EMPIRES AND IMPERIAL TRADITION IN THE EXTERNAL AND INTERNAL ARABIC SOURCES BEFORE THE NINETEENTH CENTURY

The literary Arabic sources for the western Sahel are external until almost 1500.Footnote ⁴ The texts about the Bilād al-Sūdān, i.e. the area south of the Sahara, were translated and published in compendia, the most famous of which are the Recueil by Joseph Cuoq and the Corpus by Nehemia Levtzion and John Hopkins.Footnote ⁵ Together, they present roughly 600 pages of patchy and selective sources dating from the eighth to the fifteenth century. While relatively rich compared to sources for many other regions of Africa, the corpora are nevertheless highly limited given the time span and the connection of the region to the Islamic worlds.

Manuscript culture fully developed in West Africa in the postmedieval period.Footnote ⁶ The first internally written histories of the past of the western Sahel were produced in the middle of the seventeenth century; they are the Tārīḫ al-Sūdān by al-Saʿdī and the Tārīḫ by Ibn al-Muḫtār. This is where our search for empires begins. As Paulo Fernando de Moraes Farias demonstrates, the authors of the chronicles drew from various written and oral materials sourced from different cultural contexts. This heterogeneous material, carrying different regimes of truth, was rearranged into linear narratives reconstructing and interpreting the West African past.Footnote ⁷ Put simply, such narratives were created by historians.

Like modern historians, the authors of the chronicles devised narratives about the Sahelian past with little documentation at hand. Thanks to extant written and oral narratives, they were aware that the region had known a considerably higher number of great states than the three great states of Ghāna, Māli, and Songhay. This reality is vivid in the chronicles. During the era of Songhay hegemony in particular, a myriad of secondary sultans and rulers are reported; West Africa was then a complex political mosaic.Footnote ⁸ Nonetheless, to devise a prestigious political tradition that the Timbuktu-based Arma power could be part of, they distinguished Ghāna, Māli, and Songhay as the upholders of a linear political trajectory that would culminate with the advent of a proper Islamic Sahelian urban society.Footnote ⁹

This construction of the ‘medieval’ past was emic in its nature. The Timbuktu chronicles circulated widely in the erudite Islamic culture of West Africa. As a result, their devisement of the past shaped the historical representations of learned men during the following centuries. As Mauro Nobili's recently published study on the Tārīḫ al-Fattāš shows, at the beginning of the nineteenth century, Nūḥ Ibn al-Ṭāhir inscribed the newly formed Islamic state of his master Aḥmad Lobbo, the founder of the caliphate of Ḥamdallāhi, into this political tradition.Footnote ¹⁰

How was ‘imperiality’ expressed in these chronicles? The authors used general and vernacular terms for positions of power (e.g. malik, Arabic for ‘king’, and koï, Songhay for ‘ruler or owner of a high function’). However, when they wanted to be more specific, they used the terms sultan (sulṭān) and sultanate (sulṭana), which had a particular meaning in the Islamic context in which they were living and writing. An Islamic sultanate was superior to a mere kingdom, for a sultan usually had other rulers under his power. After the fall of the Abbasid Caliphate in 1258, the sultanate became the new form of imperiality in Islam. Consequently, in medieval sources on West Africa produced after the thirteenth century, imperiality and imperial narratives were expressed through the conceptual framework of the Islamic sultanate. Not long after Mansa Mūsā visited Cairo's citadel in 1324, the polymath al-ʿUmarī was able to record fragments of a speech given by the Malian sultan to Mamlūk officials.Footnote ¹¹ Mūsā made such an impression on his contemporaries that few years later, al-ʿUmarī registered Māli among the great sultanates of Islam of the fourteenth century in his encyclopedia Masālik al-abṣār fi mamālik al-amṣār. Mansa Mūsā's description of his state reveals that he had intimate knowledge of the Dār al-Islām. He gave a geopolitical portrait of his lands that described his wars with barbaric infidels through jihads, mimicking that of his Mamlūk counterpart and his wars with the Mongols.Footnote ¹² As Gabriel Martinez-Gros demonstrates, fighting barbarians was a key characteristic of Ibn Khaldūn's conception of Islamic empires:

Indeed, Arabic authors’ descriptions of court etiquette at the capitals of Ghāna and Māli give the impression that courtly elites displayed a certain ‘world order’ to audiences comprised of courtiers, officials, and visitors.Footnote ¹⁴ That said, the barbaric/civilized dichotomy constructed by medieval authors must be mitigated, since intensive trade networks connected these two worlds. The mansa of Māli himself embodied several political and religious practices coexisting peacefully. We must also read such dichotic discourse through its performative intent: helping the interlocutors to understand the power and magnitude of Māli by using their ways of understanding the world. Similarly, in the case of Songhay, Bruce Hall's work shows that debates pertaining to imperiality were deeply rooted in the remit of political Islam.Footnote ¹⁵

Was political filiation in the western Sahel, as displayed in the Timbuktu chronicles, experienced by the elites of the medieval period? To a degree, cultural and political inheritance was certainly handed down from one state to the other. Ghāna was granted a special status in the Māli sultanate, and it can be argued that much of its power etiquette was passed on to the mansas of Māli.Footnote ¹⁶ Likewise, the subsequent shifting of hegemony from Māli to Songhay did not erase Māli's legacy in the Niger Bend. When the Tārīḫ of Ibn al-Muḫtār was written, the memory of Māli was still very much alive, especially its Islamic heritage.Footnote ¹⁷ The Muslim elite still remembered what they owed to Māli's dominion over the region, while Mansa Mūsā's pilgrimage was still the subject of numerous anecdotes.Footnote ¹⁸ However, if the political filiation between Ghāna and Māli appears incontrovertible, the one between Māli and Songhay is subject to discussion. The remnants of the Māli sultanate, reorganized as a smaller kingdom after the fifteenth century, were often at war with the Songhay.Footnote ¹⁹ A late medieval source points out to the use of Islamic legal and political arguments by the sultan of Songhay, Askia Muhammad, rather than the local tradition of inherited imperial political power to claim sovereignty over distant Muslim groups.Footnote ²⁰ Written sources do not allow us to demonstrate with certainty if the askias of the Songhay inscribed themselves in an imperial tradition dating back to Ghāna.Footnote ²¹ What is certain is that, in the middle of the seventeenth century, a historical narrative of great medieval sultanates spread from the Niger Bend through West Africa as a means of creating an intelligible past, and that it was still alive in local Islamic historical cultures during the early decades of the nineteenth century. At the end of the nineteenth century these states were made empires by colonial scholars — the new power producing knowledge about the past of West Africa. As we shall see, it was not merely a matter of translation. The concept of empire carried ideological and epistemological baggage.

LESSER EMPIRES: MEDIEVAL EMPIRES OF WEST AFRICA AT THE BIRTH OF AFRICAN STUDIES IN MODERN ACADEMIA (NINETEENTH CENTURY–1940s)

The constitution of medieval West African studies as a European academic discipline started in the first half of the nineteenth century and is generally dated more precisely to 1841, when the British geographer William D. Cooley published The Negroland of the Arabs. Cooley was the first to attempt a historical narrative of the Bilād al-Sūdān using the Arabic medieval texts available at the time.Footnote ²² In his work Cooley does not refer to empires. In 1858 Heinrich Barth first proposed the Ghāna-Māli-Songhay trinity that would come to frame the medieval past of the western Sahel. He based his work on fragments of the Tārīḫ al-Sūdān, linking local developments to foreign influences, but still did not speak of empires.Footnote ²³ On the French side, colonial officer Louis Gustave Binger spoke in 1886 of Māli as a ‘vast empire’, while in 1888 the linguist René Basset called Ghāna and Songhay ‘empires’.Footnote ²⁴ Between 1898 and 1900, the Arabist Octave Houdas edited and translated the famous Timbuktu chronicle Tārīḫ al-Sūdān.Footnote ²⁵ He still did not use the word ‘empire’ to translate the term ‘sultanate’, preferring ‘kingdom’. At the time, the chronicle fascinated Europeans, but as Moraes Farias shows, it was considered merely an invaluable repository of historical facts rather than a sophisticated ‘intellectual innovation and reinvention of the Sahel’.Footnote ²⁶

Flora Shaw was among the first to exploit the chronicle, together with Arabic medieval texts, in A Tropical Dependency. She qualified Māli and Songhay as empires but Ghāna as a kingdom.Footnote ²⁷ It was the French Africanist Maurice Delafosse who, in his landmark study Haut-Sénégal-Niger, was the first architect of the construction ‘Sudanic empire’.Footnote ²⁸ As Jean-Louis Triaud remarks, ‘empires are everywhere in [Delafosse's] history’:

Delafosse listed many West African polities from before his time as empires, not just Ghāna, Māli, and Songhay. He presented the Sudanic empire as the most refined political organization in West African history.Footnote ³⁰ At the time, the French École Méthodique, known for its highly political, factual, and linear historical narratives, was the dominant historiographic school.Footnote ³¹ As Triaud underscores, Delafosse's historical discourse, full of ‘dates, events, kings, and empires’, is an embodiment of the history writing of the school.Footnote ³² In 1913, when Houdas edited and translated the so-called Tārīḫ al-Fattāš, Delafosse was involved as a co-editor.Footnote ³³ His influence is striking, for in the translation of this chronicle the word ‘empire’ was used everywhere.Footnote ³⁴

The Europeans’ implicit model was the Roman Empire and its capital. Hence, the identification of African imperial capitals became their main concern, if not obsession.Footnote ³⁵ In the process, the ‘question of Ghāna’ led to the reclassification of Ghāna as an empire, since the ruins of Kumbi-Saleh, identified as its capital, were deemed worthy of empire.Footnote ³⁶ The Europeans’ encounter with such a glorious past led to the creation of a ‘historical exception’ for the medieval western Sahel in keeping with the ideological tenets of the colonial era.Footnote ³⁷ The French administrator and historian Charles Monteil, the author of the first monograph on the Māli Empire in 1929, also published in that year ‘L’œuvre des étrangers dans l'empire soudanais du Mali’, an article in which he defined Sudanic empires.Footnote ³⁸ Monteil's narrative can be puzzling since compared to his contemporaries he sometime demonstrates a critical attitude, but nonetheless it reverberates with colonial ideas, leading to some contradictions. For instance, he stated that the disappearance of Sudanic empires left societies ‘barely different from their anterior state’, and a few lines later emphasized some of their legacies.Footnote ³⁹

Monteil's reasoning was threefold. First, he showed skepticism about using the concept of empire for West African polities and advocated that, even if they were so called, they were not comparable to European empires. Second, he characterized the Sudanic empire as a ‘simple mode of exploitation of peoples’, the might of which depended solely on the emperor and his army. However he had to acknowledge the textual and material evidence suggestive of an empire (e.g. monuments, ruins). He could not deny Sudanic empires any form of accomplishment. This led him to his third argument. Drawing from both the theory of the inequalities of human races and a Kulturkreis type of diffusionism, Monteil explained the ‘historical exception’ of western Sudan with a cultural and genetic diffusion from the Islamic north.Footnote ⁴⁰

After Delafosse's death in 1926 and Monteil's publications in the 1920s, the literature on the past of West Africa became scarce during the remaining years of the French occupation. From the emic construction of an Islamic political tradition based on the Timbuktu chronicles, the Europeans periodized the age of great empires that came to define the medieval period. Epistemologically speaking, the Sudanic empires and their history were cast by a European historiographical current that transplanted European concepts and ideas of history to an African context. Ideologically speaking, Sudanic empires were considered lesser empires compared to those of Europe, and their few acknowledged achievements were attributed to the northern civilization of Islam.Footnote ⁴¹

FULLY-FLEDGED EMPIRES: THE ‘GOLDEN AGE’ OF MEDIEVAL WEST AFRICAN EMPIRES AT THE TIME OF POLITICAL INDEPENDENCE (1950s–70s)

The years following the Second World War brought the onset of a shift in ideology.Footnote ⁴² By the time of West Africa's political independence in 1960, many ideological aspects of the history written during the colonial period were already judged coarse, undignified, and racist. Decolonizing historical knowledge was the scholar's priority. A critical inflexion and an inversion of values came from West African historians who, in the wake of nation building and the reappropriation of African history, erected the Sudanic empires as the embodiment of a golden age, conceived in opposition to a medieval Europe torn by war and disease.Footnote ⁴³ The Guinean historian Djibril Tamsir Niane, whose work was prominent in the 1960s and the 1970s, was the great promotor of Sudanic medieval empires. In his narrative, they were powerful centers of attraction:

Niane mainly focused his work on the Māli Empire. His translation of a Guinean version of the Sunjata epic in 1960, Soundjata ou L'épopée mandingue, was universally acclaimed. It was regarded for more than two decades as the canonical version of the epic.Footnote ⁴⁵ He was among the pioneers who systematically incorporated orature into their work as means to Africanize historical knowledge on West Africa.Footnote ⁴⁶ Yet his work demonstrates great conceptual continuity with the previous period. As we can see in the extract above, the idea of a ‘historical exception’ characterizing the medieval western Sahel was still present. Committed to the search for the imperial capitals, he even strengthened the paradigm of Niani as capital of Māli, an idea forged in the scientific colonial literature by Jules Vidal and Delafosse.Footnote ⁴⁷ During the archeological excavations directed by Władysław Filipowiak at Niani in 1965 and 1968, Niane was the principal historical advisor. The great empires Ghāna-Māli-Songhay were the focus of his work, reinforcing them as synonyms for medieval West Africa as a whole. In the post-1960 context of pan-Africanism and the building of nationalist history, he did more than ratify this chrononym; he gave it a soul and substance.Footnote ⁴⁸

The inversion in ideology and the promotion of empires benefited from the powerful momentum of the independence movement of the 1960s. The empires, having unified in the past peoples of different cultures and languages under one domain that brimmed with peace and prosperity, were perceived in spirit as centralized yet inclusive political models from which the newly formed republics could draw inspiration. Stigma became insignia of pride, and the legacies of the imperial tradition were incorporated into national histories. The main epistemological inflexion of that time was the preeminence given to orature in the construction of historical knowledge.Footnote ⁴⁹ Yet oral narratives were not impervious to the powerful epistemological frameworks that shaped the history of medieval West Africa.Footnote ⁵⁰

RECASTING EMPIRES: RECENT DEBATES AND THE QUEST FOR RECOGNITION (1980s–PRESENT)

The new status gained by the Sudanic empires in the field of African history made them a candidate for recognition by world imperial history. In 1980 the French Africanist George Balandier contributed a chapter titled ‘Le concept d'empire dans l'histoire de l'Afrique noire’ to a comparative history book on world empires.Footnote ⁵¹ First regretting the absence of African history in debates on empire, he raised the problem of definition and the inevitable tacit reference to Eurocentric imperial models.Footnote ⁵² Paradoxically, he then proceeded to argue that African societies overall only produced what he called ‘incomplete empires’ (empires inachevés), a Eurocentric statement in itself.Footnote ⁵³ This assessment applied to the whole continent. Balandier singled out one exception:

Finally, Balandier mitigated Monteil's militarist argument by stating that if we consider empires in the longue durée, they were spaces of economic control and cultural diffusion, especially regarding religion, rather than of military domination.Footnote ⁵⁶ As we can see, even though the status of Sudanic empire improved at the time of Balandier's writing, some earlier conceptions were difficult to leave behind (e.g., the agency of the north in its development and the idea of a ‘historical exception’).

In the late 1970s, the historical field of medieval West African studies lost its momentum. The publication of the two compendia, the Recueil and the Corpus, suggested that most of the Arabic medieval sources had already been discovered, which considerably slowed down the search for new texts.Footnote ⁵⁷ Scholars working on West Africa chose other temporalities; the study of the contemporary period became most prominent.Footnote ⁵⁸ Medieval studies were limited to a handful of specialists, notably John O. Hunwick and Nehemia Levtzion. In this context, two disciplines animated the reflection on empires: anthropology and archaeology. The Māli Empire received most of the attention. Because it was so intricately interwoven with the roman national of the Republic of Mali and because the Mande people carried through time the epic of its foundation, the history of medieval Māli gradually eclipsed that of Ghāna and Songhay in terms of scientific publications. The Sunjata epic, relating the birth of the Māli Empire, became the focal point of the discussions. A critical turn affected the approach to the Sunjata epic, visible in the seminal In Search of Sunjata, whereby the focus shifted to the cultural and political remnants of the Māli Empire in postmedieval greater Manden.Footnote ⁵⁹ Anthropology has been at the forefront of challenging the historical doxa of West African imperial history. A special issue of History in Africa in 1996 addressed this particular topic. When studying the liveliness of medieval Māli's political inheritance, Jan Jansen provocatively asked, ‘Did the Mali empire still exist in the nineteenth century?’Footnote ⁶⁰ Similarly, archaeologists contributed to challenging a history of empires far too dependent on political perspectives. In this respect, Roderick J. McIntosh's Peoples of the Middle Niger formed an important milestone in 1998 by putting forward the concept of heterarchy.Footnote ⁶¹

In the Republics of Mali and Guinea, from the 1990s the Māli Empire became a substratum exploited to address questions of governance, heritage, tourism, and the place of the African past in universal history, especially regarding the question of human rights.Footnote ⁶² As Éric Jolly has observed, in the context of failing decentralization and aspirations to greater democracy arose the ‘contemporaneous invention of the medieval Manden charter’, a declaration on human rights chronologically rooted in the empire's foundation.Footnote ⁶³ Gregory Mann sheds light on what was at stake in the 1990s when this body of texts rose to prominence:

The Manden charter (or charter of Kurugan Fuga) underlines the proclivity for summoning the so-called West African golden age to produce material for present aspirations.Footnote ⁶⁵ The Māli empire appears more as a federation of kingdoms and provinces with some autonomy, and not as a monarchist powerful state centralized around the capital, giving a political precedent inclusive and respectful of its cultural and political diversity.Footnote ⁶⁶ As Mann argues, ‘the Kurugan Fuga offered a tool for thinking through this process of “returning” power to smaller communities and village confederations’ and an alternative path to the European-inspired, capital-centered, and democratic nation-state which had failed to be established in a sustainable way.Footnote ⁶⁷

At the beginning of the twenty-first century, medieval history gained a new momentum in the wake of several groundbreaking publications.Footnote ⁶⁸ The door was opened to a more critical and reflexive history willing to confront its historiographical inheritances. The last decade further accentuated this renewal of studies with contributions straying away from the tripartite imperial framework, such as the fragmented narrative laid out by François-Xavier Fauvelle in his Golden Rhinoceros, Sam Nixon's edited volume on Tadmekka, or Alain Gallais’s alternative political models for West Africa before the nineteenth century.Footnote ⁶⁹ In this dynamic context, medieval empires also returned to the foreground.

Yet world imperial history did not change its relationship with African empires from before the nineteenth century. When Jane Burbank and Frederick Cooper published their acclaimed Empires in World History, medieval empires of West Africa were only mentioned anecdotally.Footnote ⁷⁰ A worldwide renewal of interest in a ‘new imperial history’ overlooked Africa.Footnote ⁷¹ As a reaction, endeavors from both Francophone and Anglophone academia strived to insert African empires into world history.Footnote ⁷² To ‘relocate this history from the periphery to the center of world history’ was an explicit goal of Michael A. Gomez's African Dominion.Footnote ⁷³

African Dominion was the first book of considerable length on the matter in a long time. For this reason, it has sparked a wide interest and has already been thoroughly discussed since its publication.Footnote ⁷⁴ Returning to the history of empires was, however, tricky. Since the publication of the Recueil and the Corpus, the repository of sources had remained stable and had already been extensively exploited. In the last two decades, specialists on the topic consciously distanced themselves from grand narratives to avoid the pitfalls of the political history framed during the first decades of the twentieth century. Instead they explored new ways to relate a history drawing on a fragmented primary source base dominated by elite points of view.

So how did African Dominion approach this history? Gomez takes us almost immediately in medias res in the prologue, addressing what he calls the ‘influential template’ of the linear Ghāna-Māli-Songhay succession.Footnote ⁷⁵ Pointing out that considering Ghāna the starting point of West African imperial tradition was merely scholars’ uncritical repetition of the past devised by the Timbuktu chronicles, he then chooses the kingdom of Gao instead, explaining its importance using Arabic sources.Footnote ⁷⁶ His argument can be strengthened by the recent unique discovery in Gao of the first fortified palace built with stones and its archeological treasure, which bears witness to a tremendous accumulation of wealth there before the eleventh century.Footnote ⁷⁷ Indeed, Gao does not compare unfavorably to Ghāna. Nevertheless, why choose again a new starting point? In other centers of Arabic historical literature in West Africa, a similar status was given to other early kingdoms. The historiography of Sokoto, for instance, favored the historical Takrūr (valley of the Senegal River). In his history of Takrūr (i.e. Muslim West Africa) Infāq al-maysūr fī tārīḫ bilād al-Takrūr, Muhammad Bello placed the beginning of West Africa's Muslim political history with the historical Takrūr in the far west, a kingdom long forgotten.Footnote ⁷⁸

It might be more accurate to state that before the eleventh century the Sahel was home to a myriad of kingdoms — of which Ghāna, Gao, and Takrūr were the most powerful — that underwent different trajectories in both West African and European historiographies. Gao lost its identity as a separate kingdom as its history was diluted in the genesis of the Songhay Empire depicted in the Timbuktu chronicles, whereas Ghāna and Takrūr were more fortunate. Challenging this ‘influential template’ forged during the colonial period is a necessity. At the time of Songhay, for example, Māli was still a powerful kingdom acting as a model and a starting point for new political endeavors and their narratives, as in the case of the Kaabu federation or the kingdom of Gonja.Footnote ⁷⁹ If we consider the wider region of postmedieval West Africa, the so-called imperial tradition of Ghāna-Māli-Songhay was one of many political traditions that developed around these three powerful states and others in orature and Arabic literature. Each of them had their own reading of the past before 1600.

The concept of empire and its historiographical trajectory, not discussed in a dedicated section, appear scattered throughout African Dominion.Footnote ⁸⁰ The notion is discussed, for instance, upon broaching the Sunjata epic:

In order to legitimize the Sunjata epic as a source for medieval history, Gomez offers three interpretive categories: ‘historical developments corroborated by independent sources’, ‘developments posited by the oral corpus and unsubstantiated by sources of unrelated provenance, yet registering within a range of historical plausibility’, and traditions whose ‘didactic quality’ enabled them to articulate social narratives that gave historical significance to relationships between groups.Footnote ⁸² The question of the Sunjata epic as a source for medieval history is itself a subfield of study. In the last three decades, as stated before, scholars working on the epic have come to realize that its current versions were certainly formed in postmedieval times. They shed light on how Manden societies made use of the political inheritance of Māli in order to position themselves in a complex network of political and social claims and statuses.Footnote ⁸³

The Ghāna-Māli-Songhay trinity emerges almost unscathed from African Dominion. The idea of a medieval imperial age, with ‘West Africa at its pinnacle’, is even strengthened.Footnote ⁸⁴ It is undeniably a powerful argument to insert West Africa into the conversation on both world imperial history and world history, a process that only truly started in the last few years.Footnote ⁸⁵ The book certainly serves its greater purpose. On the other hand, the idea of a ‘historical exception’ and its subsequent avatars (i.e. the golden age, the apex of a history) remain the corollaries of the age of empires. First elaborated in colonial times as a means to create a void between a glorious age and the arrival of the French — who took upon themselves to revive these societies with their so-called civilization after centuries of alleged decadence — the idea of ‘historical exception’ must be challenged. It is all too dependent on a certain political reading of the past and can be counterbalanced. From the perspective of Arabic manuscript culture and Islamic intellectual production, for instance, West Africa appears at its pinnacle during postmedieval centuries. The city of Timbuktu, the showcase of West African history, experienced its intellectual peak and radiated its influence in the decades following the medieval period.Footnote ⁸⁶ The same can be said of the great city of Djenné.Footnote ⁸⁷ Until the European conquest, on the whole West Africa continued to be politically, socially, and culturally very innovative and dynamic.