If you look at the history of Kujube, the people here were really very tired. They were not having peace all the time from the history of Mansa Kalamar up till today. The only time the people of Kujube had peace was during the reign of Mansa Dambeld who was then having his palace in Kabadio … But apart from that all these other ones, including from Mangone Seye, the people of Kujube didn't have any peace … and this Sylla and the influence of Alhaji [Haidara] and the [MFDC] rebel attack on their efforts … That is the reason people cannot stay in this area. People will come, fights happen; people will flee and never come back.Footnote ¹

In April 1990, separatist agitation in the Casamance entered a more militant phase when rebels of the Mouvement des Forces Démocratiques de la Casamance (MFDC) attacked the Selety border post, leaving two Senegalese customs officers dead. Reflecting on these events, the government newspaper, Le Soleil, drew the attention of its readers to a parallel occurrence in 1917 when an immigrant marabout, El Hadj Haidara, had declared jihād and beheaded the French customs officer-in-charge of the same post. Although the enemy was different – Casamance separatists in the one instance and Muslim millenarians in the other – the Senegalese press contrived to find a common thread in the shape of irrational extremism.Footnote ² In the quotation above, an informant from the border village of Kujube similarly alluded to the events of 1917 and the insurgency of the 1990s, but found an even older precedent in the shape of slave raiding by Fodé Sylla in the later nineteenth century. The partial convergence of these very different sources underlines the extent to which observers draw explicit links between contemporary manifestations of conflict and those that existed generations ago. These are different to the encoded forms of collective memory of the slave trade that Judy Rosenthal and Rosalind Shaw have explored: these are recollections of specific events that occurred in a more recent past.Footnote ³

Although the dominant tendency in writings about contemporary conflicts has been to focus upon proximate causes – such as the sclerosis of the state and the lure of marketable commodities like diamonds – there is a growing appreciation of the ways in which some wars are embedded in deeper histories. The use of amulets and the revival of secret societies and hunters' guilds are some of the outward manifestations of continuity.Footnote ⁴ In a re-appraisal of the conflict in Sierra Leone, Paul Richards takes the analysis a stage further. He recalls how nineteenth-century warlords on both sides of what is now the Liberian border responded to the incursions of Mandinkas and coastal slave traders by building up their own armies composed largely of slaves.Footnote ⁵ Many of the latter were subsequently settled in villages which continued to be exploited for their labour long after the British had formally abolished slavery. Drawing on testimonies from recently demobilized fighters, Richards argues that it was their descendants – who continued to be exploited by chiefly authority – who took up arms in the 1990s. This account, together with other emerging research, suggests that at least some of the violence in West Africa needs to be understood through an appreciation of the longue durée.Footnote ⁶

But what does it mean to say that historical patterns recur? To pose such a question goes to the heart of what one imagines the stuff of history to be. Comparative history has had relatively few takers in Africanist circles where the preference has been for the single case study, whether the scale be national, local or indeed trans-national.Footnote ⁷ One reason is that there is a residual scepticism about the methodological validity of comparisons that do not contain an obvious element of commonality: hence comparisons of colonial administrative systems and of African systems of slavery have been pursued whereas other candidates remain as yet ‘unthought’. Temporal comparisons within a single case study are even rarer. To suggest that there are historical cycles is often deemed tantamount to saying that societies have remained essentially static. Aside from the fact that the idea of history having a linear trajectory is itself potentially historicist, the objection is arguably based on a false premise. A comparison across cases typically assumes that there are points both of commonality and of difference which is precisely what makes the exercise worthwhile. Equally, a temporal comparison has to be able to account for continuity within change.

Broadly speaking, there are two reasons why historical patterns might repeat themselves: the reproduction of a given set of structures may produce cognate responses (an example being the embedded inequality to which Richards refers), and/or actors may self-consciously seek to re-enact the past.Footnote ⁸ However, historical resonances may also reside in more twisted connections. As I will reveal, violence at the end of the nineteenth century and in the early colonial period created zones of relative depopulation in the Gambia–Casamance borderlands. When more peaceful conditions returned, Jola settlers from farther south in Buluf (or Djougoutes) moved in to occupy the spaces. Their presence seems later to have attracted MFDC rebels who proceeded to re-create some of the forms of violence and in precisely the same locations. The manner in which rebels fed off the peasantry bore some similarities with late nineteenth-century warlordism, in which fighters would typically raid when the granaries were full. Settlement patterns therefore represent a good example of a historical concatenation that has worked itself out over more than a century. But equally, continuities may take the form of ideational inversions. At the end of the nineteenth century, Islam in these borderlands was associated with the enslavement of Jolas by Mandinka ‘warlords’. A hundred years later, when almost all Jolas had been converted, Islam was re-encoded as part of the eternal quest for peace and prosperity. The signs were reversed, but the histories of Islam and conflict remained as thoroughly intertwined as ever. To say that history does not repeat itself is true in a literal sense, but there are societies – including those which form the subject of this article – for whom the possibility of recurring patterns needs to be taken seriously, not least because contemporary actors operate in the belief that they exist.

The setting for this article is the far west borderlands between the Gambia and the Casamance – specifically the areas known as Kombo, Narang and Fogny Jabangkunda (see map).Footnote ⁹ This border region has received less treatment than the areas to the south – what one might call ‘deep Jola-land’ – in part because of its supposed atypicality and in part because the very existence of the international border has meant that the sources are scattered. As far as the former is concerned, the mixing of Mandinka, Jola and Karoninka (or Kalorn) is part of what I wish to examine in this article. As for the sources, the existence of two sets of archives means that it is sometimes possible to gain different perspectives on the same events.Footnote ¹⁰ Today, a number of different groups inhabit this zone, straddling the boundary itself. Many of the Jolas and Karoninkas have only arrived in the past half-century, reflecting a high incidence of population mobility. As I will demonstrate, this is partly a consequence of forced displacement, but it is also bound up with the history of peasantization and with mass conversion to Islam. Together, these processes have not merely shaped the contours of settlement, but have influenced perceptions of violence and its remedies.

Map 1. Map of the far west borderlands between the Gambia and the Casamance.

MARABOUTS AND SONINKES: THE FIRST CYCLE OF CONFLICT

In the nineteenth century, the Senegambia was engulfed in a struggle for supremacy between established ruling families – who were either overtly hostile to Islam or were syncretic in their observance – and self-proclaimed Muslim reformers who set out to establish new polities in accordance with the Shari'a. These were merely the latest manifestations of processes which had been playing themselves out across the Western Sudan for more than two centuries. The impetus for purification of the faith came late to the western Mandinka by comparison with what is now northern Senegal and Mali. But by the 1840s, a veritable revolution was under way as the Soninke (or ‘animist’) rulers of the Mandinka settlements in the Casamance succumbed to the maraboutic onslaught.Footnote ¹¹ On the south bank of the Gambia river, the kingdom of Kombo became a battle zone over the course of the next three decades, during which the advantage swung back and forth. The Muslim list of indictments followed a familiar script, with the Mansas accused of condoning the consumption of alcohol and participating in the veneration of ancestral and other spirits.Footnote ¹² In 1862, Governor G. A. K. D'Arcy of the Gambia encapsulated the great divide as follows:

The power struggle in Kombo was finally resolved in 1875 when the last Soninke king was defeated and forced to convert. Under Muslim rule, influence gravitated towards towns in Kombo that had championed the maraboutic cause, notably Gunjur.

From their trading posts at the northern tip of Kombo and along the Casamance river, the British and the French observed the emergence of a new breed of Muslim warlords who exploded D'Arcy's clearcut distinction. The latter prosecuted holy war against neighbouring Jola populations on the grounds that, as stubborn animists, their very existence represented an affront to Allah. European correspondence depicts Fodé Sylla in Kombo and Fodé Kaba in Fogny as the scourge of the Jolas, whom they enslaved in substantial numbers.Footnote ¹⁴ Whereas Kaba has actually tended to enjoy a rather favourable press, on the basis that he was genuinely devout and advanced the cause of Islam, Sylla has tended to be depicted as a ruthless warlord who dressed up the pursuit of material gain in the language of holy war – although the oral traditions of the Mandinka towns of the Gambia do emphasize his religious devotion.Footnote ¹⁵ Amongst the Jolas of the Casmance, Sylla is remembered as having waited for the crops to ripen and the cattle to fatten before despatching his soldiers to seize as much booty and as many people as they could carry away.

In a profound sense, one might say the ‘greed-versus-grievance’ debate had been operative in the Senegambia for at least a century before the emergence of the phenomenon of modern warlordism.Footnote ¹⁶ Although the two Fodés have often been placed at different ends of the spectrum, there is arguably little to choose between them. Both claimed a legitimate grievance against those who stood in the way of a thorough-going conversion and justified violence in terms of divinely sanctioned retribution. Sylla posted talibés to the settlements he captured, founded mosques and prohibited the tapping of palm-wine, much as Kaba did.Footnote ¹⁷ But equally, both leaders operated an economy of plunder in which Jolas were the victims. If one considers the consequences of their actions, Kaba certainly caused as much devastation in Fogny as Sylla did in his sphere. Most importantly, both leaders engaged very little in the arts of Islamic state-building. Had this been their primary concern, they might have sought to incorporate slaves in order to lend a demographic helping hand. In reality, their razzias led to localized depopulation as entire settlements were enslaved or dispersed.

Where the European accounts are misleading is in portraying warfare as purely ethnic. South of Kombo lay Fogny Jabangkunda, named after the Jabang family who were either amongst the first Mandinka immigrants from Gabou (present-day Guinea-Bissau) or, as some Mandinkas freely speculate, really Bainunkas who became Mandinka.Footnote ¹⁸ Subsequent arrivals from Gabou are said to have passed through Kabadio, which became the capital of its own small ‘kingdom’, before establishing their own settlements in Kombo, including Brikama.Footnote ¹⁹ To the east of Fogny Jabangkunda lay Narang. Like Brikama, its capital of Kujube was apparently founded after Kabadio. Although people today refer to ‘kingdoms’, it is doubtful whether the Mansas exercised much in the way of governance.Footnote ²⁰ However, they were thought to command considerable spiritual power. A Soninke Mansa in effect mediated with the spirits, including those particular ones which were thought to inhabit the forests.Footnote ²¹ The fact that Kujube is located in the Narang forest is therefore no coincidence.

When Fodé Sylla embarked on holy war, he targeted Mandinka settlements as much as Jola ones, and appears to have done so according to a pragmatic calculus.Footnote ²² Hence, the Mansa at Kabadio became an ally, and eventually a convert after an exchange of wives, whereas Sylla's armies fought a prolonged campaign against Mansa Kalamar at Kujube. Furthermore, it is mistaken to regard the raiding armies as being comprised entirely of Mandinka zealots. There are villages today that claim to have provided Sylla with recruits who were chosen for the perceived physical strength of the Bainunka element amongst them.Footnote ²³ Equally, Karoninkas, from the peninsula and islands of Karone, remember having provided Sylla with fighting men. These were perhaps mercenaries, although it is conceivable that some converted.Footnote ²⁴

Whatever the composition of Sylla's forces, they undoubtedly devastated much of Fogny Jabangkunda and Narang. The French and the British were intent on purchasing the produce delivered by Mandinka and Jola peasants, especially rubber and groundnuts, but this became virtually impossible under conditions of endemic insecurity. At the same time, Jola traders were wary of bringing their goods to the European posts for fear of being waylaid en route. The Europeans initially sought to parlay and even to enlist the marabouts in their own ongoing competition. Meanwhile, villages seeking protection fuelled Franco-British rivalries. In 1888, Sylla complained to the British that the Narang villages of Diébaly, Makuda and Koubanack were flying the French flag in lands which he claimed as his own. He recalled that they had previously sided with the Wolof warlord, Birahim N'Diaye, against him.Footnote ²⁵ Narang had become a veritable thorn in his side and he recalled that ‘I advised the French to be very careful how they took words from the Naran [sic] people, for they were very deceitful.’Footnote ²⁶ News that the French were seeking to persuade him to cede part of his ‘territory’ led to British diplomatic protests and counter-intrigues. The Administrator of the Gambia maintained that the British should be more proactive:

The decision to formalize a boundary arose out of a desire to take the heat out of this rivalry and to stabilize the zone stretching from the Atlantic coastline of Kombo through to Fogny. In 1889, the border was defined by treaty and a couple of years later the work of physical demarcation commenced. Fodé Sylla, knowing full well that any such border would circumscribe his activities, set out to impede the exercise. The French meanwhile sought to consolidate by investing in a Wolof strongman, Mangone Seye. From Makuda, Mangone forced surrounding villagers to labour on his own fields with the complicity of the French commandant. He seized crops as far east as Kartiak and reputedly trafficked in captives in collaboration with Sylla.Footnote ²⁸ His brief reign of terror, which was finally ended by the French, is still preserved in the songs of older women in Narang today.

In 1894, the British decided to deal with the problem of Sylla, who was defeated and retreated into the Fogny Jabangkunda. The French intercepted him not far from his military camp at Jalong and deported him to northern Senegal.Footnote ²⁹ However, they revealed their pragmatic side when they offered his soldiers land at Bandjikaky in the hope of turning them into peaceful cultivators.Footnote ³⁰ Fodé Kaba survived longer by astutely exploiting Franco-British rivalries. The British had wanted to settle their account with him in 1890, but the French had offered him refuge. In 1891, he signed an agreement placing himself under ‘protection’ in return for recognition of his rights over Fogny and Kiang.Footnote ³¹ This enabled him to continue his raiding against Jola villages, the captives being sold eastwards into Firdou in return for firearms and horses.Footnote ³² Two years later, another treaty forced him to relinquish his hold on Fogny in return for an annual payment. But it was only in 1901, after the killing of a British officer, that the French decided to cut their losses.Footnote ³³

The British and the French both expressed concern at the very low population densities of the areas they had acquired, which narrowed the potential for economic exploitation.Footnote ³⁴ Most of Kombo had provided the bedrock of support for Fodé Sylla rather than constituting raiding grounds. Nevertheless, the British estimated that the thirty or so villages they had taken over had a combined population of less than 8,000.Footnote ³⁵ In putting this down to the consequences of endemic warfare, they were presumably referring to the earlier decades. But these estimates further underline that Sylla had done nothing to rebuild the population.Footnote ³⁶ Hence, in 1904, Gunjur was the largest town in Kombo, but it still contained a mere 1,117 inhabitants.Footnote ³⁷ On the French side of the border, there were some towns that had befriended Sylla, but many others that had been systematically pillaged. When French officials travelled through Narang in 1894, they observed that the combined effects of Sylla and Mangone had been to empty many villages of all but a few huts.Footnote ³⁸ At Diébaly, only a few structures were left standing and Commandant Farque described conditions of the most abject misery:

The villagers petitioned for the return of fifty of their number who had been seized by Sylla.Footnote ⁴⁰ Those in Makuda listed people who had fallen into his hands, presumably with Mangone's connivance. At this point, villages were teetering on the brink of famine as locust invasions threatened to consume the limited food crops that had been planted. In Fogny Jabangkunda, the picture of a landscape that had been denuded of its population was repeated:

By contrast, the French noted that the population was dense in Jola areas that had largely escaped raiding.Footnote ⁴² This was to have important consequences for future settlement patterns.

COLONIAL FIXATIONS: DEMOGRAPHY, TAXATION AND FORCED RECRUITMENT

The borders between the Gambia and the Casamance were fixed while the colonial state structures were still embryonic. When it came, state consolidation was heavily inflected by the revenue imperative: people were competed for, both because of the direct taxes they represented and because duties were levied on the cash crops they produced. The incoming colonial regimes shared a view that the newly acquired territories would be worth very little, and would constitute a drain on the exchequer, unless the population base could be rebuilt.Footnote ⁴³ The British faced less of a problem in that towns around Bathurst, such as Bakau, had become a refuge for people fleeing instability further south. The challenge was to interest the refugees in resuming farming activities in southern Kombo, within the borders of the Gambia. At the same time, the British were intent on attracting additional settlers from French territory. The French sought to hold on to the people they had, whilst luring the refugees back across the border.Footnote ⁴⁴

After the defeat of Fodé Sylla, the British laid down certain ground rules with which the losers had to comply. These included the rebuilding of Gunjur and other towns to colonial specifications and accepting the right of return of Soninkes. The possibility of introducing a land tax, presumably along the lines of that in Northern Nigeria, was floated, but was dropped almost immediately because it might have provided a disincentive to further immigration.Footnote ⁴⁵ The principle was established that settlers should be able to gain access to land through the village chiefs (or Alkalos) with a minimum of fuss, subject to the proviso that they register and that adult males pay a yard tax. Seasonal workers, or ‘strange farmers’, were similarly considered as highly desirable and were liable for a special rent. Of the revenue raised inside Kombo and Fogni Province in 1922, the yard tax accounted for 38·7 per cent of the total, while rents from strange farmers added another 12 per cent.Footnote ⁴⁶ The yard tax was deliberately fixed so as to undercut the French head tax.

Although some Casamance refugees returned home, it was extremely rare for Gambians to relocate to French territory. French rule was regarded as harsher and more demanding. In the early days, the French authorities punished resistance by burning granaries and seizing livestock in a manner which seemed all too familiar. Moreover, in the early colonial period, the chiefs who were supposed to manufacture Jola compliance were often Mandinkas, because of the absence of an indigenous hierarchy to batten onto. Maintaining accurate population rolls and ensuring the taxes were paid was an important, and unpopular, dimension of their job. The reasons why the French were so relentless in the pursuit of head taxes is only partly attributable to the desire to balance budgets. They felt they needed an explicit, repeated and measurable sign that the peoples of the Casamance had accepted their subjection. The Jolas were often depicted as hardworking – in that sense, the antithesis of ‘lazy’ Mandinkas – but they were also stereotyped as nature's anarchists who only understood the language of force. However, the military reflex tended merely to unleash fresh waves of refugees into the Gambia. Each cycle of exodus and return left a residue in the Gambia, made up of people who resolved never to go back to the Casamance.

During the first decade of the twentieth century, French officials believed they were securing a modicum of Jola compliance. However, in the run-up to the First World War, forced recruitment into the tirailleurs sénégalais undid the good work. At least one British commissioner hoped to take advantage.Footnote ⁴⁷ As war beckoned, however, British officials received instructions to send back deserters, and those seeking to avoid military service, on the grounds that France was an ally whose requests for manpower were entirely legitimate. However, the cordon was not rigorously enforced and the Gambia became the net recipient of a substantial exodus from French territory.Footnote ⁴⁸ In fact, entire villages disappeared. Hence, when the present Jola occupants of Donbondir arrived there a few years after the war, they came across the remains of deserted compounds.Footnote ⁴⁹ The perceived weakening of the French also led to renewed resistance to taxation. Mandinka chiefs and their retainers fled across the border when they were attacked by Jola subjects who believed that the French would be too weak to retaliate.

Although it was the Jola who were construed as being particularly volatile, the most dramatic challenge to French rule actually came from another quarter entirely. In the early months of 1917, a marabout from the French Soudan, El Hadj Haidara, turned up in Kujube.Footnote ⁵⁰ He proclaimed himself a Cherif, announced his intention to expel the French and launched an assault on the customs post at Selety. The British came to the aid of their embattled neighbour and the incipient rebellion was nipped in the bud. Anticipating French reprisals, many people from Selety crossed the border and settled at Jiboro where they remain to this day on the principal crossing point along the Banjul–Casamance route.Footnote ⁵¹ The elders of Kujube sent their own youth away, hoping that the French would be more lenient when faced with a phalanx of elderly men. The ploy evidently failed: the village was largely destroyed and many people were killed, despite the lack of any evidence that the inhabitants had welcomed the insurgents.Footnote ⁵²

The Selety affair merely underlined a perception that the Casamance was slipping out of French control. For this reason, the authorities decided on a military re-occupation to subjugate the Jolas and to confiscate the large number of firearms in circulation. In the 1920s, the French were finally able to establish a greater measure of effective control, although further acts of resistance continued to occur.Footnote ⁵³ The population in the border zone paid tax and the production of groundnuts increased. But as the Second World War loomed, the French resumed forced conscription, with the result that the mass exodus was repeated. Seeking to balance the desire for more settlers with the imperative of not offending the French, British officials were supposed to discern whether particular groups of refugees were conscript defaulters and deserters or ‘genuine’ settlers. When people were sent back, British officials reported instances of brutality that made them less minded to cooperate in future. In April 1940, for example, the Commissioner of South Bank Province reported that a group who were persuaded to return had been roped together and frog-marched from the border to Bignona. One man died, and on arrival many were summarily imprisoned.Footnote ⁵⁴ Once Senegal joined the Vichy camp, the two administrations became adversaries and closed their common border.Footnote ⁵⁵

A TALE OF TWO MARABOUTS

Before turning to the repopulation of the borderlands, we pause on the issue of conversion which is crucial to a deeper understanding of these processes. At the end of the nineteenth century, there were still some Mandinkas who were Soninkes, while almost all of the Jolas continued to practise their own beliefs. In the 1920s, conversion accelerated, and by the 1950s the status of Islam had become hegemonic north of the Casamance river.Footnote ⁵⁶ Whereas Islam and Jola religion were to some extent reconciled in Buluf, the impact of conversion went deeper in Kombo and Narang and parts of Fogny Jabangkunda.Footnote ⁵⁷ Why Jolas who had been the victims of maraboutic raiding later converted to Islam is something which has intrigued researchers.Footnote ⁵⁸ Here, we seek to build on the platform of existing research through an analysis of current interpretations in Narang.

The inhabitants of Kujube have a rich sense of history born of a sense of repeated misfortune. The original village is now farmland, its occupants having moved to a new site a few kilometres away after they converted. The reason was that the old village was located too close to the sacred grove of Mansa Kalamar that is thought to be inhabited by dangerous spirits. This is a landscape saturated with historical signification. The people of Kujube tell an epic account of the struggle between Fodé Sylla and Mansa Kalamar that was fought as much on the spiritual terrain as on the field of battle. At one point in the story, the marabout's soldiers are driven away by bees – a standard trope signifying spiritual prowess. Having tried direct force and subterfuge, Fodé Sylla is forced to concede that he is unable to get the better of Kalamar. The story then goes that he provided a cow as a peace offering. After feasting, the carcass was buried and Sylla prayed. It is said that he prophesied that the people of Kujube would voluntarily embrace Islam within six generations. Interestingly, the fact that this was achieved in less time is not taken as evidence of divine intervention on the side of the marabout. There is a clear tension between the desire to identify with those who resisted Sylla, and the simultaneous urge to validate the ultimate triumph of Islam. The ambivalence is embodied in the material reality of the sacred grove itself. Although Mansa Kalamar's drinking well and grave may be visited, the grove is strictly off-limits. The few individuals who have entered this space are said to have died almost immediately because Islam and the forest spirits are thought to be incompatible.Footnote ⁵⁹ Significantly, the grove has not been reinscribed with fresh meanings, as was already happening in some parts of the Casamance in the later nineteenth century.Footnote ⁶⁰ On the contrary, it has been preserved, possibly out of a reverence for Kalamar, but also through apprehension about what might happen if it was interfered with.

The first person in Kujube to convert allegedly did so in order to redeem his sister who had been seized as a hostage by Sylla. However, this did not produce any immediate breakthrough. The real credit is accorded to the Mauritanian marabout, Cheikh Mahfoudz, as it is throughout Fogny Jabangkunda, Narang and much of the Lower Casamance.Footnote ⁶¹ Mahfoudz was a follower of Saad Bou and spent the last decades of the nineteenth century travelling through Senegal, Mali, Guinea and Guinea-Bissau – periodically returning to his master to deliver gifts – before entering the Casamance in 1901.Footnote ⁶² He sought French permission to settle, and spent the next four years in a village in Narang which he named Darsilami, signalling its status as a haven of peace in a region scarred by violence. This village was located in the middle of the forest, underlining the point that spirits posed no threat to the pious. After a spell in Guinea-Bissau, he returned and installed himself at Binako, near the Portuguese border. But because his relations with neighbouring Balantas were never cordial, he shifted his attention back to Darsilami(-Cherifkunda).

Mahfoudz is remembered as a peripatetic preacher who would spend time in a village seeking to convince people of the need to embrace the one true faith. Once he made headway, he would leave a few talibés behind to keep up the work, and move on. The memories of Fodé Sylla and Cheikh Mahfoudz are mnemonically linked, with the former standing for violence and the latter representing the way of peace. Indeed, legend has it that they once met and that Mahfoudz sought to convince Sylla of the folly both of fighting the Europeans and of trying to convert people through force.Footnote ⁶³ Whereas Sylla is remembered as being antipathetic towards Jolas, Mahfoudz actively sought them out. And while Sylla ultimately failed – being unable even to breach the spiritual defences of Mansa Kalamar – Mahfoudz is credited with complete success. It is said that he patiently explained how successive calamities were the consequence of not living according to Islamic principles, whilst managing to avoid offending Jola sensibilities. Crucially for our purposes, his message linked Muslim teaching to the peaceful pursuit of agriculture.

PEOPLE AND PLACE: REPOPULATION OF THE BORDERLANDS

In the half-century after 1945, the borderlands assumed a more peaceful aspect. There were no mass flights, and in fact the sparsely populated areas in the Casamance began to fill up with Jola settlers from further south. Karoninkas, who had long been migrating to Kombo on a seasonal basis to tap palm-wine, also began to establish a more permanent presence on both sides of the border. Finally, ‘strange farmers’ continued to enter Gambian Kombo in significant numbers.Footnote ⁶⁴ Settlers proper were attracted by the prospect of abundant land, but push factors were also operative in that they tended to originate from land-hungry areas. The population densities in Buluf were far higher than in the borderland regions, which explains why so many came from there.Footnote ⁶⁵ According to one informant, such status was attached to the possession of land, that matters of personal dignity were what led families to migrate.Footnote ⁶⁶

The incomers typically came in search of land that was suitable for wetland rice as well as fields to grow groundnuts, millet and vegetables and places to graze cattle. One of the added attractions of Narang was that it was suited to the planting of mangoes and oranges, which were purchased by merchants from Dakar. Jola settlers criss-crossed the international border in search of combinations of land which best balanced their overall requirements.Footnote ⁶⁷ On the Gambian side, the authorities perpetuated the British policy of welcoming strangers from the Casamance. Settlers normally approached the village Alkalos who were generally receptive. In settlement narratives, a common observation is that an Alkalo granted permission to look for an uncultivated area where the migrants could stake out their patch. In the Casamance, the chefs de canton did the same, but after 1960 it was the chefs d'arrondissements (government bureaucrats) who became the final arbiters.Footnote ⁶⁸ However, the chefs de village brokered access to land in practice.

In historic Mandinka towns, like Kabadio and Abené in the Casamance and Sifoe in the Gambia, Jolas and Karoninkas became a distinct presence, but constituted a numerical minority. Elsewhere, their influx was substantial enough to transform the character of existing communities. A classic example is Makuda which never really recovered from the combined effects of Sylla and Mangone. Here the incomers could imagine that they were occupying vacant land. The first settler was actually a Karoninka who enjoyed friendly relations with people from Mlomp. Some of the latter decided to join him and their numbers were later augmented by other Jolas from places like Balingore, Kartiak, Dianki and Tenduk.Footnote ⁶⁹ The result was a highly mixed village in which no single Jola sub-group was dominant. The Gambian border town of Darsilami, which had been renamed by Mahfoudz, was similarly transformed by successive waves of immigrants, blurring memories of the original founders.Footnote ⁷⁰ Karoninkas made up the largest contingent, but Jolas from across the Casamance were also well represented. The land between the town and the border was occupied by successive arrivals from the Casamance. One individual had been so harassed by the chef d'arrondissement in Diouloulou over tax that when he built his small hamlet inside the Gambia he called it ‘Tranquil’.Footnote ⁷¹ His place was later taken by settlers from Kaniabo, and they were subsequently joined by natives of Mlomp who relocated from nearby Touba. Darsilami also attracted a considerable number of Manjago refugees from Guinea-Bissau. Because they were mostly non-Muslim, like the majority of Karoninkas, Darsilami ended up being more religiously diverse than most towns in Kombo.

In other instances, villages were founded from scratch, often by settlers who came from the same village groupings in the Casamance. Mahamouda was founded in 1936 by people who came from Mlomp after first settling at Berending in Gambian Kombo. The village survived the war and attracted fresh migrants thereafter. Nearby Touba was also founded by settlers from Mlomp in 1954. They had first tried Nyofelleh, but, because of land litigation, they turned back into the Casamance and settled, initially at Makuda and finally at their present location.Footnote ⁷² Again, the people of Dimbaya originally came from Sindian in the Casamance, moved to Bunto and then to Sifoe in the Gambia, before settling for some time at Darsilami. After a few seasons, they finally relocated to establish a new village, having ascertained that there was fertile land to be had.Footnote ⁷³ An intra-family dispute led to some of them moving a couple of hundred metres southwards to found a separate settlement just inside Senegal.

The mutual reinforcement of conversion and Mandinkization has been examined by previous researchers. Linares tracks their combined effects on the sexual division of labour amongst Jola communities in three different Casamance locations.Footnote ⁷⁴ In an earlier study, Pélissier observed that Jola converts borrowed from a Mandinka template, constructing new towns centred around the mosque, whereas Jola villages had been loose collectivities of extended family compounds (fank).Footnote ⁷⁵ If the observation has any validity, it is especially true of the new settlements which were founded by incomers who had no particular attachment to local space. In the Mandinka towns, both space and social hierarchy had already been codified by the Europeans. The British established a chiefly hierarchy based on the village Alkalos and the Seyfos above them. The founding families provided the chiefs while other lines tended to produce the Imams. But where Jolas outnumbered Mandinkas, they were able to demand more recognition. In 1969, the Seyfo of Brikama had cause to complain about immigrants who were seeking to oust the Alkalo of Kitti.Footnote ⁷⁶ However, his attempts at insisting on the rights of indigenes received short shrift, presumably because the Gambian authorities did not wish to do anything to discourage immigration. Because there were only two tiers of chiefs in the Gambia, new Alkalos enjoyed the same status as those who had granted them land. Hence Dimbaya became an autonomous village despite being built on Darsilami land.Footnote ⁷⁷

In a village like Makuda, where the original inhabitants were effectively invisible, Islam provided a charter for social equality. The village was constructed around a community of believers following the same religious observances and sharing in a common calendar of social activities.Footnote ⁷⁸ This underlines a broader point about the centrality of religious ideas to repopulation of the border zone. The model that Mahfoudz exemplified took mobility as in the natural order of things. In Donbondir, it was Mahfoudz himself who drew attention to the abandoned compounds and advised recent converts from Diatock to seek permission from the chief of Kujube to settle. Conversion and peasantization combined to produce centrifugal effects. Hence, while Mahfoudz's village was a site of annual pilgrimage to mark his birth date, production took place elsewhere on family farms.Footnote ⁷⁹

The reality of a relatively underpopulated landscape, the legitimating discourses of devotion through hard work and the importance of a new concept of community based on the umma produced a credible fusion of thought and practice. For local intellectuals, it was as if the anomalies of history had been righted. Whereas Fodé Sylla had wrought demographic chaos, Mahfoudz had demonstrated how Muslims could conduct their lives in a way that built communities and bettered the lives of individuals.Footnote ⁸⁰

ARMED INSURGENCY: A HISTORICAL REPRISE?

By the early 1980s, rural communities in the Casamance, and to some extent the Gambia, were faced with two challenges. The first was creeping desiccation as reflected in the shrinkage of the wetlands. The changing hydrology contributed to the emergence of an alternative migratory pattern centred on the cities. Although rice farming remained important, Jolas came to depend heavily on income streams from Dakar and Banjul.Footnote ⁸¹ The second challenge was the rise of Casamance nationalism which threatened the conditions under which the peasant option had prospered. These processes were intimately connected in the sense that Jolas who travelled to Dakar became more acutely aware of their marginality in a country where Wolofization was being actively promoted. The relatively better-educated Casamançais felt the sense of exclusion all the more acutely and, in textbook fashion, gave nationalism both a vocabulary and a mythology.Footnote ⁸² They claimed that their region had never really been a part of greater Senegal, and that the two halves of the country were culturally divorced. The authorities blamed the problem on the very existence of the Gambia which divided the country in two, frustrating their efforts at nation-building.Footnote ⁸³

In 1990, the attack on Selety signalled the start of a low-intensity war that is estimated to have killed between 3,000 and 5,000 people and provoked a substantial flight of refugees into the Gambia and Guinea-Bissau.Footnote ⁸⁴ In 1992, the MFDC fractured into a Front Nord and a Front Sud, with the former effectively reaching a deal with the government, while the latter resolved to carry on fighting. There were relatively few pitched battles during the 1990s because rebels preferred hit-and-run tactics. These assumed the form of nocturnal attacks, targeting civilians plying the routes between northern Senegal (via the Gambia) and Ziguinchor. This activity blurred into banditry as rebels typically took what they could before releasing the passengers. In between attacks, the guerrillas secreted themselves in the broken terrain. The analysis here is concerned with the operational zone of the Front Nord, which had its headquarters at Diakaye in northern Buluf. Along the border, the initial peace accord produced an unusual dynamic. Evans writes that the Front's leaders agreed to stop fighting ‘in informal exchange for which they were allowed to retain de facto control of much of the northwest of Bignona department’.Footnote ⁸⁵ In practical terms, this meant that fighters took up residence along the border and, provided they did not trouble the soldiers at Diouloulou, were free to carry arms.

Initially, there was some sympathy for the MFDC. Villagers near to the rebel camps contributed food and money more or less voluntarily. But, over time, the relationship became strained. Rebel demands waxed more onerous, while their behaviour became more capricious. Hence young women in the villages surrounding Mahamouda were expected to turn out at ‘parties’ and to provide sexual services. The chastisement of elders made the rebels exciting, but local youths increasingly found themselves on the receiving end. For example, when those of Touba were discovered dancing to a popular song by Youssou Ndour, the rebels forced them (at gunpoint) to dance to the same track all the way through to daybreak. This was their punishment for daring to listen to ‘Senegalese music’.Footnote ⁸⁶ When the villagers of Kujube sent an appeal to the army at Diouloulou, following a series of physical assaults, the soldiers promised to talk to the rebels, but no more was heard.Footnote ⁸⁷ This led to some conjecture that the army and the rebels were in cahoots which, in a sense, was true.

Given the troubled history of the border, it was perhaps inevitable that memories of Fodé Sylla should be invoked. On the face of it, Sylla and the MFDC could not have been more different. After all, the one was allegedly prosecuting a holy war, while the other was pursuing the secular goal of self-determination. Again, there is nothing that one could equate with the slaving activities of Fodé Sylla. Unlike some modern rebel movements, such as the Lord's Resistance Army in Uganda and the RUF in Sierra Leone, children were not abducted to work for the MFDC. Nevertheless, there are some striking historical resonances. The rebel rank-and-file required peasants to furnish the means of their subsistence. As time went on, they reputedly seized cattle, crops and money at will. Like the razzia of old, rebel activity followed the rhythms of the agricultural cycle, with seizures being heaviest after harvest. What is even more striking is that the MFDC tended to gravitate towards precisely the same areas in Narang and Fogny Jabangkunda that had been conflict zones during the era of Fodé Sylla. One camp was established at Bandjikaky, close to the latter's former military base.Footnote ⁸⁸ Others were located in the forested areas of Narang where Sylla had met his match – in particular around Mahamouda and Touba.Footnote ⁸⁹

Whether this was purely coincidental or there was an underlying logic is worthy of closer consideration. The decision to set up bases in the forest made perfect military sense because of the opportunities for concealment and escape into the Gambia. The forests also provided just about the only valuable commodity in the shape of timber. The flight of the Senegalese forest guards from Narang enabled peasants to chop firewood, manufacture charcoal and sell wood for fencing fields. Most of it was sold in the Gambia, with the proceeds being used to pay for imported rice, given that families were unable to grow their own domestic requirements. The rebels demanded fees from those who felled trees, as well as being directly engaged in the trade themselves.Footnote ⁹⁰ Although MFDC ideologues were inclined to romanticize the Jola attachment to sacred groves, these forests had no great cultural resonance for the rebels, who felled trees at will. What was equally important for the choice of location was the underlying demography. As we have seen, the relative lack of people had created opportunities for land-hungry farmers from Buluf. Casamance nationalists made much of the claim that the belated application of the 1964 National Domain Law had permitted the alienation of land to nordistes and to French tourist operators.Footnote ⁹¹ What is forgotten is that the same legislation facilitated Buluf–Jola colonization of the northwest Casamance. It seems likely that the rebels followed the trail of the settlers in the belief that they would be natural supporters, unlike resident Mandinka populations.

However, the synergy between conversion and peasantization did not make Jola settlers ideal receptors in the long run. Although Marut is correct that the MFDC contained Muslims as well as Christians, the taking of alcohol and a perceived lack of sexual decorum alienated many devout Muslims.Footnote ⁹² Moreover, the MFDC's championing of the symbols associated with Jola religion was at some odds with the perception of local Muslims who saw peace as having come at precisely the moment in the past when the Jolas had turned their backs on ‘idolatry’. Even to nod at traditional practices was, for some, tantamount to inviting a return to anarchy.

The insecurity created by the rebel presence led to a renewed flight across the border in a by-now-familiar fashion. Because of its border location, Darsilami was a major recipient. Others headed for towns in Kombo that had served as places of sanctuary in the later nineteenth century, notably Bakau. Most never registered as refugees, but gravitated to kin and friends, placing a burden on already strained family resources. The rebels did not cross the border themselves because of the need to remain on the right side of the Gambian authorities. President Yahya Jammeh, who is a Jola of Casamance extraction, was sympathetic to the MFDC, but wished to avoid a confrontation with the Senegalese and anything that might jeopardize the Gambian tourist industry. Hence when the looting of stores in Senegalese Dimbaya strayed into Gambian territory, the rebel commanders realized their mistake and returned the goods.Footnote ⁹³

The manner in which people in the border zone responded to violence in the 1990s reflected their perception that history was, to some extent, repeating itself. The case of two associations illustrates some of the ways in which actors dealt with the challenge. The first is that of Kayong Kalorn (KK),Footnote ⁹⁴ which has worked at constructing a sense of Karoninka identity inside the Gambia. Although an earlier association had been founded in Ziguinchor in 1974, KK was the initiative of second-generation Karoninka immigrants in the Gambia. It has been active since its formation in 1994, but only on that side of the border.Footnote ⁹⁵ It has produced a manifesto, pioneered a Karoninka-language discussion programme on Banjul radio and holds meetings in rotation in those Gambian towns with significant Karoninka populations.Footnote ⁹⁶ The leaders have sought acceptance for the Karoninkas as indigenous Gambians, despite their homeland being well inside the Casamance. This is justified on the basis that Karoninkas had been migrating to Kombo long before any colonial borders were drawn.Footnote ⁹⁷

KK has pitted Karoninka identity against both Mandinka and Jola referents, but in contrasting ways. The Karoninkas are said historically to have enjoyed good relations with their Mandinka hosts despite being non-Muslims. Indeed, relations of amity were built around the renting out of palm trees to Karoninka tappers. In the later twentieth century when increasing numbers of Karoninkas converted to Islam this resulted in the adoption of a Mandinka identity. KK's mission has been to persuade its constituency that it is possible to be both Muslim and Karoninka. But the intention was never to pursue an anti-Mandinka agenda. By contrast, KK has been rather less well disposed to the Jola, who are depicted as an unrelated ethnic group with whom the ‘Kalorn’ have been erroneously lumped.Footnote ⁹⁸ In areas where high levels of immigration from Buluf became a matter of concern to Mandinkas, the Karoninkas have tended to side with the latter. The MFDC insurgency became a crucial factor in that many Karoninkas were inclined to blame Jolas for having brought trouble with them.Footnote ⁹⁹

At about the same time, a Fogny Jabangkunda association was formed in the Casamance. To a greater extent than its Karoninka counterpart, its origins lie in recent experiences of violence. It was established following a series of incidents in the vicinity of Bandjikaky. After a number of people died from wounds inflicted at the hands of local rebels, a meeting of people from surrounding towns was convened with the objective of defending the various communities against further attacks and expelling the rebels. Interestingly, the dormant memory of precolonial Fogny Jabangkunda was invoked as a rallying point for diverse communities living between the international border and Kafountine. The sub-text was that the indigenes had joined forces against Fodé Sylla's invasion and needed to close ranks again. Settlers from Buluf had threatened to upset the local power structure in the colonial period, and they were thought to be encouraging rebel activity once again. When the rebels were finally expelled after 2000, the association continued in the shape of a cultural organization that arranges a cultural festival in different villages in rotation.

In short, the mobilization of Jola identities in the Casamance has been rebuffed both by Karoninkas – denying membership of such an ethnic group whilst claiming Gambian citizenship – and by Mandinkas and Jolas in Fogny Jabangkunda who have perceived the MFDC as a cover for Buluf–Jola domination. The identity politics that is playing itself out is rooted in readings both of nineteenth-century history and of migration patterns in the twentieth century. These webs of interconnectedness underline the importance of not being mesmerized by the presentist discourse of MFDC nationalists, one that is not terribly strong on history despite its claims to be righting past wrongs. The quotation with which this article began underlines the fact that local intellectuals have a clear understanding of their troubled history over the past 120 years. What I have sought to demonstrate is that the concatenations linking recent events with nineteenth-century precedents are broadly credible. In making a case for taking into account the longue durée, I have not been dealing with deep structures, mentalities or even encoded memories: this has been a paper about lived and living memory.