David Robinson
The Holy War of Umar Tal: the Western Sudan in the mid-nineteenth century

Oxford. Clarendon Press. 1985. 420 pages

Chapter 8
Maasina: Conquest and Revolt, 1862-4

In 1862 Umar embarked upon an expedition that he had not originally intended: the conquest of the Maasina Caliphate of Hamdullahi. The confrontation was full of ironies. It pitted a Muslim Fulbe jihaad against a Muslim Fulbe state only recently established by jihaad. The two societies had much stricter attitudes towards social mores than the regimes in Nigeria and Futa Jalon. And yet the Umarian threat enabled “Muslim” Hamdullahi and “pagan” Segu to surmount an old and deep antagonism, and it subsequently overcame the enmity that had always marked relations between Hamdullahi and Timbuktu, the “fundamentalists” and “cosmopolitans” within West African Islam. The struggle put a sudden end to the Maasina experiment in Muslim government and then to the Umarian jihaad itself 1.

The Maasina campaign did not enjoy the active or nominal support which the Islamic clergy had given to the jihaad against the “pagans”. Rather, it provoked strong controversy and opposition, and that opposition transformed the stunning Umarian victory of 1862 into the surging revolt of 1863. The Maasina sequence is consequently composed of three sharply demarcated phases: conquest and initial consolidation, the uprising and virtual elimination of the invaders, and the long, draining struggle of reconquest which Tijani initiated in 1864.

The controversy engendered an Arabic documentation which far outweighs the material for all of the “jihaad against paganism”. At the core are the briefs and correspondence of the principal protagonists: Umar and his secretariat, Amadu III and his counsellors, and Ahmad al-Bekkay, the dominant Kunta figure 2. The abundance of material does not make the problem of narrating and interpreting events any easier. Umar and al-Bekkay, the most prolific writers, were engaged in polemic. They wrote not only to the opposite party but also for the larger learned community and for posterity, and they assume a knowledge of the chronology behind their arguments. Another problem which affects this sequence of events stems from the enormous destruction which eliminated some important archives, especially those of the Hamdullahi Caliphate 3, and took tens of thousands of lives. Very few Umarian, Maasinanke, or Kunta participants lived to write or tell their story after the events. What the historian finds, then, are oral and written fragments from Maasina, and slightly more co-ordinated views from Timbuktu and Segu, the cities that stood just outside the fray. The fragmentation and discontinuity make the monographs written by French administrators at the turn of the century, on the basis of extant oral traditions, all the more critical 4. Recently collected tradition can also contribute, despite the complex chains of transmission involved 5.

A. Maasina Between The Jihads

At first glance the Hamdullahi Caliphate might seem impregnable. Only forty years old, it followed the social charter laid down by Seku Amadu. The centralized command from capital to province, the rigid stratification into classes, and the limitations on the movement of pastoralists were all in effect. A strong Maasinanke identity had developed inside the Middle Delta, corresponding to a reputation in West Africa for militant Islam 6. But there were serious fissures in the state and society governed by Amadu III in the 1850s—divisions known to Umar and without which he might never have attacked, and certainly never have defeated Maasina in one bold stroke.

The first fissure consisted of the ethnic and vocational cleavages within the society. Tuareg warriors continually raided the northern provinces of the state, while some Fulbe chafed at the restrictions on their transhumance and autonomy. Many Marka were disenchanted by the taxes on commerce and restrictions such as the one against the consumption of tobacco. A “pagan” regime like Segu was preferable to the fundamentalism of their fellow Muslims. The merchants in Jenne were particularly unhappy about the eclipse of their city, singled out as the example of urban corruption by Seku Amadu 7. In similar ways the Somono and Bozo fishermen, the slaves who produced much of the food, and the “pagans” who lived on the eastern fringes of the state resented the demands which the government placed upon them. Some of these “pagans” were Dogon and Tombo farmers who went by the name of Habe or “non-Fulbe” in the Maasina lexicon. Others were Fulbe pastoralists who practised Islam in only the most nominal way.

The second division ran through the dynasty. Seku Amadu died in 1845. His eldest son, Amadu Seku or Amadu II, ruled until his death in 1853, whereupon his oldest son took the reins of power as Amadu mo Amadu, or Amadu III. The second succession was hastily contrived and bitterly resented by older members of the Bari lineage, who appealed to traditions of collateral as opposed to filial inheritance 8. Amadu III could never be sure of the loyalty of the royal family, especially that of his father's cousin, Balobbo (see Table 1). At the same time the new caliph could not eliminate his influential rivals nor do without their experience in administration and war.

The third major split involved age, ideology, and learning. The oldest generation, advisers to Seku Amadu in the early years, died in the 1840s and 1850s. Alfa Nuh, the architect of the caliphal forgery and engineer of the 1845 succession, was about the last to go, in 1858 9. A second group, age mates of Amadu II, had grown up under the tutelage of the founding fathers and in the halcyon days of expansion, prosperity, and rigorous education. Some had studied with the Kunta. Some had subsequently joined the Tijaniyya cell started by the pilgrim Umar. They judged the conduct of the regime by their own training and the high standards of an earlier day. The final cohort was that of Amadu III, a group of men in their twenties who were first and foremost warriors. They had much less serious Islamic education and less commitment to the whole experiment of Seku Amadu, and they clashed with both of the older generations.

Table 8.1
Genealogy of the Bari Family

The Tijaniyya community of the mid-nineteenth century poses a problem for the historian. The destruction of lives and records in the 1860s makes it very difficult to establish the history of these Maasinanke after the Shaikh's passage in 1839. What have survived are short poems and statements about the community, associated with the name of Yirkoy Talfi but with few chronological references 10. They indicate that the Tijaniyya formed a significant portion of the second generation of scholars (see Table 2). Normally they would have succeeded to positions in the council and the secretariat, but their dispute with Seku Amadu and the Kunta made them suspect and put their religious and national loyalties in conflict with each other. By 1848 Yirkoy Talfi was engaged in a continuing polemic with Ahmad al-Bekkay. In 1852-3 the members of the community suffered some confiscation of property and imprisonment at the hands of Amadu II, and these practices continued under his successor. Amadu III needed these scholars at court, especially as he got into deep water in the exchanges with Shaikh Umar, but he could not trust them in the confrontation which increasingly absorbed his energies 11. For the Tijaniyya the Shaikh became an increasingly important guide and inspiration to the point where, perhaps as early as 1858, some were calling upon him to bring the jihaad to their land 12. They maintained contact with him, provided invaluable information during the mounting crisis of 1860-2, and probably welcomed the conquest of Hamdullahi 13. They apparantly did not, however, furnish military assistance to the campaign.

Table 8.2
Possible Members of The Maasinanke Tijaniyya Community a

  1. For details of references in this table see Sources: A.1; A.1.a; A.2.2; A.4 (for Barry); and B.3 (for Marty, Bâ and Daget and Monteil).

The chorus of criticism of the regime of Amadu III which emerged from the Tijaniyya and other scholars of the older generations becomes a statement of moral decline in the dominant traditions of Maasina. A representative version runs as follows:

Amadu III was still a boy. The laws which his father [and grandfather] had established at Hamdullahi, he had them lifted. Woman were able to promenade without veils, to ride around on horseback chewing kola nuts every afternoon. The traditional guitars came out. He abandoned the path of his grandfather, that is why God ceased to smile upon them . . . They did exactly what they wished, and that did not please God 14.

The indictment is also nurtured by the widely circulated letter which al-Bekkay wrote to Amadu III in 1854, to rebuke him for his intransigent attitude towards the European traveller, Heinrich Barth. The Kunta scholar said:

You depose the great and give responsibility to the incompetent. You exile the learned and surround your self with the base. You glorify the children of sin and demean those of good character. How do you want me to submit to your rule, when those who are already under it despise it? 15

Whether the criticisms amount to a conviction of moral decline is hard to say. Certainly Hamdullahi was unsettled, confused, and unable to call on the indigenous human resources which had made it so successful in an earlier day. The Tijaniyya colony added to the confusion. Certainly the society was in transition, but to what kind of future it is not clear. The comments of an older generation or an external arbiter like al-Bekkay, drawn from the naturally tumultuous time after the succession, should not be taken as objective nor applied to a whole reign.

In fact, the cleavages of mid-century Maasina might very well have been overcome by success on the battlefield. But Amadu III, despite his reputation for personal bravery, did not obtain great distinction in war. The province of Jelgoji fell under Mossi control 16. The Habe of the east and the inhabitants of Farimake, a province in the north-west, were suppressed with great difficulty in the mid-fifties. On the western frontier the results were yet worse. The intervention in Karta in 1856 had brought the humiliating defeat at Kassakeri and helped make Maasina's old vassal Sambunne into a hero of the jihaad 17. It was precisely on the Segu side that Amadu III was beginning to enjoy limited success, but only because the Bambara regime was distracted by the oncoming jihaad. The submission of Segu in 1860, an event which in earlier times might have been trumpeted as a great triumph, only served to crystallize the generational and ideological cleavages, to arouse the ambition of contenders for the throne and, above all, to make Umar determined to extend his mission in a new direction.

B. The Diplomatic Game

In the wake of Umar's victory at Koghe, Amadu III adopted a new attitude towards the jihaad. His alliance with Ali had turned to ashes and his advisers pressed him to drop the hard line in favour of a more conciliatory tone. In the next letter which he sent to Umar, he acknowledged that Maasinanke interests in Bakhunu and Segu had not justified the dispatch of armies against the Futanke. He would not, however, hand over the Bambara king, who had converted and was now his protege, and he expressed the hope that Umar would withdraw from Segu and leave him the task of consolidating the practice of Islam 18.

This response underlined the Caliph's ineffectiveness. The Maasinanke Tijaniyya could have helped him understand Umar's personality, make better arguments and write more correct Arabic, but they were not available for such a task. The claim that Ali had converted was quickly nullified by the Bambara “fetishes”, brought out to show the delegation from Hamdullahi 19. The confession of error fitted perfectly into the rationale which Umar was creating for the Maasina campaign. In fact, the Shaikh had probably decided to press his case against Hamdullahi by the beginning of 1861, during the movement of letters and armies around Sinsani 20. He would demand the surrender of Ali, reparation for the losses in battle at Kassakeri, Tio, and Koghe, and some sign of reform in the Maasina capital itself. He counted on the disarray in Hamdullahi, his appeal to the Tijaniyya and other learned Muslims, and the strength of his arguments to sway his opponents, but he was prepared to decide the case on the battlefield as well. As the leading Islamic authority and missionary in West Africa, he would not tolerate the clumsy resistance of Maasina. The rest of the Shaikh's actions in 1861 should be read in this light.
At some point in the rainy season of 1861 Umar sent a delegation to give his official response to Amadu's apology and appeal for withdrawal. Like the group sent from Sinsani, these envoys were instructed to direct their appeal to the whole Maasina court. The chief ambassador was Mamadu Cerno Alassan, a member of the prestigious Baro lineage of Hayre in Western Futa. He made a strong impression on the scholars of Maasina as he delivered his Shaikh's stinging reply:

you have acknowledged your wrong, you must pay reparations for attacks on Muslims and hand over the infidel king; if you refuse, we can submit the case to arbitration 21.

The arguments had some of the desired effect. The Segu chronicler said that tears came to the Caliph's eyes; some of the scholars were deeply moved. But about the ultimate response there could be little doubt: no reparations, no surrender of a guest, and no submission to arbitration, which could only embarrass Hamdullahi. The die was now cast, and the subsequent exchange of delegations only confirmed it 22.

Umar could not, however, move against the Caliphate without a careful written justification. Since he had waged the jihaad against “paganism”, he must now show that the Maasina ruler was in reality “pagan”, and persuade his followers and some external authorities to that effect. He turned routine administrative chores over to Amadu Sheku and devoted his energies to a brief for the bar of West African scholars. By marshalling the works of classical and contemporary scholars and excerpting from his correspondence with Hamdullahi, the Shaikh wrote a compelling case, the Bayaan or “Explanation of what happened between us and Amadu mo Amadu”, for convicting his opponent of “infidelity”. The first section reviewed the exchange of letters and the sequence of events, beginning in Bakhunu in 1855-6. Umar maintained that Maasina had no effective control over the contested areas and had made no serious effort to establish Islam in portions under its jurisdiction. He showed the overwhelming deficiencies in Amadu's knowledge of the faith, the Koran and the Traditions, and ended with the “idols” of the Segu temple and the admission of wrongdoing. The second section drew on the writings of al-Maghili, Uthman dan Fodio and Muhammad Bello to argue that Amadu's alliance with Ali and attacks on the “Muslims” constituted an overwhelming case for apostasy. The Caliph was no longer a Muslim and those who fought under his banner were polytheists, the term reserved heretofore for the Bambara of Karta and Segu 23. Umar worked on the Bayaan through most of 1861. He used its arguments in his letters to Hamdullahi, and he probably circulated it to Hamdullahi and other audiences. He had composed a crushing indictment to force compliance or achieve confrontation.

C. The Conquest of Maasina

While the Shaikh was intransigent in his arguments, he was quite cautious in his plans for the military campaign. He had defeated Maasinanke forces on three occasions, but he had never faced their full army at home. He could not assume that disunity in deliberation would translate into weakness at war, when the defence of the homeland was at stake. He waited until the hot dry season, when soldiers could move with maximum efficiency in the Middle Delta, and then mobilized every available resource. He took a large supply of powder, bullets, guns, and cannon. He reduced the Segovian garrison to about 1,500 talibe under the command of Amadu. Most of the other sons and his nephew Tijani would accompany him. All of the other soldiers, some 30,000, were poured into the campaign 24. It took about two weeks to assemble the convoy between the Niger and the Bani. When finally deployed, it had a somewhat different composition from the force which marched out of Markoya in 1860, reflecting the influx of sofas and the garrison duty around Segu (see Table 3 below).

Table 8.3 — The Umarian Army in 1862 a

New Name Old Name Garrison Duty in 1861/2
Ngenar regiment Ngenar Banancoro
Dugassu regiment Murgula Dugassu
Bamabugu regiment Yirlabe Bamabugu
Toro regiment Toro Koghe
Jomfutung or guard Jomfutung Segu Sikoro
Sofa (not a separate unit in 1860) Segu Kow
  1. This nomenclature emerges in BNP, MO, FA 5457, fos. 1-4 (letter of Makki and Tijani to Amadu Sheku after the Cayawal campaign. This source is to be preferred over the more conventional sources mentioned in chapter 7, table 1, because it was written soon after the battle of Cayawal and reflects the understanding at the time.

Maasina was waiting. Balobbo and a large force were based at Poroman, the garrison capital of Bobolo province on the southern frontier. Amadu III had an even larger army in nearby Jenne. Together these two bodies constituted most of the military might of the Caliphate some 50,000 cavalry and infantry, armed with spears, swords, bow and arrow, but very few firearms 25. The Umarian troops crossed the Bani and headed directly for the capital. En route they inspected the village where Ali had taken refuge, skirmished with Balobbo and seized his signal drum. On Saturday May 10, near the Bani, they encountered the combined Fulbe armies in a small wooded depression or cayawal, the name by which this epochal confrontation has come to be called. The battle raged throughout the afternoon and much of the evening, exacting a heavy toll on both sides without any demonstrable victor. For one commentator, the intensity of the struggle made the earth shake, while the sweat of the horses swelled the river to flood stage 26.
The examples of courage abounded. Two of the most frequently mentioned are found in a document written by Makki and Tijani to Amadu in Segu. In one instance Batu, a Massassi sofa who had enlisted in the jihaad in 1855, almost single-handedly fought off the Maasinanke advance. Covered with blood and dirt, he finally succumbed to the eighteen spear wounds in his body. His fellow sofas, on Umar's orders, then fought back to the point where they could recover his body and give it a proper burial 27. The story of the courage of Amadu III is even better known. Armed with the swords of his grandfather and father, as well as his own, he took personal charge of the fighting on the western flank, mowed down the talibe in an awesome display of will, and momentarily turned the tide in Maasina's favour 28.

On Sunday the Caliph opted for a strategy of siege. For four days he surrounded the jihaadists and harassed them with sorties. He sent messages throughout the land to announce his “victory” and bring in reinforcements for the final charge. But his strategy was not up to the standard of his bravery. The siege was not sufficiently tight to prevent Umar from getting local iron and making bullets to replenish his exhausted supply 29. A member of the Hamdullahi court, who had sought to dissuade Amadu from the confrontation, joined the Futanke forces and provided precious information on Maasina deployment 30. With him the Shaikh planned a surprise attack for dawn on Thursday morning, 15 May. All the talibe, accustomed to riding their mounts, would become infantry in order to use their guns more effectively. The cannons, useless in such close encounter, were packed on camels. Then the bonds of solidarity were tightened. In the words of Makki and Tijani,

“we bound everyone by compact and covenant until death. We swore allegiance unto death, that we would not flee regardless. We agreed that Segu Sikoro was far, that Nioro was farther still, that we would never flee even if we had to die, and that the Shaikh would indeed enter Hamdullahi 31.

The attack caught the Maasinanke by surprise. Mage provides a vivid impression of the discipline of the Umarians as they broke out of the siege:

Al-hajj kept advancing. He forbade his men to shoot, despite the intense fire of the Maasinanke and the hail of spears on their heads. Finally, at less than 50 paces, after the Maasinanke fired again, al-hajj raised his hands in the air and bellowed, “Awa! Awa! (to the front! to the front!)”.

Then came the shock of confrontation—violent, irresistible. The Maasina infantry was trampled and more than half of the cavalry fled 32.

Amadu III and his brother Mamudu sought to stem the tide, but to no avail. Mamudu was killed, while the Caliph had to be taken away, wounded, by his servants. They put him and his immediate family in boats and dispatched them by the Bani to Timbuktu.

Cayawal effectively ended Maasinanke resistance. It also exacted by far the heaviest toll of any single battle of the jihaad: some 30,000 Maasinanke dead and 10,000 Futanke 33. Superior weapons, organization and strategy had prevailed. Underneath these factors lay the contrast between the solidarity of the Umarians and the division of the other side. While it is probably not accurate to accuse Balobbo of treason, he made little effort to stop his cavalry's flight on that fateful Thursday morning 34.

On Saturday 17 May, after a long prayer outside the walls, Shaikh Umar entered the caliphal capital 35. The regiments filed into the city to carefully designated quarters while the inhabitants looked on. One of Yirkoy Talfi's students wrote a poem in honour of the occasion, and some members of the court voiced their submission. Most of the rest of the leadership followed suit in the coming days. The mujaahid himself settled in Amadu's palace and distributed the royal treasury to his followers.

The first order of business was to capture the Caliph. Alfa Umar Baila gave pursuit and caught up with the party in northern Maasina, not far from Timbuktu and the protection of al-Bekkay, in early June. Amadu III was killed, his family and counsellors imprisoned, and the family documents confiscated 36. How he was killed is a matter of controversy. Some sources, in order to absolve Umar, attribute the death to the Tuareg who blocked his way or to a caliphal slave settling an old score. The most likely version is the one Mage heard in Segu:

Alfa Umar wrote for instructions on his return journey, carried out the execution at Mopti and buried the body in a secret location 37. It is clear that the Shaikh wanted his antagonist dead but did not want to carry out the act himself. He did carry out the execution of some of Amadu's counsellors, including the former Lam Toro Hamme Ali, who had deserted from Segu, apparently because he believed that Umar was seeking new dominions to give outlets to his sons 38. Other members of the caliphal entourage were sent to prison in Segu.

The Shaikh also put a high priority on establishing once again the religious identity of Bambara Segu. In essence, he was repeating the brief of the Bayaan and driving for a sharp conviction. His expert witness was present in the form of Ali, the fugitive king captured in the aftermath of Cayawal. The mujaahid had Amadu Sheku bring the “idols” of the temple to Maasina in early August. He convoked a large assembly and made his point by interrogating his witnesses:

— What is this, next to us?
— These are idols.
And he [Ali] gave their names that they be known.
The Differentiator [Umar] then said to them [Ali and other Segu Bambara]:
— Now break them, crush them, and build mosques in all of Segu.
Ali said:
— You mock me. You alone can smash them and survive. Anyone else would not live to tell the tale.
Then the evil bowed their heads in shame, they did not look up or speak because of their lies.
Then the Unique One [Umar] rose up and crushed [the idols] with his powerful hand, imitating the action of the Elected One [Muhammad] at Medina 39.

Like Muhammad, like Moses, Umar had dared to grapple with the most prestigious and powerful rivals to Islam. Not only had he survived, he had won. The smashing of the “idols” in Hamdullahi was the final step of the jihaad against “paganism”.

The Shaykh, after settling scores and winning his case, gave some attention to the administration of his dominions. He repeated the installation of Amadu Sheku. After discussions with family and talibe, he had the Futanke and Maasinanke communities swear allegiance to his son as successor and ruler of all of the territory of the jihaad, from “Timbuktu to Futa” 40.

Umar did this with intention and visibility, in the full knowledge of Amadu's inexperience and the ambitions of the younger generation of Tal.

The August ceremony helps to make sense of the fragmentary impressions of the plans of the Shaikh in the last half of 1862. Between June and November, from the Mossi states to St Louis, indigenous and European observers give glimpses of a vast Islamic empire policed by perpetual jihaad. Mossi and Kong were called upon to submit 41. A messenger went to Dingiray in July to tell the community that Umar would soon come, while others visited Sierra Leone a bit later to prepare the inhabitants to elevate Islam and subjugate “paganism” 42. At the same time armed contingents were moving into the Sahel as far as Walata, demanding submission and tribute from chiefs, traders, and pastoralists alike. One of them briefly retained Aliun Sal, a mulatto lieutenant on an exploratory mission for St Louis. Sal, on the basis of what he overheard, claimed that the Shaikh intended to sweep down both banks of the Senegal and drive the “infidel” French into the sea 43. The lieutenant had always been hostile to the jihaad and probably expanded what he heard to increase his own importance, but his testimony cannot be completely dismissed in the light of the time and context. Umar, having accomplished his goals against “paganism” and apostasy, having installed his oldest son as the administrator of the empire, saw the jihaad continuing in every direction, whether by his example or though his own intervention. This portrayal is consistent with the Shaikh's neglect, in all of the conquered areas, of the task of creating enduring institutions of government and islamization.

This is not to say that Umar took no steps to control Maasina. He confiscated the property of many who had been killed at Kassakeri, Tio, and Koghe, and distributed it to his followers. He built a heavy stone fortress for administration and defence and refurbished the walls around the city. For this work he called on the local citizens as well as the Futanke. He made a few appointments of provincial commanders, but left most of the incumbents—after they had sworn allegiance in place. He allowed Balobbo and other members of the royal family to stay in the capital under limited surveillance. One of his most important decisions was to dispatch a contingent to Timbuktu in June, to restore the arrangements whereby Hamdullahi received a share of the city's taxes and customs. The fragmentary sources suggest that the new landlords allowed less autonomy than the previous regime, which provoked Ahmad al-Bekkay to intervene with the Shaikh on Timbuktu's behalf 44.

The sources permit some glimpses of the jihaadic community in the last half of 1862. Umar was surrounded by his personal bodyguard. He was probably less accessible to the talibe than before. He did receive his principal generals and administrators regularly. He established an archive and secretariat in the palace, where he placed the documents seized in the capital and from the Caliph's canoes and put some of his scholars to work. In particular, he charged Yirkoy Talfi and another local Tijaniyya, Alfa Umar Mamadu-al-Awsi, to go through the materials and respond to letters 45. The most challenging correspondence came from Ahmad al-Bekkay, in reference to the status of Timbuktu and his own relations, past and present, to the jihaad. In late 1862 Alfa Umar Mamadu and Yirkoy Talfi composed substantial treatises, patterned on the Bayaan, in which they documented Kunta deception, contradiction, and hostile action towards the Umarian enterprise. Yirkoy Talfi was particularly articulate in his brief, which he entitled “What makes al-Bekkay weep.” He made good use of the Kunta leader's letters to Amadu III and Ali, which had now fallen into Umarian hands, to argue that Bekkay had encouraged “pagans” to resist Muslims 46.

Makki, the senior son in Hamdullahi, became the patron of this scholarly endeavour. His name appeared as the owner or commissioner of documents and occasionally as author. Along with Tijani he wrote the account, used extensively above, of the battle of Cayawal, and he recorded the names of friends and enemies during the revolt of 1863 47. He served as a kind of deputy leader, more accessible than his father and more aware of day-to-day events. He was probably being tutored to assume the governorship of Maasina under the overall jurisdiction of Amadu Sheku. The sources originating in Bandiagara at the turn of the century, under the influence of Makki's younger brother Agibu, certainly give this impression 48.

D. Revolt and Demise 49

While the internal sources suggest that the Shaikh's yoke rested very lightly on Maasina, a precious document from the under-represented Maasinanke perspective presents a different picture 50. This Bayaan, the “Explanation of what happened between the Commander of the Faithful Ahmad and al-hajj Umar”, emphasizes the abuses inflicted on the royal family, the execution of chiefs and deportation of others, the confiscation of property and the forced public works. It claims that the women of the royal entourage were sent to the Segu harem while Aja, the wife of Seku Amadu and a leading force in the Caliphate, was enslaved and sold, with the proceeds going to the invaders. This insult to the Maasinanke heritage provoked the intervention of al-Bekkay:

— It has reached me that your agents have subjected her [Aja] to the treatment of a slave, justifying their conduct by claiming that she is a pagan. Is there anyone among all the Fulbe there, let alone [the family of] Seku Amadu, who is pagan? 51

The abuses of 1862 were not yet sufficient to generate a revolt. What was required was the crystallization and convergence of three forces: widespread popular resentment against domination, leadership by members of the Bari family, and committed support and co-ordination from outside. The first two elements had been present in Karta and Segu, in uprisings that finally failed. The last, in the form of the mastermind of Ahmad al-Bekkay, would make this revolt succeed in its immediate goal of eliminating the invaders. The three elements came together in the early months of 1863.

The general resentment of the Maasinanke, accustomed to independence and proud of the role which they had played in institutionalizing a certain form of Islam, grew as time went by. The taxation, confiscation, and foreign occupation constantly reminded them of their weakness and subordination. The event which finally provoked the internal and external leadership came in February. The Shaikh brought Amadu Sheku to Hamdullahi for a second time. He may have wanted to communicate more Tijaniyya mysteries. He may have wanted to provide more clear instruction in administration before resuming the jihaad in other areas. He had received a letter, accompanied by a gift of seven pure-bred horses, in which al-Bekkay called upon him to restore Maasina to the rule of the Bari, in the person of Balobbo 52. Umar rejected the counsel and specifically entrusted the administration of Maasina to his eldest son. Although Amadu soon left to put down a smouldering rebellion in Segu, never to return, the February installation finally convinced Balobbo and his relatives that they had no political future in the new order. They communicated to al-Bekkay their desire for his support in exchange for an undefined Kunta hegemony over the Middle Delta. This was just the encouragement which the shrewd Timbuktu leader needed: an internal movement to coalesce with his forces and a promise of power in an area which had consistently eluded his grasp.

Al-Bekkay had been in direct correspondence with Umar and the secretariat since the expedition to Timbuktu the previous June. In four letters written in late 1862, copies of which constitute a small corpus found in many West African libraries, the scholar adopted a cool but conciliatory tone towards his arch rival 53. He apologized for some earlier slanderous remarks. He interceded for the caliphal family, especially Aja. He argued for the autonomy of Timbuktu and the Tuareg. He withdrew from Timbuktu, and thus from submission to the jihaad, until Umar clarified their relationship: surely the Shaikh did not expect the Kunta leader to come to Hamdullahi and swear allegiance, like the political leaders of Maasina? Al-Bekkay also went to some length to show his distaste for jihaad and his fear of its consequences:

The companion of Muhammad Bello said to me:
— Why do you not declare jihaad, so long as you are capable of it, and thereby forgo the need to beseech tyrants and robbers?
I said to them:
— You know the virtue of jihaad. But jihaad also leads to kingship, and kingship to oppression, and our state as it is now is better for us than jihaad, and safe from the error to which it leads 54.

It was partly this reluctance to engage in direct political and military activity that had made the Kunta acceptable to Muslim and non-Muslim leaders and successful in developing a network of disciples and trade arrangements in the Western and Central Sudan.

Now al-Bekkay used the networks to declare a war of his own. Yirkoy Talfi's treatise had not “made him weep”, Umar's Bayaan had not convicted him at the bar of judgement. He cast over board the historic Kunta position of non-involvement, or peripheral involvement, and used his carefully nurtured contacts in one supreme effort to wage jihaad against jihaad. At one level he was an ambitious man who feared a talented rival and seized the opportunity presented by Maasinanke resentment. At another he was a well-trained traditional scholar fearful that an intolerant ideologue might destroy the material and social base of the Niger region.

Unlike his antagonist, al-Bekkay did not prepare a long justification of his actions. He labelled Umar “the impostor who does evil and against whom God has ordained jihaad55. He sent his nephew Sidia to co-ordinate operations with the Fulbe. In March he dispatched contingents of Tuareg and Kunta into Maasina and wrote to his partisans throughout the Middle Delta. In the case of chiefs who could not read, he reputedly sent a given number of pebbles to indicate how many days to wait before rising in revolt 56. His awesome reputation worked to his advantage, and in itself served as a kind of counterpoise to the now awesome prestige of Umar. This Kunta power is often reflected in the griots' accounts of the revolt:

The uncles of Amadu III [viz. Balobbo and others] were unhappy . . . and went to Timbuktu to see Kalle [viz. al-Bekkay], a person skilled in the occult sciences, to ask him to capture the soul of Shaikh Umar. For eight days Kalle worked on the problem, preparing [Maasina's] revenge. He finally killed the snake who had been supporting the army of Shaikh Umar. When Umar learned that the Kunta Kalle had killed the snake, he rose and spoke to Alfa Umar and the others:
— They found a cleric who killed the snake, we will advance no further 57.

Umar did not discover the plans until late March. At that time one of his Maasinanke followers intercepted al-Bekkay's request for information about the deployment of Futanke forces 58. The Shaikh convoked Balobbo and his relatives and forced them to acknowledge their participation in a plot, but he did not yet appreciate their determination. He apparently believed that he could still dissuade his adversaries from any action. Consequently, he did not execute the Bari but put them in shackles under guard. He sent Yirkoy Talfi and other prestigious Maasinanke to reason with the leaders of the rebellion in the northern province of Gimballa. The negotiations failed in April, at about the same time that Balobbo and one other member of the Bari family escaped from Hamdullahi. They would provide the royal standard for revolt 59.

At this time the Shaikh finally realized the inevitability of the uprising. He executed all surviving members of the dynasty in the capital. He sent contingents to Gimballa to counter the growing coalition of forces. With the help of the Habe they captured a number of Bari and brought them to Hamdullahi for execution. By now it was May and the mobilization synchronized by al-Bekkay was going into effect. Amadu Sheku could no longer get guns and powder through by water or land, so that the precious weapons advantage which the jihaad had always enjoyed was slipping away. Umar still controlled most of the south and centre, but he was isolated, cut off from his traditional support in the “west” and from his potential Habe allies in the east 60.

In early June two stunning defeats tightened the noose around the Futanke. Alfa Umar Baila, after beating the coalition in Gimballa, marched against Timbuktu to punish al-Bekkay. There he took a considerable booty and a number of hostages, and then headed back to Hamdullahi. Some of the troops prevailed upon him to swing by the rich pastures near Lake Debo, where Maasinanke cattle were concentrated in the hot dry season. At a spot called Mani-Mani the army was ambushed and annihilated almost to the last man. Mani-Mani ended all Umarian influence in the north and unleashed a deluge of rebellion in the centre and south 61.

The coalition soon destroyed several garrisons and pressed in on the capital. The Shaikh sought to break out of the shrinking circle by sending out his best remaining troops under Alfa Uthman, the hero of Murgula and Bangassi. At the village of Sege, in nearby Kunari province, the entire force was demolished by Balobbo and Sidia 62. At this point Umar no longer had the resources in men and weapons to even attempt to leave Hamdullahi, while Amadu Sheku, soon to be tied down by revolts in Segu and Sinsani, was in no position to bring relief 63.

The Maasinanke and Kunta invested Hamdullahi from the middle of June to early February 1864. The besieged, perhaps 10,000 at the outset, were forced to eat horse, donkey, and dog flesh and finally to consume human cadavers 64. Umar imposed a draconian discipline, executing all caught trying to escape, but he could not stem the flow of people and information to the besieging forces nor raise the energy and morale of the Futanke.

His own morale was none too high. He wrote a brief note to his son and successor which can be dated, by internal evidence, to this period of haunting isolation, hunger, and anticipation of death. Dictated to a disciple caught between “sleep and wakefulness”, it reads as follows:

This then is from the Shaikh, Umar ibn Sa'iid, to his son who drinks from the wells of grace, Ahmad.
The reason for this letter is what is happening as a result of enemies near and far, white and black. He [Umar] is not afraid of it, he is in its grip like a pure slave. He thinks that it is now decided ... I see all our land, even to the whites [Moors] and their learned and their saints giving allegiance to Ahmad, the righteous and just prince 65.

It is curious that the sources mention no effort to send for help during the first seven months of the siege. Perhaps Umar did not know of the extent of revolt in the Segu heartland, or the extent of the blockade between Sinsani and Jenne. The flow of information in, like the flow of food, had been cut off. Perhaps unsuccessful attempts were simply not recorded. In any case, the rapidly weakening community did succeed in January 1864 in smuggling out Tijani to secure aid among the Dogon, Tombo, and Fulbe who lived east of Maasina and harboured a sharp antipathy to the old ruling class 66. Tijani was not able, however, to return with reinforcements before the coalition assaulted the gates of the city and drove the survivors into the palace in early February. The end was very near.

During the night of Saturday 6 February, Umar and a small party of about 100 relatives, talibe and sofas escaped. How they escaped is the subject of debate. What seems likely is that the coalition, after lighting fires around the palace, failed to watch closely at one exit. They discovered the error the next morning, killed those who remained behind and pursued the refugees to the east. It was at this moment that the first signs of the fragility of the alliance emerged: Balobbo and Sidia disputed the right to enter the palace first 67.

The scene now shifted to the high ground and cliffs of the Habe, where the Shaikh hoped to meet the troops mobilized by Tijani. The exhausted band reached Goro and Degembere, small villages of Tombo “pagans” whom they might once have despised and destroyed. Balobbo and Sidia pressed hard on their heels, forced the Tombo to expel their guests, and drove them into the cliffs on the night of Wednesday 10 February. The Umarians used their position and precious ammunition to fend off the Thursday attacks. On Friday morning most of the talibe and sofas came down, surrendered, and transmitted a message from their chief: the Shaikh would give up the fight if his pursuers would allow him one more day to pray and prepare. When Balobbo and Sidia discovered, perhaps from an Umarian informer, that the offer was a ploy to gain time for Tijani's return, they pushed up the cliffside and set fires at the mouths of the caves 68. The end came quickly.

Exactly how the end came and what happened to Umar's body have been the subjects of continuing controversy. Suffocation, explosion, suicide, mutilation, disappearance into another world—all have their advocates. The Maasinanke version is rather straightforward: the fire suffocated the survivors or exploded their gunpowder, whereupon the attackers recovered and buried the Shaikh's body in an unknown location. This approach protects the dignity of the attackers. It also gains a certain revenge for the execution and secret burial of the last Maasinanke Caliph 69.

The Umarian versions are legion, stemming from fragmentary observation and variety of motivation. For the true disciple the Shaikh should have some control over his manner of dying, rather than exiting by suffocation or explosion. Unlike the traditional warrior, he could not take his own life. He should “escape” in order to inspire future generations, but his body should also be present at an earthly shrine for the faithful. Mutilation of that body could be attributed to the barbarity of the enemy, but no limb should actually fall into their polluted hands.

The historian can probably conclude that the fire forced the survivors out of the caves, that it finally ignited the dwindling powder reserves and that the explosion killed Umar and the approximately fifteen persons who were with him. The blast may have dismembered the bodies and made them unrecognizable. Some pieces were then taken as “trophies” by the attackers, while others were buried at the Bandiagara palace of Tijani, where they formed a place of pilgrimage 70.

E. Reconquest and Exhaustion

Tijani was able to gain early access to the caves because he arrived on the scene within twenty-four hours of that fateful Friday morning. Supported by considerable numbers of Fulbe, Dogon, and Tombo, and aided by the increasingly open divisions between the Kunta and Maasinanke, he defeated the coalition and quickly went on the offensive. Within a year he secured the southern and eastern edges of the Middle Delta. In the process Yirkoy Talfi and al-Bekkay were both killed 71.

Over the next twenty years Tijani engaged in a three-cornered struggle for control of the inland delta 72. Around a core of perhaps 100 talibe and sofas who had survived the siege and flight, Tijani built his administration and army. The bulk of his troops came from the Habe and Fulbe of the east who had always resented Maasinanke domination. Tijani made his capital at Bandiagara, in the midst of these allies and near the hallowed cliffs of Degembere. Bandiagara allowed him to control areas as far east as Hombori and to compete for the domination of Maasina. As long as his indigenous allies provided support, Tijani was the strongest contender. He effectively used Fulbe cavalry and the canoes manned by his Bozo and Sorko supporters. He transferred many of the people he conquered into the Kunari region, where they joined the Habe in providing the grain to feed the army. Every year he spent at least six months fighting; much of the rest of the time went into preparations for the following year's campaign.

Map 8.1 — Maasina, the Umarian Conquest and Reconquest
Maasina, the Umarian conquest and Reconquest
Sources: C. Monteil, Djenné; Underberg; Briquelot; Menvieille

The pattern of annual expeditions eventually exhausted the inland delta. Maasina increasingly resembled a devastated march caught between forces which had made warfare into a way of life. The Maasinanke Fulbe, led by members of the royal house, had their greatest strength in the south, where many had taken refuge after Umar captured Hamdullahi. Their first base of operations was Jenne. They subsequently retreated to Fion, where Bobo allies provided protection with their poisoned arrows. Balobbo was the first leader, while Amadu Abdul, a cousin of Amadu III, took over in the 1880s. The Maasinanke were the weakest of the three contenders for power, but they could always rouse some support by appealing to the residual regional loyalty. They quarrelled constantly with their presumed allies, the Kunta of Timbuktu, over the arrangements made at the inception of the revolt in 1863: were the Kunta to succeed to the throne of Maasina, or had it been agreed that they would control some portion of the kingdom under Bari rule?

Ahmad al-Bekkay appointed his nephew Bekkay Ntieni to co-ordinate the Kunta, Tuareg and Fulbe forces in Maasina. Ntieni made his base in Tenenku in the west. In the 1870s his son Abidin led the campaigns, but in the 1880s he was ousted by another Abidin, the son of Ahmad al-Bekkay. This Abidin soon retreated into Farimake as Tijani gained strength in the central part of the inland delta. It was left to Tijani's successor Muniru to wipe out the last vestiges of Kunta control in the north-west in 1889.

The lines on Map I suggest roughly the areas that corresponded to different periods of Tijani's and Muniru's hegemony. They are not intended to suggest effective administrative control, but only a kind of preponderance of influence.

Table 8.4
Selective Kunta Genealogy

Until the very end it was possible for any one of the contenders to sweep into the domain of another for a punishing raid, especially during the hot dry season when cavalry could move easily. In the absence of further research, it is difficult to estimate the cost in lives of the Umarian reconquest. It is probable that the figure rivals the appalling casualties for the 1862-6 period (see Table 5), which would mean that Maasina suffered far greater losses, in absolute and relative terms, than any other area of the jihaad. Even such a figure would not show the disruption brought on by war, flight and the policy of forced migration implemented by Tijani.

Umar's nephew was the only member of the second generation of Tal who continued to fight in the fashion of the original jihaad. It is ironic that he waged this fight with by far the smallest number of Senegambian Fulbe. During the struggle he took the title of Commander of the Faithful and had his chronicler, Abdullay Ali, make the reconquest into the sequel to Umar's jihaad 73. He also possessed a kind of last will and testament of the founder: some objects that Umar had transmitted to one of his followers during the last days at Hamdullahi or Degembere. These objects included a metal box, two small books, a cap, a talisman, and probably Umar's seal 74.

Amadu Sheku obviously did not accept Tijani's claims. He pointed to the several occasions when Umar designated his successor. He asserted, probably with some justification, that Tijani had confiscated the sacred objects intended for the oldest son. He was never able, however, to establish any control over his cousin. The records suggest that there was almost no contact between Bandiagara and Segu in the late nineteenth century 75. Amadu did eventually achieve a symbolic victory, but it came four years after Tijani's death. At this time he was a refugee from the French who were gobbling up the old Umarian capitals. When Nioro fell in 1891 Amadu fled to Bandiagara, had Muniru deposed by the talibé community, and was acknowledged by all as Commander of the Faithful. When the French arrived in 1893, Amadu fled to the east again. This time he and his followers called the act hijra, the movement away from pollution in imitation of the Prophet and the Senegambian migration of 1858-9. This flight became part of the holy history of the Umarian movement, and it appears just behind the Segu 2/anonymous version of the jihaad in many West African libraries 76.

Table 8.5 — Approximate Casualties and Prisoners in The Maasina Campaigns, 1862-6 a, e

  Date Battle Umarian toll Opponent toll
    Statement on killed Figure or estimate Statement on killed Figure or estimate
Conquest 5/62 Konihu     some 100
5/62 Cayawal extremely heavy 10,000 b extremely heavy 30,000 b
Revolt 4/63 Gimballa heavy 500 heavy 500
5/63 near Hamdullahi heavy 500    
  Jenne heavy 500    
6/63 Mani-Mani very heavy 3,000 c some 100
  various heavy 500    
  Sege very heavy 2,000    
6/63-2/64 siege of Hamdullahi extremely heavy 12,000 d    
Reconquest 2-6/64 various very heavy 1,000 extremely heavy 5,000
early 65 various very heavy 2,000    
early 66 various very heavy 2,000    
30,000   39,700
  1. The figures on the campaigns through February 1864 come from the whole body of sources used in chapter 8. The single most helpful one is Mage.
  2. The estimates on Cayawal run very high. Ouane (Enigme, p. 183) gives 50,000 Masinankes and 15,000 Umarians. I have opted for what I consider a reasonable figure.
  3. I have here followed Tauxier (Peuls, p. 198, with no indication of his source), rather than Mage's figure of 15,000.
  4. I have here taken the figures offered in two widely separated interviews: with Hamady Sayande N'Diaye at Seno Palel (Futa Toro) on 14 Feb. 1974, and with Ba Cekoro Kulibali in Segu on 15 Aug. 1976.
  5. For the 1864-6 campaigns I have followed the indications in Bandiagara/Abdullay Ali and Briquelot, “Notes”.

1. The only internal chronicle which makes a claim that the Umarian jihaad continued after the death of Umar in 1864 is Bandiagara/Abdullay Ali. And here the claim takes the unusual form of Tijani's effort to retake the Maasina initially conquered by his uncle. See below, note 73.
2. See chapter I, section D. Hamdullahi provided the least important contribution to the debate, or less of its contribution has survived. Of the internal narratives, Segu l/Y. Saint-Martin, Segu 3/Cam and Bandiagara/Abdullay Ali are the most consistently valuable. Nioro 2/Adam has some important detail.
3. Most of the archival destruction occurred in 1863-4 and during the reconquest of Tijani.
4. The important monographs are those by Capt. Underberg, Segu Resident in 1890 (“Notes sur l'histoire du Macina”, in ANS IG 122, nos. I and 2); Capt. Bellat, Resident at Sinsani in 1893 (“Renseignements historiques sur le Sansanding et le Macina”, in ANS IG 184); Capt. Briquelot, Resident at Segu (“Notes sur l'histoire du Macina”, 1893; in the main he copies Underberg); and Capt. Menvielle, Resident at Bandiagara, 1896 (“Notice sur les états d'Aguibou” ANS 15675; his report is extensively quoted in Tauxier, Peuls). See also ANM ID 47, “Renseignements historiques sur le Macina, 1864-75”. 5. The fullest collection of oral tradition from Maasina, without citation of informants, is Bâ and Daget, Empire peul, I. Diarah (“Maasina”) has worked with Bâ's materials for the 1850s. Brown (“Caliphate”) has extensive oral material, with citations, as does Ibrahima Barry (“Royaume Peul”). Barry permitted me not only to read his thesis but also his interview notes for Allaye Mamoudou Sekou of Hamdallaye (Christmas 1975), Sekou Bella Koita of Jenne (Christmas 1975), Abou Bokary Koita of Sofara (Christmas 1974), Tahirou Cisse of Tenenkou (Christmas of 1974), and Mamoudou Cisse of Mopti (Christmas 1974). In August 1976 I was able to interview a number of important informants in the Maasina region: Bokar Sango at Segu on 14 Aug. 1976; Bougouboly Alfa Makki Tal at Bandiagara on 19 and 20 Aug. 1976; Alfa Baba Thinbely at Bandiagara on 20 Aug. 1976; and Amadou Oumar Bâ at Bandiagara on 20 Aug. 1976. Most of the contemporary French observations (from the Bakel and Medine reports in ANS 13G 167, 177 and 15G 108) are very confused.
6. Cf. Person, Samori, I, p. 169, and II, p. 810; Raffenel, Nouveau voyage, I, pp. 358, 430, 493.
7. Johnson, “Economic foundations”, especially, pp. 483-4. On the ethnic and other cleavages at mid century, see Bâ and Daget, Empire peul, I, passim. For a view of the Caliphate from Jenne, see F. Dubois, Timbuctoo the Mysterious (1896), pp. 158-60.
8. Bâ and Daget, Empire peul, I, pp. 72, 248 ff., 285-6. Diarah, “Maasina”, pp. 300-32. See also Mage in ANS IG 32, piece 35, pp. 10-12, and Gaden's remarks in Tyam, Qacida, p. 185n.
9. The Ta'riikh Fittuga gives the death dates of several members of the older generation. Copies may be found at CEDRAB, Document 177/RB 1, and at IFAN (Fonds Vieillard, Maasina, cahier 4); it forms the basis for W. Brown, “Toward a chronology for the Caliphate of Hamdullahi”, CEA 8.3 (1968).
10. Most of the extant Tijaniyya material is found in the BNP, MO, FA: 5361, fos. 4-6; 5519, fos. 38-49, 92-7; 5575, fos. 215-17, 219-22; 5599, fos. 58-63; 5605, fos. 65-73; 5606, fos. 95-7; and 5681, fos. 91-6. The name “Yirkoy Talfi” should actually be applied to several brothers: al-Mukhtar, Abu Bakr and Malik, all of whom were sons of the original Yirkoy Talfi or in the Arabic rendering of the Songhay expression, Wadrat Allah. Al-Mukhtar is the most prominent brother, whom I call Yirkoy Talfi in conformity to most of the documents. He is the one who wrote Taa'idait al-Rabbaaniyya (on the Tijaniyya; 5599, fos. 58-63 and 5605, fos 67-73) and the Tabakkiyyat al-Bakkay (the accusations against the Kunta leader; 5697, fos. 29-42). Another brother, Abu Bakr, wrote poetry and served as a messenger between Hamdullahi and Segu in July 1862 (5575, fo. 217; 5713, fo. 49). He was also the author of the short invitation to Shaikh Umar to come wage the jihaad in Maasina (5519, fo. 49; the same lines are found in 5605, fo. 66). Al-Kansusi, the Moroccan Tijaniyya leader, knew about all three brothers (see 5573, fos. 62-3). For an overview of the work of al-Mukhtar, see J. Willis, “Al-hajj 'Umar al-Futii and Shaikh Mukhtar b. Wadrat Allah: literary themes, sources and influences”, in Willis, Cultivators (1979).
11. Amadu III did make some use of the Tijaniyya. He sent Amadu Tafsir on the mission to Shaikh Umar at Sabusire in 1857 (Maa Jaara) and he may have sent Modibo Dauda to Nioro in 1855 (Y. Saint-Martin, Voyage, p. 177). He apparently ordered some books from Shaikh Umar and sent the payment with Amadu Tafsir, perhaps on the same occasion as the Sabusire mission (ANS 15G 74, section L, no. 4). On the difficult relations of the Maasinanke Tijaniyya with al-Bekkay and the Caliphs, see BNP, MO, FA 5519, fos. 96-7; 5606, fos. 96-7; Umar's reference in his Bayan, 5605, fo. 26; Brown, “Caliphate”, pp. 149-50, 230; Bâ and Daget, Empire peul, I, 240 ff.
12. See the reference to the invitation to jihaad in note 10. This may have been delivered to Umar in Bundu about May, 1858, when Bandiagara/ Abdullay Ali reports a letter arrived from “Yirkoy Talfi”. A strong Maasinanke tradition reports that many scholars, perhaps 100, wrote letters to Umar to come set their society straight. Interview with Almamy Malik Yettara, Mopti, 22 Aug. 1976; Brown, “Caliphate”, pp. 232-3. There may have been some kind of hijra from Maasinanke society in the late 1850s as an expression of dissatisfaction with the regime; see BNP, MO, FA 5713, fos. 65, 134-5. This would obviously fit easily with the moral decline argument made below.
13. Amadu Tafsir and Modibo Dauda, the Maasinanke Tijaniyya cited in note 11, remained with Shaikh Umar throughout the Segu campaign and undoubtedly helped him interpret messages from Maasina and plan his action. Yahya, a pupil of al-Mukhtar b. Yirkoy Talfi, wrote two poems of praise to Umar at the time of his entrance to Hamdullahi. BNP, MO, FA 5519, fo. 51.
14. Interview with al-hajj Mamadou Lamin Soumbounou, Bamako, 3 Aug. 1976. Compare this with Umar's criticism of Futa Toro in chapter 6, section D. For a full exposition of the “decline” argument, based on the Bâ materials, see Diarah, “Maasina”, pp. 300-32.
15. Kamara, Zuhuur, 1, fo. 88. Copies of this or similar letters may be found in MAMMP 8.1. U.3; 8.4.M; 9.15; in Barth, Travels, III, 650-6; and A. Boahen, Britain, the Sahara and the Western Sudan (1964), pp. 251-2. See also Bâ and Daget, Empire peul, 1, pp., 72, 248 ff., and 285-6, and Brown, “Caliphate”, pp. 231-5.
16. For discussions about when and to what extent Jelgoji fell under the control of the Caliphate, see M. Izard, Introduction a l'histoire des royaumes mossi (2 vols., 1970), vol. 2, pp. 332-51; P. Riesman, Société, pp. 53-4.
17. On Sambunne, see chapter 5, sections B and C, and Segu 3/Cam, p. 182. On Amadu III and his struggles with the provinces generally see Ta'riikh Fittuga and Underberg, “Notes”, pp. 18-27.
18. Umar apparently sent back the first copy of this letter, which may correspond to BNP, MO, FA 5681, fos. 6-12 and MAMMP 10.6 (see the French summary in ANS 15G 74, last section, no 3). The second letter, actually written by a Futanke named Alfa Sulaiman Amadu, who had immigrated to Maasina some time before, is only available in excerpts contained in Umar's Bayan (esp. 5605, fo. 15). See also Underberg, “Notes”, p. 27; Bellat, “Renseignements”, pp. 133 ff.; and Segu 3/Cam, pp. 171-2. Sulaiman Amadu may be Cerno Sule Amadu Ali Sidi Ba of Mbolo Ali Sidi, a man from Futa Toro who joined the jihaad and later served as a cleric back in his home in central Futa. Kamara, Zuhuur, vol. 2, fo. 176; ANS 13G 211 (letter of 20 Mar. 1866), 123 (letter of 19 Apr. 1866) and 140 (passim).
19. Bellat, “Renseignements”, p. 134.
20. Umar had already threatened Hamdullahi in a letter to Balobbo early in 1861 (Y. Saint-Martin, p. 167) and his letter to Amadu III at about the same time contained the main elements of the grievances found in the Bayan (see chapter 7, note 51).
21. On the delegation, see Segu l/Mage, p. 171; Segu 3/Cam, pp. 172-3; Underberg, “Notes”, p. 27; Bellat, “Renseignements”, 135; and Ouane, Enigme, p. 183.
22. Cam was part of the first delegation. A second exchange of delegations occurred, probably in January and February 1862. Umar at this time wrote to the Maasina community to explain that his quarrel was with Amadu III alone. CEDRAB 947. He may have sent a copy of his Bayan. See ANS 13G 177 (selections for January to March 1862) for French confusion about what was going on. Their information was probably influenced by Sambala. The French believed that Maasina was the stronger party and would soon defeat Umar.
23. Henceforth Umar would call his opponent “Amadi”, a “pagan” equivalent of Ahmad. The key argument turned on muwaalaat, an affective tie or alliance with “pagans” which created a situation of takfir, the process of creating or becoming an “infidel”. It drew heavily on al-Maghili's responses to Askiya Muhammad, especially as they are found in Uthman dan Fodio's Najm al-Ikhwaan, which Umar quoted at great length. Uthman wrote the Najm during the debate with Bornu about the legitimacy of the jihaad. Umar may also have drawn on Muhammad Bello's treatise on his struggle with the Hausa leader Abdul Salam (Sard al-Kalam fimaa Jaara baini wa baina Abdu'l-Salam). The argument of muwaalaat had been mentioned in Umar's 1855 letter (see chapter 4, section F) and used very explicitly in the 1858-9 recruitment (see chapter 6). See Mahibou and Triaud, Voilà.
24. In addition to Segu l/Mage, p. 172, the most useful sources on the mobilization of the Umarian forces and the whole Cayawal campaign are two short works found in the BNP, MO, FA: 5457, fos. 1-4 (a description of Cayawal written by Muhammad al-Makki and Ahmad al-Tijani for their older brother, Ahmad b. Umar) and 5713, fo. 49 (a description of Cayawal and of events at Segu from April to July 1862). 5713, fo. 49 uses “Amadi” to refer to the Caliph.
25. On Maasina deployment, see Segu l/Mage, p. 173; Maa Jaara; and the interview with Ali Gaye Thiam cited in chapter 1, note 49.
26. Segu 2/anonymous.
27. BNP, MO, FA 5457; Segu 3/Cam, p. 178. See chapter 5, section D.
2828 Amadu III's demonstration of bravery is usually ascribed to the Thursday, May 15 battle (Segu l/Mage, p. 174; Underberg, “Notes”, p. 29), but the Makki/Tijani document (5457) shows that it occurred on Saturday, May 10.
29. Mage ( Voyage, p. 173) is the only source to describe the manufacture of bullets. There were also some skirmishes on the Sunday and Monday, but nothing which disrupted the preparations of the Umarians.
30. There is rather wide agreement that this man, Hamadi Sanfulde, played a critical role in Umar's success. See Segu l/Mage, p. 173; Segu 3/Cam, pp. 178-9; 5457, fos. 1-4.
31. BNP, MO, FA 5457, fo. 4.
32. Segu l/Mage, p. 174. The same tactic was used with devastating effect by Amadu Sheku in January 1865 at Toghu. Mage, Voyage, pp. 290-8.
33. This estimate is rather conservative. Larger figures can be found in the Bougouboly Tal interview (see note 5, above) and Ouane, Enigme, p.l83. See also Maa Jaara and Mage, Voyage, p. 298. A list of many of the prominent Maasinanke killed at Cayawal is given in 5457, fos. 3-4.
34. See the information which Henri Gaden collected from Usman Salif in Tyam, Qacida, p. 186n.
35. The date of entry is given by the two short works cited in note 24 (5457 and 5713).
36. Bandiagara/Abdullay Ali has Alfa Umar kill Amadu III at the moment of capture

37. Segu l/Mage, p. 175.
38. The information on Lam Toro Hamme Ali comes from Toro and is found in Kamara's ethnohistory, Zuhuur, vol. 2, fos. 298-301.
39. Segu 3/Cam, pp. 184-5. Cam puts some of the dialogue in Bambara. The idols were actually brought to Hamdullahi by Amadu Sheku and the 60 talibe and 40 sofas who accompanied him, on orders from the Shaikh, in late July 1862. 5713, fo. 49. Ali was executed during the revolt, probably about April 1863.
40. Segu 3/Cam, pp. 187-8. The fullest account of the installation is found in a legal opinion solicited by Amadu Sheku some years later, at a time when he was struggling with his brothers over the succession, from al-hajj Sa'iid (BNP, MO, FA, 5561, fos. 66-9). It reads in part:

In Maasina Shaikh Umar assembled all the elders of Futa and consulted about who would succeed him—they are the people who make decisions—and they all consented to his making you his successor. The next day he went to the mosque where he gathered the people of Futa and the people of Maasina. Here he rose among them, raised you and put his hand on you and said to the people:
— I inform you that this is the khaliifa.
They answered:
— Yes, we have been informed.
Then Shaikh Umar said:
— Whatever I have conquered he inherits; who ever asks of me for the blessing of the Envoy of Allah and of Shaikh Ahmad al-Tijani, let him ask for their blessing from him.
I was also told that he called you, seated you in his seat and sat between your hands and swore his allegiance, then he commanded the people to do likewise. And they did. And he swore that whoever did not swear allegiance from the descendants of Sa'iid [Umar's father] would be destroyed. This, he said would be the fate of all who disagree, whether they were present or not. This we heard and it was also told to us by those who were present.”

41. On Mossi see Gaden's notes in Tyam, Qacida, p. 185n. On Kong, see Binger, Du Niger au Golfe de Guinée, I, p. 341. On Dyula concern about the demise of the caliphate, see Person, Samori, II, p. 810. 42. For Dingiray, see BNP, MO FA 5713, fo. 49. For Sierra Leone, see a letter written by T. G. Lawson on 3 Oct. 1884, recalling messengers from Umar in Nov. 1862. D. Skinner, Thomas George Lawson (1980), pp. 191-2.
43. Umar's triumph did produce temporarily more militant attitudes in Futa Toro towards the French. For that and Sal's message, see Robinson, Chiefs and Clerics, pp. 62-6, and ANS 177, letters of 1, 11, 15 and 22 Nov. and 20 Dec. 1862.
44. For the confusing Timbuktu story of June 1862, see BNP, MO, FA 5259, fos. 66-73 (al-Bekkay's 1862 letters) and 5716, fos. 32 ff. (Umar al-Awsi's description of the relations with al-Bekkay); Pefontan, “Tombouctou”, p. 99; Gaden's notes in Tyam, Qacida, 190n. A different and largely erroneous version is found in Mage, Voyage, pp. 5-7, 178, and Delafosse, Haut-Sénégal-Niger, II, pp. 321-2. For a Timbuktu perspective on the whole jihaad, see 0. Lenz, Timbouctou (2 vols., 1886), vol. 2, pp. 173-5.
45. It was at this time, between June and August of 1862, that the Umarians found the incriminating documents cited in chapter 7, note 47. Awsa means left bank in Songhay, in contrast to Gurma or right bank, and should not be confused with Hausa. See Abdel Kader Zebadia, “The career and correspondence of Ahmed al-Bekkay of Timbuctu from 1847 to 1866”, London University, School of Oriental and African Studies, Ph.D. thesis, 1974, 416 ff.
46. See the citation for Tabakkiyyat al-Bakkay, “What makes al-Bekkay weep”, in note 10 above.
47. Makki was the joint author of the document in 5457 cited in note 24; his list of friends and foes is found in IFAN, Fonds Brevié, document 12a. See also chapter 7, section B. Makki also wrote, presumably on Umar's instructions, to the relatives of the family in Sokoto, including Caliph Ahmad b. 'Atiq and the father of Amadu's mother Aisha. BNP, MO, FA 5737, fo. 59.
48. De Loppinot, “Souvenirs”, p. 27; Sal in J. Ancelle, Explorations, p. 221.
49. Beginning with the revolt and 1863, most of the internal narratives cited in note 2 are of much less utility.
50. The Maa Jaara first introduced in chapter 1, note 42.
51. From al-Bekkay's fourth letter to Umar, some time late in 1862. BNP MO, FA 5259, fo. 70.
52. I have not been able to locate a copy of this letter. Its contents are discussed in Gaden's notes in Tyam, Qacida, pp. 190-1; Mage, Voyage, pp. 178, 287; and Delafosse, Haut-Sénégal-Niger, II, p. 321.
53. For this corpus or diwaan see BNP, MO, FA 5259, fos. 66-70; CEDRAB 699; MAMMP 8.4.E and 9.15. See also al-Bekkay's letter to a Tuareg ally of Umar (BNP, MO, FA 5259, fos. 70-1). It appears that Umar was not very communicative in return. Al-Bekkay's fourth letter makes this point. In 5716, fo. 40, there is a very short note from Umar to al-Bekkay which may be the response to al-Bekkay's first letter. On the other hand, Alfa Umar al-Awsi did respond for the Shaikh: see 5606, fos. 182-4; 5716, fos. 32-40 and 182-5 (these last folios contain al-Awsi's response to the last of al-Bekkay's letters excerpted in 5716, fos. 38-9). For the Kunta perspective, see Saad, “Social history”, pp. 380-2.
54. From al-Bekkay's first letter to Umar. BNP, MO, FA 5259, fo. 67.
55. Al-dajjal al-mufassid alaadhi awjaba Allabu jihaadihi. This comes from Bekkay's letter to al-hajj Buguni found in the Arabic original and French translation in ANS 15G 77, piece 52. Similar language is used in letters to the Tuareg and the Fulbe of Maasina found in BNO, MO, FA 5259, fos. 72-3. See also Abun-Nasr, Tijaniyya, p. 125. Further study of Bekkay should reveal the relative roles of ambition and conservatism in his actions. Saad (“Social history”, pp. 377-85) suggests that Bekkay's withdrawal from Timbuktu at the time of the Umarian takeover in June 1862 already constituted a hijra in preparation for jihaad.
56. De Loppinot, “Souvenirs”, p. 29. See also Maa Jaara; Menvielle in Tauxier, Peuls, pp. 186 ff.; Mage, Voyage, p. 180.
57. Interview with Ali Gaye Thiam noted in chapter 1, note 49. Compare it with the account cited in chapter I, note 50.
58. Mage (Voyage, p. 177) has al-Bekkay writing directly to the Umarian disciple, Modibo Dauda, who had once been a Kunta pupil, in this scenario al-Bekkay was betting that he could invoke Dauda's old loyalties against his new ones. Underberg, Bellat and Maa Jaara are more reliable on this point.
59. The other royal was Abdussalam, son of Seku Amadu. He was captured and executed shortly after this.
60. One contingent sent to obtain powder from Segu was exterminated near Jenne at the end of May 1863. Mage, Voyage, p. 180; Delafosse, Haut-Senegal-Niger, II, p. 322; Bandiagara/Abdullay Ali.
61. Mage (ANS IG 32, piece 35, p. 13) estimated the loss at 15,000 men. I have taken the figure of 3,000 offered by Tauxier (without explanation, in Peals, p. 198) as a more reasonable figure for table 5. On Mani-Mani, see Maa Jaara; Underberg, “Notes”, pp. 35-6; Menvielle in Tauxier, Peuls, pp. 187-8. See also Mage, Voyage, p. 437.
62. Umar probably lost at least 2,000 men at Sege. See the sources cited in note 61 and the notes of Gaden in Tyam, Qacida, pp. 192-4.
63. For Amadu Sheku's situation see chapter 7, section E, and Mage, Voyage, pp. 180 ff. The Sinsani revolt started in August 1863, and tied Amadu's forces down through November.
64. The best general accounts of the siege are:

Umar apparently amputated the feet of some Maasinanke women to prevent escape. Underberg, “Notes”, p. 36. Kamara heared rumours of cannibalism. Samb, “Omar par Kamara”, p. 172.
65. BNP, MO, FA 5713, fo. 56. The letter requires the context of impending doom which only the siege of 1863-4 can supply. I have assumed that it was written early in 1864 in Hamdullahi, rather than during the flight to Degembere, simply because a Hamdullahi letter would have had greater chance of getting into the Segu archives; a Degembere letter would probably have fallen into the hands of Tijani or the Maasinanke.
66. The best account of Tijani's recruitment comes from Gaden and his informants in Tyam, Qacida, p. 195n.
67. At least one Maasinanke woman and one of Umar's wives did escape and reach Segu. Mage, Voyage, pp. 436 - 9, Maa Jaara. On Umar's escape I have generally followed Bandiagara/Abdullay Ali and Mage, Voyage, p. 437. Underberg (“Notes”, p. 37) has the Umarians light a fire to provide light for the escape. Nioro 2/Adam (p. 122) has the Yirlabe and Ngenar regiments betray information to the attackers so that Toro and Umar have to fight their way out.
68. The Umarian informer is often identified as Mamadu Ismaila Samba Sire of Giray in Ngenar province. Gaden in Tyam, Qacida, p. 197n; Nioro 3/Delafosse; Bellat, “Renseignements”.
69. For the Maasinanke versions, see Maa Jaara; Samb, “Omar par Kamara”, p. 399.
70. Some of the more imaginative Umarian solutions can be found in Bandiagara/Abdullay Ali (p. 141 of the Cissoko translation); Abun-Nasr Tijaniyya, p. 141; Soleillet, Voyage, pp. 359-60. Amadou Hampaté Bâ has formulated what is probably the most satisfying version for all concerned: a faulty gun in the hands of Umar's slave ignites the last bit of powder and kills everyone in the cave. It can be found in the Diallo interview cited in chapter 4, note 59. The dead included:

71. On the reconquest, the basic text is the last part of Bandiagara/ Abdullay Ali. On Yirkoy Talfi's death, see Cissoko's translation and note (p. 145n). Yirkoy Talfi had written a lament of Umar's death (BNP, MO, FA 5565, fos. 65-6). On al-Bekkay's death, see Marty, Soudan, II, p. 92; Briquelot, “Notes”, p. 16.
72. On Maasina in the late nineteenth century, see E. Caron, De St. Louis a Tombouctou (1891), pp. 278, 306, 321; Marty, Soudan, II, pp. 149, 179, 194, 201; M. Forget, “Population et genres de vie dans le Kouniary”, in P. Galloy et al., Nomades et paysans d'Afrique noire occidentale (1963), pp. 162-4. The best account of Tijani's state, written just before Tijani's death in 1887, is Caron, Tombouctou, pp. 139 ff. See also Caron's material in ANS ID 91 and I G 83. Good information is also contained in my interview with Amadou Oumar Bâ cited in note 5 above. The principal chronology for the wars is found in the monographs cited in note 4 above.
73. Bandiagara/Abdullay Ali, pp. 140-3 of the Cissoko translation. Abdullay Ali has Tijani designated as Umar's successor before he slips out of Hamdullahi. For the songs sung by griots in Tijani's honour, See Fonds Vieillard, Maasina, cahier 5 (IFAN).
74. Interview with Pate Sy at Bandiagara, 18 Aug. 1976. Sy is part of the “Agibian” entourage which claims that Agibu rescued the objects when he was appointed chief of Bandiagara by Archinard in 1893. See also Briquelot, “Notes”, p. 15. The disciple who is often presented as having received the objects is Yirkoy Talfi.
75. Amadu apparently did record a dream in which he conquered Maasina. It is tentatively dated to 1877/8. BNP, MO, FA 5713, fos. 130 and 175.
76. French translations may be found in Samb, “Omar par Kamara”, pp. 401 off., and in Bendaoud Mademba, “La dernière étape d'un conquérant”, BCEHSAOF 192 1.