Empire of Horses

by John Man

  He proceeded not with a head-on assault, but with the plot suggested by his pro-war adviser – a plot so complicated that an experienced tactician should have told the emperor it was doomed to fail. An official named as Old Nie Yi was ‘fired’, a ruse that allowed him to ‘defect’ to the Xiongnu and claim he would murder the top men in Mayi, 100 kilometres south of the Great Wall. This (he claimed) would allow the Xiongnu to pillage with impunity. Old Nie Yi returned, killed some condemned prisoners, hung their heads outside the city wall, and sent a message saying that one of them was the governor’s. It was, of course, a trap. The Chinese army lined the hills on either side of a valley that was the obvious invasion route, ready to pounce. But the Xiongnu realized what was up 50 kilometres from the town, as Sima Qian relates with his usual flair for the dramatic (and flagging up our usual suspicions about his agenda and how he could possibly have known the details, like the chanyu’s exact words – in what language, one wonders).

  The chanyu, leading 100,000 troops, noticed that, although the fields were full of animals, there was not a single person in sight. Growing suspicious, he attacked one of the Great Wall’s beacon towers, perhaps to see if it was manned or not. By chance, an official who knew of the planned ambush had taken refuge there.

  When the chanyu captured him and was about to kill him, he told the chanyu where the Han forces were. The chanyu was very alarmed and said, ‘Just as I thought!’ And so he led his troops back. When he had gone beyond the border he said, ‘I captured a military official. It was Heaven, yes Heaven, that sent him to warn me.’

  Since in Chinese philosophy Heaven is on the side of the virtuous, Sima Qian was putting the chanyu forward as morally superior – to whom? The implication is that it was the emperor – Sima Qian’s own emperor. Men had been executed for such treasonable words. But he had a point. Someone had to take the blame, someone who had advised the emperor to back a disastrous strategy. The imperial finger pointed at Wang Hui, leader of the pro-war faction at court and the commander of the force that failed to engage. He committed suicide.

  The peace faction sensed victory. Three scholars wrote memos to the emperor pointing out the disadvantages of war: its expense, the sorrow, the hardship, the despair, the dangers of rebellion.

  Still the emperor vacillated. It took him another four years to decide, with roads being built as if to prepare the infrastructure for all-out military action, but with the border markets still open for trade, as if peace was everlasting.

  Then, finally, the events of 129–8 BC began to undermine the log-jam in his mind. In a series of tit-for-tat campaigns, the violence slowly escalated. A raid on Shanggu (in the northeast) drove Emperor Wu to order an attack by four battalions of 10,000. The first was victorious, capturing or killing 700. The second failed to make contact. The third was defeated. The fourth, under the famous Li Guang, the so-called ‘Flying General’ with the long arms who shared his men’s hardships, was also defeated, and Li Guang himself briefly captured, until he managed to escape. The next year the pattern repeated itself: an attack, a response, a few defections from leading Xiongnu and a mass defection by one of the Xiongnu’s vassal tribes.

  In 127 BC, a campaign into Shuofang (to the north-west of Ordos, just outside the Great Bend of the Yellow River) and Ordos retook lands once conquered, and then lost, by Qin, thus ‘re-establishing the old border that Meng Tian had created in Qin times, and the river was fixed as the border’. It was in this context that Sima Qian placed his criticisms of Meng Tian, saying that he had cut through the ‘veins of the earth’, ignored the suffering of the common people, and failed to stand up to the emperor. All of this was a backhanded way of criticizing Wu’s imperialist agenda, pushed forward by his general Wei Qing – invasion, the building of fortifications, road-building – a mirror image of what the First Emperor and Meng Tian did in the same area.

  By this time, Emperor Wu had given up on the Yuezhi. But the information about peoples and established trade routes that Zhang brought back in 126 BC would change his country’s history, turning attention westward to cultures of which the Chinese had had no previous inkling – thirty-six in the Western Regions alone – all leading like stepping-stones along what would eventually be called the Silk Road, to India. Wu was happy to learn that, in military terms, these powers were ‘feeble at best’. So, given a decent army, ‘all the lands within the realms of the four seas and all the people therein would be under the beneficence of Han’. In particular, Zhang told of wonderful horses raised in Ferghana, the fertile valley of what is now eastern Uzbekistan. Tall, standing at 16 hands, these ‘celestial horses’ – or ‘blood-sweating’ horses, as they were known, from the pin-prick wounds caused by local parasites – were just what Wu needed to strengthen his cavalry. Such tempting prospects would inspire China’s conquest of Central Asia and in effect determine its modern borders. From the seeds planted by Zhang sprang both the great trade routes joining China and the West, and their defence: the future Silk Road and the Great Wall’s western extension.

  In the year of Zhang’s return came a swift response from the Xiongnu, despite a troubled succession – Gunchen’s death, a power grab by his younger brother Ichise (Yizhixie in Chinese) – and despite the defection of the true heir and two Xiongnu princes. That year saw a catalogue of assaults: a provincial boss killed and 1,000 captives taken in one raid; another 1,000 in a second; more thousands in a third, fourth and fifth; a full-scale invasion across the Gobi that took 15,000 prisoners (well, a lot anyway); a counter-attack; a counter-counter-attack, with many more prisoners taken and unlikely numbers killed on both sides. This all happened in 126 BC.

  This was not the kind of stability Wu could tolerate, with its occasional raids, expensive appeasements, and the eating of much humble pie. No emperor of a unified China could put up with a ‘barbarian’ neighbour who was liable to launch raids whenever he felt like it. The only answer was to set the boundaries of China ever wider. In the words of Owen Lattimore, Mongolist, linguist, traveller and historian, who was forced to leave his native America for Britain during the McCarthy anti-Communist witch-hunt of the 1950s, Wu needed ‘a closed economy, a self-sufficient world, and an absolute Frontier’.4 It was in pursuit of this absolute that Wu finally escalated the rumbling rivalry into full-scale war.

  He had to grapple with two problems. The first was the grasslands of Mongolia, which started just north of the Yellow River and ran eastwards and northwards, fading in the wastelands of the Gobi and picking up again to the north. The other problem was the oasis states – the thirty-six minikingdoms ruled by the Xiongnu – of the Central Asian badlands, which lay westwards. The difference between them was this: the Gobi and the steppe could not be conquered and held; the oasis kingdoms could. Once the Chinese had taken them, they could run them like their own cities, whereas nomads could benefit from them only by giving up nomadism. The war aim that Wu adopted, therefore, was to ‘cut off the right arm’ of the nomads (the west, since the dominant direction for nomads was southward) – i.e. to pick off the tribal kingdoms one by one and garrison them in order to deny them to the Xiongnu. With the Western Regions in Chinese hands, it would be possible to isolate, invade and destroy the Xiongnu.

  There were two keys to success in this venture. The first was Ordos, reoccupied a few years before in 127 BC. The second, starting only 150 kilometres west of Ordos, was the narrow stretch of land running westward through what would later be known as Gansu. The Gansu Corridor, also known as the Hexi (hé xī , ‘river-west’, i.e. west of the Yellow River) Corridor, is another geopolitical keystone, crucial to understanding China’s relations with Inner Asia. The Corridor is hemmed in by the Qilian Mountains to the south and deserts to the north, with icy rivers from the Qilian’s snowy heights forming fine pastures down the middle. It was this bottleneck, only some 25 kilometres across at its narrowest point, through which nomads galloped to invade north China from the west. From now on, Chinese emperors imbibed a great truth with their mothers’
milk: Whoever wishes to rule China must rule the Gansu Corridor.

  To close off this open frontier, Wu had only one option: total, all-or-nothing commitment to a range of tactics, all interlinked, all leading step by inevitable and very expensive step to an extension of the Great Wall. He had the manpower (a population of 60 million, a million-strong conscript army, some 10–13 million available for forced labour). He had the firepower. He needed horses by the tens of thousands, and these would have to be raised in China, or bought. There had to be an enduring relationship with the oasis kingdoms to the west, which meant great expenditure on gifts, especially silk. There had to be invasions of Xiongnu lands, and a conquest. Trade goods had to flow. There could, perhaps, be an invasion of the steppes, but that would be of no use unless the borderlands were secured. This meant garrisons, which would have to be fed, which meant sending in colonists to grow grain. And there would have to be fortresses, and overnight places, and houses, and lookout points, and an administrative apparatus to supervise the whole thing. This was the iron logic that drove Wu’s decision to conquer the west.

  In 121 BC, Han attacked in the far west and the centre (Dingxiang, the region adjoining the Yellow River where it turns south near Hohhot), two vast invasions, each with 50,000 cavalrymen and ‘several hundred thousand’ infantry complete with baggage trains and fodder. Commander-in-chief and commander of the central force was Wei Qing. In command of the western force was Huo Qubing. The two need an introduction, because they, and especially the young Huo, are about to play vital roles in the story of the Xiongnu wars.

  The most remarkable thing about Huo was that he was just nineteen, all the more remarkable because of his unusual origins, to understand which means looking into events in the palace of the Princess Pingyang, elder sister of the emperor. It is an extraordinary sequence of random events, strong evidence for those who think history is all about luck.

  This is the story in brief, without names:

  A servant girl in the princess’s palace had a baby girl. The husband died. The mother had two more affairs and produced two more children, a boy and a girl. The first-born girl became a song-and-dance entertainer, the other two grew up as servants. Eventually, the princess put on a show for her brother the emperor, who was unhappy with his empress, a woman eight years his senior and childless, which to his embarrassment was blamed on his impotence – bad news for him, given that he was young and politically weak. During the show, the first-born girl caught his eye, and then his heart. He took her into his palace, along with her half-brother to act as a stable boy. The empress was bitterly jealous, made the girl’s life a misery, and ensured she was employed as a maid, not a concubine. A year later, depressed, the girl applied to leave the palace. By chance, the emperor saw her tearfully waiting her turn to leave. By now, he had gained in authority. Re-smitten, he told her to stay, took her to bed, made her pregnant, and thus proved his virility. The bitter empress sought revenge by kidnapping the pregnant girl’s half-brother, who was saved by friends. When the emperor heard of the incident, he demoted the empress, made the girl his No. 1 consort and the half-brother his chief-of-staff. What meanwhile of the other girl, the half-sister? Another palace affair, another illegitimate child, a boy.

  Now to summarize where we stand, and add the names. All the main characters belonged to a single family, named Wei. The emperor’s pregnant concubine, Wei Zifu, became Empress Zifu, and remained the emperor’s consort for the next forty-nine years. The ex-stable-boy half-brother, Wei Qing, became one of the two most famous generals of his generation. The son of the other sister, Wei Shaoer, was named Huo Qubing, who as a teenager was taken under the wing of his uncle General Wei Qing. At the age of seventeen this ‘exemplary horseman and archer’5 made a name for himself fighting the Xiongnu, capturing the chancellor and two relatives of the chanyu, which is why, two years later, he was leading an army on his own account with the poetic rank of Agile Cavalry General, and on his way to match his uncle in fame. Not that he was a popular figure, because he was noted for his lack of care for his men. Once when his soldiers needed rest, he ordered them to make a field where he could watch them play taju, a game with a leather ball stuffed with feathers. Sources do not record the rules, but it sounds like a sort of football.

  The 121 BC campaign westward marked a turning point, because Huo Qubing was aiming to clear the way through the Gansu Corridor. He left Ordos by its far north-west corner, where the Yellow River is deflected by mountains into a right-angle bend. It is a region of lowlands and meandering streams, as if the river hasn’t quite made up its mind to head east. He made his exit through the Gaoque Pass, guarded by a mountainside fortress, the ruins of which can still be seen today. He ‘trampled five Xiongnu vassal kingdoms’, marched on for 500 kilometres, captured a Xiongnu chancellor and several commanders, and killed or enslaved 32,000 Xiongnu (as usual an unbelievable figure from Sima Guang that simply suggests victory). He returned with his prisoners to rich rewards and a towering reputation that, two years later, would win him a special place in the history of the Xiongnu wars.

  His victory had quite an impact. Five more Xiongnu princes defected, all being granted the title of marquis, along with ‘several tens of thousands’ of ordinary Xiongnu, who were rewarded with cash, transported to the capital in 20,000 carriages (not really: that’s two wan, short for ‘many’) and supplied with horses requisitioned from local Chinese. One of the defecting princes, Hunye, settled down near Chang’an and went into business, to the fury of a top official named Ji An. In summary, he said: We’re draining our coffers to service Hunye, just because he surrendered! We’re honouring him as if he’s the Son of Heaven! We should be using his wealth to repay ourselves for the Xiongnu attacks! The Emperor disagreed. Ji An was ‘babbling nonsense’. It was better to honour enemies to encourage others to defect.

  And to protect conquests with the Wall. It was not that the Wall itself would stop the Xiongnu – it never had in the past – but its soldiers and settlers might. Besides, the Wall was a mighty project that would allow him to control his own population. Hence the 1,000 kilometres plus of rammed earth that ran westwards from the Yellow River. In Wu’s day, this was a frontier marked by hard-packed, whitewashed earthworks (yes, whitewashed – that’s why Mongolians call it ‘the White Wall’). Today, from car or train, the weathered, saw-toothed remnants mark nothing but itself.

  Several years ago, I had a chance to get close to the Han-Xiongnu war. My guide, Xu Zhaoyu, or Michael, as he called himself, was the best guide in Gansu. He told me so, several times, and I was inclined to believe this was more than a boast, because he had a knowledge of history that was astonishing.

  We headed north-west out of Zhangye, which guards the Gansu bottleneck, to a site that Michael said was important in the struggle between the two empires. Autumnal corn lay drying in the fields and the radio blared pop songs – number one at the time was ‘Ta bu jidao’, ‘She doesn’t know (I love her)’. We turned down a farm track, and parked by a field of chilli peppers. Michael led the way through the peppers and another field of corn stalks awaiting harvest.

  Ahead loomed a sand dune, dotted with camel thorn and red willow bushes, above which rose an earthen wall. This, Michael explained, was once a Xiongnu fortress-city, their advance base, dominating the Gansu Corridor. We climbed, and the full thing came into view: eroded walls making a huge square, 250 of my paces on each side, with the stub of a guard-tower in the north-east corner. This was a revelation to me. I had known of the fortified town of Ivolga, but that was way north, in southern Siberia. I had no idea that they had fortresses on the border, of which several others have since been revealed by archaeologists. All (like this one) were built close to springs or rivers. They were probably maintained by prisoners and defectors used to a settled lifestyle and able to grow food enough for themselves and their Xiongnu garrisons.

  In this case, Michael explained, the river was the Black Water (Hei Shui), which flowed from the heights of the Qilian M
ountains, and still does, though with a different course. The Xiongnu must have thought they were there for keeps: fine farmland, their own river, a stranglehold on trade through the Corridor. But they had reckoned without Emperor Wu and his brilliant young general Huo Qubing.

  It was Huo who assaulted the Black Water fortress. He at once saw the weakness of the Xiongnu position: there were no defences right across the Corridor, and the fortress was utterly dependent on its water supply. So he diverted the river – easily said, but an immense engineering achievement of which nothing is known – isolated the fortress, destroyed it, and moved on into territory once occupied by the Wusun, before the Xiongnu kicked them out.

  Emperor Wu was so delighted with the victory that he sent Huo a huge flagon of wine, which caught up with him 150 kilometres further on, in the town now known as Jiuquan. Huo said that it was his soldiers, not he himself, who deserved the wine, so he poured it into a spring to share it with all his men. Though he was noted for his utter contempt for the lives of ordinary soldiers, this gesture entered folklore. Jiuquan means ‘Wine Spring’, recalling his act for today’s residents and tourists.

  Huo went on to follow the Black Water northwards to what was then Lake Juyan, part of a vast and well-watered delta where the river vanished into the desert. It has all changed now. Lake Juyan has been a dusty plain since the mid-twentieth century, though the lake’s last remnant still appears on maps as Lake Gaxun. In Huo’s day, the lake lapped another of the Xiongnu’s border fortresses, which bore the same name as the lake, Juyan, where he besieged a Xiongnu force under the Xiongnu general Xiutu (or something similar in Xiongnu). Somewhere nearby was the chanyu, because Huo seized a golden statue that had been central to the worship of Tengri, the Xiongnu’s highest deity. So Sima Guang tells us, though this is the only mention of a focal point for the religion and the details of ceremonies are unknown. A dispute between Xiutu and another Xiongnu general, Hanye, ended with Hanye killing Xiutu and surrendering with both their armies. Sources claim Huo’s army killed tens of thousands, figures that are beyond literal belief, being as always shorthand for ‘lots and lots’. The chanyu, down but by no means out, escaped, returning to his tent-city 750 kilometres to the north, beyond the Gobi, in the heart of Mongolia. As a result, wrote Sima Guang, the whole region from the Gansu Corridor into the Taklamakan Desert ‘had become a no man’s land, in which there were no Xiongnu to be seen’.