Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī

Wen-shing Chou

Despite the legendary role ascribed to Shaolin monastery少林寺it is probably not an exaggeration to say that it has been considered sacrosanct within Chinese Chan Buddhist discourse [since at least] the mid-8th century that legitimacy comes from the south, and not the north. Since the tenth century, the rhetoric of the so-called ‘five schools’ has perpetuated peculiarly southern lineages; in practice, both the Linji and Caodong lineages (in China and beyond) propagate stories of celebrated patriarchs against a distinctively southern Chinese backdrop. What are we to make of Chan monasteries or cloisters in Ningbo, Fuzhou Jiangning, and of course, Hongzhou, apparently named to reflect the enduring significance of Mount Wutai五臺山, a notably northern sacred site? In the first part of this article I outline the less than marginal – or peripheral – role Mount Wutai appears to have played in ‘core’ Chinese Chan Buddhist sources. Then I proceed to explain how four Qingliang monasteries 清涼寺in southern China attest to the preservation and dissemination of a lineage of masters who supported what looks like a ‘Qingliang cult,’ with a set of distinctive teachings and practices that appears to collapse several longstanding assumptions about what separates Chan from the Teachings in Chinese Buddhism.

Qian Qianyi (1582–1664) was a prominent late Ming scholar-official who fashioned himself as a Buddhist in the early Qing. Despite Qian's claims that he had taken refuge in Buddhism, however, fellow literati close to him still inclined to emphasize his other identities fostered by a classical Confucian education—poet, scholar, and politician. Focusing on Qian's commentary on the Shoulengyan jing, i.e., the ragama Stra, and his compilation of Hanshan Deqing's (1546–1623) anthology, this paper unravels Qian's efforts in promoting Buddhism and documenting its history. It brings to light Qian Qianyi's hitherto overshadowed Buddhist endeavors, their historiographical and historical implications, as well as the complexities of interreligious dynamics during the Ming-Qing Transition.

This article focuses on the relationship between the controversial late-Ming playwright Tu Long and the contemporaneous Tiantai school of Buddhism represented by Youxi Chuandeng. It aims to contribute on the one hand to studies of Tu Long, which have paid little attention to Tu's connection with Tiantai, and on the other hand to studies of Chuandeng and the late Ming Tiantai revival, which have not yet explored the nexus between Chuandeng's philosophical teachings and his social network. I argue that it is most likely through Tiantai Buddhism, rather than "mad Chan" or Yangming philosophy, that Tu justified his antinomian deeds. This case study of Tu Long indicates that the success of the Tiantai school in the late Ming may be attributed to its fulfillment of the literati appetite for reading and writing, as well as to its tolerance of their unconventional lifestyle and syncretic spirituality. Unveiling some disagreements between Tu and Yunqi Zhuhong, whom many regard as Tu's guru, this paper highlights the agency of a lay Buddhist amid the tensions between rival schools and teachers.

,PSHULDO$SSDULWLRQV0DQFKX%XGGKLVPDQGWKH&XOWRI0DxMX‫ء‬Uά :HQVKLQJ&KRX $UFKLYHVRI$VLDQ$UW9ROXPH1XPEHUVSS $UWLFOH 3XEOLVKHGE\8QLYHUVLW\RI+DZDL L3UHVV '2,DDD )RUDGGLWLRQDOLQIRUPDWLRQDERXWWKLVDUWLFOH KWWSVPXVHMKXHGXDUWLFOH Access provided by Princeton University (2 May 2016 14:37 GMT) Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī Wen-shing Chou Hunter College Abstract This essay reconsiders the Qing imperial appropriation of the sacred mountain range of Wutai Shan through a study of three Manchu monasteries, Baodi Si, Baoxiang Si, and Shuxiang Si, built at the court of the Qianlong emperor between 1750 and 1775. Qianlong’s consuming interest in the vision cult of Wutai Shan’s resident deity Mañjuśrī is displayed in the building of the three monasteries, which were all modelled after famed temples at Wutai Shan. An investigation of the ritual, architectural, and artistic productions surrounding the three monasteries reveals the crafting of a distinct Manchu Imperial Buddhist identity centered on Qianlong himself as the apparition of Mañjuśrī at Wutai Shan. keywords: Manchu Buddhism, the Qianlong emperor, Wutai Shan, Baodi Si, Baoxiang Si, Xiang Shan, Shuxiang Si, Chengde, Mañjuśrī, Tibetan Buddhism, Chinese Buddhism, divine kingship. Q uite unlike the development of any other Chinese Buddhist sacred site, the holy mountain range of Wutai Shan 五臺山 (the Mountain of Five Terraces) (Fig. 1) in northeast China has, from the inception of its fame during the Tang dynasty, captivated the imaginations of ruling elites in China. During the last millennium and a half, numerous rulers of reigning dynasties, with the help of their religious advisers, enlisted Wutai Shan’s resident deity Mañjuśrī, as the protector of their nation, and sought to reinforce legitimacy for their rule through an alignment with Mañjuśrī’s earthly abode.1 The religious and worldly sagacity of Mañjuśrī, regarded as the Chinese bodhisattva par excellence and most often associated with qualities of wisdom, also became linked with Indian Buddhist models of religious kingship in both Chinese and Tibetan traditions.2 Sovereigns who identiied themselves or became identiied as the wheel-turning king (Skt. cakravartin) or the ruler of law (Skt. dharmarāja) also evoked ties to Mañjuśrī, sanctifying their secular role with a spiritual mission and condition. The Qing Manchu emperors added a new level of signiication to this millennial tradition of Buddhist kingship at Wutai Shan when they merged their own identities with that of Mañjuśrī—promoting themselves as emanations of Mañjuśrī through the uniquely Tibetan Buddhist notion of bodhisattva reincarnation.3 The Qianlong 乾隆 emperor (1711–1799), in particular, employed unprecedented visual, material, ritual, and rhetorical means to assert, over and over again, his identity as the wheel-turning Mañjuśrī-incarnate. This essay examines the Qianlong’s creative impersonation of Mañjuśrī through the construction of three temples around Beijing that were built to imitate (Ch. fang 仿) important monasteries at Wutai Shan. These three temples—Baodi Si 寶諦寺 (Temple of Precious Truth) and Baoxiang Si 寶相寺 (Temple of Precious Form) in Xiangshan 香山 (Fragrant Hills), the imperial park at the foot of the Western Hills just west of Beijing, and Shuxiang Si 殊像寺 (Temple of Mañjuśrī’s Image) in present-day Chengde 承德, the imperial summer retreat located 140 miles northwest of Beijing—were commissioned by the Qianlong emperor between 1750 and 1775 and were established as the irst of what eventually amounted to more than a dozen Manchu Buddhist monasteries. Although Qianlong’s forebears had supported Buddhism practiced by the multilingual constituents of his empire, and promoted the Gelukpa sect of Tibetan Buddhism (the sect of the Dalai Lamas) among them, it was Qianlong who initiated the translation of the scriptures into the Manchu language, mandated their ritual recitations, founded monasteries that were exclusively staffed by Manchu lamas with the help of his guru and state preceptor the Monguor4 reincarnate lama Chankya Rölpé Dorjé (1717–1786), and eventually undertook the monumental project of compiling the Manchu Buddhist canon. Why did Qianlong seek to create exclusively Manchu monasteries? Where was the place of Buddhism for a people who were initially forbidden to become lamas, and whose own religious tradition was preserved in imperial shamanist rituals that 140 ARCHIVES OF ASIAN ART Fig. 1. View from Central Peak, Wutai Shan, Shanxi Province, China. Photograph by the author. Qianlong himself had ordered to codify? Why and in what ways were three of the chief Manchu Buddhist monasteries derived from models of those at Wutai Shan? The term “Manchu,” though appearing to denote a unitary group of nomadic people who came to rule China through conquest, was coined in 1636 by Hong Taiji (1592–1643), Qianlong’s great-great-grandfather, in an effort to unite different Jurchen tribes on China’s northeastern frontier.5 By Qianlong’s time, this constructed ethnic marker had become an ancestral tradition that he sought to uphold in governing a largely Chinese empire through the preservation of shamanistic rituals and the use of the Manchu language.6 Since the status of ethnic Manchus as bannermen made it almost impossible for them to become monastics, Manchu lamas were selected instead from the booi (Ch. Baoyi 包衣) class of the Imperial Household Department (Neiwu Fu 內務府), rather than from the Eight Banners. Booi, which literally means “household persons,” were dependent servants who manned the Imperial Household Department, which managed the emperors’ personal affairs.7 Booi were therefore considered the emperor’s personal property.8 By the Qianlong period, they were mostly descendants of Han, Korean, Mongol, Jurchen, and even Russian groups who were previously captives of the Manchus and condemned by them to servitude.9 The so-called Manchu Lamas were thus Manchu-speaking people belonging to the court who were rarely ethnic Manchus. These monasteries did not extend beyond the court to include regular Manchu bannermen, and would have been seen only by resident lamas and close members of the imperial family. As institutions built and staffed by the Imperial Household Department, they should therefore be more accurately called Manchu Imperial Household Monasteries. My investigation into the making of these monasteries—based on extant archival, visual, architectural, sculptural, epigraphic, travel, and cartographic sources— demonstrates Qianlong’s singular preoccupation with the complex and continuously evolving projects of re-creating replicas of Wutai Shan’s temples over a twenty-ive-year span, with the goal of reenacting a particular vision cult of Mañjuśrī at Wutai Shan that had diverse followings WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī throughout Central Asia, East Asia, and the Himalayas. By restaging an apparition of Mañjuśrī10 through ritual, literary, and artistic means, Qianlong sought to craft and advance a distinct Manchu Imperial Buddhist identity centered upon himself as the Mañjuśrī of Wutai Shan. His appropriation of models and material manifestations of Mañjuśrī’s millennium-old vision cult from Chinese, Tibetan, and Mongolian sources and traditions further perfected, from the point of view of the emperor, a uniquely Manchu Imperial Wutai Shan in the original mountain range. This is not an attempt to perpetuate ixed notions of the ethnonyms “Manchu,” “Mongol,” “Tibetan,” and “Han,” nor to reify the categories of “Chinese” and “Tibetan” Buddhism in the eighteenth century. Indeed, recent scholarship has shown the extent to which these constructed categories were utilized by Qianlong at a time when the very deinition of “Manchuness” was being challenged.11 Instead of providing a static understanding of cultural and ethnic entities, this essay shows how luid the boundaries were between the perceived traditions of Chinese and Indo-Tibetan Buddhism, and how much Qianlong was responsible for the creation of a pan-Mahayana Buddhist narrative incorporating various Buddhist traditions. Qianlong’s engagement in prodigious cultural, political, and artistic enterprises, of which his Buddhist practice and activities were only a part, relected the dynamic and hybrid conditions of his empire that, despite his own heavy-handed effort to promote ixed notions of ethnicity, could not be reduced to cultural or ethnic terms. They also delineated an imperial project of cosmological recentering that places all under the emperor’s domain. The construction of the three temples at Xiangshan and Chengde lies at the nexus of imperial activities on several fronts: irst, the Qing imperial promotion of Wutai Shan—including frequent pilgrimages to the mountain range, sponsorship of its monasteries, engagement with rituals and initiations while at Wutai Shan, and the production of its gazetteers; second, Qianlong’s famously inventive “replicas” of great Tibetan monasteries and the pervasive culture of replication during his reign; third, the host of other visual and rhetorical assertions of “emperor-as-bodhisattva”; and inally, the making of the Manchu Buddhist canon, an immense task of translation that was structurally analogous to the building of Wutai Shan replicas. This essay, by situating the creation of Manchu imperial monasteries within the various all-consuming agendas and material productions of the Qing court that peaked during the Qianlong reign, brings to the fore the conceptual, geographical, and cosmological importance of Wutai Shan in the creation of 141 the Qing imperial Buddhist persona. A variety of source materials points synergistically toward Wutai Shan as an indispensable ground for Qianlong’s imperial selffashioning. Patricia Berger’s seminal work Empire of Emptiness (2003) remains the only art historical study to pay attention to Qianlong’s appropriation of the vision cult of Wutai Shan. In her study of Qianlong’s and his court artist Ding Guanpeng’s (丁觀鵬 [active 1708–1771]) copies of true images, which included a brief account of the miraculous icon at Wutai Shan, Berger reveals the transformative power of Qianlong’s copies for both the copy and the original.12 What remains to be elucidated in a thorough study here is Qianlong’s use of the potent signiications of the mountain and cult of Mañjuśrī to establish Manchu imperial Buddhism, as well as Qianlong’s comprehensive reconceptualization of Buddhist cosmology and historiography through these building projects. Following Berger’s use of the terms “copy” and “replica” to denote a range of emulative acts in the Qing court that interpret more than they duplicate,13 and in keeping with recent art historical scholarship that emphasizes the generative14 and revelatory nature15 of the copy, this essay investigates the logic and physical processes by which the past is made anew through the act of replication. Baodi Si: Founding a New Manchu Monastic Culture In 1750, immediately after Qianlong returned from a pilgrimage to Wutai Shan with his mother and his guru, the Monguor lama Rölpé Dorjé, he told the latter about his aspirations to build an exclusively Manchu Tibetan Buddhist monastery.16 (Rölpé Dorjé also accompanied Qianlong to Wutai Shan on his subsequent visits until his own death in 1786, and spent nearly all his summers from 1750 to 1786 in retreat there, frequently giving teachings and initiations.)17 Even though there had been Manchus who had become monks, an exclusive Manchu monastery would be the irst of its kind. To fulill his wish, Qianlong commissioned the building of a temple at the imperial park of Xiangshan west of Beijing that would be an imitation of Pusa Ding 菩薩頂 (Monastery of the Bodhisattva Mañjuśrī’s Peak) at Wutai Shan, a monastery that was built in the ifth century and originally named Wenshu Yuan 文殊院 (Cloister of Mañjuśrī), but renamed Pusa Ding early in the ifteenth century. Qianlong asked Rölpé Dorjé to be in charge of the new temple’s design, and named it Baodi Si.18 The monastery was completed in 1751, although by the end of 1750 two hundred Manchu lamas had already been chosen to study Buddhist scriptures in the Manchu 142 ARCHIVES OF ASIAN ART language at Baodi Si.19 Baodi Si subsequently became the headquarters for all twelve of the Manchu monasteries in and around Beijing that were built throughout Qianlong’s reign; a court-appointed oficial residing at Baodi Si oversaw all Manchu Buddhist affairs.20 What did Qianlong mean by an imitation? What aspects of Pusa Ding were copied, and what were his aims of such a material transfer? Even though no buildings from Baodi Si are extant, and no stele inscriptions survive or have been recorded, early maps, gazetteers, and court documents from the Imperial Household Department offer glimpses into the building process. They reveal a detailed attempt to re-create, and also to revise, the ritual setting of Pusa Ding. This concern for exactitude and speciicity of the ritual setting paved the way for what became the irst in a series of projects for the establishment of Manchu monasteries. The ways in which certain ritualized spaces and bodies became a medium through which imperial identity was articulated would become apparent in subsequent buildings of Baoxiang Si at Xiangshan and Shuxiang Si at Chengde. They would also clarify Qianlong’s choice and appropriation of Pusa Ding. Among Wutai Shan’s more than one hundred monasteries, Pusa Ding (Fig. 2) has been a locus of pilgrimage and imperial sponsorship since at least the Tang dynasty. Located on the summit of Lingjiu Shan 靈鹫山 (Vulture Peak Mountain, named after the Indian site where the Buddha gave many sermons), it is the highest point in the town of Taihuai 臺懷, the valley town between the ive terraces of Wutai Shan. As the name of the mountain Lingjiu Shan suggests, it is itself a Chinese transplantation of the Indian original, the source of Wutai Shan’s religious legitimacy in the irst place. According to the Expanded Record of the Clear and Cool Mountains, compiled around 1061,21 the irst temple at the summit of Lingjiu Shan was Wenshu Yuan, built by the Northern Wei emperor Xiaowen 孝文 (r. 471–499). The same record indicates that, although apparitions of Mañjuśrī were reported to have appeared on this peak frequently, it was not until the time of the Tang emperor Ruizong 睿宗 (662–716) that the temple featured a sculpted image of Mañjuśrī. This history is related in a well-known tale of the reclusive sculptor Ansheng 安生, of unknown origin, who, after many failed attempts to complete a sculpture of Mañjuśrī without cracks, appealed to the bodhisattva and then succeeded in making a perfect image by modelling it after seventy-two manifestations of Mañjuśrī that accompanied him as he completed his work.22 This temple, known thereafter as Zhenrong Yuan 真容院 (Cloister of the True Appearance), became a primary locus of pilgrimage and a conspicuous recipient of do- nations by emperors of successive dynasties. In maps of Wutai Shan from Dunhuang, for example, Zhenrong Yuan most often occupies the center of the composition. Even as the original icon had disappeared, and the temple’s name was changed to Pusa Ding during the Ming Yongle 永樂 period (1403–1424), stories of the miraculous image continued to be published in every imperial and nonimperial guidebook. In fact, the absence of the original image had in all likelihood served to enhance its allure, and contributed to the increasingly more elaborate narrative of its miraculous origin. By the early Qing dynasty at the latest, Pusa Ding became the chief Gelukpa monastery.23 The Geluk sect monopolized the mountain range after the founding of the Qing dynasty, when many of the temples were said to have been “converted” to Tibetan Buddhism.24 After the Fifth Dalai Lama’s visit to Beijing in 1652, the Qing Shunzhi emperor established the appointment of “jasagh lamas” (of Mongolian, Tibetan, and Han origins) to preside over religious affairs at Wutai Shan and installed monks from Tibet and Mongolia at Wutai Shan’s various monasteries.25 Although the position of jasagh lamas was also created at the capital in Beijing, Mukden, Hohhot, Jehol, and Dolonor, the successive jasagh lamas at Wutai Shan became especially tied to Tibet, as later regulations speciied that they should be drawn from a pool of lamas in Tibet.26 In order to house the jasagh lamas at Wutai Shan, Shunzhi renovated Pusa Ding extensively into an oficial imperial establishment (with yellowglazed tiles). Pusa Ding thus became the oficial residence of the jasaghs who oversaw all religious activities on a mountain range of some one hundred temples during the eighteenth and nineteenth centuries.27 It also housed the imperial travelling palace (xinggong 行宮), where the Kangxi 康熙 (r. 1662–1722), Qianlong (r. 1736– 1795), and Jiaqing 嘉慶 (r. 1796–1820) emperors all stayed during their numerous visits to Wutai Shan. By the reign of Qianlong, Pusa Ding housed approximately one-third of the three thousand lamas (of Tibetan, Mongolian, Manchu, and Han ethnic markers) who were residing at Wutai Shan. Because Pusa Ding was the undisputed center of worship and imperial sponsorship since the Tang dynasty, its re-creation at Xiangshan not only served as a substitute for the original monastery but also evoked the entire mountain range of Wutai Shan.28 This was relected in the couplet that Qianlong inscribed on a pair of placards hung at Baodi Si, proclaiming what the site was: a surrogate of Wutai Shan, which was a surrogate of India (by way of Lingjiu Shan) but much closer to his court than India or Wutai Shan.29 Qianlong’s choice of initiating a Manchu Buddhist monastery and housing it WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī 143 Fig. 2. Pusa Ding Monastery, Wutai Shan. From Daijō Tokiwa and Tadashi Sekino, Shina bunka shiseki vol. 1 (Tōkyō: Hōzōkan, Shōwa 14–16, 1939–1941), plate 92. in a surrogate of Wutai Shan’s most conspicuously imperial as well as Gelukpa Buddhist temple, deined by the memory of a miraculous icon, seems more than appropriate. As a sacred mountain range in China with deep roots in Tibetan Buddhism, and as the ield of enlightened activities for the deity of whom the Manchu emperors were considered incarnations, Wutai Shan was an excellent source and model for the inauguration of a new imperial Manchu monastic culture. Appropriately, as Pusa Ding was home to the jasagh lamas who oversaw all Buddhist affairs at Wutai Shan, Baodi Si, too, became the chief Manchu monastery that oversaw all Manchu Buddhist affairs. Baodi Si’s enormous scale and importance have long been obscured by its lost ediices and inscriptions. But thanks to extant maps and court records, we can reconstruct some of the precise ways in which the conceptual transfer was realized in material terms. First of all, the imitation seems to be at least partially relected in the design of the exterior architecture. A map of the imperial summer garden Yihe Yuan 頤和園 and the surrounding area (Fig. 3), which has been dated to after 1888, depicts a monastery with a stone gate at the entrance and steps leading up to it (Fig. 4). Photographs from the beginning of the twentieth century show a surviving stone gate of the same design (Fig. 5).30 This was presumed to be an imitation of the set of steps and the gate in front of Pusa Ding (Fig. 6), commissioned and inscribed by Qianlong’s grandfather, the Kangxi emperor. However, it is closer kin to the contemporaneously erected stone gate at Biyun Si 碧雲寺 (Azure Cloud Monastery) in Xiangshan, a Yuan-dynasty temple where Qianlong replicated a Tibetan-style Mahābodhi Temple in 1748 (Fig. 7)31—that is, although the 144 ARCHIVES OF ASIAN ART Fig. 3. Yiheyuan Baqi Bingying tu (Map of Eight Banners Brigade barracks and the Yiheyuan Summer Palace). Pen and ink and watercolor, 97 × 172 cm. After 1888. Original map and image are in the public domain; digital image provided by the Geography and Map Division, Library of Congress. steps clearly refer to Pusa Ding’s iconic set of 108 steps, no effort seems to have been made in its design to replicate the architectural style of Pusa Ding’s built environment. The gate is made out of stone rather than wood; moreover, the decorative details and the proportions of the architectural elements are entirely different from those on the gate at Pusa Ding. The scale of Baodi Si is conveyed only in a court document in the inancial accounts of the Imperial Household Department (Neiwu fu Zouxiao dang 內務府奏效檔) regarding its restoration beginning in 1770: a survey of Baodi Si conducted by that department, which recorded its ive-bay main hall, ive-bay rear tower, six-bay side hall, nine-bay dugang 都剛 (a large assembly hall), six-bay side hall near the front of the complex, three-bay hall of heavenly kings, three-bay mountain gate, bell and drum towers, eighteen-bay side dormitory hall, twenty-four-bay corner dormitory hall, and six-bay guard building, which makes a total of eighty-seven bays.32 The map of Yihe Yuan depicts only a single ive-bay central hall as Baodi Si’s main building, whereas the original complex at Pusa Ding would have featured four halls on the central axis, three of which had only three bays.33 Although we have limited knowledge of Baodi Si’s architectural exterior, records from the Palace Workshops of the Imperial Household Department (Neiwu fu Zaoban chu Jishi lu 內務府造辦處記事錄) reveal details about the complicated process through which Baodi Si’s interior was furnished. On April 6, 1750 (the thirteenth day of the second month of the ifteenth year of Qianlong’s reign), Qianlong issued a decree to obtain the dimensions of Pusa Ding’s Mañjuśrī dugang as well as model drawings of all its Buddhist images and ritual implements (jiang Pusa Ding Wenshu dugang dian dipan chicun foxiang [fa?]qi dengxiang ju huayang 將菩薩頂文殊都剛殿地盤尺寸佛像 [法?]器等項俱畫樣). This was undoubtedly preparatory work required for the building of Baodi Si. The term dugang is a Chinese transliteration of the Tibetan word ’du khang, a large assembly hall within a monastery where monks gather for prayer recitations. In the eighteenth-century Chinese imperial gazetteer, only one other monastery at Wutai Shan was listed as having a dugang.34 It is not clear how dugang halls at Wutai Shan actually followed the design of a Tibetan ’du khang, but they certainly refer to halls that can accommodate large monastic assembly in the Tibetan WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī 145 Fig. 4. Detail of Fig. 3. tradition. Since Manchu Buddhism was in large part the practice of Buddhism in the Manchu language following the Gelukpa tradition, according to Rölpé Dorjé, Qianlong’s guru and state preceptor, the modelling of a Manchu monastery on a Tibetan assembly hall would have made perfect sense. During the same week, however, Qianlong also ordered the measurement and construction of a model of another hall at another monastery at Wutai Shan, Xiantong Si’s Beamless Hall (wuliang dian 無梁殿), presumably also intended as a potential model for the monastery of Baodi Si.35 The so-called beamless hall refers to a vaulted masonry hall that does not require the beams of traditional post-and-beam construction. The term “beamless” in Chinese (wuliang 無梁) is a homophone for “immeasurable (wuliang 無量),” and these are the Chinese translations of the names for the Buddha of Immeasurable Light and the Buddha of Immeasurable Life (Skt. Amitābha and Amitāyus).36 For this reason, beamless halls, which only numbered about a dozen in China and were considered to have non-Chinese origins, usually carry the connotation for longevity and were therefore often used for birthday celebrations. Together, these records suggest that Qianlong was initially looking toward different temples as potential models for his replica. Qianlong’s decree of April 6, 1750, provided no further details about what Buddhist images and ritual implements were to be modelled, but records of the weeks 146 ARCHIVES OF ASIAN ART Fig. 5. Stone Gate at Baodi Si, Xiangshan (1750), 1906–09. Photograph. From Ernst Boerschmann, Chinesische Architektur (Berlin: E. Wasmuth, 1926), plate 267. to follow suggest that the copying was carried out in earnest. They also relect that the principal concern in building Baodi Si was the proper setting for rituals rather than the imitation of any particular architectural features. On April 12 (the sixth day of the third month), Qianlong ordered that the sets of Seven Treasures (qibao Fig. 6. Gate at entrance to Pusa Ding, Wutai Shan. Photograph by Ani Lodro Palmo, ca.1985. 七寶) and Seven Royal Treasures (qizhen 七珍) from Wutai Shan be repaired, and ive days later, he ordered two sets of replicas of the Five Treasures (wubao 五寶), Seven Treasures, and Eight Treasures (babao 八寶) together with offering tables, all of which were brought from Wutai Shan.37 The hall or monastery of origin was not speciied in this record, although it is clear that they would have come from a Tibetan Buddhist temple, presumably Pusa Ding’s Mañjuśrī dugang, the only one of the two dugangs at Wutai Shan mentioned in the records. As with the rest of the ritual paraphernalia Qianlong subsequently commissioned, the two sets of replicas were probably intended for speciic locations—one for Baodi Si, and the other to be sent back to Wutai Shan. The various sets of ritual offerings, which would have been placed in front of the main icons, are standard offerings within Tibetan traditions that are absent in their Han-Chinese counterparts, which would have featured only a much simpler set of Five Offerings (wugong 五供). Found throughout Qing imperial temples, these offerings were either produced at the court or given as gifts by high-ranking lamas visiting from Tibet and WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī 147 Fig. 7. Stone Gate at Biyun Si, Xiangshan, 1748. Photograph by the author, 2009. Mongolia.38 Equally important as their cultural and religious association to Tibetan Buddhism were their imperial connotations; for example, the possession of the set of Seven Royal Treasures, which originated in preBuddhist India, was one of the deining features of a wheel-turning worldly sovereign (cakravartin).39 Even though in Buddhist traditions, the set of Seven Royal Treasures later became ritual offerings to the Buddha, they still carried with them imperial connotations and were regularly used to furnish imperial chapels, especially during the Qing dynasty. Such offerings—some of which are still in their original locations, and many more of which were looted and sold, and are today scattered in museum collections around the world—have by now become visual hallmarks of Buddhism in the Qianlong reign (e.g., Fig. 8). Thanks to records from the Imperial Household Department that describe details about the material, color, pattern, and precise type of enamel, and so forth of the offerings and the frames and stands that were made for them, we know that the sets that were produced for Baodi Si and repaired and possibly remade for Pusa Ding would have looked very much like extant examples. The insistence on repairing and reproducing a Tibetan Buddhist ritual setting is consistent with an all-consuming effort at rectifying and standardizing ritual and iconography at Qianlong’s court, in each case of which an Indo-Tibetan, rather than a HanChinese model, was followed.40 A similar attempt to standardize and reintroduce Indo-Tibetan ritual and iconography can be observed here. Still, despite the fact that these offerings were so ubiquitous and nearly synonymous with Qianlong-era Buddhism, what we can glean from the records is an insistence on replicating and repairing the particular sets of ritual offerings at Pusa Ding. Again on April 26, Qianlong ordered two sets of replicas of Pusa Ding’s 148 ARCHIVES OF ASIAN ART Fig. 8. Offerings of Seven Royal Treasures (top register) and Eight Treasures on display at Treasure Gallery, Palace Museum, Beijing. Photograph by the author, 2008. mandala offerings, and various offering tables and offering bowls, together with elaborately designed sets of Five Sense Offerings (wuyugong 五欲供), and Eight Offerings (bagong 八供); and one each of these complete sets, when inished, was to be placed at Baodi Si and the other brought back to Wutai Shan’s Pusa Ding.41 Qianlong’s concern for the precision of the ritual setting can also be seen in his frequent instructions that the sets be veriied and authenticated by Rölpé Dorjé. Together, the numerous records of production undertaken within a period of three weeks indicate that not only were the temple architecture and its interior furnishings, cloth hangings, streamers, images, offerings, and ritual implements to be replicated wholesale, but also that this replication process was an occasion to make the original more perfect, and the two sites more precisely and perfectly congruent. As Patricia Berger astutely noted with regard to Qianlong’s replicas of Inner Asian temples as well as his copying of previous paintings and icons, every act of copying reinterprets and revises the original, such that “the original was also forced to live up to the expectations of the copy.”42 As the irst Manchu monastery to be built from the ground up, Baodi Si relied on the precise transferring and perfecting of Pusa Ding’s ritual setting to create a familiar albeit dis- tinctly imperial Buddhist identity centered on the deity Mañjuśrī. Available sources did not spell out Qianlong’s appropriation of Wutai Shan beyond ritual eficacy, but the persistent centrality of Mañjuśrī’s vision cult becomes apparent in his subsequent projects. Baoxiang Si: Staging an Apparition After Qianlong’s pilgrimage to Wutai Shan in 1761, the twenty-sixth year of his reign, his attention shifted from the realm of ritual and architecture to the appropriation of a famed icon: an image of Mañjuśrī on a lion. This trip—Qianlong’s third visit to Wutai Shan, and the second time he went there with his mother—coincided with the empress dowager’s seventieth birthday and Qianlong’s own iftieth birthday.43 The pilgrims were greeted with appropriate fanfare, including the performance of a six-part drama presented in honor of the double birthday celebration.44 At Shuxiang Si, the Temple of Mañjuśrī’s Image (Fig. 9), Qianlong was awestruck with the temple’s widely revered namesake image of Mañjuśrī on a lion, a sculpted igure that especially attracted pilgrims from Tibet and Mongolia and that still survives today in its repainted and restored form (Fig. 10). Qianlong was moved to make at least two WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī Fig. 9. Hall of Mañjuśrī, Shuxiang Si, Wutai Shan. Photograph by the author, 2009. Fig. 10. Mañjuśrī on a lion, Hall of Mañjuśrī, Shuxiang Si, Wutai Shan. Photograph by the author, 2015. sketches of the image plus a lengthy inscription while en route back to Beijing.45 This was a rare gesture for an emperor who wrote voluminously but was hardly known for his own paintings.46 Consider, for example, that during Qianlong’s previous trips to Wutai Shan, he had ordered court oficials Zhang Ruo’ai and Zhang Ruocheng to compose traditional landscape paintings of a snowy scene, on which he wrote lengthy colophons (Figs. 11 and 12). They represent a conspicuous identii- 149 cation with the Chinese classical tradition of gentlemanly cultivation. By contrast, here his attention was turned toward the single icon and to capturing its true trace with his own hand. According to records from the Imperial Workshop for Carvings and Paintings, known as Ruyi guan 如意舘 (Wish-Fulilling Studio), one of the sketches entered the imperial art collection and was subsequently remounted several times over the next several years. While the sketch does not appear to have survived, it subsequently became the basis for the building of an even larger temple next to Baodi Si. Qianlong’s original sketch was, according to his instructions in the colophon of the sketch, enlarged and transferred onto a stone stele.47 In 1762, Qianlong ordered a sculpted replica of the image based on the engraving from the stele, and asked Rölpé Dorjé to design a temple to house this image.48 The temple, which he named Baoxiang Si, was built immediately adjacent to Baodi Si on its western side (see Fig. 13). It was completed in 1767, and the stone stele bearing the engraving was placed next to it, although it and other steles were already fallen by the early twentieth century (Fig. 14).49 Court documents suggest that as soon as construction was under way, Manchu lamas were selected from the booi class and placed there. As early as 1763, only one year after the building project began, the monastery had expanded to include the addition of sixty lamas.50 What about this image so captivated Qianlong? The icon at Shuxiang Si featuring the image of Mañjuśrī on a lion has a complex genealogy. Shuxiang Si is located on the edge of the Taihuai village where major temples, including Pusa Ding, are clustered. It was rebuilt in 1496 after structures from preceding dynasties were burned to the ground. During the Ming and Qing dynasties, it became a large, imperially sponsored monastery and underwent major renovations. Already a prominent pilgrimage destination and a recipient of imperial sponsorship, Shuxiang Si was frequently visited by the Kangxi emperor, who wrote numerous poems about the remarkable characteristics of the image (faxiang zuiyi 法相最異) and made very generous donations for its restoration.51 Even though the monastery had always been Chinese Buddhist in afiliation, rather than Tibetan Buddhist, it became so revered among the Tibetan and Mongolian population that the Tümed Mongol prince Yéshé Döndrup (Ye shes don grub bstan pa’i rgyal mtshan, 1792– 1855) authored a text on the history and environs of Shuxiang Si with the help of the eminent Tibetan Buddhist grammarian Ngawang Tendar of the Alasha banner (A lag sha Ngag dbang bstan dar, 1759–1831).52 This Mongolian language guidebook about the exalted image at Shuxiang Si was published and translated into Fig. 11. Zhang Ruo’ai, Zhenhai Si, 1746. Colors on paper, 127.6 × 62.8 cm. Photograph provided by the Palace Museum, Taipei. WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī 151 Fig. 14. Main Hall of Baoxiang Si. Photograph. From Ernst Boerschmann, Chinesische Architektur (Berlin: E. Wasmuth, 1926). Fig. 12. Zhang Ruocheng, Zhenhai Si, 1750. Colors on paper, 103.4 × 56.9 cm. Photograph provided by the Palace Museum, Taipei. Fig. 13. Detail of Fig. 3, Boaxiang Si, Yihe Yuan Baqi Bingying tu. Tibetan around 1813, attesting to the image’s popularity among Mongol and Tibetan pilgrims. It is recorded in Rölpé Dorjé’s biography that before he passed away at Wutai Shan in 1786, he led an assembly of prayers in front of a magniicent image of Mañjuśrī in a great hall, and was joined by the emperor. It is quite likely that the icon at Shuxiang Si was the very image mentioned.53 When the Russian diplomat Dmitri Pokotilov visited Shuxiang Si in 1903, he credited the monastery’s survival well into the twentieth century to the nonstop low of donations from Mongol pilgrims at a time when donations for all other monasteries at Wutai Shan were dwindling, even though Shuxiang Si was never a Tibetan Buddhist monastery.54 This image of Mañjuśrī at Shuxiang Si (Fig. 10) can be dated to 1496, less than a decade after the main hall was erected (1489). In fact, what is referred to as an image here and in the imperial records (the Chinese word is xiang 像) probably has existed for most of its history, and exists in the current version, as a sculptural group, composed of a central igure of Mañjuśrī seated atop a lion dais, lanked by the igure of the Khotanese king as a lion-tamer (leading the lion by a leash), the youth pilgrim Sudhana from the Gaṇḍavyūha chapter of the Avataṃsaka Sūtra, and several other attendant igures.55 The iconography of Mañjuśrī riding on a lion and accompanied by a lion-tamer can be traced back to the lost sacred icon at Pusa Ding/Zhenrong Yuan, the temple that later became the model for Qianlong’s Baodi Si.56 Even though the original image from the Tang dynasty is no longer extant, iconographic assemblies similar to what is found at Shuxiang Si were popular in Dunhuang, Japan, and at Wutai Shan itself from as early as the tenth century, and even made its way to the ifteenth-century 152 ARCHIVES OF ASIAN ART iconographic pantheon of Gyantse Kumbum in central Tibet (Fig. 15).57 Surviving images from Dunhuang, Japan, and central Tibet show that they share more or less the same iconography, with Mañjuśrī on a lion as the central deity, a Khotanese king as lion-tamer, and the youth Sudhana as an attendant disciple. Even though the iconographic origin of this sculptural group is still a matter of scholarly dispute, we know for certain that it became associated with the cult of Wutai Shan; when and wherever it appeared, these Mañjuśrī igures harked back to and served as a synecdoche for Wutai Shan. Not unlike the competition for relics in medieval Christian churches, monasteries within and beyond Wutai Shan competed for ownership of Mañjuśrī’s true presence as manifested in the sculptural group in order to assert their centrality in the pilgrimage circuit. It appears that Shuxiang Si succeeded in its claim for the true presence of the bodhisattva and maintained it from the Ming dynasty onward. The sculptural group at Shuxiang Si acquired more than its iconography from Pusa Ding/Zhenrong Yuan. According to the widely recounted origin tale of Shux- iang Si, Mañjuśrī appeared in perfect form in the sky to a frustrated sculptor experiencing artist’s block.58 Apparitions, whose elusive guise is given tangible form only through miracle tales, have inherently complicated and extended genealogies. Mongolian and Tibetan recensions of the story provide more speciics for this particular image. In one account, when the deity instructed the sculptor to make an image in his likeness, the sculptor improvised by grabbing the nearest available dough in the monastery’s kitchen (it was nearly lunchtime) and molding it into the shape of the apparition’s head.59 The sculptor in another account, while holding up a piece of barley bread as an offering for the majestic apparition in the sky, received blessings from Mañjuśrī in the form of the bodhisattva’s perfectly shaped countenance in the bread, and subsequently completed the rest of the body to create a statue of exceptional beauty.60 Today, this image is still referred to as the “Buckwheat-dough-headed Mañjuśrī” in Tibetan and Mongolian sources (Tb. ’Jam dbyangs rtsam mgo, Mong. Gulir terigütü manzusiri).61 Sure enough, during the 1983 restoration, it was discovered that the Fig. 15. Mañjuśrī on a Lion with Five Attendants, main sculptural image in the ifteenth chapel, second loor of Gyantse Kumbum, Gyantse, Central Tibet, 15th century. Photograph by the author. WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī head was really made from buckwheat, with clay illings for holes created by resident mice.62 It is particularly interesting that this popular legend with “a grain of truth” is preserved in Mongolian and Tibetan languages, but is not included in Chinese-language texts, further attesting to the fact that the predominant populations venerating the image were Mongols and Tibetans during the late eighteenth and nineteenth centuries. Even though the image’s perfect form, which artists can only create through divine intervention, is a common trope for sacred images or for any work of high artistic merit, variations of the tale resonate most closely with that of the Tang dynasty Mañjuśrī on a lion made by the sculptor Ansheng, modelled after an apparition at Zhenrong Yuan. In both tales of miraculous images from Zhenrong Yuan and from Shuxiang Si, the bodhisattva comes to rescue the troubled artisan by manifesting his true form. Although the sculptural group at Shuxiang Si has a distinct local lavor relecting the Tibetan Buddhist population at Wutai Shan during the later period, it can be considered a true substitute for the miraculous image from the Zhenrong Yuan story, made at a time when the image from Zhenrong Yuan had long disappeared. In fact, it was erected right around the time the Tang dynasty Mañjuśrī disappeared from Zhenrong Yuan, during the Ming dynasty (no later than 1482), and soon earned its renown as the only “true image” of Mañjuśrī in the Taihuai valley of Wutai Shan.63 For pilgrims, the sculptural group at Shuxiang Si therefore became a sort of replacement of the original one at Pusa Ding, satisfying a thousand-year-old zeal for the bodhisattva’s true countenance. In the most authoritative Tibetan-language guidebook since the late-eighteenth century, compiled by Rölpé Dorjé and his disciples, many stories from Chinese-language gazetteers were abbreviated, whereas stories of the miraculous images of Mañjuśrī at Pusa Ding and Shuxiang Si were reiterated in greater detail than available in the Chinese source texts, attesting to their historical signiicance for the Tibetan and Mongolian populations, despite the fact that Shuxiang Si was not itself afiliated with Tibetan Buddhism. Qianlong’s court was no doubt aware of the distinction at the practiced level as well; in 1768, it was the Chinese ritual setting of Five Offerings, not the elaborate setting of a Sino-Tibetan Buddhist altar as recorded in the building of Baodi Si, that were placed in the main hall of the main altar at Baoxiang Si.64 Rituals of Transformation Beyond his spontaneous experience of awe before the image of Mañjuśrī, Qianlong was no doubt deeply 153 aware of the power of that image and its miraculous origins. Even though the stele at Baoxiang Si and the sculptural group based on Qianlong’s original sketch are either no longer extant or inaccessible (as the hall housing the sculptural group is currently in a veterans’ rehabilitation center contained within the walls of a military compound off-limits to the public), two surviving paintings and one textile from the same series of replicas shed light on his interest in and manipulation of the apparition’s many lives. As soon as he returned to Beijing in 1761, Qianlong ordered court painter Ding Guanpeng to make a large painting based on his original sketch. Documents from the Imperial Workshop record several paintings ordered multiple times through the year 1761.65 Two of the paintings, along with one of Qianlong’s own sketches, as well as a closely related textile gifted by the mother of Qianlong’s trusted oficial, entered Midian Zhulin 秘殿珠林, Qianlong’s catalogue of religious art. The two paintings and the textile are now in the collection of the National Palace Museum in Taipei.66 Matching the inscription on one of them to documents from the Imperial Workshop, the two paintings can be dated to the fourth and twelfth months of the twenty-sixth year of Qianlong (i.e., 1761), respectively (Figs. 16 and 17).67 The earlier painting is made up of many small pieces of paper, suggesting that it might have acted as a kind of large preparatory painting for the second painting, which, as Ding notes in his colophon, took seven months to complete. The three monumental images are of similar dimensions, each measuring about ten feet in height and ive feet in width. Except for some seals along the edges, the entire length of each composition is occupied by a single bodhisattva on a lion in a highly unusual backgroundless void. Gone too are Mañjuśrī’s illustrious attendants, such as Sudhana and the Khotanese King, who had been an integral part of the miraculous image in replicas from Tibet to Japan. A detailed comparison of the two paintings reveals the many subtle, calculated adjustments that were made between the painting done in the fourth month and the painting completed in the twelfth month, suggesting that the second painting was indeed a correction or modiied version of the irst. To make the matters more intriguing, a third monumental image (Fig. 18), rendered in the medium of embroidery by the mother, wife, and granddaughters of the court oficial Qiu Yuexiu 裘曰脩 (1712–1773) as a gift to the emperor, entered the imperial collection (for the catalogue of which Qiu was one of the compilers), and was based closely on the earlier of the two paintings, save perhaps for the feminization of the bodhisattva’s face.68 154 ARCHIVES OF ASIAN ART Fig. 16. Ding Guanpeng, First painting of Shuxiang Si’s Mañjuśrī on a lion. 1761. Hanging scrolls. Ink and colors on paper, 297.3 × 159.1 cm. Photograph provided by the Palace Museum, Taipei. Through these small but profound changes in the portrayal of Mañjuśrī’s physiognomy and attire, Qianlong’s manipulation of a thousand-year-old lineage of iconic production becomes clear. A consistent transformation of the igure from an idealized Chinese bodhisattva to a “humanized” tantric initiate subtly forges a link between Wutai Shan’s famous icon with the emperor himself. While the earlier version (Fig. 19) bears the rather round face and softly rounded chin of a bodhisattva igure in Ming-dynasty Chinese Buddhist Fig. 17. Ding Guanpeng, Second painting of Shuxiang Si’s Mañjuśrī on a lion. 1761. Hanging scrolls. Ink and colors on paper, 297.3 × 159.1 cm. Photograph provided by the Palace Museum, Taipei. paintings, the later painting (Fig. 20) shows Mañjuśrī with a somewhat angular, more elongated face, making him look more human; and the parallel curves just below the bodhisattva’s chin are replaced by a single curve of a leaner face with a protruding chin. Whereas Mañjuśrī’s eyelids in the earlier painting are more closed, ever-so-slightly, gently downcast with pupils undistinguished from the irises, conveying the compassionate gaze for all sentient beings that can often be seen in earlier depictions of Chinese bodhisattvas, the eyelids in the later painting appear to be opened wider through the heightened contrast between the dark pupils and the WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī 155 Fig. 19. Detail of Fig. 16. Fig. 20. Detail of Fig. 17. Fig. 18. Mother of court oficial Qiu Yuexiu, Shuxiang Si’s Mañjuśrī on a lion. Hanging scroll, embroidery, 354 × 150.3 cm. Photograph provided by the Palace Museum, Taipei. lighter irises as well as the slight thickening of the upper eyelids. These modiications create the impression of an active human gaze, set off by a noticeably wider nose and thicker, more natural, and less shaped eyebrows. In the earlier painting, bizarre snakes of hair fan out symmetrically to either side of Mañjuśrī’s neck, while largebeaded earrings and strings of small pearls hanging down from his crown lare outward and lank a circle of stiff folds in the collar with equally unconvincing animation. This highly implausible but dramatic upper portion 156 ARCHIVES OF ASIAN ART Five Buddha crown in the irst image looks like a crown worn by a deity or a priestly igure, often seen in depictions of Mañjuśrī from the Ming dynasty onward, and is likely a more accurate depiction of the Shuxiang Si image at the time. In contrast, the crown in the second image is formed of distinct lat panels receding back as it encircles the bodhisattva’s head, more in keeping with a crown worn by ritual specialists or practitioners during a Tibetan Buddhist tantric rite. (A ritual crown worn by none other than the Qianlong emperor himself when he underwent tantric initiations in 1780 displays a similar design, Fig. 21.) Atop the crown in the second image, Mañjuśrī’s previously unadorned topknot is now adorned with a small gold image of a seated Amitabha Buddha and encircled by colorful gems set within gold “lames.” Embroidered images of a seated Buddha Śākyamuni adorn two pendants that hang down from either side of the crown and over Mañjuśrī’s shoulders. The heavily cloaked bodhisattva in the irst image undergoes a change of season in the second image by wearing what appears to be a diaphanous collar above an elaborate chest plate decorated with netted beads, precious stones, and small gold plaques featuring Buddha images. Whereas the beaded chest plate of the irst bodhisattva features a single image of what appears to be Buddha Śākyamuni in an earth-touching gesture, the beaded chest plate of the second features twelve Buddhas, most visibly a cosmic Buddha Vairocana (with hands held in the dharmacakra mudra position) at the center of his netted chest plate (Fig. 22). The modiied Mañjuśrī is bedecked with Buddha images from head to toe—numerous golden Nirvana Buddhas in the crown, Fig. 21. Gilded gold ritual crown with Five Directional Buddhas used by Qianlong in 1780. From Du Jianye, Yonghegong: Palace of Harmony (Hong Kong: Yazhou yishu-Art Blooming Publ., 1995), 220. of the painting is reduced to stillness and simplicity in the second painting, where the strings of pearls curving outward are replaced by straight-hanging pendants of embroidered cloth, and the bodhisattva’s hair is now neatly tucked away behind an identical but smaller pair of earrings that also hang downward, in accordance with the law of gravity and the decorum of royalty. Other features also mark a clear shift from an idealized bodhisattva igure to a more “humanized” one. The Fig. 22. Detail of Fig. 17. WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī 157 Fig. 23. Ceremonial costume for an imperial lama: Beaded collar and apron. From Du Jianye, Yonghegong: Palace of Harmony (Hong Kong: Yazhou yishu-Art Blooming Publ., 1995), 221. in the jeweled net, and on the petals of the lotus throne. Mañjuśrī’s lion, now positioned nearly sideways to reveal the length of its body, sports a matching collar and apron made of equally ine netted beads, jewels, and bells, though (appropriately) without Buddha images. Like the depiction of the ritual crown, these nets of beads and plaques resemble those that would have been worn by those undergoing important Tibetan Buddhist tantric rites. The depiction is nearly identical to a contemporaneous set preserved at the Yonghe Gong (Palace of Peace and Harmony) in Beijing (Fig. 23).69 The pervasive appearance of multiple Buddhas on the second bodhisattva, just as on Tantric Buddhist crowns and chest plates, visu- ally reinforces the transformative capacity of tantric rituals to unite a human being with his Buddhahood. Yet other representational and iconographical changes from the irst to the second painting bring the bodhisattva from an otherworldly space to that of the viewer, further enhancing the human-like quality of Ding’s second painting. Judging from the posture of the igure and the sculpture’s current appearance, it is most likely that Mañjuśrī balanced a ruyi scepter between his hands in the original sculpture, as he does now (Fig. 10). Mañjuśrī’s hands in Ding’s irst painting are depicted in the same position, with ingers curved slightly inward, albeit without holding any implement. In Ding’s 158 ARCHIVES OF ASIAN ART second painting, however, Mañjuśrī’s right arm and palm open up completely to abhaya mudra (gesture of fearlessness/protection), and his left arm is placed on his knee as though in a gesture of royal ease. The clariied mudra of the right hand as well as the palpable weight of the left hand resting on the knee convey a presence and immediacy that is accentuated by the change in the lion’s position. Again, possibly following the sculpted image at Shuxiang Si, the lion in Ding’s irst painting stands with its head turned upward and to the left. In contrast to the dynamic upper part of the painting around the bodhisattva’s head and upper body discussed above, here the lion’s mane appears in orderly patterns, neatly combed on his back. His head is turned away from the viewer, and his legs stand on free-loating lotus blossoms, which demarcate a self-enclosed, otherworldly space. But in Ding’s second painting, the lion faces forward, its head and body are rotated clockwise to reveal a semi-proile view, and its paws are planted squarely on the ground. This perspective (combining frontal and semi-proile views), implausible for a three-dimensional form, asserts a pictorial independence from its sculptural origin. Unlike Ding’s irst painting, in which the upper portion features more movement than the lower portion, the lower portion of his second painting becomes the active center of the composition: the bodhisattva’s foot, with the ankle now exposed, presses against a tilted lotus blossom on a vibrantly ornate saddle, while illusionistic ribbons, bells, feather-ornament, hair, and lames all lutter in gusts of wind that do not affect the upper portion of the painting. Here, the lion’s frontal, animated, and grounded stance puts the bodhisattva’s calm but human and almost confrontational presence in the here-and-now right into the space of the viewer. All together, these modiications mark a substantial ontological shift—from the portrayal of the miraculous sculptural image of Mañjuśrī, with all of the earthly trappings and emotive transcendence of a Mahayana Chinese bodhisattva igure, to the intimation of divinity in a human body through ritual transformation. The idea that a person can be ritually transformed into a receptacle of the divine is a hallmark of Tibetan tantric Buddhism;70 that the person carries the trappings of royalty further marks the igure of a tantric cakravartin (a universal, enlightened ruler, whose reign brings peace and justice). In the modiications of the original icon, Ding’s second painting therefore superimposes the esoteric, and speciically Tibetan, tradition of ritual transformation and an Indian ideal of Buddhist kingship onto a Chinese Buddhist icon with a popular Mongolian cult following, visually and metaphorically reenacting the bodhisattva’s hybrid identity through Indo-Tibetan, Mongolian, and Chinese iconography and history. In light of the fact that Qianlong himself had undergone tantric initiation rituals (the implements from some of which are still visible, see Fig. 21, for example), plus the wealth of textual and visual materials produced at the Qing court that asserted his status as the cakravartin-bodhisattva incarnate, it would not be too far-fetched to see Ding’s second painting as a portrayal of Qianlong himself.71 The vitality of ritual in Qing rulership has been at the center of recent scholarship. Angela Zito, in her study of the grand sacriice—the most signiicant ceremonial occasion for the Qing emperors—showed how the performance of ritual texts “make manifest” the power of the heavens in human affairs and the power of the past in the present, and argued that the emperor, by donning a variety of ceremonial robes, “embodied” his constituencies.72 James Hevia analyzed Qing guest ceremonies and found that rituals of inclusion (guest / host rituals) and transformation (initiation rituals) were ways to “encompass and include others in their own cosmologies.”73 The painting of a revered sculptural icon in the guise of an imperial tantric initiate reiterates the primacy of the ritual reenactment as a category in the physical and metaphysical articulation of Qianlong’s imperial identity, as does the wholesale replica of Pusa Ding’s ritual space in the building of Baodi Si. But what about Ding’s second painting, which explicitly and exclusively establishes Qianlong’s identity? After all, the face of the igure in the painting does not look anything like that of Qianlong’s, as we have come to know so well from a plethora of Castiglionesque paintings of him. Considering Ding’s painted “copies” in light of their multiple origins going back to the Tangdynasty sculpture at Zhenrong Yuan, and memories of the miraculous original(s) that are kept alive in countless textual, visual, and oral iterations in the Chinese, Tibetan, and Mongolian languages, the self-referentiality of Qianlong’s interventions becomes clear: if in this particular reenactment Ding Guanpeng played the role of the skilled artisan who helped make manifest the earthly form of Mañjuśrī, then Qianlong’s sketch is the mediating force—that is, the divine intervention of Mañjuśrī that prompted and guided the image-making process.74 In Ding Guanpeng’s paintings, Mañjuśrī is therefore not only the subject but also the agent of the depiction, and that agent is none other than Qianlong himself. Ding’s paintings thus take the appositional relationship between the emperor and the bodhisattva to a level of unprecedented speciicity. Acting as a referent in the double sense of the word (one who refers and one who is referred), Qianlong implied a connection with Mañjuśrī beyond resemblance. The staging of himself WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī as a new “apparition” of Shuxiang Si’s miraculous image, and a refashioning of the image as a royal tantric initiate—as in the case of Ding’s second painting— allowed Qianlong to embody the bodhisattva of Wutai Shan, and thereby also perfect it. Much like the reparation and replication of ritual offerings at Pusa Ding, the copying of works modeled after Qianlong’s sketch was a reclaiming of the ownership of Wutai Shan; however, unlike the wholesale replica of a temple interior, this was a far more succinct assertion, one that reached the diverse pious constituencies of Shuxiang Si’s miraculous image. Forestalling any possibility that this nuanced substitution might go undetected, Ding’s unusually lengthy inscription on the second painting makes explicit that by “relying on the heavenly brush [of the emperor],” he was able to complete Mañjuśrī’s golden countenance. Ding then compares himself to the artisans who carved the sandalwood Buddha commissioned by the Indian King Udayana, but attributes the inadequacy of the inal result to his own lack of insight.75 That Qianlong’s divine intervention is analogous to the famed miraculous image of the sandalwood Buddha further sealed the identity of a Buddhist king.76 This identiication may also explain why Mañjuśrī’s entourage was eliminated in Qianlong’s copies: in this new guise of emperor as bodhisattva, these mytho-historical igures from another time and place are no longer relevant. Furthermore, if in the eighth century Zhenrong Yuan became a locus of pilgrimage on account of a miraculous icon of Mañjuśrī, it stayed as the center of pilgrimage in the Qing despite the loss of its namesake icon. What need is there for an icon when it is the very abode of the personal embodiment of the bodhisattva, the Mañjughosa emperors? Qianlong’s series of enactments reveals not a simple assertion of his identity as Mañjuśrī vis-à-vis his Tibetan and Mongolian constituents, but his role as a benevolent, universal Buddhist ruler over the vast domains of the image’s sway. By appropriating Wutai Shan’s most emblematic icon, Qianlong inserted himself in the place of both the apparition and the icon. Even if we understand the implications of this transformation from an idealized bodhisattva to a humanized one, who was responsible for it? Under whose command were all the subtle adjustments evident in Ding’s second painting completed? Was it based on a directive issued by Qianlong himself, or was it Ding’s own decision to depart from Qianlong’s sketch? While we may never know the answer, records from the Imperial Workshop reveal that Ding was asked to use several sources for his second painting, which took seven months to complete: two sketches by Qianlong, Ding’s earlier painting, and, most directly, a wax model of Mañjuśrī.77 159 In fact, it must be concluded, in light of the fact that several extant two-dimensional and three-dimensional replicas more closely conformed to one another, that Ding’s second painting was an even more striking departure, one that is exclusively asserted through the two-dimensional medium of the painting. Ding’s inscription described Mañjuśrī’s countenance with phrases that evoke the imagistic metaphors of Chan Buddhism—“radiant with the subtle glow of wisdom, [the relection of] the moon that seals the river” (yuanguang moshi, ruyue yinchuan 圓光默識。如月印川)—suggesting that the true countenance of Mañjuśrī exists beyond the ordinary external physical appearance. It follows that to make a true copy of the divine, one must not only painstakingly copy the external features but must also discern the hidden qualities. In other words, true likeness in the Buddhist context has to go beyond the ordinary external appearance. If Ding’s own confession of inadequacy in his inscription is more than the false modesty of an imperial subject, perhaps it is an acknowledgment of his struggle to reconcile these two levels of resemblance, which would help explain the strange, unsettling quality of Ding’s second painting. If, indeed, Ding’s second painting speciically portrays Qianlong, the impersonated divinity complicates what was originally an “imitation (fang 仿),” as it was called, of the sculptural image of Mañjuśrī at Shuxiang Si, and places it into the rank of Qianlong-as-bodhisattva paintings (Fig. 24).78 Among the best-known visual examples of Qianlong’s claims to bodhisattvahood, these paintings present the formal likeness of Qianlong’s Fig. 24. The Qianlong Emperor as Bodhisattva Mañjuśrī, Thangka, colors on cloth. The Freer Gallery of Art, Washington, D.C. Photograph provided by the Freer Gallery of Art. 160 ARCHIVES OF ASIAN ART face (based on a subdued modelling technique of the Jesuit painters) against a depiction of him in the Tibetan Buddhist iconographic guise of the Mañjughoṣa emperor at the center of a host of deities and teachers in a mandalic formation.79 Instead of this stark juxtaposition of two modes of representation, Ding’s painting conveys the emperor-as-bodhisattva identity through the subtle manipulation of a sacred icon.80 But the signiicance of classifying Ding’s second painting as a portrayal of Qianlong-as-bodhisattva lies beyond its employment of a different pictorial strategy: its very existence poses a challenge to the commonly perceived notion that Qianlong’s identiication with the Mañjuśrī was a project of self-fashioning that he performed primarily within the Indo-Tibetan esoteric Buddhist context.81 At least eight extant Tibetan thangkas featuring the likeness of Qianlong’s face that are found in Tibetan Buddhist inner sanctuaries of the court and at the court of the Dalai Lamas and Panchen Lamas in Tibet, together with references and addresses to Qianlong as the wheel-turning Mañjughoṣa emperor in Tibetan and Mongolian sources, have led scholars to regard the eficacy of Qianlong’s self-fashioning within a Tibetan Buddhist (and speciically Gelukpa) sectarian and courtly context.82 But this is subject to circular reasoning, as the assumption that Qianlong only accepted and promoted his bodhisattva-incarnate identity toward Mongols and Tibetans was built on a selective use of Tibetan Buddhist materials.83 The multiplicity of sources in Qianlong’s appropriation of Wutai Shan’s numinous icon, and its purpose in the establishment of a Manchu monastery, however, reveals a much more complex picture that goes beyond the appeasement of the empire’s particular ethnic constituencies: indeed, Qianlong’s selffashioning of himself as the Mañjuśrī of Wutai Shan was based on a seamless bringing together of multiple visual and devotional traditions, including Indian, Chinese, Tibetan, and Mongolian, under a single imperial domain around the emperor himself. If Ding’s revised second painting portrays Qianlong in the guise of a Tantric initiate, fashioned after Qianlong’s sketch of Wutai Shan’s celebrated Chinese Buddhist icon of Mañjuśrī, how was this translated onto the sculptural form? Even though we no longer have any evidence today as to what the Baoxiang Si sculptural group looked like, the exterior of the ruins of the main hall of Baoxiang Si (Fig. 14) offers further insights into the possible ritual or symbolic dimensions suggested above. The main hall itself, called Xuhua zhi ge 旭華之閣, was constructed as a beamless hall, featuring a square plan with ive arched openings on each of the four sides in the exterior, and probably a circular plan with a vaulted dome in the interior.84 Its majestic double friezes of glazed green-and-yellow tiles below the eaves, which can be seen from afar, still imparts a clear sense of architectural distinction. According to records from the Palace Workshops of the Imperial Household Department from 1750 that requested the dimensions of Wutai Shan’s own beamless hall, the brick barrel-arch construction of Baoxiang Si was inspired by none other than the aforementioned hall in Xiantong Si, though with one important difference: a square plan instead of the narrow rectangular one at Xiantong Si. Was Baoxiang Si a fuller realization of Baodi Si, in that it fulilled Qianlong’s wishes to re-create a beamless hall from Wutai Shan? How was this an improvement upon the original? As mentioned earlier, the Chinese term for “beamless” is a homophone of the word “immeasurable.” Beamless halls are thus often associated with the wish for longevity and are therefore appropriate for birthday celebrations. Baoxiang Si’s 1767 stele conirms this purpose, explaining that Qianlong’s primary intention for re-creating Shuxiang Si at nearby Xiangshan was to save his elderly mother from the toil of journeys to Wutai Shan, which is in Shanxi province 200 miles southwest of Beijing.85 There have been many such surrogate Wutai Shans in the long history of pilgrimages to that mountain, but perhaps none before that had been built for a single person.86 As an act of ilial piety toward the empress-dowager, this re-creation was effectively used as such, since Qianlong did not travel to Wutai Shan between his 1761 visit and 1781, four years after his mother had passed away. The use of Shuxiang Si for birthday celebrations, as expressed on the 1767 stele, suggests the use of the temple in a personal and familial context, which also contributed to a strengthened sense of Manchu imperial kinship and identity. However, in order to legitimize this re-creation, Qianlong launched into a lengthy explanation of the location of his newly created monastery in relation to Wutai Shan, here referred to by its alternate name, Qingliang (Clear and Cool). On speciically why this re-creation was both necessary and legitimate, the stele records: Mañjuśrī has long dwelled in this worldly realm, but has exclusively manifested and preached at Qingliang, or the Clear and Cool Mountains . . . Qingliang is located to the west of the capital, and Xiangshan is also to the west of the capital; in relation to Qingliang, Xiangshan is still positioned to its east; in relation to India, Qingliang and Xiangshan are both in the easterly direction. Therefore, how can one say these two mountains are not the same, or that they are different? . . . WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī Mañjuśrī can be seen with the rise and fall of phenomena; he manifests and transforms without limit . . . So, why would he insist on Qingliang as his ield of enlightenment, and not know that Xiangshan can also be? . . . In the past, we have paid obeisance to Mañjuśrī at Wutai to pray for [his] blessings. But Qingliang is more than a thousand li away from the capital. Being carried in an imperial carriage, I have only made it there three times. But Xiangshan is only thirty li away from the capital, so we can go year after year. Therefore with the aspiration for the lourishing of the Buddhist faith for ten thousand years from this point on, the temple at Xiangshan was initially built.87 Qianlong, repeatedly acknowledging that Mañjuśrī is unbounded by place and form, is paradoxically invested in locating and relocating the tangible material body that can best serve as a receptacle for Mañjuśrī. In re-creating the image of Mañjuśrī from Shuxiang Si, Qianlong sought to re-create the entire temple, and by extension, to replicate the entire mountain range of Wutai in Xiangshan, just outside the capital for the ease of frequent veneration. The authenticity of the image, as a synecdoche for Wutai Shan, rests on two seemingly contradictory claims: irst, Mañjuśrī is unconined by ixed notions of place and form; and, second, Mañjuśrī is rightfully in a speciic place (Xiangshan), and precisely in a speciic form (the image of Mañjuśrī at Baoxiang Si) because of its speciic location in relation to India and its status as a copy in relation to the original.88 It is precisely in the ambiguity caused by these two claims that Qianlong was able to derive his own identity as a Mañjuśrī-incarnate and Manchu Buddhist ruler. Carefully locating his court east of India, closer to Wutai Shan, and closer yet to replicas of them than the Dalai Lama in Lhasa, Qianlong asserted explicitly what had only been a tacit connection for previous Manchu rulers: that the successive Manchu emperors are the wheelturning incarnates of the bodhisattva Mañjuśrī. The inscriptions at Baoxiang Si do not elaborate on this association, nor do they mention the establishment of a Manchu Buddhist monastery. But the structure of the main hall itself hints at the possibility that both Qianlong’s rhetorical wordplay that justiied Baoxiang Si’s eficacy and his bodhisattva-incarnate status were not only pictorialized but also animated in architectural terms. If Ding Guanpeng achieves the emperor-as-bodhisattva portrayal of Qianlong by depicting a royal tantric initiate in the guise of Wutai Shan’s celebrated icon of Mañjuśrī, at Baoxiang Si, it is the architectural restaging of the sculpted image that imbues it with the same iden- 161 tity. The contemporaneous imperial gazetteer of Beijing, Qinding Rixia Jiuwen, speciies the structure of the main hall as square on the outside and round on the inside (waifang neiyuan 外方内圓).89 Designed to house the famed image of Mañjuśrī, from Wutai Shan, and therefore as the very structure that Berger suggests is referred to as a “mandala” in the Tibetan inscription90 Baoxiang Si’s main hall likely also evoked in form the structure of an Indo-Tibetan mandala. After all, none of the extant beamless halls from before or after Baoxiang Si is square in plan with symmetrical vaulted openings. If the main hall of Baoxiang Si were a simple copy of the beamless hall at Wutai Shan or elsewhere, the concentric square and circular plan would have been unnecessary. This intentional modiication of the original into a square/circular structure suggests a symbolic signiicance and/or ritual function beyond the usual apparent association of beamless halls with longevity and birthday celebrations. Representations of mandalas in the Indo-Tibetan tradition, which are idealized models of the cosmos with a principal power or deity residing at its center, are used for consecration rituals and meditative visualizations. Therefore, when a temple is designed after a mandala, it implies the establishment of a ground for consecration.91 Regardless of what the sculpted image looked like and whether the space indeed served ritual functions, the fact that a mandalic or mandala-like structure was built to house the sculpted replica of the Mañjuśrī on a lion traced by Qianlong’s hand underscores the symbolic potency of the newly re-created sculptural icon. Furthermore, as noted above, the Tibetan biography of Rölpé Dorjé recorded that the emperor engaged in many ritual initiations related to Mañjuśrī at Wutai Shan, and Qianlong famously occupied the central position as Mañjuśrī in aforementioned thangka paintings. The thangkas of Qianlong as Mañjuśrī were compositionally modelled after the refuge ield (Tib. tshogs zhing) paintings, which were often used as the basis of visualization practices for the devotion to one’s guru. Some scholars have even proposed that these paintings also served as tools for visualizing Qianlong as a deity in the course of meditative training practices.92 Finally, the Qianlong emperor built Yuhua Ge 雨花閣 (Pavilion of Rainy Flowers) in the imperial palace as an initiation hall in 1750 and Pule Si 普樂寺 (Temple of Universal Joy) in Chengde in 1766–67 as a mandala of the Buddhist deity Samvara. They were designed by Rölpé Dorjé, who gave Qianlong the tantric initiation into the mandala of Samvara back in 1745. Our knowledge of these various contemporaneous activities allows a certain degree of speculation about the little-known structure of 162 ARCHIVES OF ASIAN ART Baoxiang Si: when further evidence becomes available, it will not be surprising to discover that a three-dimensional, architectural mandala that would symbolically or ritually enhance and reinforce Qianlong’s Mañjuśrī status was also embedded in the design of Baoxiang Si’s main hall.93 In his initial attempts to establish Manchu Buddhism through the reiication of his bodhisattvahood, Qianlong’s stated emphasis was the re-creation of Mañjuśrī’s holy abode for reasons of ilial piety. But his more public and more personal agendas were both increasingly made known in his subsequent projects. Shuxiang Si at Chengde: Copying the Copy In 1774, Qianlong began building a Manchu Buddhist monastery at Chengde, which he named after the original Shuxiang Si at Wutai Shan (Fig. 25). Completed in just one year, this architectural replica was designed from the beginning to facilitate the translation of the Manchu Buddhist scriptural canon, a monumental project that had commenced the year before.94 Its main hall, Huicheng Dian 會乘殿, was also designed to house a copy of the Manchu canon that was inished in 1790.95 The Manchu canon, produced through translation from and consultation with existing Chinese-, Tibetan-, and Mongolian-language versions, was in a sense a linguistic parallel to the Wutai Shan replicas. Though bearing the name and look of the Tibetan Kangyur, the Manchu canon was in fact an entirely new compilation based on a synthesis of Chinese, Tibetan, and Mongolian canons while following the structure of the Chinese Buddhist canon, the Tripiṭaka (Three Baskets of Teachings).96 Likewise, it was through the close juxtaposition of Chinese and Tibetan iconographic and scriptural traditions, architectural styles, and ritual lexicons that a distinctly Manchu Buddhist culture (with mandatory Manchulanguage recitation) was created. Qianlong, on his 1775 stele inscription commemorating the completion of the monastery, explains that although the image of Mañjuśrī was to be made in the same way as the image from Baoxiang Si, the halls and pavilions (diantang louge 殿堂樓閣) were “roughly” based on the original one at Wutai Shan—which indicates that not only was Qianlong explicitly aware of the difference between Baoxiang Si at Fig. 25. Huicheng Dian, Shuxiang Si, Chengde. Photograph by the author. WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī Xiangshan and Shuxiang Si at Wutai Shan, but also that his decision for copying was a calculated one.97 Assuming that the Baoxiang Si sculpture of Mañjuśrī on a lion was, through the above-mentioned process of replication, transformed into an image of Qianlong-as-bodhisattva, the choice of modelling a new image after the copy at Baoxiang Si must be read as a way to perpetuate that identity. The Shuxiang Si replica at Chengde was built on the northern slopes beyond the Summer Palace on the western side of Putuo Zongcheng Miao 普陀宗乘廟 (Qianlong’s re-creation of the Potala Palace, erected in 1771 as part of a birthday present to his eighty-year-old mother). The monastic complex of the Shuxiang Si replica follows the central plan of a Han-Chinese monastery: the gate, the protectors’ chapel (Tianwang Dian 天王殿), and a main prayer hall are laid out on a central axis, with chapels and monks’ quarters on both sides (Fig. 26). The third building on the main axis, which is the main hall of the complex, is set at the top of a series of steps on a gently sloping hill. Comparing the layout of this Shuxiang Si with gazetteer depictions of Shuxiang Si at Wutai Shan, some have argued that it is indeed closely based on the original.98 In fact, the layout of the Shuxiang Si replica is no different than any centrally planned Chinese temple. The conscious adoption of a Han-Chinese temple plan for the building of a Manchu monastery that nonetheless follows the ritual protocols of a Tibetan Gelukpa monastery would have appeared conspicuous in light of the two Tibetan replicas that Qianlong built on that same hill before and after he built Shuxiang Si, namely Putuo Zongcheng Miao and Xumi Fushou zhi Miao 須彌福壽之廟, modelled after Tashi Lhunpo in 1780. In her study of Chengde, Anne Chayet speculated that Wutai Shan was perhaps irst and foremost “a Chinese sacred place” for Qianlong, and therefore he decided that his evocation of Wutai Shan in Chengde “had to be purely Chinese.”99 Chayet’s explanation overlooks the various ways in which cultures and traditions have been simultaneously evoked and juxtaposed in Qianlong’s series of replications. An eighteenthcentury map of Chengde (Fig. 27) shows that faux-Tibetan style buildings and stupas (similar to the blind walls with small ornamental windows of the Putuo Zongcheng Miao), were also built off to the side of the central axis. In terms of function, the main hall of Shuxiang Si at Chengde also adapted and incorporated different designs. Huicheng Dian measures seven bays wide and ive bays deep, and is designed as a prayer and gathering hall, with images at the far end (Figs. 28 and 29). The layout of the main prayer hall allows for a lexible use of space, with enough depth to accommodate both the 163 ritual needs of a Tibetan Buddhist assembly hall (’du khang), where prayer gatherings are held (in this case by resident Manchu lamas), and that of a Chinese-style hall, in which images usually occupy the central space. The main hall of the original Shuxiang Si at Wutai Shan is much smaller, measuring ive bays wide and three bays deep—just enough space to house the central image.100 Therefore, despite the repeated rhetoric that Qianlong’s Baoxiang Si and Shuxiang Si are close replicas of the halls and pavilions of the original Shuxiang Si at Wutai Shan, the actual architectural designs of Shuxiang Si’s buildings at Chengde are different. The planning and design of these temples speak much more to their ritual and symbolic purpose as mandalic architecture, in the case of Baoxiang Si, and their practical function as a place of monastic assembly, in the case of the new Shuxiang Si. What distinguished Shuxiang Si at Chengde was the miraculous image, not the temple complex; copying the architecture exactly was hardly necessary when Shuxiang Si can be referenced by a replication of the true image and by the imperial authority invested in Qianlong’s stele inscriptions. The image that was modelled after the Mañjuśrī of Baoxiang Si is housed in an octagonal pavilion called Baoxiang Ge 寶相閣 (Precious Form Pavilion) atop a hill behind Shuxiang Si’s main building complex.101 An artiicial mountain landscape (jiashan 假山) with grottoes and meandering passages leads up to Baoxiang Ge (Fig. 30).102 The entire garden landscape is reminiscent of those found at Qing imperial gardens, while the miniature mountain landscape evokes the Wutai Shan range. Even though the pavilion and the original image are no longer extant, early photographs allow us to compare this replica of a replica with the original image at Shuxiang Si and with Ding Guanpeng’s paintings (see Figs. 16 and 17).103 The Chengde Mañjuśrī is in almost exactly the same position as the igure in Ding Guanpeng’s earlier painting (Fig. 31): the bodhisattva sits in a frontal position with his right knee pointing outward and foot tucked around the nape of the lion’s head, which is turned upward to the right; the lion’s feet, stubbier than the originals at Wutai Shan (due to the transfer from a three-dimensional image to a two-dimensional one and back), are also planted on lotus blossoms. Even the low of the bodhisattva’s garb and locks of hair follow the same contour. What is added are attendant igures beside the bodhisattva, which suggests that they were not part of what would have been copied from the array of original sources. We can therefore deduce that the sculpture was a rather careful three-dimensional replica of the replica at Baoxiang Si, of the two-dimensional replica Fig. 26. Idealized Plan of Shuxiang Si. Copyright The J. Paul Getty Trust, 2009. All rights reserved. WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī 165 Fig. 27. Detail of Shuxiang Si, Map of Palaces at Jehol, 18th century. Colored, mounted on silk scroll, 122 × 226 cm. Original map and image in public domain; digital image provided by the Geography and Map Division, Library of Congress. by Ding Guanpeng, of the sketch by Qianlong, and of the original image. The imitation was not just a reproduction in name but also in a formal, material technique designed to transfer, over and over again, the true likeness of Qianlong-as-Mañjuśrī / Mañjuśrī-as-Qianlong, with each new copy reinforcing and enhancing the notion of the true form. As the deining foci of the recently instituted Manchu monasteries, these imperially mediated copies modelled after Wutai Shan’s numinous icon positioned Qianlong at the center of a newly established tradition that nevertheless traces itself back to one of Buddhism’s most illustrious bodhisattvas and his earthly realm. It should come as no surprise then that travellers to Shuxiang Si at Chengde noted the similarity between the face of Mañjuśrī at Baoxiang Ge and that of the Qianlong emperor, despite the fact that a physiognomic afinity is not apparent in available photographs of the image.104 As suggested in Ding Guanpeng’s second painting, the way in which Qianlong asserted his bodhisattva identity revealed a form of likeness that is deined through the concept of “true trace” and the visual lexicons of a royal tantric initiate, rather than through the more familiar technique of modiied chiaroscuro introduced by and demanded of the Jesuits in the Qing court. Qianlong’s stele inscriptions conirmed his increasing interest in advancing his bodhisattva identity for the promotion of Manchu Buddhism. Whereas the earlier inscriptions at Xiangshan, from 1767, stressed ilial piety, Qianlong himself proclaimed for the irst time in Shuxiang Si’s commemorative stele in 1775 the urgent need for Manchu translations of Buddhist scriptures, and for those who would study and recite them in order to spread the teachings of the Buddha. Importantly, 166 ARCHIVES OF ASIAN ART Fig. 28. Main Images of Huicheng Dian, Shuxiang Si, Chengde, ca. 1933. Copyright The J. Paul Getty Trust, 2009. All rights reserved. Fig. 29. Interior of Huicheng Dian, Shuxiang Si, Chengde. Copyright The J. Paul Getty Trust, 2009. All rights reserved. among the twelve sets of Manchu canon that were carved, many were distributed to non-Manchu monasteries, including the Potala Palace of the Dalai Lamas in Lhasa, Tashi Lhunpo of the Panchen Lamas in Shigatse, and Yonghe Gong in Beijing. 105 Their placement in key monastic establishments of the empire where no one would be able to read them suggests the performative aspect of this endeavor—that the emperor has produced a true and perfected version of the scriptural canon. Following his remarks on the propagation of the Manchu canon, he asked, “The Tibetan lamas call me an emanation of Mañjuśrī based on the near homophone of ‘Manchu’ and ‘Manju,’ but if it were really true that our names correspond to the reality, wouldn’t Mañjuśrī laugh at me for that?” Although the rhetorical question implies Qianlong’s ambivalence toward this gift of honor, expressed at least in the Chinese language, a year later, he wrote the following on another tablet at Shuxiang Si: The image of Wenshu 文殊 [Mañjuśrī] is nothing shu 殊 [extraordinary]. It’s magniicent as is. The two peaks [behind Shuxiang Si and behind the Potala] stand side by side, not more than half a li away from each other. His dharma body can manifest as a young boy, or as a tall gentleman. The vermillion edict [from the Dalai Lama] has been overly enthusiastic WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī Fig. 30. Exterior of Baoxiang Ge, Shuxiang Si, Chengde, ca. 1933. From Sekino Tadashi and Takuichi Takeshima, Jehol: The Most Glorious and Monumental Relics in Manchoukuo (Tokyo: The Zauho Press, 1934), Vol. 4, page 11. in its praise [of me as a Mañjughosa emperor]. Wouldn’t it be laughable if it were true?106 This refrain at Shuxiang Si, which would have been seen only by close members of the court, is Qianlong’s closest written acknowledgment of himself as an emanation of Mañjuśrī. It also made apparent that this selfidentiication was deined vis-à-vis Avalokitêśvara (the bodhisattva of great compassion). Directly adjacent to Shuxiang Si is Putuo Zongcheng Miao, which had been built just a few years earlier (in 1771) in homage to the Dalai Lamas, successive incarnations of whom are considered emanations of Avalokitêśvara. Because Shuxiang Si was known to house objects from Qianlong’s childhood, and was popularly referred to as Qianlong’s “family shrine,” and because of the two steles there that bear Qianlong’s own repeated suggestions of his association with Mañjuśrī, the temple would have been seen as the very embodiment of a Manchu Imperial Buddhist identity founded on Qianlong’s connection with Wutai Shan, and as a direct counterpart to the seat of the Dalai 167 Fig. 31. Interior of Baoxiang Ge, Shuxiang Si, Chengde. From Sekino Tadashi and Takuichi Takeshima, Jehol: The Most Glorious and Monumental Relics in Manchoukuo (Tokyo: The Zauho Press, 1934), Vol. 4, page 14. Lama as an emanation of Avalokitêśvara.107 Replicating the Shuxiang Si image was surely a way to unite a celebrated image that encapsulated Wutai Shan’s numinous history with Mañjuśrī’s other manifestation as Qianlong himself. Lives of an Image Consider, for a moment, the chain of transformations of that celebrated image (Fig. 32). Based on the miraculous tale of the eighth-century sculpted image of Mañjuśrī in Zhenrong Yuan (the temple later renamed Pusa Ding), a similar tale was established to account for the origin of the ifteenth-century image of Mañjuśrī in Shuxiang Si, also at Wutai Shan. The Qianlong emperor, soon after his pilgrimage to Wutai Shan in 1761, made a sketch based on the sculptural image of Mañjuśrī in Shuxiang Si, which was then transferred, in accordance with Qianlong’s instructions, onto a stone stele. That same year, he also commissioned court painter Ding Guanpeng to make a large painting from his sketch, which 168 ARCHIVES OF ASIAN ART Fig. 32. Diagram showing chain of replicas. was enlarged to about a third the size of the original sketch, and several other paintings of Mañjuśrī; a wax model and a textile of the image were also made in conjunction with the paintings. Subsequent copies were based on all earlier models. The stone stele was erected in front of Baoxiang Si in Xiangshan, completed in 1767, which housed a replica of the Mañjuśrī image that was based on three sources: the sketch, the stele, and Ding Guanpeng’s paintings. This Baoxiang Si copy of the Shuxiang Si image subsequently became a source for a further copy, enshrined in an octagonal pavilion at Shuxiang Si that was named Baoxiang Ge, a sculptural image that can be traced to the earlier of the two paintings of Mañjuśrī by Ding Guanpeng. This secondary copy became the namesake of Shuxiang Si in Chengde, the monastery that houses Baoxiang Ge. This chain of copies, as well as the earlier acts of recreating Wutai Shan in Baodi Si, suggests a luid relationship between copy and original: each copy in its speciic form and medium takes on a life of its own, and the process of replication makes something more true, and thus creates something new (in this case, a Manchu imperial identity). Through a variety of generative acts of copying—whether the repairing and re-creation of sets of ritual objects, the mapping and planning of ar- chitectural spaces, the insertion of the imperial brush trace in the sketch of the sculptural image, the commitment of the copied form to the authoritative (and longlasting) surface of a stone stele, the creative revision of its painted versions, or a repeated rhetorical act of achieving geographical equivalence—the Qianlong emperor enacted his identity as a Mañjuśrī-incarnate. The various two-dimensional and three-dimensional media, continuously imitating and informing subsequent replicas, collectively produced a lineage that not only re-created the pure land of Wutai Shan closer to the capital but also enhanced, perfected, and resituated it around the Manchu ruler himself. Replication, Translation, and the New Geography of Manchu Imperial Buddhism The conscious alignment of Manchu imperial identity with Wutai Shan’s sacred history and power puts into perspective Qianlong’s subsequent activities in connection with the mountain range, such as the translation of Mañjuśrī-related texts into Manchu and a new edition of the Wutai Shan gazetteer. On his 1781 trip to Wutai Shan, the fourth of his six pilgrimages there, Qianlong copied Dasheng wenshu shili pusa zanfo fashen li WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī 大聖文殊師利菩薩讚佛法身禮 (The Great Sage Mañjuśrī Bodhisattva’s Praise of the Dharma Body of the Buddha Liturgy),108 and translated it into Manchu. Rölpé Dorjé is said to have selected this text from the Chinese Tripiṭaka, Chinese being the only language in which the text survived.109 The text was brought back to the capital, and in addition to its Manchu translation (Amba enduringge nesuken horonggo fusa. fucihi i nomun i beye de doroloho maktacun), it was later translated into Tibetan (Byang chub sems dpa’i ’jam dpal dbyangs kyis sangs rgyas kyi sku la bstod pa) and Mongolian and incorporated into a quadrilingual edition.110 Qianlong, in a praise poem that he wrote while visiting Baoxiang Si in 1782, commented on his own translation of the text into Manchu, and on his order that the printing house produce “gold-lettered quadrilingual editions”(jinshu siti 金書四體) to be offered on Wutai Shan’s ive peaks as well as at Baoxiang Si in Xiangshan.111 In reality, many other copies were made, and their circulation was not limited to Wutai Shan and Xiangshan.112 There are also single-language translations of the text in Chinese, Mongolian, or Manchu. That this particular translation was carried out at Wutai Shan and by Qianlong himself suggests that the project’s primary importance lay in the Manchu emperor’s authority in reproducing and disseminating a previously untranslated text on Mañjuśrī. Qianlong’s virtuoso act of translation makes clear that, in this case as well as in the case of the monumental task of compiling a Manchu canon, what mattered more was not whether the texts were used and consulted for generations to come, but Qianlong’s performance of translation. As the Manchu incarnation of the Bodhisattva Mañjuśrī making a pilgrimage to his sacred abode, Qianlong asserted his own agency in translating a scriptural homage to Mañjuśrī and disseminating it throughout the key Buddhist locations of his empire. Even with a very limited audience, the ultimate aim of Qianlong’s gesture was, as Pamela Crossley argues, “to make all true expression, in any language, the property of the emperor.”113 By doing so, he not only declared his authority in the making of Manchu Buddhism, but also linked himself to Buddhism’s Indic origins. As he noted in the preface to his translations, this “Homage to Mañjuśrī” had never before been available in the languages of his Tibetan, Mongolian, or Manchu constituents. Qianlong was thus the irst to bring them this text—which was originally translated from the Sanskrit by Amoghavajra (705– 774)—and thereby connect himself to early translators who were responsible for the transmission of Buddhism to China. As a site that was from the beginning created to transplant Buddhist India to China, Wutai Shan itself became a source for translation and transplantation. 169 Qianlong’s inal major effort to seal the connection between himself and Mañjuśrī, and between Wutai Shan and the capital, reached a much wider audience than did his previous endeavors. The project began with Qianlong’s province-wide coniscation of all Wutai Shan gazetteers and their printing blocks in order to control the proliferation of “erroneous” information.114 The motivation for this order was undoubtedly to maintain control over the history of the mountain range, and moreover, to make canonical his connection to it, much like Qianlong’s other projects of compiling Buddhist iconographic scriptural and literary canons, and catalogues of objects in his collection. That the Qing court took such a step to curtail the popular circulation of such publications also conirmed their popularity among tourists, pilgrims, and the like. Subsequently, Qianlong issued his own edition of the mountain gazetteer, the Imperial Record of the Clear and Cool Mountains (Qinding Qingliang shan zhi 欽定清凉山志) in 1785 (reprinted in 1811). While more than twice the length and the number of volumes (juan) than the previous edition of the Wutai Shan gazetteer prefaced by the Kangxi Emperor, this new text reduced and eliminated much of the history of Wutai Shan to make room for lengthy descriptions of imperial restorations, steles, and Qianlong’s other writings about the mountain range. The new guidebook took on the perspective of one pilgrim—the emperor himself—which presented the mountain range as exclusively imperial.115 It solidiied Qianlong’s connection to the site, not only through suggestions of his bodhisattva identity, but also by publicizing his activities as one of the most devoted imperial sponsors. Qianlong’s heavy-handed editing, revision, and translation of Wutai Shan’s history serve as a perfect textual parallel to the series of replication projects explored in this essay. Seen as part of the micro-universes that Qianlong created at Xiangshan and Chengde, the Wutai Shan replicas anchored the Manchu imperial identity within an India-centered cosmography; everything east of India was considered the domain of Mañjuśrī and therefore of Qianlong. The replicas that derived their power from the true image of the bodhisattva emperor functioned not only symbolically, but also as the fundamental basis for the initiation of Manchu imperial Buddhist monasticism. The retracing of the steps of replication shows how Qianlong, by combining various architectural, artistic, ritual, conceptual, and semantic evocations of Mañjuśrī at Wutai Shan, created what he saw as a perfected and universal form of Buddhist teaching and practice around himself as the universal emperor, and in so doing re-created a more perfectly Manchu imperial Wutai Shan. 170 ARCHIVES OF ASIAN ART Rethinking Universal Emperorship Having conquered China from outside the Great Wall in northeast Asia, the Manchu rulers carefully crafted a multifaceted imperial persona that was absolutely central to their governance of an expanding and increasingly diverse empire. As many scholars of Qing history and religion have shown, the Manchu rulers’ statecraft depended heavily on a retelling of their origins and identity, as well as those of the peoples over whom they sought to rule, projecting themselves as “the ultimate apotheosis of righteous rulers in the recurring cycles of history and myth.”116 It was under the Qianlong emperor that the Qing empire reached its greatest territorial extent and height of power and prosperity. As the fourth Manchu emperor to rule from China proper, Qianlong inherited the identity-making enterprise from his forebears, yet a very different reality from each of them.117 Qianlong’s incarnation of the wheel-turning Mañjuśrī, alongside his zealous cultivation of an imperial Confucian persona, both of which matured through his long reign of sixty years (1736–1795), attests to his ability to embody the moral centers of all cultural and religious traditions under his domain and allowed him to recenter his imperium upon himself. Just as Qianlong rehearsed the early Qing ruler’s reenactment of the lama–patron relationship of the Yuan Mongols and the religious leaders of the Sakya sect of Tibetan Buddhism and re-created the palaces of the Dalai Lamas and Panchen Lamas in Chengde, his act of copying Wutai Shan recentered, reorganized, and reconigured the past, such that his reenactments and reappropriations produced a new imperial cosmology. Qianlong’s replicas were eventually achieved through his reenactment of an embodiment of a sacred icon that was highly venerated across North, East and Central Asia. His merging of himself with Wutai Shan’s most celebrated icon, which had ties to Chinese, Mongolian, Tibetan, and even Central Asian iconography, allowed the message he sought to convey to transcend all religious, cultural, and linguistic differences. Of paradoxical importance is that neither the images of Mañjuśrī on a lion Qianlong commissioned and received, nor the temples he built to enshrine the images, were accessible to the public, or for that matter put on display for his multicultural subjects. Moreover, the Manchu monasteries’ institutional and architectural ephemerality meant an even smaller audience over time. A court record from the end of the thirty-fourth year of the Qianlong reign (1770), twenty years after the initial construction of Baodi Si—the irst Wutai Shan replica at Xiangshan—reported that the temple complex was in urgent need of repair.118 Overall, Qianlong’s monumen- tal efforts at shaping and preserving a distinct Manchu imperial Buddhist monastic and scriptural heritage was not sustained as imperial support of monasteries at Wutai Shan waned in the latter half of the Qing dynasty.119 By the early twentieth century, with the collapse of most of the ediices at Xiangshan, the Manchu monasteries fell into obscurity. But even this history of demise is instructive. Instead of reading it as evidence of Qianlong’s failed attempt to create a lasting impact, or attribute the short-lived institution to the inevitability of Sinicization, I argue that Qianlong’s aims were elsewhere. The sophistication of these building projects showed that having undertaken the tasks (of perfecting the teachings and practices of Buddhism in the form of Manchu Buddhism) in his role as an emperor mattered more to him than the monasteries’ projected longevity within a historical timeframe. Far from serving as instruments of political or religious propaganda, Qianlong’s copies of Wutai Shan display the expansive temporality of a universal, wheel-turning Sino-Tibetan bodhisattva emperor, one whose political, religious, cultural, and artistic engagements were as much about the instrumental governance of his empire as they were aimed at the manifestation of an ideal, universal kingship, a role that Qianlong fully identiied with throughout his long reign. A careful study of the creation of the Manchu monasteries has allowed us to reconstruct the worldview of the eighteenth-century ruler on his own terms— a cosmology in which religion and politics were not separate categories. Regardless, the imperial promotion of Wutai Shan was to have long-lasting consequences on religious culture in the Qing empire: it played an important role in initiating a thriving Sino-Tibetan Buddhist pilgrimage culture at Wutai Shan, supported by visiting Mongols and Tibetans in the nineteenth century (despite the lack of imperial support), and laid the groundwork for the development of Wutai Shan as a center of Qing Gelukpa Buddhist scholasticism and a site of Tibetan Buddhist hagiographical traditions. The articulation of a Manchu imperial Wutai Shan, which synthesized the past and present in Chinese, Mongolian, and Tibetan imaginations, set the stage for Tibetan Buddhism to lourish on the mountain. Wen-shing Chou is assistant professor of art history at Hunter College. Her articles on maps and wall paintings of Buddhist sacred sites have appeared in the Art Bulletin, the Journal of Asian Studies, and the Journal of the International Association of Tibetan Studies. She is currently completing a book on the transcultural pilgrimage site of Wutai Shan in late imperial China. [chouwenshing@gmail.com] WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī Acknowledgments Parts of this essay were irst presented at the annual conferences of the Association for Asian Studies (2012) and American Academy of Religion (2014). I thank the participants of both panels for their questions and suggestions. I am especially grateful for insights and help from Patricia Berger, Isabelle Charleux, Kevin Greenwood, Johann Elverskog, Ellen Huang, Li Jianhong, Lin ShihHsuan, Wei-cheng Lin, Nancy Lin, Christian Luczanits, William Ma, Wang Ching-Ling, Wen Wei, and Yang Hongjiao. I also thank Tara Zanardi, Lynda Klich, Stephen Frankel, Stanley Abe, and the anonymous readers for their astute editorial comments and suggestions. The inal version of this essay was completed at the Institute for Advanced Study, Princeton. Notes 1. Imperial patronage at Wutai Shan is a topic that has been well studied and documented in the mountain’s own mytho-historiography. Emperors of the Northern Wei dynasty (385–534), Northern Qi dynasty (550–577), and Sui dynasty (581–618) all erected temples at Wutai Shan. Rulers during the Tang dynasty (618–907), whose ancestral home is located in the vicinity of Wutai Shan in the Taiyuan 太原 region, were especially committed to advancing Mañjuśrī as the protector of the imperial clan and of the entire nation. A special temple of Mañjuśrī was built in Taiyuan at the suggestion of Buddhist translator and tantric master Amoghavajra (705–774). See Raoul Birnbaum, Studies on the Mysteries of Manjusri (Boulder, CO: Society for the Study of Chinese Religions, 1983), 32; Stanley Weinstein, Buddhism Under the T’ang (Cambridge: Cambridge University Press, 1987), 83. The mytho-historiography of the mountain’s origins traces its history of imperial connection back even further. According to a sixteenthcentury gazetteer compiled by Monk Zhencheng 鎮澄 (1546–1617), Mañjuśrī had irst come to Wutai Shan (from India) to convert King Mu (r. 1001–947 BCE) during the Zhou dynasty (1046–256 BCE), and the deity’s presence was again recognized during the Han dynasty (206 BCE– 220 CE) through the clairvoyance of two Indian monks, Kaśyapa Matanga and Dharmaratna, who travelled to China after Emperor Ming (r. 58–75) had a dream about a radiant golden igure. See Zhencheng, Qingliang Shan zhi (Record of the Clear and Cool Mountains), juan 1, reprinted in Zhongguo Fosi shi zhi huikan 中國佛寺史志彙刊, 2nd series, vol. 29 (Taibei: Minwen shuju, 1980–85), 17, 97–98, 206. Imperial patronage of temples at Wutai Shan continued through the Song, Yuan, and Ming dynasties, and was well documented in Ming gazetteers, but it was not until the Qing dynasty that its Manchu emperors embraced the project of building Wutai Shan with unprecedented fervor. 171 2. See Pamela Crossley, Translucent Mirror: History and Identity in Qing Imperial Ideology (Berkeley: University of California Press, 1999), 233. 3. David Farquhar’s 1978 study was the irst to draw attention to the Qing imperial identiication with Mañjuśrī. According to Farquhar, the Qing emperors’ self-fashioning represented a blending of the Tibetan Buddhist “theory of bodhisattva metempsychosis” in identiiable individuals, especially rulers who spread the Buddhist teachings, and the Chinese Buddhist understanding of Mañjuśrī’s residence at Wutai Shan. See David Farquhar, “Emperor as Bodhisattva in the Governance of the Ch’ing Empire,” Harvard Journal of Asiatic Studies, 38, no. 1 (1978): 15. 4. Monguors refer to a group of Mongols who had settled in the northeast Tibetan highlands during the Mongol Yuan dynasty. During the Qing dynasty, many religious authorities from this group played an important role in mediating the relationship between the Qing court and Central Tibet. 5. Mark Elliott, Emperor Qianlong: Son of Heaven, Man of the World (New York: Longman, 2009), 53. 6. See Nicola Di Cosmo, “Manchu Shamanic Ceremonies at the Qing Court,” in State and Court Ritual in China, ed. Joseph P. McDermott (Cambridge: Cambridge University Press, 1999), 352–98. Di Cosmo observes two separate strands of developments of Manchu ritual and religion: a political and ideological one that resulted in the veneration of Wutai Shan and the emperors’ association with Mañjuśrī, and a social phenomenon rooted in Manchu religious cults that led to the incorporation of Buddhist and native Chinese deities into Manchu shamanistic rituals. See Di Cosmo, “Manchu Shamanic Ceremonies,” 375. Manchus were in fact forbidden to take monastic vows until the reign of the Shunzhi 順治 emperor (1638–1661). 7. For a study of the organization of the Imperial Household Department, see Preston M. Torbert, The Ch’ing Imperial Household Department: A Study of Its Organization and Principal Functions, 1662–1796 (Cambridge, MA: Harvard University Asian Center, 1977). 8. Booi were often inexactly described as bondservants or slaves. For more on the origins and deinitions of booi, see Mark Elliott, The Manchu Way: The Eight Banners and Ethnic Identity in Late Imperial China (Stanford, CA: Stanford University Press, 2001), 81–84. 9. Ibid., 83. 10. That is, staging Mañjuśrī’s reappearance in a vision. 11. For a recent volume that explores the making of ethnicity in the Qing, see ed. Pamela Crossley et al., Empire at the Margins: Culture, Ethnicity, and Frontier in Early Modern China (Berkeley: University of California Press, 2006); especially Mark Elliot’s contribution “Ethnicity in the Qing Banners,” 27–57. 12. Patricia Ann Berger, Empire of Emptiness: Buddhist Art and Political Authority in Qing China (Honolulu: University of Hawai‘i Press, 2003), 161–64. 13. Ibid., 126–27. 172 ARCHIVES OF ASIAN ART 14. See, for example, Maria Loh, Titian Remade: Repetition and the Transformation of Early Modern Italian Art (Los Angeles: Getty Research Institute, 2007). Loh utilizes Deleuze’s model of a rhizome, among others, to reconstruct the interdependent relationship between the original and the replica, between the imitator and the original author. 15. See Christopher Wood, Forgery, Replica, Fiction: Temporalities of German Renaissance Art (Chicago: University of Chicago Press, 2008). For Wood, the processes of substitution and replication are sites for the revelation of “deep structure of thinking about artifacts and time.” Even more fundamentally, Whitney Davis argues that to describe the process or dynamics in replication is to describe cognition, consciousness, and therefore culture itself. See Whitney Davis, Replications: Archaeology, Art History, Psychoanalysis (University Park: Penn State University Press, 1996), 4. 16. See Thu’u bkwan, Lcang skya rol pa’i rdo rje’i rnam thar (Biography of Chankya Rölpé Dorjé) (Lanzhou: Kan su’u mi rig s dpe skrun khang, 1989), 332. Thu’u bkwan does not mention an exact date for this event between the 1740s and 1750s, but the timing of the institution of the irst Manchu Buddhist monastery is corroborated both in the archives of the Grand Council (Junjichu 軍機處) and in Damcho Gyatsho Dharmatāla, Rosary of White Lotuses: Being the Clear Account of How the Precious Teaching of Buddha Appeared and Spread in the Great Hor Country, trans. and annotated by Piotr Klafkowski; supervised by Nyalo Trulku Jampa Kelzang Rinpoche (Wiesbaden: O. Harrassowitz, 1987), 320–21. See Beijing Number One Archive, documents no. 03-182-2218-15 and no. 03-182-2218-14, which were originally a single ile. I thank Lin Shih-Hsuan for helping me piece this together. Chen Qingying’s attribution of this event to after 1761 is likely erroneous. See Chen Qingying, “Zhangjia Ruobi duoji nianpu (II)章嘉若必多吉年谱（二） (Chronology of Rölpé Dorjé)” Qinghai minzu yanjiu 2 (1990): 37. 17. Many secondary sources indicate that Rölpé Dorjé entered retreat consecutively at Wutai Shan beginning in 1750, but from his biography it is obvious that he did not go there during a two-year trip to Tibet in search of the Seventh Dalai Lama’s reincarnation between 1757 and 1758. 18. See Huang Hao 黄颢, Zai Beijing de zangzu wenwu 在北京的藏族文物 (Tibetan cultural materials in Beijing) (Beijing: Minzu chubanshe, 1993), 85. Huang speculates that this monastery was Xiangjie Si, but court and gazetteer records and Thu’u bkwan’s biography Lcang skya rol pa’i rdo rje’i rnam thar, indicate Baodi Si to be the monastery in question. 19. Dou Guangnai 竇光鼐 (1720–1795), ed., Qinding rixia jiuwen kao 欽定日下舊聞考 (Imperial Edition of legends of old about the capital) (Beijing: Wuying dian, 1774), juan 103, 7. Wang Jiapeng noted the number of lamas at each Manchu monastery recorded in court documents, but did not provide speciic sources; see Wang Jiapeng, “Qianlong yu Manzu lama siyuan” 乾隆與滿族喇嘛寺院 (Qian- long and Manchu Tibetan Buddhist Monasteries), Gugong bowuyuan yuankan 1 (1995): 60. Beginning with Baodi Si, as many as thirteen Manchu monasteries were built in Beijing, Shengjing, Chengde, and the Western and Eastern Mausoleums. See Lin Shih-Hsuan 林士鉉, Qingdai menggu yu manzhou zhengzhi wenhua 清代蒙古與滿洲政治文化 (Mongolia and the Political Culture of the Manchus in the Qing Dynasty) (Kaohsiung: Fuwen, 2009), 136–38. 20. Awang Pingcuo 阿旺平措,“Qingdai Zangchuan Fojiao zai neidi de chuanbo yu yingxiang 清代藏传佛教在内地的传播与影响 (The spread and inluence of Tibetan Buddhism in China during the Qing),” Fayin (2012), accessed September 5, 2014, http://www.fayin.org/luntanjingpin/2012/0823/425.html. 21. There exist a series of such histories of Wutai Shan, each of them imperially endorsed to a certain degree; each one is a new compilation by a new editor (or editors), with illustrations produced from a new carving of woodblocks, and through the sponsorship of a new patron (or patrons), based on a selective use of the sources available at the time. A later history of Wutai Shan is mentioned later in this essay (Imperial Record of the Clear and Cool Mountains, published in 1785). 22. The various recensions of this story in connection with the sacred icon have been the focus of a number of art-historical studies. See Sun-ah Choi, “Quest for the True Visage: Sacred Images in Medieval Chinese Buddhist Art and the Concept of Zhen” (PhD diss., University of Chicago, 2012), 164–74, and Wei-cheng Lin, Building a Sacred Mountain: The Buddhist Architecture of China’s Mount Wutai (Seattle and London: University of Washington Press, 2014), 89–98. See page 23 and note 65 for Mongolian and Tibetan recensions of the story. 23. For a summary of the history of the Tibetan presence at Wutai Shan, see Karl Debreczeny, “Wutai Shan: Pilgrimage to Five-Peak Mountain,” Journal of the International Association of Tibetan Studies 6 (December 2011): 30–39, accessed September 6, 2014, http://www .thlib.org?tid = T5714. 24. This process of so-called conversion is one that requires further investigation. Even though most secondary sources speak of the ten monasteries that the Shunzhi emperor converted from Chinese Buddhist to Tibetan Buddhist temples—as Köhle pointed out in her 2008 article “Why Did the Kangxi Emperor Go to Wutai Shan?: Patronage, Pilgrimage, and the Place of Tibetan Buddhism at the Early Qing Court,” Late Imperial China 29, no. 1 (June 2008), 73– 119—none of the literature that makes this statement cites a primary source, and this process of conversion was probably a more gradual process, where the Chinese, Tibetan, and Mongolian traditions coexisted within these institutions. In Dharmatala’s Rosary of White Lotuses, the conversion is discussed in straightforward terms; it includes a description of the Shunzhi and Kangxi emperors’ construction of large imperial temples on each of the ive terraces. See Damcho Gyatsho Dharmatāla, Rosary of White Lotuses, 418–19. WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī 25. Originally used to describe a series of laws laid down by Chinggis Khan (1162?–1227), the Mongol term jasagh was subsequently used among the Manchus to denote a status of military and administrative rule, and “jasagh lamas” was used to describe high-ranking imperially appointed lama oficials. See Dorothea Heuschert, “Legal Pluralism in the Qing Empire: Manchu Legislation for the Mongols,” The International History Review 20, no. 2 (June 1998): 310–24. For more on Qing administrative documents concerning imperial sponsorship of jasagh lamas, see Vladimir Uspensky, “The Legislation Relation to the Tibetan Buddhist Establishments,” paper given at the “Wutai Shan and Qing Culture” conference at the Rubin Museum of Art, 2007. See also Jagchid Sechin (Zhaqi Siqin 札奇斯欽), “Manzhou tongzhi xia menggu shenquan fengjian zhidu de jianli” 满洲统治下蒙古神权封建制度的建立 (The Establishment of the Manchu-Controlled Mongolian Feudal System of Incarnation), Gugong wenxian 2, no. 1 (1970): 1–18. 26. See Gray Tuttle, Tibetan Buddhists in the Making of Modern China (New York: Columbia University Press, 2005), 22; and see also Qinding Lifan yuan zeli 欽定理藩院則例 (Imperially Commissioned Norms and Regulations of the Board for the Administration of Outlying Regions), in Gugong Zhenben Congkan 300 (Haikou shi: Hainan chubanshe, 2000), juan 58, 9. The three earliest jasagh lamas, Awang Laozang (Ngag dbang blo bzang, 1601– 1687), Laozang Danbei Jiancan (Blo bzang bstan pa’i rgyal mtshan, 1632–1684), and Laozang danba (Blo bzang bstan pa, [act late seventeenth–early eighteenth centuries), wrote prefaces to the imperially sponsored editions of Wutai Shan gazetteers in Chinese and Manchu and included their own biographies among the eminent monks of Wutai Shan. These prefaces are preserved in ed. Gugong bowuyuan 故宮博物院, Qingliang shan zhi. Qingliang shan xin zhi. Qin ding Qingliang shan zhi 清凉山志.清凉山新志.欽定清凉山志 (Record of the Clear and Cool Mountains. New Record of Clear and Cool Mountains. Imperial Record of the Clear and Cool Mountains) (Haikou Shi: Hainan Chubanshe, 2001). See also Tuttle, “Tibetan Buddhism at Wutai shan in the Qing: The Chinese Language Register,” Journal of the International Association of Tibetan Studies, no. 6 (December 2011): 192–94, at http://www.thlib.org?tid=T5721 (accessed September 5, 2014); Natalie Köhle, “Why Did the Kangxi Emperor Go to Wutai Shan?,” 78–79; the biographies are included in Qinding Qingliang shan zhi, juan 16, 21a–22b; and Qingliang shan xin zhi, juan 7, 21b–24b. For a partial English translation of these biographies, see Hoong Teik Toh, “Tibetan Buddhism in Ming China” (PhD diss., Harvard University, 2004), 228–37. 27. Köhle (2008) showed that Pusa Ding was probably converted into a Tibetan Buddhist temple as early as 1481. 28. The temple was then called Dafutu Si. See Huixiang 慧祥, Gu Qingliang zhuan 古清涼傳 (Ancient History of the Clear and Cool Mountains) (Tang dynasty), ed. Takakusu Junjirō 高楠順次郎 (1866–1945) and Watanabe 173 Kaikyoku 渡辺海旭 (1872–1932) et al., Taishō shinshū dai zōkyō 大正新脩大藏經 (Revised version of the canon, compiled during the Taishō era, 1912–26) (Tokyo: Taishō Issaikyō Kankōkai, 1924–32 [–1935]), T.2098: 51, 1094a25–b2. Following standard convention, references to texts in the Taishō canon are indicated by text number (T.), followed by the volume, page, register (a, b, or c), and, when appropriate, line numbers. Qianlong was by no means the irst person to “relocate” Wutai Shan elsewhere through the re-creation of a monastery at Wutai Shan. Throughout its long history, Wutai Shan has been a uniquely popular site of replication in Japan, Korea, Central Asia, Tibet, and areas close to Beijing, and such re-creations frequently involved the erection of a new temple named after a monastery at Wutai Shan. On sites in Japan, see for example, Susan Andrews’s paper, “Moving Mountain: Mount Wutai Traditions at Japan’s Tōnomine,” presented at The Mountain of Five Plateaus Conference, Wutai Shan, Shanxi, July 27–August 2, 2015. On sites in the Tangut state, see Yang Fuxue 楊富學, “Xixia Wutai shan xinyang zhenyi” 西夏五臺山信仰斟議 (Notes on Wutai shan veneration in the Xixia dynasty),” Xixia yanjiu 1(2010): 14–22. On sites in Tibet, see the Tibetan-language guidebook by ’Jam-dbyangs Mkhyen-brtse’i-dbang-po et al., Guide to the Holy Places of Central Tibet (Rome: Istituto Italiano per il Medio ed Estremo Oriente, 1958), 72; and Andreas Gruschke, The Cultural Monuments of Tibet’s Outer Provinces: Kham (Bangkok: White Lotus Press, 2004), 82. See also Shi Jinbo 史金波, Xixia fojiao shilue 西夏佛教史略 (Survey of Tangut Buddhist history) (Yinchuan: Ningxia renmin chubanshe, 1988), 118–19 and 156, cited in Robert Gimello, “Wu-t’ai Shan 五臺山 during the Early Chin Dynasty 金朝: The Testimony of Chu Pien 朱弁,” Chung-Hwa Buddhist Journal 7 (1994): 507. A smaller site known as Wutai Shan also exists in Zhangjiakou west of Beijing. Most recently, a Mountain of Five Peaks was ritually initiated at the Larung Valley in eastern Tibet by Khenpo Jigme Phuntsok. 29. Dou Guangnai, ed., Qinding rixia jiuwen kao, juan 103, 7. The original text reads: 地即清凉, 白馬貝書開震旦, 山仍天竺, 青鴛蘭若近離宮 (This is the very ground of the Clear and Cool [Wutai Shan] Palm Leaf manuscripts of the Baima Monastery that opened China [up to Buddhism]. The mountain is still Indian, but the black-tiled monastery is close to the summer palace.) By inscribing this statement on the placards, Qianlong essentially asserted that the site is a surrogate of Wutai Shan, which is a surrogate of India. 30. Poet Zhu Ziqing 朱自清 (1898–1948) documented in his travelogues the collapse of the gate in the spring of 1932. See Zhu Ziqing, “Songtang Youji” 松堂游记 (Journey to the Pine Pavilion), accessed September 6, 2014, http://www.xys.org / xys/ebooks / literature / prose / Zhu-Ziq ing / songtang.txt. 31. Anne Chayet, “Architectural Wonderland: An Empire of Fictions,” in New Qing Imperial History: The Making of Inner Asian Empire at Qing Chengde, 49; Li 174 ARCHIVES OF ASIAN ART Qianlang, “Beijing Biyun si jingang baota zuo” 北京碧云寺金刚宝塔座 (The Diamond Throne at Beijing Biyun si), Zijin cheng 9 (2009): 12–15; Zhang Yuxin, Qingdai lamajiao beiwen 清代喇嘛教碑文 (Stele inscriptions from Qingdynasty Lamaism) (Beijing: Tianjing guji chubanshe, 1987), 132–33; and Isabelle Charleux, “Copies de Bodhgayā en Asie orientale: Les stupas de type Wuta à Pékin et Kökeqota (Mongolie-Intérieure),” Arts Asiatiques 61 (2006): 120 – 42. Signiicantly for the temple’s connection to rulers, this is where revolutionary and modern China’s founding father Sun Yat-sen’s body was interred temporarily before his burial in his mausoleum in Nanjing. 32. Shan Shiyuan 單士元, Qingdai jianzhu nianbiao 清代建築年表 (Beijing: Zijincheng, 2009), 202. 33. Qinding Qingliang shan zhi, juan 10, 1. 34. That monastery was Luohou Si 羅睺寺. See Qinding Qingliang shanzhi, juan 10, 9b. It is the second-largest Gelukpa Monastery at Wutai Shan, which housed about two hundred lamas by the end of the nineteenth century. Isabelle Charleux, Nomads on Pilgrimage, Mongols on Wutaishan (China), 1800–1940 (Leiden: Brill, 2015), 110, citing Tian Pixu, Wutai Xinzhi (New Gazetteer of Wutai) ([China]: Chongshi shuyuan, 1883). 35. Jiang Xiantong Si Wuliang Dian chicun tangyang chenglan qin 將顯通寺無量殿尺寸盪樣呈覧欽. See Zhongguo diyi lishi dang’an guan, ed., Qinggong Neiwu Fu Zaoban Chu Dang’an Zonghui 清宮內務府造辦處檔案總匯 (The Complete Archive of the Royal Manufactory in the Imperial Household Department), vol. 17 (Beijing: Remin Chubanshe, 2005), 275. 36. The back shrine of the irst loor of Yuhua Ge (Pavilion of Rainy Flowers) is also called a Wuliang Dian 無量殿, referencing the Buddha of Immeasurable Light and the Buddha of Immeasurable Life. See Wang Jiapeng, “Gugong Yuhua Ge tanyuan 故宮雨花閣探源 (Inquiry into the origins of the Pavilion of Rainy Flowers),” Gugong bowuyuan yuankan 47 (1990): 52, 54–55. Located in the northwestern sector of the Forbidden City, Yuhua Ge’s complex structure was designed by Rölpé Dorjé at Qianlong’s request during the same year of 1750. See Berger, Empire of Emptiness, 97–104. 37. Zhongguo diyi lishi dang’an guan, ed., Qinggong Neiwu Fu Zaoban Chu Dang’an Zonghui, vol. 17, 431. 38. Wang Jiapeng, Cultural Relics of Tibetan Buddhism Collected in the Qing Palace (Qinggong Zangchuan Fojiao Wenwu 清宫藏传佛教文物) (Beijing: Forbidden City Press, 1992), 169. 39. Robert Beer, The Handbook of Tibetan Buddhist Symbols (Boston: Shambhala, 2003), 37–42. 40. In numerous passages from the biography of Qianlong’s guru and Qing imperial preceptor Rölpé Dorjé, he was said to have copied rituals from Lhasa. See Berger, Empire of Emptiness, 84, citing Thu’u bkwan, Lcang skya rol pa’i rdo rje’i rnam thar, 138, 225, 187, and 221. The prefaces of the Canon of Iconometry, translated into Chinese by Mongol scholar and Qing court translator Göm- pojab (Tb. Mgon po skyabs; Ch. Gongbu Chabu 工布查布, 1699–1750) state an intention to correct previous HanChinese models, which were thought to be imprecise. See also Kevin Greenwood, “Yonghegong: Imperial Universalism and the Art and Architecture of Beijing’s ‘Lama Temple’” (PhD diss., University of Kansas, 2013), 221–27. In addition to various iconographic pantheons undertaken by Rölpé Dorjé, his compilation of multilingual dictionaries that aimed to standardize the process of translation also served the same need for ritual authenticity. In Empire of Emptiness, Berger astutely shows how these various linguistic and iconographic projects were harnessed to produce an orthodoxy of form and meaning. 41. Zhongguo diyi lishi dang’an guan, ed., Qinggong Neiwu Fu Zaoban Chu Dang’an Zonghui, vol. 17, 431–33. The Chinese date is the second day of the fourth month. The commissioning of drawings for sets of Five Sense Offerings and Eight Offerings are listed in great detail. For example, among the Eight Offerings, the offering of music has a “gilt bronze vajra bell on purple sandalwood tray with cloisonné enamel stand.” 42. Berger, Empire of Emptiness, 6. 43. The entourage departed Beijing on the tenth day of the second month, and returned more than a month later; Zhongguo diyi lishi dang’an guan, ed., Qianlong di qiju zhu, vol. 20, 42–85. 44. A copy of Bitian Xiaoxia (Glowing Clouds in an Azure Sky) is in the Gest Library at Princeton University. See Wu Xiaoling, “Glowing Clouds in an Azure Sky: A Newly Discovered Royal Pageant,” Gest Library Journal 3 (1989): 46–55. In the genre of tributary dramas, the play featured “celestial deities on ive-colored clouds,” “gods of the Five Marchmonts,” and heads of “ten thousand states” arriving to pay obeisance and offer birthday wishes to the emperor and empress-dowager. 45. 是像即非像, 文殊特地殊, 亳端寶王剎, 鏡裡焰光珠, 法雨滄桑潤, 梵雲朝暮圖, 高山仰止近, 屏氣步霄衢。謁殊像寺得句, 因寫滿月容, 以紀其真, 即書於右, 行營促成, 限於方幅, 迴鑾餘暇, 將放展成大圖勒石, 須彌棗葉, 無異無同, 五於此未免著相矣。辛已暮春, 保陽行宮並識. Translation: “An image and not an image, Mañjuśrī’s abode is indeed special. The awe of the bejeweled king is at the tip of the brush, and brilliant laming light in the relection of the mirror. The rain of Buddhist teachings moistens all worldly sufferings, heavenly clouds at dawn and dusk make a marvelous sight. I gaze up at the tall mountains; holding my breath, I approach the high path. When I paid a visit to Shuxiang Si, these verses came to me. Therefore I sketch the full-moon countenance [of Mañjuśrī] in order to document its authenticity, and compose a colophon to its right. This is hastily executed while still on the road, so its size is constrained. When there is time after our return, I will enlarge it and afix it to a rock [i.e., make a relief carving]. Mount Meru and a jujube leaf are neither different nor the same. If one were to insist on this, it would be attaching oneself to form. Written at Baoyang travelling palace, at the end WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī of spring season during the Xinyi year (1761).” It is recorded in Qinding Midian Zhulin, Shiqu Baoji, xubian 欽定秘殿珠林, 石渠寳笈, 續編 (Imperially Ordered Beaded Grove of the Secret Hall and Precious Bookbox of the Stone Drain, supplement) (Taipei: Guoli Gugong Bowuyuan, 1971), 42. Baoyang Palace probably refers to the travelling palace at Baoding 保定 in Hebei province. 46. For evidence of Qianlong’s own hand inside the One or Two paintings, see Kristina Kleutghen, “One or Two, Repictured,” Archives of Asian Art 62 (2012): 37–39. 47. Ibid., 48. The exact term used in the imperial catalogue is fangzhan cheng datu leshi 放展成大圖勒石, i.e., enlarge and aix [the sketch] to a rock. 48. According to his biography, Rölpé Dorjé was in charge of building Baoxiang Si; see Thu’u bkwan, Lcang skya rol pa’i rdo rje’i rnam thar, 486. One modern-day blogger has noted seeing a stele at Baoxiang Si with an image of Mañjuśrī carved on it; however, until further access is permitted, no study of this stele can be undertaken; accessed May 2, 2014, http://blog.sina.com.cn/s /blog_512f6d690100fp9s.html. 49. Dou Guangnai, ed., Qinding rixia jiuwen kao, juan 103, 8. 50. Wang Jiapeng, “Qianlong yu Manzu lama siyuan,” 60. Wang cites Neiwufu zouxiaodang 內務府奏銷檔 (Imperial Household Agency archives, Financial accounts volumes), 319 ce. See also Qinding Lifan yuan zeli, juan 58, 16; juan 59, 25. 51. See ed. Zhao Lin’en 趙林恩, Wutai Shan shige zongji 五台山诗歌总集 (Anthology of Wutai shan poems), vol. 2 (Beijing: Zongjiao wenhua chubanshe, 2002), 407. 52. Ye shes don grub and A lag sha Ngag dbang bstan dar, Ri bo dwangs bsil gyi ’jam dpal mtshan ldan gling gi mtshar sdug sku brnyan gyi lo rgyus bskor tshad dang bcas pa dad ldan skye bo’i spro bskyod me tog ’phreng mdzes (A beautiful lower garland to rouse the faithful: the history and environs of the Beautiful statue of the Temple of Mañjuśrī’s Marks at the Clear and Cool Mountains) (Beijing: Songzhu Si, 1818), in the Collection of the Library of the Minorities Cultural Palace, Beijing. 53. Thu’u bkwan, Lcang skya rol pa’i rdo rje’i rnam thar, 615; Chinese translation by Chen Qingying, 294. 54. D. Pokotilov, “Der Wu T’ai Schan und seine Klöster,” translated from Russian into German by W. A. Unkrig. Sinica-Sonderausgabe (1935): 79. 55. Judging from available images from Dunhuang, this triad was later expanded sometime in the ninth century to include the Kashmiri monk Buddhapāli and the bearded old man. For a study of the Mañjuśrī pentad in Japan, see Wu Pei-Jung, “The Manjusri Statues and Buddhist Practice of Saidaiji: A Study on Iconography, Interior Features of Statues, and Rituals Associated with Buddhist Icons” (PhD diss., University of California, Los Angeles, 2002); Sha Wutian, “Dunhuang P.4049 ‘xinyang Wenshu’ huagao ji xiangguan wenti yanjiu” (A Study of the Sketch of the “New-Style Mañjuśri” in Dunhuang 175 P.4049 and Related Issues) Dunhuang yanjiu (March 2005): 26–32. 56. This sculptural group is also found at Nanchan Si, Foguang Si, and Yanshan Si in the Wutai Shan area from the eighth to the twelfth centuries. 57. Some scholars have argued that this iconography originated not in Wutai Shan but in Khotan, noting the obvious prominence of the Khotan King. For our present purposes, it matters less where this iconography originally came from and more that it somehow became associated with Wutai Shan. See Jiang Li, “Qianxi Dunhuang xinyang Wenshu zaoxiang chansheng de yuanyuan” (A Primary Analysis of the Origins of the Production of “New-Style Mañjuśrī” Images at Dunhuang), Mei yu shidai (January 2010): 67–69. 58. The legend was cited in the carved colophon on a 1608 stele erected by monk Zhencheng, the Ming-dynasty compiler of Qingliang shan zhi. See ed. Cui Zhengsen and Wang Zhichao, Wutai Shan beiwen xuanzhu, 289–91. See also Huanyu, “Shuxiang Si li de chuanshuo gushi” 殊像寺里的传说故事 (Legends of Shuxiang Si), Wutai Shan yanjiu, 3 (1996): 47–48. 59. In this story, the old abbot of the monastery hosted a competition for the design of the main image. Dissatisied with each and every design entry, the old abbot inally accepted the pleas from an extremely skilled sculptor and his team of artisans, who, having journeyed from afar, vowed not to return home if their work did not meet the expectations of the abbot. The project began and progressed in due time, but came to a standstill when the sculptor found himself stymied by artist’s block in attempting to come up with the perfect design for Mañjuśrī’s head. After several days of this, at around lunchtime, clouds suddenly parted, and an image of the perfect form of Mañjuśrī riding on a lion appeared in the sky. Witnessing this, all of the artisans prostrated themselves in amazement. The sculptor immediately got up, ran into the kitchen, grabbed a batch of buckwheat dough prepared for lunch, and sculpted it into the form of the heavenly apparition. Just as he was inishing it, the image of Mañjuśrī disappeared. This story of miraculous occurrence spread far and wide, and soon pilgrims rushed there from all parts of the country to pay homage to the resulting sculpture. See Ye shes don grub and A lag sha Ngag dbang bstan dar, Ri bo dwangs bsil gyi ’jam dpal mtshan ldan gling, 5a. 60. Lcang skya rol pa’i rdo rje, Zhing mchog ri bo rtse lnga’i gnas bzhad (Xining: Mtsho sngon mi rigs dpe sgrun khang. 1993), 43. 61. For a recently published Tibetan source, see Ngag dbang bstan dar, Dwangs bsil ri bo rtse lnga’i gnas bshad (Pilgrimage Guide to the Clear and Cool Five Peak Mountains) (Beijing: krung go’i bod rig dpe skrun khang, 2007), 58; for Mongolian, see Ye shes don grub and A lag sha Ngag dbang bstan dar, Ri bo dwangs bsil gyi ’jam dpal mtshan ldan gling, 5b, line 3; the name also appears as an inscription on the late-eighteenth-/early-nineteenth-century 176 ARCHIVES OF ASIAN ART map of Wutai Shan at Badgar Coyiling Süme. See Wen-shing Chou, “The Visionary Landscape of Wutai Shan in Tibetan Buddhism from the Eighteenth to the Twentieth Century” (PhD diss., University of California, Berkeley, 2011). 62. Charleux, online appendices to the book Nomads on Pilgrimage, 59. 63. Charleux notes that the true image was replaced in 1482 by a new golden statue; Charleux, Nomads on Pilgrimage, 310. 64. Diyi lishi dang’an guan ed., Qinggong Neiwu Fu Zaoban Chu Dang’an Zonghui, vol. 33, 40–41. 65. Ibid., vol. 26, 693. 乾隆二十六年四月十八日: 十八日接得員外郎安泰押帖一件, 內開本月十七日奉旨著丁觀鵬用舊宣紙畫文殊菩薩像著色工筆畫, 得時裱掛軸, 欽此。乾隆二十六年十二月十五日: 十二月十五日接得達色押帖一件, 內開十四日太監胡世傑持來御筆文殊像二幅、丁觀鵬畫文殊像一副。傳旨著觀鵬仿蠟身樣法身起稿, 仍用舊宣紙另畫三幅, 其塔門暫且放下, 先畫文殊像, 欽此。I thank Wang Ching-Ling for irst bringing this reference to my attention. 66. Qinding Midian Zhulin, Shiqu Baoji, xubian 欽定秘殿珠林, 石渠寳笈, 續編 (Imperially ordered Beaded Grove of the Secret Hall and Precious Bookbox of the Stone Drain, supplement) (Taipei: Guoli Gugong Bowuyuan, 1971), 357–58. 67. This would date the painting to the eleventh month of the year, just one month before it was presented to the emperor as a record in Neiwu fu Ruyi guan’s documents. This is noted by Wang Ching-Ling in an e-mail to the author, August 13, 2010. 68. Qinding Midian Zhulin, Shiqu Baoji, xubian, 405. The inscription reads: 臣裘曰脩之母王氏率孫媳等敬繡. 69. I thank Christian Luczanits for this observation. 70. David Snellgrove, Indo-Tibetan Buddhism: Indian Buddhists and their Tibetan Successors (Boston: Shambhala, 2002), 231–70. 71. The immensely informative Tibetan biography of Rölpé Dorjé by Tuken records the initiations. For an analysis of them, see Wang Xiangyun, “Tibetan Buddhism at the Court of Qing: The Life and Work of lCang-skya Rolpa’i-rdo-rje (1717–86)” (PhD diss., Harvard University, 1995), 293–96. 72. Zito, Of Body and Brush. 73. James Hevia, “Lamas, Emperors, and Rituals: Political Implications in Qing Imperial Ceremonies.” Journal of the International Association of Buddhist Studies 16, no. 2 (1993): 246. 74. I thank Lin Wei-Cheng for pointing this out to me at the Association for Asian Studies conference in 2012. 75. The complete inscription reads: 乾隆辛已春。上以祝釐巡幸五臺。瞻禮曼殊寶相。圓光默識。如月印川。回鑾後。摹寫為圖。水墨莊嚴。妙合清涼真面。復以稿本。命小臣觀鵬設色。齋盥含毫。積七閱月。雖華鬘珠珞。猊座蓮臺。殫竭小乘知解。而於師利本來相好。實未能裨助萬一。竊自念凡庸末技。幸得仰承天筆。擬繪金容。譬諸匠眾為優填王作旃檀像。雕鐫塗澤。無足名稱。而濁質鈍根。獲霑香國功德。歡喜信不可思議。臣丁觀鵬敬識。 Translation: During the spring of 1761, the emperor toured Wutai to obtain blessings. He visited the precious image of Mañjuśrī [at Shuxiang Si], which is radiant with the subtle glow of wisdom, like the moon’s seal on a river. After returning to his palace, he made a sketch after the image in ink splendor. The sketch wondrously matched the true countenance of [the sage of] Clear and Cool [Mañjuśrī]. Then, based on the sketch, he ordered the humble servant Guanpeng to make a colored painting. Observing ritual fasting and cleansing, I diligently held a brush for seven months. [Even though I was able to paint] the garlands and the beaded pearls, the lion throne and the lotus pedestal, exhausting all knowledge and understanding of the smaller vehicle, Mañjuśrī’s primary and secondary marks are originally excellent, therefore my brush could not enhance even one-ten-thousandth. I secretly feel my thoughts are banal and my skill is limited. Fortunately, by relying on the heavenly brush [of the emperor], I was able to lay out the golden countenance. Just like the artisans who carved the sandalwood Buddha for King Udayana, [I completed it] after much careful chiseling and modiication. It can still hardly deserve to be called anything. Even though my qualities are impure and roots dull, that I can still obtain merits of the fragrant land [paradise of Amitabha], I feel blissful beyond measure. Servant Ding Guanpeng respectfully acknowledges [this]. 76. For a thorough study of the history of sandalwood Buddhas in China, see Martha Carter, The Mystery of the Udayana Buddha (Napoli: Istituto Universitario Orientale, 1990). 77. Qianlong instructed Ding to paint an image of Mañjuśrī by “imitating a wax model” (fang lashen yang fashen 仿蠟身樣法身). 78. See Patricia Berger, “Lineages of Form: Authority and Representation in the Buddhist Portraits of the Manchu Court,” Tibet Journal 28, nos. 1–2 (2003): 109–46; and Michael Henss, “The Bodhisattva-Emperor: TibetoChinese Portraits of Sacred and Secular Rule in the Qing Dynasty,” Oriental Art 3 (2001): 1–16, and 5 (2001): 71– 83. 79. They counter, in Berger’s words, “the apparent immediacy of the imperial face with a patterned, canonical pantheon of visions.” Berger, Empire of Emptiness, 61. 80. Indeed, Berger credits Ding’s copies of true images with the integration of these two modes of representation—“while Ding faithfully copied the overall outlines of archaic vision, in an illusionist’s trick he also leshed them out, plumped them up, and made them uncannily real.” Ibid., 166. 81. “Farquhar’s inluential work may also represent the prevailing tendency to focus exclusively on Tibetan Buddhist concepts and materials.” Farquhar attributed Qianlong’s self-promotion to a political need to manage the allegiance of Mongols and, later, Tibetans, who had become subjects of their expanding empire; and he did this WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī by showing how Qianlong employed the same method of self-identiication as Mongols and Tibetans, who by the seventeenth century had all become adherents of Tibetan Buddhism and were subscribing to a growing system of ecclesiastical reincarnations in which occupants of monastic thrones were considered incarnations of speciic bodhisattvas. That is, the Qing emperors’ parallel self-identiication would therefore allow them to gain control of Tibet and Mongolia by raising themselves to the same level of divinity as Tibet and Mongolia’s own bodhisattva incarnates. Furthermore, Farquhar demonstrated how early Qing rulers irst justiied their claims on Mongol precedence by tracing their ability to claim this status for themselves to a history of priest–patron relationships between Chinese emperors and Tibetan lamas that had begun in China during the Mongol Yuan dynasty (1271–1368): the invitation of the Fifth Dalai Lama (1617–1682) by the Qing dynastic founder Abahai (1592–1643) to his court in Mukden in 1637 was modeled after previous emperors who invited religious teachers to their court, and especially the lama–patron relationship between Kublai Khan (1215–1294), the founder of the Yuan dynasty, and his imperial preceptor the ’Phags pa lama (1235–1280) from Tibet; it was upon his return to Tibet that the Fifth Dalai Lama, jointly with the Fourth Panchen Lama (1570–1662), bestowed on Abahai the title of “Mañjuśrī-Great Emperor,” a title that was maintained and solidiied by and for subsequent Manchu emperors. This identiication was made indisputable by the Third Dalai Lama’s prophecy that a great secular incarnation of Mañjuśrī would unite China, Mongolia, and Tibet, together with the homophonic similarity between the names “Manchu” and “Mañju.” See David Farquhar, “Emperor as Bodhisattva,” 15–20; the notion that Qianlong promoted his image as the Mañjughoṣa emperor exclusively toward the Mongols and Tibetans has also been implicitly substantiated by the fact that Qianlong at the same time supported Confucian state rituals and the preservation of Manchu Shamanistic rituals. For more on Qianlong’s practice of Confucian state rituals and participation in Shamanistic rituals, see Angela Zito, Of Body and Brush: Grand Sacriices as Text/Performance in EighteenthCentury China (Chicago: University of Chicago Press, 1997), and Di Cosmo, “Manchu Shamanic Ceremonies.” 82. On Tibetan appellation of Qianlong as Mañjuśrī, the Panchen Lama Blo bzang dpal ldan yes shes composed a prayer to Qianlong’s previous incarnations on the occasion of the emperor’s seventieth birthday. See Vladimir Uspensky, “The Previous Incarnations of the Qianlong Emperor According to the Panchen Lama Blo bzang dpal ltan ye shes,” in Tibet, Past and Present: Proceedings of the Ninth Seminar of the International Association for Tibetan Studies, ed. Henk Blezer (Leiden: 2000), 215–28. An album of thirteen painted leaves featuring Qianlong and his twelve previous incarnations, now in the Palace Museum Library in Beijing, is based on this incarnation lineage. It is accompanied by quadrilingual inscriptions in 177 Tibetan, Manchu, Mongolian, and Chinese, and situates Qianlong in a mountainous “pure land” of Wutai Shan. I thank Lin Shih-Hsuan for bringing this album to my attention. 83. Recent scholarship has come to challenge Farquhar’s claim (that Qing emperors’ Buddhist guise was primarily targeted at Mongols and Tibetans) by focusing on Chinese language texts that reiterate their Mañjuśrī appellation and spell out their sumptuous imperial donation in both Chinese and Tibetan monasteries at Wutai Shan. See Natalie Köhle, “Why Did the Kangxi Emperor Go to Wutai Shan?; and Gray Tuttle, “Tibetan Buddhism at Wutai Shan in the Qing: The Chinese-language Register.” Gray Tuttle’s study of imperially endorsed Chinese-language gazetteers challenges Farquhar’s assertion that the Manchu emperors’ association with Mañjuśrī was a strategy directed at Tibetan and Mongolian populations. Tuttle postulates instead that Han-Chinese adherents of Tibetan Buddhism were a principal recipient of Qing imperial patronage and were the population group especially targeted for the propagation of Wutai Shan and, in particular, Tibetan Buddhism. 84. Ibid. For the design of the mandalic structure, see Heather Stoddard, “Dynamic Structures in Buddhist Mandalas: Apradakṣina and Mystic Heat in the Mother Tantra Section of the Anuttarayoga Tantras,” Artibus Asiae 58, no. 3–4 (1999): 169–213. Stoddard relates the building of the Cakrasamvara Mandala at Pule Si in the Qing imperial summer palace of Jehol to the mandala initiations that Qianlong undertook at around the same time. The presentday ruins of Baoxiang Si are located inside a closed military compound and are therefore closed to researchers. 85. 歲辛巳, 值聖母皇太后七旬大慶, 爰奉安輿詣五臺, 所以祝釐也。殊像寺在山之麓, 為瞻禮文殊初地, 妙相莊嚴, 光耀香界, 默識以歸。即歸則心追手摹, 係以讚而勒之碑。香山南麓, 曩所規菩薩頂之寶諦寺在焉。迺於寺右度隙地, 出內府金錢, 飭具庀材, 營構藍若, 視碑摹而像設之。… 經始於乾隆壬午春, 越今丁亥春蕆工。 See Zhang Yuxin 張羽新, Qing zhengfu yu lama jiao 清政府與喇嘛教 (The Qing government and Lama Religion) (Lhasa: Xizang remin chubanshe, 1988), 409–11. 86. See note 38 above. 87. 因記之曰：文殊師利久住娑婆世界, 而應現說法則獨在清涼山, 固《華嚴品》所謂東方世界中菩薩者也。夫清涼在畿輔之西, 而香山亦在京城之西。然以清涼視香山, 則香山為東, 若以竺乾視震旦, 則清涼、香山又皆東也。是二山者不可言同, 何況云異？矧陸元暢之答宣律師曰：文殊隨緣利見, 應變不窮, 是一是二, 在文殊本不生分別見, 倘必執清涼為道場, 而不知香山之亦可為道場, 則何異鑿井得泉而謂水專在是哉？而昔之詣五臺禮文殊, 所以祝釐也, 而清涼距畿輔千餘里, 掖輦行慶, 向惟三至焉。若香山則去京城三十里而進, 歲可一再至。繼自今憶萬年延洪演乘, 茲惟其恒, 是則予, 建寺香山之初志也。寺成, 名之曰：寶相。See Zhang, Qing zhengfu yu lama jiao, 408. 88. For Qing knowledge of and policy toward India, see Matthew Mosca, From Frontier Policy to Foreign Policy: The Question of India and the Transformation of 178 ARCHIVES OF ASIAN ART Geopolitics in Qing China (Stanford, CA: Stanford University Press, 2013). 89. Dou Guangnai, ed., Qinding rixia jiuwen kao, juan 103, 8. Concerning the building of Baoxiang Si’s mall hall, the text reads: 命於寶諦寺旁，建茲寺，肖像其中，殿制外方內圓，皆甃甓而成，不施木植，四面設甕門. 90. Berger, Empire of Emptiness, 161. 91. David Snellgrove and Hugh Richardson, A Cultural History of Tibet (New York: Frederick A. Praeger, 1968), 115–16. 92. Ishihama, “Study on the Qianlong as Cakravartin, a Manifestation of Bodhisattva Mañjuśrī, Tangka,” Waseda daigaku Mongol Kekyusyo, 2, no. 24 (March 2004). Qianlong gave at least one painting to Rölpé Dorjé around the year 1784; see Wang, “Tibetan Buddhism at the Court of Qing,” 296. 93. Caroline Bodolec, “Uncommon Public Buildings with Vault with Abutments in the Chinese Landscape of Wooden Construction (Sixteenth–Eighteenth Centuries),” in Proceedings of the Second International Congress on Construction History, vol. 1 (Exeter, UK: Short Run Press, 2006), 409–16. 94. The Manchu Buddhist canon took nearly twenty years and more than ive hundred translators to complete. A Manchu translation bureau (Qingzi Jingguan 清字經館) was established in 1772 (the thirty-seventh year of the Qianlong reign). See Marcus Bingenheimer, “The History of the Manchu Buddhist Canon and First Steps towards its Digitization,” Central Asiatic Journal 56 (2012–13): 203– 19; Gao Mingdao (a.k.a. Friedrich Grohmann),“Rulai zhiyin sanmai jing fanyi yanjiu 如來智印三昧經翻譯研究” (master’s thesis, Taipei: Chinese Culture University, 1983), 1–33, 153–205; Walther Fuchs, “Zum mandjurischen Kandjur,” Asia Major 6 (1930): 388–402; and Hans-Rainer Kämpfe, “Einige tibetische und mongolische Nachrichten zur Entstehungsgeschichte des mandjurischen Kanjur,” Zentralasiatische Studien 9 (1975): 537–46. On the connection with Shuxiang Si in Chengde, see Feng Shudong 馮術東,“Shuxiang Si yu manwen dazang jing” 殊像寺與滿文大藏經 (Shuxiang Si and the Manchu Canon), Wenwu Chunqiu 1 (2005): 41–43. 95. According to Wang, Lifan yuan’s records indicate sixty-three Manchu lamas resided in Shuxiang Si; see Wang, “Qianlong yu Manzu lama siyuan,” 62. Wang does not cite the speciic passage. 96. One exception to the structure of the Chinese canon is addition of esoteric texts; see Gao Mingdao, “Rulai zhiyin sanmai jing fanyi yanjiu,” 10. 97. 莊校金容, 一如香山之制；而殿堂樓閣, 略仿五臺山。 The commemorative stele dates to the fortieth year of the Qianlong reign (1775); see Zhang, Qing zhengfu yu lama jiao, 443. 98. Meng Fanxing 孟繁興, “Chengde Shuxiang Si yu Wutai Shan Shuxiang Si 承德殊像寺與五台山殊像寺,” in Bishu shanzhuang luncong 避暑山莊論叢 (Collected essays on the summer palace) (Beijing: Zijincheng chubanshe: 1986), 450–54. 99. Anne Chayet, “Architectural Wonderland: An Empire of Fictions,” in New Qing Imperial History: The Making of Inner Asian Empire at Qing Chengde, ed. James A. Millward et al. (London and New York: Routledge Curzon, 2004), 49. 100. The current structure is ive bays wide and three bays deep. According to Qinding Qingliang shan zhi (1785), the hall measures two bays (three ying 楹) wide; see Qinding Qingliang shan zhi, juan 10, 10. 101. The name is a variation of Qianlong’s iconographic project, the Baoxiang Lou 寶相樓, located in the Cining Palace inside the Forbidden City, which was also constructed in honor of the eightieth birthday of his mother, the empress-dowager, in 1771. 102. Zhengjue Si, another Manchu monastery that began construction in 1773 near the Yuanming Yuan Summer Palace, also appears to have had an octagonal pavilion that housed a sculptural image of Mañjuśrī on a lion. See Zhou Fang, Zhengjue Si li hua jinxi 正觉寺里话今昔 (Speaking of the Past and the Present inside Zhenjue Si), accessed August 26, 2015, http://www.mzb.com.cn/html/ Home/report/220569-2.htm. This sculptural image is also referenced in Eugene Pander, Lalitavajra’s Manual of Buddhist Iconography (New Delhi: International Academy of Indian Culture and Aditya Prakashan, 1994), 40. 103. The Getty Conservation Institute has been engaged in a restoration project at Shuxiang Si since 2002, and has published old photographs of the monastery for its extensive conservation report. See Chengde Cultural Heritage Bureau, Hebei Cultural Heritage Bureau, and The Getty Conservation Institute, Assessment Report on Shuxiang Temple, Chengde, rev. ed. (Los Angeles: Getty Conservation Institute, 2009); accessed April 29, 2011, http:// getty.edu/conservation/publications/pdf_publications /shuxiang.html. 104. Xiang Si 向斯, Huangdi yu foyuan 皇帝的佛緣 (Emperors and Buddhism) (Hong Kong: Heping tushu, 2005), 297; and Banyou, “Wai ba miao yu qingdai zhengzhi 外八廟與清代政治,” Chengde minzu zhiyie jishu xueyuan xuebao 承德民族職業技術學院學報, vol. 4 (1996), 47. Sven Hedin noted, “Folk-lore says that the rider on the lion is the divine representation of the Emperor Ch’ien Lung as Mañjuśrī.” Hedin, History of the Expedition in Asia 1927–1935 (Stockholm: Elanders, 1943), 141; http://dsr.nii.ac.jp/toyobunko/E-290.9-HE01-025/V-2 /page/0149.html.en. See also a similar description of another Mañjuśrī on a lion at Zhenjue Si by Eugene Pander in Pander, Lalitavajra’s Manual of Buddhist Iconography, 40, cited in Berger, Empire of Emptiness, 226. Berger wonders whether Pander could be referring to another temple at Wanshuo Shan. Zhenjue Si was also a Manchu Buddhist monastery built around the same time as Chengde’s Shuxiang Si. It would not be surprising to ind Qianlong’s replicas of Shuxiang Si’s Mañjuśrī sculpture at other Manchu monasteries as well. 105. I thank Lin Shih-Hsuan for this observation. WEN-SHING CHOU • Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī 106. Qi Jingzhi 齊敬之, Wai ba miao beiwen zhushi 外八廟碑文註釋 (The Eight Outer Temple’s Annotated Inscriptions) (Beijing: Zijingcheng chubanshe, 1985), 92: 殊像亦非殊，堂堂如是乎。雙峰恆並峙, 半里弗多纖。法爾現童子，巍然具丈夫。丹書過情頌，笑豈是真吾。 107. These childhood objects include a silver vase, a golden bowl, ivory pillars, and porcelain plates; see Feng Shudong, “Shuxiang Si yu manwen dazang jing,” 397. 108. T. 20.1195. 109. For a more detailed study of this translation project and its signiicance, see Lin Shih-Hsuan, “Wutai Shan yu Qing Qianlong nianjian de manwen fojing fanyi” 五臺山與清乾隆年間的滿文佛經繙譯 (Wutai Shan and the translation of Manchu Buddhist scriptures during the Qianlong reign). Paper presented at The Mountain of Five Plateaus Conference, Wutai Shan, Shanxi, July 27–August 2, 2015. 110. See Lin Shih-Hsuan, Qingdai menggu yu manzhou zhengzhi wenhua, 215, quoting Yuzhi shi siji 御製詩四集 (Imperial Poems in Four Volumes), juan 89, 19. 111. The original reads: 《大聖文殊師利菩薩讚佛法身禮經》載漢經中而番藏中乃無。去歲巡幸五臺, 道中因以國語譯出, 並令經館譯出西番、蒙古, 以金書四體經供奉臺頂及此寺. 112. A study has yet to be done on how many copies were made and how widely they were disseminated. 113. Crossley, A Translucent Mirror, 266. 179 114. Beijing Number One Archive, document no. 0401-38-0015-011. 115. For a comparison of the different versions of gazetteers at Wutai Shan, see Chou, “The Visionary Landscape of Wutai Shan,” 51–53. 116. Johan Elverskog, Our Great Qing: The Mongols, Buddhism, and the State in Late Imperial China (Honolulu: University of Hawai‘i Press, 2006), 8. 117. As Stephen Whiteman shows through his study of the imperial summer retreat in modern-day Chengde during the reign of Qianlong’s grandfather Kangxi, the retreat constructed under Kangxi displayed a very different vision of rulership than that of the Qianlong era, despite Qianlong’s employment of a rhetoric of continuation from his grandfather. The transformations of the summer retreat that took place between the two reigns relected the fact that whereas Kangxi faced the “challenges of conquest and consolidation,” it was only under Qianlong’s reign that a model of universal emperorship was established. See Whiteman, “From Upper Camp to Mountain Estate: recovering historical narratives in Qing imperial landscapes,” in Studies in the History of Gardens & Designed Landscapes 33, no. 4 (October, 2013): 266. 118. See Shan, Qingdai jianzhu nianbiao, 202. 119. Di Cosmo, “Manchu Shamanic Ceremonies,” 390.

Log In

Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī

Imperial Apparitions: Manchu Buddhism and the Cult of Mañjuśrī

Related Papers

RELATED PAPERS