The Decline of Songhay: Patricide on the Niger

What Askia had built up in thirty-five years, his sons began to demolish in one generation. And on the coast appeared sails that no one yet understood.

A Deposed Old Man

In August 1528 three of Askia Muhammad’s sons gathered in secret at the edge of Gao. Askia Musa, the eldest, was the driving force, and Askia Benkan and Askia Ismail stood beside him. The father was by now nearly ninety years old, blind, largely withdrawn into his palace quarters. The question between the brothers was not whether the old man would be deposed, since that decision had already been made; the question was which of them would receive the throne after him.

Musa won that political conversation, because he brought the army with him. When Askia Muhammad was deposed that month by his own son, he did not resist, or could not resist; the sources are vague. The Tarikh al-Sudan tells the moment with a kind of poetic sadness: the old king asked for a place to pray, was given one, and was thereafter banished to an island in the Niger, Kangaba, where he would spend the rest of his days.

For nine years, from 1528 to his death in 1538, Askia Muhammad lived on that island, where friends and family could visit him. The Tarikh al-Fattash contains anecdotes from which it is clear that he remained of clear mind, received people, and retained some influence. But politically he was dead, and his sons ruled.

Askia Musa: A Crown of Three Years

The patricide was less a glorious takeover than a dangerous gamble. Musa had hardly seated himself on the throne before his brothers began to quarrel among themselves over the division of provinces, taxes and military commands: Ismail was unhappy, Benkan felt passed over, and the provincial governors, the farmas whom Askia Muhammad had appointed, took sides depending on their own interests.

Askia Musa reigned three years. In 1531 he was murdered, not by an external enemy but by a nephew, or according to some sources his own brother, in a palace plot. The Tarikh al-Sudan names the murderer as Askia Muhammad Benkan, who then claimed the throne for himself, and with this began a pattern that for the next six decades would mark the palace of Gao: no succession without blood.

What this means for an empire is considerable. The central state of Songhay had become, under the old Askia, an institution and not only the person of the king, but an institution survives only if the rules of the game are predictable. As soon as the throne becomes an object of theft, and every succession moment a risky power moment, legitimacy crumbles from the top, even while the qadis in the cities continue to dispense justice, the farmas continue to collect taxes, and the caravans keep moving. But the regime in Gao loses who it actually is.

Benkan reigned from 1531 to 1537, and was deposed by yet another brother, Askia Ismail, who himself died in 1539. Then followed Askia Ishaq I (1539-1549), Askia Dawud (1549-1582, a relatively stable period), and then in quick succession Askia Muhammad IV, Askia Muhammad Bani, and finally Askia Ishaq II, who in 1591 would experience the death blow of Tondibi. Eleven Askias in sixty-three years, whereas Askia Muhammad the Great alone had reigned thirty-five.

Askia Muhammad Benkan and the First Years

Often Askia Muhammad Benkan (reigned 1531-1537) is summarized in tellings as “yet another ambitious brother”, but his reign is interesting because it shows how the Songhay institutions, built by his father Muhammad, continued to function despite the succession crises. Benkan maintained the Hajj delegations and received ambassadors from Morocco, while he confirmed the qadis in their posts and paid the stipends to the scholars of Timbuktu. The state administration kept turning, like a machine that has built up its own momentum.

This is an important point for the understanding of institutional sustainability, since a state is a set of agreements, procedures, expectations, career paths for officials and reputation mechanisms for courts, and never just a king. Askia Muhammad had woven those into a fabric stronger than one person, so that even while the brothers murdered each other, taxes kept coming in, caravans kept moving, and qadis kept dispensing justice.

But a fabric frays, since each coup interrupted a career or forced an official to take sides, and each new Askia had to replace part of the higher hierarchy with his own loyalists. The institutional continuity was slowly, scrape by scrape, undermined.

Askia Dawud: A Temporary Recovery

Not all the successors were catastrophes. Askia Dawud, who reigned from 1549 to 1582, gave Songhay thirty-three years of relative stability: he reformed agriculture, again promoted the scholarly institutes, maintained correspondence with the Maghreb, and cultivated diplomatic relations with the Saadi dynasty in Morocco, which at that moment was just rising. Under Dawud, Timbuktu blossomed again, although now more as an intellectual center than as a growth machine.

But under Dawud something also began to shift fundamentally, and this is where we must pause.

The Atlantic Competitor

In 1482 Portugal had, long before Sunni Ali died, built a trading post on the West African coast: Elmina, on what is today called the Ghanaian Gold Coast. From there the Portuguese began to trade directly with the Akan gold fields, bypassing the trans-Saharan routes, bypassing Timbuktu, and bypassing Songhay altogether.

This was revolutionary, and its consequences were slow but irreversible. The Saharan trade had connected West Africa for more than seven centuries with the Mediterranean, since every ounce of gold that traveled from Bambuk or Bure or the Akan to Fez or Cairo went via caravan routes that were controllable by Songhay. Tax, toll and trading posts were the foundation of Songhay income, and salt from Taghaza traveled the same route; this was the system that al-Bakri ﵀ and Ibn Battuta ﵀ had described.

But a Portuguese ship in Elmina can negotiate directly with an Akan merchant, so that the caravan that once inevitably went via Djenné or Gao began to receive optional alternatives. In 1510, in 1530, in 1550 this was still only a small portion of the total, but by 1580 it was a substantial factor, and the Portuguese and subsequently Dutch and English expansion on the coast gradually pushed the profit margins of the Sahara trade lower.

For Songhay this meant that the economic foundation which Askia Muhammad had inherited began under his successors to wash away gradually, not suddenly, but irreversibly.

And this is something that no Islamic source from that time has been able to analyze well, because the Songhay elite did not see the Portuguese coast as a competitor; they saw only their own markets. This was the blind spot that the empire would carry along into its fall.

The Sahel Dries Up

Alongside that economic pressure came another pressure that is too often forgotten: the climate. A series of dendrochronological and paleoclimatological studies have shown that the sixteenth century was a period of increased drought in the Sahel, for between 1450 and 1600 the southern boundary of the Sahara slowly shifted southward. Villages that had been livable under Askia Muhammad became more difficult under Askia Dawud, and caravan routes that under Mansa Musa had had familiar water stations dried up over decades.

This is no complete explanation, but it is a factor, since an empire that experiences several successive drought years loses grain reserves, loses population in border regions, and loses the economic base for its tax system. Songhay suffered from this together with its own political instability, two forces in the same direction, each accelerating the other.

The Salt and Gold Monopolies Crumble

A specific symptom: the salt mine of Taghaza, high in the Sahara, had since the time of Mansa Musa essentially supplied the monopoly on Sahel salt, and the Songhay state controlled the Taghaza route and collected duties on the salt. But in the second half of the sixteenth century Taghaza itself came into danger, as the wells became saltier, the Moroccan Saadis began to interfere with who exactly ruled over Taghaza, and Askia Dawud had to send military expeditions to maintain the position.

In 1586 Sultan Ahmad al-Mansur of Morocco launched a small expedition to occupy Taghaza directly. He was repulsed, but it became clear to the Songhay rulers that Moroccan interest in the Sahara was not diplomatic. Al-Mansur, whom we shall meet at length in the next article, had greater ambitions than just a salt mine.

Brothers Against Brothers

While this external pressure increased, the internal pattern of brother-against-brother continued. After Dawud’s death in 1582 his son Askia Muhammad al-Hajj followed, who already in 1586 was deposed by his brother Askia Muhammad Bani; Bani died unexpectedly in 1588, and a full civil war broke out between factions supporting different Askia candidates. This was a moment of real fragmentation, since provincial governors in the coastal regions stopped paying tribute, Tuareg tribes again began to raid the northern settlements, and Mossi troops came up from the south.

Out of the chaos eventually Askia Ishaq II emerged as winner, but his position was far from firm: his own brothers and nephews continued to intrigue, his farmas were dissatisfied, and his army had endured internal succession struggles which had even divided the soldiers into factions.

And in precisely that moment of maximum internal weakness, Ahmad al-Mansur sent an army from Marrakesh across the Sahara.

For the story of the decline itself it is enough to note that between 1528 and 1591 Songhay underwent no catastrophic collapse; it was hollowed out slowly, by succession wars, by economic pressure from the coast, by climate fluctuations, and by a bureaucracy that lost balance when the top wavered.

It was an institution that lost its own administrative power, from the inside out.

The Scholars Live On

There is one aspect that stands out in every investigation of this period. However politically Songhay fell apart, the scholarly tradition in Timbuktu continued to flourish. Ahmed Baba al-Timbukti ﵀ was born in 1556, in the middle of these turbulent decades, from the Aqit family that had been restored to honor under Askia Muhammad. He studied under his father and under Muhammad Baghayogho ﵀, himself one of the great scholars of the Sankore university, and by 1580 Ahmed Baba ﵀ was himself teacher, author and prominent mufti, while by 1590 he had already written several books.

The institutional infrastructure of Islamic knowledge in Timbuktu lived on, even while the political center wavered, and this is a reminder of something seen more often in Islamic civilizations: the Ulema tradition is bound to a community of readers and transmitters rather than to one specific ruler, so that as long as there are manuscripts and as long as there are study circles, the tradition lives.

But this independence too had its limit. When in 1591 Moroccan troops took Timbuktu, the Ulema community would experience something it had never experienced, a total physical disruption: the libraries were plundered, scholars were led in chains to Marrakesh, and an entire generation of academic production was interrupted.

The physical removal of scholars is the same structural act as the later physical removal of West Africans in the transatlantic slave trade: a system of forced deportation whose goal is not extermination, but expropriation, repurposing, forced labor or forced service to a foreign elite. The context is different and the order of magnitude incomparable, and yet the logic resides in the same family.

Court Intrigues and the Queen Mother

There is one aspect of the Songhay decline that in the common tellings is rarely discussed, but which is emphatically present in the sources themselves: the role of women at court, and especially the position of the queen mother.

In Songhay tradition the magha-koi, the mother of the reigning Askia, had a formal function at court with her own income, her own staff and her own political influence. During succession crises it was often the mothers of rival princes who behind the scenes forged alliances, traded marriage gifts and bribed army officers, and the Tarikh al-Fattash contains various anecdotes about court women who were decisive in determining which prince received the support of which faction.

This is regular court politics rather than “harem intrigue” in the orientalist sense known from bad tellings, court politics as existed in all world empires of the premodern period, from the Ottomans to the Chinese imperial courts to Versailles. The absence of women from the usual Songhay narrative says more about our own narrative conventions than about historical reality.

This matters because the matriarchally weighted structure of many communities that grew out of plantation slavery is sometimes seen as a post-slavery phenomenon, a consequence of family disruption under plantation conditions. But in West Africa itself women already before the slave trade had positions of power in courts, in trade and in religious life, so that the line is no break but a continuity under disturbed circumstances.

The Role of Ahmed Baba’s Youth

While the court in Gao deteriorated through its brother-against-brother wars, in Timbuktu a child was being raised who would ultimately become the most weighty scholar of his generation. Ahmed Baba al-Timbukti ﵀ was born in 1556, and his father Ahmad ibn Ahmad was already a known scholar. The family stood directly in the line of the Aqit, who had been driven out under Sunni Ali, restored under Askia Muhammad, and now, under the weaker Askias, had to maintain a delicate balance with an increasingly unpredictable court.

Ahmed Baba ﵀ studied under his father, under Muhammad Baghayogho ﵀ (a Fulbe scholar), and under various other masters. By his thirtieth he was teacher, by his fortieth he was authorized to issue fatwas, and his bookstore in Timbuktu grew, according to his own testimony, to 1600 volumes, which for that time was an enormous private library.

What is relevant for the story of the decline is that Ahmed Baba ﵀ grew up in a city that told him his work mattered, since the Islamic infrastructure of Timbuktu was strong enough to form young scholars even while the political empire in which the city lay was falling apart. This means that Islamic civilization in the region at the level of the city had its own resilience that stood apart from the imperial crown.

But that would change in 1591, when the physical infrastructure itself would be struck, and then it appears that knowledge without a state to protect it is ultimately also vulnerable.

An Empire on the Edge

In the year 1590 Askia Ishaq II sat on the throne in Gao. He ruled an empire that on paper was still the largest of West Africa, on paper could still mobilize forty or fifty thousand soldiers, on paper still controlled the Taghaza salt route.

But in reality it was a hollow tree, whose roots were rotted by succession wars and whose branches had withered from economic pressure from the south, while from the north, from Marrakesh, came the axe: an axe of about four thousand men, equipped with a weapon that West Africa had not yet seen in great numbers, the arquebus, the firearm, under the command of a eunuch general of Andalusian origin named Judar Pasha.

That axe was at that moment undergoing the heaviest march that an army had ever undertaken through the Sahara, thousands of kilometers of desert, of hunger, thirst and losses, and according to some sources no more than a third of the original troops arrived alive at the Niger Bend. But the third that arrived still carried its arquebuses, and that was enough.

Father’s Death, Brother’s Murder, Betrayal

What began in 1528 with Askia Musa deposing his father brought us in 1590 to Askia Ishaq II, who from a divided court had to face a Moroccan invasion, and between those two dates lies the complete story of an empire that ate its own future.

It is a Greek drama, but Sahelian, in which each generation wanted to be better than the previous and each generation demolished what the previous had built, so that in the end there was no one left to defend the gates of Gao against an enemy who had come through the desert in an unimaginable way.

That is a story of maturity, for great empires always fall from within first; the external blow is only the last drop.

---

Sources

• ʿAbd ar-Rahman as-Saʿdi ﵀, Tarikh as-Sudan, chapters 13 to 21 on the deposition of Askia al-Hajj Muhammad, the reigns of Musa, Muhammad Bonkana, Ismaʿil, Ishaq I, Dawud, al-Hajj, Muhammad Bani and Ishaq II and the arrival of Pasha Jawdar, in the translation of Hunwick, pp. 116–119, 124, 126, 132, 135–137, 142, 151, 155, 166–169, 172, 175, 186, 188–189.
• Tarikh al-Fattash, the anecdote of the old Askia complaining on the island to his son Ismaʿil about the mosquitoes, and the shorter reign it assigns to Musa, as quoted in Gomez, p. 321, and in the notes of Hunwick, p. 124.
• Ahmad Baba al-Timbukti ﵀, his own testimony that at his arrest in 1593 his library of sixteen hundred volumes was seized, the smallest among his kin, as quoted in Hunwick, p. lxi.
• Al-Bakri ﵀, Kitab al-masalik wa-l-mamalik, and Ibn Battuta ﵀, Rihla, as describers of the trans-Saharan trading system the article mentions; without page reference.
• John O. Hunwick (trans.), Timbuktu and the Songhay Empire: al-Saʿdi’s Taʾrikh al-sudan down to 1613 and other contemporary documents, Brill, Leiden 2003 (2nd printing; 1st printing 1999), pp. xxix, lx, lxi, lxiii, 116–119, 124, 126, 132, 135–137, 142, 151, 155, 166–169, 172, 175, 186, 188–189, 336–337, 372.
  • pp. 116–119 (TS ch. 13–14): the Askia who went blind towards the end of his reign; Musa’s break with his father and ʿAli Fulan; the deposition on the prayer-ground on 10 Dhu al-Hijja 935, 15 August 1529, when Musa swore that nobody would pray before he held power and let his father remain in the palace; the reign of thirty-six years and six months; Musa who began killing his brothers and the qadi Mahmud ibn ʿUmar ﵀ who refused to look at him.
  • pp. 124, 126, 132 (TS ch. 14–15): the assassination of Musa by his brothers on 12 April 1531, at the initiative of the Shaʿ-farma ʿAlu, after two years, eight months and fourteen days; the succession of his cousin Muhammad Bonkana ibn ʿUmar Komadiakha, who banished the old Askia to the island of Kangaga near Gao; Bonkana deposed after six years and two months by the Dendi-fari Mar Tumzu.
  • pp. 135–137 (TS ch. 15–16): Ismaʿil who brought his father back from the island in June 1537; the death of Askia al-Hajj Muhammad on the night of 31 January to 1 February 1538; Ismaʿil dead at the end of 1539 after two years and nine months; Ishaq I proclaimed on 27 December 1539.
  • pp. 142, 151, 155 (TS ch. 16–17): the Moroccan demand for Taghaza to Ishaq I; the Taghaza-mondyo killed in 964/1556–1557 with the authorisation of the sultan of Marrakesh and the opening of Taghaza al-ghizlan; the death of ʿAbd al-Malik in 1578 and the succession of Ahmad al-Dhahabi, to whom Askia Dawud gave ten thousand mithqals of gold, from which a friendship grew.
  • pp. 166–169 (TS ch. 18–19): the emissary who spied out the land; the army of twenty thousand men sent to Wadan that fell apart through hunger and thirst; the two hundred musketeers who found Taghaza deserted and the ban of September 1586 on going there; Askia al-Hajj deposed on 16 December 1586; the famine under Muhammad Bani and the revolt of the Balmaʿa al-Sadiq in March 1588.
  • pp. 172, 175 (TS ch. 20): Ishaq II in power on 10 April 1588 after Bani’s reign of one year, four months and eight days; the Balmaʿa acclaimed askia by his army, defeated, fled and put to death at Kanatu.
  • pp. 186, 188–189 (TS ch. 21): Jawdar as a eunuch from Cuevas del Almanzora in Almería; the three thousand musketeers with twice as many support troops who set out in October or November 1590; the arrival at the Niger on 28 February 1591; the Askia at Tondibi, forty-eight kilometres north of Gao, with twelve thousand five hundred cavalry and thirty thousand infantry.
  • pp. xxix, lx, lxi, lxiii, 336–337, 372: Muhammad Baghayogho ﵀, a Wangari from Jenne, as the shaykh of Ahmad Baba ﵀, with whom he read the Mukhtasar of Khalil ﵀ eight times; the sixteen hundred volumes; the return in 1608 and the death in 1627; the regnal dates of the Askias from 1493 to 1592, with Dawud who died on his estate at Tondibi; Ahmad Baba ﵀ (1556–1627).
• Michael A. Gomez, African Dominion: A New History of Empire in Early and Medieval West Africa, Princeton University Press, Princeton 2018, pp. 218, 297, 309–310, 315, 321, 324, 332.
  • pp. 309–310, 315, 321, 324: the old and blind Askia, whose blindness ʿAli Fulan concealed; the eighty-eight-year-old who gave way on 15 August 1529; the twenty years of intrigue and bloodshed and the absence of succession rules that proved fatal to the empire; the banishment to Kangaga; the rise of the court eunuchs and the growing influence of royal women in rivalries.
  • pp. 218, 297, 332: the Portuguese who reached Elmina by 1471 and by the middle of the sixteenth century had exported some two hundred and fifty thousand Africans across the Atlantic; the title waiza for daughters of the Askias, possibly an office.