Ibn Khaldun on why dynasties die in four generations

A fortress in the Algerian hills, 1377

Ibn Khaldun spent twenty years drafting letters for sultans in Tunis, Fez, Granada and Bougie. Several of them lost their thrones while he was on the payroll. In 1364 Muhammad V of Granada sent him over to Seville as an envoy to Pedro of Castile. Then he quit. In 1375 he withdrew to Qalʿat Ibn Salāma, a fortress in the hills of what is now Tiaret province in Algeria, and stayed about four years. The first draft, he says, took him five months. What came out is Al-Muqaddimah, the long prologue to his universal history, the Kitāb al-ʿIbar, finished in 1377, which is 779 in the Islamic calendar.

The Maghrib he worked in was a machine for making and unmaking dynasties. The Almohad empire came apart in the thirteenth century. It left the Hafsids at Tunis, the Zayyanids at Tlemcen and the Marinids at Fez, each founded by a tribal confederation, each rotting from the middle outward. Plague took both his parents in 1348, in Tunis, the city where he was born on 27 May 1332. He was sixteen. Why does the same shape keep coming back, and roughly on schedule? For a chancery secretary, that is an odd thing to want to know.

"Plato, Aristotle and Augustine were not his peers" (Robert Flint, 1893)

ʿIlm al-ʿumrān, and the solidarity that makes states

Al-Muqaddimah opens with a claim of originality so flat that it's easy to miss. Ibn Khaldun announces a new science, ʿilm al-ʿumrān, the science of civilisation, or of human social organisation, and says nobody has written it before him. Its subject is human association itself. He splits that subject in two: ʿumrān badawī, the organisation of rough rural and pastoral life, and ʿumrān ḥaḍarī, the organisation of settled cities. Both, he insists, are natural objects with regularities you can state in advance. Events are the surface. Without such a science a historian can't sort a possible report from an impossible one. That, he says, is how al-Masʿūdī came to hand Moses an army of 600,000 men in a wilderness. The Persian empire at its height, by Ibn Khaldun's own reckoning, could field 200,000.

The first principle is ʿaṣabiyya. English usually gives group feeling, or solidarity. What it names is the readiness of a body of men to fight and die for one another unpaid. Ibn Khaldun grounds it in kinship, real or invented, and in shared hardship. Badāwa, the rough life outside the cities, manufactures the stuff, because out there nobody else does the defending: no walls, no garrison, no watch. Ḥaḍāra, settled urban civilisation, dissolves it for the same reason running backwards. Religion can multiply ʿaṣabiyya, as he thought it had under Muhammad and the first caliphs, though it can't conjure the stuff out of nothing. Mulk, royal authority, is what a group with enough of it can take and hold.

Three moving parts and a tax curve

A group with strong ʿaṣabiyya takes a city and founds a dynasty. Then power buys taraf, luxury, and luxury pulls the ruler's interests away from his kinsmen's. Ibn Khaldun treats that split as mechanical, never as a moral failing. Kinsmen have claims on him. They took the throne with him, and they keep saying so. Clients and imported mercenaries have no claims at all, so the ruler swaps solidarity for salary. That works. It also costs money every year, forever. Call it part one of three.

Part two is jāh, rank and nearness to patronage. In a city, income follows jāh, so the man handing out the offices draws in the talent that would otherwise be making something. Part three is fiscal, and it is the sharpest stretch in Chapter Three of the Muqaddimah. Early dynasties levy small assessments on a wide base and collect a great deal. Late ones levy heavy assessments and collect little, because the rates have killed the activity being taxed. Ibn Khaldun's rulers then go into trade themselves to cover the gap, undercutting the subjects they tax. The base shrinks again.

The lifespan falls out of this, and the arithmetic wobbles. Chapter Two allows a noble house four generations of prestige. Chapter Three gives a dynasty three, at forty years each, so about 120 years, the forty borrowed from the years the Israelites spent in the wilderness. The steps are the same either way. The builder knows the desert and knows what the thing cost. His son met the founder, heard it first-hand, and imitates. The third has the tradition and no memory of anyone who fought for it. The fourth takes the founders for primitives and rank for a natural property of itself, and a fresh ʿaṣabiyya from the periphery removes him. The two counts never quite meet. Two questions test the model on a live state: who pays for its defence, and is anything besides money holding them?

Al-Azmeh's objection, and why I think it lands

Robert Flint set the tone in 1893. In his History of the Philosophy of History he wrote that "Plato, Aristotle and Augustine were not his peers". Arnold Toynbee went further in 1934. In the third volume of A Study of History he ranked the Muqaddimah above anything of its kind produced anywhere. Later admirers made Ibn Khaldun the first sociologist, the first economist, the Laffer curve six centuries early. Arthur Laffer's 2004 Heritage Foundation paper hands him the credit in as many words. Aziz al-Azmeh went after the whole apparatus in Ibn Khaldūn in Modern Scholarship: A Study in Orientalism (1981) and Ibn Khaldūn: An Essay in Reinterpretation (Frank Cass, 1982). His case has two halves, and the second is the one that does real damage.

Provincialism is the first half. What Al-Muqaddimah describes, on al-Azmeh's reading, is one Maghribi arrangement, a tribal hinterland leaning on a handful of coastal cities, promoted to a law of civilisation as such. Reading Ibn Khaldun as a proto-social-scientist also cuts him out of the Maliki jurisprudence and Ashʿarite theology he wrote inside. In that theology secondary causes work within a created order, and never stand in for it. That is fair, up to a point. Every general theory starts out provincial somewhere, and Ernest Gellner built much of Muslim Society (1981) on the model without apology. The second half is harder. Al-Azmeh points out that Ibn Khaldun barely uses his own model in the ʿIbar, the narrative volumes Al-Muqaddimah exists to introduce.

That's the objection I find strongest. It comes from inside the book. Determinism can be argued with, since he allows for variation and is describing tendencies with rough rates attached. Romanticising the tribes is a small fault, and Tacitus did worse with the Germans around AD 98. But a man who builds an apparatus for explaining history, then writes several thousand pages of history without once reaching for it, has told you how much weight he thought it would bear. A beautiful derivation, never checked against his own archive.

What survives once you take the desert out

Take the camels out and two checkable claims remain. First, political authority runs on a solidarity that is expensive to produce and mostly produced by hardship. Second, success raises the price of that solidarity while lowering the felt need for it, so regimes buy substitutes. Substitutes are loyal to the payment, and a payment can always be outbid. Edward Gibbon worked the same ground in 1781 and called Rome's collapse "the natural and inevitable effect of immoderate greatness", then gave his best chapters to Christian enervation. Ibn Khaldun's version needs no vice. Both of his claims have consequences you can measure, the fiscal one in a revenue series, which is more than most answers to why civilisations fall can offer.

What the model can't do is imagine a state getting out of the cycle. Constitutions, standing institutions, taxation by consent: all absent. He had no name for the possibility. In 1401 he was lowered by rope over the wall of besieged Damascus to meet Timur. He spent about thirty-five days in the camp. Much of it went on Timur's questions about the Maghrib, which he wrote up for him afterwards. He had been chief Maliki qadi of Cairo since 1384, sacked and reappointed repeatedly. He died in office on 17 March 1406 and was buried in the Sufi cemetery outside Bab al-Nasr.