Hujrid Dynasty

The third family

Drafted from Fisher, Oxford Classical Monographs (2011)formation

Three Arab dynasties were on retainer in late antiquity, and it is worth being careful about their names. Ghassan, Lakhm and Kinda are confederacies — groupings of tribes. The Jafnids, the Nasrids and the Hujrids are the ruling families inside them, and they are what the empires actually dealt with; later Arabic tradition merged the two so that a family with a following became a nation with a royal house.

The pattern is the same in all three cases and the paymasters are different. Rome paid the Jafnids to hold the Syrian steppe. Sasanian Iran paid the Nasrids at al-Hira to hold the Euphrates frontier. And Himyar — the South Arabian kingdom at the height of its power in the fifth century — paid the Hujrids to hold central Arabia.

That third relationship is the one usually left out, because Himyar is usually left out. A reader who knows the Jafnids and the Nasrids as the Arab clients of the two great empires has been given a picture with a third empire missing from it: South Arabia in this period was a substantial state with an interest in the tribes to its north, and it managed them the way Rome and Iran managed theirs.

What all three families held, they held on the same terms. Their money came from somebody else's treasury, and their right to lead came from somebody else's grant — which is a different kind of authority from either conquest or descent, and it is the same kind whether the grantor was the Roman emperor or a king in the Yemeni highlands. It works for exactly as long as the grantor can pay and wants to.

A power in the middle

Drafted from Fisher, Oxford Classical Monographs (2011)peak

Central Arabia is not a place a state governs. It is a place tribes move through — a plateau of pasture and oases where a confederation can be assembled and a population cannot be taxed, and where the significant question for any outside power is not who holds it but who can guarantee that a caravan crosses it.

That is what the Hujrids were for. Himyar, at the height of its power in the fifth century and interested in the routes north towards the Hejaz and the Roman frontier, installed and backed a family of the Kinda to exercise authority over the tribes of the Najd. The family traced itself to Hujr, called Akil al-Murar, and their position rested on Himyarite recognition, Himyarite money, and their own ability to hold a coalition together.

At its widest the Hujrid position extended over a great deal of central and northern Arabia and reached towards both imperial frontiers, and by the early sixth century Roman and Persian diplomacy had to take account of it. Roman embassies dealt with the Hujrids directly; a Kindite chief appears in the Byzantine sources negotiating over the tribes of the desert frontier, which is a thing Constantinople did not do with people it considered marginal.

Qaryat al-Faw, on the western edge of the Empty Quarter, is the archaeological site most associated with Kinda: a substantial town with markets, houses, temples, tombs and inscriptions, excavated since the 1970s and one of the richest pre-Islamic sites in the peninsula. Whether it was ever a Hujrid seat is another matter: nothing consulted says so, and the association is with Kinda in general rather than with this family.

What the Hujrids never acquired was anything that would outlast the subsidy and the man. There was no city they governed from, no administration, no revenue of their own. This is the standing condition of the three client dynasties and it is why all three ended the way they did: they were arrangements, and an arrangement can be cancelled.

Al-Hira

Drafted from Fisher, Oxford Classical Monographs (2011)contraction

Around 525 al-Harith ibn Amr, the Hujrid chief, took al-Hira.

Al-Hira was the seat of the Nasrids, the Arab dynasty that held the Euphrates frontier for Sasanian Iran — the eastern counterpart of the Jafnids, whose own chief would destroy the Nasrid king in battle in 554. It was a real town with churches and a court and a literary reputation, and it was Persia's instrument in Arabia.

A chief from central Arabia holding it, however briefly, is the most striking thing the Hujrids ever did. It is the only occasion on which one of the three client dynasties displaced another, and for a few years it made this family something more than the object of somebody's Arab policy: Constantinople and Ctesiphon both had to deal with them as a party rather than manage them as an asset.

It did not hold. Al-Mundhir of al-Hira recovered his capital, and in 528 he defeated al-Harith and killed him, along with — the accounts say — a large part of his family. The Persians backed the recovery, and the whole episode reads from Ctesiphon as a client being restored rather than a war being won.

The Hujrid position in central Arabia did not survive it. The confederation broke into its constituent tribal groupings, which spent the following decades fighting each other and were still doing so within living memory of the rise of Islam. The sons and grandsons of the house became claimants to something that no longer existed, and one of them wrote about it.

Caisus the fugitive

Drafted from Caesarea, History of the Wars, Books I-II: The Persian War (1914)contraction

Around 531 the emperor Justinian sent an ambassador named Julianus down the Red Sea with two requests, and the second of them is the only contemporary outside notice of what had become of the Hujrids.

The first request went to the Ethiopians: buy silk from India and sell it to the Romans, so that Roman money stops going to Persia. The second went to the king of Himyar, and it was this. Install Caisus as captain over the Maddeni, and then invade Persia with a great army of your own people and of the Maddene Saracens.

Maddeni is Ma'add — the tribes of central Arabia. Caisus is Qays. And the way Procopius describes him tells you the whole position of the family:

This Caisus was by birth of the captain's rank and an exceptionally able warrior, but he had killed one of the relatives of Esimiphaeus and was a fugitive in a land which is utterly destitute of human habitation.

A man of the ruling stratum, personally formidable, wanted for a killing at the Himyarite court, living in an empty desert — and being discussed by the Roman emperor as an asset that the Himyarite king might usefully reactivate.

What the request assumes

Two things, and both are informative.

It assumes that the leadership of the central Arabian tribes was in the gift of the king of Himyar. Not a matter of who the tribes followed, but of whom Zafar appointed. That is exactly the relationship the Hujrid position had always rested on, and here it is stated as a plain fact by an outside power making plans on the strength of it.

And it assumes the office was vacant, or vacant enough to be filled by proclamation. Nobody in Constantinople thought there was a functioning confederation in central Arabia with a leader who would have to be dealt with. They thought there was a job, and an exile who could be put back into it.

That is the condition of the house a few years after its chief was killed recovering al-Hira. Not destroyed, not conquered, and not there either: a qualification without a position, held by a man who could not go home.

Nothing happened

Neither king did what he had promised.

The Ethiopians could not buy Indian silk, because Persian merchants sat at the harbours where the Indian ships first came in — they lived next door — and bought whole cargoes before anyone else could bid.

And the Himyarites found that crossing the desert to reach Persia was harder than it had sounded in Constantinople: a country empty of people, extending so far that the march would take a very long time, at the end of which they would arrive among men far more warlike than themselves.

Later, Procopius adds, once Abraha had made his own position in Himyar completely secure, he repeatedly promised Justinian that he would invade Persia — and started once, and turned straight back.

So a Roman emperor spent years planning a war in which Arab client tribes under a restored Hujrid would open a second front against Persia, and the plan died of distance. The dynasty's last appearance in the record of the empires is as an instrument that was considered, agreed to, and never picked up.

Three generations of ambassadors

Drafted from Constantinople, The Library of Photius, Volume I (1920)contraction

The Byzantine empire employed the same family as its ambassadors to the Arabs for three generations, and the last of them wrote a book about it. The book is lost; a ninth-century patriarch's summary of it survives, and it is the fullest outside account of what the Kindite chieftaincy was worth to an empire.

The line runs like this. Nonnosus's grandfather was sent to Arethas — al-Harith the Kindite — under the emperor Anastasius, to conclude a treaty of peace. Nonnosus's father Abramos was sent under Justin to al-Mundhir of Hira, and got two captured Roman generals released. Then Justinian sent Abramos to Qays the Kindite, and a treaty was concluded on condition that Qays's son Mu'awiya be taken to Constantinople as a hostage. Then Nonnosus himself was sent to the same Qays.

Photius identifies him in a way the Arabic tradition would recognise: Qays was chief of two of the most illustrious Saracen tribes, the Kindites and the Ma'add. That is the Hujrid position stated from outside — not a kingdom, but the leadership of two confederations, held by one man and negotiable with.

Nonnosus's instructions were threefold: to persuade Qays to come to the emperor in person, and then to go on to Kaleb king of Aksum and to the Himyarites. He got through treacherous tribesmen, wild animals and a great deal of difficulty, and came home safely.

What Qays did is the end of the story, and Photius gives it in two sentences. After Abramos had been sent to him a second time, Qays set out for Constantinople, having first divided his chieftaincy between his brothers Amr and Yazid. He took a large number of his people with him. And the emperor appointed him administrator of Palestine.

A man who led the tribes of central Arabia partitioned his following among his brothers, emigrated with a substantial part of it, and accepted a provincial office inside the Roman empire. Whatever the house had been, that is what it was worth by then: a career, transferable, and better exercised somewhere else.

Photius keeps one other thing from the lost book, and it is the most valuable observation in it about the society the Hujrids led. Most of the Saracens, says Nonnosus, have a sacred meeting-place consecrated to one of the gods, where they assemble twice a year. One assembly lasts a whole month, ending about the middle of spring when the sun enters Taurus; the other lasts two months and is held after the summer solstice. During these meetings complete peace prevails, not only among themselves but with everyone else in the country — and, he adds, even the animals are at peace with each other and with people.

That is a description of the Arabian sacred months, recorded by a Greek envoy about a century before Islam, as an institution a diplomat needed to know about: there are periods in which nobody in Arabia is at war, they are fixed astronomically, and they belong to a shrine.

Photius then notes that other strange and more or less fabulous information is also given, and declines to repeat it — which is how most of Nonnosus was lost.

What the clients were for

Drafted from Fisher, Oxford Classical Monographs (2011)end

By the middle of the sixth century the Hujrid confederation had dissolved. Nobody conquered it, because by then there was nothing to conquer: it came apart.

That is the third of three endings and they are worth setting beside each other, because between them they are the state of Arab politics on the eve of Islam.

The Hujrids ended in the 530s and 540s, broken by a Persian client and then by their own tribes, with Himyar — the power that had backed them — collapsing at the same time under an Aksumite invasion. The Nasrids of al-Hira ended in about 602, when the Sasanian king executed the last of them and put a Persian governor in his place. The Jafnids ended in 582, when the Roman emperor arrested the supreme phylarch and divided his following among fifteen chiefs.

Within about seventy years, all three empires dismantled their own Arab frontier arrangements, for reasons that were sound in each case and unrelated to each other. Rome's client had acquired an independent religious constituency. Persia's had become unreliable. Himyar's patron ceased to exist.

Then, in the 630s, armies came out of Arabia and took the Sasanian empire entirely and the Roman east permanently.

The connection between those two facts is one of the oldest arguments in the field and it will not be settled here. The caution is worth stating: the Roman east in the 630s had just survived twenty-six years of Persian war and plague, there is more than one thing wrong with it, and an argument that runs from a decision in 582 to a battle in 636 has to explain fifty-four years.

What can be said is narrower and is not in dispute. Three empires spent two centuries building Arab political authority on their frontiers — courts, subsidies, titles, churches, a poetry — and then ended all of it within a lifetime. The people those arrangements had organised did not stop existing, and the next thing they organised themselves into was not somebody's client.

The prince who went looking for a kingdom

Drafted from Fisher, Oxford Classical Monographs (2011)aside?What aside means

Imru' al-Qays is the first poet of the Arabic canon. His mu'allaqa — one of the seven odes tradition says were hung at the Ka'ba — opens with the line every Arabic-speaking schoolchild knows, about halting at an abandoned campsite to weep over the remembered beloved, and the convention it established governed Arabic poetry for a thousand years afterwards.

He was a Hujrid. His father Hujr ibn al-Harith was the last Kindite king over the Banu Asad, who killed him; and the tradition has Imru' al-Qays receiving the news at a drinking party and saying that today is for wine and tomorrow for serious matters, and then spending the rest of his life trying to recover his father's kingdom. He went to the tribes, and to the Jafnids, and finally to Constantinople to ask Justinian for an army — and died on the way home, in Anatolia, of a poisoned robe in the legend and of something else in fact.

Almost none of that can be verified. The biography comes from Arabic sources written three centuries later, in a tradition that assembled poets' lives out of their own verses, and Fisher's book is careful about the difference between what the pre-Islamic poetry can be used to establish and what it cannot. What is reasonably secure is that a Kindite prince of that name existed, that the dynasty had lost its position, and that the poetry which became the foundation of Arabic literature was composed in and about that world.

Political weight and cultural output come apart constantly, and the Hujrid case is about as extreme as the separation gets. A client dynasty that held nothing, built nothing, lasted perhaps eighty years and left no institution is remembered because the man who lost it wrote well about having lost it.

What it transmitted was not a religion, a script or a claim that anyone inherited. It was a way of beginning a poem — the halt at the abandoned campsite, the weeping over the traces — and every Arabic poet for the next thousand years began that way, including the ones writing in Cordoba and Isfahan who had never seen a desert.

The men who wrote it down

Drafted from Lyall, Translations of Ancient Arabian Poetry, Chiefly Pre-Islamic (1885)afterlife

Everything the Hujrids left is poetry, and the poetry did not survive by being written down at the time. It survived by being memorised, professionally, for two hundred years.

Pre-Islamic Arabia had an institution for this. Every professed poet had a rawi — a reciter — to whom he gave his poems as he composed them and who transmitted them onward. Many rawis were poets themselves, and many poets had served as somebody else's rawi first. Imru' al-Qays, the Hujrid prince, was the rawi of a poet of al-Hira before he was anything else.

That is the whole apparatus. No archive, no library, no manuscript: a chain of named men, each holding a body of verse in his head and passing it to the next.

The collectors

The chain reaches paper in the eighth century, in Iraq, under the Umayyads and the early Abbasids, when scholars began hunting down what survived of the old poetry. Four names matter, and the two that matter most for this dynasty are uncomfortable ones.

Hammad al-Rawiya — a man of Persian descent, son of a Shapur — is the source of the first redaction of Imru' al-Qays's poems, and it was Hammad who assembled the selection that became the seven Mu'allaqat. Everything the canon holds about the Hujrid prince passed through him.

Hammad is suspected of having put forward compositions of his own as ancient poems. His contemporary Khalaf al-Ahmar, another foreigner of enormous learning, is known to have done it.

That is the transmission of the pre-Islamic canon stated honestly: a two-century oral chain, terminating in eighth-century Iraq with a small number of collectors of non-Arab origin, at least one of whom forged, and the most important of whom is under suspicion.

The suspension

The tradition that gave the poems their name says they were so admired that Arabs had them written out in gold on fine white Egyptian cloth and hung inside the Ka'ba, which is why the seven are also called the Golden.

It is a story about status rather than a report of an event, and it is doing something precise. The odes are being placed, retrospectively, in the one building that everybody afterwards agreed mattered — before Islam, in a shrine Islam kept.

What the dynasty came to

Put the two things together and the Hujrid afterlife is stranger than the ordinary case of a state remembered through its literature.

A client confederation in central Arabia held a position granted by a South Arabian kingdom for perhaps eighty years, lost it, and left nothing — no city, no institution, no successor, no inscription of its own. Its last prince's poetry became the foundation of the literary language of a civilisation that had not yet begun.

And it did so by being carried in the memories of professional reciters through the rise of Islam and the conquests, into the hands of Persian and Turkish scholars in Iraq who were compiling a canon of Arab antiquity for an empire that needed one — some of whom were writing their own contributions to it.

The Hujrids survive. What survives of them was selected, arranged and possibly in part invented by people from outside Arabia, two hundred years after the last of them died.