by John Man
How did they get there, these Chinese chariots with their iron-rimmed wheels? They would not have travelled well through mountains, but they were suited to the steppe, and fortunately there was a steppe close by that led due north-south – Ordos – where the difficult parts had been made easy by Meng Tian’s road. Perhaps even under Xiongnu rule (206–120 BC) it was kept in good condition to speed the delivery of gifts and princesses.
To return to Erdenebaatar’s museum: the most valuable objects (found in tomb No. 20) are fourteen gold-coated plaques used to decorate horse-straps – gorgeous little cartouches and 13-centimetre discs with bas reliefs of two mythological creatures, a sort of a unicorn with a deer’s body, a horse’s head and a single huge, curling horn; and a bers (as it is known in Mongolia today), with the face of a snarling snow leopard, a single horn, a dragon’s neck, wings and camel’s feet. The Xiongnu liked gold, of which Mongolia has an abundance, probably collecting it in minute traces from rivers, so these are almost certainly a Xiongnu product, not Chinese. Other objects reveal international connections with China and beyond. Small gold ornaments with turquoise inlays recall similar ones found in Afghanistan (ancient Bactria), and were quite possibly made there.5
The most intriguing of his discoveries, found in a large circular grave beside the main tomb, is a small blue glass bowl, only 7.5 centimetres high. It has white lines painted round the rim, and the delicate bulge of the body is strengthened with raised ribs. Glass is not a material used by the Chinese at this time. So where did it come from and when? The surprising answer is: Rome in about AD 50. That’s certain because there are six other similar pieces in western museums.6 But how and why was it in the grave of an eminent Xiongnu?
‘I thought it had been traded all the way across Asia,’ said Erdenebaatar, ‘but when I was in Berlin a professor there said no, it was too rare and delicate for trade. It must have been a gift.’ His words set the imagination to work. Could it be evidence for those ‘Romans’ saved from the ruins of Zhizhi’s fortress in 36 BC? Could a Roman have travelled all the way across Asia, like a precursor of Marco Polo? It seems impossible. Remember that Herodotus, in his description of the Scythians, said that it took seven sets of interpreters to cross Asia. More likely, local leaders had handed on this prize possession as a gift, until it landed in the hands of some leading Xiongnu, perhaps the chanyu himself.
A few days later, I set off for the two Gol Mod cemeteries. There were three others with me – Batmõnkh, a guide with a passion for history, Tsend, one of Erdenebaatar’s PhD students, and the driver, Mönkhöö, master of our 4 × 4 Forgon van. We headed west from Ulaanbaatar on tarmac, then north over the open steppe. They call this an ocean of grass, with good reason – distant horizons, a vast blue sky, no fences and a track-network hundreds of metres wide, as if a giant had run a comb across the grass. After 400 kilometres and a night in a friendly tent, we splashed through a stream, crunched through a snow-filled gully, wove between firs and arrived at Gol Mod 2.
Dominating a plateau surrounded by hills, what had once been a deep pit had been completely filled in. Stones outlined a shape like a giant paddle, the shaft marking what had once been a descending ramp and a square over the grave itself. Archaeologists commonly refer to tombs with entrance passageways as ‘terrace’ tombs. Larger ones are said to be those of the ‘élite’, while smaller ones are termed ‘aristocratic’, and the smallest, whether circular or rectangular, are ‘satellite tombs’. The élite tomb we were looking at is the biggest yet discovered.
‘It wasn’t like this when I first saw it,’ said Tsend. ‘Those stones over there’ – he pointed to piles of them a few metres away under fir trees – ‘were all on top of the grave,’ which raised it to make a platform. It had taken a team of about 200 soldiers, released from duty, to clear the rocks and prepare the tomb for excavation. Twenty-one metres down, they found a coffin, broken by looters and then crushed under a protective carapace of rocks. The only human remains were a skull and a femur.
The main grave was part of a huge necropolis. There were 190 smaller terrace tombs nearby, along with 85 circular burials and 250 others. To one side of the main grave were twenty-seven circular tombs, forming a regular, gentle curve, for officials or relatives killed to attend their leader in death. Excavations showed that the coffins were laid out to follow the line of the arc, suggesting that this was a single ritual, all the killings and burials done together. Today, seeing the size of the main grave and the carefully placed subsidiary graves, you get an idea of the power of the chanyu to command in this world and the next, matching that of the Chinese emperor he fought and emulated.
Concealment would have been impossible. This burial site covers a square kilometre, made over many decades, by a people confident that they could protect their tombs from looting or destruction by enemies. The way to do this, it seems, was to build big and deep. What looter with a basic pick made of deer-antlers could dig down 15 or more metres? Quite a few as it happens, but the Xiongnu could not have known that. So this was chosen as a fine place for aristocratic funerals – plenty of pasture, protected from winds by mountains, a nearby stream for drinking water and cooking, and lots of trees for fires, tomb-props and coffins. And beautiful, whether in high summer with the sandy earth unfrozen and soft for digging, or on a crisp autumn day, as this was, with the smell of the firs and a cold breeze sighing through the pine needles. Mongolians called this mysterious place Balgasin Tal, the City Steppe. I imagined myself as a time traveller, leaving the present for the world 2,000 years ago, watching sixty generations of Mongol herders flicker past, and then as the centuries slip by, seeing ghostly parades from earlier cultures – Jurchens, Kitans, Uighurs, Turks and half a dozen others – all pausing to wonder at this evidence of long-vanished predecessors.
I paced the outline of the main grave, not to record it – all the measurements are in the reports – but to feel its scale. Six paces across at the entrance, a 40-metre passageway, sometimes called ‘the pathway to the other world’, becoming wider as it approached the 40-metre-per-side grave, which (as the excavation showed) dropped 21 metres in a series of steps. Imagine a stepped pyramid with a sloping ramp leading almost to the top, and then in your mind turn it upside down. The stepped structure and the sloping walls were to prevent the soft soil collapsing. A back-of-the-envelope calculation suggests that the diggers – probably Chinese prisoners of war and criminals, not ordinary Xiongnu – had to shift 14,000 cubic metres of earth, about 21,000 tonnes. How did they do it? With wooden shovels, loading slings carried on a yoke by two men? Today, a manual worker can dig 5 tonnes a day. The numbers of diggers and sling-carriers would have been constricted as they worked their way down, but in round figures it suggests 400 men working for 10 days, or 200 working for about three weeks.
Nothing to match the First Emperor’s tomb (16,000 men working for 10 months),7 but still quite an operation, with no way to keep it secret. Digging was just the start. The tomb would have been fitted with timbers, dressed with tapestries and finally filled with the objects now in Erdenebaatar’s museum, before the coffin was carried down the entranceway and lowered into its final resting place. The grave, once filled in, was made even more obvious by covering it with hundreds of stones and giving it a low masonry wall, roughly cut into blocks. These people or their captives knew how to dress stones, a skill normally connected with city-dwellers. A layer of cinders near the surface showed that fire had played some part in the funeral rites.
Sunset and a biting wind drove us away. With the low sun turning car-tracks into shadowy claw-marks, we arrived at the nearest town, a scattering of plywood shacks either side of a stretch of steppe corrugated by car-tracks. A two-storey flat-pack building proclaimed: ‘Food Place. Hotel. Karaoke. People’s Shop’. Someone came from the nearby village to cook mutton and rice. We were the only occupants. Next morning, we were off at dawn. Two hours later, splashing through the half-frozen Khünüi River, climbing a steep and forested ridge, descending thr
ough trees flaming an autumnal orange, we came to Gol Mod 1.
This was a place with a long history of archaeology. A Mongolian scientist, Dorjsuren (1923–97), started work here in 1956, listing over 200 graves, half a dozen terrace tombs and several other satellite graves, opening twenty-six of the smaller circular ones. An attempt to excavate the massive main tomb ended when the sides of his 8-metre hole collapsed. Then in 2000 came the French, headed by Jean-Paul Desroches and the Mongolians under Erdenebaatar, who continued Dorjsuren’s work, identifying 316 tombs (since then the count has risen to over 393), of which 214 were aristocratic or élite ones, with entrance pathways. The entrances are aligned roughly north-south, so that as you descended into the grave you headed north, in the direction of the seven stars that make the Great Bear, though whether this is significant is anyone’s guess. Perhaps we should switch the point of view. Perhaps the dead were supposed to exit the tomb uphill and enter the afterlife by heading south, which (if traditional Mongolian practice is anything to go by) is the direction their tents faced.
The team cut the intruding firs, cleared wind-blown sand, and excavated two élite tombs and another seven satellite burials. The most significant was the largest, labelled T1, which was carbon-dated between AD 20 and 50. With the occasional help of a mechanical digger, 800 volunteers carefully scooped out 4,000 tonnes of soil, revealing two layers, bronze and iron relics in the upper one, gold and semiprecious objects below. A rock floor was apparently designed to frustrate looters, but a deer-horn, used as a pickaxe, showed that looters had entered anyway, presumably by digging a long-gone tunnel. Seventeen metres down, the team found a double coffin of larch-wood planks, one inside the other – empty, looted, except for a few small bits of jewellery.
Why loot it? Was there a market in buried jewellery? Unlikely, given that nomads did not use cash. And why go to all the trouble of digging up old stuff when new stuff was available more easily by trade or pillage or from local workshops? Some scholars suggest that the looting was driven by the ideology of successor groups eager to possess or destroy the symbols of prestige displayed by the predecessors. We have no answers.
Anyway, in this case, the looters must have given up, because from the area round the coffin, the team gathered hundreds of items – fragments of vases, gold plaques, a bit of a bronze mirror, bones, silver ornaments, iron wheel-rims, chariot decorations, horse accoutrements, arrowheads, and several pieces of material – silk, wool and felt. A Chinese jade pendant drilled by a local artisan to make two holes for gold decorations. A multi-coloured bead made by fusing pieces of different coloured glass – a complex operation, perhaps done in Europe, involving the use of manganese, copper, iron and barium.
The fragment of a bronze mirror deserves a note. Some thirty bits have been found in other tombs, most of them belonging to women. They were highly polished on one side and decorated on the other. Only fragments, not whole mirrors. They had been broken deliberately, for burial. Why? Many objects were broken, a shamanistic practice intended perhaps to ‘release the essential spirit of the object, thereby rendering it useful in the after-life’.8 Mirrors, though, seem to be in a class of their own. The fragments suggest a combination of Chinese and nomadic rituals. Scholars speculate that mirrors represented the cosmos (many mirrors from Chinese graves have sky symbols, like clouds and dragons, on their reverse sides) and were thought to capture images, as the eye does, or to emit the light ‘absorbed’ above ground. In China, they were commonly placed on the dead, facing upwards, to lighten the darkness and drive away evil spirits. I once saw a house near Xian with a mirror over the door, so that evil spirits would see their own reflection and be scared away. Mirrors were objects of ‘reverence and prestige within nomadic cultures and, moreover, retained sacred symbolism’, being passed down the generations or smashed into pieces that were divided between different burials.9 Even today, when burying a deceased woman in Mongolia, her beloved items are often shown to the mirror before it is placed in the coffin, so that she can have them with her in the afterlife.
(The belief in the magic of mirrors is not limited to Mongolia and China. As everybody who has watched a vampire movie knows, vampires do not get reflected in mirrors. In Sleeping Beauty, the evil stepmother’s mirror reveals truths. In Aachen, in the early fifteenth century, gullible pilgrims used little concave mirrors to ‘collect’ the healing power of Christ’s robes when they were shown to the public. Johannes Gutenberg, inventor of printing with moveable type, planned to make his fortune by turning out these mirrors by the thousand, before a plague cancelled the pilgrimage and forced him to turn his attention to something more significant.)
A lacquered tray offers rare support for written sources, like a torch spotlighting a find in a cave. The tray had a Chinese inscription recording its origins: made by, or under the supervision of, someone called Wu, working under seven other layers of supervision10 – seven! This was the product of a vast and complex bureaucracy in the Imperial Workshop, Chang’an, in 16 BC. As the authors of the article deciphering the inscription say, ‘This is important as it is only the fifth absolutely dated object excavated from Xiongnu élite tombs.’ It is also further proof that the Xiongnu aristocracy was in thrall to the Chinese after their submission in the mid-50s BC (more about that shortly). If this correlates with the burial, it dates from the reign of a chanyu named Seuxie (or Souxie, transliterations vary), who ruled 20–12 BC. Perhaps it is even the tomb of Seuxie himself.
After removing everything they found, they re-covered the grave and replaced the low surrounding wall, stone by well-dressed stone, each one carefully numbered.
The second élite tomb – T20, just over 18 metres deep – contained the partial remains of sixteen horses, notably their heads, all arranged to look north. Their teeth reveal that they were killed during the summer, supporting the idea that tombs were dug then. Their legs were placed to suggest that they were walking, as if to accompany the dead into the afterlife, perhaps also to provide them with food, because the limbs were laid on a bed of burning charcoal. One small item was missing. ‘It is important to note,’ writes Hélène Martin in a report on the finds,11 ‘that no trace of the hyoid bone was found,’ whereas in the case of a human (see p. 231), it had been carefully preserved – a symbol of something that will probably forever remain a mystery.12
For an overview, we drove to the top of a little hill, which turned out to be a sand dune fixed in place by a covering of grass. From there it was obvious that the two cemeteries were remarkably similar, of equivalent size, in places of equal beauty, set in forests and protected by forested mountains. Sixty kilometres apart as the crow flies, they were linked by culture, family and ritual. Perhaps leaders from Ordos came here with other delegates from across the empire to take part in the burials and memorial rites.
Surely, with cemeteries of this size, there would have been a capital city nearby?
‘Yes,’ agreed Tsend, ‘because, of the ten thousand known Xiongnu graves, about one-third were in this area, near the Tamir River.’ Within 100 kilometres, river valleys and natural highways connect to the Orkhon River, the main artery for later empires. This was the Xiongnu heartland. But, Tsend added, even at its height when there was little chance of an invasion, the capital must have been no more than a tent-city. If there are remains of stone – well cut like those surrounding the tomb in front of us and final proof that this region was in fact the administrative heart of the Xiongnu empire – they await discovery.
There was a mystery here. The two Gol Mod cemeteries, Noyon Uul and the other élite tombs were radically different from the graves in Ordos, where there were none of the big, deep terrace tombs with sloping entrances and Chinese chariots; radically different too from the thousands of other Xiongnu tombs in Mongolia.
Back in Ulaanbaatar, Tsagaan Törbat of the Mongolian Academy of Sciences gave me an explanation.
‘All Mongolian élite graves date from the end of the second half of the first century soon after the Xiong
nu submitted to Han, to the first century AD,’ he said. We were in his office, which was cluttered with the books and papers to do with his current research into the Xianbei, successors to the Xiongnu. ‘So we’re talking about a short period, only a few decades. In Mongolia, we have only eight places with these élite graves. They are in the shape of Chinese royal graves’ – though the Chinese tradition was to make mounds over the graves, as did the Scythians, something that the Xiongnu did not do. ‘We have to ask why we suddenly get these big Chinese-style graves full of Chinese – or Chinese-style – objects: chariots, buttons, jade discs, mirrors, ceramics.’
He paused, then answered his own question.
‘I think something happened to Xiongnu ideology. It was to do with the balance of power. In 51 BC, the Xiongnu under Huhanye accepted peace, acknowledging China as the superior culture. It must have been a severe shock, tolerable only because they received so much from China. I think that’s why they adopted Chinese-style tombs and burial rites.’
It sounds right. Written evidence supports the suggestion of a radical change. The chanyus started to call themselves ruodi (or jodi), which may be a Chinese version of the Xiongnu for ‘filial’ or ‘loyal’. Members of the aristocracy, who like most Xiongnu had names that were hard to transliterate into Chinese, started to give themselves simple Chinese names: Zhi, Dong, Zhu, Bi. As the American scholar Bryan Miller writes, ‘They chose to appease the Chinese with whatever language became necessary, in exchange for a peaceful southern frontier and a steady increase in gifts.’13