Most historians believe that the erection of expansive walls became possible only with material prosperity, the emergence of class differentiation, and the rise of political power. Without doubt, raiding and pillaging would not have been worthwhile until people and places began accumulating goods and wealth or, as the Tao Te Ching notes, “When gold and silver fill the hall, robbers and thieves will come.” While counterarguments need not be raised, tribal societies apparently clashed for reasons other than plunder, including hatred, revenge (the most often mentioned), threat reduction, domain expansion and resource control, and seizing prisoners for enslavement. Nevertheless, even in early agricultural settlements or villages where hunting remained primary, theft and forceful seizure may still have represented a comparatively efficient expenditure of time, threatening the undefended.27

Whenever troops were inadequate or slaves unavailable, the requisite labor had to be co-opted from the local settlement or within the greater sphere of political dominance. Diverting numerous workers from the crucial tasks of farming, hunting, and fishing implies that a minimal surplus in foodstuffs such as millet had been achieved, some sort of need perceived, and a basic plan formulated. A chieftain or village council with sufficient power to coerce the inhabitants into undertaking the project also had to exist. If successful, the effort would certainly enhance the chieftain’s authority and contribute to consolidating the ruling clan’s power.

Because great strength was not necessary, merely determined effort on a vast scale, nearly anyone could perform at least some of the tedious labor required for excavating ditches and constructing walls. Warring States writings indicate that people were expected to participate in accord with their physical abilities, women as well as men, a practice that probably dated back to antiquity. For example, the Mo-tzu states that those who can build should do so and that those who can carry earth or measure should also be appropriately employed.28 The Huai-nan-tzu, although composed fifteen hundred years after the event, notes that, “in employing the people to construct earthworks, Yi Yin had the lanky dig, those with strong backs carry earth, those with discriminating eyes determine level, and the bent-over work on plastering.”29 Brief statements in theoretical military writings such as the Wei Liao-tzu indicate that women participated in Warring States siege defense by constructing and repairing the walls.

Although constructing earthen walls might seem to be simplicity itself—merely pile up immense amounts of soil and rely on gravity to keep it in place—reality proves otherwise. Massive mounds constitute formidable obstacles, but if basic engineering and construction principles are not observed, the work will lack integrity and may collapse under its own weight. Moreover, to function effectively a wall must have sufficient height to preclude jumping over it, be nearly vertical to make climbing it difficult, and have well-compacted soil so as to provide a solid footing for any soldiers engaged in mounting an active defense atop the ramparts.

Comprehensive planning and thoroughgoing organization are therefore required to construct conjoined walls and moats. General dimensions have to be determined, topographical features analyzed, and the terrain’s load-bearing capability estimated. The construction has to proceed in sequential steps; the laborers must be organized; and the diverse tasks of excavating the soil at designated sites, transporting it in reed baskets, depositing it in appropriate locations, watering as necessary, and then pounding for the lengthy period required to attain concretelike hardness within prepositioned wooden forms must be systematically assigned.

An entry in the Tso Chuan for 510 BCE, five centuries after the demise of the Shang but timeless in its essentials, describes the process of wall building:30

First they assessed the length as the number of chang,31 estimated the height [of the ground] as high or low, measured the thickness and thinness, plumbed the depths of the moat and ditch, graded the soil, determined [where to acquire the required materials], estimated the time required for the work, calculated the number of soldiers and ordinary workers to be employed, pondered the materials that would be used, and recorded the amount of provisions in order to mandate the amount of work for each of the feudal lords.

Apart from the earliest walls created by simply piling up dirt and loosely packing it down without the use of retaining forms, multiple discrete layers were always employed that sometimes interspersed dry sand, pebbles, organic materials, pottery shards, and even straw, twigs, and branches in arid areas. After the wooden components were fabricated, the forms had to be correctly positioned and temporarily secured at each stage before being repeatedly shifted upward or sideways as the filling and pounding proceeded layer by layer and section by section. As the walls became more massive over the centuries, experience taught that a solid foundation that penetrated well below the surface and was capable of sustaining the immense weight had to be rigorously prepared. Deep trenches were therefore excavated and the foundation systematically built up with layers of well-pounded soil in a process identical to erecting the wall itself.

Although later practices suggest that workers were probably responsible for furnishing their own provisions and perhaps finding local shelter, whether numbering just a few hundred or soaring to thousands, they had to be fed and sometimes temporarily housed. Compelled to exert themselves from dawn to dusk irrespective of the weather, to endure dusty and muddy conditions, the conscripts must have found the work exhausting. Some sense of their misery may be gleaned from an ode preserved in the Book of Odes, a compilation that includes authentic material from the early Chou dynasty as well as compositions no doubt dating to the end of the Western Chou period. Despite having been composed long after the Neolithic ended, the depiction retains its validity because the basic techniques had remained unchanged:32

Crowds brought the earth in baskets;

They threw it with shouts into the frames;

They beat it with responsive blows;

They pared the walls repeatedly, and they sounded strong.

Five thousand cubits of them arose together,

So that the roll of the great drum did not overpower (the noise of the


Of particular interest is that the work effort involved in building the Chou capital was not only supervised, but also controlled by the beat of a drum.

Anyone who has labored with a shovel or pick, whether digging foxholes or landscaping a backyard, can easily envision the physical stress involved in constructing the simplest tamped earthen walls. Even when they were exploiting naturally occurring veins of loose soil or digging along riverbeds, ancient workers wielding cumbersome bone and stone shovels with less than ideal shapes and minimum cubic capacity must have found the effort extremely taxing. Highly compressed soil, clay, intermixed pebbles and stones, even surface vegetation and entangling roots that had to be cut and cleared away, would have immeasurably lengthened the process.

Although most walls were built with soil excavated from adjoining ditches, technological acumen early on mandated the interspersed employment of different types of soil, some of which had to be sought out, excavated, and then transported back over considerable distances and possibly difficult terrain. The simplest encirclement would have required thousands of worker-days, evidence of the magnitude of their dedication, as the population base in the average Neolithic settlement would rarely have exceeded several hundred to perhaps a few thousand. (In comparison with the work’s requirements, such low numbers immediately suggest that slaves, prisoners, or nominally subjugated peoples were coercively employed on larger projects, prima facie evidence that localized warfare was already proliferating.)

Defensive fortifications require regular maintenance and occasionally extensive repairs to prevent deterioration. The exposed surfaces seem to have frequently been finished with an exterior coating of sticky clay that congealed in the sun or was occasionally directly fired, as well as normally rendered as smooth as possible to prevent the enemy from ascending.33 Nevertheless, earthen walls unenclosed by brick or stone were always susceptible to weathering. Rain wreaked the greatest havoc from direct impact and runoff, but the drying effects of wind and the sun, coupled with repeated cycles of heat and cold, expansion and contraction, could rapidly degrade the integrity of exterior portions. Minor cracks might rapidly expand, causing the wall to quickly erode under unrelenting seasonal moisture or the occasional torrents unleashed by typhoons in coastal regions.

Vegetation (other than slippery grass) that might have provided a measure of surface protection had to be removed due to the invasive nature of roots and because bushes and shrubs might conceal stealthy assailants or provide a hand grasp for midnight enemies. Moats had to be similarly cleared of rapidly spreading vegetation and ditches of newly sprouting shrubs that might momentarily shelter enemy troops from archers shooting down upon them. Keeping the requisite angle of fire clear and preventing blind spots created by vertical precipices certainly account for the gradual slope that marks most ditch walls that abut fortifications. In contrast, the walls on the far side were normally rendered as perpendicular as possible to impede enemy descent, and most early ditches were characterized by wide mouths and narrow bases. Debris from the ever-eroding walls and accumulating leaves that could reduce a ditch’s effectiveness over time also had to be periodically cleared.

A number of historians have pondered the amount of time required to construct Neolithic and later fortifications in China. Although the relatively small site of Pao-tun Ku-ch’eng required moving 250,000 cubic meters of dirt, the walls at Cheng-chou are estimated as having required somewhere between 870,000 and 1,439,000 cubic meters of soil, depending on assumptions about the wall’s average width.34 However, the size of the city’s interior space allows some crude population and labor estimates to be conjured that coincidentally suggest the site’s immensity and the Shang’s astonishing power.

In an early experiment it was found that the surprisingly meager amount of 0.03 cubic meter of soil per hour—roughly a cubic foot—could be excavated from the surrounding terrain using bronze tools appropriate to the early Shang and only 0.02 cubic meter per hour with stone implements, drastically less than anyone experienced in digging with modern, relatively sharp shovels might have expected.35 Assuming that 10,000 laborers working 330 ten-hour days annually were impressed for the task and organized into 3,000 assigned to digging, 3,000 to transporting, and 4,000 to pounding, roughly eighteen years would have been required.36

One of the questions besetting this estimate is whether the tasks are properly apportioned, whether the number assigned to the pounding work is too low or high. Thinness being the secret to compacting the layers and realizing the extraordinary hardness required, very small-diameter branches or dowels, often bundled, were employed. Moreover, modern reconstructions have repeatedly shown that each layer requires many hours of consistent work rather than quick tamping with heavy stone weights to achieve true, virtually sedimentary solidity.37

A larger labor force would naturally accomplish the task more quickly, while fewer men and more women would take correspondingly longer. Whether slaves, prisoners of war, soldiers, lower-class members of the Shang or its subordinate proto-states, or all of these were impressed for the task, they must have required a large number of support personnel to provide the food, water, and clothing necessary for the builders and themselves, suggesting that up to 13,000 people might have been engaged in the project.

Based on the oft-suggested estimate that the average five-person household occupied roughly 155 square meters of space,38 20,000 families may have lived within the confines of Cheng-chou’s approximately 3 million square meters. Even excluding the large numbers of slaves, artisans attached to the workshops, and farmers tending small suburban gardens who would have been housed outside the walls, this translates into 100,000 people. If each family furnished one person for labor service, voluntarily or otherwise, the requisite number could easily have been assembled.

Without further replication efforts, these projections must be deemed highly conjectural. A number of other calculations have resulted in estimates that range from 5 through 8 to 12.5 years for the work at Cheng-chou.39 Even though still requiring massive amounts of labor and great organizational effort, the more reasonable assumption of 0.5 cubic meter per day per worker would significantly cut the likely time from 18 years, depending on how many were actually employed in digging.40

Several complicating factors and technical complexities that stem from the nature of the soil, the design and shape of the wall, and the wall’s exterior surface could also adversely affect the construction rate. Soft, loose soil facilitates excavation work but makes pounding more difficult and complicates the preparation of a sufficiently strong and stable foundation. (Deep excavations for ditches and moats also pose significant risks because of the danger of collapse.) Core walls were sometimes first prepared; many older walls were expanded more than once, often right out over previously excavated defensive ditches or moats; and exterior knee walls were added for strength and protection.

Neither stone, which would have required enormous preparatory work including quarrying, dimensioning, and transporting but constituted an essentially impervious surface (including to enemy diggers), nor mud blocks or bricks were employed for wall facings in the Neolithic period and Shang dynasty. However, sometimes pebbles gathered from nearby rivers and embedded in the face provided minimal additional protection against water’s scouring action, and readily collected stones were employed by peoples in the Northern complex.

In response to these issues, external threats, increasing experience, rising population, and greater wealth, the defensive measures implemented at every level not only became more extensive and complex over time, but also more regularized. The first ditches, whose dirt was often used to raise the entire community or provide early building foundations, evolved into ditches conjoined with internal ramparts created by simply mounding up the excavated soil. Eventually it was discovered that pounding the earth as it was piled up would compact and harden it, improving its defensive characteristics while simultaneously raising its resilience.

The need for excavated foundations was quickly realized and the practice of layering quickly emerged, possibly because pounding the soil as it is laid down proved the only effective method. Experience taught that intermixing layers of soil with different characteristics could produce a stronger, more stable wall, a practice that was applied to the excavated foundations and the massive building foundations then being constructed. Thereafter, Chinese walls and platforms consistently show improvement in the uniformity and number of their layers right through Erh-li-t’ou and Erh-li-kang, when thin layers of ten to eleven centimeters were routinely pounded to uniform consistency with bundles of small rounded tools. The placement of additional dry ditches and moats on the exterior and occasionally interior ultimately produced the integrated defensive systems visible in the late Hsia and early Shang at such sites as Yen-shih and Cheng-chou.

The earliest fortified towns generally assumed two basic shapes, circular and rectangular, though portions of a wall’s course might be modified to resolve or exploit abnormalities in the terrain, resulting in trapezoids, crushed corners, and odd indentations. In addition, ancient fortified sites invariably took advantage of heights provided by tumuli and mounds and exploited natural depressions, lakes, marshes, and flowing water. In fact, one of the primary conceptualizations of Chinese military science—that configurations of terrain convey strategic advantages—had already been recognized and was being actualized, at least in rudimentary form, in the Lungshan period.41 Moreover, gate openings were fixed neither in number nor in position, but located largely in accord with exigency despite a preference for centering on each of the walls. With the addition of inner citadels or royal quarters demarked by internal walls, the final form of the Chinese capital, one theoretically distinguished by inner and outer walls rather than the thickness of the fortifications or integrity of the moat-encircled defensive compound, would be realized.

If you find an error or have any questions, please email us at Thank you!