Introduction
The inscriptions and texts used in this article document some of the steps through which Shang kings established a direct line, father to son, of the transmission of royal power. These steps, observable in the inscriptions of the reign of Wu Ding 武丁can be understood, I argue, as an answer to a series of political crises linked to the process of royal succession.
The Shang king Wu Ding was given, in the received texts, the title of High Ancestor (Gaozong 高宗).Footnote 1 His reign came after a series of royal successions named in the later Shiji 史記 as the “nine generations chaos” (九世亂).Footnote 2 This “nine generations chaos” was characterized by a transmission of royal power not in a direct line father-to-son, but through an adelphic, brother-to-brother system. Those brothers were sons of the same father, but from different mothers. I argue in this paper that this led, through a reworking of royal power succession, to a situation where two royal lineages alternated royal power between them.Footnote 3 I analyse this in terms of political crisis that had lasting effects on later reigns.
It reignited when Xiao Yi 小乙, the last reigning king of four brothers, Yang Jia 陽甲 the elder, Pan Geng 盤庚, Xiao Xin 小辛, and Xiao Yi himself, left the royal power to his own son, Wu Ding: as Cai Zhemao 蔡哲茂 remarked, the successor of Xiao Yi should have been a son of Yang Jia. The circumstances surrounding Wu Ding's accession to royal power bear witness to this enduring tension among the members of the royal lineage.
Historical sources allude to difficulties at the beginning of his career, when his father, Xiao Yi, ordered him to leave Shang-controlled territory. Archaeological information available since the discovery of the Shang city Huanbei 洹北, immediately north of Anyang, in 1999, add weight to what received sources only alluded to: the start of the reign of Wu Ding was a very dynamic process. In this paper, I argue that part of this process was linked with intra-lineage problems inherited from the previous “nine generations chaos”.
It led Wu Ding to implement a policy of recruiting new allies, even from outside the Shang clan, through multiple matrimonial alliances; one of these, with the Lady Jing 婦妌, who was not part of the Shang clan, is particularly significant. The establishment of a powerbase through marriage was only one part of the measures needed for the stabilization of the Shang polity. Wu Ding took the initiative to name one of his sons a direct successor, when he bestowed him the title of young king, or king in waiting 小王. This initiative was repeated in subsequent reigns.
Wu Ding also implemented a series of ritual experiments, initiating one of the first attempts to identify, among all the royal ancestors, a principal line of royal succession. This attempt failed but, after his reign, his successors benefitted from the ritual roadmap that he first designed.
The aim of my article is to understand the origin of the political situation at the beginning of the reign of Wu Ding. One specific marker given to a limited number of ancestors in the inscriptions of this and subsequent periods is germane to this question. I interpret this marker in the context of political instabilities brought about by the polygamous alliances of the Shang kings and the subsequent difficulties for a king in choosing an heir.
I. The “nine generation chaos” and the crisis of the two sub-lineages
a. The Shiji testimony
Sima Qian gives the following interpretation of this period:
自中丁以來，廢適而更立諸弟子，弟子或爭相代立，比九世亂，於是諸侯莫朝。
From the reign of Zhong Ding on, the (principle of the) main succession line has been abandoned and the sons of younger brothers became kings, those sons of younger brothers fought with each other and there was disorder during those nine generations [of kings] and the territorial lords did not come to royal court.Footnote 4
The nine generations mentioned by the Shiji are the following kings:
(1) Zhong Ding 仲丁(中丁) → (2) B Wai Ren 外壬 (Bu Ren卜壬) → (3) B He Dan Jia 河亶甲 (Jian Jia 戔甲) → (4) S Zu Yi 祖乙(祖乙) → (5) S Zu Xin 祖辛(祖辛) → (6) B Wo Jia 沃甲 (Qiang Jia 羌甲) → (7) BS Zu Ding 祖丁(祖丁/son of Zu Xin 祖辛) → (8) FBS Nan Geng 南庚 (南庚, son of Wo Jia 沃甲) → (9) cousin second degree Yang Jia 陽甲 (Hu Jia 虎甲/son of Zu Ding 祖丁).Footnote 5
Sima Qian gives only one reason for the disorder: the struggle for royal power, provoked by the abandonment of “the (principle of the) main succession line” (di 適), the succession of a king by the elder son of the principal wife. Di is used in received sources, not in Shang documents. Moreover, the rule it designates did not apply in the royal succession of the Shang dynasty, as the well-documented example of the succession of Wu Ding shows: he was succeeded by two of his sons, Zu Geng 祖庚 and Zu Jia 祖甲.Footnote 6 Nevertheless, the historical reliability of the Shiji should not be over-doubted, for two reasons:
• This list of ancestors in the Shiji corresponds almost exactly to the list of royal ancestors reconstituted from the Zhouji 周祭, a series of cyclical sacrifices presented briefly below.
• Shang oracular inscriptions confirm that the transmission of the royal power was not a straightforward process.Footnote 7
What Shang oracular inscriptions provide, particularly those given by the Zhouji 周祭, is the mention of spouses of the kings.Footnote 8 This element gives a clearer picture of intra-lineage relationships.
b. The oracular evidence
The Zhouji is the name given by Chinese scholars to a series of sacrifices organized in a cycle, discovered first by Dong Zuobin 董作賓 in 1945.Footnote 9 This cycle is composed of a series of five types of sacrifice, dedicated to Shang non-royal and royal ancestors, and was first implemented during the reign of the king Zu Jia 祖甲, son of Wu Ding (GZ 1202–1169 bce, TDC 1191–1148 bce).Footnote 10 The cycles, after an interruption, have been continued until the end of the Shang dynasty.Footnote 11 Oracular inscriptions from those cycles provide the names of the spouses of the kings belonging to the “nine generations chaos”:
• Zhong Ding 中丁 has two spouses, Bi Ji 中丁奭妣己 and Bi Gui 中丁奭妣癸;Footnote 12
• Zu Yi 祖乙 is associated with two spouses, Bi Ji 祖乙奭妣己 and Bi Geng 祖乙奭妣庚;
• Zu Xin 祖辛 has only one spouse, Bi Jia 祖辛奭妣甲;
• Qiang Jia 羌甲 is given one spouse;Footnote 13
• Zu Ding 祖丁 is associated with two spouses, Bi Ji 祖丁奭妣己 and Bi Geng 祖丁奭妣庚.Footnote 14
The last king, Yang Jia 陽甲, has no recorded spouse.
This sequence is exactly the same in the Shiji and in the reconstituted list from the cyclical sacrifices: Zhong Ding was succeeded by his brother, Bu Ren 卜壬; royal power then went to the two sons of Zhong Ding, first Jian Jia 戔甲 and after Zu Yi 祖乙.Footnote 15 Therefore, Jian Jia had succeeded his father's brother (FB). Zu Yi was succeeded by two of his sons, first Zu Xin 祖辛and after, Qiang Jia 羌甲. The royal power went then to the only recorded son of Zu Xin, Zu Ding 祖丁, who also succeeded his father's brother (FB). After Zu Ding, the royal power went to Nan Geng 南庚, the only recorded son of Qiang Jia. Nan Geng therefore succeeded his father's brother's son (FBS), Zu Ding, to whom he was a paternal cousin.Footnote 16 In this sequence, the mention of kin relationship between alternating kings is not the most important information, as Figure 1 shows.Footnote 17
Figure 1. The alternating nine kings and the succession of Yang Jia
Zhong Ding 中丁 is accompanied by two spouses and has two sons; only one of them, Zu Yi 祖乙, is also recorded with two spouses and two sons. The number of spouses corresponds to the number of sons at the next generation.
For the reigns of Zu Xin and Qiang Jia, only one spouse and one son are mentioned. From the reign of Zu Yi 祖乙 to that of his two sons and grandsons, the constitution of two parallel lines of kings can be observed. Their ascension to royal power alternates: a king from one line is succeeded by a king from the other line. This appears clearly by taking into account the generation to which each king belongs.
The first royal line is:
Zu Yi 祖乙 → Zu Xin 祖辛 (son of Zu Yi) → Zu Ding 祖丁 (son of Zu Xin);
The second is:
Zu Yi 祖乙 → Qiang Jia 羌甲 (son of Zu Yi) → Nan Geng 南庚 (son of Qiang Jia).
These were two emerging royal unilineal lines of descent from the same ancestor, Zu Yi, mirroring one another, each composed of a “unit” comprised of grandfather, father and son.Footnote 18 The royal power was passing from one line to another. The reason for such a combination does not appear clearly either in the Shiji or in the oracular inscription. Was it the result of a “gentleman's agreement” between the two lines? I think that the alternation between kings belonging to the two lines suggests such an arrangement. This agreement would then have been made to the detriment of the other sons of Zu Yi; it might also have been sealed with an exchange of women between the two lines, at least between Zu Xin and Qiang Jia.Footnote 19
These two parallel royal lines are characterized by a patrilateral, unilineal line of descent, with the brothers Zu Xin and Qiang Jia each associated in the Zhouji with only one spouse, corresponding strictly to the number of their reigning offspring.
The agreement should have led, after the reign of Yang Jia 陽甲, to the succession of one of the sons of Nan Geng, that is why I have given the same succession number (10) to two kings, Pan Geng 盤庚 and an X, son of Nan Geng.
Who was this prince X? In the many inscriptions dating from the long reign of Wu Ding, Cai Zhemao 蔡哲茂 noticed the presence of a Bing (pronunciation hypothetical) Father Ren 
父壬, mentioned for example on the fragment 合集 1823正, group 賓 1.Footnote 20 A series of inscriptions mentions three ancestors: Qiang Jia 羌甲, Nan Geng 南庚 and this Father Bing Ren, presented as capable of inflicting harm on the king (害王). Based on the fact that Qiang Jia and Nan Geng are father and son, Cai Zhemao hypothesizes that those three ancestors belonged to the same lineage: the mention of Bing Father Ren would be suggestive of such a link. I agree that he was probably a son of Nan Geng.Footnote 21
His hieronym, Bing Fu Ren (the character bing being interpreted by Cai Zhemao as the personal name of this individual), indicates that he belonged to the generation of Yang Jia, Pan Geng, Xiao Xin, and Xiao Yi.Footnote 22 Referring to the above genealogical schema, Father Bing Ren was exactly at the position X would be vis-à-vis Wu Ding. It is therefore possible that Bing Father Ren is the son who should have succeeded the king Yang Jia. It is not surprising that inscriptions mentioning the royal line of Qiang Jia – Nan Geng – Bing Father Ren is in a context of the oracular testing of the harm done to Wu Ding himself: Wu Ding was indeed the heir of another line that, from Yang Jia on, excluded the other line from royal power.
Those elements allow us to understand why Sima Qian made the first reigning son of Zu Ding, Yang Jia, the last monarch belonging to what he called the “nine generations chaos”: Yang Jia should have been succeeded by a son of Nan Geng. The son of Nan Geng who, according to what happened before (the alternation between the two lines of Zu Xin and Qiang Jia), should have acquired royal power after Yang Jia, did not. Sima Qian noted that 帝陽甲之時，殷衰 “In the time of the monarch Yang Jia, the Yin (Shang) became weak”.Footnote 23 I hypothesize that it is during his reign that the process of succession (alternation) was called into question. Even if the reasons for the turmoil do not appear clearly in the received sources, what happened after the reign of Yang Jia indicates that the intra-lineage accord came to an end: the successor of Yang Jia, Pan Geng, decided to move the capital.
c. Pan Geng 盤庚 and the moving of the Shang capital
During the reign of Pan Geng, the king who became king, possibly in lieu of one of the sons of Nan Geng, Bing Fu Ren, the Shang capital was moved. Two sources, the Shiji and three chapters of the Shangshu 尚書 (“Pan Geng” 盤庚 I, II, III) attribute the move directly to Pan Geng, who moved the capital from Xing 邢, set by Zu Yi 祖乙, to Bo 亳, the ancient capital of the first Shang king, Tang 湯.Footnote 24
Recent archaeological discoveries seem to confirm such a move. A walled site, Huanbei 洹北, situated north-west of the later Shang city of Yinxu 殷墟, and north of the Huan river 洹河, was discovered in 1999; it is dated from middle Shang first and second phase.Footnote 25 A large consensus of scholars identifies Huanbei with the city of Pan Geng 盤庚, Xiao Xin 小辛, and Xiao Yi 小乙.Footnote 26
The move of the capital was motivated, as the “Pan Geng” II has it, by the necessity to “stabilize the polity” 安定厥邦.Footnote 27 The Shiji adds that Pan Geng established his capital on the site of the old city of Tang, the first king of the dynasty 復居成湯之故居. This passage makes Pan Geng's move a foundation anew of the Shang monarchy. This foundation might have been necessitated because of the conflict of legitimacy between the two royal lines descended from Zu Yi: the moving of the capital would have then solved the tension arising when Pan Geng became king instead of one of the sons of Nan Geng.Footnote 28
Pan Geng's move of the capital might not have extinguished intra-lineage problems of succession, involving the four sons of the king Zu Ding 祖丁: Yang Jia 陽甲, Pan Geng 盤庚, Xiao Xin 小辛, and Xiao Yi 小乙, if Cai Zhemao 蔡哲茂 is to be followed: he hypothesized a crisis of succession involving Xiao Yi and his son Wu Ding 武丁.Footnote 29 In other words, he treated Wu Ding's accession to royal power as an anomaly.
II. The “anomalous” reign of king Wu Ding
According to the Shiji, Wu Ding succeeded his father, Xiao Yi小乙, the last of the four brothers (Yang Jia 陽甲, Pan Geng 盤庚, Xiao Xin 小辛, and finally Xiao Yi), sons of Zu Ding. When Xiao Yi died, royal power should have been given to a son of the elder brother Yang Jia. It was not the case since it is the son of Xiao Yi, Wu Ding, who became king.
a. The nature of the crisis
Cai Zhemao hypothesized that Wu Ding, son of Xiao Yi, has been able to “bypass” the adelphic way of succession (the rule according to which a son of Yang Jia should have become king after the death of Xiao Yi) because the descendants of two previous kings, Qiang Jia 羌甲 and Nan Geng 南庚, offered him support. His work is all the more provocative since the Shiji does not mention any dynastic conflict for the reign of Wu Ding. Nevertheless, what is not in dispute in the genealogy of the Shang kings is the fact that, after the crisis of the nine generations of kings, four sons of the king Zu Ding became kings, one after the other.
A case in the Shiji, in the annals of the territorial house of the Wu 吴, the “Wu taibo shijia” 吴太伯世家, is presented by Cai Zhemao as a proximate historical illustration of an intra-lineage conflict between brothers.Footnote 30 Considering Shang information and reasoning by analogy, Cai Zhemao surmises that the rule would have been, in the case of the succession of several brothers, that royal power should be bestowed on the son of the elder brother after the end of the reign of the younger, last reigning brother.
To my knowledge, Cai Zhemao is the first to have questioned the ascension to royal power of the king Wu Ding by taking into account the possibility of a power struggle between royal sub-lineages. He has detected traces of this struggle in oracular inscriptions: several reconstituted inscriptions (dating from the reign of Wu Ding) mention three kings: Father Jia 父甲, Father Geng 父庚. and Father Xin 父辛, corresponding to Yang Jia 陽甲, Pan Geng 盤庚, and Xiao Xin 小辛.Footnote 31 In some inscriptions, there is the mention of “three Fathers”, 三父, a collective to whom sacrifices are offered. This collective is sometimes called san jiefu 三介父 “the three collateral fathers”.Footnote 32 While in some inscriptions (where the two other recipients are named Father X (case of the 合集 903 group 賓 standard), Yang Jia is also named Father Jia 父甲, in others, he is called Yang Jia 陽甲. Cai Zhemao explains this phenomenon as a mark of disrespect, a result of the political struggle Wu Ding had engaged in against the male heirs of Yang Jia.Footnote 33 Therefore, in order to reinforce his own powerbase, Wu Ding then allied himself with the descendants of Qiang Jia.Footnote 34 I do not think this is the case, due to the fragment 合集 1823 正 analysed above showing that the Qiang Jia royal line was interpreted by Wu Ding as a source of possible harm.
Cai Zhemao's questionable interpretation of the relationship between Wu Ding and the Qiang Jia royal line does not void his ingenious hypothesis: the reign of Wu Ding was, succession-wise, within the frame of adelphic succession, an anomaly. Other sources allude to the difficulties the monarch encountered before and at the beginning of his reign.
b. The debut of Wu Ding: a very dynamic process
Sima Qian's Shiji mentions that Wu Din, after the death of his father, was 思復興殷 “pondering on [the best way] to revive the fortune of the dynasty”.Footnote 35 Other received texts note that Xiao Yi, the father of Wu Ding, played a very important role in the circumstances leading to the reign of his son:
• chapter “Wu yi” 無逸 of the Shangshu: 其在高宗，時舊勞于外，爰暨小人。 “As for Gaozong (=Wu Ding), he began to live and toil outside, with the common people.”Footnote 36
• Zhushu jinian: 小乙: 六年，命世子武丁居于河，學于甘盤。 “the sixth year [of his reign], the king Xiao Yi ordered his heir Wu Ding to live inside the Yellow river and to learn the ways of Gan Pan.”Footnote 37
• chapter “Yue ming” 說命下 of the Shangshu: 王曰：「來！汝說。台小子舊學于甘盤，既乃遯于荒野，入宅于河。 “The king [Wu Ding] said, ‘Come, O Yue. I, this child, a long time ago learned with Gan Pan. Afterwards I fled in the wild, and then I went to (the country) inside the Yellow river, and lived there’.”Footnote 38
Of course, two of those texts (chapters “Yue ming” and “Wu yi”) are very late in provenience. The chapter “Wu yi” in particular has another anecdote about Zu Jia, one of the reigning kings of Wu Ding, mentioning that 舊為小人 “he was at first one of the lower people”.Footnote 39 The reference to the “lower people” might only be a trope to show the virtue of the monarchs, able to sympathize with the difficulties of commoners.Footnote 40
The text from the Zhushu jinian only says that the king Xiao Yi, having designated Wu Ding as his heir, ordered him to go (presumably) outside the Shang capital – the rationale given for such a decision being to learn with a man who was probably an ally of Xiao Yi.Footnote 41 While the chapter “Yue ming” cannot be considered a reliable historical testimony, it does mention the same Gan Pan.Footnote 42
There might have been another reason for Xiao Yi to send his son outside the Shang capital. If the probable intra-lineage tensions resulting from the “ejection” of the sons of Yang Jia from the succession are taken in account, Wu Ding would have become a target of another powerful royal lineage.Footnote 43 To become king and effectively exert royal power would have required a series of drastic measures. The discovery of the site of Huanbei provides an important clue to that.
The site of Anyang is characterized by the fact that, in the present state of documentation and after a century of digging, no material belonging to reigns prior to Wu Ding has been found.Footnote 44 The site of Huanbei has been abandoned in the third stage, and the two palace compounds excavated (1 and 2) have been deliberately burnt (almost no material remains have been found above the burned foundations).Footnote 45
In other words, it is probable that the site of Anyang was founded by Wu Ding himself. It is surprising that no historical record makes mention of such a move, but move there was.Footnote 46 This move would have allowed Wu Ding to start with a “clean slate”, outside (literally) the tangled web of intra-lineage politics. Another way for Wu Ding to consolidate his power was his marital alliance with a non-Shang lineage, examined below.
The clean slate I have hypothesized was nevertheless only relative, since the move from Huanbei to Anyang proper (as the abandonment of Huanbei shows) suggests that at least a portion of the Shang people went with Wu Ding when he made the move to Anyang proper. He might, therefore, have had to cope with the sons of the previous kings when they were alive, but also after their deaths, as the oracular inscriptions of his reign show (posthumously they are known as the xiong 兄, the elder brothers, twelve of them being mentioned). Since the term xiong is classificatory, it designated his own brothers (from the same father, but from other spouses of his father) and cousins, that is to say male children of the brothers of his father.Footnote 47 Some of those “brothers” could have accessed royal power by virtue of their being sons of kings, for example the sons of Yang Jia.
Wu Ding had to offer sacrifices not only to dead members of the royal lineage, he had also to deal with those who were alive. The case of the war chief Que 雀 is the subject of a long and fascinating article by Zhang Weijie 張惟捷.Footnote 48 This Que was active during Wu Ding's middle reign period.Footnote 49 Zhang Weijie noticed (pp. 714–22) that he was sometimes ordered to perform rain-making sacrifices dedicated to the deities of mountains and rivers, usually reserved to the king. For the author (pp. 750–55), Que was not called “Zi” (prince), because he has acquired a great deal of power and autonomy.Footnote 50 Que, as the inscription 合集 6946 (period 1, group 賓 standard) shows, had command over the royal host (wangzu 王族), and other chiefs of noble houses reported to him (gao 告). In light of this evidence, Zhang Weijie (p. 755) makes the hypothesis that Que was a half-brother of Wu Ding himself, born of a different mother.Footnote 51 I think he might also have been a son of one of the paternal uncles of Wu Ding.
Que's tasks show that he was a major helper in royal undertakings. Zhang (p. 756) rightly emphasizes the exalted status of Que, attributing it to his blood relationship to the king and his military prowess; he also suspects another set of hidden reasons: a “tradeoff” for royal succession. In other words, Wu Ding had to “compensate” at least one of his (classificatory) brothers in order to reign. This compensation associated Que to the ritual and military exercise of royal power, thus “neutralizing” a possible rival.
Wu Ding also relied on multiple marital alliances to consolidate his power. One of the important spouses of Wu Ding was Lady Jing 婦妌. Her case is significant, since this woman came from outside the Shang clan.Footnote 52
c. The case of Lady Jing 婦妌 and the marital alliances of Wu Ding
Lady Jing, Fu Jing 婦妌, was one of the three spouses of Wu Ding mentioned in the later iterations of the cyclical sacrifices. Her hieronym was Bi Wu 妣戊 Lady Jing; she died at the end of Wu Ding's reign.Footnote 53
This Lady Jing is mentioned in many inscriptions dating from the reign of Wu Ding.Footnote 54 The king, as was the case for Lady Hao 婦好, another important spouse of Wu Ding, divined her pregnancy.Footnote 55 She was involved in many important affairs, such as agriculture and war.Footnote 56
This spouse came from Jing 井 territory: her name is composed of the character 井 to which the character nü 女 (woman) has been added.Footnote 57 This territory, situated west of Shang-controlled territory, is attested in the oracular inscriptions of Wu Ding; most inscriptions present it as an ally.Footnote 58
Was Lady Jing kin-related to the king Wu Ding? Meng Shikai 孟世凱 and Zhu Zhen 朱楨 both answer in the negative: Lady Jing was a woman from an allied (most of the time) territory.Footnote 59 Roderick Campbell, while positing that “Shang kings took wives from closely affiliated lineages” as a kind of endogamous alliance between “high-ranking Shang clans”, adds that some spouses came from allied territories.Footnote 60 The alliance with Lady Jing and her lineage was therefore highly significant.
The distant, non-kin relationship between the Jing territory and the Shang can be illustrated by the inscription 合集 1339, period 1, group 賓 1:
癸卯卜賓貞井方于唐宗彘
The day guimao cracks, Bin tested the oracular proposition: [let the] people of Jing [present] a wild boar in the temple of Tang (The Victorious = Da Yi 大乙).
There is no sacrifice mentioned in the inscription, hence our supplying the term “present”. From this inscription and the absence of sacrifice offered to Da Yi, it is possible to deduce the absence of any previous blood ties between the people of Jing and the Shang king. The alliance with the non-Shang lineage from the Jing territory bears witness to the contacts the Shang dynasty established with non-Shang people. Archaeology confirms the reality of those contacts.
The comparison between the structures and the material dug in the sites of Huanbei and Anyang gives clues to the changes the Shang dynasty underwent immediately after the abandonment of Huanbei, the city founded by Pan Geng. The two cities have developed according to two different models, Huanbei being walled and Anyang not.Footnote 61 What is more significant is the fact that the Wu Ding phase of Anyang suggests “a high degree of heterogeneity of material culture and population”, with lots of imports of objects “totally different from those typical ‘Shang’ traditions”, that is to say material from earlier phases of Shang dynasty.Footnote 62 From those observations, it is easy to deduce that the reign of Wu Ding saw a phase of growing cultural contacts with non-Shang entities. The marriage with Lady Jing is just one example of that.
The link between Lady Jing and her territory of origin was so strong that an inscription, 屯南 4023, period 3, group 無名, after the reign of Zu Jia, mentions a sacrifice to a Bi Wu Jing 妣戊妌.Footnote 63 To the normal hieronym was added the character designating the territory of origin, even after her death and a timespan of a minimum of two generations.
In light of tensions arising within the Shang royal lineages, it would have made sense for the young Wu Ding to make powerful allies outside of those lineages; one of those alliances was then sealed through a marriage with Lady Jing.
One of the conditions of this alliance was probably to ensure that the fruit of this union would be given a special status and the attribution of the title of “young king” or king in waiting 小王 to a prince son of Lady Jing is a clear indication of that.Footnote 64 While the naming of an heir by Wu Ding cannot be regarded as institutional, since it is the only recorded instance in oracular inscriptions, it had the double advantage of potentially avoiding the pitfalls of adelphic succession and the entanglements of Shang intra-lineages’ rivalries.Footnote 65 It might also have been a source of tensions. That is why Wu Ding also pursued throughout his reign a multi-pronged strategy of alliances, manifested through:
1. the impressive number of spouses of Wu Ding recorded in inscriptions dating from his reign; and
2. his marriage to Lady Hao, daughter of Yang Jia 陽甲.
On the first point, the inscriptions dating from the reign of Wu Ding mention numerous spouses, some (a dozen or so) belonging to Shang nobles. While the exact number of spouses of Wu Ding varies according to the different tallies, it is clear that the king was not monogamous.Footnote 66 Only a very limited number of spouses came from non-Shang polities (case of Lady Jing). This means that Wu Ding, through this network of alliances, associated to him a large number of Shang-related lineages, constituting thus an important part of his powerbase.Footnote 67
The second point is linked to the information available through the recently discovered Huayuanzhuang 花園莊 inscriptions. The prince of Huayuanzhuang was the young son of Wu Ding, the future king Zu Jia.Footnote 68 Due to the abundance of mentions of Lady Hao in Huayuanzhuang's corpus and the fact that this lady was one of the honoured spouses of the king Wu Ding, it is plausible that she was the mother of the prince.Footnote 69
An ancestor, designated as Zu Jia 祖甲, was the third most mentioned in Huayuanzhuang's inscriptions. From the context of those inscriptions, it can be inferred that Zu Jia/Yang Jia 祖甲/陽甲 was the father of Lady Hao.Footnote 70 Since he was the first reigning paternal uncle of Wu Ding, what was the reason for his importance? I infer that Wu Ding, taking as a wife Lady Hao, that is to say a parallel paternal cousin (FBD), established through her an alliance with the lineage of Yang Jia, whose son should have been the successor to the royal power after the death of Xiao Yi.Footnote 71 This marital alliance would have extinguished, so to speak, one source of discord inside the Shang royal lineages.Footnote 72
Marital alliances were a tool to establish a powerbase, but Wu Ding also needed to reshape the Shang polity through a series of ritual processes.
III. The organization of the royal ancestral cult during the reign of Wu Ding
We do not have access to palaeographic documents dating from before the reign of Wu Ding. Nevertheless, since this monarch left the site of Huanbei to establish a new capital, it would not be surprising if such a move also entailed a reworking of whatever ritual system existed before. Some oracular inscriptions dating from Wu Ding's reign show that this reworking was taking place.
Certain ancestors were given a pre-eminent place; this is the case for the ancestor Shang Jia 上甲.
a. The choice of the great ancestor Shang Jia
Received sources mention Shang Jia as a most important ancestor for the Shang dynasty:
• Zhushu jinian 竹書紀年: 武丁十二年，報祀上甲微。 “The twelfth year [of his reign, the king Wu Ding] offered sacrifices of gratitude to Shang Jia Wei.”Footnote 73
• Guoyu 國語: 上甲微，能帥契者也，商人報焉 “Shang Jia Wei, he was the one who could lead [Shang lineages] to bond with each other, the Shang expressed thus their gratitude to him.”Footnote 74
The second text uses the term shuaiqi 帥契 that I have translated as “leading the bond”. The character qi 契 is indeed used in later Zhou sources as designating a bond or a contract. Since the royal lineages have been through a series of crises of succession, one immediately before the reign of Wu Ding, it seems that Shang Jia has been given pre-eminence in order to provide the Shang with a great, unifying ancestor. This pre-eminence was even more obvious during the reign of Zu Jia, as the inscriptions of the Zhouji show: he was honoured at the beginning of each type of ceremony included in the cyclical sacrifices (all periods), using the formula “(sacrifice X offered to) Shang Jia and all his descendants” 自上甲至于多毓. This formula was already employed during the reign of Wu Ding as the inscription 合集 10111 (group 賓 3) shows:
癸亥卜，古貞: 
年自㘡（上甲）至于多毓 。九月。
The day guihai cracks, Gu tested the oracular proposition, praying for the harvest (by offering sacrifices) to Shang Jia and his numerous descendants (made in) the ninth month.
This ancestor was not accompanied by any spouse. Therefore, he “avoided” the entanglements of Shang royal succession: the lack of spouses “neutralized” in this ancestor any entanglements with lineage-based politics.Footnote 75 Indeed, the crisis immediately before the reign of Wu Ding was linked to the four sons of the king Zu Ding 祖丁 who had four spouses: each son was by a different mother.Footnote 76
There is a further example of the ritual work done during the reign of Wu Ding, pertaining to the specific status of Shang Jia. Chao Lin quotes the example of the following sentences from the inscription 合集 34047 (period 1–2, group 歷 1):Footnote 77
• 己丑卜在小宗升歳自大乙 “The day jichou cracks, in the little temple, offering the sacrifices sheng and sui (decapitation with an axe) to the [royal ancestors] starting with Da Yi/(Tang)”;Footnote 78
• …亥卜在大宗又升伐三羌十小
自[上]甲 “The day … hai cracks, in the great temple, offering by the sacrifices you and sheng three decapitated individuals of the Qiang people and ten young penned sheep to the [ancestors] starting with Shang Jia”.Footnote 79
自[上]甲 “The day … hai cracks, in the great temple, offering by the sacrifices you and sheng three decapitated individuals of the Qiang people and ten young penned sheep to the [ancestors] starting with Shang Jia”.According to Chao Lin, the little temple was a place where the tablets of the kings could be placed for specific ceremonies dedicated to them. The ancestor mentioned as the first of the series was often Tang 湯, the first king according to received sources. By contrast, the first of the ancestors in the “great temple” was Shang Jia, who also became the head ancestor of the cyclical sacrifices implemented during the reign of Zu Jia son of Wu Ding.
Those inscriptions indicate that there were attempts to separate non-royal from royal ancestors at the end of Wu Ding's reign, before the inception of the more sophisticated system of the cyclical sacrifices, implemented after his reign.Footnote 80 In any case, it is significant that the temple containing the first (non-royal) ancestors and all the ancestors following him has been named the “great temple” in contrast to the other “small temple” containing “only” the royal ancestors.
The mention of temples is not the only domain where ritual experimentations took place during the reign of Wu Ding. Inscriptions of this reign mention sets of characters such as dashi 大示 “great tablet”, xiaoshi 小示 “lesser ancestral tablet”, and xiashi 下示 “secondary ancestral tablet”; there is also the association of shi (示 tablet) with a numeral, six, seven, ten… .Footnote 81 This type of regrouping is linked with the order of the royal succession, but since this order chronologically fluctuated and thus redefined, a posteriori, those categories of da shi 大示 and xiao shi 小示 must be seen only as an attempt to “work out” different classifications and not as indicative of the existence of iron-clad rules. Those attempts are documented during the reign of Wu Ding and the subsequent reign of Zu Geng.Footnote 82 Their importance cannot be overlooked: it prepared the implementation of a rule of unilineal transmission of royal power from father to son.Footnote 83 This rule might have been implemented after the reign of Wu Ding but for the death of the “young king” which happened before his father's demise. Instead, two of the other sons of the king succeeded him: Zu Geng, followed by Zu Jia. The implementation of the rule took place only after the reign of Kang Ding 康丁, one of the sons of Zu Jia.
The two sons of Wu Ding, who effectively became kings after him, came from a different mother; I surmise that the identity and the status of those mothers, royal spouses, had a tremendous importance. Under the reign of Wu Ding, the mention of a specific marker bears witness to the importance of those spouses in the process of giving birth to a unilineal line of succession. This marker was also employed in the inscriptions after his reign.
b. Who are my ancestors: a question of markers
The inscriptions of the reign of Wu Ding mention a specific marker of lineage, analysed first by Qiu Xigui 裘錫圭 and more recently by Liu Huan 劉桓. Liu Huan interprets the character of the marker (usually transcribed as yu 毓, to give birth to or to raise) as zhou 胄 (designating in Zhou-era texts the posterity of kings).Footnote 84 In the inscriptions of the cyclical sacrifices, there is mention of the duo yu/zhou 多毓/胄, the “numerous descendants”, in formulas type: … 自上甲至于多毓/胄 “[sacrifice X offered to] Shang Jia and his numerous descendants”. This designated the entire group of the most important ancestors, all descendants of the one who was the first ancestor, Shang Jia 上甲. It so indicated that this group was regarded as a whole, under the common reference of Shang Jia. When the character yu/zhou was applied to a singular royal ancestor, singling him or her out, the general meaning of the character changed. It became a specific marker and, to understand what was then at stake, one must examine in detail the chronology of the relevant inscriptions, the nature of the ancestors receiving the marker, and the characteristics of the reigns under which the marker has been attributed.
c. The marker yu/zhou as an identifier of direct line of descent
The marker yu/zhou appears in a context where kinship terms were classificatory and not descriptive. In other words, the same term was applied to all the individuals of the same sex in the same generation. I present here the essential results of Qiu Xigui and Liu Huan, adding generational precisions for each reign.
As Liu Huan noted (p. 65), only royal ancestors who in the sacrificial cycles were accompanied by spouses were given this quality. It is also the case for royal spouses, with a nuance: not all spouses of those kings received this marker. During the reign of Wu Ding 武丁, and in the present stage of the documentation, no male ancestor received the marker individually. The individual marker yu/zhou was attributed only to Bi Ji 妣己 (yu/ zhou Bi Ji 毓/胄妣己), one of the spouses of Zu Ding 祖丁.Footnote 85 This ancestor had four sons succeeding him: Yang Jia, Pan Geng, Xiao Xin, and Xiao Yi. Zu Ding had more than one spouse but only Bi Ji was called yu/zhou Bi Ji (毓/胄) 妣己. Through this marker, Wu Ding distinguished among the spouses of Zu Ding the one from whom his own lineage emerged: Liu Huan concludes that this spouse was the mother of Xiao Yi, his own father. I concur: by attributing the marker, Wu Ding “singled out” his own grandmother. The marker was attributed at the level EGO+2.
The marker yu/zhou has also been attributed after Wu Ding, during a limited number of reigns, to the following ancestors and spouses:Footnote 86
• yu/zhou Zu Yi 毓/胄祖乙= Xiao Yi 小乙, marker attributed to the level EGO+2 in the inscriptions of the second period (reigns of Zu Geng/Zu Jia);Footnote 87
• yu/zhou Zu Ding 毓/胄祖丁 = Wu Ding 武丁, marker attributed to the level EGO+2 in the inscriptions of the third period (reigns of Lin Xin 廪辛/Kang Ding 康丁, most probably Kang Ding);Footnote 88
• yu/zhou Fu Ding 毓/胄父丁 (Xilie p. 63, 屯南 629 + 647). Since those inscriptions are classified in the 歷無類 group (reign of Zu Jia 祖甲), this ancestor is Wu Ding 武丁. The marker is attributed at the level EGO+1.
• There is, in inscription 屯南 3186, a record of sacrifices offered to yu/zhou Zu Bi Geng毓胄祖妣庚.Footnote 89 This inscription indeed mentions a royal spouse (Bi Geng), but the king is not identified: following the date of the inscription and the EGO+2 rule observed in most other cases, it should be Xiao Yi 小乙, the father of Wu Ding 武丁. The relevant portion of the inscription should be read 毓祖[乙][毓]妣庚. The celestial stem name of the king has been omitted as well as the second yu/zhou, the association of an ancestor with his spouse being enough to identify him with certainty. Therefore, it is either Zu Geng 祖庚 or Zu Jia 祖甲 who gave the marker yu/zhou to their grandmother, Bi Geng, the mother of their father Wu Ding 武丁.
• yu/zhou Bi Xin (毓/胄) 妣辛, marker attributed to the level EGO+2 in the inscriptions of the reign of Kang Ding 康丁.Footnote 90 Bi Xin, one of the spouses of Wu Ding 武丁, the famous Lady Hao 婦好, was one of the three spouses of Wu Ding honoured in the later iterations of the cyclical sacrifices.Footnote 91 Based on the EGO+2 rule and since the inscriptions dates from the reign of Kang Ding, the marker yu/zhou identifies Lady Hao as the grandmother of Kang Ding and, indirectly, the mother of Zu Jia 祖甲, father of Kang Ding.
Liu Huan (p. 66) interprets correctly the role of the marker yu/zhou as a device through which a king identified his own direct line of descent. This line of descent was through male and female, but in Figure 2 I introduce a nuance with regard to the level at which the marker was given.
The three kings, and their spouses who have received the marker yu/zhou, form a direct transmission line to the exclusion of other spouses.
Figure 2. The line Zu Ding – Wu Ding and the presence/absence of markers
During the reigns of Zu Geng 祖庚 and Zu Jia 祖甲, Xiao Yi 小乙, the father of Wu Ding, received the marker at the EGO+2 level. The reign of Zu Jia saw an exception to the rule: Zu Jia gave the marker to his own father. There are, to date, only two inscriptions with the association of the marker and “Father Ding” and constitute an anomaly only if the character yu/zhou is mainly a marker at the level EGO+2. I do not deny that in most cases it functions indeed as a seniority marker. Nevertheless, all of the existing cases and the two inscriptions of Xiaotun (屯南 629 + 647), dated from the reign of the son of Wu Ding (level EGO+1), show that it has another characteristic: it also distinguishes qualitatively between ancestors. Zu Jia also honoured specially the only spouse associated in the sacrificial cycles to Xiao Yi; Bi Geng.
During the reign of Kang Ding 康丁, two ancestors, Wu Ding and Bi Xin, received the marker. Both were honoured at the level EGO+2 and were husband and wife.
The different cases involving the marker given individually at the level EGO+2 can be classified thus:
• Individual spouse: Bi Ji.Footnote 92
• Ancestors identified as a couple: Xiao Yi and Bi Geng; Wu Ding and Bi Xin.
• Father: Wu Ding.
Are all these cases alike? Wu Ding was singled out by one of his sons, Zu Jia. According to the received sources, the reign of Zu Jia was marked by a crisis.Footnote 93 Following this crisis, I think that he had to reaffirm his legitimate status by recalling his direct link to his father. Wu Ding himself singled out Bi Ji, through whom he affirmed the legitimacy of his own father: Bi Ji was most probably the mother of Xiao Yi. As Cai Zhemao showed, Wu Ding emerged victorious from a power struggle pitting himself against the descendants of several kings who were his paternal uncles.Footnote 94 Identifying the mother of his father, one of the four spouses of the king Zu Ding, allowed him to identify his own line of descent.
Such a struggle also took place after his reign: his two sons, Zu Geng and Zu Jia, also singled out their grandparents, and the logic of such a move does seem to be linked to their own power struggle. At least during the reign of Zu Jia, under whom the first sacrificial cycle was implemented, the reference to Bi Geng, the spouse of Xiao Yi, would have been important: while in the list of male ancestors in this first cycle, Zu Geng, brother of Zu Jia, was the last king honoured, this Bi Geng was the last royal spouse mentioned in the cycles: none of Wu Ding's spouses were mentioned.
The next case, that of Bi Xin 妣辛 = Lady Hao 婦好, is quite straightforward: Kang Ding, by singling out both Wu Ding and Lady Hao, established the legitimacy of his own lineage. He was the successor of a direct line initiated with Wu Ding and Lady Hao. Therefore, his elder brothers, Zu Ji and Zu Geng, with their respective mothers Bi Wu and Bi Gui, were sidelined.
The marker yu/zhou, when considered qualitatively, is interpreted by Liu Huan as a marker of legitimacy. The ancestors receiving the marker, either at the level EGO+2 (in the majority of the cases) or EGO+1 (in the case of Wu Ding) were those whose son (or sons) succeeded them. The marker yu/zhou designated and distinguished kings who in the cyclical sacrifices were accompanied by a spouse, but also among those spouses and in the case where there was more than one spouse to a given king, only one of them. The example of Wu Ding and his three spouses honoured in the cycles is quite obvious: Kang Ding distinguished, among those spouses, the mother of his own father, Lady Hao, and therefore acknowledged the line from which he emerged directly. Therefore, I deduce that Kang Ding “constructed” (that is to say retrospectively) a direct line of transmission of royal power, as shown in Figure 3.
Figure 3. The direct line Wu Ding – Kang Ding
Hence, yu/zhou in fact functioned as an indicator of direct transmission of royal power, from father to son.
The singling out of one, and only one, royal spouse must be interpreted as a selective instrument that was necessitated by the existence of non-monogamous unions. Therefore, in cases where it was bestowed on one individual ancestor, I translate the expression yu/zhou as “the ancestor X, from whom I descend directly”. The yu/zhou method of singling out a direct line of descent, starting during the reign of Wu Ding, was paralleled with another one which is more complex to interpret but was probably linked to a “streamlining” of the royal lineage.
d. Agnatic and uterine descent
One marker, duosheng 多生, has been observed in inscriptions dating from the reigns of Zu Jia and Kang Ding.Footnote 95 For Chao Lin, the term duosheng 多生 would designate a child by his/her maternal ancestry, the character zi 子 being used to connote the paternal ancestry.Footnote 96 According to Chao Lin, each member of the higher order lineage Zi had a double system of descent: agnatic and uterine, this last identity marked by the character sheng, 生 “born” (of a mother).Footnote 97 Chao Lin adds that, while there is an obvious contrast between zi and sheng, the quality of zi, as the author noted (Yinqi pp. 28–31), was not automatic and had to be conferred by the king. If this definition, according to which sheng denoted the uterine descent, can be accepted, was it necessarily applied to children whose mothers belonged to the royal sub-lineages but whose fathers did not? It might also be that the sheng 生 or the duosheng 多生 were children not conferred the title zi “prince” by the king. This operation was not necessarily directly linked to the constitution of a unilineal line of transmission but I take it as indicating at least that the intervention of the king, conferring a title or not, was essential to the process of royal succession.
The term duosheng 多生 was employed as a marker applied to living persons and was probably linked to a clearer distinction of pedigree within the royal lineage, since this title was different from the one given to the duozi 多子, the numerous princes. The other marker, yu/zhou, was clearly used to delineate a unilineal line of transmission of royal power from father to son, but it did not come to fruition immediately. The cyclical sacrifices implemented during the reign of Zu Jia did not sanction such a line: indeed, they illustrate all modes of succession, unilineal but also adelphic (brother to brother).
Conclusion
Numerous problems in the transmission of royal power in the Shang dynasty emerged from the practice of polygamous marital alliances. Those alliances were in part conditioned by the promise to male heads of the most powerful lineages (the wife-givers) that their grandsons or nephews would become kings. One of the results of this system of alliances was the adelphic rule: heirs to the king, benefitting from the help of the lineage of their mothers, became kings in succession. The potential (and real) instability of the royal transmission was provoked by the very instrument through which a given king established stability, but for his personal reign only. In other words, the adelphic system led to a vicious circle, since a son of a king had to, vis-à-vis his own brothers, establish his own powerbase by establishing as many marital alliances with powerful lineages as possible.
The king Wu Ding accessed royal power after a long period of instability concluded by the move of the capital from Huanbei to Anyang. This period was marked by the “nine generations of chaos” during which the difficulties linked to the adelphic system gave birth to a configuration I have interpreted as an attempt to keep power alternating between the two lineages initiated by Zu Xin 祖辛 and Qiang Jia 羌甲. After the death of the son of Qiang Jia, Nan Geng 南庚, and the death of his immediate successor, Yang Jia 陽甲, royal power did not revert to a son of Nan Geng but to a brother of Yang Jia, Pan Geng 盤庚, who was succeeded by two other brothers, Xiao Xin 小辛 and Xiao Yi 小乙. When Xiao Yi gave his son Wu Ding royal power, he precipitated another crisis: within the logic of this implementation of the adelphic system, it should have been inherited by one of the sons of Yang Jia, the elder brother of Xiao Yi.
The long reign of Wu Ding was quite successful. One of the means he used to consolidate his reign was the establishment of a vast network of marital alliances with non-Shang (e.g. Lady Jing) and Shang lineages (the Lady Hao is one of the most prominent). He even tried to promote a direct line of transmission father-to-son (a system that had benefitted him when he succeeded his father Xiao Yi) through the institution of an official heir, the young king, son of a non-Shang woman, Lady Jing. Nevertheless, the implementation of the direct line way of transmission of royal power was quite an uneven process.
After the death of the young king, Wu Ding was succeeded not by one son but by two: Zu Geng and Zu Jia. A passage of the Zhushu jinian suggests that Zu Jia had not been able to transmit the royal power to one son only:
二十七年，命王子囂、王子良。
The twenty-seventh year of his reign, he [Zu Jia] named the prince Xiao (= Kang Ding) and the prince Liang his heirs.Footnote 98
Zu Jia was probably doing what his father, Wu Ding, had done, with one difference. The text of the Zhushu jinian indicates that Zu Jia named two sons, contrary to Wu Ding who only designated one (the young king) as heir. If the pattern established before the reign of Zu Jia is followed, it means that those two sons were born of two different spouses. Nevertheless, the later iterations of the cyclical sacrifices only associate one spouse (Bi Wu 祖甲奭妣戊) to Zu Jia. This probably reflects the tendency of later reigns to favour a unilineal line of descent, with the elimination of the other spouses of Zu Jia and one of his reigning sons, Lin Xin.
The cyclical sacrifices first implemented during the reign of Zu Jia are a quasi-genealogical system which did not start with a king but a distant (mythical?) ancestor, Shang Jia 上甲. The examination of those cycles, the subtleties of their structure and evolution should benefit from what has been understood in this article of the difficulties and tensions in Shang royal lineages. I will endeavour to study it in detail in another article.
父壬身份的探討, Guwenzi yu Yin Shang shi 古文字與殷商史, 3, 2012, 125–48.
also named Bing 
) who was engaged in different activities (for example, warfare) at the service of the king. Cai Zhemao also quotes (“Wu Ding buci zhong Bing Fu Ren shenfen de tantao”, 142–4) an inscription from Huayuanzhuang (H 290–3) in order to suggest that the prince of Huayuanzhuang, Zai/Zu Jia, offered sacrifices to Bing. Adam Craig Schwartz (“Huayuanzhuang East 1: A study and annotated translation of the oracle bone inscriptions”, PhD, University of Chicago, 2013, 504–05) interprets this sentence as the record of a feast offered to a person named Bing. It is not certain that it is the same person.
(=帥)般; Yan Yiping interprets the character gan 甘, a mistranscription of the character shuai 帥. There are more than 30 inscriptions mentioning this name.
田夫 “The day jichou, the prince made the cracks and tested: the young king (personal name X, not deciphered yet) goes hunting to Fu (name of a place)” gives his personal name. The young king died before the end of Wu Ding's reign as the inscription 合集 39809, period 1, group dui 
small characters shows: … 㞢小王己 
“… offering one bull to the young king Ji by the sacrifice you”. The inscription also shows that the young king was indeed Zu Ji. The identification of Lady Jing as his mother is made through elimination: in the cyclical sacrifices post-Zu Jia, Wu Ding is accompanied by three spouses, Bi Xin 妣辛= Mother Xin 母辛 = Lady Hao 婦好, Bi Gui 武丁奭妣癸 = Mother Gui 母癸 (her living name remains unidentified), and Bi Wu 武丁奭妣戊 = Mother Wu 母戊 = Lady Jing 婦妌. I will show below that Lady Hao was the mother of Zu Jia. Yan Yiping (in Yin shiji, t. 1, 175), based on the inscription 合集 2580, period 1, group 賓 standard, has identified Mu Gui 母癸 as the mother of Zu Geng. Therefore, the mother of the young king must have been Lady Jing/Mother Wu 母戊. Another clue is the complete absence of sacrifices offered to Mu Wu in the inscriptions of the reigns of Zu Geng and Zu Jia. The alliance between the king Wu Ding and the Jing territory seems (cf. above, n. 58) to have entered a difficult period after his reign. It is possible that the death of the young king led to a reassessment of this alliance.
, which was the powerbase of the prince. Dong Zuobin (in “Jiagu duandai yanjiu li”, 甲骨斷代研究例, article published in 1933, republished in Dong Zuobin xiansheng quanji 董作賓先生全集, vol. 1, Taipei: Yiwen, 1977, 427–9) interpreted the character 
as an archaic form for Zai 載, the personal name of the king Zu Jia as given in the Zhushu jinian 竹書紀年. A prince of this name, son of Wu Ding, is documented in the Anyang royal inscriptions.
