The Character and Function of Ancient Chinese Walls and Fortifications

The Great Wall of China / Photo by Jakub Halun, Wikimedia Commons

The concepts and technology of defensive fortifications fitfully but continually evolved.

By Dr. Ralph D. Sawyer
Senior Research Fellow
University of Massachusetts

The Fortifications

LONG VIRTUALLY DEFINED by the mythic aspects of its Great Wall, China’s tradition of wall building far exceeds the most exaggerated claims spawned by its famous icon. As early as 7000 BCE, protective ditches had already appeared in scattered settlements along the two great river systems and their tributaries. However, rather than being employed to construct defensive fortifications, the excavated soil furnished the raw material for structural foundations and for raising the entire settlement above the surrounding terrain, thereby preventing inundation from pooling rainwater and overflowing streams and providing a slight tactical advantage.

In response to escalating threats, the concepts and technology of defensive fortifications fitfully but continually evolved over the next five millennia until what has been claimed to be the distinctive form of the double-walled Chinese city protected by an external moat was finally realized. Several stages can be discerned: shallow perimeter ditches; simple ditches augmented by mounded earthen walls, the latter simply the by-product of ditch and moat excavation; earthen walls deliberately constructed with early pounding techniques, generally augmented by perimeter ditches or early moats; massive, rigorously constructed hang-t’u (stamped or pounded earth) walls coupled with expansive protective moats; and the culmination of Chinese martial engineering, massive rammed earth walls faced with stone or brick, buttressed with waist walls, and systematically augmented by conjoined exterior and interior moats.

The massive moat and wall surrounding The Forbidden City / Photo by degreeero, Flickr, Creative Commons

Although low stone walls had been erected in conducive, semiarid regions such as Inner Mongolia as early as the Neolithic period, solid walls of stone, brick, and even marble were not constructed in China until recent centuries, and then only in limited areas. Moreover, regional practices and localized disparities would persist across the centuries despite consistent progress in understanding the engineering principles required for fortification construction, the development of work methods, and the evolution of administrative measures. Some towns in the late Shang and early Chou continued to employ nothing more than ditches long after massive fortifications had become commonplace.1 Conversely, peripheral states such as Chao that faced external threats began constructing lengthy defensive structures, appropriately termed “long walls,” along their proclaimed borders late in the Warring States period, a practice that would eventually culminate in the colossal Ming dynasty edifice.

Ditches and walls have always been the first defensive measures adopted in response to environmental threats and human violence. In consonance with the Marxist dictum that civilization evolved from a matriarchal, egalitarian society through patriarchal-based structures and an inherent tendency to warfare, Chinese scholars have long claimed that the earliest Neolithic ditches were intended simply to prevent domesticated livestock from venturing forth and thwart incursions by wild animals. However, although they may have been minimally effective in achieving the former, the ditches would never have deterred agile predators from entering a settlement. Therefore, they must have performed a defensive function, one that was enabled by the raised profile of most early villages.

Many premodern civilizations fabricated highly formidable walls from readily available rocks and laboriously quarried stone blocks, but others resorted to more easily worked, though perishable, materials such as wood to erect functional barriers, ranging from hedgehogs and simple palisades through complex log fortresses. Those dwelling in environments bereft of readily accessible trees and stone were compelled to construct primitive earthworks, employ sun-dried mud blocks, or exploit the evolving technology of kiln-fired bricks. Although China was still heavily forested throughout the Neolithic period, the soil deposited in the alluvial plains, even the famous sticky “yellow earth” washed down by the Yellow River, lent itself far more easily to digging, carving, and shaping with the basic tools of the time—axes, scrapers, and short shovels cobbled together by affixing appropriately shaped pieces of stone or bone to a wooden handle—than did hewing trees or chiseling stone from local outcroppings. Concerted pounding could compress and harden the soil to an almost concretelike substance roughly equivalent to sedimentary rock that would, with minimal maintenance, endure for millennia.

A number of strategically advantageous locations near rivers or amid mountains were continuously occupied from Neolithic times. As these settlements evolved into fortified towns in the Lungshan period and labor became available or the degree of threat increased, their massive walls were enlarged. Preexisting structures were generally reshaped and significantly augmented rather than simply rebuilt, often deliberately expanded out over the very ditches that had protected them. Many Shang sites subsequently served as the nucleus for resplendent cities when the early Chou established the fortified power centers that eventually evolved into heavily fortified Spring and Autumn and Warring States capitals.2

A section of the Yangtze River flowing through Yunnan / Photo by BrokenSphere, Wikimedia Commons

Unfortunately, despite their critical role in civilization’s evolution and impact on warfare, the history and technology of Chinese fortified settlements have yet to be systematically studied.3 Nevertheless, discoveries over the last several decades have pushed the defensive horizon well back into the Neolithic period, permitting a tentative recounting of the salient features within the context of China’s military history.4 However, contemplation of more profound questions, including whether increasingly urbanized, fortified sites stimulated warfare or simply reflected warfare’s increased scope and intensity, must be foregone.

Particularly in the earliest periods, when the populace was comparatively small and competition for land minimal, burgeoning groups invariably sought to exploit the terrain’s strategic advantages. Later ages might be able to employ vast labor resources to overcome natural obstacles and reshape the terrain’s configuration, but Neolithic groups had to choose their sites for their environmental benefits and protective features. Except for mountainside villages constructed in semiarid locations, migrant settlers seeking well-watered areas generally established their communities alongside marshes, lakes, ponds, streams, and rivers. The many early towns that were not abandoned despite suffering extensive flood damage show the overwhelming importance of immediate proximity to active water sources.

Still bodies of water may have presented little flood risk, but the danger of being inundated by a sudden increase in the volume of an adjacent river or stream prompted the implementation of basic protective measures. Insofar as embankment building and preventative channeling did not begin until well after the technology of rammed earth had been reasonably perfected, the sole options were choosing somewhat raised terrain, especially natural terraces, or artificially increasing the village’s overall height to produce a sort of mound settlement, the precursor of the platform city.

Cognizant of water’s effectiveness as a defensive barrier, settlement planners—there is considerable evidence that even early settlements were built according to design rather than haphazardly evolved—frequently chose sites amid multiple bodies of water. Many early towns were surprisingly positioned within the confluence of two or three rivers or between a river and nearby lake or marsh rather than simply above or along the outer bank of a single river, despite the high flood risk and the inconvenience of having to cross these watercourses on a daily basis.

The Walls

Even in the Lungshan period (3,000 to 2,000 BCE) immediately prior to the Hsia dynasty, Chinese walls had already reached astonishing dimensions and sometimes exceeded twenty-five meters (eighty-one feet) in width. Constructing these fortifications must have required a Herculean expenditure of energy that inflicted intense misery on the workers, yet the motive for their massiveness remains unknown. Later periods would exploit the resulting top surfaces to deploy contingents of soldiers, set up countersiege engines or large trebuchets, and erect low protective walls or crenellations for archers and crossbowmen. However, combatants in the Yangshao, Lungshan, and Shang lacked large defensive devices, employed very limited infantry forces, and didn’t otherwise require a large platform to repel attackers because city assaults were rarely undertaken, making the ranks to the rear largely ineffective.

Lungshan Temple in Lukang / Photo by David Stanley, Flickr, Creative Commons

Because these ancient peoples never built watchtowers or constructed bow emplacements on top of these ancient walls except possibly above a few gates, another justification for their inordinately great width must be sought.5 One commonly advanced explanation claims that this massiveness was required to withstand the scouring effects and relentless pressure that might be exerted by nearby creeks and rivers, that only substantial walls and deliberately emplaced dikes could prevent the disastrous inundations that plagued settlements along the Yellow and Yangtze rivers, as well as their more vigorous tributaries, throughout Chinese history.

Despite the assertion frequently seen in the later military writings that “in walls, massiveness is best,” this explanation remains inadequate for several reasons. First, many cities protected by imposing walls were located far from problematic rivers and streams. Second, height rather than width has always proven the most formidable psychological deterrent and greatest physical obstacle to external forces, yet China’s walls were almost invariably more expansive than necessary to sustain their height. Third, the problem of erosion could be solved in part by facing the exterior portion with smaller rocks, overlaying exposed surfaces with hardened clay, constructing stone knee walls, or encapsulating the earthen core in stone or brick, as witnessed from the Han onward. Truly thick walls would only become necessary in the Spring and Autumn period, when aquatic warfare commenced and powerful assaults effected by damming rivers and diverting mountain streams could not be withstood without this massiveness.6

Nevertheless, indestructible bulk creates a sense of awesomeness, and insurmountable fortifications foster a sense of security. Apart from functioning as a military bastion and serving as a refuge against such natural onslaughts as typhoons, walls define communities and nurture a sense of uniqueness by physically isolating the members of the settlement. Western psychoanalytic literature has long pondered the significance of this inclusiveness and the symbolic sense of sanctuary that walls can create, and popular analysis continues to attribute China’s ethnocentric tendencies (or supposed “citadel mentality”) to their deliberate employment over the millennia, including to segregate the much-denigrated “barbarians.” Compelling the populace to engage in organized wall building would have enhanced this sense of common identity while presumably solidifying the chief’s authority and absorbing excess disruptive energy.

Whatever the validity of these insights, it should be remembered that geopolitically defined states did not appear prior to the late Western Chou. Walls were first constructed around villages and towns, then tentatively between states to demark borders as much as thwart incursions, and only late in the Warring States period between so-called sedentary China and the nomadic steppe peoples. Internal compartmentalization slowly developed in late Lungshan cities, then accelerated dramatically in the Shang and thereafter, increasingly segregating the royal quarters from the privileged populace, defining the artisanal sectors, and segmenting the residential and industrial areas into smaller and smaller units, ostensibly for defense but more often to prevent people from moving about freely and to thwart evildoers. Apart from impeding external intrusions, the walls came to be deemed an essential measure for regulating the populace, controlling their activities, and managing intercourse with the outer world, functions that they inherently performed from inception even in the absence of explicit theorizing.

More important, rather than just preserving the inhabitants and their property, walls provided the means to protect military forces, project power, and control the countryside. Prior to the advent of cannon and explosives, fortified walls constituted tremendous combat multipliers in early China because they allowed small but determined garrisons to successfully withstand virtual hordes of aggressors. Accordingly, in a much misunderstood and frequently misquoted statement, Sun-tzu condemned wasteful assaults on fortresses as the lowest possible tactic and incisively postulated that even minimal forces ensconced in a defensive posture would prove more than adequate. Fortifications could thus be exploited to maximize the value of limited power rather than simply provide a place of refuge in times of weakness and necessity.7

Statue of Sun Tzu in Yurihama, Tottori, in Japan / Photo by 663highland, Wikimedia Commons

The growth of settlements, largely synonymous with the evolution of civilization in ancient China, fundamentally affected the history of warfare. As populations increased, class differentiation arose, and ruling elites emerged, a few developed into sizable military and administrative centers. Material wealth began to accumulate in them; ceramic, bronze, and other early craft industries developed in their immediate environs; significant surpluses produced by the burgeoning agricultural sector were stored within them; and they became centers for consumption. Even if (as vociferously claimed) their growth was power based rather than economically derived, rudimentary trading practices also appeared that further stimulated the evolution of certain fortified towns into regional centers as well as highly desirable military targets.

The nature of warfare in the Neolithic remains unknown, but based on Shang tactical practices and the lack of destructive evidence at the sites so far investigated, it can be concluded that these massive fortifications proved effective in deterring attacks. Rather than seeking to conquer and annex them, chieftains in search of the spoils necessary for self-aggrandizement and motivating their followers targeted contiguous areas and unprotected settlements. Extensive grain pits discovered at Wu-an Ts’u-shan dating to the second half of the sixth millennium BCE, which still held the equivalent of some fifty metric tons of millet, provide incontrovertible evidence of agriculture’s ever-increasing productivity and that Neolithic settlements were already generating astonishing grain surpluses, one of the foundations of power and vitality in ancient China.8

Although ditches, then walls, developed in accord with the rise of agriculture and animal husbandry, other factors apparently stimulated their comparatively rapid evolution, including environmental degradation, climatic changes, population increases, and the ever-intensifying struggle for edible natural resources and wild animals in the less fertile and more arid areas. A significant shift to a threat-dominated context is generally envisioned as having occurred during the transitional early Lungshan period. However, the recent discovery of astonishingly early Neolithic defensive measures and numerous studies of ancient weapons and skeletons that predate this demarcation indicate China had a long heritage of violence that may well have commenced with its famous Beijing Man.

The oldest presently known Chinese settlements, mostly scattered individual hamlets, date back prior to the Neolithic Age that commenced in parts of China by approximately 10,000 BCE.9 For the next 3,000 years primitive yet constantly evolving agricultural practices spawned slowly growing settlements that were generally concentrated near rivers and accessible marshes. Unsurprisingly, so-called Hua-Hsia or Chinese civilization thus developed primarily (but not exclusively) in the middle and lower reaches of the Yellow and Yangtze rivers, both well-watered areas with soft, alluvial soil.

Although barter and other forms of minimal interaction within a day’s walking distance seem to have occurred with the inception of these settlements, life in that era of “ten thousand villages” probably differed little from the Tao Te Ching’sidealized depiction:10

Although neighboring states look across at each other,

And the sounds of roosters and dogs be mutually heard,

Unto old age and death, the people do not travel back and forth.

The earliest among the several sites that have been at least minimally excavated and radiocarbon dated is Nan-kuang-t’ou at Hsü-shui in Hebei, which flourished in the ninth or eighth millennium BCE.11 Next in succession appears to be the P’eng-t’ou-shan cultural site on the plains just north of the Li River in Hunan, essentially located in the middle reaches of the Yangtze river basin. Although dates stretching from 7800 to 5795 BCE have been suggested for this site, it was probably occupied from 6900 to 6300 BCE.12 This settlement is particularly important for the history of Chinese warfare because it was already protected by a circular ditch, and the dwellings, which average some thirty square meters and are much larger than formerly, were probably intended for extended family use.13 Moreover, the benefits of rice cultivation were already evident, satisfying one of the oft-stated conditions for true warfare.

The Yellow river basin area, marked by several cultural clusters including the P’ei-li-kang, has also yielded numerous sites with dates ranging from 6100 to 5000 BCE, as has the middle Yangtze River area. One of the oldest ditched settlements from this era, the now well-known Yangshao village of Pa-shih-tang in Hunan, reportedly dates back to about 6000 BCE and was continuously occupied for about a thousand years before finally being abandoned for another two thousand.14 Located near a marsh, Pa-shih-tang was defined by an encircling protective ditch that had been augmented for defensive purposes by a low interior wall and therefore marks the earliest stage of conjoining walls with ditches. Despite meeting a river that flows in the north and west (which would have provided additional protection at a distance), from the terrain’s contour and evidence that the ditch was periodically cleared of debris, it never functioned as a moat but may have provided drainage for the settlement.

A model of Jiangzhai, a Yangshao village / Prof. Gary Lee Todd, Wikimedia Commons

The site, which exploited a slight natural rise in the terrain, spans some 200 meters north to south and 170 meters east to west. The low interior wall was constructed by simply mounding up soil excavated from the ditch, none of the pounding or other work that characterizes tamped earth walls having been employed. Two stages are visible. The initial effort apparently created a slight wall some 0.5 to 1.0 meter high and 6 meters wide that was characterized by a gradual pitch of 20-30 degrees. The accompanying ditch had a width of 1 to 2 meters but a very shallow depth of less than 0.5 meter, barely enough to impede aggressors standing at the bottom as they confronted the low wall. However, without any increase in height the base of the wall was subsequently widened to about 7.5 meters and the mouth of the ditch opened to 4 meters (with a bottom expanse of 1.5 meters), as well as deepened to a functional 1.5 to 2.0 meters, sufficient to impede aggressors and thus clearly defensive in intent, despite opinions to the contrary.15

A terrace settlement at Ta-ti-wan in Gansu’s eastern portion, reportedly the earliest Yangshao site yet discovered in the province, eventually evolved to cover about a million square meters. Continuously occupied for some 3,000 years starting about 5800 BCE, it too was protected by a circular ditch (or possibly moat) that ranged from 5.5 to 8 meters wide at top and from 2.5 to 3.5 meters at the bottom, and had a depth of 3.5 to 3.8 meters.16

The period from 5000 to 3500 BCE is generally considered the late Neolithic. Although discussions of copper’s discovery and bronze’s development remain highly speculative, their incipient beginnings are said to fall within this era, even though the products of bronze technology do not become truly noticeable until the appearance of the Hsia. Defensive ditches continue to feature prominently in such famous, well-defined sites as Pan-p’o and Lin-t’ung Chiang-chai.17 Both were continuously occupied for several centuries in the Lungshan period, but based on structural remains and numerous artifacts including arrowheads, Pan-p’o is now considered one of the defining sites for the preceding early Yangshao culture.18

Located approximately 800 meters from the Ch’an River, which has now shifted from one side of the village to the other, and in the immediate vicinity of the strategically critical city Hsi-an (Xian), Pan-p’o stands about 9 meters above the nearby river plain. A typical prehistoric village, its earliest stages have been variously dated to between 5000 and 4000 BCE.19 A massive protective ditch some 6 to 8 meters wide at the top and 1 to 3 meters at the bottom, with an average original depth of 5 to 6 meters and a full circumference of 600 meters, surrounded the site, delimiting a total area of about 50,000 square meters. This ditch was marked by a strongly defensive contour: the outer wall has a nearly vertical drop that would have made controlled descent difficult, but the inner wall near the village gradually slopes away, preventing blind spots below the interior rim and fully exposing aggressors to bow shots from above.

Remnants of a meter-high wall formed by simply piling excavated dirt into a continuous mound have been detected just inside the town’s perimeter. These somewhat extemporaneous fortifications seem to mark an intermediate stage between simply employing a protective ditch and deliberately conjoining ditches and moats with carefully erected, rammed earth walls. Moreover, insofar as the soil excavated from protective ditches had previously been used for building platforms and raising a settlement’s overall height (and Pan-p’o seems to have been constructed on a 0.5-meter platform),20the wall seems to have been intentionally constructed and was apparently augmented late in Pan-p’o’s period of occupation, contrary to assertions that it lacks a deliberate character.21

A second, semicircular interior ditch that defines and protects about a third of the settlement apparently was constructed about the same time. Much reduced in scale, its remnants vary between 1.4 and 2.9 meters in width at the top and 0.45 and 0.84 meter at the base, and it has a depth of 1.5 meters, or about a man’s height. It may have been a functional precursor to the walled-off royal quarters that would become visible in Shang and other double-walled cities. At least two guardhouses seem to have controlled access to the interior realm and the outer compound, presumably additional evidence of class differentiation having emerged.22 Wooden palisades appear to have been erected on both the interior and exterior of the main ditch, implying great concern with security problems. These remnants reinforce the idea that Pan-p’o’s defenses mark a transitional stage between simply relying on ditches or palisades and erecting the rammed earth walls typical of fortified towns and cities.

Pan-p’o archaeological site / Creative Commons

Another early, fully excavated Yangshao site identified with the Pan-p’o culture that has attracted considerable attention is Chiang-chai, located near Hsi-an.23 Sitting atop a slight plateau defined by the two rivers that flow around three of its sides, it was backed on the fourth by Mt. Li to the southeast. The well-defined dwelling area, roughly 18,500 square meters, consisted of a central square surrounded by houses that were inwardly oriented, probably for defensive as much as psychological purposes. Nearly circular at some 160 meters east to west and 150 meters north to south, it is partially defined by four ditch segments that run from the east to the south, presumably the remnants of a continuous perimeter system that encircled the entire core area. Its population has been estimated at about 400 to 450.

Although it is generally claimed that these ditch remnants were intended for both drainage and protection, their defensive function probably would have been primary, because the site seems to have been chosen for the enhanced security offered by the Wei River some four kilometers to the north (but apparently somewhat closer when running in its old course in antiquity) and the Lin River to the southwest. The southernmost ditch segment appears to have once been connected with the Lin River, but its height above the river’s level would have prevented water from filling it. (Ditches become moats when they penetrate the level of the local groundwater or are lowered sufficiently to derive water from a nearby source, whether a river or lake.) The dimensions of the ditches vary considerably, but they would have been sufficient to impede aggressors: 1.5 to 3.2 meters in width at the top, tapering down to between 0.5 and 1.3 meters at the bottom, and about 1 to 2.4 meters deep.

Defensive ditches dating back to the middle Neolithic period, or 7000 to perhaps 5000 BCE, have also been found throughout southern Inner Mongolia, western Liaoning, and northeast Hebei, including the general area of Beijing. Somewhat oversimplified, it can be stated that ditches constituted the primary defensive measures in the Hsing-lung-wa (variously dated as 6200/6000 to 5600/5500 BCE), Hung-shan (3500- 3000 BCE), and Hsia-chia-tien (2000-1500 BCE).24 However, because these were hunting cultures, the ditches need not have been particularly formidable to effectively demark the settlement’s bounds and sufficiently retard enemies.

At the definitive site of Hsing-lung-wa the ditch had a radius of 80 to 100 meters, was about 2 meters wide and 1 meter deep, and was broken only by a single entrance on the northwest. Another site, Peich’eng-tzu, is bordered by a river on the western side and therefore required ditches just on the remaining three.25 Among the other small, circular sites composed of well-arrayed dwellings found in Inner Mongolia and western Liao-tung that have been identified as Hsing-lung-wa, one at Ao-han-chi in Inner Mongolia stands out. The dwellings, which were concentrated into eight well-ordered rows of ten, encompassed a surprising 50 to 80 square meters each (with two even attaining 140 square meters), much in contrast to contemporary Yellow River dwellings of a mere 4 square meters.

Ditches continued to be employed as the sole defensive measure at many sites even after wall building began to emerge. For example, an immense ditch varying between 15 and 20 meters in width and marked by depths of 2.5 to 3.8 meters has recently been discovered in Hubei near Sui-chou. Its somewhat oval shape of 316 meters north to south and 235 east to west enclosed a site of 57,000 square meters. Constructed in the Ch’ü-chia-ling around 3000 BCE, it was used well into the Shih-chia-ho, even though the ditches were allowed to deteriorate relatively early. A well-established settlement with a wet rice agricultural basis, it is one of at least ten such settlements in the middle Yangtze River area known to have had comparable protective ditches.26


  1. Two examples would be Chi-kung-shan and Wu-chia Ta-p’ing at Wei-ning in Kuei-chou, both dating to about 1300 to 700 BCE (Kuei-chou-sheng WWKK YCS et al., KK 2006:8, 11-27 and 28-39, and Chang Ho-jung and Luo Erh-hu, KK 2006:8, 57-66).
  2. A classic Lungshan site of some 140,000 square meters known as “the defensive ancient city” (“Fang-ku-ch’eng”), located in Shandong, furnishes a particularly good example of fortification continuity; it was employed right through the Warring States, when it served as a stronghold on Lu’s eastern border. (See Fang-ch’eng K’ao-ku Kung-tso-tui, KK 2005:10, 25-36.)
  3. Even Paul Wheatley’s erudite but now outdated examination of city growth, The Pivot of the Four Quarters, never ponders the craft of wall building. Furthermore, the two volumes in Needham’s Science and Civilisation in China series that contemplate essential aspects of fortification—Civil Engineering and Nautics and Military Sieges and Technology—barely mention Neolithic and Shang fortifications.
  4. Despite many hundreds of archaeological reports, only a few synthetic overviews such as P’ei An-p’ing, KK 2004:11, 63-76, and Shao Wangping, JEAA 2 (2000), 195-226, have appeared. (See also Shao’s “The Interaction Sphere of the Longshan Period.”) A few works have discussed city history, including Ch’ü Ying-chieh, Ku-tai Ch’eng-shih, 2003; Ning Yüeh-ming et al., Chung-kuo Ch’eng-shih Fa-chan-shih, 1994; and Yang K’uan, Chung-kuo Ku-tai Tu-ch’eng Chih-tu-shih Yen-chiu . Valuable materials also appear in Li Liu, The Chinese Neolithic: Trajectories to Early States, and Chang Kwang-chih and Xu Pingfang, eds., The Formation of Chinese Civilization.
  5. Other than at Ma-chia-yüan Ku-ch’eng, no evidence has yet been reported for defensive works atop walls, not unexpectedly since erosion has affected almost every wall so far excavated.
  6. For the history of aquatic warfare in China, see Ralph Sawyer, Fire and Water.
  7. In “Military Disposition,” Sun-tzu states: “One who cannot be victorious assumes a defensive posture, one who can be victorious attacks. By assuming a defensive posture your strength will be more than adequate.” In “Planning Offensives” he further observes: “The highest realization of warfare is to attack the enemy’s plans; next is to attack their alliances; next to attack their army; and the lowest is to attack their fortified cities.” The passage continues with a sweeping condemnation of the wastefulness of citadel assaults, though not their complete exclusion. (For further discussion and a complete translation of the Art of War with historical introduction and textual notes, see Sawyer, Sun-tzu Art of War.)
  8. The eighty-eight pits at Ts’u-shan near Wu-an in Hebei held varying amounts of desiccated grain, which, when freshly stored, would have weighed over 50 metric tons. (See Jen Shih-nan, KK 1995:1, 38-39.) Similar reserves (such as 120 metric tons at Ho-mu-tu) have been found at other sites. However, rather than a surplus, they may have been the basic subsistence requirement for the coming year.
  9. The Neolithic is generally taken as encompassing 10,000 to 3500 BCE, though others extend it to 2100, the reputed date of the Hsia’s inception. Bronze metallurgy emerged during the Hsia; nevertheless, as it remained primarily a stone culture, the first two centuries are sometimes subsumed within the Neolithic as well.
  10. “A Small State with Few People” in the traditionally received text. (For a complete translation and contextual discussion, see Sawyer, The Tao of War.)
  11. See Jen Shih-nan, 37. (Various dates have been derived from the recovered artifacts, resulting in controversy.)
  12. Jen Shih-nan, 37-38.
  13. Yen Wen-ming, KKWW 1997:2, 35.
  14. Site descriptions are based on P’ei An-p’ing, KK 2004:11, 66-67, and Chang Hsüeh-hai, KKWW 1999:1, 36-43. There are numerous difficulties reconstructing the earliest stage and disagreements about how to interpret the archaeological evidence. Some archaeologists have dated the site to nearly 6000 BCE, but the more commonly recognized range is Yen Wen-ming’s 5450 to 5100 BCE (cited on page 35 of The Formation of Chinese Civilization).
  15. Another P’ei-li-kang platform site in Henan was not only surrounded by water on three sides, but also further protected by two shallow ditches about half a meter deep that could have functioned for drainage or demarcation as much as defense. (One of the ditches varied between a functional 1.65 and 5.15 meters in width, but the other was only 0.75 to 1.1 meters wide.) However, the location’s desirability is evident from its continuous occupation into the Erh-li-t’ou cultural phase (Chang Sung-lin et al., KK 2008:5, 3-20).
  16. For site reports see Kan-su-sheng WWKK YCS, KK 2003:6, 19-31, and Lang Shu-te, KK 2003:6, 83-89. Only minimal information on the ditch has yet been provided.
  17. Lin-t’ung Chiang-chai, which was continuously occupied right into the Warring States, was protected by a circular ditch that encompassed approximately 55,000 square meters or slightly more than Pan-p’o. However, as the ditch itself has not yet been analyzed, little can yet be said about its profile or overall significance. (For a basic report, see Honan Sheng Kung-yi-shih Wen-wu Pao-hu Kuan-li-suo, KK 1995:4, 297-304, as well as Ning Yüeh-ming’s appraisal, Ch’eng-shih Fa-chan-shih, 13.)
  18. For a basic discussion of the Yangshao culture that was originally defined by its red-tinged pottery (in contrast to the black pottery of the later Lungshan period), see K. C. Chang, 1986 (which covers Pan-p’o on pp. 112-123) or Yen Wen-ming, Yang-shao Wen-hua Yen-chiu. Yangshao culture, generally dated from 5000 to 3000 BCE, was centered in the Kuan-chung area and included Gansu around the T’ien-shui River and the upper reaches of the Ch’ing-shui, the middle and upper reaches of the Luo River, the upper portion of the Han River in Shaanxi, the southern portion of Ching-hsi, and the Yü-hsi area, with the Wei River as the focus (Chang Hung-yen, KKWW 2006:5, 66-70, and WW 2006:9, 62-69, 78). Our discussion of Pan-p’o is primarily based on Ch’ien Yao-p’eng’s two articles, KK 1998:2, 45-52, and KK 1999:6, 69-77. However, also see Yen Wen-ming, SCKKLC, 146-153. (Evidence indicates that Lin-t’ung Chiang-chai was occupied well into the Warring States period.)
  19. For example, Ch’ien Yao-p’eng believes the site flourished from 4770 to 4190 BCE or almost 600 years, but others such as Ning Yüeh-ming et al., 1994, 12-13, date it as late as 4000 BCE.
  20. Ch’ien Yao-p’eng, KK 1999:6, 69.
  21. Both Pan-p’o and Lin-t’ung Chiang-chai are marked by raised platforms and well-smoothed earth. It should be noted that Ch’ien Yao-p’eng, 46, believes that simply widening the walls, as was also done at P’eng-t’ou-shan, should be understood as an afterthought to ditch excavation rather than a deliberate attempt to solidify them and increase their height. In contradiction, Chang Hsüeh-hai concludes that the interior mounded wall was actually the result of a later, more deliberate effort, and notes that there is evidence for a third ditch some ten meters beyond the main one that may have partially furnished the dirt for the inner wall or a no longer visible outer wall. (See Chang Hsüeh-hai, KKWW 1999:1, 41-43.)
  22. See Ch’ien Yao-p’eng, KK 1998:2, 48-52.
  23. For a recent discussion see Pi Shuo-pen et al., KKWW 2008:1, 9-17. Yen Wen-ming, WW 1990:12, 21-26, also briefly discusses the site’s significance, but cites somewhat different measurements, including 160 by 210 meters for a total area of 33,600 square meters. (The settlement’s probable appearance is depicted in Chang Kwang-chih and Xu Pingfang, eds., The Formation of Chinese Civilization, 68-69.)
  24. For three illustrative sites, see SHYCS Nei Meng-ku Ti-yi Kung-tso-tui, KK 2004:7, 3-8.
  25. Liu Kuo-hsiang, KKWW 2001:6, 58-67. In contrast, the settlement at Hou-ma Tungch’eng-wang identified with the Miao-ti-kou culture has a ditch about 11 meters wide and 2.6 meters deep. (See Kao T’ien-lin, KK 1992:1 62-68, 93.)
  26. Hu-pei-sheng WWKK YCS, KK 2008:11, 3-14.

From Ancient Chinese Warfare, by Ralph D. Sawyer (Basic Books, 03.01.2011), published by Erenow, public open access.