Navigation – Plan du site
Entre servitude et liberté : un presque rien

The Concept of Honour and the Persistence of Servility in the Western Soudan

Martin A. Klein
p. 831-852


La notion d'honneur et la persistance de la servitude au Soudan occidental.
Cet article tente d'expliquer pourquoi certains vestiges de l'esclavage subsistent dans le Soudan occidental (Sénégal, Gambie, Burkina Faso, la plus grande partie du Mali et de la Guinée). Il examine la politique française vis-à-vis de l'esclavage et le départ massif des esclaves de leurs lieux de servitude au début du XXe siècle. Bien que leur changement de statut leur ait alors donné la possibilité de maîtriser leur vie familiale et leur mode de travail, beaucoup sont restés soumis. L'absence d'histoire a créé la base idéologique de leur soumission. Les anciens maîtres sont aussi restés les plus forts dans les régions où ils ont préservé leurs droits sur les terres, mais la clé de leur pouvoir reste la question de l'honneur. Les sociétés hiérarchiques de la région avaient des codes qui exigeaient du noble de la générosité et un certain contrôle de soi. Il était souvent avantageux pour les anciens esclaves d'accepter le système en mendiant et en se comportant de manière vulgaire, mais, en agissant ainsi, ils n'ont fait qu'affirmer leur infériorité. Beaucoup ont lutté contre cette position d'infériorité en refusant de mendier, en adoptant une conduite noble, où de manière plus importante, en se donnant une légitimité en tant que musulmans.

Haut de page

Texte intégral

  • 1 A new native law code was proclaimed on 10 November 1903, Archives de la République du Sénégal, M 7 (...)

1Over a century ago, in 1903, administrators in French West Africa were told that they could no longer take slave status into consideration in legal matters1.   They could neither return a runaway to his or her master nor consider slave status in cases of inheritance or marriage.   In 1905, transactions in persons were banned.   This meant that slaves could no longer be sold, exchanged, given as gifts or bequeathed.   Enforcement of these two acts was often uneven.   They were totally ignored in Saharan circles, but in most of French West Africa, they ended slavery as a legally enforced system of compulsory labour (Klein 1998: ch. 8).   With the approval of the French administration, masses of slaves simply left their masters, perhaps a million in all of French West Africa.   Most of them walked home, but others stopped in towns or formed new settlements.   Many experienced great hardship.   They lacked food and survived either with what little food they could carry or by scavenging.   Massive as the exodus was, most slaves remained with their masters, though relationships changed, quickly in some places, gradually in others.   The more harshly treated trade slaves were largely gone.   The greatest losses were often in plantation areas (Klein 1998: ch. 10; Roberts & Klein 1981).

  • 2 See MEILLASSOUX (1986: ch. 6) for a description of the slaves' stages of integration: from esclaves (...)
  • 3 Report from Dubreka, Archives de la République du Sénégal, 7 G 67.   In Fuladu in the upper Gambia, (...)

2Those who remained were mostly "born in the house", and thus already enjoyed a somewhat ameliorated status2.   Most masters recognized that they could no longer push their slaves as hard as they once did.   Slaves generally acquired control over their own work lives and family relations.   Particularly important was the right to marry whom and when they wanted.   If masters had any doubts about their control over their slaves, intermittent acts of resistance brought home the fragility of their control.   Thus, there were large-scale departures when the tirailleurs returned from the First World War. The force that went to France was about 75% slaves (Echenberg 1991). Many were not willing to return to a servile relationship and were often assertive enough to lead friends and family out of master-controlled settlements to new locations (Clark 1990: 230-233, 1994: 66-67; Manchuelle 1989: 101-102; Klein 1998: 216-219).   "They have acquired under the colours", one administrator wrote, "a rather extensive notion of their rights..."3. Even where slaves continued to accept serious obligations, they were free to seek work elsewhere and to leave home in the quest for respect.   Whether they left, merely moved out of the household or remained under a master's tutelage, freed slaves everywhere distinguished themselves by their willingness to work hard.

3In spite of the erosion of slavery, slave status remained very important in much of the western Soudan.   Former slaves played a subservient role in ritual life, for example, cooking on festive occasions or repairing their master's houses.   In some cases, they continued to bear responsibility of onerous jobs like cleaning the latrines.   Many were expected to give gifts like the assaka, a tax which was based on the Muslim obligation to give charity, but had somehow been converted into an obligation of the slave to his master.   Islam was an important battleground.   Slaves often sought an Islamic education.   It seems probable that many of the slave masters most hurt by the post-1905 exodus were able to use their Quranic knowledge to recover some of their losses.   In spite of this, former slaves could not become imams, except possibly in mosques built by and for former slaves, and those seeking to make the pilgrimage to Mecca had to buy their freedom all over again under traditional law.   Former slaves were still expected to be deferential, to address their one-time masters as "father", and in some cases, to sit or stand in a properly deferential way.   This was true not only in conservative areas, but in regions like the Saalum region of Senegal, where former slaves were able to acquire land and wealth and competed quite effectively with nobles.   The question I would like to address is why.

Staying where They Were

  • 4 On the reconstruction of Wasulu, the most systematically ravaged part of the Soudan, see KLEIN (199 (...)
  • 5 The Mandinka word woloso means "born in the house".   There are numerous words for slave in the sav (...)

4During the period between the conquest and the First World War, at least a third of the slaves in French West Africa picked up and went home. With little help from either the colonial state or the missions, they cleared land, re-created communities, re-attached themselves to natal lineages and began life once again as free men and women4.   A majority of slaves, however, remained behind.   Most of them were slaves "born in the house", who worked land in exchange for a labour rent or a percentage of the crop5. For them, there was a complicated process of renegotiation.   I use this term not in any sense of formal negotiation, but in the sense that changes in the social situation of freed slaves came out of a process of give and take, in which masters had to recognize that the power they once had over their slaves no longer existed.   Masters, who prided themselves on their ownership of slaves, found themselves pushed to moderate their demands. The hours of labour were often reduced and so in some areas were the more onerous demands.   The masters were concerned, however, with status and with the symbols of submission.   The slaves were usually willing to continue many of the rituals of submission if they could have control over their family and work life.   Control over family life meant negotiating their own marriages and being free of the sexual demands of their masters.   Control of work life meant having one's own piece of land, albeit subject to various obligations, of which the most important was the payment of part of the crop to the former master.   It seems that in most areas there were soon only captifs de case (Meillassoux's esclaves casés), who worked lands allocated to them by their masters.   The trade slaves, captured during their lifetime and more harshly treated, were generally gone, or had their situation transformed into that of captifs de case.   Some writers, therefore, have used the term serf instead of slave for these freed, but still subservient, slaves (Derman 1973; Botte 1994: 112, 1999a).

5The slaves who went home were usually persons raised in the master's society and were indoctrinated from infancy to deference and to hegemonic ideology.   They spoke the master's language and participated in rites of passage in his society.   In spite of the tendency to deny the slaves any kinship position—the master was their father—the slaves had kin, but even the affective ties to parents, siblings and spouses tied them into the community.   Whether or not the slaves internalized their submission, they knew how to speak to the master or other nobles, what to expect and how to beg.   The slave's status was ameliorated, but pressure could be brought to bear.   The nobility had many ways to persuade the former slave to adapt. Yaya Sy (2000) tells a story which demonstrates the ways in which, even today, the nobles can bring pressure on a servile person who forgets his place.   This story took place in Paris in the 1990s and tells of two lovers who violated the ultimate taboo, that of a sexual union between a male slave and a female noble, in this case the daughter of a marabout.   The male lived in one of the dormitory apartments that housed Soninke immigrants.   He carried out menial tasks for the family of his master, but soon got involved with a daughter of the master's family and was seen leaving a hotel with her.   Called to appear before an assembly of nobles, he refused to present himself and refused to pay a 5,000-franc fine.   When the young woman got pregnant, she wisely moved out of her father's residence and even sought the protection of the police.   She clearly thought that both she and her lover were in physical danger.   The young man was denied all access to mutual aid, as were his uncles.   Like many immigrant groups, Soninke immigrants put a regular sum into a fund that can be used to deal with problems.   Being cut off deprived the lovers of any possible assistance if struck by ill health or unemployment.   His family in the village was excluded from village events, including the community prayer at the mosque.   They were also denied access to a cemetery, denied the right to get married in the village, and some in the village were deprived of land allocated to them.   His supporters in Paris were cut off from the village association there.   The young woman was totally excluded from her family, and her sisters were sent back to Mauritania to be educated according to Soninke values.   The young couple were able to sustain their relationship only because they were in Paris and because she escaped her father's apartment before something more drastic was done.   In rural Mauritania or Senegal, a dissident could leave, but that could involve cutting himself off from his family or from networks that provided work and security.   What is striking about this story is while fears of physical coercion were present, the most important threats were of exclusion from the material and psychological benefits of being part of the community.   Group sanctions were also used.

6There was a difference between the more conservative societies like the Soninke and those in areas where wages and cash crops created options. In many communities, the only option for the slave who resisted was to leave, and yet, even those who left often found that in old age, their only security was where their roots were.   Mirjam De Bruijn and Lotte Pelckmans tell of Maman, a woman who fled a village in the bend of the Niger after being raped, and married a wealthy man in Abidjan.   She lived well and was accepted as a free woman, but her husband died young and when her resources declined, the daughter she left behind brought her back to the village, where she had to re-adapt to servile status.   Judith Irvine (1973: 100-110) argues that in Kajoor, even a man who paid for his freedom was regarded as a slave.

7Of course, there were often tensions.   Slaves were not always docile. David Ames (1953: 16-18) wrote in the 1950s about a village in the Gambia:

"Jam are still expected to show deference in the presence of the free born.   However, some of them have taken advantage of the freedom given them by the British and as one elderly freeborn man put it, they throw their hats as far as they can. Jam still take off their shoes when greeting their masters and respectfully greet them as ‘grandfathers'.
Not all jam, of course, follow these patterns; some of them avoid such behaviour to the extent that they are permitted to do so by society."

8When doing research on slavery, one hears countless stories about people of slave descent who refuse to play the role.   Thus, a friend told me that his father had an old slave who came from time to time to greet his father and often to beg from him.   I asked if the man had children. Yes, I was told, he had a son, who was a schoolteacher.   The son never came to visit, even on special occasions.   For the masters, the freedom slaves had under the law encouraged potential troublemakers to either leave or invest their energies in work or entrepreneurial activities.

9The process of negotiation meant that there was a great variety of social situations.   Situations changed over time and varied from one region to another.   Slaves effectively remained slaves in many desert areas and were more dependent in desert-side areas, but even there, slaves made choices. Bella, for example, were "freed" when Tuareg revolted against the French, but sometimes, a Bella fraction found it useful to reaffirm a traditional relationship.   In 1946, there was a large-scale movement of liberation of Bella and Harratin in desert-side areas.   It involved at least 80,000 people and probably freed people, who had been freed at least once before (Klein 1998: 233-235).   Slavery became for many a clientship relationship.   Slaves often continued to give a part of their crop to their masters, and still do, but this usually gave them benefits.   They could turn to the master if in need, and since independence, could get access through their masters to development agencies.   Among the Soninke, they participated in migratory networks and group self-help.

Control of Land

  • 6 The advantage of a desert-side location was that the exchange between nomads and agriculturalists w (...)

10One very important question was control over land.   Most of West Africa had surplus land until after the end of colonial rule.   This meant that freed slaves could easily move out either in family groups or in larger communities and find land.   Thus Roberts (1988: 293-295) describes the process of desert-side slaves moving in groups to create villages in slightly better watered areas of the Sahel.   A desert-side location was no longer advantageous and rainfall was essential to survival6.   This process has continued to the present largely because of the droughts that have frequently troubled the savannah lands since 1968.   In 1988, I saw communities of Bella and other desert-side people, who had moved to the well-watered but under-populated lands of Wasulu because of the droughts.   Though the moves were undoubtedly motivated by climatic change, one effect was to free the migrants from the demands of Tuareg fractions.   In the commercially important Saloum region of Senegal, former slaves could often find land and create hamlets, administratively linked but physically separate, from their natal villages.   There were, however, areas where historical or ecological conditions left little land for new settlements.   In these areas, former slaves depended on their former masters for land to cultivate.   This dependency locked them into social relations that perpetuated their servility.

  • 7 Vieillard left behind valuable notes, which are available as the Fonds Vieillard at the Institut fo (...)

11This is striking in Guinea, where control of slaves broke down early in the coastal part of the country (Klein 1988).   It happened somewhat later in upper Guinea, where there was a large movement out of Mandinka villages after the First World War, usually led by returning veterans.   Servility remained important, however, in the Futa Jallon, where the existence of a powerful state facilitated the accumulation of large slave holdings.   The 18th century civil war ended with nobles living in foulassos on the ridges and their slaves in hamlets (runde) in the valleys.   Wars of expansion in the 19th century and the Futa's central position in trade between savanna and coast increased the accumulation of slaves (Barry 1988; Botte 1991). At the moment of conquest, the Futa had a slave majority.   There was significant slave flight during the period of conquest and the tempestuous period after it.   While Futa Jallon was seen as an unfriendly state, the French military encouraged flight (Botte 1999b).   As elsewhere, these flights probably involved the departure of many of those who knew another home.   The Futa was conquered in 1896 by a very small force of 150 men only because the French commander was able to play on internal rivalries and ally himself with one of the factions.   Conscious of their weakness and in spite of the frequent intransigence of the chiefs, many French officials were convinced that the kingdom could only be ruled in collaboration with them.   They ended up reinforcing and relying on chiefly power, which meant recognizing the control of slaves.   As a result, former slaves (riimaybe) found themselves in a position where access to land involved their submissiveness to the authority of men who remained masters.   Accounts of the Futa written during and after the Second World War underline this. Oswald Durand's novel, Terre Noire (1935), makes the rather naive argument that the introduction of ploughs and progressive agricultural methods will free the riimaybe, but he describes a degree of control that had changed rather little since the conquest.   The riimaybe were submissive; the Fulbe arrogant.   The riimaybe handled all forced labour levies, paid high rents, were sometimes robbed, and their women were subject to the demands of Fulbe males.   Gilbert Vieillard, one of the first professional researchers in Guinea, described a situation in which the riimaybe worked more hours for their masters than for themselves—and bore all of the burden of forced labour for the French7.   In 1956 and 1958, Sekou Toure's Parti démocratique de Guinée was able to prevail in an election and a referendum by organizing the servile masses against their masters (Botte 1994).   To be sure, many riimaybe had left.   Some others had done well because of their willingness to try different kinds of work.

12In the Futa, many riimaybe remained in spite of the overcrowded, eroded land.   The conservatism of the French had the effect of tying much-needed labour to an economy that was not very productive.   In Masina, it was just the opposite.   The region dominates the inner delta of the Niger, an area where floods water and depose silt on fields which produce yields twice as large as the rain-fed area adjacent to the flood plain (Gallais 1967). In spite of the wars that divided the area during the period between 1864 and French conquest in 1893, population was high and the riimaybe had good reasons to stay.   Like Futa Jallon, the Masina state was created by a jihad, which ironically was in many ways a rising of Fulbe pastoralists against the depredations of the slave trade (Sanankoua 1990; Brown 1969). Fulbe slave holdings were not large when the jihad was begun in 1818, but the creation of a defence against slavers necessitated development of a tax base.   This was done largely by developing slave-based agriculture.   Thus, when the French conquered the area, there was a slave majority with a sharp division between Fulbe, who were scholars, functionaries and pastoralists and riimaybe, who worked the land.   This was true in spite of almost 30 years of civil war in the area.   The Soudan (then Haut-Sénégal et Niger) had pursued a more active anti-slavery policy than Guinea, largely because of William Ponty, who administered the area, first as délégué-général and then as Lt. Governor (Conklin 1997: ch. 3 & 4).   In Masina, many administrators feared that if they undercut Fulbe authority, massive slave departures would leave Bamako without adequate grain supplies.

  • 8 Salmana CISSé's unpublished thesis (1978) is the best description of the situation in Masina in rec (...)

13In Masina, riimaybe resistance involved not departure, but protest and a refusal to pay the jamgal, the tribute they owed their masters.   The jamgal was about twice what slaves who farmed for themselves elsewhere owed their masters.   Normally, such slaves paid about 250 to 300 kilos of grain a year, which was as much as was necessary to feed an active man for a year.   The riimaybe of Masina paid about 550 kilos a year, about twice as much.   The French tried to arbitrate between riimaybe and Fulbe on the basis of sharecropping contracts (Klein 1998: 179-185).   Their logic was that the riimaybe were free, but the Fulbe owned the land.   Essentially, French policy was to replace servile obligations with rent.   In general, the French were not opposed to exploitation.   They were concerned to distance themselves from slavery, but not to free riimaybe from an oppressive economic relationship.   Their model for a contract was a sharecropping relationship, métayage, which was common in France.   Riimaybe representatives agreed to the proposed settlement, probably because it cut their obligations in half. The Fulbe, however, were intransigent.   The riimaybe were their slaves. They owned them.   That was more important to them than ownership of land.   The Fulbe were particularly unhappy that Lt. Governor Clozel had suppressed two concessions made to them, the right of masters to the services of riimaybe children and to the bride price paid for riimaybe girls. Both would have confirmed the slave status of the riimaybe.   The French started out sympathetic to the property rights of the masters, but by 1910, many of the administrators were so frustrated by Fulbe intransigence that they simply wanted to suppress all Fulbe rights.   That did not happen. With the coming of the war and a shift to the right after the war, the subject virtually disappears from archival sources.   The Fulbe, however, kept the land and collected rent, but still regarded riimaybe as their slaves.   Some riimaybe managed to acquire land, but those who did not acquire land remained subservient, addressing their masters as father, sitting in the back of the mosque and doing unclean tasks like cutting up sheep.   As in Guinea, the issue erupted again in the early days of Modibo Keita's radical Union soudanaise regime8.

  • 9 Islam was capable of serving as the slaveholder's religion, but it could also be a religion of revo (...)

14We see the question of control of land even more clearly if we look at Senegal's bassin arachidier, the area which saw the earliest and most striking development of a market economy in French West Africa.   It was an area in which the French and the chiefs were both interested in the fullest development of productive resources.   In many areas, this meant giving slaves a great measure of freedom though masters sometimes resisted such actions.   There were, however, striking differences between Kajoor in the old peanut basin and Saalum, which eventually became the most productive region, but only began developing export agriculture after conquest by the French in 1887.   During the 1880s and 1890s, there was a significant movement of slaves from the Soudan to Kajoor, where they worked the peanut fields.   Many came even after the defeat of Lat Dior in 1886 gave the French complete control of Kajoor.   The rapid development of peanut cultivation in areas near the line of rail meant that there was little free land and no place for freed slaves to find land.   The salvation of the slaves came from the Mourides of Amadu Bamba, who colonized eastern Bawol and eastern Saluum with groups that consisted primarily of former slaves (Babou 2002; Copans 1980; Klein 1998: 200-203).   A doctoral thesis by Judith Irvine gives a picture of the situation in a village in southern Kajoor in the 1960s.   The jaam, persons of slave descent, were only about 12% of the village population, but everyone said that jaam were more numerous in earlier times.   Some had moved into the cities, which local people thought to be made up mostly of slaves.   Most probably went to Mouride dahra9. In these agricultural colonies a young man would work for about eight years and then be freed to marry.   For a former slave, it was an opportunity to get land.   Mourides of servile origin often speak of Amadou Bamba as their liberator.   The remaining jaam in the southern Kajoor village still did onerous tasks like the construction and cleaning of latrines.   Many still lived in the households of their masters.   Irvine cites only one jaam family budget, but it allotted 5% for assaka, a considerably larger sum than budgets in the Saalum area.

15Pélissier's thesis (1966) on Senegalese peasants clearly describes the difference between Kajoor and Saalum.   In Saalum, two areas were of interest.   Eastern Saalum had a low population until Mouride colonization and French well-digging techniques opened up what came to be known as the terres neuves.   Mouride colonies moved into the area as the railroad inched its way from Kaolack to Kayes after 1904 and often fought with Fulbe herders for control of the land.   Most of it was not accessible to small or unprotected groups.   To the degree that freed slaves moved into the area, it was mostly as members of Mouride colonies.   In areas south of Kaolack, there were more villages between lots of land between them.   Free land was available.   There was thus a pattern of jaam seeking land.   Thus, one freed slave family from a Tokolor village moved down into the Gambia and then back to Senegal, where it found land in the village of Thyssé-Kaymore, about 20 km from their original location.   The patriarch of the family, Al Haj Musa Ba, became one of the most successful peasants in the village. His household of over 60, the largest in the village, split up only in 1972. Similarly, Al Hajj Biraan Turé moved in the 1930s to a hamlet outside of the village of Sonkorang and as he cleared land, gradually moved his family there.   At one point, he had 15 sons or labour migrants (navetaans) working under his direction.   He sold seven cows to finance the purchase of his freedom and his pilgrimage to Mecca.

  • 10 VENEMA (1978: ch. 11 & 12) identified 126 men for a study of village elites. Of these, 26% were jaa (...)

16Jaam are still limited in many ways, but play a major role in modern institutions.   In 1980, five of eleven members of the cooperative council of the village of Sonkorang were jaam and three of the first four presidents, all of them prosperous peasants10.   But the most interesting was the question of assaka.   I searched the village monographs done by students at the École nationale d'économie appliquée.   They included budgets, though not always with indication of who was jaam and who paid assaka.   Four jaam budgets, however, listed assaka at from 0.6% to 3% of household expenses. A number of other budgets included assaka in a category that included gifts to marabouts and other ritual expenses.   The total was never over 8% and often a good bit less.   Assaka had become for many a token payment, but it was still paid.   This contrasted to higher payments in southern Kajoor, and even higher in the bend of the Niger.   As elsewhere, the willingness to work hard has served freed slaves well and has contributed to their integration into the institutions of modern life.

Codes of Honour

17The question of land was important, but it did not fully explain slave-master relations.   When I first began doing research on slavery in French West Africa, one of my working hypotheses was that economic change had eroded the content of slavery.   I found that in Senegal, where I started, slavery no longer existed in its classical form, that is to say, as a labour system. Former slaves were often as well off as their former masters, but much to my surprise, I discovered that even in the heart of Senegal's bassin arachidier, important vestiges of slavery still existed.   Former slaves often paid assaka to their former masters and they could not make the pilgrimage to Mecca without buying their freedom under traditional law.   Buying a freedom previously given by the colonial state was called among the Gambian Mandinka "paying for his/her neck" (Bellagamba 2005).   The most difficult problem I faced was reconciling my materialist assumptions about historical change and the persistence of institutions that confirmed the social hegemony of the former masters.   I was struck by the fact that not only did former slaves accept aspects of their own submission, but that masters clung most passionately not to the economic perquisites, but to the status that slave ownership gave them.

  • 11 There is an important question of terminology.   I often speak of the masters, but the masters were (...)
  • 12 POLLET & WINTER (1971: 432) make the same point.

18De Bruijn and Pelckmans (forthcoming) stress two variables in explaining the perpetuation of servility.   The first is pulaaku, a Fulbe code of honour (Riesman 1974, 1992).   The second is the absence of history (Klein 1989).   Pulaaku is one of several such codes found in the western Soudan. Bazin (1975) speaks of Bambara horonya, Bellagamba (2005) of horomaa among the Gambia Mandinka.   These codes demand that noble be brave and generous11.   He is also restrained in behaviour.   He does eat in front of others and is discrete about bodily functions.   He is not crude, does not speak in a loud voice.   The noble woman is chaste.   The slave is loud and vulgar, often coarse and shameless.   The slave has the license to do many things a free man will not do.   Ames writes that "It is said that the slaves are fearless and without shame, even able to accuse a person of adultery in public, which a freeborn could never bring himself to do" (Ames 1953: 17)12.   Bellagamba (forthcoming) writes:

"Slaves lacked ‘honour' [. . .] and all those positive social and personal qualities, which the master's ideology attached to the freeborn identity, i.e. self-control, shame and initiative.   According to the masters, the negative moral qualities inherent in the slave's condition declined, but never disappeared with the passing of generations [. . .]."

  • 13 Role-playing has been probably important to slaves in many cultures, for example, ancient Rome, whe (...)

19Of course, much of what the slave was licensed to do was to play the role of a Sambo, the licentious, happy-go-lucky slave.   This is a stereotype that exists in many slave societies (Elkins 1959).   Riesman, Ames and Vincent describe the attraction of the slave village for freeborn Fulbe.   Riesman (1977: 121) speaks of Fulbe men enjoying the freer atmosphere of riimaybe villages and visiting them regularly.   Ames (1953: 17) tells of lessened restraint when freeborn men were with slave women: "They may caress them in public, and tell risque stories in their presence as they never would with women of their own class."   This, of course, means that social relations between the two groups often depend on role-playing, which underlines the nobility of the freeborn and the moral inferiority of the slave13.   Vincent (1963) opens up another dimension of servitude when he describes slaves tearfully pouring their problems in sessions which suggest either psychological dependence or role-playing.   For the former slaves, playing a role was a way to deal with and entertain the nobles, but the effect was to confirm the difference between the two cultures.

20The ultimate expression of honour and dishonour lay in the realm of sexuality.   The slave woman was owned.   She could be used.   She did not have the option of chastity.   Women being traded were often raped and once bought, could be given to whomever the purchaser wanted.   For female slaves, a key aspect of the end of formal slavery was often control over their own sex lives, but in many areas, slave women were still available.   Meillassoux writes that in Gumbu, members of the master's family could always have sexual relations with slave women (Meillassoux 1975: 220-230; Pollet & Winter 1971: 26, 250; Olivier de Sardan 1976: 19, 1984: 35-39).   The denial of the slave woman's control over her own body meant that this was a major concern for both slave men and women during emancipation.   In spite of this, in many areas, sexual exploitation continued into and after the colonial period.   Orsini (1966) suggests that in the Sahara, a guest would often be provided with a slave girl for the night, a practice that was expressly prohibited by Islamic law.   Robin Maugham, who went to Timbuctou to buy a slave in 1958, describes a group of young male Tuareg sending a slave girl out to collect firewood, then following her and raping her (Maugham 1961: 167-168).

  • 14 A former slave woman described such dances to Maria Grosz-Ngaté.   Personal communication.

21One of the most vivid expressions of slave sexuality was the obscene dances slave women performed in many West African societies.   Robert Arnaud (1910) describes such a dance in a novel based on his pre-World War I service in the Futa Jallon.   In it three young slave women, stripped down to several pieces of cloth, did an intense frenetic dance to obscene refrains, simulating sexual intercourse and making lascivious movements (Arnaud 1910; Pollet & Winter 1971: 246)14.   These dances have long since disappeared, but dancing and in some cases, music, have been subjects of concern in many communities as a commitment to stricter forms of Islam has deepened.   Dances done in African villages today are no more erotic and usually much less so than what is done in European discos, but stricter Muslims often see dancing as immoral.   In some villages, dancing is totally banned.   In others, it is banned only for the freeborn, thus setting a boundary between the freeborn and the slaves.   Freeborn persons may gather and watch, but they do not participate.   Baz Le Coq (forthcoming) writes that among the Tamachek, free persons danced, but they did so in a stiff, slow and restrained manner whereas slaves danced without restraint and with much movement.   Similarly, it is usually only griots that make music. The one exception is slaves (De Bruijn & Pelckmans 2005).   The innocent enjoyment of music thus serves to perpetuate boundaries between the noble and the servile.

22The second thing De Bruijn and Pelkmans insist on is the absence of history: "Without a deep genealogy or without a tarikh one is considered a non-person."   Slaves did, of course, have memories.   They also had parents, siblings and children, and in the decline of slavery often demonstrated love and affection.   Maman's daughter brought her back from Abidjan when she got old and sick, but juridically, slaves did not form lineages.   The problem was that slaves had none of the rights in offspring, progenitors or spouses that came with recognized marriage.   If a slave died, his master could claim his property.   Slaves are often reluctant to talk about their origins if they know them at all.   This means that any family the slave has is the master's family and any history is the master's history.   Former slaves being interviewed often know the master's history better than their own.   To be sure, many slave-master ties are real.   The more wealthy or powerful the family, the more likely the master's most trusted family member is a slave.   Slaves have privileged relations in the households of the powerful and the wealthy because the master can trust the slave.   He cannot always trust his brother or even his son.   The master and the slave are often also kin.   Though only the most powerful slave official can ever marry a free woman, masters traditionally had slave women as a fourth wife or as a concubine, sometimes referred to as a fifth wife.   Here too, the inferior woman's slave status might make her a favourite.   Any children born to the slave wife or concubine inherited the father's status.   De Bruijn and Pelckmans discuss a local politician who inherited his father's status, but used his mother's servile roots to seek votes in the ex-slave community.

  • 15 This was from interviews conducted for me by Aly Kampo in Masina.

23Begging also sets boundaries.   When doing research on slavery, one is often told by nobles that they give more to their slaves than they receive from them.   This is probably true, at least for those areas where assaka has become only a token payment.   Begging is a sanctioned behaviour for both griots and slaves.   Noble codes stress generosity as much as courage and restraint.   The beggar gives the noble the opportunity to prove his nobility.   Even the poorest noble feels obligated to reach into his pocket for a few coins.   Peter Weill is convinced that slaves consciously exploit the noble's code of honour by inviting him to weddings and naming ceremonies and turning to him in moments of need (Weill, personal communication). This is also true of slaves attending the master's ceremonies.   When asked if he participated in his master's ceremonies, a riimaybe from Masina responded, "If there is something to be gained, I go15".   Similarly, Wynkoop (1991), writing about Koubia in the Futa Jallon, suggests that when Guinea's socialist economy collapsed, nobles were more successful in getting credit. This meant that credit was likely to be extended to riimaybe who accepted their status.   This, in fact, may have been a factor throughout the colonial period.   Peasants operated with a narrow safety margin.   Any peasant was well advised to maintain links that could provide a safety net.

The Quest for Honour

24Emancipation for the freed slave involves constant struggle.   Masters everywhere sought some way to maintain control.   The responses of former slaves to their dependence were diverse.   Freedom gave former slaves diverse options and involved constant debate over whether to struggle or to conform. In any study of emancipation, there is a range of behaviours from a total refusal to conform to a total acceptance of the legitimacy of the master's claim.   Slave struggles start with an effort to assert their autonomy.   This involves not only control over their own work, but also control over family life.   This means the right to negotiate their own marriages, to receive bride price for their daughters, and to be free of the sexual demands of the master and his kin.   The value placed on discretion in sexual matters makes the history of the sexual interface between slave and free difficult to write, but ending the right of the master to command female sexual behaviour was undoubtedly one of the most important concerns of the emancipated slave.

  • 16 When doing research in Saluum, I was several times given the names of slave villages, probably orig (...)

25Former slave groups often create history, even creating genealogies or chief lists, which sometimes are patently false16.   Often, more elaborate histories are generated.   Sometimes, these may be beliefs long and quietly held, but often they are more interesting as memory than as history.   They often involve some claim to noble status.   J.-P. Olivier de Sardan (1976) has described a Songhoy slave village's reconstruction of the history of its conquest by Tuareg.   De Bruijn and Van Dijk (1993: 103) have discussed traditions of riimaybe in the Bend of the Niger, who claimed to be original inhabitants cheated of their land, which would change their relationship to the Fulbe.

  • 17 In many areas, the slave-owners who suffered the biggest losses were also the most learned Muslims. (...)

26Islam has been an important battleground in the slaves' quest for honour.   One of the striking characteristics of early colonial history is that there was a dramatic expansion in Islamic practice.   It involved both the slaves and the freeborn, but a major part of it was that freed slaves, who had been given only the elements of a Muslim education, now sought it out.   One of the ways in which slave-owners who had lost most of their wealth in the slave exodus, recuperated, was to teach the Qu'ran, perhaps to the very same slaves who had left their plantations17.   Former slaves sought religious education and practiced Islam with increasing dedication. Where former slaves were not welcome in the mosques of their ex-masters, they built their own.   At the same time, they were surrounded with prohibitions that circumscribed their roles.   A freed slave could not be an imam, could not travel to Mecca, and often could not kill animals without being freed under traditional law.   Groups of masters varied in their determination to resist the Islamization of the riimaybe.   De Bruijn and Van Dijk suggest that it was an effective way for the riimaybe of Hayre to acquire ndimu or nobility.   In the Futa Jallon, resistance was particularly intense.   Riimaybe were seen as dirty and ignorant, but even after they studied the Qu'ran and put on a white boubou, they had to fight for access to the mosque.   De Bruijn and Pelckmans describe a trusted retainer, who wanted to pay for his freedom so he could participate more fully in Islamic ritual.   When he asked advice from a cleric, he was told that it would be more admirable to take a collection so his master could make the pilgrimage to Mecca. He would bask in the reflected glory, but in the process, confirm his servility.   He ended up doing nothing.


  • 18 DE BRUIJN & VAN DIJK (1993: 97-103) see the different segments of Hayre society becoming like each (...)

28In his Honour in African History, John Iliffe (2005) has a chapter entitled "The Honour of the Slave".   In it, he makes the argument that slaves too had their sense of honour.   Iliffe sees honour as the "chief ideological motivation of African behaviour".   To extend this argument to slaves, he cites cases in which slaves displayed courage, bore punishment or refused to bend before the system.   In seeing a sense of honour as that which marked the nobles and differentiated them from slaves, I do not argue that slaves were never noble or brave, only that the concept of honour and the denial of history were used to convince slaves to accept their status.   It is important for the master to deny the slave's honour.   If the master recognized a form of slave honour, it was largely in loyal and faithful service.   Iliffe is, in fact, on more solid ground in the slavery chapter than in many of the earlier ones.   He constructs his argument on honour in ruling elites largely from epics, which define an ideal rather than practice.   His chapter on slave honour is to a much greater degree based on biographical data and court cases.   This is understandable.   Slaves struggled.   They fled.   They resisted.   They often also embodied the ideals of the societies they inhabited. That is why male slaves became Africa's most effective soldiers whether fighting for Samory or for the French or British colonizers.   For the French officers who commanded tirailleur units, they were the best soldiers possible because they were loyal and brave.   But slaves have also often showed pride in other accomplishments.   They often did as well as the nobles economically, and sometimes better, because they were willing to do all kinds of work and also, were often more entrepreneurial18.   I confronted many cases of slaves proud that they had provided a level of economic security for their families.   "I am the one who cleared these lands", Layen Njobu Sow claimed.   In an interview in a village within sight of Senegal's Transgambian highway, an old man pointed to trucks carrying part of the harvest up the road and claimed that many of the trucks were owned by jaam.   "My only Master", he claimed, "is Senghor", referring to Leopold Senghor, then President of Senegal.

29One of the interesting changes taking place at the time of writing is that the extension of democratic politics into the villages is contributing to the breakdown of hegemonic ties.   De Bruijn and Pelckmans begin their article with a discussion of a local election in which the chief of Douentza is angered that a group of former slaves of his family refused to vote the way he instructed them.   He threatened to take away their land, but did not do so.   When they gave him his share of their crops, he returned it, effectively breaking the relationship.   That, of course, is not necessarily the end of the story.   One angry action does not shatter customs rooted in years of practice.   They end the article with a comment made by the mayor of the village of Dalla, himself the son of a slave woman: "Il y a toujours la corde."   De Bruijn and Pelckmans conclude that the nobles still have the upper hand, but democratization has shifted the balance between the two groups.   The former slave is now a voter and has become less obedient with time.

30Department of History, University of Toronto, Canada.

Haut de page


— 1953 Plural Marriage among the Wolof in the Gambia, Ph. D. Thesis (Evanston: Northwestern University).

— 1910 Le Commandant et les Foulbe (Paris: Sansot).

BABOU, Cheikh Anta
— 2002 Amadu Bamba and the Foundation of the Muridiyya: The History and Development of a Muslim Brotherhood in Senegal, 1853-1927, Ph. D. Thesis (E. Lansing: Michigan State University).

— 1988 La Sénégambie du XVe au XIXe siècle: Traite négrière, islam et conquête coloniale (Paris: L'Harmattan).

— 1975 Guerre et servitude à Ségou, in C. MEILLASSOUX (dir.), L'esclavage en Afrique précoloniale (Paris: Maspero): 135-181.

— 2005 "Slavery and Emancipation in the Colonial Archives: British Officials, Slave-Owners and Slaves in the Protectorate of the Gambia (ca. 1890-1936)", Canadian Journal of African Studies 39: 5-41.

— 1987 Église et pouvoir colonial au Soudan français: les relations entre les administrateurs et les missionnaires catholiques dans la Boucle du Niger, de 1885 à 1945 (Paris, Karthala).

— 1991 "Les rapports nord-sud, la traite négrière et le Fuuta Jaloo à la fin du XVIIIe siècle", Annales ESC XLVI: 1411-1435.
— 1994 "Stigmates sociaux et discriminations religieuses: l'ancienne classe servile au Fuuta Jaloo", Cahiers d'Études africaines XXXIV (1-3), 133-135: 109-136.
— 1999a "Riimaybe, Haratin, Iklan: les damnés de la terre, le développement et la démocratie", in A. BOURGEOT (dir.), Horizons nomades en Afrique sahélienne (Paris: Karthala): 55-78.
— 1999b "L'esclave, l'almami et les impérialistes.   Souveraineté et résistance au Fuuta Jaloo", in R. BOTTE, J. BOUTRAIS & J. SCHMITZ (dir.), Figures peules (Paris: Karthala): 101-139.

— 1969 The Caliphate of Hamdullahi c. 1818-1864: A Study in African History and Tradition, Ph. D. thesis, mimeo (Madison: University of Wisconsin).

— 1978 L'esclavage "domestique" dans la partie Gourma du Moyen Niger: structure sociale et comportement de classe, thèse de 3e cycle (Paris: Université Paris VII).

— 1990 Economy and Society in the Upper Senegal Valley, 1850-1920, Ph. D. Thesis, mimeo (E.  Lansing: Michigan State University).
— 1994 "Slavery and its Demise in the Upper Senegal Valley, 1890-1920", Slavery and Abolition 15: 51-71.

— 1997 A Mission to Civilize: The Republican Ide of Empire in France and West Africa, 1895-1930 (Stanford: Stanford University Press).

— 1980 Les marabouts de l'arachide, la confrérie mouride et les paysans du Sénégal (Paris: Le Sycomore).

— 2005 "Facing Dilemmas: Former Fulbe Slaves in Modern Mali", Canadian Journal of African Studies 39: 69-95.

— 1993 "State Formation and the Decline of Pastoralism: The Fulani in Central Mali", in J. MARKAKIS (ed.), Conflict and the Decline of Pastoralism in the Horn of Africa (London: Macmillan): 122-142.
— 1994 "Drought and Coping Strategies in Fulbe Society in the Hayre (Central Mali): A Historical Perspective", Cahiers d'Études africaines XXXIV(1-3), 133-135: 85-108.

— 1973 Serfs, Peasants and Socialists: A Former Serf Village in the Republic of Guinea (Berkeley: University of California Press).

— 1943 Terre noire (Paris: L. Fournier).

— 1991 Colonial Conscripts: The Tirailleurs Sénégalais in French West Africa, 1857-1960 (Portsmouth, NH: Heinemann).

— 1959 Slavery: A Problem in American Institutional Life (New York: Grosset & Dunlap).

— 1967 Le Delta intérieur du Niger.   Étude de géographie régionale (Dakar: IFAN).

— 2005 Honour in African History (Cambridge: Cambridge University Press).

— 1973 Caste and Communication in a Wolof village, Ph. D. thesis (Philadelphia: University of Pennsylvania).

— 1988 "Slave Resistance and Slave Emancipation in Coastal Guinea", in S. MIERS & R. ROBERTS (eds.), The End of Slavery in Africa (Madison: University of Wisconsin Press): 203-219.
— 1989 "Studying the History of Those who would rather Forget", History in Africa 16: 209-217.
— 1998 Slavery and Colonial Rule in French West Africa (Cambridge: Cambridge University Press).

— 2005 "The Bellah Question: Slave Emancipation, Race, and Social Categories in Late 20th century Northern Mali", Canadian Journal of African Studies 39: 42-68.

— 1989 "Slavery, Emancipation and Labour Migration in West Africa: The Case of the Soninke", Journal of African History 30 (1): 89-106.

— 1961 The Slaves of Timbuktu (New York: Longmans).

— 1975 "État et conditions des esclaves à Gumbu (Mali) au XIXe siècle", in C. MEILLASSOUX (dir.), L'esclavage en Afrique précoloniale (Paris: Maspero): 221-251.
— 1986 L'anthropologie de l'esclavage: le ventre de fer et d'argent (Paris: Presses universitaires de France).

— 1976 Quand nos pères étaient captifs (Paris: Nubia).
— 1984 Les sociétés Songhay-Zarma (Niger-Mali).   Chefs, guerriers, esclaves, paysans (Paris: Karthala).

— 1966 Survivance des "non-libres" au Sahara et particulièrement à Tindouf, CHEAM, mémoire no 4074.

— 1966 Les Paysans du Sénégal (St. Yieix: Imp. Fabrègue).

— 1971 La société Soninke (Dyahunu, Mali) (Bruxelles: Éditions de l'Université de Bruxelles).

— 1971 "L'abolition de l'esclavage au Sénégal.   L'attitude de l'administration française (1848-1905)", Revue française d'Histoire d'Outre-Mer LVII, 210: 5-80.

— 1974 Société et liberté chez les Peuls Djelgôbé de Haute-Volta (Paris-La Haye: Mouton).
— 1977 Freedom in Fulani Social Life (Chicago: University of Chicago Press).
— 1992 First Find Your Child a Good Mother.   The Construction of Self in Two African Communities (New Brunswick, N. J.: Rutgers University Press).

— 1988 "The End of Slavery in the French Soudan, 1905-1914", in S. MIERS & R. ROBERTS (eds.), The End of Slavery in Africa (Madison: University of Wisconsin Press): 282-307.

— 1981 "The Banamba Slave Exodus of 1905 and the Decline of Slavery in the Western Sudan", Journal of African History 21 (3): 375-394.

— 1990 Un empire peul au XIXe siècle: la Diina du Maasina (Paris: Karthala).

SY, Y.
— 2000 "L'esclavage chez les Soninke: du village à Paris", Journal des Africanistes 70 (1-2): 43-70.

— 1978 The Wolof of Saloum: Social Structure and Rural Development in Senegal.   Agricultural Research Report 871, Wageningen.

— 1963 Pasteurs, paysans et pêcheurs du Guimballa (Partie centrale de l'Erg du Bara).   Nomades et paysans d'Afrique noire occidentale (Nancy: Université de Nancy): 36-157.

— 1944 Black Boy (New York: Harper).

— 1991 The Political Economy of Migration: A Case Study of the Impact of Migration in Rural Guinea, Ph. D. Thesis (Boston: Boston University).

Haut de page


1 A new native law code was proclaimed on 10 November 1903, Archives de la République du Sénégal, M 79.   When Secretary General Martial Merlin distributed the code to administrators, his instructions were explicit.   He told them to reject any claim by "would-be masters" to other persons claimed as slaves. He also told them that any effort to seize such slaves would expose the masters to legal action.   Circular, 10 December 1903.   Archives nationales, Section Outre-mer, Sénégal XIV 28.   Part of the circular is reprinted in RENAULT (1971: 100).   It is not clear that these instructions had much effect right away.

2 See MEILLASSOUX (1986: ch. 6) for a description of the slaves' stages of integration: from esclaves de peine, who worked full-time for a master, esclaves mansés, who also had a small plot, to esclaves casés, who worked on their own and paid their masters a part of their crop, and finally to esclaves manumis, free, but still clients.

3 Report from Dubreka, Archives de la République du Sénégal, 7 G 67.   In Fuladu in the upper Gambia, a similar response takes place with the return of veterans from the Second World War.   See BELLAGAMBA (2005).

4 On the reconstruction of Wasulu, the most systematically ravaged part of the Soudan, see KLEIN (1998: 121-122).   The Catholic missions had the best of intentions and were very important for runaway slaves, but the period of the great exodus was the same as the disestablishment of the French Church.   It deprived missions of resources that might have been used to help returning slaves.   In fact, many of those gathered around mission stations took advantage of the increased security to return to homes in southern Mali or in what is now Burkina Faso (DE BENOIST 1987: ch. 4 & 5).   De Benoist understates the Church's difficulties during this period (KLEIN 1998: 173-76).

5 The Mandinka word woloso means "born in the house".   There are numerous words for slave in the savanna, which have the same root: woroso among the Soninke, woloso among the Bambara, horso among the Soninke.

6 The advantage of a desert-side location was that the exchange between nomads and agriculturalists was very profitable.   The nomads maintained transhumant cycles and depended on desert-side agriculturalists for cotton textiles and grain. Some of their needs were met by their slaves, but desert-side communities also found production for the nomad market very profitable and acquired slaves to produce for it.

7 Vieillard left behind valuable notes, which are available as the Fonds Vieillard at the Institut fondamental d'Afrique noire in Dakar.

8 Salmana CISSé's unpublished thesis (1978) is the best description of the situation in Masina in recent decades.   It is a particularly vivid depiction of the tensions in the area.

9 Islam was capable of serving as the slaveholder's religion, but it could also be a religion of revolt and social justice.

10 VENEMA (1978: ch. 11 & 12) identified 126 men for a study of village elites. Of these, 26% were jaam, though jaam were 42% of the total population.   He was struck by the under-representation, but I was struck by the breakdown of disabilities in the economic sector.

11 There is an important question of terminology.   I often speak of the masters, but the masters were part of a larger community of the noble or freeborn, many of whom were quite poor.   The term freeborn is correct, but many of them are dependent in different ways.   I use all three terms, depending on the context, but prefer the term noble, though its connotation is very different from the one it has in Europe.   Being free is for many not being a slave.   Those who are free are defined by the codes of honour.

12 POLLET & WINTER (1971: 432) make the same point.

13 Role-playing has been probably important to slaves in many cultures, for example, ancient Rome, where a Sambo-type stereotype existed, and perhaps even more so, in post-slavery societies.   In Black Boy, his autobiographical account of growing up in Mississippi, Richard Wright describes how his seriousness and the absence of "happy talk" threatened whites.   He had to use subterfuges to get books and never mentioned his reading in front of a white.

14 A former slave woman described such dances to Maria Grosz-Ngaté.   Personal communication.

15 This was from interviews conducted for me by Aly Kampo in Masina.

16 When doing research in Saluum, I was several times given the names of slave villages, probably originally slave hamlets that had become autonomous villages.   Often, I knew that the tradition cited was false because it had too much time depth or because there was a change in the names of the chiefs.

17 In many areas, the slave-owners who suffered the biggest losses were also the most learned Muslims.   In spite of this, the families that suffered the greatest losses are often today very wealthy.   On a recent visit to Boké in Guinea's Rio Nunez, I interviewed two members of such a community, the Tubakayes.   Both talked about diverse strategies: hard work, shrewd investment and the teaching of the Qu'ran.

18 DE BRUIJN & VAN DIJK (1993: 97-103) see the different segments of Hayre society becoming like each other.   They describe the flexibility of former slaves in coping with the years of drought.

Haut de page

Pour citer cet article

Référence papier

Martin A. Klein, « The Concept of Honour and the Persistence of Servility in the Western Soudan »,  Cahiers d’études africaines, 179-180 | 2005, 831-852.

Référence électronique

Martin A. Klein, « The Concept of Honour and the Persistence of Servility in the Western Soudan »,  Cahiers d’études africaines [En ligne], 179-180 | 2005, mis en ligne le 01 janvier 2007, consulté le 16 octobre 2019. URL :  ; DOI : 10.4000/etudesafricaines.15018

Haut de page

Droits d'auteur

© Cahiers d’Études africaines

Haut de page