Chapter 6 - Violence and Shang Civilization

Violence is surely one of the most contested terms in the English lexicon. The reason is that violence sits at the volatile intersection of morality and power, its interpretation concerns positioned judgments about transgression and harm, and its legitimacy and visibility are historically constructed. While violence has traditionally been seen as antithetical to history (Arendt 1970) – the breakdown of order and reversal of developmental sequences – it could instead be argued that both harmful practices and the discursive efforts to vilify or justify, commemorate or ignore, ban or normalize them, reveal a great deal about a time and place. Whether talking about collateral damage from drone strikes, the trans-Atlantic slave trade, Roman circuses or Shang human sacrifice, moral economies of violence are deployed to distance, justify, or glorify what in other contexts would have been impermissible, horrific or unthinkable. These regimes or institutions of violence, moreover, in each case fit into larger complexes of politics, hierarchy, morality and community. The challenge for a twenty-first-century interlocutor into the world of late second millennium BCE North China then, is to understand how vastly different moral economies of violence and order, such as those of the Shang, could have functioned, and what they can tell us about this ancient Chinese kingdom.

Civilization, as discussed in previous chapters, is another highly charged term. While Western political theory has traditionally seen civilization as either begetting violence (e.g. Rouseau) or mitigating it (e.g. Hobbes), more recent approaches (Foucault 1995, Campbell 2014b) have come to recognize the mutable nature of violence, the intimate entanglements of its particular forms with the specific constellations of normativity, hierarchy, power and technology from which civilizations are made.

One of the most influential theorists of “civilization” and its relationship to violence is Norbert Elias. Elias (1994) saw civilization or the “civilizing process” as having three central characteristics: 1) monopolization of the means of violence, 2) the removal of physical violence to behind the scenes, and 3) lengthening chains of interdependence between people – which he claimed, in combination, led to increasingly pacified and civil societies. The historical and logical relationship between these factors was posited to be that the initial process of internal pacification brought about a monopolization of violence, which, in turn, allowed the creation of increasingly extensive webs of interdependence among people who no longer needed fear their neighbors. This then lead to reinforced affective connections. Moreover, the restrictions on physical violence were then seen by Elias to lead to more restraint in expression of affect more generally, and ultimately, to a rationality that is based on foresight and self-restraint. As some authors have noted, however, (Burkitt 1996; Bauman 1989), there is a tendency in Elias’ work to conflate the etiological myths of the West with a general theory of civilization.¹ Indeed, some of the worst instances of violence in the last century could be seen as consequences of monopolies of violence, the removal of violence behind the scenes, and the lengthening chains of interdependence, which become so tenuous and abstract as to create what Burkitt (1996) calls moral invisibility. Moreover, if, as Sontag (2003) has suggested, “being a spectator of calamities taking place in another country is a quintessential modern experience,” our modern “civilizing process” may not imply a reduction or removal of violence from view so much as its transformation into spectacle at a safe distance, virtualization, and commodification as entertainment – not to mention the advent of invisible, structural violences² that are now global in extent. While it may be tempting to dismiss the large-scale human sacrifice and endemic warfare of the Anyang period as Bronze Age savagery, I would argue that we are not separated from this violence by some radical socio-evolutionary break, but rather that the networks and connections between social agents have grown so long and abstract that violence that was once immanent now appears mostly virtual, and the relationships between violence and socio-political order are now obscured by the complexity of our increasingly compartmentalized knowledgeability. Indeed, as even Elias himself suggests, violence does not disappear, but rather is transformed into economic and other forms (Elias 1994) with internal pacification (which, in any event, is never perfect). Shang practices of violence then, rather than being mere exotic historical spectacle, can be better understood as part of civilizing, if ambivalent, processes. Thus, rather than see “civilization,” or “civilizations,” as static, elite cultural contexts in which early states were embedded, understanding civilization as a process and violence as its accompanying shadow, allows us to see at once the dynamic interconnections between institutions and practices of authority, the production of normativity, and the politics of harm.

From the point of view of studying early complex polities like that of Shang Anyang, a critical use of Elias’ “civilizing process” suggests some interesting lines of analysis. For instance, to what degree and through what mechanisms was violence monopolized by the Shang kings at Anyang? How was the work of internal pacification figured in their practice and discourse? Moreover, if we modify Elias’ first characteristic (borrowed from Weber) to read “monopoly over legitimate violence” then we must also take into account the social or moral economy through which monopolized violence was sanctioned.³ The second characteristic, the removal of violence behind the scenes, raises the question of what Elias would call the sociogenesis of personality structures with respect to violence in the wake of internal pacification. Given that internal pacification is a process rather than an absolute state, and may take different forms in different societies, the question is raised concerning how institutions of pacification and their attendant symbolic economies of legitimacy are figured in the inculcation of dispositions toward violence and its socio-political uses. Moreover, considering the prevalence of spectacular violence in many early complex polities (e.g. Roman circuses, Aztec sacrifices, Neo-Assyrian massacres, etc.), as well as our own society’s predilection for violent spectacle in the form of movies, videogames, and contact sports (or even the nightly litany of distant suffering on the evening news), the idea that violence is removed behind the scenes during processes of internal pacification may be too simplistic. It might be more useful to note that as socio-political entities get larger, more integrated and internally diversified, direct participation in inter-subjective violence tends to become specialized (police, professional criminals, soldiers, terrorists), distanced, and/or bracketed into “exceptional” social fields, even while violent and competitive predilections find different social outlets, and take on different structural forms. The particulars of the specialization, forms and arenas of legitimate violence are thus potentially revealing lines of socio-political analysis. The third characteristic, the lengthening of chains of interdependence, gives rise to questions concerning the development, nature and extent of the Shang community. Moreover, as expanding networks of mutual dependence are supposed to act as a structural constraint on inter-personal violence, the issue of violence and social identity is raised. Against whom, and in what circumstances is violence permitted or restricted? How is interdependence and mutual obligation practically and discursively structured in the Shang? Since interdependence occurs within hierarchical structures, and across asymmetries of power, how do identity, community, power and legitimate violence combine in the Shang social order?

If Elias is optimistic about the positive value of the civilizing process, other authors such as Foucault (1995), and, more recently, Agamben (1998), have seen much the same process in markedly negative terms. Though Foucault’s vague and all-pervasive notion of power has been abundantly critiqued (e.g. Giddens 1982, Wrong 1988), his analysis of the prison in Discipline and Punish effectively turns Elias’ civilizing process on its head, seeing internal pacification and the removal of violence behind the scenes, in terms of pervasive social control and discipline, enacted on the docile bodies of social subjects. If the creation of pacified bodies and minds receptive to authority structures is part of the process that attends increasing socio-political integration and scale, then the question arises of what early regimes of social discipline looked like, and how they related to hierarchical orders.

Agamben (1998), for his part, argues that violence is fundamental to political relations, and has underlain “sovereign” power since antiquity. Indeed, the most crucial aspect of sovereign power for Agamben is its alienating ability to create bare, killable life. Rather than seeing state monopolization of violence as a positive development, Agamben sees grave danger in the increasing extension of state “biopower” over the lives of its citizens. Indeed, the flip side of universal citizenship, and increased integration into the body politic, seen in modern nation states, is the potential power of the state to alienate through the suspension of rights. For early complex polities like the Shang, Agamben’s thesis raises questions concerning the creation of non-persons, or more generally, the production of hierarchies of being, and their relationship to unequal exposure to harm. If Agamben is correct in positing the ability to exclude “citizens” from the realm of internal pacification, to banish them literally, or figuratively, beyond the protection of law and community, and expose them to violence as a fundamental aspect of sovereign power, then the particulars of this social reduction, and its relationship to the social economy of being and power, become critical issues for understanding the nature of polities.

One of the most striking features of Shang Anyang is the scale of its violence. From the thousands of sacrificial victims in the royal cemetery and palace-temple area, the tens of thousands of functional and symbolic weapons buried with the dead in Shang tombs, to the Shang kings’ divinatory focus on the spilling of blood in ritual, warfare, or the hunt – it is difficult to escape an overwhelming impression of violence. Indeed, the later Zhou accounts of the depravity and cruelty of the last Shang kings form a ready-made explanation for those inclined to be credulous of much later historical accounts written by the victors. For those more inclined to evolutionary narratives, Shang savagery fits with popular ideas concerning the founding violence of early civilizations. A third choice, and by far the most common, is to pass over Shang violence as incidental, and focus on their brilliant achievements in art and technology. On closer inspection, however, while the third option is a non-explanation, the first two options are also demonstrably problematic – it was the virtuous Wu Ding, not the evil Di Xin who presided over the greatest episode of human sacrifice known to Eurasian prehistory, and, as noted in earlier chapters, Anyang and its ancestral-sacrificial complex were new and elaborated variations on a much older Central Plains Civilizational theme. While we may never know with certainty the historical reasons behind the dramatic increase in evidence for war and sacrifice at Anyang, the contemporaneous importance afforded to these practices suggest that we cannot adequately understand this time and place without reconstructing the social economy of its violence. In other words – what were the cultural logics that made these inhuman practices human?

War, Sacrifice and the Polity

Returning to the questions with which we began this chapter, a crucial issue in understanding the polity centered at Anyang and its local world concerns the issue of internal pacification. In Chapter 4 we have already characterized the King’s coercive network as relying on both direct and indirect networks of capital, in a political landscape where the practical and material resources of violence were dispersed among a myriad of local rulers and lineage heads. From the oracle-bone inscriptions we get a sense of near-constant raiding, campaigns and rebellions – at best a dynamic balance requiring constant maintenance. If internal pacification generally appears more as an endless process rather than a state of affairs for the Shang kings at Anyang, it does not mean that this was equally true of all parts of the polity, or that nothing changed over this approximately 200-year period. As mentioned earlier in Chapter 4, there was a central zone of direct control – the royal demesne – over which a relatively high level of internal pacification was maintained through touring, hunting, gifting and sacrifice, as well as more routine social, political and economic networks – ruptured only by the occasional rebellion. It is in the indirect zone of allies, clients and buffer polities, and beyond their shifting borders, that the ruling of the four quarters required the constant work of campaigning, alliance-making, and, where possible, subjugation and incorporation. There were, moreover, some trends in the development of the King’s technologies of domestication over the Anyang period.

In the oracle-bone inscriptions, the king’s use of coercive power could be indirect, as when missions were delegated to subordinates, direct, as when the King himself participated, or a combination of both as in the period I oracle-bone examples below.

癸亥卜,貞: 令倉征壴. (6)

Guihai day cracked, Bin divined: (We should) order Xi, Lord of Cang to launch an expedition against Zhu.

辛未卜,爭貞: 帚好其比沚伐巴方, 王自東伐戎(陷)于帚好立. (6480)

Cracked on Xinwei day, Zheng divining: Fu Hao should perhaps join with Zhi Guo to attack the Ba Fang, the King from East Shen(?), will attack Rong trapping (them) at Fu Hao’s position.⁴

貞: 王而白龜比伐□方.

Divined: it should be Xu Bo Gui that the King joins to attack the … Fang.

This latter example refers to at least three different participating forces (and possibly four, if the King joins Gui the Bo-chief of Shu), with allies, subordinates and the King himself joining the endeavor, suggesting the frequently complicated strategic and logistic calculus involved in major Shang military action. Indeed, it is difficult to be certain that even when there is only a divination about a subordinate being ordered to undertake a coercive endeavor, that the king was not also involved, since it is rarely possible to reconstruct the total context. On the other hand, while it is highly likely that allies, and possibly some subordinates, frequently acted on their own initiative and to their own ends, since the oracle-bones are mostly records of royal divinations, they rarely provide information concerning events not directly concerning the King. Nevertheless, in early period inscriptions, and especially in the reign of Wu Ding, there are some divinations concerning the actions of allies which suggest the King’s limited control of military endeavors within the four quarters.

6623: 癸卯卜, 貞:甫 (乎) 令 沚方. 七月.

Guimao cracked, Bin tested: it is Bu (who we should call upon) to order Zhi to harm the Qiang Fang.

6995: 沚其戎。

Zhi will perhaps make war upon X.

6377: 貞戉弗其伐〔方〕.

Tested: Yue will perhaps not attack the Gong [Fang].

4279 (1)貞戉其乎來.

Tested: (It should be) Yue perhaps (whom we) call upon to come.

4280 (1)戉其來.

Yue will perhaps come.

戉不其來

Yue will perhaps not come.

In example (4) we can see the tortuous route through which the king might exercise authority, divining about whether or not to call upon someone to order a third party to attack an enemy. Example (5) shows that the King is uncertain as to whether or not an ally will attack, while the use of the modal particle qi 其⁵ suggests that this attack was not something the King desired. Example (6), on the other hand, shows the opposite situation with an ally possibly not carrying out an attack against a Shang enemy. As examples (7) and (8) show, allies might not necessarily be counted on to come (to court?) if called, suggesting the practical limitations undercutting the King’s discursive claims of universal authority.

Indeed, one of the main differences between the early and later period military inscriptions is the relative frequency of divinations concerning allies and allying. In period V military oracles, in contrast with period I inscriptions, the majority of divinations concerning war seem to concern campaigns in which the King is directly participating, and individual allies are mentioned more rarely. Indeed, in some period V campaign inscriptions, other political actors are mentioned only as categories, as in the following example.

乙丑王卜, 貞, 今巫九: 余作遣告、田, 冊方、方、羞方、方. 余其比、田, 伐四邦方. (36528 reverse)

Cracked and tested by the King on Yichou day, (meaning of phrase unclear):⁶ I made zun and qian sacrifices announcing to the hou and dian-lords to pierce the Qie(?) Fang, Qiang Fang, the Xiu Fang and the Sui(?) Fang. I will join with the hou and dian-lords in X attacking these four enemy polities.⁷

丁卯王卜, 貞, 今巫九: 余其比多田于多白, 正盂方白炎. 衣翌日步, 亡尤. 自上下于(祭)示余受祐. 不〔〕. 告于茲大邑商, 亡才. 〔王曰〕: 引吉. 才十月, 遘大丁翌. (36511)

Cracked and tested on Dingmao day by the King, (meaning of phrase unclear): I will join with the many dian and bo-lords to mount an expedition against Dan, the bo-lord of Yu. It should be on the day of the yi and yi-rituals that we march, there will be no trouble. From the upper and lower spirits through the sacrificial altars I will receive divine aid. There will certainly be no harm [or disasters]. Announce this to this Great Settlement Shang, there will be no harm in the omen-cracks.⁸ The King examining the cracks said: extended auspiciousness. (Recorded) in the tenth month, upon the day of Da Ding’s Yi ritual.

The first example tests the auspiciousness of making an announcement to the many hou and dian-lords of an attack on the Qie(?), Qiang, Xiu and Sui(?) Fang, and that the King will be joining forces with them against those four polities. In the second example, we again see the King joining with the many bo and dian-lords, this time against the rebellious Dan, bo-lord of Yu. This statement is followed by a divination that the King will receive divine blessing, and that the campaign will be announced to the Great Settlement Shang. The final phrase records the time in terms of place in the cycle of royal ancestor worship.⁹ In these examples we can see a late period trend toward consolidation and standardization that has long been noted in royal divinatory and sacrificial practices, but is apparently symptomatic of wider changes. What is not entirely clear is whether joining forces with the many bo, hou and dian-lords as categories, rather than specific individuals, bespeaks mobilizations of a greater scale,¹⁰ or simply a change in the conventions of royal divination. Nevertheless, if an analogy may be drawn from divination about sacrifice and the reformation of ritual in periods III–V, a systematization of (at least) the discursive practice of war is in keeping with broader changes in the organization of practices of authority in Anyang’s latter phases. Moreover, just as Anyang’s last kings structured time with the cycle of their ancestral sacrifice, so too the King’s campaigns stood as landmarks in the topography of social memory.

癸未王卜貞: 旬亡. 王來正人方. (36500)

Cracked and tested by the King on Guiwei day: the week will have no disasters. (During) the King’s expedition against the Ren Fang.

Not only do we frequently see divinations recording time in terms of the month and the King’s ritual cycle in late period inscriptions, but major campaigns also marked the passage of time as we can see in the period V inscription above. Moreover, royal divinations were not the only place where this pattern can be seen – the bronze inscriptions that begin to appear at the end of the Anyang period, although usually marking time with the King’s ritual cycle, sometimes also refer to the King’s campaigns, as in the following example,

丁巳, 王省夒X. 王易小臣俞夒貝. 唯王來正人方. 隹王十祀又五, 肜日. (小臣俞尊 The Minor Retainer Yu Zun)

On Dingsi day the King inspected Nao X (a place). The King presented Minor Retainer Yu with Nao cowries. It was during the King’s campaign against the Ren Fang. It was the King’s fifteenth ritual cycle on the day of the rong-ritual.

The impression given by Shang practices of authority in the second half of the Anyang period in general, is thus one of at least discursive consolidation, incorporation and systematization. Allies and subordinates tended to be referred to in corporate rather than individual terms, sacrificial practices were routinized in a cycle, and oracle-bone divinatory inscriptions came to look more like reflections of a structured system for vetting royal decisions than the ad-hoc technology of ancestral communication and negotiation they once were. Moreover, while Wu Ding appears to preside over a larger territory and network of alliances, his ancestral sacrifice was distinguished from that of other elites more by its monumental scale than qualitative difference. Kings Di Yi and Di Xin on the other hand, seem to have ruled a smaller domain with fewer allies,¹¹ but with the King’s authority systemically differentiated with the near monopolization of royal/state divination,¹² and the structuring of time with the routinized cycle of royal sacrifice, and punctuated by the King’s expeditions.

Nevertheless, the basic structure and cultural logic of war seems to have remained more or less the same over the period covered by the oracle-bone inscriptions. Sacrificial offerings to the ancestors and spirits were undertaken to assure their approval and assistance both before¹³ and during campaigns, forces were levied from lineages under the King’s direct control and allied forces were mustered. Campaigns could extend over the course of a year or more, and see the King traveling, sacrificing, hunting and bestowing gifts¹⁴ as he moved across the landscape at the head of his forces – suggesting that the King’s exercise of coercive power was largely personal, direct and intense, if sporadic and unsystematic, even while tied to a broader panoply of diffuse practices of authority.

One example of a late period campaign is that against Dan the bo-lord of Yu which took place over at least five months in the King’s ninth and tenth ritual cycle. As we saw in example (50) in Chapter 4, the first divination records the announcement that Yu raised forces and was attacking Gao. Example (10) above then shows the King sacrificing to the ancestors, joining with allied lords, and announcing to the capital the commencement of the campaign. While we do not know the precise details of the campaign or its results, we do know that over the months of the campaign the King hunted, divined about the weekly fortune and continued to perform the cycle of sacrifices to his ancestors.

…隻白兕, 于… 才二月. 隹王十祀,日. 王來正盂方白□. (37398)

… catch white buffalo, nai-offer to … in the second month, during the King’s tenth ritual cycle, on the day of the rong-ritual to the sun. During the King’s expedition against … the bo-lord of Yu.

In other campaigns, records of the outcome are preserved.

…小臣比伐, 危美…人二十人四, 而千五百, 百…丙, 車二丙, 百八十三, 五十, 矢…用白于大乙, 用白印… 于祖乙, 用美于祖丁, 甘亭, 易 … (36481)

… minor retainer X joined with (someone) to attack, capturing Mei of Wei … people, 24 people; scalps(?), 1,570; captive women(?)¹⁵ 100 … bing; chariots, 2; Y 183 armor¹⁶ 50, arrows … use You, bo-lord of Wen(?) to Da Yi; use Yin, bo-lord of Z … the captured women(?) to ancestor Yi; use Mei (in an offering) to Ancestor Ding; sai-lodges,¹⁷ 20 buildings, reward …

Though there remain many uncertainties in the decipherment of these inscriptions, we can nonetheless see what was at stake in Shang inter-community violence. For the losers there was subjugation, servitude and perhaps death, either on the battlefield or on the altars of the victor. For the King, the campaign served to make tangible his authority all along the route of his campaign as he observed the cycle of ancestral ritual, domesticated the unruly spirits of the land through his hunting and sacrifice, and gathered the forces of those allied to him. Discursively, campaigns were figured as exercises in the King’s power to punish transgressors and domesticate the dangers that threatened the order created through the regular rhythm of royal sacrifice. They were also critical occasions for the maintenance and negotiation of allegiance, community and identity. Not only was service rewarded and disloyalty punished, with the fates of political agents and whole communities potentially decided, but the production of supralocal identity, or its dissolution, was also intimately tied to the strong affective practices of inter-community violence.¹⁸ As seen above, the forces of political actors are described in terms of their polity or leader, not as a dynastic army as Han and Warring States texts would have us believe. Nevertheless, the raising of large forces undoubtedly combined normally independent groups into supra-community units as we can see both in the King’s divinations about joining with allies, and the list of bo-lords 伯 captured along with Mei, bo-lord of Wei 危伯美 seen in example (14) above.

Thus, if war potentially brought together smaller, more basic, socio-political units, it was nevertheless these units that formed the basis for, and informed the basic logic of, inter-community violence. As argued in Chapter 5, the basic socio-political unit at Anyang (and presumably elsewhere in the Central Plains) was the zu-lineage. In addition, the correlation between weapons and burial status suggests an ideology of social violence beyond the royal court, while the widespread presence of weapons in tombs suggests the participation of relatively low-status people in this symbolic economy of violence. Based on a statistical analysis of Shang lineage cemeteries at Anyang (see Chapter 7), 20 percent of the tombs were equipped with weapons. While we can’t infer that 20 percent of the population participated in war (in fact, the percentage may have been higher), we do know that 20 percent of the burial population went to the next life armed and that mortuary symbols of violence were second in quantity and expense only to equipment for ancestral feasting.¹⁹ There was, moreover, a significant correlation (0.01, two-tailed, r = 0.6) between the number of weapons in a tomb and the size of the tomb, suggesting that there was a relationship between quantities of weapons and burial status even as jade blades, yue-axes, pole-sabers and chariots qualitatively distinguished the elite (see Chapter 7, Tang 2004). The importance of weapons in Anyang assemblages, their broad distribution and the care lavished on the making of the finest examples of them, all argue for the extraordinary importance of inter-community violence in the imagination and practice of Anyang society (see Figures 6.1–6.3).

Figure 6.1 Ge Dagger-Axe with Jade Blade, Bronze Backing and Turquoise Inlay

(after Institute of Archaeology 2005: 187)

Figure 6.2 Jade and Bronze Weapons from Anyang: (left) jade-bladed, bronze-socketed spear (after Institute of Archaeology 2005: 184); (right) Fu Hao’s yue-axe

(after Rawson 1996: 104, fig. 46)

Figure 6.3 Bronze Weapons from Anyang Tombs: (top left) spears (after Institute of Archaeology 1994: 313, fig. 173); (bottom left) ge dagger-axes (after Institute of Archaeology 1994: 309, fig. 170); (top middle) yue-axes (after Institute of Archaeology 1994: 312, fig. 172); (top right) pole saber (da dao) (after Institute of Archaeology 1994: 314, fig. 174); (bottom right) shield, helmet and chariot riding equipment

(after Institute of Archaeology 2003: 402, fig. 7-7)

If the oft-quoted Zuo Zhuan 左傳 passage that states “the great affairs of the polity are sacrifice and war” (隠公五年 Cheng Gong 13th year) was in some sense more true for the Anyang period than any other time in Chinese history, it would probably be more accurate to say that for the Shang people, sacrifice and war were the great affairs of the lineage, and operated materially, practically and discursively through its hierarchies of being – incorporating the living and the dead, the straight and the crooked, the pacified and the wild. At the same time, however, as argued in Chapter 5, the Shang polity formed a kind of super-lineage, and the king acted as a kind of lineage leader of the world.

The question then, is through what mechanisms were individual lineages integrated into the king’s practices of sacrifice and war, offering their capital, labor and blood up the hierarchy of Shang authority?²⁰ As argued above, part of the answer to this question resides in the king’s real or fictitious bonds of kinship with other important lineage leaders, creating a further ramification of the basic social order. In terms of social practices producing dispositions favorably orientated toward the king’s order, participation in the king’s punishment of rebels and enemies, or defense of Shang lands would have provided the opportunity for the creation of powerful affective bonds of supra-lineage identity. Participants in the capture and sacrifice of enemies, in a moral economy of divinely sanctioned punishment, would also have garnered ancestral honor and merit – to a degree equal to the social actors’ physical, and/or ideological involvement – even as sanctioned participation in the social death of the enemy simultaneously conferred a relative rise in the hierarchy of being, a position within the re-defined circle of the King’s community,²¹ a sense of awe at the terrible majesty of royal sacrifices, and, to the extent that captives could be identified with, perhaps also terror.

Moreover, as seen in Chapter 4, the sheer numbers of people levied for war (3,000–5,000, but sometimes exceeding 10,000), strongly suggests that ordinary lineage members made up the bulk of the Shang kings’ armies, even as the advent of the chariot in the Anyang period meant that Shang forces were divided between those who went to battle on foot, and those who rode to war. Chariots, then, as major symbols of elite status, functioned not only as mobile fighting platforms, but also as stages for the enactment of hierarchy on the battlefield – creating differential practices and experiences within a general participation in organized violence. Shang warfare was a pervasive structuring mechanism, a major “tournament of value” (Appadurai 1986) where “the dispositions of the central tokens of value” (21) were at stake. It was, moreover, a game played with human lives, for unequal material, symbolic and existential rewards, where both ancestral place and politico-religious authority could be decisively negotiated.

The Logic of Sacrifice

As noted above, war in the Shang context was embedded in a larger set of practices of pacification which prominently included sacrifice. Indeed, just as the approval of the ancestors and powers were necessary for victory, so too failure was seen in terms of the total structure of authority (see examples from Chapter 4).²² Moreover, not only was the approval of the ancestors sought, but successful campaigns ended with the sacrifice of captives – the ancestral capital of one lineage consumed for the sustenance of another. War in the Shang was, thus, not simply diplomacy by other means, but a set of fundamental practices of authority feeding the sacrificial economy and maintaining the king’s world order in a turbulent and dangerous spiritual and political landscape. Nevertheless, and to a degree even greater than war, the basic practices of Shang royal sacrifice changed dramatically over the course of the Anyang period with broad implications for the structural bases of royal authority.

Although many scholars have noticed the trend toward routinization and standardization in the royal cycle of five sacrifices that became increasingly prominent by oracle-bone period III (e.g. Chang 1987, Chen 1988, Keightley 1999, etc.), by examining occurrences of the ritual verbs gao 告 (to announce), yu 禦 (to exorcise/lustrate), and bin 賓 (to host) across the five periods of the oracle-bone inscriptions, I have sought to further illuminate these ritual developments and their structural ramifications for the Shang polity at Anyang.

Gao-announcing, as noted in Chapter 4, was also used by allies announcing enemy raids, as well as by the King in making declarations to the great settlement Shang (see example (10) above).²³ Ritually announcing to the ancestors and spirits of the land, thus formed an important tool of communication with the upper echelons of socio-physical power. They were, moreover, overwhelmingly directed toward royal ancestors or former ministers (see Table 6.1). This is in accordance with the general impression from period I and II inscriptions, that ritual, as intervention in the world, was predicated on a logic of real or fictive kin relations, a network of patronage and dependency that was an extension of, rather than mirroring, (as Durkhiem would have it), the social world.

Table 6.1 Gao-Announcing Recipients

Father	Ancestor	Mother	Ancestress	High Ancestor	Former Minister	Other
10 (15%)	18 (28%)	2 (3%)	4 (6%)	21 (32%)	1 (2%)	9 (14%)

Inscribed divinations about announcing to the ancestors appear to be a basically early phenomenon with period III–V examples becoming increasingly rare. While it is entirely possible that announcements to the ancestors continued in the second half of the Anyang period, inscribed divinations about whether or not to perform them decreased, perhaps as announcing rituals became a fixed (but unmentioned) part of the ritual cycle, or were rendered obsolete as the weekly ritual cycle essentially routinized communication with the ancestors.

Turning to yu-exorcism 禦 (Table 6.2), a similar pattern emerges, in that this ritual fades from the inscriptions after period II, disappearing entirely from the period V inscriptions. Exorcism, moreover, appears to be a broad heading for ritual action under which many different sacrifices might fall (with liao-burning 燎, mao-splitting 卯, chen-drowning 沉, and you-cutting sacrifices all associated with yu-exorcism).

Table 6.2 Yu-Exorcism Sacrifices

Mao 卯	Liao 燎	Ci	You	Chen 沉
5 (31%)	9 (56%)	4 (25%)	7 (44%)	1 (6%)

Exorcism was apparently performed both post-calamity, for relief of illness or ill fortune (as in example (19) below), and in a more routinized, proactive mode (as examples (20) and (21) show).

貞: 疾止禦于匕己. (40373)

Tested: (as for) the sick foot, perform an exorcism to Ancestress Ji.

乙亥卜, 貞: 乍大禦自上甲. (14860)

Cracked on Yihai day, Bin tested: Make a great exorcism (starting) from Shang Jia.

貞: 禦自唐, 大甲, 大丁, 祖乙百羌, 百. (300)

Tested: exorcise (starting) from Tang, (through) Da Jia, Da Ding, (and) Ancestor Yi (using) one hundred qiang-captives, one hundred specially-reared sheep.

貞: 禦, 牛三百.

Tested: exorcise, it should be cattle, three hundred (that we use).

As these examples also show, exorcisms could be large or small, and target single entities, or groups of ancestors or powers. Exorcisms sometimes involved diverse offerings as in example (21), although human beings, cattle and specially-reared sheep were the most common offerings (Table 6.3).

Table 6.3 Yu-Exorcism Offerings

Humans	Cattle	Sheep/Goats	Specially-Reared Cattle	Specially-Reared Sheep	Pig	Dog	Millet Liquor
9 (17%)	10 (19%)	6 (12%)	3 (6%)	13 (25%)	7 (13%)	1 (2%)	3 (6%)
p = 56.6	p = 75.9	p = 4	p = 7.7	p = 111.9	p = 9.3	p = 1	p = 67.7

The fact that exorcism disappears from late period inscriptions is in keeping with the reduction of ad-hoc ritual in the case of the more specifically instrumental curative exorcisms, while the collective, preventative great exorcisms may have been rendered redundant with the advent of the five-ritual cycle. The majority of yu-exorcisms were directed toward the nearer, more anthropomorphic ancestors, rather than ancestralized powers of nature (see Table 6.4), showing, as Keightley has argued, that both the source and resolution of ill fortune was seen as lying largely with the nearer ancestors, as they served as the link between the living community and the unseen powers of the world.

Table 6.4 Yu-Exorcism Targets

Father	Ancestor	Mother	Ancestress	High Ancestor	He (River Spirit)	Tu (Earth Spirit)	Directional Powers
11 (11%)	20 (21%)	7 (7%)	26 (27%)	21 (22%)	5 (5%)	2 (2%)	5 (5%)

Bin-hosting on the other hand, unlike exorcisms and announcing to the ancestors, appears through all periods of the oracle-bone inscriptions. Bin-hosting, like exorcism, appears to be a broad category of ritual, or an event, under which a wide variety of specific rituals or sacrifices could take place.²⁴ It was generally performed for near ancestors, but more distant ancestors and powers were also hosted (see Table 6.5).

Table 6.5 Bin-Hosting Recipients

Father	Grandfather/Ancestor	Elder Brother	High Ancestor	Ancestress	Former Minister	Nature Spirits
17 (22%)	27 (36%)	2 (3%)	14 (18%)	2 (3%)	3 (4%)	11 (14%)

As Liu (2004) has insightfully noted, Shang ritual can be divided into that which was goal directed, and that which was not. Liu (2004) further divides goal-directed ritual into praying for favor, and eliminating ill-fortune, and notes that the trend toward the systematization of ritual throughout the Anyang period is also a trend toward non-goal directed ritual. Like gao-announcing, but unlike yu-exorcism, bin-hosting can be seen as non-goal oriented ritual, being more relationship maintenance than goal-specific intervention. And yet, gao-announcing shares with yu-exorcism both an ad-hoc nature dependent on external events, and a disappearance from the late period inscriptions. Bin-hosting on the other hand, does not appear to have been performed on an ad-hoc, event-specific basis, and lasts throughout the Anyang period. Thus, if a generalization can be made from these three cases, it seems that it is not so much that goal-directed ritual was gradually replaced with non-instrumental ritual, but rather that the core goals of receiving ancestral favor, and avoiding misfortune, came increasingly to be sought through a systematic cycle of ritual that rendered ad-hoc interventions unnecessary. If Wu Ding and his immediate successors sometimes ritually intervened on behalf of favorites, in the latter half of the Anyang period, ancestral blessing was more systematically negotiated and distributed, thus changing the structural basis of the social economy of ancestral power. While Wu Ding’s ancestral sacrifices were distinguished from those of other elites by their magnitude and the prestige of his genealogical and political position, Di Yi and Di Xin’s ancestral sacrifices – through their systematized, routine nature – had become “sacrifices of state,” patterning the passage of time with their ceaseless ritual cycle, representing the king as steward of a complex apparatus of world pacification.

One aspect of this structural change is reflected in the nature and quantity of the sacrificial victims themselves. Two general trends can be seen: the gradual reduction of human sacrifice, and the decrease in the size of offerings generally over time, as can be seen in Table 6.6.

Table 6.6 Sacrificial Victims Over Time^a

	Period I	Period II	Period III	Period IV	Period V	Cumulative Averages
People	26.5%/40	25.5%/16	33%/16	30%/3	5%/18	24%/19
Specially-Reared Cattle	5%/8	0.5%/1*	22%/5	39.5%/4	52%/3	24%/4
Specially-Reared Sheep	21%/18	29%/3	8%/5	1.5%/3*	0	12%/6
Cattle	26%/10	37%/7	24%/2	23.5%/5	39.5%/1	30%/5
Sheep	8%/5	4.5%/2*	12.5%/14	4.5%/4	0	6%/5
Pigs	7%/6	0	4%/29*	1%/5*	0	2%/8
Dogs	2%/2*	0	2.5%/25	0	0	1%/5
Horses	1%/1*	0	1.5%/3*	0	0	0.5%/1
Millet Wine	1%/50*	3.5%/4*	2.5%/NA	0	4.5%/3*	2%/11
Human/Animal Ratiob	1.4: 1	1.1:1	1:1	1:3	1:2	1:1
Mean Offerings /Sacrifice	18.8	7.8	10.8	3.9	3.0	8.5

^a This table was constructed from a stratified sample of 1,200 sacrificial inscriptions (400 period I and 200 per period for each subsequent period) selected at random from pages of the Heji Shiwen. The percentage score indicates the percentage of inscriptions for that period that involved the sacrifice of that category of offering. The number that follows the percentage is the mean number of victims per sacrifice for that category as calculated from those examples that recorded a number in addition to the type of sacrifice. The asterix signifies that the number of examples that recorded a number of offerings was less than five and so the mean was calculated from a very small sample in those cases.

^b The human/ animal ratio is the ratio of the sum of human offerings to the sum of animal offerings in each period. This is arrived at by multiplying the percentage of occurrences of human sacrifice multiplied by the mean number of victims and comparing it to the aggregate of animal sacrifice occurrences multiplied by their mean number of victims.

The reduction in the size of sacrificial offerings over time, if it is not simply an artifact of the narrowing scope of the oracle-bone inscriptions,²⁵ could stem from two aspects of Anyang period developments, the first being the shrinkage of the King’s networks of allies, and, thus, web of resources, and the second being a lack of rationale for huge impressive sacrifices with the institutionalization of royal sacrifice as structurally differentiated from the practices of other elites. These two possibilities are not mutually exclusive, and, in fact, it is even possible that economic constraints were the driving force behind the late period ritual reforms. Against this hypothesis, however, is the fact that the Great Settlement Shang continued to grow, while its industries dramatically increased in scale and output in the latter part of the Anyang period (Li 2003a, Institute of Archaeology 2003, Li et al. 2011, Campbell et al. 2011, Campbell 2014a) suggesting that the “Great Settlement Shang” was an undiminished hub of economic, political and ritual activity throughout its existence. The decrease in the use of human sacrifice over the Anyang period could be related to the general decrease in the quantity of sacrificial offerings. As will be argued below, human sacrifice appears to employ two separate logics: that of exaltation, and of trophy taking, and that of reduction, and sacrificial consumption. Example (14) above records the late period use of the enemy leaders captured in the campaign against Mei, bo-lord of Wei, and example (22) divines about the use of three enemy leaders in one sacrifice.

甲申, 貞: 其執三邦 白于父丁. (32287)

Cracked on Jiashen day, tested: perhaps the captured three polity bo-lords (should be sacrificed) to Father Ding.

In addition to those examples, the following inscribed human skull fragments (Figure 6.4, examples (23), (24) display and commemorate the ultimate fate of captured enemy leaders,

Figure 6.4 Inscribed Human Skull Fragments Recording Sacrifice (heji 38759, 38758)

… 方伯用 … (38759)

… bo-lord of … fang, use …

… 方伯 … 祖乙伐 … (38758)

… bo-lord of … fang, … Ancestor Yi decapitate …

These skull fragments clearly demonstrate that the Shang sacrifice of leaders followed a logic of display and commemoration – punishment for the losers, and glorification for the victor – analogous to the inscription of hunting divinations on the skulls of prey like deer or tigers (see Figure 6.5).

Figure 6.5 Inscribed Deer Skull from Xiaotun, Anyang

(from Shijie 2008: 64, image 22)

Nevertheless, the vast majority of the early inscriptions regarding human sacrifice appear to follow another logic, that of reduction from political agent and enemy, to captive and sacrificial capital, interchangeable with animal offerings.

… 祖甲升 □□牢一牛, 用. (T2343)

… (what is to) Ancestor Jia offered up should be … specially-reared cattle and one (regular) head of cattle, use.

其三牢. 吉

Perhaps X three specially-reared head of cattle. Auspicious.

其五牢.

Perhaps five specially-reared head of cattle.

羌十人.

Qiang-captives, ten people.

十人五.

Ten people and five.

廿人. 大吉. 茲用.

Twenty people. Great Auspiciousness. Use this.

In the above example the offering originally divined about was specially-reared sacrificial cattle, but then shifted to the auspiciousness of offering qiang-captives²⁶ instead. Not only were qiang-captives a potential substitute for specially reared cattle in this instance, but comparing the numbers of victims offered, the relative value of human offerings appears to be lower than that of cattle (this pattern is generally supported by Tables 6.6 and 6.7).

Table 6.7 Numbers of Victims in Sacrifices where Qiang-captives and Animals Co-Occur^a

Qiang-Captives	Cattle	Sheep	Pigs	Specially-Reared Animals	Other
30					10
30					10
30	30
10	20
10	5
10	3
9	1
9	9
9	9				1
3				1
30	10
10					8
10					5
10				5
10				1
5		4	4
30			30
3				2
15				3
5		4	4
3				10

^a In the 21 divinations where qiang-captives and animals are offered together, there are fifteen cases (71 percent) where the qiang are more numerous and four cases (19 percent) where the numbers are matched.

A process of reduction from political agent, and potentially dangerous enemy, to anonymous victim, can be seen in the pattern of inscriptions concerning war, capture and sacrifice. While in war inscriptions enemies are referred to as fang, or by the place/lineage/polity names of their leaders, when capture is divined about, captives are sometimes referred to without reference to their political affiliation (Table 6.8). Perhaps the most remarkable pattern that emerges, however, is the relative number of divinations concerning qiang-captives 羌 (42 percent of the divinations concerning capture). Despite the fact that there is a Qiang Fang 羌方, inscriptions concerning it do not account for an equally important percentage of the warfare inscriptions – certainly not more than twice the aggregate of all other named enemies, as in the capture inscriptions.

Table 6.8 Capture Verbs and their Targets

	Huo 獲	Qin 摩	Zhi 執	Ji及	Totals
Specific fang	20 (15%)	1 (11%)	23 (29%)	4 (9%)	48 (18%)
Fang (generic enemies)	16 (12%)	3 (33%)	13 (16%)	35 (80%)	67 (25%)
Qiang	93 (68%)	4 (44%)	13 (16%)	3 (7%)	113 (42%)
Captives (yin, zhi, fu, etc.)	7 (5%)	1 (11%)		1 (2%)	9 (3%)
Other 芻			30 (38%)	1 (2%)	31 (12%)
Totals	136	9	79	44	269

Examining divinations concerning human sacrifice, an even more striking pattern can be seen. Aside from leaders who are frequently listed by name, title, and or polity (e.g. Mei, bo-lord of Wei 危伯美), human sacrificial victims (although presumably taken in war like their leaders), are almost never referred to by their political reference, but rather, as Yao (1979) pointed out, by some variant of the term “captive” (e.g. yin 印, fu 孚), by the method of their sacrifice (e.g. fa 伐), or simply by a general categorical term (ren 人 “man/person,” nu 女 “woman,” etc.). The sole apparent exception is qiang 羌. The “qiang-paradox” has hitherto generally been explained by assuming that for one reason or another, the Shang despised the Qiang people and especially ear-marked them for capture and sacrifice.²⁷ Recently, some scholars have proposed that the Qiang represented a fluid category of barbarian Other whose ethnic identity shifted over time rather than being a specific ethnic group as many scholars have assumed (Wang 1992, Shelach 1996a, Fiskesjö 2001).²⁸ If this is correct, then the Qiang would be a forerunner of Zhou terms for barbarian outsiders like Yi 夷, Man 蠻, Rong 戎 and Di 狄. While there is a Qiang Fang, in the oracle-bone inscriptions (and perhaps an Yi Fang 夷方), enemy political entities are generally termed fang irregardless of their ethnic/cultural affiliation – thus, Yu 盂 becomes Yu Fang 盂方 after it levies forces and threatens to attack Gao 高 (see Chapter 4 example (50). This is in marked contrast with the Zhou practice of referring to polities within its cultural domain by terms such as bang 邦 (“polity”/ “state”), and barbarian outsiders by collective ethnic terms. Moreover, this cannot explain the discrepancy between the relatively modest number of war inscriptions concerning the Qiang Fang with the overwhelming frequency of the word qiang in capture and sacrifice divinations. If capturing enemy populations was part of Shang Anyang’s generalized practice of inter-community violence, then why were qiang mentioned so frequently in contexts of capture and captive sacrifice, but not as frequently in war? The Wu Ding period oracle-bone inscription example below suggests an answer to this question.

乙未卜,貞:隻. 十二月. 允隻十六,以羌六. (258)

Yiwei cracked, tested: X will capture Y. Twelfth month. Indeed, (X) captured sixteen, (and is) bringing qiang-captives, six.²⁹

In the charge, the issue divined about concerned the capture of Y (or people of Y) by X. The verification confirms that indeed sixteen were caught, but then they are termed “qiang.” This strongly suggests that the word “qiang” in the captive and sacrifice divinations is not, in fact, referring to people of the Qiang Fang, but rather denotes a more general term for captive or slave. This explanation would go a long way to explaining why there are so many examples of the word qiang in capture and sacrifice divinations, not to mention the graphic variant of qiang written with a rope around the neck. A remaining question then, is whether or not this sense of qiang is merely a semantic extension of Qiang Fang to a generalized barbarian Other, and term for captive, or whether the graph used to write Qiang Fang represents more than one word, one of which means “captive” or “slave.” The latter hypothesis is supported by the fact that the word qiang 羌 was a homophone of the verb 強 in Old Chinese, and the latter has both verbal and adjectival uses as in “strong/hard” and “to force” (強迫).³⁰ Although there are no known cases of either 強 or 羌 being used in later transmitted texts in the sense of “slave” or “captive,” it would not take a great semantic leap of faith to argue for a meaning of “slave/captive” as a derivative of “to force” (or visa versa).³¹ If 強 was written 羌 in the oracle bones it might also explain why Qiang Jia 羌甲, a Shang royal ancestor, nevertheless apparently bore the name of an ethnic group that, in the traditional account, was so despised as to be especially marked out for capture and sacrifice. In this re-reading, Qiang Jia would be rendered something like “Mighty Jia” instead of “Loathsome Barbarian Jia,” which would be much more in keeping with the practice of referring to ancestors by either a genealogical marker (father, ancestor, big, middle, small) or terms of glorification such as Tang “the Accomplished,” Ding “the Martial,” Ding “the Martial and Cultivated,” etc. The argument for reading qiang as a generalized term for barbarian, on the other hand, is supported by examples such as the following,

乙卯卜, 爭貞: 王□伐馬羌. (6624)

Yi Mao day cracked, Zheng tested: the King (should) attack the Horse Qiang.

己酉卜, 王北羌伐. (6626)

Ji You day cracked, Ke (tested:) it should be the Northern Qiang that the King attack.

□亥卜,羌二方白其用于祖丁, 父甲. (26925)

… Hai day cracked, the qiang two Fang bo-lords should perhaps be used (in sacrifice) to Ancestor Ding,

In the first two examples, the enemy is neither referred to as a Fang nor as a specific polity/political actor but rather as “qiang” of different type/location. In the third example two bo-leaders are about to be offered to the royal ancestors, but they are designated simply as “qiang” rather than by their polities or personal names. Given that there generally seems to be one bo-leader per polity in the oracle-bone inscriptions, “qiang” does not appear to refer to a polity here, and is either a loose political/ethnic term or simply designates that the leaders are captives or “barbarians.”³²

Whether the “qiang” that appears in the sacrificial inscriptions derives from semantic extension of Qiang Fang or represents a phonological loan for qiang 強 (“to force”< “those who are forced”< “captives/slaves”), the important issue is that it does not have a specific political referent in these cases, but rather instantiates a logic of sacrificial reduction, and erasure of agency³³ – a process of “pseudo speciation” rendering captives available for ancestral consumption along with cattle, sheep, pigs and dogs.³⁴ From dangerous enemies of specific, named polities, to nameless captives, and, finally, sacrificial capital expended in the tens to hundreds (Figures 6.6 and 6.7).

Figure 6.6 Sacrificial Pit in Royal Cemetery

(after Institute of Archaeology 1994: plate 15, 1)

Figure 6.7 Sacrificial Pit with Headless Victims in Royal Cemetery

(after Institute of Archaeology 1994: plate 15, 4)

Returning to the issue of the reduction of human sacrifice in the last half of the Anyang period, this phenomenon may be tied to the wider ritual changes outlined above. In Wu Ding’s times, a logic of excess pervaded royal ritual, creating distinction through scale. The King and his network of allies were also very active militarily, and, as noted in Chapter 4, the chief forms of tribute or booty sent in by allies, and divined about by the King, were cattle and captives. The debasement and sacrifice of conquered enemies, moreover, bolstered the King’s discursive claims to a privileged place in the hierarchy of being. Royal punishment for rebels and enemies provided post-factum verification of the King’s divine favor, even while the outcome of the negative dialectic of inter-subjective power played out on the battlefield was cemented through the transmutation of once-dangerous enemies into faceless captives whose “use” among the livestock of the King’s rituals marked their final domestication, and ultimate place within the hierarchy of being. As the King gained with each sacrifice, so too his enemies were reduced, for the King, nearly deified in life, upon receiving a royal burial, would be apotheosized by his descendants, even while his sacrificed enemies were denied places among their kin – bodies hacked, burned or buried alive to feed the glory of the royal line – their social, physical and ancestral deaths dialectically enhancing the King’s social, political and ontological being.

In the later reigns of the Anyang kings, the royal ritual had become a systematic, institutionalized complex of techniques of world domestication insinuated into the rhythm of life to an extent unprecedented in earlier reigns. It may have been that the excessive gestures of the earlier kings like Wu Ding – distinguishing their ritual from that of other high elites – were no longer necessary. It may also be that the later Kings simply did not have the same coercive networks, or as frequently used them, as Ding “the Martial” (Wu Ding), or that while enemy leaders were still sacrificed, their followers were more frequently put to other uses – relocated, enslaved, or incorporated into lineages as dependents of some sort (Huang 2004).

Ultimately, the theme of internal pacification reveals the structural instability built into the Shang world with the dispersion of coercive capital, segmental organization and the possibility of other ancestral orders and hierarchies of authority – as the Zhou conquest demonstrated. If the position of the king was vouchsafed during his lifetime significantly through success in war and sacrifice, then his position in the afterworld depended on his descendants, both in terms of their mortuary and sacrificial ritual, and in terms of their relative position in the world. The ancestors of overthrown dynasts, after all, are no longer royal ancestors. The seemingly stable Shang hierarchy of authority then, was actually one of dynamic balance, negotiated with the living, the dead and the powers of the land through the perpetual work of pacification. In this sense, war and sacrifice were two aspects of the same structuring institution, creating order and a measure of ontological security. While later kings such as Di Yi 帝乙 and Di Xin 帝辛 seem to have presided over a less dynamic and more systematic apparatus of pacification, their complex of practices of authority remained fundamentally the same as that of their ancestors: warring, inspecting, hunting, allying, sacrificing and divining: in constant motion across the landscape. From the point of view of the polity, the relationship between forms of social violence and authority meant that Shang warfare was not an unfortunate aberration of failed communication, or even diplomacy by other means, but rather a key practice of authority, glorified through such elite symbols as ritual weapons, chariots and monumental human sacrifice. Yet, if Shang practices of social violence involved both glorification and internal pacification, it is important to distinguish their particular forms and logics. The glory won through battlefield and sacrificial success has to be seen within the total structural context of ancestral place, and the lineage-based organization of social life. The kudos passing from defeated to victor was not merely inter-subjectively constitutive of enhanced personal honor, but that honor was fundamentally predicated on a structuring of social place based on ancestral merit.³⁵ Internal pacification in the Shang case likewise does not merely refer to the elimination of military threats within a territory, but rather to the pacification of sources of danger in general, and the creation of order out of the anarchic forces of the world. Neither should pacification be seen solely in instrumental, monolithic “state” terms: war and sacrifice were not monopolies of the Anyang kings, but rather general, hierarchically structuring practices differentially participated in by all. These structuring practices, moreover, constituted the authority of other elites as well, their power stemming from their position as ritual and military leaders, and representatives of their lineages. Nevertheless, as will be discussed more fully in the next chapter, the great disparities of symbolic and material capital lavished on royal burials in contrast to those of other lineage leaders, the concentration of the vast majority of sacrificial offerings in the palace-temple area and royal cemeteries, and the size of the forces levied by the Shang kings, all suggest that the ritual and coercive capital of subordinate lineages was largely channeled into the dynastic enterprise, even as the capital of subordinate lineage members was harnessed by their lineage heads.