Stūpa to Maṇḍala: Tracing a Buddhist Architectural Development from Kesariya to Borobudur to Tabo

Swati Chemburkar

Borobudur, Java, Indonesia. The latest discoveries and astonishing revelations, which rewrite the history of the kings Sailendras and Borobudur. Who built this temple, this extraordinary mandala? The precise dates of construction of Borobudur. Why and how are the three Buddhist temples, Candi Borobudur, Candi Pawon and Candi Mendut linked? Who were the Sailendras kings of Java and what were their connections to the kingdom of Srivijaya and Palembang, Sumatra? Where was the palace of the Sailendras kings located in Java? And a lot of other completely new informations. Keywords : Indonesia, Java, bouddhisme, Borobudur, Mendut, Pawon, Sailendra, candi, temple, mandala.

The stupa of Borobudur in Indonesia is remarkable and intriguing for its magnificent architectural and iconographic conception. Since the other massive stupa structures of earlier or similar date, with which it could be compared, have reduced to ruins, it now appears unique. It has been suggested that Indian prototypes surely existed, such as ruined stupa of Nandangarh or the stupa of Kesariya in Bihar. With the recent excavations and restorations at Kesariya in east Champaran, Bihar, these architectural linkages are emerging stronger. The paper compares the sacred architectural space of Kesariya in Bihar, India, Borobudur in central Java and the main temple of Tabo monastery in the Spiti valley, India. It also addresses the larger question of how and when within this chronological context a sacred geometry evolved from a stupa to mandala and how the architectural ideas travelled within the connected Buddhist world.

I became a member of the Triratna Buddhist Order 45 years ago. Since then, I founded a charity to fund social work in India, I co-ran the London Buddhist Centre and I co-founded the North London Buddhist Centre. I am currently the President of the North London Buddhist Centre. My academic interests are the early development of Buddhism in India and its spread out of the subcontinent. I studied the historical development of Buddhist traditions as well as the archaeology of Buddhist art and architecture at the School of Oriental and African Studies. My special interest is in the rock-cut Buddhist monasteries of Western India, since they have the best-preserved remains of early Buddhist monasticism. Of these, I chose Bhaja, Bedsa and Karla for the subject of my PhD research, because their close proximity to each other enabled me to make a detailed study of local Buddhist history from the regional physical remains. I finished this study and graduated in 2022.

Stūpa to Maṇḍala: Tracing a Buddhist Architectural Development from Kesariya to Borobudur to Tabo1 Swati Chemburkar Jnanapravaha, Mumbai INTRODUCTION There were occasions for the direct transfer of Southeast Asian Buddhist developments to India, and there is evidence of at least two specific moments when this occurred. Both instances provide opportunities for a range of interpretative analyses.2 Hiram Woodward, in his “Esoteric Buddhism in Southeast Asia in the Light of Recent Scholarship,” singles out the moment when Bālaputradeva, an exiled scion of the Śailendra dynasty, the builders of the Buddhist Borobudur monument in Central Java, established a 1. This article is based on a paper presented at the conference “Cultural Dialogues between India and Southeast Asia from the 7th to the 16th Centuries” at the K.R. Cama institute, Mumbai, in January 2015. The Kesariya-Borobudur part of this article appears in Swati Chemburkar, “Borobudurs Pāla Forebear? A Field Note from Kesariya, Bihar, India,” in Esoteric Buddhism in Mediaeval Maritime Asia: Networks of Masters, Texts, Icons, ed. Andrea Acri (Singapore: ISEAS, 2016). I owe a special word of thanks to Prof. Tadeusz Skorupski for introducing me to esoteric Buddhism and generously sharing his deep knowledge of texts. I appreciate the critique of my draft by Hiram Woodward and Max Deeg. Despite their feedback, errors may still remain and they are no doubt mine. My sincere thanks to Yves Guichand and Christian Luczanits for graciously providing me the aerial images of the Kesariya stūpa and the layout of Tabo Monastery along with the photos. 2. Hiram Woodward, “Review: Esoteric Buddhism in Southeast Asia in the Light of Recent Scholarship,” Journal of Southeast Asian Studies 35, no. 2 (2004): 346–347. 169 170 Pacific World, 3rd ser., no. 20 (2018) monastery at Nālandā, Bihar in 850 or 860 CE.3 A verse inscribed on a small stūpa at this monastery is taken from the Bhadracarīpraṇidhāna (Vows of Bodhisattva Samantabhadra). The same text informs the ninth-century reliefs of the topmost galleries at Borobudur.4 To Woodward, this suggests that there were either long-standing similarities between Nālandā and central Java or it was Bālaputra’s monastery that brought new emphasis to Nālandā from abroad. Deciding between these two possibilities is not an easy option, and Woodward tends to favor the latter. The new emphasis in design—the circular arrangement of deities in certain numerological configurations on the upper three terraces of Borobudur—appears to reflect a characteristic of the yoginī-tantras that developed at Nālandā in the late eighth to early ninth centuries.5 The distinctive architecture of Borobudur is still being debated. Scholars have looked at Indian prototypes in the ruined stūpa of Nandangarh6 and the partially excavated stūpa of Kesariya7 in Bihar. The unique, almost circular arrangement of deities in the external niches of Kesariya suggests an architectural linkage with Java and the possibility of the new emphasis having some earlier currents in the Buddhist world of Nālandā. 3. Hirananda Sastri’s text of the inscription can be found in “The Nālandā Copper-Plate of Devapāladeva,” Epigraphia Indica 17 (1923–1924): 310–327; and in Hirananda Sastri, Nālandā and Its Epigraphic Material: Memoirs of the Archaeological Survey of India (Delhi: Archaeological Survey of India, 1942), 95. 4. Gregory Schopen translated the text in “A Verse from the Bhadracarīpraṇidhāna in a 10th Century Inscription Found at Nālandā,” Journal of the International Association of Buddhist Studies 12 (1989): 149–157. See also Hiram Woodward, “The Life of the Buddha in the Pāla Monastic Environment,” Journal of the Walters Art Gallery 48 (1990): 15–17. 5. Ronald Davidson, Indian Esoteric Buddhism: A Social History of the Tantric Movement (1st Indian ed., Delhi: Motilal Banarasidas, 2004), 118, 302. 6. For a detailed account of Nandangarh stūpa and its possible influence on Javanese monuments, see J. E. van Lohuizen-de Leeuw, “South-East Asian Architecture and the Stūpa of Nandangaṛh,” Artibus Asiae 19, nos. 3–4 (1956): 279–290; Joyanto Sen, “The Colossal Stupa at Nandangarh: Its Reconstruction and Significance,” Artibus Asiae 75, no. 2 (2015): 179–220. 7. Based on the overall measurements and the architecture, Caesar Voûte and Mark Long list similarities and differences between Kesariya and Borobudur in Borobudur: Pyramid of the Cosmic Buddha (New Delhi: D. K. Printworld, 2008), 187–191. Chemburkar: Stūpa to Maṇḍala 171 The second historical moment of immediate contact between Southeast Asian Buddhism and India, which Woodward alludes to, came two centuries later. In 1012 CE, a learned Buddhist monk from northeast India went to live in “Śrīvijaya” to study esoteric Buddhism under Dharmakīrti.8 He was born Candragarbha, renamed as Dīpaṃkaraśrījñāna when he entered the sangha, and after initiation into yoginī-tantras he received the name Atīśa. After studying for twelve years somewhere in the maritime federation known as Śrīvijaya, he carried up to Tibet the oldest surviving Śrīvijayan Buddhist commentary Durbodhāloka (Illuminating the Unfathomable), composed by his teacher, Dharmakīrti.9 This text, extant only in its Tibetan translation, says that it was written “in the city of Śrīvijaya in Suvarṇadvīpa” under the patronage of the Śailendra monarch Cūḷāmaṇivarman.10 Besides this text, certain concepts regarding inner and outer maṇḍalas were picked up by Atīśa during his Śrīvijayan sojourn and possibly carried to Tibet.11 Among the surviving Buddhist temples of India, Tabo in Himachal displays a complete sculptural maṇḍala of the life-size clay figures of the Vajradhātu Maṇḍala deities. Atīśa visited Tabo in 1042 CE when the 8. Bimalendra Kumar, “Contribution of Ācārya Dharmapāla of Nālandā,” in The Heritage of Nālandā, ed. C. Mani (New Delhi: Aryan Books/Asoka Mission, 2008), 103; B. B. Kumar, “Nālandā: Its Significance,” in ibid., 185. 9. Alka Chattopadhyaya, Atīśa and Tibet: Life and Works of Dipaṃkara Śrījñāna in Relation to the History and Religion of Tibet with Tibetan Sources (Delhi: Motilal Banarsidas, 1996), 84–95; Peter Skilling, “Geographies of Intertextuality: Buddhist Literature in Pre-Modern Siam,” Aséanie 19 (2007): 94. 10. J. A. Schoterman, Indonesische Sporen in Tibet (Leiden: Brill, 2008), 185; Peter Skilling, “Dharmakīrti’s Durbodhāloka and the Literature of Śrīvijaya,” Journal of the Siam Society 85, parts 1–2 (1997): 187–194. According to John Miksic, Śrīvijaya could be Palembang, Jambi, Chaiya, or Kedah at different times in the connected maritime Malay world of the peninsula and Sumatra; see Singapore and the Silk Road of the Sea 1300–1800 (Singapore: NUS Press, 2013), 110. 11. Alex Wayman, “Reflections on the Theory of Barabudur as a Maṇḍala,” in Barabudur: History and Significance of a Buddhist Monument, ed. Luis O. Gomez and Hiram W. Woodward (Berkeley: Institute of Buddhist Studies, 1981), 140–2; Max Nihom has disputed this in Studies in Indian and Indo-Indonesian Tantrism: Kuñjarakarṇadharmakathana and the Yogatantra (Vienna: De Nobili Institut für Indologie der Universität Wien, 1994), 72n192. 172 Pacific World, 3rd ser., no. 20 (2018) monastery was undergoing major renovation.12 An exactly contemporaneous set of Vajradhātu Maṇḍala bronzes survives from East Java.13 At the time of Atiśa’s departure from Śrīvijaya, esoteric Buddhist sites sprouted in several parts of Sumatra, especially at Muara Jambi. The majority of temples are in ruins today, but the objects found from the site of Caṇḍi Gumpung contain four vajras and gold sheets from the tenth century inscribing the deities of the Vajradhātu Maṇḍala.14 The Buddhist tradition of Java and Śrīvijaya probably shared many elements. Hudaya Kandahjaya urges us to keep in mind that the Javanese island wasn’t a blank sheet when Sumatra was bustling with Buddhist 12. Deborah Klimburg-Salter et al., Tabo: A Lamp for the Kingdom: Early IndoTibetan Buddhist Art in the Western Himalaya (Milan: Skira, 1997), 91, 105. 13. The Nganjuk bronzes, discovered in 1913 and now split between the National Museum Jakarta and other collections and museums around the world, belong almost entirely to the Vajradhātu Maṇḍala described in the eighth-century Sarvatathāgatatattvasaṅgraha and Sarvadurgatipariśodhanatantra as well as in maṇḍala 19 in the later Niṣpannayogāvalī. Lokesh Chandra (in collaboration with Mrs. Sudarashana Devi Singha), “Identification of the Nanjuk Bronzes” and “The Buddhist Bronzes of Surocolo,” in Cultural Horizons of India: Studies in Tantra and Buddhism, Art and Archaeology, Language and Literature, Vol. 4 (New Delhi: International Academy of Indian Culture and Aditya Prakashan, 1995), 97–107 and 121–147 respectively; Benoytosh Bhattacharya, ed., Niṣpannayogāvalī of Mahapāndita Abhayākaragupta (Baroda: Oriental Institute, 1972). 14. The largest concentration of Buddhist sites appeared in Muara Jambi in the eleventh century. See John Miksic, “The Buddhist-Hindu Divide in Premodern Southeast Asia,” Nalanda-Sriwijaya Working Paper Series 1 (2010): 27. S. Nagaraju speculates that Caṇḍi Gumpung was “the principal monastery in the region.” S. Nagaraju, “A Central Sumatran Metropolis at Muara Jambi and Its Buddhist Connection: Some Reflections,” in Śrī Nāgābhinandanam: Dr. M. S. Nagaraja Rao Festschrift, ed. L. K. Srinivasan and S. Nagaraju (Banglore: Dr. M. S. Nagara Rao Felicitation Committee, 1995), 2:750. The gold foil sheets found in ritual deposit boxes in the ruins of Muara Jambi bear the names of five tathāgatas, sixteen vajrabodhisattvas, and sixteen vajratārās of the Vajradhātu Maṇḍala. Along with the gold sheets, there were stūpikas found among the ruins of Caṇḍi Gumpung that were placed on the platform in a pentad arrangement of the key Sarvatathāgatatattvasaṅgraha buddhas. See M. Boechari, “Ritual Deposits of Caṇḍi Gumpung (Muara Jambi),” Final Report: Consultative Workshop on Archaeological and Environmental Studies of Srivijaya (Bangkok: SPAFA, 1985), Appendix 7d, 229–243. Chemburkar: Stūpa to Maṇḍala 173 activities.15 The Śailendra-period gold foil unearthed from Ratu Boko near the Prambanan temple complex and a lead bronze foil with inscribed dhāraṇī unearthed during the restorations of Borobudur16 display elements of the Vajradhātu Maṇḍala.17 The murals of Tabo and Borobudur both illustrate pilgrim Sudhana’s wanderings around India as described in the Gaṇḍavyūha-sūtra, and the sacred space of the two monuments is arranged on similar principles. This paper therefore looks at the development of architectural space at Kesariya in east Champāran, Bihar, India (ca. seventh to eighth centuries CE); Borobudur in Central Java, Indonesia (ca. eighth to ninth centuries CE); and the main temple of Tabo Monastery (founded in 996 CE and renovated in the eleventh century) in the Indo-Tibetan sphere, 15. Hudaya Kandahjaya, “Saṅ Hyaṅ Kamahāyānikan, Borobudur, and the Origins of Esoteric Buddhism in Indonesia,” in Esoteric Buddhism in Mediaeval Maritime Asia, Networks of Masters, Texts, Icons, ed. Andrea Acri (Singapore: ISEAS, 2016), 85. 16. Arlo Griffiths, “The Greatly Ferocious Spell (Mahāraudra-nāma-hṛdaya): A Dhāraṇī Inscribed on a Lead-Bronze Foil Unearthed near Borobudur,” Epigraphic Evidence in the Pre-Modern Buddhist World: Proceedings of the Eponymous Conference Held in Vienna, ed. K. Tropper (Wien: Arbeitskreis für tibetische und buddhistische Studien, Univ. Wien, 2014), 1–36. The foil is presently preserved at the Borobudur site museum. This unearthed dhāraṇī has displayed close inter-textual connections to the Sarvatathāgatatattvasaṅgraha, the root text of yoga-tantra that defined the Vajradhātu Maṇḍala. 17. The Buddhist mantra oṁ tạkī hūṁ jaḥ svāhā inscribed on gold foil was unearthed sometime during or just after the Second World War. Its description occurs in the reports of Archaeological Service of the former Netherlands East Indies (Oudheidkundig verslag, 1950). The first analytical commentary was offered by the late Indonesian archaeologist Kusen in 1994, but since I don’t read Indonesian I have referred to Jeffrey Sundberg, who dates it to 784–803 CE in “A Buddhist Mantra Recovered from the Ratu Baka Plateau: A Preliminary Study of Its Implications for Sailendra–Era Java,” Bijdragen tot de taal-, Land- en Volkenkunde 159 (2003): 165, 170, 171; Arlo Griffiths, “The Greatly Ferocious Spell,” 177–180, pointed out more Old Javanese inscriptions containing the same mantra the and its occurrence in the Gūhyasamāja-tantra, a tantric Buddhist text. For the most recent work on the implications of the Ratu Boko mantra, see Andrea Acri, “Once More on the Ratu Boko Mantra: Magic, Realpolitik, and Bauddha-Śaiva Dynamics in Ancient Nusantara,” in Esoteric Buddhism in Mediaeval Maritime Asia, Networks of Masters, Texts, Icons, ed. Andrea Acri (Singapore: ISEAS, 2016), 85. 174 FIGURE 2. Model of Borobudur stūpa kept at the site museum. FIGURE 3. Kesariya east elevation with brick niches housing life-size Amitābha and Akṣobya Buddhas. FIGURE 4. Borobudur east elevation with stone niches housing lifesize Akṣobhya Buddhas. Pacific World, 3rd ser., no. 20 (2018) FIGURE 1. Aerial view of Kesariya stūpa. Photo courtesy Yves Guichand. Chemburkar: Stūpa to Maṇḍala 175 Spiti Valley, India. It weighs similarities among the three monuments and reflects on whether a particular type of architectural form, which had its origin in the eighth century, was circulated and enhanced by the cross-cultural exchanges of religious teachers. Comparative study of Kesariya, Borobudur, and Tabo presents a body of evidence in support of inter-Asian connections. These sites reflect a consistent pattern of religious, cultural, and ritual ideas that defy geographical boundaries, suggesting a need for scholarship to examine the architectural and compositional interactions between South and Southeast Asia and comparative analysis of architectural models that have possibly a common textual and ritual basis. COMPARING KESARIYA AND BOROBUDUR Based on the overall measurements and the architecture of the two stūpas, K. K. Muhammed compares the structure of Kesariya to Borobudur.18 Mark Long also observes the similarities and differences between the two structures.19 The aerial photographs of the huge brick structure at Kesariya have a distinct, almost circular maṇḍala form resembling the rather more squared terraces of Borobudur (see figs. 1–2). Kesariya’s terraces, with large external buddhas in niches, have no known precedent as far as I am aware and are a marked departure from the smooth hemispherical stūpas of Sanchi, Bhahrut, and Amaravati. Six half-excavated concentric terraces of Kesariya, beneath what was originally a high and bulbous stūpa, are built on a natural hill, like Borobudur. The four lower terraces of Kesariya are more circular than those of Borobudur, but close examination reveals the upper two terraces to be square—something like an inverted combination of the square and circular terraces found on Borobudur. Like Borobudur, Kesariya’s design combines three elements: natural hill, stūpa, and maṇḍala. Both monuments present themselves to the viewer as horizontally somewhat flattened. Anyone standing at the base of either monument cannot see the crowning stūpa. Much like the stūpa of Borobudur, Kesariya has rows of chambers on each terrace at regular 18. K. K. Muhammed, “Evolution of Terraced Stupa in India with Special Emphasis on Kesariya,” unpublished paper presented at the Allahabad Conference in 2005. 19. Voûte and Long, Borobudur: Pyramid of the Cosmic Buddha, 187–191. 176 Pacific World, 3rd ser., no. 20 (2018) FIGURE 5. Kesariya stūpa: probable arrangement of buddhas in the exposed and restored brick chambers. Only basic dimensions are provided in the drawing. intervals holding a life-size buddha statue (see figs. 3–4). Above the fifth terrace the stūpa rises to a height of 9.38 m and is 22 m in diameter. The exposed terraced structure of the monument is 123 m in diameter and 37.5 m in height.20 The dimensions of Borobudur are almost the same. On the top fifth terrace of Kesariya, just below the stūpa, there are four single brick chambers facing the cardinal directions establishing a fourfold overall structure of the monument.21 The chamber on the eastern side contains an image in the bhūmisparśamudrā of Akṣobhya Buddha. Given the damage and the only partial excavation of the monument, it is at present impossible to determine the identity of the 20. Indian Archaeology: A Review 1999–2000 (New Delhi: Archaeological Survey of India, 2005), 11. 21. Ibid., 17, 19. Chemburkar: Stūpa to Maṇḍala 177 images in the other three chambers. The highest level of Borobudur, the top three almost circular terraces, houses 72 buddhas (16+24+32) in small, latticed stūpikas seated in dharmacakra-mudrā. The fourth terrace of Kesariya has triple chambers facing the cardinal directions, and the lower three terraces have in addition triple brick chambers facing the sub-cardinal directions. All the chambers have a raised platform to house a buddha image. The entire monument from the fifth terrace to the lower-most terrace would have housed 32 (4+4+8+8+8) brick chambers and would have once contained 88 (4x1+4x3+8x3+8x3+8x3) buddha statues.22 Figure 5 shows the buddhas from the top level of the monument to the bottom level, based on the ASI report of 1999–2000. It assumes symmetry in the unexcavated sections. The excavated chambers at Kesariya show a combination of statues in bhūmisparśa- (of Akṣobhya) and dhyāni-mudrā (of Amitābha) on the same side of the stūpa, whereas Borobudur houses 108 images of the Four Jinas, displaying their respective mudrās on four sides of the monument. The total number of buddhas in the niches at Borobudur is much more than Kesariya, but both monuments generate number grids and circular arrangements of buddha figures in their architecture, indicating the presence of the yoginī-tantras (possibly in a nascent stage) that Davidson sees and a shift in the design of stūpas. Only the upper two terraces of Kesariya are connected by a staircase (80 cm wide), concealed in the southwest corner within the polygonal designs between the chambers.23 Since the excavations are not yet complete, it is difficult to determine the number and exact nature of the staircase(s). Borobudur is connected from the ground level to the topmost stūpa by a set of four staircases, rising from the middle of each side. The circumambulatory paths on all the terraces at Kesariya are today devoid of reliefs, but there is enough space to have housed them. Whether there were any narratives in stucco, plaster, or paint 22. The topmost level has a single chamber in all four cardinal directions, containing an image of Buddha in each chamber (4x1=4). The fourth floor terrace has four chambers facing the four cardinal directions and each chamber has three compartments, thus containing 4x3=12 images. The lower three terraces have eight chambers facing the cardinal and sub-cardinal directions. Each chamber has three compartments housing (8x3) 24 images. The total number of buddha statues is therefore 88 (4+12+24+24+24). 23. Indian Archaeology: A Review 2000–01 (New Delhi: ASI, 2006), plate no. 8. 178 Pacific World, 3rd ser., no. 20 (2018) is impossible to determine from the present archaeological evidence. Borobudur is of course renowned for its kilometers of carved stone reliefs, which were presumably plastered and painted. At Kesariya there are three brick chambers on the eastern side (as seen in fig. 5) beyond the base of the lowest terrace and rammed earth base. Due to the incomplete excavation, it is not yet possible to ascertain whether they were part of the stūpa structure, but their alignment and size suggests they were. They seem to be a later addition to the main structure and may indicate another terrace below the lowermost terrace, positioned somewhat like the hidden foot of Borobudur. This hypothesis can only be tested by further excavation. The excavators have unearthed a number of finely carved bricks with geometrical patterns and kīrtimukhas (faces of glory); tiles; vases; and many small, red earthenware ritual pots with lids, spouts, and sprinkler heads that are presumed to have been used in consecrations. The scale of Kesariya seems to imply that it was part of a large ceremonial center, but its relationship to a dynastic center is so far unknown. The ruined structures around Kesariya suggest it was part of a vihāra or a temple monastery,24 where senior monks would have performed daily rituals. Borobudur is aligned with a small fire ritual temple called Caṇḍi Pawon and the regal Caṇḍi Mendut, forming the monumental state ceremonial center of the Śailendra Kingdom; it extended over 3 kms and presumably was situated at the center of a large city.25 Archaeological 24. See Alexander Cunningham, Four Reports Made during the Years 1862–63–64– 65 (Government Central Press, 1871; repr., New Delhi: ASI, 2000), 67 and plate XXIII. 25. Theodoor Van Erp was the first person to recognize the significance of the alignment of the three structures; see “Eenige mededeelingen betreffende de beelden en fragmenten van Boroboedoer in 1896 geschonken aan Z. M. den Koning van Siam,” Bijdragen tot de Taal-, Land- en Volkenkunde van NederlandschIndië 73 (1917): 285–310a. N. J. Krom believed that the three temples would have functioned as a part of a single plan (Barabuḍur: Archaeological Description, vol. 1 [The Hague: Martinus Nijhoff, 1927]); Paul Mus, Barabudur; esquisse d”une histoire du Bouddhisme fondée sur la critique archéologique de (Hanoi: Imprimerie d’Extrême-Orient, 1935), 418–420, talked about the ritual dependency of the three structures that J. L. Moens supported (“Barabadur, Mendut en Pawon en hun onderlinge samenhang,” Tijdschrift voor Indische Taal-, Landen Volkenkunde uitgegeven door het Bataviasch Genootschap van Kunsten en Chemburkar: Stūpa to Maṇḍala 179 finds made in a 5 km radius of Borobudur indicate a large monastic complex.26 BEGINNING OF A NEW STYLE IN STŪPA ARCHITECTURE Dating the Kesariya monument has hardly begun. The structure that is only partly visible today suggests that there were various stages of construction, and the sheer size implies that it was funded by royal resources at each stage.27 Xuanzang’s seventh-century account mentions a stūpa built in the area of Champāran, Bihar, where Licchavis of Vaiśālī took leave of the Buddha on his way to parinirvāṇa. Here the Buddha left his alms bowl as a memento for them. The record mentions the stūpa, possibly built in the location of Kesariya, as a memory of the event28 to be one of the principal Buddhist sanctuaries of the region and notes that the Buddhists referred to it as cakravartin stūpa—a monument that commemorates the abhiṣeka ceremony of a Wetenschappen 86 [1951]). See English trans. by Mark Long, “Barabudur, Mendut and Pawon and Their Mutual Relationship,” Tijdschrift voor de Indische Taal-, Land- en Volkenkunde (2007): 7, 8, 67. It was also supported by Lokesh Chandra, “Borobudur as a Monument of Esoteric Buddhism,” The Southeast Asian Review 5, no. 1 (August 1980): 35–36; and Voûte and Long, Pyramid of the Cosmic Buddha, 98–99. 26. That Borobudur was a vihāra is attested in the Karangtenah inscription of 824 CE. See line 15: āstāṁ vihārah, in J. G. de Casparis, Inscripties uit de Çailendratijd (Bandung: A.C. Nix, 1950), 40. Based on M. Boechari, “Preliminary Report on Some Archaeological Finds around the Borobudur Temple,” in Pelita Borobudur. Reports and Documents of the Consultative Committee for the Safeguarding of Borobudur. 5th Meeting April 1976 (Jakarta: Departemen Pendidikan dan Kebudayaan Republik Indonesia, 1982), 90–95. John Miksic writes about the monastic complex placed next to the monument in Borobudur: Golden Tales of the Buddhas (Singapore: Periplus, 1991), 34–35. A. J. Kempers, Ancient Indonesian Art (Cambridge: Harvard University Press, 1959), 45 has written about the remains of the vihāra to the northwest of the monument. 27. The structure clearly shows two phases of construction activity; see “Sunga/Kushana and Late Gupta Period” [late seventh, early eighth century], Indian Archaeology: A Review—1998–99 (New Delhi: ASI, 2004), 11. In a telephone conversation on January, 16, 2014, Dr. K. K. Muhammed stated that the slopes are strewn with late Gupta period bricks or may be even later bricks. 28. Cf. Cunningham, Four Reports, 66. 180 Pacific World, 3rd ser., no. 20 (2018) Buddhist king-of-kings.29 The Licchavi stūpa was possibly expanded by king Harṣa (ca. 606–647) at some stage, the first great post-Gupta king in the region.30 He patronized several monastic buildings along with thousands of stūpas, each over 100 feet high.31 Gupta and late Gupta period bricks from the seventh century were found on the slopes of the Kesariya stūpa.32 Harṣa was the first Indian king to cement ties with the Tang court of China, notably through his personal friendship with the well-connected Xuanzang. After ruling from Kanauj (Uttar Pradesh) for decades, he moved his capital to Magādha in 641 CE and announced the event by sending a delegation to the Tang court in China.33 In re29. Xuanzang describes a stūpa built at approximately 200 li to the northeast of Vaiśalī that Cunningham identifies with Kesariya (Cunningham, Four Reports, 65–66). Xuanzang writes: “In the city there is a stupa at the place where Buddha had told an assembly of various Bodhisattvas and men and heavenly beings about his past events of cultivating Bodhisattva deeds. He was once a universal monarch [cakravartin] named Mahādeva (known as Datian or great city in Chinese), in this city, possessing the seven treasures and being competent to rule over the four continents of the world.” See Xuanzang’s The Great Tang Dynasty Record of the Western Regions, trans. Li Rongxi (Berkeley: Numata Center for Buddhist Translation and Research, 1995), 214; and Thomas Watters, On Yuan Chwang’s Travels in India, 629–645 A.D. (London: Royal Asiatic Society, 1905), II:71–72. 30. This is my hypothesis based on the ASI findings of the post-Gupta period bricks at the site. The sheer scale of the monument wouldn’t have been possible without royal funding. Champāran was part of Harṣa’s vast kingdom. See Chemburkar, “Borobudur’s Pāla Forbear?” 31. See Watters, On Yuan Chwang’s Travels, 164; and Li Rongxi, Great Tang Dynasty Record of the Western Regions, fascicle V:144. Even though Xuanzang mentions Harṣa’s building activity, the only architectural evidence from his reign may be sought at Nālandā. The archaeological remains of Nālandā date from the fifth century CE to the end of twelfth century CE, and during Harṣa’s reign the monastery and university were certainly at the height of their fame. 32. Based on the findings during the excavations and the size and the nature of the bricks, ASI has tentatively dated the structure to the late Gupta period. Indian Archaeology: A Review 1998–99, 11. 33. Based on her understanding of the Chinese sources, Devahuti mentions that Harṣa was the king of Kannauj for a long time, but by the time the Chinese mission arrived in 641 CE, he had already claimed the throne of Magādha. See D. Devahuti, Harsha: A Political Study (Oxford: Oxford University Press, 1970), 84, 214, 217; based on his readings of the Xin Tang shu 221a (New History of the Tang [Beijing: Zhonghua shuju, 1975], 6237). Tansen Sen (“In Search of Chemburkar: Stūpa to Maṇḍala 181 sponse, the court dispatched an embassy in 643 CE,34 presumably to attend his Buddhist cakravartin ceremony. Did Kesariya play a part in this ceremony? The site remained active in later centuries: The recent excavations by the Archaeological Survey of India at this site have discovered a Pāla period stūpa dating from the eighth century. The excavations have revealed the terraces of the stūpa, with “Prādakshīnā Path,” which follows the pattern of those reported from Pahārpur in East Bengal and Nandangarh [in east Champāran]. The stūpa has been found with several [life-size] stucco figures of Lord Buddha in Bhumīsparśā posture in the cells provided all over the terraces.35 A late Pāla period structure was added to the stūpa summit in the eighth century, but the exact nature of the construction is as yet very difficult to determine.36 The Pālas inherited the territory that was previously ruled by Harṣa and the later Guptas.37 Champāran, the site of the Kesariya stūpa, played a significant role under the Pālas, where massive stūpa sites such as Lauriya Nandangarh, Lauriya Areraj, Bettiah, Rāmpurva, and Pipariya were constructed.38 The region has a key position on the royal road from Pataliputra (Patna) to Nepal and has produced a huge number of Pāla period images. The arrangement of a crowning stūpa over a fourfold symmetry at Kesariya along with the radiating chapels housing buddha images is in line with features that were developed later during the Pāla period.39 The heartland of the Pālas in northeast India became the most sigLongevity and Good Karma: Chinese Diplomatic Missions to Middle India in the Seventh Century,” Journal of World History 12, no. 1 [2001]: 7) concludes the same. 34. Sen, “Search of Longevity and Good Karma,” 8. 35. Dilip Chakarabarti, Archaeological Geography of the Ganga Plain. The Lower and the Middle Ganga (Delhi: Permanent Black, 2001), 203. 36. Ibid., 206. 37. Fredrick Asher, The Art of Eastern India: 300–800 (Minneapolis: University of Minnesota, 1980), 69. 38. Study of the construction of these massive stūpa sites, along with Kesariya, awaits excavation. 39. John and Susan Huntington, Leaves from the Bodhi Tree: The Art of Pāla India (8th–12th Centuries) and Its International Legacy (Seattle: Dayton Art Institute, 1990), 90–91. Claudine Bautze-Picron (p. 283) supports this in her review: 182 Pacific World, 3rd ser., no. 20 (2018) nificant international center of Buddhist learning and was the major source of teachers, authoritative texts, and Buddhist iconography.40 Apart from its soteriological religious function, the Buddhist temple or stūpa in this period became a political statement. Kesariya, with its new stūpa-maṇḍala model design, marks a crucial post-Gupta and pre-Pāla shift in the Buddhist monumental architecture according to my judgment. What was the maṇḍala model? MAṆḌALA MODEL: TEXT, RITUAL, KINGSHIP, AND POLITICS New forms normally arise in religious architecture when there are significant changes in belief and/or ritual. The architecture of Kesariya resembles a Buddhist maṇḍala that we see on many Buddhist thangkas, although the specific maṇḍala cannot yet be determined. This new stūpa-maṇḍala model was then spread in the Pāla domain to the contemporary monasteries of Uddanḍapura (Odantapurī) and Vikramśīla in South Bihar, of Somapura Lālmai and Maināmatī in present day Bangladesh, and other Buddhist sites in Odisha. The central structures of these monasteries share a cruciform plan, crowned with a stūpa or a temple, and rising stepped terrraces. Archaeological research has unearthed several monuments with similar plans in Bihar and Bengal41 showing an identical arrangement of sacred space that could have “As the author emphasizes, a special feature of the architecture was then the niches on the outside walls of the temple. Those niches were occupied by the sculptures as we know from temple 2 at Nālandā, still adorned with stone panels, or from the Maniyar matha at Rajgir or the temple at Aphsad where stucco images used to adorn the niches.” See Claudine Bautze-Picron, “Crying Leaves: Some Remarks on ‘The Art of Pāla India (8th–12th centuries) and Its International Legacy,’ ” East and West 43, no. 1/4 (1993): 277–294. 40. Huntington and Huntington, Leaves from the Bodhi Tree, 70, 84–85. 41. Abu Imam mentions that further cruciform temples “in the 7th–8th century time bracket” have been discovered in recent excavations at Savar near Dhaka. Maināmatī monasteries (the Salbān, Bhojā, Aṇandā, and Rupbān vihāras) in Comilla district in Bangladesh show an identical cruciform structure at the center of the temple. Some of these monuments display a central temple instead of a crowning stūpa, possibly to take care of the expanding ritual systems. See Abu Imam, Excavations at Mainamati: An Exploratory Study, Studies in Bengal Art Series 2 (Dhaka, Bangladesh: International Centre for Study of Bengal Art, 2000), 133. Chemburkar: Stūpa to Maṇḍala 183 FIGURE 6a. Central cruciform structure of Kesariya on the topmost level. FIGURE 6b. Vikrama-śīla Vihāra, Antichak. End of eighth century. Adapted from B. K. Jamuar, The Ancient Temples of Bihar (New Delhi: Ramanand Vidya Bhawan, 1985), 87. FIGURE 6d. Caṇḍi Sewu central shrine. FIGURE 6c. Rupban Mura Vihāra, Maināmatī. End of eighth century. Adapted from Abu Imam, Excavations at Mainamati: An Exploratory Study, Studies in Bengal Art Series 2 (Dhaka, Bangladesh: International Centre for Study of Bengal Art, 2000), 66. FIGURE 6e. Caṇḍi Kalasan central shrine. FIGURE 6. Pāla and Śailendra monuments displaying the fivefold central structure and identical space arrangement. (Drawings are not to scale.) 184 Pacific World, 3rd ser., no. 20 (2018) served parallel functions in Borobudur, Sewu, Kalasan, Lumbung, Bubrah, and Plaosan in the Śailendra domain (see figs. 6b–6e).42 Text Adding to the fourfold structure of these monuments a central buddha, this yields the fivefold structure of the Five Jina Buddhas found in the seminal Yogatantra text Sarvatathāgatatattvasaṅgraha.43 In the Sarvatathāgatatattvasaṅgraha, the five buddha family scheme becomes a dominant structure after Vairocana consecrates himself as a buddha. He then draws in a number of personages, beginning with Samantabhadra, who is crowned and consecrated with the name Vajrapāṇi. Later, the other thirty-six figures of the maṇḍala are consecrated with names conferred on them by Vairocana, before they are positioned in the maṇḍala.44 Certain numerical configurations occur in the late eighth-century text Saṃvarodaya-tantra, describing the course of the moon and the sun with respect to the astronomical body and the human body. A tendency to identify the individual with the universal and the internal or corporeal with the global or cosmic through the medium of their qualitative and structural similarities is noticeable in this text. Ultimate reality, which is attained through the human body, is then identified with the universe and the maṇḍala deities of the text.45 The number of terraces and the buddha groupings seen at Kesariya (4+12+24) and at Borobudur (16+24+36) might be suggestive of this textual source.46 These texts contain explicit references to divine kingship. 42. Leeuw, “South-East Asian Architecture,” 297–401; Geoffrey Samuel, “Ritual Technologies and the State: The Mandala-Form Buddhist Temples of Bangladesh,” The Journal of Bengal Art 7 (2002): 39–56. 43. David Snellgrove, Indo-Tibetan Buddhism: Indian Buddhist and Their Tibetan Successors, 2 vols. (London and Boston: Serindia, 1987), 175, 189, 198. 44. Ibid., 8, 203. 45. Tsuda Shinichi, “Saṃvarodayatantra: Selected Chapters” (PhD diss., Australian National University, 1970), 1, 66, 231. 46. Hiram Woodward (“Review: Esoteric Buddhism in Southeast Asia in the Light of Recent Scholarship,” 343, 346) proposed that the numerology of the cakra system in the Saṃvarodaya-tantra might be connected with that of the three circular terraces at Borobudur, leaving open the question of which system had chronological priority. In a later article (“Bianhong: Mastermind Chemburkar: Stūpa to Maṇḍala FIGURE 7. Diagram of the Vajra Realm Maṇḍala, P. 4518(33), originally from Dunhuang, Gansu Province, China, tenth century, 17x12in. Kept at Bibliothèque Nationale de France. Ink and light colors on paper. From Michelle C. Wang, “Changing Conceptions of ‘Maṇḍala’ in Tang China: Ritual and the Role of Images,” Material Culture 9, no. 2 (2013): 202. © Bibliothèque Nationale de France. 185 186 Pacific World, 3rd ser., no. 20 (2018) FIGURE 8. Vajradhātu Maṇḍala of basic thirty-seven deities according to Sarvatathāgatatattvasaṅgraha. Structure adapted from Do-Kyun Kwon, “Sarva Tathāgata Tattva Saṃgraha: Compendium of All the Tathāgatas” (PhD diss., School of Oriental and African Studies, London University, 2002). Ritual and Kingship David Snellgrove establishes intimate connections between maṇḍala, kingship, abhiṣeka ritual, and Vairocana as the cakravartin buddha in Vajrayāna Buddhism.47 The hallmark aspect of esoteric Buddhism is the representation of maṇḍalas in various media, especially in paintings that depict the universe as a perfectly ordered and harmonious system where enlightenment can be attained. The most usual representations of maṇḍala paintings comprise formations of buddhas, bodhisattvas, and associated guardian deities or symbolic forms positioned of Borobudur?” Pacific World, 3rd ser., no. 11 [2009]), he argued that an alphabet diagram (prastara) lay behind both systems. 47. David Snellgrove, “The Notion of Divine Kingship in Tanric Buddhism,” in La Regalità Sacra- Contributi al Tema dell’ VIII Congresso Internazionale di Storia delle Religioni (Leiden: E. J. Brill, 1959), 206. Chemburkar: Stūpa to Maṇḍala 187 in circles and squares around a central buddha in a certain hierarchy as mentioned in texts such as the Sarvatathāgatatattvasaṅgraha. Several ninth- to tenth-century maṇḍala drawings from Dunhuang, China kept at the museums represent this structure along with the ritual implements, highlighting their ritual significance (fig. 7).48 The most important ritual performed with a maṇḍala is abhiṣeka. During the abhiṣeka a lustration vessel is placed at the center of a maṇḍala, which is a visual representation of a sanctified place or a perfect universe. The properties of the buddhas and bodhisattvas of the maṇḍala are understood to gather into the water of the lustration vessel. When anointed with this water, the monarch would acquire all the powers embodied in the central deity of that maṇḍala to become cakravartin or earthly ruler. He would then be able to exercise the powers of the central buddha, whether mundane (e.g., producing rain) or supramundane (e.g., deepening one’s store of wisdom and compassion), and be responsible for the spiritual as well as the temporal wellbeing of his geographical maṇḍala or the kingdom.49 Detailed accounts of abhiṣeka rituals are given in the Sarvatathāgatatattvasaṅgraha.50 By the early eighth century, we gain a sense of increasing importance of the maṇḍala consecration and a systematic metaphorical association with the kingship. Constructing a Ritual-Political Center The vocabulary used to read these painted maṇḍalas is related to the construction of a palace and not a temple. Several terms assigned to the residences of the maṇḍala divinities are exactly those employed 48. See 9th–10th Century Maṇḍala Ritual Drawing from Dunhuang at National Museum, New Delhi (Ch00379), at Musée Guimet, Paris (PC 2012), at British Museum (1919,0101,0.174). 49. Snellgrove, “The Notion of Divine Kingship in Tanric Buddhism,” 208. 50. Do-Kyun Kwon, “Sarva Tathāgata Tattva Saṃgraha: Compendium of All the Tathāgatas” (PhD diss., School of Oriental and African Studies, London University, 2002), and Steven Neal Weinberger, “The Significance of Yoga Tantra and the Compendium of Principles (Tattvasaṃgraha Tantra)” (PhD diss., University of Virginia, 2003); cf. Ryuichi Abe, The Weaving of Mantra: Kūkai and the Construction of Esoteric Buddhist Discourse (New York: Columbia University Press, 2013), 133–149. 188 Pacific World, 3rd ser., no. 20 (2018) for palaces and pavilions in medieval architectural manuals with the identical architectural terminology suggesting their intended construction.51 The Vajradhātu Maṇḍala that was first described in the Sarvatathāgatatattvasaṅgraha52 found its way into the architecture as a concrete arrangement of deities, on a basic fivefold or a ninefold model. Akṣobhya and his attendants in the east, Ratnasambhava in the south, Amitābha in the west, and Amoghasiddhi in the north made up a mandalic arrangement around Vairocana or Mahāvairocana (fig. 8, above). This pentad and the attendant deities demarcating respective buddha-fields and one thousand buddhas of Bhadrakālpa found prominent places in architecture.53 51. See Bruno Dagens, trans., Mayamatam: Treatise of Housing Architecture and Iconography, 2 vols. (New Delhi: Indira Gandhi International Centre for Arts, 2007), 119, 148, 176, 203. For example, the central buddha resides in the pavilion called kūṭāgāra—not the garbhagṛha with its four entrances dominated by arched gateways (toranas) and not the assembly halls or jangha in the shape of scepters (vajra) and guarded by an adamantine wall (vajrapanjara). Harṣacarita uses the term vajrapanjara (a cage or a citadel) as a metaphor in its identification of Harṣa’s body with specific parts of the citadel, clearly indicating the relationship between esoteric Buddhism and imperial metaphor that Snellgrove discusses in his “The Notion of Divine Kingship,” 204–218. See Harṣacharita by Bāṇabhaṭṭa, Uchchvāsas I–VIII, ed. with an Introduction and notes by P. V. Kane (Bombay, 1918), 33–34. 52. There are two extant Sanskrit manuscripts of Sarvatathāgatatattvasaṅgraha from Nepal. Guiseppe Tucci obtained a nineteenth-century manuscript of the tantra, and in 1956 David Snellgrove and John Brough discovered an Indian palm-leaf manuscript that they identified as a ninth- or tenth-century work from Bihar, India. David Snellgrove and Lokesh Chandra (Sarva-tathāgatatattva-saṇgraha: Facsimile Reproduction of Tenth Century Sanskrit Manuscript from Nepal [New Delhi: Sharada Rani, 1981]) published a photographic reproduction of this manuscript; Do-Kyun Kwon (“Sarva Tathāgata Tattva Saṃgraha: Compendium of All the Tathāgatas,” 22, 28, 29) and Steven Neal Weinberger (“The Significance of Yoga Tantra and the Compendium of Principles [Tattvasaṃgraha Tantra],” 47, 61, 62, 72, 73) have described the formation of the Vajradhātu Maṇḍala in the Sarvatathāgatatattvasaṅgraha in the light of its Indian, Chinese, and Tibetan commentaries. 53. For a detailed description of the maṇḍalas see Adrian Snodgrass, The Matrix and Diamond World Mandalas in Shingon Buddhism, vols. 1 & 2 (New Delhi: Aditya Prakashan, 1988), 634; for their use in the architecture of Caṇḍi Sewu, Mendut, and Borobudur, see Chandra, “Borobudur as a Monument of Esoteric Chemburkar: Stūpa to Maṇḍala 189 Brajdulal Chattopadhyaya sees these textual developments of the maṇḍala with the strong center and subsidiary sets in relation to the hierarchical structure of “samānta feudalism” of mediaeval India.54 The idea of a maṇḍala with the central figure representing a supreme deity and directional figures as subordinate deities reflects the idea of the supreme king at the center, surrounded by vassals who are expected to exercise power as local landlords rather than independent rulers. The nature of a maṇḍala is therefore to map the social and political interests and designate levels of hierarchy. Ronald Davidson argues that “the central and defining metaphor for mature esoteric Buddhism is that of an individual assuming kingship and exercising dominion … through a combination of ritual and metaphysical means, thereby becoming a supreme overlord (buddha) or universal ruler (cakravartin).”55 These textual developments in Buddhism served the interests of imperial figures in organizing political and social landscapes with the assistance of their spiritual advisors. Architecture made a key contribution to the ceremonial or ritual center in these developments. The terraced architectural design of Kesariya and Borobudur, along with the arrangement of buddha statues, clearly displays the hierarchical organization of a maṇḍala structure. The question we must ask is how these ideas concerning architectural forms circulated between India and Indonesia and whether they formed a part of a shared culture in the connected Buddhist world.56 WIDE WEB OF BUDDHIST MONKS AND EXCHANGE OF IDEAS IN PĀLA AND ŚAILENDRA DOMAINS Nālandā prospered even after the chaotic period, which had begun with the death of Śaśānka (628 CE) and Harṣa (647 CE). The pro-Buddhist Pāla dynasty came to power in the late eighth century when Gopāla Buddhism,” 8; for Sewu, see F. D. K. Bosch, “Buddhist Data from Balinese Texts and Their Contribution to Archaeological Research in Java,” in The Selected Studies in Indonesian Archaeology, English trans. (The Hague: Martinus Nijhoff, 1961), 111. 54. Brajadulal Chattopadhyaya, The Making of Early Medieval India (Delhi: Oxford University Press, 1994). 55. Davidson, Indian Esoteric Buddhism, 121. 56. Woodward (“Review: Esoteric Buddhism in Southeast Asia in the Light of Recent Scholarship,” 353) has already advanced an argument for “treating Indonesia and India as an integrated unit well into the ninth century.” 190 Pacific World, 3rd ser., no. 20 (2018) (ca. 750–775) gained control of northeastern India and established the Odantapuri Monastery at the new city of Odantapuri, some 10 km from Nālandā. The political and military ambition of his son Dharmapāla (ca. 775–812 CE) was matched by unprecedented generosity to Buddhist establishments that provided a platform for generating texts, sacred art, and architecture. He sponsored Vikramaśīla, in present-day Bhagalpur in Bihar, and fifty other monasteries.57 Many of their successors, including Devapāla (ca. 812–850), Mahipāla (ca. 992–1042), and Ramapāla (ca. 1087–1141), patronized the chain of the large principal monasteries at Vikramaśīla, Nālandā, Pahārpur, Jaggadala, and Odantapuri.58 These monasteries were not just powerful, well supported, and interconnected, but were also influential in setting the institutional format for the Buddhist monasteries. The cosmopolitan allure of Nālandā is evident in the temples built from the eighth to tenth centuries bearing the name of Nālandā just north of Kandy in Sri Lanka. Nālandā played a major role in the transmission of artistic motifs to Southeast Asia since the eighth century and continued to be an inspiration.59 With Pāla patronage, Buddhism blossomed again, attracting monks from all over the world to these monasteries. The prominent Indian and Chinese travelers in this period played a crucial role in transmitting new religious thoughts. There is much evidence that they spread the texts and ritual techniques and introduced iconographic and stylistic forms, which were to merge with the local artistic idioms.60 57. Puspa Niyogi, Buddhism in Ancient Bengal (Calcutta: Jinasa, 1980), 102; for Buddhist monasteries under the Pālas, see Sukumar Dutt, Buddhist Monks and Monasteries of India (Delhi: Motilal Barnasidass, 1988), 344–66. 58. The timeline for Pāla kings is adopted from Huntington and Huntington, Leaves from the Bodhi Tree, 542, chart 1. 59. Bernet-Kempers, The Bronzes of Nalanda and Hindu-Javanese Art (Leiden: Brill, 1933); Pauline Scheurleer and Marijke Klokke, Divine Bronze: Ancient Indonesian Bronzes from A.D. 600 to 1600 (Leiden: E. J. Brill, 1988). Peter Skilling (“King, Sangha, and Brahmans: Ideology, Ritual, and Power in Pre-Modern Siam,” in Buddhism, Power and Political Order, ed. Ian Harris [London & New York: Routledge, 2007], 97) has argued for Nālandā style imagery re-appearing in eleventh- to twelfth-century Angkor, Bagan, the Malay Peninsula, and east Java. 60. Edward Schafer, The Golden Peaches of Samarkand: A Study of T’ang Exotics (Berkeley: University of California Press, 1963), 268, points out that “a prime objective of Chinese pilgrims in the holy lands of the Indies was the acquisition Chemburkar: Stūpa to Maṇḍala 191 FIGURE 9. Buddhist shrine at Seattle. Storeyed Pyramidal Monument in Miniature, ivory, 4.5 in, 33 oz. Early Pāla period, ninth to tenth century, Seattle Art Museum, Eugene Fuller Memorial Collection, accession no. 48.166. Taken from Margaret F. Marcus, “Sculptures from Bihar and Bengal,” The Bulletin of the Cleveland Museum of Art 54, no. 8 (Oct. 1967): 240– 262. © Seattle Art Museum. The delegation that attended King Harṣa’s Buddhist ceremony in 643 CE visited Rājagṛha (Rājgīr) and the Mahābodhi complex in Bodhgaya, where the artist Song Fazhi made drawings of Buddhist of holy statues, and images to edify the faithful at home and adorn the rich temples of T’ang.” 192 Pacific World, 3rd ser., no. 20 (2018) architecture and artefacts to carry to the Tang court.61 A model of the Nālandā monastery, an image of the Mahābodhi shrine, and other Buddhist illustrations were also taken to China at this time by the monk Huilun. Many images of the popular Pāla Buddha in the earth-touching mudrā and seated atop a throne were made in Tang China. Buddhist monk Divākara (612–687) studied at the Mahābodhi temple, went to China, and integrated the iconography associated with the images into rituals at the Tang court.62 The biography of Japanese monk Ennin notes how the five esoteric buddha images of the Jinge monastery on Mount Wutai were modeled after the Nālandā images by Amoghavajra in the eighth century.63 The Chinese monk Jiye visited India from 966 to 976 and recorded the truly international character of Nālandā,64 which seems to have remained active up to the thirteenth century. There were certain architectural and sculptural models circulating across the seaways. The Seattle Art Museum has an interesting small, Pāla period, ivory object (fig. 9, above). It could have been used in private rituals, but it could also have been an architectural model. It represents in miniature a monument composed of four levels with Buddhist figures oriented to the four quarters along with supporting figures in the niches. The crowning member is missing, but Dr. Lee suggests that to complete the cosmological formula, this may have been a position of the supreme Buddha Vairocana.65 From an architectural point of view, 61. Sen quotes “a painting of Maitreya drawn in India by Song Fazhi [that] seems to have used as a blue print for a sculpture at the Jing’ai monastery in Luoyang” from Lidai minghua ji (Records of the Famous Painters of All the Dynasties) authored by Tang dynasty scholar critic Zhang Yanyuan in 847 CE. See Sen, “Search of Longevity and Good Karma,” 9. For the exchange of architectural ideas, see Ernst Boerschmann, Baukunst und die Religiöse Kultur der Chineses (Architecture and Religious Culture of the Chinese), vol. 3, part 1, Pagoden Pao Tà (Berlin, Leipzig: verlag Walter De Gruyter and Co., 1931). 62. Michelle C. Wang, “Changing Conceptions of ‘Maṇḍala’ in Tang China: Ritual and the Role of Images,” Material Religion 9, no. 2 (2013): 198. 63. Ennin’s Diary: The Record of a Pilgrimage to China in Search of the Law, trans. by Edwin Reischauer (New York: Ronald Press, 1955), 253. 64. Jiye Xiyu xingcheng, T. 2089: 982a.2–b.5, in E. Huber, “L’itinéraire du pélerin Ki ye dans l’inde,” Bulletin de l’École française d’Extrême-Orient 2 (1902): 256–259; É. Chavannes, Les inscriptions chinoises de Bodh-gaya (Paris: Ernest Leroux, 1896). 65. Sherman Lee, “An Early Pāla Ivory,” Journal of the Indian Society of Oriental Art 17 (1949): 1–5. Chemburkar: Stūpa to Maṇḍala 193 this place would have been reserved for the stūpa or a kūṭāgāra. There were quite a few portable shrines circulating in the Buddhist world during the eighth to twelfth centuries.66 John Guy has traced around twenty late Pāla-Sena period architectural models of the Mahābodhi temple that were dispersed from eastern India to Nepal, Tibet, Arakan, and Myanmar, indicating continuity in Buddhist travels.67 Buddhism became a bridge that fostered dialogue between the Chinese and Indian courts. Prominent monks in this period played a crucial role in transmitting new religious developments.68 This twoway sea traffic of monks and pilgrims interacting with each other was part of a single symbolic language69 in which the Śailendras played a part as cultural brokers.70 By the ninth century, the shoreline of the Bay of Bengal, nourished by its river networks, had acquired a vibrant new commercial identity.71 Strong links, mostly Buddhist, provided connections with eastern India, Java, and Sumatra. Indonesia’s Śrīvijayan port at Palembang, Sumatra, became a center of Sanskrit language study for monks travelling to the sacred sites and institutions of India. Palembang lay halfway between India and the Chinese capital of Changan (Xian today), where international scholars congregated and consolidated the growing Buddhist network (see fig. 10). There are several indicators of the growing importance of Sumatra and Java. The monk Śubhākarasiṃha from Odisha (637– 735) arrived in Changan in 716, bringing paintings of the maṇḍalas 66. Phyllis Granoff, “A Portable Buddhist Shrine from Central Asia,” Archives of Asian Art 22 (University of Hawai’i Press, 1969), 80–95. 67. John Guy, “The Mahabodhi Temple: Pilgrim Souvenirs of Buddhist India,” The Burlington Magazine 133, no. 1059 (1991): 362–364. 68. The Chinese Buddhist imagery of the late Tang period also shows signs of increased interaction with northern and southern Indian art. See Marylin M. Rhie, Interrelationship between the Buddhist Art of China and the Art of India and Central Asia from 618–755 A.D. (Napoli: Istituto Universitario Orientale, 1988), 39–40. 69. A. H. N. Verwey, “A Distant Relative of the Silver Mañjuśrī from Ngemplak Semongan,” Mededelingen van het Rijksmuseum voor Volkenkunde 15 (1962): 141; Alessandra Lopez Royo, Prambanan; Sculpture and Dance in Ancient Java; A Study in Dance Iconography (Bangkok: White Lotus, 1998), 9. 70. Roy Jordaan, “The Śailendras, the Status of the Kṣatriya Theory, and the Development of the Hindu-Javanese Temple Architecture,” Bijdragen tot de Taal-, Land- en Volkenkunde 155, no. 2 (1999): 228. 71. Himanshu Prabha Ray, “The Archaeology of Bengal: Trading Networks, Cultural Identities,” Journal of Economic and Social History of the Orient 49, no. 1 (2006): 78. 194 Pacific World, 3rd ser., no. 20 (2018) FIGURE 10. Connected Buddhist world of India, China, and Indonesia during the eighth to twelfth centuries. Chemburkar: Stūpa to Maṇḍala 195 of the Sarvatathāgatatattvasaṅgraha to China.72 Vajrabodhi (671–741) from Kāñcī in southern India studied at Nālandā, sailed to the Malay Peninsula on his way to Sumatra and Java, and eventually reached China. Amoghavajra (705–74), who also became a patriarch of Chinese Buddhism, met Vajrabodhi in Java and accompanied him to Changan.73 Abhiṣeka became the defining feature of the rituals extracted from the major esoteric Buddhist texts translated by these three monks. The success of their magical powers in Chinese military operations is celebrated. During their time in China, in the middle of the eighth century, the ritual practices of Famen Monastery underwent significant changes. Han Jinke’s meticulous study of the excavated objects, many of them made of silver and gold, from the monastery illustrates their arrangement in the form of a maṇḍala, a concentric layout symbolizing the Buddhist universe.74 Mandalic diagrams or altars were employed by these most influential esoteric masters in some of the rituals to link a patron to the cosmic reality. Many engravings of the Famen Monastery crypt display esoteric imagery, and the objects found were 72. N. Iyanaga, “Récits de la soumission de Maheśvara par Trailokyavijaya d’après les sources chinoises et Japonaises,” in Tantric and Taoist Studies in Honour of R. A. Stein, vol. 3, ed. Michel Strickman (Brux-elles: Institue Belge des Hautes Etudes Chinoises, 1985), 724–725. 73. This generally accepted view is based on Yuanzhao’s biography as the most reliable source. Yi-Liang Chou, Tantrism in China (Cambridge: Harvard-Yenching Institute, 1945), 321. Jeffrey Sundberg and Rolf Giebel are in agreement with Chou over Java being the meeting place of Vajrabodhi and Amoghavajra. See “The Life of the Tang Court Monk Vajrabodhi as Chronicled by Lü Xiang: South Indian and Śrī Laṇkān Antecedents to the Arrival of the Buddhist Vajrayāna in Eighth-Century Java and China,” Pacific World: Journal of the Institute of Buddhist Studies, 3rd ser., no. 13 (2011): 148. However Woodward (“Review: Esoteric Buddhism in Southeast Asia,” 339) maintains that Amoghavajra never went to Java (on this trip) and never met Vajrabodhi there. He follows biographies of Amoghavajra by Zhao Quian (T. 2056) and Fexi (T. 2120). 74. “Most recent and detailed discussion of the esoteric influences on relic veneration at the Famen Monastery is Famensi digong tang mi mantuluo zhi yanjiu, ed. Wu Limin and Han Jinke (Hong Kong: Zhongguo fojiao wenhua chuban youxian gongsi, 1998),” Tansen Sen, Buddhism, Diplomacy and Trade: The Realignment of India-China Relations, 600–1400 (Honolulu: Association for Asian Studies and University of Hawai’i Press, 2003), 72. Patricia Eichenbaum has also discussed the issue in her work “Esoteric Buddhism and the Famensi Finds,” Archives of Asian Art 47 (1994): 78–85. 196 Pacific World, 3rd ser., no. 20 (2018) FIGURE 11. Kesariya plan showing the overall layout of the structure. Drawing by Swati Chemburkar. FIGURE 12. Borobudur plan. Drawing by Swati Chemburkar. Chemburkar: Stūpa to Maṇḍala 197 used in esoteric ceremonies. These monks became the most influential monks of Chinese esoteric Buddhism and made the major contribution of weaving esoteric Buddhist concepts through the increasingly connected international Buddhist world.75 The new emphasis of the yoginītantras and the circular arrangement of deities and mandalic architectural elements of Kesariya would have been part of this wide web of the Buddhist world. BOROBUDUR AND ŚAILENDRA BUDDHISM Like Kesariya, Borobudur constructs the fourfold buddha system along with the supreme Buddha Vairocana in its architecture. We don’t know which images were housed in the top four cardinal chambers of partly excavated Kesariya except Akṣobhya, but it is quite possible that it will eventually be shown by archaeologists to embody the fourfold buddha system. There are enough common elements in the architecture of both the monuments at present to indicate the use of a common theme (see figs. 11 and 12, above). Borobudur has been described as a stūpa, a multi-storied palace (prāsāda), Mount Meru, and a maṇḍala.76 Each of these descriptions is 75. Why else are there so many translations and explanatory texts of Sarvatathāgatatattvasaṅgraha? All the texts that got translated and transferred in the Buddhist world of India, China, Japan, and Indonesia were part of Sarvatathāgatatattvasaṅgraha. Across the Chinese, Japanese, and Tibetan traditions a number of variations of the Vajradhātu Maṇḍala based on Sarvatathāgatatattvasaṅgraha and its explanatory texts are known. Lokesh Chandra, “A Comparative Study of the Tibetan, Japanese, Indonesian and Khotanese Maṇḍala of the Tattva-samgraha,” Amala Prajna: Aspects of Buddhist Studies: Professor P.V. Bapat Felicitation Volume, ed. N. H. Samtani and H. S. Prasad (New Delhi: Sri Satguru Publications, 1989), 187–200. Although Vajrabodhi had begun translating the Sarvatathāgatatattvasaṅgraha into Chinese in 723 CE, the continuing importance of the Sarvatathāgatatattvasaṅgraha at the end of tenth century is signaled in the fact that the entire twenty-six chapters were translated into Chinese and re-translated in Tibetan. See Rob Linrothe, Ruthless Compassion: Wrathful Deities in Early Indo-Tibetan Esoteric Buddhist Art (Boston: Shambhala, 1999), 155. 76. For some arguments regarding the nature of Borobudur, see A. Foucher, “Notes d’archéologie bouddhique: I, Le stupa de Boro-Budur; II, Les bas-reliefs de Boro-Budur; III, Iconographie bouddhique à Java,” Bulletin de l’École française d’Extrême-Orient (1909): 4; A. J. Bernet Kempers, Borobudur mysteriegebeuren in steen; verval en restauratie; oudjavaans volksleven (Wassenaar Servire, 1970), 198 Pacific World, 3rd ser., no. 20 (2018) true, and scholars now agree on the multivalent nature of the monument. While agreeing on this multivalent nature, this paper has focused on the complex nature of stūpa-maṇḍala structure of the monument. If it is a stūpa, then as Woodward describes, “It is just as a souped-up, hoodless car with gleaming engine parts is an automobile.”77 If it is a maṇḍala as argued by several scholars, then which maṇḍala is a question that is still open for debate. Based on the certain geometrical perforations on the stūpikas of the three circular terraces of Borobudur and the 504 buddhas adorning the square terraces of the monument, Chandra argues that these buddhas are not, in fact, the Five Jina Buddhas but the thousand buddhas of the Vajradhātu Mahāmaṇḍala: The Vajradhātu becomes a mahāmaṇḍala because of the thousand Buddhas, which is its unique and contradistinctive attribute in the world of Tantric maṇḍalas. The Vajradhātu mahāmaṇḍala added to the already existing five forms of the Buddhas a sixth one with the dharmacakramudrā. These six forms of Buddhas were repeated 168 times to the auspicious number of 1008. That is why there are 504 Buddhas, and the niches and stupas enclosing these 504, symbolically make up the requisite double number: 504x2=1008.78 The text, Sarvatathāgatatattvasaṅgraha, and its explanatory Sarvadurgatipariśodhana-tantra that described the Vajradhātu Maṇḍala, have not been found in any form in Indonesia, but a small selection of verses from both these texts have been identified in the old Javanese Saṅ Hyaṅ 133; Moens, “Barabadur, Mendut en Pawon en hun onderlinge samenhang”; Chandra, “Borobudur as a Monument of Esoteric Buddhism”; Marijke Klokke, “Borobudur: A Mandala?: A Contextual Approach to the Function and Meaning of Borobudur,” International Institute for Asian Studies Yearbook 1 (1995); Hiram Woodward, “Barabudur as a Stupa,” in Barabudur: History and Significance of a Buddhist Monument, ed. Hiram Woodward and Luis O. Gómez (Berkeley: Asian Humanities Press, 1981); Kandahjaya, “A Study on the Origin and Significance of Borodudur.” For a good summary of all the interpretations, arguments, and counterarguments regarding the architecture of Borobudur, see Julie Gifford, introduction to Buddhist Practice and Visual Culture: The Visual Rhetoric of Borobudur (New York: Routledge, 2011); and Hudaya Kandahjaya, “Saṅ Hyaṅ Kamahāyānikan, Borobudur, and the Origins of Esoteric Buddhism in Indonesia,” in Esoteric Buddhism in Mediaeval Maritime Asia, Networks of Masters, Texts, Icons, ed. Andrea Acri (Singapore: ISEAS, 2016), 101–105. 77. Woodward, “Barabudur as a Stupa,” 121. 78. Chandra, “Borobudur as a Monument of Esoteric Buddhism,” 28. Chemburkar: Stūpa to Maṇḍala 199 Kamahāyānan indicating the knowledge of some of the key concepts embodied in these texts.79 Kandahjaya has observed an occurrence of “Tantra Bajradhātu” in the third version of Saṅ Hyaṅ Kamahāyānan and has argued that by the time this version (roughly 929 to 947 CE during the reign of King Sindok) was compiled, tantric teachings relating to Vajradhātu—apparently the Vajradhātu Maṇḍala—had already been circulated amongst the Javanese Buddhists.80 An inscribed object now kept at the site museum in the Prambanan complex and recently published by the Bureau of Conservation of Archaeological Remains of the Yogyakarta region shows a remarkable parallel to the Sarvadurgatipariśodhana-tantra.81 As we have already seen, dhāraṇī with close inter-textual connections to the Sarvatathāgatatattvasaṅgraha82 and a gold foil unearthed from Ratu Boko near the Prambanan temple complex in central Java also display elements of the Vajradhātu 79. There are three versions of Saṅ Hyaṅ Kamahāyānikan, and the text consists of two sections. The first section is titled Saṅ Hyaṅ Kamahāyānan Mantranaya and the second is Saṅ Hyaṅ Kamahāyānan Advayasādhana. See J. Kats, Sang hyang Kamahâyânikan: Oud-Javaansche tekst, met inleiding, vertaling en aanteekemingen (‘s- Gravenhage: Martinus NIjhoff, 1910), 30, 70; J. de Jong, “Notes on Sources and the text of Sang Hyang Kamahayanan Mantranaya,” Bijdragen tot de Taal-, Land- en Volkenkunde 130, no. 4 (1974): 465–482; Ishii Kazuko, “Borobudur, the Tattvasamgraha, and the Sang Hyang kamahayanikan,” in The Art and Culture of South-East Asia, Satapitaka Series, Indo-Asian literature 364, ed. Lokesh Chadra (New Delhi: International Academy of Indian Culture and Aditya Prakashan, 1991), 151–164. A comprehensive study on two sections of the text has been made by Lokesh Chandra, Cultural Horizons of India, Vol. 4: Studies in Tantra and Buddhism, Art and Archaeology, Language and Literature (New Delhi: International Academy of Indian Culture and Aditya Prakashan, 1995), 295–434. Recently Hudaya Kandahjaya’s paper compares and analyzes the relationship of this unique scripture with tantric Buddhist texts in Sanskrit, Tibetan, and Chinese. Kandahjaya provides preliminary answers to some key questions concerning its date and doctrinal inspiration, the milieu of its authorship, and its relationship with the Central Javanese Buddhist monument of Borobudur. See Kandahjaya, “Saṅ Hyaṅ Kamahāyānikan,” 67–112. 80. Kandahjaya, “Saṅ Hyaṅ Kamahāyānikan,” 99. 81. Arlo Griffiths, “Written Traces of the Buddhist Past: Mantras and Dhāraṇīs in Indonesian Inscriptions,” Bulletin of the School of Oriental and African Studies 77, no. 1 (2014): 167–169. 82. Arlo Griffiths, “The Greatly Ferocious Spell,” 1–36. 200 Pacific World, 3rd ser., no. 20 (2018) Maṇḍala.83 There are several Vairocana84 and Vajrasattva85 bronzes found in Java from the nine to eleventh centuries. From the above data, what we can safely say is that Borobudur definitely houses a hierarchical organization of maṇḍala in its architecture along with the basic elements of the Sarvatathāgatatattvasaṅgraha that were current during the construction period of the monument. Under the Śailendras in the eighth to ninth centuries, Javanese architecture changed rapidly to embody the maṇḍala system. Caṇḍi Sewu underwent an enlargement in a cruciform structure, probably to represent the Vajradhātu Maṇḍala.86 Two important architectural changes that occurred in Central Javanese temples during the Śailendra period were the transformation of a central sanctuary from a square to cruciform plan and inclusion of four separate rooms,87 presumably to follow the fourfold structure of the Pāla monuments that began at Kesariya. Some textual material derived from Sarvatathāgatatattvasaṅgraha likely served to demarcate the ground plans of central Javanese Buddhist monuments.88 A tenth- to thirteenth-century Javanese abhiṣeka rite of the Vajradhātu Maṇḍala is detailed in the Saṅ Hyaṅ 83. Sundberg, “A Buddhist Mantra Recovered from the Ratu Baka, ”165, 170, 171; Griffiths, “The Greatly Ferocious Spell”; and Acri, “Once More on the Ratu Boko,” 85. 84. Most of the Vairocanas found from Central Java are in Bodhyagiri mudrā. The first chapter of Sarvatathāgatatattvasaṅgraha has a section on four kinds of mudrā such as mahā-jñāna, samaya-jñāna, dharma-jñāna, and karma-jñāna. In karma-jñāna, Vairocana sits in the bodhāgrī mudrā. Cf. Kwon, “Sarva Tathāgata Tattva Saṃgraha,” 54. 85. J. van Lohuizen-de Leeuw, Indo-Javanese Metalwork (Stuttgart: Linden Museum, 1984), 38; Scheurleer and Klokke, Divine Bronze; J. Fontein, Sculpture of Indonesia (New York: Harry N. Abrams, 1990), 220 nos. 64 and 65. Also see MET Museum, New York, Vajrasattva no. 1987.142.169 from Samuel Eilenberg collection and 1984.409.1; British Museum Vajrasattva no. 1859.1228.18 and 1859.1228.17, V&A Museum Vajrasattva no. IS.38-1994. 86. Bosch (“Buddhist Data from Balinese Texts,” 111) identified Sewu as a Vajradhātu Maṇḍala drawn from the Sarvatathāgatatattvasaṅgraha. Lokesh Chandra has demonstrated it in detail as how a Śailendra ruler, as an aspirant to the status of cakravartin, dedicated the temple to Vairocana. See “Borobudur as a Monument of Esoteric Buddhism,” 8. 87. Jacques Dumarçay, Temples of Java (Cambridge: Oxford University Press, 1986), 20. 88. Arlo Griffiths, “Written Traces of the Buddhist Past,” 159–166. Chemburkar: Stūpa to Maṇḍala 201 Kamahāyānan Mantranāya text, which shows many similarities to Sarvatathāgatatattvasaṅgraha and contains the phrase cakravartyabhiṣeka in one of the closing chapters. Helmuth von Glasenapp notes its similarities to the rites described in present-day Japan and Tibet.89 The dominant theme is consecration to secret knowledge.90 The consecration rituals for the universal emperor interested many Buddhists in South, East, and Southeast Asia. The Javanese monk Bianhong, who was ultimately headed for India, arrived in Changan in 780 CE to undergo the Garbhadhātu Maṇḍala consecration of the Vairocanābhisaṃbodhi tantra.91 He also received Vajradhātu Maṇḍala consecration of the Sarvatathāgatatattvasaṅgraha system.92 His arrival in China coincides with the Śailendra period and the early construction phase of Borobudur. He joined the enormously influential group of monks in China that had been formed earlier by Śubhākarasiṃha, Vajrabodhi, Amoghavajra, and Huiguo. These masters were all experts in state protection sutras and maṇḍala 89. Helmuth von Glasenapp, “Ein Buddhistischer Initiationsritus des Javanischen Mittelalters,” in Tribus: Jahrbuch des Linden Museums (Stuttgart: Museum für Länder-und Völkerkunde, 1952–1953), 260; J. de Jong, “Notes on Sources and the Text of Sang Hyang Kamahayanan Mantranaya,” Bijdragen tot de taal-, Land-en Volkenkunde 130, no. 4 (1974): 465–482; Kandahjaya, “Saṅ Hyaṅ Kamahāyānikan, Borobudur, and the Origins of Esoteric Buddhism in Indonesia,” draws attention to a practice related to the anuttarapūjā, its possible depiction at Borobudur, and its knowledge in Japanese, Tibetan, and Chinese Buddhism. 90. Glasenapp, “Ein Buddhistischer Initiationsritus des Javanischen Mittelalters,” 263. 91. Yutaka Iwamoto, “The Śailendra Dynasty and Borobudur,” in Proceedings of the International Symposium on Chandi Borobudur (Tokyo: Executive Committee for the International Symposium on Chandi Borobudur, 1981), 85; M. Coquet, Le Bouddhisme ésotérique japonais (Paris: Vertiges, 1986), 89. The brief Chinese biography of Huiguo states that Bianhong had already studied esoteric texts in Ho-ling (Java) and after arriving in Changan expressed his interest in the teachings of the womb maṇḍala. See T. 2057.50.295b16–18, trans. Hudaya Kandahjaya, “A Study on the Origin and Significance of Borobudur,” 65, 94–96, 108, 165. 92. Based on Haiyun’s report Iain Sinclair discusses the teachings received by Bianhong at the Daxingshan monastery. See “Coronation and Liberation according to a Javanese Monk in China,” in Esoteric Buddhism in Medieaeval Maritime Asia, ed. Andrea Acri (Singapore: ISEAS, 2016), 29–66. 202 Pacific World, 3rd ser., no. 20 (2018) consecration rituals. Bianhong’s arrival in Changan, with official gifts such as a pair of conches, a brass object, and four vases for his master, Huiguo indicates his familiarity with tantric protocols93 but also implies familarity with these objects in his homeland of Java. Though we have a Tang record of only one Javanese monk to venture abroad in search of esoteric knowledge, his presence among Huiguo’s top disciples suggests not only that the Javanese were schooled enough in the esoteric discipline to send their elite monks to China, it implies the esoteric doctrines and deities being taught by Vajrabodhi, Amoghavajra, and Huiguo were available to Javanese temple architects. Whether Bianhong returned to Java and played any role at the Śailendra court, or in the construction of Borobudur, is not known, but it seems likely.94 In any case, news of tantric abhiṣeka rites of the Tang emperors would have reached the Javanese Buddhist circle. Bianhong’s manual on performing initiation that he compiled in China deals with the same subject matter as the Javanese text Saṅ Hyaṅ Kamahāyānan.95 NORTH-EASTERN INDIA, JAVA-ŚRĪVIJAYA, AND THE INDO-TIBETAN SPHERE There are at least six pieces of epigraphical evidence highlighting Pāla-Śailendra connections. Even though Śailendras were well connected and had access to Chinese Buddhist circles, their participation in Indian Buddhism is even more evident. The 778 CE Caṇḍi Kalasan inscription mentions a temple dedicated to Tārā by the Śailendra king.96 The construction of this temple 93. Ibid., 33. 94. The hypothesis that Bianhong did return to Java and was involved in the design of Borobudur was proposed by Kandahjaya, “A Study on the Origin and Significance of Borobudur,” 165, 251 and supported by Hiram Woodward, “Bianhong: Mastermind of Borobudur?” 25. Iain Sinclair refutes this claim in “Coronation and Liberation,” 35. 95. Sinclair, “Coronation and Liberation,” 39, 49. 96. The Kalasan Inscription opens with a laudatory verse to Tārā and also mentions the erection of the Tārā image. This twelve-stanza Sanskrit inscription written in early Nāgarī script was translated by F. D. K. Bosch, “Çrîvijaya de Çailendra- en Sañjayavamça,” Bijdragen tot de Taal-, Land-, en Volkenkunde 108 (1958): 113–123; also see H. B. Sarkar, Corpus of the Inscriptions of Java (Corpus Inscriptionum Javanicarum, up to 928 A.D.), vols. 1–2 (Calcutta: Mukhopadhy, 1971), 35–36. Chemburkar: Stūpa to Maṇḍala 203 corresponds with the rise of the Pālas (ca. 775–1214) and the construction of Tārā temples at Chandradvīpa, Nālandā, and the Somapura vihāra built by them.97 Tārā was one the most celebrated deities of the Pālas and possibly exerted some influence on the Śailendras, and she continued to remain important in both domains. Three centuries later, when Atīśa went to Tibet, after spending twelve years in Śrīvijaya to renew Buddhism, he gave fresh impulse to Tārā.98 The 782 CE Buddhist inscription of Kĕlurak in central Java mentions Bengali guru Kumārghoṣa from Gauīḍvīpa (Bengal), who consecrated an image of Bodhisattva Mañjuśrī.99 A 792 CE inscription of Caṇḍi Sewu also mentions the construction of a Mañjuśrī temple. Mañjuśrī evidently played a significant role in the Śailendra, Pāla, and Tang courts. In fact, Amoghavajra made him the national deity of China.100 A mid-ninth-century Caṇḍi Plaosan inscription mentions the central Javanese temple being visited by people arriving from Gurjardeśa, which refers either to Gujrat in western India, the Valabhī domain of 97. D. C. Sircar, “The Tārā of Candradvīpa,” in The Śakti Cult and Tārā, ed. D. C. Sarkar (Calcutta: University of Calcutta, 1967), 113, 128; Mallar Ghosh, Development of Buddhist Iconography in Eastern India: A Study of Tārā, Prajñās of Five Tathāgatas and Bhṛikuṭī (New Delhi: Munshiram Manoharlal, 1980), 9, 30. 98. See J. G. de Casparis, Prasati Indonesia II: Selected Inscriptions from the 7th to the 9th Century A.D. (Bandung: Masa Baru, 1956), 175ff.; Sarkar, Corpus of the Inscriptions of Java, 48; J. A. Schoterman, Indonesische Sporen in Tibet, 23; S. Ch. Das, Indian Pandits in the Land of Snow (Calcutta: Baptist Mission Press, 1893), 53–83; F. D. K. Bosch, “Een oorkonde van het groote klooster te Nālandā” [A charter from the large Nālandā monastery], Tijdschrift van het Bataviaasch genootschap voor Kunsten en Wetenschappen 65 (1925): 559. 99. This twenty-stanza inscription is written in pre-Nāgarī script. F. D. K. Bosch, “De inscriptie van Kěloerak,” Tijdschrift voor Indische Taal-, Land-, en volkenkunde uitgegeven door het Koninklijk Bataviaasch Genootschap van kunsten en Wetenschappen 68, nos. 1–2 (1928): 29–30; and H. B. Sarkar, Corpus of the Inscriptions of Java, 41, 100. For the Sewu inscription see John Miksic, “The Mañjuśrīgṛha Inscription of Candi Sewu, Śaka 714/A.D. 792,” in Texts and Contexts in Southeast Asia: Proceedings of the Texts and Contexts in Southeast Asia Conference, Yangon, 12–14 December 2001 (Yangon: Universities Historical Research Centre, 2003), 19–42; for Mañjuśrī’s role, see John Miksic, “Manjusri as a Political Symbol in Ancient Java,” in Anamorphoses: Hommage à Jacques Dumarçay, ed. Bruno Dagens and Henri Chambert-Loir (Paris: les Indes savants, 2006), 186. 204 Pacific World, 3rd ser., no. 20 (2018) the Matrīka kings, or to the kingdom of the Gurjara-Pratiharas in central north India.101 Indian textiles are referred to as gifts in central Javanese sīmā charters, which record tax and labor rights granted for religious foundations from the ninth to the fifteenth centuries. They mention buat kling putih, i.e., “white cloth made in India/Kālinga.”102 In India, the most important source on Indonesian history, the Nālandā copper plate of Devapāladeva (850 CE), mentions that the ornament of the Śailendra dynasty, Bālaputra, established a Buddhist monastery at Nālandā. The inscription refers to “bodhisattvas well-versed in tantras.”103 The inscription also provides important details about the ancestry of Śrīvijaya’s ruler at the time. Along with the northern Indian connection, there were strong connections with Southern India. Śaiva Coḷa rulers also patronized Buddhism, and in 1006 CE Rājarāja Coḷa granted permission to the Śailendra king Cūḷāmaṇivarman to build a Buddhist monastery (Cūḷāmaṇi Vihāra) at the coastal town of Nāgapaṭṭiṇam, which was supported by Coḷa grants.104 Two other Inscriptions dated to 1014–1015 and 1015 CE, during the reign of Rajendra Coḷa, also refer to the grants made by a Śrīvijayan agent.105 A detailed discussion regarding the political alliance between Java and Śrīvijaya under the Śailendras is beyond the scope of this paper, but the most in-depth study of the subject by Roy Jordaan and Colless states that Śrīvijaya was an allied kingdom of the Śailendras, who were 101. The ninth-century fragmentary stone inscription in Sanskrit is now kept in the Jakarta Museum (no. D82). See de Casparis, Prasati Indonesia II, 188–89, 202; Sarkar, Corpus of the Inscriptions of Java, 48. 102. Jan Wisseman Christie, “Texts and Textiles in ‘Medieval Java,’ ” Bulletin de I’École française d’Extrême-Orient 80, no. 1 (1993): 199. 103. Hirananda Shastri, “The Nālandā Copper-Plate of Devapāladeva,” Epigrahia Indica 17 (1924): 310–327; de Casparis, Prasati Indonesia II, 297. 104. Epigraphia Indica 22, no. 34 (1933). For the translation of the Sanskrit text see Noboru Karashima and Y. Subbarayalu, “Ancient and Medieval Tamil and Sanskrit Inscriptions,” in Nagapattinam to Suvarṇadvīpa: Reflections of the Coḷa Naval Expeditions to Southeast Asia, ed. Hermann Kulke, K. Kesavapany, and Vijay Sakhuja (Indian ed., Delhi: Manohar, 2010), 272–273. 105. Ibid., 275–276. Chemburkar: Stūpa to Maṇḍala 205 FIGURE 13. Tabo Monastery’s overall layout showing the modest mud structure in the central courtyard. the mahārājas of the Malay-Indonesian archipelago. In this case, the Buddhist tradition of Java and Śrīvijaya would have been very close.106 There is evidence of the impact of Pāla styles upon Javanese bronzes during this period.107 The iconographic material found at eighth-century Ratnagiri mahāvihāra in Odisha built by the Pālas has many parallels in temple sites of Java and Sumatra, especially Śailendra temples like Caṇḍi Mendut.108 Excavations at Udayagiri in Odisha have brought to light remains of a huge Buddhist monastic complex of Mādhavapura mahāvihāra where a seven-meter-high stūpa with four Jina buddhas in all four cardinal directions was constructed. The southern part of the Udayagiri has revealed important images such as Tārā, Mañjuśrī and Avalokiteśvara, dated to the eighth century.109 Based on the tenth106. Roy Jordaan and B. E. Colless, The Mahārājas of the Isles: The Śailendras and the Problem of Śrivijaya (Leiden: Department of Languages and Cultures of Southeast Asia and Oceania, University of Leiden, Semaian 25, 2009), ap. X. 107. For the Nālandā inscription of Bālaputra and the Leiden copper plate inscription dated to 1006 CE, see Epigraphia Indica 22, no. 34 (1933). For the influence on bronzes see Pauline Scheurleer and Marijke Klokke, Divine Bronze: Ancient Indonesian Bronzes from A.D. 600 to 1600 (Leiden: E. J. Brill, 1988), 27–30. 108. Natasha Reichle, “Imagery, Ritual and Ideology: Examining the Mahāvihāra at Ratnagiri,” Esoteric Buddhist Networks in Maritime Asia, 7th–13th Centuries CE, ed. Andrea Acri (Singapore: ISEAS, 2016). 109. See http://asi.nic.in/asi_exca_2005_orissa.asp. 206 Pacific World, 3rd ser., no. 20 (2018) century Intan shipwreck cargo of several ritual bronzes and vajras found off the Sumatran coast, John Miksic has hinted the possibility of Nālandā-Java-Sumatra connections.110 A large Pāla period black stone slab with a tenth- to twelfth-century inscription in Siddhamātṛkā script was found at Kesariya.111 It is the same script that was introduced to Java by the builders of Borobudur, the Śailendra kings.112 From an architectural point of view, a monument like Borobudur can only have been the culmination of a long period of artistic gestation. Given the Śailendra-Pāla contacts and the construction of the eighth-century Śaiva temples on the Dieng Plateau, it is not beyond the bounds of possibility in this connected Buddhist world that a breakthrough development in the Pāla domain, which transformed a stūpa into a maṇḍala of life-size buddhas, was enhanced with narrative reliefs at Somapura and Vikramaśīla before reaching its ultimate form of expression on Javanese soil. Just a century after Borobudur’s construction we find a perfect knowledge of the yogini-tantras in Java in the Nganjuk and Surocolo bronzes (last quarter of the tenth century or later) exhibiting the deities of the Vajrasattva and Hevajra maṇḍalas.113 Eleventh-century ritual deposits of Caṇḍi Gumpung inscribe the Vajradhātu Maṇḍala deities. Frequent appearance of the vajra motif and reliefs of dancers wearing elephant and ox masks at Biaro Pulo, Padang Lawas, Sumatra might be local variants of the sacred dances of Buddhist ceremonies performed in Nepal, Tibet, Mongolia, and Laos.114 The use of animal 110. John Miksic, “The Buddhist-Hindu Divide in Pre-Modern Southeast Asia,” 20–21. 111. According to D. R. Patil, The Antiquarians Remains in Bihar (Patna: Kashi Prasad jayaswal Research Institute, 1963), 201, the stone slab was found by J. B. Elliot in 1835 that had a representation of Viṣṇu, but the exact nature of the representation is not known. 112. Roy Jordaan, “The Śailendras,” 212. 113. Lokesh Chandra, “Identification of Nañjuk Bronzes,” in Cultural Horizons of India, vol. 4 (Delhi: Aditya Prakashan, 1995), 97–107; and Lokesh Chandra and Sudarshana Devi Singhal, “The Buddhist Bronzes of Surocolo,” Cultural Horizons of India, vol. 4 (Delhi: Aditya Prakashan, 1995), 121–147. 114. The relief panels around the base of the temples of Biaro Pulo and Biaro Bahal I depict dances in vigorous postures. Out of the five panels from Biaro Pulo, kept today at the Jakarta Museum, two clearly display masked dancers— one with an elephant head and another with an ox head. Biaro Bahal I temple Chemburkar: Stūpa to Maṇḍala 207 masks is common in the ‘cham dances of Tibetan monasteries. It gives us some idea of the Buddhism practiced in Java/Śrīvijaya post-Borobudur, during the time of Atīśa.115 The Buddhist traditions would have flourished post-Borobudur and possibly had new concepts that Indian circles lacked, as Atīśa went to Śrīvijaya from India in search of certain Buddhist practices. COMPARING BOROBUDUR AND TABO The massive monuments of Kesariya and Borobudur in brick and volcanic rock bear no similarity in external form to the modest mud architecture of Tabo (see fig. 13, above). But despite the disparate geography and outward appearance, Borobudur and Tabo have much in common, for they share a common religious philosophy, a sacred geometry, and fusion of the maṇḍala with an architectural space.116 The main temple of the Tabo Monastery was founded in 996 CE by King Ye-shes’-od under the religious supervision of Rin-chen-bzanpo.117 The king enjoyed launching missionary campaigns throughout the Indo-Tibetan sphere, with the help of his preceptor Rin-chenreliefs of the dancers are not masked but demonstrate vigorous postures. These masked reliefs may depict masked dancers taking part in Buddhist sacred dance. See John Miksic et al., Art of Indonesia (London: Tauris Parke, 1994), 75. For a discussion of the masked dance tradition in Mongolia see Patricia Berger, “Buddhist Festivals of Mongolia,” in Mongolia: The Legacy of Chinggis Khan, ed. Patricia Berger and Terese Tse Bartholomew (New York: Thames and Hudson, 1995). 115. The statuary found on sites of Biaro Pulo and Biaro Bahal I is possibly from the same period as Atīśa or just after his departure from the island. Miksic, “The Buddhist-Hindu Divide,” 26, 28. 116. Natasha Kimmet has recently compared the sacred space of Tabo and Borobudur in “Sharing Sacred Space: A Comparative Study of Tabo and Borobudur,” in Connecting Empires and States: Selected Papers from the 13th International Conference of the EurASEA, vol. 2, ed. D. Bonatz, A. Reinecke, and M-L. Tjoa-Bonatz (Singapore: National University of Singapore Press, 2012), 93–102. 117. Based on the inscription on one side of the cella, known as renovation inscription, the temple was founded in a monkey year (996 CE) and renovated forty-six years later (1042 CE) by the great nephew of the king (Klimburg– Salter, Tabo: A Lamp for the Kingdom, 18). The inscription has been translated and edited by Helmut Tauscher, “The Admonitory Inscription in the Tabo Du khan,” in Inscriptions from the Tabo Main Temple: Texts and Translations, vol. 83 208 West Amitābha Ratnasaṃbhava Amitābha Ratnasaṃbhava South Akṣobhya Amoghasiddhi 72 Vairocana North Amoghasiddhi Akṣobhya East FIGURE 14. Tabo assembly hall plan and Borobudur plan: arrangement of sacred space. Tabo plan © Christian Luczantis. Borobudur plan © Kern Institute collection, Leiden University Library. Pacific World, 3rd ser., no. 20 (2018) 4 Vairocana 209 Chemburkar: Stūpa to Maṇḍala FIGURE 15. The Sarvatathāgatatattvasaṅgraha pentad of Tabo and Borobudur. 210 Pacific World, 3rd ser., no. 20 (2018) bzan-po, commonly known as the Great Translator. The latter translated the Sarvatathāgatatattvasaṅgraha text to introduce the Vairocana Maṇḍala into the monasteries of Tabo and Alchi. Like the three distinct vertical structural levels of Borobudur and Kesariya that comprise of square and circular terraces with a crowning stūpa, the main temple (gtsug-lag-khang) of Tabo comprises three horizontal levels: an entry hall (sgo-khang), an assembly hall (‘du-khang) housing three-dimensional Vajradhātu Maṇḍala deities,118 and a cella (dri-gtsang-khang) surrounded by an ambulatory path (skor-lam). The entrance doors of the assembly hall of Tabo are protected by the guardian deities in a similar manner to the kāla heads above the four entrances of Borobudur. In a maṇḍala, a human seeking enlightenment must move symbolically from the violent and unconscious periphery towards the sacred center. The arrangement of the narrative reliefs of Borobudur is similar to the outer periphery depicting the worldly scenes, then the terraces closer to the central stūpa depict the world of bodhisattvas, and ultimately the center is reserved for the supreme buddha. To absorb all the doctrines, texts, and concepts embedded in the reliefs, adepts had to circumambulate the monument ten times in a clockwise direction. While doing so, they are sanctified by the presence of buddha icons in the balustraded niches of the upper gallery. At Tabo too, after crossing the entry hall, a practitioner circumambulates horizontally along the narrative murals of the assembly hall and moves towards the ambulatory and cella, into the realm of fully developed buddhas and bodhisattvas. While circumambulating, life-size clay images of the buddhas, suspended on the walls of the assembly hall around 1 m height from the floor, bless a practitioner. Traditionally the practitioner would circumambulate at least three times around the main Vairocana image. In Tabo he [or she] progresses through the spiritual geography of the maṇḍala and of Serie Orientale Roma, ed. Petech Luciano and Christian Luczanits (Rome: Is. I. A. Q, 1999), 9–28. 118. David Snellgrove, Buddhist Himalaya: Travels and Studies in Quest of the Origins and Nature of Tibetan Religion (Oxford: B. Cassirer, 1957), 66–67, 185; KlimburgSalter, Tabo: A Lamp for the Kingdom, 203; Laxman Thakur, Buddhism in Western Himalaya: A Study of Tabo Monastery (New Delhi: Oxford University Press, 2001), 98–126; Christian Luczanits, Buddhist Sculpture in Clay: Early Western Himalayan Art, Late 10th to Early 13th Centuries (Chicago: Serindia Publications, 2004), 72. Chemburkar: Stūpa to Maṇḍala 211 simultaneously identifies with the spiritual pilgrimage accomplished in the narratives, first by Sudhana and then by Siddhartha, the Buddha Śākyamuni. Thus through meditation and ritual circumambulation he [or she] performs a symbolic pilgrimage, which also leads to successively higher levels of consciousness.119 While physically moving through the space of these two monuments, a practitioner literally activates the narrative and experiences the dynamic space of the maṇḍala.120 The square terraces of Borobudur house multiple directional Jina buddhas along with the seventy-two Vairocanas of the top three circular terraces, thus forming the core of a unique form of the pentad at the heart of the Vajradhātu Maṇḍala.121 At Tabo, the rectangular plan of the assembly hall is an unusual shape for a maṇḍala, but by organizing the space of the hall in four directional quarters and placing the directional buddhas in each quarter, the builder overcomes the lack of symmetry of the maṇḍala (see fig. 14, above). THE SUPREME BUDDHA VAIROCANA OF THE VAJRADHĀTU MAṆḌALA AT BOROBUDUR AND TABO The central Vairocana at Tabo is placed at the back of the assembly hall to allow for daily rituals and the congregation of monks. The Vairocana sculpture is unique in consisting of four separate, complete, and identical human bodies seated back to back and facing the cardinal directions. This aspect of the sarvavid “all-seeing” Vairocana is elsewhere conventionally represented with four faces above a single body. The off-center placement of Vairocana at Tabo is a deviation from the textual maṇḍalas. Borobudur too, is utterly unique, with seventy-two Vairocanas in a dharmacakra or wheel-turning mudrā seated in bell-shaped, latticed stūpikas arranged around a large central stūpa on three circular terraces. Here the symbolic center of the maṇḍala has also been shifted from the actual center of the monument. There is 119. Klimburg-Salter, Tabo: A Lamp for the Kingdom, 108. 120. Ibid., 132–133; Klimburg-Salter also notes that this viewing experience is similar to viewing narrative sutra scrolls in East Asia, although in this case the viewer is stationary but activates the narrative through the unfolding and viewing of the scrolls. 121. Bosch, “Buddhist Data from Balinese Texts,” 109–118; Chandra, “Borobudur as a Monument,” 24–25. 212 Pacific World, 3rd ser., no. 20 (2018) also an emphasis in each case on the multiple Vairocanas emerging from the center and radiating across the whole maṇḍala. Around the Vairocanas, sculptures of the four buddhas of the Vajradhātu Maṇḍala (Akṣobhya, Ratnasaṃbhava, Amitābha, and Amoghasiddhi) form the key component of the maṇḍala in each temple (see fig. 15). At Tabo, the four directional Jina buddhas are differentiated by their respective colors and slightly larger size than the other deities of the maṇḍala. The entire assemblage makes up a configuration of thirty-three of the thirty-seven main deities of the Vajradhātu Maṇḍala. The maṇḍala of Borobudur, on the other hand, only incorporates the four directional Jinas, multiplied by 108 life-size images of each on the four sides of the pyramid. Lokesh Chandra prefers a more complex exegesis and claims the 504 buddha figures housed on the terraces of Borobudur are not, in fact, the Five Jinas but are morphological types that represent the thousand buddhas of the Vajradhātu Maṇḍala through their symbolic doubling (504x2=1008).122 He contends the presence of the one thousand buddhas is the distinguishing feature of the Vajradhātu Maṇḍala among the yoga-tantras.123 The inside wall of the ambulatory at Tabo depicts the hierarchy of the bodhisattvas, mahābodhisattvas, and one thousand buddhas of bhadrakalpa.124 Both Tabo and Borobudur went through at least two phases of construction activity, but the original iconography of both the monuments was based on the Vajradhātu Maṇḍala.125 The later phase evidently expanded the original concept. Despite the differences between these monuments, Klimburg– Salter sees a significant parallel in theory and practice at Borobudur and Tabo: The existence of Borobudur in Java is particularly interesting from our point of view for several reasons. 1) We have the fusion of the Vajradhātu-maṇḍala with an architectural space. 2) The elements 122. Chandra, “Borobudur as a Monument,” 24–5. 123. Ibid. 124. Christian Luczanits, “In Search of the Perfection of Wisdom,” in From Turfan to Ajanta: Festschrift for Dieter Schlingloff on the Occasion of His Eightieth Birthday, ed. Eli Franco and Monika Zin (Nepal: Lumbini International Research, 2010), 573. 125. Deborah Klimburg-Salter, Tabo Monastery: Art and History (Vienna, Austria: Austrian Science Foundation, 2005), 48; Jacques Dumarçay, Borobudur (Singapore, Oxford, & New York: Oxford University Press, 1978). Chemburkar: Stūpa to Maṇḍala 213 of the iconographic program are the same as those at Tabo: the Vajradhātu-maṇḍala, and the narratives from the Gaṇḍavyūha and the life of the Buddha.126 The lack of symmetry in the Tabo assembly hall and unconventional and incomplete set of the maṇḍala deities at both the monuments must be acknowledged. However, enough of the fundamental elements of the maṇḍala are present to indicate a conscious choice by the patrons. Java and Spiti were well grounded in the Sarvatathāgatatattvasaṅgraha and other yoga-tantras. GAṆḌAVYŪHA-SŪTRA: THE NARRATIVE PROGRAM OF THE TWO MONUMENTS The entire iconography program of the assembly hall at Tabo, including the story of the pilgrim Sudhana, is from the second phase in the eleventh century.127 Tabo and Borobudur both house the narrative stories in an identical manner between the lower and upper registers of the respective walls of the monuments. At Borobudur, the main focus of the narratives is the Gaṇḍavyūha of the Avataṃsaka-sūtra, especially the last chapter of its sutra, the Vows of Bodhisattva Samantabhadra. At Borobudur, this text has been accorded far more space than the other narratives reliefs. The sutra describes Sudhana’s spiritual journey in search of the ultimate reality by visiting 150 sacred places and spiritual guides or kalyāṇa mitra (good friends). The journey ends when Sudhana attains a vision of the Bodhisattva Samantabhadra and realizes that his own nature, and those of all the buddhas and bodhisattvas are, in fact, one and infinitely interpenetrate one another. The sutra concludes with Samantabhadra reciting the verses known as Bhadracarī. At Borobudur, special attention is paid to Sudhana’s encounters with the future Buddha Maitreya, the compassionate Bodhisattva Avalokiteśvara, and the bodhisattva of wisdom Mañjuśrī, but Samantabhadra plays the major role in the ultimate sections of relief.128 It is Samantabhadra’s direction of the pilgrim’s path that is accorded the place of honor on the highest levels of Borobudur. 126. Klimburg-Salter, Tabo: A Lamp for the Kingdom, 105. 127. Thakur, Buddhism in Western Himalaya, 148; Klimburg-Salter, Tabo Monastery, 39; Luczanits, “In Search of the Perfection of Wisdom,” 569. 128. Miksic, Borobudur, 127. 214 Pacific World, 3rd ser., no. 20 (2018) In the assembly hall of Tabo, the murals are organized along the lower registers of the wall in a more or less horizontal progression clockwise from the east wall of the entrance to the western wall of the ambulatory entrance. On the other side of the entrance of the ambulatory, the west wall bears narrative murals from Lalitavistara, a text that is also important to Borobudur. The upper register of the wall depicts several buddha realms, including that of the ten directional buddhas and their bodhisattva attendants. Several extracts from the Tibetan version of the Gaṇḍavyūha are inserted and correspond, though not precisely, with the narrative murals.129 The Gaṇḍavyūha stretches over 460 panels of the bas-reliefs at Borobudur,130 whereas in Tabo, the spiritual journey of Sudhana is compressed into 56 mural panels. At both monuments, the emphasis is on the search for “perfection of wisdom.” The Gaṇḍavyūha had been popular for some centuries even before its depiction at the Tabo monastery. But what is unique about the depiction at Tabo is the arrangement of its individual scenes accompanied by textual panels. The extensive use of cartouches possibly indicates Chinese influence rather than Indian.131 The last few reliefs of the top gallery of Borobudur are difficult to understand, but they depict Sudhana, sitting beside Samantabhadra with a halo, suggesting he has reached the state of an advanced bodhisattvahood (see fig. 16). At Tabo, the narrative Gaṇḍavyūha murals of the assembly hall continue inside the ambulatory path of the cella, but instead of Bhadracarī, stories from the last chapters of Aṣṭsāhasrikā Prajñāpāramitā are depicted. Some scenes are easily recognized and some have explanatory texts, but as at Borobudur many are difficult to interpret. Christian Luczanits says of these reliefs (see fig. 17): Principally, the ambulatory narrative is very similar to that of Sudhana, with its protagonist wearing the same dress and apparently also journeying from one teacher, commonly a Buddha or Bodhisattva, 129. Ernst Steinkeller, A Short Guide to the Sudhana Frieze in the Temple of Ta pho (Vienna: Arbeitskreis für Tibetische and Buddhistische Studien, Universitat Wien, 1966), 6, fig. 1. 130. Miksic, Borobudur, 127. 131. Pointing towards the use of cartouches at Dunhuang cave murals and their absence in Indian art, Dorothy Wong has suggested the influence of Chinese Buddhism. See Dorothy Wong, “The Huayan/Kegon/Hwaŏm Paintings in East Asia,” in Reflecting Mirrors: Perspective on Huayan Buddhism, ed. Imre Hamar (Weisbaden: Harrassowitz Verlag, 2007), 353. Chemburkar: Stūpa to Maṇḍala FIGURE 16. Borobudur gallery IV-53. Samantabhadra with his three-stemmed flower is elevated on a lotus cushion, and Sudhana is seen with a halo. 215 FIGURE 17. The protaganist at Tabo can be taken as reinforcing the quest of Sudhana on a higher level as per Luczanits. Here he is seen with a crown instead of a halo, offering himself to the bodhisattva. Tabo Ambulatory, N-wall. Photo © J. Poncar. to the next. The scenes are set against cloud-like mountains or within simplified architecture, as is also typical for the Sudhana narrative. However, the protagonist is now invariably crowned, as if Sudhana would have retained as exalted spiritual state after receiving blessings from Samantabhadra under the eyes of Vairocana in the last scene of the assembly hall narrative.132 THE SACRED CENTER: BOROBUDUR AND TABO The cella of the Tabo Monastery main temple houses images of the buddha seated on a double lotus cushion on a lion throne against the wall. The bodhisattvas Avalokiteśvara/Padmapāṇi and Vajrapāṇi/ Vajrasattva stand beside the central figure on the south and north walls respectively (see fig. 18). The central buddha has been interpreted as both Vairocana and Amitābha. Giuseppe Tucci identified the central figure as Amitābha because he is painted red and is seated in his dhyāna-mudrā position.133 Deborah Klimberg-Salter, however, disputes 132. “The ambulatory cella narratives are based on the last chapters of Aṣṭsāhasrikā Prajñāpāramitā, namely the story of Sadāprarudita in search of “Perfection of Wisdom.” See Luczanits, “In Search of the Perfection of the Wisdom,” 569–570. 133. Tucci Giuseppe, Indo-Tibetica, Vol. III: I Templi del Tibet Occidentale e il loro Simbolismo Artistico, Parte I, Spiti e Kunavar; Parte II, Tsaparanag (Roma: Reale 216 Pacific World, 3rd ser., no. 20 (2018) this, saying it is Vairocana because of the lion throne and because he is seen in Dunhuang and Ropa caves with his hands in this position holding an upright wheel of the law.134 If we accept the central buddha as Vairocana, then an unanswered question arises about two Vairocanas, one in the cella and the other in the assembly hall. The unusual presence of the two Vairocana images in the main temple may possibly be attributed to the eleventh-century temple renovation, which incorporated elements of the original artistic program with the new one.135 The question of two Vairocanas arises even at Borobudur. The 64 buddha images seated in the vitarka-mudrā on the topmost square terrace balustrade have been called Vairocana136 and/or Samantabhadra.137 If these images are assumed to be Vairocana, then like Tabo, this puts in question the identification of the partly visible Vairocana-like images in dharmacakra-mudrā in the 72 perforated stūpikas on the circular terraces. It is conceivable that at both Tabo and Borobudur, the sixth buddha is a representation of Vajrasattva of the yoginī-tantra, as claimed by the UNESCO restorers in panels at the base of the monument. (UNESCO calls the vitarka-mudrā Buddha Vairocana and the dharmacakra-mudrā Buddha Vajrasattva, in an ascending hierarchy. See the upper right-hand panel in the sign at the base of the monument.) Avalokiteśvara and Vajrapāṇi accompany the main buddha in the cella at Tabo. At Borobudur, the debate about whether there was a Accademia d’Italia, 1935), 78. Supporting Tucci’s interpretation, Thakur (Buddhism in Western Himalaya, 115) identifies the cella triad as AvalokiteśvaraAmitābha-Mahāsthānaprāpta along with Kṣitigarbha and Ākāśagarbha. 134. Klimburg-Salter (Tabo: A Lamp for the Kingdom, 143) identifies the cella triad as Avalokiteśvara-Vairocana-Vajrapāṇi/Vajrasattva based on similar figures from Dunhuang and Ropa. 135. Ibid., 91. 136. Moens/Long, “Barabudur, Mendut, and Pawon,” 22. 137. J. E. van Lohuizen-de Leeuw, “The Dhyāni Buddhas of Borobudur,” in Bijdragen tot de Taal-, Land- en Volkenkunde 121, no. 4 (1965): 408, 416; Frédéric Louis, Borobudur (New York: Abbeville Press, 1996), 184. According to Alice Getty, in the group of bodhisattvas, Samantabhadra displays varada-mudrā or vitarka-mudrā with his right hand while the left hand holds the cintāmaṇi, but at Borobudur the left hand of the Buddha is empty. See Alice Getty, The Gods of Northern Buddhism: Their History, Iconography and Progressive Evolution through the Northern Buddhist Countries (London, New York, Bombay: Oxford University Press, 1914), 46. Chemburkar: Stūpa to Maṇḍala 217 FIGURE 18. Tabo main cella with triad of Avalokiteśvara, Vairocana/ Amitābha, Vajrapāṇi. buddha in the main stūpa is not entirely resolved.138 A small four-armed Avalokiteśvara bronze was found in the main stūpa of Borobudur and is now kept in the National Museum of Ethnology in Leiden (Rijksmuseum Volkenkunde).139 There are also unconfirmed sources mentioning 138. Reports about the nineteenth-century discovery of a damaged and incomplete Akṣobhya within the broken and looted main stūpa led to claims that this was the main image of Borobudur. Moens (“Barabudur, Mendut and Pawon,” 33) thought this was an unfinished reject statue. This image came to light in 1842 during the excavations by Hartmann. Neither Thomas Raffles, Cornelius, nor Crawfurd had seen the image as per Louis Frédéric (Borobudur, 184), who suggested the possibility of the statue being placed in the stūpa by Hartmann or one of his subordinates in good intention. Stutterheim, van Lohuizen-De Leeuw, Bernet-Kempers, De Casparis, and Soekmono believed its position was authentic. For the summary of all the arguments and counter arguments, see Moens/Long, Barabudur, Mendut and Pawon, 32–35; and Nandana Chutiwongs, “Pieces of the Borobudur Puzzle Re-Examined,” in Indonesia: The Discovery of the Past, Exhibition Catalogue (Jakarta: National Museum Jakarta, 2005), 40–48. 139. N. J. Krom and Theodoor Van Erp, Beschrijving van Barabudur, s’Gravenhage (The Hague: Martinus Nijhoff, 1920), 652–654. The publisher produced an 218 Pacific World, 3rd ser., no. 20 (2018) a golden buddha statue from the main stūpa of Borobudur.140 If this was the case, then like Tabo, Borobudur too would have possibly had a bronze triad placed at the sacred center of the monument. The presence of a triad is seen in the iconography of contemporary Caṇḍi Mendut. Here, the central buddha is flanked by Avalokiteśvara and Vajrapāṇi, much like that of the Tabo cella. No one has yet offered a satisfactory reading of the last few reliefs of Borobudur’s fourth gallery. In the panel below (fig. 19) we see Sudhana sitting beside Samantabhadra below Amitābha’s western paradise. Amitābha, sitting in dhyāna-mudrā, is unusually accompanied by Vajrapāṇi as well as his own Bodhisattva Avalokiteśvara, forming an uncanny resemblance to the cella triad of Tabo. The standing bodhisattvas at Tabo display similar mudrās to the bodhisattvas seated in the Borobudur panel. The sculptures and reliefs of Tabo and Borobudur show several common features and support the arguments for seeing the Vajradhātu Maṇḍala embodied in the architecture of both. EXPERIENCING A MAṆḌALA The architectural designs of Tabo and Borobudur imply that sacred art requires activation through ritual movement in order to apply and validate their religious and political messages. The architectonic maṇḍalas need to be experienced through spatial movement that is vertical in the case of Borobudur and horizontal in case of Tabo. Gery Malandra sees a maṇḍala as a cosmic diagram in painting, sculpture, or architecture that is transformed to embody supernatural power by adept movements in rituals. The conception of the maṇḍala as a diagram is extended into visualization of concrete architectural space, and was transformed into actual temple architecture and sculpture. The universe-in-themaṇḍala is thus described and represented as a palace and, at the same time, the maṇḍala as a whole is conceived as being located in kūṭāgāra, a three-storied caved palace resting on the top of mount Sumeru…. [S]uch maṇḍalas as these include layers, or galleries in English translation of the two Dutch volumes entitled, Barabudur: Archaeological Description, along with three portfolios of illustration, in 1927. 140. Chandra, “Borobudur,” 3; Chutiwongs, “Pieces of the Borobudur Puzzle,” 44. Chemburkar: Stūpa to Maṇḍala 219 FIGURE 19. Borobudur gallery IV-50 top register relief with depiction of Avalokiteśvara, Vairocana/Amitābha, Vajrapāṇi. Sudhana is seen with a halo as well as Samantabhadra, who is not yet elevated to full buddhahood as he is not yet seated on a lotus cushion. which reside numerous manifestations of Buddhas, Bodhisattvas, and other deities….141 Many questions remain open. Did Atīśa play any role in the iconographic program of Tabo? Did he introduce anything from Śrīvijayan soil? Deborah Klimberg-Salter argues in an exhaustive study that the Tabo chapel was finished before Atīśa’s arrival;142 however, Christian Luczanits maintains that the renovation phase of Tabo is indebted to the eleventh-century commentary of Ānandagarbha on Sarvatathāgatatattvasaṅgraha.143 This is the same commentary that is invoked as the source of Nganjuk bronzes from the same period in 141. Gery Malandra, Unfolding a Maṇḍala: The Buddhist Cave Temples at Ellora (New York: State University of New York Press, 1993), 18. 142. Klimberg-Salter, Tabo: A Lamp for the Kingdom, 108. 143. Christian Luczanits, “The Clay Sculptures,” in Klimburg-Salter, Tabo: A Lamp for the Kingdom, 193–195. 220 Pacific World, 3rd ser., no. 20 (2018) East Java.144 Irrespective of Atīśa’s actual presence at either Tabo or Borobudur, the iconography of both the monuments certainly bears witness to some of his teachings propounding the integration of certain tantric practices. CONCLUSIONS This paper traces a paradigm shift in architecture from stūpa to maṇḍala at the ritual center of the royal Buddhist sphere. The mandalic architecture of Kesariya, Borobudur, and Tabo, with a central supreme deity and subordinate deities, reflects the political structure of samānta feudalism. The maṇḍala model thus provided a metaphor for earthly governance reflecting a celestial order. It contributed to spiritual enlightenment as laid out in texts and two-dimensional maṇḍala paintings but also employed a sacred model for realization of political ideology.145 For the Pālas of north-eastern India, the Śailendras of Central Java, and the royal lamas of Spiti, the maṇḍala designated levels of hierarchy for organizing the political and social landscapes of their kingdoms. How this maṇḍala model was used in the ritual or architecture is difficult to determine, as many of the teachings associated with the Vajradhātu Maṇḍala were oral, secret, or esoteric. The narrative of King Indrabhūti receiving the hidden scriptures in the important commentary of Prajñāpāramitā Nayaśatapañcāśaṭikā (150 Line Perfection of Insight) throws light on the preaching and practice of such esoteric scriptures. The narrative shows how the royal chief priest represented the court of princes, princesses, and ministers on a maṇḍala board. Thus the Vajradhātu Maṇḍala of the Sarvatathāgatatattvasaṅgraha text is physically enacted by the members of the court.146 144. K. W. Lim, “Studies in Later Buddhist Iconography,” in Bijdragen tot de Taal-, Land- en Volkenkunde 120, no. 3 (1964): 335–337. He was inclined to suggest that they correspond closely with details supplied within the Tattvālokakarī, an important commentary on the Sarvatathāgatatattvasaṅgraha written by Ānandagarbha (ca. tenth to eleventh century). 145. David Snellgrove recognizes the structural similarities between the maṇḍala and political systems. See Indo-Tibetan Buddhism, 199. Davidson, Indian Esoteric Buddhism, 131–144, systematically develops the argument. 146. Jñānamitra’s commentary on Prajñapārāmitā Nayaśatapañcāśaṭikā is found in the imperial catalogue of the Denkar library of ca. 810 CE (canon no. 2647, bs Tan-‘gyur, rgyud, ju, fols. 272b7–294a5); quoted from Davidson, Indian Esoteric Buddhism, 242–244. Chemburkar: Stūpa to Maṇḍala 221 This paper is a preliminary attempt to weave together some of the scattered strands of important new conceptions passing through a connected Buddhist world. A comparative architectural study of the three monuments presents a body of findings in support of seeing strong Indian and Southeast Asian enhancements resulting from the travel of architectonic ideas crossing geographical boundaries from the eighth to twelfth centuries. This brief encounter with the three key monuments of north-eastern India, Indonesia, and the Himalaya offers a cursory view of a connected Buddhist world of maritime Asia.

RELATED PAPERS

RELATED TOPICS

Log In

Stūpa to Maṇḍala: Tracing a Buddhist Architectural Development from Kesariya to Borobudur to Tabo

Stūpa to Maṇḍala: Tracing a Buddhist Architectural Development from Kesariya to Borobudur to Tabo

Related Papers

RELATED PAPERS

RELATED TOPICS