by Ling Zhang
This last number is the size of nearly twenty Olympic-sized swimming pools put together. But it refers to only one embankment. How many embankments of such an average size were built in Hebei between 1048 and 1128? The multiple courses of the Yellow River alone demanded multiple stretches of dykes, each 400–700 kilometers long; moreover, there were other rivers and other waterbodies to deal with. I have found no records of a precise number for the embankments in existence at any given moment during 1048–1128, but there must have been hundreds. The total material costs would have been mind-boggling.
We must remember that the fascine sites were not constructed once to last forever. Even the modern Three Gorges Dam, which was built by most advanced technology and construction materials in the present millennium, cannot be sustained permanently. According to some hydro-engineers, it can only promise a lifespan of stability for about a hundred years. As early as in 2003, before the completion of its construction, the Three Gorges Dam had already seen some minor fractures over its concrete surface. Made of wood and mud, the eleventh-century embankments wore out rapidly and were frequently torn apart by the Yellow River's torrents. Given their very short lifespan, they required constant maintenance and replacement. For instance, a massive flood in the summer of 1056 swept away in a single day the embankments that were constructed in eastern Hebei over the previous years. Beginning in the late tenth century, the maintenance and reconstruction of embankments became a daily task. Look at a rare inscription engraved on a stone incense burner in Huazhou by a lay Buddhist practitioner, some Mr. Zhang, in 1085. Zhang specified that in Huazhou a hydraulic site on the Yellow River's northern bank required constant maintenance. In 1082, eight rounds of renovation were done by using wood and grass; in 1083, three rounds of renovation used wood and grass and another five rounds used grass only; and in 1084, three rounds of renovation used wood and grass and another three rounds used grass only.71
In order to fulfill such constant hydraulic maintenance, the Song government established a set of regulations that commanded its hydraulic officials and regional civil officials to collect vegetative materials every year. As early as the end of the tenth century, over ten million bundles of materials were gathered annually and distributed to various fascine sites for their normal maintenance.72 This amount continued to increase year after year; the government could not suspend the hydraulic works so as to let the laborers go home or pause the extraction of raw materials. It was this routine, endless demand that made the collection of materials so burdensome, making it a compulsory part of the annual taxes on commoners throughout north China.73 Taxpayers were compelled to chop down trees, cut bushes, and uproot grass all over north China. In a sense, without knowing the purpose of their actions, they continued to feed and fuel the hydraulic mode of consumption that was centered in Hebei.
Among the various materials needed, a considerable amount of lumber came from the northwestern part of the empire, in particular from the mountainous terrains of Shanxi and Shaanxi. According to Wen Ji 文洎 (died in 1037), who was charged with the transportation of lumber in the 1030s, trees felled in the northwest were packed up in one to two million bundles every year. They were then shipped via the Yellow River on a journey across several hundreds of kilometers, to arrive at downstream areas where they were loaded and distributed to various hydraulic sites.74 The shipments increased to 3,760,000 bundles in 1029 and then sharply to 7,800,000 bundles in 1030. Compared to this “mountain lumber” that was hard to obtain, materials of smaller sizes such as bamboo, wooden sticks, and grass were easier to gather. These materials came from all over north China, and were collected year-round. All together, the quantity of the smaller vegetative materials might amount to several times that of the lumber.
After 1048, although various north-China regions continued to provide vegetative materials, Hebei overtook those regions as the nearest and most important provider of materials. The river floods, the focus of the state's hydraulic efforts, and the flood-control facilities had largely moved to Hebei. Along with them was the physical movement of vegetative materials; most no longer ended up in Henan, but were shipped to Hebei. Meanwhile, material production sites began to emerge in Hebei. The Taihang Mountains in western Hebei, for instance, began to supply an enormous amount of wood. Of the forty-seven major fascine sites whose names were recorded in Song documents between 1048 and 1128, all but four – that is, 90 percent – were located inside Hebei.75 Assuming they were each of an average size (the volume of twenty Olympic-size swimming pools) according to Shen Li's hydraulic standards, we are envisioning a gigantic mass of wood and grass of the size of 940 swimming pools. This volume of materials spread along the river courses in Hebei. In reality, there were certainly more embankments than forty-seven; some were named and perhaps many were nameless. Over time, large embankments expanded until they were split into several sub-sites.76 As the river developed multiple courses in the following decades, the embankments also multiplied. The local streams became chaotic as well, and necessitated more embankments and the corresponding consumption of materials. For example, in one case of repairs made to the banks of the Zhang River in the early 1070s, more than a hundred thousand pieces and bundles of elm and willow wood were logged seemingly overnight from southwestern Hebei.77 Bringing together the Yellow River's embankments and countless small levees guarding the local streams and ponds, between 1048 and 1128, the entirety of Hebei's hydraulic infrastructure had collected and consumed a massive amount of vegetative materials, so large that it defies calculation.
All of the analysis above concerns the demands of routine hydraulic maintenance. In reality, the rivers and ponds flooded unexpectedly, and the emergencies required additional materials. The emergency consumption was contingent upon how critical the situation was. The annual provision of 10 million bundles that the government ordered to gather at the end of the tenth century was perhaps not enough to handle a single flooding event. After a catastrophic flood in 1019, the government reportedly used 16 million bundles of vegetative materials to fix the river bank's rupture over a time span of nine years.78 The situation was even worse in 1048. The flood and shift of the river course alone prompted the government to acquire 18 million bundles.79 During the 1056 incident, due to the miscalculation and poor management by the hydrocrats, one to two million bundles of materials were destroyed or washed away by the flood within a day.80 Between 1077 and 1078, in order to return the southerly surging river to Hebei, within six months the government assembled nearly 13 million bundles of materials to block up the river's bank rupture.81 In the late 1080s, the state promoted the hydraulic plan of “returning the river” to southeastern Hebei. The project alone anticipated a budget of 20 million bundles of raw materials; in 1089, the number rose to 30 million and even to 50 million.82 The project was not carried out at full scale and was eventually dropped. Yet, even the primitive stage of its preparation in 1087 and 1088 had already necessitated the collection of 24 million bundles of materials.83 This outrageous amount of materials was collected from all over north China and delivered to southern Hebei, where part of it was used in actual hydraulic works, and the rest was abandoned after the abolition of the project. The materials sat at the hydraulic sites and exposed to air and floodwaters; due to bureaucratic negligence, they gradually rotted away.84 Hence, when trying to imagine the total volume of materials consumed in Hebei, we must not neglect the tremendous random costs and waste. A great deal of the trees and grass taken from the increasingly depleted landscape never actually served the hydraulic works or improved Hebei's environmental situation.
The above numbers are so large that it is hard to comprehend the impact of their consumption on the socio-economic life of the people, as well as on the physical environment. Where did these immense quantities of materials come from? From the point of view of ordinary people, the taxation for such materials imposed a tremendous burden. Usually, farmers paid their taxes with straw left from various crops, handing it in to the local govern
ment after harvests. A great portion of straw was then delivered to the hydraulic sites in the winter to prepare for dyke maintenance that usually took place in the early spring. If farmers had a bad harvest and failed to produce enough straw, which happened fairly often in the eleventh century, they would have to resort to other means to fulfill their tax obligations. They had to sacrifice their cash crops by chopping down their fruit and mulberry trees.85 Alternatively, they were forced to buy the materials from markets, which were often dominated by speculative merchants who manipulated prices to maximize their profits.86 Paying steep prices in order to fulfill their tax payments for straw placed many farming households under tremendous hardship.
In Shaanxi in northwestern China in the 1060s, people were forced to go deep into mountains to log. Year after year, the more trees they felled, the farther the forests retreated, and the more difficult the job became. The hard work plus the difficulty in shipping the lumber drove many households into bankruptcy.87 In Hebei, the External Executive of the Water Conservancy set up an annual quota of vegetative materials for every local district; the local government was responsible for producing and submitting enough to fulfill the quota. In Qinghe and Qingyang 清陽 counties in central Hebei, county governments appointed nineteen households as leaders to supervise other local families in collecting reeds and grass. These leading households were exhausted by the job; for six years they were unable to meet the quotas and owed 140,000 bundles of materials to the External Executive of the Water Conservancy.88 They suffered enormous pressure and harassment from the External Executive and the local governments.
The soaring demand for prodigious amounts of raw materials became a tremendous financial burden on the state, too. Taxation did not provide all of the necessary materials; the government often had to use extra funds to purchase a large portion of them. It named this purchase the “harmonious purchase” (hemai 和買), meaning that peasants sold their materials to the government at their own will, rather than by force. As the demand for vegetative materials rose quickly, the government constantly went over budget and experienced financial stress. In 1093, for instance, the Fiscal Commissioner of Jingxi 京西 in western Henan reported that its district suffered from poor harvests and it failed to collect taxes, and at the same time, its military consumption was exceptionally high. It fell far behind in its payment of the “wood-grass money” that the Water Conservancy demanded in order to carry out the hydraulic works inside Hebei.89 As regional governments failed to produce materials, the central government had to use state treasuries.90 In the late eleventh century, expenses related to Yellow River flood control, including the purchase of vegetative materials, equalled the state's military and bureaucratic expenses, and became a major financial burden on the Song state.
The essential question remains: was the land of north China, especially in Hebei, able to produce the materials in such gigantic quantity, and in a sustainable way? In 1088, the project to return the river to southeastern Hebei anticipated the use of 58,848,082 bundles of vegetative materials. Treating this as a critical project, the government spent three months promoting the “harmonious purchase” and levying taxes on ordinary people. As a result, it obtained only 49,000 bundles, merely 0.8 percent of what it needed; this severe lack of construction materials turned out to jeopardize the hydraulic project.91 The failure in acquisition was not because hydrocrats and civil officials did not work hard enough, or did not coerce or terrorize ordinary people into compliance. Rather, it was due to the reality that the land was unable to produce so much wood and grass within such a short time.
While draining human society and the imperial state, the hydraulic mode of consumption rapidly depleted the vegetation across north China. As early as the 1030s, forests in the mountains of Henan, Shanxi, and Shaanxi had been “gradually thinned out.”92 In Hebei, the vast plain saw no hills or forests; human settlements were “far from mountains and valleys” where natural vegetation was relatively thick. Under tremendous pressure to make a living, Hebei people often “cut trees at the wrong time and destroyed what was created by Heaven and Earth.”93 In the worst scenarios, when the imperial state issued urgent orders for wood collection in the 1070s and 1080s, people were driven to sell their property to acquire materials from markets and went bankrupt, or they were forced to fell trees in graveyards that were meant to protect ancestral graves.94 In some harsh winters, many could hardly get firewood to meet their basic needs.
Not surprisingly, natural forests in the Taihang Mountains on Hebei's western border increasingly came under the axe. As early as the late tenth century, the government set up timber workshops on the edge of the mountains to conduct logging. Most of the products went to the state-monopolized iron mining sites nearby as firewood or they were used as construction materials and for tools. After 1048, a considerable portion of the wood went to various hydraulic sites along the Yellow River to build embankments. Due to such large-scale, continuous wood collection, the Taihang Mountains began to lose much of their pine forests. When Shen Gua, Hebei's Fiscal Commissioner, traveled there in the mid-1070s, he lamented that the mountains had become largely bald.95
Equally falling victim to the hydraulic mode of consumption was a peculiar kind of vegetation. Trees, bushes, and grass grew along riverbanks and dykes: some were natural born, but many were deliberately cultivated by human labor. From the late tenth century, the Song government ordered the planting of elms and willows along the banks and dykes of both the Yellow River and local rivers. Around the frontier ponds, it is said that the military and corvée laborers planted three million trees.96 The roots of the plants spread widely and interconnected with each other to fasten the loose, silt-based earth, to hold together the banks, and to prevent erosion and rupture.97 The plants were government property, or more explicitly, the property of the hydraulic infrastructure administered by the External Executive of the Water Conservancy. Any damage to them by private parties (e.g., theft) would incur severe punishment.98 Occasionally Hebei's regional and local governments used those trees to construct city walls. As river floods intensified, those plants assumed important roles in protecting the banks, so the state prohibited local governments from using them for other purposes.99 The hydraulic works controlled by the hydraulic leadership took priority over the civil business conducted by local governments. Despite such regulation, as the environmental problems escalated after 1048, and as the shortage of vegetative materials became more and more evident in north China, the riverside plants could no longer stand alive and remain untouched. One after another, they were chopped down by hydraulic workers and used as construction materials to repair bank ruptures.100 In 1089, Liang Tao, a statesman at Emperor Zhezong's court, lamented: “The trees previously planted on river banks and dykes have been uprooted completely.”101
Hence, a painful irony ruled the environment of not only Hebei but also the whole of north China. Inside Hebei, the plants cultivated expressly for the purpose of strengthening the dykes were felled and used for embankment construction. The newly built embankments certainly helped to ease floods, but the old embankments, once deprived of their protective plants, were substantially weakened and became more vulnerable to any future flooding. As a result, the flooding problem was merely shifted from upstream or downstream, or from one river to another. The new floods, once again, triggered another round of logging of the riverside plants and thereby caused more floods elsewhere. For the eighty years between 1048 and 1128, the hydraulic works in Hebei were caught in a downward spiral: although their practices often defeated their own purpose, the hydraulic works could not stop; they had to continue felling the riverside plants.
This irony also dominated the pattern of resource production, distribution, and consumption across all of north China. As I explained in Chapter 1, the key reason behind the Yellow River's floods lay in deforestation and soil erosion in the river's middle reaches. There, in Shaanxi and Shanxi, the untimely, inappropriate depletion of vegetation, either through log
ging or dry-land farming, uprooted a tremendous amount of loess and sand. The latter contents became river silt and traveled downstream to deposit on the plain of Hebei and Henan, causing the riverbed to rise, blocking the river channel, and forcing the river to flood and shift its course. Song contemporaries had some sense of the causal relationship between the vegetation cover, land erosion, and tranquility of the river. Many were aware that silt coming from the far west had become the ultimate enemy of the river's tranquility. Yet, they were unable to connect the two phenomena that north China was experiencing at the time: the unceasing deforestation and the generation of massive silt in the empire's northwestern frontier, and the flooding disasters and hydraulic works in the empire's northeastern frontier, Hebei. The two geographic units were so far apart that people in the eleventh century, even at the state level, failed to conceptualize their environmental connection.102 The logging industry in the northwest that the imperial state relied on intensified soil erosion and the river's siltation, and thus contributed to the river's turbulence in Hebei. Yet, in order to tame the river's violent torrents and build extensive dykes in Hebei, the state's increasing demand for construction materials had to turn to the northwest again, causing more trees to fall, more forests to disappear, and more silt to travel downstream to choke the river course. So, the vicious cycle was perpetuated. Once again, we witness the interplay among various environmental spaces – local,regional, and transregional – that defied the stable structure of the autonomous, self-sufficient macroregions G. William Skinner conceptualized for China. The interplay among multiple environmental spaces swirled north China into the unstoppable hydraulic mode of consumption of resources.
