Hafsid Sultanate

Keeping the province when the doctrine went

Drafted from Abun-Nasr, A History of the Maghrib in the Islamic Period (1987)formation

The Almohad empire rested on a claim about a man: that Ibn Tumart had been the infallible mahdi, and that the caliphate his followers built was the political expression of that. After the defeat at Las Navas de Tolosa in 1212 the later caliphs quietly abandoned the mahdist doctrine rather than defend it, and in doing so removed the argument for their own authority.

The Hafsids are what the abandonment produced at the other end of the empire. The family descended from Abu Hafs Umar, one of Ibn Tumart's inner circle of companions, and had governed Ifriqiya for the Almohads since the beginning of the century. In 1229 the Almohad caliph al-Ma'mun formally repudiated Ibn Tumart's doctrine, and Abu Zakariya Yahya, governing at Tunis, declined to follow him — on the grounds, exactly inverted from the usual secession, that the centre had left the movement rather than that the province was leaving the centre.

Provinces that break away usually have to invent a justification. This one had its justification handed to it. A governor whose family's standing came from proximity to the mahdi could keep the province simply by remaining loyal to a doctrine the capital had discarded, and did.

What he kept was worth keeping. Ifriqiya had been the Aghlabids' province and before them a Roman one, with the agriculture of the Tell, the ports, and Tunis sitting where the caravan routes running north from the Sudan reached the Mediterranean. The Aghlabids had farmed the same ground for tribute four centuries earlier and lasted a hundred and ten years on it. The Hafsids held it for three hundred and forty-five.

Seville writes to Tunis

Drafted from Khaldun, Histoire des Berbères et des dynasties musulmanes de l'Afrique septentrionale (1852)expansion

Abu Zakariya kept Ifriqiya in 1229 and spent the next twenty years finding out how much else would come to him without being asked. The answer, briefly, was al-Andalus.

What was happening on the other side

Spain had gone to pieces. The Almohad princes were quarrelling over the shreds of the monarchy; Ibn Hud had usurped the east and Ibn al-Ahmar had the west in revolt; in 1229 Ibn Hud drove the Almohads out of Seville by force and imprisoned all of them. A few years later the Sevillans expelled Ibn Hud's brother and swore allegiance to one of their own notables, a descendant of a famous jurist, who took a regnal title and was killed with his vizier outside the walls by a force Ibn al-Ahmar had sent against his camp. Ibn al-Ahmar held the city for less than a month before the population rose and called Ibn Hud back. Then Ibn Hud died and they recognised the Almohad at Marrakesh instead and put a local man in charge, and when that Almohad died in 1242 they had to think again.

Six changes of allegiance in thirteen years, none of them producing anything that could defend the city.

The deputation

When Abu Zakariya took Tlemcen, the people of Seville judged the kingdom of Morocco to be finished, and decided to recognise the sovereignty of the sultan of Tunis, as Valencia and Murcia had already done. Jerez and Tarifa followed. Then all these towns together named a deputation and sent it to Tunis, to ask Abu Zakariya for one of his relatives to govern them.

He chose a cousin, Abu Faris, and sent him. That is an Ifriqiyan dynasty barely fifteen years old being asked by the cities of Andalusia to supply them with a ruler — the Friday sermon said at Seville in the name of a man in Tunis, across the whole width of North Africa and a strait, with two hostile states in between.

And then

It did not hold, and the way it failed is instructive. In 1245 the notable who had handed Abu Faris the command took the power back for himself, deported him to Ceuta, made an alliance with the Christian king, and struck off the army rolls the names of every man who persisted in raiding Christian territory. Those men murdered him shortly afterwards. Their captain took over the city, recalled Abu Faris and restored the Hafsid authority.

The Christian king heard this with lively displeasure, broke the truce, took Carmona and Marchena and sat down in front of Seville. The city refused his peace terms and handed the command to a council of five — the captain, three other notables, and a man named Yahya ibn Khaldun, whose great-grandson wrote this account and mentions the relationship in passing.

Seville fell in 1248. What the episode leaves behind is a fact about the Hafsid state at its height that no map shows: for a few years in the 1240s the most credible Muslim authority in the western Mediterranean was in Tunis, and everybody from Murcia to Tarifa knew it, and it made no difference at all to what Castile was going to do.

A caliph in Tunis, for about a year

Drafted from Abun-Nasr, A History of the Maghrib in the Islamic Period (1987)peak

In February 1258 Hulegu had the last Abbasid caliph killed at Baghdad, and for the first time since the 630s there was no caliph. The following year the sharif of Mecca — the guardian of the holy cities, whose acknowledgement was the most valuable recognition available — acknowledged the Hafsid ruler al-Mustansir as caliph, and had the Friday sermon at Mecca preached in his name.

It did not last and it never became general. The Mamluks installed a surviving Abbasid at Cairo in 1261, and Meccan recognition followed the sultanate that could supply the holy cities with grain. But it happened. For a stretch of about a year, the office that had been in Medina, Damascus and Baghdad was located, by the man best entitled to locate it, in Tunis.

The state underneath the title was solid and unusual for this region in being commercial rather than tributary. Tunis stood where the trans-Saharan caravans from the Sudan met the Mediterranean, and the Hafsids ran both ends deliberately: treaties with Pisa, Genoa, Venice and the Crown of Aragon giving European merchants their own quarters and consuls in Tunis and Bougie, and corsairing against those who had no treaty, which the sources here treat as a branch of revenue rather than as disorder. Gold came north, cloth and metal went south, and the customs were the state's.

That combination is what brought a crusade. Louis IX of France landed at Carthage in 1270 in an expedition that had been aimed at Egypt and was diverted here, and it collapsed almost immediately: dysentery went through the camp and the king died in it in August. The Hafsids had not beaten a crusade so much as outlasted one, and the treaty that followed put the relationship back on the commercial footing it had had before.

Five months in a castle

Drafted from Abun-Nasr, A History of the Maghrib in the Islamic Period (1987)aside?What aside means

Ibn Khaldun was born at Tunis in 1332, into a family of Andalusi origin that had left Seville for Ifriqiya when Castile took the city and had served the Hafsids since. He held office under them, and then — because the Maghreb in his century was four or five competing dynasties and a scholar with administrative skill was portable — under the Marinids at Fez, the Zayyanids at Tlemcen, the Nasrids at Granada, and the Hafsids again. He was imprisoned twice, changed sides repeatedly, negotiated with Pedro the Cruel of Castile, and eventually left for Egypt, where he was a Maliki chief judge under the Mamluks and, in 1401, went out from besieged Damascus to talk to Timur, who questioned him for several weeks about the Maghreb.

In 1377, in a castle in what is now western Algeria, between employments, he wrote the Muqaddima — the introduction to a universal history — in about five months.

What it argues is why he belongs in the history of the Maghreb and not only in the history of scholarship. Ibn Khaldun asked why dynasties rise and fall, and answered it with a mechanism rather than a moral. Cohesion — asabiyya, the solidarity of a group that has to hold together to survive — is strongest among people living hard lives on the margins: desert tribes, mountain confederations. That cohesion lets them take a settled state from a dynasty that has lost it. Having taken it, they settle: they acquire cities, taxes, luxury, a professional army in place of kinsmen, and the solidarity that brought them in dissolves over roughly three or four generations. Then a group from the margins that still has it arrives, and the cycle runs again.

He was describing what he could see out of the window. The Almoravids came out of the Sahara and were replaced by the Almohads out of the Atlas, who disintegrated into the Hafsids, Zayyanids and Marinids, who were exhausting themselves against each other while he wrote.

The window was not the limit of it. The Tahirids, Saffarids, Samanids and Ghaznavids run the same pattern across Khurasan four centuries earlier. The Seljuks come off the steppe and are Persianate within a generation. So do the Karakhanids, the Qara Khitai, the Turkmen confederations, and the Ottomans out of the smallest of the Anatolian beyliks. A Tunisian civil servant sitting out an unemployment in a castle in western Algeria produced, in five months, a description of state formation that fits most of the Islamic world for most of a millennium — and he produced it from inside one of the dynasties it applies to, which is the part that ought to be hardest.

The Hafsids left a state that was solvent, commercial and long-lived, and it is not why anyone reads about them. The most consequential thing produced under their rule was a book about why states like theirs do not last.

A jerboa hole

Drafted from Khaldun, Histoire des Berbères et des dynasties musulmanes de l'Afrique septentrionale (1852)contraction

In 1347 the Marinid sultan of Fez took Ifriqiya, and the Hafsid state ceased to exist for about twelve years. Ibn Khaldun was fifteen and in Tunis for all of it.

How a country changes hands without a battle

Abu'l-Hasan had wanted Ifriqiya for years and needed a pretext. He got one when the Hafsid heir was assassinated: he had personally signed the document naming that prince successor, so he could declare the act violated and the war just. He distributed money to his troops in the camp outside Tlemcen and marched east with an immense army.

Nobody fought him. The Arab emirs of the Ifriqiyan interior had sent to ask for his support in a feud of their own and got a favourable answer, after which the tribes declared for the Marinids as far as the furthest provinces. Then the lords of the south came in: Gabès, Tozeur, Gafsa, El-Hamma, Nefta, all at once, meeting him at Oran and swearing allegiance — some willingly, others from fear, Ibn Khaldun says, and he was in a position to know which. They brought the homage of the lord of Tripoli, who had been unable to travel that far. Bougie thought about resisting and submitted. Constantine submitted. Algiers sent a deputation and went home with a set of tax collectors.

The Hafsid princes were handled with care rather than cruelty, which is more efficient. The amir of Bougie was sent to Morocco with his brothers and given Nedroma to live in, with a portion of the town's taxes for maintenance. The sons of another prince went to Oujda on the same terms. At Constantine Abu'l-Hasan opened the prison, released the Hafsid princes who had been shut up there by their own family, took them to Tunis and shipped them to Morocco with the rest.

The horse

One of them ran. Abu Hafs left Tunis with the tribes that would still ride with him and made for the desert. A detachment caught up in the territory of El-Hamma, near Gabès; the tribes scattered after a feeble resistance; Abu Hafs turned to flee, and his horse put a foot in a jerboa hole and came down under him.

When the dust settled they saw the amir and his freedman Zafir going away on foot. The commander arrested them both and put them in irons, and at nightfall had them killed, in case the Arabs came for them. The heads were carried to Abu'l-Hasan, who had just entered Béja.

He made his entry into Tunis in September 1347 with a magnificent procession. Ibn Khaldun's sentence for what was left is the best one in his account of the century: nothing remained of the brilliant empire of the Hafsid Almohads but a feeble spark, whose gleam could still be made out at Bône.

The spark was there because Abu'l-Hasan had left it. He installed a Hafsid prince as governor of Bône as a courtesy to a family he had married into.

Then he went sightseeing

Having finished, the conqueror went to Kairouan, then Sousse, then Mahdia. He looked at the monuments in those towns and the remains of the older buildings the Fatimid kings and the Zirids had put up. He sought a blessing at the tombs where certain Companions of the Prophet and their immediate successors are said to lie. Around December he went back to Tunis.

Ifriqiya had been a Roman province, then the Aghlabids', then the Fatimids', then the Zirids', and a Moroccan sultan who had just taken it spent his first winter as its owner visiting the graves.

Twelve years

It did not hold either. Marinid Ifriqiya broke on the same Arab tribes that had handed it over; the Hafsid princes came back from Morocco; Mahdia went over to Tunis before the fleet sent to recover it arrived; and in 1359 the people of Bougie, who loathed the governor Fez had given them, secretly invited the Hafsid chamberlain to come and take the city. The garrison commander and his household were put on a ship to Tunis by the mob, and the sultan Abu Ishaq occupied Bougie the following year.

Hafsid rule was restored and lasted another two centuries, and the shape of it had changed. The sultan who reoccupied Bougie exercised almost no authority there, because his chamberlain Ibn Tafragin governed him from the capital and directed him as he pleased.

Rain-water out of cisterns

Drafted from Africanus, The History and Description of Africa (1896)contraction

A Granadan diplomat who travelled the Maghreb for the sultan of Fez in the 1510s called Tunis the capital of the richest kingdom in all Africa, and then described a city whose citizens could not farm the ground outside their own walls and drank rainwater out of cisterns.

Both halves are true, and the gap between them is the Hafsid state.

The city

Leo counts. The families of the city together with the suburbs come to about ten thousand. Each trade has its own quarter. There are linen-weavers in great numbers and the cloth they make is exceedingly fine and sells at a high price all over Africa, and the linen-drapers are the wealthiest men in Tunis.

He watched the women spin and thought the method worth recording: standing on an upper floor, they let the spindle down through a window or a hole in the floorboards into the room below, so that its own weight draws the thread even.

Three suburbs had been built outside three gates by two successive sultans. One holds three hundred families; one holds more than a thousand, and in it live all the Christians of the king's guard. The third is at the sea gate, half a mile from the gulf:

Hither doe the Genoueses, Venetians, and all other Christian merchants resort, and here they repose themselues out of the tumult and concourse of the Moores: and this street is of so great bignes, that it containeth three hundred families of Christians and Moores, but the houses are verie low, and of small receit.

The two sentences that matter

Then, in the middle of the inventory, Leo explains why a city that rich has so few rich men in it.

Rich men here are but few, by reason of the scarcitie of all kinde of graine: for a man cannot till a piece of ground, be it neuer so neere the citie, in regard of the manifold inuasions of the Arabians.

Grain came from other places — Urbs, Béja, Bône — and the citizens who did farm did it inside walled fields in the suburbs, each with a pit and a wheel turned by a mule or a camel to water the upper part of the ground. Leo invites the reader to consider what crop can be reaped from a field that small, walled round, and watered by such industry.

The second sentence is about water.

In this citie they haue no fountaines, riuers, nor welles of fresh water: but they all vse raine-water taken out of cesterns.

There is a fountain in the suburbs from which porters carry salt water into the city to sell, which the inhabitants think more wholesome to drink than rainwater. The wells that give good water are reserved for the king and his court.

The bread is excellent, he says, though they leave the bran in the flour. The merchants and most of the citizens live on besis, a barley-meal dumpling with oil or citron broth poured over it, and there is a place in the city where nothing is sold but barley made ready for it. All the mills except a few on a river outside are turned by mules and asses.

What that is a description of

A state whose capital sits on a coastal plain it does not control.

Everything in the Hafsid century that looks like strength is in the first list: the linen, the customs, the Christian merchant quarter, the king's Christian guard, the court that Leo says was organised on the Moroccan model and that produced the most impressive chancery in North Africa. All of it is revenue taken from things passing through a port.

Everything in the second list is what the dynasty could not do. A capital that cannot be provisioned from its own hinterland is a capital living on grain convoys and good relations with whoever holds the road, and a city without fresh water is a city that can be besieged by anybody willing to sit down in front of it. Within twenty years of Leo's visit an Ottoman corsair had taken the city and Charles V had taken it back and reinstalled the Hafsid as a Spanish client; forty years after that the Ottomans took it for the last time, and the last Hafsid was carried to Constantinople and killed there.

The men with the stones

Leo leaves one thing standing outside all the arithmetic.

There are certaine people in this citie whom a man would take to be distraught, which goe bare-headed and bare-footed, carrying stones about with them, and these are reuerenced by the common people, for men of singular holines.

And the king of Tunis built one of the city's monasteries for one of them — a man Leo names, and whose society he calls fond — and endowed it with the most ample revenues.

That is a Hafsid caliph, in the capital of the richest kingdom in Africa, putting state money into a lodge for barefoot holy men carrying stones. In a dynasty that had spent three centuries arguing that it held the Almohad caliphate by right, it was a reasonable investment: the marabouts were the only authority in the countryside the state had lost.

A prize in somebody else's war

Drafted from Abun-Nasr, A History of the Maghrib in the Islamic Period (1987)end

The Hafsids lasted three hundred and forty-five years — longer than the Almoravids, the Almohads, the Marinids and the Zayyanids put together at any one place — and the last stretch of it is the least like the rest.

The fourteenth and fifteenth centuries were the ordinary difficulties of a Maghribi state: Arab tribal confederations in the interior that the centre taxed when it could and negotiated with when it could not, rival branches of the family holding Bougie and Constantine, and a commercial relationship with Europe that continued through all of it because both sides wanted it to.

What ended the dynasty was not a Maghribi problem at all. In the sixteenth century the western Mediterranean became the frontier between two empires that had nothing to do with Ifriqiya, and Tunis was on it. An Ottoman force took the city in 1534; Charles V retook it in 1535 and put the Hafsid back on the throne as a Spanish tributary with a garrison at La Goulette. From that point the dynasty was not a party to the contest but an instrument in it, restored, deposed and restored again as the two powers needed a local name to rule through.

In 1574 the Ottomans took Tunis for the last time and stopped needing one. The last Hafsid was carried to Constantinople and killed there.

The histories do not agree on which Hafsid he was. Several of the men carried between Tunis and La Goulette in those forty years bore the same regnal names as their fathers and uncles, and the numbering of the last of them differs from one account to another. It is a small thing and an appropriate one: a line that had said the Friday prayer in its own name for three and a half centuries went out without the sources being able to agree who was in the room.

Forty sections, and a chief for each

Drafted from Rousseau, Annales tunisiennes, ou aperçu historique sur la régence de Tunis (1864)afterlife

The Ottomans took Tunis in the summer of 1574 and did not put a Hafsid back. What they put there instead ran the country for the next three hundred and fifty years, which is about as long as the dynasty it replaced, and it was assembled in a few weeks by a man who then sailed home.

The taking

Sinan Pasha — a Milanese renegade said to be of the Visconti family, at least according to the story attached to him — arrived to find that the pashas of Algiers, Tripoli and Kairouan had already agreed to bring their contingents to him. The Spanish commander Serbelloni evacuated Bizerte and put its garrison into La Goulette, and manned the fort on the lake with the troops who had been holding down the population of Tunis. The Turks were investing both before he had finished.

La Goulette fell first. Everything in it was killed except two or three hundred men who were put in chains, the governor among them. The fort between the city and the lake fell a few days later; when the citadel of Tunis went on 3 September its garrison was down to thirty men still on their feet. Serbelloni was one of them, and was spared because his ransom would be large; Rome offered to exchange him for several Muslims held in the Castel Sant'Angelo, one of them a son of Ali Pasha taken at Lepanto.

The Tunisian historians say the two sides lost about ten thousand men each. The Spanish historians say the Turks lost thirty-two thousand. Rousseau prints both figures and asks how they are to be reconciled, and does not answer, which is the correct thing to do with them.

The machine

Before returning to Constantinople, Sinan built a government. He left four thousand troops, divided into forty sections, and put at the head of each a chief who took the title of dey. Over them he set a pasha, and named the man. He instituted a qadi to judge suits between private persons by the civil law, and named the first one. He organised a Divan on the model of the ones already working at Algiers and in Egypt. Then he fixed the pay of the troops, the salaries of the senior officials, the details of the administration and the official costumes.

The public prayer in the mosques was said in the name of the reigning Ottoman sultan, and the coin was struck in his name.

That last sentence is the whole transfer. The two things a Maghribi ruler possessed that could not be delegated were the khutba and the mint. For three hundred and forty-five years both had carried a Hafsid; a caliph's name for a few months of 1259, a sultan's name for the rest. After 1574 they carried Constantinople's.

What it turned into

Rousseau, who despises the arrangement and says so, notes the obvious flaw: it armed men without control who owed their sudden elevation to nobody, and like every institution founded on violence it did not last long. In 1590 the soldiery decided to be rid of the bölük bashis whose administration had become insupportable to the population and to the army alike. The minister of war, who was one of the conspirators, arranged to be away from his post that day. They broke the gates of the Kasbah, looted the arsenal, and went through the city killing their officers.

Out of that, and out of the century of adjustments after it, came hereditary rule in Tunis again — twice. Power became, Rousseau writes, in some sort hereditary in the family of Murad, until a coup in 1702 sent the heads of five of them to the capital to announce it. Three years later the militia elected as bey an agha named Husayn ben Ali, the son of a Greek renegade recruited in the Levant to fill out the ranks of the Tunisian soldiery, whose austerity and ability had carried him up through it. With him, Rousseau says, began a dynasty that had succeeded itself on the throne of Tunis by heredity down to the year he was writing in.

An occupying force sent to end a dynasty in Tunis had, within a century and a half, produced another one — governing the same ground from the same city, paid for by the customs on the same trade, and founded from inside the garrison that was supposed to be the alternative to having a dynasty at all.