Emperor Xiaowen and his entourage worshipping the Buddha

A limestone procession relief, nearly four metres wide, that was cut out of the wall of the Central Binyang Cave at Longmen in the 1930s and reassembled in New York. It shows the Northern Wei emperor Xiaowen and his court carrying offerings towards the Buddha; its companion, the empress and her court, is in Kansas City. The pair were chiselled into more than six thousand fragments on the orders of a Beijing dealer named Yue Bin, working to instructions from buyers abroad, and about four thousand of those fragments reached the United States.

The Metropolitan Museum's published provenance for the panel is one line long. It names the cave. It names nothing and nobody between the cave and the museum's own purchase in 1935. When a Chinese documentary team asked the museum for its restoration records in the course of work broadcast in May 2026, the account they published says very little was provided.

Roughly 2,300 fragments of the two reliefs never left China and are still in the storerooms of the Longmen Grottoes Research Institute. In 2026 that institute matched a fragment to the wall of the cave, by 3D scanning and printing, at a reported 93.3 per cent fit: the face of Emperor Xiaowen.

What it is, and where the wall it came from still stands

The object is a limestone relief panel 208.3 cm high and 393.7 cm wide, with traces of the pigment that once coloured it. It shows a procession: a figure in court robes holding a tasselled baton leads, followed by a smaller armoured attendant and another bearing an incense burner, and behind them the emperor with his entourage, several of whom carry lotuses and other offerings. The Metropolitan Museum reads the whole procession as an act of merit-making directed at the Buddhas inside the cave, an offering intended to continue in perpetuity and to improve the future lives of the people shown making it.

It is not a free-standing sculpture and was never meant to be looked at as a panel. It was carved into the northeast wall of the Central Binyang Cave at Longmen, twelve kilometres south of Luoyang, positioned between a register of protective semidivinities above and a narrative debate scene below. The trees carved into it mark the point at which the procession turned a corner, from the north wall onto the east wall. It wrapped around masonry. What is in New York is flat, because the corner was cut off and the two planes were flattened into one when the fragments were put back together. The wall it was cut from is still there, and it is still empty.

An imperial commission with a documented labour cost

Emperor Xuanwu of the Northern Wei ordered the excavation of three caves at Binyang from 500, in honour of his father Xiaowen, who reigned from 471 to 499, and his mother, the dowager whom the Metropolitan's catalogue names as Wenzhou and dates to 494. The emperor and empress in the two reliefs are understood to represent those two people. Only the central cave was finished within the Northern Wei campaign, in 523; work on the north and south caves stopped and they were brought to completion only under later dynasties.

The carving of the Binyang group ran for twenty-four years and is recorded as having involved over 800,000 workers. Figures of that kind almost never survive for an object of this age, and it belongs on the page: this is one of the very few works of art in any museum whose commissioning cost is expressed in human labour rather than in money. Note that the number is a total for the three-cave campaign, not for this panel, and that the more precise figure of 802,366 which circulates in the literature is not carried by the source cited here.

The relief is a political document about clothes

Xiaowen is the emperor who moved the Northern Wei capital from Pingcheng to Luoyang in 494 and drove through a programme of sinicisation: the Xianbei ruling elite were required to adopt Chinese surnames, Chinese speech at court, and Chinese dress. The Binyang commission is explicitly commemorative of that reign and those reforms.

So the costume in this relief is the argument, not the decoration. The court that files across the panel wears the long robes and wide sleeves of Chinese officialdom, not the riding coats and trousers of the steppe from which the dynasty came. A Northern Wei emperor is shown to posterity dressed as a Chinese emperor, performing a Buddhist devotion in a cave paid for by his son. The offering is real piety and the clothing is real policy, and the panel does both at once. That is what a viewer standing in Gallery 206 in New York is not told, and it is among the most important things about the image.

How it was taken off the wall

The two donor processions were destroyed in the 1930s, in the sense that they were reduced to rubble on purpose so that the rubble could be sold. A Beijing antiquities dealer named Yue Bin organised the systematic chiselling of the two panels into more than six thousand fragments, working on orders from buyers overseas. Around four thousand pieces went to the United States. Roughly 2,300 stayed in China and are held in the storerooms of the Longmen Grottoes Research Institute.

The looting was not confined to these panels. The wider modern despoliation of Longmen concentrated on heads and faces, which is itself evidence of who the market was: in the visual culture of Chinese Buddhism the hands and their mudras carry the doctrinal weight, and a severed head is a Western collecting preference, not a local one. Chinese looters removed the parts that Western buyers would pay for.

Historical records of Yue Bin's confession, cited in a 2026 Chinese documentary on the reliefs, add a detail that matters for every fragment now in a museum: the looters deliberately mixed genuine pieces with fakes in order to inflate the count and the price. Nobody could then say with certainty which fragments were authentic, and that uncertainty was manufactured at the point of sale.

The reassembly, and how much of this object is plaster

The fragments were bought abroad and put back together abroad. At the Nelson-Atkins Museum in Kansas City, the curator Laurence Sickman, assisted by the Peking-based dealer Otto Burchard, acquired large numbers of pieces of the empress panel; the fragments were reassembled by Sickman with the sculptor Wallace Rosenbauer, and the gaps were filled with plaster. That panel went on public display in 1941. The Nelson-Atkins bought it in 1940 and gives its own provenance as the Binyang Cave until the 1930s, then the Chinese art market between 1932 and 1939, where the relief was, in the museum's own words, vandalised and sold in fragments.

The Metropolitan's emperor panel came earlier, in 1935, on the Fletcher Fund. It is the more heavily restored of the two. Scholarship summarised by Harvard's Whose Culture? project states plainly that the faces of both processions were salvaged and collected piecemeal, and that the remainder of the Metropolitan's emperor procession is mostly plaster infill, while the Nelson-Atkins empress procession is more intact.

That is a striking sentence to find nowhere on the museum's own label. A visitor is shown a Northern Wei imperial relief; a reader of the scholarship is told that much of the surface in front of them is twentieth-century filler. When a documentary team from China asked the Metropolitan for its restoration records while making a film broadcast in May 2026, their published account is that very little was provided. This record does not assert that the museum holds records it is withholding; it records that the request was made, that the answer was thin, and that the museum has published no breakdown of original stone against restoration.

The provenance the museum publishes is one line long

Read the Metropolitan's own provenance field for object 35.146 and this is the whole of it: Binyang Cave, Longmen, Luoyang, Henan Province. There is no dealer. There is no date of removal. There is no intermediate owner, no consignor, no agent, no port of entry. The panel appears in the museum's history in 1935 as a purchase on the Fletcher Fund and has no recorded life between the sixth century and that year.

For most objects a one-line provenance is a gap. Here it is the story. Everything documented about how this relief travelled - Yue Bin, the six thousand fragments, the market of 1932 to 1939, the confession about salted fakes - is documented outside the museum that owns it, mostly by the Chinese institution that guards the empty wall. The companion museum in Kansas City publishes a chain for its half of the same act of destruction: the cave until the 1930s, the Chinese art market between 1932 and 1939, the acquisition of fragments by Sickman through the dealer Otto Burchard, and the purchase of 1940. The Metropolitan publishes one line.

The claim, and why it has not become a case

There is no litigation over this panel and, in the sources consulted, no formal restitution demand from the Chinese state naming it. That absence is worth stating precisely, because it is often assumed the other way round.

China's cultural-property claims run into a hard chronological problem: the 1970 UNESCO Convention and the 1995 UNIDROIT Convention are not retroactive, and a removal of the early 1930s falls outside both, as it falls outside almost every national limitation period that might otherwise apply. What China has instead is scholarship, publicity and patience - and, increasingly, scanning equipment. The Longmen Grottoes Research Institute's strategy has not been to sue for the stone. It has been to prove, fragment by fragment, exactly what was taken and exactly where each piece belongs, and to publish the result.

Chen Yi's documentary How I Miss Her, broadcast on China Central Television's documentary channel in May 2026, is that strategy made public. In it a researcher holds a 3D-printed fragment against the weathered wall of the Central Binyang Cave and it seats into a groove cut fifteen centuries earlier; the data are reported as a 93.3 per cent match. The fragment is the face of Emperor Xiaowen. Work of this kind used to take years for a single head; this took fourteen months.

The reunification, if it comes, is likelier to be digital than physical. That is not a lesser outcome for scholarship and it is a significantly lesser one for the source community, and both things are true at once.

The man who has been sorting the fragments since 1965

Wen Yucheng joined a three-person team investigating the looting of Longmen in 1965, shortly after graduating in archaeology from Peking University. He was still at it sixty years later, appearing in the 2026 documentary as honorary director of the Longmen Grottoes Research Institute at the age of eighty-seven.

That is the labour side of the provenance chain, and it almost never appears in a catalogue entry. Against the reassembly of the empress panel in Kansas City, completed for its display in 1941, stand six decades of one archaeologist's cataloguing in Henan. A record of this object that lists only the museums that own it has described the smaller half of the effort.

What a reader should not confuse this with

Three things are routinely run together in writing about this relief and they are three different objects.

The first is the companion panel. The empress procession is not in New York and never has been; it is at the Nelson-Atkins Museum of Art in Kansas City, accession 40-38, 203.2 by 278.13 cm, bought in 1940. Text and photographs of the two are frequently swapped. The Metropolitan holds the emperor; Kansas City holds the empress.

The second is the rubbing. The Metropolitan also holds object 68.166.1, catalogued as a twentieth-century rubbing of the relief and titled almost identically - Emperor Xiaowen with his entourage worshipping the Buddha. It is a different accession, it is not in the public domain, and the museum publishes no image of it. A rubbing taken from the wall is a document of what the carving looked like at the moment the ink was pressed onto it, which makes it potentially the best comparison material there is for judging the restoration. It should never be cited as a photograph of the panel.

The third is the wall. Photographs captioned as this relief sometimes show the northeast wall of the Central Binyang Cave as it is now, with the register missing. That is a picture of an absence, and the absence is the point, but it is not a picture of the object.

What it is worth, and why no figure exists

There is no auction record for this panel, no insurance value in the public domain, and no purchase price published by the Metropolitan for the 1935 Fletcher Fund acquisition. The Nelson-Atkins publishes no figure for 1940 either. Both museums bought fragments, not a finished object, which means that even a recorded outlay would not correspond to the thing now on the wall.

The absence is structural rather than accidental. Objects removed from monuments do not have market histories in the ordinary sense: they are assembled from many small purchases, they are almost never resold, and their value is never tested. Anyone doing diligence on a comparable piece should note the consequence. A Longmen fragment with a clean pre-1930 photograph attached is a different asset from one without, and the photographic record held in Chinese and American university archives is what separates them.