下载掌阅APP，畅读海量书库
立即打开
《唐研究》第二十卷
2014年，63—106頁
Andrew Eisenberg
This essay will explore the collapse of the Northern Wei regime as a function of the disintegration of the political consensus that had knit together the Northern Wei court．As a self-conscious ethnic conquest regime the throne had established with its Xianbei and Xianbeized elites of various status levels a tacit，nearly explicit“covenant”granting these elites political perquisites not readily available to the conquered populations．For these elites，the members of the original conquering host，their descendants，and others who attached themselves to this regime over the course of time，the court was the lodestone of their political and administrative careers and for their social connections，as well．The linkage with the court world was more intense and visceral than for the much broader and diversified elites of previous（and contemporary）native Chinese dynasties.Dating from the early years of the Northern Wei regime，the court was apparently structured along apartheid lines with the key political and military positions reserved for Xianbei，especially the Northern Wei policy making“inner court.”The court also made a point of guaranteeing entry level court postings to Xianbei youth as well as access to military service in palace guard units and Xianbei cavalry units for this constituency．Mark C．Elliott referred to this mode of conquest political domination as an exercise in“ethnic sovereignty.”The manner by which the throne maintained such intimate relations with its Xianbei elites（as well as multiplying numbers of increasingly distant collateral royal relatives）marks the Northern Wei dynasty as a“corporate”regime．Consequently，the throne had significant restrictions placed upon its policy making capabilities and its personnel decision making capabilities to an extent unknown to its Chinese counterparts．This paper will argue that subsequent efforts by the throne to alter，indeed，to wriggle out from under this ethnically based“covenant”will trigger a series of political crisises at court after 500.
The later court structural reforms of Xiaowendi from 490 through 499 eliminated the formal apartheid structure of the court．Nevertheless，personnel allocation decisions continued to adhere to the practice of court apartheid personnel placement．Beginning with the reign of Shizong（Xuanwudi，r.500-515）efforts were undertaken to break through the apartheid mentality，particularly with regard to the“ecumenical”factionalism which Shizong encouraged and manipulated．Ironically，these very efforts would result in a secular trend of destabilizing the pre-existing political and social comity of the court，culminating in the 528 dominance of the Erzhu family satraps over the Northern Wei court at Luoyang.
Xiaowendi's court structural reforms were complemented by a massive re-orientation of the Northern Wei geo-strategic stance，made manifest in the relocation of the capital from Pingcheng to Luoyang in 493．First，this essay will address the geo-strategic reorientation of the regime，which will also play a role in the subsequent discussion of the long term，inadvertent de-stabilization of the Northern Wei court beginning with the reign of Shizong in 500.
Warfare was crucial to Xiaowendi's imperial agenda，and in my opinion was a key factor in enabling Xiaowendi to partially shake off the influence of more conservative Xianbei factions and proceed with his court re-structuring agenda．From a geo-strategic perspective，by re-locating the capital to Luoyang，it was only approximately eighty miles from the closest border zone with the Southern Qi-the Nanyang area．In view of this situation Xiaowendi's immediate military goal seems to have been to turn the area between the Huai River and the Yangzi River into a contested buffer zone in order to ensure tight security north of the Huai thus relieving the strategic pressure on the Luoyang capital.
Xiaowendi conducted three major campaigns to this effect in late 494-95，mid-497-late 498，and early 499．After this latter campaign，in the fourth lunar month，Xiaowendi died of illness on the North China Plain．The first campaign became focused on the south bank of the middle Huai River and attempted to seize the major Southern Qi fortifications there．None were seized，though the campaign did shake Nanjing．The second campaign was strategically the most significant．It focused on the zone between Nanyang and Xiangyang（the latter being on the Han River north of modern Wuhan）and successfully seized Nanyang and forts to its south．This was an unprecedented breakthrough which was confirmed by the 499 campaign which beat back a Southern Qi counter-offensive．Xiaowendi clearly understood，as he said in a conversation to Li Zuo李佐，one of the Li related to Li Chong，“the [Nanyang] region is the southern door to Luoyang.”（WS 29.895）Nanyang is also the back door route up the Dan River and over the Qinling Mountains to Chang'an．Notably，Xiaowendi's Nanyang success was also achieved by purely military force．In the past（and near future，as well）strategic Northern Wei gains against its southern opponent were initially achieved by strategic defections which were subsequently consolidated and then expanded．The Northern Wei acquisition of the North China Plain between the Yellow River and the Huai River was first achieved in 467 when several Southern governor generals，fearful of being implicated in a violent succession dispute in Nanjing，defected to the Northern Wei．The later Northern Wei acquisition of a major Southern Qi fortified city on the south bank of the Huai，Yuzhou，or Shouchun（re-named Yangzhou by the Northern Wei），in 500 occurred in the context of the accele-rating political disintegration of the Southern Qi and the decision of the commanding officer to defect to the Northern Wei．Once entrenched in Yuzhou the Northern Wei was able to briefly expand its presence to forts south of the city（a number of these extended positions were lost to the Liang regime in 507）．It was also able to expand west of Yuzhou and link-up with its Nanyang holdings．In 505 the city of Hanzhong，deep in the Qinling Mountains on the Han River defected to the Northern Wei and was turned into a base area for sporadic forays into Sichuan．（See the citations by Li Wencai，90）
By virtue of his military success at Nanyang Xiaowendi was able to realize his immediate strategic goal and partially secure the area north of the Huai from attack from the southern regimes．The strategy and the large scale use of military force to attain it were bequeathed as an honored legacy to his successor，Shizong．
What makes Xiaowendi's military campaigns interesting is that he personally accompanied his armies on all three campaigns．In between attending to field campaigns he staged two northern tours．Most of 494 was taken up by his tour of Pingcheng and the Six Garrisons in Inner Mongolia．This tour was politically crucial because at Pingcheng Xiaowendi directly confronted elite Xianbei critics of his Luoyang move and intelligently responded to their concerns．Then，immediately after returning to Luo-yang in late 494 he undertook the first southern campaign．The second northern tour occurred in 497 immediately following the Mu Tai affair at Pingcheng．Xiaowendi returned to Pingcheng，personally approved the various punishments meted out to the Mu Tai rebels and then proceeded from Pingcheng to the western border of modern Shanxi Province to receive the surrender of tribal groups in the area．From there he went on to tour Chang'an．Upon returning to Luoyang in the sixth lunar month of 497 he started organizing the second southern campaign of 497 through 498．Xiaowendi's military campaigns were inextricably intertwined with his ethnically sensitive court based“domestic”political agenda．Xiaowendi was an incessant，frenetic presence on the Northern Wei political“topographical”scene．His behavior was similar to that of the battle tested founder of a new dynasty than that of the sixth emperor of an established regime．Xiaowendi clearly understood that he had re-set the entire imperial agenda-the move to Luoyang was similar to establishing a new dynasty．His personal presence was crucial in driving the regime forward，consolidating loyalty to the throne，and enabled him to dominate his Tuoba agnates and Xianbei court elite houses．The war-making itself was a key means of keeping critical conservative factions under pressure to support the throne-it had direct political ramifications in addition to its direct military strategic and tactical implications．Ironically，the behavior of this so-called“sinicizing”emperor was very much in tune with the traditional Xianbei warrior ethos.
In the context of this discussion of military affairs the Six Garrisons issue is relevant．They were portrayed as bastions of the original Xianbei military ethos and thus worthy of attention and concern from the central court．Unfortunately for the Six Garrisons，their strategic value had been reduced to a secondary status．This was due，in part to the removal of the capital to Luoyang，distant from the Inner Mongolian frontier and the focus on the Southern front．More significant，however，was the fact that the Rouran nomad confederacy had simply receded as a strategic threat（Hamaguchi 137-8）．Xiaowendi staged a major successful campaign against them in 492 and almost ignored them thereafter．Indeed，in the tenth lunar month of 506，during the reign of Xiaowendi's successor，Shizong，a Rouran mission arrived at Luoyang seeking a peace treaty，presumably on terms of parity with the Northern Wei．Shizong turned them away stating that the Rouran confederacy was not the power that they used to be．He added that if they came as tributary supplicants he would not leave them orphaned and alone．（ZZTJ 146.4568）
Earlier，in 472 during the reign of the“retired”emperor Xianzu when the northern border was more active，the senior Xianbei courtier，Yuan He源賀 proposed a revamping of the northern defenses（WS 41.922）．In his account，each year from the fall through the end of winter large cavalry forces were mustered to aggressively patrol the frontier and strike deep into nomad territory in order to disrupt nomad grazing cycles and military preparations．These forces were mustered from both the Six Garrisons and cavalry forces based around Pingcheng and it was considered an expensive endeavor．Yuan He proposed a permanent stationary defense force comprising pre-existing garrisons with intervening defensive walls equipped with troop emplacements and heavy artillery．The northern defense force would be partially self-sufficient．This suggestion was tabled，but it was very similar to Chinese imperial defense models.
In 503 and again in 504 Yuan He's second eldest son，Huai懷，was dispatched to the Six Garrisons frontier zone to investigate their situation and report back to the throne．He found the usual corruption，recommended the equitable re-distribution of irrigated farmland and submitted the following military recommendation：given the lack of significant military activity on the border the garrison staffs should be reduced by forty percent．In 504 following a Rouran raid he returned to the area and noted increasing lawlessness among formerly submitted border tribal groups，manpower and materiel shortages，and drought conditions experienced by the garrisons．Yuan Huai then resurrected his father's policy proposal of creating fixed defensive structures between the major garrisons with full logistics and command support．The text notes that these suggestions were implemented（ZZTJ 145.4533-34，4544；WS 41.926-28），This is supported by archeological evidence showing increased building of defensive walls，subordinate outlying fortified posts and warning stations，and the probable strategic re-location of the Woye 沃野Garrison．（Sagawa 5-8，25-6）Zhang Jinlong（vol.9，245-46）notes that at some point during the reign of Shizong for the first time one general was put in charge of two or more garrisons as a frontier governor-general（dudu都督）．Sagawa dates an early version of this practice to the Xiaowendi period（Sagawa 43）．Overall，this would seem to reflect the personnel reduction and systemic command re-organization policy recommendations of Yuan Huai．The Northern frontier was re-organized into larger sectorial commands that could coordinate more readily than individualized garrison commands of equal status.
The proposals of both Yuan He and subsequently his son，Huai，would impose a much more routinized，administratively disciplined approach to securing the Six Garrisons frontier zone（Hamaguchi Shigekuni notes the direct relationship between the earlier proposal of the father and the expanded later proposal of the son．See Hamaguchi 128）．Assuming this system could continue to deny nomads easy access to the grasslands of Inner Mongolia，then the very expensive and manpower intensive central cavalry fall and winter border patrols and the traditional periodic，aggressive cavalry strikes deep into nomad territory north of the Gobi which supported the vaunted Xianbei martial ethos were anachronistic and would be displaced by a more“capital intensive”fixed defensive infrastructural system．Very importantly，Sagawa points out that beginning in 467 when the Huaibei region was acquired during the reign of Xianzu，and accelerating dramatically after the 493 re-location of the capital to Luoyang during the reign of Xiaowendi，the active Southern frontier on the Huai River siphoned off large amounts of manpower and materiel．Yuan Huai's policy would enable the Northern Wei to maintain the Northern frontier with reduced manpower．（Sagawa 21-2，37）Given this emerging trend in Northern Wei strategic thinking combined with the new focus on the Huai River military front，then it was clear that the Six Garrisons were slated for secondary strategic status.
The argument that the Six Garrisons were ignored by the Luoyang court on a long term basis and that Garrison personnel were deprived of advancement opportunities to serve at the court，points made by Wei Langen 魏蘭根（Bei Qi shu 23.329-30）and the Guangyang Prince Yuan廣陽王淵（also referred to as Shen深．WS 18.429-30）after the outbreak of violence in 523，seem overdrawn．Xiaowendi personally toured the Garrison zone and as noted above his successor，Shizong，paid close attention to Northern Garrison developments．Even the Suzong court，consumed by its factional concerns，was aware of the situation in the Northern Garrisons．In 519 it sent a member of the Department of State Affairs，Zhangsun Zhi長孫稚，to inspect the Garrisons．（Zhang，vol.9，104）Further，during the Yuan Cha period of court dominance the various Northern Garrison commands were directly involved in facilitating the defections of the Rouran kaghans，Anagui 阿那瓌and Poluoman婆羅門．The opening comments by both of the above courtiers that in the early Northern Wei the best and the bravest served in the Northern garrisons and viewed this service as beneficial to their future advancement，but that this tradition and accompanying martial ethos was undermined by the middle period of the Northern Wei（an oblique reference to the reign of Xiaowendi，in particular）seems to be winsome nostalgia．Warfare had become both more complex and more focused on one particular front in the south．Fast moving cavalry campaigns and individual heroics typical of steppe warfare were no longer as central as they once were．Siege warfare，infantry campaigns，and riverine warfare were much more relevant to the Huai River combat zone．Court institutions were becoming more complex and placing greater demands on administrative abilities.
Individual biographies indicate that access to court positions from the Garrisons was still possible．The biography of Xi Kangsheng奚康生，who became a leading commander of palace guard forces before his death in 521 traces a career from a Northern Garrisons local officer to exemplary combat performance on the Huai River．This was recognized by Xiaowendi and Xi Kangsheng was promoted to high court position．（WS 73.1629）Hulü Ping 斛律平，a Tieluo Turk from the Rouxuan Garrison 柔玄鎮served on the palace guard sometime between 500 through 504，fairly close in time to the outbreak of the Six Garrisons violence in 523．（Bei Qi shu 17.228；Zhang vol.9，422）
The Six Garrisons rebellion was triggered by extended droughts and food shortages that inflamed relations between the Garrisons and the Rouran who were suffering from the same conditions-thus accounting for the damaging Rouran raid of 523．Desperate people made the situation even more desperate．This ecological crisis was exacerbated by the rigid centralized control the Northern Wei court exercised over all aspects of its military affairs．The first outbreak of violence，which occurred in 523 at the Huaihuang Garrison懷荒鎮，was a food riot by the local Garrison inhabitants demanding that the Garrison Commander，Yu Jing于景，open the emergency granaries，which he refused to do．Yu Jing，who was killed by the rioters，was a high level courtier，brother of the deceased Yu Zhong于忠，who dominated the early years of the Suzong regency．Most probably，Yu Jing lacked authorization from Luoyang to open the granaries and was afraid to act upon his own initiative．As an example，prior to the outbreak of the Six Garrisons violence the Commanding Officer of the Wuchuan Garrison 武川鎮，Ren Kuan任款，unilaterally opened the granaries even after a request to implement food subsidies to the court was turned down．He was impeached for this action and had to be defended at court by a high ranking courtier，Li Ping李平，before charges were dropped and his actions recognized as appropriate for the si-tuation．（WS 65.1453）The subsequent，more organized rebellion begun in 524 may have been an opportunistic power grab by Poliuhan Baling破六韓拔陵，a local tribal leader，possibly of Xiongnu descent，and a junior officer in either the Woye Garrison沃野鎮 or the Huaishuo Garrison懷朔鎮．（Zhang vol.9，230 ff）The initial violent impetus came from outside the Xianbei community.
The geo-strategic situation discussed above（prior to the outbreak of the Six Garrisons Rebellion）was bequeathed to Xiaowendi's successor，Shizong．Beginning in 500，a few months after Shizong ascended the throne，through 509 there was incessant combat along the length of the Huai River and farther west in the Qinling Mountains border zone．Shizong was able to expand the basic military posture by retaining the Nanyang region and expanding fortified positions on the south bank of the Huai River in a line to the east of Nanyang．On the southwest border of the Northern Wei it obtained the strategic defection of modern Hanzhong on the Han River in the Qinling Mountains bordering on Sichuan．However，in 507 the regime suffered a strategic set-back regarding Northern Wei positions previously seized in the middle of the southern Huai combat zone．From the perspective of offensive operations against the south this combat zone was viewed as the most strategically crucial and militarily prestigious．Combat on the Huai abated from 510 through 514 and picked up again in this latter year when Shizong ordered an invasion of Sichuan that would take off from Hanzhong in the Qinling Mountains．He was probably seeking out a military front with possibilities for expansion since further Northern Wei advances south of the Huai were stymied and frustrated．This Sichuan campaign never became operational because Shizong died in 515 and the campaign was cancelled.
Shizong carried forward or at the very least，maintained the court re-structuring agenda of Xiaowendi and his geo-strategic vision．Where the new emperor exhibited extraordinary originality and openness was in his conduct and manipulation of court power intercourse，bringing court factional political behavior closer to conduct in a standard Chinese court．Structure and process were becoming congruent．The traditional Northern Wei monochromatic factional scene of agnatic power manipulation was opened up to an entire polychromatic array of factional constellations actively manipulated by the throne．Matrilateral relatives and affines，Chinese courtiers，a screen of random imperial favorites，not to mention the usual agnatic and Xianbei groups were all invited to participate and compete with each other．A multi-colored shifting kaleidoscope of competing factions，cliques，individualized social groupings all jump out of the pages of the staid dynastic annals.
Under Shizong，for the first time females designated as Empress actually gave birth to royal children．Shizong had two serial empresses and both gave birth to male offspring who died young or in infancy．The heir to Shizong，Suzong，was born from a concubine，Ms．Hu 胡（later Empress Dowager Hu，also titled Empress Dowager Ling 靈），in 510 and declared heir apparent in 512．These women were not forced to commit suicide and they functioned like Chinese style Empresses．Shizong's first empress，Ms．Yu 于，of an elite Xianbei family，was crowned empress in 501 and died in 507（rumored to have been poisoned，along with her son，by Shizong's maternal uncle，Gao Zha高肇，see WS 13.336；154.2431）．The next empress，Shizong's matrilateral cross cousin，Ms．Gao，of Korean descent（Li Ping，164-79，has a detailed study of the ethnic origins of the Gao family）with a thin，possibly non-existent，family political base at the court，was installed in 508，was forced to enter a convent in 515 after Shizong died，and then was executed in 517 by the reigning Empress Dower Hu（the biological mother of the reigning emperor，Suzong）．
Shizong not only had real empresses-he actively utilized his affinal male relatives（in the case of the Gao family they were simultaneously his matrilateral relatives）as court factional leaders and policy advisers．The extent to which Shizong incorporated these new kin elements was boundary breaking．In the post 490 period Xiaowendi had kept away from his biological matrilateral relatives，probably out of concern that he would offend the entrenched Feng 馮 family and their associates．In contrast，Shizong appointed his maternal uncle，Gao Zhao，to high imperial service in the Department of State Affairs in 501 and Zhao's younger brother and nephew to high inner court postings seven years before he installed his maternal cousin，Ms．Gao，as Empress．（Kubozoe 393）Shizong's mother died suddenly in 497 probably poisoned in accord with standing Northern Wei Inner Palace policy up to that time．In that same year（501）Shizong installed his first Empress，Ms．Yu．She was the niece of Yu Lie于烈，a senior powerful Xianbei courtier and a crucial figure in crushing a potential power seizure by Shizong's eldest uncle，Xi．Additionally，Lie's son Zhong忠，was a close personal companion and adviser to Shizong and he detested Gao Zhao．After Shizong died in 515 Yu Zhong would see to it that Gao Zhao was killed．From the very beginning of his reign Shizong was actively manipulating his affines against his matrilateral kin.
Shizong's form of actively encouraged competitive factional politics was very different from the stolid relatively non-competitve courtier blocs that his father，Xiaowendi，preferred to maintain and dominate through his personal charisma．What accounts for this dramatic stylistic difference？It may be due to the fact that Shizong was chronically ill（WS 91.1969）．For the entirety of his reign Shizong with one exception（a 502 journey to Ye）never left the vicinity of Luoyang．In my opinion，he lacked the direct personal presence and charisma of Xiaowendi and so compensated for this lack by devising elaborate forms of divide and rule factional manipulation in order to keep his court in line．There is also a related issue-Xiaowendi never engaged in the wholesale culling of his agnates as did many of his predecessors．（Eisenberg 28-91）The sixteen year old Shizong inherited a complete panoply of politically ambitious agnates of all categories，most especially his uncles．Xiaowendi may not have wanted to kill his brothers，but his son might have to do just that-which is exactly what occurred．In 500 the new emperor was seventeen years old and relatively mature by the standards of the time so he was not going to be easily manipulated．By 501 Shizong was actively constructing his byzantine factional apparatus．Indeed，Xiaowendi prior to his death in 499 probably directly contributed to the formation of Shizong's policy of multiple factional manipulation．The“Council of Six”六輔mandated in Xiaowendi's pre-mortem political will seemed purposely staffed to encourage the Council members to mutually check each other's ambitions and thus give the new monarch more maneuvering room（see the analysis and vivid discussion by Zhang Jinlong，2008 vol.8，16-31；1989 1-7）．The creation of the“Council of Six”also simultaneously catered to the Tuoba/Xianbei tradition of dynastic conciliar rule，especially prominent in the previous ascensions of the youthful Gaozong，Xianzu，and Xiaowendi，himself.
As noted above，Shizong encouraged multiple factional groups and competitive factional sub-groupings with changing membership over time．However，one group that was specifically targeted early in his reign were his uncles．Shizong originally had six uncles，one of whom，however，died in 499 at the age of thirty．In 500 there were five remaining uncles，but only one would survive beyond 508．Xi 禧，the eldest uncle was executed in 501 for an unsuccessful power seizure，another was beaten to death by a cuckolded husband（Feng Junxing馮俊興，ZZTJ 144.4486，a younger brother of the two previous Feng empresses under Xiaowendi）．Apparently Feng went unpunished by Shizong.The youngest uncle，Xiang 詳，was executed in 504 after losing a factional struggle with Shizong's in-law，Gao Zhao．Additionally，at this time（504）in the immediate aftermath of the execution of Xiang，following a recommendation by Gao Zhao，Shizong placed the residences of princes of the blood under strict palace guard（Zhang Jinlong，vol.8，56；ZZTJ 145.4540）．Finally his uncle，Xie 勰，who was a close confidante of Xiaowendi in his last years，was executed in 508 following trumped up charges leveled by Gao Zhao of being implicated in the revolt of Shizong's brother，Yu 愉，which occurred on the North China Plain at that time．Gao Zhao was a prime instrument Shizong used to advance his policy of close agnate surveillance for which Gao Zhao would pay with his life after Shizong died in 515.
The most definitive expression of Shizong's attitude towards his uncles prior to their sequential deaths was made in very early 501 immediately after emerging from his mourning responsibilities to Xiaowendi．Shizong ordered Yu Lie，commander of palace guard units and father of his close companion，Yu Zhong，to round up his uncles，many of whom were formal members of the council of six，and bring them to the royal presence．The biography of the youngest uncle，Xiang，notes that the uncles were arrested and placed in a mobile pen，similar to one used for the transport for sacrificial animals．Xiang's mother witnessed the arrests and prematurely panicked，assuming that her son and his brothers were to be executed．（WS 21.562）After the drama of the round-up and its implied threat，Shizong personally informed the uncles that they were temporarily retired from active court service and that he，Shizong，would directly take charge of all court policy making．A purposeful exception was made for Xiang，who was retained on active service．Shizong was purposely pitting the uncles against each other．（WS 20.577-78，580；ZZTJ 144.4482，4502）Then，to round affairs out，he pitted his affinal relatives，the Yu family，against his uncles．The role of Yu Lie in arresting the uncles was noted above，and when the eldest uncle，Xi，rebelled later in 501，Yu Lie was responsible for the capital defense and the security measures which resulted in Xi's rapid capture．In late 501 Lie's son，Yu Zhong，was appointed to be the chief of staff 長史 to Xiang，whom he publicly denounced for embezzling government supplies for his personal use．（WS 8.194；ZZTJ 144.4502）They actively disliked each other．Simultaneously，Yu Zhong was also pitted against Shizong's matrilateral relative，Gao Zhao，whom Yu Zhong would have the pleasure of executing in 515 following the death of Shizong．Gao Zhao tended to be utilized in implementing policies designed to circumscribe the power of the brothers and uncles，and policies identified as detrimental to Altaic court households．As an outsider he was the perfect scapegoat．Finally，a random screen of personal favorites was erected as a defense against all the above-mentioned individuals．This explosion in factional activity was abetted by Shizong，and he seems to have kept careful tabs on all the court personnel whom he was manipulating．
Shizong never let individuals feel too comfortable in their positions as imperial favorites．Shizong's factional manipulations took on dialectical proportions of incessant change and paradoxical transformations．In late 503 he decided to remove senior members of the Yu family from their positions of trust．From 497 through 501 the position of Palace Guard Commander（ling jun 領軍）was held by Yu Lie．Lie was succeeded in this position by his brother，Yu Jin 于勁from 501 through，nominally，506．Beginning in late 503 Shizong moved against the Yu family members．In circa the eleventh lunar month of 503 Yuan Huai toured the Northern frontier on behalf of Shizong as a imperial commissioner with executive authority．As part of his duties Yuan Huai impeached and had relieved of duties Yu Zuo 于祚，Commander of the Woye Garrison and the eldest brother of Yu Zhong．The Zizhi Tongjian lauds Yuan Huai as a brave servant of the throne for daring to confront the nephew of the powerful Commander of the Palace Guard，Yu Jin．（ZZTJ 145.4533）Yu Jin's daughter was the Empress Yu．However，Yu Jin himself will shortly be under attack，his assumed court influence but a hollow shell．The following month（actually early 504）the authority of Yu Jin was indirectly undermined in the Zhao Xiu 趙脩 Affair which has been incisively analyzed by Hu Hong.（Hu 122-23；ZZTJ 145.4535-36）Zhao Xiu was one of the random personal favorites of Shizong whom Gao Zhao eagerly supplanted，though Shizong would subsequently rotate in other personal favorites．Zhao Xiu had become a political and social burden at court so Shizong simply dropped him-there were always others to replace him with．As Hu Hong notes，Shizong authorized units of the Palace Guard to seize Zhao Xiu in the residence of Yu Jin where they were discussing mundane business arrangements．Zhao Xiu was then taken to the Palace Guard headquarters where he was tortured and eventually died from the mistreatment．Yu Jin was completely unaware of these looming developments and was not involved in the interrogation of Zhao Xiu even though it occurred in his own headquarters establishment．This can only be interpreted as an insult to the authority of Yu Jin．Shizong hit two targets simultaneously．In 506 Yu Jin was sent out as Governor General of two provinces in the western portion of the Wei River valley．（ZZTJ 146.4561）It is not clear who commanded the Palace Guard between 506 through 508，but in the latter year a Tuoba，Yuan Zhen 元珍，a known sympathizer of the Gao Zhao faction，was appointed ling jun from 508 through 512．（Zhang Jinlong 1995，58）
At the same time that senior Yu family members were being downgraded Shizong was actively courting a junior member of the Yu family，Yu Zhong，his companion and entourage member as an apparent counterweight to Zhao Gao．Contrary to Gao Zhao's requests，Shizong retained Yu Zhong at the court and among his titles during the period 505 through 512 was command of the Left Palace Guard（ling zuowei jiangjun 領左衞將軍．WS 31.742）This command would make him a direct subordinate of the ling jun from 508 through 512，the pro-Gao Zhao Tuoba，Yuan Zhen（see Hamaguchi 1971）．In 512 Yu Zhong will be appointed ling jun，a position he will hold until 515 into the reign of the succeeding emperor，Suzong．This sort of“bracketing”of personnel from opposing cliques within the same work unit or chain of command was a typical tactic employed by Shizong．
While Shizong was a master of factional manipulation，the high echelon personnel of the Department of State Affairs，Secretariat，and Chancellory were deceptively stable over extended periods of time．（Ershiwushi Bubian，vol.4，4509-12）．Nevertheless，the regime was staggered by a series of setbacks in 507 through 508 which were reflected in policies adopted and strategic personnel shifts．A second temporal node of change was the personnel changes in 512.
For all the discussion of the influence of Gao Zhao at the court，in actuality from 501 through the ninth lunar month of 507 he was the Right Deputy Department of State，ranking below the Left Deputy．From 502 until his death in 506 the Left Deputy was Yuan Huai，the same who conducted the 503 Northern border tour of inspection．Yuan Huai was a Xianbei and a deeply trusted policy adviser to Shizong．From 506 through 512 the Left Deputy was Shizong's younger brother，Yi．Interestingly，in 507 Zhao Gao was promoted over Yi to be the head of the Department of State Affairs（shangshu ling 尚書令），a position he held until 512 while Yi remained Left Deputy until that same time．Yi detested Gao Zhao，and again we see Shizong's“bracketing”approach to court personnel assignments（WS 22.591；83.1830）．
The events of 507 and 508 merit close attention as a crisis period that triggered personnel and policy decisions that influenced subsequent events in an almost tandem fashion．By the fourth lunar month of 507 an extended siege of the Liang fort of Zhongli 鍾離，located on the south bank of the Huai River to the east of the Northern Wei position at Yangzhou（Northern Wei name for the city）failed disastrously．If Zhongli had been taken then the Northern Wei would have been in position to absolutely dominate the region south of the Huai．With Zhongli still standing，the southern Huai region remained a combat buffer zone．However，the failed siege resulted in more than simply a standstill，it was a strategic failure that compromised the Northern Wei position in the region．Northern Wei outposts in the middle of the southern Huai region were lost to Liang counter-attacks．The Northern Wei military losses at Zhongli itself were considered staggering，with figures of 40％ to 50％ manpower losses for an army allegedly totaling 100，000 men．（WS 19.500；59.1315）As of the eighth lunar month of 507 the commanding officers of the campaign，some of the most experienced field commanders of the Northern Wei，including a leading member of the royal family，Yuan Ying 元英，were either cashiered or dramatically demoted in rank（ZZTJ 146.4574．The demotions would last only one year since the onset of another crisis resulted in their reinstatement）.
That the throne and the entire court were deeply shaken by the military setback followed by the dramatic personnel demotions is manifested by the imperial edict of the ninth lunar month of 507 granting all central court personnel（including supernumerary personnel）a universal one degree（yiji 一級）promotion in rank．（WS 8.204）To scholars specializing in Northern Wei administrative and institutional history this 507 edict is famous，or perhaps infamous，marking a watershed in how the throne appointed，evaluated，and promoted court personnel．The administrative and long term political ramifications of this edict will be discussed below，but it is important to note the political rationale behind its promulgation – to settle and appease the court after a military disaster on the Huai River front．As part of the 507 edict the occupant of the position of head of the Department of State Affairs，a senior elderly Tuoba，Yuan Jia 嘉，was promoted to the high honorific post of Minister of Works（sikong 司空），he will die in early 511．Gao Zhao was then put in his stead．Gao Zhao was already a controversial figure to the court elite，a non-Xianbei who was identified with factional assaults against members of the royal family．In the eyes of Shizong，though，Gao Zhao was a pliant tool and scapegoat to be used to further the power policies of the throne．（Zhang Jinlong 1992，120）In view of the delicate political situation，Shizong apparently wanted the Department of State Affairs secured by a client whom he could absolutely trust to both implement his will and screen him from any criticism resulting from unpopular policies．
The tension at court was raised the following month（tenth lunar month of 507）when the Empress Yu died suddenly under unexplained circumstances，possibly from disease．Later，her one year old son will also die from disease in the third lunar month of 508，leaving Shizong bereft of an heir apparent．Court rumors blamed Gao Zhao for the death of both individuals．（ZZTJ 146.4575；147.4581）In the midst of this swirl of events a sign that the Zhongli defeat still reverberated at the court and even among the populace is the imperial edict of early 508 granting a three year land tax exemption to households whose men had fallen at Zhongli．（WS 8.205）
The following year，508，would turn into the climax year for the throne，triggered by the brewing factional tensions in the court and then followed by a dangerous resurgence of Liang military efforts on the Huai River and beyond．Once again，the Zhongli fiasco and the 507 personnel policies of Shizong are directly related to the events of 508．At approximately the same time that the 507 edict was promulgated along with its associated personnel transfers，the second eldest brother of Shizong，Yu，was transferred to be governor of Ji Province 冀州 on the North China Plain．From 501 until 507 Yu was Inspector of the Secretariat（zhongshu jian 中書監），a position that he apparently chafed under since he felt that his younger brothers received higher status posts and treatment．（Bubian 4508-09；WS 22.590）Yu was becoming a problem for Shizong exacerbated by the circumstances of birth．The official empresses of Xiaowendi，the two Feng sisters who were nieces of the deceased Empress Dowager Wenming，had no known children and seemed to have functioned as“placebo”empresses．Shizong and his brothers were the offspring of ranked concubines．Consequently，the type of status preference granted to uterine brothers born of the Empress common in Chinese regimes，which placed half brothers in a distinctly inferior political status position was irrelevant．Birth seniority seems to have been particularly important to the brothers，Yu being the second eldest after Shizong．Yu used his status at court to build a social image as a patron of the arts，Confucian scholarship，liberal donor to Buddhist institutions，and all around free spender．Shizong，allegedly disgusted with Yu's profligate ways had him arrested，flogged，and then sent out as governor of Ji Province in 507．（WS 22.590）Yu had also crossed the throne by ignoring his main wife，a sister of the Yu empress in a marriage arranged by the throne，in favor of a Chinese singer he had met in the provinces．What may have concerned Shizong more than Yu's prodigal behavior was the combination of his brother's known frustrated ambitions in conjunction with his salon patronage and the creation of an elite court clientage network．Given Yu's status ranking as the next eldest brother to the Emperor，the throne could only view his emerging client network with alarm．（Li Wencai 86）Needless to say，Yu's disregard of the throne's preferred marital arrangements exacerbated the situation-it was disrespectful.
In the seventh lunar month of 508 Shizong decided to appoint the niece of Gao Zhao，who was then head of the Department of State Affairs，as his new Empress．Close Tuoba agnates of the throne despised Gao Zhao．Xie，Shizong's prestigious uncle，strongly opposed this choice．（WS 21.582）．At about the same time Gao Zhao implemented personnel policies which reduced the income from noble titles and he also demoted or cashiered allegedly unqualified court personnel who had previously earned military merit and had had the merit points converted into court ranking and appointment to court positions．（ZZTJ 147.4581-82）These policies were probably the outgrowth of the impact of the 507 edict on court personnel policy and the ongoing effort by the throne to cope with large numbers of individuals seeking limited number of court positions．Nevertheless，Gao Zhao's yeomen efforts on behalf of his monarch were viewed by the court as gross violations of court tradition，complaints were aired，the policies were viewed as outrageous，and of course，Gao Zhao received all the blame as intended by Shizong.
In this context，in the eighth lunar month of 508，Yu rose up in rebellion in Ji Province，called himself emperor，and claimed that Gao Zhao was intent upon killing Shizong and seizing the throne．Yu was attempting to capitalize upon the disdain and fear that the court had of Gao Zhao．Unfortunately for Yu，the court and the provincial garrisons held firm．Yu was isolated in Ji Province and quickly captured and executed within a month．Nevertheless，the failed rebellion had significant ramifications．Prior to the suppression of the rebellion Shizong，operating through Gao Zhao，had his uncle Xie falsely accused of collaborating with Yu and then killed．With the exception of one remaining uncle，Yong 雍，Shizong had killed or tolerated the elimination of four of his five uncles．The individual who led the Palace Guard contingent that arrested Xie and oversaw his death was Yuan Zhen 元珍，Commander of the Left Palace Guard（zuowei 左衞），who immediately after was appointed Commander of the Palace Guard（ling jun）（Zhang Jinlong 1995，58．The ling jun position may have been vacant between the exit of Yu Jin in 506 and the appointment of Yuan Zhen in 508）．Zhen was a known sympathizer of Gao Zhao．（WS 21.582-83）At approximately this same time command over the fortified defenses surrounding Luoyang（si zhongfu 四中府）were transferred to the Palace Guard Command under Yuan Zhen，making the Palace Guard Commander a much more influential position than ever before．（Hamaguchi 103）This probably reflected Shizong's concerns，in the light of the Yu affair，that armed uprisings in the provinces not be able to penetrate the capital district．This personnel arrangement will last until 512 when Shizong arranged a new“rotation.”
The impact of the failed Yu rebellion was not limited to the court．Between the ninth lunar month of 508（immediately following the repression of Yu）through the end of the year，the Northern Wei suffered a massive series of defections to the Liang along the entire southern front accompanied by Liang attacks against Northern Wei positions．The worst losses were in the previously secured zone north of the Han River and east of Nanyang．（ZZTJ 147.4587）The timing of the defections suggests a perception that the Yu rebellion indicated a mortal weakness in the Northern Wei court．This perception was inaccurate and the Northern Wei fought back and regained the locations that had defected．The cashiered commanders of the Zhongli campaign were restored to their former ranks and were important participants in the Northern Wei“reconquista.”In fact，the Huai River front stabilized at its new post-Zhongli equilibrium until after 515．Nevertheless，the fragility of the Northern Wei position was graphically illustrated，and the subsequent Northern Wei posture on the Huai River front was increasingly defensive in nature．（Zhang Jinlong，vol.9，124）Indeed，by 514，Shizong was seeking to open a Sichuan front while the Huai River front maintained its stasis．
The year 512 was the second nodal point in the Shizong reign marked by significant personnel changes．During the Period of Disunion，the leadership positions of the Department of State Affairs were both powerful executive positions as well as providing direct input in high imperial policy making discussions．In very early 512（long before the announcement regarding the heir apparent was made）Gao Zhao，uncle to the reigning Empress，was promoted from the potent position of minister（ling 令）of the Department of State Affairs to the higher ranking but relatively honorific post of situ司徒．The“promotion”was viewed by Gao Zhao himself as a diminishment of his political influence at court．（ZZTJ 147.4601）．Interestingly，Gao Zhao's political opponent，Shizong's brother，Yi，was also“kicked upstairs”from the powerful position of Left Deputy of the Department of State Affairs to a similar honorific appointment．Beginning from 512 through Shizong's death in very early 515 the entire complexion of the leadership of the Department of State Affairs was altered．From 501 through 512（with the exception of 508）the three top leadership positions of the Department of State Affairs were always filled．（Bubian 4510-11）The pattern was that two of the posts were occupied by Tuoba or high level Xianbei，and the third post by Gao Zhao．From 512 through 515 only one of the three posts was filled，usually the Left Deputy，and the Ministerial post（ling）was left vacant．The occupant of the Left Deputy post during this time period was an elite，administratively very competent Chinese courtier，Guo Zuo 郭祚．For the Shizong reign period the Wei Shu biography of Guo Zuo emphasizes his expertise in imperial law and court administrative regulations．（WS 64.1422-26）In particular，the two major memorials by Guo Zuo featured in the biography focus on the vexing issue of the job performance evaluation and promotion system，established by Xiaowendi，and that Shizong wrestled with throughout his reign period．Other personnel appointments also indicate significant personnel movement at this time．In early 512 Yu Zhong was first appointed to head the Northern Wei version of the Board of Works（duguan shangshu 都官尚書）and slightly later that same year，he was appointed Commander of the Palace Guard，replacing Gao Zhao's supporter，Yuan Zhen．（WS 31.742；Zhang Jinlong 1995，58）
The trigger prompting the above personnel alterations is not clear．One possibility is that it may have been the announcement of the two year old heir apparent（the future Suzong）in the tenth lunar month of 512．The previous Empress Yu had given birth to a son who died at one year of age in early 508．The current Empress Gao had earlier given birth to a son while still a ranking concubine，but the child died in infancy（WS 13.336）．One of Shizong's ranking concubines，Ms．Hu（the future Empress Dowager during the Suzong period），gave birth to the current child in 510，upon which the infant was placed in special protective custody to avoid any mishaps that could be delivered by a human hand or by disease．（WS 13.337）I assume that the announcement of the tenth lunar month of 512 was planned in advance and that the earlier personnel changes discussed above reflected the political alignments that Shizong wanted in place to support a future imperial succession．The situation，as it stood in late 512 was awkward，in that the reigning Empress Gao had not produced an heir and was politically forced to tolerate an imperial heir borne by a subordinate concubine．The presence of an“alien”heir apparent also had obvious implications for Gao Zhao's own position at the court．Another possibility is that Shizong，who was confronted with an unending series of droughts and floods on the North China Plain and on the Northern frontier where the Six Garrisons were located，needed administratively competent personnel to manage the developing emergency situations for him．Guo Zuo would certainly be a qualified administrator capable of coping with the rolling food crisises．It certainly appears that the figures associated with factional in-fighting at the court were removed from executive leadership positions and experienced administrative personnel put in their stead．In the case of Yu Zhong's appointment to be Commander of the Palace Guards，his family（and his own personal career track）provided him with deep ties to the institution．Perhaps it was both of the above two factors coming together which nudged Shizong into implementing these personnel arrangements.
Then in the twelfth lunar month of 512（actually early 513）Shizong ordered that all leading provincial officials who had previously been impeached by the Censorate but had been pardoned due to the promulgation of a general edict of grace，and，provincial officials who received a performance evaluation grade of average（zhong 中）were all to be replaced．（WS 8.213）The evaluation grades were probably the result of a broad re-evaluation of all personnel performance spanning a number of years that was conducted in very early 512．（WS 8.211）The overall impression is that Shizong engaged in a major re-shuffling of court personnel at all administrative levels along with a partial purge of courtiers who had associated themselves with Gao Zhao．In late 514 Gao Zhao was sent out from the court as the Commanding Officer for a massive invasion of Liang controlled Sichuan which would take off from the Hanzhong area controlled by the Northern Wei．Gao Zhao had no known military command experience and had been effectively uprooted from his court political base．As far as Shizong was concerned，Gao Zhao's utility at court had been exhausted．However，one can state that Shizong was offering Gao Zhao an opportunity to replenish his court“political capital”by achieving military success．（Li Wencai 91-2）While Gao Zhao had no military experience，he was surrounded by excellent field commanders．All he had to do was ride on their coattails and hopefully emerge successful from the crucible of an active military command．The invasion，however，was cancelled due to Shizong's death in early 515．When Gao Zhao returned to court to pay mourning respects to his deceased monarch he was promptly strangled in a pre-arranged assault orchestrated by Tuoba and Xianbei court elites．（WS 83.1831）
Shizong may have been one of the most talented imperial political manipulators in Chinese history．（Li Wencai 83-4）Zhou Yiliang，for different reasons，also rated Shizong as an outstanding and activist emperor who successfully carried through the policies initiated by Xiaowendi．（Zhou Yiliang 1985，317-20）
The five years old Suzong ascended the throne in 515 and given his age，the situation demanded that a regency council be established to rule in his name．In the process the Gao family and some of its most prominent supporters were executed or purged．The Empress Gao was sent to a convent and then executed in 517．The conciliar regency front seems to have been structured based on experience culled from previous practice．The Empress Dowager Hu（also referred to as Empress Dowager Ling）was a ranking concubine of the deceased Shizong and the biological mother of the five year old Suzong．She was granted significant political authority and functioned as a political equal with other Xianbei male elites，though she does not seem to have tried to elevate her agnates to high posts besides the honors bestowed upon her father．To an extent，Hu functioned much like the previous Empress Dowager Wenming，though，in ritual affairs she was much more aggressive，insisting upon personally participating in high imperial rituals previously closed to females．Indeed，if there is a Northern Wei precedent to the prominent ritual role Wu Zetian of the Tang assumed for herself and court females（and as far as I know there is no specific reference in Tang sources to such a precedent），then that precedent would be Empress Dowager Hu，not the earlier Empress Dowager Wenming 文明（see Gernet's brief comment in this regard，285）．However，as an independent political actor，Hu may have been less powerful than Wenming during the latter's extended regency period from 476 through her death in 490．Hu seems to have functioned more as a unifying and legitimist consensus figure for various Tuoba and elite Xianbei interest groups throughout the course of her political career．Later，in 528，when she no longer fulfilled this role she was brutally eliminated．In this sense，Hu was a more activist version of the early Wenming（465-471）when the latter responded to Tuoba interest groups in authorizing，as the formal Empress Dowager，the suppression of the Yi Hun clique in 466，thus assuring the position of the boy emperor，Xianzu．Wenming of the 476 through 490 regency period clearly had much more independent political cachet and authority．
Operating under the formal authority of Hu was Suzong's eldest and“closest uncle，”Yi懌，who was highly regarded at court and may have been personally close to Hu．（WS 22.591）．Another uncle，Huai 懷，was also prominent at the court．Two of the remaining three uncles had been executed during the reign of Shizong，and one was relatively inactive at the beginning of Suzong's reign．Yi，in particular，was a powerful policy maker and it seems one of his tasks was to keep an eye on the Tuoba agnates，particularly those closely related to the throne．Hu also saw to it that her brother-in-law，Yuan Cha 元叉，a distant Tuoba agnate married to her younger sister was given high level responsibilities along with his father，Ji 繼．Both men，Ji in particular，can be identified as supporters of the concerns and interests of more distant Tuoba agnates at the court．Below this level of Tuoba elites were Xianbei and Chinese courtiers who ran the court on a daily basis．Two，in particular were very influential，Yu Zhong and Cui Guang崔光，a Chinese policy adviser．Both men were crucial in securing Suzong's rapid succession，protecting the life of then concubine Hu，and in eliminating the Gao family faction．For purposes of stability and public image daily court affairs were publicly managed by now senior and somewhat distant Tuoba elites，Cheng 澄 of the Jingmu 景穆royal line and Xiaowendi's tangshu 堂叔（Xiaowendi's father's paternal first cousin）and Yong雍，Suzong's great-great uncle and sole surviving son of Xianzu．Both men were in their forties．Within the palace there were key eunuchs who were also important collaborators in seating Suzong on the throne.
Following the model set by Shizong，the regency was itself riven by factional competition which the Empress Dowager Hu attempted to manipulate in order to maintain a dynamic equilibrium that would not threaten the new court balance of power．Yi was at loggerheads with Cha（WS 16.404；22.592）and Yu Zhong at odds with Yong and a group of Chinese courtiers behind Yong．Yu Zhong initially commanded the palace guard（lingjun）but by late 515 had lost his court prominence and palace guard command was given to Yuan Cha's elderly father，Ji，from 515 through early 519，and then to Yuan Cha from 519 until his execution in 525．（Zhang Jinlong 1995，58；Zhang 2004，762-67；WS 16.404；ZZTJ 149.4656）Zhang Jinlong has consistently pointed out the crucial importance of the Palace Guard units in stabilizing disputed successions or enabling successful coup d'etats．（also Tanigawa 125）In the Northern Wei context the Palace Guard commanded by a lingjun fully emerges with Xiaowendi's 493 provisional court office restructuring document．（Zhang 1995，56）．Prior to 515 Yu Zhong，as lingjun，had proven to be an effectively aggressive court politician．Perhaps fearful of a recurrence of such a personality，Hu had opted to appoint an elderly but respected distant Tuoba agnate to the position，her sister's father-in-law，whose loyalty to her status quo could be assumed.
This moderately peaceful court factional environment was shattered by the violent coup of Yuan Cha in 520，who utilized his new position as lingjun，held since 519，to assert his dominance at court．Cha was from a distant collateral line of the royal family，a descendant of one of the sons of the founding emperor，Daowudi．Due to his marriage to Hu's younger sister he had been rapidly elevated to some of the highest court positions．To a degree，Tuoba political alignments in the Northern Wei court seemed to have been partially congealed around royal lines of descent which also served as a means of determining how far one's court career could advance．Relatively distant agnates were never dropped from the court，but simply relegated to lesser career paths．That is，a sense of lineage structured elite Tuoba and Xianbei court behavior to a degree unknown in the more open environment of the Chinese court．This rigidity was also reinforced by Xiaowendi's earlier“sinifying”court status determining legislation which focused on the relative court perquisites of specified lines of Tuoba and Xianbei descent．The biography of Yuan Tianmu元天穆，a sixth generation grandson of a pre-Daowudi Tuoba confederacy ruler，notes，“… because he was a distant agnate he had no hope of an elevated career and so threw his support to Erzhu Rong.”（WS 14.356）Yuan Zhao元昭 was considered incompetent at court and was cashiered by Xiaowendi，but his career later benefitted from the authority and influence of his cousin Hui 暉，an imperial favorite and high level councillor．Both individuals were descended from the sons of the father of Daowudi．One of Yuan Cha's closest factional supporters was his cousin，Yuan Faseng元法僧．Later，in 525，Yuan Faseng defected to the Liang from a strategic governorship on the North China Plain．This defection severely undermined Yuan Cha's factional dominance at the court．（WS 16.406）When Yuan Faseng announced his defection to the Liang he was，at first unsuccessfully，attacked by Northern Wei forces．In the course of events he captured a member of the Tuoba elite，Yuan Xianhe元顯和．Xianhe was a descendent of the line of the Jingmu emperor，specifically，the son of one of Xianzu's uncles．Xianhe was repelled by Faseng's behavior and stated，“We come from the same source，but different lineage segments（biepai别派）… we are all the bedrock of the royal clan.”（WS 19.449）
Given this status sensitivity to lines of descent，the（somewhat limited）violence incurred by Yuan Cha's sudden coup in 520 seems to have been literally shattering because it was directed against lineal status superiors．In part，violence against fellow agnates in the Northern Wei was not new．It could overflow all boundaries in the course of succession disputes．In terms of pre-emptive violence by sitting emperors，Northern Wei emperors had a tendency to cull the categories of brothers and uncles．Shizong's reign is a good example of an anti-avuncular focus．However，these same emperors were rather chary in making pre-emptive moves against moderately more distant agnates，like tangshu，even though this latter category was politically active and sensitive．Likewise，violence in the course of“normal”factional struggles seems to have been limited to Chinese courtiers and eunuchs．For example，before Yu Zhong was stripped of much of his court power in late 515 he had become enraged at Yuan Yong，who was managing daily court affairs in conjunction with Cheng．A clique of ranking Chinese courtiers were attempting to work with Yong to topple Yu Zhong．Zhong succeeded in ordering the execution of the Chinese courtiers but was dissuaded from killing Yong by his Chinese colleague，Cui Guang（WS 31.743）．Yu Zhong was not a Tuoba，but he was a scion of one of the highest ranking Xianbei lines at the court．Even the violence against the Chinese courtiers seemed to have disturbed the court．A month after the incident Empress Dowager Hu，who had by then been formally recognized as regent，orchestrated a court consensus resulting in a significant demotion for Yu Zhong．I assume if Yong had been executed，Yu Zhong would have paid with his life．High Tuoba agnates could only be manhandled by the throne，and under normal circumstances the throne had to have good reasons for doing so.
Yuan Cha's violent coup in 520 would put him in a dominant position at the court until his own violent demise in 525．The price of Cha's factional dominance at the court would be a fragmenting of Tuoba agnatic comity and uncontrolled factional splintering．A question to be asked is why Yuan Cha was able to even attempt such a coup．I believe that there was deep dissatisfaction at court with court personnel policy and the ongoing opening up of the court exacerbated by external military developments．In 519，a Chinese courtier，Zhang Zhongyu 張仲瑀，proposed that all military personnel be excluded from“pure”court postings，a policy very much in line with Xiaowendi's earlier status legislation．This proposal would also have placed significant restrictions upon promotion possibilities for young serving Guardsmen．This enraged the palace guardsmen from the elite yulin 羽林 and huben 虎賁 units，most of whom were probably Xianbei or members of other Altaic groups．（Zhou Yiliang 1997，132）Allegedly，approximately one thousand guardsmen rioted，damaging government buildings，burning down the residence of the Zhang family，beating and then immolating Zhang Zhongyu's brother and beating his father to death（the latter died from his wounds later，see WS 64.1432）．The court was shocked by the riot，but immediately dropped consideration of the proposal，and given the large number of participants in the riot，only charged and executed a small number of individuals，the remainder were pardoned．That frustration over general personnel policy was at the root of this violent outburst is indicated by the fact that the Minister of Personnel was immediately changed．A new policy of immediate rotation and promotion based on time in service for provincial serving court personnel was also implemented and apparently satisfied court personnel（WS 66.1479）．The Guardsmen riot triggered significant alterations to overall court personnel policy that went beyond the specific issue raised by the Zhang Zhongyu proposal．Ever since Xiaowendi's revamping of court structure and personnel qualifications based on both status and ability，the Northern Wei court was vexed by personnel problems related to implementing institutionalized procedures for promotion and demotion，reasonably rapid rotation of officials，and effective policies for opening up court postings to younger personnel.
Despite the multitude of laws and administrative regulations issued by the Northern Wei court（particularly after the onset of the mid-480's structural reforms）it often functioned in a slow，sometimes improvised fashion testifying to generalized organizational problems，personnel over-commitment，and the political expediences of the throne interfering with routine administrative functions．（Tao 96-7）This is not surprising given the multitude of orientations encompassed by a patrimonial court and specifically discussed by Yan Buke in terms of the significant social status concerns of the China based court．This aspect of the Northern Wei court probably does not make it any different than any number of other dynastic regimes established in the course of Chinese history．However，the Northern Wei situation，especially regarding personnel issues，takes on a particular salience due to the ultimate commitment of the regime to its“corporate”foundation and the guarantees it made to its elite courtier lineages and households.
Beginning in 493 and extending through 499 Xiaowendi issued a series of legislative acts which defined the positions and ranks of court offices and serving court personnel in terms of an elaborate and moderately inter-locking working court ranking and honorific title system．The importance that the throne placed upon the newly created administrative and personnel system is reflected in the unprecedented status granted to the Board of Personnel（libu shangshu 吏部尚書）as re-structured by Xiaowendi and maintained by his successors．Generally only highly literate and competent Chinese and Tuoba courtiers were appointed to head the Board of Personnel and its subordinate bureaus．（Yan Gengwang 1948，290，306，310）
Intimately linked with the reconstructed formal administrative structure and court ranking regulations was the famous elite surname fixing legislation enacted in 495．This legislation identified elite Altaic lineages with consistent records of court service dating back to the founding dynast of the Northern Wei，Daowudi（this starting point was very consciously enunciated by Xiaowendi and placed bottom limits on eligible Altaic households），and Chinese houses with related records of court service whose offspring would be eligible for guaranteed court service and appropriately ranked entry level positions and subsequent higher level court postings．Note that Xiaowendi created a service nobility-simply being high born was not a sufficient qualification for the above guarantees．（Tang Zhangru，vol.2，“…Shiyi，”80-3）
Xiaowendi also formally committed the court to distinguishing between“inner ranked”elite court posting and titles that were considered“pure，”some of which were even ranked in terms of three degrees of“purity”（a Northern Wei innovation），followed by“outer ranked”postings and titles appropriate for functional clerk positions and not appropriate for court gentlemen．Specifics regarding the above“purity”rankings are scarce and it is not clear to me if other“inner ranked”court posting and titles were specifically ranked as“turbid”or simply not given a specific purity ranking．Based on Southern Dynasties precedent，“turbid”rankings applied to inner ranked court posts that due to their burdensome nature or simply their practical administrative functions were considered unsuitable for a cultured gentleman．The term“turbid”did not necessarily refer to outer ranked posts，which were also unsuitable for a gentleman．An incident occurring at the Northern Wei court between 512 and 515 involving a serving courtier illustrates this ambiguity．（Tao 234-36；WS 88.1904）The courtier，Ming Liang 明亮，was personally promoted by Shizong one rank from a civilian supernumerary title that had a specific purity ranking to a military supernumerary title（these supernumerary titles，civilian or military，established the personal ranking of the individuals involved）．Ming Liang protested that this was a“turbid”title as established by prior imperial statutes and then requested that the throne offer him a different，acceptable，military title，which was subsequently granted to him．Shizong's initial reaction was interesting，in that he first replied to Ming Liang that all“inner ranked”titles and postings were suitable for accomplished gentlemen who advised the throne without distinction between“pure”and“turbid.”That is，it appears to me that the initial military title offered to Ming Liang probably did not have a specific purity rating and it was assumed at court to be a“turbid”title．The 519 proposal by Zhang Zhongyu seems to have been intended to make the“pure”versus“turbid”distinction among“inner ranked”titles and positions much more legally explicit．Xiaowendi also pioneered the creation of a rough system of“outer ranked”titles and positions for central court personnel employed in routine clerical positions，thus contributing to a long term trend of marking distinctions between the serving courtier and the mere literate clerk．In my opinion，Xiaowendi's intensive drive for“Huaxia”style legitimation impelled him in the direction of the semi-aristocratic fashion common since the onset of the Period of Disunion in the Southern Dynasties and assuming the guise of a“traditional”mode of proper social organization．On the other hand，this status legislation offered him a reasonable means by which to both honor the corporatist commitment of the dynasty to its elite Altaic members while at the same time limiting and defining their privileges．It was an extraordinarily astute political maneuver.
Xiaowendi，in the course of creating this massive status oriented administrative structure also enacted legislation designed to evaluate and promote all court personnel on a periodic basis．The court evaluation process has a long，hoary tradition in the Chinese canonical literature as well as in the Han dynasty historical tradition，all of which was fodder for the legitimating drives of Xiaowendi．（Dai 7-14）As established in legislation culminating in 494，all central court and provincial serving courtiers would receive a cumulative evaluation every third year which would determine their immediate career prospects in terms of promotion，demotion，or lateral movement at the same rank．Evaluation procedures would function in a vertical fashion with reports ultimately being sent to the Board of Personnel from subordinate levels for final decisions on all evaluations received．This would provide the court structure with the orderly personnel dynamism and movement necessary to rotate personnel and evaluate work performance and so also provide openings to new personnel entering the system．However，even under its creator，Xiaowendi，this evaluation and promotion system functioned on a somewhat ad hoc basis，being supplemented by alternative modes of evaluation，such as the use of central court dispatched inspectors to the provinces with full executive powers to evaluate and promote or demote．The evaluation system also had to cope with the existing varying tenures of currently serving officials（Tao 12-18，168；Dai 83）
Following the Xiaowendi period the evaluation and promotion system was constantly re-worked in order to accommodate larger numbers of personnel flowing into the court and the impact of political interference by the throne upon the normal functioning of the system．The most prominent and significant example of this development was Shizong's 507 edict of grace granting all central court“inner ranking”personnel a one degree promotion in rank．The legislative follow up to this edict of grace established significant precedents that lasted until the end of the dynasty．Those officials promoted by the edict of grace ended up having their subsequent promotion schedule lengthened compared to officials who entered the court personnel stream after the edict of grace．This seems to have been an effort to stem a flood of triennial promotions to ever higher ranks．Working central court officials continued to be evaluated and promoted every three years as under the original Xiaowendi statutes．Civil and military supernumerary courtiers（who were ranked and salaried but had ambiguous court functions）were now to be evaluated and promoted every four years with extra conditions attached to their promotions relevant to the impact of the 507 edict of grace．The Wei shu biography of Yuan Yong，21.553-54，the longest surviving younger brother of Xiaowendi，contains significant information regarding the impact of the 507 edict of grace on supernumerary courtiers．（On salaried supernumeraries see Yan Buke 1999，36 endnote 14）The central palace yulin and huben guard units were military ranked supernumeraries and so directly affected by the post- 507 personnel policies regarding court supernumeraries．（Tao 212）This post- 507 personnel policy development would appear to reflect the growing number of elite young men entering court service via the supernumerary route and awaiting an appointment to a functional office at the central court or in the provinces．Additionally，serving functional courtiers who were between appointments were often subsidized by the court by being given supernumerary ranks and salaries until a new appointment materialized．Shizong was trying to cope with an unwieldy number of such personnel．Finally，serving provincial official were not included in the 507 act of grace．Thus，by 507 Shizong had differentiated three major status constituencies in the court system：working central court courtiers，central court supernumeraries（civil and military），and serving court provincial officials．（Tao 55-9；Dai 251-53）．Then，in 512，for reasons that are not fully clear，perhaps related to previous delays in actually promoting and demoting provincial serving officials，perhaps also related to a purge of courtiers associated with Gao Zhao，Shizong ordered a sweep of all provincial officials who had not received a rating of average（zhong 中）and an entire new cohort of provincial officials were put in their place．（Tao 59-63）.
The herky-jerky functioning of the court hiring and job evaluation system was given its last major re-structuring in the 519 legislation triggered by the Guardsmen riot．The entire Shizong legacy was retained and additional adjustments were made to serving provincial officials．Serving provincial officials would receive their cumulative evaluation after their sixth year of service（no longer the third year，however，a total of six years in provincial service was not unusual even prior to the reign of Shizong．See Dai 119），after which they would have to step down from provincial service for another six years．This latter clause seems to be an innovation of the Suzong period．During this latter period former provincial personnel could either be directly re-appointed to a central court working post，or to a central court supernumerary post with an appropriate rank and salary（which is where most of them probably ended up），or simply reside at home．The second major readjustment associated with the 519 legislation was that re-appointments to provincial postings would be based on seniority-whoever was in line and had been waiting the longest would be automatically appointed based on their personal ranking and irrespective of their previous evaluation history（assuming no major gaffes or criminal activity）．（Tao 203-31）The goal，in the wake of the Guards riot，seems to have been to get personnel out of the supernumerary categories as quickly as possible and avoid tensions generated by intense scrutiny of personnel records regarding qualifications for provincial postings．In this instance the Wei shu text is very explicit in noting that overwhelming numbers of court personnel and a massive dearth of available postings were driving this policy．（WS 66.1479-80）
The consistent focus on differentiating serving provincial officials from central court officials and supernumeraries since 507 may not be as discriminatory as it seems．It appears to be based on the assumption that provincial assignments were viewed as an investment opportunity．If practiced cautiously and not too outrageously，after six years of provincial service one should have accumulated a reasonably secure nest egg．Stories of“corrupt”Northern Wei provincial officials are numerous and some of the legislation designed to counter this activity was draconian，but the issue continued to appear throughout the history of the regime．（Dai 198-212，218-19）As late as 522，during the Yuan Cha period of dominance，the court ordered its censorial staff to look into reports of provincial officials living lives of luxury and indulgence and engaging in monopolistic commercial ventures in their localities．（Tao 175；WS 9.233-34）Presumably，if these provincial officials had been able to control their ostentatious tendencies then this type of edict would not have had to be issued.
The result of the creaky manner in which the court personnel system functioned subjected elite houses at all levels of court society to enormous strains-they needed to funnel their adult members and their progeny into the system．The tensions engendered by this surplus of court personnel were most manifest in the supernumerary status groups，which included the central palace yulin and huben guard units．Finally，the regime could not renege completely on the guarantees it made to its elite court houses，though，it would try mightily to do so.
This personnel issue was made more complicated by the expectation that members of growing and ramifying royal descent lines expected postings commensurate with their lineal status-an expectation shared by members of non-royal elite Xianbei lines（Kubozoe focuses incisively on this issue on pages 488-9）．This basic expectation，which involved the categorization of households according to the various status categories created by the 495 legislation，was nonetheless a difficult and drawn out process．As late as 505 in the reign of Shizong households labeled as“Dai”（meaning their immediate forebears had moved from Pingcheng to Luoyang under Xiaowendi in 493 and later．This group was mostly Altaic but could also include Altaicized Chinese households）were still litigating their proper status category in terms of the 495 legislation．（WS 31.742）However，there were disturbing efforts by the throne to limit the access distant collaterals had to the court．The goal was probably to create more space for autonomous imperial decision making and personnel appointments．These efforts，not surprisingly，were causing increasing anxiety among distant royal collaterals．In 492 under Xiaowendi those collaterals who were not descendents of the sons of Daowudi（the founding Northern Wei emperor）or subsequent emperors were reduced one grade in noble rank and the practice of permitting progeny to inherit military rank titles was abolished（WS 7.169）．This not only reduced the status of the rank holder，but influenced the hereditary access their children had to higher court ranks at the entry level positions．（Yan 2011，383）Additionally，in 493 Xiaowendi initiated imperial ancestor worship services limited solely to the four previous emperors（the standard Confucian position advocated ancestral worship and mourning responsibilities for five generations counting ego）．Participants in these services were limited to the descendants of these emperors，more distant descendants were excluded from the ceremony（WS 7.171-72；Kubozoe 485）．
In early Suzong，in the seventh and eighth lunar months of 517，a series of edicts were issued and then partially countermanded that once again eliminated distant collaterals from participation in imperial ancestor worship ceremonies due to their distance from the reigning emperor（WS 9.226；19.446；183.2763；Kubozoe 485-86）in line with the precedent established by Xiaowendi in 493．This action precipitated a series of partially successful protest memorials to the throne by members of the lines affected（the latter two Wei shu entries listed above）．The author of the most prominent protest memorial was Yuan Cha's father，Ji，clearly acting as a lobbyist for the interests of the distant Tuoba royal lines，and just as clearly contravening the legacy of Xiaowendi．Empress Dowager Hu reversed herself and permitted distant collaterals to continue participating in imperial ancestral sacrifices，however，other exclusionary conditions may have remained in place．The Wei shu entry 183.2763 is dated the seventh lunar month of 517 and in the course of the author's objections to the ritual restrictions on participation in ancestral sacrifices that were apparently being discussed at this earlier time，raised the fear that the children of relatively close imperial collaterals would also be denied access to appropriate entry level court postings．This seems to be a response to a proposal that was articulated in an edict issued a month later in the eighth lunar month（WS 9.226）stating that children of all distant royal collaterals under the age of fourteen would no longer be eligible for entry level positions in the court．This is significant because such early access to the court in nominal or supernumerary posts was an important acculturating factor and enabled these young people to make social and political connections of great value to their future elite careers．（Yan Buke 2000，42-3）In the eighth lunar month of 517 the throne（Empress Dowager Hu's regency council）issued an edict（apparently the same one just discussed above）that also acknowledged the proliferation of royal collaterals，the fact that many were living in humble conditions，and noting the past contributions of the collateral royal lines in supporting the dynasty and protecting the throne．The edict stated that，in a demonstration of imperial beneficence，depending on the proximity or distance of the lines of descent from the throne，their members would all be employed by the court in appropriately ranked positions．（WS 9.226）It is not clear if the restriction relevant to distant collateral adolescents was revoked．This edict was designed to re-assure distant agnates of the court's concern for their welfare and perhaps to balance off the negative impact of the stricture affecting young distant collaterals．The edict's strict sense of status consciousness was very much in line with the status legislation of Xiaowendi，and despite its effort at fairness within these status confines，it also could reinforce the sense of political frustration in people like Yuan Tianmu．A key issue that confronted even Xiaowendi，was how to cope with entrenched Tuoba and Xianbei elites at all levels of the court．This issue never seems to have been fully resolved．The 519 riot was an extreme expression of what appears to have been growing frustration among Xianbei courtiers．In the context of these elite Tuoba and Xianbei concerns regarding access to status appropriate court postings，the Zhang Zhongyu proposal was politically obtuse and absolutely inflammatory.
The Guardsmen riot occurred in the second lunar month of 519．The Guard Commander（lingjun）was the elderly Yuan Ji，the father of Yuan Cha．In the fifth lunar month Yuan Ji was removed from this posting and given a high level honorific posting．Empress Dowager Hu then appointed Yuan Cha to the lingjun position，assuming that while the Guard Command needed a forceful and aggressive personality she could depend on the continuing loyalty of her sister's husband，whose very career was owed to her favor．（Zhang Jinlong 1995，58；WS 9.228-29）Assuming the position in the immediate aftermath of the Guard riot，Cha must have been under tremendous pressure from the Guardsmen，their families，and general court opinion to ensure that young guardsmen，starting out their careers，and previous serving officials occupying an interim supernumerary status would have access to subsequent court postings．（Yan 2011，238-39，244）
The court employment issue regarding Xianbei youths from families of various levels of elite status did not disappear but continued to be an albatross around the neck of policy makers．In fact，one of the major criticisms of Yuan Cha's period of governance was that he failed to maintain administrative discipline and appointed unqualified individuals to positions of provincial responsibility．（WS 16.405）This was the same criticism leveled against the 519 alteration in the court personnel policy following the guardsmen riot by individuals disgruntled with the consequent lowering of personnel qualification standards．（see Tanigawa 122-23）This illustrates the degree to which Yuan Cha became committed to and then politically identified with the full implementation of the more liberal，less rigidly status oriented post-519 personnel policy．In 524，a year before he was toppled and in the midst of the increasing Six Garrisons violence，Yuan Cha successfully pushed through a policy to hire Xianbei youths from lower ranking Luoyang houses as imperial edict messengers，which was not a very exalted position．The Wei shu specifically notes that at this time the opportunities for court employment were negligible．（WS 81.1792；Zhou Yiliang 1997，133）This employment policy was implemented in the face of an opposition lobbying campaign（literally，letters of opposition were sent to the court）from elite houses，specifically those whose members were serving as provincial level officials．This would imply somewhat distant Tuoba collaterals and lower ranking Xianbei and“other”serving courtier houses were concerned that their children would lack entry level court positions．Since the policy was not cancelled，these same houses memorialized that an elite imperial bodyguard and entourage unit be established for their young men．Yuan Cha satisfied this demand and incorporated these young men into the Palace Guard under his command（as lingjun）．The unit was subsequently known as the shuzi庶子 and was one of three such special units created at different points in time to accommodate young men from elite families into the court employment and status stream．（Hamaguchi 91-4，147-48）Yuan Cha's motivation in both cases was to ensure Luo-yang Xianbei loyalty to his regime in the context of the ongoing Six Garrisons crisis．The fracas over the court employment policy probably damaged Yuan Cha's political position at court since he was toppled a year later．The situation also strikingly reveals the necessity of the throne（or the controlling court clique）to be sensitive to the needs of elite Altaic court houses and their sense of lineage prerogatives．Regarding this issue，I wish to propose that elite Chinese court houses under a Chinese imperial order would not be as direly affected by court employment policy as were these Northern Wei Altaic elites．Chinese elites would have had other status satisfying routes in the cultural and social realm（the socially prominent recluse trend，extended self-flagellating exercises in filial piety at the grave of deceased parents or social equivalents）that could，at least temporarily，compensate for lack of employment opportunities at the court．The Altaic court elite houses seem to have had no options beyond entry level service at the court for their young men．He Dezhang has written on the adaptation of Chinese high cultural practices by elite Luoyang Xianbei houses（He 263-82）．However，the one thousand Guardsmen who brutalized the Zhang family members do not seem to have shared such high cultural horizons．This points，again，to the relatively rigid structure of the Northern Wei court and the social world built around this structure.
Returning to events leading up to the Yuan Cha coup，rapid changes in court personnel policy，while welcomed by many（but not all，see the collected essays in opposition to the relaxed standards for hiring and rotating officials in Tanigawa 122-23），probably made the Hu regency appear lost and directionless．This would be in addition to the earlier volte faces with regard to distant agnates participating in imperial ancestral services．Yuan Cha's coup occurred one year later，in the seventh lunar month of 520．Cha seems to have been riding a wave of court dissatisfaction with the way the Hu regency was handling what can be dubbed“court domestic policy.”Regardless of Cha's personal disputes with other members of the regency，the extraordinary nature of the coup，led by a lineal Tuoba inferior，could only have succeeded if elite court opinion was willing to swing in favor of Cha's dramatic action．Consequently，Xianbei and Tuoba court seniors were grudgingly willing to work with a Cha regime．On the other hand，Cha's violence would create rifts at the court which would be his undoing once he was perceived as no longer viable．In the course of the coup，Suzong's uncle，Yi，was executed，the Empress Dowager Hu was placed under house arrest，and the ten year old Suzong nominally ruled in his own name，but was a front for Cha and key eunuchs who worked in the inner palace in close physical proximity to Suzong．It was the execution of Yi that shocked and splintered the court．Yi appears to have been rigidly，almost ideologically committed to the full blown court status legislative legacy of Xiaowendi．He strongly opposed contemporary entry level court hiring trends as gross violations of Xiaowendi's specific status oriented legislation from 495．Yi showed no sign of flexibility in the face of the court's personnel dilemmas．（WS 22.595）Yi's tomb biography also depicts an individual who could almost be described as a Confucian stickler for decorum．（Zhao Chao 172）Yi was not only an obstacle to the personal ambitions of Yuan Cha，but a hindrance to the effort to flexibly resolve the growing court personnel problem.This may explain general court willingness to tolerate Yuan Cha's coup.
Nevertheless，there was a pungent reaction from segments of the Tuoba elite．Yuan Xi，元熙 the provincial governor of Xiangzhuo相州，based in the strategic city of Ye，and a close friend of Yi，rebelled．（WS 19.503）．Xi came from the line of the Jingmu emperor，a descendent of one of Xianzu's uncles．This segment of the royal clan had maintained high status positions in part because of their extraordinary numbers and in part because key members，like Cheng under Xiaowendi，maintained high policy advising positions with the throne．Members of this segment tended to work closely with close royal agnates as Suzong's uncles．The rebellion was quickly crushed，Xi and two of his three brothers were executed．The third brother escaped to the Liang regime in the south．Consequently，Cha，a distant agnate who had advanced to high positions due originally to his marriage to Hu's younger sister had the blood of some of the most elite Tuoba lineage members on his hands．Yong，the sole surviving son of Xianzu，also quietly opposed Cha and would later help engineer his downfall and restore the Empress Dowager Hu to power in 525（WS 13.339）．
An aborted putsch by a high palace guard commander（Xi Kangsheng，as youwei jiangjun 右衞將軍，or Right Commander of the Palace Guard），directly subordinate to Yuan Cha（as lingjun）and a former close collaborator，in 521 also illustrates the fissures created by the Cha coup．（WS 73.1630）The positions of Right（and Left）Commanders of the Palace Guard，each position having two slots（apparently prior to 520 each position only had one slot each，see WS 113.2979，2995），was created in the seventh lunar month of 520，the exact same time as the coup．（WS 113.3004；Hamaguchi 148）In 520 only Xi Kangsheng at the Right position，and Hou Gang 侯剛，a long time but no longer fully compliant servitor of the royal family，at the Left position（WS 93.2005）are mentioned by the standard sources．That is，these high placed individuals had to be properly rewarded for cooperating with Cha and the quid pro quo was also sealed by marriage ties between the three men（WS 73.1632）．The creation of additional slots for each position may have been an effort to dilute the influence of Xi Kangsheng and Hou Gang，but it is not clear if these additional slots were actually filled．Nevertheless，relations between Yuan Cha and Xi Kangsheng were tense．Xi was essentially a professional warrior and officer-a rough man with a strong martial ethos who had risen through the ranks by means of demonstrated military valor and field command experience．He was a professional killer of men．Yuan Cha was much more the professional courtier and political operative．This divide may also be reflected in the manner by which Yuan Cha seized power in the court by arresting and executing Yuan Yi，Suzong's eldest surviving uncle．Yuan Yi was seized by men from units of the zongshi 宗士and zhizhai 直齋（WS 16.404；Hamaguchi 92，94）．Both of these units were supernumerary military units under the command of the ling jun．The zongshi was a specialty entourage unit comprised of young men from elite houses．When members from this unit would transition to formal court employment they would generally be employed in court offices ranking relatively low，between ranks seven and eight（WS 113.3004；Hamaguchi 92）Two other such units were created later．The shuzi mentioned above was created by Yuan Cha in 524 specifically to employ young elite men，and the wangshi 望士，composed of men from elite houses with martial skills，was created in 526 during the second Empress Dowager Hu regency period．The zhizhai was one of several similarly named units that were part of the regular Palace Guard Command but ranked below the yulin and huben units in terms of military qualifications．（WS 21.554；Hamaguchi 85）Additionally，the zhizhai units were not even ranked and salaried supernumerary units but granted rank equivalents，meaning that they also did not receive a formal salary（bishi guan 比視官，WS 111.2885-86；Tao 245；Yan 2011，244）．Yan Buke views them as participants in the court world of“inner ranked”courtiers．Presumably，members of this unit would，for the most part，also be very young men，as opposed to the higher ranking regular，perhaps more professionalized，yulin and huben units of the Palace Guard．The implication seems to be that Yuan Cha could control and trust young inductees into the court world to carry out the most sensitive part of his coup，seizing Yuan Yi．This is speculative，but perhaps a clientage relationship existed in the aftermath of the 519 riot where these young men were far more dependent upon the good will of Yuan Cha in terms of furthering their careers than members of other Palace Guard units．The fact that the regular Guard units were not utilized for this purpose may indicate a certain lack of trust．Presumably，such regular units were under the immediate command of Xi Kangsheng and Hou Gang who may not have wanted their hands dirtied with the seizure and execution of a Tuoba noble．Yuan Cha would have to do the dirty work himself．However，exactly why Xi Kangsheng decided to attempt a countercoup in 521 is unclear．Did he perhaps feel that other elements of the Palace Guard would spontaneously rally to his side drawn by the charisma of a martial hero？Nevertheless，the resulting violence incurred by Cha's coup alienated an entire segment of the Northern Wei court．
Political developments after 520，aside from the temporarily successful arrangement of Rouran and Gaoche relations on the Northern frontier in favor of Northern Wei interests，generally tended to steadily undermine Cha's questionable legitimacy and efficacy，in part due to events that long pre-dated his coup．The Northern Wei position on the south bank of the Huai River had been slowly deteriorating since the early years of Shizong．The strategies of Xiaowendi and his successor，Shizong，were relatively conservative．The goal was to simply dominate the border zone between the Huai and Yangzi Rivers．No attempt was made to engage in an all out assault against Nanjing．Indeed，as Zhang Jinlong has explicitly noted，the factional infighting that occurred in 500-501 during the early years of Shizong's reign precluded any such dramatic military venture．（Zhang，vol.8，224）．This was despite the fact that in 500 a major Southern Qi urban fort on the south bank of the Huai，Yuzhou（re-named by the Northern Wei as Yangzhou，also known as Shouchun），defected to the Northern Wei，a result of the imminent collapse of the Southern Qi.
In 507 the Northern Wei suffered a major setback in a failed effort to seize a key Liang fort on the south bank of the Huai River（Zhongli）located east of Northern Wei Yangzhou，and in the process lost advance positions in the center of the area between the Huai and Yangzi Rivers．From 508 through 515 the area was relatively quiescent and stable．Then，in 515 to 516 the Liang staged an unprecedented direct attack on Northern Wei Yangzhou．The attack was beaten back with expeditionary troops reinforcing the pre-existing defensive garrison（a typical Northern Wei approach to extended frontline positions south of the Huai dating back to Xiaowendi's initial troop placements and support strategy）．This was a severe compromise of the Northern Wei position in the frontier combat zone．Then，beginning in the middle of 523 the first of the Six Garrisons revolts broke out followed by ongoing，increasingly well organized violence in the north and northwest through 528．Cha's faction was in charge of affairs when the Six Garrisons violence began．The uprisings were the result of a mixture of food riots following consecutive years of drought afflicting Northwest China and the North China Plain，and a major Rouran raid in 523 in the Six Garrisons zone（the raid itself was triggered by the desperate ecological situation）that devastated an already suffering local population．The result was local banditry，bandit armies，local militias fighting bandit gangs，and then the subsequent arrival of imperial armies fighting whomever opposed them（and initially losing in the process，too）．The Six Garrisons zone became the scene of a mini-civil war in the context of larger government and bandit armies opposing each other.
In this confusing and distracting context the Liang resumed their campaigns in the middle and eastern parts of the Huai River in 524．In the east Liang forces pushed north of the Huai River onto the North China Plain，heading for the prefectural seat of Xuzhou，the city of Pengcheng．By the end of 524 the entire Northern Wei position south of the Huai River was under siege．Then，in early 525，Yuan Cha's cousin，Yuan Faseng，defected to the Liang，turning over the city of Pengcheng to occupying Liang forces．In a surprising series of events the Northern Wei would regain the city，but the cumulative damage from all of these developments to the Cha faction at court was significant．Yuan Faseng's defection to the Liang was not due to military considerations，rather，as a close partisan of his cousin，Cha，he feared that by 525 Cha's“overweening arrogance would bring disaster upon himself.”（ZZTJ 150.4690）．Specifics are not provided by the standard sources but clearly Cha's factional stability had been eroded and his prior alienation of segments of the Tuoba elite was starting to have an impact．
In the fourth lunar month of 525 Empress Dowager Hu staged a counter coup，and brought back to power the lineage segments that had been previously alienated by Cha．Her close Tuoba supporters and policy makers were Yong，the son of Xianzu，and Hui徽，from the line of Xianzu's uncles．In a prisoner swap that same year，the surviving brother of Yuan Xi was returned to Luoyang and immediately began working closely with Hu's faction（WS 19.506）．An entire anti-Cha front focused on key elite Tuoba lineages had returned to power．In addition to these agnates，Hu also had her personally dependent servitors，as well．Shortly after，Cha and some of his close supporters were executed by a somewhat reluctant Hu，whose hand was forced by elite court pressure（ZZTJ 150.4695；WS 27.672）.
The counter coup did not occur at a propitious time．The original Six Garrisons revolt by the loop of the Yellow River was quelled by 524 and the affected population was transferred en masse to the northern portion of the North China Plain，where conditions did not improve for them．In 525 refugee Six Garrisons bandit gangs and armies surged back with greater violence and increasing organizational capabilities．Simultaneously，the Liang regime increased its pressure on Northern Wei fortifications south of the Huai River．In late 526，a major setback occurred when the Northern Wei's premier fortification south of the Huai，Yangzhou，fell to the Liang．In the past when such threats appeared to these Huai River positions the court would send large numbers of expeditionary forces to bolster local defenders．In 526 there is no mention of such support being delivered．Distracted by the growing violence on the North China Plain，which had earlier also spread to Guanzhong（the Wei River valley），the throne could not or would not provide extra troops to the Huai River．In 527 the Hu faction continued to be buffeted by high level defections．Yuan Jian鍳，based in Ye and a member of the Jingmu lineage segment dominating the court defected to an approaching Six Garrisons army under Ge Rong．His rationale is not clear，it may have simply been based on a sense of inevitable military defeat．Jian miscalculated，and a month later Northern Wei forces re-took Ye and Jian was executed．One month after Jian had been executed，in the ninth lunar month the governor general of Guanzhong，based in Chang'an，Xiao Baoyin 蕭寶夤，declared a separatist regime．Xiao，a scion of the Southern Qi royal family had loyally served the Northern Wei as a frontline general and ranking courtier since his defection from the south in 501 at the age of fifteen following Xiao Yan's toppling of the Southern Qi and establishment of the Liang．Xiao Baoyin's rationale was fear of factional attacks against him at the court while he was trying，with only partial success，to cope with the violence in Guanzhong．Xiao Baoyin's fears were real．Yang Chun 楊椿，a high ranking Chinese courtier and former governor based in Chang'an and recently retired，wrote to the court regarding what he viewed as Xiao Baoyin's unilateral command activities（WS 58.1288）．He argued that Xiao Baoyin needed to be placed under surveillance，an argument the court accepted and was in the process of implementing．Similar to previous incipient Northern Wei warlord attempts，（Yuan Faseng in 525，Yuan Jian in 527）Xiao Baoyin's Chang'an base was re-taken by Northern Wei troops in early 528．Xiao Baoyin then fled with his family to rebel forces holding Gaoping in the northwest．These rebel forces were later defeated and Xiao Baoyin captured and executed by Erzhu forces in 530．The weak link in the Northern Wei system was not its military，but its increasingly factionalized and politically splintered court.
In the context of these strategic military blows and the defections of elite Tuoba and the Xiao Baoyin affair which undermined the viability of the dominant court faction，another factional crevice appeared．By 528 Suzong was eighteen years old and fully mature by the standards of the time，yet his mother and her faction refused to yield power to him（ZZTJ 152.4736-37）．In early 528 the Zizhi Tongjian alleges that Suzong sent a secret message to Erzhu Rong（which he shortly after countermanded）to bring his forces to Luoyang as leverage that Suzong could use to assert his authority and remove his mother's clique from power．This was the fundamental legitimating base Erzhu Rong needed to eventually occupy Luoyang and assert his satrap authority over the court（ZZTJ 152.4739；Tanigawa 136）．Immediately after Erzhu Rong began the process of moving troops from Shanxi towards Luoyang，Empress Dowager Hu's faction poisoned Suzong and set up a series of child emperors in his place．The dysfunctional nature of the court factional situation was obvious．Erzhu Rong's emissaries at the court then approached Yuan Ziyou 子攸，Suzong's tangshu，and son of Xiaowendi's esteemed（and deceased）younger brother，Xie，who agreed to ascend the throne backed by warlord troops．He would become the first（and very reluctant）Erzhu puppet emperor，known as Xiaozhuangdi．Interestingly，Yuan Ziyou the deceased Suzong's tangshu，came from an agnatic group that had been quiescent throne loyalists throughout the post Xiaowendi period．In the third lunar month of 528，the Shanxi based satrap，Erzhu Rong and his puppet Northern Wei Emperor，walked across the pontoon bridge crossing the Yellow River，welcomed by a faction riven court seeking political stability．What the court received in that same month was the bloody discipline of the Heyin bridge massacre，the exacerbation of pre-existing factional tensions and the creation of more blood debts among the splintered Xianbei/Tuoba elites．The independent Northern Wei monarchy had come to a de facto termination.
In a patrimonial empire the court provides a demarcated world for the work，political competition，and socialization of members of politically elite households and their associates．Unlike a modern operating government the court world is not necessarily interested in establishing dense networks of interaction with the non-court commoner populace，though，indirect networks certainly exist．Domestic social policy as understood in the modern sense is alien to the functions of the pre-modern court．The right of the populace to a living livelihood（usually agrarian based）is broadly recognized by the court for both ideological and practical power reasons．As a result，courts often maintain emergency granaries and engage in basic efforts at famine relief.
Insofar as an empire is a structure of domination the court is the center of the life world of those elites fortunate enough to be in the legally separate status of imperial courtier（whatever their ranking）．The court gave their lives structure and legitimacy．This understanding is even more accentuated in the case of ethnically distinct conquest dynasties．For their elites，the court is their Jerusalem in a foreign land．In such regimes，the perceived need to stand against an occupied population resulted in certain built-in rigidities in the court structure which made the dangers of court factional fragmentation an issue of salient concern for these regimes.
The Chinese throne，due to its long，continuous development over the course of the centuries from the 700's BC onward was no longer constrained by the mandated presence of specific households or lineages．Needless to say，elite court members had no legitimate voice in interfering in standard vertical imperial succession affairs．Chinese emperors had to garner support in the court for political and/or policy purposes and they often attempted to manipulate the various leading court personnel and existing factional groupings．However，the Chinese throne was not a priori committed to supporting any particular court grouping，it was free to pick and choose as circumstances demanded．This can be referred to as a lineal dynasty，secure in its own prerogatives.
The Northern Wei throne，as an ethnic minority conquest regime，was never able to attain a similar degree of freedom typical of the Chinese throne．As long as the Northern Wei adhered to a form of“ethnic sovereignty”（to use Mark C．Elliott's term）then the action of the throne would always be significantly constrained．This is most obvious regarding court personnel utilization wherein de jure or de facto ethnic segregation and ethnic court post slotting limited the ability of the throne to shape and structure court personnel．The 519 riot of Altaic guardsmen in Luoyang is a dramatic example of the barriers facing the throne in this type of court．As revealed by the 496 aborted coup by Xianbei conservatives based in Pingcheng against Xiaowendi's structural reforms，the throne had to tread very carefully with regard to how it dealt with conservative ethnic elites．There seems to be a greater need for elite consensus in such regimes when compared to a Chinese regime.
In discussing ethnicity，I want to emphasize that as relative foreigners in a conquered land，the throne and its court power elites perceived their own political survival and military dominance as being premised on maintaining some degree of separation of the conquering host from the conquered population．The various linguistic and material culture markers of ethnicity were lauded and valued and political institutions were rapidly designed to maintain this“ethnic suzerainty.”The point I wish to make is that power politics drove the effort to maintain and define ethnicity，and not the other way around．Northern Wei rulers who attacked and degraded affinal tribal groupings，not to mention their willingness to kill close agnates，even sons，as perceived threats to the throne，would have little emotional sympathy for ethnicity unless it served important political purposes．As a result，the Northern Wei court had built-in rigidities which constrained the conduct of both the throne and the court．By 471，viewing the retirement of Xianzu as a watershed，one can state that the throne definitively asserted its prerogatives to determining the vertical succession independent of collateral Tuoba interests or the interests of other elite court players．But in other areas the structural rigidities remained in place．Thus the Northern Wei can be referred to as a“corporate regime.”One aspect of this“corporate”characteristic was the need to maintain a degree of normal factional comity and respect for the status ranked entrenched interests of various elite ethnic lineages．The structural reforms of Xiaowendi began the process of undermining this comity and the modes of factional freedom permitted in the court of Shizong（500-516）opened up unprecedented political possibilities．The result of these developments was the destructive factionalism and political volatility of the 520's（reign of Suzong，516-528）exacerbated by multi-front military crisises.
Bibliography
Bei Qi shu．Beijing：Zhonghua shuju，1983.
Bei shi．Beijing：Zhonghua shuju，1983.
Cefu Yuangui．Beijing：Zhonghua shuju，1982.
Ershiwushi Bubian，vol.4．Beijing：Zhonghua shuju，1986.
Tong Dian．Beijing：Zhonghua shuju，1984.
Wei shu．Beijing：Zhonghua shuju，1984.
Wenxian Tongkao．Beijing：Zhonghua shuju，1986.
Zizhi Tongjian．Beijing：Zhonghua shuju，1976.
Secondary Sources
Armstrong，John A．Nations before Nationalism．Chapel Hill：University of North Carolina Press，1982.
Dai，Weihong 戴衛紅．A Study of the Northern Wei Personnel Evaluation and Promotion System 北魏考課制度研究．Beijing：Zhongguo Shehui Kexue chubanshe，2010.
Deng，Yiqi鄧奕琦．Study of the Legal Institutions of the Northern Dynasties北朝法制研究．Beijing：Zhonghua shuju，2005.
Eisenberg，Andrew．Kingship in Early Medieval China．Leiden：Brill Press，2008.
Elliott，Mark C．The Manchu Way：The Eight Banners and Ethnic Identity in Late Imperial China．Stanford：Stanford University Press，2001.
Gernet，Jacques．Buddhism in Chinese Society：an Economic History from the Fifth to the Tenth Centuries，trans．by Franciscus Verellen．New York：Colombia University Press，1995.
Hamaguchi，Shigekuni 濱口重國．Studies in the History of the Qin，Han，Sui，and Tang Dynasties 秦漢隋唐史 の研究．Tokyo：Tokyo Daigaku shuppankai，1971 reprint.
Hansen，Valerie．Negotiating Daily Life in Traditional China：How Ordinary People Used Contracts，600-1400．New Haven：Yale University Press，1995.
He，Dezhang 何德章．Collected Essays on the History of the Wei，Jin，and Northern and Southern Dynasties魏晉南北朝史叢稿．Beijing：Shangwu Yinshuguan，2010.
Hu，Hong 胡鴻．“Small Individuals，the Big Historical Picture – Three Stories From the Tomb Epitaphs of Yuan Zan and His Wife 小人物，大歷史——北魏元瓚夫婦墓誌中的三個故事，”Wenshi文史 2：83，2008，115-128.
Kubozoe，Yoshifumi窪添慶文．Studies in the Administrative Systems of the Wei Jin and Southern and Northern Dynasties 魏晉南北朝官僚制研究．Tokyo：Kyūko Shoin，2003.
Li，Ping 李憑．Collected Papers on Northern Dynasties Research 北朝研究存稿．Beijing：Shangwu yinshuguan，2006.
Li，Wencai 李文才．“Gao Zhao's Sichuan Campaign and Gao Zhao's So-called Arrogation of Court Power”，高肇伐蜀與“高肇專權”，Beichao Yanjiu 北朝研究 1：1999，82-94.
Matsushita，Kenichi 松下憲一．Formation of the Northern Wei Nomadic State 北魏胡族體制論．Sapporo：Hokkaido Daigaku shuppankai，2007.
Piggott，Joan R．The Emergence of Japanese Kingship．Stanford：Stanford University Press，1997.
Sagawa，Eiji 佐川英治．“Research into the History of the Six Garrisons of the Northern Wei 北魏六鎮史の研究，”in Sagawa Eiji ed.，Research on the Remains of Northern Wei Fortifications in the Daqing Mountain Zone 大青山一带の北魏城址の研究．Tokyo：Tokyo University Department of Humanities and Social Sciences，6.2013.
Sinor，Denis．“The Establishment and Dissolution of the Türk Empire，”in Denis Sinor，ed.，The Cambridge History of Early Inner Asia．Cambridge：Cambridge University Press，1990.
Sneath，David，ed．Imperial Statecraft：Political Forms and Techniques of Governance in Inner Asia，Sixth-Twentieth Centuries．Bellingham：Center for East Asian Studies，Western Washington University，2006.
Tang，Zhangru 唐長孺．“On the Fixing of Elite Surname Groups and Lineages by the Northern Wei Emperor Xiaowendi 論北魏孝文帝定姓族，”in Supplemental Essays on the History of the Wei，Jin，Northern and Southern Dynasties 魏晉南北朝史論拾遺 incorporated in The Collected Essays of Tang Zhangru 唐長孺文集，vol.2．Beijing：Zhonghua shuju，2011.
Tanigawa，Michio 谷川道雄．Thesis Regarding the Historical Emergence of the Sui and Tang Empires 隋唐帝國形成史論．Trans．to Chinese by Li Jicang李濟滄．Shanghai：Shanghai Guji chubanshe，2004.
Tao，Xinhua 陶新華．A Study of the Personnel Administrative System of the Northern Dynasties Following the Reign of Xiaowendi 北魏孝文帝以後北朝官僚管理制度研究．Chengdu：Sichuan chuban jituan Ba Shu shushe，2004.
Wang，Su王素．A Draft History of Gaochang Politics高昌史稿——統治編．Beijing：Wenwu chubanshe，1998.
Yan，Buke閻步克．“A Preliminary Study of“zhiren”in the Northern Wei and Northern Qi 北魏北齊“職人”初探，”Wenshi 3.48.1999，19-37．
Yan，Buke．“The Transformation of Military Ranks into Supernumerary Military Titles in the Course of the Wei，Jin，and Northern and Southern Dynasties' Period 魏晉南北朝軍號散階化進程，”Wenshi 2.51.2000，33-52.
Yan，Buke．An Introduction to the Traditional Chinese Official Ranking System 中國古代官階制度引論．Beijing：Peking University Press，2011.
Yan，Gengwang 嚴耕望．“The Institution of the Department of State Affairs of the Northern Wei 北魏尚書制度，”in Selected Historical Essays of Yan Gengwang 嚴耕望史學論文選集．Taibei：Lianjing chuban shiye gongsi，1990.
Yan，Gengwang．“A Study of the Northern Wei Institution of the Department of State Affairs 北魏尚書制度考，”Lishi Yuyan Yanjiusuo Jikan 歷史語言研究所集刊 18.1948，251-360.
Zhang，Jinlong張金龍．“Xiaowendi's High Councillors and the Political Situation of the Northern Wei Court in the Early Years of the Reign of Xuanwudi 孝文帝的顧問大臣與宣武帝初年北魏政局，”Lanzhou Daxue Xuebao 蘭州大學學報 3.1989，1-7.
Zhang，Jinlong．“Gao Zhao's Arrogation of Power and Internal Schisms Within the Ruling Elite During the Reign of the Northern Wei Emperor Xuanwudi 高肇專權與北魏宣武帝時期统治集團内部矛盾，”Lanzhou Daxue Xuebao 3：1992，113-120．
Zhang，Jinlong．“The Palace Guard Command [ling jun] and Northern Wei Politics 領軍將軍與北魏政制，”Zhongguoshi Yanjiu 中國史研究 1：1995，54-62.
Zhang，Jinlong．A Study of the Palace Guard System of the Wei，Jin，and Northern and Southern Dynasties 魏晉南北朝禁衛武官制度研究．Beijing：Zhonghua shuju，2004.
Zhang，Jinlong．Political History of the Northern Wei北魏政治史，vols.5-9．Lanzhou：Gansu Jiaoyu chubanshe，2008-2011.
Zhao，Chao 趙超．Collected Tomb Biographies from the Han，Wei，and Nanbeichao Periods 漢魏南北朝墓誌匯編．Tianjin：Tianjin Guji chubanshe，1992.
Zhou，Yiliang．周一良．Jottings on the History of the Wei，Jin，and Northern and Southern Dynasties 魏晉南北朝史劄記．Beijing：Zhonghua shuju，1985.
Zhou，Yiliang．Collected Essays on the History of the Wei，Jin，and Northern and Southern Dynasties 魏晉南北朝史論集．Beijing：Beijing Daxue chubanshe，1997.
艾安迪（Andrew Eisenberg）
本文探討了北魏政權毁滅的原因。我認爲北魏政權的問題不是軍事方面的内爭威脇，而主要是朝廷内越來越激烈的派系鬥爭的結果。孝文帝的結構改革奠定了政權的新框架。世宗的人事政策和派系操縱的方法點燃了導火線。世宗之後朝廷的政治緊張局勢加劇了。皇帝的近親和疏遠皇親的矛盾，鮮卑/代人和孝文帝495年以來的定身份立法的矛盾，加上這批人和拓跋貴族的矛盾顯示出來。最後，500年以後北魏朝廷吸收北人的容納能力有了很明顯的限制——這個新的局勢對北人的身份感和前途有了很大負面影響。520年的元叉暴力政變把以上的矛盾都暴露出來，進一步推動了北魏朝廷的派系鬥爭，而最後導致北魏的政治解體。