Mamluk Sultanate

Ayn Jalut

Drafted from (ed.), The Cambridge History of Egypt, Volume 1: Islamic Egypt 640-1517 (1998)formation

In 1260, two years after Baghdad, a Mongol army under Kitbuqa moved into Syria and took Aleppo and Damascus. The Mamluk sultanate in Cairo — ten years old, founded by a coup, on its third sultan — marched out to meet it and won at Ayn Jalut in the Jezreel valley.

The strategic circumstances were favourable: Hulegu had withdrawn most of his force east on news of the great khan's death, and the army left in Syria was a detachment rather than the main body. But the victory was real, it was the first significant Mongol defeat in the west, and Baybars — who had commanded the vanguard and murdered his way to the sultanate within months of it — spent the next seventeen years making Syria a Mamluk province and reducing the Crusader states to a coastal strip.

Baybars also did the thing that made the sultanate legitimate. In 1261 he installed a survivor of the Abbasid house in Cairo as caliph, revived the office the Mongols had extinguished three years earlier, and had himself invested by him. The caliph had no authority, no territory and no army. He had legitimacy, which was the one thing a Kipchak slave soldier who had killed two sultans could not manufacture for himself.

Baybars builds the machine

Drafted from Irwin, The Middle East in the Middle Ages: The Early Mamluk Sultanate 1250-1382 (1986)expansion

Baybars had commanded the vanguard at Ayn Jalut, murdered the sultan on the way home, and ruled for seventeen years. What he built in them is the reason the sultanate lasted another two and a half centuries after him, and almost none of it was a battle.

He quadrupled the standing cavalry, from something like ten thousand to forty, by buying on a scale no previous ruler in the region had attempted — the Kipchak steppe was being emptied by the Mongols, and the slave markets of the Black Sea were full. He rebuilt the Syrian citadels the Mongols had slighted and garrisoned them. He laid out the barid, a relay post system with fixed stages between Cairo and Damascus that carried a message in four days, and supplemented it with pigeons; a sultan in Cairo learned of a raid on the Euphrates within a week, and this, more than any single victory, is what let one government hold two countries separated by Sinai.

The diplomacy was of the same kind. Baybars allied with Berke of the Golden Horde — a Muslim Chinggisid at war with his Ilkhanid cousins — which secured the Black Sea slave route that the army depended on, and negotiated with Byzantium for passage through the straits to keep it open. He made the Nubian kingdoms tributary and took the Red Sea ports. The state's inputs were as deliberately engineered as its army.

Against the Franks he was methodical rather than dramatic. Caesarea and Arsuf in 1265, Safad in 1266, Jaffa and Antioch in 1268, Krak des Chevaliers in 1271 — taken one at a time, and where the position could not be held it was demolished so that no fleet could reoccupy it. Antioch had been Christian-ruled for a hundred and seventy years and was sacked with a thoroughness that shocked even sympathetic chroniclers. The kingdom of Jerusalem was reduced to a coastal strip before he died in 1277, reportedly by drinking poison he had prepared for somebody else.

The end of the coast

Drafted from Irwin, The Middle East in the Middle Ages: The Early Mamluk Sultanate 1250-1382 (1986)expansion

Qalawun was another of al-Salih Ayyub's purchases, an emir of Baybars' generation who took the sultanate in 1279 and held it eleven years, and he is the nearest thing the polity has to a founder of a house: his descendants held the title, with interruptions, for a century. He beat the Ilkhanids at Homs in 1281, signed commercial treaties with Genoa, Aragon and Byzantium, and took Tripoli in 1289. He died in 1290 on the march to Acre.

His son al-Ashraf Khalil finished it. Acre fell in May 1291 after six weeks; Tyre, Sidon, Beirut and Athlit were abandoned within months, and the Frankish presence on the Syrian mainland ended after a hundred and ninety-two years.

What followed is the part that mattered to the polity. The Mamluks did not garrison the coast — they destroyed it. Harbours were blocked, walls thrown down, and the ports left ruined on the explicit reasoning that a rebuilt Frankish landing would be worse than an empty shore. Coastal Syria was still visibly derelict two centuries later. It was a defensive judgement made by a land power against a sea power it could not meet at sea, and it cost the Syrian littoral its trade for the rest of the Middle Ages, which is one reason the Egyptian route through Alexandria mattered as much as it later did.

The victory also removed the sultanate's public justification. A state built on jihad against Franks and Mongols had disposed of the first and would shortly come to terms with the second, and the emirs of the fourteenth century had to find something else to fight about. They found each other.

The frontier that would not close

Drafted from Irwin, The Middle East in the Middle Ages: The Early Mamluk Sultanate 1250-1382 (1986)expansion

Ayn Jalut did not settle anything. For sixty years the Ilkhanate raided into Syria and the Mamluks raided back, and the sultanate's border ran along the Euphrates with client Turkmen and Armenian principalities in between, changing sides according to who had campaigned most recently.

The Mamluks lost the largest of those battles. At Wadi al-Khazandar near Homs in December 1299, Ghazan — an Ilkhan who had converted to Islam four years earlier, and who therefore arrived preaching against them rather than at them — broke the Egyptian army and took Damascus. The citadel held out; the Ilkhanid occupation lasted about a hundred days and dissolved when Ghazan withdrew, but the shock in Cairo was severe and the sultanate's claim to be Islam's shield had been answered by a Muslim.

The rematch at Marj al-Suffar south of Damascus in April 1303 went the other way and settled it. There were no further serious Ilkhanid invasions, and in 1322 the two states signed a peace. Sixty-two years after Ayn Jalut, the Mamluks and the Mongols recognised each other's frontiers, exchanged embassies, and let the merchants through.

The Ilkhanate dissolved into fragments within fifteen years of that treaty, and the Mamluks outlived it by nearly two centuries. Both things were true at once: the sultanate's founding claim was that it was the state that stopped the Mongols, and the way the Mongol wars actually ended was a negotiated border between two Muslim powers who had run out of reasons.

Al-Nasir Muhammad

Drafted from Irwin, The Middle East in the Middle Ages: The Early Mamluk Sultanate 1250-1382 (1986)peak

Qalawun's son al-Nasir Muhammad was made sultan three times — as a boy in 1293, again in 1299, and finally from 1310 to 1341 — and the third reign is the sultanate at its height. Thirty-one years without a deposition is by itself remarkable in this polity's record.

He got there by breaking the men who had twice removed him. The senior emirs of Baybars' and Qalawun's households were killed, exiled or bought, and al-Nasir then declined to build a comparable faction of his own: he kept his purchased mamluks numerous but junior, promoted through marriage and money rather than through household seniority, and made himself the only fixed point in the system. Irwin's account of how he did it is the best argument in the literature that the Mamluk polity had no constitution — only a distribution of force that a sufficiently ruthless man could rearrange.

The instrument was fiscal. The rawk of 1315, a cadastral survey and general redistribution of iqta' across Egypt, took land out of the hands of the great households and reassigned it — enlarging the sultan's own share to something near a third of the whole and cutting the emirs' independent revenue. A survey is not a dramatic event and it did more to his rivals than any of his campaigns.

What the money bought is still standing. His reign is the great age of Mamluk building: the mosque in the citadel, the enormous complexes along the Bayn al-Qasrayn, the aqueducts, the hippodromes, an expansion of Cairo that made it plausibly the largest city west of China. He married into the Ilkhanid and Golden Horde houses, received Rasulid and Ethiopian embassies, and sent the kiswa to Mecca every year.

The weakness was the one he had designed. By refusing to let any faction grow, he left none capable of governing after him: twelve of his sons and grandsons held the title in the forty-one years after his death, most of them children, while the emirs who had been kept junior fought over which of them would hold the child.

A state that would not become a dynasty

Drafted from (ed.), The Cambridge History of Egypt, Volume 1: Islamic Egypt 640-1517 (1998)aside?What aside means

The Mamluk system worked like this. Boys were bought on the steppe or in the Caucasus, brought to Cairo, converted, trained for years in barracks, freed on completion, and attached for life to the household of the man who had bought them. Advancement ran through that household. A sultan's sons, born free and Muslim, were not mamluks and generally could not inherit his position; power went to another purchased soldier.

That last feature is the unusual one. Nearly every state of the period was a dynasty, and the recurring way a dynasty failed was the succession: the Mongols, the Seljuks, the Carolingians and the Ayyubids all fragmented on an inheritance. The Mamluks removed heredity from the equation and got something else instead: violent, frequent, factional turnover at the top, with sultans lasting months, alongside remarkable institutional stability underneath, because the barracks, the households and the administration continued regardless of who was currently sultan.

They held Egypt and Syria for two hundred and sixty-seven years, stopped the Mongols, ended the Crusader states, and kept a caliph in Cairo whose only function was to make the arrangement lawful.

It is the one state of its kind that solved the standing problem of slave-soldier regimes — that an army which can make a ruler will eventually make one it prefers — by conceding the point completely and building itself around it.

The plague, and after it the Circassians

Drafted from (ed.), The Cambridge History of Egypt, Volume 1: Islamic Egypt 640-1517 (1998)contraction

The Black Death reached Alexandria in the autumn of 1347 and Cairo in 1348. No usable figure for the mortality survives, and the ones in circulation are guesses dressed as counts. The qualitative record is unambiguous: villages in the Delta abandoned, canals not cleared, the sugar plantations short of hands, and the plague returning roughly every decade for the rest of the sultanate's life rather than passing once.

The fiscal consequence ran straight through the state's design. Mamluk revenue was iqta' revenue, iqta' revenue was agricultural, and agricultural revenue was labour. When the labour died the assignments were worth a fraction of their assessed value, and an emir's ability to maintain the mamluks his rank required fell with them. The army contracted. So did the number of men any household could buy, at exactly the point when the households were the only durable institutions left.

What followed was two generations of factional war among the descendants of al-Nasir Muhammad's officers, with sultans installed as children and removed as adolescents. A regent, Yalbugha al-Umari, tried to restore the training regime and was murdered by his own mamluks in 1366. Meanwhile the awlad al-nas, the freeborn sons of mamluks who were barred from the top by the system their fathers had joined, were being promoted into it by sultans who trusted family over faction — the one reform that would have made this a dynasty, attempted repeatedly and never carried.

In 1382 Barquq deposed the last Qalawunid child and took the sultanate himself. He was Circassian, bought from the Caucasus rather than the Kipchak steppe — that route had closed as the Golden Horde broke up — and he recruited his successors from the same place. Historians conventionally call everything after him the Burji sultanate. The system did not change; the men in it came from somewhere else, and the fiction of a Qalawunid line was dropped.

Then Timur. In 1400 he took Aleppo and Damascus, sacked both, deported the craftsmen of Damascus to Samarkand, and left. The sultanate did not fall, and Syria did not recover its position for the rest of the century.

The boy who left his army

Drafted from Lane-Poole, A History of Egypt in the Middle Ages (1901)contraction

At the end of 1400 the sultan of Egypt was thirteen years old and standing in Syria at the head of a great army, trying to stop Timur. Aleppo had already been sacked. Damascus was next.

The name

He had been born during the rebellion of Mintash and named accordingly. After his father Barquq beat the rebels the name was changed to Faraj — deliverance.

Barquq had been the strongest Mamluk sultan in decades and had handled Timur by refusing to be handled. When Timur sent an embassy, Barquq answered in language as haughty as he had received; he sheltered Timur's enemies, invited the expelled sultan of Baghdad, and built a coalition on the Euphrates. It held, partly because Timur was busy elsewhere.

Then Barquq died, more worried at the end about the rising Ottomans than about Timur, and the coalition passed to a boy of thirteen.

The battle

The Egyptians at first seemed to be driving the invaders back. Lane-Poole's assessment is dry: the retreat was probably strategic.

When the Egyptians attacked they were heavily repulsed. And then the thing that decided the campaign happened inside the Mamluk camp rather than on the field. Faraj found that the defeat had bred sedition among his emirs, who — as Lane-Poole puts it, with no irony — naturally desired a more experienced leader at such a crisis.

So he withdrew in haste to Cairo, leaving his army to its fate.

That is the Mamluk system's central weakness stated in one movement. A sultanate in which the ruler is made by the emirs and can be unmade by them cannot afford a lost battle, because a lost battle is immediately a succession crisis. The thirteen-year-old was not being cowardly; he was doing the arithmetic correctly. Staying with a defeated army surrounded by men who wanted somebody else in his place was more dangerous than Timur.

Damascus

The city surrendered on terms, and was sacked, ruined and burnt anyway. All northern Syria was cruelly devastated.

Afterwards, when Timur had finished in Anatolia and destroyed the Ottoman army at Ankara, his envoys came to Cairo with demands. Faraj agreed to all of them: he surrendered his prisoners, and he agreed to strike coins in the conqueror's name.

Lane-Poole adds the detail that redeems the episode slightly, and it is the kind of thing only a numismatist notices. No such coins have ever been found. Timur never entered Egypt and never controlled it. He died in February 1405 while Faraj was raising yet another army to resist whatever came next.

The submission, in other words, was made and then quietly not performed — which is exactly what a state does when it has agreed to something it cannot survive refusing and cannot bear to carry out.

What it cost

Faraj had lost his credit by these proceedings.

A struggle for power among the leading mamluks followed, in which the sultan was treated with contumely and beaten in the field, and ended in his sudden disappearance in 1405.

The material damage was real — Damascus burnt, northern Syria wrecked, the Syrian tax base broken for a generation — but it was survivable, and the sultanate survived it by another century. What did not recover was the thing Barquq had briefly possessed: a ruler whose emirs could not remove him in the middle of a war. Every crisis for the next hundred years would be conducted by men who had watched a sultan ride away from his own army and had learned the lesson from the other side.

What the sultanate paid for

Drafted from (ed.), The Cambridge History of Egypt, Volume 1: Islamic Egypt 640-1517 (1998)aside?What aside means

The politics were violent and the city was magnificent, and a page that records only the first has described the sultanate the way its chroniclers' index entries do rather than the way it looked.

Cairo under the Mamluks was the largest city in the Mediterranean world and probably in Islam, and it was rebuilt continuously for two centuries by men who could not pass their wealth to their sons. That is not incidental. A mamluk's iqta' reverted to the state when he died and his children were outside the system, so the reliable way to convert a career into something durable was the waqf: an endowed mosque, madrasa, hospital or tomb, legally inalienable, with a salaried staff that could include one's own descendants. The peculiarity of the succession rule is therefore visible in the skyline. Qalawun's hospital complex, Sultan Hasan's madrasa, Barquq's and Barsbay's and Qaytbay's funerary foundations — the architecture that still defines historic Cairo is the output of a ruling class working around its own childlessness.

The money came through Egypt because the alternatives had been closed. With the Syrian coast deliberately ruined and the Persian Gulf route disrupted, the Indian Ocean spice trade reached Europe through the Red Sea, and the Karimi merchants who ran it — a syndicate of long-distance traders operating between Aden, Jeddah and Alexandria — were rich enough to lend to sultans. Venetian and Genoese factors bought at Alexandria under treaty. Sugar was refined and exported. The customs on all of it were the sultanate's second revenue.

The scholarship was of a piece with the wealth. Cairo took in what Baghdad had lost: Ibn Khaldun taught and judged here and wrote the Muqaddima's final version here; al-Maqrizi wrote the topography of the city and a treatise on its famines; Ibn Hajar and al-Suyuti made it the centre of hadith study and of almost everything else. A Turkish-speaking military caste that did not intermarry with the population it ruled endowed, in Arabic, the largest concentration of religious learning in the Sunni world.

The keys sent to Cairo

Drafted from Lane-Poole, A History of Egypt in the Middle Ages (1901)contraction

The Mamluk sultanate was killed by an Ottoman army in 1516, and the sixty years before that are the record of Cairo watching it come and being unable to do anything about the frontier where it would arrive.

The thorn

Northern Syria and the Taurus marches were held by Turkman principalities that were technically Egyptian vassals — the White Sheep, the Dhu'l-Qadr of Elbistan, Karaman — and were in practice whatever their own succession disputes made them.

Lane-Poole is precise about why this mattered. The trouble was not the chronic insubordination of the vassals in itself. It was that every disturbance among them furnished a pretext for intervention by the sultanate's most dangerous neighbour, whose power was about to be crowned by the conquest of Constantinople.

Successive sultans tried every available method. Al-Mu'ayyad and his son reduced the border states to temporary submission more than once. Barsbay went to war with the White Sheep in 1433 and besieged Amid without success. Jaqmaq took the other approach entirely — cultivated the border chieftains, married their daughters, and received their homage at least in form. Inal continued that conciliatory line, kept a friendly if nominal overlordship, and had to send an army in 1456 to recover Tarsus and Adana from the prince of Karaman anyway.

What vassalage had become

Then Uzun Hasan of the White Sheep found the method that exposed the whole arrangement.

He captured Egyptian cities and fortresses — Kharput, Karkar — and then sent the keys to Cairo, on the theory that this made it all correct.

Lane-Poole's sentence ends with an exclamation mark, and it deserves one. A vassal was taking his overlord's cities by force and presenting the conquest as an act of homage, and the overlord had no answer, because the alternative to accepting the keys was a war Cairo could not be sure of winning. Uzun Hasan went on sending presents of camels, coats of mail and Circassian slaves for years.

The neighbour stops pretending

Egypt and the Ottomans had been formally friendly. Inal decorated Cairo for several days in 1453 when Mehmed II's embassy brought news that Constantinople had fallen.

The signs that this was over came as protocol before they came as war. An Ottoman embassy in 1464 pointedly disregarded some of the customary etiquette — which in this world is a message, sent deliberately, and read correctly.

Then the two states took opposite sides over the succession to Karaman and nearly came to blows. Then, in the dispute between two brothers for Dhu'l-Qadr — a state technically tributary to Egypt — the Porte secretly supported Shah Suwar, the candidate Cairo opposed. The mamluks were beaten with heavy loss at Ayntab, and beaten again near the Ceyhan.

Egypt eventually won that one. Deprived of Turkish support, the rebel had to surrender, was brought to Cairo and hanged. But Lane-Poole's conclusion is the point of the episode: the course of the war had shown the danger of Ottoman intervention.

Winning the argument, losing the position

Qaytbay reigned longer than any other Circassian sultan and was, by every account, one of the capable ones. His buildings are still the finest things in Cairo. He held the frontier, hanged the pretender, kept the vassals nominally his.

And at the end of it the northern march was a place where an Ottoman sultan could choose a candidate, back him quietly, beat an Egyptian army twice through a proxy, and lose nothing when the proxy failed.

That is a different kind of decline from a lost battle or an empty treasury. The sultanate could still win engagements on that frontier for another forty years. It could no longer make the frontier mean anything, because the state on the other side of it had worked out that Egyptian suzerainty in the Taurus was a form of words — and had discovered, at no cost, exactly how much force it took to make Cairo march.

The monopoly, and then the Cape

Drafted from (ed.), The Cambridge History of Egypt, Volume 1: Islamic Egypt 640-1517 (1998)contraction

Barsbay, sultan from 1422 to 1438, responded to a shrinking agricultural revenue by taking the trade revenue directly. He forced the Indian Ocean traffic away from Aden and through Jeddah, where he could tax it; he then made the state the sole buyer and seller of pepper and other spices at Alexandria, fixing the price the Venetians paid. He occupied the Hejaz to enforce it, raided Cyprus in 1426 and captured its king, and struck gold coin again from the ransom.

It worked, in the short term, and it damaged the thing it taxed. The Karimi syndicate that had carried the trade for two centuries did not survive the state becoming its competitor. European buyers, facing a monopoly price at Alexandria, had a standing motive to find another way to the pepper, and the Portuguese crown was already looking.

Vasco da Gama reached Calicut in 1498. Within a decade Portuguese squadrons were in the Indian Ocean intercepting the Red Sea traffic at its source. Qansuh al-Ghawri built a fleet, had timber hauled overland to Suez because Egypt has none, sent it to India in alliance with Gujarat, and lost it at Diu in 1509 to ships and guns his shipwrights could not match. The spice revenue never recovered, and the state was in permanent fiscal crisis for its last twenty years — debasing coin, confiscating waqf endowments, and taxing the same merchants it had bankrupted.

The military problem was the same problem. Firearms existed and the Mamluks knew it; al-Ghawri raised an arquebus corps and it was despised, because the mounted archer trained from boyhood was the whole identity of the caste and a gun could be handed to a peasant in a fortnight. Historians have argued about how much this mattered ever since. What is not in dispute is that the sultanate met the Ottomans in 1516 with a cavalry army and lost to artillery.

Marj Dabiq

Drafted from (ed.), The Cambridge History of Egypt, Volume 1: Islamic Egypt 640-1517 (1998)end

Selim I had beaten the Safavids at Chaldiran in 1514 with field artillery and janissary firearms, and the Mamluk position — a neutral third power between two larger states, holding the Syrian corridor between them — stopped being available. Qansuh al-Ghawri marched north in 1516, whether to fight the Ottomans or to mediate is disputed by the sources.

At Marj Dabiq north of Aleppo on 24 August the Ottoman guns broke the Mamluk centre, the governor of Aleppo changed sides, and al-Ghawri died on the field — of a stroke, by most accounts, at about seventy-five. Syria was Ottoman within weeks. Tuman Bay II was raised in Cairo, refused Selim's terms, and fought again at Raydaniyya outside the city in January 1517 with entrenched cannon he had had no time to learn to use. Cairo was taken after street fighting. Tuman Bay was hanged at the Bab Zuwayla in April.

Two things went to Istanbul. One was the last Cairo Abbasid caliph, al-Mutawakkil III, taken north and eventually returned; the Ottoman claim to have received the caliphate from him is a much later story, constructed in the eighteenth century when it became useful. The other was the guardianship of Mecca and Medina, which the sharif of Mecca transferred to Selim, and which the Ottomans did put at the centre of their self-description for the next four hundred years.

The mamluk households were not abolished. Egypt was governed as an Ottoman province by a pasha from Istanbul alongside a mamluk beylicate that kept buying boys from the Caucasus, training them in Cairo, and running the countryside; by the eighteenth century the beys were the effective government again, and it took Muhammad Ali's massacre of them in 1811 to end the system. The sultanate stopped in 1517. The institution it was built on did not.

The mayor

Drafted from Lane-Poole, A History of Egypt in the Middle Ages (1901)afterlife

The sultanate ended in 1517 and the system did not.

Selim took Egypt, hanged the last sultan at the Zuwayla gate, and installed as his first governor the Mamluk commander whose defection had helped him win. Egypt became a province administered from Constantinople and separated from Syria and Arabia beside it.

But the mamluks were still there — still being bought, still being trained, still holding the land grants and the offices. Lane-Poole's sentence is the one to keep: the power of the mamluks was not extinguished.

The slow reversal

As the decades passed, the authority of the Turkish pasha and his janissaries shrank, and the strength of the mamluk households grew back into the space.

Their chief emir acquired a title, and the title is almost comic given what the office actually was: shaykh al-balad, the mayor. The man who really governed Egypt on behalf of a sultan in another country was formally the head of the town.

By the eighteenth century the mayor was the government. In 1768 Ali Bey expelled the Ottoman pasha outright, declared Egypt independent, took part of Arabia and tried to annex Syria — which is exactly the programme of every Egyptian ruler since Ibn Tulun, attempted eight hundred and eighty years later by a purchased soldier from the Caucasus.

He was betrayed, beaten and killed in 1772 by his own favourite general, whom he had raised. That is also the programme.

How it actually ended

The mamluk households went on fighting each other over Egypt, more or less subject to the Porte, until Napoleon landed in 1798 and beat them outside Cairo, turning Egypt into a French province for three years. The British ended that; the Ottoman authority was restored on paper; and the mamluks and the pasha resumed their quarrel where they had left it.

It was finished by Muhammad Ali, who expelled the pasha in 1805 and massacred the leading mamluks in 1805 and again in 1811, and founded the dynasty that ruled Egypt until the twentieth century.

Five hundred and fifty years

Count it from the Bahriyya killing the last Ayyubid sultan in 1250 to the citadel massacre of 1811 and you get five hundred and sixty-one years in which Egypt was governed, or effectively governed, by men bought as children in the Black Sea or the Caucasus, trained as cavalry, freed, and made into an aristocracy that renewed itself by purchase rather than by birth.

Two hundred and ninety-four of those years came after the sultanate had been formally abolished by a foreign empire, which is the part worth sitting with. The Ottomans conquered the state. They did not touch the institution, because the institution was how Egypt worked, and for the next three centuries it went on working that way underneath a governor sent from somewhere else.