In all fictional works, each time a man encounters different al­ternatives, he chooses one and rejects the others; but in the case of the almost-undecipherable Ts’ui Pen, he chooses—simulta­neously—all of them. In this way, he creates different futures and different temporalities that also, in turn, bifurcate and multiply in their own right. It is in this that we find the explanation for the novel’s contradictions.

—Jorge Luis Borges, “The Garden of Forking Paths” (1941)

At the center of King Hu’s classic 1967 film, Dragon Gate Inn—one of the most influential Chinese wuxia (sword-fighting or knight- errant) films ever made—there is a wall.1 Positioned in front of the remote inn of the film’s title, this dilapidated structure provides a convenient backdrop for many of the work’s sword-fighting se­quences. The wall is nondescript, only a few meters long and easily overlooked in the action-packed film, though there is one slightly extraneous element that subtly tugs at our attention: a large white circle painted on the outward side of the wall (each of the outer walls of the inn is also marked by a similar circle). This circle is lit­erally a cipher, a mysterious element that draws attention to itself despite the fact that the film offers no explanation of its meaning. Precisely because it remains unexplained, however, the mysterious circular mark invites us to try to make sense of it.

[To view this image, refer to the print version of this title.]

The circle on the wall in Dragon Gate Inn, directed by King Hu (Union Film Company, 1967).

We might, for instance, see this wall circle as a symbol of periph­eries, and specifically the geographic periphery of the Chinese impe — rium. The Beijing-born King Hu filmed Dragon Gate Inn in Tai­wan, to which he had just relocated from Hong Kong; while the work’s opening sequence is set in Beijing, the rest takes place in an unspecified location along China’s northern frontier. Alterna­tively, we could see the wall circle as a symbol of vacated political centers. King Hu made and released Dragon Gate Inn in 1967, as the destabilizing Cultural Revolution was at its peak on the main­land, and the film’s story unfolds in the immediate aftermath of a critical crisis of imperial authority during the mid-Ming. More spe­cifically, we could even see the circle as a sort of spectral anticipa­tion of the Ming Wall, for the iconic brick and stone Wall would subsequently arise out of a partial collapse of imperial authority that could be traced back to the precise historical moment in which the film is set.

King Hu’s film opens with a voice-over noting the year: “China’s Ming dynasty, in the eighth year of the Jingtai reign, which is to say the year 1457 ad.” Specifying the date in relation to the current im­perial reign was conventional during the imperial period, though in this case the (technically correct) reference to the Jingtai reign is complicated by the fact that the events described in the prologue actually took place only after the Jingtai emperor had been de­posed by the same half-brother—the former Zhengtong emperor— he himself had previously replaced. The origins of this imperial re­versal can be traced back to a crisis seven years earlier, when the Zhengtong emperor was captured by Mongols and was replaced by his half-brother, who became the Jingtai emperor. This historical narrative is well known, and King Hu does not spell it out explic­itly. Instead, he proceeds to introduce the emperor’s chief eunuch and describes the eunuch’s imminent execution of the minister of war, General Yu Qian, for his alleged betrayal of the Zhengtong emperor following his capture. The prologue then cuts away to the opening credits just as the executioner’s sword is about to slice off Yu Qian’s head, and the main body of the film follows two of the general’s adult children as they are exiled to the northern frontier and pursued by secret guards whom the chief eunuch has sent on a mission to assassinate them, for fear they might return to avenge their father’s death.

King Hu was a notorious history buff, and it would be easy to view his film’s depiction of executions, attempted assassinations, and secret guards as an allegorical commentary—in the Chinese tradition of “using the past to critique the present”—on the Cul­tural Revolution that was under way in China when King Hu was making his film. Here, however, we will consider the film’s treat­ment of history as a reflection not on the present but on the histori­cal period in which the film is set—and specifically as a reflection on the relevance of that period for the subsequent construction of the Ming Wall.

The mid-fifteenth-century moment in which King Hu’s film is set constitutes the virtual origins of the Wall-building project that would increasingly absorb the court’s attention for the remainder of the dynasty. This project, furthermore, did not constitute a direct continuation of a tradition of border-wall construction dating back to the Han and Qin, but rather emerged out of what had become a complex web of endlessly bifurcating trajectories of Wall-building practices and the traditions they inspired.

With the exception of the short-lived Sui dynasty, none of China’s unified regimes since the fall of the Han had evinced much interest in border walls. The Tang, for instance, the first dynasty to succeed in unifying China after the fall of the Han, had strong ethnic and cultural roots in the northern steppe, and rather than build defen­sive walls to protect itself from its northern neighbors it sought to expand its influence in Central Asia. The Tang’s successors, the Song, meanwhile, are generally perceived as having been almost too weak to build and maintain defensive walls. The Song was mili­tarily overmatched by the Khitan-ruled Liao dynasty to its north and was forced to sign a treaty positioning itself in a subservi­ent tributary relationship with the Liao. Roughly a century later, the Song was partially defeated by another northern group, the Jurchens, who forced the Song to concede the entire region of northern China where the Wall had traditionally run. Finally, the Mongols, after they established the Yuan dynasty, already con­trolled the entire Central Asia region and therefore had little need for defensive walls to protect them from foreign invaders.

Several of the kingdoms and lesser dynasties during this period from the Han to the Ming, however, were in fact enthusiastic wall builders. Following the collapse of the Han dynasty in 220 ce, there was a roughly three-century-long period in which the region corre­sponding to modern China was controlled by a series of overlap­ping kingdoms and minor dynasties, many of which were ruled by partially Sinicized pastoral peoples from the northern and western border regions. A northern tribe called the Tuoba, for instance, es­tablished a kingdom known as the Northern Wei and proceeded to unify northern China. A Tuoba prince is recorded as having con­structed a “long wall” in the early fifth century ce to protect the border from raids from the north, and about sixty years later a Chi­nese official in the Tuoba court recommended that the court build additional border walls on a massive scale, both for the practical purpose of defending against attacks from the north and also to po­sition the dynasty within a tradition of Chinese border-wall con­struction dating back to the Qin and earlier.

Several of the other northern dynasties that followed the fall of the Northern Wei are also recorded as having built border walls through the region, including the Northern Wei’s own immediate successors, the Eastern Wei, together with the Northern Qi and the Northern Zhou. Shortly after the Sui succeeded in briefly uniting China in the sixth century, it constructed approximately 350 kilo­meters of walls along its northern border. While each of these pre­Tang regimes had strong ties to the northern steppe, their con­struction of border walls functioned to separate them from those geographic and ethnic origins while symbolically allying them with a practice of governance associated with the Chinese dynasties of the Central Plains.

The last regimes to pursue these sorts of northern border forti­fications prior to the Ming were the Liao and Jin dynasties, between the eleventh and thirteenth centuries. The Liao was founded by the northern tribe known as the Khitan, and its military superior­ity over the Song dynasty to the south allowed it to obligate the wealthy Song to make large tributary payments, which the Liao then invested in building border walls deep in Central Asia to de­fend not against the Song but against other tribes even farther north (remnants of these walls are still discernible in Outer Mongolia and eastern Russia, where they are sometimes referred to anachronisti­cally as the Wall of Genghis Khan). After one of the Liao’s vassal peoples, the Jurchens, succeeded in overthrowing the Liao and es­tablishing a dynasty of their own, the Jin, they proceeded to build a network of border walls through the same general region to defend against the Mongol forces. Like the Northern Wei and their pre-

Tang successors, the Liao and Jin dynasties were both ruled by peo­ples from the northern steppe who were in the process of trying to reinvent themselves as orthodox Sinicized dynasties. While the Liao and Jin border walls were not explicitly called long walls and were positioned significantly farther north than the border walls built during other periods, these structures can nevertheless be seen as part of a broader long-wall tradition—reflecting a process wherein peoples from the periphery of the Chinese imperium used border walls to ally themselves symbolically with the Wall-building tradi­tion associated with the Chinese interior.

In contrast to the familiar vision of the Wall as evidence of a di­rect historical lineage linking contemporary China to its Qin dy­nasty origins, what we find in the post-Han period is a complex network of parallel and overlapping wall-building traditions, inter­spersed with lengthy periods during which there was little or no border wall construction at all. If we look beyond these bifurca­tions and interregna, however, there is suggestive evidence for the continuity of the notion of a unified Wall. Not only did legends such as those of Lady Meng Jiang and Wang Zhaojun help keep the memory of a unitary Wall alive within the popular imagination, there was also a rich body of Tang dynasty poetry that prominently featured the Wall as a familiar topos. Known as “frontier” poetry, these verses often revolve around the nostalgia experienced by gov­ernment officials assigned to remote outposts at the margins of the empire, and they evoke the Qin Wall as a rhetorical anchor for their remote setting—despite the fact that the actual Qin Wall was by that point a mere memory.

Perhaps the most compelling evidence for the persistence of an abstract notion of the Wall, even when the construction of actual border walls was falling out of favor or being relegated to the politi­cal margins of China “proper,” can be found in cartography. An iconic representation of a unitary Wall appears unambiguously on several maps from the Southern Song period, including one called

[To view this image, refer to the print version of this title.]

Northern Wei, Northern Qi, Sui, Liao, and Jin dynasty walls.

Meridian Mapping.

Map of Chinese and Barbarian Nations that was originally carved in stone in 1040, which presents the Wall as a continuous, crenel­lated structure extending from the Ordos region to the Gulf of Bohai. Similar representations appear on several other Song maps (including the Geographic Map discussed in Chapter 1), suggesting that the notion of a unitary Wall spanning China’s entire northern frontier was still very much alive during this period, despite the fact that neither the Song nor the Tang had demonstrated any interest in border-wall construction, and the fact that the Southern Song no longer even controlled the territory through which the Wall ostensi­bly ran. While it is true that the Liao and Jin had built border walls of their own through the region, the general trajectory of the Wall as it appears on the Song maps corresponds more closely to that of the Qin/Han structure (as traditionally conceived) than that of the shorter and much more northerly Liao and Jin constructions. Ac­cordingly, it would appear that what is being represented in the Song maps is not the actual Wall but instead a historical memory, and it was precisely through this cultural fantasy that the Wall per­sisted until construction resumed during the latter half of the Ming.

Although the Ming is known for its vast material investment in building its defensive Wall, the dynasty actually began with a very different strategic orientation. After rebel forces, led by a former peasant and temple boy by the name of Zhu Yuanzhang, overthrew the Mongols and founded the Ming in 1368, the court was initially not at all interested in building border walls. Instead, it attempted to follow in the tradition of the preceding Yuan dynasty and extend its influence over the various peoples living along its frontier. The Ming approached these groups through a combination of diplo­macy and aggression, either granting them tributary status or at­tempting to defeat them militarily. The Ming court entered into nonaggression pacts with many of the surrounding polities while also establishing a series of border forts and garrisons along its northern border.

[To view this image, refer to the print version of this title.]

Map of Chinese and Barbarian Nations [Hua yi tu], Song dynasty map of China (1040/1137).

Reprinted with the permission of Cambridge University Press.

The Ming’s expansionist aspirations were most evident under the dynasty’s third ruler, the Yongle emperor. The fourth son of the dy­nasty’s founder, Yongle came to power by seizing the throne from his own nephew, the Jianwen emperor, and proceeded to become one of the dynasty’s most powerful and dynamic leaders. Between 1405 and 1421, for instance, he directed six of the court eunuch Zheng He’s seven naval expeditions to destinations throughout

East Asia, the Indian Ocean, and even as far as the African coast. Yongle also transferred the dynasty’s capital from Nanjing back north to Beijing, where it been located during the Yuan, and from this location he devoted considerable energy to extending the court’s influence and authority into the region of the northern steppe. Be­tween 1410 and 1424—coinciding closely with Zheng He’s mari­time expeditions—the emperor launched a series of military cam­paigns against the Mongols. These attacks did not fundamentally alter the balance of power between the Ming and the Mongols, however. After the emperor died during the fifth and final cam­paign, in 1424, Yongle was succeeded by a series of rulers who gen­erally lacked his strength and ambition, and the court shifted from a strategy of military offensives to a reliance on tributary relation­ships with its neighbors.

The diplomatic crisis that provides the immediate historical backdrop for King Hu’s Dragon Gate Inn was the product of a se­quence of events that could be traced back to the appointment, in 1439, of a Mongol by the name of Esen Tayisi as leader of a tribal confederation known as the Oirats. Esen moved quickly to expand his influence within the region, and he also began dispatching in­creasingly elaborate tributary missions to Beijing, thereby requir­ing the Ming court to devote proportionally more resources to hosting the missions and reciprocating with “gifts” in return. The strain these tributary missions placed on the Ming court was exac­erbated by a series of natural disasters in the 1440s, including mul­tiple droughts, floods, famines, and bouts of pestilence and locust plagues, which affected virtually all regions of the country. The final straw came in 1448, when the Zhengtong emperor’s tutor and chief eunuch, Wang Zhen, rejected an Oirat tribute, owing to what he felt to be the excessive compensation the Mongols were asking in return. During the ensuing negotiations a Chinese interpreter sug­gested that perhaps a solution could be reached that would involve having one of Esen’s sons marry a Ming princess, but when Esen himself raised the possibility of this sort of heqin-style arrangement with the Ming court, he was unceremoniously rebuffed. He conse­quently resolved to launch an attack on Beijing the following year, and when the Ming learned of this threat it responded with a force of half a million troops, led by the twenty-two-year-old Zhengtong emperor himself.

The Zhengtong emperor’s troops proceeded northwest from Beijing, through the Juyongguan Pass and past the Datong garrison, whereupon they encountered the remains of a Chinese advance guard that had been slaughtered by the Mongol forces just days ear­lier. Confronted with this corpse-strewn battlefield, the Ming com­manders reevaluated their mission and concluded that it would be more prudent to return to the capital and simply declare victory. On their way back, however, the Ming forces set up camp at the Tumu postal relay station northwest of Juyongguan, where they were then routed by the Mongol forces and the emperor was taken prisoner.

The capture of the emperor had a devastating impact on the Ming court and immediately led to calls to abandon the Beijing cap­ital altogether and relocate to the south. The emperor, meanwhile, was negotiating furiously in an attempt to secure his own release, to the point of agreeing—in what would have been an intriguing re­versal of Han dynasty-style heqin arrangements—to marry Esen’s sister and take her with him back to the capital. In Beijing, however, the court acted quickly to cut the Mongols’ advantage by replacing the Zhengtong emperor, Zhu Qizhen, with his younger half brother, Zhu Qiyu. As a result, the newly promoted minister of war, General Yu Qian, was able to reject the Mongol attempts to ransom the cap­tured emperor’s life.

Realizing that his imperial hostage had become useless as a bar­gaining chip, Esen finally agreed to release the Ming leader the fol­lowing year, in exchange for a token ransom. To secure his freedom, Zhu Qizhen agreed to formally renounce all claims to the throne, essentially bartering his imperial status for his personal freedom. Even his freedom, however, proved fleeting, as his younger brother, the acting emperor, imprisoned him as soon as he returned to the capital and kept him under house arrest in the Forbidden City for the next seven years. The political awkwardness of the resulting ar­rangement was highlighted by the Mongols’ insistence on includ­ing, with each tribute they sent to the court during this period, a separate donation designated specifically for the former Zhengtong emperor—clearly intended to remind the Jingtai-led court of the unorthodox circumstances underlying its claim to power. After Zhu Qizhen managed to regain the throne in 1457, one of his first ac­tions was to order the execution of General Yu Qian that is featured at the beginning of Dragon Gate Inn.

The Tumu incident, as the crisis came to be known, presented a fundamental challenge to the authority of both the emperor and the Oirat leader. Even after the Zhengtong emperor regained the throne, the underlying authority of the imperial institution contin­ued to be seriously compromised by the political legerdemain that had allowed the court to preserve its power following the emperor’s original capture. At the same time, however, things were going equally poorly for the Mongols. While in theory the capture of the Chinese emperor should have been a coup for Esen, it became in­stead a manifestation of his own political inefficacy. He had not been able to use his hostage to extract tangible concessions from the Ming court, and while he proceeded to declare himself khan in 1453, he was nevertheless assassinated following an internal revolt only two years later.

The imperial crisis, combined with poor economic conditions at the time, contributed to a pivotal reevaluation of the Ming’s earlier expansionist policies. As early as the 1550s, following the Tumu de­feat in 1449, the Ming court set about reconstructing and strength­ening the walls and fortifications already in place in Juyongguan just outside Beijing, and over the next few decades it began to em­brace the wall-building strategy that would increasingly dominate its attention and resources. Just as the Han dynasty Wall was built in response to the Xiongnu’s virtual capture of Emperor Gaozu dur­ing the 200 bce Baideng debacle, it was the Mongols’ capture of the Zhengtong emperor at Tumu that provided the catalyst for con­structing the Wall we see today. Both military setbacks underscored China’s comparative military weakness vis-a-vis its northern neigh­bors, thereby encouraging the shift to an increasingly defensive and diplomatic strategy.

As had been true of Meng Tian’s original Qin dynasty Wall, the strategic challenge that motivated the initial Ming dynasty Wall construction was centered along the southern edge of the Ordos re­gion. By the mid-fifteenth century, the Oirat Mongols controlled much of the territory inside the northern loop of the Yellow River, and one of the Ming court’s central concerns was how to contain the threat they posed to the Chinese communities to the south. Sev­eral proposals were made in the 1470s for launching expeditions to drive the Mongols back beyond the Yellow River loop, but they were all deemed prohibitively risky, and even had they been suc­cessful, the Ming court would not have been able to afford to provi­sion the military bases that would have been necessary to con­trol the region. During the Han and the early Ming, the court experimented with establishing permanent military compounds in the frontier region, with soldiers and their families farming the land themselves so that they would not need to rely on the court for pro­visions. In theory, the children of these frontier soldiers would in­herit their fathers’ positions, yielding a self-replicating population that would guard the border in perpetuity. The problem, however, was that these border regions were barely arable, and even ef­ficient farmers (which the soldiers, presumably, were not) would have had considerable difficulty living off the land. Consequently, by the mid-Ming this model had been effectively abandoned.

Given that the Ming court deemed it militarily impractical to at­tempt to retake control of the Ordos, and politically undesirable to try to reopen formal trade relations with the Mongols, they there­fore decided that one of their few remaining options was to con­struct a border wall along the southern edge of the Ordos region. The first such wall was proposed in 1471, when the newly ap­pointed magistrate of Xi’an, Yu Zujun, petitioned the emperor to have a ten-meter-high wall constructed at the southern end of the Ordos to help defend the strategically important town of Yulin. This 1,700-Zi tamped-earth wall was completed in 1474, and was followed over the next couple of decades by a network of earthen walls throughout the southern edge of the Ordos (some of which are still visible in western China). In 1485, Yu Zujun recommended building a similar wall farther east, but this structure ended up be­ing abandoned due to debates within the Ming court.

In addition to providing the foundation for the Ming’s subse­quent brick and stone Wall, these fortifications along the Ordos coincided with a broader set of diplomatic shifts that made sub­sequent wall building all but a foregone conclusion. In particu­lar, the Ming court became increasingly disinclined to engage the Mongols through either diplomacy or trade, thereby—ironically— motivating the Mongols to acquire through military raids the goods and provisions they had previously obtained through tributary ex­change. Those raids, in turn, drove the Ming court to build even more walls, making it even more disinclined to renew large-scale tributary relations with the Mongols.

In the mid-sixteenth century, the Ming began shifting from earthen walls to brick and stone constructions. These new walls were not only significantly more durable than the tamped-earth structures that had preceded them, but they also required a dramat­ically greater investment of resources. It has been calculated, for in­stance, that it would have taken approximately a hundred men to construct the same length of stone wall that a single man could build using the old tamped-earth method. These brick structures date to as early as the 1530s, and others were still being built when the dynasty collapsed a century later. While construction proceeded in a piecemeal fashion, generally speaking wall construction in western regions would push the threat of Mongol raids to areas far­ther east, creating the need to build additional walls.

The vast expense of these new walls had the effect of locking the Ming court into a defensive orientation wherein one of its only op­tions was to continue building more walls. In the 1550s, the Ming court again debated whether to renew its attempts to drive the Mongols from the Ordos region but concluded that the cost of such an offensive would be unacceptable. The court instead focused its energies on continuing the wall building that was already under way, ultimately yielding the brick and stone Great Wall we see today.

Despite the Ming’s vast investment in Wall construction, the Manchus continued their raids deep into central China up until the fall of the dynasty. The proximate cause of the dynasty’s even­tual collapse, furthermore, was not an invasion from without, but rather a rebellion from within. First, a peasant soldier by the name of Li Zicheng led an internal revolt and managed to take over Beijing. When he heard the news, a Chinese general by the name of Wu Sangui, who was guarding the Shanhai Pass, decided to allow the Manchu forces to pass through—apparently hoping that they would remove Li Zicheng from the throne. As it turns out, the Manchus did precisely that, but then they proceeded to establish themselves as China’s new dynastic house. Thus, despite the vast re­sources the Ming had invested in building its Wall, its defenses were ultimately breached not on account of any material weakness of the structure, but as a result of the weakness of will of those assigned to guard it.

We may find a concise articulation of the logic underlying the construction of the Ming Wall in Walt Disney Studios’ 1998 ani­mated feature Mulan, which retells a famous legend about a young

girl who secretly dressed as a man in order to take her invalid fa­ther’s place when he was conscripted to serve in the imperial army. At one point in the movie, the Xiongnu chanyu—or Shan-yu, as he is called in the film—states that by building the Wall, the Chinese emperor was merely challenging his (the chanyu’s) strength, thereby virtually inviting him to try to attack. Though set in a period ap­proximately a millennium before the Ming built its Wall, this scene articulates quite succinctly the feedback loop that would eventually drive the Ming’s Wall-building project, in that the construction of the Wall directly inspired even more of the same raids that it was ostensibly defending against in the first place.

While the Disney feature aptly summarizes the logic underlying the construction of the Ming Wall, the film’s representation of the Wall itself is more problematic. The Hua Mulan legend is generally set in the period between the Han and the Tang, and this historical setting is corroborated by the film’s specification that the northern forces are led by a Xiongnu “Shan-yu.” The Wall that appears in the film, however, is essentially the brick and stone construction we see today and not the more modest tamped-earth structure that would have existed at the time.

This historical anachronism is not unique to Disney, and indeed an astonishing array of contemporary texts project a version of the Ming dynasty’s brick and stone Wall back onto pre-Ming peri­ods. For instance, when Eugene O’Neill (a well-known Sinophile who had repeatedly been tempted to write a play about China’s First Emperor) wrote his 1926 play Marco’s Millions a year af­ter visiting China for the first time, he described the legendary thirteenth-century explorer Marco Polo encountering “the Great Wall of China with an enormous shut gate.”2 In John Ford’s 1938 film on the same subject, the peripatetic Venetian, played by Gary Cooper, similarly enters the Central Kingdom through a gate in the Wall (with his father’s burly accountant slung over his shoul­ders), and a more recent 1982 television miniseries directed by

Giuliano Montaldo—which happened to be the first Western pro­duction filmed on location in China since World War II—also has the explorer entering through a massive Wall.

The problem with each of these images of Polo’s arrival in China, however, is that Polo himself made no mention of any such great stone Wall, or indeed any Wall at all. Commentators have long puz­zled over Marco Polo’s failure to mention the Wall in the travelogue he composed after returning to Italy, and some have even argued that this sort of omission is evidence that the Venetian must not have made it to China in the first place. I would suggest, however, that the more interesting question is not what this “omission” tells us about Marco Polo and his travels (or possible lack thereof), but what our fascination with the omission reveals about our own as­sumptions about the Wall.

Our knowledge of Marco Polo’s journey is derived almost en­tirely from a text he dictated in a Genoa prison in the latter half of the thirteenth century. Replete with extraordinary descriptions of the life and customs of the Orient and of the magnificence of the Mongol court, this volume became one of the most popular and influential books of its time. Columbus, for instance, took a heavily annotated copy with him on his voyage to the New World, and Polo’s description of Kublai Khan’s opulent summer palace in Shangdu inspired Coleridge’s famous description of the khan’s “stately pleasure dome.” Polo’s travelogue helped plant the seed for a more general fascination with the Orient that would burgeon over the next few hundred years, and therefore it is fitting that his own entry into China is marked—in each of these modern adaptations— by his traversal of that most famous of Chinese icons, the Great Wall.

Marco Polo’s family were merchants, and in the 1260s his fa­ther and uncle, Niccolo and Matteo Polo, made their way along the Silk Road across Central Asia to the Yuan capital of Cambaluc (Beijing). They were received by Kublai Khan himself and remained in China for three years until Kublai sent them back, along with his own ambassador and a letter to the pope. The pope, as it turned out, had passed away the preceding year, but as soon as a new pope was elected in 1271, he was given the letter the Polos had brought from Kublai and responded by sending them back to China bearing gifts for the Mongol leader. The Polos, this time accompanied by Niccolo’s son Marco, again followed the Silk Road to China. They would remain there for the next seventeen years, during which time Kublai Khan allegedly appointed Marco to his Privy Council, and then made him a tax inspector in Yangzhou. In 1291, the Polos were permitted to return to Venice, where Marco regaled his friends and acquaintances with stories of his experiences in China. When, in 1298, he was imprisoned in Genoa during a military skirmish be­tween Genoa and Venice, he dictated the account of his travels to his cellmate, Rusticiano de Pisa, and it was de Pisa who subse­quently composed (in old French) the text we now know as The Travels of Marco Polo.

Although Marco Polo’s volume came to be known in Italian as Il Milione (“The million [lies]”)—so called by contemporary readers skeptical of its veracity—it nevertheless quickly became one of the best-selling and most influential books of the period. Initial doubts about the text’s truthfulness, however, were reinforced as subse­quent travelers bought back more detailed information about China, leading readers of Polo’s volume to puzzle over its apparent omissions, including its lack of any reference to such distinctively Chinese elements as calligraphy, tea drinking, chopsticks, or foot binding. Some of these apparent oversights no doubt had plausible explanations. Given that Marco Polo was not well educated, for in­stance, it is not surprising that the Chinese writing system might not have made a big impression on him. It has also been observed that tea drinking, in Marco Polo’s time, was popular in southern China but less so in the central and northern regions where he would have spent most of his time. Similarly, during the Yuan it was primarily elite women who bound their feet, and Marco Polo presumably would have had little opportunity to meet such women in person.

The omission that has caused the most consternation, however, was Polo’s failure to make any reference to the one structure that had become virtually synonymous with the Chinese nation. As early as 1747, Thomas Astley asked skeptically: “Had our Venetian been really on the Spot. . . how is it possible he could have made not the least Mention of the Great Wall: the most remarkable Thing in all China or perhaps in the whole World?”3 Half a century later, George Staunton, Lord Macartney’s second in command during his historic trip to Beijing from 1792 to 1794, did some outside re­search of his own and came up with the following explanation:

A copy of Marco Polo’s route to China, taken from the Doge’s Li­brary at Venice, is sufficient to decide this question. By this route it appears that, in fact, that traveller did not pass through Tartary to Pekin, but that after having followed the usual track of the caravans, as far to the eastward from Europe as Samarcand and Cashgar, he bent his course to the south-east across the River Ganges to Bengal, and, keeping to the southward of the Thibet mountains, reached the Chinese province of Shensee, and through the adjoining province of Shansee to the capital, without interfering with the line of the Greater Wall.4

Staunton’s meticulous account of Polo’s route, however, is at odds with Polo’s own account, which describes him traveling through the province of Tenduc, north of the Ordos. More recently, the British librarian Frances Wood has echoed and responded by making more explicit the suspicion that Astley had already articu­lated centuries earlier—namely, that Polo must not have gone to China at all, and instead was merely repeating and elaborating on stories he had heard from Arab merchants who had traveled to the region.5

Irrespective of whether or not Marco Polo ever reached China,

there is a straightforward explanation for why his narrative makes no mention of the Wall: namely, that for all practical purposes there would have been no Wall for him to have seen. Polo made his journey more than two centuries before construction began on the Ming dynasty brick and stone structure we see today. The Yuan had displayed no interest in border wall construction, and even if Polo had encountered the remains of, say, the earlier Jin dynasty walls, those dilapidated packed-earth structures would not neces­sarily have captured his imagination. The recurrent astonishment at the absence of any reference to the Wall in Polo’s travelogue, there­fore, speaks not so much to the question of the authenticity of Polo’s text as to the powerful anachronistic pull of our contempo­rary notion of the Wall.

Around the time Marco Polo’s travelogue was published, a stream of Westerners began traveling to China as the Silk Road be­came an increasingly important conduit of overland trade between Europe and Asia. It was not until around the sixteenth century, however, that Western visitors begin alluding to the Wall, and even the initial accounts tended to be comparatively restrained. In 1559, for instance, Gaspar da Cruz noted that “the Chinas have an hun­dred leagues (others saying there are more) of a Wall betweene them and the other,” while the legendary Jesuit Matteo Ricci, who was in China from 1583 until his death in 1610, notes that “to the north the country is defended… by precipitous hills, which are joined into an unbroken line of defense with a tremendous wall four hundred miles long.”6

Even as Ricci and the other earlier Jesuits were describing a Wall of comparatively modest dimensions (at least as compared with the ultimate length of the Ming structure by the time the dynasty col­lapsed about half a century later), we find a much more impressive version of the Wall in a Jesuit map from the same period. Dated circa 1590, the map is considered the first modern European map of China and features an iconic representation of the Wall stretching

[To view this image, refer to the print version of this title.]

Sinarum Regni alioru[m]q[ue] regnoru[m] et insularu[m] illi adiacentium descriptio, anonymous, possibly after Matteo Ricci and Michele Ruggieri (ca. 1590).

Courtesy of the Hong Kong University of Science and Technology.

continuously from the Korean Peninsula straight across northern China to Shaanxi Province in the west. The map, titled Sinarum Regni alioru[m]q[ue] regnoru[m] et insularu[m] illi adiacentium descriptio, is unsigned, but the prominent designation of Ricci and Ruggieri’s Jesuit church in Canton, labeled on the map “ecclesia patrum societatis” (the Church of the Fathers of the Society [of Je­sus]), suggests that it was probably based on Ricci’s own surveys and notes.

The paradox that this Jesuit map presents the Wall as being much more extensive than it appears in the descriptions by Ricci and the other Jesuits from the same period, may be explained by the fact that this Jesuit map was also drawn from indigenous Chinese sources. The descriptive table on the right side of the map, for in­stance, lists the administrative and regional divisions for each prov­ince and appears to have been borrowed from the sixth (1579) edi­tion of Luo Hongxian’s influential atlas of China, Enlargement of the Terrestrial Map, which also includes a distinct icon of the Wall (though earlier editions actually did not). Structured on a grid and visually rather different from the Sinarum Regni, Luo’s map presents roughly the same configuration of China, including the same curiously narrow strip of Gobi Desert in the upper left (ren­dered in black in Luo’s map). Most tellingly, both works present a similar representation of the Wall.

One collection of influential European maps that may have been partially derived from the cartographic tradition inspired by the Enlargement of the Terrestrial Map was the Atlas Sinensis created by Joan Blaeu in 1655, which was based on the maps provided by Jesuit cartographer Martino Martini. This atlas, the first European atlas of China, contains maps of each of China’s provinces—several of which clearly feature an icon of a crenellated Wall—together with a map of the country as a whole, which similarly features a Wall stretching from the Gulf of Bohai to the Ordos. Martini him­self noted that

[To view this image, refer to the print version of this title.]

Enlargement of the Terrestrial Map [Guang yu tu], Luo Hongxian (1579).

Reprinted with the permission of Cambridge University Press.

this celebrated wall is very famous. . . longer than the entire length of Asia. . . . The person who began this work was the emperor Xius. . . . He built this wall starting in the twenty second year of his reign, which was 215 years before Christ. In the space of five years, which is incredibly short, it was built so strongly that if anyone was able to slip a nail between the cut stones, the builder of that part would be put to death. . . . The work is magnificent, huge, and admi­rable, and has lasted right up to the present time without any injury or destruction.7

Martini was, of course, mistaken when he claimed that the Wall constructed by the First Emperor had “lasted right up to the present time without injury or destruction,” though his underlying vision of a “magnificent, huge, and admirable” Wall remains alive and well

[To view this image, refer to the print version of this title.]

Imperii Sinarum Nova descriptio, Joan Blaeu and Martino Martini (1655).

Courtesy of the Hong Kong University of Science and Technology.

even today. Transcending any specific physical incarnation or ab­stract understanding of the structure, it is precisely this vision of the Wall as an enduring monument that anchors its historical continu­ity and resilience.

By the time Lord Macartney led his historic mission to China in 1792, the Wall had already become a potent image in the European imagination. After the Macartney expedition visited the Gubeikou section of the Wall northeast of Beijing, several members wrote de­tailed accounts of the structure. Macartney, for instance, recorded in his diary:

The wall is built of bluish coloured brick, not burnt but dried in the sun, and raised upon a stone foundation, and as measured from the ground on the side next to the Tartary, it is twenty-six feet high in the perpendicular. The stone foundation is formed of two courses of granite equal to twenty-four inches or two feet. From then to the parapet including the cordon which is six inches are nineteen feet four inches, the parapet is four feet eight inches. From the stone foundation to the cordon are fifty-eight rows of bricks and above the cordon are fourteen rows; and each row, allowing for the interstices of the mortar and the insertion of the cordon may be calculated at the rate of four inches per brick.8

The extraordinary precision of these measurements reflects Macartney’s intellectual and diplomatic training but may also have stemmed from a desire to assert a sort of intellectual mastery over this daunting structure. To the extent that the significance of the Wall traditionally lay not only in its status as a physical barrier but also in its assertion of symbolic mastery over the territory in ques­tion, Macartney’s emphasis on the hyperprecise measurements of the structure could be seen as mirroring the symbolic function of the Wall.

The Macartney expedition’s descriptions of their trip to Gubeikou addressed not only the Wall itself but also the reactions of their Chinese hosts. Macartney and Staunton note, for instance, that while they were investigating the Wall, their escorts merely gazed at it “with perfect indifference; and few of the mandarins who accompanied the Embassy seemed to pay the least attention to it.”9 While the “perfect indifference” that Macartney and Staunton attribute to their Chinese escorts is, of course, contrasted with the amazement that they themselves felt upon viewing the structure for the first time, it may well have reflected their hosts’ ambivalent atti­tude toward the Wall that the Ming had tried to use to defend against the same Manchus who now ruled China.

A rather different perspective on the Wall’s political connotations during this period, however, may be found in Staunton’s description of an interaction the expedition witnessed: when “a Tartar, one of the attendants, [was] ordered to be punished by some of the Chi­nese mandarines, for misbehaviour, the man made a vigorous resis­tance, and exclaimed in a loud voice, that no Chinese had a right to inflect punishment on a Tartar after having passed the great wall.”10 If accurate, this anecdote underscores the Wall’s contemporary sta­tus as the product of a double political reversal. First, the original strategic function of the Ming Wall was obviated after the Manchus overthrew the Ming and established the Qing. Second, the Man — chus themselves immediately almost began constructing an exten­sion of the Wall that commenced at the traditional eastern terminus at Shanhaiguan and extended east along the borders of the tra­ditional Manchu homeland of Manchuria. Consisting of parallel earthen levees separated by a trench and planted with densely ar­ranged willow trees—and overlapping in some sections with exist­ing eastern sections of the Ming Wall—these Qing fortifications were designed to restrict Han (and Mongol) entry into Manchuria. In the Macartney anecdote, therefore, we find a Manchu servant re­monstrating his Han superiors (who were themselves working un­der the ethnically Manchu Qing court) about respecting the sanctity of a border barrier that had originally been intended to keep the Manchus themselves out of China, but that had subsequently been partially reinvented by the Manchus as a barrier to keep the Han Chinese out of Manchuria—though it should be noted that the Gubeikou section of the Wall where the Macartney expedition wit­nessed this incident was actually significantly to the west of the Manchurian defense line.

Western and Chinese discourses on the Wall continued to diverge through the nineteenth century, with the West increasingly perceiv­ing the monument as a transcendental symbol of the strength and resilience of the Chinese civilization, while in China the structure carried more ambivalent associations of the Ming’s failed defense against the Manchus together with the First Emperor’s notorious tyranny. A reflection on the Wall’s bifurcated trajectory in China and the West can be found in a curious speculation in one of Henry Yule’s notes to his extensively annotated 1870 translation of Marco Polo’s thirteenth-century travelogue. Commenting on a passage de­scribing Polo’s arrival in the province of Tenduc, just north of the Ordos region, Yule observes that “it has often been cast in Marco’s teeth that he makes no mention of the Great Wall of China, and that is true; whilst the apologies made for the omission have always seemed to me unsatisfactory.” Yule then proposes what might ap­pear to be a rather bizarre explanation of his own for this omission, noting that Polo makes a curiously mediated reference to the his­tory of the “Tartars” by referring to what he calls “the country of Gog and Magog.” Yule notes that Polo says, “Here also is what we call the country of Gog and Magog; they, however, call it Ung and Mungul, after the names of two races of people that existed in that Province before the migration of the Tartars. Ung was the title of the people of the country, and Mungul a name sometimes applied to the Tartars.”11 Polo suggests that the biblical terms Gog and Magog refer to two “races of people” that are known in the East as Ung and Mungul—with the latter corresponding to the Mongols or, as Polo prefers to call them, the Tartars.

Yule finds Polo’s reference here to Gog and Magog bewildering if taken at face value, and argues that the passage must instead be read as an elliptical commentary on the Wall: “Yet I think, if we read ‘between the lines,’ we shall see reason to believe that the Wall was in Polo’s mind at this point of the dictation, whatever may have been his motive for withholding distincter notice of it. I cannot con­ceive why he should say: ‘Here is what we call the country of Gog and Magog,’ except as intimating ‘Here we are beside the Great Wall known as the Rampart of Gog and Magog,’ and being there he tries to find a reason why those names should have been applied to it.”12 If we set aside the peculiarity of Yule’s claim that Polo, rather than simply discussing the Wall directly, would instead have chosen to substitute for it with a highly elliptical reference to two biblical figures who bear at most a tangential association to another legendary wall, he does succeed in raising a very interesting issue re­garding our understanding of the identity and historicity of the Great Wall.

Yule argues that Marco Polo was using the legend of Gog and Magog to suggest an equivalence between China’s Wall and what Yule calls the “Rampart of Gog and Magog.” This latter is located in the narrow pass of Derbend in the Caucasus where, according to legend, Alexander the Great built an enormous iron gate to block out the “Tartars” and prevent them from invading Europe. Marco Polo himself refers to this legendary Iron Gate at another point in his text, but without mentioning the Gog/Magog connection: “Al­exander caused a very strong tower to be built there, to prevent the people beyond from passing to attack him, and this got the name of the Iron Gate. This is the place that the Book of Alexander speaks of, when it tells us how he shut up the Tartars between two moun­tains; not that they were really Tartars, however, for there were no Tartars in those days, but they consisted of a race of people called Cominians and many besides.”13 Yule claims, in other words, that the Wall is present in Polo’s text precisely as a conspicuous absence, arguing that Polo speaks here of the Gog and Magog in order to “intimate” an equivalence between China’s Great Wall (which is not mentioned anywhere in Polo’s text) and Alexander’s Iron Gate (which Polo does mention, though in an unrelated section of his book).

Although Yule’s argument about Polo’s silent conflation of the First Emperor’s Wall and Alexander’s Iron Gate is probably sheer fantasy, this theory nevertheless points to the entirely plausible pos­sibility that the legend of the Qin dynasty Wall and its successors may have helped inspire the legend of the immense iron gate that began to appear in the Alexander Romance corpus around the sixth century ce. Just as apparent cognates of Qin used as the name of China’s first dynasty began appearing in Hindi, Greek, and Latin as early as the first and second centuries ce, it is not unreasonable to suppose that the legend of an enormous Qin Wall separating China from the Xiongnu might have made its way to Europe via Asian and Middle Eastern traders, where it could have helped provide the inspiration for a similar legendary Wall separating Europe from the Central Asian “Tartars.” Furthermore, this possibility that the Qin Wall may have helped inspire the legend of the Alexandran wall centuries later and thousands of kilometers away suggestively mir­rors the way in which the Ming Wall provided the catalyst for the subsequent development of the Western notion of an iconic “Great Wall of China.”

This point about the potential relationship between the Qin and Alexandran walls is, of course, mere speculation. I use it here to suggest one plausible bifurcation in the tradition inspired by the original Qin dynasty, and also I use it to illustrate more generally the critical role played by these sorts of speculative processes in the constitution of the historical reality of the Wall. The vision we have inherited of the Ming Wall as the product of an unbroken historical continuity dating back to the Qin is itself, in a very real sense, the product of a continual process of (unconscious) speculation—a tra­dition of speculation that has gradually come to assume the status of social reality.

We find an interesting commentary on this speculative recon­struction of the Wall’s origins in an essay by Borges. Inspired quite possibly by Kafka’s 1917 parable about the Wall as a unity of gaps (which Borges once described as Kafka’s “most memorable” work), this 1950 essay takes as its starting point the relation between the First Emperor’s virtually simultaneous book-burning and wall — building projects.14 Borges notes that these acts mirror each other quite precisely—with one attempting to erase the past and the other intended to help secure the dynasty’s future. He considers several possible ways of reconciling these two visions of the First Em­peror—speculating that the Qin sovereign could be seen as either “a king who began by destroying and then resigned himself to conserv­ing, or… a disillusioned king who destroyed what he had previ­ously defended”—and decides that both of these conjectures “are dramatic but lack, as far as I know, a historical basis.”15 He con­cludes that the Wall itself may be best approached as merely “a metaphor,” in the sense that its significance lies in its symbolic, rather than its strictly material, status.

In his speculative reconstruction of the origins of the Wall, Borges is in effect applying a version of a narrative model he devel­oped a decade earlier in another China-themed work, “The Garden of Forking Paths.” This seminal story (his first to be translated into English) describes a London-based Chinese professor whose ances­tor had dreamed of creating a monumental labyrinth in which “all men would lose their way.”16 This labyrinth turns out to be a novel that attempts to detail all possible futures for every present mo­ment, yielding a “garden” of infinitely bifurcating counterfactual alternatives. In Borges’s subsequent essay on the First Emperor’s Wall, he presents a similar garden but in reverse—taking a concrete historical eventuality (the First Emperor’s legendary acts of wall building and book burning) and working back to reconstruct all of the possible counterfactual scenarios that could explain the logical relationship between these two actions.

Beyond its relevance to Borges’s fable of the Wall, this forking — garden metaphor also provides a useful model for understanding the history and historicity of the Wall. Although the Wall is fre­quently imagined as a paradigmatically linear entity, in reality it is characterized by a continual series of bifurcations. During the pe­riod from the Han to the Ming, for instance, the Wall’s historical trajectory repeatedly branched off in different directions, as strate­gies of border-wall construction were appropriated by a variety of peoples and regimes positioned along the nation’s northern periph­ery. The symbol of the Wall also diverged from the material struc­ture, as the abstract ideal of a frontier Wall continued to retain significant purchase even during regimes that had no interest in building border fortifications. After the fall of the Ming an idealized vision of the Wall developed in Western discourses, largely indepen­dent of how the structure was understood back in China, and it would not be until the twentieth century that the Western and Chi­nese visions of the Wall would begin to reconverge.

While theoretically it would be possible to separate the Wall into each of its geographic, historical, and conceptual strands— to speak, for instance, of the Badaling section of the Ming Wall as imagined during the early twenty-first century—this approach would not accord very well with our own intuitions about what the Wall really is. We tend to imagine the Wall as a unitary and continu­ous entity, even while consciously recognizing the physical, his­torical, and conceptual specificity of its components. In practice, therefore, the Wall is generally conceived as the sum total of its in­dividual parts—as a Borgesian garden that encompasses all of its di­vergent strands.

While King Hu’s Dragon Gate Inn is positioned at a critical junc­ture in the Wall’s history, we find another perspective on the Wall’s contemporary significance if we trace one of the subsequent strands of the film itself. In 2003, the Taiwan-based director Tsai Ming — liang released an homage to King Hu’s film. Entitled Goodbye, Dragon Inn, this recent work is structured around a screening of Dragon Gate Inn at Taipei’s historic Fu-ho Theatre, on the eve of the theater’s scheduled demolition.17 Simultaneously reflecting on the transience of physical constructions (the theater) and the resil­ience of cultural productions (King Hu’s film), Tsai Ming-liang’s Goodbye, Dragon Inn explores the way in which a promise of im­pending destruction may provide the basis for an anticipatory sense of spectral return.

Though Tsai’s homage is set in contemporary Taipei, it opens with an embedded clip of King Hu’s original prologue set in Beijing. King Hu’s mysterious wall circles do not appear anywhere in Tsai’s homage, though two sets of suggestively similar circular wall mark­ings had in fact begun to appear on dilapidated Beijing walls in the years leading up to the release of Tsai’s film. First, in the 1990s it had become increasingly popular, in Beijing and other major Chi­nese cities, to mark buildings slated for demolition with the Chinese character chai, meaning “to demolish,” circumscribed by a white circle. Second, during this same period the artist Zhang Dali began anonymously drawing white circular markings on many of those same Beijing buildings—the markings being iconic images of his own profile, which originated out of his sense of isolation and cul­tural alienation. Through the fortuitous coincidence of the chai characters and Zhang Dali’s autographs, we may discern the out­lines of a logic underlying the structure and function of the contem­porary Wall. While the chai characters anticipate the imminent de­struction of the buildings on whose walls they appear, Zhang Dali’s graffiti, by contrast, developed out of the artist’s attempt to answer his sense of cultural alienation with an anonymous assertion of identity and presence.

These intersecting themes of destruction and preservation are brought together in a popular joke that contemporary China has become a nation of Chai-na—literally, a nation of “demolish that.” Punning on both the English word China and the late-nineteenth — century Japanese term for China, Shina, this contemporary neolo­gism uses the same character, na, to render the second syllable of China. Although in the Japanese term this na was used strictly for its phonetic value, in the Chai-na neologism it is also used for its se­mantic value, as the pronoun that. Na belongs to a category of words linguists call “shifters,” meaning that their concrete referent is contingent on the specific context in which they are uttered. We could, by extension, also see the neologism Chai-na as a sort of shifter—reflecting the fact that our understanding of the nation ulti­mately depends on the perspective from which we happen to per­ceive it.

The joke that China has become a nation of “demolish that” presents a version of what Foucault, in a passage discussed in Chap­ter 1, calls “the stark impossibility of thinking that.” In other words, in contrast to a conventional vision of the nation as a funda­mentally unified and historically continuous entity, the Chai-na ap­pellation reimagines China as the product of a continual process of destruction and reconstruction. It is precisely in this challenge of at­tempting the impossibility of “thinking that,” however, that we find a potential explanation for the conventional assumptions about (national) identity. In presenting the nation as a space of demo­lition, the Chai-na joke suggests a view of identity as grounded not on continuity but on a continuous process of destruction and reinvention. By a similar logic, the Wall itself could be seen as a product not so much of historical continuity and physical unity as of continuous divergence and rupture—with the identity of the Wall lying not in any single historical strand but in the collective “garden” that contains all of these intertwined “forking paths.” These contemporary mural markings, therefore, bring us back full circle to the wall circles in King Hu’s film. Each set of inscrip­tions symbolizes the processes of destruction, transformation, and erasure that have characterized the Wall throughout its history, even as the structure’s survival and resilience is ultimately predi­cated on these same transformative processes. It is, in other words, precisely in the Wall’s ability to branch off in different directions that we find the key to its coherence as a transhistorical entity.