Ziyadid Dynasty

A governor who stayed

Drafted from Bosworth, The New Islamic Dynasties: A Chronological and Genealogical Manual (1996)formation

In 819 the caliph al-Ma'mun sent an officer named Muhammad ibn Ziyad down to the Yemeni coast to put down a rising among the Alid partisans of the Tihama. This was an ordinary piece of Abbasid administration and there is no reason to think anybody in Baghdad expected it to produce a dynasty. Governors were sent to Yemen regularly. They were recalled, or they died, or they failed.

Ibn Ziyad did the job and then did the thing that turns an appointment into a state: he built somewhere to govern from. In 820 he laid out Zabid on the coastal plain, forty kilometres inland from the sea and set among the seasonal watercourses that come down off the highland escarpment. It was a working administrative seat rather than a captured city, and his family held it for two hundred years.

A provincial family usually becomes a state at the moment it acquires a centre of gravity that belongs to it rather than to whoever appointed it. Most of them take a generation or two to reach that point. Ibn Ziyad reached it in his second year, and did it while calling himself the caliph's servant, which he went on doing for the rest of his life.

That is the tension the dynasty never resolved and never seems to have wanted to. The Ziyadids were an Abbasid governorship in their own account of themselves from beginning to end. They named the caliph in the Friday prayer for two centuries, including the long stretch after Baghdad had lost the power to appoint or remove anyone in Yemen. No secession was ever declared. What happened instead was that the empire receded and the governorship stayed where it was, which is a different thing and much harder to date.

The Tihama

Drafted from trans.), Yaman: Its Early Mediaeval History (1892)expansion

The Tihama is a strip of hot coastal plain between the Red Sea and the wall of the Yemeni highlands, and it is not obvious country to build a state on. It has almost no rain. What it has is the water that comes off the escarpment in the monsoon and runs down a set of great seasonal wadis, and a system of diversion works that spreads that flood across fields of grain, cotton and dates before it reaches the sea. Zabid sat on that system.

Ziyadid authority ran up and down the plain and, in the accounts, at times a good deal further — over the ports, out to the islands, and inland towards Hadramawt. How far it reached at any given date nobody can now say in figures. Umara, writing three centuries later from within the tradition, is describing a remembered greatness, and the boundaries in his account move without warning between what was taxed, what was garrisoned and what merely read the ruler's name in the Friday prayer.

What can be said is what the plain was for. The Red Sea run to Egypt passed the Tihama ports, and the pilgrim traffic came through them, and a state sitting on the diversion works and the customs houses had two sources of money where the highland powers above it had one. The grain and the cotton paid for the government; the sea paid for what the government could not otherwise have afforded.

The highlands were another matter. Above the escarpment were the Zaydi imams at Sa'da from 897, and tribes that had been governing themselves for a very long time, and the Ziyadids never held that country in the way they held the plain. The division outlasted them by nine hundred years. Every dynasty that afterwards ruled the coast ran into it, and so did the Ottomans and the Egyptians when their turn came.

A million dinars

Drafted from trans.), Yaman: Its Early Mediaeval History (1892)peak

Abu'l-Jaysh Ishaq ibn Ibrahim held the Tihama for something like eighty years, which is long enough that a historian writing two centuries later could take his reign as the shape of the dynasty rather than an episode in it. Umara, who had the court records of Zabid to work from, does something better than praise him. He copies out an account.

He had seen, he says, a statement of Ibn Ziyad's revenues for the year 366 — 976 or 977 — and notwithstanding the reductions they had undergone by then, they came to a million 'Aththari dinars.

Then he lists what the million did not include. Duties levied on ships from India. Contributions of musk, camphor, ambergris, sandalwood and Chinese goods. Taxes on ambergris gathered along the shores of Bab al-Mandab, at Aden, at Abyan, at Shihr and elsewhere. Imposts on the pearl fisheries. And the tribute of the ruler of Dahlak, the island group off the African coast, which was paid partly in people: a thousand head of slaves a year, of whom five hundred were Abyssinian and Nubian women.

The kings of the Abyssinians, on the far side of the sea, sent him presents and sought his alliance.

What that describes

It describes a customs state.

The million in coin is the land revenue, the ordinary tax on the irrigated grain and cotton of the coastal plain, and Umara notes it as the diminished figure. On top of it sat everything the Red Sea produced: a percentage of the Indian trade at the point where it entered the Egyptian route, the specialised commodities of the ocean, the ambergris washed up along four hundred kilometres of coast, the pearls, and an island tribute levied in human beings.

The slave figure is the part to sit with. Five hundred women a year, from one tributary, as a standing item in a revenue account, alongside camphor and pearls. The dynasty's own army and administration were built out of the same traffic in the same direction, and the men who eventually took the state came up that way. It is one commerce, and the Ziyadids were both its customer and its tax collector.

Whose name was on the coin

The revenue statement is dated to the eighth decade of the reign, and Umara is explicit that by then some of the provinces had separated themselves. What that separation consisted of is the interesting part, and it is easy to misread.

Abu'l-Jaysh did not govern most of Yemen. The king at San'a, a Himyarite named As'ad ibn Ya'fur, held San'a and Bayhan and Najran and Jurash. Sulayman ibn Tarf held 'Aththar on the northern coast, a territory seven days' journey long, with revenues of half a million dinars of his own. Al-Harami held Hali. In the northern mountains the Zaydi sharif at Sa'da had raised a revolt that no lowland army was going to reach.

And every one of them recited Abu'l-Jaysh's name in the Friday prayer and struck his name on their coinage.

Ibn Tarf paid tribute and sent presents but would not come to court in person. As'ad ibn Ya'fur paid nothing at all — no tribute, no contributions of stores, not even the customary gifts — and put the coastal king's name on his money anyway.

That is what supremacy in Yemen looked like at its height, and it is worth taking seriously rather than reading as a fiction. Nobody in the country was going to be conquered by anybody: the highlands were fortresses, the plain was hot, and the distances were long. What could be had was precedence — the acknowledgement, in the two places where a medieval Muslim ruler's standing was published, that somebody else was the government. An old man at Zabid had that from every prince in Yemen while taxing a fraction of it, and a million dinars besides.

Husayn ibn Salama

Drafted from trans.), Yaman: Its Early Mediaeval History (1892)peak

The most consequential figure the Ziyadid state produced was not a Ziyadid.

Husayn ibn Salama was a Nubian slave — the property of an Abyssinian eunuch who was himself the property of the ruling family — who rose to be vizier in 981 and governed for thirty years in the name of a dynasty that had stopped governing. What he is remembered for is not politics. He built: congregational mosques, cisterns, wells and marked stages the length of the road from Hadramawt to Mecca, and the tradition that reaches Umara is a tradition about infrastructure rather than about campaigns.

This is worth pausing on, because it is the point at which the state was at its most effective and the point at which the family that owned it had become irrelevant to it. Both things are true at once and they are the same fact seen from two ends. The men who could raise revenue, move troops and build across five hundred kilometres of road were the household slaves. The men whose name was in the khutba were minors in a palace at Zabid.

Governments run by purchased men were the normal arrangement across the Islamic world in this century — at Samarra, at Ghazna, at Cairo, at Aleppo — and they failed in a consistent way. An army that can make a ruler will eventually make one it prefers. What is unusual at Zabid is that the failure took no violent form for two generations. There was no coup, because there was nothing to overthrow: the household simply went on running the state until there was nobody else in it.

The frontier is the other unfamiliar part. These men were bought across the strait from the Horn of Africa rather than on the Kipchak steppe, which is what a Red Sea state does, and it is why the dynasty that eventually took the name was Abyssinian.

The name nobody knew

Drafted from trans.), Yaman: Its Early Mediaeval History (1892)contraction

Husayn ibn Salama died in 1012 after thirty years in charge, and the state passed to a child of the Ziyadid family. Umara, who is the best-informed writer on this dynasty and had spent his youth in Zabid, cannot tell us the boy's name.

He says so plainly: sovereignty passed to a child of the family whose name I do not know, but I believe it to have been 'Abd Allah. He offers Ibrahim as an alternative elsewhere. He is not being careless — he names hundreds of people in this book, down to the jurists of small towns and the men who told him anecdotes about a vizier's piety. He simply could not establish what the last kings of Zabid were called, a century and a half after they lived, in the city where they lived, from the records of the state they nominally headed.

That is the most exact measurement of the Ziyadid position available. The dynasty had become an item of protocol.

Slaves of slaves of slaves

What had happened to the government by then is best set out as a chain.

The boy was placed under the guardianship of a paternal aunt and of a eunuch named Marjan, who held the office of vizier. Marjan had been a slave of Husayn ibn Salama. Husayn had been a Nubian slave of an Abyssinian eunuch named Rushd, who had been a slave of Abu'l-Jaysh and had held the guardianship of the previous child-ruler alongside his aunt. And Marjan in turn had two Abyssinian slaves whom he had brought up from childhood and, when they came of age, appointed to run the administration: one named Nafis, who took the capital, and one named Najah, who took the four best provinces north of Zabid.

Four generations of servitude, each one owning the next, each one exercising more of the state than the family that owned all of them.

There is no coup in this and no moment where anybody seizes anything. Each step is a household appointment, made for the ordinary reason that the man was competent and the master trusted him, and the cumulative effect is a state in which every person who can raise a soldier or collect a tax is the property of somebody who is himself property.

The arrangement holding

It is worth saying that the arrangement worked, for a while, better than the family had.

Husayn had inherited a kingdom whose outlying provinces had fallen into decay and whose highland governors had helped themselves to the districts they were posted to. He made war on the mountain chiefs and brought them back to submission; Ibn Tarf and Ibn al-Harami re-entered obedience; he recovered, in Umara's phrase, the original limits of the kingdom, and founded two new towns while doing it.

None of that was done in the name of a policy or a dynasty. It was done by a slave on behalf of a child, and when the slave died the reconquest held only as long as the next slave could hold it.

The weakness in it is obvious in retrospect and was probably obvious then. Power that rests on a household rather than on a family or a law has no rule for what happens when two members of the household want the same thing. Marjan preferred Nafis. Najah had the northern provinces and, Umara says, the affection of the people, where Nafis was feared. The aunt of the little prince was said to favour Najah and to be in correspondence with him.

That is the whole of the situation in 1018, and everything that followed came out of it.

The wall

Drafted from trans.), Yaman: Its Early Mediaeval History (1892)end

Someone told Marjan that the prince's aunt was writing to Najah. Nafis, who was already the favourite and already feared, complained of it to their master, and Marjan laid hands on the princess and on her nephew — the boy whose name nobody recorded — and handed them both over to Nafis.

Nafis had them immured. They were walled into a chamber alive, and they stood praying for mercy and adjuring him in the name of God for as long as an aperture remained, until the wall was closed.

That is the end of the Ziyadid dynasty of Zabid. It had lasted two hundred and six years from the founding of the city.

What happened next

Nafis took the royal umbrella, struck the coinage in his own name, and made himself the ruler of the Tihama.

Najah, in the northern provinces, heard what had been done to his masters and called in his neighbours, Arab and non-Arab, and marched on Zabid. Four battles are named. At Rima' and at Fashal he was beaten. At al-'Ukda he won. At al-'Irq, outside the gate of Zabid, Nafis was killed, with five thousand dead between the two sides. Najah took the city in the month of Dhu'l-Qa'da 412, which is February or March of 1022.

He then asked Marjan what Nafis had done with their masters. They are in that wall, Marjan said. Najah had the bodies taken out, prayed over them, and built a mausoleum over the place where he buried them. Then he had Marjan walled up alive, in the same wall, with the corpse of Nafis.

Afterwards he took the umbrella and the coinage himself, wrote to Baghdad offering his submission, and received a title, the right to appoint the judges, and the administration of Yemen. He ruled the Tihama and much of the highlands and was addressed as king.

The four years

The gap between 1018 and 1022 is not an interregnum in which nothing happened, and it is not, as the tidier version of this story has it, a house that quietly outlived itself. Two Abyssinian slaves of the same master fought a civil war across the coastal plain for the possession of a state whose owners were bricked into a wall in its capital.

What is true in the tidier version is that nobody outside the household was involved. No highland army came down, no Egyptian fleet arrived, no caliph intervened. The entire succession — murder, usurpation, four battles, five thousand dead and a new dynasty recognised from Baghdad — was transacted between men who had all been bought, by masters who had themselves been bought, inside one government at Zabid.

The wall is the right image for it. A ruling family that had been ornamental for two generations was put out of the way by being enclosed in the fabric of its own palace, and the man who did it was buried in the same masonry four years later by the man who replaced him.

The man who wrote it down

Drafted from trans.), Yaman: Its Early Mediaeval History (1892)aside?What aside means

Almost everything known about the Ziyadids comes from one book, and the man who wrote it was hanged in Cairo for trying to bring back the Fatimids.

Umara ibn Ali al-Hakami was born on the northern Yemeni coast, in the country of the Banu Hakam, around 1121. He went to the college at Zabid — the Ziyadid capital — as a student in 1136, and became a jurist and, more profitably, a poet. He was in Cairo repeatedly on Yemeni business, and by the late 1150s he had settled there under the protection of the Fatimid vizier Tala'i ibn Ruzzik. The vizier was a committed Isma'ili and spent some effort trying to bring Umara into the sect; Umara declined, and Tala'i went on protecting and paying him anyway for the rest of his life, which is a fair description of how a court poet's career is supposed to work.

Then the Fatimid state came apart around him. Tala'i was killed; his son was assassinated; the dissension gave Nur al-Din a pretext and he sent Shirkuh, and for five years Egypt was fought over by Syrian troops and Crusaders. When the Fatimid caliphate was finally abolished in 1171, the change reached most Egyptians only as an alteration in the wording of the Friday sermon.

It did not reach Umara that way. A conspiracy was formed in Cairo to restore the Fatimids with the help of the king of Jerusalem; it was found out and suppressed; Umara was among the accused, was convicted, and was executed on 2 Ramadan 569 — April 1174. His body was gibbeted and left on public view for three days.

The book survived him by a thin margin. Kay worked from a manuscript of eighty-four folios in the British Museum with neither the author's name nor the title on it, and the History of Yemen exists as a text at all because later Yemeni historians quoted it so heavily that its absence would have been obvious. Its deficiencies are real: Umara is writing three centuries after the Ziyadid foundation, he is a poet by trade and interested in the things a poet is interested in, and where he touches events in his own lifetime he is a participant rather than an observer.

Which is worth stating plainly, because the alternative is worse. The list of Ziyadid rulers, the million dinars, the wells and the mosques of Husayn ibn Salama, the name nobody could establish — all of it rests on one manuscript by one interested man who was killed for a conspiracy. A polity that lasted two hundred years is known to us because a poet from its province got a job in Cairo and wrote about home before the rope.

The road and the book

Drafted from trans.), Yaman: Its Early Mediaeval History (1892)afterlife

Two things the Ziyadid state made outlasted it by a long way, and one of them is the reason anybody knows about the other.

The road

Husayn ibn Salama's chain of mosques and wells ran from Hadramawt to Mecca, and a hundred and fifty years after his death a young man from the Tihama walked large parts of it and wrote down what he found.

Umara ibn Ali al-Hakami was a jurist of Zabid, trained in the Shafi'i law there, and his account of the road is an eyewitness one. Some of these works I have seen, he writes, either in good order or in ruins. He had visited al-Makhnak on the coastal road, one night's journey from Aden, and looked at its well of eighty fathoms and its ruined mosque. He had seen the great mosque of al-Hawali on the highland road, still in good order as Husayn built it. He had seen the mosque at al-Janad, which was the size of Ibn Tulun's at Cairo, and where the country people had come to believe that a visit in the first week of Rajab was worth the pilgrimage to Mecca itself — so that a man in debt would ask his creditor to leave him alone until after the pilgrimage, and mean the trip to al-Janad.

He lists the stages the whole way north, on the mountain road and the coast road and the royal highway between them, naming the wells and their depths. Every one of these, he says, he has seen himself, either in repair or fallen down.

It is one of the more concrete pictures anyone has of what a medieval Islamic public work amounted to on the ground, and it exists because a lawyer from the right town happened to travel it and to be interested. A day's stage, a mosque, a minaret, a well: repeated for sixty days' journey, and by his time about half gone.

The book

Umara left Yemen and made his career in Cairo, where he arrived as an envoy and stayed as a poet at the Fatimid court. The qadi al-Fadil, chief secretary of the Fatimid government, asked him to write the history of his own country, and he did — the dynasties of Yemen from the Abbasid governor who founded Zabid down to his own day, with the court records of Zabid behind him and, for the earliest part, a book on the history of the city written by one of the kings who had followed the Ziyadids. That earlier book is otherwise lost. What Umara quoted of it is what survives.

Then Saladin deposed the Fatimids, and Umara would not let it go. He wrote a lament for the fallen dynasty — you have blighted, O Fortune, the noblest of hands — and defended them loudly enough and long enough to become dangerous company. When a conspiracy to restore them was discovered he was named in it, tried and condemned. Sentence was carried out on 2 Ramadan 569, in the spring of 1174, and the body was gibbeted for three days.

The qadi who had commissioned the history was by then an adherent of Saladin's and no longer a friend. He was seen speaking privately to the sultan as the sentence was pronounced, and Umara called out not to listen to what he was saying. He was interceding for your life, Saladin answered. Umara put his head down and said nothing.

What is left

Zabid outlived all of it. The town Muhammad ibn Ziyad laid out in 820 to govern a coastal plain from went on being the capital of the Tihama under the Abyssinian kings who replaced him, and under everyone who replaced them, and its mosques and its schools of law went on producing the jurists whose lives fill the Yemeni biographical dictionaries — arranged, as those books are, town by town, with Zabid always among the fullest entries.

So the dynasty's afterlife is a town that kept working, a road that was half in ruins within two centuries, and a book written in exile by one of the town's lawyers, who was hanged for loyalty to a different dynasty in a different country and whose history is the only reason the four hundred and twelfth year of the Hijra at Zabid is remembered at all.