Song Empire

Who more worthy

formationDrafted from Boulger, The History of China (1898)

The Tang fell in 907, and for fifty-three years no government held all of northern China for long. Five dynasties succeeded one another in the north alone, each installed by an army that had just finished unmaking the last one, while the south split into a scatter of separate kingdoms. Zhao Kuangyin, born near modern Beijing to a family with several generations of minor military office behind it, served in the guards of the Later Zhou, the fifth and last of the northern dynasties, and distinguished himself in its wars of reunification — on one occasion, Boulger records, his presence of mind alone saved the day at the battle of Kaoping.

The task left half finished

The Later Zhou's ruler, Chai Rong, had begun the work of putting the north back together and died with it incomplete. Zhao Kuangyin, by then a senior general, took it up. The country, Boulger writes, "prayed for peace, and was anxious to give all the support it could to a man acting for its interests" — public spirit had gone dead during the years when the throne had been "the lottery of soldiers," changing hands too fast for anyone to feel invested in its survival.

The moment came while Zhao Kuangyin was on campaign. His own troops surrounded him and put the matter plainly. "The Empire is without a master," they told him, "and we wish to give it one. Who is more worthy of it than our general?" Whatever the choreography behind it — and a soldiery announcing to its commander that it has already decided to make him emperor is not, in any army, a wholly spontaneous event — Zhao Kuangyin accepted, returned to the capital, and took the throne without having to fight for it. Signs, his contemporaries said, had already appeared in the heavens confirming that Heaven willed it.

The first acts

He named his dynasty Song, took red as the imperial colour, and styled himself Taizu. His first measures were a general pardon and a proclamation, sent even into provinces still defiant of the new centre, ordering the restoration of law and the preservation of domestic peace. He restored the privileges of the scholar class, stripped from them during the preceding disorder, and — though not himself a learned man — funded and protected learning on the plain conviction that it served the state. He kept the doors of his own palace open day and night, wanting his house to resemble his heart, "which was open to all his subjects."

National Palace Museum, Taipei · Boulger, The History of China (1898) · Unknown · Public domain

To the army that had made him he applied the same instinct for institutionalising what had been personal. Where the Five Dynasties had run on the loyalty of individual soldiers to individual commanders — the exact mechanism that had put five different men on the throne in half a century — Taizu built an examined, promoted, salaried force answerable to the office of emperor rather than to whichever general held its immediate allegiance. Officers had to demonstrate physical fitness and a working knowledge of military subjects before promotion; a soldier's advancement ran through the state's own ranks, not through a warlord's household.

Reunification, city by city

The task that remained was the actual reconquest of a divided country, and Taizu pursued it province by province rather than in one campaign. Lieouki Yuen, Prince of Han and the last of the great regional strongmen to hold out, weighed the odds, judged further resistance futile, and surrendered his forces to the emperor's camp in person — absorbed afterward as a minor prince at court rather than executed, a mercy Boulger is careful to note was not Taizu's universal practice but was characteristic of it. By such accumulations of siege, negotiation and outright submission, the fragments Zhao Kuangyin had inherited from a hundred years of fracture were reassembled into a single administration for the first time since the Tang's own decline.

Boulger, The History of China (1898) · Qian Xuan · Public domain

The reunification brought a fresh problem into view immediately: the reassembled Chinese state now had a direct, long border with the Tartar kingdom of Liao, which had been meddling in Chinese affairs for a century and had no intention of stopping because the meddling now had a stronger neighbour to answer to. That confrontation, and the peace it eventually produced, would occupy the Song's founder and his immediate successors for decades — but it was a problem a reunited empire could at least face as one government, which was the entire point of the work Taizu had just finished.

Dangerous cabals

expansionDrafted from Boulger, The History of China (1898)

Renzong came to the throne at thirteen in 1022, and for the first eleven years of his reign the real ruler of China was his mother, the regent empress dowager - a capable administrator in her own right, who repealed the wartime tea and salt taxes and set up a board to keep local officials from using taxation to oppress the people they collected it from.

National Palace Museum, Taipei · Boulger, The History of China (1898) · Unknown · Public domain

A war he did not choose well

Renzong took the government into his own hands on her death in 1033, and within a year found himself facing a threat neither he nor his ministers had taken seriously enough. Li Yuanhao, ruler of the frontier state of Xia in the northwest, had spent years quietly drilling his own army on the Chinese model while the Song court assumed a long peace meant a permanently weakened neighbour. In 1038 he declared himself emperor in his own right rather than the Prince of Xia the Song court had recognised him as, and the war that followed went badly for China from the start: a three-day battle at Sanchuan went to the Xia army, and a second engagement near Yang Moulong ended with the Song commander Jinfou refusing to leave the field even mortally wounded - "I ought to die," he is recorded saying, and did, his body found among the thickest of the slain. The Khitan Liao, watching Song's weakness with their own interest, used the moment to seize ten frontier cities of their own.

Renzong bought the peace his army could not win. The 1042 treaty with Liao raised the annual tribute in silk and silver Song had already been paying; the settlement with Xia the same year restored Li Yuanhao's full dignities and committed the court to an annual gift of a hundred thousand bolts of silk and thirty thousand pounds of tea, on top of conceding his right to fortify his own frontier. It was not a settlement anyone at court could take pride in. It did buy the twenty years of peace the rest of the reign is actually remembered for.

A circle of scholars in government

What that peace let Renzong build was less a military legacy than an intellectual one. On the recommendation of his minister Fan Zhongyan, he ordered a college built in every town of the empire, staffed with lecturers and examiners, and personally prostrated himself before the doors of the capital's own Confucian academy to mark how seriously he meant it. Late in his reign he commissioned a full official history of the Tang dynasty, assembled by a team that included Sima Guang - the finished work running to two hundred and twenty-five volumes, one of the genuine monuments of Chinese historical writing.

Boulger, The History of China (1898) · Unknown · Public domain

Fan Zhongyan, Sima Guang, Ouyang Xiu and the other scholars this policy brought into government worked closely enough together that courtiers complained to the emperor about a faction operating in the palace's own corridors - "men in public service should not form a party amongst themselves," ran the accusation Renzong himself half-credited. Ouyang Xiu's own written defence survives: their association, he argued, produced results good for the nation and was not to be confused with cabals formed for an unworthy purpose. The twenty peaceful years that followed the settlements with Xia and Liao gave that argument room to be tested, and by most later judgments, Renzong's own long, unspectacular reign is remembered as having won it.

What outlived him

Renzong died in 1063 after thirty-one years on the throne, credited by the record with greater virtue than ability but real success in the harder task of keeping his ministers' loyalty and his subjects' affection without any single dramatic achievement to point to. The scholarly circle his own patronage had assembled - Sima Guang above all - went on to shape the argument the next reign would actually be remembered for: whether the state that Fan Zhongyan's generation had staffed with examination graduates should also be the state that ran the economy those graduates administered.

Boulger, The History of China (1898) · Gary Todd · CC0

A sure and certain market

peakDrafted from Giles, The Civilization of China (1911)

By the time Wang Anshi turned forty-eight he had risen, on the strength of a formidable memory and an even more formidable pen, to confidential adviser of the emperor Shenzong. What followed, in Giles's account, was one of the most sweeping programmes of state intervention any Chinese government had yet attempted — dressed, for the benefit of a population that distrusted anything without precedent, in the language of a return to older and more correctly understood readings of the Confucian classics. The dressing convinced almost nobody in the senior bureaucracy and did not need to: it had the emperor.

Giles, The Civilization of China (1911) · Unknown · Public domain

The middleman state

The centrepiece was a scheme aimed, in its own stated terms, at lowering taxation while raising revenue at the same time — a promise Giles treats with open scepticism, though he lays out the mechanism without editorial comment. A district's produce was to be divided three ways: enough set aside to meet its tax obligation, enough to meet its own needs, and the remainder bought up by the state at a low fixed rate, to be held against a price rise or shipped to wherever demand happened to be running high. The farmer got a guaranteed buyer. The state got to keep the margin a private merchant would otherwise have taken as middleman.

National Palace Museum, Taipei · Giles, The Civilization of China (1911) · Unknown · Public domain

Beside it ran a system of compulsory state credit: advances to cultivators secured against their standing crops, issued whether or not the farmer had asked for one. A militia reform required every family with more than two adult males to furnish one for military service, and separately obliged every household to raise a state-subsidised cavalry horse on the government's behalf. The old corvée obligation — forced labour drafts for public works — was converted into a money payment instead. State warehouses for bartering and pledging goods were set up empire-wide.

What did not work

The most ambitious single measure, and the one Giles singles out as the scheme's clearest overreach, was a plan to remeasure the entire country's farmland into equal taxed squares and assess each according to its fertility. It proved unworkable in practice and drew, like everything else Wang Anshi proposed, bitter resistance from the officials who had to administer it.

A separate crisis followed his decision to lift all restriction on the export of copper: the state's own copper cash began disappearing into melting pots to be recast and sold as raw metal abroad, and a currency panic followed. Wang Anshi's fix was blunt rather than elegant — he simply doubled the nominal value of each coin still in circulation, which stopped the bleeding without addressing why it had started.

The examinations, and his own reading

He also went after the civil service examination itself, pushing candidates toward practical subjects — history, geography, political economy — and away from ornamental style for its own sake. A contemporary author, quoted by Giles, recorded the effect at the grassroots with some astonishment: "even the pupils at the village schools threw away their text-books of rhetoric, and began to study primers of history, geography, and political economy." Wang Anshi described his own reading habits, by way of justifying the reform, as omnivorous to the point of eccentricity — medical texts, botanical treatises, agricultural manuals, anything that might illuminate "the great scheme of the Canon itself."

He fell from favour before the reign was out, was dismissed to a provincial post, briefly recalled, and then retired from public life to die not long after — having lived, Giles notes dryly, to see nearly the whole of his own policy reversed by the men who came after him.

The living Buddha

aside?What aside meansDrafted from Giles, A History of Chinese Literature (1901)

As a boy, Sima Guang saved a playmate from drowning in a water jar by the simplest method available to him: he picked up a stone and smashed the jar, letting the water out. Giles offers the anecdote as the first evidence of a mind that solved problems by attacking the actual obstacle rather than the conventional one, a habit that shaped both his politics and the four decades he spent, on and off, compiling the single largest work of history his country had produced.

Metropolitan Museum of Art, New York · Giles, A History of Chinese Literature (1901) · Unknown · CC0

Opposition

Sima Guang rose through the official examinations to the rank of Minister of State, and used the position to oppose Wang Anshi's reform programme from the inside for as long as he could. In 1070, unable to check it any further, he resigned the chancellorship rather than continue administering policies he believed would fail. He did not return to office until 1085, the year Shenzong died and the reforms lost the one patron who had kept them running against the bureaucracy's own judgment. Within months of his return Sima Guang had dismantled nearly the whole of Wang Anshi's programme; within months after that, he was dead — reportedly still labouring for the state's good on his return to the capital, undone by illness rather than politics. His countrymen called him, on his death, the Living Buddha.

Giles, A History of Chinese Literature (1901) · Unknown · Public domain

The Mirror

His resignation in 1070 was not idleness. He used the years out of office to complete the Zizhi Tongjian — the "Comprehensive Mirror for Aid in Government" — a chronological history of China running from the fifth century BC to the founding of the Song itself in 960, a title the emperor bestowed on the finished work in 1084 on the reasoning that "to view antiquity as it were in a mirror is an aid in the administration of government." As a student he had rested his arm on a rounded wooden pillow designed to roll and wake him the moment he began to doze over his books; as an old man compiling a work that size, the same discipline against half-measures that had smashed the water jar kept him at it for the better part of two decades.

He was, by the account of his own household, fanatically careful with the physical books themselves — his library still looked new after years of use, because he forbade his students from turning pages by scratching at them with a fingernail and made them use the pads of two fingers instead.

What he wrote about watching the watchmen

One surviving fragment of his prose, on the office of Censor, reads less like institutional history than like a warning aimed past his own century. Before the Han dynasty, he wrote, admonishing the throne had been every subject's right, "from the highest statesman down to the artisan and trader." The Han turned that right into an office, charged with weighing the empire's successes and failures for the throne's benefit and not the censor's own. In 1017 six men were formally appointed to the post; by 1045 their names were written up on public boards; by 1062 the names were being cut into stone, so that posterity could point at one and say "there was a loyal man," and at another, "there was a traitor." Sima Guang's own comment on the practice was a question rather than an endorsement: "Does not this give cause for fear?"

The fall of Kaifeng

contractionDrafted from Boulger, The History of China (1898)

The Jin had begun their revolt as vassals of the Liao, the Tartar power that had sat on the Song's northern border since before the dynasty's founding. Song diplomacy helped bring the Liao down, on the reasoning that a weaker, still-forming Jin was a safer neighbour than a strong, entrenched Liao. The Jin destroyed the Liao inside eleven years and turned on their former ally almost as soon as the job was finished.

The first siege

A Jin detachment crossed into Song territory in 1125. Emperor Huizong fled south to Nanjing for his own safety and left his son to manage what resistance could be organised. The Jin general Walipou drove on the capital, Kaifeng, but the wider Jin leadership had not yet settled how far to press the advantage — a faction among them argued the Yellow River marked the natural limit of Jin ambition. Negotiations produced a treaty, and Huizong returned to his capital believing the crisis over.

Palace Museum, Beijing · Boulger, The History of China (1898) · Zhang Zeduan (attributed) · CC0

Walipou did not share the restraint of his own colleagues. He wanted Kaifeng occupied outright and refused to honour the settlement. Forced into one retreat, he came back in greater strength, and this time the armies the new emperor, Qinzong, sent to meet him were beaten badly enough that the Jin could lay close siege to the capital itself.

National Palace Museum, Taipei · Boulger, The History of China (1898) · Unknown · Public domain

The siege

Kaifeng's garrison numbered some seventy thousand men, reinforced by another thirty thousand marched up from the south, and there was a real prospect — Boulger allows — that want of supplies might have forced Walipou to withdraw if the city could simply hold. Prince Kang Wang and the more resolute officers argued for exactly that course. Qinzong did not share their nerve. When a Jin assault briefly took the ramparts and gates, it seemed to confirm his instinct for a negotiated settlement, and only Walipou's own reluctance to spend lives on the street fighting the Song had prepared for kept the city from falling by storm there and then. Qinzong went to the Jin camp himself to arrange peace terms.

Boulger, The History of China (1898) · Gary Todd · CC0

The Jin demanded a payment Qinzong knew, even as he promised it, that he could not raise. When his followers discovered how little of the promised sum would actually materialise, the outcry among them was loud enough that Qinzong was detained rather than allowed to return to his own capital. Huizong and the rest of the imperial family still in Kaifeng were induced to seek shelter in the Jin camp and were carried off into Tartary; both former emperors eventually died there, at long intervals from each other and from home.

What was left

Walipou was not satisfied with the ransom of the capital and the removal of the ruling house. He installed a court creature as a puppet emperor over what remained of Song territory, but the arrangement lasted only as long as the Jin army stayed close enough to enforce it — the moment Walipou withdrew, the puppet was set aside.

The dynasty survived in the person of Kang Wang, who took the throne as Emperor Gaozong and restored Song rule over what territory remained to him. But the terms of the restoration were permanent: China now had two empires where it had had one, the Jin holding the north with a stronger, more vigorous state apparatus than the Liao they had replaced, and the Song confined — for the rest of the dynasty's existence — to the lands south of a frontier it would spend the next century and a half trying, and failing, to push back.

The general who would not make peace

contractionDrafted from Mayers, The Chinese Reader's Manual (1874)

Yue Fei rose from a lieutenant under Zhang Jun to lieutenant-general within a single year, 1131, and for the next five years was the one commander the restored Southern Song court could rely on to reverse rather than merely survive its losses to the Jin.

Mayers, The Chinese Reader's Manual (1874) · Liu Songnian (attributed) · Public domain

A soldier the court could not fully use

His early service brought a warlord's submission and the recovery of substantial territory from insurgent leaders operating in the chaos the Jin invasion had left behind. In 1136 he asked the court's permission to attack the Chinese provinces the Jin had occupied outright - not another defensive campaign, but a genuine attempt to reverse the settlement of 1127. The permission was refused, on the advice of the chancellor Qin Hui, who had already committed himself to negotiating a peace with the Jin that a victorious general in the field would only complicate.

National Palace Museum, Taipei · Mayers, The Chinese Reader's Manual (1874) · Emperor Gaozong of Song · Public domain

Yue Fei's own devotion to the war Qin Hui was trying to end made him, in the chancellor's eyes, the single greatest obstacle to the peace he wanted. Qin Hui had him removed to a lesser post first, and when that did not silence him, engineered a formal accusation of treason against Yue Fei and his son Yue Yun. Both men were committed to the state prison. Their trial proved nothing but the falsity of the charge - and changed nothing. Despite the protests of the statesman Han Shizhong and others at court, Qin Hui secured an imperial mandate for Yue Fei's execution, and it was carried out.

What the verdict became

Chinese historians never forgave it. The act was met, in the judgment later generations settled into, with what one nineteenth-century biographical dictionary calls "the undying execration of historians and of the Chinese people" - and the record's own treatment of the two men who made the decision could not be more lopsided. In 1179 Yue Fei was posthumously canonised; in 1204 he received the title Prince of E. Qin Hui's own name, by contrast, became a byword in Chinese usage for any official willing to deal too accommodatingly with a foreign power - a synonym still current centuries after his death, reserved for exactly the kind of pragmatic peacemaking he had believed, not unreasonably, was the only realistic policy available to an exhausted state.

National Museum of China, Beijing · Mayers, The Chinese Reader's Manual (1874) · Gary Todd · CC0

The peace it was traded for

The peace Qin Hui achieved by removing Yue Fei came the following year, in 1141 - the Shaoxing treaty, in which Gaozong renounced Song's claim to the lost north and agreed to an annual tribute in silk and silver to the Jin court that had once held his own father captive. Whether that peace could have been won without Yue Fei's death, or whether it needed to be won on those terms at all, was exactly the argument the general's own execution foreclosed - and the argument China's popular memory has never really stopped having anyway, four generations of dynasties later, over which of the two men actually served the country he claimed to be protecting.

National Palace Museum, Taipei · Mayers, The Chinese Reader's Manual (1874) · Unknown · Public domain

No home, no money

aside?What aside meansDrafted from Giles, A History of Chinese Literature (1901)

Feng Dao served ten emperors across four dynasties in the fifty-three years of chaos before 960, and described himself to a Liao court willing to employ him anyway as a man with "no home, no money, and very little brains" — a statement the Tartar monarch apparently found appealing enough to appoint him grand tutor to the heir on the spot. Chinese tradition remembers him less kindly, as a survivor who gave himself the sobriquet later English writers would translate as "the Vicar of Bray." To foreign readers his more durable claim is technical rather than moral: some crude form of block printing had existed since early in the Tang, but it was Feng Dao who first put it to the production of books. Six years after his death in 954, the Song took the throne he had spent his whole career outlasting other dynasties to keep, and printing from blocks became, for the first time, an ordinary trade.

What the printed word made possible

A literary class scattered by half a century of dynastic churn found itself, almost as soon as the Song was founded, able to reproduce a text without copying it out by hand — and Giles is explicit that the effect was immediate: history, classical scholarship, general literature and poetry were "again filled with enthusiastic workers," backed by a succession of rulers eager to encourage them.

Kaifeng Museum · Giles, A History of Chinese Literature (1901) · Gary Todd · CC0

The first large project was a re-write of the Tang dynastic history from scratch. A version already existed, compiled by a scholar of the Later Jin named Liu Xu, but it was judged unsatisfactory on enough grounds that the court commissioned a replacement rather than live with it — and the older work, Giles notes, was never actually superseded: both still stand side by side among the recognised dynastic histories today.

Ou-yang Xiu

The new commission went to Ou-yang Xiu and Song Qi together. Ou-yang Xiu had been raised in poverty, taught to write with a reed by a mother who could not afford better, and by fifteen was already attracting notice for his ability; he placed first among the candidates for the highest examination degree. His public career was turbulent, because he kept taking the position he believed correct without much regard for what it cost him — a habit that produced repeated friction with the court he served. His private output, alongside the dynastic history, ran to an exposition of the Book of Poetry, a study of ancient inscriptions, anecdotes of his contemporaries, an elaborate treatise on the peony, and poetry and essays without apparent end.

Giles, A History of Chinese Literature (1901) · Unknown · Public domain

Song Qi's own career had its lighter moments. Giles records one occasion on which, feeling the cold at an imperial gathering, he was offered a tippet by about a dozen court ladies at once — and, unwilling to seem to favour any one of them over the rest, chose to sit and shiver rather than accept from a single hand.

Giles, A History of Chinese Literature (1901) · Unknown · Public domain

A principle, not a person

aside?What aside meansDrafted from Giles, The Civilization of China (1911)

The scholar who eventually settled how the Song's own philosophical output would be read, for every dynasty that came after it, did not appear until long after the dynasty's political and literary height had passed. Zhu Xi's official career, between 1130 and 1200, was successful and his literary output enormous, but Giles is explicit that his lasting importance rests on something narrower than either: he re-read the Confucian canon and insisted, against centuries of Han-dynasty interpretation accepted as settled, that the same words could not mean one thing in one passage and something else in another.

Fujian Museum, Fuzhou · Giles, The Civilization of China (1911) · Gary Todd · CC0

A method, not a doctrine

The principle was almost aggressively simple — consistency, applied without exception to a body of text that earlier commentators had been willing to read two ways when it suited them. Its effect, Giles writes, was "magical." From that date forward, the teachings of Confucius were understood, across the whole of educated China, in the way Zhu Xi said they ought to be understood — not because his readings were self-evidently correct, but because the method that produced them refused to leave room for a competitor.

What it cost the old God

The consequence Giles traces furthest is theological rather than literary. The God in whom Confucius himself had believed, as the earliest form of the written character for the word shows — a rough picture of a man — had been personal, closer in kind to the God of the Psalms than to an abstraction. Zhu Xi discarded that reading outright. The Chinese word for God, he declared in the plainest terms available to him, meant nothing more than "abstract right": a principle, not a person. He had passed briefly through Buddhism as a young man, found it unsatisfying, and returned to Confucian orthodoxy determined to exclude the supernatural from it as completely as he had excluded it from his own belief. What he built to replace both the old personal God and the foreign religion he had tried and rejected was a moral system with no supernatural content left in it at all — a reading that, Giles notes without much enthusiasm for it, has been accepted as the correct one more or less ever since.

Zhu Xi worked and taught entirely within the reduced, southern state the dynasty had become after 1127; the "magical" uniformity Giles describes was still spreading through Chinese education, on his account, centuries after Zhu Xi's own death and long after the Song itself was gone.

Four years at Xiangyang

contractionDrafted from Boulger, The History of China (1898)

The alliance that destroyed the Jin in 1234 lasted less than a year before the two victors turned on each other, and the war that followed ran for four decades before it finally reached Hangzhou.

An alliance neither side trusted

Song had joined the Mongols against the Jin for the oldest reason a weaker power allies with a stronger one against a nearer enemy: a Jin envoy had warned, accurately, that a Song court which let the Mongols finish off the Jin unopposed would find itself next, but the temptation to see a generations-old enemy destroyed outweighed the warning. The promised reward, the restoration of Henan once the Jin fell, evaporated the moment victory was won - Mongol garrisons simply stayed where they were, and when a proposal to divide the province in two proved as unworkable as it sounds, both sides understood the war that followed was not really about Henan at all.

The first wave, 1235

Ögödei put an army estimated at half a million men into three columns in 1235: one crossing from Shaanxi into Sichuan, two others striking into the lower Yangtze provinces. The northern half of Sichuan fell within the year, its garrison commander Kao Kia dying at his post at the town of Mien rather than surrender it - but the Mongol force that took it also withdrew the way it had come once the campaigning season ended, "a passing scourge" rather than an occupation, and Song forces reoccupied what had been depopulated in the meantime. Ögödei's own attention was elsewhere - the campaigns in Russia and Hungary his nephew Batu was fighting interested him more than the slow grinding war in the south - and for the rest of his reign the Song front received little more than raids.

Xiangyang, 1268-1273

The decisive campaign waited for Kublai, and for a defector's advice on where to strike. A renegade Song general convinced him that the war-city of Xiangyang, paired across the Han river with its twin fortress Fanching, was the single position whose fall would open both the river route and, eventually, the Yangtze itself to Mongol shipping. In 1268 a Mongol army of sixty thousand veteran troops, later reinforced, invested the two cities, building an entrenched camp ten miles long and constructing their own war fleet from nothing to contest the Han.

Xiangyang's governor, Lü Wenhuan, held for four years. His garrison had provisions, by his own claim, for a decade; when the Mongols demanded surrender he answered by threatening to drag their own renegade guide back to his abandoned master in chains. In 1270, with the siege already three years old, the Song commander at nearby Anlu, Li Tingzhi, gathered three thousand volunteers and a convoy of supplies and ran the Mongol blockade by river, breaking the chains strung across the Han and fighting through the Mongol fleet while a second detachment under Zhang Shun held the enemy's main force occupied for hours. The supply convoy reached Xiangyang to loud celebration; Zhang Shun's own body was found floating past the walls days later, the garrison's reminder that the siege itself had not lifted.

A second relief attempt under the same officers' companion, Zhang Gui, ended the same way - a night run through the blockade, a fighting retreat against overwhelming numbers once a deserter had warned the Mongols he was coming, and a last stand with no survivors. The Mongol commander sent his body back to Xiangyang under flag rather than keep it, whether from admiration or calculated cruelty the record does not resolve; Lü Wenhuan had it laid beside Zhang Shun's.

Persian siege engineers brought in for the purpose finally broke Fanching's defenses with counterweight trebuchets capable of hurling stone at a scale neither Chinese siege engine had matched; its garrison fought street to street and chose suicide over surrender once resistance became hopeless. With Fanching gone and no further relief coming from the corrupt chancellor Jia Sidao, who had sent an incompetent general and then done nothing more, Lü Wenhuan accepted terms and surrendered Xiangyang in 1273, entering Mongol service himself after four years of holding the line for a court that had largely abandoned him to it.

Boulger, The History of China (1898) · Unknown · Public domain

What the fall actually opened

Xiangyang's loss did what its capture was meant to do: it put the Han river, and with it the approach to the Yangtze, in Mongol hands. Emperor Duzong's own court, informed of the disaster, is recorded retreating further into drink and indifference rather than reckoning with the minister whose neglect had produced it. The war that followed moved quickly by the standard of the four decades that preceded it - Hangzhou itself fell within six years of Xiangyang's surrender, and the flight to the southern coast that ended at Yamen was already, by 1273, the only road left open to a dynasty that had just lost the river its last defensible line had depended on.

Yamen

endDrafted from Boulger, The History of China (1898)

By the end of 1276 the Song cause held nothing but a handful of districts in Guangdong and Fujian, and even that shrank further the following winter when the Mongol armies returned from campaigning against Kublai's rebellious nephew Kaidu and drove the last Song forces off the mainland entirely, leaving them nothing but their ships and a scatter of islands off the Canton estuary. The reigning child emperor died in that extremity. A small, resolute remnant of officers proclaimed another Song prince in his place — a child given the name Tiping — in 1278, less as the start of a recovery than, in Boulger's phrase, "the prelude to its fall."

The last harbour

The general Zhang Shijie took the boy and the fleet to a natural harbour on an island called Yai, entered only on a favourable tide, and prepared as best he could for a battle he knew was coming. When the Mongol fleet found them there, the engagement was stubbornly contested, and there is little doubt, Boulger writes, that not a single Song ship would have escaped at all but for a heavy mist that fell over the battle and let Zhang Shijie break out to sea with sixteen vessels behind him.

The emperor's own vessel had no such luck. Trapped in the press of battle, it fell into Mongol hands. One of the Song's most loyal ministers, Lu Xiufu, resolved in that moment to save what he could — not the dynasty, which was plainly beyond saving, but its honour. He threw his own wife and children into the sea, took the emperor in his arms, and jumped overboard with him. Most of the remaining officers followed the same course rather than be taken. Tiping, the last emperor of the Song, died that way, by drowning, with his minister.

Boulger, The History of China (1898) · Gu Yuanji · Public domain

What Zhang Shijie tried afterward

Zhang Shijie, still at sea with what was left of the fleet, learned of the emperor's death and asked the boy's mother to name another Song prince to succeed him. Grief proved stronger than the appeal to public duty; she threw herself into the sea rather than be consoled, and no successor was named. Zhang Shijie refused even then to treat the cause as finished. He sailed south, was given refuge, ships and fresh troops by the ruler of Tonkin, and returned within the year intending to seize Canton by surprise and reopen the war — over the objections of his own followers, who judged, correctly, that Kublai's power was by then too completely secured for a single squadron to achieve anything but a local, temporary success before being ground down.

The renewed attack never happened. What ended the Song's last military effort was not a battle but a storm that scattered the fleet before it could strike, and Zhang Shijie drowned with it — the same sea that had taken the emperor and his mother closing, within months, over the last man still willing to fight in the dynasty's name. A Sung minister once asked, of the army mustered to save the child emperor, "If Heaven has not resolved to overthrow the Sungs, do you think that even now it cannot restore their ruined throne?" Heaven, on the evidence Zhang Shijie's own death supplied, had resolved exactly that.