Why Could the Han Crush the Xiongnu but the Song Lost to Liao and Jin?

Why Could the Han Crush the Xiongnu but the Song Lost to Liao and Jin?

In 200 BCE, Emperor Gaozu of Han, Liu Bang, personally led an army north and was surrounded for seven days and nights at Mount Baideng by Modu Chanyu's 40,000 cavalry, narrowly avoiding becoming China's first captured emperor. In 1004 CE, the pressed Emperor Zhenzong of Song signed the Chanyuan Treaty, sending Liao 100,000 taels of silver and 200,000 bolts of silk a year in exchange for a century of peace. Facing the same northern steppe peoples, why could the Han beat the Xiongnu so badly that there was "no royal court south of the desert," while the Song could not recover even its lost northern homeland? More puzzling: Song economic output was nearly 30% of world GDP, the army peaked at over one million men, and revenues were ten times Han levels. Such a "wealthy Song" was repeatedly bullied by Liao, Western Xia, Jin, and Mongols. The answer is not who was "better at fighting," but two completely different national logics.

Han Wudi's Three Tools Against the Xiongnu

The Han-Xiongnu war was not a single event; it lasted nearly 150 years. From the Wen-Jing prosperity to Emperor Xuan and Chanyu Huhanye's submission, the Han used several generations to put strategic pressure on the Xiongnu.

Tool one: warhorses. The Han early on lacked cavalry. When Liu Bang reigned, "the emperor could not put together four matching horses; ministers rode in ox carts." From Wen-Jing on, the Han implemented horse policy, establishing 36 horse pastures along the frontier with state land and funds. By Wudi's time the official horse count reached 400,000; with private encouragement, national warhorse reserves topped one million. In the northern campaigns, Wei Qing and Huo Qubing led pure cavalry deep into the steppe.

Tool two: strategic depth. Wudi did not play passive defense; he attacked. From 129 to 119 BCE, ten major northern campaigns: Wei Qing and Huo Qubing took the cavalry deep into Xiongnu territory — Mobei south killed 50,000, the Hexi campaign took the Hexi Corridor, Mobei achieved "sealing Langjuxu." Huo Qubing's cavalry could march 2,000 li in one move and drove the Xiongnu main force north of Lake Baikal. The goal was not to defend the Wall but to break, drive away, and destroy the opponent.

Tool three: long-term attrition. Wudi waged war on the Xiongnu for decades at enormous national cost, leaving the realm "empty, households cut in half." But he played for state staying-power: as long as the Han did not collapse, the Xiongnu would crack first. It worked: Xiongnu population fell from about 1.5 million to perhaps 400,000 through long Han military pressure and political splits (the North-South Xiongnu division). In 89 CE Dou Xian's northern campaign sent the Xiongnu west; the northern steppe was emptied. The Han used two centuries to solve a northern threat that had lasted centuries.

The Song's Hardware Problem: No Horses

The Song's passive position against Liao and Jin was set by hardware.

Warhorse numbers. The Song's biggest geographic problem was losing the Sixteen Prefectures of Yan-Yun (to Liao) and the Hexi Corridor (to Western Xia) — precisely the most important horse-raising regions in ancient China. Without the right grasslands, the Song could raise horses only in the south, where climate was wet and hot, fodder unsuitable, and horses small and weak. Under Shenzong, official horse stock never topped 200,000 — sometimes under 100,000, mostly poor stock. Liao's elite "Pishi" army alone had 300,000 cavalry; the Jin invasion across the Yellow River brought over 150,000 cavalry and over 500,000 horses.

What is it like to be a Song foot soldier facing Liao or Jin cavalry? You are infantry; they are cavalry. Your horse, if any, is a southern pony; theirs is steppe-bred. Your weapons are short blades; theirs are bow plus saber and high mobility. Almost no battlefield conditions allow infantry to suppress cavalry head-on, save in mountains, walls, or rivers.

Northern strategic terrain. The Sixteen Prefectures of Yan-Yun were the natural northern shield for every Central Plains dynasty — roughly today's Beijing, northern Hebei, and northern Shanxi, controlling the Yan and Taihang ranges. Holding Yan-Yun meant the Central Plains could be defended; without it, the North China Plain was indefensible, and steppe cavalry could reach the Yellow River in three days. The Song never retook Yan-Yun in 300 years — 300 years of strategic disadvantage.

Emperor Taizu Zhao Kuangyin once said he would "save three million strings of cash to ransom Yan-Yun." Later, Emperor Taizong twice campaigned north and lost both times; at Gaoliang River Zhao Guangyi himself was shot and fled; the Yongxi northern campaign was a complete rout. The Song gave up strategic initiative and turned to pure defense.

The Real Difference: Institutional Logic

Hardware was one face; the deeper difference was the two dynasties' utterly different state logics.

The Han was a "military empire." From the founding, it was a regime of the merit-cake and Guanzhong military aristocracy. Liu Bang himself rose from war; his founding men were all generals. In the Han, military merit was an important promotion factor. Emperor Wu's outward expansion came with mobilization mechanisms — the military-merit rank system, frontier resettlement, military farms. The whole state was organized around fighting.

The Song was a "civilian-official empire." Founder Zhao Kuangyin was a military man who took the throne by the "yellow robe coup." The first thing he did on the throne was "loosen military power over a cup of wine" — invite generals to dinner and lift their commands in one meal, switching to civilian rule. The Song established a deep "value the literati, suppress the military" policy: civilian officials far outranked generals; military men with great feats still had to look up; the powers to command troops and to mobilize them were split (the Three Bureaus controlled troops; the Bureau of Military Affairs issued orders); "a general doesn't keep his troops; troops don't know their general" — generals and units rotated to prevent armies becoming private property.

This defended against internal rebellion (the Song had no real local warlords or military coups in 300 years) but castrated combat capability. Command was fragmented, generals afraid to decide, soldiers unfamiliar with commanders, orders clogged by layers — fine for easy wins, terrible for hard fights.

Official evaluation. Han generals were rewarded for heads and fought to the death; Song generals were impeached if they won too much, because "meritorious generals tend to get arrogant." Behind Yue Fei's unjust death was the Song civil-official faction's structural fear of generals — give up northern campaigns rather than let a general grow strong.

Han's Strategic Depth vs. Song's Point Defense

The Han's thinking against the Xiongnu was "destroy your core." The Song's thinking against Liao and Jin was "don't let you come to me."

Han thinking: the Xiongnu's heart was at the desert royal court — strike there. Their economy was in the Hexi Corridor — take that. Wei Qing and Huo Qubing pushed into the Xiongnu interior on annihilation campaigns, not for show but to systematically destroy Xiongnu war-mobilization capacity. After each blow it took decades for the Xiongnu to recover.

Song thinking: when Liao comes, defend the city; when Jin comes, negotiate; when Western Xia comes, build forts. The Song built endless border forts (the "fort village" policy), hoping to wear down the enemy in layered defense. But this had no strategic initiative — the enemy chose if, when, and how much to attack. The Song spent vast sums on troops and forts, and Liao or Jin came whenever they pleased; the Song was always reactive.

The sharpest contrast is the Yongxi campaign vs Mobei. Wudi's Mobei: Wei Qing and Huo Qubing took 100,000 cavalry 2,000 li north and sought the Xiongnu in the steppe. Taizong's Yongxi: three armies totaling 200,000 men aimed to retake Yan-Yun but had no coordination, no unified strategy, generals working against each other — all three routes were defeated with heavy losses. Both were offensives; the Han found and crushed the enemy, the Song fell into chaos before even finding them.

Wealth Cannot Solve Everything

The Song was economically powerful — close to 30% of global GDP (scholars debate the exact share, but it was high). Tax revenue was about ten times the Han's (adjusting for inflation and population), urbanization, technology, and commerce far exceeded the Han. But economic strength did not translate into military advantage; it became a heavy burden.

Money can't buy strong soldiers. The Song's standing army peaked at 1.3 million; military spending was over 70% of fiscal outlay. But over 30% of the central elite force (the "Imperial Army") were decorations — broad recruitment standards let in the old, weak, and sick; less than half could fight. The local xiangjun was mostly used for chores. The northwest fanbing and xiangbing were the only real fighters.

Money buys peace. The Chanyuan Treaty sent Liao 500,000 in silver and silk per year — under 1% of Song revenue but huge for Liao. The Song thought it a good deal — buy a century of peace with cash. Side effect: the "buy peace" model gradually drained the entire military's motivation. If you can buy off the war, who wants to train and risk death?

Money fattens the gentry. The Song saw severe land concentration; great landlords held vast lands and people. The state's call for general mobilization ran into the fact that half the population were tenant farmers and semi-bondsmen of others. This is utterly unlike the Han's "register every household, every man a soldier" mobilization.

A Deeper Civilizational Choice

The Han victory over the Xiongnu and the Song's defeat by Liao and Jin reflect two very different civilizational choices.

The Han chose "expansionist agrarian empire": use the economic base of farming civilization to support a standing army with nomadic capability, push out into the steppe, and destroy enemies in the cradle. The price was huge national wear and protracted war.

The Song chose "inward-looking literary-mercantile empire": give up military expansion, pour resources into commerce, science, culture, and education; use economic advantage to buy peace; rely on civilian rule to keep stability. The upside was domestic prosperity and cultural high tide (Song poetry, ci, painting are still unsurpassed); the downside was long military passivity, ending in Mongol conquest.

Neither choice is absolutely right or wrong. The Han model brought glory and Wang Mang's late chaos; the Song model brought prosperity and the disgrace of Jingkang. How a nation chooses between "projecting power" and "holding within" is the most fundamental strategic choice a civilization makes.

Looking back, the Han's military feats and the Song's civil achievements are both precious legacies. But if you only ask "which fits the survival strategy of agrarian civilization?" — the answer is already in history. The tragic Battle of Yamen, the Mongol cavalry trampling Lin'an — these are the final balance sheet on the Song's three centuries of strategic choice.