The Kingdom of Pagan

Ten thousand temples

Drafted from Aung-Thwin, Pagan: The Origins of Modern Burma (1985)formation

On about forty square kilometres of dry plain beside the Irrawaddy there are around two thousand standing brick monuments, and the inscriptions and survey work indicate that something over ten thousand were built. They went up in roughly two hundred and fifty years, between Anawrahta's accession in 1044 and the collapse in 1287.

Pagan is the first state to hold the Irrawaddy from the delta to the dry zone. It is also, on Aung-Thwin's reading, a state that built itself to death.

Anawrahta and Thaton

The Burmese chronicles give the founding story a clean shape. Anawrahta, having united the dry zone, asked the Mon king of Thaton in the south for a copy of the Pali scriptures, was refused, took the city in 1057, and brought back the king, the monks, the craftsmen, the texts and the script. Theravada Buddhism and literate administration arrive together, in a cart, in a named year.

Aung-Thwin has spent a career arguing that this did not happen in anything like that form, and that the Mon contribution to Pagan was substantially constructed by chroniclers writing centuries later. It is a genuinely contested position and other specialists do not accept it.

The dispute does not erase 1057 as a hinge, because whatever happened that year, Pagan writes Old Burmese in a script adapted from Mon afterwards and did not before. Something came north. The argument is about how much, and about who wrote the account of it.

Merit

Drafted from Aung-Thwin, Pagan: The Origins of Modern Burma (1985)expansion

The reason there are ten thousand temples is that building one was the most effective thing a person with wealth could do with it.

Theravada Buddhist merit-making, at Pagan, meant endowing a religious foundation: a temple, a monastery, an ordination hall, with land to support it and bonded cultivators to work the land. The donation was recorded on a stone placed at the site, listing the donor, the fields, the people, the animals and the intention. Several hundred of those inscriptions survive, in Old Burmese, saying with legal precision who gave what.

The system worked for everyone inside it. Kings acquired legitimacy; officials and merchants acquired standing; the sangha acquired an institutional base; and the plain acquired the largest concentration of Buddhist architecture ever built.

The tax base

Land given to the sangha was given permanently and was free of tax.

That is the whole of Aung-Thwin's argument. The crown's income was a land tax on irrigated rice land in the dry zone, worked by bonded labour. Every act of royal merit-making transferred some of that land, and the labour attached to it, out of the taxable economy for good. Two hundred years of the most pious kings being the most generous ones is two hundred years of the state's revenue base shrinking, one donation at a time, in a direction with no mechanism for reversal.

By the thirteenth century the inscriptions show kings attempting to buy land back, to reclaim endowments, and to purge the order — moves that a state with a sound fiscal position does not need to make.

Whether this is the explanation for what happened in 1287 is argued about. It is the explanation the next chapter follows.

The chronicle, and the stones

Drafted from The Glass Palace Chronicle of the Kings of Burma (1923)peak

The Burmese court's own account of how Pagan became a Buddhist kingdom is a good story, and the inscriptions do not support it. Both of those facts are worth having, and the second one is the more interesting.

What the chronicle says

The Glass Palace Chronicle, compiled at Ava in 1829 from older chronicles, tells it like this.

Anawrahta, converted by a monk from the south, sent envoys to Manuha, king of Thaton in the Mon country, asking for a copy of the Pitakas — the Buddhist canon. Manuha refused, and spoke ill to the envoys. Anawrahta marched, took Thaton in 1057, and brought back the scriptures, the monks, the relics, the craftsmen and the king.

The chronicle is precise about the numbers and the furniture. He kept thirty sets of the Pitakas in a pyatthat richly fraught with gems and had the Order give instruction in them. He put the relics in a ruby casket and kept it by his bed where his head lay.

And of Manuha it says that his glory was such that whenever he spoke, a wheel issued radiant from his mouth — so that when he bowed before Anawrahta, that king was aghast, and his hair stood on end.

What kind of document that is

Not a chronicle in the sense of a record kept at the time. It is a nineteenth- century commission compiling much older material, and the material is heroic.

Kyanzittha, the greatest of the Pagan kings, is hidden as an infant by a Naga while soldiers kill six thousand people looking for him, and then five thousand more. He is recruited into the king's service as one of four heroes chosen for physical feats — one could plough a whole field with sixty yoke of oxen keeping a straight furrow; one could run down the cliff at Nyaung-u, swim the Irrawaddy and swim back without touching the far bank. Their horses are demons, and the chronicle names all five of them.

To its credit, the chronicle knows it is handling traditions. It quotes the Great Chronicle against the Thaton Chronicle, notes where the two disagree, and says so.

What the stones say

Michael Aung-Thwin's work on the Pagan inscriptions put the Thaton conquest under pressure and it has not fully recovered.

There is no contemporary inscription recording a conquest of Thaton. There is no archaeological horizon at Thaton matching a sack in 1057. The Mon element in early Pagan's material culture, on this reading, comes from a longer and less dramatic contact rather than from a single carried-off library, and Theravada Buddhism was present in upper Burma before Anawrahta.

The counter-case is not dead and the argument is live. What is not disputed is that the earliest evidence for the story is much later than the event.

Why a court would invent it

Because it does exactly the work a founding story has to do.

It makes Burmese Theravada a possession legitimately acquired — asked for first, and taken only after an insult. It makes the dynasty the guardian of a canon rather than a group of Upper Burma warlords who got lucky. And it supplies a defeated king whose radiant mouth proves the value of what was taken.

What actually happened is that a state on the dry zone of the middle Irrawaddy unified the irrigated basin, acquired a religion that was already in the country, and spent two hundred and fifty years building.

What is certainly true

The buildings. Something over two thousand brick temples and stupas still stand on the plain at Pagan, out of a great many more, built between the eleventh century and the thirteenth.

There is no dispute about those, they need no chronicle, and they are the reason the argument about Thaton matters so little to what Pagan actually was.

A throne taken in a privy

Drafted from Harvey, History of Burma: From the Earliest Times to 10 March 1824 (1925)peak

Narapatisithu, whose reign is the one against which Pagan's own architecture and administration are usually measured, did not come to the throne quietly. His elder brother, already king, wanted his wife and sent him off on a manufactured pretext of a war in the north to clear the way for taking her. Narapatisithu discovered the deception before he had gone far, turned his column around, and marched back on the capital. His brother was killed by soldiers sent ahead of him, and Narapatisithu took the throne — and his wife back — in 1173.

A trade dispute that ended in a raid

Sri Lanka had long kept a resident agent in Burmese ports to manage a profitable trade, and Burmese kings had long supplied him with rice and lodging as a matter of custom. Narapatisithu cut it off: he stopped the customary support, restricted the export of elephants, imprisoned Sinhalese merchants, mistreated the Sri Lankan king's envoys, and seized a Sinhalese princess travelling through his waters. The Sri Lankan king answered with a naval expedition in 1180 that reached the Burmese coast, killed a governor, burned villages and carried off captives before the two courts made peace through the monastic channels that connected them — a reminder that a kingdom built on rice and temples was not entirely insulated from a sea it rarely otherwise dealt with.

The buildings that carry his name

His reign supplied Pagan with two of its most celebrated monuments, the Gawdawpalin and Sulamani temples, alongside a wider programme of pagodas, weirs and canals stretching from the Kyaukse irrigation works to an ambitious and ultimately abandoned canal scheme further south. An inscription he raised as far off as the Shan hills marks how far his authority actually reached, at least in name.

An heir chosen by accident

Narapatisithu died in old age, in the thirty-seventh year of his reign, after naming as his heir not his eldest son but a younger one — the story handed down is that the boy's mother had endured a private discomfort of the king's without complaint, and that the king rewarded her quietly by giving her son the succession over his elder brothers. Whatever actually decided it, the chosen son took the throne without apparent resistance from his rivals, and the dynasty that had itself begun in a private killing changed hands, this time, without one.

The merit that would not come back

Drafted from Aung-Thwin, Pagan: The Origins of Modern Burma (1985)contraction

Pagan was dismantled by its own piety, on a mechanism nobody involved could have described as a problem, and the process took about a century and a half.

How a temple was paid for

A Burmese king or noble who wished to make merit built a temple, and endowed it. The endowment was land, and with the land went the people who worked it and the revenue they produced, granted to the monastery in perpetuity, recorded on stone.

The crucial word is perpetuity. Property given to the sangha was religious property. Taking it back was not a fiscal decision; it was an act against the religion, and no king could do it without putting his own merit — and his legitimacy, which rested on being a defender of the faith — at risk.

The arithmetic

Do that for two hundred years in a state whose entire tax base is irrigated rice land in a fairly small dry zone, and the arithmetic is inexorable.

Aung-Thwin's reconstruction from the inscriptions is that by the thirteenth century a very large share of the cultivated land of the Kyaukse and Minbu irrigated basins — the granary of the kingdom — was in monastic hands and off the tax rolls, along with the labour attached to it. The crown's own domain shrank while the demands on it did not.

And the incentive ran the wrong way at every level. A king who wanted merit gave land. A noble who wanted merit gave land. A landholder who wanted to keep land out of the king's reach could dedicate it and hold it as a monastic steward. Piety and tax avoidance produced identical documents.

What was tried

Periodic purifications of the sangha — reforming the order, expelling monks judged lax, and returning improperly held property to the crown. Several Pagan kings did it, and it worked temporarily and at a cost, because each purification was an argument with the institution the dynasty's legitimacy rested on.

The building did not stop. The temples of the twelfth and thirteenth centuries are the largest and most elaborate at Pagan: Dhammayangyi, Sulamani, Htilominlo, Gawdawpalin. Each of them is an endowment.

The state that met the Mongols

So when Mongol envoys arrived in the 1270s demanding submission, the kingdom they were addressing had a magnificent capital, an enormous religious establishment, and a shrinking fiscal base with which to raise an army.

Narathihapate is said to have had the envoys executed, which is the decision of a ruler who does not understand what is in front of him. What followed — the campaigns of 1277 and 1283, the fall of the capital in 1287 — is the subject of the next chapter, and it is best read as what happens to a state that has spent its revenue on merit when somebody arrives who wants tribute.

The general case

This is not a Burmese peculiarity, and it is worth saying so, because the story is often told as though excessive Buddhist devotion were the specific flaw.

Land alienated in perpetuity to a religious institution that the state cannot politically reclaim is one of the most reliable ways for a pre-modern government to hollow itself out, and the same mechanism turns up in the Gupta copper-plate grants, in the monasteries of medieval Europe before the dissolutions, and in the waqf endowments of the Islamic world.

What makes Pagan the clearest case is the evidence. The endowments were cut in stone with the acreage and the labour on them, and they are still standing next to the temples they paid for.

1287

Drafted from Aung-Thwin, Pagan: The Origins of Modern Burma (1985)end

A Yuan army came down from Yunnan. Narathihapate abandoned the capital, and was killed by one of his own sons. The Irrawaddy broke into pieces held by Shan chiefs and no single authority held it again for two and a half centuries.

The chronicles call him Tarokpyemin — the king who fled from the Chinese. That is a verdict rather than a description, delivered five centuries later by people writing for a court that needed the fall explained.

Not a conquest

The Mongols did not take Burma, and it is worth being exact about that, because "the Mongols destroyed Pagan" is the version everyone knows.

Aung-Thwin's account follows from the chapter before. Pagan was already fiscally insolvent when the Mongols arrived: the taxable land had gone to the monasteries, the crown could not raise or pay a force proportionate to what it faced, and the invasion finished a state that had stopped being able to defend itself rather than defeating one that could.

And the Yuan did not stay. They did not annex Burma, they did not garrison it, they did not administer it, and they left. What replaced Pagan was not Mongol rule but two hundred and fifty years of nobody — three Shan brothers who had been Pagan commanders, then Myinsaing, Pinya, Sagaing, Ava, and a separate Mon kingdom in the south.

It is a contested reading and the alternative is respectable: the campaigns of 1277 and 1283 were real, Narathihapate abandoned his capital and is remembered as Tarokpyemin, the king who fled from the Chinese, and a state does not usually cease to exist in the same decade an empire invades it twice by coincidence.

What both readings agree on is that no outside power governed Burma afterwards. Whatever the invasions did, they did not deliver the country to anybody.

What Pagan did not export

Drafted from Aung-Thwin, Pagan: The Origins of Modern Burma (1985)aside?What aside means

Ten thousand Buddhist temples on one plain, and Pagan did not carry Buddhism anywhere.

That is worth separating out, because scale of devotion and religious expansion get conflated constantly, and they are different things. Pagan received Theravada Buddhism — from the Mon in the south, from Sri Lanka by sea, and by whichever route the long argument about 1057 eventually settles. In every version of the story, it arrived. What Pagan then did was to make itself the largest single monument to it in the world, on a plain that already believed.

Other courts in the region received a religion from abroad and then became the route by which it spread further still, through their own merchants and marriages. Pagan did not. Nothing moved on from it. The Theravada tradition that later reshaped Cambodia and Siam travelled through Sri Lanka and the Mon world, not through Burma, and the Burmese reformist ordination lineages that did eventually carry outward are later than this kingdom.

The same shape, with letters

The Burmese script tells a similar story from the other side.

It was adapted at Pagan from the Mon script, for a language of an entirely different family — a substantial piece of work, and the reason Burmese is written the way it is today. But the parent was Mon, and what Pagan did was take a system and use it, rather than build one that others then took from it.

A kingdom can be the greatest patron a religion ever had and still be a terminus rather than a route. Pagan is the clearest case of it anywhere in Southeast Asia: everything came in, almost nothing went out, and what it built with what came in is still standing.

Two thousand temples

Drafted from Aung-Thwin, Pagan: The Origins of Modern Burma (1985)afterlife

The kingdom ended in 1287. Almost everything it was for is still standing.

What survived the state

Something over two thousand brick monuments remain on the plain at Pagan today, out of an estimated ten thousand or more originally built. They survived because brick is worth nothing to a looter, the site was never resettled at scale, and the merit attached to them made destroying them unthinkable to everyone who came afterwards.

They also went on being repaired. Pagan remained a pilgrimage site continuously for seven hundred years; kings of Ava, Toungoo and Konbaung endowed restorations of temples built by a dynasty they did not descend from, because the merit belonged to the act rather than to the ancestry.

An earthquake in 1975 brought down a great deal, and another in 2016 did more. The restoration after the first was thorough and archaeologically careless; the argument about it delayed UNESCO recognition of the site until 2019.

The religion

Pagan is where Burma becomes and stays Theravada, and that has not changed in nine hundred and fifty years.

What the kingdom established was not simply a set of beliefs but an institutional arrangement: a monastic order holding endowed land, a king whose legitimacy rested on defending and purifying that order, and a relationship between the two that involved periodic conflict over property. Every subsequent Burmese dynasty inherited the arrangement, including the purifications — the sangha reforms of the eighteenth and nineteenth centuries are recognisably the same institution being managed the same way.

The script

Burmese is written in an alphabet that took its shape here, derived from Mon, which came from a south Indian script.

The rounded letterforms are a technical consequence rather than an aesthetic choice: the writing surface was palm leaf, incised with a stylus, and straight strokes across the grain split the leaf. So the letters became circles.

The earliest substantial Burmese text is the Myazedi inscription of 1113, carved in four languages — Pyu, Mon, Burmese and Pali — on one stone. It is the Rosetta stone of Burmese studies and the reason the Pyu language could be approached at all.

The three sisters

The political afterlife is less tidy.

After 1287 upper Burma passed to three brothers of Shan origin who had been Pagan commanders, and then to a succession of kingdoms — Myinsaing, Pinya, Sagaing, Ava — while lower Burma went to a Mon kingdom at Pegu. The country was not reunified for two hundred and fifty years.

A rump line at Pagan continued as vassals for a few decades and then stopped being mentioned. There is no dynasty afterwards that claims to be the Pagan dynasty continuing.

What every later Burmese state did claim was the inheritance in a different form: to be the defender of the religion Anawrahta was said to have brought from Thaton, in a kingdom whose model was the one on the plain at Pagan. The Konbaung kings styled themselves accordingly and the Glass Palace Chronicle was commissioned by one of them, in 1829, to write the whole sequence down as a single continuous story with Pagan at the start of it.

That chronicle is why the Thaton conquest is the version everyone knows. A dynasty five hundred years later needed a founding, and gave the country one.