Bavand Dynasty

The mountains nobody wanted enough

Drafted from (ed.), The Cambridge History of Iran, Volume 4: From the Arab Invasion to the Saljuqs (1975)formation

The Elburz range runs east to west between the Iranian plateau and the Caspian, and the country on its northern face is unlike anywhere else in Iran: heavy rainfall, forest, rice, and valleys reachable through passes that a small number of determined people can hold. Every empire that ever governed Iran held the plateau. None of them securely held this.

The Bavands are what lived there. The family claimed descent from Kawus, a son of the Sasanian king Kavadh — a claim modern scholarship treats with caution, since it may compress several figures of the great Parthian house of Ispahbudhan into one convenient ancestor — and they were still asserting it seven hundred years later. They used the title ispahbadh, which is a Sasanian military office, and continued to use it four centuries after there was no Sasanian state to hold an office in.

They also stayed Zoroastrian for a long time. The Arab conquest took the Caspian plains in the seventh century and stopped at the mountains; Tabaristan kept its own princes, its own calendar and its own religion into the ninth. The Bavand ruler Qarin I accepted Islam around 842 — two centuries after the Arab armies had destroyed the Sasanian field army at Qadisiyya — and the dynasty carried on using the Sasanian title and the Sasanian descent claim regardless. By the tenth century the family was Twelver Shi'i, which in the Caspian was the ordinary confession and not a statement.

What they did politically for seven centuries was acknowledge whoever was strongest and remain in the mountains. The list is a table of contents for this region: Abbasid governors, then Tahirids, then Samanids, then the local Alid imams, then Ziyarids, then Buyids, then Ghaznavids, then Seljuks, then Khwarazmshahs, then Mongols. Each of them received recognition and, when convenient, tribute. None of them replaced the family, because taking the Elburz was expensive and holding it was worse, and the Bavands never gave anybody a sufficient reason.

Three lines, two gaps

Drafted from (ed.), The Cambridge History of Iran, Volume 4: From the Arab Invasion to the Saljuqs (1975)peak

The dynasty is conventionally divided into three, and the divisions are not a historian's tidiness — they are two interruptions during which the family did not rule.

The Kayusiyya held the mountains from the seventh century until 1006, when the Ziyarid ruler Qabus ibn Vushmgir ended them. That is the first gap: sixty-seven years in which Tabaristan was administered by other people.

The Ispahbadhiyya began in 1073, moved the seat down out of the highest country to Sari, and became something larger than their predecessors had been — holding Gilan, Rayy and Qumis as well as Tabaristan, as vassals first of the Seljuks and then of the Khwarazmshahs. Shah Ghazi Rustam, in the middle of the twelfth century, is the one the sources treat as the peak: he fought the Nizari Isma'ilis in the neighbouring valleys, held the Caspian provinces against Seljuk claimants, and struck his own coin. The line ended in 1210 with the murder of its last ispahbadh. That is the second gap.

The Kinkhwariyya appear in 1237, after the Mongols, and ruled Tabaristan as Mongol vassals for a further hundred and twelve years, until a local rival killed the last of the main line in 1349 and took the province.

Dynasties are usually ended once. The Bavands were ended in 1006 and came back; ended in 1210 and came back; ended in 1349 and did not.

What kept being available across those gaps was not a state. It was a claim — a family in known valleys, with a genealogy nobody local disputed, which could be put back together whenever the power on the plateau was distracted for long enough. That is a durable thing to be, and it is not the same thing as a durable state, which the chapter after next is about.

Shah Ghazi and the neighbours in the next valley

Drafted from (ed.), The Cambridge History of Iran, Volume 4: From the Arab Invasion to the Saljuqs (1975)peak

The second Bavand line reached its height under Shah Ghazi Rustam, who ruled from about 1142 to 1165 and is the one member of this dynasty the wider sources treat as a figure in his own right rather than as a name in a list of vassals.

He held the Caspian provinces against Seljuk claimants at the point when the Great Seljuk sultanate was coming apart into family possessions administered by atabegs, and he extended Bavand authority over Gilan and into Rayy and Qumis. He struck his own coin. The Seljuk sultan Mas'ud campaigned against him and did not finish him.

What he is remembered for is the Isma'ilis.

The Nizari Isma'ili state held Alamut and a network of fortresses in the Elburz — the same mountains, a few valleys away. The two were neighbours in the most literal sense available, and the relationship was murderous. Nizari agents killed Shah Ghazi's son. He responded with a campaign of extermination against the Isma'ili settlements he could reach, and the sources report him building a tower of their heads, which is the kind of detail medieval chroniclers report about a ruler they approve of.

That is worth dwelling on, because the Nizari state is normally described from a great distance — a network of scattered castles that no imperial army could ever finish, which is true. This is the same thing seen from the next valley, and it looks nothing like that. It looks like a neighbour whose method is to kill your son, against whom a local ruler with local knowledge could do considerably more damage than a sultan arriving from the plateau with an army.

Shah Ghazi died around 1165. His line lasted another forty-five years and ended in 1210 with a murder, and the province passed to the Khwarazmshahs — briefly, because the Mongols were nine years away.

The guest

Drafted from Isfandiyar, An Abridged Translation of the History of Tabaristan (1905)peak

A mountain dynasty in Tabaristan spent the late twelfth century being courted by larger powers on both sides of it, and Ibn Isfandiyar — who was writing in Tabaristan, for its ruling house — preserves two episodes that show what the courting was worth.

The feast on the plain

When Il-Arslan Khwarazmshah died in 1172 and his son Tekish took the throne from his brother Sultan Shah Mahmud, the losing brother and his mother came to Tabaristan for refuge.

The ispahbad Ala al-Dawla Hasan went out to Tammisha to meet them, sent his governors and officers from Gilan and the districts of Ray ahead with presents, and on the plain of Ganjina caused a feast to be spread as far as Ispid Daristan, for a distance of a parasang — such a feast, Ibn Isfandiyar says, as no man ever saw before or since.

Three miles of tables for a defeated claimant to somebody else's throne. It is a statement about the host rather than the guest: a ruler who can provision that much hospitality at short notice in a mountain province is telling the whole region what his revenues are.

The offer

The next ispahbad, Husam al-Dawla Ardashir, who reigned thirty-five years, was put to a harder test.

The Seljuk sultan Tughril, quarrelling with his brother Qizil Arslan, wrote from Iraq asking for help. Ardashir sent his amirs and nobles to Ray to meet the Seljuk king and went forward himself to Lar — where, on meeting him, he alighted from his horse, brought his royal guest to Fulul, and there enthroned and entertained him royally.

Qizil Arslan heard about it and sent an envoy to make Ardashir an offer. He was reminded of the favours he had received from Qizil Arslan's father and brother, and asked to arrest Tughril, bind him and hand him over.

The price offered for that was:

the rule of Ray, Sawa, Qum, Kashan and Qazwin, and full authority over Iraq and Adharbayjan.

That is not a bribe. It is an offer to multiply the Bavandid state several times over — the whole of the Jibal and the north-west, handed to a dynasty whose territory was a strip of forest and mountain between the Elburz and the Caspian.

Ardashir indignantly rejected the proposal that he should betray his guest. After a while he sent Tughril on to Damghan and Bistam, with instructions to his officers there to supply everything the sultan needed, day by day, until he reached his own capital.

The letter from India

The other episode is stranger and says more about how the dynasty was seen from outside than any campaign could.

In 1183-4 an ambassador arrived from the Maharaja of India. A Shi'i missionary, he reported, had come to that country arguing for his doctrine and had beaten their own scholars in debate. So the Maharaja had written for a ruling — and the address on the letter is the point. It was directed to

the just and upright king in Tabaristan, who is descended from the Kisras, and who holds this doctrine.

Descended from the Kisras — the house of Sasan. The Bavandids claimed descent from the pre-Islamic Persian kings and here is an Indian court using that claim, unprompted, as the reason to write to them rather than to Baghdad or Cairo.

The embassy had set out forty strong. Thirty-eight died on the road; two arrived.

Ardashir had the reply composed at length, in florid Arabic, by a sayyid of Amul, and it grew into a treatise with a title of its own — the Epistle to the Indians in reply to the pretensions of the obdurate.

What the dynasty actually had

No Bavandid ever ruled much. The whole point of this house is that it held one defensible province for six hundred and ninety-eight years while everything around it was conquered repeatedly.

These two episodes show what that bought. The ispahbads were rich enough to feed an army for entertainment, independent enough that a Seljuk sultan fleeing his brother thought Tabaristan a safe place to be, honourable enough that refusing to sell a guest was worth more to them than seven provinces — and old enough that a king in India believed the Sasanian line still held a throne in the mountains south of the Caspian, and wrote to it for a ruling on doctrine.

A province writes itself down

Drafted from (ed.), The Cambridge History of Iran, Volume 4: From the Arab Invasion to the Saljuqs (1975)aside?What aside means

Around 1210 a man named Ibn Isfandiyar, whose father had been a senior official of this dynasty and who had been at the Bavand court himself under Ardashir I, started collecting material for a history of Tabaristan. He finished it a few years later, in Persian, and it survives.

The Tarikh-i Tabaristan is a local history in the strict sense: not a dynasty's chronicle and not a universal history with a province in it, but an account of one region from its legendary origins through its Sasanian princes, its resistance to the Arab conquest, its conversion, its Alid and Zaydi imams and its ruling families, to the writer's own time. It preserves material found nowhere else, including a Persian translation of a letter attributed to a Sasanian-era Tabaristani ruler, and it is the principal source for most of what has been said here so far.

That is worth stating directly. Almost everything a reader can know about the Bavands comes through a book written by a courtier of the Bavands, at the end of their second line, about his own province.

There are three positions a dynasty can be in with respect to its own record, and this is the third and much the commonest.

The first is to have a friendly local source that nobody else can check: the Tarikh-i Sistan does that for the Saffarids, who are bandits in every other account and are not bandits in that one. The second is to have no source of your own at all, so that the whole account is left to your enemies, which is what happened to the Nizaris when their library at Alamut was destroyed.

Ibn Isfandiyar is neither. He is not hostile and he is not independent, and he is the only witness there is. Almost everything anybody can know about the Bavands comes through a book written by a courtier of the Bavands, at the end of their second line, about the province he had grown up in — including the descent from Kawus, and the seven-century span, and most of what the earlier chapters here assert.

That is not a reason to disbelieve him. It is a reason to say so.

What six hundred and ninety-eight years means

Drafted from (ed.), The Cambridge History of Iran, Volume 4: From the Arab Invasion to the Saljuqs (1975)aside?What aside means

The Bavandids ran from 651 to 1349, which is six hundred and ninety-eight years and makes them one of the longest-lived dynasties anywhere in the medieval Islamic world — longer than the Ottomans managed.

The figure encloses two periods, of sixty-seven and twenty-seven years, in which the family ruled nothing at all.

Whether that counts

Historians answer this kind of question two different ways, and the difference is not carelessness on anybody's part.

The Zaydi imams of Yemen are conventionally given a single span from 897 to 1962, a thousand years with substantial interruptions inside it, because Yemeni historiography treats the imamate as one institution that was sometimes in abeyance. The Ibadi imamate of Oman is counted the other way: the same office was reconstituted in the same mountains repeatedly across a millennium, and Omani historiography numbers those imamates in the plural, as separate constitutions rather than one thing with gaps.

The Bavandids belong with the Yemeni pattern, and for a reason about the sources rather than about tidiness. The three Bavand lines are one family with a continuously asserted descent, named collectively as a single dynasty by the works that record them. Three lines of one house are not three polities in the way that two Omani imamates constituted by two separate elections are.

What the number cannot say

Six hundred and ninety-eight years, stated flatly, implies a continuity that a reader should know was interrupted twice. Any single figure for how long something lasted assumes the thing had one beginning and one end, and a family that loses its mountains and comes back is not that shape.

It is the second measurement on these pages that strains in this way. Territorial extent cannot describe Srivijaya, which held a strait rather than an area, or the Nizaris, who held scattered castles and nothing between them. Duration cannot describe a polity that stops and starts. Both measures were built for states with a single life, and the better course is to give the number with the argument attached than to leave out the polities that break it.

The bath at Amul

Drafted from Isfandiyar, An Abridged Translation of the History of Tabaristan (1905)afterlife

Ibn Isfandiyar stopped writing in the early thirteenth century. Somebody else picked the chronicle up and carried it to 1349, and the last pages of that continuation are the only close account of how the dynasty ended.

A haircut

Start with the detail that has nothing to do with the ending and explains the period better than anything that does.

In 1343 the Sarbadar amir Mas'ud came out of Khurasan to take Mazandaran. The country united against him — the chronicler has an old noble, Kiya Jamal al-Din Ahmad Jal, sending word to his kinsmen to defend the honour of Mazandaran and not to trouble themselves about what happened to him, since he was an old man who had lived his life. Mas'ud was driven out and killed, and his head was cut off, stuffed with straw, and displayed on the spot where he had lately camped with a great army.

Then this. The people of Tabaristan, except the learned and the religious, wore their hair long and loose, sometimes a yard in length. A few days after the execution, Malik Jalal al-Dawla noticed that his dead enemy had shaved his head, and adopted the practice himself, along with the turban. His kinsmen copied him; then everybody else did. The long locks called kalalak and the head-dress called kala-band were abandoned, and that was the end of them.

A province gave up its national hairstyle because a king liked the look of a man whose head was on a pole.

The bath

The end of the dynasty is six years later and is a small, domestic murder.

Mazandaran had come apart into factions — the kiyas of Chalav on one side, the kiyas of Jalal on the other, and the ustandar of Rustamdar holding the mountains behind them. Into that, the chronicler says, came the dervish orders, the pirs and the murids, which had already appeared in Khurasan. Kiya Afrasiyab of Chalav and his kinsmen became enthusiasts for the new devotion and tried to bring the king to it. Their saint and spiritual director, Amir Sayyid Qiwam al-Din, disliked Malik Fakhr al-Dawla for what the text calls his real or supposed heretical tendencies.

On 17 April 1349 Fakhr al-Dawla was murdered in the bath at Amul by Ali Kiya son of Afrasiyab and his brother Muhammad Kiya.

Seven centuries of holding the Elburz against Arabs, Samanids, Ziyarids, Buyids, Ghaznavids, Seljuks, Khwarazmshahs and Mongols, and the thing that finished the house was a religious quarrel among its own nobles, conducted with a knife, in a bathhouse, in its own capital.

The last attempt

He left four young sons, the eldest ten years old, and they had nowhere to go except to Malik Jalal al-Dawla Iskandar, the ustandar of Rustamdar — the same man who had taken up the shaved head. He sheltered them, brought them up like princes of his own house, and when they were grown undertook several campaigns to put them back.

On one of them he attacked Kiya Afrasiyab and the sayyid at Maran-Dih near Amul and killed Afrasiyab. Then he was outnumbered by the Mazandaranis and driven back, and three hundred and thirty of his men of Rustamdar were killed.

That, the chronicle says, was the last attempt of the kings of Bavand to assert their power.

Seven hundred and five years

The continuator does the arithmetic himself and gets a different answer from the one usually given. He dates Baw's independence to AH 45, which is AD 665 or 666, and reckons the dynasty at seven hundred and five years in Mazandaran — because he is counting from a slightly later founding and to the murder rather than to anything else.

Six hundred and ninety-eight, seven hundred and five: the disagreement is the kind that arises whenever anyone tries to put a single number on something that lasted this long and stopped twice on the way. Neither figure is wrong. Both are answers to slightly different questions, and a house that could be reconstituted after sixty-seven years of not existing is not really the sort of thing that has a length.

What is not in doubt is the last line of the chronicle, which is that the house of Bavand had endured in Mazandaran since a Sasanian noble made himself independent there, and that it ended with a man being killed in a bath.