Images of devotion and power in South & Southeast Bengal, in: Esoteric Buddhism in mediaeval maritime Asia, Networks of Masters, Texts, Icons, ed. Andrea Acri, Singapore: ISEAS Press. pp. 163-91. Chapter 7 Images of Devotion and Power in South and Southeast Bengal c l au di n e bau t z e-pic ron T he importance of buddhist art in Eastern India from the 8th up to the 12th century has been recognized since the late 19th century as a major source of inspiration for the arts of Tibet and Southeast Asia. his artistic period has been associated with ‘esoteric’ practices and rituals and shows aspects which can difer from region to region over the centuries. Eastern India is a vast geographical area that includes Bangladesh and the modern Indian states of Bihar, Jharkhand, Odisha, and West Bengal. Bengal (Bangladesh and the state of West Bengal) and Odisha stretch along the Bay of Bengal, and the former region—in particular South and Southeast Bengal, an area which extends from the south of Dhaka up to Chittagong— was instrumental as the source of various aspects of Esoteric Buddhism in Southeast Asia. his art is usually labelled as ‘Esoteric’ or ‘Tantric’ and belonging to the Vajrayāna,1 and in inscriptions which mention them or in colophons of manuscripts which they had ordered to be produced, the Buddhists of Eastern India deined themselves as followers of the ‘excellent Mahāyāna’ (pravaramahāyāna; see below). he creation and development of a rich pantheon of characters, male and female, from the 6th century onwards, culminated in 12th-century Eastern India.2 But how was it really perceived and viewed? What were 1. I am thankful to Christian Wedemeyer and other colleagues for having replied to my query on the use of this term in the H-Buddhism Log for March 2012 (http://h-net.msu.edu/cgi-bin/logbrowse.pl?trx=lx&list=h-buddhism&user=&pw=&month=1203); see also Wedemeyer 2013: 9–10. 2. See Linrothe 1999 for an in-depth study of the development of the iconography of wrathful deities and its subsequent periods. Bautze-Picron mockup.indd 163 its functions? And were all its diferent types of production, i.e. stone, terracotta, stucco or cast images, manuscript illuminations, cloth-paintings or murals, regarded in the same way, without forgetting that images could also be visualized? It is beyond the scope of this chapter to try to answer these questions, which have already been assessed by authors in various diferent ways. For a long period, a major interpretation was the sectarian view, according to which the images are seen as relecting a conlict between the Buddhist community and Brahmanical or Hindu society. More recently, Rob Linrothe has put them in a new perspective, as relecting or symbolizing philosophical or religious experiences.3 his should however not lead us to abandon completely the more basic interpretation of some of these images as picturing deep tensions occurring between the diferent faiths existing in the region, which could surface and ind expression in the art—thus acting as a form of religious propaganda or helping to ease of the pressure.4 If images could be appreciated from diferent perspectives depending on the social position of the viewer, they also most probably relect on the situation of the community of monks within society at large. Due to the extremely rich amount of material, we shall conine our attention here to art-historical material. From its very early period, the Buddhist community had entertained close relations with the royal power. his connection, which can already be surmised from the biography of the Buddha 3. Linrothe (1990: 16–18 and passim) summarizes previous views on the topic. 4. See, for instance, Verardi 2011: 284–93, who analyses images of Cāmuṇḍā clearly betraying features of aggression towards the Buddhists. 11/7/16 10:16 PM 164 Esoteric Buddhism in Mediaeval Maritime Asia as recounted in the sources, survived through all the periods and found echoes in the imagery which emerged and developed as a language used within the Buddhist monastic community as well as among the laity. In the course of time, this visual language probably acted as intermediary between the diferent social groups: for instance, the image of the Buddha seated in bhadrāsana commonly met with on the façades of excavated monuments in Maharashtra in the 5th and 6th centuries could convey diferent meanings when conceived, commissioned, and seen by monks on the one hand, and when seen by laypeople on the other. For the monks it referred to the Sermon on Mount Meru—where the Buddha taught while seated on the throne of Indra, king of the divine universe, signifying the Buddha as a universal and divine king ruler over all gods of the Brahmanical pantheon, and hence also his power over human rulers. For the laity it was an image showing the Buddha dispensing his wisdom to humans like them. I doubt that laypeople could consciously have read more into such an image. However, it is true that such iconographic features, i.e. a major character seated in bhadrāsana on a lion throne, date back to the early centuries of our era, when it was used in depictions of Kushan rulers; it is thus possible to argue that these features remained lying in the collective imaginaire as pertaining to the iconography of a ruler. Images thus bear multiple purposes and meanings; their understanding occurs at diferent levels, some conscious, which I would qualify as ‘exoteric’, some unconscious, which I would regard as (to a certain degree) ‘esoteric’. What is exoteric here is the immediate perception of the depiction of the Buddha, seated on a lion throne, with hands joined in a speciic gesture in front of the breast, which leads to the equally immediate identiication of the Buddha as teacher. he esoteric element would be the subconscious knowledge that this image is one of royal power, portraying as it does a saintly man who became ruler of the universe as he could sit at the top of the Meru, occupying the throne of the king of the gods. his, I would add, is an understanding of the image that was willingly used by the Buddhist community on the façades of their monuments in order to transmit a message of power to lay society at a certain period and in certain regions. Bautze-Picron mockup.indd 164 his is not only the Buddha as a teacher, but it is the Buddha at the centre of the universe and thus teaching to the whole of society. his is not the use that is traditionally made of the term ‘esotericism’ in scholarly research, and one could argue that this interpretation actually concerns the symbolic message of the image. As a matter of fact, if one goes beyond the materiality of images—and this is the aspect with which I shall mainly be concerned here—then a mental, non material, conception of them emerges: the initiate creates an image which is part of an evanescent continuity and with which he either identiies (sādhana) or constructs a maṇḍala. Such spiritual practices have been described in texts such as the Sādhanamālā or the Niṣpannayogāvalī, whose rich iconographic material can help us to identify, i.e. to name, images actually cast, carved, or painted. But we should keep in mind that these texts describe another, purely visual, category of images, aimed as they were at practitioners involved in spiritual practices and not at serving as ‘user manuals’ for artists. Turning to Eastern India,5 and in particular to Bengal, where mainstream Buddhism lourished in the 11th and 12th centuries, we cannot ignore the strategic geographical position of the region, which makes it a crossroads between the Indian Subcontinent and mainland (as well as insular) Southeast Asia. his region, being traversed by monks not necessarily adept in the Mahāyāna, not to speak of its late esoteric phase, became a point of convergence of all monks visiting Eastern India and especially Bodh Gayā, which was and still is today a place where Buddhists from all over Asia meet. he main image of the temple was and still is venerated by Buddhists of every obedience, which implies that it bears diferent layers of interpretation: it can be Śākyamuni, but it can also be Vairocana (BautzePicron 2010a: 33, 37–38 and passim). his may lead us to wonder about the context in which images of the Buddha in South Bengal were produced and worshipped, and by whom (see below). Buddhists excelled in the production of images of power. Images are basically of a material nature. 5. For a general presentation of Buddhism in Eastern India from the 8th century onwards, see Niyogi 1980, Sen Majumdar 1983, and Chatterjee 1985: 230–356. 11/7/16 10:16 PM Images of Devotion and Power hey can, however, also be immaterial, such as those described in sādhanas, being created mentally and imbued with deep magical energy. hese images have let no material traces; being ‘esoteric’ and ‘subtle’, they were created by monks during their meditation. More than any material icon, a visualized immaterial image allows the monk to identify with the deity evoked. Material images likewise become active when brought to life through speciic rituals performed ater being carved or cast by artists. hus, material images always lie in a grey area, bearing an aura which is not materially perceptible but is nonetheless existent. Some images are more powerful than others; such are the cult images which are the objects of regularly performed rituals. Other images, e.g. those distributed on the façades of monuments, do not undergo similar treatment and their meaning is altogether of a diferent nature: they may, for instance, illustrate various aspects of the deity, i.e. of the Buddha in a Buddhist context; they may help in deining the sanctuary as a place inhabited by the Buddha; or they may contribute to relecting the universal presence of the Buddha.6 bengal, an ‘international’ region here are diferent ways of approaching the vast topic that is the presence of similar Buddhist images in the countries bordering on the Bay of Bengal and the Andaman Sea. he relations between images and monuments dating back to the 8th century and later from Indonesia, mainly Java, but also Sumatra or Borneo, and the homeland of Buddhism were noted a long time ago. Bihar—more precisely the 6. Such was the case of the temple at Bodh Gayā: before its massive restoration in the 19th century, stucco images of the Buddha—more rarely of Bodhisattvas (e.g., Mañjuśrī)—were distributed in the niches now occupied by stone images (Cunningham 1892: pls. XII-B, XIV–XV). Monument 1 in Nālandā also had its façades with niches containing stucco images not only of the Buddha, but also of Mañjuśrī and Avalokiteśvara, and this major diference between the two monuments (and sites) can be accounted for by the fact that the Mahābodhi temple at Bodh Gayā is the ‘house’ of Śākyamuni par excellence; Mañjuśrī being an integral part of the Buddha iconography at Bodh Gayā (see below), his presence among the stucco images is not out of place. Bautze-Picron mockup.indd 165 165 old Magadha—was considered to lie at the source of the ‘esoteric’ iconography noted in the Southern Seas. Recent research has, however, contributed to a more discerning appraisal of the situation during this fairly long period. Land routes within the Indian Subcontinent connected diferent regions and maritime routes crossed the Bay of Bengal and the Andaman Sea in all directions.7 Bengal is a vast region located at the convergence of diferent routes, from Assam and Yunnan in the east, Bagan and Arakan in the southeast, Odisha in the southwest, Bihar in the west, and Nepal and Tibet in the north and northwest. It is also a region of transition between sea and dry land, between South and Southeast Asia, with inland ways leading to harbours like Tamluk or Chittagong which were connected with maritime routes along which merchants and monks travelled back and forth.8 Monks came from abroad to join the Mahāvihāras and participate in the teachings that were dispensed there,9 but also to go on pilgrimage to the places visited by the Buddha. hey came from all over Asia, but relations between the Himalayan range, Bihar, and North Bengal were particularly intensive in the 11th and 12th centuries.10 In this context, monks from various countries and sects converged in particular 7. See Huntington 1989 and 1990 for a presentation of the ‘international legacy’ of the art of Eastern India, and Pande and Pandya Dhar 2004, which includes a number of papers dealing with various cultural, in particular iconographic, aspects shared by South and Southeast Asia. 8. Mukherjee 2011 is an in-depth study of the position of Bengal within the Asian network. 9. Besides monks from Tibet—as one could expect— others from Sri Lanka professed their faith in Mahāyāna, i.e. esotericism; such was the case of Jayabhadra who came to study at Vikramaśīla (A. Chattopadhyaya 1981: 325). See also Frasch 1998: 76–77. 10. Some inscriptions inform us about the movements of monks. See for instance the 9th-century Ghosravan inscription mentioning the monk Vīradeva who originated from northwest India (Sastri 1942/1989: 89–91; BautzePicron 2014a: 163, n. 4 for detailed references); the Bodh Gayā inscription of Vīryendra, native of Samatata (Southeast Bengal) and monk from Somapura (Paharpur) (Dutt 1962: 375–76 with further references); and the 12th-century Nālandā inscription of Vipulaśrīmitra, related to monks from Somapura (Sastri 1942/1986: 103–5; Dutt 1962: 376). For a study of the Bodh Gayā inscriptions and of the foreign presence at the site, see Leoshko 1987: 26–75. 11/7/16 10:16 PM 166 Esoteric Buddhism in Mediaeval Maritime Asia at Bodh Gayā, which was at the centre of a centripetal movement (Leoshko 1987: 40–56). Conlicts between monks originating from Sri Lanka and from the Himalaya are evoked in Tibetan historical sources, relecting conlicts between heravāda and Esoteric Buddhism.11 Similar conlicts seem to have arisen in Bagan during the early phase of the Bagan kingdom (Frasch 1998: 78), but from the second half of the 11th century onwards, heravāda deinitely grew stronger than Mahāyāna there.12 he Bagan rulers took deep interest in the Bodhi tree and the attached temple, sending missions to Bodh Gayā to preserve and restore the monument.13 In the reverse direction and as a inal stately gesture vouching for the depth of the relation between the two sites, a copy of the Mahābodhi temple was erected most probably during the reign of King Nadaungmya (r. ad 1211–ca. 1231) (Frasch 1998: 79–80, 2000b: 41). Travelling all the way from Bagan up to Bodh Gayā, pilgrims had to cross Bengal, leaving artistic evidence of various natures: bronzes cast in Eastern India were discovered at Bagan,14 and stone images of the Buddha discovered in Bengal illustrate an iconographic programme encountered in the murals of Bagan and share stylistic similarities with the 11th- and 12th-century stone carvings from this site.15 11. Apparently, the site was also visited and inhabited by monks originating from Sri Lanka; the existence of a heravādin monastery in the close vicinity of the Bodhi temple is attested from the 4th century up to the 12th century (Frasch 2000a: 58, n. 7; 1998: 73–74). he Tibetan monk Dharmasvāmin, who visited the site in the 13th century, tells how he was advised by a śrāvaka to throw the manuscript of the Aṣṭasāhasrikā Prajñāpāramitā, which he was carrying, into the river and not to worship images of Avalokiteśvara or the Tārā (Roerich 1959: 73–74). And Tāranātha records how ‘a large silver-image of Heruka … [was] smashed … into pieces … used … as ordinary money’ (A. Chattopadhyaya 1981: 279). 12. his does not mean that Esoteric Buddhist monks were altogether absent from Bagan, but they did not hold a major function in the site. 13. As Frasch (1998: 78–79) reminds us, there were three such missions during the Bagan period; see also Frasch 2000b: 41–43. 14. See Luce 1969–70 III: pls. 445a, 446, 447a–d. 15. Leoshko 1990; Allinger 2002; Lee 2009: igs. 24, 74, 82–83, 90; Huntington 1984: igs. 222–23; A. Sengupta 1993: igs. 54–55. Bautze-Picron mockup.indd 166 Besides this intense relation with nearby Burma, which is attested by various sorts of evidence, the region’s contacts extended beyond the seas. As mentioned below, Atīśa let his homeland to study in Suvarṇadvīpa, where he remained twelve years before returning to the continent and pursuing his brilliant academic career; and in the 11th to 12th century the worship of Heruka, so important in the Delta, found direct continuation at Padang Lawas, a Buddhist site in North Sumatra. Likewise, East Java established itself as heir to iconographic topics illustrated in Bengal in the 13th century.16 he presence of Buddhism in the region is well documented with sources of diferent natures, all showing how Eastern India in general, and Bengal in particular, held a fundamental position in the development of Buddhism around that period: many names of scholars survived in the Buddhist canon (Niyogi 1987, 1988), and the rich poetic work of the Mahāsiddhas was not only preserved in old Bengali but also translated into Tibetan.17 Concerned with 16. With regard to the names Hevajra and Heruka, see my n. 15 in Bautze-Picron 2014b. Mallmann (1986: 182–86) considers the names as being interchangeable when applied to some images, including the one encountered in Southeast Bengal; Linrothe (1999: 250), and following him Lee (2009), prefers retaining the name ‘Heruka’ for the images under consideration, underlining that the name refers to a ‘type’ but also to the ‘ultimate krodha-vighnāntaka’ depicted by these images. For the sake of easiness, but not necessarily being correct in view of Linrothe’s analysis, I shall retain here the name ‘Hevajra’ as in my previous study of the Padang Lawas material (BautzePicron 2014b), but remain aware that more attention should be devoted to the topic. 17. Shahidullah 1928; Moudud 1992; Dasgupta 1976: 3–109; Jackson 2004, with numerous references; Bagchi 1982: 64–75. Mahāsiddhas, whose iconography is well developed in Tibet (Robinson 1979; von Schroeder 2006; Linrothe 2006), found only a (timid) echo in the artistic representation of the region; for an example showing Śavaripa and probably from North Bengal or East Bihar, see Bautze-Picron 2007b: 85 and pl. 10.9; for an individual portrait, probably from Lakhi Sarai: Bautze-Picron 1991–92: ig. 34 and Linrothe 2006: cat. 4 188–89. Siddhas could also be depicted on the outer surface of petals of lotus maṇḍalas; see, for instance, Huntington and Bangdell 2003: cat. 68; Linrothe 2006: cat. 5 190–93; Weissenborn 2012b: pl. 52. Emaciated ascetics, possibly situated in a rocky landscape, can be found in some stone images (for instance, see Fig. 7.6), a situation which might be related to the tradition of the Mahāsiddhas 11/7/16 10:16 PM Images of Devotion and Power the history of Buddhism, Tibetan authors furnish ample information about Bengali scholars and the monasteries built in the region, as well as the local rulers who sponsored donations made to the Buddhist community or were instrumental in fostering the foundation of monasteries.18 Although these sources must always be approached gingerly, they do fortunately provide information that is not available from the Indian side. Due to the nature of the available documents, i.e. royal copperplates ratifying donations of land revenues, no details are indeed given concerning the real function or position of Buddhist monks at the court, or the genuine spiritual beliefs of the rulers.19 Data collected from local archaeology, epigraphy, and art history will be summarized here with a view to reappraising this part of the material and place it within its geographical and historical contexts. and of the Aris in Bagan (Bhattacharya 1994). An isolated and distant group of low reliefs showing siddhas and carved on the façade of caves is to be seen at Panhāle Kājī in Maharashtra (Deshpande 1986 and 1989). 18. See in particular Tāranātha (D. Chattopadhyaya 1980) and Sumpa (Das 1908). 19. he presence of the Buddhist seal at the top of the copperplates of the Pāla rulers and the fact that they label themselves as paramasaugata cannot be considered to be deinite evidence of the ailiation of the kings to the last esoteric phase of Buddhism. As seen in the inscriptions also, the recipients or donators respect the ‘excellent Mahāyāna’ (see below), and this is perhaps how they really felt. No local information is preserved which would vouch for the interference of monks in the royal function. Quite on the contrary, I would suggest that, at least in South Bengal—in the region of Vikrampur-Maināmatī—it was the Brahmins who might have held a major position in speciic rituals. Remains of only one Hindu temple were recovered on the Devaparvata, i.e. Maināmatī, which recalls the situation encountered at Bagan from the 11th century onwards, where we know that the Brahmins were involved in rituals related for instance to the foundation of the royal palace (see Bautze-Picron 2009b: 434). An important piece of information is however given in a biography of Śākyaśrībhadra (1140–225), who acted ‘as chaplain to the king of Jayanagara for some time’ (Jackson 1990: 11, 18 n. 1) and had visions of Maitreya while at court (ibid.: 10). Jayanagara was most probably the modern village of Jaynagar located south of Lakhi Sarai, where 12th-century Buddhist images were recovered in large numbers (Bautze-Picron 1991–92: 239–41). Bautze-Picron mockup.indd 167 167 political landscape Numerous inscribed copperplates have been recovered throughout Bihar and Bengal; the information that they contain is manifold and brings light to various aspects of the regional history, their main purpose being to record donation of land. hey mention kings and their genealogy, name the recipient, and describe the land that is given. While naming the rulers who authorized the donation, the texts also mention their religious ailiation: the Pālas, mainly ruling in Bihar, West and North Bengal from the 8th up to the 12th century, and the Candras, ruling in Southeast Bengal from the 10th century up to ca. ad 1050 are two dynasties that oicially claimed to be Buddhist. he king was described as paramasaugata, and the seal bearing the Buddhist symbol of the wheel lanked by two deer was aixed at the top of their copperplates, above the name of the ruler. he Pālas—When considering the recipients of the donation which these copperplates record, it should be noted that they are very rarely mentioned as being Buddhist: of all the oicial inscriptions of the Pāla rulers, only ive out of twenty-two concern a donation made to a Buddhist institution, all the others referring to donations towards Brahmins or Brahmanical temples dedicated to Viṣṇu or Śiva.20 hese ive donations which were made during the reigns of the early Pālas (8th–9th century) in Bihar and North Bengal are recorded in the two Nālandā inscriptions of Dharmapāla and his son Devapāla respectively, the so-called Murshidabad inscription of the irst ruler, the Jagjivanpur inscription of his grandson Mahendrapāla, and the Mohipur inscription of Gopāla II, son of Śūrapāla, the second son of Devapāla.21 Already in that period, the rulers might 20. G. Bhattacharya (2000: 441–42) lists nineteen inscriptions. See the following n. for the inscriptions regarding donations to the Buddhist community and which might have been discovered ater the publication of Bhattacharya’s paper. See also Sen Majumdar 1983: 144–50. 21. Nālandā inscription of Dharmapāla: see G. Bhattacharya 2000: 441 no. 1 and S. Bhattacharya 1985: 124, 154–55 no. 31 for further references. So-called Murshidabad copperplate of the same ruler: G. Bhattacharya 2000: 441 (second group) no. 1; see Furui 2011: 145 on the ind-spot of the plate. Nālandā copperplate of Devapāla: see G. Bhattacharya 2000: 441 no. 4 and S. Bhattacharya 1985: 125, 154–55 no. 33 11/7/16 10:16 PM 168 Esoteric Buddhism in Mediaeval Maritime Asia also have recognized donations made towards Brahmins or a Brahmanical temple, a practice which was to be preserved by their descendants. At the same time, Brahmins also appear as lay Buddhist donors responsible for the production of manuscripts as shown by Jinah Kim.22 he oicial inscriptions of the irst Pāla rulers relect a continuous tradition of donations made to the saṅgha by the king but also by his subjects, the sovereign ratifying the donation of land revenues for the upkeep of the institution while his subjects had provided the funds for the initial donation. Tibetan sources, however, mention that Dharmapāla had been involved in the establishment of the great monasteries of Vikramaśīla and Somapura (Paharpur, North Bengal),23 and seals were discovered in situ at Paharpur naming the monastery śrīsomapure śrīdharmapāladevamahāvihārīyāryabhikṣusaṁghasya.24 he recent discovery of a copperplate inscribed in the year 26 of the ruler’s reign conirms this link in referring to a donation of land plots for the upkeep of a community of monks at the Somapura-Mahāvihāra.25 for further references. Jagjivanpur inscription of Mahendrapāla: G. Bhattacharya 2000: 442 (second group) no. 2. Mohipur copperplate of Gopāla II: Furui 2008. 22. See Kim 2013: 236–47. It is perhaps within this context that one should consider the contents of tortoise-shell inscriptions discovered at Vajrayoginī (Vikrampur): both the Buddha and Vāsudeva, i.e. Viṣṇu, are praised side by side (Sircar 1949; Biswas 1995: 58). See also Prasad 2011a: 130, noting that in Southeast Bengal, donations were made to Brahmins rather than to Buddhist monasteries and that donations made to the monasteries had been done in a speciic limited area; and Prasad 2010 for a study of donors of images in Southeast Bengal who rightly remarks ‘a general absence of donation of any Buddhist deity by any member of society in early medieval Comilla, Sylhet, Noakhali and adjoining parts of Tripura’ (ibid.: 34). 23. A. Chattopadhyaya 1981: 274–83; Niyogi 1980: 52–59; Majumdar 1971: 115; Sen Majumdar 1983: 125–27, 128–31. 24. Dikshit 1938: 90; Niyogi 1980: 21–22, 52–59; Furui 2011: 155, n. 26. 25. See G. Bhattacharya 1994 and 2000: 442. See Furui 2011: 145 for a reading of the plate and for discussion of the ind-spot: this was initially thought to be in the Murshidabad district but the author suggests that—considering the information given in the inscription—it must have originated further north (South Dinajpur district in West Bengal; Dinajpur or Bogra districts in Bangladesh). Bautze-Picron mockup.indd 168 Even though the eulogy may start with the evocation to the Buddha and the ruler mentioned as paramasaugata, its text reveals that the world of Hindu gods was an integral part of the ruler’s culture.26 In the case of the oicial inscriptions of the Pāla and Candra rulers, who acknowledge their Buddhist faith through the seal ixed at the top of the copperplates and the label paramasaugata, we hardly ind any mention of female or male characters of the Vajrayāna. his does not imply that the individuals involved in the act of donation were not aware of this form of Buddhism: for instance, Mahendrapāla’s army chief who had the monastery built for which the donation was made is named Vajradeva (G. Bhattacharya 2000: 436). In the Devapāla Nālandā, the Mahendrapāla Jagjivanpur, and the Gopāla II Mohipur inscriptions, the donations were similarly made for the worship, copying etc. accordingly of the lord Buddha-bhaṭṭāraka, of the abode of all the leading virtues like the Prajñāpāramitā etc., of the multitude of the noble Avaivarttika Bodhisattvas, [and] for garments, food, lying and sitting accommodation, meditation and personal belongings etc. of the community of noble Buddhist monks [belonging to the] Eight classes of great personages [and] for repairing work [of the monastery] when damaged and broken.27 26. Let us briely observe that, as one could expect (see Bautze-Picron 2010a: 39, n. 91), it is the Vaiṣṇava mythology that is predominant in all inscriptions, with numerous analogies being drawn between the king and the god. he transition from one belief to the other at this ‘oicial’ level is made during the rule of the Candras in South Bengal (see below). 27. G. Bhattacharya (2000: 444–45) remarks how similar the formulation is between the Nālandā and Jagjivanpur texts, which he translates as ‘noble Buddhist monks (belonging to the) Eight classes of great personages’ (aṣṭamahāpuruṣapudgalāryabhikṣu saṁghasya). Cf. Mukherji 1992: 172 (whose translation is, however, unreliable). Furui (2008: 70) writes: ‘making the grant…. to the Buddha, the abode (sthāna) of all Dharmanetrīs beginning with the Prajñāpāramitā, a group (gaṇa) of non-returning (avaivarttika) Bodhisattvas, and to the bhikṣu-saṁgha as an embodiment of the eight great persons (aṣṭa-ārya-puruṣa)… according to their ranks. his was for the following purpose: worship, oferings and rituals…’. I wish to quote here an email by Peter Skilling, dated 16 March 2009, re- 11/7/16 10:16 PM Images of Devotion and Power Southeast Bengal—Out of the thirty-ive inscriptions that Swapna Bhattacharya had collected in 1985 in her study of the epigraphic material from Bengal and Bihar28 and which were found in Southeast Bengal—from Faridpur district up to Chittagong, most having been discovered in Maināmatī and the area of Vikrampur—only nine refer to a Buddhist recipient, all other beneiciaries being Brahmins.29 As mentioned above, the Candras who ruled from Vikramapura in the 10th and 11th plying to a question formulated by myself: ‘Avaivartaka is a quality of a bodhisattva, usually connected with the eighth stage (bhūmi)—a signiicant point because he cannot turn back, but must go on to full awakening.… One interesting thing about the inscriptions—only a handful, which belong to the Pāla period in Eastern India—is that they seem to take us from the ideal world of the texts—in which the ascent though the bhūmis takes many lifetimes—to the ‘‘real world’’ community of the monks, the saṁgha. Usually the inscriptions record donations to the mahāyanika-avaivartaka-bhikṣu-saṁgha—the community of avaivartaka Mahāyāna monks. In at least one case, the compound is preceded by the “eight noble individuals” (aṣṭa-ārya-pudgala) or similar terminology of Śrāvaka attainments…. he question is: were the monks of the vihāras in question actually regarded as having these attainments? Did they so proclaim themselves? Usually Buddhist tradition is discreet—or in some contexts restrained by Vinaya—about proclaiming one’s spiritual achievements. Or is the phrase simply a rhetoric that expresses the perceived worthiness of the recipients of dāna by the donor(s), or, more properly, by the eulogist(s)? In any case the term does not mean ‘‘the Mahāyāna Avaivartaka Buddhist sect’’ as it has sometimes been translated. In fact, the inscription(s) that mention the āryapudgala and the avaivartaka in the same breath are good examples for the fact that Śrāvakayāna (Hīnayāna, if one likes) and Mahāyāna lived under the same monastic roof and shared the same ideological bed. Again the idea of ‘‘sect’’ is quite inappropriate’. 28. See S. Bhattacharya 1985: 150–61. More copperplates have been discovered since S. Bhattacharya submitted her PhD, but their content does not invalidate the observations based on her list. Importance is to be attributed to the content of the mid 9th-century inscription discovered at Jagjivanpur, a site located in Malda district, West Bengal: it records a donation made towards the Nandadīrghika Udraṅga Mahāvihāra by Mahendrapāla’s general Vajradeva for the worship of (Buddhist) deities and the performance of rituals including the copying of manuscripts (G. Bhattacharya 1992). 29. See Furui 2013 for a study of the position of the Brahmins in Bengal. However, this is not decisive with regard to the situation of Buddhism among the society: Brahmins Bautze-Picron mockup.indd 169 169 centuries30 claimed to be paramasaugata and had the Buddhist seal and their names aixed to the top of their copperplates (Sircar 1973: 20, 51). All inscriptions of Śrīcandra start with devotion being paid to the Jina, i.e. the Buddha, the Dharma, and the saṅgha (Sircar 1973: 51), whereas his grandson Laḍahacandra (r. 1000–1020) claims his ailiation to Viṣṇu while preserving the Buddhist seal aixed to his copperplates (Sircar 1969–70: 199; 1973: 45–49, 51–54). As for Laḍahacandra’s successor, Govindacandra, he also preserves the seal and is named paramasaugata, but the content of his inscription is uniquely Brahmanical.31 From around the middle of the 11th century, the Candras were followed by the paramavaiṣṇava Varmans who also ruled from Vikramapura (Majumdar 1971: 197–204) and at a still later period, the site became the capital of the Senas, themselves devotees of Sadāśiva. A clear shit in the religious ailiation of the rulers in the region thus took place towards 1050.32 he Buddhist community could still, however, enjoy royal support in the region of Comilla at a later period: an inscription dated 1141 Śaka, i.e. ad 1219, was incised in the 17th regnal year of Raṇavaṅkamalla Harikāladeva, who claimed to be Buddhist; the inscription conirms the donation of land for the upkeep of a vihāra erected in the city of Paṭṭikerā and dedicated to Durgottārā.33 As a matter of fact, although no major donation to the Buddhist community can be anymore ascribed to a ruler afterwards, Buddhism remained present in Paṭṭikerā, a region that entertained close contacts with nearby could also become Buddhist and commission the production of manuscripts (see Kim 2013: 238–40). 30. See S. Bhattacharya 1985: 125–26, 154–56, inscriptions no. 38–50. See also Biswas 1995: 11–26 for a presentation of the history of the region. 31. Sircar 1969–70: 199; 1973: 49–51, 54–55; S. Ghosh 2008–9: 112–13; Prasad (2011a: 129) suggests ‘a gradual loss of Buddhist hegemony in Samataṭa’ in the 11th century. 32. For a survey on the ‘Religious condition of Vanga and Samatata’, see Biswas 1995: 51–60. 33. See D.C. Bhattacharya 1933: 283–84. he region, kingdom, or city known as Paṭṭikerā was part of the region of Samataṭa and is already mentioned in inscriptions of the Candras from Ladahancadra’s reign; it was located east from Comilla in Tripura district (old Tippera) (S. Bhattacharya 1985: 264; see also D.C. Bhattacharya 1933: 285–86). 11/7/16 10:16 PM Esoteric Buddhism in Mediaeval Maritime Asia 170 Bagan.34 And in an inscription dated 1158 Śaka, i.e. ad 1236, Gautamadatta, minister of the Vaiṣṇava ruler Dāmodaradeva, was ‘said to be “devoted to the feet of Śrī-Gautama”’ (Dani 1954: 186–87, line 16). As we will see below, even in the absence of royal patronage, the Buddhist community was active in the region, with manuscripts being produced around the beginning of the 12th century. Political relations and trade with the neighbouring kingdom of Bagan also characterize this period (Bautze-Picron, 2015), these connections being part of a wider network linking Bagan to Bodh Gayā, where the Burmese undertook restorations of the temple (Frasch 2000b: 41–42). lay practitioners and their manuscripts Fig. 7.1: Mañjuśrī as Dharmadhātu Vāgīśvara (detail of ms. from Harivarman’s regnal year 8; Baroda Museum inv. E.G.121). (Photo courtesy of Birgit Breitkopf) From the numerous manuscripts produced during this period, many are richly illustrated with images of male and female characters, and many are described in texts such as the Sādhanamālā.35 Few colophons give any information as to the place of production; when given, this site is oten said to be Nālandā, whereas a number of manuscripts were produced in various sites.36 As Jinah Kim’s research shows, most manuscripts produced in the 12th century relect a dramatic change in the nature of the donors—henceforth, lay practitioners. he production still, however, seems to be important in Bihar and probably North Bengal; although no indication of the site of production is given for them, a small number of manuscripts can be considered to have been produced in the region of Vikramapura-Maināmatī, since their colophons refer to rulers of Southeast Bengal. A manuscript was probably produced at Vikramapura during the reign of Govindacan- dra;37 three others were donated during the reign of Harivarman, two being illustrated38 and dated in the regnal years 8 and 19 respectively whereas the third, not illustrated, bears the date of 39.39 he irst reproduces the Pañcaviṁśatisāhasrikā Prajñāpāramitā and has for donor ‘Rāmadeva who is said to be a follower of the excellent Mahāyāna (Figs. 7.1, 7.16–17). he writing of the manuscript was completed in the eighth year of the prosperous reign of Mahārājādhirāja Parameśvara, Paramabhaṭṭāraka, Paramavaiṣṇava Śrīmad Harivarma Deva in the month of Kārttika, on the twelth day of the moon, on Wednesday in the constellation of Uttarāphālgunī’ (B. Bhattacharya 1944: 18). he second, preserved in the Varendra Research Museum in Rajshahi, contains the Aṣṭasāhasrikā Prajñāpāramitā.40 he third, kept in the library of the Asiatic Society of Bengal in Kolkata, includes the Vimalapra- 34. See Bautze-Picron 2014b; on this kingdom and its location, see Majumdar 1971: 257–59. 35. Exhaustive bibliographies on illuminated manuscripts are included in the works of Karen Weissenborn (2012b) and Jinah Kim (2013). 36. As reported by Weissenborn (2012a), the site is unrecorded in thirty-three manuscripts; it is given as Nālandā in seven cases and various other sites in ten further examples, among which a manuscript may have been produced at Somapura as suggested by J. Losty 1989b (as quoted by Weissenborn 2012a: 298 no. 26). 37. See Weissenborn 2012: 308 no. 43 for further references. 38. Bautze-Picron 1999 and 2009a; Weissenborn 2012a: 301–2 nos. 31 and 33, with further references. 39. See Majumdar 1971: 201, n. 1 for a discussion concerning the two dates given in this third manuscript, which would imply that Harivarman ruled at least forty-six years, starting his reign in 1073 or 1074. 40. See Siddhanta 1979: 383–84, where the manuscript is dated in the 19th regnal year; the six paintings are reproduced in Bautze-Picron 1999: pls. 13.32–35, 40–41, and pp. 192–93; see also p. 160 for further references. Bautze-Picron mockup.indd 170 11/7/16 10:16 PM Images of Devotion and Power bhā, a commentary on the Kālacakratantra.41 A further illustrated manuscript donated in the regnal year 47 of Lakṣmanasena was possibly produced in this part of Bengal, considering the fact that the ruler was then reigning from Vikramapura.42 he same can be said of an illuminated manuscript of the Kāraṇḍavyūhasūtra:43 although no precise date or location is given in the colophon, the style of the paintings includes features noted in the manuscripts of Harivarman’s reign, and the subdued colours profoundly difer from the vibrancy relected by the manuscripts produced at Nālandā.44 Two basic features seem to characterize the paintings of South Bengal: stylistically, line clearly prevails over volume, i.e. the surfaces are lat and covered with plain colours, and, as far as iconography is concerned, we observe a tendency towards a narrative rendering rather than the clear iconic images seen in manuscripts from Bihar or North Bengal. his is illustrated particularly well in a manuscript—the present location of which is unfortunately unknown, but which was photographed in the monastery of Nor in the 1930s—which Eva Allinger analysed, concluding that it is to be related to the art ‘of Bagan and eastern Bengal’.45 Depicted in the pedestals of images of the Tārā, Avalokiteśvara, and Mañjuśrī (Figs. 7.2, 7.21, 7.23) that are carved in the region is the worship of the manuscript which lies on a stand, beside which sits a monk wearing a ‘pointed cap’ and holding the two classical attributes of the ritual, i.e. the bell and the vajra.46 From comparison of images of Mañjuśrī 41. Shāstri 1917: 79–82: the author supposes that the river ‘Veng’ mentioned in the post-colophon was located in Jessore district (also mentioned by Majumdar 1971: 201, n. 1). 42. Kim 2013: 59, Ms. D10; Weissenborn 2012a: 294 no. 24, with further references. 43. Losty 1989a; Weissenborn 2012a: 311 no. 48; Kim 2013: 57, Ms. B5. 44. Bautze-Picron 1999: 162; Kim 2013: 106 ater Losty 1989a: 5. 45. See Allinger 2010: 36–38 and ig. 6. Two narrative texts that were illustrated in the region and in Nepal are the Kāraṇḍavyūhasūtra and the Gaṇḍavyūhasūtra, e.g. Weissenborn 2012a: 310–11 and Allinger 2008. 46. Bautze-Picron 1995: 61–62 and igs. 5–7, 10–11, 17, 22; for illustrations, see Lee 2009: igs. 16, 19, 27, 31, 95–96; ig. 80 shows the monk alone with no depiction of the manuscript; Bautze-Picron mockup.indd 171 171 Fig. 7.2: Mañjuśrī (Chandimura, Lalmai, Comilla District; Rammala Library, Comilla). (Photo: J.K. Bautze) in North (Fig. 7.3) and Southeast Bengal (Fig. 7.2) it immediately emerges that this double motif, i.e. the manuscript on the stand and the monk with pointed cap, is not only repeatedly encountered in the art of the region, but also occupies a predominant place in the ornamentation of the pedestal. he manuscript is also one of the Bodhisattva’s major attributes from an early period on; besides, Mañjuśrī is one of the Bodhisattvas encountered by Sudhanakumāra in his quest for wisdom and this is also a rare example of this iconographic motif at the bottom of a Buddha image. 11/7/16 10:16 PM 172 Esoteric Buddhism in Mediaeval Maritime Asia from inscriptions and written sources as a monastery, is said to have been founded in the second half of the 8th century by Dharmapāla, probably on the remains of an earlier religious structure.48 East from Paharpur, Mahasthangarh was a major city in the region for a very long period. hough apparently occupied up to quite a late period, none of these sites seem to have taken a very active part in the late phase of Esoteric Buddhism, embodied in Paharpur by the fragment of a small image of Hevajra embracing his Prajñā.49 Besides Somapura, another ‘Great monastery’ (Mahāvihāra) was located at Jagaddala, the foundation of which is attributed to Rāmapāla in the second half of the 11th century (Niyogi 1980: 59–60). Its precise location in North Bengal had long remained uncertain; however, excavations led at Jagdal in Naogaon District, North Bangladesh, a site situated some twenty kilometres from Paharpur, revealed Buddhist remains, including another—very delicately carved—small image of Hevajra and his Prajñā embracing each other, which might support the identiication of the site with the famous monastery.50 Large 11th- and 12th-century images of Mañjuśrī teaching (Fig. 7.3), of the Tārā (Fig. 7.4) and of Mārīcī (Fig. 7.5) were recovered in a vast area that roughly coincides with the Division of Rajshahi and the Dinajpur District in Bangladesh, and the districts of South Dinajpur (Dakshin Dinajpur) and Malda in the Indian state of West Bengal.51 West Fig. 7.3: Mañjuśrī (Niyamatpur, Rajshahi District; National Museum of Bangladesh). (Photo: J.K. Bautze) knowledge (see below). He is thus clearly an image relecting the fundamental importance attributed to this quest gained through the production, reading, and worship of manuscripts. geography and the archaeology of monasteries North Bengal—Major Buddhist sites are located in North Bengal:47 Paharpur or Somapura, known 47. See Niyogi 1980: 50–61 for a survey of the monasteries in North Bengal; see ibid.: 61–65 for the monasteries in West Bautze-Picron mockup.indd 172 Bengal. To her survey, one may add the monastery discovered and excavated recently at Jagjivanpur (Roy 2002). See also Dutt 1962: 328–80 for a survey of the Mahāvihāras and the monasteries during the Pāla period. Ranjusri Ghosh carried out intensive ieldwork in South and North Dinajpur districts (R. Ghosh 2006–7, 2008–9, 2012). 48. See Lefèvre 2012: 239–40 for a summary of the various hypotheses; the author himself suggests that the Buddhist monument was built on the remains of a Brahmanical temple. 49. Today preserved in the Indian Museum, Kolkata: see D. Mitra 1989: 182 and igs. 1–2; A. Sengupta 1993: ig. 40; and Niyogi 2001: igs. 39–40. 50. See Zakariah 1994 and Miah 2003: 153 pl. 11.6; see also Nazimuddin Ahmed in Banglapedia. 51. Today preserved in the Varendra Research Museum, Rajshahi. For the Mañjuśrī images, see Bautze-Picron 1993a: 151–52, n. 15; see also the following references: Rahman 1998: 23 and pls. 28–29, 39, and A. Sengupta 1993: ig. 41. For the Mārīcī images, see Bautze-Picron 2001a: 268–69, 11/7/16 10:16 PM Images of Devotion and Power Fig. 7.4: Tārā (Jagdal, Naogaon District; Paharpur site Museum). (Photo: J.K. Bautze) from this wide region was located the monastery of Vikramaśīla (today Antichak, on the south bank of Ganga River, Bhagalpur District, state of Bihar), the foundation of which is also attributed to Dharmapāla.52 he site opens the way to Magadha and its fundamental seats of learning, like Nālandā for instance, or to the pilgrimage route connecting all sites visited by the Buddha. As revealed by a very large number of stone images carved in the 12th century and found around the city of Lakhi Sarai, a place situated on the route between the Mahāvihāras of Nālandā and Vikramaśīla (Bautze-Picron 1991–92), the area held a crucial role in the region 288 no. 35–38, ig. 18, and also Rahman 1998: 45–46 and pls. 62–63. 52. See above, n. 23. Bautze-Picron mockup.indd 173 173 Fig. 7.5: Mārīcī (Narikelbaria, Paba, Rajshahi District; Varendra Rersearch Museum). (Photo: J.K. Bautze) during this period. One of the villages located in the southern suburbs of the town is Jaynagar, where important images were recovered,53 and which might well be the old Jayanagara where once stood the royal palace of a Pāla ruler, at whose court the Kashmiri scholar Śākyaśrībhadra resided for some time in the second half of the 12th century (see above, n. 19). As neighbouring regions, North Bengal and Bihar are closely related. In the 11th and 12th centuries, the Buddhist artistic production of both regions shared profound aesthetic similarities that are clearly distinguished from the artistic trend 53. To the references listed in Bautze-Picron 1991–92 is to be added the recently published paper by J. Kim (2012) on the nun who donated an Avalokiteśvara image (cf. BautzePicron 1991–92: 256, A.9). 11/7/16 10:16 PM 174 Esoteric Buddhism in Mediaeval Maritime Asia characterizing the production in South Bengal. he iconography is also directly in the circle of inluence of Bihar, whereas speciic iconic formulations emerge in the southern regions. South and Southeast Bengal54—South of Dhaka, the site of Bikrampur, i.e. Vikrampura (also named Vikramapurī), covers a very large area over which some seventeen or eighteen villages are scattered and which is limited by the Dhaleshvari River to the north and the Meghna to the east.55 he village of Vajrayoginī is said to be the birthplace of Atīśa (982–1054), who belonged to a royal family.56 Atīśa, also known as Dīpaṅkaraśrījñāna, let Bengal, probably in ad 1012, to study under Dharmakīrti ‘of Suvarṇadvīpa’, oten identiied with Sumatra although this was recently interpreted as a general name for the region of Southeast Asia.57 He returned to Bengal in ad 1025, having spent twelve years under the guidance of Dharmakīrti in the kingdom of the ruler Cūḍāmaṇivarmadeva, who is also remembered for having had a monastery built in Nākappaṭṭiṉam during the reign of Rājarāja around ad 1005.58 54. See Niyogi 1980: 65–86 for a survey of the monasteries located in the region. Special attention should be paid to the PhD hesis on the Buddhist art of the Dhaka region which was presented in 2009 at the University of Texas, Austin, by Eun-Su Lee, whose life tragically ended prematurely. 55. As mentioned and described by A.M. Chowdhury in Banglapedia, the area has greatly sufered and much transformed in the course of centuries through the evolution of the rivers, mainly the Padma, which runs south of the area (see also Abu Musa 2000: 1–4). 56. See Abu Musa 2000: 11–12; on Atīśa, see Eimer 1979: 182–96 and A. Chattopadhyaya 1981: 56–66. 57. Eimer 1979: 182–96; A. Chattopadhyaya 1981: 84–96; Cœdès 1964: 259, 264. See Skilling 1997: 188 for his and Wheatley’s opinions concerning the location of Suvarṇadvīpa. For the traditional views on the location, see BautzePicron 2014b, n. 79. ‘Śrīvijayapura (or Vijayanagara) of Suvarṇadvīpa’, where Dharmakīrti lived, can thus be tentatively located in South Sumatra (Palembang) (ibid.: n. 80; Schoterman this volume) or in the Malay Peninsula (Kedah) (Skilling 1997: 188–91 sums up the evidence), both regions being practically equidistant from the region of Padang Lawas, another major Buddhist site of the 11th and 12th centuries (on the Buddhist remains of this Sumatran site, see Bautze-Picron 2014b). 58. A. Chattopadhyaya 1981: 88; Cœdès 1964: 259–64; Skilling 1997: 188; Sen 2009: 67; Seshadri 2009: 125. See Guy Bautze-Picron mockup.indd 174 Vikramapura was the capital of various Hindu kings who ruled in the region. he initial centre of power of the Candras who ruled in the 10th and 11th centuries was located on the Devaparvata, i.e. Maināmatī, before Vikramapura became their administrative seat during Śrīcandra’s reign (ca. ad 930–75). hey were followed by the Varmans, who ruled in the 11th and the irst half of the 12th century from Vikramapura, before being ousted by the Senas who reigned till the 13th century. hroughout this period, from the middle of the 10th up to the beginning of the 13th century, the site proves to have been a very intensive Hindu place: numerous images of Viṣṇu, Sūrya, and Śiva, mainly as a dancing god, were produced then, all relecting very high aesthetic quality. No major architectural remains similar to those uncovered at Maināmatī or Paharpur could, however, be recovered; this may be due to the changing landscape of the region provoked by its rich and intricate water system.59 Going east from Vikramapura in the direction of Comilla, one is struck by the Devaparvata (today Maināmatī), a hill that rises out of the lat country and reveals abundant archaeological remains.60 Besides numerous architectural structures adorned with terracotta panels, a large number of small cast images and some, very rare, of large dimensions, were recovered during the excavations. Stone images were not found in abundance at the top of the hill but some were discovered in the countryside from here as far as the region of Chittagong, where a large group of 9th-century bronzes was recovered at Jhewari—now preserved in the Indian Museum (A.K. Bhattacharya 1989; D. Mitra 1982); an equally important 12th-century stone image of the Buddha was found at Betagi, where it is now worshipped. Tibetan authors mention 2004 for a survey of the Buddhist data attesting to contacts between South India and Southeast Asia. 59. Excavations recently revealed, however, a monastery at Dholagaon: http://bangladeshunlocked.blogspot. de/2012/05/atish-dipankars-vihara-dholagaon.html; http://bdnews24.com/bangladesh/2013/03/23/ancient-buddhist-vihara-found-in-munshiganj (last accessed December 2015). 60. See Bhuiyan 2008–9 for a small site located northwest of Comilla. 11/7/16 10:16 PM Images of Devotion and Power the existence of a monastery, the Piṇḍa or Paṇḍita-vihāra, in or near Chittagong: the monastery owes its fame to the debate which saw Buddhists and Brahmins opposed, and to the victory of the former ater putting on ‘pointed caps’, which have since remained part of the clothing of monks, especially in Tibet.61 As recalled by Puspa Niyogi (1980: 69), the Mahāsiddha Tilopa (988–1069) and at a later period, Vanaratna (1384–1468), the teacher of Tāranātha, were also native to the region.62 iconography and the function of the images From the immensely rich world of images which emerged within Vajrayāna in eastern India and which was inherited and further developed by the Tibetans, only speciic characters found their way into the visual iconography of Southeast Asia. While this phenomenon has yet to be rightly appreciated or even studied, we may suggest that the creation of such a rich iconography and the extremely abundant production of images in the monasteries of Bihar, Bengal, and Odisha might have partly resulted from the presence of Tibetan monks and their probable involvement in the transformations that the iconography underwent. A second aspect that the study of iconography brings to light is the geographical distribution of the images: not all ‘gods’ and ‘goddesses’ were represented evenly everywhere. his observation helps in localizing the speciic South Asian geographical area where images carved or cast in Southeast Asia had the source of their inspiration. he iconography is closely intermingled with the way of depicting these characters, i.e. with ‘style’: as demonstrated by Pauline Lunsingh-Scheurleer, the earliest socalled Javanese bronze images were in fact cast in southeast Bangladesh, and more particularly in the region of Maināmatī (see n. 64 below). Maināmatī was at the centre of an active Mahāyāna community from the post-Gupta period onwards. Stone images were rarely produced, 61. See Niyogi 1980: 69 and Bautze-Picron 1995: 62, n. 35. 62. See Pal 1989 and Niyogi 1988: 44. Like most monks, Vanaratna was widely travelled, going to study in Sri Lanka where he stayed six years before returning to India and Tibet. Bautze-Picron mockup.indd 175 175 whereas the production of metal images seems to have been very important, with numerous small bronzes from an early period (late 7th up to 9th century) having been recovered in the site and the region, depicting the main Bodhisattvas Avalokiteśvara and Mañjuśrī or the Tārā among others.63 he advance of Vajrayāna can be appreciated thanks to the discovery of a human-size bronze of Vajrasattva. Most of this material should be dated prior to the 9th century and is contemporary with images from the region exported to Java, where they would inspire a rich production of cast images in the subsequent centuries.64 Studies of iconography rely heavily on textual sources.65 However, the visual iconography oten presents features of which we ind no echoes in the texts. Producing an image is a creative process not locked in dogma and meets very speciic religious or spiritual needs at the time of its fabrication; monasteries were not entities withdrawn from the world, but were interacting with the space around them, and they were also open to monks coming from faraway countries. he fact that the monasteries of Bihar and Bengal were engaged in extremely dynamic religious and spiritual activities may also have overshadowed the fact that they might have been inluenced by developments which had taken place in other countries. Clear evidence is, for instance, provided by the representation of Sudhanakumāra standing or seated near the Bodhisattva Avalokiteśvara in images from Bihar and Bengal dated from the 10th century onwards (Fig. 7.6).66 Sudhanakumāra is no god and no image of him alone has, to the best of my knowledge, ever been produced: he is an attendant to the Bodhisattva forming a pair with Hayagrīva, whereas the 63. Lee 2009: igs. 127 to 162 reproduce such early bronzes discovered at Maināmatī or in the region. 64. See Lunsingh-Scheurleer and Klokke 1988: 66–69, cats. 14–17. 65. When researching on the late Buddhist art of South and Southeast Bengal, one is confronted with an abundance of publications which address various diferent topics; outstanding bibliographies on the topic of Esoteric Buddhist art have been prepared by Ulrich von Schroeder in his publications of 1981 and 2001, to which one can add Utpal Chakraborty’s book published in 2006. 66. Bautze-Picron 2014a: 99–100 (with further references). 11/7/16 10:16 PM 176 Esoteric Buddhism in Mediaeval Maritime Asia Fig. 7.6: Avalokiteśvara (Mahakali, Vikrampur area; National Museum of Bangladesh). (Photo: J.K. Bautze) Tārā is paired with the Bhṛkuṭī. Likewise, he can attend on Mañjuśrī in images of the latter.67 But he is the main character of a spiritual quest which leads him to Bodhisattvas like Avalokiteśvara and Mañjuśrī and which is narrated in the Gaṇḍavyūhasūtra.68 he text was illustrated in the 9th century on Borobudur in Central Java and in the 13th century on Candi Jago in East Java (see O’Brien, this volume); other versions are known from murals at Tabo (Western Himalayas), dated before 1042, and from a Nepalese manuscript tentatively dated around the mid 12th century or slightly later.69 67. See Bautze-Picron 1993a: 152–53 and Casey 1985: cat. 22. 68. Bautze-Picron 2014a: 99 (n. 116 for references). 69. See Allinger 2008: 153 (date of the manuscript as proposed by Jerry Losty), 154 and Klimburg-Salter 1997: 120–24. For a comparison of the narrative in Tabo and on Borobudur, see Kimmet 2012: 98–99. For other icono- Bautze-Picron mockup.indd 176 here is no doubt that the text was known in Bihar and Bengal as it was in these regions bordering India, but the main character was here singled out of the narrative and introduced as an attendant to the two major Bodhisattvas to be worshipped in Bihar and Bengal. Here he becomes a igure of knowledge, his main and sole attribute being the manuscript which he holds clasped under his let armpit, both hands joined in the gesture of veneration in front of his chest. We cannot exclude the possibility that the importance of the text in Java—which cannot be doubted, considering the fact that it is illustrated through 460 panels on Borobudur (Soekmono 1976: 20)—stretched out to Eastern India and beyond. In India, it is included in the iconography of two Bodhisattvas of diferent but to some extent complementary functions, although in both cases the fact that this young man in search of wisdom carries the manuscript symbolizing the spiritual knowledge which he accumulates through his encounters with wise men and Bodhisattvas may show how fundamental the book cult was within the monastery itself, but also beyond its limits, among the lay community (Kim 2013: 223–24, 236–50). he 8th and 9th centuries constitute a major period in the history of Buddhism in Java, marked notably by large constructions like Borobudur and Candi Sewu. Around the middle of the 9th century, Bālaputradeva, a Śailendra ruler, had a monastery built in Nālandā for monks originating from Java, and as a diplomatic gesture of good will, the Pāla ruler of the time, Devapāla, had the revenue from land and villages donated for the upkeep of this institution.70 he years around ad 830 have been considered by J. Dumarçay as pivotal between two phases of development of Javanese architecture, showing a ‘new cultural impulse’ which originates from India and brings new architectural techniques, and which is characterized by the creation of major construction sites (ater Klokke 2006: 51). For different but complementary reasons, Marijke Klokke (ibid.: 52) reached the same conclusion. In this context we should mention the dedication of an graphic similarities between Java and Western Himalaya, see Lokesh Chandra and Singhal 1999. 70. See Furui 2011: 155, n. 25 for the references. 11/7/16 10:16 PM Images of Devotion and Power image of Mañjuśrī made by a monk from Bengal or ‘Gauḍīdvīpa’ in the year 782 at Kelurak,71 i.e. in the vicinity of Candi Sewu, a monument whose date is much debated, but can be roughly dated between the second half of the 8th and the irst part of the 9th century.72 An inscription found in the Candi Sewu compound mentions the enlargement in 792 of a mañjuśrīgṛha, which could very likely be Candi Sewu itself (ibid.). Candi Sewu, like Borobudur, ofers a majestic three-dimensional illustration of a topic generalized in the sculpture and the painted manuscripts of Eastern India from the 9th century on, i.e. the depiction of the ive Tathāgatas, which was favoured mainly in the iconographies of Avalokiteśvara, Mañjuśrī, and the Tārā.73 As we will see below, the motif of the ive Tathāgatas saw a particular treatment in the iconography of the Potala mountain on which the Bodhisattva Avalokiteśvara resides (Fig. 7.6); this iconography is encountered in particular ater the 9th century in Bihar and South Bengal,74 which suggests that the visual concept ‘Mountain cum Tathāgatas’ might have been inspired by the impressive Javanese monuments. Images of speciic deities were more favoured than others and may not have found the same degree of religious fervour everywhere. hus, we witness a radical change in the 11th and 12th centuries in the regions of Vikrampur and Maināmatī, when large stone images of Mañjuśrī75 and Heruka/ Hevajra are produced (Figs. 7.2, 7.20–21; 7.9–10).76 his period follows a phase where the monastery of Nālandā seems to have exerted a major function: 71. See Miksic 2006: 188–90 and Klokke 2006: 53–54. 72. Klokke 2006: 53–54 (with further references). 73. But not exclusively; it crowned, for instance, the tall image of the Buddha standing at Jagdishpur, near Nālandā (Bautze-Picron 2010a: ig. 125a). 74. One of the most accomplished examples was found at Kurkihar (Bihar) and can be dated to the late 9th or 10th century; see Bautze-Picron 2014a: ig. 110. 75. Casey 1985: cats. 22, 33; M. Mitra 1999; Lee 2009: igs. 95–96. 76. Bhattasali 1929: 35–37 and pl. XII; Saraswati 1977: ill. 172; Huntington 1984: ig. 215; Biswas 1995: 40 and pls. 38–39; Linrothe 1999: 249–66 and ig. 188 (on Heruka, as named by Linrothe) and 267–75 (on Hevajra); Haque and Gail 2008: 279, 301–2 and pls. 506–7; Lee 2009: igs. 66–67, 69; Bautze-Picron 2014b: igs. 1–2. Bautze-Picron mockup.indd 177 177 Fig. 7.7: Mārīcī (Bhavanipur, Vikrampur area; National Museum of Bangladesh). (Photo: J.K. Bautze) images that are outstanding in terms of iconography and aesthetic quality have been recovered in the area of Vikrampur, which clearly relate on the evidence of their style to the ateliers located around Nālandā/Ghosravan/Tetravan in Bihar. Such is the case of a unique depiction of a speciic aspect of Mārīcī (Fig. 7.7) or a representation of Avalokiteśvara seated on the Potala (Fig. 7.6).77 More77. See Bautze-Picron 2001: 272–75 on the Mārīcī found at Bhavanipur; see also: Bhattasali 1929: 54–56 and pl. XIX; Huntington 1984: ig. 206; Samsul Alam 1985: ig. 69; Haque and Gail 2008: 131 and pl. 518; Lee 2009: 168–81 and ig. 51. Two further images of the elephant god Gaṇeśa found in Southeast Bengal and dated through inscriptions in the 10th century could likewise have been ‘imported’ from Bihar (Bautze-Picron 2014a: 160–61, nos. 15–17); it is most interesting to read that the donors of these two images and of a third one, of Viṣṇu, also sharing features with the art of Bihar, were donated by merchants who might indeed have been travelling between Southeast Bengal and Bihar 11/7/16 10:16 PM 178 Esoteric Buddhism in Mediaeval Maritime Asia aspects in Bengal (Hevajra; Saṁvara; Cakrasaṁvara; Buddhakapāla; Figs. 7.9–10, 7.15) reminds us that Atiśa, who originates from the region of Vikrampur, ‘had the visions of Tārā, Avalokiteśvara, Trisaṁvaravyūha (?) and Hevajra’.79 Hevajra/ Heruka as seen in Southeast Bengal (Figs. 7.9–10), i.e. dancing and at times trampling on a corpse, alone or at the centre of a circle of eight Yoginīs, was also worshipped in North Sumatra (Padang Lawas) and Cambodia. Other similar isolated images are found scattered in various sites of Bihar and Odisha, like Nālandā or Ratnagiri,80 and the deity can also be depicted with even more frightful aspects known as Saṁvara or Cakrasaṁvara, which seem to have been favoured in North Bengal, with images showing twelve or sixteen arms, four faces and embracing the Prajñā (Vajravārāhī).81 Veneration of Hevajra/Heruka and Mahākāla, another major wrathful male deity present in Bengal and Bihar,82 extended far beyond the South Asian Subcontinent; Mahākāla was protector of the state in the Dali kingdom and under the Yuan (Howard 1996: 235; Bryson 2012), and is also present in East Java and Sumatra. Hevajra was worshipped in North Sumatra and was a major deity in Cambodia,83 Fig. 7.8: Mahāpratisarā (Kurkihar, Bihar; Patna Museum). (Photo: J.K. Bautze) over, a 9th-century bronze image of Mahāpratisarā found at Kurkihar in Bihar is a clear product of the ateliers located around Maināmatī (Fig. 7.8).78 he importance of Heruka under diferent (Prasad 2010: 31–33). See Bautze-Picron 1991–92: 249–50 on similarities shared by the image of the Bodhisattva on the Potala and sculptures from Bihar, and also: Bhattasali 1929: 27–28 and pl. VIIa; Saraswati 1977, ig. 59; Shamsul Alam 1985: ig. 78; Haque and Gail 2008: 129 and pl. 512; Lee 2009: 54–55, 179–90 and igs. 54–55. 78. Bautze-Picron 2014a: 18 and ig. B36, and 135, n. 358 for further references. See Mevissen 1999 for a study of this iconography in Bengal and abroad. Bautze-Picron mockup.indd 178 79. See A. Chattopadhyaya 1981: 378 and Eimer 1979: 92: ‘Als Dīpaṁkaraśrījñāna studierte … sah er im Traum Heruka vor sich am Himmel und hörte ihn sagen… er solle Mönch werden und viele Schüler anleiten’. 80. hey were apparently to be found everywhere, but seem to have remained isolated in the sites where they were discovered (D. Mitra 1989, 1997–98). We also know of a silver image of the deity which must have stood at Bodh Gayā (above, n. 11). 81. See Weissenborn 2012b: 69–76 and igs. 51–55 on this iconography (also illustrated in manuscripts, 2012b: igs. 41a, 42c, 45a, 47–48). See Davidson 2002: 206–11 and Herrmann-Pfandt 2006–7 on the Cakrasaṁvaramaṇḍala, assimilated with Jambudvīpa and showing the universal power of the god over all Hindu, mainly Śaiva deities and sites of pilgrimage. 82. See Lee 2007: 204–8 and igs. 70–71 on two examples found in the Vikrampur area; the presence of a tiny Buddha image at the top of the slab, probably Amoghasiddhi (as recognized by Lee, whereas the author in Haque and Gail 2008 sees here Amitābha), makes it a Buddhist and not Śaiva image (also reproduced by Haque and Gail 2008: 163, pl. 462, and by Rahman 1998: cat. 72 and pl. 46). 83. See Bautze-Picron 2014b with further references on both gods in Sumatra and beyond. 11/7/16 10:16 PM Images of Devotion and Power 179 Fig. 7.10: Hevajra (Barkanta, Comilla District; National Museum of Bangladesh). (Photo: J.K. Bautze) Fig. 7.9: Hevajra (Lajjair, Comilla District; National Museum of Bangladesh). (Photo: J.K. Bautze) Bautze-Picron mockup.indd 179 Fig. 7.11: Wrathful three-headed male deity (Paschimpara, Vikrampur area; Varendra Research Museum, Rajshahi). (Photo courtesy of G.J.R. Mevissen) while his image found at Padang Lawas in North Sumatra is evidently based on those carved in the region of Vikramapura—another evident element of contact between the Buddhist centres of these countries. A still enigmatic three-faced wrathful god standing in the victory posture and trampling corpses was found in the region of Vikrampur (Fig. 7.11): although doubt remains as to the proper identiication of the wrathful three-headed image, this is evidently visually related to the more ‘clas- 11/7/16 10:16 PM 180 Esoteric Buddhism in Mediaeval Maritime Asia sical’ image of Mahākāla as found for instance in Lakhi Sarai but also in Vikrampur, or to further images of Yamāntaka and Krodhas from Bihar.84 he two gods, Hevajra and Mahākāla, are physically very diferent from each other: Hevajra/Heruka, in his images found in South and Southeast Bengal, has a beautiful well-proportioned human body, although he frowns and has fangs. his is not the case for Mahākāla and the wrathful image from Vikrampur: both are pot-bellied, have bulging eyes, have a heavy body with short legs, and do not display the lightness of movement which the dancing Hevajra shows. Other aspects of Heruka, like Cakrasaṁvara, found more particularly in North Bengal, partake in the same type of image, with a heavy body that expresses violence and aggressiveness, and whose obvious aim is to frighten. Images exported by monks evidently relected concepts developed in Eastern India, but they merged with local Southeast Asian spiritual and political considerations. Clearly, both Hevajra and Mahākāla were images of political power in regions outside the Indian Subcontinent, symbolizing the duty of the ruler to protect the country by destroying its enemies (Mahākāla), while illustrating the presence of the political power as universal (Hevajra/Heruka standing at the centre of a maṇḍala) and destructive of negative forces (Hevajra/Heruka dancing on corpses). One may thus wonder whether in India, as was the case in Southeast Asia or China, they had the leading position of protecting the state; there is no clear evidence that the rulers of Eastern India ever placed themselves and their states under the protection of Mahākāla or Hevajra.85 Taking this into consideration we may surmise that these gods had the function to protect the saṅgha: the monastery becomes the maṇḍala on which Hevajra rules.86 How are we to assess the function of these images in their historical context? Can we approach the carved and cast images, and the illuminations inserted in manuscripts alike? Probably not, since their basic function, which justiies their creation, difers. Large stone or cast images are to be displayed publicly, even if only to the monks, whereas small images can be carried as private belongings of monks, or donated to the monastery by monks and laypeople. Small objects could easily be transported, which makes their study rather diicult: were they produced in the site where they were discovered? Were they carried away from another site? Were they part of a set of images? Studying them thus means taking on many doubts and unknowns. he production of a manuscript relects a completely diferent situation.87 When the eye and the mind focus on a material image—for simple worship or identiication with it—the movement is centripetal, going from outside (the viewer) towards one single object. But opening and reading a manuscript opens entirely diferent vistas, the eye wandering through many diferent divine images and the mind following the spatial directions in which the deities depicted in the manuscript are distributed. he creation of a richly illuminated manuscript allows the insertion of practically as many images as desired; these male and female characters show a wide range of physical features known from their descriptions in sādhanas and can even be rarely shown in stone (Figs. 7.12–15). heir insertion all throughout a manuscript transforms it into a frame for these images, which are most oten unrelated to the text, usually of the Aṣṭasāhasrikā Prajñāpāramitā. In short, they do not illustrate the text, or only rarely do so, such as in the case of narrative texts or of the Pañcarakṣā text (Mevissen 1989, 1990, 1991–92, 1992, 1999). One manuscript of the Pañcaviṁśatisāhasrikā Prajñāpāramitā donated in the early period of Harivarman’s reign as seen above has a rich and peculiar iconography88 where Avalokiteśvara appears 84. See Bautze-Picron 1991–92: igs. 16, 25, 17 for a Krodha (Lakhi Sarai). For Yamāntaka, see Linrothe 1999: 162–76. For the Vikrampur image, see above. n. 82. 85. To which remark one should add the observation made by B.N. Prasad on the absence of inscriptions on images which would refer to a donor (compare above, n. 22). 86. See Bautze-Picron 2010a: 22–23 and Davidson 2002: 113–68. 87. See Jinah Kim’s book (2013), which ofers a range of new and challenging considerations regarding such manuscripts. 88. For Kim (2013: 104–6), the folios preserved in Baroda and those kept in diferent Western collections, private and public, do not belong to the same manuscript, whereas for Weissenborn (2012a: 301–2 no. 33) they all belong to one single manuscript. Bautze-Picron mockup.indd 180 11/7/16 10:16 PM Images of Devotion and Power Fig. 7.12: Ḍākinī (?) (Manuscript dated N.S. 393 or ad 1273, folio 401a; Varendra Research Museum, Rajshahi). (Photo: J.K. Bautze) Fig. 7.13: Eight-armed red goddess (folio 297a; Varendra Research Museum, Rajshahi). (Photo: J.K. Bautze) Fig. 7.14: Eleven-armed blue goddess (folio 173b; Varendra Research Museum, Rajshahi). (Photo: J.K. Bautze) Bautze-Picron mockup.indd 181 181 Fig. 7.15: Four-armed blue god (Hevajra?) (folio 388b; Varendra Research Museum, Rajshahi). (Photo: J.K. Bautze) as a major Bodhisattva, showing many diferent aspects (Figs. 7.16–17); his importance is consistent with the attention paid to diferent types of images of the Bodhisattva—for instance residing on his mountain, the Potala—found in the region (Fig. 7.6).89 he motif of the mountain quite naturally enhances the divine position of the Bodhisattva, this also being the place of residence of a major god like Śiva (Lee 2009, ig. 64), or the seat of the Tārā in a unique and outstanding carving found around Maināmatī.90 In this image, the Tārā preaches, evoking the ‘Potalake Bhagav[at]ī Tārā’ depicted in two Nepalese manuscripts of the 11th century.91 he fact is that these manuscripts, like the illustrated Kāraṇḍavyūhasūtra kept in the British Library, take part in the very same iconographic tradition relected by the manuscript dated in the regnal year 8 of Harivarman, i.e. they depict ‘famous’ images 89. See above, n. 77. Lee 2009: ig. 59 is related to this iconography, a trefoil arch being the only reminder of the niche within the mountain where the Bodhisattva sits: see Bautze-Picron 1999: 183–85. For other images of the Bodhisattva in the region, see Bhattasali 1929: 25–26 and pl. VIa; Saraswati 1977: ill. 60; A. Sengupta 1993: ig. 56; Haque and Gail 2008: 158, 160–61 and pls. 513–14; Lee 2009: 121–29, 185 and igs. 27, 31, 58–60. 90. See Lee 2009: 190 and ig. 65. For other images of the goddess in the region, see Bhattasali 1929: 56-58 and pls. XX–XXII; Shamsul Alam 1985: igs. 49, 65; Biswas 1995: 32, 35 and pl. 28; Niyogi 2001: ig. 50; Haque and Gail 2008: 146–47 and pl. 46; Lee 2009: 96–99, 105–12 and igs. 15, 17–19. 91. See Foucher 1900: 192 no. 16, 210 no. 18, and pl. VII.16. 11/7/16 10:16 PM 182 Esoteric Buddhism in Mediaeval Maritime Asia of the entire Buddhist world, with images of deities located in faraway countries like China, Java, Sri Lanka, or Maharashtra to quote only a few, whereas the images in the Harivarman manuscript are those of many aspects of Avalokiteśvara, the Tārā, or Mārīcī for instance.92 Looking back at the history of the Buddhist community in South Asia, more particularly at its carved artistic production, it is obvious that the dividing line which one would wish to draw between the community and the non-Buddhist society within which the saṅgha was evolving was never deeply anchored: Buddhists were well aware of the existence of Hindu gods and goddesses, introducing them in their own artistic imagery but having a perception of them which could thoroughly change over the centuries (Bautze-Picron 1996, 2010b). hus, Brahmā and Śakra (Indra) are present from the very beginning, appearing as peaceful characters fully submitted to the Buddha, an aspect that they preserve up to the 13th century. But from perhaps the late 10th century onwards, they also appear as demons belonging to Māra’s army, thus sufering the fate of deities like Gaṇeśa/Gaṇapati,93 Śiva, Pārvatī, and others who were perceived as malevolent characters and hence could be trodden on by frightful Buddhist characters whose major function was to destroy them. Going through the centuries, one thus attends a permanent encounter between the Buddhist and the Brahmanical (Hindu) imaginary worlds that inds its visual expression in a Buddhist context.94 When considering ‘Buddhist’ images, one tends too easily to forget that they were conceived in a context where the religious imagery was not exclusively Buddhist. Quite on the contrary, Buddhist 92. See Bautze-Picron 1999: 188 and Kim 2013: 56–57 and 93–109. 93. Lancaster 1991: 278 mentions a 5th-century Chinese translation of a text referring to him as Vināyaka, an ill-intentioned character ‘who keeps the practitioners from progressing and is thus a negative force that must be overcome.’ On Gaṇapati/Vināyaka in the Śaiva and Buddhist realms, see Acri, this volume. 94. his is not to say that the Hindu side was unaware of the existence of the Buddhists, but the main, if not only, character who was shown in a demeaning position was the Buddha; see Verardi 2011, passim. Bautze-Picron mockup.indd 182 Fig. 7.16: Avalokiteśvara (detail of ms. from Harivarman’s regnal year 8; private collection). (Photo: J.K. Bautze) Fig. 7.17: Twelve-armed Avalokiteśvara (detail of ms. from Harivarman’s regnal year 8; Baroda Museum inv. E.G. 122). (Photo courtesy of Birgit Breitkopf) monasteries were surrounded by a landscape which was fundamentally inhabited by ‘Hindu’ gods and goddesses and which might have felt dangerous, if not threatening. In a wide movement which swept over Bihar and Bengal from the 10th century 11/7/16 10:16 PM Images of Devotion and Power 183 Fig. 7.18: Buddha and his life (Betagi, Chittagong District; Ratnankur Vihar, Betagi). (Photo: J.K. Bautze) onwards, Hindu deities were considered to belong to Māra’s army where they held a leading position, replacing the traditional monsters of the army (Bautze-Picron 1996, 2010b: 111–16), and possibly having their outer appearance modiied—Brahmā becoming for instance a threatening character showing fangs (Bautze-Picron 1996: 126–27 and ig. 20). his perception of the Hindu pantheon apparently arose in the 10th century, an outstanding example being the image of Jagdishpur, a site located in the direct vicinity of Nālandā (BautzePicron 2010b: 108–11). From there, it is found up to Bagan in Burma, following a line that crosses Lakhi Sarai in East Bihar and the entire Delta (Figs. 7.18–19; below: he Buddha). Bautze-Picron mockup.indd 183 Fig. 7.19: Buddha and his life (Sibbari, Khulna District; Kamalapur Buddhist Monastery, Dhaka). (Photo: J.K. Bautze) Images are not only carriers of religious or spiritual values, they also deeply relect daily concerns; images could be conceived as an answer to other images which were produced by men of different beliefs, and as such could be acting as echoes to them. Images were clearly part of a ‘campaign’ that was aimed against those belonging to the ‘other’ side. Some examples can be given here, drawn from various parts of Eastern India and belonging to different periods: images of the Bodhisattva Mañjuśrī produced in Magadha in the 7th to 9th century, more particularly in the region of Bodh Gayā, 11/7/16 10:16 PM 184 Esoteric Buddhism in Mediaeval Maritime Asia clearly borrow elements from images of Skanda, such as the necklace and the hair-dress, whereas the lion serving as vāhana was probably borrowed from Durgā’s iconography.95 At a later date, the 12th century, outstanding images of Mahākāla closely resembling contemporary images of Bhairava were carved in the area of Lakhi Sarai.96 And, turning to South Bengal in the 11th and 12th centuries, it is probably no coincidence that most Hevajra images (Figs. 7.9–10) were produced in this region where the cult of Śiva Naṭarāja was particularly important, or that Avalokiteśvara shows various ‘Śaiva’ aspects in the manuscripts of the region, notably the rare form of Padmanartteśvara.97 Another factor might lie in the proximity of Assam, more particularly of the Kāmākhyā-pīṭha located north of the area here considered, and in the vicinity of which ‘a cemetery called Heruka’ would be located according to the Kālikāpurāṇa.98 Hevajra images receive a ighting stamp in being a wrathful character sharing features with the frightful Hindu goddess Cāmuṇḍā, whereas images of this goddess and of Śiva-Bhairava can also convey a strong message of violence towards the Buddhist community, wearing for instance a long garland of Buddha’s heads or trampling on a bowl full of similar heads.99 95. Not forgetting that Skanda is chief of the divine armies and that Durgā is a ighting goddess; see Bautze-Picron 1989: igs. 1, 4–15, 18 and 1993a: 151–52 (on the lion). 96. Mahākāla: Bautze-Picron 1991–92: igs. 16 and 35; Bhairava: Kumar 2011: igs. 15a–b. On the similarities, see Lee 2009: 206–7. Davidson 2002: 211–17 relates Heruka to Bhairava, ‘Heruka [being] formed in imitation of Maheśvara’, and Heruka being a ‘divinity of a cremation ground’; the situation is indeed very complex as the author also notes on p. 214. It is indeed likely that the (Buddhist) cult of a god named Heruka owes a lot to the (tribal and Hindu) Tantric tradition which had emerged in Assam (ibid.). 97. See Bautze-Picron 1999: 184. As Saṁvara, Heruka rules over the universe from Mount Meru, reminding us, as mentioned by Davidson (2002: 210), that this image echoes the motif of Śiva on Mount Kailāsa; but it also echoes the presence of the Buddha on Mount Meru, where he teaches the Dharma to his mother and the gods (Bautze-Picron 2010a: 28–35). 98. Davidson (2002: 213) mentions that the cemetery is today named Bhairava. 99. As shown by Verardi (2011), visual language was a major vector of propaganda and could relect deep conlicts Bautze-Picron mockup.indd 184 Also, Mañjuśrī is armed with a sword and represented in the gesture of using it, holding it high above his head.100 he development of the Bodhisattva iconography at Nālandā in the 7th and 8th centuries inds an echo in his importance in Java in the late 8th century. As seen in an earlier paper (Bautze-Picron 1993a), his function is to teach when Śākyamuni becomes Buddha, which makes the two iconographies—of the Buddha’s enlightenment and of the Bodhisattva teaching—profoundly complementary.101 Mañjuśrī possesses the sword with which darkness is slit and light (of the Bodhi) pervades the mind: such a major image used to stand at Bodh Gayā till it was stolen from the site,102 and the very characteristic gesture of holding the sword above the head as if going to be used occurs in three outstanding sculptures from Southeast Bengal (Lee 2009: 239–53 and igs. 95–96, 101). hese include a rare depiction of Arapacana Mañjuśrī (Fig. 7.20)103 which foreshadows a 13th-century image of the same iconography carved in East Java,104 and two images of the also very rarely illustrated Mañjuvajra (Fig. 7.21) where the Bodhisattva is three-faced and six-handed, and has attributes which illustrate his function as slayer of which were tearing the society apart. And in the ‘Buddhist camp’, this goddess was really perceived as a dangerous deity, belonging to the army of Māra and ferociously attacking the Buddha (Bautze-Picron 2010b: 111–16). 100. See Bautze-Picron 1999: igs. 1, 9–10. he weapon can be symbolic of destroying ignorance and darkness, opening the way to wisdom and light, but it might also be considered at a more basic level, being a visual rendering intended for non-Buddhists; the same can be said of the sword being held by Mārīcī (and other deities), whereas in the opposite camp it is Cāmuṇḍā who is depicted in an even more aggressive mood (see above, and n. 4 ). 101. It is worth noting that the Mañjuśrī illustrated in Fig. 7.2 is accompanied by Maitreya and Avalokiteśvara, thus partly reproducing the triad so oten encountered in Bihar of having the Buddha lanked by these two Bodhisattvas (Bautze-Picron 2010a: 77). 102. Bautze-Picron 2014a: 117–18, 214 n. 170 (see also ig. 170). 103. Bhattasali 1929: 28–29 and pl. VIIb; Saraswati 1977: ill. 27; Shamsul Alam 1985: ig. 89; A. Sengupta 1993: ig. 62; Haque and Gail 2008: 132 and pl. 491; Lee 2009: 248–53 and ig. 191. 104. As rightly suggested by Pauline Lunsingh-Scheurleer; see Bautze-Picron 2014a, n. 30 for further references. 11/7/16 10:16 PM Images of Devotion and Power 185 Fig. 7.21: Mañjuvajra (probably from Southeast Bangladesh). (Photo: MMoA, New York, www.metmuseum.org) Fig. 7.20: Arapacana Mañjuśrī (Jalkundi, northeast of Vikrampur; Nat. Mus. of Bangladesh). (Photo: J.K. Bautze) Bautze-Picron mockup.indd 185 Fig. 7.22: Mārīcī (Salban vihara, Mainamati; formerly at the Mainamati Museum, present location unknown). (Photo: J.K. Bautze) ignorance—a main attribute still being the manuscript—and as bestowing his compassion.105 he Bodhisattva is not the only character who makes a reference to Bodh Gayā and the Bodhi. A major goddess is then Mārīcī, whose numerous images were found all through the region (Fig. 7.22)106 and hardly difer from those wor105. See Lee 2009: 239–48 and igs. 95–96. An even rarer depiction of the Bodhisattva as Nāmasaṅgitī has been discovered in the Vikrampur area; the particular feature of this image is that it is the real female Nāmasaṅgitī who is depicted and not the usually seen male one (Akman 1999; Bautze-Picron 2000: 108–11; Lee 2009: 145–55 and ig. 37). 106. See my article of 2001 for a study of the iconography of the goddess in Eastern India and in particular, 11/7/16 10:16 PM 186 Esoteric Buddhism in Mediaeval Maritime Asia Fig. 7.23: Aṣṭamahābhaya Tārā (Sompara, Vajrayoginī, Vikrampur area; National Museum of Bangladesh). (Photo: J.K. Bautze) shipped in Bihar or North Bengal. his is not a peaceful deity: she carries weapons in her eight hands, presenting the vajra in the upper-right hand;107 four sow-faced female deities surround her. All of them form thus a maṇḍala, a structure which is encountered in more than one iconographic type; see for instance the images of Arapacana Mañjuśrī, Mañjuvajra (Figs. 7.20–21), Hevajra (Fig. 7.9), the ive Tathāgatas, and the Aṣṭamahābhaya Tārā (Fig. 7.23),108 or even the image of the Buddha’s life (see below). Mārīcī symbolizes the sunlight which penetrates the entire universe when Śākyamuni becomes Buddha; with her position of victory and her numerous armed hands, she is depicted as a warrior-goddess chasing away the darkness, i.e. ignorance. he very fact that she also stands within the womb of a stūpa is there to prove that light resides in the bosom of the Buddhist monastery. Her numerous images also testify to the importance of the female element as a dynamic, creative force. Movement is what opposed her to the images of the Tārā but brings her close to depictions of Parṇaśabarī (Fig. 7.24),109 a frightful goddess of folk origin, clad with leaves, who protects from diseases but also dispels hindrances to the spiritual quest, using the vajra and the aṅkuśa as weapons. Her images are rare but they are the relection of the way chosen by the Mahāsiddhas, drawing their inspiration from non-canonical literature, refusing to be part of the ‘oicial’ Buddhist community, and opting for a freedom of mind which they found in the solitude of the forest.110 As far as one can surmise from the existing material remains, Buddhism in this part of Bengal was not solely conined to the monasteries where monks were for the images produced in Southeast Bengal, see pp. 268–69 and 288–89: no. 39–47; cf. also the two images in the Khulna Museum published by Mevissen (2009) and Lee (2009: 155–68). The image reproduced here was stolen from the Maināmatī Museum (for further references, see Mevissen 2009: 280) and reappeared on the Belgian art market in 2006–7 (52e Foire des Antiquaires de Belgique: 209 where the provenance is said to be a ‘private English collection’); I myself saw it with the art-dealer K. Grusenmeyer (Sablons, Brussels). See also: Bhattasali 1929: 43–45 and pls. XIIIb–XIV; Huntington 1984: fig. 217; A. Sengupta 1993: figs. 27, 67–68; Haque and Gail 2008: 83, 146, 172–273, 282–83 and pls. 519, 523–25; Lee 2009: 155–68 and figs. 23, 42–49. 107. In Bihar, she can present the sword (Bautze-Picron 2001: igs. 6, 9, 14–15). 108. Bhattasali 1929: 56–57 and pl. XXI; B. Bhattacharyya 1958: ig. 375; Saraswati 1977: ill. 103; M. Ghosh 1980: 42 and ill. 11; Shamsul Alam 1985: ig. 90; A. Sengupta 1993: ig. 31; Haque and Gail 2008: 146 and pl. 530; Lee 2009: 99–105 and ig. 16. 109. Bhattasali 1929: 58–61 and pls. XXIIIa–b; B. Bhattacharyya 1958: igs. 173–74; Saraswati 1977: ills. 188–89; Huntington 1984: ig. 210; A. Sengupta 1993: ig. 70; Haque and Gail 2008: 147 and pl. 541; Lee 2009: 129–44 and igs. 33–34; see also M. Mitra 2000. 110. On the siddhas, see Davidson 2002: 169 f. Bautze-Picron mockup.indd 186 11/7/16 10:16 PM Images of Devotion and Power Fig. 7.24: Parṇaśabarī (Naynanda, Vikrampur area; National Museum of Bangladesh). (Photo: J.K. Bautze) mainly active: as Jinah Kim’s research recently demonstrated, the lay practitioners were more and more active notably in the production and worship of the book; besides, the Mahāsiddhas were also very active as ‘independent’ spiritual seekers throwing a critical look on the ‘oicial’, mainstream Buddhist way. Lay practitioners and these ascetics relect two further aspects of Buddhism, the study of which is more diicult to grasp, but which had to be taken into consideration by the saṅgha—hence the existence of these large images of Parṇaśabarī Bautze-Picron mockup.indd 187 187 in what was the capital, i.e. Vikramapura, where many other images were collected side by side with those of Sūrya, Viṣṇu, or dancing Śiva, and where, let us remember, the political power was openly Brahmanical. he community was evidently outnumbered and this severely reduced position might also explain the multiplication of Mārīcī images; emerging as seen above in the womb of a stūpa, but also belonging to the kula of Vairocana, it is the monument as body of the latter which is here standing, but it is also the monument as symbol par excellence of the Buddhist way, and thus of the community, which is here depicted. he Buddha—Considering the rich pantheon which we just briely surveyed and which developed during nearly two centuries, it must be noted that the iconography of the Buddha is rather uniform, depicting him touching the earth, i.e. at the very moment of his enlightenment. However, details added to this model can vary from image to image, and features can be inserted which are not encountered in Bihar, for instance the protome of an elephant below the Buddha.111 he main image remained, however, the one of Śākyamuni at Bodh Gayā, i.e. evoking the earth-goddess while touching the soil with the ingers of his right hand; as such, it could be venerated by any monk. A very speciic iconographic programme depicted on book covers and in a series of illuminations traditionally distributed at the beginning and the end of the text echoes the narrative rendering of the Buddha’s life, which is carved in the lower part of images produced in the region in the 12th and 13th centuries, and which relects a boundary between the ‘eight scenes’ model encountered in Bihar and the detailed rendering of the biography painted in various temples of Bagan.112 As also observed by Eva 111. See Lee 2009: 217–32 and igs. 74–80, 82–88, 90 for a lengthy discussion of the Buddha image in the region of Dhaka. he motif of the elephant protruding below the Buddha is also encountered on the images of Viṣṇu or Śiva in Bengal (Bautze-Picron 1999b: n. 16). Further, see Bhattasali 1929: 30–34 and pls. VIII, IXb; Saraswati 1977: ills. 196, 199; Huntington 1984: igs. 208–9, 223; Shamsul Alam 1985: igs. 55, 72, 77; A. Sengupta 1993: ig. 52; Biswas 1995: 36, 40–41 and pl. 40; Haque and Gail 2008: 129–30, 157 and pls. 486, 503. 112. Allinger (2010) wrote a detailed analysis of the question. 11/7/16 10:16 PM 188 Esoteric Buddhism in Mediaeval Maritime Asia Allinger, this iconographic model is also illustrated in early cloth-paintings that have been preserved in Tibetan monasteries.113 Even if one surmises that the carving of the images is posterior to some murals at Bagan, I personally do not think that their presence in the lower part of stone images from the Delta speciically relects a Burmese inluence, but rather illustrates a conscious choice of presenting the biography. he model retained is based on a composition which was elaborated in Magadha around the 9th century and which shows that the life of the Buddha reached its peak at Bodh Gayā, where the Bodhisattva became Buddha: this is the moment from where no return is possible (Figs. 7.18–19). Around the central large image, seven scenes are distributed following a very speciic visual pattern, paired as they are according to their iconography and not in a chronological order.114 All scenes, apart from the inal decease, are depicted as if they were icons standing on their own, in their own shrines, and not as part of a continuous narrative. Now, the motif of the shrine is typical of the art of Southeast Bengal, but the proile of the tower is apparently that of the Bodh Gayā temple and only certain godly igures sit within such a structure, i.e. the Buddha and Mañjuśrī. In these images only the central Buddha sits in a niche supporting this particular type of spire, whereas the secondary scenes are distributed in niches having their arch fully inserted in a series of lat and broad recesses which support an āmalaka, a structure well known in manuscript illuminations and which is considered to be a motif of (northern) Bengali origin.115 he spire of the Bodh Gayā temple, being only noted above images of the Buddha and Mañjuśrī ,is a consistent feature since, as mentioned above, the 113. Allinger 2010: 34–35; Bautze-Picron 1995–96. he place of manufacture of these cloth-paintings is still a matter of debate; I personally tend to consider the thangkas reproduced by Eva Allinger, igs. 5–6, to be of Bengali origin but made for a Tibetan patron (the painting of ig. 5 includes the depiction of heavily clad characters and monks). 114. See Bautze-Picron 2002: 222–23, with further references. 115. But which could occur in manuscripts produced at Vikramaśīla, a site located in East Bihar; see Losty 1989b: 89, 95. Bautze-Picron mockup.indd 188 Bodhisattva holds a major position in the iconography of the enlightenment.116 Not only does the presence of a tower transform the image into a shrine, but through the presence of four subsidiary shrines distributed around the spire the sculpture also illustrates a maṇḍala. he secondary scenes can also be seen as relecting speciic aspects of the Buddha, since apart from the Birth scene—always depicted in the lower part—all refer to Śākyamuni as a Buddha. he Birth scene is symmetric to the ofering of madhu by the monkey which took place at Vaiśālī; both scenes refer to inal moments in a long spiritual quest, the Birth scene coinciding with the end of the career of Gautama as a Bodhisattva, whereas the Vaiśālī event marks the moment when Śākyamuni took the decision to deinitively depart from this life, thus putting an end to his life as a Buddha (Leoshko 1993–94: 258–59). Similarly, the irst sermon held at Sarnath illustrates a ‘real’ event, whereas the sermon at Śrāvasti relects the magic powers possessed by the Buddha—and it is also there that the Buddha takes the decision to go on Mount Meru where he will teach his mother and the gods (Bautze-Picron 2010a: 28, n. 50). he Parinirvāṇa scene topping the composition indicates that the Buddha is now out of reach, in undeined realms. Now, the ‘perfect’ life which is here illustrated rests upon a high base covered with scenes carved in low-relief; what is depicted in the lower part of the image, being in contact with the soil beneath it, is located at a human level and thus is of a lower essence per se than the divine world seen above and which it has also for function to support (BautzePicron 1995, 2007a: 93–94). As a matter of fact, the scenes carved in the lower part of the image mostly show Śākyamuni as a Bodhisattva: the irst scene depicts his father presenting the newborn to Asita, followed by the great departure, the cutting of hair, the long period of asceticism up to the moment where Mucilinda protects the meditating monk, with further scenes being inserted. Now, what could appear to be a ‘simple’ narrative depiction of this 116. Moreover, these two images of the Bodhisattva also include a depiction of the Seven Jewels of the Cakravartin, a theme which was well favoured at Bodh Gayā (on which topic see Hsu 2008). 11/7/16 10:16 PM Images of Devotion and Power phase of the Buddha’s life hides or reveals speciic aspects of the Buddha’s personality. As seen in a previously published paper (Bautze-Picron 2008: 89–91), the depictions of the emaciated Buddha and of the Buddha being sheltered by Mucilinda refer in fact to two contradictory but complementary aspects of the Buddha’s personality which have been exempliied since the beginning of our era in Gandhāra, i.e. the Buddha as master of ire and water. In small images carved in the so-called andagu stone and which were mainly, but not exclusively, discovered in Burma, these two scenes are symmetrically distributed below the main cycle of eight events; in cloth-paintings or on painted book covers, they are usually seen side by side (BautzePicron 2008: ig. 5 and p. 89, n. 21). he image of the Buddha, being elaborated so as to include the canonical topic of the ‘Eight great events’—seven being distributed around the main one—and adding to it further moments of this biography, is addressed to any Buddhist. hus, this ‘perfect’ life inds its conclusion in the imperceptible moment of the enlightenment, which is the aim to be reached by the devotee, even if it is through radically diferent ways. Such images, besides narrating the initial phase of the Buddhist way, in some way also speak for the eternity of the Dharma. he ive Tathāgatas and the maṇḍala—he motif of the ive Buddhas is commonly encountered topping the images of the Buddha (whoever he is), Avalokiteśvara, Mañjuśrī, and the Tārā. However, it occurs irregularly in the images just considered which depict the biography of Śākyamuni, where it is the set of the seven Buddhas of the past joined by Maitreya that always crowns the image (BautzePicron 1995–96: 360–61). he depiction of the ive Buddhas is also seen in illuminated manuscripts and/or on their book covers.117 he distribution of their ive images118 all through the manuscript transforms the linear text into a three-dimensional structure.119 heir presence crowning stone images 117. Kim 2013: 83–85; Weissenborn 2012b: 25 and ig. 9. 118. Even if there are only three, they still suice to create a direction, i.e. a three-dimensional volume: see BautzePicron 1999a: 186–87. 119. See Kim 2013: 132–48 on the three-dimensional space that painted images confer to the manuscript (besides the fact that the very act of handling the manuscript transforms Bautze-Picron mockup.indd 189 189 allows them to achieve the same result, particularly when carved in a rock-like landscape that symbolizes the Potala Mountain on which Avalokiteśvara resides; they constitute the basic maṇḍala, four of them marking the four cardinal points around the ith one located at the centre. However, one should mention that nowhere in Eastern India were architectural compositions such as those achieved in Central Java in the 8th and 9th centuries ever created in stone. he concept of the maṇḍala is observed in numerous images produced in the region: Mārīcī, the Aṣṭamahābhaya Tārā, Arapacana Mañjuśrī, Mañjuvajra, or Hevajra, as mentioned above. Even the topic of the Buddha’s life has been reproduced according to a maṇḍala-like model that refers to Bodh Gayā, centre of the Buddhist universe, where Śākyamuni became buddha—a place and moment in time which shows the start of a new beginning and the emergence of a new spiritual thought. his model also integrates all around this fundamental scene smaller depictions of major events of Śākyamuni’s life. What is shown here is a pilgrimage that can be done through visualization. he idea of the non-dual universe with images emerging out of its centre found a particular achieved form in cast lotus maṇḍalas, found in the Faridpur district in Southeast Bangladesh up to Bihar, where the deity presiding over the maṇḍala can be Hevajra, Vajratārā, Cakrasaṁvara, or the Buddha.120 And beyond Bengal, this type of object was also produced in Bagan, having as central image the Buddha, a stūpa, or the spire of the Mahabodhi temple bearing a depiction of the eight great events, while monks are distributed on the eight petals.121 hese images of gods and godthe lat two-dimensional surface into a three-dimensional space). 120. Lee 2009: 340–44 and igs. 192–94: Hevajra, Vajratārā; Kim 2013: 65–68 and igs. 2–6, p. 67: Vajratārā, Buddha surrounded by eight Bodhisattvas; Weissenborn 2012a: ig. 52; Huntington and Bangdell 2003: cat. 68; Linrothe 2006: 190–93, cat. 5: Cakrasaṁvara. 121. See Luce 1969–70: pls. 425–28. Although it might appear diicult to consider these cast lotuses as being ‘esoteric’ or ‘maṇḍalas’, the fact is that the main topic which is depicted within the inner space shows the eight great events drawn from the Buddha’s life either distributed around a central shrine or ixed on the eight petals around 11/7/16 10:16 PM 190 Esoteric Buddhism in Mediaeval Maritime Asia desses each ofer in their own way a subtle version of this symbolic vision of the universe, and similarly these cast lotuses perfectly illustrate the oneness of the universe with its central point out of which all peripheral deities emerge. But the maṇḍala also represents—in a more prosaic manner—the monastery; such images are then subdued ways of visually expressing this spiritual stranglehold that the monastery pretends to possess.122 Certain deities can be magically invoked or worshipped within this context, having for function to protect from any danger, like the Pañcarakṣā goddesses,123 or from dramatic situations encountered while travelling, like the Aṣṭamahābhaya Tārā (who had inherited this function from Avalokiteśvara),124 or to protect from diseases, like some aspect of Avalokiteśvara (Kim 2012: 210). Asserting such powers and declaring its own sectarian vision of the universe—thought to be merged with the Buddhist community—were a necessary reaction in a religious landscape where highly powerful Hindu images were created in the 11th and 12th centuries, relecting the strength of the Brahmanical temples sustained by the political power.125 a central stūpa. hus it would be as if these various aspects of the Buddha either irradiated out of the central shrine or all merged into the central and inal monument. 122. Davidson 2002: 206–11 and Herrmann-Pfandt 2006–7 study the Cakrasaṁvaramaṇḍala, which includes deities linked to speciic pilgrimage sites distributed all over India and which are not speciically Buddhist, but mainly Śaiva or Śākta—a particular way of ‘assimilating’ the world of the other, of ‘magically reclaiming the whole of India for Buddhism’ (Herrmann-Pfandt 2006–7: 16). 123. See Mevissen 1999. For an image found in the Vikrampur area, see: Bhattasali 1929: 61–62 and pl. XXIX; B. Bhattacharyya 1958: ig. 185; Niyogi 2001: ig. 80; Haque and Gail 2008: 131 and pl. 5; Lee 2009: 112–13, 143–45 and ig. 36. 124. Allinger 2000; Bautze-Picron 2004: 239 and n. 105 for further references. 125. For instance, the images of Viṣṇu generally show the god surrounded by tiny images of his avatāras, and those of Sūrya present him with similar images of the Grahas or of the Ādityas: these three models illustrate the gods as a power which is eternal (the avatāras) or which rules throughout the universe (the Grahas) (see Mevissen 2006b and 2010; Bautze-Picron 1985: 473–74, 2007a: 101). Śiva is worshipped in the region mainly as dancing, displaying his power to destroy the universe, which he achieves through this action. Bautze-Picron mockup.indd 190 conclusion Artistic production in Bihar its within a long tradition; ateliers were numerous and very active and they were closely related to monasteries and sites of pilgrimages. his presence might account for stronger and stricter iconographic rules. he situation difers in the Delta: there is no site of pilgrimage and the monasteries do not seem to have played the same intensive ‘cultural’ role of drawing monks from abroad to remain and study there. Rather, the region seems to have been the bastion of the last ramiications of Esoteric Buddhism as strongly practised in, and originating from, Bihar and North Bengal. his type of Buddhism underwent deep transformations in Bengal, allowing the lay society to perform rituals involving the production of manuscripts, and being confronted with the countercultural ways of life advocated by the Mahāsiddhas. Moreover, being an area to pass through on the way to Bihar when coming from Burma, it was thus visited by heravādin monks and not exclusively by those advocating Mahāyāna. As we may surmise from the short survey presented here, this late (and last) phase of Buddhist art in Bengal relects a deep intricacy which makes the study of its iconography diicult but also stimulating. Whereas images of Mārīcī and Mañjuśrī relate to the Bodhi and thus to the homeland of Buddhism, i.e. old Magadha or Bihar, the history of the wrathful representations of Heruka is deeply interwoven with the iconographies of Śiva and Cāmuṇḍā as they emerged in Bengal and Assam, and the iconography of the Buddha’s life betrays deep similarities with its representation in Burma and Tibet. his chapter of the history of Buddhist art would undoubtedly deserve to be written anew. 11/7/16 10:16 PM