Empire of Horses Read online

  One summer day in 1972, an old herdsman named Wang Shun was tending his flock of sheep in sandy semi-desert west of the region’s capital, Ordos City, where pasture merged into desert. A sudden storm caught him out in the open. Rain and wind combined to shift some loose earth from a nearby slope. As the skies cleared, he was astonished to see something glinting in the newly fallen soil. He dug, and found bits and pieces of gold, mixed in with bones. The storm had washed open a tomb, one of two, as later excavations revealed. Among the finds was a golden coronet with a separate golden skullcap topped by an eagle made of turquoise. It has since become the greatest of the Ordos Bronzes, the objects that define not only the culture of Ordos but much of Scythian culture all the way across the continent.

  The coronet, the most impressive of 218 golden items found at the tomb-site, is unique. Though almost all the thousands of Ordos’s little medallions are of bronze, the golden coronet is treated as the centrepiece of the Ordos Bronzes, and was the inspiration for Ordos City’s Bronze Museum, which opened in 2015. The building is circular, as the coronet is, with three storeys, and it has a domed roof, decorated with clouds, which could be both the coronet’s skullcap and the Mongolian deity, the Blue Sky.

  The museum’s curator, Wang Zhihao, who wore one of those sleeveless jackets favoured by archaeologists in the field, introduced me to the collection. He was helped by an English-speaking aide, the Deputy Director of the Education Department, who had adopted a surprising pseudonym. He admired the American basketball player (now retired) and rapper, the vast (104.3-kilogram), towering (2.16-metre) yet remarkably agile Shaquille O’Neal, who shortened his name to Shaq. With a little linguistic licence, my translator, Zhang Ziyang, called himself ‘Mr Shark’.

  On the way in through the displays, the coronet, of course, had pride of place, well lit and glittering in its glass case. I was thrilled to see this gorgeous object up close. Gold braid forms a circle that links a horse and a goat on opposite sides. Perhaps they symbolize the wealth of the herding community. A braided semicircle accentuates the forehead section, with a pair of tiger heads on either side, looming like a threat above the horse and goat motifs. On the separate little skullcap is a golden eagle with a head of green turquoise. The eagle’s head was attached to the body with gold wires, so that it swayed as the wearer moved. The skullcap itself is divided into four embossed sections, each of which – you have to look closely, because it’s hard to see the markings – contains a stylized wolf intertwined with a goat or ram. Dating probably from the late Warring States period (403–221 BC), it must surely have proclaimed its owner to be ruler of these grasslands. Nearby the figure of an imperious nomad chief showed how it would have looked in life.

  So it was a bit of a let-down when Mr Shark said: ‘It’s good, isn’t? For a replica.’

  ‘What?’

  ‘Yes. It’s perfect in every way. But the real one is in the museum in Hohhot.’ That’s Inner Mongolia’s capital.

  It seemed unfair on the Bronze Museum, since the coronet has become the symbol of Ordos. But Hohhot had it long before the Bronze Museum was built, so no doubt it will stay there.

  We moved on to the bronzes. Usually only a few centimetres across, they are typically belt-buckles, but also include tops of tent poles, decorations for horse-harnesses, knives, daggers with decorated handles, arrowheads, bronze mirrors, hooks for hanging things in tents, and buttons. Two intriguing little objects seem to be a pair of weights that were attached to ropes, which were swung and thrown to entangle the legs of fleeing animals – a trick that no other Asian group adopted, but which is widely used in South America.

  Belts were important to nomads, signifying status, power, adulthood and identity. In The Secret History of the Mongols,4 the only Mongolian source for the rise of Genghis Khan, Genghis’s mother asserts her status by putting on her high hat and belt. When Genghis renews his vows of friendship with his blood-brother (anda) Jamukha, he gives him a golden belt. When Genghis punishes his brother Khasar, he seizes his belt.

  The designs reveal a fascination with animals – horses, sheep, deer, birds. Some finds have an eagle fighting a tiger for a goat, monsters of various sorts, dogs, and warriors killing captives, any of which may have a face-to-face double. They liked their belt-buckles made of mirror images. Sometimes the maker included an odd mythological creature with a beak-like nose, which seems to have been unique to Ordos. (Other Inner Asian groups liked griffins – half-eagle, halfman – and winged lions, but that fashion did not spread to Ordos.) A favourite motif was the tiger – as the golden coronet showed – often shown eating a sheep or carrying a deer over its shoulder. There were no tigers in north China. The nearest ones were in the forests of eastern Siberia. It seems that to have a tiger on one’s belt-buckle was a statement of status and power.

  The buckles and plaques could also be used as gifts, which made them a sort of currency to assert equality of status (you see similar behaviour at English middle-class dinner parties today. Guests bring chocolates, wine or flowers as ‘house gifts’ which may cost more than their meal). Some were in the form of horse-decorations, which were usually discs attached to the reins and bridles. Wagons, too, were decorated, for example with little triangles set around the edges of wheels.

  Most of the bronzes come from graves, not just in Ordos but from hundreds of graves scattered across north China and Inner Asia. Some have even been found in Iran. Specialists try to make sense of the stylistic variants, tracking the trade routes and dating them within their 1,000-year span, between about 800 BC and AD 200, which overlaps the Xiongnu empire. In fact, one cemetery dating from the sixth to the fifth century BC, found in 1979 in Maoqinggou (Liangcheng County, Inner Mongolia, just east of Hohhot), is described by some scholars as the cradle of Xiongnu culture because its contents suggest that the people who used it were changing from farmers to nomads. It had seventy-nine graves, all simple pits, with 229 bronze plaques, fairly evenly distributed between men and women (though elsewhere belt-plaques were mostly used by men). It is tempting to use the motifs – dragons, griffins and other Animal Style images – as evidence for clan or tribal connections. But that’s not possible, because the plaques were made and traded too widely. In a system that suggests modern parallels, many were made for the Chinese market, with Chinese characters on them – ‘Chinese styles of execution with steppe iconography’, as the American scholar Bryan Miller puts it.5

  Graves were often destroyed by bad weather, as Wang Shun’s discovery of the golden coronet showed. In addition, the bronzes were made to be worn and attached to horses and wagons, and must have often been dropped. A century ago in Ordos, these items could be found lying about in the sand. Locals gathered them by the bucketful and sold them for a pittance. No one took much notice of them. In fact, only a few people, mostly foreigners, were interested in Chinese art and antiquities, which made the early twentieth century a golden age for foreign collectors. Explorers – Aurel Stein from Great Britain, Petr Kozlov from Russia and others – acquired vast quantities of manuscripts for practically nothing. As Mr Shark said ruefully, ‘Some people might call this robbery.’

  Among the collectors was the remarkable Isabel Ingram. Isabel was the daughter of an American missionary who had good connections with Puyi, the boy-emperor who had been allowed to stay on in the Forbidden City when China became a republic in 1911. At the age of twenty, Isabel graduated from the US, returned to Beijing and became tutor to Puyi’s new wife, the sixteen-year-old, missionary-educated, English-speaking Wanrong. A picture of them together shows them looking sweetly beautiful and delicate as porcelain. They enjoyed their similarities, and even swopped clothes to look alike. In her position as royal tutor, Isabel met many visiting scholars and officials. One of them was Horace Jayne, Curator of Oriental Art at the Philadelphia Museum of Art. Another was her future husband, William Mayer, military attaché at the US embassy. In 1924, the emperor and Wanrong were expelled from the palace, Isabel’s role as tutor ended and she got
a job back in the USA in the Philadelphia Museum. When her boss became director of the Pennsylvania University Museum, she followed, and began to publish scholarly articles on Chinese art. In 1930, she married William Mayer in Beijing, and the two began buying art. Although few people were interested in the bronzes – they did not even have a collective name – there was huge international interest in Scythian art, and she recognized that the Ordos artefacts shared motifs with Scythian ones. Criss-crossing north China and haunting antique shops, she and her husband gathered over 500 ‘Scythian’ bronzes and other items, most of which they sold to the Pennsylvania Museum in 1941.

  As experts came to recognize the value of these finds, Europe and the USA built their collections, buying them in small numbers wherever they could. The objects were not just found in Ordos and were made of gold, silver and tin as well as bronze, but from the 1950s they have been referred to as the Ordos Bronzes. Those found locally are now stored in the Ordos Bronze Museum. ‘We have a collection of about ten thousand,’ said Mr Wang. ‘People still find them today.’

  ‘What would one sell for?’ I answered my own question online later. Bronzes rarely come up for auction. When they do, they go for £500 and up. In 2015, a bronze of the sixth or fifth century BC sold for 123,000 euros.

  ‘Nothing,’ said Mr Wang. ‘It is illegal to sell them.’

  Which makes one wonder where the sellers got theirs.

  Who made them? No one knows. There are no accounts of their manufacture. A few carbon-14 dates suggest that the business was flourishing in 500 BC, which means it must have arisen considerably earlier, when pastoral nomadism and mounted archery began to mature as a way of life. That idea is supported by a reference in about 661 BC to two grassland tribes in what is now Inner Mongolia named Rong and Di, who had a reputation for violence. An adviser to the north-eastern border state of Zhao, Guan Zhong, noted that ‘The Di and Rong are like wolves, and can never be satisfied.’ Chinese historians often referred to them as hu, but this was no more than a generalized term for ‘barbarians’. They could have become a real threat, for a coordinated attack by an army of mounted archers could work on infantry like a chainsaw on a tree-trunk. What it took was a leader with a vision. That was what the Rong and Di lacked. The problem for a would-be conqueror was that nomads were famously independent. They were more likely to vanish back to their herds in a cloud of dust rather than do as they were told. Still, there was a threat to be countered.

  The first written evidence that a new way of life had emerged survives from Zhao. In 307 BC, a Zhao king named Wu Ling, overcoming opposition from his more conventional uncle, ‘decreed that the entire kingdom should adopt the hu attire, training his subjects to ride and shoot arrows from horseback’.6 The kit included a long slit coat for ease of riding (the forerunner of trousers) and no floppy Chinese-style sleeves that might interfere with archery on horseback. It worked. Zhao defended itself from nomadic raiders because it was half-nomad itself, and as a result built a powerful kingdom that lasted almost a century.

  No one recorded any further details of the Di and Rong. Sima Qian – a historian of the second century BC who will appear often in this book – also mentions them along with a further ten tribes of ‘barbarians’: ‘These nomadic tribes settled in the riparian valleys and mountainous pastures; each had its respective chieftain and was to each its own.’7 But from the second half of the fourth century BC, all of them were, it seems, in the process of being conquered, or absorbed, or in some way coming together to create a single tribe: the Xiongnu.

  What this new group called themselves in their unwritten language is unknown, but other cultures suggest their name. In 313 BC, a trader from Sogdia (in today’s Uzbekistan and neighbouring countries) abandoned some letters – that is, strips of bamboo that were used to write on – in a tower near Dunhuang. One of the letters complains bitterly of the destruction caused by people the writer calls the Xwn, Hun. That is the core of their name in both Mongol and Chinese. In the Latin script versions of both languages, they are often simply ‘Huns’, implying that they were the ancestors of Attila’s people who helped bring down the Roman empire in the fifth century AD – a controversial idea to which we will return in the final chapter.

  The most common name reflects Chinese usage. To represent foreign names, Chinese chooses syllables that sound vaguely like the original. Since each syllable has many written signs, Chinese commonly selects a character that suggests something suitable. ‘Hun’ in Chinese is represented by the sign that is transliterated in today’s pinyin system as xiōng. The sign means ‘terrible, horrible, bad, fear-inducing’. Then, for some unknown reason, a Han-dynasty scribe chose to add a second sign – nú, with an ascending tone, meaning ‘slave’. The two elements together – ‘bad slaves’ – must have seemed a suitable name for the ‘barbaric’ northerners. That’s what Sima Qian called them, writing in the late second century BC. Since China is the region’s dominant culture, the new rulers in Ordos and neighbouring steppes are almost universally known as Xiongnu (pronounced Shiung-noo), in the pinyin transliteration system. In the old Wade-Giles system, now superseded, they are Hsiung-nu.

  The first mention of the Xiongnu in Chinese sources refers to a date some time before the mid-third century BC. Sima Qian says they had been around for 1,000 years before that, ‘a vast period during which the tribes split up and scattered into various groups, sometimes expanding, sometimes dwindling in size’. Like all Inner Asian groups, leaders always tried to grow their clans, but if they succeeded their efforts evaporated on their deaths. There were no empires on the steppe before the Xiongnu. Genetic studies suggest that they were ‘probably a local development from within Mongolia dating back to at least the Bronze Age slab-grave burials (1200–400 BC).’8 If so, they migrated into Ordos and somehow by the late third century BC they had become a significant force. We have no idea if the Xiongnu were a separate tribe or the name of a confederacy or what other unmentioned sub-groups they absorbed. The few details given by Sima Qian suggest they resembled the Scythians in their habits. Like the Scythians, they turned the skulls of slain enemies into wine-cups, sewed their scalps into clothes or tied them on to the reins of their horses and mixed wine with the blood of an animal with a sword when swearing an oath. On their origins, Nicola Di Cosmo, of the Institute of Advanced Study, Princeton, summarizes the problem: ‘After several decades of debates, questions relating to the ethnic and linguistic identity of the Xiongnu are still unanswered.’9

  Here’s a possible scenario. Imagine the Xiongnu as the Chinese first heard of them – a small group of nomads in the pleasant pastures (as they were then) of Ordos. They are herders, leaving few archaeological records, no large tombs for instance, because their leaders were too poor to amass any riches for grave-goods. That would come later, in Mongolia. Life for them is a succession of seasons, herding, horse-raising and occasional raids on their settled neighbours to the south. ‘They move about in search of water and pasture, and have no walled cities or fixed dwellings, nor do they engage in any kind of agriculture,’ says Sima Qian. ‘Their lands however are divided into regions under the control of various leaders.’ There is a lot of intermixing and overlap between grasslands and farmlands. Herding, raiding, trading, nomadizing from season to season, intermixing: that was normality, for many unrecorded generations, until the 250s BC.

  Now they get their first proper mention, in the records of 244 BC, which look back several years, perhaps as many as fifteen. The Mirror of Good Governance (Zizhi Tongjian, the massive 354-volume tenth-century history by Sima Guang, which provides much of the information about the Xiongnu)10 records the story of the great Zhao general Li Mu, who ‘often stayed in the Dai and Yanmen [prefectures on the northern border] to protect them from the Xiongnu … he taught his soldiers to shoot arrows and ride horses, and he carefully maintained the beacon towers’ (in the Zhao frontier walls, which like those of other states preceded the Great Wall). His plan was to use the Zhao wall not for defence
, but for deception. ‘He issued a regulation that said: “If the Xiongnu invade the border to plunder, you must quickly enter the fortifications, and unauthorized capturing of enemies will be punished by decapitation.” ’ This went on for some years. As a result, the Xiongnu concluded that Li Mu was a coward, as did his own troops. So did the king himself. He replaced Li Mu with a general who had more aggression but less sense. ‘Every time they went into battle, they suffered many setbacks and had many people killed or injured. Therefore they could not cultivate the land or raise animals on the border.’ The king saw the error of his judgement, and recalled Li Mu, who accepted only on condition that he could use his former tactics: retreat to safety when threatened. Having re-established his reputation for cowardice, Li Mu prepared a great army of 1,300 charioteers, 13,000 cavalrymen, 50,000 infantry and 100,000 archers. When the Xiongnu, unimpressed by the numbers, attacked next, Li Mu feigned defeat and let them capture ‘a few thousand men’. The Xiongnu chief, now certain of victory, sent a larger force, on which Li Mu pounced from two flanks, ‘killing hundreds of thousands of men and horses … Ten years after this, the Xiongnu still did not dare to come close to the cities on the border of Zhao.’

  There is no need to take these figures seriously. The point is the Xiongnu were growing into a formidable force. A few years pass, and they begin to come more sharply into focus. They get their first named leader, Tumen. His title is chanyu. This is short for ‘Chengli Gutu Chanyu’. The meaning is obscure. Probably chengli is the equivalent of Tengri, the Mongol sky god. Chinese sources say that gutu means ‘son’, but more likely, given the Xiongnu’s cultural roots, it derives from the Turkic qut, meaning ‘good fortune’.11 No one knows the origin of the word chanyu, but the sense is clear: ‘Heavensent Supreme Leader’.

  Not supreme yet. The Xiongnu were subservient both to the Dong in today’s north-east China, and to the Yuezhi, a tribe ruling the west, in what is today Gansu. The Yuezhi form an important part of our story, but their origins too are largely a mystery. In a detailed study of them, Craig Benjamin, the Australian-born Professor of History at Grand Valley State University, Michigan, suggests that they migrated from southern Russia, arriving in Gansu around 2000 BC. They established themselves in Xinjiang and Gansu – today’s north-west China – and were recorded by surrounding cultures as traders in jade and horses. Their name in Chinese means ‘Clan of the Moon’, but what they called themselves is unknown. They had a king, and they probably spoke an Iranian language. Some scholars equate them with the Tocharians, who lived in the oasis communities round the great deserts of today’s Xinjiang, forming commercial dots on the early versions of the Silk Road. Later, they were expelled by the Xiongnu, undertook a vast migration westward and founded the Kushan empire, which between about 50 BC and AD 200 was a crossroads linking Rome and China (among others). In brief, the Yuezhi formed a wide-ranging, long-lasting, highly significant culture that was virtually unknown to the wider world because it had no narrative history in written sources. The Xiongnu dealt with an eastern sub-section, attracted to them as moths to a flame by their commercial links westward.