentieth century, they remained in a grey zone as regards status. This was true to the extent that the Sabinko integrated some of the migrants settling in their village as komo. Since these migrants were free to come and go, understanding how they were ranked as ‘slaves’ will provide us with an opportunity to examine the social mechanisms that, as in other regional locations, affected the mobility of slave descendants after legal abolition.
Most forms of mobility in this region are founded on hospitality.Footnote 52 Host-guest relations were also crucial for immigrants who wanted to settle down and integrate. Although Marie Rodet has argued that some settlers in the peanut basin came with, or were later joined by, their spouses and children, it appears that the majority of the migrants were young men who came on their own.Footnote 53 Like most other Sabinko, immigrants did not maintain relations with their places of origin and their genealogies mostly consist of marriages with local people (see below). Not only was marriage an expensive venture which lone migrants could seldom afford, but it also implied lengthy negotiations between the two extended families. Soninke freeborn men tended and still tend to marry mostly women within their own family circles, entering into external marriage relations only with prestigious families. Since many migrants came on their own, there were no kinsmen to assist them. If either they or their children wanted to marry a local woman, they were subjected to close inspection. Unless they could demonstrate family links in the village or bore clear signs of high status, such as cultivated manners and an Islamic education, they were considered as outsiders, and hence potentially slaves.Footnote 54 The fact that numerous former slaves were among the migrants hailing from French territories probably exacerbated the villagers’ concern about social origins. As one slave descendant put it: ‘if they [the villagers] don't know where you come from [socially and geographically], they will give you a slave woman’.Footnote 55
Some of today's slave descendants protest that ‘my father was not a slave, we were nobles!’Footnote 56 Whether such claims to noble origins are factual or whether they are meant to hide more dramatic and shameful stories of enslavement is hard to ascertain. Interestingly, many freeborn do not object to these claims; rather, they cynically rejoin that: ‘yes, [that slave descendant's] grandfather was a noble, but then he married a slave woman…’.Footnote 57
Restricting immigrants to the lower rungs of the social hierarchy helped the ruling elite to hold on to power under rapidly changing demographic, economic, and sociopolitical conditions. Furthermore, by supporting seasonal migrant farmers, some landlords were able to enlarge their followings.Footnote 58 Since slave descent is matrilineal, the children of immigrants who married slave women became komo. It is unlikely that either the immigrants or their children were integrated as slaves proper; they were probably closer to the condition of manumitted komo or second generation slaves during the emancipation process. Nevertheless, the Sabinko did not make any distinction between real and classificatory komo. On a par with ‘real’ komo, migrants’ descendants were usually bound to their mothers’ masters, carried out some work for them, and later retained customary ties with them.Footnote 59 More rarely, immigrants were classified as komo but did not become connected to any host/master.
In the absence of unequivocal documentary evidence, it is difficult to know the reasons why some migrants decided to settle down and marry spouses of slave descent. It is likely that among the immigrants there were people who could not return, such as runaway slaves, outcasts, and refugees fleeing wars in the late nineteenth century and probably reaching the Gambia basin after years of itinerant travel.Footnote 60 During the first two decades of the twentieth century, abolition and worsening ecological and economic conditions continued to drive many people from Upper Senegal to the peanut basin.Footnote 61 Other migrants could, however, go home. The oral accounts I have collected reveal that it was sometimes opportunities and sometimes misfortunes, deception, or other unforeseen circumstances, which induced them to stay. More than the reasons for settling, what matters to us is to understand the implications of ranking for the emigration of the settlers’ descendants. Given that these people were free to move, how did slave status affect, if at all, their mobility?
From immigration to emigration: two kome households
Rather than trying to account for all possible trajectories of migration and rank, I shall focus on two households formed by two men who arrived in Sabi during the emancipation period. In addition to describing their dynamics of immigration, I will follow their trajectories through to subsequent generations, drawing attention to the way mobility has been shaped by persistent structural socioeconomic inequalities. I have chosen two very different stories in order to show the diversity and gradation of the effects of ranking. Although this type of classification strongly correlates with relative immobility, it should not be seen in overly deterministic terms.
The first story begins at the turn of the twentieth century, when Tamba Samagare left Gidimaxa and travelled to Sabi to work as a seasonal farmer.Footnote 62 He was said to be a hard-working farmer and he made the most of the burgeoning peanut sector in the Upper River region during the first quarter of the century. There was land available on the southern bank, and instead of returning home, Tamba stayed in Sabi. He managed to amass enough capital to start trading with other Sabinko and traders along the route to Futa Jallon (in north-east Guinea Conakry). He invested in cattle, a growing niche market along the Gambia river.Footnote 63 As Tamba had become fairly wealthy, he could afford to pay bridewealth. He wanted to marry a noble woman, but the Sabi nobles refused. He then married the daughter of another newcomer (who had married a kome woman as a second wife). He and his wife moved to the outskirts of Basse, about eight miles from Sabi, where he cleared land and continued to farm and trade. He apparently employed seasonal migrant farmers, too.
By the 1920s, Basse became a bustling place, hosting European companies and some Lebanese traders.Footnote 64 Although Tamba never became a prominent figure in the region, he had close contacts with this business community.Footnote 65 From around 1920 to the 1930s, he reportedly possessed large herds, and gave presents to some of his contacts. Oral accounts do not say whether he suffered from any misfortune, though severe cattle epidemics such as those in 1917, 1921, and 1932 may well have damaged Tamba's wealth.Footnote 66 Most probably, contacts with the Basse traders provided some social and economic security, and this, in turn, enabled his family to navigate through the vagaries of farming and cattle breeding.
Through his trading expeditions, Tamba maintained strong relations with Sabi. He was especially close to another wealthy newcomer (also classified as kome). When Tamba died in the mid-1940s, his friend helped the Samagare family move to Sabi, and the two families established marriage links thereafter. In Sabi, Tamba's five sons continued their father's activities. The second generation was engaged in agriculture, and at least one Samagare man used to go to Guinea Bissau to trade cloth. After the Second World War, production expanded in the Upper River (see above); yet the colonial administration began to centralise the marketing structure, thereby undermining an already shrinking indigenous trade.Footnote 67 When ‘diamond fever’ reached Sabi in the mid-to-late 1950s, one of the Samagare men left for Sierra Leone, followed shortly by two other third-generation adults. They fared well in the diamond trade, and in the 1970s and 1980s they managed to help more household members to migrate to Europe and the United States, or to set them up in business in Sierra Leone. Today, Samagare-kunda is a large household and several of its members are in West-Central Africa, Europe, and the United States.Footnote 68
Ever since Tamba's marriage, the Samagare have married komo, either in Sabi or in nearby villages. On some occasions, relations with the mothers’ former masters have been maintained. The Samagare have performed at ceremonies and received presents from them, or have actively sought their support. Except for such ritual occasions, however, the Samagare have never worked as labourers for them.
The second biography is that of Musa Singate and his descendants.Footnote 69 In the late 1920s, Musa and his brother left eastern Senegal as seasonal migrant farmers and never returned home. Musa settled in Sabi, while his brother settled in a nearby Soninke village. In Sabi, Musa married a slave woman. The reasons for his marital choice are uncertain. His descendants today complain that he was not a slave and he was discriminated against; they say that when his family sent a messenger to look for him, the man was apparently astonished to see Musa married to a slave woman. Musa was so ashamed that he did not go back. The former masters of his wife do not remember the dynamic of Musa's marriage either. It appears that they lodged Musa in their household and helped him to build a compound near the ‘slave quarters’. One of them remarked that many people fell into slavery or dependency because of poverty, and added that seasonal migrants who had borrowed seeds and had to repay their landlords at the end of the season were particularly exposed to crop failure or other calamities.Footnote 70 Peanut cultivation was a volatile business indeed: epidemics, droughts, and market fluctuations affected production, causing chronic levels of indebtedness among farmers.Footnote 71 Cumulative indebtedness and poverty probably continued to induce destitute immigrants to accept a degree of dependency on their host for subsistence and reproduction. Musa's hosts were not the elites of the community, but they were among the first settlers, and had a large family and resources to attract immigrant farmers (and to buy slaves before that).Footnote 72
Musa stayed in Sabi within the sphere of influence of his patrons. Like many other people in the 1930s and 1940s – a period marked by declining prices for produce and depressed trade – he almost certainly had to rely on extra support before his three sons were able to help him out in the fields. Two of his sons came of age in the 1940s-50s, when subsistence and commercial agriculture became favourable again. Yet Musa's family was probably not in a position to take full advantage of economic growth. Relying on a patron and maintaining relations with former masters came at a price; as late as the 1950s, his two boys still worked for their patrons, albeit occasionally.
Musa's youngest son – Ibrahima – came of age in the mid-1960s, and he soon began to nurse an ambition to travel. He spent a dry season labouring in the nearby town of Tambacounda. At the end of the season, having heard of job opportunities in Mali, he decided not to return to Sabi. He stayed in Bamako for a few years, working as a watchman in the compound of a prominent government official. Then he went to Ghana and Nigeria, where other Soninke migrants introduced him to the African art trade. In the early 1980s, Ibrahima managed to help one of his brother's sons to migrate to the United States, and the latter in turn helped another brother. Since the 1990s, remittances have provided a fairly stable income for Singate-kunda. The household is however divided internally, so that not all household members benefit from the remittances. Despite having achieved economic autonomy, the Singate have not interrupted customary ties with former patrons/masters.
Needless to say, Singate-kunda exemplifies a more common pattern of immigration/emigration than Samagare-kunda. Tamba Samagare was one of the most successful farmers and traders in the Sabi area. Although the exchanges he had with his first host in Sabi are not wholly clear, it is certain that he did not have many binding commitments with villagers thereafter. He moved to another village and the wealth he amassed there allowed him to choose a bride without relying on any patron. Tamba could not marry a noble woman, but at least his wife was a stranger with no apparent bonds to local masters. Thus his children did not work for anybody.Footnote 73 Tamba participated in trading networks and this directly benefitted his descendants, who used these contacts in order to emigrate. Musa Singate arrived later, when slave-master relations were weaker, and yet he entered into a binding relationship with his host. He did not have enough time or opportunities to develop an autonomous household. The first migrant in his household was his youngest son, who emigrated only in the mid-1960s, and it was a long time before this son could take up a profitable occupation. Only then was he able to send money to his household and help other household members to migrate.
If the Samagare-kunda case shows that slave descent was not an insurmountable barrier, the Singate-kunda one demonstrates that entering migratory circuits from a marginal position might pose problems. To be more accurate, Singate-kunda suggests that one effect of ranking might be the postponement of this entry (the rank effect). Neither Tamba nor Musa lacked freedom of movement. Rather, how and when their descendants could emigrate partly depended on how and when they were integrated into the kome group.
Ranking people, effecting immobility
One important factor that impinged on the komo's ability to emigrate was the size of their household (ka), specifically the size of the household's active male population. Whereas married women worked in individual fields, and were asked to make limited contributions to collective meals, men worked primarily on the communal field (tee xoore) of the household in order to replenish the granaries with millet. After emancipation, manpower was a problem for many kome households. As a kome elder told me, his father refused him authorisation to leave for Sierra Leone in his youth because ‘There were only the two of us in the fields – my father and me. He said I should stay and help him.’Footnote 74 Seasonal migrant farmers could be employed as a substitute for emigrants, but this was usually done by households with enough foodstuff to feed them.
The compounds of the freeborn, especially those of the nobles, were and still are often more populous.Footnote 75 While in 2006 the number of hoore (noble) compounds (70) and kome compounds (72) was roughly the same in Sabi (the nyaxamalo have 35 compounds), on average noble compounds were larger.Footnote 76 In addition, it should be noted that many noble Soninke moving to the Upper River valley in the late nineteenth century were clerics and created Quranic schools. Quranic students worked for their teachers, thus relieving some of their household members from farm work and enabling the latter to undertake long-distance trade.Footnote 77Komo have been discriminated against in Quranic education, and have rarely developed Quranic schools.
Certainly, some kome households were relatively large when emigration intensified. A limited number of slaves followed their masters to Sabi in the late nineteenth or early twentieth century, and formed separate households from the start. In the 1950s, at least one of these households had a sizeable population, possessed draught animals and implements, and harvested conspicuously large crops.Footnote 78 In contrast, as mentioned, later immigrants who were ranked as slaves rarely came with their families. By truncating family ties with home communities, single migrants were forced to start new (small) households. This also seems to be true of domestic slaves who were granted the possibility of creating their own households outside those of their masters only after legal abolition.
Analysing the composition of the Paris-based expatriates from a Soninke village along the Senegal river, Mahamet Timera has also pointed out that household size and fragmentation is reflected in migration rates.Footnote 79 The village komo have 53 people from 21 households in France, while the freeborn list almost twice as much (99) from only 16 households. Timera has commented that:
We first of all notice that the size of the ‘houses’ (ka) decreases as we follow the order of precedence in power relations … In spite of their numbers (it is the largest status group), [the komo] experience a fragmentation that has, without a doubt, an effect on their capacity of cohesion and collective action. The more recent formation of the ka of the komo might explain their reduced size. These compounds have sometimes existed for four generations only after becoming autonomous vis-à-vis the families upon which they depended some decades ago.Footnote 80
Both Timera's observations and mine lend support to the argument that the size of the active male population in each household is a decisive factor in long-distance migration. Writing about Upper Senegal during the 1970s and 1980s, Sally Findley has argued that migrating internationally (particularly to France) requires considerable resources and long-term planning, something which favours large households.Footnote 81 Those households with many men have a large enough workforce to cope with the loss of a migrant's labour until he is able to send remittances in order to compensate. International migration is a long process, involving the bureaucracy of visa applications, becoming settled at one's destination, and finding work. Short-distance migration is by contrast a more immediate response to economic emergencies but it offers less security and smaller returns in the long run.
Having small households with few links to others increased the vulnerability of komo. As the Singate show, some households were dependent on their ‘masters’ and worked for them on an occasional basis until after the Second World War. This further affected the ability of these komo to wield control over their labour and impinged on their production of surpluses with which they could sponsor emigration within the region, and then to West and Central African destinations. In addition, in many communities of the region, the elites controlled access to key economic assets.Footnote 82 In Sabi as elsewhere, access to land was important for subsistence and trade (sales of peanuts generated merchant capital).Footnote 83 It is true that while slaves did not generally obtain the more fertile land near the swamps, emancipated slaves like Tamba Samagare could clear and own a forested area and farm peanuts on it. As was noted, however, peanut agriculture turned some into successful farmer-traders but many more into indebted peasants. More importantly, clearing new land requires a considerable labour force, that is, a large household. Finally, it appears that the patronage of former masters facilitated access to the land but did not play a substantial role in sponsoring the emigration of slave descendants. Especially in the first period of migration, ‘going on adventure’, as migration was characterised, was somehow considered an ‘extra’, which young men undertook by relying on their own efforts.Footnote 84 The household would help only if subsistence had already been secured.
In other words, poverty can be just as much a propelling factor of emigration as a reason why people are unable to move.Footnote 85 The initial costs of migration tend to be too high for vulnerable people such as ex-slaves, and this can direct them to migratory patterns that are subsistence-oriented and fail to generate long-term social mobility.Footnote 86 In the 1950s and 1960s, small and poor households could temporarily allow their young men to go to neighbouring cities to find work or trade in goods across the border in the dry season; but at the end of the season, they were likely to return to work in their families’ fields. Sometimes, young men like Ibrahima Singate did not return. While Ibrahima may well epitomise the young man who migrated against his father's will, it is significant that he took a long time to reach a profitable migrant destination. He had to stop at several intermediate stations, where he took up minor occupations. Once he had become sufficiently established, he was able to send remittances and help some of his household's younger members to emigrate, regardless of the fact that household production was still deficient.
Samagare-kunda is different. The success story of its founder, and the fact that he managed to exempt his sons from working for any patron or nominal master, enabled him to accumulate resources, and to employ seasonal migrant farmers too. Tamba's sons and grandsons were integrated into trade networks at an early age, and when the opportunity emerged, they were quick to migrate to Sierra Leone and other West African countries, as many of their freeborn counterparts did.Footnote 87
For poorer immigrants and for ‘slaves’, integrating into local networks had some advantages as well. Social capital is a crucial factor in reducing the costs of emigration, especially during the initial stages of migration when resources and connections have to be mobilised to sponsor travel.Footnote 88 During and after emancipation, slave descendants could rely on fellow villagers to partly compensate for the lack of human and economic resources in their small households. However, in Sabi, the strongest networks are based on kinship.Footnote 89 Prior to settling in the Gambia, many freeborn (particularly those with noble status) had wide-ranging commercial and matrimonial alliances across the region which they maintained when they moved into the Gambia. Though they gradually severed matrimonial relations with eastern communities, for some time settlers coming from the east were still recognised as kinsmen, and thus married into an already tightly-knit community.Footnote 90 From the 1950s onwards, freeborn would-be-migrants could thus count on large and often prestigious families allied to each other through marriage. This was not the case for slave descendants. Since status groups are endogamous, they could not forge matrimonial alliances with freeborn villagers, and thus they mostly found themselves amidst unrelated, equally isolated strangers. Consequently, their descendants’ kin networks had limited outreach and were relatively fragmented.
Social networks became more and more important during the third phase of international migration, after the 1960s. In the early stages of this migration boom, entrepreneurial young men like Ibrahima Singate could sometimes raise the necessary money to travel to West African countries. They sold peanuts or their mother's jewellery, or migrated seasonally. With migrants arriving in large numbers at the same destination, however, business niches became saturated. Starting up a business required higher initial investments and, most often, influential relatives who could provide credit. In Sierra Leone, for instance, by the early 1960s shallow mineral deposits that could be easily exploited became scarce, and mining operations shifted to higher terraces, which required larger teams of diggers and expensive machinery.Footnote 91 Most important of all, the emergence of Western Europe and the United States as major destinations meant that migrants needed not simply money to pay for the airfare, but also good contacts in situ to arrange for work contracts and entry visas. Both the Samagare and the Singate cases show that having relatives abroad able to help younger relatives to emigrate was crucial to ensuring the continuity of migrations (and remittances) within households. Nonetheless, for this chain of migration to start, someone had to make the first move. Ex-slaves with small and dependent households, and with weak and fragmented connections, started migrating later and in smaller numbers.
CONCLUSION
Various scholars have argued that the transition from a slave- to a migrant-dependent agricultural system in the Gambia valley – one of the key areas of the colonial economy – marked the beginning of a free labour market.Footnote 92 Indeed, when slavery was abolished, a heterogeneous mass of West African slaves began to leave their masters. They migrated to the Gambia in order to start a new life, to overcome landlessness, or simply to improve their socioeconomic condition back home. In turn, incoming waves of migrants provided commercial farmers with a cheap replacement for slave workers, thus favouring the emancipation of those local slaves who had stayed put until then.
Without discarding such arguments, this article has questioned the evolutionary logic that underpins them. Far from being a linear trajectory, the relationship between legal abolition, social emancipation, and mobility was fraught with contradictions. I have employed the notion of ‘rank effect’ to describe some of these contradictions in the history of Sabi, a village situated in the leading peanut-exporting region of the Gambia. In the first place, by distinguishing classification from formal enslavement, the notion of rank effect sheds light on the paradoxical experiences of those immigrants who settled in Sabi and eventually became ranked as slave descendants. This is a paradox because some of these immigrants were surely the beneficiaries of legal abolition in British and French territories. Yet whether they fled from their masters or chose to earn a living elsewhere, they did not ultimately free themselves of their slave status. Notwithstanding the emergence of free-labour markets and sites of emancipation like cities, post-abolition mobility occurred within a socially structured landscape, where moving without the necessary social capital could result in long-lasting forms of exclusion and subordination.
Moreover, neither freedom of movement nor economic opportunities automatically triggered emigration. How can one account for the fact that slave descendants have on average performed poorly as economic migrants in a milieu like the Soninke one where migration is an accepted means to socioeconomic advancement? A focus on ranking has allowed me to detect the linkages between the ascription of slave status and the socioeconomic liabilities that have prevented slave descendants from pursuing profitable mobility strategies. Even though some of the immigrants who were ranked as ‘slaves’ never lacked freedom of movement, they were relegated to the margins of society, forced to depend on patrons, and obliged to join a status group affected by atomisation. Their households were often recent and small, and struggled to achieve self-sufficiency. When in the 1950s, young men of freeborn descent began to leave for West African countries, those of slave descent were surely willing to go as well. Yet they had to balance their aspirations for mobility against the labour requirements of their households. Young slave descendants had to ensure that granaries in their compounds were full before they left. Furthermore, they could only rely on a few relatives to sponsor their travel and to help them settle quickly at their destination, so that they could send money home to make up for their absence in the fields. Of course, there were success stories in what was (and still is) a very diverse status group. In general, however, in the aftermath of emancipation, slavery's rank effect deferred entry into the business of migration.
