A Ruin Is Not a Draft
Myanmar put Bagan forward for World Heritage listing in the 1990s and was turned down, because the government rebuilding its temples was inventing the parts that had fallen off. It was finally inscribed in July 2019, after an earthquake had gone through the plain and demonstrated, brick by brick, which of the work was medieval and which was not.
The plain at Bagan holds something over three and a half thousand stupas, temples and monasteries across roughly 50 square kilometres of the Irrawaddy bend. Nearly all of them went up between the middle of the eleventh century and the end of the thirteenth, when Bagan was the capital of a kingdom that ran most of the Irrawaddy valley, and chronicles claim ten thousand were built. An eight-volume inventory compiled through the 1990s catalogued what survived one monument at a time, and it is the document everything since has argued over.
The argument is not about money and never was. It is about what a ruin is. If a temple is a damaged building, then the responsible thing to do with a collapsed upper storey is put it back. If it is a finished object — evidence, with a shape that includes the missing parts — then putting it back destroys the thing you were trying to save. Bagan is the place where both positions were held with complete sincerity by people who cared about it, and where one of them was carried out at scale before the other could stop it.
Finishing the ruins
From the mid-1990s the military government ran a mass reconstruction programme on the plain. Donors could sponsor the rebuilding of a stupa and have their names recorded on it, and over a thousand structures were raised, re-roofed or completed. The work used modern brick and cement, and where nobody knew what the missing superstructure had looked like, one was designed. A golf course was laid out inside the archaeological zone. A paved road was driven across it. A 61-metre viewing tower opened in 2005, taller than every temple on the plain, in a landscape whose entire visual argument is that nothing stands above Thatbyinnyu. Myanmar nominated Bagan for the World Heritage List in 1996 and it was not inscribed. The reason given, in substance, was that too much of what was there had been built recently and nobody could tell which parts.
The merit argument
It would be easy to file the rebuilding programme as vandalism by a government with no taste, and that reading is too comfortable. In Theravada Buddhism, repairing or raising a pagoda is an act that generates merit for the person who pays for it. Bagan was built in the first place by kings and merchants doing exactly that, and it has been repaired continuously ever since by people doing the same. Shwedagon in Yangon is not a preserved eighteenth-century object; it is a structure that was rebuilt to its present height after the 1768 earthquake and has been re-gilded by public subscription in living memory, and no Burmese Buddhist regards it as compromised. The proposition that a broken stupa should be left broken so that foreign specialists can read its stratigraphy is not obviously the pious position. It is a claim that the building belongs to the study of the past rather than to the practice of the present, and it is worth stating plainly that this is a claim, not a self-evident truth.
The fight was never about how much to spend on Bagan. It was about whether a ruin is a damaged building waiting to be finished, or a finished object in its own right.
What the 2016 earthquake showed
On 24 August 2016 a magnitude 6.8 earthquake with its epicentre near Chauk — deep, about 84 kilometres down, which is why it was felt across most of the country without levelling the towns above it — shook the plain hard. Close to four hundred monuments were damaged. Conservators who went over the wreckage reported a pattern that settled a good deal of the argument without anybody having to win it: a great deal of what had come down was the recent work rather than the medieval fabric. Cement does not move the way lime mortar does, and superstructures grafted onto old cores in the 1990s sheared off them. The response after 2016 was different in kind — emergency stabilisation, documentation before intervention, reassembly using original fallen material where a piece could be identified, and a firm preference for doing less. In March 2018 climbing the monuments was prohibited outright, which removed the sunset crowds from the terraces and put them on artificial earth mounds instead.
Baku, July 2019
Bagan was inscribed on the World Heritage List on 6 July 2019, at the committee session in Baku, twenty-three years after the first attempt. The inscription came with commitments attached: a management plan, action on the hotel zone that had been permitted inside the site, the future of the viewing tower and the golf course, and the reversal of inappropriate reconstruction where reversal was possible. That last clause is the interesting one, because it is an admission that the 1990s work is not simply going to be lived with, and also that most of it cannot be undone — you cannot un-pour cement into a thirteenth-century core without taking the core with it. The listing did not resolve the argument. It recorded which side had won, and left the losing side's work standing on the plain.
The fault moved again
The Sagaing Fault runs roughly north to south for well over a thousand kilometres, more or less down the middle of the country, and Bagan sits near enough to it that the plain has been shaken repeatedly through its recorded history — in 1975 severely, in 2016, and before that often enough that the chronicles treat rebuilding as routine maintenance. On 28 March 2025 the fault ruptured again, a magnitude 7.7 event centred near Sagaing and Mandalay that killed several thousand people — the administration's count settled just below four thousand, independent estimates run well above it, and no full survey has been possible in a country at war — brought down the old Ava bridge and left monasteries and housing across two cities shored up rather than repaired. Fresh damage was reported at Bagan as well. What is different this time is that nobody can survey it properly: the country is in the middle of an armed conflict, the international missions that worked on the plain after 2016 are largely not there, and the money is not there either. The argument about what a ruin is has not gone away. It has simply been overtaken by a period in which the choice between conserving and rebuilding is not on offer, and the plain waits — as it has before — for someone to be in a position to make it.