What Edward Luttwak Doesn’t Know About Ancient China (Or a Short History of Han-Xiongnu Relations), pt. 1

A Mongolian stamp depicting Maodun, founder of the Xiongnu Empire.
Image source.

A few weeks ago a friend passed along one of the least correct essays I have ever had the misfortune to read. It was written by  Edward Luttwak, secret agent  author of classic titles in the field of strategic studies like Coup D’état: A Practical HandbookGrand Strategy of the Roman Empire, and Strategy: The Logic of War and Peace. I was disappointed to find out that this particular piece, published in the Hoover Institute’s online magazine Strategika, closely mirrors a passage in Mr. Luttwak’s most recent best-seller, The Rise of China vs. the Logic of Strategy. In it Luttwak suggests contemporary Chinese foreign policy follows a pattern first seen in the foreign relations of the Han Dynasty two millennia ago [1]. To quote:

What is peculiar to China’s political culture, and of very great contemporary relevance is the centrality within it of a very specific doctrine on how to bring powerful foreigners—indeed foreigners initially more powerful than the empire—into a tributary relationship. Specialists concur that this doctrine emerged from the very protracted (3rd century BCE to 1st century CE) but ultimately successful struggle with the Xiongnú (匈奴) horse-nomad state,  just possibly remote ancestors of Attila’s Huns, but definitely the inventors of the Steppe State political system that would be replicated by all their successors, and more adapted than replaced even by the Mongols.

Formidable mounted archers and capable of sustained campaigning (a primary objective of the Steppe State), the Xiongnú ravaged and savaged and extorted tribute from the perpetually less martial, and certainly cavalry-poor Han until the latter finally felt able to resist again. Even then, 147 years of intermittent warfare ensued until Huhanye (呼韓邪), the paramount Chanyu (Qagan, Khan) of the Xiongnú, personally and formally submitted to the emperor Han Xuandi in 51 BCE, undertaking to pay homage, to leave a son at court as a hostage, and to deliver tribute, as befitted a vassal. That was a very great downfall from the familial status of earlier Chanyus of the epoch of Xiongnú predominance, who were themselves recognized as emperors, whose sons and heirs could have imperial daughters in marriage, and who from 200 BCE had received tribute from the Han, instead of the other way around. It is this successful transformation of a once superior power first into an equal (signified by imperial marriages) and then into a subservient client-state that seems to have left an indelible residue in China’s tradition of statecraft. [2]

I am a fan of the analytic approach Mr. Luttwak uses here. History is important. Ancient history is important. It might seem silly or frivolous to examine ancient polities in order to understand modern politics, but the insights this lens of analysis makes possible are hard to get through other means. Many of these insights come from seeing the world through the long view. The political and social structures civilizations are built on emerge on a timescale far longer than the lifespan of any individual human being. Many of the constraints societies face—be they physical or cultural—can only be seen clearly by examining centuries of conflict, competition, and collapse.

Just as important as these recurring patterns of history are the perceptions today’s decision makers have of the past. China’s 3,000 recorded years of war and high politics offer many different lessons to the Chinese statesmen of the modern era. The lessons they choose to draw from this history shape the decisions they will make tomorrow.

Thus if Edward Luttwak wants to talk about how the echoes of the Han-Xiongnu war can be heard in modern China’s foreign policy, I am all ears.  Long term readers of The Stage know that there are few conversation starters I would find more thrilling to hear. Too many contemporary controversies cannot be understood until we step back and look at world affairs from the long view of history.

But there is a catch in all this: the history has to be correct. It must accord to the facts. If one uses the past to interpret the present then your reading must be based on events that actually happened.

This cannot be said for Mr. Luttwak’s essay. The story he tells simply did not happen. 


 To understand why Luttwak’s narrative of the Han-Xiongnu war is so flawed we must look at a longer excerpt of his essay:

It is this successful transformation of a once superior power first into an equal (signified by imperial marriages) and then into a subservient client-state that seems to have left an indelible residue in China’s tradition of statecraft. It was achieved with a specific “barbarian-handling” tool box first described by its early practitioner, the scholar and imperial advisor Lou Jing (婁敬) 199 BCE. His method was first applied when the Xiongnú were still very strong and the Han were not only tactically inferior (their chariots were totally obsolete for fighting mounted archers) but also beset by political divisions, so much so that a 198 BCE4 treaty required the payment of an annual tribute in kind (silk, grain, etc.), and the formal attestation of equality for the Chanyu embodied in a marriage alliance, formalized by imperial letters that make the equality fully explicit.

The first barbarian-handling tool is normally translated as “corruption” in English translations, but perhaps “addiction,” or more fully “induced economic dependence” are more accurate: the originally self-sufficient Xiongnú were to be made economically dependent on Han-produced goods, starting with silk and woolen cloths instead of their own rude furs and felt. At first supplied free as unrequited tribute, these goods could still be supplied later on when the Han were stronger, but only in exchange for services rendered.

The second tool of barbarian handling, is normally translated as “indoctrination”: the Xiongnú were to be persuaded to accept the authoritarian Confucian value system and the collectivistic behavioral norms of the Han, as opposed to the steppe value system, based on voluntary allegiance to a heroic (and successful in looting) fighting and migration leader. One immediate benefit was that once the Chanyu’s son and heir married an imperial daughter, he would be ethically subordinated to the emperor as his father-in-law—remaining so when he became Chanyu in turn.

The much larger, longer-term benefit of the second tool was to undermine the entire political culture of the Xiongnú, and make them psychologically well as economically dependent on the imperial radiance, which was willingly extended in brotherly fashion when the Han were weak, and then contemptuously withdrawn when the Xiongnú were reduced to vassalage. [3]

What Luttwak is describing is the heqin, or the “peace marriage” system. The system was put in place during the reign of Han Gaozu, first emperor of the Han dynasty, shortly after his disastrous defeat at Pingcheng (200 BC) The empire was at that time still quite new and the Han did not have the resources to wage large-scale war against the Xiongnu. Gaozu was left with the difficult task of finding a permanent resolution to threat the Xiongnu posed to his realm. The heqin system was the best his court could come up with.

The heqin system had four components:

  • The Han Emperor and Xiongnu Chanyu would address each other as equals and brothers. The Xiongnu would acknowledge Han suzerainty of all people of the plow; the Han, for their part, would acknowledge Xiongnu lordship over the ‘people of the bow.’ 
  • The Han Emperor would provide an imperial princess to be a wife of every Chanyu (Note: In reality the ‘princess’ was always a concubine taken from the imperial palace. A multitude of legends and love stories about these concubines have made this a strong and enduring memory in Chinese popular culture for the last two millennia).
  • The Chanyu—or one of his subordinates—would take regular trips to Chang’an (the Han capital) to pay his respect to the emperor. During these visits the emperor would bestow lavish gifts upon the Xiongnu retinue as a sign of their friendship. 
  • Trade between the Xiongnu and the Han commoners would be allowed at select border stations across the frontier. [4]

The benefits of this system for the Xiongnu were obvious. One of the prime motivations behind Xiongnu raids were the economic benefits each raid offered–Xiongnu incursions were almost always marked by mass kidnappings and the theft of thousands to hundreds of thousands of Chinese livestock. Opening trade on the frontier allowed regular Xiongnu to increase their household wealth through trade instead of theft. On the other hand, the luxury gifts from the capital were distributed among the Xiongnu’s closest allies, retainers, and vassals as a way to maintain prestige and improve loyalty ties among the Xiongnu elite. Without a system of trade in place the Chanyu faced immense pressure from his impoverished subjects; without a stream of luxury items that could be distributed to the Chanyu’s favorites the anarchic dynamics of steppe politics threatened to tear the Xiongnu empire apart. Thus “raids in China were a profit making enterprise that served to wield the Xiongnu into a single unit.” [5]; open plunder was the simplest way to reduce tensions when trade was not on the table. It usually was not on the table:

“The right to trade at the frontier had to be wrung from an unwilling Han court because China opposed it for political reasons. Although the Xiongnu were a natural market for the grain surpluses and craft production of the northern areas, such trade would orientate the population away from the Han court and Chinese interests. The Han court attempted to tie the frontier regions to the center even if it meant hardship for the local population. Its policy was to create as much cleavage between the steppe and China as possible; the Great Wall was to be a barrier against all contact with the steppe. The Shanyu was forced to overcome this reluctance by extorting trade rights the same way he extorted subsidies–by raiding, or threatening to raid, China.”[6]

The Han decision to accommodate the Chanyu and adopt the heqin system was made in a time of crisis. It was proposed in the immediate aftermath of the disastrous Battle of Pingcheng and the mass desertion of Han generals and kings to the Xiongnu. [7] Ending the war before the situation spiraled out of the court’s control was its paramount priority and immediate peace was the greatest boon the heqin system could provide. This was not, however, the only objective of the original heqin treaty. The grand historian narrates a conversation between Gaodi and the Lou Jing that reveals the policy’s hidden purpose:

“I can only suggest a plan whereby in time Maodun’s descendants can be made subjects of the Han. But I fear Your Majesty will not be able to carry it out…”

“If it will actually work, why should I not be able to carry it out?” asked the emperor. “Only tell me what I must do!”

“If you could see your way clear to send your eldest daughter by the empress to be the consort of Maodun, accompanied by a generous dowry and presents, then Maodun, knowing that a daughter of the emperor and empress of the Han must be generously provided for would, with barbarian cunning, receive her well and make her his legitimate consort and, if she had a son, he would make him heir apparent. Why would he do this? Because of his greed for Han valuables and gifts. Your majesty might at various times during the year inquire of his health and send presents of whatever Han has a surplus of and the Xiongnu lack. At the same time you could dispatch rhetoricians to expound to the barbarians in a tactful way the principles of etiquette and moral behavior. As long as Maodun is alive he will always be your son-in-law and when he dies your grandson by your daughter will succeed him as Shanyu. And who has ever heard of a grandson trying to treat his grandfather as an equal? Thus your soldiers need fight no battles, and yet the Xiongnu will gradually become your subjects.” [8]

Luttwak’s descriptions of the heqin policy’s aim is basically correct. It was designed to corrupt the Xiongnu and slowly ‘Sinicize’ them. It was designed, through the power of Confucian family norms, to subordinate the Xiongnu ruler to Han Emperor.

What Luttwak neglects to mention is that the policy was a complete and utter failure.


The first heqin treaty was signed in 200 BC. Let us jump forward 44 years into the future and see how successful it was:

“In the fourteenth year of Emperor Wen’s reign the Shanyu led a force of 140,000 horsemen though the Chaona and Xiao passes, killing Sun Ang, the chief commandment of Beidi province, and carrying off large numbers of peoples and animals. Eventually he rode as far as Penyang, sent a surprise force to break into and burn the Huizhong Palace, and dispatched scouts as far as the Palace of Sweets Springs in Yong [bringing them within eyesight of the capital].

…The Shanyu remained within the borders of the empire for a little over a month and then withdrew. The Han forces pursued him beyond the frontier but returned without having been able to kill any of the enemy.

The Xiongnu grew more arrogant day by day crossing the border every year, killing many of the inhabitants, and stealing their animals. Yunzhong and Liaodong suffered most severely, while in Dai Province alone over 10,000 persons were killed. The Han court, greatly distressed, sent an envoy with a letter to the Xiongnu, and the Shanyu in turn dispatched one of his household administrators to apologize and request a renewal of the peace alliance.” [9]

This was one of only of a dozen major incursions between 200 BC, when the first treaty was signed, and 133 BC, when Han Wudi decided to launch the empire into open war with the Xiongnu. It also points to the heqin treaties’ greatest failure: they did not provide a meaningful sense of security for the Han. Under the heqin system the initiative belonged to the Xiongnu. The peace was theirs to keep—or break. The Han lacked an effective deterrent to ward off Xiongnu attacks and possessed little power to retaliate once attacks came.

Not all of these raids were organized or even condoned by the Chanyu. One of the inherit weaknesses of the ‘co-equal’ treaties between the Shanyu and Emperor was that the two sovereigns did not exercise equal amounts of control over their respective empires. [10] The Xiongnu confederation was an amalgamation of different tribal groups forged together by the Xiongnu ruling line. The Chanyu’s power was proportionate to the threats his people faced or the rewards he could promise them. If he tried to control the tribal autocracy with too tight a hand the inevitable consequence would be flight (to China or to other steppe confederations on the steppe) or open rebellion [11]. If one of his subordinate leaders decided to campaign against the Han on his own accord there was little the Chanyu could do stop him.

The other reason the Chanyu was willing to turn a blind eye to these attacks (and in occasion lead them himself) was that they strengthened the Xiongnu position at the negotiating table. “Almost each time a new pact was signed something was lost by the Han, and gained by the Hsiung-nu.” [12] Far from being an instrument the Chinese used to sap the Xiongnu elite’s political cohesion (as Luttwak suggests), the treaty system was a way for the Xiongnu elite to extort the Han Dynasty.  Thomas Barfield describes the logic behind the system from the Xiongnu point of view:

“The Xiongnu alternated periods of war with periods of peace in order to extract ever increasing benefits from China. Plundering invasions were followed by envoys from the Shanyu who always suggested that the current troubles could be resolved by a new treaty. Building upon each broken treaty as a foundation for new demands, the Shanyu extracted larger subsidies and gained trade benefits in return from the promise of peace. The length of the peace was determined in part by the adequacy of the newly formed treaty. The initial agreements providing subsidies but not trade lasted only a few years. After trade [at the border posts] was added the periods of peace became substantial. Nevertheless, behind even the most peaceful relations lay the implicit threat that the Xiongnu could cause the Han Empire serious trouble if their demands were not satisfied, and that no peace treaty could bind them permanently.” [13]

There are obvious parallels between Xiongnu-Han relations and China’s troubles with European imperialists in the 19th century. When the rhetorical garnish used in imperial proclamations to the outside world are peeled away the nature of both the unequal treaties and the heqin system can be seen for what they truly were: means through which foreign powers laid claim to China’s wealth. Both were systems of extortion imposed by the threat of martial force.  And in both cases the decision to cede to the foreigners’ demands instead of mobilize the population for general war was a decision to relinquish sovereignty and claims of cultural superiority for the sake of political stability.

Political stability was what the Han needed in the early days of their dynasty, during which time rebellions, uprisings, and attempted coups absorbed most of the court’s attention. It was not until the reign of Han Wudi that the dynasts felt secure enough within China to turn their attention outside of it. There was also an important ideological shift that occurred during the reign of Han Wudi. Before Wudi’s time the dominant political philosophy at court was Huang-lao Daoism, which privileged emergent order over conscious creation and stressed that the best rulers governed with a light hand. The ascendance of classicist [14] political ideals was an important element in the Han’s path to war. The classicists of Wudi’s day had an unyielding commitment to hierarchy, ritual, and decorum. They believed that these principles should govern the conduct of all men. They were unforgiving towards all who disregarded them. To a mind steeped in the classicist ideology of the early Han, there was no breach of decorum more scandalous than a barbarian who called himself the Emperor’s brother and extorted gifts from him as if he were the superior. The  heqin system was an affront to Confucian political philosophy. Jia Yi, perhaps the most brilliant classicist thinker of that age, expressed the outrage of it all by comparing “the situation of the empire [to a] person hanging upside down… To command the barbarians is a power vested in the Emperor at the top, and to present tribute to the Son of Heaven is a ritual to be performed by vassals at the bottom. Now the feet are at the top and the head at the bottom” [15].

It is difficult to tell just how aware the Xiongnu were of the cultural crises their strategy created in China. In contrast to Buddhism, Christianity, Islam, and other belief systems of more Western origin, Confucian norms and philosophy never caught hold in Inner Asia. This explains why attempts to “corrupt” the Xiongnu and subordinate the Chanyu to the emperor failed so dismally.  The nomads were eager to seize Chinese goods and luxury items, but were never economically dependent on them [16]. Unlike in Japan, Korea, Vietnam, and the “barbarian” statelets of South China, the economic structure of the Xiongnu empire was radically different than agrarian China’s. There simply was not enough common ground between the two peoples for Chinese lifestyles to be grafted onto the Xiongnu people. This is a recurring theme of Inner Asian history: nomadic elites could adopt the trappings of Chinese culture, but without moving off of the steppe they could never absorb its substance. The social worlds of the two peoples were simply too different. The ritualized and hierarchical relationships of the traditional Chinese family had no analogue in the egalitarian family life of the steppe. The first recorded Chanyu seized power by murdering his own father. The notion that ‘filial piety’ – or any of the other civilized virtues China hoped to ‘tame’ the barbarians with – could be instilled in the Xiongnu by giving them wives and clothes and toys was a political fantasy.

It took the Chinese eight decades to come to the same realization. By that time influential voices in the court and men with experience on the frontier argued that the existing system could not meet its original aims. It could achieve relative peace and stability, but the Chinese no longer wanted that. They wanted superiority. That could not be had without switching to a much bloodier strategy.

Part I of a series. Part II, which covers how the Han actually defeated the Xiongnu, is here.


[1] Edward Luttwak,”The Cycles–or Stages–of Chinese History,” Strategika, Issue 11 (14 February 2014), pp. 5-11; The Rise of China vs. The Logic of Strategy (Cambridge, MA: Belknap Press, 2012), pp.  25-28.

[2] Luttwak, “Cycles,” 6-7.

[3] ibid., 7-8.

[4] Thomas Barfield, The Perilous Frontier: Nomadic Empires and China, 221 BC-1757 AD (Cambridge, MA: Blackwell Publishers, 1989), p. 46; Nicola Di Cosmo, Ancient China and its Enemies: the Rise of Nomadic Power in East Asia (New York: Cambridge University Press, 2004), p. 193; Yu Ying-chih, “Han Foreign Relations,” Cambridge History of China, vol 1: Ch’in and Han, 221 BC-220 AD (Cambridge: Cambridge University Press, 1986), p 386.

[5] Barfield, Perilous Frontier, 46.

[6] ibid., p. 47; see also the discussion in Sechin Jachid and Van Jay Simmons, Peace, War, and Trade Along the Great Wall: Nomadic-Chinese Interaction through Two Millenia (Indianapolis: Indiana University Press, 1989), pp. 30-51.

[7] The primary source I rely on for this section of the essay is Sima Qian’s Shiji or Historical Records. Luttwak’s citation of this same source is needlessly obtuse; the relevant portions of the Shiji were translated and published in English by Burton Watson was over 30 years ago, and a newer, more scholarly translation of the relevant sections by a William Neihauser has been available for the last five years. The only purpose Luttwak’s note that the text “is increasingly available in English translation” serves is to make it more difficult for those not versed in Sinological convention (i.e. all of the readers of Strategika) to investigate the sources Luttwak uses for themselves. 

For an account of the Battle of Pingcheng and its immediate aftermath, please see SJ 93 (Records of the Grand Historian: Han Dynasty, vol I, trans Burton Watson, p. 187); SJ 99 (Records, vol I, trans Burton Watson, p. 238); SJ 110 (Records, vol II, p. 138).

[8] SJ 99 (Records, vol I, p. 239).

[9] SJ 110 (Records, vol 2, p.145).

[10] Di Cosmo, Ancient China and its Enemies, p.224.

[11] Barfield, Perilous Frontier, 24-28. See also Philip Salzman Pastoralists: Equality, Hierarchy, and the State (Boudler, Co: Westview Press, 2009), 50-51.

[12] Yü Ying-shih “The Hsiung-nu,” The Cambridge History of Early Inner Asia (Cambridge: Cambridge University Press, 1994), 125.

[13] Barfield, Perilous Frontier, 50.

[14] That is, the ru (儒 ), in later ages to acknowledged as “Confucians.” The label “Confucian” is probably anachronistic for this time period however, for Confucian ideology had not coalesced into a unified philosophical school.

[15] Quoted in Yu Ying-shih, Trade and Expansion in Han China (Berkley: University of California Press, 1969) p. 11.

[16] Nicloa Di Cosmo, “Ancient Inner Asian Nomads: Their Economic Basis and Its Significance in Chinese History,The Journal of Asian Studies, 53, no. 4 (Nov. 1994), 1092-1126.

Leave a Comment


Thank you for this post. The USA- China relationship is not analogous to this historical situation, and it gives us little guidance as to what is happening in China, what the decision-makers there are doing, what their strategy is (if they have a unitary strategy at all — I suspect they don't that different factions are doing inconsistent things) or otherwise help us respond in the present day. That said, not allowing ourselves to be overly dependent economically on a political and military competitor is common sense, and we should move away from it.

Thanks for the comment Lex.

You said "That said, not allowing ourselves to be overly dependent economically on a political and military competitor is common sense"

The real question is though: is our economy dependent on them or is their economy dependent on us?

I tend to think it is the second. The Chinese demand for treasury bonds is not a gun pointed at Washington's head, but at Beijing's.

"The Chinese demand for treasury bonds is not a gun pointed at Washington's head, but at Beijing's. "

Even so, if the whole thing blows up everyone is going to get hurt.

Speaking of the Hsiung-nu, seen this?

The Union of Two Worlds: Reconstructing Elements of Proto-Athabaskan Folklore and Religion

"The Union of Two Worlds: Reconstructing Elements of Proto-Athabaskan Folklore and Religion"

"Archaic roots for motifs are assumed, yet wider Inner Asian Buddhist and Xiongnu/
Hun cultural contexts reveal other possible interpretations. Navajo insect messenger guardians (notably Dontso) reside in the eastern quadrant of drypaintings. Dontso
provides primary assistance to the hero, Slayer of Enemies. Apart from the guise of the
insect itself, these guardians are essentially identical to members of a broad class of
Vajrayana Buddhist messenger-guardians known as dGra-lha (or dGra-bla), ‘enemy gods’,
in Tibetan. Enemy gods, like the Navajo Dontso, are guardians and messengers residing
on the shoulders of heroes, whispering in the ear and aiding in the destruction of enemies
(Waddell 1894, 354). dGra-lha are more than simply battlefield-guardian spirits, but (like
Tlingit dragonflies) embody primordial intelligence and life force, to the extent that
severance from the dGra-lha is certain death for the warrior (Gibson 1985)"

He argues that the Navaho are Hsiung-nu.