此外，北魏早期传入代地的汉文化元素多经迂回路径——非直接源自中原（西晋崩溃后该地基本停建墓葬及大型组织化活动）——而是来自东西两侧边缘社会（其自身已消化吸收并发展这些风格技艺）。111在丝绸之路东段要冲河西走廊，混居族群将华夏风格与中亚影响结合，由此输出大夏（Bactrian）柱式及早期云冈"佛祖即皇帝"刚健造型。112另条区域发展脉络位于辽东及以东地区（新兴燕国与高句丽）。113（更东方的边缘政权——日本大和政权——正以独特方式吸收改造华夏元素渐成雏形。）因诸路（至少暂时）皆汇平城，这些元素被纳入当地迅速演变的墓葬设计（至少在承担得起的精英阶层）。如沙岭葬母的富裕将领案例所示，部分平城居民雇佣营造团队建造砖室墓（此制渐成东北亚精英身份新标志）。114而更保守的鲜卑群体仍延续内亚传统竖穴土坑葬俗。115
源自东北地区（慕容诸燕与高句丽）的借鉴包括对墓主形象更为生动的描绘——这在代地沙岭墓男女主人像中得到充分体现。尽管男主人可能另葬他处，但如平城墓葬常见形式（虽高度程序化），116夫妇像仍共绘于墓室后（东）壁：女左男右。他们统御着画面呈现的所有场景。此风格似为生前死后共通的范式。某丘穆陵氏代地男子（与崔浩同朝为官）的传记记载其"夫妻并坐共食"——不过该故事继而讲述二人傲慢地仅允夫家叔伯食余羹，遂遭时人讥讽。117
回到沙岭M7后壁场景：夫妇端坐经典汉式观景阁台基之上，脊饰鸱尾。二人身着宽袍，显然沉浸于其独特的中国风。118但其核心身份通过所绘鲜卑黑帽得以彰显。如同族群身份认同般重要（甚或更甚）的是等级制度——现代学者林圣智指出此墓（及同类墓葬）壁画中生活场景普遍渗透着等级观念。119侍者分立两侧，亦戴黑帽、着北方式紧身衣袴。但刻意处理为其体型仅半。120
除侍从外，男女主人目及之处显然为早期平城精英两大要务（其借以获取、维系与彰显地位的作为）：夫妇右侧（北壁）程序化展现威仪男性率军行进、猎取战利；左侧描绘盛大户外宴飨，由另一观景阁中人物主持。由于壁画剥蚀，此人身份难辨。但本文主张主持者即破多罗夫人。121此观点至少获得两类证据间接支持：其一为前文所述平城宴飨中女性掌权的可靠文献记载（可追溯至太武帝时期）；其二为墓葬画面构图——现代学者张帆称其"对称布局"，"其余壁画围绕后壁夫妇像展开"。122笔者完全认同此观点：若后壁与破多罗夫人同绘之男性同时现身侧壁（既主持宴飨又率领仪仗），则构图将丧失对称性。
Replica of the Shaling M7 mural exhibited at the Sackler Museum, Peking University. Photo, Fan Zhang, 2017.
反之，我们认为这些环绕石榻上破多罗夫人遗体的壁画（墓中唯一明确实体存在），以某种方式呈现了界定其生平及周遭人活动的统一隐含叙事——如同充斥美国中产家庭墙面的文凭、奖状与运动奖杯。123
代地的权势地位首要源自成功的军事统帅。北魏早期皇帝的核心角色是战场统帅，而几乎所有重臣（无论职能）皆在军府持有将衔（无论是否实际领兵）。124沙岭壁画北壁描绘该夫人族人率领建制部队：其乘舆居中，舆后幢旗或为本部标识。如高句丽安岳三号墓所示，此处非仪仗而是行军布阵。125尽管场景元素承袭汉制，但规模、组织度与行动统一性显着提升。北魏早期军队核心自是骑兵——沙岭壁画虽绘矛兵等步兵，骑军更为突出：轻骑兵戴特色鸡冠盔；重骑兵（在东亚仅存在约一个世纪）人马皆披甲。126舆驾前列阵列中鼓手号手（其融入行军场景的描绘似源自辽东新发展）。127击鼓鸣号之举，无疑在诸多社会中有助整肃军纪、提振士气。128
尽管木兰传奇为此议题增添趣味褶皱，但我们可推定沙岭墓女主人并未易裙为甲效力族军（亦未如木兰般卸甲归家）。129然作为部族核心成员，其声望源自族人功业。整个部族亦借墓室壁画彰显代地社会地位——既为沿墓道致哀的送葬者，亦为冥界居民宣示。现代学者Timothy Davis在研究北魏晚期（尤以洛阳地区）渐趋普及的墓志铭时指出：此类铭文既服务于墓主，亦裨益整个家族。130
沙岭墓女主人显然更直接参与北魏早期墓葬壁画描绘的另一大事业：聚集人群举办宴飨。在代地，正如所有社会，食物不仅是果腹之物——借用人类学家玛丽·道格拉斯（Mary Douglas）的核心观点131——更是凝聚家庭与社群的重要纽带。食物似乎能将"社会"消化为"共同体"；或更直白而言，"征服男人心需先征服其胃"，或许也能赢得其忠诚。132宴飨长久以来在界定（或重构）群体与个人在更大整体中的地位方面发挥关键作用，将等级制度与共同性天衣无缝地交织于美味飨宴。133
"狩猎"与宴飨在许多社会紧密关联。例如大英博物馆藏著名"乌尔军旗"（Standard of Ur，约公元前2600年）一面绘战争、另面绘宴饮。134在华夏世界观中，二者可视为构建等级制度的阳与阴之互补。正因如此，尽管多数社会中女性通常负责备膳供食，男性往往掌控重大场合，主导缔结盟约、宣示等级与化解冲突。135汉墓此类场景描绘中，女性作为二层观者出现；三国时期汉人政权墓葬则设女性独立餐区。136但在五世纪代地，女性已掌控宴飨：正如太武帝皇后执掌宫廷膳房、文明太后与嗣皇帝共宴时主桌领唱，本文主张沙岭墓壁画宴飨主持者即为该墓女主人。此时此地，至少部分女性深度公开参与此类场合形成的互动网络；在正式允准的领域内，她们从构建的关系中汲取权力。（如同文明太后，作为权臣之母的破多罗夫人亦由此获得个人权威。）北壁画面中，男性他者占据征伐大军核心，其荣耀反照于破多罗夫人；南壁则其独居主位——她端坐观景阁，统御宴席（座下或为其族人麾下将领，如文明太后灵泉宴般跪拜应和）。137尽管此系女性葬礼仪式的程序化呈现，但无疑映像更广阔的现实图景。138
Replica of the Shaling M7 mural exhibited at the Sackler Museum, Peking University. Photo, Fan Zhang, 2017.
正如大多数人类社会的情形，这些女性掌管着家庭内部事务（主家事）。139至少部分女性还走出家门担任家族使节——据载文明太后常与权门主母退居后室进行"闺阁密谈"，共商大小事宜。140破多罗夫人若欲扮演此角色（于观景阁帷幔后闲谈、祝酒或更有深意地磋商要务），自需通晓代地通用的克里奥尔语（creole）。至于宴飨中的等级确立，最直观体现于座次规则。141但即便在排座之前，赴宴义务已显端倪。142我们难以断言缺席破多罗夫人宴席者将遭何处置（或许她会授意其子褫夺缺席者晋升资格），但知晓两百年前拒赴力微宴请的部酋下场，以及卫王死后诸王对道武帝宴邀的惶恐。
沙岭宴饮场景中，主事者虽居于固定建筑内，但整体为露天环境，周边环饰卡通式山水树木——此风格再次令人联想到（并确与）长江流域绘画技法发展相关（如传顾恺之作品）。不过现代学者曾庆盈有力论证了此类主题在代文化（故代地墓葬艺术）中的独特性。143壁画中的宴饮者似乎沉醉于户外烧烤。观者几可闻席间盛赞羊肉羹美味，或愈加急切的索添马奶酒（kumis）之声——这些人虽多数已非牧民，仍嗜食羔肉、发酵马乳。食物常盛于大型漆器，如沙岭M7宴饮壁画中宾客前置的圆盘所示。144
平城宴飨包含多种活动：射箭竞技、猜谜游戏等。145如预期般，文明太后灵泉宴所见祝酒致意乃核心环节。此外（虽无墓葬壁画描绘）或存更随意的争执斗殴消遣。但沙岭壁画定格时刻，宴饮者正观赏歌舞——或为远方新至曲目。关于道武帝宫廷乐舞的记载显示，除歌颂王朝起源神话的乐曲外，亦演奏源自辽东、渭河谷地、河套匈奴及长江以南的曲调。146遗憾沙岭夫人早逝数年，未及听闻北凉征服后乐工携至平城的中亚乐潮——其或更风靡一时。
遗憾的是，我们永远无法亲闻激荡这些先民心魂的乐曲。但在沙岭M7墓后数年（太武帝攻取长安打通丝路直联后），墓葬中渐增外来器物。其中包括萨珊波斯全新服饰纹样——袍缘饰以圆形纹章与珍珠。147作为代地精英身份标识，珍珠在相对朴素的沙岭墓中仅以天花板装饰形式出现（采用中亚佛教艺术吸纳的摩尼教样式）。148其他身份标志包括银器与玻璃器皿（多自中亚或伊朗输入），代地显贵以此餐饮。149此类器物极受追捧，而私造行为被视为威胁可汗对物资分配的掌控，故太武帝444年颁诏严禁贵族豪强私蓄匠人（违者处死）。150破多罗夫人墓中虽未见此类器物（略早于西域奇珍涌入高峰），但壁画显示其备有多组餐盘待客。墓中出土陶罐陶壶或曾用于宴飨。151至于盘中盛物，虽谷物更易获取，肉类仍为主食（多为羔羊肉，偶有生切羊肉片（sashimi））。152乳制品丰富，包括酸奶、黄油与马奶酒。现代学者吕一飞论及华夏与内亚饮食传统差异时指出：前者重谷物而后者嗜肉；华夏讲究五味调和，北人则偏好食材分置以保本味（尤重肉香）。153
但异乡人总能适应，如南人王肃（463–501年）轶事所示。493年北投受孝文帝礼遇，初时难忍羊肉与羊乳，仍嗜鱼饮茶。数年后宫宴，帝察其已食羊饮酪，询其故。王答："羊乃陆产之最，鱼为水族之长，各有所美。"154即便无法如毛修之般调和南北风味，至少可兼品其妙。
将视线从破多罗夫人南壁壁画左侧（宴饮场景）转向右侧织物屏风另侧，可见同等乃至更引人注目的内容。此处临时厨房帐篷内，劳碌的仆役们为宴饮者服务。155观景阁紧邻屏风东侧，而仆役区停放着阁内贵人的华贵车驾（朱盖红帷，马匹在侧啃草）。厨房近在咫尺的事实进一步佐证观景阁中人即破多罗夫人：尽管背对仆役，但她掌控庖厨事务，如同赫连皇后般洞悉内情；难以想象对面率军之男性角色会如此行事。若接受此前提，破多罗夫人自仆役区下车后，或先向管家做最后指示，继而穿屏风间隙迎候宾客。既已端坐观景阁主位，除非重大疏失，她应无需折返。若遇此情形，她需如文明太后般抉择是否宽宥。
仆役工作繁杂多样。画面顶部绘三座移动粮仓，右侧排列数辆运载食材器皿的停车——府邸女主人掌控所有这些家产。156其下四顶小帐篷分列，其中两顶内有侍女劳作。场景最生动处位于帐篷下方：两名男子立于散置瓶罐间，刚宰杀绵羊悬空放血入釜。157
现代学者林圣智提出有趣观点：平城墓葬壁画中宴猎并置的特征根植于代地缺乏稳定经济基础的不确定性。158至少初期，这是个饥饿的帝国，食物生产充满变数。对无宫廷深窖储粮者而言，牲畜或猎获乃最可靠热量来源。破多罗夫人庖厨所宰之羊（宴客所享羊肉）获取途径多样：或购自平城市集（牧民自草原驱来），或来自皇家鹿苑（兼作巨型牧场）。159但更可能其家族自有畜群。北魏崩溃之初（六世纪初），汉人官员杨椿训诫子孙时追述先祖投效道武帝获赐"田宅、奴婢、马牛羊，遂成富室"。160在此背景下，混合经济形态应运而生：部分奴婢于大同盆地耕作，另些（或为安置牧民）于山丘放牧，而杨氏安居宅邸，不时外出野宴啖食羊肉羹。161
当然，至少多数奴婢与依附民的生活境遇大相径庭。2000年大同东郊雁北师范学院工地发现的某平城墓葬（编号M2）出土一组引人注目的随葬品。除砖室墓常见的陶制马、牛、骆驼俑外，另有三件游牧帐篷陶模型：一件高逾7英寸的典型圆形毡帐（yurt或ger，类似今蒙古所用），旁置稍大的两件方形帐篷模型（顶部设排烟活板）。162汉人宴饮描绘于宅邸内，而代地宴飨呈现为盛大户外烧烤，此差异不足为奇。代地精英虽拥有庄园宅邸（宴后归返，巨釜由牛车运自彼处），但仍有牧羊奴仆（随时驱羊至宴宰杀）。163据现代学者古贺明峰推测，这些牧人仍以亲缘组织自治单元生活，毕生多居毡帐——春季驱畜至丰茂高地草场，冬季迁入山谷（穿越无人戍守的古长城遗迹）。另一显赫家族匹娄氏"奴婢千数，牛马以谷量"。164
此类牧人生计艰难，身处边缘。部分成为最终倾覆洛阳的起义首领。165但在所述时期，他们仅为照料主家牲畜，游牧于旷野。其不用壁画所见巨釜，而使用更紧凑的内亚炊具（北朝墓葬常见，但沙岭M7未出）——汉语称"鍑"。166此器铜铁所铸，器型较小、阔口深腹，双耳可插杆悬于火上。部分专为骑乘设计：单侧扁平，便于挂靠马身，适应游牧、狩猎或征战途中野炊。
1. Borrowing in part at least from the Chinese phrase, “chasing the deer” (zhu lu 逐鹿), which beginning with the power struggles that followed the Qin collapse, ca. 210 bce, has been used in connection with predatory warfare undertaken in order to take control of taxable populations: SJ 92.2629. Among other places, it has appeared in a study of the later Northern Dynasties: Jiang Lang 姜狼, Zhu lu tian xia: Bei Qi he Bei Zhou si shi nian zheng ba shi, 526–581 逐鹿天下: 北齐和北周四十年争霸史, 526–581 (Taibei: Da di chu ban she, 2012).
2. A phrase taken from the caption on an exhibit of 17th- and 18th-century Hanoverian dining silver, displayed in Gallery 250 of Boston’s Museum of Fine Arts. Similar remarks are provided by Lothar Ledderose regarding the Han elite, which “defined and displayed social status through luxury tableware, just as the aristocracy in medieval Europe would” (Ten Thousand Things, 177–78).
3. See Chapter 9 note 22.
4. Treager, Geography of China, 213, points out persistent danger of famine in the loess lands.
5. Luoyang’s rapid rise and rapid fall (493–534) were recorded by Yang Xuanzhi 杨衒之 (d. ca. 555) in his Luoyang qie lan ji 洛阳伽蓝记 (published as Luoyang qie lan ji jiao zhu 校注 [Shanghai: Shanghai gu ji chu ban she, 1978]); translated by Jenner in his Memories of Loyang; and by Yi-t’ung Wang as A Record of the Buddhist Monasteries in Lo-yang (Princeton, NJ: Princeton University Press, 1984). As for Pingcheng, there was a poem, “Mourning Pingcheng” 悲平城, written shortly after the capital had been moved to Luoyang (see Lei Bingfeng 雷炳锋, “Wang Su ‘Bei Pingcheng’ shi chuang zuo shi jian kao bian” 王肃《悲平城》诗创作时间考辨, Suzhou xue yuan xue bao 30.8 [2015]: 64–66). The four-line poem, by Wang Su, whom we shall discuss in more detail in Chapter 16, was, however, more complaint about the region’s weather than grief at the city’s abandonment (WS 82.1799): “the Shadow Mountains (Yin shan) are always dark and snowy; desolate pine, wind without cease.” It must also be made clear that Pingcheng was not entirely abandoned; it continued to play a role in East Asian history, as “western capital,” for instance, of the Khitan state. In archaeological work done at Datong, Liao tiles are found atop those of the Wei (Joy Yi, Yungang, 23ff.).
6. WS 2.31, 23.604. Cao Chenming 曹臣明, “Bei Wei Pingcheng bu ju chu tan” 北魏平城布局初探, in Bei Wei Pingcheng kao gu yan jiu: gong yuan wu shi ji Zhongguo du cheng de yan bian, ed. Wang Yintian (Beijing: Ke xue chu ban she, 2017), 1.
7. See the statement made more than fifty years later by Gao Yun 高允 to the emperor Wencheng (r. 452–465), in opposition to a building project, that it would take 40,000 men six months to complete the project, bringing suffering upon the common people: WS 48.1073. In the same passage, Gao Yun expressed disapproval of the fact that Daowu had scheduled construction of Pingcheng without taking account of farmers’ slack seasons.
8. WS 23.604.
9. WS 2.33–34.
10. WS 105C.2392; Cao, “Bei Wei Pingcheng bu ju chu tan,” 3; Yin, “Bei Wei Pingcheng shi lüe,” 193.
11. Yin, “Bei Wei Pingcheng shi lüe,” 195.
12. WS 105C.2392.
13. See the figures given in Yin, “Bei Wei Pingcheng shi lüe,” 196. A suggested reconstruction of the layout of the walls can be seen in Duan and Zhao, Tian xia da tong, 19; in this theory, the outer walls went around and contained the palace city. For wall height, see p. 28.
14. Duan and Zhao, Tian xia da tong, 17–18.
15. NQS 57.985. Duan and Zhao, Tian xia da tong, 20–21, point out that this is the first textual mention of a system of walled wards, which went on to influence Wei’s Luoyang, and from this the larger East Asian world. Dien, Six Dynasties Civilization, 31–32, cites Miyazaki Ichisada 宫崎市定 (“Rikuchō jidai Kahoku no toshi” 六朝时代华北の都市, Tōyōshi kenkyū 20.2 [1961]: 53–74) in discussion of use of walls in these cities as a control for restive, transplanted populations. Though the main NQS quote raises the issue with the guo cheng (“barbicans”), Yin Xian states this began with the city walls, the wai cheng: “Bei Wei Pingcheng shi lüe,” 196.
16. Duan and Zhao, Tian xia da tong, 21.
17. Originally built in 416, the drum was added later: Duan and Zhao, Tian xia da tong, 67, citing Shui jing zhu shu 2: 13.1144–45.
18. NQS 57.985. Note suggestion for correction of this passage by the editorial authors of the Zhonghua shu ju edition.
19. WS 3.62; Yin, “Bei Wei Pingcheng shi lüe,” 196. For estimates of the barbican’s height see Duan and Zhao, Tian xia da tong, 27.
20. Song, Bei Wei nü zhu lun, 21; Sakuma, Gi Shin Nanbokuchō suirishi kenkyū, 363–64.
21. Segawa, “You mu yu nong geng zhi jian,” 104–5; Yin, “Bei Wei Pingcheng shi lüe,” 198.
22. For discussion of the broader topic, see among many possibilities Louis Nelson’s Architecture and Empire in Jamaica (New Haven, CT, and London: Yale University Press, 2016).
23. WS 102.2275; Bonnie Cheng, “Exchange across Media in Northern Wei China,” in Face to Face: The Transcendence of the Arts in China and Beyond (Lisbon: Centro de Investiga··o e Estudos em Belas-Artes [CIEBA], 2014), 138.
24. WS 2.36; NQS 57.984. Fragments of this have been discovered: see Yin Xian 殷宪, “Datong Bei Wei gong cheng diao cha zha ji” 大同北魏宫城调查札记, Bei chao yan jiu 4 (2004): 153. And for discussion of belief that mica could prolong life, and prevent decomposition of the dead’s flesh, see the article by Edward H. Schafer and E. H. Snafer, “Notes on Mica in Medieval China,” TP 43.3/4 (1955): 265–86.
25. NQS 57.984; WS 2.42; Zhang Zhizhong 张志忠, “Bei Wei Pingcheng shuang que kao” 北魏平城双阙考, rpt. in Bei Wei Pingcheng kao gu yan jiu, 24–27.
26. https://www.hrp.org.uk/tower-of-london/explore/white-tower/#gs.WaiQ7xY; accessed 19 May 2018.
27. Liao shi 41.506.
28. Zhang, “Bei Wei Pingcheng shuang que kao,” 26.
29. Fu Xinian, Traditional Chinese Architecture: Twelve Essays (Princeton, NJ, and Oxford: Princeton University Press, 2017), Chapter 4.
30. See the comments of Dien, Six Dynasties Civilization, 15, on why “[r]elatively little work has been done on urban archaeology in China.”
31. For the loss of remains under new constructions, see the comments of Zhang, “Bei Wei Pingcheng shuang que kao.” On the walls in particular, see Duan and Zhao, Tian xia da tong, 28–29.
32. See Ren Yuan’s article “Back to the Future: The Fake Relics of the ‘Old’ Chinese City of Datong,” in the Guardian, 15 October 2014. Insight into the changing nature of the town in the 21st century is given in Chris Buckley’s New York Times article, “In China’s Coal Capital, Xi Jinping’s Dream Remains Elusive” (21 October 2017).
33. On difficulties involved in reconstructing city plans, see inter alia the general comments made by Dien at the beginning of his chapter on “Cities and Outposts,” in his Six Dynasties Civilization, 15; and for Pingcheng the introductory comments in Zhang’s “Bei Wei Pingcheng shuang que kao,” 24; and Duan and Zhao, Tian xia da tong, 46. The latter is the most ambitious effort this author has seen to attempt, with much speculation, to describe Pingcheng, and follow it from Wei into later ages.
34. I have relied on the more detailed attempted reconstructions of Duan and Zhao, Tian xia da tong, 52–53. These are, however, tentative: somewhat different attempts are given in Yin, “Bei Wei Pingcheng shi lüe,” 193; and Cao, “Bei Wei Pingcheng bu ju chu tan,” 1, 5.
35. WS 2.42, 3.52, 3.58. Ren Aijun 任爱君, “Bei Wei Xianbei ren woluduo yi zhi ling shi” 北魏鲜卑人斡鲁朵遗制零拾, Bei chao yan jiu 3 (1996) (cited in Sagawa, “You mu yu nong geng zhi jian,” 106 note 2), suggests that Daowu’s “palaces” at Pingcheng may in fact have been tents (see also Sagawa, 111). The issue of the “peripatetic rulership” of the Taghbach is discussed at length in Chapter 2 of Chin-yin Tseng’s Making of the Tuoba Northern Wei (Oxford: BAR Publishing, 2013).
36. SoS 95.2322; He Dezhang 何德章, “‘Yinshan que shuang’ zhi su jie” “阴山却霜”之俗解, Wei Jin Nan bei chao Sui Tang shi zi liao 12 (1993): 102–16.
37. This can, of course, be connected with the forced sedentarization imposed upon the Helan and other groups, assignment to fixed location in the fixed space of the imperial domain; see this volume’s Chapter 7.
38. For use of fortified cities as storage centers, see Chapter 5 note 36.
39. WS 2.44.
40. Duan and Zhao, Tian xia da tong, 54.
41. WS 112A.2910; Duan and Zhao, Tian xia da tong, 54, 56. The style did not, of course, originate on the archipelago but was borrowed from architectural developments on the continent.
42. See detailed discussion of timber frame architecture in Dien, Six Dynasties Civilization, Chapter 3.
43. Duan and Zhao, Tian xia da tong, 48–49.
44. Yin Xian 殷宪, “Bei Wei Pingcheng zhuan wa wen zi jian shu” 北魏平城砖瓦文字简述, rpt. in his Pingcheng shi gao, 147.
45. Duan and Zhao, Tian xia da tong, 35, 37; and more general discussion of such influences in the columns depicted at Yungang by Kate·ina Svobodová, Iranian and Hellenistic Architectural Elements in Chinese Art, Sino-Platonic Papers No. 274 (Philadelphia: Department of East Asian Languages and Civilizations, University of Pennsylvania, 2018). The most famous example of the knockers would be found on the tomb of Song Shaozu (buried 477): see Liu Junxi 刘俊喜, ed., Datong Yan bei shi yuan Bei Wei mu qun 大同雁北师院北魏墓群 (Beijing: Wen wu chu ban she, 2008), Chapter 5; and Annette L. Juliano, Unearthed: Recent Archaeological Discoveries from Northern China (Williamstown, MA: Sterling and Francine Clark Art Institute, 2012), 35–53, esp. 43–46. For suggestion of origin, see Lin, “Bei Wei Shaling bi hua mu yan jiu,” 17–20, who proposes possible links to the depictions of animals on the well-known belt buckles of the Xiongnu.
46. WS 30.725. He went on to serve in a high office in the guard units.
47. Duan and Zhao, Tian xia da tong, 58.
48. Li Ping, Bei Wei Pingcheng shi dai, Chapter 2. Kubozoe Yoshifumi 洼添庆文, “Guan yu Bei Wei de tai zi jian guo zhi du” 关于北魏的太子监国制度, Wen shi zhe (2002.1): 124–29, paints a little more complex picture of the situation, with the emperor having ultimate authority over the administration, and the heir some roles in military activity.
