Andrew Eisenberg-中国中古的太上皇制度

Retired Emperorship in Medieval China: The Northern Wei Author(s): Andrew Eisenberg Reviewed work(s): Source: T'oung Pao, Second Series, Vol. 77, Livr. 1/3 (1991), pp. 49-87 Published by: BRILL Stable URL: http://www.jstor.org/stable/4528517 . Accessed: 19/07/2012 21:58 Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp . JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org. . BRILL is collaborating with JSTOR to digitize, preserve and extend access to T'oung Pao. http://www.jstor.org T'OUNG PAO LXXVII (1991) RETIRED EMPERORSHIP IN MEDIEVAL CHINA: THE NORTHERN WEI BY ANDREW EISENBERG* Northeastern Illinois University Chicago, Ill. I. Introduction The broad issue to be addressed in this paper is the politics of succession in medieval China, as reflected in the institution of the retired emperorship (taishang huangdi, Ltt XJV) of the Northern Dynasties (386-581). Statements made regarding the Northern Dynasties' precedent have direct implications for the occurrence of similar phenomena in the subsequent early Tang. This paper will argue the hypothesis that the retired emperorship was first suc- cessfully utilized by the Northern Wei (and imitated by subsequent regimes) as a means of stabilizing the imperial succession. The position of this paper is, in part, inspired by the argument rendered by Hurst for the Japanese retired emperorship and also by the specific reasons offered for the Northern Qi retirement of 565 (Zizhi Tongjian 169: 5251 ).' * I wish to thank the following individuals for their valuable criticism and advice. Prof. Jack L. Dull, University of Washington, Seattle; Prof. He Ciquan, Beijing Normal University; Profs. Tian Yuqing & Li Ping, Beijing University; Profs. Tang Changru & Mou Fasong, Wuhan University. Research in China during the 1988-1989 academic year was made possible by a grant from the Committee on Scholarly Communication with the People's Republic of China of the National Academy of Sciences. l I borrowed the term retired emperor from the laterJapanese institution as it has been discussed by G. Cameron Hurst (1976). Perhaps a more accurate term would be ranked co-emperorship. In the cases of the "junior" emperors of the Chinese Northern Dynasties, they were all minors and did not exercise significant political authority beyond serving as the living symbol of the stability and continuity of the primogenital imperial succession. Analogous to the case to be presented for the Northern Wei, Hurst (1976: 47-50) argues for a strong correlation between the early Japanese use of the 50 ANDREW EISENBERG In discussing the power politics of the medieval period, we should bear in mind the structure of political action peculiar to this period. The formal institutionalization and differentiation of the social, political, and economic structures of the period were often rudimentary. Where rationalization of such structures did occur, the result usually tended to the formation of what can be termed the substantive political economy (Weber 1978: 641-900, 1006- 1069). In the arena of high politics, there was frequently little distinction between public governance and the private needs of the imperial family. In this regard, the politics of succession was an extraordinarily sensitive area that gave full play to the melding of public weal and private needs within the context of a patrimonial regime. In a lengthy discussion of corporate kin groups and offices associated with pre-modern kin oriented politics Meyer Fortes noted the following: Succession . . is of key importance in corporate group structure as well as ... in the continuity of corporation sole .. . With kinship corporation and hereditary offices or statuses, the connection is patent ... Office perpetuated by succession, hereditary or appointive, can serve as the focus for the corporate structure of a heterogenous or internally differentiated aggregation." (Fortes 1969: 305-306). Such being the case, serious study of the politics of succession affords us access to, and new interpretive vistas of, power relations within such regimes. Power Power may be defined as the ability to obtain compliance (this discussion is drawn from Luttwak 1976: 195-200; Poggi 1982: 1-15; and Etzioni 1961; 1985: 108-123). I feel that the term is axiomatic and defies precise definition. Power is exercised in a variety of fields of endeavor in a variety of ways. Political power may be taken to mean the ability to command the allocation of social resources on a society-wide basis. That which comprises the accepted field of political action and that which can be identified as "social resources" varies with time and place. Under normal circumstances, whether or not an American President attends religious services is not considered as being within the proper realm of political activity, nor is the President's church-going considered retired emperorship in the 600's to 700's, with the existence of an imperial succession that vacillated between a horizontal mode and a vertical, primogenital mode, with the former mode appearing much more frequently than the latter mode. In Japan, imperial retirement developed as a means of insuring that the pre-selected heir made it safely to the throne. RETIRED EMPERORSHIP IN MEDIEVAL CHINA 51 a social resource. In contrast to this, in the pre-modern agrarian empires the religious services performed by the Emperor were an integral part of the realm of political action and were considered a definite social resource. Failure to perform the proper rites at the customary times, it was felt, could have the direst of social and political consequences. In point of fact, the distinction between the private needs (religious or otherwise) of the princely household and the public exercise of political power was blurred and indistinct. Political power properly legitimated may be referred to as au- thority. Legitimation of power implies that such power is exercised through duly recognized institutional structures. Such institutional structures may be characterized as formal or non-formal. Formal institutions have some sort of legal charter, rules of conduct, and are often associated with a specific physical structure. The Ameri- can Congress would serve as an example of such a formal political institution. An example of a non-formal political institution would be a faction. Particularly in ancient societies, factional activities frequently became an accepted, almost stereotyped mode of politi- cal action. Recognized as customary, factions were never fully, legally constituted or chartered. In this study I will be concerned with the exercise of legitimated power (authority) by means of the retired emperorship. Unfortunately, under patrimonial regimes, the distinction between formal and non-formal political institutions becomes indistinct. The exercise of imperial will, excluding cases of sacrilege and violation of tradition, is ipso facto valid. The emperor is the ultimate source of law. He is not bound by the laws which he enacts though he may voluntarily opt to abide by these laws. The emperor may even issue an edict to that effect, insofar as it suits the imperial purpose. That which the emperor acknowledges is thus formally sanctioned by law. From this perspective, the retired emperorship can be considered a formal position within the impe- rial administrative structure. Patrimonial Rule Patrimonial rule refers to the exercise of territorial rule by the incipiently differentiated household staff of the prince (Weber, 1978: 226-241, 1006-1069). The differentiation of the imperial household staff may even be rationalized into a structure approxi- mating a formal bureaucracy (as in China or the Late Roman Empire), but the prerogatives and needs of the imperial household, which is at the root of this precocious bureaucratization, are never obscured. Thus, in the Chinese case, we see the frequently posited dichotomy of the Inner and Outer courts (though, Bielenstein, 52 ANDREW EISENBERG 1980: 154-155, has pointed out that personnel were often rotated through both courts as part of the normal course of a courtier's career). The inner court being the direct tool of the emperor and the outer court maintaining a slightly more autonomous, institu- tionalized existence vis-a-vis the emperor. Similarly, the political importance of eunuchs, imperial agnates and affines, and personal friends of the emperor is made possible solely by virtue of their direct access to the emperor and members of the imperial house- hold. These latter categories of kin and clients came to exercise empire-wide policy-making and executive privileges. This situation existed because all legitimate, traditional author- ity was appropriated, insofar as was possible, into the hands of the prince. Indeed, the existence of an incipient household staff and the concentration of political powers into the hands of one man are the key diagnostic traits that distinguish patrimonial rule from the lack of functional differentiation and the diffusion of authority typical of familial and clan patriarchalism. In the case of patrimo- nial rule, outside of the bounds of traditional comportment a considerable domain of political action was left open to the arbi- trary exercise of princely will. Actually, under patrimonial condi- tions, it was not necessary to utilize arbitrary, unprecedented commands in order to realize political goals. This is because the most startling innovations could be cleverly implemented under the guise of re-introducing the pristine ways of the past. Thus, the exercise of authority under patrimonial conditions implies unique forms of institutionalized expression, which are intimately bound to the emperor's role as the formal head of the imperial household. Consequently, there is an interpenetration of imperial household needs and the requirements of public governance. In a sense, the emperor was the patron for the entirety of his territorial subjects, who were his filial clients. It is through such an interpretive framework that I intend to investigate the phenomenon of the retired emperorship. Retired Emperorship The appearance of the retired emperorship in the Northern Dynasties and its extension into the early Tang is strongly corre- lated with succession difficulties faced by these regimes. This early use of imperial retirement as a tool to stabilize imperial successions is also strongly correlated with regimes attempting to institute a primogenital form of succession over against an earlier possible tradition of horizontal succession. This is particularly true of the cases from the Northern Dynasties and also the later retirement RETIRED EM1ERORSHIP IN MEDIEVAL CHINA 53 cases from Japanese history. The Tang cases may represent a developmental variation on this trend. In the case of the Northern Dynasties' emperors, I will argue that the utilization of the retired emperorship represented a posi- tive move by the mature imperial incumbent (at least 16 years of age) to ensure that his young son would not encounter a legitimacy crisis upon his demise. Quite simply, the reigning monarch an- nounced his "retirement" and (often) his removal to a separate palace complex. The young heir designate was then formally installed as emperor. In the Northern Dynasties all the boy-emperors were under ten years of age at their ascension. Under such conditions, the young heir was formally and legitimately ensconsed upon the throne long before the demise of his father. In the Northern Dynasties' cases, the mature retired emperor continued to manage all affairs of state until his demise. Under the conditions of the Northern Wei retirement, the boy-emperor was obligated to pay a formal court visit upon his father once a month (Wei Shu 7: 138). In the Tang cases the retired emperors appear to have voluntarily relinquished authority to their already mature sons. The Tang cases are very interesting but will not be directly discussed in this paper (see Twittchet and Fairbank 1979: 326-328, 334-345). In this introductory discussion of Chinese retired emperors, it is necessary to distinguish between nominal and active retired emper- ors (a distinction also implicitly made by Zhao Yi, 1963: 14: 255). To understand why such a distinction is made we must quickly review the historical development of the Chinese institution.2 Prior to the establishment of the Western Jin regime (265-317), the retired emperorship was a strictly honorific appellation. Qin Shihuangdi first bestowed this title upon his deceased father. During the early years of the Former Han, Gaozu bestowed this title upon his living father as a means of expressing filial piety and preserving the customary status hierarchy between his father (who was a commoner) and the son (the founding dynast). In 301, as part of the complex political infighting accompanying the Insurrection of the Eight Kings, Sima Lun (IJ Tf0) staged a coup and forced the puppet emperor, Huidi (;Ai;j) to "retire" with the title of taishang huang. Lun assumed imperial responsibilities and the retired em- peror Hui was sequestered in the Jinyong Fortress in Loyang. The use of the retired emperorship occurred again in 400 when Li! Guang (E 5), ruler of the Later Liang regime in Gansu retired in favor of his son, and then died immediately thereafter. Following 2 This narrative is derived from Boodberg 3: 3, 4: 1938, 235-237; Anthony B. Fairbank 1986: 140, 162 footnote #45; Wenxian Tongkao juan 251-252. 54 ANDREW EISENBERG this occurrence are the key Northern Dynasties cases (Northern Wei 471, Northern Qi 565, 576, 577 and Northern Zhou 579). In 617 Li Yuan (Tang Gaozu) bestowed the title of retired emperor upon Sui Yangdi in absentia. In 626, following the palace coup engineered by his son Li Shimin (Tang Taizong), Gaozu retired with the title of taishang huang. He held this title until his death in 635. Following this is the Ruizong retirement in favor of Xuanzong in 712. Finally, in 756 in the midst of the An Lushan rebellion, Xuanzong, who was then in Sichuan, retired in favor of his son Suzong, who then directed the effort from his base in northeastern Gansu. The early Southern Song made continuous use of this institution. The last case of imperial retirement was that of the Qing Qianlong emperor. Based upon the above historical outline, I believe that a develop- mental pattern can be established. Prior to the Western Jin coup of 301, the retired emperorship was an honorific title. In 301, how- ever, this title was directly involved in the factional politics of the time. The important point is not that Jin Huidi used his position as retired emperor to exercise political power, he did not. Historical opinion considers Huidi to have been incompetent. From the perspective of this paper, the mere use of the originally honorific, otiose title of the retired emperor as a tool of factional in-fighting was unique at the time. Thus, the Western Jin utilization of the retired emperor may be viewed as a crucial transitional usage, a harbinger of the active role the retired emperor would play in the later Northern Dynasties and the Tang. Indeed, in the three major cases from the Northern Dynasties the retired emperors continued to "manage important affairs of state, especially matters of foreign policy." (Boodberg 1938). The second developmental stage in the use of the retired emper- orship appeared during the 16 Kingdoms Period. In 399 AD Li! Guang, emperor of the Later Liang kingdom in western Gansu, retired in favor of one of his younger sons (the eldest son of his main consort was not the eldest male of all the sons propagated by Guang). Guang died immediately after retiring. Maintaining the stability of the succession in the face of the clear threat posed by his older "bastard" sons was clearly uppermost in Guang's mind when he retired (Jin Shu 122: 3063; Zizhi Tongiian 111: 3487, 3504-3506; WS 95: 2086). At the time, Li! Guang was an old and mortally ill man of about 62 years of age (Jin Shu 122: 3063-64). The strategem failed to secure the stability of the succession, however, this incident marked the first time that imperial retire- ment and primogenital succession to the throne were directly RETIRED EMPERORSHIP IN MEDIEVAL CHINA 55 (albeit unsuccessfully) linked together. Commencing with the cru- cial and influential Northern Wei retirement of 471, this linkage was successfully effectuated. In 471, two courtiers participating in the discussion regarding the retirement issue at the Northern Wei court were originally from western Gansu (the eunuch, Zhao Hei, M [WS 94: 2016], and the Xianbei general, Yuan He, R [WS 41: 919]. They may very well have been directly aware of the Later Liang precedent. One of the reasons the Northern Wei monarch offered for desiring to retire in 471 was that he was physically ill. The claim does not appear to have been true, but it may have been a direct imitation of the circumstances which Li! Guang (truth- fully) used to justify his retirement in 399. The statements issued by the Northern Wei court justifying the 471 retirement referred exclusively to the Former Han precedent. The Hou Liang case was never mentioned. Events in this second developmental stage also seem to be closely tied to efforts by non-Han regimes to shake loose from a tradition of horizontal succession and to establish a pattern of vertical, primogenital succession to the throne. The third developmental stage can be discerned in the Tang usage of the imperial retirement for reasons that may not be related to the issue of conflicting traditions of imperial succession. The succession issue remained, however, at the root of many of the Tang retire- ments, too. The Tang usage would appear to have marked a further rationalization of court politics and factional infighting in the direction of standard Han Chinese beliefs and practices.3 An interesting aspect of the retired emperorship during the Period of Disunion is that it was unique to the Northern Dynasties. This strategem was not used by any of the Southern Dynasties as a means of resolving their own manifest succession problems. The question that naturally arises is why this particular mode of power manipulation was chosen by the Northern emperors, who pion- eered the use of this institution as an important tool in the exercise 3 This trend would tie in well with statements regarding other aspects of Tang society that point to a growing Confucianization of this society. See Ch'u 1961: 281; Dull 1978: 23; Wong 1979: 270-274. In the early Tang period, however, direct North Asian political influence should not be ignored. Fletcher (1978: 66) explicitly makes this point: " . . . in 626 the pressures of bloody tannistry are to be seen in the fact that Li Yuan abdicated the throne in favor of his son Li Shih-min, who had played a leadership role from the beginning, enjoyed a large personal following, and had used it to eliminate his brothers. Having established his personal authority over the various magnates of the empire, he was forced-in order that his realm should not disintegrate-to spend the rest of his life waging wars of expansion. Bloody tannistry's traces and the traits of the ruler