Shirvanshahs

Arabs who became Sasanians

Drafted from Minorsky, A History of Sharvan and Darband in the 10th-11th Centuries (1958)formation

Yazid ibn Mazyad al-Shaybani was an Abbasid general of the Banu Shayban, sent to govern the Caucasus frontier around 799 and to hold the pass at Derbend against the Khazars. His descendants held Shirvan afterwards, and by 861 — with the caliphate in the middle of its collapse at Samarra, where four caliphs were killed by their own Turkish guard in nine years — they had stopped behaving like governors and taken the old local title of Shirvanshah.

What happened to them over the following two centuries is the interesting part and it is what Minorsky's book is for. An Arab tribal family in a Persianate country, ruling a population that was neither Arab nor entirely Persian, wrote itself into the landscape it had been sent to hold. The names go first: by the eleventh century the shahs are called Manuchihr, Kavus, Farrukhzad, Faridun — Sasanian and Shahnama names, in a family whose founder was a Shaybani Arab. Then the genealogy: descent from Bahram Gur, and from Khusraw Anushirvan, asserted publicly and carried in the titulature.

Almost nobody believes it. The Arab descent is well attested and the Sasanian one is a construction — and it is worth being careful about what that does and does not settle. Whether the family was descended from Bahram Gur is a question about genealogy, and the answer is no. What ground they claimed the right to rule on is a different question, and the answer to that one is not in doubt at all: they asserted this descent publicly, in their titles, for half a millennium, and they had every opportunity to choose something else and never did.

Minorsky's history exists because he reconstructed the local chronicle of Shirvan and Derbend out of a later compilation. Without it this dynasty would be a column of names and dates in a genealogical manual — which is worth saying on a page about a polity whose distinction is that it lasted, because lasting is precisely the thing that leaves no dramatic record.

Eating the mountains

Drafted from al-Masudi, El-Mas'udi's Historical Encyclopaedia, entitled 'Meadows of Gold and Mines of Gems', Volume I (1841)expansion

Al-Mas'udi passed along the Caspian in the 940s and wrote down who owned what in the eastern Caucasus, and his list is the only contemporary account there is of how the Shirvanshahs got large.

The country he is describing is not a country. The Caucasus, he says, holds seventy-two nations, every one of them with its own king and its own language, and there are passes and valleys in it inhabited by peoples who have no acquaintance with each other at all, because the communication between them is impeded by the height and roughness of the mountains, by marshes and forests, by the water coming down off the summits and by immense rocks. Anushirvan had dealt with that by appointing a chief to each of them, and the titles had stuck: a Layzan-shah at Layzan, a Khursan-shah, a Tabarsaran-shah, and so on down the range, each a hereditary king of one valley.

Mas'udi's Shirvanshah is Muhammad ibn Yazid, and the first thing he records about him is what the family had been a generation earlier: he and his father held Layzan, and no other kingdom. What they had besides Layzan was a claim — descent from Bahram Gur, which Mas'udi reports without objection and adds that nobody who knows genealogy will contradict it.

Then the list of what Muhammad ibn Yazid had eaten. Layzan, taken recently. The Muqaniyya, subjugated. Al-Lakz, whose king was his vassal, in a kingdom whose mountain population Mas'udi calls innumerable. And, in summary: this man has subjected many kingdoms in these mountains which had been given to various chiefs by Anushirvan, and they are now all under his sceptre — among them Khursan Shah and Rawan Shah.

He had also taken the Gate. The city of Bab al-Abwab, Mas'udi notes, had been governed since the first Arab conquests by the descendants of a man of the Ansar, and on the death of the last of them it passed into the Shirvanshah's possession.

Two things in the account are worth holding on to. The first is the method, which is not campaigning. It is one valley lordship absorbing the next over two or three generations by inheritance, marriage, vassalage and the occasional short war, in country where no outside power could reach far enough or fast enough to object — the Abbasids having stopped being able to, and nobody having replaced them yet.

The second is what the method could not do, which Mas'udi records in the same breath. Among the peoples of al-Lakz are the Dudaniyya, who are pagans, who have no king at all, and who do not acknowledge the Shirvanshah as their master. A polity assembled out of mountain valleys is exactly as large as the number of valleys that have agreed, and the ones that have not agreed sit inside it, unreachable, for as long as anybody cares to look.

How to last seven hundred years

Drafted from Minorsky, A History of Sharvan and Darband in the 10th-11th Centuries (1958)aside?What aside means

The Shirvanshahs ran from 799 to 1538. Almost nothing else in the medieval Islamic world lasted that long, and the ones that did were empires.

Ask how they governed, how they raised troops, or how the throne passed, and the answer is that it depends entirely on which century you mean. Across those seven hundred years this polity was an Abbasid governorship, an independent principality, a Seljuk vassal, a tributary of the Ilkhanate, a Timurid client and a Safavid dependency. The way it raised troops in the tenth century against Rus raiders coming down the Volga has nothing in common with the way it did in the fifteenth. There is no arrangement that held for most of it.

The method

What can be described is not an institution but a technique, and it is consistent even when nothing else is.

Submit early, before being attacked. Pay. Be useful on the frontier to whoever is currently the power — against the Khazars for the caliphate, against the Georgians and the Kipchaks for the Seljuks, against the Golden Horde for Timur. Hold the mountains and the passes, which are expensive to take. Be cheap enough to leave alone that nobody bothers. Do not fight for independence, and do not lose the country.

It is the opposite of everything a dynasty is usually praised for, and it worked for longer than any of the empires it submitted to.

Others who tried it

The Artuqids acknowledged nine sovereignties in two hundred and seventy-five years, and recorded them on a rock at Mardin. The emperors of Trebizond married their daughters to whoever might protect them, and outlived Constantinople. The Bavandids in the Caspian mountains persisted through six centuries of everybody.

What these have in common, besides longevity, is that they are hard to say much about. A polity that survives by never committing to anything leaves behind no characteristic institution, no reforming ruler, and very little that a historian can hold up and describe. The Shirvanshahs are the most successful practitioners of the method and the least visible dynasty in the region — and those two facts are the same fact.

The twelfth-century court

Drafted from Minorsky, A History of Sharvan and Darband in the 10th-11th Centuries (1958)peak

Shirvan in the twelfth century, under Manuchihr III and his son Akhsitan I, was one of the significant literary courts of the Persian-speaking world, which is not what a reader would expect of a small principality on the Caspian.

Khaqani Shirvani was its poet — born at Shamakhi, employed at the court, imprisoned by it at least once, and one of the most technically difficult poets in Persian, whose qasidas are still set as the hard case in the tradition. His Tuhfat al-Iraqayn is among the earliest Persian verse travel accounts, and the poem he wrote on the ruins of Ctesiphon — the arch of the Sasanian palace, seen from the road — is one of the best-known meditations on fallen empire in the language. It is worth pausing on which court paid for that poem: a dynasty that claimed the Sasanians as ancestors, employing a poet who wrote about their ruins.

Falaki Shirvani and Mujir al-Din Baylaqani were of the same court. And Nizami Ganjavi — the poet of the Khamsa, working just to the west at Ganja — dedicated Layli and Majnun to Akhsitan I in 1188, at the shah's request and, by his own account in the poem's introduction, with some reluctance about the subject.

Political weight and cultural output come apart constantly, and the Caucasus and upper Mesopotamia are full of examples — the Artuqids, the Marwanids, Trebizond, the Kara Koyunlu. What distinguishes Shirvan is duration. This was not one generous ruler in one generation. It was a court doing this while its neighbours were being conquered, and still there afterwards.

An earthquake destroyed Shamakhi in 1191 and the seat moved to Baku for a time. Then the Mongols came, and Shirvan submitted, and continued.

Baku

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

Ibrahim I took the throne in 1382, founded the line historians call the Derbendids, and did the single most consequential thing any Shirvanshah ever did: when Timur came into the Caucasus he went to him in person, submitted before being attacked, and offered service.

He was confirmed in his own country and used as an ally against the Golden Horde. Shirvan came through the Timurid period intact — no siege, no deportation of craftsmen, no towers of skulls — while Isfahan, Baghdad, Delhi, Damascus and Sarai were being destroyed. Submitting before being attacked is the technique the dynasty had been practising for six hundred years, and this is the occasion on which it paid the largest dividend.

What followed is the peak. Khalilullah I ruled from 1417 to 1465 and Farrukh Yasar from 1465 to 1500, and between them they made Baku a capital and built the complex that is still the thing visitors to the city go to see: the palace of the Shirvanshahs, with its divankhana, its mosque and minaret, its mausoleum and its bath, cut in the pale local limestone on the hill inside the walled town.

The money came from three things and no source ranks them. The land tax of a fertile province. The tolls of the Caspian coast road and the Derbend pass, which is the only practicable land route between the steppe and Iran. And extraction: Shirvan had the silk of Shamakhi, which Genoese, Venetian and later Muscovite merchants came for, and it had the Absheron peninsula, with its salt pans and its naphtha seeps and hand-dug oil wells.

A medieval principality drawing revenue from petroleum is the kind of detail that reads as an anachronism and is not one. The oil was collected, sold, burnt for light and heat, used in war, and taxed, four hundred years before anyone drilled for it.

The debt Isma'il collected

Drafted from Newman, Safavid Iran: Rebirth of a Persian Empire (2006)end

The Safavid order's move from Ardabil into politics went through Shirvan twice, and both times it went badly.

Shaykh Junayd led his Turkmen followers north on a raid in 1460 and was killed in battle against the Shirvanshah Khalilullah I. Twenty-eight years later his son Shaykh Haydar did the same thing, and was killed in battle against Khalilullah's successor Farrukh Yasar, who had Ak Koyunlu help. Two generations of the Safavid house died fighting the same dynasty in the same province.

Haydar's son Isma'il came out of hiding in 1499 with the Qizilbash behind him, and the first campaign he fought — before Tabriz, before the proclamation of Shi'ism, before any of the things the Safavid empire is remembered for — was against Shirvan. In 1500 he defeated Farrukh Yasar and killed him. The accounts of what was done to the Shirvanshah's body, and to the tombs of the dynasty at Baku, are not pleasant and are not in dispute.

Seven centuries of holding the corridor by submitting early to whoever was coming ended against an opponent for whom the country was not a corridor. It was a debt.

The dynasty continued for another thirty-eight years as Safavid vassals, and in 1538 Shah Tahmasp deposed the last of them and made Shirvan a province governed from Tabriz. Baku's walls and palace stayed where they were; the province was fought over by the Safavids and the Ottomans for the next century and a half, lost to Russia at Gulistan in 1813 along with the rest of the eastern Caucasus, and is the capital of a state today.

The ground on fire

Drafted from Strange, The Lands of the Eastern Caliphate: Mesopotamia, Persia, and Central Asia, from the Moslem Conquest to the Time of Timur (1905)afterlife

The dynasty ended in 1538 and the province was governed from Tabriz, then fought over by Safavids and Ottomans, then taken by Russia in 1813. What happened to Shirvan afterwards is not really a political story at all. It is about the thing its own geographers kept mentioning in passing.

What Istakhri thought worth a sentence

The Arabic geographers describe Baku as a port south of Derbend with naphtha springs. Istakhri refers to them as well known. Yaqut and others go into detail: the produce was worth a thousand dirhams a day, the naphtha flowed continuously, and all the ground was on fire round about.

Mustawfi, later, mentions the castle of Bakuyah standing so high above the town that it kept the town in shadow at midday.

That is the entire medieval notice of the place: a fortress, a shadow, and ground that burned. It is filed between a bridge of boats over the Samur and a district near the mouth of the Kura where they grew corn and cotton.

A thousand dirhams a day

Take the figure seriously for a moment, because the geographers did.

A thousand dirhams a day is a revenue, from a commodity, being collected and counted in the tenth century. The oil was gathered from seeps and hand-dug pits, carried away, sold, burnt for light and for heat, used in war, and taxed. The Shirvanshahs drew part of their income from petroleum for five hundred years, and nobody at the time thought it remarkable enough to write more than a sentence about.

What made Shirvan wealthy, in every account of it, was the silk of Shamakhi and the tolls of the coast road. The naphtha is a footnote about a curiosity: fields that were on fire, which travellers went to look at.

The footnote takes over

In the nineteenth century somebody drilled. The fields of the Absheron peninsula became one of the great oil regions of the world within a generation, and the Baku most people have seen — the boulevard, the opera house, the mansions of the oil millionaires — was built with the money.

The palace of the Shirvanshahs is still inside the walled town on the hill, in its pale limestone, with the divankhana and the mosque and the mausoleum. It is what visitors go to see. It was paid for by silk and tolls and a fertile province, by a dynasty that had lasted seven hundred years by never fighting anybody it could avoid fighting.

Underneath it was the asset that would eventually make the country a state, and for the whole of that seven hundred years it was a line in a geography book about ground that would not stop burning.