by Ling Zhang
4.1 Why Hebei?
A brief survey of historical sources leading up to the 1048 flood indicates that the river naturally tended toward Henan – the land in the south – rather than toward Hebei. From various Tang, Five Dynasties, and Song sources, I single out forty-two years from a span of 350 years prior to 1048, in which the river's floods, bank ruptures, and course shifts reportedly took place.5 Among these forty-two years, thirty-six give us some rough indication of the locations where the events occurred or the areas that the floods affected. They demonstrate the river's tendency to shift between two trajectories, northward into Hebei or southward into the heart of Henan (see Table 2). What is immediately obvious from a cursory glance at this information is that by the late tenth century, the river consistently flooded the area to its south. By the end of the tenth century, however, the river began to flood the area to its north with greater frequency. Does this mean that the river developed a “natural” tendency toward the north, and the 1048 event was merely the culmination of the centuries-long tendency?
Table 2. The Yellow River's Floods before 1048
YearDirection of FloodingYearDirection of FloodingYearDirection of FloodingYearDirection of Flooding
722 North 941 South 972 South 1000 South
813 South 944 South 973 South/North 1011 North
829 South 946 South and North 977 South 1012 North
891 South 954 South and North 978 North 1014 North
918 South 964 South and North 979 North 1015 North
923 South 965 South 982 South/North 1019 South
931 South 966 North 983 South 1028 North
932 South and North 968 North 992 North 1034 North
939 South 971 South 993 North 1041 North
A close examination of these records leads to quite a different picture. First, the records suggest that most northerly floods happened on a small scale; by contrast, the southerly floods led to more destructive, widespread damage. In many cases, the river even shifted its entire course toward the south to affect the Huai River valley. In terms of scale and severity of floods, the river seemed to tend toward the south. Second, the exacerbation of the river's flooding problems went hand in hand with the establishment of strong central authorities in north China centering on the land of Henan. Their increasing commitment to flood control led to the construction of hydraulic works along the southern banks of the river. Such efforts helped prevent southerly floods. But as their unintended consequence, these efforts pushed the river's hydrological force to concentrate on the weaker northern banks, thereby increasing the chance of northerly floods. If the river indeed developed a northward tendency from the end of the tenth century, it was most likely a product of the deliberate hydraulic efforts, which prioritized the south at the expense of the north.
A scrutiny of a few floods and hydraulic practices demonstrates how the states before and including the Song intervened more and more in the movements of the Yellow River. In 918 and 923, two different generals of the Late Liang Dynasty 後梁 (907–923) commanded their armies to breach the river's dykes in order to create floods that would halt the advancement of their enemies from the north.6 The floods let loose by these actions were supposed to submerge the enemy's land in the north, but in reality, they plagued the core territory of the Late Liang, Henan. Embroiled in the war, the Liang never even tried to ameliorate the disasters. In the following decades, the destabilized Yellow River flooded Henan in 931, 932, 939, 941, 944, and 954, with catastrophic effects.7
Only the flood of 954 led to the beginning of strong flood-control efforts from the Chinese state then in power, the Late Zhou Dynasty 後周 (951–960). The Zhou, unlike the negligent Late Liang, immediately recruited corvée laborers to repair the bank rupture and to stop the flood from extending farther south.8 The Zhou emperor, Shizong 世宗 (921–959), was young, energetic, and capable, and he sought to pacify the violent river as he had conquered his military enemies.9 Through his efforts, the broken dykes were repaired, the river's tendency toward the south was blocked and its water was pushed northward. These actions created a vast stretch of waterlogged land in southern Hebei and northern Henan. Because these waters did not threaten the political core area of his state, the emperor made no further effort to handle them. Year after year, these stagnant waters ravaged the northern land unchecked, hurting the livelihood of the people, particularly those in southern Hebei.
The troubling situation remained unchanged through 964, four years after the Northern Song Dynasty had replaced the Late Zhou. Local residents and officials in southern Hebei petitioned the Song court to have the water problem fixed.10 Such requests failed to attract any substantial attention from the court. The court claimed that it could not afford the immense financial and labor costs required for the hydraulic work. It also noted that if the stagnant waters were to leave southern Hebei, it would need to flow somewhere else. Who wanted to take over the waters? Certainly not the core political area in central Henan on the southern side of the river. Like the Late Zhou before it, the Song court chose to ignore the problem and stick with the status quo, much to the detriment of southern Hebei. A hydraulic approach, “protect the south and ignore the north,” emerged. Althought it was not intended as a deliberate assault on the people in the north, it more or less generated the same effect.
The Song's southern bias dominated its later hydraulic practices. In 965 alone, a series of bank ruptures and floods broke out in the metropolitan area of Kaifeng as well as Mengzhou 孟州, Chanzhou, and Yunzhou 惲州 – all within a distance of 200 kilometers from Kaifeng. Later, floods occurred in Huazhou in 966 and 967, in Chanzhou in 971, and in Chanzhou, Puzhou 濮州, and Kaifeng in 972.11 The majority of the territories of these districts were located in Henan, suggesting that the river continued to flood toward the south, thus threatening the capital at Kaifeng. In each case the Song government followed the Late Zhou precedent of 954: it immediately repaired the bank ruptures without any hesitation despite heavy costs. It is obvious that the state had become very sensitive to the river's threat to Henan, the land to the south of the river, and would do whatever it could to protect it.
The Late Zhou and Song states’ treatments of the floods suggest four things. First, the river's bank ruptures and floods became increasingly frequent through the tenth century, of which the Song state was well aware. Second, serious floods continued to occur on the southern side of the river, and on several occasions the river even shifted its entire course into Henan. The state fixed the problems promptly and by doing so, pushed the river back to the north. There is no historical or modern scientific evidence to suggest that the river had naturally shifted away from the south. Rather, it was human forces – the imperial state's deployment of certain policies and various resources – that prevented the river from remaining in the south. Third, the state's interventions did far more than block the southerly floods. As historical geographer Li Xiaocong has specified, by repairing and constructing dykes, the state's technical solutions forced the river to press northward, thereby increasing both the chance and severity of northerly floods.12 The increasing number of the northerly floods, which Table 1 exhibits, was the consequence of both the southerly floods and the state's hydraulic practices – the latter sacrificed the land in the north for the safety of the politically important Henan. Lastly, one may argue about the creditability of the historical records. The records for the southerly floods were perhaps better constructed and preserved; yet, those for the northerly floods might be poor, implying that there were perhaps more northerly floods than that recorded in real history. This argument, in effect, lends credence to my argument by pointing out the state's neglect of the land north of the river, a neglect that left Hebei poorly protected and vulnerable to disasters.
As we can see, behind all of these “natural” disasters lurks a series of responses of the imperial state, whose calculus of sacrifice and salvation came to have dire consequences for Hebei. These early approaches of the state to disasters and envi
ronmental management drove the further development of the state's hydraulic policies and practices over the next few decades. During that period, Hebei loomed large as the state's chosen environmental victim who, by becoming the river's flooding ground, saved the land in the south and guarded the greater good for the state.
4.2 To Flood South or to Flood North – A Politico-Hydraulic Enterprise
The State Perception of Geopolitics and the River
The Song state was involved with flood control from the moment of its founding. The first emperor, Taizu, immediately assumed the role of “hydraulic leader” (following Karl Wittfogel's notion13) and set up official positions in every district along the river's lower reaches to take charge of flood-control works. His government went to great lengths to prevent floods by investing large sums of money, recruiting corvée labor, planting trees along riverbanks, and building and repairing river dykes. When floods and bank ruptures did occur, the government treated them as emergencies, and took prompt actions to cope with them.14 To Taizu, the institutional and technical efforts were to pacify the river's ferocious torrents and restore its peaceful, tranquil state before the tenth century.
The more the state labored to prevent floods, however, the more hopeless the situation became: flooding events grew more frequent and more severe. The river continued to send down tremendous amounts of silt, and in the lower stretches of the river the extraordinary speed of sedimentation and the resulting increase in the force of water flows overwhelmed any human efforts. The years of 965, 966, 967, and 971 saw the explosion of serial floods and bank ruptures, with most of these floods extending southward toward the capital Kaifeng. 972 was a particularly devastating year. During the fifth and sixth lunar months, north China suffered extensive rain, which caused water levels to soar and to breach banks at several locations. “There are severe floods in various prefectures on both the southern side and the northern side of the river.”15 Once again, the river's southerly floods threatened the capital. Over all of north China, agriculture was decimated, large numbers of people were forced to flee their land, various supplies ran short, and food supplies in the capital could barely sustain its population for half a year. The demand for food diverted grain through long-distance transportation from the lower Yangzi valley to the north. The flood, the harvest failure, and the difficulty of shipping goods made 972 a year of “great hunger (daji 大饑).”16
Faced with a crisis of such magnitude, Emperor Taizu and his young imperial state discovered that more than a decade of battling the river had yielded only frustration, insecurity, and anxiety. Before long, the state came to grasp its vulnerable position within north China's unstable environmental system. Such awareness urged the ruling members of the state to evaluate the significance of the Yellow River against the state's geopolitical context. River disasters should no longer be treated as individual, random incidents to be contained by individual technical solutions; they should now be seen as grave threats to the stability of the fledgling state. The river must be incorporated into the state's holistic geopolitical structure as something so critical as possibly to jeopardize the fate of the newly established Song state. This new realization forced the state to accept that the power of the physical environment often surpassed that of the state, to ground the state's geopolitical considerations within a broader environmental context and to take serious any environmental challenges imposed by the recalcitrant river upon its political and socio-economic advancement.
To understand how this realization of the entwined relationship of the state and the Yellow River shaped the state's hydraulic policies, let us read the geopolitics in the late tenth century through the state's eyes. In the late tenth century, the Song state, led by Emperor Taizu, saw the world around it as complicated and hostile. Following the precedent set by its short-lived predecessors of the Five Dynasties, the Song settled its political core in the Henan area. There, the plain of Henan provided an economic, political, and military base to the state, with Kaifeng in northern Henan as its capital. The great strength of this imperial city was its proximity to convenient water transportation. The Bian Canal, whose northern end was connected with the Yellow River and southern end with the Huai River and the Yangzi valley, ran directly through the city. As the central artery of an empire-wide transportation system, the canal shipped wealth and goods from the south to supply the heavily populated metropolitan area of Kaifeng and the enormous military forces spread throughout north China. But the tie between Kaifeng and the canal also posed a serious danger.17 Located less than a hundred kilometers south to the Yellow River, the city faced increasing flooding threats from the latter. Any danger the river inflicted toward the south would not only harm the city itself, but also jeopardize the canal. Flooding that damaged the canal and clogged its channel with silt would render the canal unnavigable.
Illustration 10. A Geopolitical Map of the Early Song
A decade into its rule, Henan was still the sole region solidly under Song control. Beyond this political core area, the land all around remained a politically and militarily contested zone. In south China, the Wuyue 吳越 (907–978) and Southern Tang 南唐 (937–975) kingdoms in the Yangzi valley were still independent from the Northern Song. In the southwest, Sichuan had just surrendered to the Song but was not yet under the Song's solid control. In north China, the political situation was even less clear and stable. In the far northwest, Shaanxi had fallen in the hands of the semi-nomadic Tangut, who later grew to be a major enemy of the Song. Shanxi remained under the control of the Northern Han 北漢 (951–979) kingdom, whose ruler relied on the overlordship of the nomadic regime of the Khitan for protection. In 969, Emperor Taizu attempted to conquer the Northern Han, but his military expedition ended in defeat. In the northeast, from the mid-tenth century onward, the central government had developed better control over Hebei, whose warlords gradually submitted to the state. Yet, as I discussed in Chapters 1 and 2, Hebei's autonomous tradition remained strong; regional leaders, if dissatisfied with the Song state, might easily divorce the state and claim independency. Beyond the northern border of Hebei, the Khitan was the Song's fearsome enemy. Nearly every year through the end of the tenth century the Khitan cavalry marched southwards to plunder Hebei, posing a constant threat to the Song's frontier security.18 Besieged on almost all sides, the leaders of the Song must have wondered if the young state could survive the threats posed by the hostile neighbors or if it would follow in the unfortunate footsteps of its five predecessors and become the sixth short-lived regime.
Adding to this worrisome geopolitical situation, the state could not escape the dangers of a destabilized environment. In the 970s, in order to survive any inter-state conflicts, the Song required, at the very least, a strong home base. Yet, its core area lay in a low-lying plain, without natural barriers. To make the situation even worse, the monstrous Yellow River attacked this area with constant floods. The southerly floods and course shifts undermined Henan's – hence the state's – economic growth, labor capacity, and political solidarity. The environmental challenges at the heart of the Song's territory, just like the political and military enemies surrounding its borders, were a major component of the hostile world that the state was born into in the late tenth century. Hence, the Yellow River became a major player in the state politics. Any hydraulic management that the state conducted with regard to the river would invariably incur serious political implications. The river had become a matter of “national security.”19
Given such geopolitical-environmental realities, the state was forced to reflect on its existing flood control practices, which had heretofore dealt with individual floods by means of individual technical solutions, be it constructing dykes to block water or opening a small channel to drain water. These practices needed to be embedded in a systematic politico-hydraulic enterprise, which balanced the state's political demands and its environmental constraints. Hence, in order for a hydraulic plan to work, the state needed not only to understand the pla
n's technological challenges but also to rationalize the plan's potential political benefits and costs. This meant that the state had to make hydraulic choices and decisions in both environmental and political terms.
Emperor Taizu's Politico-Hydraulic Landscape
Against this backdrop, after the outbreak of flood and famine in 972, Emperor Taizu issued an edict to lay out his understanding of the Yellow River issues:
Lately, it has been raining in the prefectures of Chan, Pu, and others. The torrential river has caused disasters. We have suffered so much flooding, which has severely troubled the people. Therefore, whenever reading past literature, I attentively study river issues. As for the records of the Xia period, they mention only diverting the river to the sea and smoothing its stream in accordance with mountainous terrains. It is never heard [in those records] that [people] managed the torrents by force and by constructing extensive high embankments. In the Warring States period, various states pursued their own interests by clogging the old river courses [in Hebei]. Small interests obstructed large affairs, and the private harmed the public. Therefore, the system of the “Nine Rivers” [i.e., various branches of the river's lower reaches in the Hebei Plain that were said to be dredged by the legendary Yu] was eliminated. Since then, river disasters continued without an end throughout later dynasties.20
The emperor expressed his regret over the river's endless damage, and criticized the uselessness of conventional approaches in dealing with the floods. Using the conduct and achievements of the legendary sage king Yu the Great 大禹 as a yardstick for contemporary flood-control activities, Emperor Taizu conveyed five logically interconnected messages.21
