Scholars, Secrets, and Sultans: Clerical Authority in West Africa, 1450–1650

Abstract

Available historical sources for West Africa's Middle Niger c. 1450–1650 reveal that the ‘indigenous’ (non-Arab) Islamic scholarly class was already a self-conscious, independent social entity long before the clerical revolutions of later centuries. The influence of Muslim scholars was not limited to urban environments like Timbuktu, and clerical elites claimed a number of mostly independent communities throughout West Africa by the end of the sixteenth century. Mostly based on a reading of Arabic texts such as Muḥammad al-Kābarī's Bustān al-fawāʾid (‘Garden of Beneficial Prayers’) in dialogue with the Tārīkh Ibn al-Mukhtār and Tārīkh al-Sūdān (‘Timbuktu Chronicles’), this article argues that Muslim scholars were engaged in a spiritual war for independence clearly on display since the beginning of the Songhay empire. Scholarly texts display deep concern for tempering unjust political power and the protection and attraction of women, discourses that reveal a perilous clerical struggle to assert community independence. Later armed jihads were thus not so much a break from earlier traditions of clerical pacificism, they were the natural evolution from this earlier spiritual jihad.

On the eve of the collapse of the Songhay empire, the distinguished West African Muslim jurist Maḥmūd al-Kaʿti (d. 1593) advised the Songhay Askiya Isḥāq II to evacuate the population of Gao in advance of the invading Moroccan army. A political advisor protested, ‘The jurists know nothing of the matters of war, they only know the Book and recitation’. Al-Kaʿti warned, ‘If you refuse and discard my advice, there will come upon you a grim day’. And later, after hearing of the defeat of the king (askiya) the scholar remonstrated, ‘If the askiya had taken my advice, he would not have suffered such loss’. Indeed, the source from which this account derives, the Tārīkh Ibn al-Mukhtār, concludes, ‘It was when the affairs of Songhay had become corrupt that God dissolved their polity … the people of Songhay neglected the rights of God and oppressed His servants (ʿibādihi)’.¹

There is an essential corrective in this key passage to a century of mistranslation, which previously rendered ʿibād (God's worshipful servants) as ‘slaves’ (which would have been ʿabīd in Arabic) and thereby blamed the fall of Songhay on ‘the wickedness of slaves’.² Both the Tārīkh Ibn al-Mukhtār and Tārīkh al-Sūdān, the two most important chronicles of the Songhay empire, in fact frequently mention famous Muslim scholars as ‘among the virtuous servants of God’,³ and Muslim scholars were certainly the ‘servants of God’ most capable of fulfilling the rights of God's worship. The message could not be clearer: the fall of Songhay, according to the narrative of medieval Muslim scholars in West Africa, was the result of misguided rulers who abandoned standards of justice and oppressed the ʿulamāʾ (Islamic scholarly class). The Songhay chronicles, then, situate themselves in a much larger Islamic scholarly discourse linking the injustice of rulers to political decline.⁴ The Moroccan invasion of 1591 and the collapse of West Africa's last ‘grand empire’, these chronicles suggest, was thus God's punishment for the rulers’ betrayal of the clerical establishment.

It is no doubt true that the seventeenth-century Arabic histories of Ibn al-Mukhtār and al-Saʿdī center the perspectives of scholars at the expense of the political establishment, and thereby diminish the very real power of one of the most significant empires in the premodern world. According to Michael Gomez, these texts exaggerate the ‘autonomy’ of scholarly centers ‘while granting them a measure of political clout they did not actually wield’. Gomez admits, however, that religious elites did ‘exercise extraordinary influence’, mostly through their ‘performance of knowledge’ and possession of baraka (worldly blessing).⁵ Given the later history of clerical revolutions in West Africa from late seventeenth century through the late nineteenth century, this observation should give historians pause. In truth, the performance of knowledge was inseparable from ritual authority and could be a revolutionary act; and certainly, the claim to baraka was also a claim to social and political power. The Arabic Songhay ‘chronicles’ are thus much more than ‘straightforward political histor[ies] of the Askias’⁶: they consciously made the claim to Islamic scholarly authority that may have laid the foundation for later clerical revolutions.

The available historical record for the Middle Niger from roughly 1450 to 1650 reveals that the ‘ulamā’ was already, prior to the age of clerical revolutions beginning in the late seventeenth century, a self-conscious entity with a very real social and political positionality. The influence of Muslim scholars was not limited to urban environments like Timbuktu, and clerical elites claimed a number of mostly independent communities throughout West Africa by the end of the sixteenth century.⁷ Indeed, according to Bruce Hall, the debate over the political marginality or relative influence of scholars mistakenly presumes that ‘Islam was a distinct social and political force that was fundamentally foreign to the cultures and societies of the Sahel’.⁸ Sources from this period demonstrate that the Muslim clerisy, like other indigenous elites negotiating a balance of power with the Songhay empire, hoped to secure a degree of both influence and political autonomy that could temper or outlast perceived tyranny and political instability. Although later clerical revolutions would prove that this attempt was ultimately unsuccessful through any other means besides armed confrontation, the narrative of clerical independence at the height of the Songhay empire provided a foundation for later attempts to create clerical states.

This article situates a rereading of Islamic scholarly positionality in medieval West Africa within a number of sources. First and foremost is a roughly seventy-page manuscript of ‘secrets’, the Bustān al-fawāʾid wa l-manāfiʿ associated with the foundational Timbuktu scholar Muḥammad al-Kābarī (fl. 1450),⁹ probably the earliest extant Arabic text from West Africa. This source is read in dialogue with internal histories of West African state formation, namely the Tārīkh al-Sūdān and the Tārīkh Ibn al-Mukhtār. Significantly, the latter is based on the forthcoming Arabic reconstitution and new translation of this key source, unavailable in its original form since the text's reformulation as the Tārīkh al-fattāsh to support the revolution of Aḥmad Lobbo in the early nineteenth century.¹⁰ Epigraphic and archeological evidence can also suggest insights into state-society relations in the Songhay empire, and it is used where appropriate. Moreover, recorded oral traditions of various clerical lineages are employed for substantiating evidence, although it should be admitted such traditions, like the earlier textual record, obviously represent the perspective of Muslim scholars or their clients. Given the ascendancy of the clerical establishment in many regions of West Africa since the latter half of the nineteenth century, we must expect that earlier or alternative voices have been mostly marginalized in the predominant textual or oral records. Generally, then, our sources cannot hope to represent a comprehensive picture of state-society relations in the Songhay empire. But they can offer significant insights into the self-perceived positionality of the clerical class as an influential faction within that society.

The overall conclusion from this reading is that Muslim scholars claimed independence from, and perhaps even ultimate authority over, the political establishment. Indeed, they were effectively in competition with political elites for social influence in West Africa even at the height of the Askiya dynasty. This argument is based in two types of evidence: Islamic scholarly spiritual warfare against political corruption, and aspirations for augmented social influence through building communities, mostly through attracting and protecting women. In sustained analysis of Timbuktu's earliest extant Arabic manuscript, this study enriches academic understanding of state-society relations in the environs of Timbuktu at the dawn of the Songhay empire in important ways. Prior research — whether situated in reading of the ‘Timbuktu chronicles’, recorded oral traditions, or the archeological and epigraphic record — has had difficulty understanding the unmediated concerns, aspirations, and ritual expertise of the Muslim scholarly class. Moreover, earlier scholars such as John Hunwick subtly privileged Sanhaja or Berber writings — represented in works like al-Saʿdī's Tārīkh al-Sūdān and Aḥmad Bābā's Kifāyat al-muḥtāj — over the articulations of preexisting Soninke/Mande Muslim scholarly communities, found in the writings of Muḥammad al-Kābarī and Ibn al-Mukhtār al-Qunbili. The ‘esoteric’ or ‘talismanic’ practices recorded in al-Kābarī's Bustān al-fawāʾid thus provide unique insight, when read in dialogue with other sources, into the stakes of indigenous community independence, as reflected in concerns over political authority and female sexual agency. Before discussing this text, some background to the positionality of the Muslim clerisy in West Africa is warranted.

Muslim scholars in the western Sudan to the seventeenth century

Muslim scholarly lineages and knowledge genealogies both predated and outlasted the rise and fall of dynastic powers in West Africa. For example, the ‘five maraboutic clans of the Malinke’ (Berté, Touré, Cissé, Jané, and Saghanogho/Baghayoro), which were instrumental in the rise of Sunjata Keita and the Malian empire, likely had prior connections to the ancient Soninke Ghana/Wagadu kingdom.¹¹ Many of these same lineages — ‘extended family networks with unique spatial and social range’¹² — retained influence within the Songhay empire, and some remain productive of Islamic scholarly expertise to the present day.¹³ The appearance of famous scholarly names on tombstones in the Middle Niger from the thirteenth century — even those imported from elsewhere, like that of al-Ghazālī¹⁴ — may in fact speak to the power of Muslim scholarly attribution to exert independent ritual or even political authority. If this were true, the longevity of certain family names associated with Islamic scholarship speaks as much to clerical clientage networks as it does to genealogical descent. The claim to an Islamic scholarly lineage might have ensured a degree of independence from the caste regime of West African societies,¹⁵ but it no doubt required a genuine commitment to the craft of Islamic scholarship. This in turn allowed for the successive reenactment of Islamic learning and ritual expertise from one generation to the next even in the context of scholarly diasporas throughout West Africa.

Indeed, there seems little evidence that powerful empires and monarchs, including even the renowned Mansa Mūsā of the Mali empire, played a determinative role in the settlement or diffusion of scholarly communities. If the Sudānī (‘Sudanese’) scholar Muḥammad al-Kābarī, for example, was said to have settled in Timbuktu in the mid-fifteenth century as a result of Mansa Mūsā's ‘indigenization project’ initiated a century earlier,¹⁶ why did he choose to stay under the Tuareg takeover of the city from 1434? More likely, scholars from Kābara (or Kabora) expanded into Timbuktu as part of a larger scholarly strategy of founding and participating in purified ‘cities of God’ in the model of Kābara, Dia/Diagha (Arabized as Jaʿba in the Tārīkh Ibn al-Mukhtār), and Kunjawr. Such towns were places of refuge from the perceived injustices of sultans, where jurists could govern affairs according to the precepts of Islamic law. According to the Tārīkh Ibn Mukhtār, ‘Timbuktu did not permit any ruling except the ruling given by the one entrusted with the sacred law. It had no sultan, for the judge was the sultan, and in his hand was the sole authority to loosen and bind’.¹⁷

The establishment of such scholarly communities provided continuities in the articulation of Islamic identities that outlasted the fall of empires. After the collapse of Songhay and the imposition of the increasingly indigenized Moroccan ‘Arma’ rule, scholars returned to Timbuktu and ‘restored its glory and prosperity … blessings multiplied until the people almost forgot the Songhay dynasty’.¹⁸ Through the blessing of such righteous habitations, ‘God brought plentiful rain and caused agriculture to flourish’. Peace prevailed: ‘God extinguished the fire of enmity and rancor between the people and the Arma’.¹⁹ If Timbuktu scholars so readily ‘forgot’ the last great empire of West Africa and reconciled themselves to foreign rule, it was because West African Muslim scholarly communities had their own agendas that had existed in dynamic tension with political authorities for centuries.

A comprehensive history of various Muslim scholarly lineages is beyond the scope of this discussion, but one example of a scholarly community active outside of Timbuktu during the Songhay dynasty should suffice in demonstrating the longstanding relationship of prominent clerical lineages to political elites. The Soninke/Mande lineage of Mori Muḥammad Hawgāro plays a particularly important role in the Tārīkh Ibn Mukhtār in illustrating the dynamic tension between Muslim scholars and state power. The scholarly heritage of Mori Hawgāro, likely originating in the village of Yara on the southern bank of the Senegal river,²⁰ is absent from the Timbuktu-Sanhaja-centric account of al-Saʿdī's Tārīkh al-Sūdān. But the Tārīkh Ibn Mukhtār — emphasizing as it does state-society relations in the Songhay empire outside of Timbuktu²¹ — appears to largely mirror oral traditions emphasizing the lineage's influence at the foundation of the Songhay empire. Mori Hawgāro is described in the text as a ‘pious and righteous jurist’ and a great-grandfather of scholars contemporary to Askiya Muḥammad, thus suggesting a lifespan in the first half of the fifteenth century.²² According to oral traditions collected in the town of Morikoyra, said to contain Hawgāro's grave,²³ ‘Mori [Hawgāro] was the most famous of all saints’. He endowed later scholars of Morikoyra with the enviable position as advisors to the Askiya dynasty: ‘We are not soldiers with bows, rifles, or spears. We are only people [armed] with words. We are only students. We are the marabouts of the Songhay. They inform us of everything they do’.²⁴ As a ‘source of potent spiritual power’,²⁵ the Morikoyra scholars represented the ‘southern factor’ of ‘early clerical communities known in West Africa from which later and better-known traditions, such as that in Timbuktu, have benefitted’.²⁶

While it appears unlikely that Mori Hawgāro was the direct maternal cousin of Askiya Muḥammad as claimed in oral traditions,²⁷ the scholars of Morikoyra shared an apparent Soninke ancestry with Askiya Muhammad.²⁸ Proximity to the Songhay elite meant that even the alleged tyrant king Soɲi ʿĀli could not avoid sending gifts to Morikoyra, in the form of a captive princess who would be married to Mori Hawgāro's brother.²⁹ Those incensed at the injustices of Soɲi ʿĀli also turned to the scholars of Morikoyra, at one point asking them to declare Soɲi ʿĀli an unbeliever.³⁰ This they finally did when Soɲi ʿĀli had two prominent descendants of Mori Hawgāro, Mori al-Ṣādiq and Mori Ḥabīb placed in chains and marooned to die on a deserted island in the middle of the Niger river. The prayer of these two Morikoyra scholars against Soɲi ʿĀli is recorded as the immediate cause of the Soɲi's death.³¹ The subsequent kindness of Askiya Muḥammad,³² who allegedly cursed the reign of any of his own descendants who contravened the askiya's promise of inviolability to the Hawgāro scholars,³³ may in fact have been the askiya ‘repaying them for their active opposition to the Sunni [Soɲi]’.³⁴

But the warning of Askiya Muḥammad was not enough to protect his descendants from violating the sanctity of the Hawgāro lineage. A later scholar associated with the Morikoyra tradition, Mori Magha Kankoi,³⁵ was said to have cursed the apparent miscreant Askiya Mūsā (r. 1529–31), first warning him, ‘If you bring our efforts to naught and reject our ḥurma (sanctity), then those hands which are still raised to God Most High in prayer for you will be raised to Him to curse you’.³⁶ Despite witnessing otherwise unseen lions perched on both of Kankoi's shoulders ready to pounce on him, Mūsā continued his murder of Muslim scholars and other innocents,³⁷ thereby realizing Kankoi's curse against him. Such curses may have been part of the fawāʾid (powerful prayers) that Kankoi taught his many students after having studied in the town of Kābara;³⁸ indeed al-Kābarī's Bustān al-fawāʾid contains specific prayers to destroy unjust sultans (see below).

The Soninke/Mande Islamic scholarly tradition appears as invested with significant spiritual power. A constant refrain in oral accounts from Morikoyra thus reminds the listener to respect such esoteric authority:

Morikoyra, at the time of Songhay … was always a village of marabouts. Their rifles, lances, and arrows were comprised in their pens. Because of their pens, they were called ‘right away’ … [this was because] every time these marabouts invoked God for someone, the result was instantaneous … No one could achieve victory [on the battlefield] without their having shared in the victory.³⁹

In other words, Morikoyra scholars could call upon the direct action of God through their pens, their ability to write secret prayers invoking the power of the Qur’ān. For their prayer services and sanctity, scholars were to be respected and provided sustenance. Some oral traditions insist that the askiyas were ‘disciples’ of the Morikoyra, and ‘for one's master, who taught him the religion, the disciple must give everything he has. Even his own self belongs to the master. Nothing can compensate the master for the work he had done for the student’.⁴⁰ The Tārīkh Ibn Mukhtār emphasizes Askiya Muḥammad's discipleship to the chief judge of Timbuktu, ʿUmar Aqīt, who reprimands the sultan for forgetting his earlier entreaty to the scholar: ‘Remember the day you came to my house? You grasped my feet and held on to my garments, and you said, “I have come seeking refuge in your sanctity and entrusting myself to you, that you may come between me and the Hellfire. Help me!”’ Arabic histories and oral traditions may differ on the details, but the message is the same. The success of rulers is found in turning to scholars, and their destruction is found in turning away from them.

The Islamic esoteric sciences in West Africa

Relations between sultans and scholars in West Africa cannot be unraveled without reference to the ‘prayer services’ of Muslim scholars, expertise for which they were both courted and feared. Studies of West African Muslim societies have generally used the term ‘Islamic Esoteric Sciences’ to refer to the set of practices glossed elsewhere as the ‘occult sciences’,⁴¹ corresponding to prevalent regional invocations of the Arabic word sirr (plural asrār) to describe ʿilm al-ḥurūf (‘lettrism’ or letter magic), ṭilsamāt (talismans), ruqyā (incantations), awfāq (‘magic squares’), al-kimiya (alchemy), ʿilm al-nujūm (astrology), ʿilm al-raml (geomancy), and ṭibb (particular methods of healing), among other specializations.⁴² Muslim Salafi reformists and French colonial officials often associated such practices with Sufism or the allegedly syncretistic practices of Islam noir (‘black Islam’),⁴³ but in fact they represented a separate field of knowledge that has been ‘an integral part of the vibrant Islamicate intellectual and scientific enterprise over many centuries and over vast territories’, and considered by philosophers such as Ibn Sīna (d. 1037) and Abū Ḥāmid al-Ghazālī (d. 1111) to be part of the ‘natural sciences’.⁴⁴ According to al-Ghazālī, ‘The fifth [branch of natural sciences] is the science of talismanic things, namely the combining of celestial powers with the powers of some terrestrial bodies so as to have composed from this a power that enacts some strange acts in the terrestrial world.”⁴⁵ Talismanic science is thus differentiated from siḥr (sorcery) or nīranjāt (magic) as it combines the ‘celestial’ with the ‘terrestrial’ rather than magic's practice of ‘mixing of the powers of [only] earthy substances so as to produce strange things’.⁴⁶

There is no doubt that Islamic esotericism engaged in a robust dialogue with non-Islamic ritual knowledge in West Africa, with Islamic symbols influencing non-Islamic practices, and Muslims learning from non-Muslim experts. As an example from a later period, the great grandson of the West African scholar-warrior ʿUmar Tāl, Cerno Bokar Tāl (d. 1939), studied and worked with a ‘pagan healer’ of the Malian Dogon people named Ancamba: ‘From the time he met Cerno Bokar the two of them began to cure people together, Cerno from the Arabic perspective, and Ancamba from the Dogon perspective, with his knowledge of plants. They joined the two approaches to cure quite a few illnesses’.⁴⁷ According to another reading, Bokar Tāl thus ‘translated his [Ancamba's] knowledge of plants and cures into Islamic analogies, thus integrating this knowledge within a Muslim framework’.⁴⁸ A comparative study of Yoruba Ifa religion and Sufism observes: ‘Ifa bears a striking resemblance to the Islamic occult sciences … Ifa could be seen as a pre-Islamic tradition of asrār, as an occult science. Alternatively, Ifa could also be seen as a creative Yoruba adaptation of these Islamic divinatory sciences’.⁴⁹ More fascinating still is the propensity for different traditions to position important Muslim leaders differently within their own esoteric traditions. ʿUmar Tāl's mystical powers were read as karamāt (saintly miracles) by Muslim intellectuals; but the caste of woodworkers, of which Tāl's father may been a member, interpreted his ability to summon an army of jinn as testimony to his ritual powers as part of their special caste knowledge, linked to pre-Islamic relations with the spirits of trees.⁵⁰ Similarly, the Arabic ‘History of Ibn al-Mukhtār’ positions the esoteric power of Askiya Muḥammad as a result of his Islamic piety;⁵¹ while the Songhay language oral account, ‘The Epic of Askia Mohammed’, claims that the first askiya derived his esoteric power from his father, who was a chief of the jinn ruling a town under the water in the Niger river.⁵²

The Askiya dynasty appears to have drawn on both Islamic esoteric power as well as preexisting ritual experts. The later Moroccan historian Muḥammad al-Ṣaghīr al-Ifrānī (d. 1745) narrates Askiya Isḥāq II's battle against the invading Moroccan force as follows:

He [Askiya Isḥāq] was not satisfied with the forces he had assembled until he had added to them the leaders in witchcraft and those who blow on knots, as well as the masters of talismans and magic, thinking that these would save him. How wrong he was! The sword speaks more truthfully than books … their lances and spears were of no avail against gunpowder.⁵³

These loaded words appear to bear an internal contradiction: the reason for Songhay's defeat was more likely the lack of gunpowder rather than a misuse of the Islamic esoteric sciences. In fact, Moroccan scholarship of the period was well versed in this tradition,⁵⁴ and West Africans were no more prone than their North African counterparts to draw upon ritual practices outside of the Arabic textual tradition. Indeed, if we are to accept al-Ifrānī's testimony about the Songhay state's use of witchcraft, we should also accept the testimony of al-Saʿdī a century earlier that the Moroccan scholar who accused Muḥammad al-Kābarī of kufr (infidelity) himself died the death of an infidel for having resorted to eating human hearts to cure leprosy.⁵⁵ Although the internally contested nature of occult practices certainly lend themselves to theological condemnation, the reality this occasional contestation obscures is the fact that Muslims on both sides of the Sahara participated in a wide range of occult practices.

Indeed, the field of Islamic esotericism appears almost unique among the Islamic sciences for its ecumenicism. Prominent Muslim jurists such as Muḥammad al-Shāfiʿī (d. 820, Cairo) and Shihāb al-Dīn Aḥmad al-Nafrāwī al-Mālikī (d. 1713, Cairo) licensed the use of occult healing practices from non-Muslim sources, with the latter observing, ‘whatever is proven beneficial cannot be kufr’.⁵⁶ This idea may have first been expressed by the Prophet Muḥammad himself, who told his companions: ‘Your [pre-Islamic] incantations (ruqyā) have been made known to me. There is no harm in what does not contain idolatry’.⁵⁷ Such permeability seems to have influenced the exchange of Islamic ‘secrets’ across the boundaries of Sufi orders. When the noteworthy Moroccan Tijānī scholar Aḥmad Sukayrij (d. 1943) was criticized by fellow Tijānīs for sharing secrets with the Qādirī-Fādilī Saharan scholar Māʾ al-ʿAynāyn (d. 1910), he replied, ‘The litanies are specific to the respective Sufi orders, but the secrets are shared between the saints’.⁵⁸ The idea of ‘sharing secrets between saints’ underlines another important element of the Islamic occult sciences: they were meant to be transmitted through initiation. Even if Islamic esotericism was a ubiquitous and vibrant tradition, practitioners normally insisted on the withholding of secrets from the unworthy, those of polluted intentions, and the uninitiated.⁵⁹ Most academics, following anthropologists such as Benjamin Soares, have subsequently avoided ‘discuss[ing] the contents of such sciences except in rather general terms … largely for ethical reasons’.⁶⁰ An opposing trend in the emerging field of ‘Islamic magic’ — based almost entirely on the reading of texts disconnected from the living communities who produced those texts — freely discusses the initiatory secrets of these traditions in academic publications.⁶¹

Muḥammad al-Kābarī's Bustān al-fawāʾid wa l-manāfiʿ, a title which loosely translates as ‘The Garden of Beneficial and Useful Prayers’, is mostly a work concerning the Islamic esoteric sciences.⁶² Analysis of this manuscript is based primarily on the 71-page copy contained in the Northwestern University Herskovits Collection, with supporting reference to a partial 29-page copy in Timbuktu's Ahmad Baba Collection.⁶³ The two manuscripts consulted here, the former probably copied in Nigeria in the nineteenth century and the latter transcribed in Mali in the early twentieth century, are not identical.⁶⁴ Nonetheless, they are remarkably similar for a single manuscript that has never been published and has circulated throughout West Africa for over five hundred years. While most of the included prayers are based solely on the Qurʾān, others appear to have a non-Islamic correlation (see below). Paradoxically for an apparent work on ‘secrets’, al-Kābarī's title, ‘Beneficial and Useful Prayers’, appears to target a wider, more general audience beyond an initiated elite of fellow esotericists. In his introduction to the text, al-Kābarī writes of his book, ‘may God grant benefit in it for us and for all Muslims’.⁶⁵ This leaves the researcher (who might ordinarily be inclined to respect the initiatory preferences of esotericists) in a curious ethical dilemma. Some of the included prayers closely mirror prayers designated elsewhere as ‘secret’, and would clearly be subject to censure by contemporary Muslim reformists. On the other hand, al-Kābarī did not seem to limit the audience of his work in his own time, suggesting that there was a wide currency of ‘special prayers’ in medieval West Africa, and that al-Kābarī's own scholarly credentials would have silenced any criticism. Alternatively, he may have relied on the difficulty of acquiring and reading the text to limit the audience. A similar precept was communicated by the Egyptian esoteric master Aḥmad al-Shinnāwī (d. 1619), who reportedly said of his lengthy commentary on Muḥammad al-Ghawth's al-Jawāhir al-khams, ‘Anyone who finds this book has permission to use it’.⁶⁶ Despite my own preference to keep secret the precise content of Islamic esoteric sciences in accordance with the wishes of practitioners, this analysis of al-Kābarī's text takes the author at his word that a lay readership might appreciate the prayers it contains. Nonetheless, the reader is reminded that the practices referenced below, no matter how ‘strange’ to modern sensibilities, were laid out by one of the most prominent experts of Islamic law in his time, described as follows in al-Saʿdī's Tārīkh al-Sūdān: ‘Modibbo Muḥammad al-Kābarī attained the very pinnacle of scholarship and righteousness’.⁶⁷

Al-Kābarī claims that he is not really the author of an original text, but that he has only ‘gathered the beneficial and useful knowledge that I found among the wise men of divine provision, and which I transmitted from the books of scholars and the righteous’.⁶⁸ While al-Kābarī does not mention his exact textual references, some exploration of earlier Islamic esoteric influences in West Africa is warranted. One obvious source could have been various works attributed to Aḥmad al-Būnī (d. 1225, Cairo),⁶⁹ or perhaps the ‘encyclopedic lettrist opus’ Shams al-āfāq fī ʿilm al-ḥurūf wa-l-awfāq of ʿAbd al-Raḥmān al-Bistāmī (d. 1454, Aleppo).⁷⁰ Arabic compilations of ‘secrets’ may reasonably have been among the ‘texts [Mansa] Mūsā brought back from Cairo’, especially given that Mansa Mūsā's son and successor Maghā was later known in oral traditions as ‘Maghan the Sorcerer’,⁷¹ presumably for his talismanic expertise rather than a return to Islamically-prohibited practices of sorcery. Al-Kābarī's text demonstrates some theological influence from the North African Almohad leader Ibn Tūmart (d. 1130),⁷² suggesting that esoteric practices associated with Ibn Tūmart as the legendary successor of al-Ghazālī may have been available to al-Kābarī, perhaps contained in al-Ghazālī's text, al-Sirr al-maknūn.⁷³ As previously mentioned, the name al-Ghazālī appears on West African tombstones no later than the fifteenth century, included in ‘unusually long genealogies, which must have been also understood as chains of transmission of knowledge’.⁷⁴ Al-Kābarī's own Bustān al-fawāʾid contains reference to a three-by-three ‘budūḥ’ magic square⁷⁵ rendered famous by al-Ghazālī's transcription of the square in his autobiographical al-Munqidh min al-ḍalāl. ⁷⁶ Given al-Kābarī's apparent predilection towards Ibn Tūmart, it is tempting to associate such an alleged Ghazalian inheritance with another collection of esoteric sciences called Kanz al-ʿulūm wa-l-durr al-manẓūm, sometimes attributed to Ibn Tūmart, but more likely authored in the thirteenth century.⁷⁷ Notwithstanding such possible external influences, al-Kābarī's text must be considered an original contribution to the field of Islamic talismanic sciences, drawing from a variety of Arabic texts and oral transmissions in the Middle Niger available in fifteenth-century West Africa.

The authority of scholars over sultans

There is no doubt that Ibn al-Mukhtār's Tārīkh, more so than al-Saʿdī's Tārīkh al-Sūdān, endeavored to portray the state and scholars in a mutually constructive relationship both within and beyond Timbuktu.⁷⁸ But even he makes it clear that the best of kings are those who listen to scholars. Of course, this coincides with scholarly articulations elsewhere in the Muslim world in the medieval period:

All authors of (Islamic political) advice literature are likely to stress the respect and humility that kings should show in their dealings with scholars. Kings should hold the ʿulamāʾ in the greatest esteem and know that their own worldly greatness is as nothing before knowledge and learning … honouring everyone who wears the clothes of the scholar and righteous man is a cause of the soundness of the state.⁷⁹

The Songhay chronicles thus celebrate sultans who showed respect to scholars and gave them generous gifts of land, gold, salaried positions, books, and slaves/concubines. Conversely, the chronicles chide those who were stingy or haughty with scholars. While Moroccan conquering leaders are originally presented in less than flattering portraits, even they ultimately recognized the authority and saintliness of Songhay scholars: seeking their advice, and visiting the tombs of West African holy men.⁸⁰ But later scholarly alienation, and the accented narrative of scholarly resentment in the Tārīkh al-Sūdān and the Tārīkh ibn Mukhtār demonstrate that this attempt to inform and discipline the Moroccan invaders as to the inviolability of West African ʿulamāʾ was as unsuccessful on the Arma regime as it had been on Songhay authorities.

Scholars portrayed themselves as the real authorities in the Songhay empire, with the power to discipline unjust rulers. For example, Ibn al-Mukhtār describes Askiya Ishāq I as ‘a pleasant man, both righteous and blessed, and prone to giving plentiful charity. He was fastidious in keeping to the congregational prayer. He was intelligent, discerning, and sagacious’. But even this just king cannot escape the censure of scholars. The askiya visited the city of Djenné in 1539, and held court in the mosque for several days:

The mosque was filled to capacity. Then he ordered his spokesperson to make known to the people his words: ‘By God, I did not travel on this journey here except to set right the lands [of my kingdom] and to rectify [the states of] God's worshipful servants. So now, make known to us anyone who is causing harm to the Muslims, and who is oppressing people in this city. Whoever knows such a person and remains silent, this [unfulfilled] obligation to himself and to God's servants will hang from his neck [on the Day of Judgement]’.

The askiya's spokesperson walked up and down the rows of people in the mosque saying these words, but everyone remained silent. Among those present in that congregation was a jurist, the judge Maḥmūd b. Abū Bakr Baghayogho,⁸¹ who was seated close to the askiya. After some time had passed and no one had responded, the jurist Maḥmūd said to the askiya, ‘Do you mean what you say, O Isḥāq?’

‘Yes, by God, it is the truth’, he said.

‘And if we were to inform you of such an oppressor, what will you do to him?’

The askiya replied, ‘I would punish him as he deserves, either by executing him, having him beaten, imprisoned, or exiled. And if he caused the loss of wealth or property, I would see it returned to the original owner’.

Then the jurist Maḥmūd Baghayogho, may God be pleased with him, said, ‘We are not aware of any greater oppressor here than yourself. You are the source of and reason for every injustice. No one here unjustly expropriates wealth except for your sake, by your command, and by the force of your authority. If you would see an oppressor executed, you should start with yourself, and get on with it quickly. This money that you carry away from here, did you earn it through trade? Are we your slaves who harvest the land here for you? Does the commercial wealth (of this city) belong to you?’

Askiya Isḥāq was startled and bewildered at these words. He heaved deeply and began to cry out of remorse for what he had been told.⁸²

The elderly jurist thus effects the public embarrassment and repentance of the askiya, demonstrating that moral authority remained in the hands of scholars and not kings. But later, the askiya had his revenge, appointing Baghayogho to the office of chief justice in Djenné, causing the scholar to pray against the sultan: ‘Isḥāq has deprived my eyes of rest, and bound me to sleepless nights. May God also disturb his wellbeing and inflict him with preoccupations’.⁸³ This example demonstrates that the Baghayogho scholarly family, likely connected to the Suwarian Jakhanké tradition,⁸⁴ exhibited a well-defined (no doubt agriculturally-based) independence even in the urban context of Djenné.

The Tārīkh Ibn Mukhtār is riddled with the specific words of supplications with which scholars brought down tyrants. Scholarly descendants of the famous Mori Hawgāro killed Soɲi ʿĀli (d. 1492) when God answered their prayers, ‘O God, save us from him, and destroy him before he can rise from where he sits. Make him die in a state of denying God, outside of Islam’.⁸⁵ Later in the text, Ibn al-Mukhtār is more revealing in the esoteric work scholars could do to discipline political leaders who got out of line. In the 1580s, the governor of Timbuktu's port city of Kabara (distinct from al-Kābarī's native town of Kābara) disputed with the elderly scholar Maḥmūd al-Kaʿti (Ibn al-Mukhtār's maternal grandfather) and knocked him to the ground. The scholar himself remained patient, but his student could not bear the insult:

One of the shaykh's students, a close disciple, was present with him when this happened. After the governor had left, the student said to the shaykh: ‘If not for God's words in the Qurʾān, “Whoever kills a believer intentionally will be punished with Hell”, I would surely have killed the governor this very day, even without a sword or spear’. The shaykh responded, ‘God said, “Whoever kills a believer”; He did not say, “Whoever kills a corrupt oppressor’”. The student promised, ‘Thus will the governor of Kabara be destroyed’. Then the student took a piece of paper and wrote on it some things and some letters. He folded it and wrapped it in some black fabric and hung it from the neck of a male goat. Then, he took a spear and impaled the goat beneath its shoulder. The goat fell dead. When the day came marked by this killing of the goat, God caused the general Ṣādiq to overpower the governor of Kabara and impale him from beneath his shoulder in exactly the same way. And he thus died by the power of God.⁸⁶

This particular prayer, consisting of ‘some things and some letters’, was undoubtedly a specific talismanic formula, probably verses of the Qurʾān with various magic squares.⁸⁷ The ability of the esoteric expert to kill an enemy was attested by the earlier Persian scholar Fakhr al-Dīn al-Rāzī (d. 1210, Herat), summarized as follows: ‘Once the mind has been trained to a high level of meditative focus, it acquires both knowledge of occult matters and the power to act directly on bodies and to influence souls: it can even influence the rain and kill a man remotely’.⁸⁸ Continuing rumors of the ability of West African Muslim scholars to kill adversaries through slaughtering a goat or sheep has been recorded in ethnographies conducted in Mali.⁸⁹ But more generally, Islamic esoteric experts were probably influenced by the Prophet Muḥammad's statement, ‘Supplication is the weapon of the believer’.⁹⁰

While the specific prayer used to kill Kabara's governor is not found in al-Kābarī's Bustan al-fawāʾid (none mention impaling a goat) there are several others that reveal a veritable spiritual arsenal that could be deployed against enemies and unjust oppressors. The following prayer is presented as a protection from a corrupt aggressor:

As a protection from a corrupt person, an enemy, or whoever wishes you evil … [let a person] write ‘they compete in worldly accumulation’ until the end of the chapter [of the Qurʾān] in the milk of a female dog and hang it on his body. Whoever should seize … his shirt, his turban or his sleeve, that person will be speedily destroyed if God wills.⁹¹

This is a simple enough talisman, but its ingredients would not be so easy to procure. If nothing else, milking a female dog demonstrates a pronounced commitment to acquiring the type of spiritual armor that would destroy anyone who laid hands on the scholar. Another method requiring simpler ingredients, but more prayerful exertion would be the following:

If you want to destroy an enemy, an infidel, or a corrupt person, ask for clean ashes and spread them out [on the ground]. Next pray two cycles of prayer, with the opening chapter of the Qurʾān and after ‘Say He is God the One’ three times. Then write the name of the corrupt person in the place your forehead [touched the ground], and then sit and recite over it [the verse], [mention] ‘And Jonah, when he went away in anger, thinking that We had no power over him, but then he cried out in the deep darkness, There is no god but you, glory be to You, I was among the wrongdoers’. Read this [verse] …, and then ‘Say He is God the One’ until the end of the chapter … Surely the enemy will be destroyed quickly by God's permission.⁹²

Such prayers would require a significant investiture of time and effort. Given the elite status of scholars such as al-Kābarī in Timbuktu, the enemies or supposedly corrupt persons targeted by these prayers would undoubtedly have been competing elites from the Middle Niger political establishment.

Al-Kābarī's text also contains specific mention of scholarly relations with sultans. The first reference is early in the manuscript, with a heading, ‘What to recite when entering in the presence of the sultan’. Al-Kābarī connects such a preoccupation with the early Muslim community, reporting a hadith from the Prophet Muḥammad:

As reported by Abū Hurayra, God's Messenger, God's blessing and peace upon him, ‘Whoever is singled out by the despotism of a deliberately [unjust] sultan or arrogant oppressor, let him read upon entering his presence, “In the Name of God, the Mighty, the All Powerful. Enter by the gate, for if you enter by it, surely you will be victorious. Put your trust in God if you are indeed believers. [And the verse] He said, Fall back [into the Fire], and do not speak to Me. [And the verse] I have taken refuge in my Lord and your Lord, lest you should assault me. [And the verse] Enter here in peace and security. [And the verse] Their plots will not harm you in the least. Surely God encompasses all that you do. Indeed, God the exalted will defend you from his evil and turn his plots back against his throat, and God will replace your fear with security”’.⁹³

This seemingly falls in the same category of several others employed in al-Kābarī's text: it cannot be located in the canonical collections of Prophetic narrations, but it was no doubt considered authentic enough to be acted upon by al-Kābarī's students.⁹⁴ The Qurʾān verses employed here are noteworthy for their relevance to the theme of confronting injustice. The first verse — Put your trust in God if you are indeed believers — describes God's words to the followers of Moses when entering the ‘promised land’ despite its being populated by powerful and oppressive people. The second verse — Fall back [into the Fire], and do not speak to Me — represents God's words to castigated miscreants who demand to be removed from their punishment in hellfire. The third verse — I have taken refuge in my Lord — relates the statement of Moses when he enters the presence of Pharoah to demand the release of the Hebrew people. The fourth verse — Enter here in peace — expresses God's assurance to the believers to enter Paradise after having been protected from the Satan. The last verse — Their plots will not harm you — refers to some Jewish conspirators who, according to Muslim accounts, plotted to kill the Prophet Muḥammad in Medina. Such scriptural contexts leave no doubt that scholars in medieval West Africa rarely considered sultans to be benefactors, but perhaps more often thought of them as sources of oppression and political intrigue.

That confrontations with unjust political power could sometimes turn physically violent, aside from ‘remote killing’, is attested to in the knocking down of Maḥmūd Kaʿti or the existence of prayers to punish anyone who seized a scholar's robes, both as earlier mentioned. Al-Kābarī's text also contains prayers to overcome opponents while wrestling, to ease the pain of beatings, and to protect the body from weapons in battle.⁹⁵ This last appears to include a number of opaque combinations of Arabic letters that should be read over water that is later used for bathing and drinking: ‘No weapon will harm you in the course of war or other [conflict], if God wills’. Coincidentally perhaps, such prayers reveal that prayerful exertions did not always exempt scholars from physical altercations with political elites.

Paradoxically, such prayers to protect scholars would have made clerical expertise highly valued, most of all by kings. An example from the Timurid sultanate in Central Asia proves the point. A scholar put a certain prayer in his turban before he was sent to be beheaded by the sultan. When the executioner's sword bounced off the scholars neck a number of times, the scholar's life was spared provided that he gave the prayer to the sultan.⁹⁶ Similarly, West African clerics were often called upon to provide prayer services to both Muslim and non-Muslim political elites.⁹⁷ In his questions to the North African jurist, Muḥammad al-Maghīlī (d. 1505, Touat, Algeria), Askiya Muḥammad expressed consternation about the power of such ritual experts: ‘Some assert that they can write [talismans] to bring good fortune, such as material prosperity or love, and to ward off ill fortune by defeating enemies, preventing steel from cutting or poison from taking effect’.⁹⁸ However, al-Maghīlī's ruling that such ‘liars’ should be punished was ‘generally ignored’ in West Africa, as the core texts of Mālikī school of law in North and West Africa ‘explicitly give permission for the use of charms’.⁹⁹ Oral traditions sometimes narrate the rise and fall of political power according to which leader employed the more powerful scholar.¹⁰⁰ Sultans thus both recruited and feared the prayer services of scholars.

The war for women

The Tārīkh Ibn Mukhtār connects the sudden death of the ‘corrupt tyrant’ Soɲi ʿĀli to the curse of a ‘righteous man’ whose daughter had been raped by the unjust king.¹⁰¹ The ruling establishment's appropriation of women seems to have been a consistent preoccupation of scholars (and others) in the Songhay empire. When a companion pilgrim of Askiya Muḥammad entered the tomb of the Prophet Muḥammad in Medina, Arabia, he declared to the sultan, ‘Now I will make some requests of you. The first is that you do not take my daughters into your palace except with the contract of marriage’.¹⁰² This request appears to suggest that kings had a habit of appropriating the women of their subjects as concubines.

On the other hand, sultans appear to have restricted the marriage prospects of scholars. Askiya Muḥammad's document of inviolability to the descendants of the scholar Mori Muḥammad Hawgāro contained the following curious language concerning marriage rights, which seems to prove that such permissibility was not normally available:

I grant permission for them and their descendants to marry among the women of my entire kingdom … The exception is the Sorko and the Arbi. I forbid them to take wives among these people, for these two groups belong to us. Nevertheless, if one of them transgresses and begets children by a woman among them, then such offspring are free on account of the sanctity of their ancestor [Hawgāro].¹⁰³

In other words, this particular scholarly lineage could freely marry among the women of the empire, except for the professional fishing people (Sorko) and the palace entourage (Arbi), which were servile casted groups directly subject to the sultan. There seems little doubt that the askiya's words, ‘May God curse the reign of anyone among my descendants who wrongs them or unjustly expropriates anything from them’, meant to prevent future rulers attempting to steal the wives of this scholarly class.

The fact that a scholar unrelated to Mori Hawgāro would record such a document as a testimony to Askiya Muḥammad's exemplary justice demonstrates a prevailing concern. It meant that scholars were not always free to marry women outside of their own professional circles, that the children of such ‘inter-caste’ marriages not directly sanctioned by the Songhay administration could be taken as slaves, and that unjust rulers did sometimes expropriate the wives and daughters of subjects, including the women of Muslim scholars. This was a great threat to Muslim scholars: predatory rulers undermined scholars’ ability to sustain religious communities and to expand knowledge networks and spread Islam through marriage. It is not surprising, then, to find that protecting and attracting women were significant concerns in West African texts of Islamic esotericism.

The single greatest preoccupation of many collections of Islamic esoteric sciences, including al-Kābarī's Bustān al-fawāʾid, appears to have been women. Maintaining women was of course essential to accumulating ‘wealth in people’.¹⁰⁴ And marriage as a foundation of constituting communities was tested in times of political and social upheaval: ‘Control over women became a central feature in political power both in kin groups and between communities. These were processes that were intensified by warfare, state formation and slave trade’.¹⁰⁵ But concepts of male control only tell part of the story: the Muslim scholarly preoccupation with being patient with women and meeting their sexual expectations demonstrates that men were in competition with each other for women, and that women sometimes abandoned some men for others.

Prayers that concern women were not only to assist with men's inadequacies; they were sometimes tailored to specific requests that women would have made of Muslim scholars. Al-Kābarī's Bustān al-fawaʾid thus contains reference to prayers to cure women's infertility,¹⁰⁶ to prevent miscarriages,¹⁰⁷ to regulate excessive menstrual bleeding,¹⁰⁸ to prevent unwanted pregnancies,¹⁰⁹ to facilitate marriage,¹¹⁰ to ease pain in the breasts,¹¹¹ or ‘to return the appearance of virginity’ to a woman.¹¹² Some of these prayers take the form of a straightforward supplication, such as the following aforementioned prayer to get married quickly, which is perhaps logically interspersed with verses of the Qurʾān normally recited at weddings:

Praise be to God. We seek His assistance and forgiveness. We seek refuge in God from evil and from our own lethargy. Whomever God guides, no one can misguide him. Whomever God allows to stray, there is no guide for him.¹¹³ I bear witness that there is no deity save God and that Muḥammad is His servant and messenger. Fear God as He should be feared and die not except in a state of submission.¹¹⁴ Fear God, through whom you claim rights of one another, and (reverence) the wombs (that bore you). Ever does God watch over you. ¹¹⁵ Fear God and speak the truth. ¹¹⁶

But other prayers require more effort, such as that to make a woman appear to be a virgin. Here, a woman ‘must take, by the blessing of God the exalted and His messenger, the gall bladder (marāra) of a bull and rub it on her genitalia. She will become [like] a virgin again, by the permission of God the exalted, of this there is no doubt’. Such prayers for women demonstrate that Muslim scholars were called upon to address the direct concerns of women, not only to meet the needs of men in relation to women. Indeed, women may have been among the most frequent clientele of Islamic esoteric experts. According to ethnographic research in contemporary Nioro (Mali): ‘Many Malians readily assert that women make up the main clientele for practitioners of the Islamic esoteric sciences’.¹¹⁷ It is not difficult to imagine that al-Kābarī's text was similarly deployed to meet the needs of women in the Songhay empire.

Medieval West African Arabic texts display a certain anxiety over the sexual gratification of women. The Prophet Muḥammad himself warned men that they are considered deficient if one has sex with his wife and ‘fulfills his needs before she fulfils hers’.¹¹⁸ The famous Timbuktu scholar Aḥmad Bābā (d. 1627) authored a text of erotica, Durar al-wishāḥ li-fawāʾid al-nikāḥ, discussing different sexual positions, ways of delaying male orgasm, and the importance of pleasing women; itself an abridgment and commentary on the Egyptian scholar Jalāl al-Dīn al-Suyūṭī's (d. 1505) al-Wishāḥ li-fawāʾid al-nikāḥ.¹¹⁹ Such works contain chapters on ‘what women love from men and what they dislike’, with citations from pious Muslim women describing their preferred styles of intercourse,¹²⁰ and even reference to a narration that God gave women ‘ninety-nine percent’ of all sexual desire and men ‘only one percent’.¹²¹

There are undoubtedly more prayers to increase male virility, or power in copulation, than there are any other type of prayers in al-Kābarī's Bustān al-fawāʾid. Most of these are based in the writing of various sections of the Qurʾān and drinking the water that has been used to wash off the ink, a practice the earlier Damascene scholar Ibn Qayyim al-Jawziyya (d. 1350) declared licit — attributing its performance to many of the righteous early Muslims, thereby perhaps explaining the hadith, ‘The man without any bit of the Qurʾān in his belly is like a broken-down house’.¹²² Here is an example of the practice from al-Kābarī's text:

If you want to strengthen yourself in the act of copulation, write [the Qurʾān verses] ‘He said, my Lord, open my breast’ until the words ‘your request is granted, O Moses’; and ‘Have we not expanded for your breast’ until the end of the [Qurʾān] chapter; and the chapter ‘For the protection of the Quraysh’. Wash off [the writing] and drink [the water] and do this often. By this you will have strength in copulation if God wills.¹²³

Other variations of the same method to increase male virility include writing the Qurʾān chapter Quraysh three times in the bottom of a dish used for eating, and then washing the writing with water infused with tamarind before drinking.¹²⁴ Another version calls for writing the Qurʾān chapter ‘The Merciful’, and then mixing the Qurʾān-infused water with hot pepper: ‘let it sit for a little bit, then drink; surely it will help you as God wills’.¹²⁵ The frequency of such prayers suggest that male scholars understood the stakes of female sexual gratification. Indeed, most Muslim jurists would grant a woman a marriage annulment or divorce on account of her husband's impotence.¹²⁶

Al-Kābarī's text also includes a number of prayers to attract women more generally and ‘bind’ them to a man. For example, the Bustān explains:

If you wish that your woman does not love any man except for you, dig a hole [in the ground], and light a fire or place some embers in it. Make a complete ablution and pray two prayer cycles. Then [take off your clothes] and sit naked while writing [the Qurʾān verse], Have you observed the fire that you light … Wash off the verse [in water] and give [the water] to the woman to drink. She will hate all men except for you, by the power of God the exalted.¹²⁷

This extraordinary prayer demonstrates the great lengths a reputable scholar would go to keep a wife from running away with other men. Another prayer for the same purpose calls for writing certain letters on one's dhakar (‘manhood’).¹²⁸ It would have been logical to write both prayers at the same time, as the scholar would already have been unrobed. Certainly, an eminent jurist, such as al-Kābarī or his students, might be risking his reputation should he be seen sitting naked over a fire in the middle of the night writing, perhaps even on his own body. More explicitly, such prayers suggest that women leaving men for other men was a frequent enough phenomenon to be of great concern for scholars. Although women's voices are mostly silent in the Arabic literature of medieval West Africa, their agency is thus not entirely erased. Such agency was the subject of considerable anxiety for scholars who were attempting to build communities in a West African social space.

Powerful political leaders were also in the habit of stealing women, as the anecdotes at the beginning of this section attest. Such leaders even took women from each other, not only from subjects. The mother of Askiya Mūsā (r. 1529–31), named Jāra Kurbu,¹²⁹ had been captured by Askiya Muḥammad I (r. 1493–1528) during a military campaign, after which she bore him Mūsā. She had originally been the concubine of a local governor. Later, the ruler of Busa revolted against the askiya and captured Jāra for himself, after which she bore him a son who later inherited his authority in Busa.¹³⁰ Askiya Mūsā, perhaps on account of the affronts his mother had experienced at the hands of his father for stealing her from her earlier household, behaved with ‘shameless insolence’ towards his father, and after deposing and exiling him, took his father's concubines for himself in blatant violation of Islamic law.¹³¹

If scholars were politically restrained from engaging in military confrontation over women in the Songhay empire, they were not entirely powerless. Al-Kābarī's text demonstrates that scholars could, and likely did, fight back against the appropriation of women by political elites. One prayer provided, requiring the writing of a supplication on the leg of the one asking, is specifically for the purpose of ‘freeing a concubine from the palace of the king’.¹³² Should an unjust man refuse to free a captured woman, scholars could use their prayers to strike a man directly. The Bustān thus contains a number of prayers to ‘tie down erections’. For example, one prayer to make a man's erection ‘die and pass away’ (māta fāta) includes reading the Qurʾān verses, ‘Truly you will die, and they will die. Then on the Day of Judgement will your disputes be settled’. The verses are read together with some other supplications over knots tied in a string of red silk, after which the string is buried.¹³³ The similarity of this prayer with explicitly condemned forms of sorcery in the Qurʾān — ‘the evil of [sorcerers] blowing on knots’ ¹³⁴ — is difficult to miss. It also seems to parallel the reported non-Muslim practice, attributed to Bambara people, of ‘tying knots in pieces of string’ in contemporary Mali, sometimes including the burying the string in question;¹³⁵ as well as talismanic practices among the ‘weaver’ caste of Fulani in northern Senegal who tie knots in string ‘onto which incantations are muttered by the knotter’.¹³⁶ A subgroup of this Fulani weaver caste, specializing in music and singing, were associated with the ability to cause impotence among male adversaries.¹³⁷ This certainly begs the question of the exact line between sorcery/magic and Islamic esotericism,¹³⁸ but there is no such remuneration in the text. Rather al-Kābarī's exhortation is that the scholar should not lose track of the string once buried, so that the man could be released from being tied up once the original reason for the prayer had been resolved.¹³⁹ In any case, employment of verses from the Qurʾān — a ‘celestial’ element — would have technically satisfied al-Ghazālī's distinction between magic and talismanic science. Whatever the judgment, this type of work represented a significant investment in what can only be described as a type of spiritual warfare waged against those who, in targeting (Muslim) women, threatened the viability of Islamic scholarly communities.

This section inevitably privileges male agency in relation to women. But further research on female Islamic esotericism in the region may balance this gendered portrait. There is some evidence of female prayer experts elsewhere in the Muslim world. The Abbasid Caliph Hārūn al-Rashīd was said to have been seduced by successive servant girls who employed the same talisman hidden in their hair.¹⁴⁰ The agency of women through prayers is not entirely absent in the West African Arabic sources. The Tārīkh al-Sūdān explains Askiya Muḥammad's wresting of power from the Soɲi dynasty as the result of prayers made by Nānā Tinti, the daughter of a famous Timbuktu scholar, who had been entreated to pray against Soɲi ʿĀli by Askiya Muḥammad's mother, Kasay.¹⁴¹ This suggests that (Muslim) women themselves were instrumental in marshalling the spiritual power of the Islamic esoteric sciences for the constitution of Muslim communities — communities they felt reasonably assured would protect them from corrupt political authorities.

Conclusion

After entering Timbuktu and killing a number of scholars in 1593, the Moroccan army dragged out the leading clerics of the city, later marching them off across the desert to spend their exile in Marrakesh. Only the celebrated Aḥmad Bābā would return. ‘This attack on the scholars was the greatest violence against Islam’, Ibn al-Mukhtār remonstrated. As the soldiers brought the elderly judge, ʿUmar Aqīt, a young boy began to cry at the sight, and was promptly impaled on the sword of a Moroccan solider. The judge then laughed, saying, ‘I used to think I was better than this boy, but now his excellence over me has become apparent, for he has entered Paradise before me’.¹⁴² Although Timbuktu scholars would later be reconciled to the Arma regime, the West African scholarly class had already identified, by the late sixteenth century, a cause for which they would lay down their lives. The ability of scholars to build and protect communities was synonymous, in their minds, with the ability of Islam to take root and flourish. Defending these communities was worthy of martyrdom.

While Ibn al-Mukhtār attempts to end his narration of the Songhay empire in the sixteenth century, a curious detail escapes his account from the mid-seventeenth century. This was the violation of Askiya Muḥammad's document of inviolability granted to the descendants of Mori Hawgāro: ‘I have witnessed with my own eyes many of the descendants of Mori Hawgāro being sold in the market of Timbuktu. And they were sold as slaves even as they cried out that they were the children of Mori Hawgāro’.¹⁴³ Ibn al-Mukhtār uses this account to make a larger argument that subtly castigates both the later Songhay dynasty and the successive Arma regime, making of them one political entity that had contravened the rights of scholars: ‘The decline of Songhay was thus caused by the actions of its own people. This contravention of the askiya's command, along with the people's exhibition of enmity and hatred [to the scholars], began in the year 1075 [AH, 1664–5 CE]’. In the selling into slavery of the sanctified scholarly lineage of Mori Hawgāro, Ibn al-Mukhtār had perhaps witnessed the limits of scholarly cohabitation with corrupt rulers. This may not have been a declaration of clerical jihad against the ancien régime, but it certainly set the stage for what would come later. Indeed, the ‘age of the jihads’ that broke out in West Africa only a decade following Ibn al-Mukhtār's testimony, consistently invoked the call to resist the enslavement of Muslim scholarly communities.¹⁴⁴ The slave trade undermined the carefully cultivated independence of Muslim clerical establishments, putting both women and men at risk of capture.

These later armed jihads can be better understood by situating them in a context of spiritual jihad, made through powerful Islamic prayers, that had been simmering from the beginning of the Songhay dynasty. Scholars went to war in the age of jihads from the late seventeenth century not so much as a break from an earlier tradition of pacifism, but because they had already been battling in an esoteric ‘prayer jihad’ from at least the beginning of the Songhay empire. Later scholars gave up armed jihad under colonial occupation, but they retained an emphasis on prayer as the ‘weapon of the believer’ and warned rulers to leave them alone.¹⁴⁵ West African scholarly texts, at least from the fifteenth century, display deep concern for tempering unjust political power and the protection and attraction of women, discourses that reveal a perilous clerical struggle to assert community independence. This struggle would only increase in the age of jihads and subsequent colonial occupation.