7.
DEFINITIVE EVIDENCE FOR the Shang’s military and economic power may be seen in several important archaeological sites, including Yen-shih, Cheng-chou, Anyang, and P’an-lung-ch’eng. (The first two are critical for understanding the early history of the Shang, the last for its projection of power southward, while Anyang was the capital over the final two centuries.) Although much of Yen-shih and Cheng-chou still remains buried despite several probing excavations, enough has been uncovered to identify them both as major sites dating back to the Shang’s initial period. However, their relative sequence and importance have become matters of acrimonious contention. Proponents argue over whether Yen-shih was the initial capital of Po or at least a secondary or western capital known as Hsi Po, or Cheng-chou was Hsi Po or perhaps Ao, the site to which Chung Ting moved the Shang administrative headquarters in the middle period, as well as other possibilities.1
The most reasonable explanation—which will be argued below after examining their basic features but need not be accepted to understand the military character and significance of Yen-shih—is that the multilayered site at Cheng-chou, with its vast dimensions and numerous, impressive artifacts, was a fully developed royal capital constructed at a continuously occupied Shang site, but from the outset Yen-shih was a military bastion. It was probably erected shortly after the conquest in the heart of enemy territory to serve as a fortress even though Erh-li-t’ou itself had been occupied because, contrary to traditional accounts, many Hsia groups remained unsubmissive, as might be expected if the Shang campaign had focused on eliminating the ruling clan.
In addition to these critical cities, several hundred Shang era sites have now been located, suggesting that the total number of substantial fortified towns or cities, including those of other peoples, may have numbered about seventy. In consonance with the historical materials furnished by the oracle bones from Anyang, a few such as Wu-ch’eng and Tung-hsia-feng have greatly contributed to a much revised understanding of the dynasty’s power, extent, and dynamics.
All the Shang cities bear witness to increasingly sharp class distinctions and the evolution of royal groups with immense power, including life and death, over the general populace; the development of massive, sophisticated buildings; segmented inner quarters for the highly privileged; the slow emergence of burgeoning economies based on increased agricultural yields, animal husbandry, handicraft production, conspicuous consumption, and limited trade; the further perfection of bronze technology, resulting in large workshops capable of casting massive cauldrons, fine symbolic items, and weapons, the latter in multiple molds; and clear differentiation in the dwellings within these citadels and across their expanded sites.2
The Shang was also capable of mobilizing the large labor forces required to undertake ever greater fortification work, vast communal projects, and deeper, more complex tombs. Although Yen-shih and Cheng-chou embody important Lungshan characteristics, including expansive wall systems, they are undoubtedly early Shang fortified capitals. However, just as at Erh-li-t’ou, no defensive fortifications apart from a single moat have yet been discovered amid the opulent remains at Anyang, immediately raising the question of whether those rulers were too immersed in the pursuit of pleasure to undertake them or simply felt that surpassing military power rendered them unnecessary. If so, the absence of defensive walls proved a fatal conceit, because the last emperor lacked a defensible refuge after his forces were vanquished at Mu-yeh.
Although traditional scholars have long acrimoniously but futilely argued over the identity and location of the preconquest capitals, recent discoveries showing the preconquest forcible displacement of Hsia cultural elements apparently confirm that the incipient Shang seized terrain early on and erected citadels to block its nemesis at peripheral points of confrontation. In particular, the remnant structures at Chiao-tsuo in Fu-ch’eng, Henan, appear to comprise a preconquest bastion that was constructed in a crucial area where the Hsia had been exerting, possibly even expanding, its influence and thus clashed with the Shang, then an increasingly dynamic post-tribal entity intent upon projecting power and broadening its domain.3
Abutting the T’ai-hang Mountains in the Shen River area, Chiao-tsuo’s roughly 300- by 310-meter-long walls form a virtually perfect square that encompassed somewhat more than 90,000 square meters, impressive but too limited for the site to have served as anything more than a detached city or secondary bastion despite suggestions that it might have been the predynastic capital of Yung. Based on remnants, the original walls apparently averaged 15 meters in width and stood at least several meters high. Although marked by the usual 8- to 12-centimeter-thick layered construction typical of the period—nineteen layers being required for the 2.1-meter-high eastern wall—a somewhat unusual method of pounding with bundles of fifteen to twenty comparatively small sticks was employed. All the walls were erected on ample excavated foundations, ensuring their stability.4Several major structures have been discovered, and the citadel’s interior was divided into two compounds.
Apparently maintained as a militarized city until somewhat after Cheng-chou was abandoned, postconquest Chiao-tsuo presumably continued to function as a control point over the Hsia populace and then a bulwark against the Eastern Yi. By the middle of the dynastic Shang period, after the Yi had been defeated and Hsia opposition had evaporated, it may have become an unnecessary expenditure5 or a casualty of the internal turmoil supposedly responsible for Cheng-chou’s demise.
Another obvious military bastion, the Shang-fortified city at Yen-shih is located some ten kilometers east of Luoyang in the Luo River watershed just above the present-day river, close on a marsh that encroaches from the southeast. First discovered and excavated in 1983, it continues to be the subject of ongoing investigations in the quest to determine its historical identity.6 Several possibilities continue to be argued despite the findings of the Xia-Shang-Zhou dynastic project: Chen Hsün, the last Hsia capital; Hsi Po, the first Shang capital; an early Shang secondary capital, perhaps paired with Po (which is then said to be Cheng-chou); an early Shang military bastion; T’ai Wu’s new capital of Hsi; T’ai Chia’s T’ung Palace;7 or even P’an Keng’s new capital, known as Po Yin.
As the diagram shows, the city eventually assumed the shape of a contorted rectangle that tapers slightly downward in the upper northeast corner and is severely squashed by the marsh in the lower southeast. With a remnant circumference of 5,330 meters, the outer walls span some 1,700 meters from north to south and 1,215 meters from east to west, although the southernmost portion extends only 740 meters before turning upward. A total of seven gates have been identified; eleven large roads oriented along the cardinal directions crisscross the interior; and a formidable ditch some 20 meters wide and 6 meters deep encircles the walls at a distance of approximately 12 meters. Clearly functional in intent, its near side gradually slopes downward, facilitating archery fire while precluding protective concealment, but the far side is a daunting, nearly vertical, 19 feet. When filled with water and functioning as a moat, its 20-meter expanse, although easily negotiated by swimmers, would have required rafts or boats to convey any sort of siege or assault equipment.
The remnant walls of the greater enclosure presently stand 2.9 meters over the outside terrain and 1.5 to 1.8 meters above the interior, but were certainly higher when first erected. They are 17 to 18 meters wide at the base and taper to just under 14 meters at the top and have a perpendicular outer face to preclude easy ascension. Constructed by the usual pounded earth technique that produced layer thicknesses varying between 8 and 13 centimeters, they were underpinned by a similarly compressed, multilayer foundation excavated down between 1.3 and 1.7 meters, with an expanse slightly narrower than the ground-level wall base.8 However, the consistent use of a single soil type rather than intermixed layers of distinctively different soil is interpreted as evidence of either less sophisticated engineering practices or haste in constructing the defenses. Nevertheless, the layers are described as hard and dense, as they must have been for the walls to have still protruded some 2 meters above the terrain as late as the Han and Wei periods,9 with the outer walls being erected upon a carefully leveled and formed inner core.
More recent excavations have uncovered the interior enclosure shown on the diagram as well as a self-contained palace area slightly south of its midpoint and two other segregated areas. The royal quarters were set off in a roughly square enclosure measuring about 210 meters east to west and 230 meters north to south, creating a segmented area of somewhat over 40,000 square meters that contains evidence of palace foundations.10 As the wall averages only 3 meters in width, it clearly was more symbolic than functional in intent.
The royal quarters were initially surrounded by fortifications that assumed a fundamentally rectangular shape that extended 1,100 meters from north to south and 740 meters from east to west. Far narrower than the walls of the greater enclosure, they average 6 to 7 meters in width and were erected upon a shallow, meter-deep foundation. The freestanding portions retain a remnant height of only 0.2 to 0.6 meter, but they were certainly higher in antiquity. Constructed before the outer fortifications subsumed them, they demark the approximately 800,000 square meters that constituted the first Shang enclave at this site. However, within decades the highly dynamic Shang more than doubled the city’s bounded area to 1,900,000 square meters, even building right out over their bronze workshops. The final configuration thus assumed the palace, inner city (ch’eng), and outer confines or suburbs (kuo) widely discussed in late Chou and later literature as the highly idealized form of traditional Chinese capitals. As a result of this evolutionary expansion, Yen-shih is sometimes said to be the first to manifest it.11
Two walled enclosures, each with dimensions somewhat more than half the royal city, have also been discovered within the greater fortifications. Evidence suggests they may have been military barracks or storage areas for weapons and provisions. An 800-meter-long underground ditch roughly 2 meters wide, ingeniously constructed from wood and stone and linked to all the buildings in the palace compound, runs from the royal quarters out below the city’s walls. Originally characterized as a drainage ditch, its connection to a large artificial pool some 128 meters in length, 20 meters wide, and 2 meters deep suggests it was designed to provide water rather than simply drain it away. (This pool is the earliest artificially constructed urban reservoir yet discovered in China.)12
The inner walls, roads, gates, and artificial pond, being highly regularized and expertly constructed out of tamped earth, clearly show the systematic execution of a well-crafted strategic design that emphasized structural organization and active defense. Additional features such as two gatehouses that could control access to the entire settlement and the interior quarters, coupled with the narrowness of the gates, emphasize Yen-shih’s martial character, one highly appropriate for a new power center that had to be forcefully imposed in the enemy’s very core.
The citadel’s outer walls, which were erected within a few decades of the inner enclosure, and most of its functional remains are said to date to the fourth Erh-li-t’ou and lower Erh-li-kang phases, evidence that Yen-shih was not an ordinary site that gradually evolved over time, but was deliberately erected.13 Conversely, although austere, the site is too palatial to have simply been a military bastion, suggesting it must have been employed for some decades as the first capital (known as Hsi Po) before royal power was shifted back to the new capital at Cheng-chou. 14 Key radiocarbon dates for artifacts from the site fall just about the time traditionally claimed for the Shang’s ascension of 1600 and the astronomically derived year of 1588.15 Late geographical and eclectic writings such as the Lü-shih Ch’un-ch’iu and Ch’un-ch’iu Fan-lu preserve abundant, albeit much later, textual evidence that suggests King T’ang established his first capital of Hsi Po in this area, perhaps even roughly at this site.16
Apart from being located in the Hsia heartland, the site could also control vital passes to the east and west and project power to the troublesome northwest, where the Shang would continually experience severe challenges. Armies going east out through the Hei-shih and Huchien passes would move directly into the Hua-pei plains; marching westward through Han-ku Pass and T’ung Pass, they would cross into the critical Kuan-chung area; advancing northward, they would enter Chin and Shaanxi after fording the Yellow River; and going south on campaign through Yin-ch’üeh, they would move into the Ying and Ju river basins. From this single location all the military campaigns spoken of in Mencius and other writings, as well as any to the west and northwest, would have been feasible.17
Thus, the identification of Yen-shih as T’ang’s first capital, in conjunction with Erh-li-t’ou being the last capital of the Hsia, seems well founded, particularly if Ta-shih-ku is considered to have been Po after being wrested from the Hsia or predynastic Cheng-chou fulfilled that role. Moreover, the discovery of wheel tracks on an interior road, though no longer evidence of the very first use, shows that by 1600 BCE the Shang was already employing primitive vehicles that would have provided considerably greater earthmoving capability than laboriously toted baskets. Molds, bronze fragments, charcoal, and crucibles indicate that a significant bronze workshop was operated on the site even though vessel production continued at Erh-li-t’ou. However, the bronzes and other items lack inscriptions and are generally simpler than those subsequently identified with the Shang, befitting an earlier, more martial site. The few tools and agricultural implements recovered indicate that rather than being a major production or economic center, Yen-shih depended upon the greater Shang domain for provisions and logistical support.
The now famous Shang capital at Cheng-chou, a city near the Luo River in Henan province in the middle lower course of the Yellow River area, has also been the subject of numerous excavations since its discovery. 18 Before Yen-shih was identified and explored, many scholars vociferously argued that Cheng-chou had to be the first Shang capital of Po, others that it was Ao or even a Hsia capital.19 However, in contrast to Yen-shih’s comparative austerity, the extensiveness of its bronze and other production facilities, numerous opulent bronze cauldrons recovered, the stage of cultural and artistic development, and similar factors clearly indicate a more politically and economically developed stage. Based on radiocarbon datings and an analysis of the underlying layers, the walls at Yen-shih were probably constructed some sixty to eighty years before those at Cheng-chou, vital evidence that the former was probably the Shang’s initial postconquest capital.20 Thus, even though many analysts still argue that both were initial capitals and should be identified with King T’ang, the defining Erh-li-kang manifestation of Cheng-chou may well have been erected as a new palatial and ritual center and assumed the functions of the Shang capital under Ta Keng.21 Claims that Cheng-chou was most likely the capital of Ao, mentioned in numerous traditional texts such as the Shih Chi, Bamboo Annals, and Shang Shu associated with Chung Ting, are certainly balked by Cheng-chou’s lengthy occupation, the hundred years far exceeding the brief reigns associated with Ao.22
Exploratory excavations indicate that Cheng-chou was a vast site whose city walls encompassed some 25 square kilometers within their total length of 6,960 meters. Situated roughly along the compass directions on a height between two minor rivers, the city obviously functioned as a royal administrative center as well as the dwelling area for important members of the populace. Evidence of extensive royal quarters in the northeast portion of the city, apparently protected in part by an internal moat, has been uncovered, and several large production sites for bronze, bone, and ceramic products have been identified in virtually every direction outside the city.23 In both quantity and quality the artifacts unearthed at each site greatly exceed the limited amount recovered from Yen-shih and include rather specialized items such as wine vessels. The inscribed bronzes preserve evidence of the development of written language, and the ritual cauldrons that continue to be recovered, some as large as a meter high and weighing hundreds of pounds, are so magnificent as to be identifiable only with the royal household. They provide incontrovertible evidence that the markedly advanced cultural site of Cheng-chou, which obviously traded with other areas, must have been the Shang capital during its occupation.24
The originally massive and unusually high walls apparently towered some 10 meters over the outside terrain and had a formidable width of 20 meters at the base but just 5 meters at the top. (Excavated portions actually run from 1 to 9 meters high and 4.8 to 22.4 meters wide, with the corners being about 30 meters thick.) More specifically, the eastern wall is 1,700 meters long, the western 1,870 meters, the southern 1,700 meters, and the northern 1,790 meters, with openings for gates in all but the eastern wall. Generally constructed over an unusually narrow excavated foundation of only 2.5 meters in width, but with the usual downward trapezoidal shape and pounded, leveled ground, the fortifications actually consist of a 10-meter-wide core wall to which protective waist walls of some 5 meters in width were appended on either side.
Constructed in 3.8-meter-long sections framed by wooden boards between 2.5 and 3.3 meters long and 0.15 to 0.30 meter high, the layers are clearly defined, well leveled, and generally uniform, being between 8 and 10 centimeters thick but occasionally dipping to as little as 3 centimeters and bulging up to 20. However, little variation is evident in their composition, all the layers being composed largely from a mixture of sticky red soil and grayish red sand pounded to maximum hardness. The tools employed, bundled rods roughly 2 to 4 centimeters in diameter, left permanent impressions up to 2 centimeters deep on the surface of each layer. The two external protective walls were similarly pounded, and the outer one was coated with a layer of protective pebbles, presumably to forestall erosion by falling rain and perhaps buttress it against floodwaters.25
More recent excavations have discovered remnants of another 5,000 meters of substantial walls that externally encircled the southern portion and an external ditch or moat. Discernible wall heights presently range from 1.2 to 2.3 meters, with widths in the various sections reported of 12, 17, and an expansive 25 meters, all constructed on an excavated foundation trench.26 In the northeast corner of the city, evidence of an even earlier 100-meter-long wall some 8 meters in width has also been reported. Together with artifacts from predynastic Shang and other cultures, it indicates that Cheng-chou was an important location even before it became a Shang capital, perhaps King T’ang’s initial city of Po.27
Assessments of late Hsia and early Shang sites generally differ in how they identify the fortified cities of Yen-shih and Cheng-chou rather than in how they interpret the basic archaeological data. Whether Yen-shih was just a secondary capital or the first, primary capital of the Shang, though historically important, is somewhat less significant for military history than its definitive martial character. Moreover, when coupled with the evidence that Erh-li-t’ou is almost certainly Chen-hsün, the last capital of the Hsia, Yen-shih’s strategic function becomes obvious and virtually requires that Cheng-chou be viewed as the second Shang capital. This does not deny a martial function to Cheng-chou as well, especially since its role as a weapons production center has been clearly established by the discovery of numerous molds for casting bronze axes, halberds, and arrowheads in workshop ruins.
Being just six kilometers away from Erh-li-t’ou, Yen-shih must have been constructed to project power and consolidate control, to nullify the influence of the old Hsia center at Chen-hsün, which continued to flourish. In contrast, although Cheng-chou certainly retained an important military role in controlling people to the east and south, it has been characterized as a relatively open site that lacked the internal controls and defensive measures found at Yen-shih, and thus more of a ritual and administrative center.28 The walls, though massive, are also marked by gradually sloping protective waist walls, hardly a desirable feature against assaults.29
A simple comparison of the relative might of each of these sites will aid in differentiating their character and functions. As already noted, Yen-shih lacked the productive basis for independent existence and had only minimal workshops and agricultural fields. In contrast, Cheng-chou included not only numerous large, specialized workshops, but also loosely integrated agricultural fields in nearby areas, accounting for the recovery of farming tools. The area enclosed by Cheng-chou’s walls totals some 3 million square meters, whereas the original compound at Yen-shih, a citadel of 740 meters by 1,100 meters before expansion, totaled about 800,000 or slightly more than a quarter of Cheng-chou’s size.
Whatever the dating of their walls, both sites were occupied over long periods, suggesting that Cheng-chou, much as Ta-shih-ku, may have been a pre-Erh-li-kang site initially occupied by the predynastic Shang as they expanded out of their core area in Chi-chung and undertook to subjugate the peoples in the nearby Luo-tung area, neutralize the Eastern Yi, and vanquish the Hsia’s last ally, the K’un-wu, over the four years chronicled by the Bamboo Annals.30 After overcoming the Hsia, the Shang required a new military citadel at Yen-shih to consolidate and impose its power in the Hsia heartland even as it continued to maintain its presence and project power in the east through some eleven additional campaigns.31 Having completely subjugated the various Hsia groups, ensured the area’s security with a strong bastion, and expropriated the Hsia’s wealth, Cheng-chou was apparently fortified on a grand scale to become an opulent administrative and ritual center for the newly aggrandized rulers, Yen-shih functioning as secondary capital and strongpoint.32
Until recently many analysts believed that all the archaeological evidence pointed to a major disjuncture between the third and fourth periods of Erh-li-t’ou’s occupation, justifying the traditional accounts of the Shang emerging through military conquest rather than simple evolutionary displacement.33 The four distinct layers clearly visible at Erh-li-t’ou have generally been correlated with four stages of activity: establishment, expansion, flourishing, and decline. The first three display evolutionary continuity, but the fourth, in addition to the apparent abandonment of palace buildings and other evidence of contraction, suddenly shows an admixture of lower Ch’i-t’an Shang culture, very similar to that characterizing the first of Yen-shih’s three periods.34
However, further excavations and the publication of additional reports indicate not only that Erh-li-t’ou continued to flourish in phase 4 (now dated as 1564 to 1521), but also that a new palace surrounded by tamped earth walls and other buildings were constructed.35 In addition, the site seems to have continued as a major bronze production center until Erh-li-t’ou suddenly declined to the level of an ordinary village late in the sixteenth century, when much of the populace may have been forced to migrate to Cheng-chou.36
No evidence of a massive disaster or destruction by fire has yet been uncovered at Erh-li-t’ou. This coheres well with traditional, highly idealized accounts that the Shang targeted the chief miscreant, the one man, and that the actual fighting occurred away from the capital. Somewhat surprisingly, despite the importance of fortifications in Lungshan cultures and their derivatives, rather than enclosing the city with protective walls, the Hsia had opted to rely on a fortified palace at Erh-li-t’ou. 37 Perhaps the founders thought that the sheltering hills surrounding them and their location in the center of the Loyang basin would provide an adequate defense. In marked contrast, Yen-shih’s initial occupants immediately opted to construct the massive walls typical of Shang fortified cities even though Yen-shih equally enjoyed significant advantages of terrain, including Mt. Mang to the north and the Luo River to the south.
The unprotected city, with most of its population intact, was probably compelled to capitulate. Moreover, having refrained from vengefully destroying it, the Shang could then exploit Erh-li-t’ou’s productive capabilities with, as attested by the paucity of Erh-li-kang artifacts, only a minimal presence. However, they opted to impose significant control from the security of the nearby bastion of Yen-shih, perhaps because they felt Erh-li-t’ou to be indefensible given its lack of walls. Yen-shih’s initial period thus not only coincides with Erh-li-t’ou’s fourth period, but also displays some Hsia elements intermixed among predominantly Shang artifacts.38
A number of aspects clearly indicate an ongoing evolution in the capabilities and practices of material culture, particularly in the technology and styles of bronze casting, one of the hallmarks of Shang civilization. Although some magnificent ritual bronze vessels have been recovered, many of the artifacts discovered at Erh-li-t’ou are comparatively simple and, having been produced from an extremely high copper alloy, thin walled. Even allowing for the continuity that should be expected given that Erh-li-t’ou continued as a ritual production center for decades, the bronze vessels and implements found at Yen-shih early on show evidence of an advanced mixture that mingled tin and a large lead component while reducing the copper to approximately 80 percent, allowing larger, heavier, thicker objects (including weapons), marked by better curves and more refined surfaces, to be cast from multiplepart molds.
Significant differences in burial practices and ceramics can also be discerned at the two sites, even though ceramics were already being de-emphasized with a shift in focus to bronze objects. A second definitive cultural aspect, the use of symbols or early writing, also advanced, if only in terms of the number of symbols appearing on the artifacts. However, despite claims of essential continuity the numbers remain minuscule, hardly the language of the oracle bones identified with Anyang.
The palace quarters similarly show distinctive Hsia and Shang placements. The essentially self-fortified compound at Erh-li-t’ou, which extends some 100 meters on a side, contains a highly systematized central hall with a surrounding portico built in the third period of fluorescence that faces south and, apart from being slightly offset toward the east, is fundamentally centered and thus surrounded by industries and living quarters. However, the palace compound at Yen-shih, which apparently also faced south, was located in the southern portion of the fortified city, slightly offset to the west, between two other walled compounds that seem to have been reserved for defensive functions.39
Yen-shih thus resembles Cheng-chou and other well-known Shang cities and fortified towns in which the markets, living quarters, and industries are usually found “behind” or to the north of the inner compound or administrative center, within the enclosure created by the outer walls. The construction techniques employed at Yen-shih, including the incorporation of a drainage ditch with a very slight but sufficient pitch, the solidity of the walls and foundations, and the overall systematic organization of the city thus all manifest perceptible differences and incremental advances over Erh-li-t’ou practices.40