Prometheus Brings Water: Development and Fix-Fixing in China
Nowhere is the Chinese Party-State’s Promethean thinking more vividly apparent than in its continuous proclivity to build more and bigger water projects. And where these projects create problems, the solution is not to remove the projects but to build further projects, to construct ‘fixes to fix the fixes’. From Yellow River conservancy in the first years of the People’s Republic of China to ever-expanding water projects in the early twenty-first century, storing, releasing, and moving water have brought Chinese social-ecological systems to a condition of severe technological lock-in, decreasing the resilience of these systems to extreme weather, extreme policy, and other ‘natural’ disasters. When Deng Xiaoping proclaimed in 1992 that ‘only development is a solid truth’ (发展才是硬道理) (Guang’an Ribao 2016), he was not saying anything new. Not only did his remarks accord with those of capitalist development theorists like W.W. Rostow (1960) and with the ambitious goals of both late-Qing self-strengtheners and the Republican modernists who followed them; they were also in complete agreement with the ambitions of his own party in its earlier, more explicitly socialist age. The 1949 Common Programme of the Chinese People’s Political Consultative Conference (全国政协基础纲要, CPPCC) proclaimed:
The basic principle for the economic construction of the People’s Republic of China is … to achieve the goal of developing production and a flourishing economy … to promote the development of the entire society and economy. (CPPCC 1949: Article 26)
Even in 1973, at the height of the Cultural Revolution—a period when, later narratives argue ‘revolution’ precluded ‘development’ (Zweig 1989: 192)—Michel Oksenberg could write: ‘One major purpose of the Chinese Revolution is to build a prosperous and strong country. Economic development is a major aim of any program undertaken in China’ (1973: 13–14).
Chinese Communist leaders have thus always shared Promethean ideals with postwar capitalism. Assuming human ingenuity can solve all problems, including those of extracting more value—measured in calories, water, or money—from natural ecosystems, both capitalists and communists have often failed to recognise that those systems themselves have limits. What differed between the likes of Rostow and the likes of Mao was not the desirability of development but the means deployed to accomplish it (Schmalzer 2014: 78–79). If, for Rostow and similar thinkers, the primary obstacle to development was technological, for Mao, it was political: if the relations of production could be rearranged to ‘liberate the productive forces’, those forces could be channelled into growth, rather than into parasitic exploitation by national elites and Western imperialists. The goals were the same.
Two other inheritances have contributed to Prometheanism in Chinese Communist thought. One of these is very old: Confucian humanism. When the Chinese Communist Party (CCP) adopted the slogan ‘Human determination will triumph over nature’ (人定胜天), it was not, as many have assumed, to say that human destiny is to conquer nature, but rather that human plans (人定) can overcome natural obstacles (Wu 2014). Communists used this age-old saying to implicitly deny that nature (天) could ever place limits on human activity. During the Great Leap Forward (1958–60), the failure to recognise these natural limits was encapsulated in the slogan ‘The product of the earth will be as great as the courage of the people’ (人有多大胆, 地有多大产) (Liu 1958), prompting all sorts of well-known nutty schemes, from ploughing up the earth to a depth of more than 1 metre (Zhong 2011) to melting down cooking utensils to ‘make’ steel (Wagner 2011).
Marxism, through its critique of capitalism, reinforced and focused Chinese Communist Promethean thinking. For example, commentators demonstrated through impeccable logic that environmental pollution was a capitalist problem. Capitalists, in pursuit of profit, had to cut costs and thus neglect the environmental harm of their activities (which was true enough in many, many cases, and something that Marx himself pointed out; see Schneider 2015), while communist economic planners, in pursuit of the public interest, would not degrade the environment (Whitney 1973: 107). The only problem with this argument was that the logicians neglected to look out the window. The first 20 years of development in the People’s Republic of China (PRC) not only brought forth pollution, but also, more importantly, failed to align development plans to maintain the resilience of China’s ecological and social systems, leaving them continuously vulnerable to natural and political disturbances.
The Trade-Off between Development and System Resilience
Systems ecologists define resilience as the ability of a system to withstand a disturbance or perturbation and continue functioning (Holling and Gunderson 2002: 28). Resilience is maintained by buffers, which can be ecological, infrastructural, institutional, or cultural. For example, an agro-ecosystem can be buffered against extreme weather events by wetlands, fallowing, crop diversity, dikes, regulations, access to outside products, or ethical values. Typically, development removes ecological buffers by reclaiming unused lands, planting monocrops for their higher caloric or market values, discharging more pollutants than air or water can dilute, and so on. This increases stress on institutional and infrastructural buffers. If previously, for example, overflow basins contained Yellow River floods (Pietz 2015), reclaiming those basins for agriculture or industrial development necessitates dikes, drainage canals, dredging, and institutions to regulate these structures and processes.
Development does not diminish resilience in a linear way, however. Low levels of development can sometimes increase resilience, as when formerly farmed steep slopes are converted to terraces, thereby both increasing agricultural yields and reducing erosion. Or a reservoir can conserve water for irrigation use in times of drought while holding back potential floodwaters in times of unusual rainfall. But past a certain level of growth, development almost always decreases resilience. Terracing will not work on the steepest slopes, and encroaching on the edges of a reservoir by cultivating new fields will diminish the amount of rainwater the reservoir can hold, meaning that a smaller storm could bring just as big a flood. Figure 1 shows the curvilinear relationship between productivity and resilience in agricultural development.
CCP Promethean thinking is thus not just about failure to recognise natural limits on productivity. It is also about failure to recognise that as development alters ecosystems past a certain level, it also diminishes their resilience. Irrigation canals built in the 1950s to increase water supply to fields on the North China Plain raised the water table and made agricultural areas more vulnerable to waterlogging after excess rainfall. Later, having basically exhausted surface water supplies in the same area, the turn in the mid 1960s to groundwater irrigation using tube-wells drew down the water table and created cones of subsidence, making urban buildings more vulnerable to cracked foundations and causing sinkholes to open in urban streets decades later. Monocrop agriculture made rural communities more vulnerable to crop diseases, as when Panama disease (or Fusarium wilt) wiped out large areas of banana plantations in Guangxi in the 2010s (Liu 2018).
If development, however, is ‘the only solid truth’—if increasing productivity is the sole or primary goal of agricultural, industrial, and land-use policy—then it is easy to ignore limits on productivity and diminishing resilience and apply the Promethean principle that ‘humans left to their own devices will automatically generate solutions to problems’ (Dryzek 2005: 67), even to those problems that humans have created by their own dismissal of natural limits and ecosystem feedbacks. Thus, the strategy of the fix to fix the fix.
Fixing the Fix
Recognising that development can lead to decreased resilience and vulnerability to disasters, one solution might be to retreat from the development that brought about the diminished resilience. But this contradicts both the slogan about solid truth and, more substantively, the legitimacy of the regime, which since 1980 has rested ever more on its ability to bring material prosperity. Hence, nothing can be de-developed, no matter how much vulnerability it has occasioned. So, the solution to the problems caused by a technological fix is another fix, and fixing the fix has become the primary strategy of environmental planning in the PRC. This strategy is evident across the board, but particularly in waterworks and water conservancy. Three examples are emblematic of fix-fixing and even fix-fixing-fixing.
In 1953–54, Chinese planners formulated the General Plan for the Yellow River to prevent flooding and convert many of the rainfed fields of the North China Plain to irrigated agriculture (Pietz 2015: 158–74). The plan included several dams, including two on the Yellow River, at Liujiaxia in Gansu and at Sanmenxia (三门峡, or Three Gate Gorge), a narrow spot on the Yellow River in Henan, about 100 kilometres downstream of where the river turns eastward and begins its flow across the North China Plain. Soviet engineers helped plan the dam, which was begun in 1957 and completed in early 1961.
By late that year, it had become evident that planners had seriously underestimated the amount of silt that would be trapped behind the dam. Less than a year after the dam was completed, sediment deposits had already reduced the capacity of the Sanmenxia reservoir by half, causing the area of the reservoir to expand far beyond its planned limits and the channel of the river to widen upstream near the confluence of the Wei River, which flows out of Shaanxi and into the Yellow River right at the bend. Water backed up into the Wei River Valley around Xi’an, flooding farmland and causing widespread salinisation (Pietz 2015: 225; Shapiro 2001: 62–63). Siltation raised the level of the Wei River so much that it buried an important bridge, and a new one had to be built. By 2003, some stretches of the Wei River had become—like much of the Yellow River further downstream—a ‘hanging river’ (悬河), with its bed higher than the surrounding plain (Tan 2003), increasing vulnerability to floods; and, indeed, floods displaced Wei Valley farmers for several months in the autumn of 2005 (Wang and Wang 2011). Reduced flow, along with sediment that clogged the turbines, rendered the dam almost useless for power generation.
The adverse effects of Sanmenxia were not limited to upstream areas and tributaries. Sediment trapped behind the dam slowed the flow of the river, so that in addition to reducing power generation, the Yellow River downstream slowed so much that more sediment was deposited in its bed rather than being carried to the sea (Bi et al. 2019: 973).
Even advocates of big water projects admitted that building Sanmenxia was a mistake. Zhang Guangdou, one of the engineers who worked on the original project (and also on the giant Sanxia 三峡, or Three Gorges Dam project on the Yangtze), admitted in a 2004 television interview that ‘Sanmenxia was a mistake, and I am not the only one to say so’ (Pietz 2015: 302). Qi Pu, a senior engineer at the Yellow River Resources Institute and usually a big advocate of water projects, told a group of touring ecology students that ‘the idea to construct Sanmenxia was fundamentally flawed’ (Wang and Wang 2011).
However, the presence of the dam itself had created a technological lock-in (Perkins 2003) or path dependence (Arthur 1994)—the short-term costs of removal would far exceed the costs of a fix. So, the solution was not to undo the mistake and remove the dam, but rather to build another dam to fix the fix. The Xiaolangdi (小浪底) dam, 130 kilometres downstream, was completed in 2000, and a process called water–sediment regulation opened sluice gates annually there and at Sanmenxia, creating a series of artificial floods to flush out concentrated sediments (Li et al. 2017: 162). This process has resulted in a buildup of sediment at Xiaolangdi and reduced sediment further downstream, changing the sediment regime there once again, from deposition to erosion. As dams and devices were built upstream, even the size and shape of the river’s delta changed, first growing, then shrinking, and after water–sediment regulation began, growing again (Wang et al. 2010; Wu et al. 2017). Whether there will be a further fix to fix the fix to fix the fix remains unclear at the time of writing.
2. Poyang Lake
Lying just south of the Yangtze River in north Jiangxi, Poyang (鄱阳湖) is invariably described as ‘China’s largest freshwater lake’. But that depends on the season. During the dry seasons of autumn and winter, Poyang is primarily fed by five rivers draining into it from the south, while its waters flow out into the Yangtze. The lake’s autumn surface area in recent years has been about 900 square kilometres. When the Yangtze runs high, however, the flow is reversed and the river backfills the lake, quadrupling its size to an average annual maximum of about 3,700 square kilometres (Wan et al. 2018: 99). Even this yearly maximum is smaller than the lake’s historical size, which was reported to be about 5,100 square kilometres in 1949.
Flooding in the Poyang region increased after the founding of the PRC in 1949, as deforestation along the lake’s tributaries increased sediment flow into the lake and seasonal high flows in the Yangtze also increased with upstream deforestation, adding to the summer backflow of the Yangtze into Poyang (Shankman and Liang 2003).
Enter the great Sanxia (三峡) or Three Gorges Dam. Before Sanxia was completed, hydrological forecasters predicted that building the dam would increase flows in the Yangtze in the early summer, when the Sanxia reservoir level was being lowered to make room for possible upstream floodwaters, at the same time as heavy rainfall in the Poyang watershed would raise the lake level. In combination, these would increase the danger of summer lake floods, even as disastrous river floods like those of 1998 would become less likely. But the forecasts were inaccurate. After Sanxia was built, the average annual summer and autumn lake surface dropped about 2 metres (Mei et al. 2015: 2), and the dry season began to start earlier in the autumn and extend later into the winter than previously (David Shankman, personal communication). As a result, wetlands shrank and grasslands expanded, potentially affecting habitat for large numbers of migratory birds that overwintered in the area (Burnham et al. 2017). A long-submerged Ming-period causeway emerged and was touted as a tourist attraction, dubbed the Thousand Eye Bridge (千眼桥) (Mei et al. 2015: 1).
Some of the technological fixes that brought about this situation could be fixed by removing the cause. In particular, sand mining around the lake mouth, which increased the wintertime gradient between the lake and the river and thus increased the outflow and lowered the lake level, could be halted (Lai et al. 2014). But sand, as a component of concrete, was a key raw material for building China’s cities, as the urban population increased from about 20 per cent in 1980 to 65 per cent today. And, anyway, Sanxia was there to stay, meaning there would need to be a fix to fix that gigantic fix. Thus, authorities proposed building a 3-kilometre-long sluice wall across the lake outlet, retaining water in the lake during the winter and purportedly protecting fishing, shipping, and wetlands (Ives 2016).
Controversy over the sluice wall was still raging when the middle-Yangtze region experienced its highest-ever rainfall in early June 2020, bringing about major floods and raising the heretofore declining lake level of Poyang to a record high (BBC Chinese 2020). Contrary to hydrological experts’ previous predictions, lake floods and river floods could happen at the same time after development had removed ecological buffers. As a result, in early 2021, the Jiangxi Government’s detailed plans for the sluice wall were included on the list of national large-scale water projects (Diao 2021). Whether this latest fix will fix the multifarious harms brought about by the former fix remains to be seen.
3. The National Water Network
As early as 1953, Mao Zedong pontificated: ‘There is a lot of water in the South and not much water in the North; if it’s possible, it would be OK to borrow a bit of water [南方水多, 北方水少, 如果可能, 借点水来也是可以的]’ (Crow-Miller 2015: 180). By the 1990s, previous fixes—including increased irrigation, growing municipal water use, and groundwater drawdown—had exacerbated the disparity, and clearly a further fix was needed. It came in the form of the South–North Water Transfer (南水北调) Project, which would transfer water from the Yangtze watershed to that of the Yellow River by two routes: one from the Danjiangkou Reservoir, where the Dan and Han rivers meet in northern Hubei, and another along the Grand Canal in the east (a third route through the mountains in the far west was postponed for the time being). Critics expressed considerable doubt that the project was necessary or desirable, pointing out that conservation and more efficient use could save a lot of water, and that delivering Yangtze water to the north at low cost would create a moral hazard and disincentive to conserve or economise (Berkoff 2003). Besides, there would be problems with pollution along the eastern route especially (Tang et al. 2014), and the central route could deprive farmers in the sending area of some of the water they needed for their crops (Webber et al. 2017).
Nevertheless, the fix-fixing mentality prevailed and, by 2014, the north was ‘borrowing’ an annual total of about 27 cubic kilometres of southern water. The lowered flow of the Han River was a particular object of concern, so between 2010 and 2014, Hubei authorities built one of the first fix-fixing fixes: the ‘Taking the Yangtze to fill the Han’ (引江济汉) project—a 67-kilometre open tunnel from Jingzhou between the Sanxia Dam and Wuhan to Gaobeishi Township, about 100 kilometres upstream.
This fix fixed only the final stretch of the Han between Gaobeishi and Wuhan, doing nothing for the middle reach between Danjiangkou and Gaobeishi. In addition, the Wei River Valley was becoming increasingly water-deficient, and the best way to get water to that region would be the ‘Taking the Han to fill the Wei’ (引汉济渭) project, which will take about 1.8 cubic kilometres per year from two tributaries of the upper Han in the Hanzhong region of Shaanxi, well above Danjiangkou, and divert it through a 75-kilometre-long tunnel as much as 2,100 metres deep under the Qinling Mountains to serve the Wei Valley cities of Baoji, Xianyang, Xi’an, and Weinan, as well as surrounding farms. So that fix will soon exacerbate the shortages in the middle reaches of the Han.
Never fear: a fix to fix the fix to fix the fix began construction in July 2022 (Hubei Daily 2021). The ‘Taking the Yangtze to replenish the Han’ (引江补汉) project will divert about 3.8 cubic kilometres annually from the Longtan River, a left-bank tributary of the Sanxia reservoir about 7 kilometres above the dam, to the Han River immediately below the Danjiangkou reservoir—a site chosen to avoid lowering the quality of water carried to northern China along the Middle Route (Danjiangkou Online 2021). Figure 2 illustrates the relationships among all these water fixes.
These water fixes are China’s most spectacular, but they are far from unique; the national agenda contained 172 large-scale water projects in 2014 and the State Council announced another 150 in 2020 (Zhang et al. 2022). These included big transfer projects in Zhejiang, Yunnan, and even Guangdong, which receives more than 1,500 millimetres of rainfall every year (People’s Government 2019; Guangdong 2019; Central Yunnan Water Delivery 2018). Although it is not yet approved, there is even a serious proposal to bring desalinised seawater 2,000 kilometres to solve ‘water shortage problems’ in Xinjiang (CCTV 2019). The total preliminary budget for what is now officially dubbed the ‘national water network’ (国家水网) is about RMB3 trillion (Zhang et al. 2022).
Promethean solutions to water allocation problems were not inevitable, and whether completing the national water network will ultimately increase or decrease resilience against extreme weather events is not clear. After all, we can only evaluate the resilience of a social-ecological system in retrospect, after a disturbance happens. What we do know is that the livelihoods of citizens in Hubei and Shaanxi are ever more dependent on infrastructural buffers—dams, canals, and tunnels—and on the institutions that regulate and maintain that infrastructure. Dependence on the path of an ever-expanding water transfer network has brought the water system to what resilience ecologists call a ‘rigidity trap’ (Holling and Gunderson 2002), where continuation of the system is dependent on infrastructure and institutions. If the institutions break down, the infrastructure will follow, and even though the costs of maintaining the infrastructure continue to rise, there is no choice but to maintain it; and, given that it has brought about further problems, there is no choice but to build further fixes.
It would be remiss and simplistic to assert that further fixes have been the only approach to the problems brought about by rapid development in the PRC. Authorities have addressed certain serious environmental problems effectively by removing their causes; air pollution has decreased dramatically since 2015 through reducing coal-burning, regulating the sulphur content of gasoline, and shutting down many of the most polluting sources. After deforestation reached dangerous levels from the 1950s to the 1970s, forest cover has more than doubled in the past few decades. The state has recognised that technological fixes will not fix everything and embraced a general goal of building an ‘ecological civilisation’ (生态文明). But the Promethean mentality remains, especially when water is involved.
It would also be simplistic to condemn the Promethean mentality generally. Chinese citizens now enjoy more comfortable, secure lives than ever before, primarily due to development. But overemphasis on material growth through technological fixes can lead to the pursuit of short-term goals and ignoring long-term consequences. The Yuan Dynasty scholar Lin Yuan (林远) was clearly worried about Prometheanism and its effect on resilience when he critiqued a design for—yes—a water project: ‘Aim for small advantage and wreck a great plan, fret over short-term success and leave behind long-term suffering’ (射小利, 害大謀, 急近功, 遺袁患) (Lin n.d.).