Hostname: page-component-745bb68f8f-v2bm5 Total loading time: 0 Render date: 2025-02-11T20:32:24.294Z Has data issue: false hasContentIssue false

RETHINKING THE ANCESTRAL SHRINES IN THE EARLY EMPIRES

Published online by Cambridge University Press:  17 October 2022

Tian Tian
Affiliation:
Tian Tian, 田天, Peking University; email: tiant@pku.edu.cn
Zhou Wen
Affiliation:
translated by Zhou Wen 周雯.
Rights & Permissions [Opens in a new window]

Abstract

This article examines the development of early imperial ancestral shrines by exploring the Liye and Yuelu 嶽麓 Qin slips, along with other excavated texts and historical documents. It argues that Qin Shihuang's 秦始皇 court was the first to specify the regulations for the early imperial ancestral shrine, a crucial part of which was the establishment of the Taishang huang 太上皇 shrines throughout the realm, making the imperial ancestral cult part of the daily local administrative affairs. The Western Han courts largely adopted the regulations stipulated by Qin Shihuang in their commandery and kingdom shrines until late Western Han, when ritual reforms brought the imperial ancestral shrines closer to what Michael Loewe calls the Reformist vision, entailing potential conflicts between bloodlines and the hereditary order of succession. By no means did the early empires simply continue in the stipulations for the imperial ancestral shrines the royal practices of the pre-imperial period; instead, the precedents transmitted to Eastern Han reflected two major ritual reforms, with local ancestral shrines and personal participation by the emperor key subjects of debates.

提要

提要

本文探討里耶、嶽麓秦簡與其它出土、傳世文獻,重述早期帝國宗廟的發展歷程。本文認為,秦始皇創立了帝國宗廟制度,其中重要的部分之一是在縣級行政區劃設立太上皇廟,使宗廟祭祀成為地方日常行政的一部分。西漢時代的郡國廟制度承自秦始皇。西漢晚期的宗廟改革,使帝國的宗廟制度更近於儒家禮制的規定,也將血緣關係與繼承關係的衝突帶回宗廟制度中。文章結論,早期帝國的宗廟制度並非前帝國時代的延續,而是經歷了兩次變革才變為東漢以後的形態,地方宗廟與皇帝親祭即其中的核心問題。

Type
Festschrift in Honor of Michael Loewe on his 100th Birthday
Information
Early China , Volume 45 , September 2022 , pp. 167 - 201
Copyright
Copyright © The Author(s), 2022. Published by Cambridge University Press on behalf of The Society for the Study of Early China

Established no later than the Eastern Zhou (771–256),Footnote 1 the ancestral shrines were in place until the end of imperial China in 1911,Footnote 2 and vigorous debates over aspects of the ancestral shrines recurred throughout antiquity.Footnote 3 During Eastern Han (25–220) and the post-Han period, the imperial ancestral shrines were nearly always located in the capital environs, but for most of Western Han (206 b.c.e.–8 c.e.), imperial ancestral shrines were also erected in the administrative seats of the commanderies and subordinate kingdoms, in addition to the capital shrines.Footnote 4 Such local ancestral shrines, generally called “commandery and kingdom shrines” (jun guo miao 郡國廟),Footnote 5 were abolished as a result of ritual reforms in late Western Han, as noted in Michael Loewe's second book, Crisis and Conflict in Han China.Footnote 6 Scholars used to assume that a system of “commandery and kingdom shrines” was first implemented during Western Han, but newly unearthed texts show that this tradition followed precedents from the Qin empire, as Qin Shihuang 秦始皇 (r. 221–210), First Emperor of Qin, established shrines dedicated to Taishang huang 太上皇, the Grand Emperor, throughout his empire.Footnote 7

What does the discovery of local ancestral shrines mean to us today? This article attempts to answer this question. The article suggests that during the Qin and Western Han dynasties, the local ancestral shrines played a key role in the development of the ritual institutions for the imperial ancestral shrines.Footnote 8 To furnish proof of this, this essay will begin with a fresh overview of the excavated Liye 里耶 and unprovenanced Yuelu Academy 嶽麓 bamboo slips and boards from Qin, to reconstruct the local ancestral shrine rulings established by Qin Shihuang. Then, it will discuss the ancestral shrine reforms undertaken during the reign of the Second Emperor, Ershi 二世 (r. 209–207), and their probable impact. The next section will analyze how the Western Han rulers both adopted and altered the local ancestral shrine institutions they had inherited, although eventually, in late Western Han, reformers persuaded the court to de-commission the imperial ancestral shrines located in the commandery and kingdom administrative seats. Debates regarding such reforms highlight a discursive tension between competing ideas of political legitimacy, in relation to the succession to the throne and to the duties owed to the family lineage within the imperial ancestral shrines.

Qin Shihuang and the Establishment of the Imperial Ancestral Shrines

Like most kingdoms during the Eastern Zhou,Footnote 9 Qin during the Zhanguo (a.k.a. Warring States period, roughly 475–221 b.c.e.) held their important ancestral ceremonies in the ancestral shrines in their capital at Yong 雍 (present Baoji county, Shaanxi Province), and they continued to do this even after the capital was moved to Xianyang in 350.Footnote 10 Immediately after the unification under Qin, in 221, a large number of Taishang huang shrines dedicated to King Zhuangxiang 莊襄 of Qin (r. 249–247), the father of Qin Shihuang, were established in all the counties and marches (dao 道), with the latter the administrative units at the county level where great numbers of “barbarians” lived.Footnote 11 This information is revealed for the first time in the recently published Yuelu shuyuan cang Qin jian 嶽麓書院藏秦簡.Footnote 12 Seven slips among the Yuelu Qin slips relate to the ancestral shrines, although they do not form a single coherent text. I translate the slips which are relatively intact as follows:Footnote 13

泰上皇祠廟在縣、道者:

The shrines of Taishang huang are to receive cult offeringsFootnote 14 in the counties and marches (0055[2]-3/325 front)Footnote 15

The title “Taishang huang” refers to the father of Qin Shihuang, Zhuangxiang of Qin.Footnote 16 The content of this slip is consistent with that recorded on a piece of board in the Liye Qin slips, which have been published since 2002.Footnote 17 The Liye board, which was broken into four parts and later restored by scholars, mentions “traveling to inspect a shrine” (xing miao 行廟). The following translates the front side of the board:

廿六年六月壬子,遷陵□、[丞]敦狐爲令史更行廟詔:令史行⧄ I

失期。行廟者必謹視中□各自署所質日。行先道旁曹始,以坐次相屬。

On the day renzi, the sixth month of the twenty-sixth year. . .of Qianling, and Prefect's assistant Dunhu issued a regulation that a member of the Prefect's staffFootnote 18 should take turns traveling to inspect the ancestral shrine; the Prefect's staff member should inspectFootnote 19.. . .Footnote 20

. . .skipped a periodic inspection. Those who were asked to travel to inspect the ancestral shrine should personally sign the date of the inspection in a calendar after they have carefully looked inside. The order of the inspection should start from the adjacent office/department,Footnote 21 rotating in the order of the official seats. (II8–138+8–174–522+8–523)Footnote 22

When the Liye Qin slips were first published, scholars were not sure what “shrine” the slips referred to, but comparison of the Liye material with the Yuelu Qin slips makes it likely that the term refers to the Taishang huang shrines. According to the Yuelu Qin slips, the Taishang huang shrines were located in the counties and marches, which suggests that Qin's plan to establish ancestral shrines involved all regions under its rule. The Liye Qin slips provide a clear date for this: at the latest, during the sixth month of 221, the very year the Qin empire was founded, the Qin court requested local officials to inspect the local ancestral shrines on a schedule. Presumably, in the year or so following unification by Qin, local Taishang huang shrines were already being set up, for while today we have no direct evidence for multiple county-level Taishang huang shrines, the excavation site at Liye represented a relatively remote area in the Qin empire, if an important military outpost, as part of Qianling 遷陵 county (present-day Liye Township, Longshan county, Xiangxi Tujia, and Miaozu Autonomous Region).Footnote 23 Plainly, Qianling would not have been in any way exceptional in establishing its shrine for Taishang huang, the father of Shihuang; it would have been a county obeying a decree issued from the capital administration, and since there was a Taishang huang shrine in Qianling county, probably county-level Taishang huang shrines were erected in all the counties throughout the country.Footnote 24

Local officials of Qin empire were required to make regular inspections. The Yuelu Qin slips record a related regulation:

更,五日壹行廟,令史旬壹行⌞,令若丞月行廟□□□⧄。

… in shifts, travel to inspect the ancestral shrines every five days. A member of staff of the Prefect is to travel to inspect the shrines every ten days. The county Prefect or Prefect's staff member is to travel to inspect the shrines once a month.Footnote 25

(J47 front / 322 front)Footnote 26

Based on the parallel phrasing about inspections, there should have been some lower-level functionaries who were required to inspect the shrines every five days. But due to our lack of understanding about Qin local administration, as well as the fragmentary nature of the text, the position of the officials charged with this duty cannot be clearly identified.

令都Footnote 27吏有事縣道者循行之,毋過月歸(?),當繕治者輒繕治之,不□□者□□□□有不□□

Functionaries attached to the central and Metropolitan officials on official business who happens to pass by a county or a march while on official business should also inspect the [ancestral shrines].. . .Footnote 28 If the shrine is found to need repair, then repair it.. . .If not.. . .not.Footnote 29 (0327 front/326 front).Footnote 30

On the verso of the quoted Liye Qin board mentioned above is a set of dates and the names of those Prefect's staff members who had traveled to inspect the shrines, confirms the stipulation in the Yuelu Qin slips. The dates start in the eleventh month and the last entry is from the sixth month, with the full text as follows:

十一月己未,令史慶行廟。

十一月己巳,令史行廟。

十二月戊辰,令史陽行廟。

十二月己丑,令史夫行廟。

□□□□□,令史韋行。

端月丁未,令史行廟。

□□□□,令史慶行廟。

端月癸酉,令史犯行廟。

二月壬午,令史行行廟。

二月壬辰,令史莫邪行廟。

二月壬寅,令史釦行廟。

四月丙申,史戎夫行廟。

五月丙午,史釦行廟。

五月丙辰,令史上行廟。

五月乙丑,令史□□□

六月癸巳,令史除行廟。

In the eleventh month, on the jiwei day (day 12 in the sexagenary cycle),Footnote 31 the Prefect's staff member Qing traveled to inspect the ancestral shrine.

In the eleventh month, on the jisi day (day 22 in the cycle), the Prefect's staff member Si traveled to inspect the ancestral shrine.

In the twelfth month, on the wuchen day (day 2 or 4 in the cycle), the Prefect's staff member Yang traveled to inspect the ancestral shrine.

In the twelfth month, on the jichou day (day 13 in the cycle), the Prefect's staff member Fu traveled to inspect the ancestral shrine.

… . the Prefect's staff member Wei traveled to inspect.

In the first month, on the guiyou day (day 27 in the cycle), … . the

the Prefect's staff member Fan traveled to inspect the ancestral shrine.

In the second month, on the renwu day (day 7 in the cycle), the Prefect's staff member Xing traveled to inspect the ancestral shrine.

In the second month, on the renchen day (day 17 in the cycle), the Prefect's staff member Moye traveled to inspect the ancestral shrine.

In the second month, on the renyin day (day 27 in the cycle), the Prefect's staff member Kou traveled to inspect the ancestral shrine.

In the fourth month, on the bingshen day (day 22 in the cycle), the staff member Rongfu traveled to inspect the ancestral shrine.

In the fifth month, on the bingwu day (day 2 in the cycle), the staff member Kou traveled to inspect the ancestral shrine.

In the fifth month, on the bingchen day (day 12 in the cycle), the Prefect's staff member Shang traveled to inspect the ancestral shrine.

In the fifth month, on the yichou day (day 21 in the cycle), the Prefect's staff member … .Footnote 32

In the sixth month, on the guisi day (day 20 in the cycle), the Prefect's staff member Chu traveled to inspect the ancestral shrine.

(D IV8–138back+8–174back+8–522back+8–523back)

Among these records, the twelfth month, the first month, plus the second month and the fifth month constitute three records, and the interval between inspections is roughly ten days, which is consistent with the provisions transcribed on the Yuelu Qin slips. However, the slips that have already been transcribed may not constitute a full record, as the third month is missing its notation and the fourth month has only one record, which means some records apparently allude to intervals much longer than ten days. That said, the restored bamboo board is in good condition and the time span is largely continuous, suggesting the document may be relatively complete. Besides, the dates of the inspection are consistent with the calendar (zhiri 質日) of 220 in the Yuelu documents cache.Footnote 33 Conceivably, then, this irregularly timed record should be the actual record for a series of shrine inspections made on behalf of the Prefect of Qianling in the year 220. If this surmise is correct, then at least during the first year in which the regulation was issued, Qianling county generally complied with the mandates of the central administration.

The purpose of regular inspections was, at least partly, to ensure the cleanliness and good repair of the local imperial ancestral shrines, as recorded in the Yuelu Qin slips:

如下邽廟者,輒壞,更為廟便地潔清所,弗更而祠焉,皆棄市。各謹明告县道令、丞及吏主。

If the shrine has fallen down, as happened with the ancestral shrine in Xiagui county, it should be rebuilt in a cleaner place in a convenient area. If the officials hold the rituals without rebuilding the shrine, they are all to be sentenced to public execution in the marketplace. Notify all the Prefects, Prefect's assistants, and officials in charge of the counties and marches respectively, carefully and clearly.

(0624 front/ 321 front)Footnote 34

If the imperial ancestral shrines lacked careful management, they would be liable to decay, collapse, and even theft (all conditions manifesting profound disrespect, which was legally a capital crime), and this was no less the case during most or all of Western Han, as we surmise from the following case. During the reign of Han Wendi (r. 179–157), a jade offering was stolen from the shrine of Gaozu, the Han founder, and the Commissioner of Trials Zhang Shizhi 張釋之 proposed that those who stole the property of the ancestral shrine be punished by “public execution in the marketplace” (qishi 棄市).Footnote 35

Combined with the records in the Shi ji, such records allow us to reconstruct the process by which Qin Shihuang set up the imperial ancestral shrine in several steps. First of all, when Qin Shihuang was still alive, he built himself “the highest shrine” (jimiao 極廟) in his capital city of Xianyang.Footnote 36 Second, Qin Shihuang conferred upon his father, Zhuangxiang of Qin, the title of “Taishang huang,” which forced a partial reorganization of the Qin ruling clan lineage, and thus should be seen as one aspect of the Qin shrine reform. Significantly, Qin Shihuang left the royal ancestral shrines in Yong unchanged, but he began the new sequence of emperors with his father, Taishang huang.

Third, the commandery-county-march administrative system mandated by Qin Shihuang for the newly unified empire reformed the way the imperial ancestral shrines were used and cult offered during the pre-imperial era. During the Eastern Zhou period, it was always the kingdom's ruler who presided over the royal ancestral shrine sacrifices.Footnote 37 However, as seen from the previous quotation, at the local level officials were to conduct the rituals and oversee the physical state of the shrines, a situation that tallies with the contents of the Yuelu Qin slips. And when the Qin Shihuang mandated a shrine to be erected at the county and march levels, he thereby made it physically impossible for the ruler to offer sacrifices personally at all the imperial ancestral shrines.

The reform of the imperial ancestral shrine rulings was a part of Qin Shihuang's efforts to build a unified empire. Needless to say, the promulgation of many policies by Qin Shihuang was designed to impose a new type of administration upon the inhabitants residing in the areas once held by the pre-unification Six Kingdoms, the main rivals of the Qin before 221. The establishment, maintenance, and provision for regular cult offerings at the Taishang huang shrines in each county and march were expressly designed to integrate the imperial affairs into the daily routine administration of the localities, bridging the difference between center and periphery and insistently proclaiming the imperial presence in daily life of the Qin subjects. Judging from the Liye example, the imperial ancestral shrines were implemented on a vast scale along with the commandery-county administrative system, precisely so that the shrines would become an incarnation of imperial power and the legitimacy of the ruling family in the localities.

The Establishment of Seven Ancestral Shrines: The Shrine Reform of Ershi

During his very short reign of three years, the Second Emperor imitated his father as often as possible, in order to strengthen his own legitimacy. For example, he followed Qin Shihuang's example by making an inspection tour to the east; he commissioned another of the inscribed stone steles; and he continued to have his court oversee the erection of the Epang Palace, whose foundations had been barely laid during Qin Shihuang's lifetime.Footnote 38 In addition, by an edict issued by the Second Emperor in his first year of rule (209), the Second Emperor reinforced the unification of the standard weights and measures, decreed by Qin Shihuang in 221.Footnote 39 Only in the case of the imperial ancestral shrine rulings did the Second Emperor diverge from, and possibly defy the intentions of his father.

According to the Shi ji, Ershi's edict on the ancestral shrines was the first edict issued by the Second Emperor upon his accession to the throne; the edict asked his officials to discuss how best to honor the shrine of Qin Shihuang. Among the suggestions in relation to the imperial ancestral shrine were the following:

1. 天子儀當獨奉酌祠始皇廟。

According to the ritual for the Son of Heaven, he alone should personally offer sacrifices at the shrine of Qin Shihuang.

2. 自襄公已下軼毀,所置凡七廟,群臣以禮進祠。

The shrines from the time of Duke Xiang (r. 777–766) on [literally, “below”] should be gradually disestablished. Leave only seven shrines in toto. Have the various officials offer sacrifices, according to the stipulated rituals.Footnote 40

3. 以尊始皇廟为帝者祖廟。

The shrine of Qin Shihuang should be honored as the “founder's shrine” for all his successors.”Footnote 41

Conceivably, Ershi and some of his officials had in mind a classicizing set phrase “seven ancestral shrines of the Son of Heaven” (tianzi qimiao 天子七廟), which associates the number seven with the ruler, but that is far from certain.Footnote 42 And due to the limited sources at our disposal for Ershi's reign, scholars have generally ignored Ershi's reform of the imperial ancestral shrine.Footnote 43 The discovery of the “county and march shrines” introduced by Qin Shihuang focuses our attention on the imperial ancestral shrine reform at the beginning of Ershi's reign.

The “Seven Shrines” system that the Second Emperor tried to put in place required the demolition of a large number of ancestral shrines established by the previous kings and by Qin Shihuang, Ershi's father. In pre-imperial times, so far as we know, the kingdoms made no provision to decommission ancestral shrines gradually. Indeed, the Shi ji plainly stated, “the imperial ancestral shrines of the previous kings were either in Xi and Yong or in Xianyang” 先王廟或在西、雍,或在咸陽.Footnote 44 The Qin clan had enjoyed noble patent, since early Eastern Zhou under Duke Xiang, so if the Second Emperor followed the suggestion to only acknowledge Seven Shrines, he would have to personally select which shrines would be disestablished. In texts of Qin and Han, whenever the words “above so-and-so” (i.e., from earlier times down to so-and-so) and “below so-and-so” (i.e., from the time of so-and-so on) are used, the texts usually stipulate the class of things to be enumerated, as many examples in the Shi ji attest.Footnote 45 The saying that “below Duke Xiang,” all the shrines should be gradually decommissioned should be taken to mean the demolition of all the royal ancestral shrines, including that of Duke Xiang of Qin. Therefore, the Second Emperor's shrine reform was radical.

Moreover, in his shrine reform, the Second Emperor paid extra respect to the shrine of Qin Shihuang, perhaps because, as rumor had it, Ershi had connived at the death of the rightful heir to the First Emperor: the shrine of Qin Shihuang is the only shrine where the emperor personally offered cult. At the same time, the shrine of Qin Shihuang was singled out by a special name as the “founder's shrine” for all subsequent emperors in the Qin ruling line (di zhe zumiao 帝者祖廟). Qin Shihuang was elevated above all other Qin rulers, past and future, due to his status as founder of the imperial fortunes, and even among the “Seven Shrines,” his shrine was to be in a different location than the royal shrines dedicated to the previous Qin kings. From the name of “founder's shrine” we may infer, too, that, after several generations, the ancestral royal shrines of the old Qin rulers might be abolished and, if so, the “founder's shrine” would then become the sole, as well as the highest, ancestral shrine sanctioned by the Qin empire, which was never to be abolished.

The Taishang huang shrines, to which Qin Shihuang attached great importance, are not mentioned in this discussion, nor do the sources at our disposal say whether the Second Emperor established widespread county and march shrines for Qin Shihuang. The core feature of the strict “Seven Shrines” design is that the number seven should be the upper limit for the number of ancestral shrines at the capital,Footnote 46 even if the number of local shrines could expand over time. Possibly, then, the Second Emperor did not demolish the local Taishang huang shrines, and it is equally possible that Ershi abandoned the idea of continuing to build local imperial ancestral shrines for Qin Shihuang, since building imperial ancestral shrines was fundamentally at odds with the “Seven Shrines” proposal.

Still, the shrine reform carried out by the Second Emperor had a considerable impact, we learn. The “Imperial Edict, first year of Ershi” (209) (Qin Ershi yuannian zhaoshu 秦二世元年詔書) excavated in recent years at the Tuzishan 兔子山 site in Yiyang 益陽 (Heshan District, Yiyang City, Hunan Province) begins with the following statement:

朕奉遺詔,今宗廟事及著以明至治大功德者具矣,律令當除定者畢矣。

We have upheld the testamentary edict left by Qin Shihuang [to assume the throne]. Now, the matters relating to the imperial ancestral shrine and the advertisement of Qin Shihuang's great merits and virtues are all in place, and we have completed the processes of getting rid of and fixing, as specified by the Statutes and Ordinances.Footnote 47

That this edict mentions the shrine affairs before the laws in suggests that Ershi saw the shrine reform as of great importance.Footnote 48 We would add to this, two quotations in the Shi ji of Jia Yi's 賈誼 (200–167) “On Faulting the Qin” (Guo Qin Lun 過秦論) that deal with the imperial ancestral shrines:Footnote 49

1. 一夫作難而七廟墮,身死人手,爲天下笑者。

When a fellow fomented rebellion, all the Seven Shrines collapsed, and the ruler assassinated, so that it [Qin] became the object of ridicule throughout the whole realm.

When Jia Yi wrote his essay “Faulting the Qin” in early Western Han, no Seven Shrines rule had as yet been implemented, so Jia Yi seems to have been referring not to a total number of generic shrines, but to the Qin dynasty.Footnote 50

2. (二世)重之以無道,壞宗廟,與民更始,作阿房宮,繁刑嚴誅,吏治刻深。

[Ershi] carried out twice as many immoral acts: he abolishedFootnote 51 the ancestral shrines, and he made a new compact with the people; he built the Epang Palace. His punishments were numerous and harsh, and his officials were far too strict in their administration.Footnote 52

Jia Yi lists many evil policies of the Second Emperor, the first of which appears to be a change to the old traditions for the ancestral shrines.Footnote 53 This is consistent with the “Imperial Edict, first year of Ershi” unearthed at Tuzishan, as cited above.

To recapitulate: both Qin emperors paid particular attention to the reform of the Qin ancestral shrines. Although the directions of their reforms were diametrically opposed, the two emperors had two things in common: both tried to distinguish the Qin empire from the Qin state by creating a new imperial shrine sequence, and both were aware of the close relationship between the ancestral shrines and the legitimation of imperial order. Understanding the purposes and differences of their reforms makes it possible to better assess the Western Han ancestral shrines in their turn.

Following Qin Shihuang: Imperial Ancestral Shrines in Early Western Han

In 201, Liu Bang 劉邦, a.k.a. Han Gaozu (r. 202–195), honored his father Taigong 太公 as Taishang huang.Footnote 54 In the month after Taigong's death in 197, Liu Bang ordered the subordinate kingdoms to set up shrines in honor of Taishang huang in their capitals.Footnote 55 Notably, Liu Bang's Taishang huang shrines were to be set up in some areas outside of the Han court's direct control (see below).Footnote 56 All this appears to show the Han founder trying hard to imitate Qin Shihuang.Footnote 57

However, in some respects the Taishang huang shrines in the early years of Western Han differed from those of the Qin dynasty. First of all, the spatial distribution of the shrines established by Liu Bang was limited to the capitals of the subordinate kingdoms, whose number was much smaller than the total number for the Qin counties and marches. By the eighth month of 197, near the end of the Liu Bang's reign, there remained but nine subordinate states and eight kingdoms,Footnote 58 four rulers of whom did not bear the imperial Liu-clan name. During the first seventy years or so of Western Han, from the reign of Liu Bang to that of Han Wendi, the subordinate kingdoms proved to be so independent that the historical texts often juxtapose their powers and privileges with those of Han.Footnote 59 Secondly, far more than the Qin, the Han needed impressive rituals to confirm its legitimacy, since Liu Bang, the Han founder had been a commoner, so the kings tended to regard him more as primus inter pares than as absolute ruler. This situation became clear soon after Liu Bang's death in 195, for his crown prince, the future Huidi (r. 194–188), on the very day after the funeral went to the Taishang huang shrine in Chang'an to formally assume the title of emperor, at which point the assembled officials conferred upon Liu Bang the honorific title of “Highest Emperor” (gao huangdi 高皇帝).Footnote 60 All the ceremonies performed at this one site were expressly designed to underscore the superior legitimacy of the Han throne over that of the rival kingdoms.Footnote 61 Unlike Qin Shihuang, Liu Bang, the founder of the empire, had had no ancestral shrine erected for his father before him.

Liu Bang and his successors clearly reflected on the reasons for Qin's rapid demise and the Second Emperor's reform. Jia Yi's harsh denunciation of Ershi for destroying the ancestral shrines broadly reflected the elite consensus in early Western Han. The Second Emperor's Seven Shrines rule was not only considered a failure, but a major factor in Qin's rapid collapse, and therefore it had to be abandoned altogether. Qin Shihuang's establishment of Taishang huang shrines throughout his territories was a major innovation that erstwhile local officials like Liu Bang would have recalled, particularly as the local officials were required to “inspect the ancestral shrines” on a regular schedule. The Han courts reckoned it was far better to revert to Qin Shihuang's path and abandon the Second Emperor's approach, when it came to the imperial ancestral shrines.Footnote 62 Accordingly, after Liu Bang's death, Huidi ordered ancestral shrines dedicated to Gaozu to be set up in all the administrative seats of the commentaries and kingdoms.Footnote 63 (Perhaps the Han court lacked the resources to establish shrines at the county level, as had been the case with Qin's Taishang huang shrines.)

For the rest of Western Han, always the highest status among the imperial ancestral shrines was given to Gaozu. In 156, his shrine was consecrated as the “temple of the imperial line's supreme ancestor” (di zhe taizu zhi miao 帝者太祖之廟).Footnote 64 Note the similarity between this temple name and that of the “founder's shrine” the Second Emperor conferred upon the shrine of Qin Shihuang. From Huidi onward, all new Han emperors, during their accession ceremonies, solemnly paid their respects at the shrine of Gaozu, in order to gain the Han founder's approval and blessings.Footnote 65 When Huo Guang 霍光 (d. 68) engineered the deposal of Liu He 劉賀 (d. 59), he said, for example,

宗廟重於君,陛下未見命高廟,不可以承天序,奉祖宗廟,子萬姓,當廢。

The ancestral temple is of greater weight than any individual ruler. His present Majesty never made an appearance to receive his orders at the founder's temple, and for that reason, he cannot ascend the throne of the imperial line, offer sacrifices in the ancestral temples, or treat as his sons the masses under his rule. He must be deposed.Footnote 66

Similarly, the installation of an empress also took place at the shrine of Gaozu,Footnote 67 as did all other important ceremonies, including the imperial heir's “coming-of-age ceremony” (jia yuanfu 加元服) and ceremonial changes to the regnal title (geng ming 更名).Footnote 68 Important archives were also placed within in the Gaozu shrine complex, so that successive generations would have access to the carefully transcribed records.Footnote 69 Leading up to the usurpation by Wang Mang 王莽 (45 b.c.e.–23 c.e., r. 9–23 c.e), Wang reportedly manipulated the supreme authority of the Gaozu shrine for his own political advantage. For example, he imitated other Western Han emperors by going to the shrine of Gaozu to accept the throne after the abdication by the current Han throne-holder.Footnote 70 When his regime became unstable, Wang traveled to the shrine of Gaozu, in an effort to suppress the angry spirits of the Han imperial ancestors, whom he believed to be collectively opposing him.Footnote 71 The foregoing all suggests that during Western Han the function and status of the founder's shrine conformed to the basic design of the Qin emperors for the ancestral shrine dedicated to the spirit of Qin Shihuang.

Local Ancestral Shrines during the Early Empire

In Western Han, the “commandery and kingdom shrines” referred to the shrines of Han Taishang huang, plus the shrines of Gaozu, Wendi, Wudi, and two “re-established shrines” (yuan miao 原廟) dedicated to Gaozu and established by Huidi, one of which was located in the greater metropolitan area of the Han capital of Chang'an, north of the Wei 渭 River,Footnote 72 and the second, far from Chang'an, in Pei 沛 county, Gaozu's old home town.Footnote 73 When the two shrines were re-established in c. 190, Shusun Tong 叔孫通 (d. c. 188) said:

願陛下為原廟渭北, … 益廣多宗廟,大孝之本也。

I would have Your Majesty rebuild a shrine [dedicated to Gaozu] on the north side of the Wei River. . .to further increase the number of ancestral shrines and expand the foundations of filial piety.Footnote 74

Shusun Tong's statement need not have been based on any of the so-called ‘Confucian’ Classics, as he was simply following a trend. From the tone of this remark, it appears that initially, in early Western Han, the establishment of ancestral shrines was anything but systematic. Later local ancestral shrines were established according to the administrative divisions, with no more ancestral shrines built at random. Han Jingdi (r. 156–141), to honor the achievements of his father Wendi, established temples dedicated to Wendi in all the commanderies and kingdoms, following the model established at the Gaozu shrines.Footnote 75 Han Xuandi (r. 73–49) later set up shrines to Wudi in the commanderies and kingdoms that Wudi had visited during his imperial progresses.Footnote 76 With the establishment of the commandery and kingdom shrines, the “temple names” (miao hao 廟號) of the emperors were also decided, so that Gaozu (Founder) became Taizu 太祖, while Wendi and Wudi became Taizong 太宗 and Shizong 世宗, respectively.Footnote 77 By the end of Western Han there were altogether 167 imperial ancestral shrines in the localities spread out over sixty-eight commanderies and kingdoms, including the Taishang huang shrines erected to Han Gaozu's father.Footnote 78

What is the relationship between the commandery and kingdom shrines in the localities and imperial ancestral shrines in the capital, Chang'an? I believe that the commandery and kingdoms shrines in Western Han can be seen as replicas of the Chang'an imperial ancestral shrines (possibly smaller), except that with them there was no expectation that emperors offer sacrifices in person, in the localities. The evidence at our command is this: when the commandery and kingdom imperial shrines were in danger, the Han emperors regarded the impending crisis with equal gravity. Having gained the upper hand during the Rebellion of the Seven States in 154, Jingdi issued an imperial edict accusing Liu Ang 劉卬 (d. 154), King of Jiaoxi 膠西, in the following language:

今卬等又重逆無道,燒宗廟,鹵御物,朕甚痛之。朕素服避正殿,將軍其勸士大夫擊反虜。

Now, Liu Ang, among others, has repeatedly engaged in acts of great treason; he has torched the imperial ancestral shrines [outside the capital] and seized goods belonging to the throne. So enormous is Our grief that such acts have occasioned that We will don the plain clothes of a mourner and avoid holding the morning court council session in the main audience hall. May my generals encourage their men to attack and overturn the rebels.Footnote 79

During the Seven Kingdoms’ Rebellion, Wu and Chu led rebel armies to fight against the Han, but since the rebels never advanced as far north as the Hangu 函谷 Pass, the ancestral shrines to which Jingdi referred in his edict must refer to the local ancestral shrines far from the capital at Chang'an. Similarly, in 70, a great earthquake shook forty-nine commanderies and kingdoms along the eastern seacoast and destroyed the local imperial ancestral shrines in the two commanderies of Beihai 北海 and Langya 瑯琊. Forthwith, Xuandi issued an imperial edict of amnesty, which mentioned the collapse of the ancestral shrines (zongmiao duo 宗廟墮) in language that did not distinguish those local ancestral shrines from the capital shrines. In his edict, Xuandi announced that he, too, would assume mourning and avoid holding court for five days (sufu bi zhengdian wuri 素服避正殿五日).Footnote 80 This form of ritual practice was plainly designed to express imperial anguish, as it often came in response to major calamities or crises.Footnote 81 Thus, the way that Wendi and Xuandi reacted to the destruction of the commandery and kingdom ancestral shrines was no different from the way they responded to any damage of the Chang'an shrines, albeit slightly stronger.Footnote 82 Besides, the term yu wu 御物 (imperial items), used in the Shi ji entry, refers specifically to the objects belonging to the Son of Heaven. Although the shrine was located in the commanderies and kingdoms, the shrine buildings and their interior furnishings still belonged to the emperor and his ruling line, because the imperial his privy purse paid for their building and upkeep. Secondly, the rituals held at the local imperial shrines were also very similar to those of the Chang'an ancestral shrines. A large number of bamboo slips were discovered in 2016 at the unlooted tomb of Liu He.Footnote 83 One set of slips records the way that Liu He offered imperial ancestral sacrifices at his local shrine when he was King of Changyi 昌邑, in the northeast, prompting the conclusion that the ceremonies offered by a local king to the imperial ancestors were similar to the capital rituals performed by the emperors in their ancestral shrines, as detailed in the “Treatise on Rites and Music” in Han shu.Footnote 84

With two reasons supplied for believing that the commandery and kingdom imperial ancestral shrines enjoyed some degree of parity, a follow-up question surfaces: Did the imperial ancestral shrines retain throughout Western Han the same paramount influence that they had had in Liu Bang's era? To answer this question adequately, it may be necessary to return to an even more fundamental question: What, after all, is an “ancestral shrine”? As a place of ancestor worship reserved for families belonging to the governing elite prior to unification, the “ancestral shrine” simultaneously promoted at least three core elements: ancestor worship as a solemn duty, celebration of the patriline, and the patriarchal clan system. By the Han compilation entitled Record of Rites (Li ji 禮記), commoners had no ancestral shrines at all; to have an ancestral shrine was a signal privilege granted solely to the king and his high-ranking officers, whose number was determined by rank.Footnote 85 Supposedly, too, by the retrojections in the Li ji systematizing account, the pre-imperial ancestral shrine was inseparable from the feudal and patriarchal clan system.Footnote 86 Indeed, the core of the ritual “order” was said to derive ultimately from the patriarchal clan system, for it determined the order in which the ancestors’ tablets were arrayed, the level of sacrifices each ancestor enjoyed, and the person who presided over each ceremony. Today's Zuo zhuan 左傳 asserts that the rituals ascribed to the royal Zhou ancestral shrines in the pre-unification period were first and foremost related to the kinship order:

凡諸侯之喪,異姓臨於外,同姓於宗廟,同宗於祖廟,同族於禰廟。

In the case of the mourning for all vassal lords, if the deceased was of a different clan, the lord would lament his passing outside the city wall. If he was of the same clan, the lord would lament his passing at the Ancestral Temple. If he was of the same line, the lord would lament his passing at the founding ruler's temple. And if he was of the same royal house, the lord would lament his passing at his father's temple.Footnote 87

Besides, the Li ji has Confucius say, “The rituals of the ancestral shrines are the means to arrange the zhaomu” (宗廟之禮,所以序昭穆也).Footnote 88 It was the basic premise of classical learning during the early empires that the monarch, as the patriarch of his clan, was to personally offer sacrifices in a shrine where the ancestral tablets were arranged in the appropriate order.

The county and march shrines created by Qin Shihuang still employed the name “shrine,” but, in fact, it redefined the meaning of “ancestral shrine.” After all, Qin Shihuang completely abolished the feudal kingdoms, so that imperial sons, aside from the imperial heirs, no longer could enjoy independent or semi-independent domains and therefore, in some sense, the old patriarchal clan had ceased to exist, which destroyed one of the three legs of the stable pre-unification royal ancestral shrines.Footnote 89 Moreover, within the framework of the county and march shrines erected to the Qin emperor, each Qin emperor, in theory, would have shrines dedicated to him all over the country upon his death, which eliminated the problem of deciding the proper number of shrines, not to mention the proper sequence implied by the zhaomu. By Qin Shihuang's alterations, local officials who had no blood relationship with the emperor were to preside over the rituals in the county and march shrines. Thus, the ancestral shrines designed by Qin Shihuang for his empire can be viewed simultaneously as an extreme expansion of the earlier meaning of ancestral shrine and also a complete reversal and transformation of it.

Due to the imperial centralization of powers, the local imperial ancestral shrines from the reign of Jingdi (r. 157–141) on, gradually lost the significance they had had earlier in the localities during Qin and early Western Han; instead, the local shrines became an institution dedicated to commemorating the imperial virtues and merits.Footnote 90 When Wendi died, he left an edict asking his subjects to simplify the mourning rituals in his honor, partly by shortening the traditional mourning period for him.Footnote 91 However, as soon as Jingdi assumed the throne, he issued an edict asking his officials to discuss the title of Wendi, and the final decision was:

世功莫大於高皇帝,德莫盛於孝文皇帝。

Through the ages, there can be no merit greater than that of the Highest [Founding] Emperor, and no greater virtue than that of the filial Emperor Wen.Footnote 92

Wendi's merits apparently qualified him to have imperial cult offered to him in the commandery and kingdom shrines. As readers will recall, Wendi and his predecessor, Huidi, were brothers. By the pre-imperial customs, as reconstructed by the Li ji, brothers should be ranked on the same zhaomu side, a situation that could create difficulties when determining the proper zhaomu sequence.Footnote 93 By establishing commandery and kingdom shrines dedicated to Wendi, Jingdi's intention may have been to confer a higher status upon Wendi in the ancestral shrines than upon Huidi.

In any case, Xuandi's establishment of the commandery and kingdom shrines dedicated to Wudi may have been designed to arrange his own place within the shrines order. Soon after his accession to the throne, Xuandi considered elevating the lineage position of his grandfather, Liu Ju 劉據 (128–91), as Liu Ju had once had the honor of being crown prince for Wudi. In 73, Xuandi issued an edict “to discuss his [Liu Ju's] posthumous title and allocate a cemetery and mausoleum town for his burial” (yi shi, zhi yuan yi 議謚,置園邑).Footnote 94 However, Xuandi's request was rejected by a group of officials (probably led by Huo Guang), who quoted a line from the Gongyang Commentary on the Annals: “a person's successor acts as a son for him” (wei ren hou zhe wei zhi zi 為人後者為之子);Footnote 95 by this rationale, Xuandi was the successor of Zhaodi, and therefore he should not offer imperial sacrifices to his own birth parents or birth grandparents. Perhaps to appease Xuandi's feelings, those very officials proposed that his father be given the honorary title of Shi Huangsun 史皇孫, and his mother, the posthumous title of “Grieving (dao 悼), while a mortuary park was established for them, following the specifications for the Han kings, with the taxes from three hundred households dedicated for the upkeep. Xuandi's grandfather, Liu Ju, the former crown prince, was only to be given a posthumous title of “Unrepentant” (li 戾), with two hundred families to “serve” him.Footnote 96 This was the first time in a long time that an emperor and his court during Western Han had disagreed on such a fundamental question of kinship and succession.

A year later (72), Xuandi announced that Wudi was henceforth to be honored as Shizong 世宗 (Generational Ancestor), and the shrines dedicated to Wudi were established in all the commanderies and kingdoms he had ever visited.Footnote 97 Considering that Xuandi had first proposed to confer a posthumous title for Liu Ju, it is possible that he saw the establishment of the shrine of Shizong as his next best option; by the logic of the time, Xuandi, having failed to find a place in the imperial ancestral shrines for his own grandfather, went back to his great-grandfather, Wudi, in order to underscore his own legitimacy. In addition, Xuandi faced a similar problem as that of Wendi in terms of succession. Xuandi was the successor of Zhaodi. However, in terms of kinship, Xuandi was grandson of Zhaodi's elder brother, Liu Ju, who was not, in theory, of the same lineage as Zhaodi. Additionally, both belonged to the zhao sequence in the zhaomu order.Footnote 98 Thus a likely factor in Xuandi's attribution of extraordinary achievements to Wudi was Xuandi's anxiety to establish his own legitimacy.

Over time, from Wendi to Xuandi, the functions of the commandery and kingdom shrines gradually diverged from the rationales given for the same shrines during Qin and early Western Han, insofar as the occasions for building shrines became times to evaluate the merits of individual emperors and to confirm their positions within the imperial lineage.Footnote 99 Meanwhile, the zhaomu arrangement, which purportedly created problems during the pre-unification period, was gradually “restored.” The imperial shrines continued to be maintained throughout the realm, and the Chang'an ancestral shrines never adopted the Eastern Han solution (to eventually drop some imperial ancestors from the sequence to be worshipped). That did not preclude controversy over the central issue of “order” in the ancestral shrines during the late Western Han debates over ritual reforms.

The Significance of Personally Offering Cult: Abolition of the Commandery and Kingdom Shrines

Repeated debates in late Western Han, from the reign of Han Yuandi (r. 48–33) on, led to several amendments to the imperial ancestral shrines.Footnote 100 Scholars generally compare the abolition of the commandery and kingdom shrines with the institution of the imperial ancestral shrines in Chang'an,Footnote 101 but in fact these are two distinct phases in the restructuring of the ancestral shrines, in terms of the timeline and the perspectives of the reformers and counter-reformers.

Yuandi's court, during the reign period 43–39, abolished throughout the realm 475 of the 683 imperial cults to honor local gods and heroes, at Kuang Heng's 匡衡 (d. 30 or 29) prodding.Footnote 102 In 40, it seems, this abolition of the commandery and kingdom ancestral shrines did not occasion any debate, judging from the Han shu and Han ji, and the court's decision seems to have been carried out almost immediately. This swift action with regard to the imperial ancestral shrines outside the capital contrasts sharply with the hesitation shown by successive courts in repeated cycles of abolishing and re-instating the licit cults established to honor local gods and heroes.Footnote 103 For shortly after the imperial ancestral shrines in the localities were abolished under Yuandi, many of the local hero cults were abolished only to be restored during Chengdi's reign. Then, shortly after Chengdi's death, in 5, Aidi revived the entire range of licit cult sites, whose total number of sites exceeded the number that had once been abolished.Footnote 104 The need to maintain the old cults was strongly felt whenever a potential crisis loomed, since the emperors were understandably afraid to anger the spirits, and that explains why Liu Xiang 劉向 (77–6), the great classical scholar and staunch defender of the imperial clan, to which he belonged) defended the previous iterations of the imperial cults in no uncertain tones, deeming caution to be the better part of valor.Footnote 105

So tenacious were the ruling house's precedents for the local cults for local spirits that the established cults at the imperial ancestral local shrines seemed almost laughably weak by comparison. For at least by the records of Han shu and Hanji, no official rose to protest the disestablishment of the commandery and kingdom ancestral shrines, aside from Liu clan itself. In 39, we learn, basely a year after the commandery and kingdom ancestral shrines were abolished, Yuandi, then in poor health, dreamed that his ancestors condemned him for abolishing the commandery and kingdom shrines, and his younger brother, King Xiao of Chu 楚孝王 (d. 25), purportedly had the same dream in the same year.Footnote 106 After all, many still thought of the de-commissioned commandery and kingdom shrines as places where the ancestors’ spirits had taken up residence, so naturally Yuandi and his brother, as descendants of the Liu clan, were bound to question whether they had made the right decision in abolishing the local imperial shrines. Nevertheless, the local shrines were not restored. Many of Yuandi's officials believed that the prime function of the commandery and kingdom shrines at the beginning of the dynasty had been to encourage “identification with the supreme ruler” 一民之至權, and thus securely “tie the hearts of the realm's subjects” to the throne 繫海內之心.Footnote 107 In their view, given the inherent stability of the realm, these were no longer the paramount problems confronting the court; greater by far were the financial constraints and ritual controversies triggered by the phenomenal expenses incurred in connection with the commandery and kingdom shrines.

Obviously, huge outlays were needed to sustain the frequent sacrifices and maintain the buildings, robes, and implements at the commandery and kingdom shrines. Referring to this financial burden, sometime after Yuandi's shrine reform, Kuang Heng went to pray on Yuandi's behalf at the ancestral shrines of Gaozu, Wendi and Wudi, even as Yuandi's illness persisted; in those prayers, Kuang Heng insisted that “there was no way to restore and maintain the commandery and kingdom shrines” 郡國廟無以修立, given the poor agricultural harvests in recent years.Footnote 108 Contra Kuang Heng's insistence, there existed other ways to reduce the expenditures devoted to the commandery and kingdom shrines; no ancestral shrine reform was absolutely needed, because well before the reforms of 43–39, Yuandi had already cut the number of ancestral shrine guards in the Liu-clan kingdoms by half, on the advice of Gong Yu 貢禹 (d. 44).Footnote 109 At the same time, the cost of offering those sacrifices was not so great that it could not be paid out of the imperial funds, if doing this would relieve the emperor's anxiety and allow him to worship in the old way to the spirits. This much is clear from the court's action later: because Aidi's court feared the absence of a successor to Aidi, his court decided to suddenly restore more than 700 imperial cult sites that had been abolished, which meant 37,000 sacrifices offered annuallyFootnote 110— at a rate that far outstripped that of the combined total of offerings made at the Chambers of Rest 寢Footnote 111 in the imperial mausoleum complexes and at the imperial ancestral shrines in Chang'an. So while the need to introduce economies was certainly one factor in the decisions the court made vis-à-vis the abolition of the commandery and kingdom shrines, only digging more deeply into the sources will help us find the most important underlying reasons.

As I see it, in late Western Han the central problematic of the commandery and kingdom shrines reform was the inability of the emperor to perform the rituals himself. In 40, Yuandi issued an imperial edict discussing the proposal to abolish the commandery and kingdom shrines, suggesting that while preserving the shrines would, in theory, “enable the distantly related Liu-clan members and the low-ranking [officials, perhaps commandery governors] to come together to offer sacrifices” 令疏遠卑賤共承尊祀, nonetheless “if I do not offer sacrifices myself, it is as if the sacrifice has not taken place!” 吾不與祭,如不祭, quoting an Analects line to defend his position.Footnote 112 His thinking was clear enough: it was only the Son of Heaven who should personally offer sacrifices at the imperial ancestral shrines; no others of lower-rank were fit to perform such duties on his behalf. A group of officials, led by Wei Xuancheng 韋玄成 (d. 36), then added two additional arguments in support of the emperor's position:

1. 立廟京師之居,躬親承事。

The ancestral shrine should be located in the capital and the emperor should personally conduct the sacrifices himself.

2. 《春秋》之義,父不祭祀於支庶之宅,君不祭於臣僕之家,王不祭於下土諸侯。

According to the Annals classic, a father should not be offered cult by anyone in the lateral branch of the clan outside the patriline or by a concubine's son. By analogy, a ruler should not be offered cult by his official or servant; nor a true king by his noble or vassal lord.Footnote 113

Originally, in early Western Han, the idea was that the Son of Heaven himself would personally attend the Chang'an imperial ancestral shrines, and the Liu-clan kings were to send envoys to Chang'an to attend the ceremonies.Footnote 114 Eventually, as the territories of the empire expanded, the number of commandery and kingdom shrines became so large that it was impossible for the emperor to personally offer the sacrifices himself. Yuandi and his ministers understood this crucial point, when they tied the imperial ancestral shrines to the Son of Heaven's personally offering cult at the shrines. By abolishing the commandery and kingdom imperial shrines in response, they altered the basic function of the ancestral shrines that had been accepted ever since the First Emperor of Qin. His institution of imperial shrines in the commandery and kingdom shrines had given his princes and commandery officials (probably governors)Footnote 115 the privilege of offering cult to the imperial line, in effect dispersing the strong patrilineal tie binding the ruler to the shrines.Footnote 116 By contrast, the dual insistence on the part of the late Western Han reformers on the sanctity of the capital and the necessity for the emperor's personal participation in the sacred ritesFootnote 117 made it nearly inevitable that the commandery and kingdom shrines would be eliminated, as would some sites associated with the imperial mausoleum complexes (e.g., the Chambers of Rest) (see below). The reformer's same insistence also provoked debates about the number of imperial ancestral shrines to be sponsored at Chang'an. As reimagined by these reformers, local or capital officials would no longer be officiating at the rituals at the imperial ancestral shrines, because the ancestral shrines belonged to “the one [ruling] family,” and the blessings they conferred were the sole possession of the imperial clan.

Rites and Relatives: The Rise and Fall of Ancestral Shrines in Chang'an

Only a month or so after the abolition of the commandery and kingdom shrines, Yuandi discussed with his officials how to gradually de-commission most of the ancestral shrines in Chang'an. This second phase of the reform had two focal points of debate. The first concerned the abolition of the worship sites located in the side halls, ancestral shrines, and mortuary complexes in or near Chang'an, and the second, the question whether some imperial shrines might not better be “gradually de-commissioned in their turn” (die hui 迭毀). Obviously enough, the resolution of both debates rested upon reaching consensus about the basic import of key passages within the Classics (some in contradiction with others), so as to devise good guidelines for choosing the honorific posthumous titles to be granted deceased emperors. To some degree, these were subjects taken up by the Qing evidential scholars and also in important work by Michael Loewe in his Crisis and Conflict.Footnote 118 For that reason, this essay will only sketch for readers the main storyline for the events with the help of previous secondary scholarship and passages recorded in Han shu. In the capital of Chang'an, the ancestral shrine reform was not implemented until 39, nearly two years after Yuandi initially broached the subject, because of the difficulty of reconciling the disparate arguments to the court's satisfaction.Footnote 119 Finally, during a protracted process of debate, Wei Xuancheng proposed to the court that the ancestral shrine dedicated to Yuandi's grandfather be retained, while the two imperial ancestral shrines dedicated to Taishang huang, Liu Bang's father, and to Huidi be de-commissioned. Wei's proposal won broad acceptance and was duly implemented,Footnote 120 since all aspects of Wei's proposal were based on the single coherent notion that “there should be but five [court-sponsored] ancestral shrines, with the others de-commissioned in their turn” (wu miao er diehui 五廟而迭毀).Footnote 121 Wei's proposal was upheld until Liu Xin 劉歆 (50–23 c.e.) formally proposed the implementation of his Seven Shrines theory during the reign of Aidi.Footnote 122

Perhaps more than the gradual de-commissioning of some imperial shrines, the cycles abolishing and restoring the Side Halls, Chambers of Rest, and imperial mortuary complexes reflects the subtleties of these ancestral shrine reform debates.Footnote 123 For a clearer presentation, the relevant events during Yuandi's and Chengdi's reigns have been placed in chronological sequence:Footnote 124

  1. 1. In 40, in the ninth month, there were removed from the list of court-sponsored sites the cemeteries of Zhaoling hou 昭靈后 (Liu Bang's mother), Wu'ai wang 武哀王 (Liu Bang's elder brother), Zhao'ai hou 昭哀后 Liu Bang's elder sister), Weisi hou 衛思后 (Wudi's deposed empress, Wei Zifu 衛子夫), Liu Ju and his wife (Wudi's crown prince, tried for treason, and his wife).Footnote 125

  2. 2. One month later, in the tenth month of 40, the commandery and kingdom shrines were abolished.Footnote 126

  3. 3. The following year, in 39, the court decided not to maintain at its expense the Chambers of Rest, and shrines, nor to support the mausoleum complexes dedicated to the Han founder's father, Taishang huang, or to Huidi.Footnote 127

  4. 4. Then, in the following year, in 38, the court abolished the Chambers of Rest and funerary parks dedicated to the mothers of Han Wendi and Han Zhaodi (Xiaowen taihou 孝文太后 and Xiaozhao taihou 孝昭太后), sites that escaped the earlier decrees of the preceding year.Footnote 128

  5. 5. In the sixth month of 34, the decision was made to restore the mausoleum of Liu Ju to its former status.

  6. 6. In the seventh month of 34 (i.e., in less than a month's time), the court ordered the restoration of all structures dedicated to Liu Bang's father, and the Chambers of Rest and funerary parks that enclosed the tombs of the Han founder's father, mother, elder brother, and elder sister, and Wudi's deposed Wei empress.Footnote 129

  7. 7. Nearly a year later, in the third month of 33, Chengdi's court restored the sites formerly dedicated but then de-commissioned that had belonged to Huidi and the mothers of Wendi and Zhaodi.Footnote 130

  8. 8. Only two months later, in the fifth month of 33, Chengdi's court for the second time de-funded the shrines dedicated to Liu Bang's father, to Huidi, and to Jingdi, even as it destroyed the Chambers of Rest and funerary parks dedicated to the mothers of Wendi and Zhaodi, and the Han founder's mother, elder brother, and elder sister.Footnote 131

  9. 9. Some five years later, in 28, Chengdi's court decided to restore the whole complex of sites dedicated to Liu Bang's father.Footnote 132

As this timeline shows, the court first chose to de-fund the grand cemeteries dedicated to the relatives of several emperors before abolishing the commandery and kingdom shrines. With few exceptions (Huidi, less surprising, and Jingdi, quite surprising), the imperial ancestors whose cult sites lost funding were neither emperors nor empresses, which may explain why the initial reforms encountered few objections. Because he continued to lack an heir, after 38, Yuandi (with the members of his court) came to the momentous decision to restore some worship sites. The shrine of Huidi was restored last, even after the shrines dedicated to Liu Bang's mother and father. At this remove, why this was the case is difficult to say. Presumably Huidi's lack of an heir meant that the cult site dedicated to him was considered expendable, and indeed, it was re-abolished during Chengdi's reign, never to be restored again. Upon Yuandi's death, the officials led by Kuang Heng abolished for the second time almost all the cult sites that Yuandi had first abolished but later restored; to this group, they even added the shrine dedicated to Jingdi which had not been abolished during Yuandi's reign. Apparently, Kuang Heng's group of officials was more radical than Wei Xuancheng's group had been. However, presumably for tactical reasons (to please the throne), Wei Xuancheng preserved the shrine of Huangkao 皇考 dedicated to Yuandi's grandfather, while Kuang Heng preserved the shrine dedicated to Huangkao, plus the cemeteries of Wudi's deposed empress and deposed imperial heir, as well as those dedicated to the direct male line of ancestors for Xuandi, Yuandi, and Chengdi.Footnote 133 Learning from Yuandi's indecisiveness, and anxious to reduce resistance to the abolition of shrines, Kuang Heng and the members of his group took the initiative to preserve some cult sites that did not, in point of fact, conform to classical rituals articulated in the Liji; mindful that the institution of the ancestral shrines held great emotional significance for the emperor and the members of his family, Kuang Heng was content to jettison the strictest construction of the imperial shrines articulated in one of the Rites classics.Footnote 134

Like Yuandi, Chengdi came to suffer from the lack of an imperial heir, at which point he and his court decided it was best to restore the ancestral shrines again. Why was only the shrine of Liu Bang's father restored? After all, Taishang huang was a very distant relative. By way of an answer, Han shu states that the cults of Liu Bang's mother, elder brother, and elder sister were “worshipped as before” (ru gu 如故), i.e., by the old Han precedents.Footnote 135 I suspect that since the spirit tablets of Jingdi and Huidi were arranged according to zhaomu sequence in the shrines of Taizu and Taizong (i.e., Gaozu and Wendi), restoration of the cult site dedicated to the father of the Han founder gave to each of the imperial relatives in the patriline and matriline their “proper place,” by the ideas of the time. In the turbulence occasioned by these incessant ritual reforms and counter-reforms, Chengdi still saw to it all the imperial ancestors would be duly worshipped.Footnote 136 During the reign of Yuandi, for a period of time the court had set aside a precedent established by Empress Lü in an edict mandating that a sentence of public execution in the marketplace would be meted out to anyone who “in unauthorized ways, without permission, debated any issues pertaining to the imperial ancestral shrines” (shan yi zongmiao zhe qishi 擅議宗廟者棄市).Footnote 137 But in 28, Chengdi restored the death penalty for those who discussed the ancestral shrine without express permission to do so, showing his determination to revive the old imperial ancestral shrine rituals.Footnote 138 Both Yuandi and Chengdi proved to be supporters of the classicizing ritual reforms, but the function of those shrines as protective buffer or as bane for the emperor and his ruling line could not but generate debates, once such a limited number of cult sites remained in the capital environs. No matter how radical the proposals of the ritual reformers, an emperor's feelings and judgment as head of his clan always carried considerable weight when debating the affairs relating to the imperial ancestral shrines.

Finally, under the auspices of Wang Mang, the last Han courts issued a general order about how the ancestral shrines and the emperors’ temple names were to be arranged for Western Han: Emperor Gao was to be honored as Taizu (Founder); Wendi, as Taizong (Great Ancestor); Wudi, as Shizong (Generational Ancestor);Footnote 139 Xuandi, as Zhongzong 中宗 (Mid-dynastic Ancestor) and Yuandi, as Gaozong 高宗 (High Ancestor).Footnote 140 (Note that this made a total of five imperial ancestors, who were to be commemorated in perpetuity.) The ancestral shrine order for Western Han, in turmoil for decades, was “finalized” for a very short time, however, before Liu Xin proceeded to upend it.

Conclusion

Worship at the ancestral shrines in the vassal kingdoms of the pre-unification period were part and parcel of the feudal patriarchal clan system. Ideally, the heir to a polity should be the first-born son who would offer sacrifices to the ancestors on behalf of the polity's ruling line. Descendants of a ruling line in the lateral branches soon forfeited the privilege of worshipping their distant ancestors as descendants multiplied and branches of the clan subdivided. By idealized prescriptions drawn from the Classics, in the relatively rare cases when a member of a lateral branch in a minor clan succeeded to the throne of the main patriline, the new ruler of the polity must renounce his duty to worship his biological father. In real life, few rulers were willing to do so. Quite often in history there arose conflicts over the ruler as biological son or as successor to a ruling line.Footnote 141 An early prototype for this sort conflict is found in today's Zuo zhuan, which has an entry about “raising (ji 躋) the rank of Lord Xi of Lu 魯僖公 (r. 659–627)” two years after Lord Xi's death.Footnote 142 By rights, Lord Min 閔公 (r. 661–660), the elder brother and predecessor of Lord Xi, should have been ranked above Lord Xi, but he reigned at an early age for only two years, and died without siring descendants. Therefore, Lord Xi, who had descendants to inherit the title, was allowed to take precedence in the worship ceremonies, a usurpation of privileges severely criticized by the “Noble Man” (junzi 君子, commonly assumed to be Confucius).Footnote 143 Similar controversies rocked the imperial era as well, as emperors, wanting the privilege to worship their own biological fathers, found ways to get supporters on their side.Footnote 144 For that reason alone, the imperial ancestral shrine regulations never could be fully stabilized.

Since the Qing, scholars have cast the Han's commandery and kingdom shrine regulations as problematic, and believed the late Western Han reforms to be consistent with Western Zhou rituals. The picture is far more complicated. As have seen, Qin Shihuang destroyed the basis of patriarchal clans dating back to the pre-unification era, styling the ancestral shrines in some sense the possession of the imperial family. Then, heedless of the cost of frequent inspections and offerings of cult, the First Emperor established a large number of ancestral shrines in the localities of his realm. Had his idea been fully implemented, worship of the imperial ancestors would have become a line-item in the imperial budget and a part of local administration. At the same time, the existence of local ancestral shrines evidently was thought to undermine the older duties to offer cult at the ancestral shrines near the capital. Reformers in late Western Han concentrated the ancestral shrines in and near the capital of Chang'an, but they found it impossible to finally resolve problems pertaining to the feudal patriarchal order that predated the more centralized imperial order. Nonetheless they not only changed the number of imperial ancestral shrines and the way these ancestors were to be worshipped; they also redefined the very rationale for the imperial ancestral shrines.

By the very end of Western Han, reforms of the imperial ancestral shrines had dramatically altered the earlier forms established by Qin Shihuang and Ershi. In the successor state of Eastern Han, debates about the order of the ancestral shrines continued, always within the context of what were mistakenly viewed to be the classical pre-unification rituals. Whether by design or unwittingly, this new set of reformers never mentioned the seminal roles the Qin emperors had once played in establishing the imperial ancestral shrines.

References

1. All dates are b.c.e. unless otherwise noted. All mortuary terms (e.g., Chamber of Rest) follow Michael Loewe’s renderings and explanations, as noted in the notes below, to obviate the need for unnecessary digressions.

2. On ancestral worship in Eastern Zhou, see Constance A. Cook, “Ancestor Worship during the Eastern Zhou,” in Early Chinese Religion, Part One: Shang through Han (1250 BC–220 AD), 2 vols, ed. John Lagerwey and Marc Kalinowski (Leiden: Brill, 2009), vol. 1, 237–79.

3. Loewe, Michael, Problems of Han Administration: Ancestral Rites, Weights and Measures, and the Means of Protest (Leiden: Brill, 2016), 1107CrossRefGoogle Scholar.

4. For the period 197–40 b.c.e. in Western Han, see Michael Loewe’s The Men Who Governed Han China (Leiden: Brill, 2004).

5. The phrase “commandery and kingdom ancestral shrines” (jun guo miao 郡國廟) is common in the Han shu 漢書. See Han shu (Beijing: Zhonghua, 1962), 25B.1253, 73.3116, 3117, 3121. This article contrasts the “county and march shrines” of Qin with the “commandery and kingdom ancestral shrines” of Western Han, although both fall under the rubric of “local ancestral shrines.”

6. For details, see Loewe, Michael, Crisis and Conflict in Han China, 104 BC to AD 9 (London: Allen & Unwin, 1974; rpt. New York: Routledge, 2019), 154–92Google Scholar; Nylan, Michael, “Han Views of the Qin Legacy and the Late Western Han ‘Classical Turn,’Bulletin of the Museum of Far Eastern Antiquities 79–80 (2018), 73122Google Scholar.

7. As both Qin and Han established shrines for Taishang huang, this article will label a shrine or shrines as “Qin” or “Han” in brackets, as necessary, to avoid misunderstanding.

8. Michael Nylan and Nicholas Constantino show that there was no ritual “system” in place during Western or Eastern Han times, although there were court rulings (zhi 制). See their “On the Rites in Mid-Eastern Han,” in Autour du Traité des Rites, De la canonisation du rituel à la ritualisation de la société, all about the Rites: from canonized ritual to ritualized society, ed. Anne Cheng et Stéphane Feuillas (Paris: Maisonneuve & Larose, 2021), 241–83.

9. For relevant examples, see Zuo zhuan zhushu 左傳注疏 (Shisan jing zhushu 十三經注疏 ed., 1815; rpt. Beijing: Zhonghua, 1980), 21.1867, 21.1869, 30.1943, 51.2110.

10. Archaeologists believe a building site at Majiazhuang 馬家莊 (No. 1) in Yong city may represent a Zhanguo shrine dating to the pre-imperial kingdom of Qin. See Shaanxi Yongcheng kaogu dui, “Fengxiang Majiazhuang yihao jianzhu qun yizhi fajue jianbao” 鳳翔馬家莊一號建築群遺址發掘簡報, Wenwu 1985.2, 3–31, 100.

11. Han shu, 19.742. For dao, see The Cambridge History of China, Volume 1: The Ch’in and Han empires, 221 B.C.–A.D. 220, ed. Denis Twitchett and Michael Leowe (Cambridge: Cambridge University Press, 1986), 475; China’s Early Empires: A Reappraisal, ed. Michael Nylan and Michael Loewe (Cambridge: Cambridge University Press, 2010), chap. 4 and p. 263.

12. All transcription and photos of the seven slips appear in Chen Songchang 陳松長, ed., Yuelu shuyuan cang Qin jian (si) 嶽麓書院藏秦簡 (肆) (Shanghai: Shanghai cishu, 2015). All Yuelu Qin slips are marked with two sets of serial numbers, the first being the number it was originally assigned and the second, the number referring to the edited and collated version included in the volume. Because some of the earlier published articles still use the original number, two sets of numbers are indicated in this article to facilitate cross-checking with other studies.

13. Chen, Yuelu shuyuan cang Qin jian (si), photos and transcription 201–2; transcription without notes 249.

14. The phrase “ci miao” 祠廟 is often understood as a single expression by scholars, but in the Qin-Han context, “miao” 廟 always refers to ancestral shrines, while “ci” 祠 means either a noun (a shrine) or the verb “to worship.” The meaning of “ci miao” here is somewhat uncertain, so I tentatively render it as “shrines [designated] to receive cult offerings,” hoping that future excavations will provide texts that will allow me to refine my tentative translation.

15. Chen, Yuelu shuyuan cang Qin jian (si), photos and transcription 201–2; transcription without notes 249.

16. Yuelu shuyuan cang Qin jian (si), 226n34. In 221, Qin Shihuang began to call himself “Emperor” and gave his father the title of “Taishang huang.” Shi ji (Beijing: Zhonghua, 1959), 6.236.

17. Hunan sheng wenwu kaogu yanjiu suo, Liye fajue baogao 里耶發掘報告 (Changsha: Yuelu, 2006) [hereafter Liye], 8. See also Robin D. S. Yates, “The Qin Slips and Boards from Well No. 1, Liye, Hunan: A Brief Introduction to the Qin Qianling County Archives,” Early China 35/36 (2012–2013), 291–329.

18. This does not follow Yates’s translation of lingshi, but rather that of Michael Loewe.

19. In Shi ji, the Chancellor (chengxiang 丞相) was responsible for “inspecting the cemeteries every season” (sishi xingyuan 四時行園), here “xing” means inspect also. Shi ji, 122.3142.

20. The remaining text is missing.

21. For the interpretation of this sentence, see Lu Jialiang 魯家亮, “Liye Qin jian lingshi xing miao wenshu zai tan” 里耶秦簡’令史行廟’文書再探, Jianbo yanjiu 2014, 47.

22. Chen Wei 陳偉, ed., Liye Qin jiandu jiaoshi (di yi juan) 里耶秦簡牘校釋(第一卷) (Wuhan: Wuhan daxue, 2012), 78. The full text of this imperial edict has been translated by Robin D. S. Yates, in Yates, “The Qin Slips and Boards from Well No. 1, Liye, Hunan,” 322–25. My translation differs from that of Yates.

23. NB: many scholars have overstated the remoteness of Liye, since it was an important military outpost; they also tend to fail to see that there were military rituals, as well as civil, to be carried out at such outposts.

24. If there is an ancestral shrine of Taishang huang in every county, the number of the shrines could be around one thousand. The number of counties in the Qin dynasty cannot be fully determined. Yan Gengwang believes that there were about 1,000, while Hou Xiaorong deduces that there were at most 800–900. See Yan Gengwang 嚴耕望, Zhonguo difang xingzheng zhidu shi 中國地方行政制度史 (Taipei: Institute of History and Philology, Academia Sinica, 1991), 35; Hou Xiaorong 后曉榮, Qindai zhengqu dili 秦代政區地理 (Beijing: Shehui kexue wenxian, 2009), 449–55.

25. The second half of the sentence is missing, and the translation here is the first half of the sentence.

26. Chen, Yuelu shuyuan cang Qin jian (si), 201. Translation tentative, because conceivably geng 更 means gengzu 更卒 (commoners who perform their duties in shifts).

27. The second character was originally read as bu 部, but Chen Wei correctly changed it to du 都. See Chen Wei 陳偉, “Yuelu Qin jian si jiao shang (san)” 岳麓秦簡肆校商(三), March 29, 2016, 簡帛網 www.bsm.org.cn.

28. The meaning of the words 毋過月歸(?) is not clear, so I provide no translation here.

29. The last sentence is too fragmented to be translated.

30. Chen, Yuelu shuyuan cang Qin jian (si), 201.

31. Lu Jialiang calculated the day of the month according to the ganzhi 干支 sexagenary cycle. To make it easier to understand the intervals between inspections, the results of Lu’s calculation have been added to this translation. See Lu, “Liye Qin jian,” 44.

32. The remaining text is missing.

33. Lu, “Liye Qin jian,” 45.

34. Chen, Yuelu shuyuan cang Qin jian (si), 201.

35. Shi ji, 102.2753. Loewe has recently changed his Superintendents to Commissioners, due to changes in English of the connotations attached to those words. The tingwei 廷尉 was one of the ministers.

36. Shi ji, 6.241. See also Tian Tian 田天, Qin Han guojia jisi shigao 秦漢國家祭祀史稿 (Beijing: Shenghuo dushu xinzhi, 2015), 70–73; and Loewe, Problems of Han Administration, 182–84.

37. Zuo zhuan zhushu (Shisan jing zhushu), 5.1743.

38. Shi ji, 6.267, 269.

39. Guojia jiliang zong ju ed., Zhongguo gudai du liang heng tu ji 中國古代度量衡圖集 (Beijing: Wenwu, 1981), 65, 66.

40. NB: This article has modified the way the Zhonghua version punctuates the sentences.

41. Shi ji, 6.266.

42. This phrase can be found in several chapters of today’s Record of Rites (Li ji) and in Guliang zhuan, an Annals commentary, two Western Han compilations that plainly postdate Ershi’s time. We cannot know whether any of the passages in these two works represent pre-Qin or Qin teachings. See Li ji zhushu 禮記注疏 (Shisan jing zhushu), 12.1335, 18.1393, 23.1431, 46.1589; Chunqiu Guliang zhuan zhushu (Shisan jing zhushu), 8.2397. Most early commentaries agree with Zheng Xuan’s 鄭玄 (127–200) interpretation that “the seven shrines” includes three shrines built in honor of the first three ancestors of the ruling family and four for the most recent four male ancestors. If Zheng is correct, with each new accession to the throne, the earliest of the latter four shrines would be de-commissioned, to make room for a shrine of the most recently deceased ruler, but the other three shrines would be permanent. At present, no evidence exists to show that the Seven Shrines was applied before the Qin dynasty, though some would retroject the group back in time to Western Zhou.

43. Both Brashier and Li Kaiyuan 李開元 note the importance of this reform. See K. E. Brashier, Ancestral Memory in Early China (Cambridge, MA: Harvard University Aisa Center, 2011), 104–7. Li Kaiyuan, “Qin Shihuang di yi ci xunyou dao Xi xian gao miao jizu shuo: jianji Qin tongyi hou de miaozhi gaige” 秦始皇第一次巡遊到西縣告廟祭祖說——兼及秦統一後的廟制改革, Qinhan yanjiu 10 (2016), 10–17.

44. Qian Xuan cited the examples of Lu and Jin states from the Zuo zhuan, which can also prove that the number of ancestral shrines within the Eastern Zhou states was more than that recorded in the ritual books. See Qian Xuan 錢玄, San li tonglun 三禮通論 (Nanjing: Nanjing shifan daxue, 1996), 446–47.

45. For example, “Starting from Fei Lian, he gave birth to Ji Sheng, and five generations later, to Zao Fu, and at that time the family moved to Zhao kingdom” (自蜚廉生季勝,已下五世至造父,別居趙). This refers to the five generations of Fei Lian, Ji Sheng, Meng Zeng 孟增, Heng Fu 衡父, and Zao Fu. See Shi ji, 6.175. “When the Changle Palace was completed, officials from the Chancellor down moved to Chang’an to handle their administrative affairs” (長樂宮成,丞相已下徙治長安). See Shi ji, 8.385.

46. See Li ji zhushu (Shisan jing zhushu), 12.1335.

47. Hunansheng wenwu kaogu yanjiu suo, Yiyang shi wenwu ju, “Hunan Yiyang Tuzishan yizhi jiu hao jing fajue jianbao” 湖南益陽兔子山遺址九號井發掘簡報, Wenwu 2016.5, 43, photo 40.

48. Kanli, Chen 陳侃理, “Sima Qian yu Guo Qin pian” 司馬遷與過秦篇, Lingnan xuebao (fukan di shi ji) 10 (2018), 152Google Scholar.

49. Li Kaiyuan and Chen Kanli have noted and discussed these two passages, respectively. See Li Kaiyuan, “Qin Shihuang di yi ci xunyou dao Xi xian gao miao jizu shuo,” 16n1; and Chen Kanli, “Sima Qian yu Guo Qin pian,” 150.

50. Additionally, in a paragraph attached at the end of the present version of “Annals of Qin Shihuang” in the Shi ji, Ban Gu 班固 (32–92 c.e.) presented a memorial that also mentioned that “Ziying . . . visited the seven ancestral shrines.” The “seven ancestral shrines” here also refer to the Seven Shrines set up by Ershi. Only Ban Gu’s evaluation of Ziying differs from that of Jia Yi (Shi ji, 6.292). Here I would note that Ban Gu’s Han shu, written during Zhangdi’s era, mainly focuses on the late Western Han reforms, insofar as those limited the number of important shrines that the Eastern Han ruling line could build, by precedent, and this, by Zhangdi’s era, had become a looming issue.

51. Or: let fall to ruin.

52. Shi ji, 6.284. While one anonymous reviewer of this article thought the justification for this reform might lie in Huidi’s lack of an heir, I believe that Huidi’s accession to the throne would weigh more heavily than his lack of an heir in most debates relating to the ancestral shrine hierarchy.

53. NB: in Jia Yi’s essay (here summarized), abolishing the ancestral shrines (huai zong miao 壞宗廟) heads the list of all the crimes of Ershi.

54. Shi ji, 8.382.

55. Han shu, 1B.68.

56. As is well known, the Han founder initially had little control over much of the empire nominally ruled by him, since the kingdoms had their own administrations, and considerable independence over affairs of state.

57. See note 6 above.

58. They were Ying Bu 英布 (d. 196) of Huainan 淮南, Lu Wan 盧綰 (d. 194) of Yan 燕, Liu Jia 劉賈 (d. 195) of Jing 荊, Liu Jiao 劉交 (d. 179) of Chu 楚, Liu Fei 劉肥 (d. 189) of Qi 齊, Peng Yue 彭越 (d. 196) of Liang 梁, Liu Ruyi 劉如意 (d. 194) of Zhao 趙, Wu Chen 吳臣 (d. 193) of Changsha 長沙, and Dai 代, which was under the jurisdiction of Zhao and thus had no king of its own. Han shu, 13.377–79, 14.397–406.

59. Chen Suzhen 陳蘇鎮, Chunqiu yu Han dao: Liang Han zhengzhi yu zhengzhi wenhua yanjiu 春秋與漢道: 兩漢政治與政治文化研究 (Beijing: Zhonghua, 2020), 72–107.

60. Shi ji, 8.392. Han shu contains basically the same information, except that it adds the fact that on the day of Liu Bang’s funeral, the officials and the prince returned to the Han shrine for Gaozu’s father, Taishang huang. See Han shu, 1.80. Lu Jia 陸賈 also blamed Ershi for the collapse of Qin dynasty, as we see from Wang Liqi 王利器, Xin yu jiaozhu 新語校注 (Beijing: Zhonghua, 2012), 1.28.

61. Of course, Liu Bang may have had additional reasons for his measure.

62. The Han administrative system and not a few Han rituals basically continued those of the Qin dynasty, and so these were inherited from the time of Qin Shihuang. See Chen Suzhen, Chuqiu yu Han dao, 38–72; Nylan, “Han Views of the Qin Legacy and the Late Western Han ‘Classical Turn,’” 80–101.

63. Shi ji, 8.392.

64. Han shu, 5.138.

65. See Shi ji, 9.399, 10.418, 436. And Han shu, 7.217, 8.238, 9.277, 10.301, 11.334, 12.347.

66. Han shu, 68.2945–46.

67. Han shu, 7.221, 12.357, 97B.4010.

68. Han shu, 7.229, 12.352.

69. Han shu, 3.96.

70. Han shu, 99A.4095, 99B.4113.

71. Han shu, 99C.4169. For Wang Mang’s accession to the throne, see the contribution by Béatrice L’Haridon in this volume.

72. Han shu, 43.2131.

73. Shi ji, 8.392.

74. Shi ji, 39.2725–26. We know the shrine at Pei was re-established in 190, and the first shrine between 194–190.

75. Shi ji, 10.436.

76. Han shu, 8.243. To add to this, a Western Han tile inscribed with the word “zongmiao” was found at the site of the Han Wei Luoyang imperial city in 2020, and archaeologists speculate that this indicates a Western Han local ancestral shrine was built there. The brief archaeological report has not yet been published; a short introduction can be found on Jan. 13, 2021, at Luoyanggang 洛陽網 (www.lyhand.com/n/1034435, accessed in early 2022).

77. Translations of these posthumous titles appear below.

78. Han shu, 73.3115. Zhou Zhenhe 周振鶴 estimates the number and location of the shrines in different commanderies and kingdoms; however, Zhou’s calculation is based on an estimate of the total number of ancestral shrines in 167. See Zhongguo lishi wenhua quyu yanjiu 中國歷史文化區域研究, ed. Zhou Zhenhe (Shanghai: Fudan daxue, 1997), 73–78.

79. Shi ji, 106.2834.

80. Han shu, 8.245.

81. Han shu, 72.3091, 81.3357. For example, Han Aidi (r. 4 b.c.e.–1 c.e.) avoided holding court in the main palace due to a series of unusual disasters. Likely Aidi was advised to avoid court, because to do so might strengthen Wang Mang’s hand. Moreover, several Han emperors declared their intentions to don mourning after fires broke out in the Chang’an ancestral shrines or their affiliated buildings (Han shu, 6.159, 7.230, 8.269).

82. Probably Jingdi and Xuandi responded with special solemnity, because both had faced severe crises at the beginning of their reigns.

83. Jiangxi sheng wenwu kaogu yanjiu yuan et al., “Jiangxi Nanchang Xi Han Haihun hou Liu He mu chutu jiandu” 江西南昌西漢海昏侯劉賀墓出土簡牘, Wenwu, 2018.11, 87–96.

84. Tian Tian, “Xi Han Haihun hou Liu He mu chutu zongmiao yi lei wenxian chutan” 西漢海昏侯劉賀墓出土宗廟 ‘儀’ 類文獻初探, Wenwu, 2022.6, 65–67, 74.

85. Li ji zhushu (Shisan jing zhushu), 12.1335.

86. Most scholars have construed the shrine worship as part of the patriarchal clan system. See, e.g., Yang Kun 楊坤, Liang Zhou zongfa zhidu de yanbian 兩周宗法制度的演變 (Shanghai: Shanghai guji, 2021), 3–17.

87. Zuo zhuan zhushu (Shisan jing zhushu), 31.1951. This translation follows Stephen Durrant, Wai-yee Li, and David Schaberg, Zuo tradition = Zuozhuan: Commentary on the “Spring and Autumn Annals” (Seattle: University of Washington Press, 2016), 995, on 12.3(4) (slightly mod.). As readers may recall, today’s Zuo zhuan is not identical with the Han-era Zuo shi chunqiu, as confirmed by William Hung demonstrated long ago, in his “Combined Indices to the Titles Quoted in the Commentaries on Ch’un-ch’iu, Kungyang, Ku-liang, and Tso-chuan (Taipei: Chinese Materials and Research Aids Service Center, 1966), Introduction, and Pauli Wai Tashima, “Merging Horizons: Authority, Hermeneutics, and the Zuo Tradition 左傳 from Western Han to Western Jin (2nd c. b.c.e.–3rd c. c.e.)” (Ph.D. thesis, University of California at Berkeley, 2013).

88. Li ji zhushu (Shisan jing zhushu), 52.1629. On the zhaomu 昭穆 system, see Loewe, Problems of Han Administration, 4–7. Li Hengmei 李衡眉, Zhaomu zhidu yanjiu 昭穆制度研究 (Jinan: Qi-Lu shushe, 1996).

89. Decades ago, Jin Jingfang realized that Qin completely destroyed the patriarchal clan system, but he gave no explanation for Qin’s establishment of the ancestral shrines. Jin Jingfang 金景芳, “Lun zongfa zhidu” 論宗法制度, Dongbei renmin daxue renwen kexue xuebao, 1956.2, 222. Reference to the “three legs” makes use of the antique metaphor by which a stable realm is likened to a bronze tripod.

90. Meguro Kyoko 目黒杏子 suggests that the theory emphasizing the achievements of the emperor was not introduced before Aidi and Pingdi. See her “Zen Kan kouhanki ni okeru sōbyōsei no henyō” 前漢後半期における宗廟制の變容, Tōhō gakuhō 東方學報 95 (2020), 21–26.

91. Shi ji, 110.434. On the translation of this imperial edict and on mourning in general during the Han dynasty, see Brown, Miranda, The Politics of Mourning in Early China (New York: SUNY Press, 2007), 2429Google Scholar.

92. Shi ji, 110.436.

93. For more examples, see Loewe, Problems of Han Administration, 40–45.

94. Han shu, 8.242.

95. See Chun qiu Gongyang zhuan 春秋公羊傳 (Shisanjing Zhushu), 1.2196.

96. Han shu, 63.2748.

97. Han shu, 8.243. There was some debate with Wendi as well.

98. This is how the relationship between the two was handled during the reform of Yuandi. See Han shu, 73.3120.

99. In mid-Western Han, the idea that the establishment of commandery and kingdom shrines could be used to commemorate imperial merit seems to have been the consensus view. During the reign of Zhaodi (r. 86–74), Liu Dan 劉旦, King of Yan 燕, submitted a petition alleging that Wudi was “so very virtuous and splendid” 德甚休盛 that local commandery and kingdom shrines should be established in his honor. While Huo Guang rejected this proposal, Huo offered Liu Dan great rewards for his suggestion (Han shu, 63.2751).

100. Loewe, Problems of Han Administration, 50–56; Brashier, Ancestral Memory, 130–41.

101. For example, Lin Congshun divides the development of commandery and kingdom shrines of Western Han into three parts. The third part treats the abolition of the commandery and kingdom shrines during Yuandi’s reign as a single phase. See Lin Congshun 林聰舜, “Xi Han junguo miao zhi xingfei: lizhi xingge yu tongzhi zhixu weihu de guanxi zhi yi li,” 西漢郡國廟之興廢——禮制興革與統治秩序維護的關係之一例 Nandu xuetan (renwen shehui kexue xuebao), 27.3 (2007), 1–8.

102. Han shu, 25B.1257–58.

103. Han shu, 73.3124. No attempt was ever made to restore the commandery and kingdom ancestral shrines, so far as we know.

104. Han shu, 25B.1264. See also Tian Tian, “The Suburban Sacrifice Reforms and the Evolution of the Imperial Sacrifices,” in Chang’an 26 BCE: An Augustan Age in China, ed. Michael Nylan and Griet Vankeerberghen (Seattle: University of Washington, 2015), 263–91. For a different view of many aspects of this story (discussed in detail in my earlier contribution in Chang’an 26 BCE), see Marianne Bujard, Le sacrifice au ciel dans la Chine ancienne: théorie et pratique sous les Han Occidentaux (Paris, EFEO, 2000), or the shorter English version: “State and Local Cults in Han Religion,” in Early Chinese Religion, Part One: Shang through Han (1250 BC–220 AD) (Leiden: Brill, 2009), vol. 1, 777–811.

105. Han shu, 25B.1258–59.

106. Han shu, 73.3121.

107. Han shu, 73.3116, 3121.

108. Han shu, 73.3121.

109. Han shu, 72.3079. For the expenses incurred with the commandery and kingdom shrines, see also Loewe, Crisis and Conflict, 179–82.

110. Han shu, 25B.1264.

111. The translation of 寢 follows Loewe, Problems of Han Administration, 9.

112. Han shu, 73.3116–17, citing Analects 3/12.

113. Han shu, 73.3117.

114. Shi ji, 10.436.

115. The word “governors” is omitted from the passage cited earlier from Han shu, so the translation put a question mark and square brackets there to indicate that. As there were imperial ancestral shrines in the commanderies, one might deduce that commandery governors were the likely people deputed to oversee the sacrifices at the ancestral shrines in the commanderies.

116. Editor’s note: This suggests that the princes had more power than the Han sources would credit them with.

117. For example, in his prayer speech, Kuang Heng repeatedly emphasized that “It is appropriate for the ancestral shrines to be in the capital and for the Son of Heaven to serve them personally” 廟宜一居京師,天子親奉. See Han shu, 73.3121.

118. This process has been discussed in detail by Loewe in his “The Imperial Tombs of the Former Han Dynasty and their Shrines,” in Divination, Mythology and Monarchy in Han China (Cambridge: Cambridge University, 1994), 267–99. For a discussion of the targets of gradual obliteration and the number of ancestral shrines, see Guo Shanbing 郭善兵, Zhongguo gudai diwang zongmiao lizhi yanjiu 中國古代帝王宗廟禮制研究 (Beijing: Renmin, 2007), 120–36. Qing scholars were extremely enthusiastic about the patriarchal clan system, and they discussed the number of ancestral shrines and their references in the Classics in the Western Han dynasty from the perspective of classical learning. For an overview, see Gao Jingcong 高婧聰, “Qing ren dui Zhou dai zongfa zhidu de yanjiu” 清人對周代宗法制度的研究, Gudai wenming, 2019.1, 102–12.

119. Han shu, 73.3120.

120. Wei Xuancheng did not say a word about the shrine of Jingdi, but from the discussion that follows, it is clear that the shrine of Jingdi was not abolished. See Han shu, 73.3119–20.

121. At this time there were in fact seven ancestral shrines preserved, namely the shrines of Gaozu, Wendi, Jingdi, Wudi, Zhaodi, Xuandi, and shrine of Huangkao 皇考. From Wei Xuancheng’s last memorial, he chose to focus on the arrangement of zhaomu, and the number of ancestral shrines was treated vaguely.

122. Han shu, 73.3125–27. Also see Brashier, Ancestral Memory, 130–41.

123. For the locations of the imperial mortuary complexes near Chang’an, see Chang’an 26 BCE, Map I. 06, 34. All these locations have been confirmed by the latest drills performed by archaeologists.

124. See also Loewe, Problems of Han Administration, 54–55.

125. Han shu, 9.292, 73.3117.

126. Han shu, 9.292.

127. Han shu, 9.293. While one anonymous reviewer of this article thought the justification for this reform might lie in Huidi’s lack of an heir, I believe that Huidi’s accession to the throne would weigh more heavily than his lack of an heir in most debates relating to the ancestral shrine hierarchy.

128. Han shu, 9.294.

129. Han shu, 9.297.

130. Han shu, 9.298.

131. Han shu, 9.298.

132. Han shu, 10.309.

133. With the matriline, only the cult site of the deposed Empress Wei was preserved. Shi Huangsun’s 史皇孫 real name was Liu Jin 劉進. The imperial line went from Wudi-Liu Ju (a.k.a. Li taizi 戾太子)-Liu Jin-Xuandi-Yuandi.

134. NB: This suggests either that the court reformers were pragmatic and not strict Confucians or that the Rites classics had relatively less authority in late Western Han than they would come to assume in late Eastern Han and the post-Han period, thanks in large part to Zheng Xuan.

135. Han shu, 73.3125.

136. During the time of Aidi, there were also discussions about restoring the ancestral shrines of Huidi and Jingdi, but they came to nothing in the end. See Han shu, 72.3082.

137. Hou Han shu, 2.142n.

138. Han shu, 73.3125.

139. Han shu, 73.3127.

140. Han shu, 12.357.

141. Because this observation is not relevant to the main line of reasoning advanced here, I direct readers to the clear statements made on this point in the authoritative Gongyang Commentary for the Annals classic. There Gongsun Yingqi 公孫嬰齊 was renamed Zhong Yingqi 仲嬰齊 after he became his elder brother’s heir (1.2196). The Gongyang Commentary explains the rule for the name change: “a person’s successor acts as the son for him” 為人後者為之子. By this rule, Gongsun Yingqi had to change his family name and he could no longer sacrifice to his biological ancestors in the Gongsun line. Earlier this essay cited the same line (“a person’s successor acts as the son for him”) in connection with Xuandi. Xuandi, as the successor to Zhaodi, could not properly offer imperial sacrifices to his own parents or grandparents by birth.

142. Zuo zhuan zhushu (Shisan jing zhushu), 18.1839. This event is also recorded in the Guo yu 國語. Loewe discusses the relationship between this event and zhaomu; see Loewe, Problems of Han Administration, 40–45.

143. Henry, Eric, “‘Junzi Yue’ Versus ‘Zhongni Yue’ in Zuozhuan,” Harvard Journal of Asiatic Studies 59.1 (1999), 125–61CrossRefGoogle Scholar.

144. The most well-known ones should be the “pu yi” 濮議 of the Northern Song dynasty and the “Dali yi” 大禮議 of the Ming dynasty, both of which stretched for a long time and shook the elite class.