Kakuyid Dynasty

Son of the maternal uncle

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

The dynasty is named for a relationship traced through a woman, which is rare enough to be the first thing worth saying about it.

Kakuya is Daylami for the son of a maternal uncle. Rustam Dushmanziyar was uncle to Sayyida Shirin, the Buyid regent who governed the Jibal from Rayy for her son Majd al-Dawla for more than thirty years and is one of the few women this period's political narrative cannot avoid. His son Muhammad received Isfahan from her in about 1008 on the strength of that connection, took the title Ala al-Dawla in the Buyid manner, and then did what governors in this region do: he stopped being one.

By the 1020s he held Isfahan and Hamadan in his own right, minted, and was negotiating with the Buyids as a power rather than as a relative. The same movement had already been made in this generation out of a Tahirid governorship, a Samanid guard command, a Saffarid ayyar band, a Ghaznavid slave captaincy, a Seljuk pasture grant and a Khwarazmian washing-bowl appointment. The Kakuyid version is the smallest and the most domestic: a nephew given a city by his cousin's mother.

The position was never strong. Isfahan sat between the Buyids of Fars, the Buyids of Iraq, and — from the 1020s — Mahmud of Ghazna, who took Rayy in 1029 and ended Buyid rule in the Jibal, and whose son Mas'ud took Isfahan itself for a period in the 1030s. Ala al-Dawla lost the city more than once and got it back each time, which is the actual skill the sources credit him with: an ability to survive being conquered.

He was good at losing, and he used the intervals to hire a philosopher.

Two kings on one square

Drafted from others, The History of India, as Told by its Own Historians: The Muhammadan Period, Volume IV (1872)expansion

In 1029 Ala al-Dawla watched what happened to the man whose family had given him Isfahan, and appears to have drawn the correct conclusion from it.

Mahmud of Ghazna came west that year to take Jibal. On the way, Majd al-Dawla — the Buyid at Rayy, the son of the regent Sayyida Shirin, and the cousin through whom the Kakuyids held anything at all — sent a messenger to Ghazna complaining about his own commanders and asking for help against them. Mahmud sent troops towards Rayy. Majd al-Dawla joined them, and they arrested him.

Khwandamir preserves the interview. Mahmud had the prisoner brought into his assembly and asked him whether he had read the Shahnama and whether he knew Tabari's history. Yes. Had he ever played chess? Yes. Then, said Mahmud: in those books, have you ever seen it written that two kings can rule in one country — and on the board, have you ever seen two kings on one square? No, said Majd al-Dawla. Then how does it come about, said Mahmud, that you handed the reins of your power to a man stronger than yourself? He sent the Buyid and his son to Ghazna under guard and gave the province to his own son Mas'ud.

Two years later Mahmud was dead, Mas'ud had blinded his brother and taken the throne, and one of the first acts of the new reign was a farman handing the governorship of Isfahan to Ala al-Dawla ibn Kakuya. Khwandamir puts the transaction in a single dry sentence and then explains the family name — in the language of Daylam a maternal uncle is called kakuya — and adds the whole of the dynasty's political history in a second sentence: this Kakuya was at first the viceroy of Iraq, and afterwards he arrogated to himself absolute dominion.

The arrogating was done by watching where the sultan was. Mas'ud spent his reign campaigning in India, which Khwandamir treats as the error that lost him everything else, and while he was on the Indus, Ala al-Dawla rebelled and drove the Ghaznavid governor of Rayy out of the city. Mas'ud came back in 1036 or 1037 to find the Seljuks established in Khurasan and his Iranian provinces gone, and by Khwandamir's account regretted the Indian expedition for the rest of his short life. A poet sent him verses on the subject: your enemies were ants and have grown into serpents; kill the ants that have become serpents, give them no rest and delay no longer, for if they are given time the serpents will become dragons.

The verses were about the Seljuks. They describe the Kakuyids exactly as well. Ala al-Dawla never won a battle that mattered and never held anything for long against anybody who came for it in force. What he did instead, for thirty years, was accept the governorship from whoever was strongest, wait for that man to be somewhere else, and take the city back — which is why the answer to Mahmud's question turns out to be that two kings can share a square perfectly well, provided one of them is prepared to be the smaller piece for as long as necessary.

The only settled years Ibn Sina had

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

Ibn Sina had refused to go to Ghazna when Mahmud demanded the scholars of Khwarazm, left before the army arrived, and spent the following years moving between the courts of western Iran — writing while employed as a physician and an administrator, imprisoned once, in hiding more than once, and never anywhere long.

In about 1024 he reached Isfahan, and Ala al-Dawla kept him until he died.

He was made a member of the court and served as physician and adviser, and by several accounts as vizier; he accompanied the ruler on campaign, which is how he came to die at Hamadan in 1037 on one of them. What the arrangement bought, in a working life that had been interrupted about every three years, was roughly thirteen consecutive ones.

The output is the reason a small principality in the Jibal is remembered at all. The Canon of Medicine was substantially finished here, and it was taught in European medical faculties into the seventeenth century. So was much of the Book of Healing, the philosophical and scientific summa that carried Aristotle, in Avicennan form, into Latin scholasticism.

And there is the Danishnama-yi Ala'i — the Book of Knowledge for Ala al-Dawla — which he wrote for his patron, and wrote in Persian. That last fact is the one to hold on to. Ibn Sina's work is otherwise in Arabic, as essentially all serious philosophy and science of that world was; the Persian revival at Bukhara and before it had been poetry and history, not logic and metaphysics. The Danishnama is one of the first attempts to do philosophy in Persian, and it exists because a Daylamite prince in Isfahan wanted to read it and the terminology had to be invented to let him.

Nobody has ever published a figure for how large the Kakuyid state was or how many people lived in it, and it does not matter much. The most influential book in the history of medicine was finished under this dynasty, and the Persian philosophical vocabulary was assembled here to please one reader.

What they built at Yazd

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

In 1036 or 1037, two military governors serving Ala al-Dawla built a domed tomb-shrine at Yazd, on the edge of the desert, and it is still standing. It is called the Davazdah Imam, the Twelve Imams, and it is the oldest dated building in the city.

Its importance is structural and it is a genuinely large point in the history of building. The problem it solves is the ancient one of putting a round dome on a square room. The Iranian answer was the squinch — an arch thrown across each corner to turn the square into an octagon — and what the Davazdah Imam contains is one of the earliest surviving dated examples of a squinch broken into a trefoil: three small arched cells instead of one plain arch. Multiply that subdivision, and repeat it upward in tiers, and you have muqarnas, the honeycomb vaulting that covers half the significant architecture of the Islamic world for the next eight hundred years — Seljuk portals, the Alhambra, the Safavid mosques of Isfahan.

The building is small, in a provincial town, put up by two officers of a minor dynasty in the year before their patron's philosopher died.

That last sentence is the point. The Kakuyids held Isfahan for forty-three years, ruled Yazd as somebody else's fief-holders for ninety more, and vanished at a battle fought for a Seljuk sultan against Khitans from north China. By any measure of power they are a footnote.

What they left is a settled decade that produced the Canon of Medicine, the first substantial philosophical writing in Persian, and a corner of a dome at Yazd on which the next eight centuries of Islamic architecture is built.

There is no way of ranking states that captures that, and there probably should not be. A squinch broken into three cells is not a quantity. It is a solution to a problem, arrived at in a provincial town by somebody whose name is not recorded, and copied for eight hundred years by people who never heard of the dynasty that paid for it.

Nine months outside the walls

Drafted from Abu'l-Faraj), The Chronography of Gregory Abu'l Faraj, the Son of Aaron, the Hebrew Physician, Commonly Known as Bar Hebraeus (1932)contraction

Bar Hebraeus tells the fall of Isfahan without mentioning the family that held it, which is the most accurate thing about his account.

Tughril Beg sat down in front of the city and stayed there. Bar Hebraeus gives nine months and puts it in a year he cross-dates two different ways and gets wrong both times, so the figure and not the date is what to keep. The people inside, worn down by a siege he describes as an imprisonment, did not appeal to their own ruler or to any neighbouring power. They sent to Baghdad and asked the caliph to intercede with the sultan on their behalf.

The caliph refused. Tughril had been asking him for some time to be honoured with titles befitting his kingship, and al-Qa'im had not been willing. The Isfahanis had now given him something he could exchange for it. He agreed, wrote to Tughril, addressed him as Lawful King, Asylum of the Muslims, Rukn al-Din Sultan Tughril Beg, and at the end of the letter asked him to be merciful to the people of Isfahan.

Tughril, Bar Hebraeus says, had been expecting exactly this. He accepted the caliph's intercession and sent back twenty thousand dinars to the caliphal treasury and two thousand for the officials who administered it. And from that time he began to have the figure of a bow drawn at the head of his seal with the titles written inside it — the tughra, and the clerk licensed to draw it the tughra'i.

Every party to that transaction got what it wanted except the one being fought over. The townspeople got relief; the caliph got twenty-two thousand dinars and a client who had been made to ask twice; the sultan got a name with al-Din in it, a seal design that outlived his dynasty by eight centuries, and the city. Nobody involved thought it necessary to consult the man who was supposed to be ruling Isfahan, and no account of the episode records what he did during the nine months.

That is what the end of a middle-sized principality looks like from outside. It is not conquered so much as handed over in the course of an unrelated negotiation between two larger parties, as the consideration in a bargain about titles.

The last one died at Qatwan

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

Isfahan fell in 1051 and became the Seljuk capital. Nizam al-Mulk would administer an empire from the city the Kakuyids had lost.

The family was not destroyed. It was moved sideways. Yazd and Abarkuh were granted to them as fiefs, and they held those for another ninety years as Seljuk governors, married into the Seljuk house, and built — mosques, canals, fortifications, and at Yazd a set of tomb towers of the kind Qabus had built at Gurgan. So it is not a conquest with a date on it: the polity stopped being independent in 1051 and the family stopped mattering in 1141, and there is no single date on which it was destroyed.

Garshasp II, the last of them, was killed in 1141 at the battle of the Qatwan steppe.

That is the battle of Qatwan: Sanjar's army destroyed by a Buddhist steppe empire outside Samarkand, Transoxiana passing under non-Muslim suzerainty, and news of it travelling west to become the Prester John expectation. A Daylamite prince from Isfahan, holding Yazd from a Turkish sultan, died in Transoxiana fighting Khitans from north China, and his dynasty ended there.

So the dynasty has two ending dates ninety years apart, and they answer different questions. The Kakuyid state — the thing that held Isfahan, defied the Buyids, survived Mahmud, and paid Ibn Sina — ended in 1051. The Kakuyid family ended in 1141, on somebody else's battlefield, in somebody else's war.

The books went to Ghazna

Drafted from Browne, A Literary History of Persia, Volume II: From Firdawsi to Sa'di (1906)afterlife

Ibn Sina died in his patron's service in 1037 and the Kakuyid state outlived him by fourteen years. What happened to his library is the better story, because it runs through three dynasties and ends in a fire.

Why he was there

Ibn al-Athir is blunt about the arrangement, and Browne repeats him without softening it: there is no doubt as to the unsoundness of Ala al-Dawla's religious views, and it was on that account that Avicenna attached himself to his court, so that he might be unmolested in the composition of his own heretical works.

That is a transaction, and both sides understood it. A minor Daylamite prince with a reputation for not caring what a man believed acquired the most famous mind of the age, who in return produced the Canon, the Danishnama in Persian for his patron to read, and a quantity of philosophy that would have been awkward almost anywhere else.

The price of a heterodox court is that it has no friends.

1034

In 1034 Ala al-Dawla was defeated by the troops of Ghazna.

They plundered Isfahan, and among the plunder they took Ibn Sina's books. The whole library was carried east and deposited in one of the libraries of Ghazna — the city of the sultan who had demanded him from Khwarazm twenty years earlier, had a portrait of him circulated when he ran, and never got him.

Mahmud got the books in the end. He had been dead four years.

And then Jahansuz

They stayed at Ghazna for a hundred and sixteen years.

In 1151 Ala al-Din Husayn of Ghur took the city and burned it for seven days, and was called Jahansuz, the World-burner, ever afterwards. Browne records what went with it: the books remained in that library until they were destroyed in the fire, in the sack of Ghazna by the King of Ghur, appropriately called the World-burner.

So the working library of Ibn Sina — the annotated copies, the drafts, whatever he had not published — was taken as loot from a small Iranian principality by a Turkish empire, held for over a century in an Afghan mountain town, and burned by a mountain chief from Ghur avenging his brothers.

What survived instead

Everything that survives of him survives because it had already been copied out and sent elsewhere.

The Canon of Medicine went to Baghdad and Cairo and eventually into Latin, and was taught in European faculties into the seventeenth century. The Danishnama survived as the first substantial philosophy in Persian. The correspondence with al-Biruni survived because two young men had argued about Aristotle in writing and other people kept the letters.

None of that was in the library. The library was the part that was his, and it is the part that is gone.

A dynasty that is a footnote in every political history of Iran paid for the conditions in which those books were written, and could not protect them for a single year after it lost a battle.