Defending Heaven: China's Mongol Wars, 1209-1370

Introduction and Acknowledgements

Victory is not glorious, those for whom it is glorious delight in killing human beings.

Those who delight in killing human beings will never control the realm. When there are mounds of dead one should weep with sorrow.

When one is victorious, observe the mourning rites.

Laozi, Dao De Jing

As an undergraduate I was warned by several tutors that it was well-nigh impossible for anybody wishing to complete a history degree at the School of Oriental and African Studies to escape the Mongols. They were right and I now find myself composing a fourth book in which they are the sine qua non. However, I also once again find myself writing more genially of their enemy than of the khans. It has been a consistent theme in my work that the Mongols were, as the authors of 1066 and All That would have had it, ‘not a good thing’. I recorded their absolute devastation of Persia and Iraq in the pages of The Ismaili Assassins and the blessed fact that they were stopped from reaching the Mediterranean by the Mamluk dynasty was a central theme of The Knights of Islam. They appeared once more as a bête noire in Sacred Swords as they choked the Tigris with bodies before moving on to massacres in Syria in an attempt to extinguish Islamic civilisation.

Mongolists would argue, I hope passionately–or perhaps dispassionately, as most of the major arguments for the Mongols being a constructive element of world history rely on a detached appraisal made through the long lens of history–against my opinion, but the rehabilitation, through books and film, of Chinggis Khan that is currently taking place would seem to suggest that my viewing him and his offspring as one of the worst man-made catastrophes ever to strike Eurasia will not harm the khans’ image too much.

The above very probably seems somewhat out of character for any writer of history or professional historian. It is integral to our line of work that we commonly recount the deaths and miseries of thousands in simple telegraphic detail before moving on to a more detailed dissection of the consequences of victories in the field or of a resulting Pax, whether that be of a Romana, Mongolica, Britannica or Americana variety. Engels was not too far from the truth when he wrote, ‘history is about the most cruel of goddesses, and she leads her triumphant car over heaps of corpses,’¹ I, however, feel that undertaking a too detached approach to history leads us into a moral void. I take Wellington’s observation that ‘next to a battle lost, the saddest thing is a battle won’ very seriously indeed.

Then there is the question of historical relativism. Following Benedetto Croce’s dictum that the historian’s task is one of ‘criticism, criticism and then criticism’ is vital for our craft if history is truly to offer modern man anything at all in the way of guidance. Accepting that any atrocity or even misguided governance can be framed as being tolerable by the standards of its time just will not do, for as Croce also said, ‘every historical judgement gives to all history the character of contemporary history, because however remote in time events thus recounted may seem to be, history in reality refers to present needs and present situations wherein those events vibrate.’² If we condone the sins of the fathers we run the risk of excusing them in ourselves, particularly if they might obtain what we consider to be desirable ends. Furthermore, it is dismissive of the humans of the past to suggest that we are any more civilised or less capable of accepting barbarism now than our forefathers were. Liu Ji, a Yuan Dynasty writer wrote that:

Weapons are instruments of ill omen, war is immoral.

Really they are only to be resorted to when there is no other choice. It is not right to pursue aggressive warfare because one’s country is large and prosperous, for this ultimately ends in defeat and destruction. Then it is too late to have regrets. Military action is like a fire–if not stopped it will burn itself out. Military expansion and adventurism soon lead to disaster.

The rule is ‘even if a country is large, if it is militaristic it will soon perish’.³

Then there is the question of how it was to live through profoundly destructive times. Discounting or not acknowledging the immediate effects on a society, and on men’s thinking, of brutality, indifference to suffering, incompetent or absent governance and inhumanity will not get us any closer to achieving the goal of every historian as dictated by Ranke: ‘wie es eigentlich gewesen’–show simply how it really was, and not how it looked later.

Even given the above, there still are undoubtedly dangers in a partisan approach to history but my ‘purely academic’ argument for its partial application in this work is that there is currently available no single-volume history of the Mongol invasions as seen from the standpoint of the Chinese dynasties they were unleashed upon. There are several studies of the Mongols’ military achievements and government, and if I had taken up a strictly neutral point of view in my recounting of the defence of China I would have risked merely creating a pale imitation of such superb books as Timothy May’s The Mongol Conquest in World History, David Morgan’s The Mongols and Morris Rossabi’s Khubilai Khan: His Life and Times. The virtue of creating a work that generally reflects the action from one point of view is that it allows for an understanding of perceptions. The perceptions of both the governing and of the governed, however erroneous they may have been and as difficult as they may be to reconstruct, are vitally important when we want to understand why some dynasties or governments last and others do not, and why some wars are lost that should have been won. Not to accept that any Chinese citizen or official would have attempted to make sense of the Mongol invasions and subsequent Yuan government through an application of Confucian morals and philosophy is frankly disingenuous, and the same must be said of every other culture that ever had contact with the Mongols. The fact that our primary sources are coloured by the often-unconscious influence of their authors’ milieux should not surprise us, and whilst we must seek truth from the facts presented to us by contemporaries of the events of the past we cannot escape the very basic idea that facts presented themselves to contemporaries as a series of impressions; this is how every human experiences their world and, more importantly for the historian, how they respond to it. It may be trite to comment that only hindsight is perfect but it is an important point for any historian interested in relating ‘how it really was’.

Much ‘quantifying’ of the Mongol invasions of both China and of the Middle East has been undertaken by historians, usually with the intent of rationalising the figures given by medieval historians. Generally speaking, the trend has been towards making far more conservatively sized mounds of dead than the original sources had portrayed. In some respects this is commendable. Historians have a duty to create hypotheses about the past: it is, after all, a social science that we are engaged in, and the question of magnitude is always a question we are likely to ask and be asked. But, even if we can ‘prove’ such suppositions as ‘each household should be multiplied by five to calculate the population affected’ or that ‘the square hectarage of the remains of a city could not possibly have supported as many people as the contemporary historians claimed were slaughtered there’, I would suggest that this is still not as useful, historically, as gauging the psychological impact of having nearly all your neighbours killed and next year’s crops destroyed. Whether the number killed was seventy thousand or seven thousand does not in fact matter; the question of magnitude of terror depends, in fact, on the numbers left alive, and the sources make it very clear that this number was a fraction of those who died.

Further to this, contemporaries of the Black Death, the single greatest killer ever to be visited upon mankind, commonly wrote not just of the volume of death they surveyed but of the fact that society’s mores had been degraded, and even hypothesised as to whether the breath, pus, sweat and stench of the ill had corrupted the souls and conscience of the healthy. Boccaccio’s Decameron was very clear on what damage mass death could inflict on those who survived:

E in tanta afflizione e miseria della nostra città era la reverenda auttorità delle leggi, cosí divine come umane, quasi caduta e dissoluta tutta per li ministri e essecutori di quelle, li quali, sí come gli altri uomini, erano tutti o morti o infermi o sí di famiglie rimasi stremi, che uficio alcuno non potean fare; per la qualcosa era a ciascun licito quanto a grado gli era d’adoperare.⁴

It is notable that the right degree of simpatico for Boccaccio’s words can only really be found in the 1620 English translation commonly accredited to John Florio, who entered into the task in a time not unused to sudden death and contagion. Perhaps we who write of the Mongols in comfort and security might take note of a how a man living in rather different times brings to us an account of what terror, whether biological or man-made, could do to pre-modern societies:

In misery and affliction of our City, the venerable authority of the Lawes, as well divine as humane, was even destroyed, as it were, through want of the lawfull Ministers of them. For they being all dead, or lying sicke with the rest, or else lived so solitary, in such great necessity of servants and attendants, as they could not execute any office, whereby it was lawfull for every one to do as he listed.⁵

We will return to this subject at the conclusion of the tale of the Yuan and how the collapse of their dynasty was recorded by Boccaccio’s peers in China.

Most Mongol historians certainly accept that terror was an acceptable practice in war for the khans (just as it is in the modern world, though now in our post-Douhet age it is delivered from the air rather than from the saddle), and if we set out to cause terror by killing enough of the populace to cow the remainder, then the question of whether the populace we slaughter is seven thousand or seventy thousand does not really matter. The end is terror, the means are proportionate to that and equally so is the fear, discord and loathing that terror is likely to arouse in populaces subjected to it.

I hope I have not–and many times my draft work was rescued from it by peer reviewers–fallen into a trap that far greater historians than I have been lured into. Even J. Saunders claimed that the Mongols suffered from ‘blind unreasoning fear and hatred of urban civilisation’.⁶ I thank Dr David Morgan for drawing to my attention to this back in 2007 and to the work of Professor Joseph Fletcher, who noted that a major cause for the huge destruction that the Mongols wrought was that they came, ‘too fast’. Fletcher was in fact discussing the damage done by the Mongols to Persia when he wrote that they had no time to acculturate themselves to the desert habitat and continued with ‘attitudes nurtured in the East Asian steppe: disdain for peasants, who like the animals that the Mongols herded, lived directly off what grew from the soil . . . With the steppe extortion pattern in mind, the Mongols did violence and used terror, reinforced by their ideology of universal dominion, to induce their victims to surrender peaceably.’ Fletcher argued that the Mongols, ‘came to understand settled society . . . quickly’, but, ‘by that time the Mongolian juggernaut had done its dreadful work’.⁷ My argument is that something very similar happened in China. I do not argue that the Mongols were inherently savage–or no more so than any other peoples ancient or modern who had been instructed that they might please their generals not with prisoners but with killing and burning –but rather that the means they were forced to apply to the conquest of China, as they were pushed so hard by the Song, were terrible and then untameable in the post-conquest period, and though they may even have wished to govern in the ‘style’ of a Chinese dynasty, the vested interests within their state–the very interests that had brought them to power–were by this point ungovernable. They had opened a Pandora’s box by taking on China, and the infrastructure of government that they brought with them into the Middle Kingdom was not up to the task of governing what they had won. Song’s long resistance revealed the fault lines in Mongol government, and the Yuan Dynasty was in trouble so early on after finally defeating the Song that its leaders never had time to adapt or to tame the forces that would soon enough begin to tear it apart. With this in mind, it is notable that the Ming Dynasty, albeit brought to power by force of arms and through an internecine civil war, grew up within the existing cast of Song government and lasted nearly three centuries.

In my recording of China’s confrontation with the Mongols I have recorded the Song Dynasty’s long war against the khans in more detail than that of the Jin and Xi Xia. I chose to do this because the history of Chinese civilisation is more extensive and complex than that of any single Western nation, and the profusion of names and the sheer ‘density’ of Chinese civilisation and society risks producing a book that is both vast and indigestible for Western readers. Furthermore, Song was China. It is no accident of editorial judgement that Jin and Xi Xia, are placed along with the Mongol Yuan Dynasty in Volume 6 of the magisterial Cambridge History of China: Alien Regimes and Border States. All these dynasties took on ‘Chinese characteristics’ but all three were transmitters of this tradition whilst Song was both a vessel of the past achievements of China and a continuing locus of production for the vital elements of Chinese civilisation. Certainly the Song viewed all these aliens in the same light; the Jurchen of the Jin and the Tangut of Xi Xia were no better than the Mongols of the Yuan in their eyes, and that essentially xenophobic perception of its enemies was certainly part of Song’s undoing.

It is also no accident of history that the rebellions which finally toppled the Mongols in China started in what had been Song lands and not in the former lands of the Jin. Song China’s long resistance to the Mongols, and the memory and myth of it, was fundamental to the Mongols’ eventual fall from power in China and to the birth of the Ming Dynasty.

Acknowledging mentors and friends is always the most pleasant part of writing any book and my thanks go out once more to Dr David Morgan, who first introduced me to the Mongols ‘at the safe distance of eight hundred years’, and to Dr Brian Williams, who is, I am sure, right now somewhere in Central Asia discovering the history of tomorrow. Their ongoing kindnesses and words of wisdom and encouragement are beyond value. I would like to thank Professor Morris Rossabi for replying to an unsolicited and decidedly cheeky email that I sent him from the depths of depair whilst trying to structure this book. The fact that the good professor responded with a full, thoughtful and insightful message whilst travelling hotfoot beween Ulan Bator and Tokyo speaks volumes both for his goodness and for his encyclopedic knowledge of the subject.

The generosity of John Man appears to be as limitless as the oceanic Mongolian grasslands of which he has written so eloquently in his marvellous renditions of the lives of Chinggis and of Qubilai Khan. The Foreword he has gifted for this book is every bit as insightful and vibrant as the piece he fashioned for my first book. Betsy Kohut of the Smithsonian Institute’s Freer and Sackler Galleries has also been unstinting in her kindnesses and invariable in her expertise in the provision of images for both this book and for my previous Middle East ‘trilogy’.

For keeping me on the straight and narrow I would like to thank my editor Kate Baker and my anonymous peer reviewers who steered me back again and again from a course of ruin and who can bear no responsibility for any of the errors that may remain. For continual kindnesses I would like to thank Jane, Liz, Wayne, the two Marks, Minna, Wael, Boris, Svetlana, Di, Melinda, Mandie, Chris, Jan Marie and Brent.

Simply priceless is the love, understanding and intellect of my dear wife Michele, she is also sine qua non.

1

Heaven Inverted China on the Eve of the Mongol Invasions

See the Southern Gate of Heaven, Deep Green, Crystalline, Shimmering Bright, Studded with Jewels.

On Either Side Stood Scores of Heavenly Marshals, Tall as the Roof Beams, Next to the Pillars,

Holding Metal-tipped Bows and Banners. All Around Stood Gods in Golden Armour . . .

Wu Chengen, the Monkey King’s first view of Heaven,

The Journey to the West, c.1580

China’s ‘Mongol problem’ was not unique to the thirteenth century. Many a Chinese dynasty had expended much revenue and blood countermanding the threat of Turco-Mongolic tribes from the north. Indeed, this dated from far back into the second century BC with the invasions of Han Dynasty lands by the Xiongnu, who may or may not have been the same individuals who were known in Europe as the Huns and who would one day be led by Attila to the gates of Rome.

Chinese policy, in brief, was constructed to a degree of appeasement called heqin or ‘harmonious kinship’, which entailed payments to the barbarian,¹ trading with him through frontier markets, and making the outlaw into an in-law through the giving of imperial brides. Along with this conciliation of the barbarian went construction of defensive lines, and agitation and intrigue among the northern tribes to foment wars among them and thereby to divide potential confederations. There was also recruitment of Central Asian cavalry to serve in Chinese armies and the application of military science in the form of early warning stations and systems for mobilising impressively large forces for the defence of the state and for punitive expeditions. The application of this policy over the period of more than a millennium was, give or take a few glitches, generally a successful one, but in the Tang Dynasty it began to crumble.

Arguably the Kitan, a Turkish confederation lying to the northeast of the Tang state, began their movement into Chinese territory as a direct and slightly panicked response to the chaos that the Tang state had collapsed into by the 870s. The Kitan were forced to undertake an occupation of Chinese territory as the Tang court had become a powerless vessel among a raft of more than fifty Chinese warlord states and was incapable of raising tribute in the form of bolts of silk and silver to pay off the Kitan and of ensuring its safe transport to the Kitan court.

The Kitan occupation was unusual, as Turco-Mongolian confederations aimed, generally speaking, to exploit China through the seeking of tribute, often disguised as ‘gifts’ by the imperial courts, through plundering Chinese wealth and through carrying off manufactured articles that could then be sold on through the Central Asian trade system. Indeed, before the establishment of their ‘state’, it was common for the Kitan to build ‘Chinese’ cities in their own lands as centres of production and commerce. Captured and refugee Chinese artisans were commonly relocated to these centres.²

Occupying Chinese territory and becoming one of the settled, as opposed to one of those who fed off the settled, was always likely to be a dangerous policy for any steppe tribe, as it would strain its political system. The organisation of these steppe tribal confederations was based on a very simple principle of exploitation of a cowed state and not on conquest of that state per se, and certainly not on the careful management, administration and husbandry of a settled state. For a tribal confederation leader to demand that his followers give up the saddle and bow and take up the administrator’s chair, as we will see, was always likely to cause dissent among his own people. Furthermore, a steppe tribe, being made up of nomadic cavalry capable of striking randomly and quickly and at multiple locations, was not suited to controlling a region, and if it did take on garrisoning and consolidation as military tasks it sacrificed its very essence. Indeed, great steppe politician though he was, Chinggis Khan’s invasion of China was arguably one of history’s greatest political blunders.³ Any chance of longevity for the khan’s steppe empire was essentially destroyed when he embroiled his nascent state in the conquest of China.

By the 890s, the Tang emperor had become a puppet of northern warlords who had formerly served as his generals. As China disintegrated, the Kitan moved into Chinese territory, and by 947 they had settled and formed their state in modern-day Inner Mongolia and taken the Chinese dynastic name of Liao. The beginning of the end of the chaos that had gripped China for most of the tenth century occurred a few years after the formation of the Liao state. China’s slow march back from anarchy began at the Battle of Gaoping in 954, which halted the Kitan and Northen Han’s invasion of Shanxi. This battle was part of a campaign that brought together a small group of commanders who would form the nucleus of the nascent Song Dynasty. The man who became the first Song emperor was General Zhao Guangyin, who after the battle was promoted to grand commander of the chief of the palace troops. It was in the palace guard that he formed the military clique that would bring him to power.

In February 960 the new Song Dynasty, having unified much of China, established its capital at Dongjing, modern Kaifeng. The verdict of one historian, that ‘Chinese Empires were built slowly at immense cost in blood, and a lot of its history is not dynastic at all but the chaos in between’⁴ seems justified if we review the bloody and destructive wars that the Song waged to put the former Tang lands under Song control and to solidify the border with the Liao.

That the Song army was a capable force at this juncture was shown in December 964, when two columns of thirty thousand men marched into Sichuan province over high mountains in the dead of winter. Meanwhile a further column to their east forced the Yangzi River defence line via pontoon bridges with a mass infantry attack under General Pan Ai, possibly the first time this had been done by any army on such a vast scale, to force the capitulation of the Shu kingdom. In 970 the army also showed itself capable of operating with the close support of a riverine navy when the Southern Han kingdom was defeated.

A peace accord with the Liao that was struck in 974 allowed Zhao Guangyin, who had taken the throne name Taizu, to reduce the remaining small kingdoms that the Tang empire had shattered into. The Southern Tang and Northern Han were conquered, and in 979 the Song turned their attentions to the ‘lost’ sixteen prefectures, which had been part of the Tang Dynasty’s lands before being lost to the new Liao state. Unfortunately, despite the army sent to undertake this task being led by General Yang Ya ‘the invincible’, it was badly mauled at the Battle of the Gaoling River, just west of modern Beijing, with the Song Emperor Taizong fleeing the battlefield, severely wounded, in a donkey cart. This unlucky ruler then faced a rebellion in his capital and he was forced to exterminate much of the royal family in order to bring it to an end.

The consequences of the year 979 on the Song Dynasty were to be immense. The crushing defeat of Gaoling and the Song’s failure to annexe the sixteen prefectures became a festering wound that blinded the Song court to its own interests in almost every subsequent strategic decision it made about the northern border. That the sixteen prefectures were Song lands and the Liao were illegitimate monarchs over them became a virtual doctrine of the Song court. This obsession with the ‘lost’ prefectures grievously misguided the Song emperors and their counsellors at key junctures during the later contests with the Jin Dynasty and then with the Mongols. Furthermore, the near annihilation of the imperial family by Taizong also required the emperor to create an extensive bureaucracy to assist him in ruling. The Chinese government exam system predated the Song, but Taizong and his descendants formalised it as the only route into government and professionalised the roles of bureaucrats. While there were undoubted benefits to this system of government, a major drawback, as we shall see, was that the Song court often acted like a veritable ivory tower: narrow in its vision and often almost deliberately cut off from the realities and dangers that the state faced.

In the novel Der Glasperlenspeil Hermann Hesse created a fictional city named Castalia, where scholars played the glass bead game, an abstract game of pure intellect. Similarly, the intelligentsia of the later Song court did historians a great service to posterity in that they anthologised and thereby saved for us much of China’s military history prior to their dynasty and maintained censuses of a standard that would challenge many modern states to replicate. Indeed, Sima Guang’s monumental ‘Comprehensive Mirror for Aid in Governments’ (zishi tongjian) was produced within the court’s cloistered walls in the late eleventh century. But this vast body of work–the censuses aside–is equally a great potential trap for the historian. There was no Caesar, Guderian or Musashi to recount the achievements and failures of the Chinese military in the field, and the civilian officials assigned to the History Office (shi guan) rewrote, condensed and altered records to suit the court’s world view and to make recorded events fit the accepted configuration of dynastic Chinese history and follow the rules of writing historiography as laid down during the Han Dynasty by the great Sima Qian. Their creations were, then, in many ways far removed from the reality on the ground.

There was certainly plenty of material for these Song historians to draw on. China was a paper-rich society, and records of eyewitness accounts, proclamations and appointments and rewards abounded. Unfortunately, once the court historians had steamrollered much of the reality out of them to create a record in harmony with the court’s view, even documents such as the lubu or announcement of victory and xingzhuang or government accounts of conduct, akin to being mentioned in dispatches, lose much of their primary source credibility. The lubu give us quite exact dates and locations, but often grossly overestimate the numbers of the enemy. They also record the number of enemy killed without recording Song losses.

In simple terms, the shidafu–the scholar-officials who recorded Song’s martial history–give us very little military detail because they were not military men. Troy had Homer, Antioch had the anonymous soldier-author of the Gesta Francorum, but the Mongol siege of Xiangyang, despite lasting some five years, had no directly involved author to record the Song’s defence of the city. The lack of military authors is not surprising, given the divorce between the Song’s general staff and the world of the bureaucrats of the court. A Song general could not aspire to be of the literati who governed and recorded the actions of the state as the Roman Ammianus Marcellinus had done. That we have virtually nothing recorded on battle, tactics or details of combat is a result of the fact that those who led in the field did not write.

The men of the pen also set a genre through their hegemony over the culture of the Song state, and their writings are loaded with conventional literary expressions rather than fact. Even Song poetry avoids any narrative about the battle and the clash of arms, whilst in the