Navigation – Plan du site

AccueilThématiques12Race and military coalitions in t...

Race and military coalitions in the Middle Niger during the second half of the 19th century

Race et coalitions militaires au moyen Niger pendant la seconde moitié du XIXe siècle
Joseph M. Bradshaw

Résumés

Dans la seconde moitié du XIXe siècle, les envahisseurs Futanke et leurs alliés renversèrent le califat de Ḥamdallāhi. Les Futanke et les Masinanke de Ḥamdallāhi partageaient une foi, une langue et une culture communes, ils avaient mené des guerres sous l’égide du djihad contre des non-musulmans et des musulmans dont ils considéraient la pratique peu orthodoxe. Malgré la rhétorique anti-« païenne » de leurs dirigeants, souvent envers des gens qu’ils qualifiaient de « noirs », l’instabilité des guerres de 1861-1890 a nécessité la construction d’alliances qui ont diminué la valeur instrumentale de l’exclusion raciale et religieuse. En accueillant des alliés non musulmans, les Futanke ont construit une forte coalition qui a fini par vaincre leurs ennemis musulmans Fulbe.

Haut de page

Entrées d’index

Géographique :

Afrique de l'Ouest, Mali, moyen Niger
Haut de page

Texte intégral

Introduction

  • 1 A pastoral people who have migrated across the African Sahel (sg. Pullo, pl. Fulbe, adj. Fulbe) The (...)
  • 2 The main reference work for the history of Ḥamdallāhi is A.H. and J. Daget, 1984.

1From 1818 until its collapse in 1862, the Caliphate of Ḥamdallāhi in the Middle Niger folded racial and religious difference into its politics. This state was founded by the Fulbe Muslim reformer Ahmad Lobbo (d. 1845) of the Barri clan, who established his capital Ḥamdallāhi in an inundated region known as the Burgu. Aḥmad Lobbo drew the support of leaders from across the Middle Niger, especially from the floodplains along the Niger, Bani, and Diaka rivers, who joined his revolutionary movement and eventually fought a war to liberate the Fulbe1 of the Niger River’s inland delta from the Arbe, local proxies of the Bambara kingdom of Segu. The Fulbe of the inland delta’s Masina sub-region, called the Masinanke, had long distinguished themselves from the Bambara and other local groups. In the process of conquering the inland delta, the caliphate constructed racial and religious distinctions that became important factors in regional politics and social hierarchy. Ḥamdallāhi, a state that initially brought enslaved and subjugated groups together to fight a war of liberation, eventually created a social hierarchy that placed the caliphate’s Fulbe population above supposedly servile populations categorized as “black.” Ḥamdallāhi contrasted the identity and legal status of free Fulbe against the “blackness” and servility of the Dogon, Bobo, and others. Ḥamdallāhi then used this category of “blackness” to cast doubt upon conversions to Islam among these people and thereby justify their continued taxation, enslavement, and conscription.2

  • 3 D. Robinson, 1985 is the main secondary source on the history of al-ḥājj ‘Umar’s movement.

2Similarly, al-ḥājj ‘Umar Tal—a powerful and influential Sufi cleric from Futa Toro, a Fulbe Islamic state along the northern bend of the Senegal River—constructed categories of unbelief to justify wars against the Bambara kingdoms of Kaarta and Segu in the 1850s, and later against the Caliphate of Ḥamdallāhi in the early 1860s. During his pilgrimage to Mecca, al-ḥājj ‘Umar had received a mandate to proliferate Tijaniyya Sufi practice in the “Sudan” from Muhammad al-Ghali, leader of the Tijaniyya tariqa in the Hijaz. On his return from Mecca, al-ḥājj ‘Umar traveled to the Fulbe-led caliphates of Sokoto (c. 1832–38) and Ḥamdallāhi (1839) where he attracted followers to his nascent movement. In 1841 the Almami of Futa Jallon Alfaya Bakar, granted al-ḥājj ‘Umar permission to establish a community at Jegunko where he taught and authored his most influential work, the Rimāḥ ḥizb al-Raḥīm.3

  • 4 D. Robinson, 1985, p. 121-126.
  • 5 I use the term jihad to describe wars, but only by way of pointing out that certain belligerents th (...)

3From around 1846–47 al-ḥājj ‘Umar began to travel throughout the Senegambia gathering followers.4 During this tour al-ḥājj ‘Umar Tal attracted a significant number of followers from Futa Toro, and consequently his movement developed a Futanke consciousness. As recruits from Futa Toro flocked to the ‘Umarian movement, it began to transition from an isolated scholarly community to a Futanke-led movement that advocated an activist approach to Islamic reform, particularly among neighboring non-Fulbe populations. Soon after he returned to Futa Jallon, al-ḥājj ‘Umar and his followers initiated their first military campaign against “non-Muslim” Mandinka at Tamba. This was followed by far more ambitious campaigns against the Bambara kingdoms of Kaarta and Segu. Eventually, as detailed below, the leaders of the Caliphate of Ḥamdallāhi opposed the Futanke invasion of Segu and as a result were branded “non-Muslim” apostates by al-ḥājj ‘Umar, who subsequently led a jihad against the caliphate.5

  • 6 The word “pagan” appears in quotes to signal that this was an historic argument about unbelief appl (...)

4The wars between the Futanke and the Masinanke of the Ḥamdallāhi are riddled with contradictions. The Futanke of the ‘Umarian movement and the Masinanke shared a similar faith, language, and culture, and both had fought wars they framed as jihads against populations they labeled as “pagan.”6 For the Futanke and Masinanke, “pagan” was a term that was often applied to Muslims whose practices were deemed to be unorthodox as well as to non-Muslims. Consequently, the Futanke and Masinanke claimed their wars against Muslims and non-Muslims alike were jihads. Such constructions of unbelief became more salient when the Futanke and Masinanke accused each other of apostasy, to justify the near constant wars they fought against one another from 1861 to 1890.

  • 7 J. Hanson, 1996, p. 9.
  • 8 D. Robinson, 1985, p. 81-89.

5The construction of unbelief is a common theme among the Fulbe jihads that established Futa Toro, Futa Jallon, the Sokoto Caliphate, the Caliphate of Ḥamdallāhi, and the ‘Umarian states.7 Through their jihads, Fulbe reformers sought to replace rule by non-Muslims or Muslims they argued were “errant” with reformed Islamic states. Consequently, populations classified as “pagan” by these states were the targets of punitive military campaigns. David Robinson argues that the Fulbe jihads transformed Fulbe identity by fostering a “sense of electedness,” backed by genealogical assertions that connected the Fulbe to the first community of Muslims and the elevation of Fulfulde as a language of Islamic learning.8 Such assertions were critical for Fulbe-led reform movements as they sought to dominate neighboring non-Fulbe communities where most people did not practice Islam, as well as others in which Islam was already well established.

  • 9 L. Alkaly and J. Alessandra, 2020; I. Khaldūn and D. Juwaydī, 2015.
  • 10 J. Webb, 1995, p. 22-26.
  • 11 B.S. Hall, 2011, p. 52-53.
  • 12 C. Hamel, 2013.
  • 13 A.H. Bâ and J. Daget, 1984; D. Robinson, 1985; M. Dupire, 1994; R. Botte and J. Schmitz, 1994; B.S. (...)

6At Ḥamdallāhi, the racial category sūdānī or “black” was also applied to certain non-Fulbe populations to cast doubt on the faith of Muslims within these groups and justify their enslavement, taxation, and conscription. The racial category of “black” had been employed in the centuries before in northwest Africa. Arab travelers as early as the 14th century used such a label to distinguish supposedly inferior foreign cultures.9 This racial category was probably used by Arab-Berber nomads along the desert’s edge to explain their domination of “black” cultivators,10 and it was frequently employed to justify the enslavement sub-Saharan Africans by Arabized groups.11 In the mid-17th century, the Moroccan ruler Mūlāy Ismā‘īl used the race and slave origins to justify his conscription of free blacks.12 Eventually, the Fulbe-led Muslim states established through jihads in the 19th century framed their own conquest, enslavement, and conscription of certain groups around such distinctions.13

7In summary, even while they made such assertions about “unbelief” and “blackness,” both the Futanke and Masinanke incorporated so-called “pagan” and “black” soldiers into their armies. In this article, I argue that despite the importance of categories such as “pagan” and “black” within Fulbe social hierarchy, and within the Caliphate of Ḥamdallāhi in particular, the instability associated with the Futanke-Masinanke wars of 1861–90 necessitated the building of alliances that diminished the instrumental value of racial and religious exclusion. Ultimately, the Futanke who founded a new emirate among the Dogon of Bandiagara would prove better coalition builders, and they conquered most of the territory formerly ruled by the Caliphate of Ḥamdallāhi.

Figure 1: The Middle Niger

Figure 1: The Middle Niger

Joseph M. Bradshaw

Enslaving and conscripting “blacks” in Ḥamdallāhi

  • 14 A. Mayor, 1997, p. 38.
  • 15 Ibid., p. 39.
  • 16 A.H. and J. Daget, 1984, p. 280.

8Evidence indicates that during the construction of Ḥamdallāhi, Aḥmad Lobbo and his early followers settled—albeit briefly—among “black” non-Muslims, the Dogon. Anne Mayor’s work on the material history of Modjodje supports oral accounts that Ahmad Lobbo and his followers lived in a Dogon village for four years during the construction of Ḥamdallāhi.14 According to Mayor, “the Barri family had brought their herds to the foot of the Dogon’s escarpment before the establishment of the village of [Modjodje].”15 Familiarity with the Barri family and concerns about Segu are likely to have played roles in the Dogon’s embrace of their Muslim Fulbe guests. This hospitable cohabitation was short-lived. Before Ahmad Lobbo’s death in 1845, Ḥamdallāhi considered the Dogon crown slaves and routinely pressed Dogon men into military service to support its wars. After Ahmad Lobbo’s death, according to Hampâté Bâ’s account, the Timbuktu scholar Aḥmad al-Bakkāy criticized Ḥamdallāhi for failing to Islamize the Dogon and continuing to enslave and conscript them.16

  • 17 Malian Arabic Manuscript Microfilm Project, Tārīkh Jenne (Yale University, 1976), reel 7.
  • 18 Jenne likely submitted to Ahmad Lobbo’s authority shortly after the second siege in 1821; see A.H. (...)
  • 19 Using non-Muslim Bobo archers would also have allowed Ahmad Lobbo to sidestep Islamic prohibitions (...)

9A similar status was applied to the Bobo. Ḥamdallāhi’s location east of the Bani River placed it to the north of the eastern reaches of the Boobola section of the Seno-Gondo Plain, which was inhabited by Bobo cultivators. By the time Aḥmad Lobbo settled east of the Bani, Bobo cultivators had migrated westward and settled both banks of the river. The historical text Tārīkh Jenne even contains the peculiar claim that the Bobo were Jenne’s first settlers.17 Jenne became an important religious center under the Songhay Empire founded by Askiya Muhammad at the end the 15th century, and its Muslim population accepted Ḥamdallāhi’s rule as early as 1821.18 However, non-Muslim Bobo cultivators remained on the farmable lands surrounding it. William A. Brown’s dissertation on the Caliphate of Ḥamdallāhi suggests that Aḥmad Lobbo had relied on formidable Bobo archers to expand his territory, and that their use of poison-tipped arrows explains how Aḥmad Lobbo’s forces managed to defeat both the well-trained cavaliers who fought for the Dikko clan of Masina, and Segovian foot soldiers armed with fusils.19 Despite its early inhabitants’ period of cohabitation with the Dogon, and Aḥmad Lobbo’s probable use of Bobo archers and other “slave soldiers,” Ḥamdallāhi eventually considered the Dogon and Bobo subalterns within the caliphate’s racial and religious hierarchy.

  • 20 B.S. Hall, 2011, p. 70-73.
  • 21 The textual history of the Tārīkh al-fattāsh was long debated by Arabists of an earlier generation. (...)
  • 22 Institut de France, Fonds de Gironcourt, ms. 2406 (56-60) / XVI, Aḥmad Lobbo, Manshūra fī istirqāq (...)

10It is important to contextualize such hierarchies in the long-standing discourse on slavery and racial difference in the Western Sahel. Within this discourse, revealed by Bruce Hall in A History of Race in Muslim West Africa, the term “blacks” became “a legal status category inferior to free Muslims.”20 In Ḥamdallāhi, “blackness” justified the enslavement of Muslims among the servile populations who were considered crown slaves. The Caliphate of Ḥamdallāhi used a forged history, titled Tārīkh al-fattāsh, to claim that Aḥmad Lobbo inherited his authority over those populations that were racialized as “black” from the founder of the Songhay Empire, Askia Mohammad.21 Thus, Ḥamdallāhi’s forgery of the Tārīkh al-fattāsh represented an effort to insert a racial category of “black” servility into the documentary record to support Aḥmad Lobbo’s claims over non-Fulbe, even if individuals from these “black” populations had practiced Islam before the establishment of the caliphate. Ḥamdallāhi extended this racial category to the Dogon and Bozo, to justify their taxation, conscription, and enslavement. Aḥmad Lobbo’s son and heir, Aḥmad II, referenced his inherited rights over “tribes of blacks” in a political decree that referenced this forged document: “[Aḥmad II] has ordered that the enslavement of blacks would be governed generally by the Tārīkh al-fattāsh, even if their mothers were free.”22

  • 23 B.S. Hall, 2011, p. 71, note 5.

11In the same ruling, Aḥmad II described his authority over some non-Muslims, using the Arabic term zanj, which was also used to describe the servile tribes in the Tārīkh al-fattāsh. Earlier, the term had been applied by Arab traders to the dark-skinned populations of East Africa and slaves taken from the East African coast, and Arab scholars had used the term to describe a 10th-century slave revolt in Iraq.23 Thus, Aḥmad Lobbo used an Arabic term associated with slavery to support his claims over the populations raced as “black” and cast doubt upon their conversions to Islam.

12The Fulbe of the Caliphate of Ḥamdallāhi also applied the Fulbe term habe, meaning non-Fulbe, to the Dogon and Tombo of the Hayre. According to Henri Gaden, a French colonial officer and ethnographer stationed in Bandiagara in the mid-1890s,

  • 24 ANOM FP Fonds Gaden, “Personal correspondence, Bandiagara 20-25 June 1895”; emphasis mine; for mode (...)

[t]he Fulbe originally applied the term Habe to anyone who, unlike them, did not have white blood in their veins. Habe meant “the blacks” in opposition to oriental populations like the Moors, Tuaregs, and Fulbe. In the time of [the] Masina [Caliphate] the term lost its generality and was applied specifically to cultivators in the mountains from Bandiagara to Doventza and along the Seno-Bankasso plain.24

13According to Gaden, habe originated as a broad racial term that eventually became the Masinanke name for the rebellious non-Muslims who lived at the caliphate’s fringes. While Gaden’s reference to “white blood” indicates he interpreted the ethnographic data he collected through a 19th-century polygenist lens, he observed that habe and black were categories of “otherness” that contrasted Fulbe identity. Indeed, for the Dogon and Tombo, the term habe marked both their non-Muslim status and their supposed servility within the Ḥamdallāhi’s social order.

  • 25 S. Zehnle, 2015, p. 173.
  • 26 U. Baumgardt, 1994, p. 295-311.
  • 27 S. Zehnle, 2015, p. 174.
  • 28 M. Ochonu, 2014, p. 9.

14Stephanie Zehnle’s dissertation “A Geography of Jihad” notes that the term habe also meant non-Muslim and servile in other Fulbe states;25 Ursula Baumgardt’s assessment of a Cameroonian Fulbe storyteller’s repertoire shows that the term habe contained both ideas of inheritable servility and religious difference; and Fulbe categories of racial and religious difference also underpinned Sokoto’s rule over non-Muslims.26 Early Fulbe dictionaries produced in the vicinity of the Sokoto Caliphate treated habe as expressing both racial and religious difference, and linguists often translated the word as either “pagan” or “slave.”27 Moses Ochonu has presented a convincing argument that the British colonial administration encountered an entrenched system of racial and religious difference that they adapted to a system of indirect rule centered around the Hausa-Fulani caliphate.28 Ḥamdallāhi developed a similar hierarchical system, in which the Dogon—called habe—were classified among the servile races that the forged Tārīkh al-fattāsh referred to as “tribes of blacks.”

  • 29 M. Johnson, 1976, p. 486.

15Racial and religious differences underpinned a social structure that treated “black” non-Muslims in the caliphate as crown slaves or vassals, confined to certain forms of labor such as agriculture, masonry, canoe transport, or fishing, and made them eligible for conscription during times of war. The slaves Ḥamdallāhi settled on farmable land owed the state a yearly payment called jamgal and thus have been argued to have lived in a state of serfdom.29 It is important, however, not to elide the fact that this so-called serfdom was based on racial and religious differences, and this article therefore refers to those servile populations as crown slaves, to highlight how the state rationalized its claims on them.

  • 30 ANS 15G 172, “Rapport sur le délimitation des États d’Aguibou et Ahmadou Abdoul, 1894”; ANS 1G 213, (...)

16After the Futanke toppled the Ḥamdallāhi in 1862, the surviving members of the Masinanke royal family continued to treat the “black” non-Muslims east of the Niger as crown slaves and “pagans.” They relocated their warriors to Fio where they lived among Bobo farmers, whom they continued to tax heavily.30 However, this strategy of expropriation based on racial and religious difference did little to augment the armies of the royal family with non-Fulbe warriors. Their Futanke adversaries, on the other hand, entered into an alliance with their Dogon hosts at Bandiagara that transgressed the hierarchy established by the Caliphate of Ḥamdallāhi. Despite their earlier positions against forming alliances with non-Muslims, the Futanke at Bandiagara built a coalition of forces that included Dogon, Bambara, Bozo, and others. The ideological flexibility of the Futanke indicates that racial and religious hierarchies were less useful during the decades of near-continuous warfare between Fulbe-led armies in Middle Niger in the period 1861–90.

Toward the Futanke conquest of Ḥamdallāhi, 1855–62

  • 31 D. Robinson, 1985, p. 282.
  • 32 In the instances where names did not appear in Arabic sources, I have written them as they appear i (...)
  • 33 D. Robinson, 1985, p. 263.

17In 1855 al-ḥājj ‘Umar overthrew the non-Muslim Bambara king of Kaarta under the umbrella of a jihad against “paganism” and then advanced toward the Bambara kingdom of Segu. Ḥamdallāhi’s relationship with Segu on the eve of the ‘Umarian conquest demonstrates Masinanke flexibility regarding racial and religious distinctions. As noted above, the caliphate’s founder Aḥmad Lobbo had fought a jihad to liberate the Fulbe of Masina from the rule of Bambara kings and their allies; but after Aḥmad Lobbo’s death, Ḥamdallāhi’s ties to its non-Muslim Bambara neighbors at Segu improved. Robinson notes that the Futanke threat enabled the Masinanke to overcome a “deep antagonism” toward both Segu and Timbuktu.31 A polemic written by Yirkoy Talfi claims that sometime in 1860 Aḥmad al-Bakkāy wrote to Aḥmad III and Ali,32 the king of Segu, to encourage them to resist the Umarian jihad.33 That November, after the Futanke invaded Sinsani, Aḥmad III wrote to al-ḥājj ‘Umar to dissuade him from pressing his attack against Segu, arguing that after years of conflict with Ḥamdallāhi its people had converted to Islam. Unconvinced, al-ḥājj ‘Umar sent a delegation to Ḥamdallāhi to challenge the legitimacy of Ali’s conversion, and in response Aḥmad III stationed Masinanke cavalry on the right bank of the Niger River, directly across from Sinsani, to contain the Futanke. The standoff erupted into a battle when a bellicose group of Futanke soldiers took it upon themselves to cross the river to attack the Masinanke. In March of 1861, Ali fled from Segu to Ḥamdallāhi, where Aḥmad III continued to offer his protection to the deposed ruler.

  • 34 For a description of Aḥmad III and al-Bakkāy’s opposition to the ‘Umarian conquest of Segu, see D. (...)
  • 35 S.M. Mahibou and J.L. Triaud, 1983.
  • 36 B.M. Perinbam, 1997, p. 102; J. Hanson, 1997, p. 85-97.

18These events strengthened Aḥmad III and Aḥmad al-Bakkāy’s accusations that al-ḥājj was guilty of sowing discord among Muslims.34 Al-ḥājj ‘Umar responded to these accusations and justified his impending attack on Ḥamdallāhi, in his treatise Bayān mā waqa‘a, which turned the argument that he had sown discord among Muslims on its head.35 In part, al-ḥājj ‘Umar accomplished this rhetorical sleight of hand by appealing to overlapping ideas of ethnic and religious difference: in explaining his seizure of Sinsani, he used the term bambara—which was both an ethnonym and a pejorative term for “pagans” in the West African Sahel at that time—to elicit an emotional response from his audience.36

  • 37 Bakhunu was a region to Kaarta’s east with significant Soninke population. Al-ḥājj ‘Umar argued tha (...)
  • 38 al-ḥājj ‘Umar constructed the Soninke as muḥarubūn (warriors). Triaud and Mahibou have focused on h (...)

19The word bambara’s connection to “paganism” lent a certain rhetorical force to al-ḥājj ‘Umar’s arguments that lands he conquered around Bakhuna37 and Segu were not Muslim. He argued that in these regions “pagans” had subjugated Muslims and that the Soninke mixed Muslim and “pagan” practice;38 this meant that the only “true” Muslims there were weakened captives of non-Muslims, and thus that the ‘Umarian army’s incursion liberated Muslims and advanced the spread of Islam. Al-ḥājj ‘Umar described the three types of people that inhabited Sinsani in the following terms:

  • 39 The following translation from Arabic is my own, from the original text published in S.M. Mahibou a (...)

The first are the bambara pagans, unbelievers who worship Godless idols, and the second are bandits (muḥarubūn) who block the roads and take tolls, use what is forbidden, and take from unbelievers and Muslims alike. They say “There is no god but God” but they had not been under the Caliphate’s rule from the time they settled the land until the day we arrived there. The third are tribes of Muslims pressed into captivity at the hands of the bambara and others.39

  • 40 S.M. Mahibou and J.L. Triaud, 1983, p. 107-138.

20In the section of the Bayān that contained his argument against Aḥmad III, however, al-ḥājj ‘Umar dropped the pejorative term bambara. Instead, he drew on passages from the Qur’an and writings of earlier scholars to argue that Aḥmad III had abandoned his duty of promoting unity among Muslims (muwālā with fellow Muslims). Al-ḥājj ‘Umar further argued that Ḥamdallāhi’s alliance with Segu was muwālā with unbelievers rather than with Muslims and claimed that by allying with non-Muslims and supporting their military actions against his army, Aḥmad III had committed acts of apostasy.40

  • 41 D. Robinson, 1985, p. 295.

21Despite this argument against alliances with “pagans,” the army al-ḥājj ‘Umar assembled for its offensive against Ḥamdallāhi included two contingents that were made up of slaves and former Bambara and Mandinka infantry. After the Umarian army captured Segu, al-ḥājj ‘Umar folded Bambara slave soldiers into an elite personal guard called the jomfutung. These Bambara soldiers filled out two regiments of non-Futanke soldiers that accompanied four other Umarian regiments on the march to Ḥamdallāhi.41 Preparations took two weeks, but once it was set in motion, the Futanke conquest of Ḥamdallāhi unfolded quickly. The ‘Umarian army marched directly toward the capital of Ḥamdallāhi. Its progress was interrupted when the ‘Umarians encountered a large Masinanke army in a dell (or cayawal in in the Pulaar language). It was in that valley that the Futanke won their most significant victory at what is now known as the Battle of Cayawal.

Ḥamdallāhi divided on the eve of ‘Umarian conquest

  • 42 D. Robinson, 1985, p. 298.
  • 43 B. Sanankoua, 1990, p. 155.

22By all accounts, Aḥmad III led a courageous effort to defend Ḥamdallāhi. After his defeat at the Battle of Cayawal, he fled downriver toward Timbuktu, where he was captured and executed by the Futanke commander Alfa ‘Umar Baila.42 Masinanke accounts to colonial officers, like those collected by Hampâté Bâ, suggest that disputes among members of the royal family over the succession weakened the resolve of Aḥmad III’s uncles who led the military.43 The available evidence also indicates that the Masinanke were plagued by divisions not only within the royal family but in the caliphate’s population more broadly. After Cayawal, the Futanke army ranged as far north as Fittuga, south of Timbuktu.

  • 44 D. Robinson, 1985, p. 287.
  • 45 A.H. and J. Daget, 1984, p. 248-249. Bâ communicated his conclusion that the Ḥamdallāhi was divi (...)
  • 46 A.H. and J. Daget, 1984, p. 249.

23The first major split in the leadership of Ḥamdallāhi occurred shortly after the death of Aḥmad Lobbo. As the founder of the Caliphate of Ḥamdallāhi neared death, two candidates emerged as potential successors: Aḥmad II, the founder’s son and preferred heir; and the military chief Ba Lobbo, Aḥmad Lobbo’s nephew (i.e. a son of his younger brother Bakari Lobbo).44 The succession of Ba Lobbo would have been in keeping with the Fulbe tradition of fraternal succession, in which generally the eldest male in the fraternal lineage succeeded rather than the eldest son of the deceased. Accounts of the contest for succession after Aḥmad Lobbo’s death indicate that there was a split between the partisans of the military leader and Islamic scholars.45 Hampâté Bâ notes that, having considered all the eligible males, the grand council narrowed the field to just these two candidates; and those who preferred the founder’s son noted his devotion and erudition, while those who preferred his nephew noted his bravery and largesse.46 This reference to Ba Lobbo’s generosity was undoubtedly connected to his ability to distribute wealth gained through warfare as gifts to his soldiers and partisans.

  • 47 C. Monteil, 1932, p. 112.
  • 48 Yousouf Diallo argues that Ḥamdallāhi’s interventions in the Boobola had a significant economic dim (...)
  • 49 A.H. Bâ and J. Daget, 1984, p. 72.
  • 50 Ibid., p. 266.
  • 51 Ibid., p. 265-266.

24The grand council voted for Aḥmad II, but Ba Lobbo remained the favorite of a noisy, bellicose contingent within the aristocracy; and the grand council gave him command over the entire army. Warfare and punitive raids predicated on Masinanke-defined racial and religious difference were an important part of Ba Lobbo’s political and military strategy. He had first served his uncle as vice-regent of the region of Fakala. His army patrolled the east bank of the Bani where “black” non-Muslim Bobo subjects cultivated fields for the Masinanke.47 For Ba Lobbo, the “black” non-Muslims of the Seno-Gondo Plains to the east of Fakala represented a target of opportunity for punitive campaigns through which he and his army could gain slaves, territory, and horses;48 and Hampâté Bâ suggested that punitive raiding against non-Muslim populations was practiced more frequently once Ba Lobbo gained complete control of the army after Aḥmad Lobbo’s death in 1845.49 The same author also posited that Ba Lobbo thought of the region of Saro as part of his birthright, because it was his mother’s home,50 and during Aḥmad II’s reign, Ba Lobbo’s numerous successful military expeditions included devastating offensives against the Bambara of the Hayre and the Bambara of Saro.51

  • 52 D. Robinson, 1985, p. 297.
  • 53 Aḥmad II’s four brothers—ʿAbd Allāh, Muḥammad, ‘Abd al-Salām, and ‘Abd al-Raḥmānwere all potentia (...)
  • 54 Citing Ba, Brown notes that Aḥmad III “was completely under the control of the military aristocracy (...)

25It is tempting to trace the fractures among the ruling family to Ba Lobbo’s failed bid to succeed Aḥmad Lobbo, and to locate the reasons for Ḥamdallāhi’s failure to repel the Futanke in the discontent of the general and his partisans. Information presented to the colonial officer Henri Gaden by Usman Salif Thiam claimed that Ba Lobbo’s actions at the battle of Cayawal were less than heroic, and that the general made little effort to keep his troops from fleeing the battlefield.52 However, this account must be balanced against the political victories Ba Lobbo had won shortly after Aḥmad II’s death. According to Masinanke traditions, Ba Lobbo played the role of kingmaker by supporting the candidacy of his nephew Aḥmad III, the grandson of Aḥmad Lobbo and the son of Aḥmad II. Ba Lobbo had used the Fulbe tradition of fraternal succession to support his own claims to the caliphate but faced new rivals within the second generation of Masinanke leadership when Aḥmad II died, and supported Aḥmad III to stymie the advances of his competitors and increase his own influence.53 According to Hampâté Bâ, Aḥmad III appointed Ba Lobbo emir of Masina to secure his support. And once he held the governorships of the regions of Masina and Fakala, as well as control of the army, Ba Lobbo could easily check his nephew.54 Thus, though it is widely assumed that Ḥamdallāhi’s defenses collapsed because of the deliberate inaction of military leaders, it should not be overlooked that Ba Lobbo lost a great deal of influence after Aḥmad III’s defeat.

26Ba Lobbo managed to escape the ‘Umarian occupation and form an alliance with Aḥmad al-Bakkāy, who led Timbuktu’s prestigious Kunta clan. He and his Kunta allies successfully besieged and occupied Ḥamdallāhi in 1863. Shortly after their victory, the Kunta and Masinanke leadership argued over who would enter the capital first and thus claim the right to rule the reconquered caliphate. The allies quickly turned on each other, and a tripartite war broke out that pitted the Kunta and their Fulbe allies against Ba Lobbo loyalists and a new Futanke-led army that regrouped under the leadership of al-ḥājj ‘Umar’s nephew, Tijānī. Throughout the nearly thirty years of warfare that followed, Ba Lobbo’s army remained hemmed into a territory largely populated by non-Muslim Bobo vassals east of the Bani, where the Barri family had first settled. Outnumbered by their enemies, the Futanke colonists formed alliances with non-Muslim Dogon and Bambara warriors and regrouped in the Hayre—the territory east of Ḥamdallāhi occupied by the non-Muslim Dogon and Tombo that the Masinanke continued referring to as habe.

Futanke alliances during the tripartite wars

  • 55 D. Robinson, 1985, p. 308.
  • 56 Ibid., p. 309.
  • 57 Field notes, Bandiagara, August 2015. See also, V. Petit, 1997, p. 215.

27The Futanke position during the 1863 siege of Ḥamdallāhi was dire. According to some accounts, food was so scarce that the city’s inhabitants were forced to eat dogs and cadavers to survive.55 Somehow, Tijānī managed to sneak out of the city with a large sum of gold that his uncle al-ḥājj ‘Umar had given him to recruit an army from among the Dogon who lived in the Hayre along what is today called the Bandiagara Plateau.56 The modern name is derived from the town of Bandiagara, a trading post that according to oral tradition was established by a Dogon hunter as early as the 14th century.57 Despite earlier objections to Ḥamdallāhi’s military relationship with the Bambara raised in the Bayān, circumstances had compelled the ‘Umarians to seek alliances with the non-Muslims of the Hayre.

  • 58 I. Barry, 1993, p. 401; E. Caron, 1891, p. 200.
  • 59 Ali Campo claimed that the Bambara prince of Bore was Tijānī’s first ally in the region, Interview (...)

28Tijānī’s coalition of auxiliary forces included Dogon and Bambara warriors, as well as Fulbe cavaliers. The first Dogon contingents were drawn from three villages along the plateaus east of Bandiagara: Kambarai, Ningari, and Kani. Through an agreement with Sala Baji, the chief of Kambari, Tijānī he drew the largest part of his forces from the Dogon villages from Kambari to Ningnari. Dogon warriors who fought for the chief of Kani, Gougouna Kansaye, also joined Tijānī’s coalition.58 Amadu Amiru, the Fulbe chief who ruled the village of De to Bandiagara’s east, provided horses and cavaliers, and the Bambara chief of Bore, Koniba Joko, provided soldiers.59 Tijānī’s new army was unable to save al-ḥājj ‘Umar, who eventually died in a cave in Degembere, but over the following two decades it would reconquer much of Ḥamdallāhi’s former territory.

  • 60 ANOM, SEN IV 114 a, “Lt. Caron to Col. Gallieni, November 1887”; see also E. Caron, 1891, p. 176.
  • 61 E. Caron, 1891, p. 183.
  • 62 Ibid.

29Tijānī’s military success was tied to the diversity of his forces. He strategically used gifts to maintain strong alliances, transgressing earlier Masinanke expectations of Dogon servitude by sharing the spoils of war with his Dogon allies and recognizing their leaders before others. For example, the Dogon chief of Bandiagara was the first to receive spoils of war after the Futanke.60 Tijānī also used gifts to court Fulbe warriors in the Hayre. In the accounts of his mission to Bandiagara in 1887, Lt. Edouard Caron describes how Tijānī was rebuffed when he attempted to give a horse to one of his local Fulbe allies. The Fulbe leader demanded an even finer horse than the one Tijānī had presented. According to Caron, this refusal offended some of the Futanke nobles present but drew no reaction from Tijānī.61 Tijānī’s gift and his diplomatic reaction to its refusal indicate that he recognized the value of his alliances with the Fulbe of the Hayre.62

  • 63 The Bozo were among the “servile” groups inherited by Aḥmad Lobbo, according to the rīkh al-fattā (...)
  • 64 E. Caron, 1891, p. 137.
  • 65 I. Barry, 1993, p. 475; E. Caron, 1891, p. 292. See also M. Johnson, 1976, p. 494.
  • 66 ANOM, SEN IV 144 a, “Report from Caron to Commandant supérieur Gallieni, Bamako, November, 1887”. C (...)

30One of Tijānī’s crucial innovations was his riverine navy, drawn from the Middle Niger’s Bozo fishermen. Members of this group were treated as crown slaves by Ḥamdallāhi. They paid a portion of their produce to the state and traded the rest of their catch for food produced by herders and cultivators.63 Tijānī transformed the Bozo fishermen’s pirogues into a fleet that could rapidly transport his troops up and down the Niger to attack his enemies. Tijānī’s admiral, Bubakar, a fisherman from the Gulf of Guinea, was also appointed as the chief of Mopti, a town at the confluence of the Niger and the Bani rivers.64 Tijānī’s fleet was instrumental in his systematic destruction of the Caliphate of Ḥamdallāhi. It brought his troops to the Burgu, where they mounted a devastating assault, burned villages, and took women and children captive.65 It is likely many of the people enslaved in the Burgu were resettled in the Jenneri: in a report to his commanding officer, Caron noted that Tijānī had settled the Fulbe of Masina and captives in and around Jenne.66 Bubakar’s fleet later traveled upriver to Lake Debo, where Tijānī’s forces burned dozens of villages and again took captives to resettle east of the Niger. Through nearly thirty years of warfare and enslavement, the hierarchical order established by the Masinanke disintegrated as the Futanke enslaved Fulbe Muslims and allied with “black” non-Muslims.

  • 67 I. Barry, 1993, p. 471.
  • 68 ANOM, SEN IV 114 A, “Lt. Caron to Col. Gallieni, November 1887”; see also E. Caron, Voyage, p. 176.

31Tijānī’s regime depended on non-Fulbe as administrators and as warriors. For example, after the chiefs of Guimballa submitted, Tijānī appointed the Bambara prince Koniba Joko as his representative in that region.67 Tijānī first approached the Dogon as mercenaries rather than crown slaves, and he continued to offer their leaders a share in the spoils of war. The Futanke colonists who settled among the Dogon at Bandiagara clearly achieved a fuller alliance with their non-Muslim neighbors than had been the norm. Dogon and Futanke leaders lived in compounds that butted up against one another, and each group was thus able to keep a close eye on the other. While the Futanke continued to draw on Dogon labor, Caron observed that Dogon leaders at Bandiagara were treated with a certain amount of respect.68

Masinanke Alliances

  • 69 W. Brown, 1969, p. 155.

32On the whole, the Masinanke royal family failed to maintain effective alliances during either the Futanke invasion or the wars that followed. The disagreement between the surviving members of the Barri family and the Kunta after their reconquest of Ḥamdallāhi reveals deep fractures within Masinanke society.69 The Fulbe from the inland delta territories of Masina, Fittuga, and Kunari were immediately drawn to Timbuktu’s Kunta clan when the Masinanke-Kunta coalition dissolved. On the other hand, the territories along the Bani close to where the Barri family first settled—Fakala, Sebere, and Jenneri—remained loyal to the royal family; notably, all had significant non-Fulbe populations that were subordinated to the royal family through claims made in the forged Tārīkh al-fattāsh.

  • 70 I. Barry, 1993, p. 407.
  • 71 For the execution of ‘Abd al-Salām, see W. Brown, 1969, p. 235, note 30. For the death of Ba Lobbo, (...)
  • 72 ANS 15G 172, “Rapport sur la délimitation des États d’Aguibou et Ahmadou Abdoul, 1894”.
  • 73 Using the Tarikh al-Fittuga, Brown dates the Battle of Nemende to 1874: W. Brown, 1969. Barry dates (...)

33After the breakup of the Masinanke-Kunta coalition, those Masinanke who had not allied with the Kunta clerics were once again split among themselves over who should become the head of the royal family. Some still felt that Ba Lobbo should lead, while others preferred Aḥmad II’s half-brother ‘Abd al-Salām, who had distinguished himself fighting the Futanke at the Battle of Cayawal. ‘Abd al-Salām and his son Ahmadou Abdoul were, along with Ba Lobbo, among the few members of the royal family to survive the Futanke conquest. According to Ibrahima Barry, those who preferred leadership to pass to ‘Abd al-Salām won only a minor victory when he was recognized as the official head of the royal family, insofar as Ba Lobbo remained the head of the military and thus the de facto chief.70 ‘Abd al-Salām was executed at Dalla in 1862, and after Ba Lobbo’s death in 1876 leadership passed to ‘Abd al-Salām’s son Ahmadou Abdoul, who took the title of Almami of Fio.71 In 1896, a colonial officer from Jenne, Lt. Montgard, visited Ahmadou Abdoul at Fio. Montgard wrote that Ahmadou Abdoul had betrayed Ba Lobbo and briefly joined the Kunta forces in the inland delta but later reconciled with his uncle and helped him support a Tombo uprising in Tijānī’s territory.72 Indeed, the Lobbo and the Kunta reconciled briefly and won a rare victory at the Battle of Nemende (c. 1872), just before Tombo and Dogon uprisings broke out near Bankass.73

  • 74 I. Barry, 1993, p. 446.
  • 75 Ibid.

34However, the Tombo uprisings backed by Ahmadou Abdoul were a short-lived and ultimately failed alliance of opportunity. Ahmadou Abdoul did not form a coalition of Fulbe and Tombo warriors or fold Tombo contingents into his regular army. The Masinanke were willing to form military alliances with non-Muslims when pressed by their enemies, but these arrangements did not transgress the racial and religious hierarchies established by the Caliphate of Ḥamdallāhi. For example, during an 1865 campaign against Ahmad al-Bakkāy in Masina, Ba Lobbo was chased across the Niger to the Bambara village of Kemitaga on the west bank of the Bani River. The Bambara understood that their hospitality toward Ba Lobbo, a prince on his heels after a losing campaign, put them in danger.74 Perhaps more importantly, other Bambara leaders had already joined Tijānī’s coalition. The village sent word to Tijānī that his enemy was with them, and the Futanke leader used this advantage to mount an offensive against the battle-weary Masinanke and push Ba Lobbo’s forces further south.75

  • 76 ANOM FM SOUD I cahier c, “Télégramme, Commandant supérieur à Gouverneur de Saint Louis, 12 Fév. 189 (...)

35Ba Lobbo and Ahmadou Abdoul took refuge in villages in the Boobola—the region inhabited by non-Muslim Bobo cultivators that the Masinanke considered crown slaves. The Masinanke maintained the status quo and taxed the increasingly restive Bobo, to the best of their ability. As the tripartite war ground on, the royal family was increasingly confined to the village of Fio in the Boobola and the Seno-Gondo Plains; and by the time colonial administrators arrived on the scene, the Masinanke at Fio were reduced to preying upon Dyula merchants who brought horses from the Upper Volta kingdom of Mossi through Fio to trade at San.76

Conclusions

  • 77 Y. Diallo, 1997, p. 87.
  • 78 A. Hubbel, 2001, p. 35.
  • 79 Y. Diallo, 1997, p. 75.

36The chaotic period that followed the collapse of the Caliphate of Ḥamdallāhi was marked by the emergence of new Fulbe-led states in the Hayre and Seno Gondo plain—Bandiagara, Dokwi, Barani, and Fio—that competed with one another for material wealth and territory. The Fulbe of the Boobola at Dokwi, Barani, and Fio continued to assert their supremacy over non-Fulbe neighbors they treated as their crown slaves. Citing colonial documents, historian Yousouf Diallo notes that at the end of the 19th century, Masinanke migrants who settled around the Fulbe villages of Dokwi and Barani on the Seno-Gondo Plain imposed their political domination over the Bwa and Bobo farmers of that region.77 Barani became a longtime ally of the Futanke settlers at Bandiagara and funneled Samo slaves from Sourodougu to Bandiagara’s slave market.78 At Dokwi, Fulbe cavaliers demanded millet, goats, chickens, and slaves from the Bobo villages under their control.79 Their expropriations were similar to the eastward campaigns of Ḥamdallāhi described above, and effectively maintained the hierarchy the caliphate established over the populations on its eastern fringe.

  • 80 ANS 15G 172, “Rapport sur la délimitation des États d’Aguibou et Ahmadou Abdoul, 1894”; ANS 1G 213, (...)
  • 81 ANS 1G 213, Lt. Montgard, “Mission à Ahmadou Abdoul…”.

37Similarly, the leaders of the Barri royal family at Fio continued to recognize the hierarchies established by Ḥamdallāhi. Ba Lobbo and Ahmadou Abdoul continued to view the Bobo as crown slaves whose labor and produce were state assets. Masinanke cavaliers at Fio harassed caravan traffic from the east and continued raiding near Barani; but the Bobo vassals did not share in the few spoils that the Masinanake managed to gain through their raids and banditry.80 Visiting colonial officers reported that the Bobo cultivators of Fio lived in abject poverty. Any wealth that the Masinanke cavaliers gained from pillaging caravans of Dyula merchants was augmented by taxes levied on Bobo farmers. Eventually, a Bobo uprising near Fio forced Ahmadou Abdoul to flee to Dokwi, and the colonial government brought his territory under its direct administration.81

  • 82 See especially H. Gemery and J. Hogendorn, 1974, p. 223-246. See also R. Roberts, 1980; C. Meillass (...)

38Scholars have described the Hayre and Seno-Gondo Plains as a kind of “slave reservoir” preyed upon by warrior states: polities in which warfare was the primary expression of state power and the primary form of enterprise.82 The present research confirms this: Fulbe-led armies conducted eastward raids that produced slaves that were sold into the trans-Saharan slave trade, but it is important to note that predation in the mid- to late 19th century was not limited to acquiring captives for external slave trading. Militarily, these Fulbe-led states also relied on non-Fulbe who were either courted as allies or coerced into service. The use of slave soldiers and vassals was a common component of warfare in Muslim West Africa, but the late 19th century Middle Niger was exceptional because several Fulbe-led armies employed strategies of alliance and coercion to augment their forces with non-Muslim warriors. As we have seen, the Futanke and Masinanke approaches to such strategies differed considerably.

  • 83 D. Robinson, 1985, p. 183.

39Aḥmad Lobbo might have relied on emancipated non-Muslim slaves in his initial campaigns, but jurists at Ḥamdallāhi eventually developed a legal category based on “otherness” to justify their conscription of Dogon and others. When it came to the use of “pagan” soldiers in warfare, the ‘Umarians proved themselves to be flexible despite their campaigns to oppose religious practice they considered anathema. Al-ḥājj ‘Umar himself established the jomfutung, and it was al-ḥājj ‘Umar who had sent Tijānī to the Hayre to recruit a contingent of Dogon warriors in the first place.83 Tijānī, Ahmadu, and other Futanke leaders after al-ḥājj ‘Umar continued to augment their ranks with trusted non-Fulbe slave soldiers called sofas, who sometimes rose to positions of authority. But the predominance of non-Futanke warriors within Tijānī’s coalition distinguished Bandiagara’s army from his Futanke contemporaries and the commanders of other Fulbe-led armies.

  • 84 IFAN, Fonds Vieillard, Cahier 5, Ibrahima Cam, “Souvenirs de Banjagara”. Bandiagara was sometimes a (...)

40For example, the Fulbe of Diankabu, who inhabited the plains just east of the Bandiagara Plateau, were willing to engage diplomatically with Mossi but continued to conscript the Dogon of the Seno-Gondo. According to Ibrahima Cam’s chronicle “souvenirs de Banjagara,” Bokari Haman Dikko, a Masinanke noble who emigrated to the Seno-Gondo region, established a residence at Diankabu by betraying Aḥmad Lobbo and allying with the Mossi—a move that defied Ḥamdallāhi’s diplomatic status quo. But once Diankabu was established, it depended on the maintenance of a hierarchical system that placed the local Dogon population under Fulbe rulers. Ibrahima Cam’s account of the founding of Diankabu suggests Bokari Haman Dikko’s son and successor Ali Bokari partnered with the Fulbe inhabitants of Seno-Gondo, who were Aḥmad Lobbo’s enemies, and convinced them that together they could subjugate the Dogon on the eastern edge of the Bandiagara Plateau.84

41Unlike their Fulbe enemies at Diankabu and Fio, the Futanke at Bandiagara grew their fighting force through alliance rather than coercion. At Bandiagara, the Dogon whom Ḥamdallāhi and Diankabu routinely conscripted were courted by the Futanke with gold. Bandiagara’s successes in its three decades of near-continuous warfare allowed Tijānī to maintain this alliance through the distribution of the spoils of war. As noted above, Dogon leaders received gifts before other allies. The Futanke also faced resistance from the Dogon and Tombo of the Hayre; but of all the Fulbe-led states that competed with one another after Ḥamdallāhi’s collapse, Bandiagara was the most successful.

  • 85 Field notes, Bandiagara, August 2015. There are photos of the Dogon mosques in the palace of Aguibu (...)
  • 86 ANM ID-5, “Notice générale sur le Soudan - Notice sur la région Sud 1895–1899”; ANM 2M-56, Cercle d (...)
  • 87 C. Monteil, 1903, p. 108; C. Monteil, 1932, p. 152.

42Under the leadership of Tijānī, the Futanke at Bandiagara strategically transgressed the racial and religious hierarchy established by Ḥamdallāhi. Their courtship of Dogon and Bambara warriors reversed their earlier position on military alliances with non-Muslims, as expressed in al-ḥājj ‘Umar’s Bayān. Tijānī made some effort to encourage the Islamization of the Dogon but was careful not to disrespect his allies. He encouraged the building of mosques in Dogon villages but allowed certain non-Muslim religious and judicial practices to continue at Bandiagara.85 In the Futanke garrison capital, a Dogon court was overseen by the Dogon’s traditional Hogon priests, while Muslim law applied to the Futanke.86 Despite this accommodation, Tijānī enforced a strict and austere practice within the Futanke community at Bandiagara and among other Muslims he ruled. At Jenne, for example, he put religious leaders in place who aggressively policed Muslim practice.87 The different approaches in these two cities suggests that the parallel religious and judicial institutions that continued at Bandiagara were tolerated for reasons of political expediency rather than of religious or political ideology. By accommodating non-Muslim allies in this and other ways, the Futanke built a strong coalition that eventually defeated their Masinanke enemies.

Haut de page

Bibliographie

Akalay, L., and Alessandra, J., 2020, Les voyages d'Ibn Battûta, Charleroi, Dupuis.

Bâ, A. H., Daget, J., 1984 [1955], L’Empire peul du Macina (1818-1953), Abidjan-Paris, Les Nouvelles Éditions africaines-Éditions de l’EHESS [Dakar, Institut français d’Afrique noire, « Études soudanaises », n° 3].

Barry, I., 1993, Le royaume de Bandiagara, 1864-1893: Le pouvoir, le commerce et le Coran dans le Soudan nigérian au 19ème siècle, PhD Dissertation, École des hautes études en sciences sociales.

Baumgardt, U., 1994, “La représentation de l'Autre. L'exemple du répertoire d'une conteuse peule de Garoua (Cameroun)”, Cahiers d'études africaines, 34 (133/135), p. 295-311.

Bazin, J., 1975, “Guerre et servitude à Sigou”, in C. Meillassoux (ed.), L'esclavage en Afrique précoloniale, Paris, François Maspero, p. 135-181.

Bott, R., Schmitz, J., “Paradoxes identitaires”, Cahiers d'études africaines, 34 (133/135), p. 7-22.

Brown, W., 1969, The Caliphate of Hamdullahi ca. 1818–1864, PhD Dissertation, University of Wisconsin.

Caron, E., 1891, De Saint-Louis au port de Tombouktou: voyage d’une canonnière française, Dijon, Imp. Darantiere.

Diallo, Y., 1997, Les Fulbe du Boobola, Köln, Köppe.

Dupire, M., “Identité ethnique e processus d’incorporation tribale et étatique”, Cahiers d'études africaines, 34 (133/135), p. 265-280.

El Hamel, C., 2013, Black Morocco: A history of slavery race, and Islam, Cambridge, Cambridge University Press.

Gemery, H., Hogendorn, J., 1974, “The Atlantic slave trade: A tentative economic model”, Journal of African History, 15 (2), p. 223-246.

Hall, B.S., 2011, A history of race in Muslim West Africa, 1600–1960, Cambridge, Cambridge University Press.

Hanson, J., 1997, “Islam, ethnicity and Fulbe-Mande relations”, in M. de Bruijn, H. van Dijk (eds.), Peuls et Mandingues: dialectiques des constructions identitaires, Paris, Karthala, p. 85-97.

Hanson, J., 1996, Migration, jihad, and Muslim authority in West Africa, Bloomington, Indiana University Press.

Houdas, O., Delafosse, M. (eds. and trans.), 1913, Tārīkh al-fattāsh fī akhbār al-bldān wa-al-juyūsh wa-akābir al-nās, Paris, E. Leroux.

Hubbel, A., 2001, “A view of the slave trade from the margin: Souroudougou in the late nineteenth-century slave trade of the Niger bend”, The Journal of African History, 42 (1), p. 25-47.

Irwin, P., 1981, Liptako speaks: History from oral tradition in Africa, Princeton, Princeton University Press.

Khaldūn, A., Juwaydī, D., 2015, Muqaddimat Ibn Khaldūn, Beirut: al-Maktabah al-ʻAṣrīyah.

Johnson, M., 1976, “The economic foundations of an Islamic theocracy – The case of Macina”, The Journal of African History, 17 (4), p. 481-496.

Levtzion, N., 1971, “A seventeenth-century chronicle by Ibn al-Mukhtār: A critical study of Ta’rīkh al-fattāsh”, Bulletin of the School of Oriental and African Studies, 34 (3), p. 571-93.

Mahibou, S.M., Triaud, J.L., 1983, Voilà ce qui est arrivé – Bayân mâ waqa`a d’al-Hâgg `Umar al-Fûtî: plaidoyer pour une guerre sainte en Afrique de l’Ouest au xixe siècle, Paris, Centre régional de publication de Paris, Éditions du Centre national de la recherche scientifique.

Mayor, A., 1997, “Les rapports entre la Diina peul du Maasina et les populations du Delta intérieur du Niger, vus au travers de traditions historiques et des fouilles archéologiques”, in M. de Bruijn, H. van Dijk (eds.), Peuls et Mandingues : dialectiques des constructions identitaires, Paris, Karthala, p. 33-60.

Meillassoux, C., 1991, The anthropology of slavery: The womb of iron and gold, Chicago, The University of Chicago Press.

Monteil, C., 1903, Monographie de Djenné, cercle et ville, Tulle, Imp. Mazeyre.

Monteil, C., 1932, Une cité soudanaise. Djenné, métropole du delta central du Niger, Paris, Société d’Éditions Géographiques, Maritimes et Coloniales.

Nobili, M., Malthee, M.S., 2015, “Towards a new study of the so-called Tārīkh al-fattāsh”, History in Africa, 4, p. 37-73.

Nobili, M., 2020, Sultan, caliph and the renewer of the faith: Aḥmad Lobbo, the Tārīkh al-fattāsh and the making of an Islamic state in West Africa, Cambridge, Cambridge University Press.

Ochonu, M., 2014, Colonialism by proxy: Hausa imperial agents and middle belt consciousness in Nigeria, Bloomington, University of Indiana Press.

Perinbam, B.M., 1997, “‘Animist’/Islamized imaging in the western Sudan”, in M. de Bruijn, H. van Dijk (eds.), Peuls et Mandingues: dialectiques des constructions identitaires, Paris, Karthala, p. 101-128.

Petit, V., 1997, “Société d’origine et logiques migratoires: les Dogon de Sangha (Mali)”, Population, 52 (3), p. 515-543.

Reyna, S.P., 1990, Wars without end: The political economy of a precolonial African state, Hanover, University of New Hampshire/University Press of New England.

Roberts, R., 1980, “Production and reproduction of warrior states: Segu Bambara and Segu Tokolor, c. 1712-I890”, International Journal of African Historical Studies, 13 (3), p. 389-419.

Robinson, D., 1985, The holy war of Umar Tal, Oxford, Oxford University Press.

Sanankoua, B., 1990, Un empire peul au xixe siècle: la Diina du Maasina, Paris, Karthala.

Syed, A., 2020, “Between jihād and history: Reconceptualizing the Islamic revolutions of west Africa”, in F. Ngom, M.H. Kurfi, T. Falola (eds.), The Palgrave handbook of Islam in Africa, London, Palgrave.

Webb, J., 1995, Desert Frontier: Ecological and economic change along the Western Sahel, 1600-1850, Madison, University of Wisconsin Press.

Zehnle, S., 2015, A geography of jihad, jihadist concepts of space and Sokoto warfare (west Africa ca. 1800–1850), PhD Dissertation, Universität Kassel.

Haut de page

Notes

1 A pastoral people who have migrated across the African Sahel (sg. Pullo, pl. Fulbe, adj. Fulbe) They are often referred to as Peul in French sources, and Fula or Fulani in British sources.

2 The main reference work for the history of Ḥamdallāhi is A.H. and J. Daget, 1984.

3 D. Robinson, 1985 is the main secondary source on the history of al-ḥājj ‘Umar’s movement.

4 D. Robinson, 1985, p. 121-126.

5 I use the term jihad to describe wars, but only by way of pointing out that certain belligerents thought of their efforts as fitting into this category. Jihad is not a fixed concept in Islam and jihads must be understood in their individual contexts. See, A. Syed, 2020, p. 107-108.

6 The word “pagan” appears in quotes to signal that this was an historic argument about unbelief applied to Muslims and non-Muslims alike.

7 J. Hanson, 1996, p. 9.

8 D. Robinson, 1985, p. 81-89.

9 L. Alkaly and J. Alessandra, 2020; I. Khaldūn and D. Juwaydī, 2015.

10 J. Webb, 1995, p. 22-26.

11 B.S. Hall, 2011, p. 52-53.

12 C. Hamel, 2013.

13 A.H. Bâ and J. Daget, 1984; D. Robinson, 1985; M. Dupire, 1994; R. Botte and J. Schmitz, 1994; B.S. Hall, 2011; M. Ochonu, 2014.

14 A. Mayor, 1997, p. 38.

15 Ibid., p. 39.

16 A.H. and J. Daget, 1984, p. 280.

17 Malian Arabic Manuscript Microfilm Project, Tārīkh Jenne (Yale University, 1976), reel 7.

18 Jenne likely submitted to Ahmad Lobbo’s authority shortly after the second siege in 1821; see A.H. and J. Daget, 1984, p. 151.

19 Using non-Muslim Bobo archers would also have allowed Ahmad Lobbo to sidestep Islamic prohibitions against using poison in warfare. See W. Brown, 1969, p. 204, note 51.

20 B.S. Hall, 2011, p. 70-73.

21 The textual history of the Tārīkh al-fattāsh was long debated by Arabists of an earlier generation. The authors on the only available edition and translation (O.V. Houdas and M. Delafosse, 1913) argued that the chronicle was a 16th century chronicle slightly modified in later centuries. Accepted by many, this theory was criticized by N. Levtzion, 1971 who convincingly argued that the available edition combined a 17th-century document and a version of it that was manipulated in the 19th century to look like an 16th-century one. Based on previously unexamined documentary evidence, however, Mauro Nobili and Mohamed Shahid Mathee argued convincingly that the Tārīkh al-fattāsh is a 19th-century manuscript in its entirety, and that its authorship by the 16th-century scholar Maḥmūd Ka‘ti was falsified by the 19th-century scholar Nūḥ b. al-Ṭāhir to lend legitimacy to Aḥmad Lobbo and forge “inherited” rights over Ḥamdallāhi’s servile populations: M. Nobili and M.S. Mathee, 2015. For a full story of the Tārīkh al-fattāsh, see M. Nobili, 2020.

22 Institut de France, Fonds de Gironcourt, ms. 2406 (56-60) / XVI, Aḥmad Lobbo, Manshūra fī istirqāq al-zanj.

23 B.S. Hall, 2011, p. 71, note 5.

24 ANOM FP Fonds Gaden, “Personal correspondence, Bandiagara 20-25 June 1895”; emphasis mine; for modern understanding of the word habe among the Fulbe in neighboring Liptako, see P. Irwin, 1981, p. 14.

25 S. Zehnle, 2015, p. 173.

26 U. Baumgardt, 1994, p. 295-311.

27 S. Zehnle, 2015, p. 174.

28 M. Ochonu, 2014, p. 9.

29 M. Johnson, 1976, p. 486.

30 ANS 15G 172, “Rapport sur le délimitation des États d’Aguibou et Ahmadou Abdoul, 1894”; ANS 1G 213, Lt. Montgard; ANS 1G 213, “Mission à Ahmadu Abdoul Roi de Fion”, 1896.

31 D. Robinson, 1985, p. 282.

32 In the instances where names did not appear in Arabic sources, I have written them as they appear in French sources.

33 D. Robinson, 1985, p. 263.

34 For a description of Aḥmad III and al-Bakkāy’s opposition to the ‘Umarian conquest of Segu, see D. Robinson, 1985, p. 262-263.

35 S.M. Mahibou and J.L. Triaud, 1983.

36 B.M. Perinbam, 1997, p. 102; J. Hanson, 1997, p. 85-97.

37 Bakhunu was a region to Kaarta’s east with significant Soninke population. Al-ḥājj ‘Umar argued that this Soninke population and others nearer to Segu were not Muslim because they mixed Islamic and “pagan” practice. See S.M. Mahibou and J.L. Triaud, 1983.

38 al-ḥājj ‘Umar constructed the Soninke as muḥarubūn (warriors). Triaud and Mahibou have focused on how al-ḥājj ‘Umar used the word muḥarubūn to emphasize his argument that Muslim practice among the Soninke was lax. In the passage below, al-ḥājj ‘Umar stresses that the muharibun are warriors who practice banditry. They “take from unbelievers and Muslims alike” references their indiscriminate theft rather than mixing religious practice.

39 The following translation from Arabic is my own, from the original text published in S.M. Mahibou and J.L. Triaud, 1983, p. 99: fol. 12 verso; emphasis mine.

40 S.M. Mahibou and J.L. Triaud, 1983, p. 107-138.

41 D. Robinson, 1985, p. 295.

42 D. Robinson, 1985, p. 298.

43 B. Sanankoua, 1990, p. 155.

44 D. Robinson, 1985, p. 287.

45 A.H. and J. Daget, 1984, p. 248-249. Bâ communicated his conclusion that the Ḥamdallāhi was divided between a military and a scholarly elite directly to William Brown: see W. Brown, 1969, p. 231, note 9.

46 A.H. and J. Daget, 1984, p. 249.

47 C. Monteil, 1932, p. 112.

48 Yousouf Diallo argues that Ḥamdallāhi’s interventions in the Boobola had a significant economic dimension and cannot be understood solely in the contexts of Islamization: Y. Diallo, 1997, p. 188.

49 A.H. Bâ and J. Daget, 1984, p. 72.

50 Ibid., p. 266.

51 Ibid., p. 265-266.

52 D. Robinson, 1985, p. 297.

53 Aḥmad II’s four brothers—ʿAbd Allāh, Muḥammad, ‘Abd al-Salām, and ‘Abd al-Raḥmānwere all potential successors, according to the practice of fraternal succession. ʿAbd Allāh, also known as Allaye, posed the greatest threat to Ba Lobbo. He was revered for his erudition and supported by Nūḥ b. al-Ṭāhir, the author of the Ta’rīkh al-fattāsh. See W. Brown, 1969, p. 231, note 10, and A.H.  and J. Daget, 1984, p. 286.

54 Citing Ba, Brown notes that Aḥmad III “was completely under the control of the military aristocracy”: W. Brown, 1969, p. 229, note 54.

55 D. Robinson, 1985, p. 308.

56 Ibid., p. 309.

57 Field notes, Bandiagara, August 2015. See also, V. Petit, 1997, p. 215.

58 I. Barry, 1993, p. 401; E. Caron, 1891, p. 200.

59 Ali Campo claimed that the Bambara prince of Bore was Tijānī’s first ally in the region, Interview with Ali Campo, Mopti, 2015. W. Brown also notes that certain Bambara villages in the Kunari enjoyed political power over Fulbe and Songhay populations until Aḥmad Lobbo established the Caliphate of Ḥamdallāhi, and that these Bambara rulers returned to power through their alliance with Tijānī Tal (Ahmad al-Tijan). W. Brown, 1969, p. 53. See also I. Barry, 1993, p. 402.

60 ANOM, SEN IV 114 a, “Lt. Caron to Col. Gallieni, November 1887”; see also E. Caron, 1891, p. 176.

61 E. Caron, 1891, p. 183.

62 Ibid.

63 The Bozo were among the “servile” groups inherited by Aḥmad Lobbo, according to the rīkh al-fattāsh. Marion Johnson notes that the Bozo who had supported Aḥmad Lobbo from the start were exempt from the jamgal tax placed on crown slaves, but that the rest were considered crown slaves, and that all Bozo were considered crown slaves by the Futanke: M. Johnson, 1976, p. 483. See also C. Monteil, 1903, p. 297, 338.

64 E. Caron, 1891, p. 137.

65 I. Barry, 1993, p. 475; E. Caron, 1891, p. 292. See also M. Johnson, 1976, p. 494.

66 ANOM, SEN IV 144 a, “Report from Caron to Commandant supérieur Gallieni, Bamako, November, 1887”. Caron also stopped at a village near Mopti inhabited by “Peuls de Macina transportés”: ANOM, SEN IV 114 a, “Report from Caron to the Chef de bataillon, Commandants de cercles, Commandant supérieur 31 October 1887”.

67 I. Barry, 1993, p. 471.

68 ANOM, SEN IV 114 A, “Lt. Caron to Col. Gallieni, November 1887”; see also E. Caron, Voyage, p. 176.

69 W. Brown, 1969, p. 155.

70 I. Barry, 1993, p. 407.

71 For the execution of ‘Abd al-Salām, see W. Brown, 1969, p. 235, note 30. For the death of Ba Lobbo, see I. Barry, 1993, p. 478-484.

72 ANS 15G 172, “Rapport sur la délimitation des États d’Aguibou et Ahmadou Abdoul, 1894”.

73 Using the Tarikh al-Fittuga, Brown dates the Battle of Nemende to 1874: W. Brown, 1969. Barry dates the same battle to 1871–72: I. Barry, 1993.

74 I. Barry, 1993, p. 446.

75 Ibid.

76 ANOM FM SOUD I cahier c, “Télégramme, Commandant supérieur à Gouverneur de Saint Louis, 12 Fév. 1891”; ANS 1G 149, “Mission Spitzer au Mossi, 1890”.

77 Y. Diallo, 1997, p. 87.

78 A. Hubbel, 2001, p. 35.

79 Y. Diallo, 1997, p. 75.

80 ANS 15G 172, “Rapport sur la délimitation des États d’Aguibou et Ahmadou Abdoul, 1894”; ANS 1G 213, Lt. Montgard, “ANS 1G 213, Mission à Ahmadu Abdoul Roi de Fion, 1896”.

81 ANS 1G 213, Lt. Montgard, “Mission à Ahmadou Abdoul…”.

82 See especially H. Gemery and J. Hogendorn, 1974, p. 223-246. See also R. Roberts, 1980; C. Meillassoux, 1991; S. Reyna, 1990; J. Bazin, 1975.

83 D. Robinson, 1985, p. 183.

84 IFAN, Fonds Vieillard, Cahier 5, Ibrahima Cam, “Souvenirs de Banjagara”. Bandiagara was sometimes able exact a tribute from Diankabu, but its leaders routinely sought to undermine Futanke control in their region.

85 Field notes, Bandiagara, August 2015. There are photos of the Dogon mosques in the palace of Aguibu Tal in Bandiagara. These were supposedly built in Dogon villages for two reasons: firstly, so the Dogon could avoid the taxes placed on non-Muslims; and secondly, because Tijānī hoped the mosques would encourage conversions.

86 ANM ID-5, “Notice générale sur le Soudan - Notice sur la région Sud 1895–1899”; ANM 2M-56, Cercle de Bandiagara, 1905–1920; I. Barry, 1993, p. 534.

87 C. Monteil, 1903, p. 108; C. Monteil, 1932, p. 152.

Haut de page

Table des illustrations

Titre Figure 1: The Middle Niger
Crédits Joseph M. Bradshaw
URL http://journals.openedition.org/afriques/docannexe/image/3243/img-1.jpg
Fichier image/jpeg, 204k
Haut de page

Pour citer cet article

Référence électronique

Joseph M. Bradshaw, « Race and military coalitions in the Middle Niger during the second half of the 19th century »Afriques [En ligne], 12 | 2021, mis en ligne le 25 décembre 2021, consulté le 22 mai 2022. URL : http://journals.openedition.org/afriques/3243 ; DOI : https://doi.org/10.4000/afriques.3243

Haut de page

Auteur

Joseph M. Bradshaw

Indiana University, Bloomington

Haut de page
  • Logo Institut des mondes africains
  • DOAJ - Directory of Open Access Journals
  • Revue soutenue par l’Institut des sciences humaines et sociales du CNRS
    CNRS - Institut national des sciences humaines et sociales
  • OpenEdition Journals
Rechercher dans OpenEdition Search

Vous allez être redirigé vers OpenEdition Search