In 2015, archaeologists opened the dam at Laohuling, northwest of Hangzhou.

Oval outlines appeared within its section. They lay above and beside one another, with their joints staggered. The earth had not simply been tipped into a heap. Each bundle had once been enclosed in plant material. More than five thousand years later, this technique—grass-wrapped clay—remained legible inside the dam.

Grass enclosed the clay; the clay held water. Ordinary plants and earth entered an undertaking far larger than a single bundle.

Laohuling belongs to the high-dam area at the mouths of valleys northwest of Liangzhu's ancient city. It was no solitary barrier. Other dams, levees and lower stretches of water extended through the valleys and foothills. People had carried bundles of earth into these places. What they ultimately changed was how water from the hills entered the plain.

I

Water here sustained life and complicated it.

Liangzhu lies in what is now Yuhang, Hangzhou, in Zhejiang, where the eastern foothills of the Tianmu Mountains meet the Hangzhou–Jiaxing–Huzhou Plain. Hills rise to the west; low ground opens eastward. Channels cross the land, with isolated hills scattered among them. Liangzhu culture lasted approximately from 3300 to 2300 BCE. Rice fields, settlements and eventually the city developed in this wetland setting.

Water descending from the hills did not wait for human permission. Greater rainfall sent it through the valleys toward the lower land, where fields and dwellings stood. Yet expelling all the water was no answer for a rice-growing society. Cultivation and transport still needed it.

The dam sites were therefore not arbitrary points on a flat map. High dams occupied gaps between hills, holding water in the spaces behind them. Natural slopes supplied part of a reservoir's enclosure; a built barrier joined the gap. People did not have to construct an entire basin from nothing. They did have to find places where hills and dams could contain water together.

Farther down, low dams on the plain and levees along the foothills carried the arrangement into another level. Retention above, storage below and regulation of flow were not merely repetitions of the same wall. Structures at different elevations had to work together. Water continued to move, but some was detained and some redirected.

Before one dam, only a stretch of earth is visible. Following the terrain reveals its relationship with other hills, barriers and channels.

II

To close a gap, the builders first had to turn earth into a dam.

Grass-wrapped clay was not a fired brick. Plant material, including reeds and grasses, enclosed the earth; plant strips bound the bundle. Loose material acquired a form that could be carried and placed. Surviving wrapping and construction traces preserve more than the origin of the soil. They preserve ways of handling it.

A dam needed foundations and successive layers. Grayish clay, yellower earth and wrapped bundles occupied different positions in its section. The vegetation did more than contain the soil. It helped hold the material together and increased its resistance to tension. Staggering the bundles kept successive layers from sharing one continuous joint. Later builders using grass sacks filled with earth would recognize this way of bringing plants into reinforcement.

Each bundle was modest; the dam was much larger. One had to be wrapped, then another. Excavated earth had to be moved, and moved earth had to be placed correctly. Wrapping, carrying and laying had to continue together before materials became a barrier across a valley. The repeated outlines preserve something more particular than a statement that ancient people were clever: people really performed these tasks, again and again.

Annual plants in the bundles also preserved clues to the work's date. They grew during short lives, were cut and bound, and entered the dam. Radiocarbon measurements of such material can help approach the age of construction.

The dates and stratigraphic relationships show that the whole network did not appear at one instant. The high dams were generally earlier; low dams and levees followed as the system expanded. Some construction probably preceded much of the city's principal building work. Certain dams also show signs of later rebuilding. The connected system visible on a modern plan emerged through stages of construction and adjustment.

A Late Liangzhu ditch lay above the Laohuling dam. An earlier structure had been overlaid by another activity. Life in the city changed, and the works entered new relationships of use. Grass-wrapped clay did not fix the entire landscape at the moment its first dam was built.

III

Beyond the dams, a city grew on earth too.

The principal city lay south and southeast of the hydraulic complex. Earthen walls enclosed an inner city of approximately three square kilometres. At its centre, Mojiaoshan occupied a great artificial platform. Earth raised the low, wet ground and provided space for buildings above. Walls, platforms and reinforced banks employed related earthworking techniques, including grass-wrapped clay.

The same material moved from valley gaps into the city. A dam had to restrain water, a platform support buildings, and a bank stabilize ground beside a channel. Different tasks confronted the same meeting of soil and water.

The city walls did not exclude water. Natural channels connected with artificial ditches, while water gates allowed waterways through the boundary. Liangzhu was not a city accessible only overland, with rivers kept outside. Waterways entered its layout and linked it to production and transport around it.

Wooden structures beside the channels take the story from earth to another material. At Meirendi, timbers laid in different directions supported upright planks forming a bank. Bianjiashan preserved a pier formed by rows of wooden stakes. The wood had positions and connections; it was not merely debris that had drifted together. The waterside became a place that could be used and had to be maintained.

Stone for the city's wall foundations came from the northern hills; waterways helped carry it into the city. Places that restrained water and places that admitted it both participated in construction. Dams retained water in the uplands. Gates and piers enabled people to use it for moving materials.

Those who needed to eat remained part of these relationships.

A storage pit east of Mojiaoshan contained abundant rice remains. These were not a handful of grains dropped separately on the ground, but evidence of concentrated storage. Grain had left the fields and entered a place where it could be kept together. Supplying a city required a harvest, followed by arrangements for storage and movement.

Dams, channels, piers and food thus belong to connected stories. Fields needed water; transport needed water; dwellings and stored grain could not simply be surrendered to flooding. Some people cultivated, some carried, some made objects. A growing settlement had to allow these activities to continue.

IV

Who brought the work together?

The works left no roster from which the laborers' names can be called. Yet earth no longer lay where it had begun. Dams occupied valley gaps, platforms rose in wetlands, and strengthened banks bordered channels. Digging, carrying and piling had accumulated into visible forms.

Nor could people be involved only at the outset. Dams and waterways required use and maintenance, while farming followed its own seasons. Gathering labor and connecting work in different places were unavoidable parts of the undertaking. Its precise commands, payments and divisions of work cannot yet be reconstructed. Its scale already exceeded the capacities of a single household.

Status within the city differed too. Raised buildings and burial goods reveal more prominent positions and access to precious jades, while other people left much less wealth. Who directed the construction, and how they obtained obedience or cooperation, cannot be read directly from a dam. Water management and power met here, but there is no known ruler who can answer for all the remains.

The complete system did not operate forever. The city's flourishing ended, and many buildings became archaeological remains. Some barriers in the hills entered later lives. Qiuwu still retains water, with an ancient dam joined to modern concrete construction. Continued storage at one location does not mean that the entire five-thousand-year-old system operated without interruption. A place could be used again and modified by different people.

The archaeologists did not recognize everything at once either. Tangshan's unusual terrain attracted attention in 1987. Surveys in 2009–2010 confirmed high dams at valley mouths. Satellite imagery subsequently helped locate low dams on the plain, followed by examination on the ground. Excavation at Laohuling and other dams in 2015 opened their internal structures. Land that could appear naturally uneven gradually disclosed human arrangements.

Early reports discussed a combination of eleven dams and levees. Later surveys found more. The extent of the engineering landscape is still being understood, and its history continues to unfold.

Today, the Laohuling section brings the bundles of clay before people again. The hands that wrapped them are gone; the bindings and layers remain. The builders did not inscribe their names on this dam. They placed grass and clay here, holding water for a time among the hills and allowing life elsewhere to continue.

Sources and editorial boundaries

An original AI-assisted bilingual historical narrative by Shi Tu / Clio Walker, checked against the archaeological research and institutional materials actually obtained. Permanent research number 004 is retained. The research question about who organized Liangzhu's dams becomes a story following wrapped clay, the dam network and the city. No ruler, worker, order, construction day or fixed work schedule is invented. The following materials were actually read on 14 October 2026.

1. Bin Liu, Ningyuan Wang, Minghui Chen, Xiaohong Wu, Duowen Mo, Jianguo Liu, Shijin Xu and Yijie Zhuang, Earliest hydraulic enterprise in China, 5,100 years ago, PNAS 114(52),2017,13637–13642, DOI 10.1073/pnas.1710516114. The version read was the 17-page author accepted manuscript deposited in UCL Discovery, not an independently acquired final journal facsimile or complete excavation report. Sections used: Defining the Liangzhu Hydraulic Landscape; Logistic Planning and Operation; Conclusion; Methods and materials; and relevant figure captions. These support construction layers, wrapping, dating and phases, the later ditch above Laohuling, Mojiaoshan's platform and rice storage, the Meirendi bank, Bianjiashan pier, stone transport by water, abandonment and partial later use. Technical sequences summarize material evidence and engineering relationships, not an eyewitness construction day. Numerical estimates for population, earthwork, workforce, work duration and reservoir capacity are not repeated as established counts. Central planning and political organization remain archaeological interpretations rather than direct readings of a named regime or proof of despotism. Version record: https://discovery.ucl.ac.uk/id/eprint/10040887/ Accepted manuscript: https://discovery.ucl.ac.uk/10040887/1/Revised%20manuscript.pdf

2. Liangzhu Museum, “流淌的古文明之歌①千秋工程——良渚古城外围水利系统,”18 April 2025. Used for the 1987 Tangshan clue, high-dam surveys in 2009–2010, recognition of low dams in 2011–2012, excavations in 2015, construction material and form, functions, and Qiuwu's joining of ancient and modern structures. The page distinguishes eleven features discovered in 1996–2014 from more than twenty subsequently recognized, giving over thirty dams. The older eleven-feature account is not today's exhaustive inventory. Continued water retention at particular sites does not establish uninterrupted operation of the whole system. https://www.lzmuseum.cn/QiTa1/2025344760349.html

3. Liangzhu Museum, “良渚先民的‘黑科技’，五千年前的‘超级建材’！” Used for wrapping, staggered laying, reinforcement and applications in platforms and walls. Modern experimental photographs and reconstructions are not ancient eyewitness images. Page and image-file dates do not date prehistoric events. Claims of experimentally optimal bundle dimensions are not repeated as universal conclusions. https://www.lzmuseum.cn/LiangBoXinWen/2026854964858.html

4. Liangzhu Museum, “流淌的古文明之歌②｜水乡泽国的都城营建,”16 June 2025. Used for the relations among channels, water gates, bank protection and wooden remains. Modern replicas, metal figures and tourist reconstructions are not treated as prehistoric observations. No cargo manifest, voyage or particular episode of well use is invented. https://www.lzmuseum.cn/QiTa1/2025554805551.html

5. UNESCO World Heritage Centre, Archaeological Ruins of Liangzhu City, property 1592 description and geographical table. The approximate 3300–2300 BCE range describes the archaeological cultural setting, not simultaneous construction of every dam throughout it. The 2017 accepted manuscript gives most high-dam samples at 5200–4800 cal BP and most low-dam/levee samples at 5000–4800 cal BP, retaining distinctions among sample groups. BP is referenced to 1950, not a subtraction of 5100 from 2026. City dates are not forced onto every construction phase. No surviving construction text supplies exact days or months. https://whc.unesco.org/en/list/1592/ ; https://whc.unesco.org/en/list/1592/maps/

Geography and map: the city and dams occupy the area now in Yuhang, Hangzhou, Zhejiang, at the eastern Tianmu foothills and western Hangzhou–Jiaxing–Huzhou Plain. Modern administrative and archaeological names are not asserted as known prehistoric place names. UNESCO component 1592-002 provides a high-dam-area reference at 30°25′13″N,119°54′12.98″E(30.4202778,119.9036056); component 1592-003 provides a low-dam/causeway-area reference at 30°24′12.89″N,119°56′58.38″E(30.4035806,119.9495500). Both identify broad heritage areas, not the exact Laohuling section, an individual dam or an ancient channel. The 3300–2300 BCE associations describe Liangzhu cultural engineering contexts. A separate 2015 association marks modern excavation within the high-dam area; discovery dates do not date ancient construction. Exact dates and locations for the Mojiaoshan storage pit, quarry, piers, banks, changing channels and individual building or repair phases remain unresolved. No guessed transport route is drawn.

Rights: the modern paper, institutional pages and images retain their respective protections; public access is not public-domain status. Facts and limitations are organized in original prose, without reproducing the paper, plates, photographs, extended modern writing or protected translations. The UCL manuscript is not republished as a site attachment. Both editions retain complete sources and matching uncertainties.