Introduction
The spread of Sanskritic Buddhism(s) across Asia has mainly been studied from a perspective focusing on transmission through the overland routes popularly known as ‘Silk Roads’, emphasizing Central Asia as an important transit corridor and contact zone between South and East Asia. However, current scholarship capitalizing on recent archaeological and epigraphic discoveries, as well as a more comprehensive and careful reading of textual evidence from various cultural areas and historical periods, has recognized the significant role that the trans-Asian maritime networks or ‘Silk Roads of the Sea’ played in shaping premodern intra-Asian connectivity. This has paved the way for an appreciation of the important contribution of the southern rim of Asia – especially South India, Sri Lanka, and Southeast Asia – to the genesis, transformation, and circulation of various forms of Sanskritic Buddhism.
This novel appreciation of the maritime networks has rectified misconceptions such as the received idea that the Southern regions of India and maritime Southeast Asia had a marginal role in the Buddhist Cosmopolis, as well as the overemphasis on the dominance of Theravāda/Pāli Buddhism in Sri Lanka and mainland Southeast Asia.Footnote 2 In fact, prior to the thirteenth century, both Southeast Asia, especially the Austronesian-speaking littoral and insular regions of Southeast Asia now called ‘Nusantara’Footnote 3 (Figure 1), and Sri Lanka hosted important (and even predominant) Sanskritic Buddhist traditions.Footnote 4 They also played a constitutive role in the genesis and transmission of both nascent and consolidated forms of Mahāyāna and Mantranaya/Vajrayāna across Asia from the fifth to the thirteenth century. Southeast Asia and Sri Lanka constituted stopovers and entrepôts for traders and voyagers. However, they were also frequented by monks and laymen alike as termini in their own right. Reasons for such visits included: collecting texts, relics, and icons, visiting pilgrimage sites, acquiring knowledge in institutionalized centres of higher learning or from renowned individual masters, and securing royal patronage.
Figure 1. Buddhist sites in Nusantara (Map by Andrea Acri).
Nusantara played an important, Asia-wide role as both a crossroads and terminus of Buddhist contacts from the early centuries of the Common Era. In one of his recent epigraphical studies, Arlo Griffiths (Reference Griffiths2014a: 137) makes a case for “the pan-Asian character of Buddhism and the integral place the Indonesian Archipelago once held in the ancient Buddhist world”. Similarly, Peter Skilling notes that “the peninsular and insular worlds of the ‘Southern Seas’ shared in a wide culture of ritual and ideas that stretched from Central Asia to East Asia” (Reference Lammerts, Griffiths, Silk and Eltschinger2015: 56). Skilling re-evaluates the important participation of premodern Siam in a much wider world of Buddhist cultural interchange than is usually assumed at present, questioning “whether ‘India’ should always be the ‘centre’, Siam the periphery – a passive recipient of ‘influence’” (2009: 42). Hiram Woodward (Reference Woodward2004: 353) has advanced an argument for “treating Indonesia and India as an integral unit well into the ninth century”, even making “a case for possible influence of Borobudur Buddhism upon subsequent developments in India”.
This survey, synthesizing and linking together evidence published thus far in secondary literature, highlights the important and constitutive role played by Nusantara in the genesis and circulation of Sanskritic Buddhism(s) across the geographically wide socio-spatial grouping of Maritime Asia (Acri Reference Acri and Acri2016a, Reference Acri2018). It presents an historical overview of the networks of sites and agents from a geographically broad perspective, emphasizing the maritime interactions that occurred across geographical and cultural boundaries in the region comprising of a web of coastal and inland polities that were connected to each other through a network of cosmopolitan ports and entrepôts from the Bay of Bengal to the South China Sea, over the course of several centuries. In so doing, it advances an alternative, and complementary, historical narrative that takes the ‘southern pathways’, i.e. the sea-based networks, into consideration, thereby revealing the limits of a historiography that is uniquely premised on land-based, ‘northern pathways’ of the transmission of Buddhism across the Eurasian landmass.
Trans-Asian Maritime Networks and the Constitution of Sanskritic Buddhism(s), from the Fifth to the Fourteenth Century
Evidence of the long-distance transfer of Buddhism from its north-eastern Indian cradle to the outlying regions of South India, Sri Lanka, Southeast Asia, and China via the maritime routes goes back to the early centuries of the Common Era. From the fifth century onwards, written and material evidence from the southern rim of Asia becomes more substantial, testifying to an efflorescence of long-distance maritime contacts that was to last for several centuries. As is shown by textual, epigraphic, and art historical materials – including icons, ritual accoutrements, dhāraṇīs, manuscripts, and monuments – Buddhist cults, imaginaries, and ritual technologies flourished across the vast swathe of littoral, island, and hinterland territory of Maritime Asia. Buddhist vestiges have been recovered from the Indian Subcontinent littorals, Sri Lanka, the Maldives, peninsular and coastal mainland Southeast Asia, and what are now called the Indonesian Archipelago and the Philippine Islands. These speak in favour of the existence of pervasive and sustained multi-directional Buddhist exchanges among interconnected nodes linking South Asia in the East to China, Korea, and Japan in the West through maritime routes (Acri Reference Acri2018; Sen Reference Sen, Wong and Heldt2014c). A polycentric, geographically wide, and maritime-based approach is necessary to fully appreciate how religious, mercantile, and diplomatic networks acted as catalysts for the transmission of Buddhism far and wide across Asia over nearly two millennia. Making a case for a multi-centric circulation of Buddhism, rather than a monodirectional transmission from a South Asian ‘homeland’ to Southeast and East Asian ‘peripheries’, recent scholarship has unveiled the multi-directional connections existing between Buddhist centres, tied to each other by overlapping networks of relations that were religious as much as economic, diplomatic, and political in nature (Sen Reference Sen2003, Reference Sen and Sen2014a, Reference Sen2014b). Therefore, in order to grasp the multifaceted, trans-regional phenomenon of the patterns of maritime Buddhist transmission across Asia, it is necessary to adopt a network approach, focusing on links between nodes and conduits, the movement of agents, and their role in dynamic processes of exchange.Footnote 5
While it is undeniable that the overland and maritime ‘Silk Roads’Footnote 6 were fundamentally interlinked and complementary, combined archaeological and textual evidence increasingly points to the predominant role of the latter in facilitating the mobility of Buddhist agents, artefacts, texts, and ideas over long distances from the early centuries of the first millennium of the Common Era. By the second century AD, the seasonal Monsoon winds were fully exploited by maritime traders plying the routes connecting the ports in the Mediterranean Sea with those along the coastal and insular areas of South, Southeast, and East Asia. The sea was a connecting factor in Asian history from time immemorial:Footnote 7 cutting across the natural boundaries and barriers of continental topography, sea-based routes formed a network of conduits that led to the formation of a mediaeval global Buddhist Asia. By the middle of the seventh century, factors such as the radical expansion of commercial maritime routes connecting South with East Asia, as well as the gradual decline of Buddhism and Buddhist exchanges in Central Asia following the Muslim conquest of Transoxiana and other socio-political contingencies, contributed significantly to the sea-based exchange, not only of mercantile goods, but also of Buddhist beliefs and ritual practices.
The existence of Buddhist monasteries near major commercial nodes and trading routes from the early centuries of the Common Era up to the thirteenth century may have facilitated the spread of Buddhism, as well as ensured its support by merchant communities. Early and mediaeval sites in the Western Deccan, the Konkan coast, Andhra Pradesh, Odisha, and Tamil Nadu were strategically located in the vicinity of ports along the trade routes connecting the mainland to Sri Lanka and further afield to Southeast Asia. This testifies to the increasing popularity of maritime travel in Buddhist communities from the sixth century onwards. The concurrent development in the same locales of ‘Saviour Cults’ – focusing on the bodhisattvas Avalokiteśvara, Tārā (especially in her aṣṭamahābhaya aspect), and Mahāpratisarā as protectors of travellers, and of sailors in particular, against the perils encountered along their journeys – may have been due to the increasing number of merchants and monks plying the commercial routes.Footnote 8 While it is undeniable that lay householders active in trade, crafts, and warfare played a role in patronizing and spreading Buddhism through pilgrimage, travel, and migration, the success of Buddhism overseas is to be attributed primarily to royal sponsorship. According to Ronald Davidson (Reference Davidson2002: 82–83, 167), an important factor in the rise and quick spread of esoteric fashions of Buddhism across Asia from the seventh century was the loss of mercantile support due to the dominance of Persian/Muslim traders on the Indian Ocean network, and the concomitant escalation in royal patronage. This was made possible through the intimate relationship between ritual specialists and the political elites who, lured by the promise of invincibility, protection for the state, and superhuman powers, often employed tantric monks as royal chaplains – thereby following the pattern that already existed between Brahmanical purohitas and the courts that they served. Significantly, many of the powerful dynasties that were instrumental in the sponsorship and spread of Sanskritic Buddhism ruled over domains located along the nodes of commercial and diplomatic maritime networks, such as the Pālas in Northeastern India, the Bhauma-Karas in Odisha, the Early Second Lambakaṇṇas in Sri Lanka, the Śailendras in Sumatra and Java, and the Tangs in China.
Historical evidence going back to at least the third century provides us with a picture of a steady traffic of itinerant monks travelling both eastwards and westwards along the sea paths linking a swathe of territory across the Indian Subcontinent and Japan, in search of texts, teachers, and patrons. It would seem that most of the monks travelling both ways between India and China preferred the maritime route to the overland one, or at least sought to include a maritime leg in their journey, which usually included stopovers in Sri Lanka and Nusantara. Seventh-century monk Yijing informs us that a significant number of the Chinese and Korean monks who went to India and Southeast Asia during his time travelled by sea on merchant ships (Pachow Reference Pachow1960: 211; Kandahjaya Reference Kandahjaya2004: 58). As recorded by a conservative scholarly estimate, 66 out of the 103 monks (of all ethnicities and geographical provenances) who were involved in the transmission of Buddhism to China used the maritime routes.Footnote 9
While the names and life circumstances of most of those anonymous agents are bound to remain unknown to us, Sino-Japanese biographies allow us to reconstruct the pedigree and social circle of some prominent monks who have gone down in history as vigorous translators, commentators, authors of original texts, initiators of lineages, royal advisers, and thaumaturges.Footnote 10 The first monk to travel from India to China via the maritime route was the Sogdian Kang Senghui, who arrived in Nanjing in AD 247. Many more monks are recorded to have reached China from South and Central Asia via Sri Lanka and Southeast Asia from the fourth to the sixth century. From the seventh century, a remarkably mobile and cosmopolitan network of monks is associated with the first wave of Mantranaya. This network includes *Vajrabuddhi (Jingangzhi 金剛智; 671–741) and his ordained pupil Amoghavajra (Bukong 不空; 704–774). We also know about an eighth-century Javanese monk, Bianhong, who went to China to study under Huiguo and composed an esoteric Buddhist initiation manual focusing on state protection (Sinclair Reference Sinclair and Acri2016a). The networks of seventh- to ninth-century itinerant monks offer a telling picture of the extraordinary period of intra-Asian maritime connectivity that became the hallmark of the rise and spread of esoteric Buddhist traditions over the course of just two or three generations.
As a result of socio-political contingencies,Footnote 11 the Buddhist traffic between India and China decreased from the ninth century, although smaller regional interlocking networks remained active, and the sea-based transmission of Buddhism between South and Southeast Asia was sustained through the following centuries at the hands of monks and other (possibly non-monastic) agents who are bound to remain anonymous due to the paucity of Chinese records of that period. Notable exceptions are the famous master Atiśa (*Adhīśa/Atīśa?, aka Dīpaṅkaraśrījñāna, AD 980–1054), who travelled from northeastern India to Sumatra (suvarṇadvīpa) and stayed there for twelve years to study with *Sauvarṇadvīpī-Dharmakīrti, one of the five most prominent Buddhist intellectuals of his time; the monk-translator Shihu (施護, *Dānapāla, d. 1018), apparently a native of Swāt (now in Western Pakistan), who knew the languages of Śrīvijaya, and must therefore have resided there for some time; and, almost five centuries later, the sixteenth-century Indian Buddhist Siddha Buddhaguptanātha, who is recorded by his student the Tibetan chronicler Tāranātha to have travelled extensively by sea from the Konkan coast to Sri Lanka, Southeast Asia (Java and Sumatra?), and back to visit Buddhist vestiges and communities (Tucci Reference Tucci1931; Templeman Reference Templeman2009).
Despite the lack of biographic records, the continuation or (re)establishment of long-distance contacts across the Indian ocean from the eleventh to the thirteenth century is suggested by certain notable facts. These include the royal-sponsored endowments sent by sea from Myanmar to Bodh Gaya (Singh Reference Singh, Singh and Dhar2014) and the (re)appearance of Nālandā-style imagery of Buddhist divinities in Angkor, Pagan, the Malay Peninsula, and East Java (Skilling Reference Skilling2007). Another example of this is the election (arguably through initiation) of transgressive and martial forms of ‘Phase Three’ tantric Buddhism as a personal and official cult by important royal figures such as Jayavarman VII in Cambodia (r. c.1181–1220), Kṛtanagara in East Java (r. 1268–1292), Kublai Khan in China (r. 1260–1294), and Ādityavarman in Java and Sumatra (r. ?–1375). The rise of these new networks of tantric Buddhism may have been triggered by unfavourable international political developments, most notably the decline of Buddhism in northern India at the end of the twelfth and beginning of the thirteenth century. When Nālandā, Vikramaśīla, and Uddaṇḍapura had been razed, scholars and artisans fled to Nepal and Tibet (von Schroeder Reference von Schroeder1981), and possibly further afield to Southeast Asia.Footnote 12 According to Tāranātha, most of the Buddhist scholars of madhyadeśa fled to mainland Southeast Asia (i.e. the kingdoms of Pegu, Campā, Kamboja, and so on) after Magadha was invaded by the Turks (Chattopadhyaya Reference Chattopadhyaya and Chimpa1990 [1970]: 330).
Sanskritic Buddhism in Nusantara–The Contribution of Nusantara to Sanskritic Buddhism
Nusantara was a strategic geographical area in the maritime Silk Roads system that has yielded significant vestiges of its glorious Hindu and Buddhist past, yet is still underrepresented in contemporary scholarship. Far from being a cultural backwater that passively received and ‘localised’ Indic influences, it had an integral place in the Buddhist Cosmopolis as both a crossroads and terminus of contacts from the early centuries of the Common Era. The contribution of Nusantara to the Indian Ocean trade network has been recognized through its provision of superior shipping technology, nautical terminology, and ship crews. However, the creative and constitutive force of Southeast Asian agents and socio-cultural milieux in the transfer, transformation, and translocation of people, texts, notions, and artefacts in the Buddhist world remains to be fully appreciated.Footnote 13 Witness the fact that a number of monks who travelled the sea routes and visited Southeast Asia – such as, for example, *Guṇavarman, *Paramārtha, Yijing, *Vajrabodhi/*Vajrabuddhi, Amoghavajra, and Atiśa – not only contributed to shaping the Buddhist paradigm in that region, but also stirred up new developments in China, Tibet, and the Indian Subcontinent itself (Kandahjaya Reference Kandahjaya2004: 79).
Sparse finds of Buddha images are documented across a vast area from Sumatra and Java to Kalimantan and Sulawesi; although their dating is uncertain, with estimates spanning from the second to the ninth century, they highlight the maritime mobility of Buddhism within the Archipelago. Vestiges of Buddhism, such as stūpikās, are also documented on the island of Bali, where communities of specialists in tantric Buddhist rituals have survived down to the present day. It is, however, in Sumatra and Java that the most Buddhist vestiges are concentrated. These two islands were strongholds of Mahāyāna Buddhist tantra from the seventh to the fifteenth century. Significant Buddhist vestiges include: the Batang Hari river sites in Muaro Jambi and Padang Lawas in Sumatra, the central Javanese temple complexes of Borobudur, the Kedu plain, and Ratu Boko; the royal temples of the Siṅhasāri and Majapahit kingdoms in east Java; and a number of textual documents preserved in inscriptions and manuscripts.
Java
Although Śaivism had been the dominant religion in Java for over a millennium in pre-Islamic times, Buddhism has also had a significant place in the religious, socio-political, and artistic life of the island. Chinese monk Faxian (337, 342–c.422) writes in his account that Buddhism in Java was not worth speaking of and that Brahmanism was very strong. However, just a few years after this, the Kashmirian monk *Guṇavarman, immediately before reaching China, succeeded in converting the royal family and their Javanese subjects to Buddhism. This information could be matched with the extensive, and nearly contemporary, archaeological remains of Batujaya in northwestern Java (Manguin and Indrajaya Reference Manguin, Agustijanto, Manguin, Mani and Wade2011), and with seventh–early ninth century epigraphic evidence featuring Buddhist formulas traceable to a fifth-century Sarvāstivāda milieu (de Casparis Reference de Casparis1956: 75). Another report by Yijing indicates that, by the middle of the seventh century, the Sarvāstivada school was prevalent in Southeast Asia,Footnote 14 and that the Chinese monk Huining visited ‘Kaliṅga’ (holing) in central Java from AD 665 for three years, to study and translate Āgama literature with a local (波凌人) śramaṇa called Jñānabhadra. Huining's pupil, Yunqi, after going back to China to present a Chinese translation of the Mahāparinirvāṇasūtra, returned to Southeast Asia, spending more than ten years in Java and living in Śrīvijaya up to the time of Yijing.Footnote 15
Art historical and textual remains point to the existence of close Buddhist links between western India, the Malay Peninsula, and Java in the eighth century. The seemingly ‘archaic’ character of eighth–ninth century Mantranaya Buddhism in Java and Sumatra, when the two islands were under Śailendra rule, suggests that the early esoteric developments that were emerging from the western Deccan promptly reached Nusantara via the maritime routes; these became the basis of, and were preserved in, later religious configurations. For instance, the ‘courtly’ eighth-century Eight Bodhisattvas known from some of the Ellorā caves, Ratnagiri, and the Malay Peninsula, where they are part of a maṇḍala arranged around a Śākyamuni, are depicted on the exterior walls of Candi Mendut and in the interior of Candi Plaosan (Bautze-Picron Reference Bautze-Picron and Eilenberg1997). Significantly, this maṇḍalic formation is described in the ‘proto-tantric’ Aṣṭamaṇḍalakasūtra, translated into Chinese by Amoghavajra (T 1167) and, a century earlier, by *Puṇyodaya (Nati 那提, T 486). Amoghavajra travelled to China via Java, and *Puṇyodaya was active in Cambodia (Woodward Reference Woodward2004: 336; Lin Reference Lin1935: 83–100). Indeed, similarities between the iconography of the Eight Bodhisattvas in Java, Shingon Buddhism, and eighth-century Chinese translations have been discerned by Bautze-Picron (Reference Bautze-Picron and Eilenberg1997: 28). A possible iconographical influence stemming from the later ‘esoteric’ phase of Ellora and other western Indian caves is detectable on the sculpted triptych of Mendut (Revire Reference Revire, Dallapiccola and Verghese2018) and, perhaps, Candi Bogang in central Java.Footnote 16 This could represent the triad of Avalokiteśvara-Padmapāṇi, Śākyamuni, and Vajrapāṇi. These deities are seen as the three heads of bodhisattva families in texts belonging to the Kriyātantra genre, such as the Susiddhikarasūtra and the Mañjuśriyamūlakalpa. This confirms the archaic, pre-systematized character of Mantranaya or the Mahāyāna Buddhist tantra in the Śailendra domains in the eighth and ninth centuries.Footnote 17 Another echo of the Susiddhikara and other early tantric texts is found in a pan-Asian pericope including the sequence namaś caṇḍavajrapāṇa-ye/-yiṣya mahāyakṣasenāpat-aye/-iṣya, found in incantations recovered from epigraphic documents from Java and Bali, such as the Mahāraudra-nāma-hṛdaya (Griffiths Reference Griffiths2014a: 183–184, Reference Griffiths and Tropper2014b). Two maṇḍala-stakes (kīla) from central Java, palaeographically dateable to the ninth or tenth century, contain versions of mantras also found in the Guhyasamāja and the ninth-century ritual manual Sarvavajrodaya by Ānandagarbha, and version B of the Sarvadurgatipariśodhana (Reference Griffiths2014a: 170–171). Although the exact relationship between dhāraṇīs preserved in written texts (manuscripts) and inscriptions is often difficult to disentangle, especially concerning the issue of directionality, the existence of close parallels in the Sarvavajrodaya suggests that the latter text might have been the source of the central Javanese material, in which case it must have reached the island soon after its composition. The availability of Mantranaya material at an early period in Java, including a gold foil from Candi Plaosan Lor containing the exact same version of a dhāraṇī as can be found in the Sanskrit texts brought to Japan by Kūkai in the early eighth century (Griffiths Reference Griffiths2014a: 165), as well as a representation of the Gaṇḍavyūha – a text that Osto (Reference Osto2009) regards as ‘proto-tantric’Footnote 18 – on the Borobudur, suggests not only that a ‘commonality of religious culture’, especially with respect to dhāraṇī-focused Buddhism, existed across Nusantara and other locales of the Buddhist Cosmopolis, but also that Nusantara may have played a role in the circulation and development of ideas and practices.
The major architectural undertakings of the Śailendra dynasty in the eighth and ninth centuries included the majestic and exquisitely crafted Buddhist monuments of Borobudur, Candi Mendut, Candi Sewu, Candi Kalasan, and Candi Plaosan (Plate 1). Borobudur remains the largest Buddhist monument ever built and stands unrivalled in its architectural complexity, symbolical depth, and artistic quality.Footnote 19 Candi Kalasan has yielded one of the biggest bronze bells ever recovered from a Buddhist monument, and may have hosted a colossal statue of Tārā that is, unfortunately, no longer extant. These temples, constituting artistic and architectural masterpieces, must have ranked among the great sacred centres of the Buddhist Cosmopolis, and attracted a steady traffic of monks and pilgrims to the island. For instance, a ninth-century Siddhamātṛkā inscription unearthed at Plaosan in the Prambanan area describes the worship at a Buddha-temple (jinamandira) by pilgrims continuously arriving from Gurjaradeśa in north India (de Casparis Reference de Casparis1956: 188–189, 202). An account by Yuanzhao compiled into the Zhenyuan xinding shijiao mulu, as well as a report by the Japanese master Kūkai, records that *Vajrabuddhi first met Amoghavajra in Java (Chou Reference Chou1945: 321; Sundberg and Giebel Reference Sundberg and Rolf2011: 152). An early eleventh-century illustrated manuscript of the Aṣṭasāhasrikā-Prajñāpāramitā (CUL Add. 1643) dedicates one of its first vignettes to an image of the Buddha Dīpaṅkara in Java (f. 2r),Footnote 20 and Java figures in the early ‘tantric geography’ exposed by the Manjuśriyamūlakalpa (51.636–640).
Plate 1. Candi Plaosan, central Java. (Photo Andrea Acri, 2016).
Links between masters from the Pāla-Sena domains and Śailendra-sponsored Buddhism may be inferred from the Kelurak Sanskrit/Siddhamātṛkā inscription of AD 782, recording Kumāraghoṣa, the royal preceptor from Gauḍīdvīpa who installed an image of Mañjughoṣa (Mañjuśrī) at vajrāsana mañjuśrīgṛha (Candi Sewu in central Java?) at the request of Śailendra King Śrī Saṅgrāmadhanañjaya (Sarkar Reference Sarkar1971, I: 37, 45). The importance of Mañjuśrī in Java, and especially its connection to political power, finds a parallel in the cult of the same deity in esoteric Buddhist circles at the Tang court during roughly the same period.Footnote 21 The rājaguru Kumāraghoṣa of the Kelurak inscription is probably to be identified with the anonymous śailendrarājaguru praised in the Siddhamātṛkā Sanskrit inscription of Kalasan of AD 778.Footnote 22 This document opens with a salutation to Āryā-Tārā, and mentions the construction of a temple for Tārā (tārābhavana), which Jordaan (Reference Jordaan1998) identifies with what is now called Candi Kalasan. In a fragmentary line of the Siddhamātṛkā Sanskrit inscription of Plaosan, the wife of an unidentified royal figure is said to “shine forth like Tārā” (tāreva virājati) (Long Reference Long2014: 244); the same simile (tāreva) occurs in the Nālandā copper plate of Devapāla, linking that goddess to the queen-mother of Bālaputra – a Śailendra royal figure known in mid-ninth-century Sumatra and Java (Sastri Reference Sastri and Krishna Sastri1923–24: 326). On the basis of this epigraphic evidence and art historical material, Jordaan (Reference Jordaan1997: 287–288) has hypothesized that the Tārā cult in Java and Sumatra was connected to the contemporary Pāla royal cult of (Śyāmā-)Tārā. Be this as it may, both areas are likely to have acted as an important place in the development of the cult of Tārā, which had an inherent maritime aspect – this goddess being a tutelary deity of travellers, and seafarers in particular (Bopearachchi Reference Bopearachchi, Dhar and Singh2014; Ray Reference Ray, Skilling and McDaniel2012: 56–60).
A steady Buddhist traffic must have linked Java and Sri Lanka, as evidenced by biographies of monks, as well as epigraphic and art historical evidence from both islands. An eighth-century foundation inscription, again in the Sanskrit language and Siddhamātṛkā script, records that a branch of the Sri Lankan Abhayagirivihāra was established by the Śailendras on the Ratu Boko promontory in central Java (Plate 2). The Ratu Boko structures were apparently intended for the use of esoteric-minded Sinhalese Buddhist ‘rag-wearing monks’ (pāṃśukūlika),Footnote 23 who were active in the eclectic monastic institution at Anurādhapura where a variety of Buddhist texts were studied. Significantly, the Abhayagirivihāra-related structures of Ratu Boko share common architectural motifs, such as the peculiar double meditation platforms, with their Sinhalese prototypes (Sundberg Reference Sundberg and Acri2016). Indeed, the area in the Kedu plain where Candi Sewu and the Prambanan temple complex were built appears to have been termed Laṅkapura (‘The Laṅkan City’) by then, as if to recreate a local ‘replica’ of Buddhist Sri Lanka (Griffiths Reference Griffiths, Acri, Creese and Griffiths2011a),Footnote 24 and multiple Abhayagiris could have existed elsewhere in Southeast Asia: namely in southern Cambodia, southern Vietnam, and peninsular Thailand (Griffiths Reference Griffiths2013: 75).
Plate 2. The Ratu Boko monumental gate, Central Java. (Photo Andrea Acri, 2016).
Evidence for the pan-Asian cult of Mahāpratisarā, a female deity related to Tārā, has been recovered from Java (Mevissen Reference Mevissen1999: 102–103; Cruijsen, Griffiths, and Klokke Reference Cruijsen, Griffiths and Klokke2012). The popularity of this deity in both Central Java and the centre of the Tang empire in China, namely Dunhuang and Changan, from the eighth to the tenth century, seems to be linked by the activities of Amoghavajra (Mevissen Reference Mevissen1999: 117), who not only recited the Mahāpratisarādhāraṇī in order to avert shipwreck while travelling to China by sea (Chou Reference Chou1945: 275, n. 19; Sundberg and Giebel Reference Sundberg and Rolf2011: 139), but actually translated it and submitted a copy of it to Emperor Suzong (T 2120.829b2–21) in AD 758.Footnote 25 A gold foil recovered from the approximately tenth-century Cirebon shipwreck off the Java northern coast, containing a dhāraṇī addressed to a goddess personifying the incantation, and paralleling material found in the Ekādaśamukhadhāraṇī and Sādhanamālā confirms that objects like this were actually worn by passengers and/or crews as amulets intended to protect against the dangers of voyages at sea (Griffiths Reference Griffiths2014a: 157–159).
Woodward has apprehended the influence of Borobudur on Buddhism in the subcontinent: according to him, this monument could represent an early phase of the system of subtle bodily centres called cakras or sthānas/ādhāras (Reference Woodward2009: 48). Furthermore, the same scholar (Reference Woodward1990: 16–17) has advanced the hypothesis that a tenth-century inscription quoting verse 46 of the Bhadracarī-praṇidhāna (Schopen Reference Schopen1989) – the text that came to be transmitted as the final chapter of the Gaṇḍavyūha depicted in the uppermost series of reliefs on Borobudur – on a memorial stūpa found at the Nālandā monastery, which was established in the ninth century by Bālaputradeva for the use of pilgrims from Śrīvijaya, might represent a new stimulus introduced by the Sumatran monarch, or in any event the existence of “longstanding similarities in religious practice” in the two areas (Reference Woodward2004: 353). This verse is notable in that it is the only passage from a Mahāyāna text so far recorded in an Indian inscription. It also survives in another inscription that forms part of the Sambas Buddhist hoard found in Western Kalimantan, which contains more verses traceable to Sanskrit texts, namely the Mahāpratisarāmahāvidyārājñī and the Vajracchedikā Prajñāpāramitā (Griffiths Reference Griffiths2014a: 146–147).
Vestiges of Vajrayāna Buddhism in ninth- and tenth-century Java may be found in the groups of bronzes from Surocolo and Nganjuk, which have been tentatively identified as representing tridimensional esoteric maṇḍalas dominated by Vajrasattva, one of the central deities of ‘Phase Two’ and ‘Phase Three’ Esoteric Buddhism (Tanaka Reference Tanaka2010: 339). Since the portrayal of this pantheon in paintings or statuary is relatively scarce in the Buddhist world, the Surocolo hoard is especially significant (Szántó and Griffiths Reference Szántó, Arlo, Silk, von Hinüber and Eltschinger2015: 372). Another hoard, found at Ponorogo, has been associated to goddesses of the Vajrasattvakulamaṇḍala and the Śrīparamādya seventeen-deity maṇḍala in its connection with the Sarvabuddhasamāyoga on the basis of commentaries in Tibetan translation (Matsunaga Reference Matsunaga2017). A small statuette has been recovered from the cargo of the Intan shipwreck salvaged from the waters of Java Sea, which may have belonged to one such tridimensional maṇḍala (Liebner Reference Liebner2014: 191–194). These findings testify to the keen interest in initiatory maṇḍalas in Java: significantly, the only text attributed to the Javanese monk Bianhong is the ‘Ritual Manual for Initiation into the Great Maṇḍala of the Uṣṇīṣa-Cakravartin’ (T 959), and most of the Sanskrit verses cited in the Old Javanese Saṅ Hyaṅ Kamahāyānan Mantranaya – a text that identifies its main Buddha with Vajrasattva, the transcendent essence of all Tathāgatas – are traceable to esoteric Buddhist initiatory manuals related to the Guhyasamāja (Kandahjaya Reference Kandahjaya and Acri2016). While the date of this text is not certain, a related Śaivized version has a colophon mentioning the name of King Mpu Siṇḍok of the Īśāna dynasty (r. 929–947), and its early core might date back to the eighth or ninth century.
Material stemming from the Trailokyavijaya cycle in the Sarvatathāgatatattvasaṅgraha, and also associated with more wrathful Krodha-vighnāntaka deities, is known from Java. Trailokyavijaya/(Caṇḍa)Vajrapāṇi is invoked in incantations that can be found in epigraphic documents from Java and Bali (Griffiths Reference Griffiths2014a: 183–184, Reference Griffiths and Tropper2014b), and at least four images of Trailokyavijaya have been recovered from central Java (probably from the tenth century). Standing in pratyālīḍha posture trampling over Śiva and Umā, these images closely follow Eastern Indian (Pāla) iconographical prototypes. However, the Mahāraudra-nāma-hṛdaya that was recovered near Borobudur attests to a major iconographical discrepancy with the known sculptural corpus all over the Buddhist world, namely the placement of the right foot on Śiva and the left foot on Pārvatī’s chest, instead of the opposite. To Griffiths (Reference Griffiths and Tropper2014b: 32), this suggests a genuine, as yet unattested iconographic variant, rather than a textual error. The ‘anti-Śaiva’ overtones of this incantation are also found in a short mantra engraved on a gold foil from Ratu Boko in central Java as well as on similar artefacts (Acri Reference Acri and Acri2016b; Griffiths Reference Griffiths2014a: 177–180), which are related to the krodha-vighnāntaka deity Ṭakkirāja and the Sarvatathāgatatattvasaṅgraha cycle.Footnote 26
An emphasis on ‘Phase Three’ Vajrayāna is detectable in the statuary and epigraphic remains associated with King Kṛtanagara of Siṅhasāri, who was apparently initiated into transgressive, non-dual Buddhist and Śaiva tantric traditions and founded a royal cult identifying the monarch with ‘Śiva-Buddha’. The later chronicle Deśavarṇana (43.2–5) indicates his commitment to Buddhism, records his initiation name Jñānabajreśvara,Footnote 27 and associates him with the teachings of the unidentified Subhūtitantra,Footnote 28 as well as the gaṇacakra rituals that are known to us via the Guhyasamāja. Indeed, Kṛtanagara had a Sanskrit-inscribed, colossal statue portraying him as Mahākṣobhya – the principal deity of the Guhyasamāja – installed at the cremation ground of Wurare in AD 1289. Another 5.5 metre-high colossal statue of Akṣobhya going back to the same period lies unfinished at the Reco Lanang site in the vicinity of Trawas in the Mojokerto District of east Java (Plate 3).
Plate 3. The Reco Lanang unfinished Akṣobhya statue, east Java. (Photo Andrea Acri, 2016).
A northeastern Indian and/or Nepalese (Newar) influence on the east Javanese Buddhist art and inscriptions of the Siṅhasāri period has been noted, for instance in the statuary and decorative features (O'Brien Reference O'Brien1993: 252–255; Lunsingh Scheurleer Reference Lunsingh Scheurleer2008: 296–298). Schoterman (Reference Schoterman, Klokke and Scheurleer1994: 168) noted that the five main statues of Bodhisattvas at Candi Jago were executed according to the teachings of the Sanskrit Amoghapāśasādhana, which was written by Śā-kya-śrī-bha-dra in northern India around the year 1200 and may have reached Java shortly thereafter. However, Sinclair (Reference Sinclair2016b: 150) has convincingly argued that the Malay and Javanese forms of eight-armed Amoghapāśa Lokeśvara may predate the textual legacy from the Subcontinent, and may have therefore inspired the Indian and Nepalese specimens – perhaps via a statue installed at Bodh Gaya by an Indonesian donor.Footnote 29 The Nāgarī-inscribed Buddhist statues from Candi Jago, and the occurrence of the words bharāla (‘god’) and bharālī (‘goddess’) in a number of inscriptions associated with Kṛtanagara, support a possible northeastern Indian, and especially Newar, link.Footnote 30 Interestingly, Nepalese artisans became popular at the courts of both Khubilai Khan in China and his sworn adversary Kṛtanagara in east Java.Footnote 31 The existence of links between Sumatra (and Java?) and northeastern India in the thirteenth century may be evinced by the Pasir Panjang rock inscription in Sanskrit and Rañjanā script at Karimun Besar in the Riau Archipelago, recording a mahāyānika-gauḍa-paṇḍita-śrīgautamaśrī: that is, a Mahāyāna scholar from Bengal named Gautamaśrī. This personality, whose identity and presence in Nusantara have been recently discussed by Iain Sinclair (Reference Sinclair2018), was active in Nepal (specifically at Guitaḥ in Lalitpur) and Tibet before the middle of the thirteenth century, at a time when northeastern India had just been raided by the Turks and Buddhism was on the decline. The fact that this inscription was written in Sanskrit in a non-local cosmopolitan script, and that it displays a short, graffiti-like (yet well-executed) character, lends credit to the hypothesis that it could have been engraved by Gautamaśrī himself en route to mainland Sumatra and/or Java (according to Sinclair, not long after the turn of the first half of the thirteenth century).
In the Majapahit period, Buddhism was a constitutive (albeit seemingly subordinate) member of the religious coalition formed by Śaivism (śaivapakṣa), Buddhism (sogatapakṣa), and the Ṛṣis. While the term ‘syncretism’ is not appropriate to describe this situation of religious pluralism and inclusivism, it must be noted that the truly hybrid cult of Śiva-Buddha introduced by Kṛtanagara appears to have become popular in this period, as it is frequently encountered in the religious literature as well as in the belles lettres (Acri Reference Acri and Lammerts2015). A fourteenth-century Old Javanese literary source, the kakavin Sutasoma by Mpu Tantular, attests to Esoteric Buddhist cults and a form of Mahāyānic tantra that, despite prescribing the worship of Śiva-Buddha (i.e. Mahāvairocana) and preaching a form of identity between the two supreme deities, keeps the Buddhist and Śaiva ways neatly separate. A similar situation is reflected in the fourteenth-fifteenth-century kakavin Kuñjarakarṇa by Mpu Ḍusun, which has been shown to contain elements from the Sanskrit Kāraṇḍavyūhasūtra (Nihom Reference Nihom1994: 136–139). From the chronicle Deśavarṇana we learn about the existence in East Java of two types of private Buddhist institutions, namely the kavinayan and a (more prevalent) kabajradharan (80.1); the former implies that monastic rules were followed, while the latter is of a tantric type where celibacy was not enforced. The text (77.3) associates a locale named buḍur (Borobudur?) with the latter type of establishment (Sinclair Reference Sinclair2013), and also informs us that Java was visited by foreign monks (bhikṣu) coming from mainland India, such as a certain Srī Buddhāditya from Kāñcī, who made an eulogy of the king “in countless verses” (93.1).Footnote 32 While Buddhism declined and eventually disappeared after the fall of Majapahit, it seems that small communities were still extant in the sixteenth century: the Buddhist Siddha Buddhaguptanātha (Tāranātha's guru), travelling from the Subcontinent to Java at that time, allegedly found the followers of the ‘Sravaka Sendhapa’ there,Footnote 33 and then proceeded to a small island in the middle of a lake, called Vaṇadvīpa (‘Forest Island’, perhaps in Sumatra?), where he saw the cave of Padmavajra (mTsho skyes rdo rje) and a two-armed image of Hevajra installed in a square-shaped temple, and heard about the existence of many tantric scriptures (Templeman Reference Templeman2009: 264). The tantric legacy and Śaiva-Buddhist coalition of Siṅhasāri and Majapahit appears to have continued in Bali to the present day, where Balinese Buddhist ritualists cover a subsidiary role in certain Śaiva ceremonies, and have preserved Sanskrit scriptures that include fragments from several authoritative works of the Mahāyāna and Vajrayāna (Hooykaas Reference Hooykaas1973: 600–603).
Sumatra
Sumatra hosted renowned centres of Buddhist activity and learning. Seventh-century Chinese monk Yijing praised the high level of Buddhist scholarship that he found on the island, where he stopped en route from Guangzhou to Nālandā, and from where he procured the shipment of numerous Buddhist texts (Takakusu Reference Takakusu1896: xxxvi). Yijing reports that Śākyakīrti, one of the five most distinguished Buddhist teachers of his time, travelled far and large across the ‘Five Indias’ and finally settled in Śrīvijaya (Śrībhoja) (Takakusu Reference Takakusu1896: 184). The continued existence of a high level of Buddhist scholarship and royal sponsorship of the religion is also suggested by the later figure of Shihu/*Dānapāla (施護, d. 1018), an exceptionally prolific monk-translator who in the late tenth century reached China with a good knowledge of the languages of Sanfochi/Śrīvijaya and Shepo/Java (Sen Reference Sen2003: 384; Orzech Reference Orzech, Orzech, Sørensen and Payne2011: 449–450). Skilling (Reference Skilling1997: 188) notes that the composition, in the early eleventh century, of the Durbodhāloka by *Dharmakīrti from Suvarṇadvīpa (which he locates in Kedah) “presupposes the existence and study in Śrīvijaya of the abstruse Prajñāpāramitā and Abhisamayālaṃkāra literature; of a high level of scholarship; and of royal sponsorship”.
Sumatra has yielded a corpus of stone tablets and other inscriptions on metal foils and statues that are notable in that they cite early Mahāyāna Sanskrit texts, including the Suvarṇabhāsottamasūtra, which is regarded as ‘proto-tantric’ by Huntington (Reference Huntington1987), and the Aparimitāyuḥsūtra (Griffiths 2011b: 169, Reference Griffiths2014a: 152–154). The Talang Tuo Old Malay inscription of AD 684 recovered near Palembang has also been interpreted as an early document of Mantranaya Buddhism (Woodward Reference Woodward2004; Kandahjaya Reference Kandahjaya and Acri2016: 82). The epigraphic legacy, coupled with the account by Yijing on texts on yoga and incantations (vidyā, dhāraṇī) available in Śrīvijaya,Footnote 34 and the Sanskrit-Old Javanese manual(s) Saṅ Hyaṅ Kamahāyānikan/Saṅ Hyaṅ Kamahāyānan Mantranaya, would seem to cast new light on the genesis of Esoteric Buddhism across the Buddhist cosmopolis (Kandahjaya Reference Kandahjaya and Acri2016).
The Intan and Cirebon shipwrecks, discovered beneath the sea lanes linking Sumatra to Java, have yielded precious data on tenth-century traffic of Buddhist bronze paraphernalia (including vajras, spear-shaped sceptres, bells, statuettes, and inscribed foils) along with other commonly traded merchandises among regional entrepôts and the larger Indian Ocean and Chinese markets.Footnote 35 On the Sumatran mainland, most of the archaeological remains and scant epigraphic documents spread over disparate locales of the island – especially along the Batang Hari river (for instance, Muara Jambi and Muara Takus) – have yielded remains of Buddhist monuments and inscriptions mostly dating back to the tenth to thirteenth centuries.Footnote 36 The 21 gold plates with inscribed seed-syllables and names of deities found at Candi Gumpung in Sumatra follow an arrangement that most closely resembles that of the Trailokyavijayamahāmaṇḍala as described in the Vajraśekharatantra, thus suggesting that Trailokyavijaya must have been the central deity of the maṇḍala (Nihom Reference Nihom1998).
From the eleventh century onwards, Sumatra appears to have been an important seat of the Hevajra cult: witness the epigraphic evidence of mantra portions directly quoted from the Hevajratantra (Griffiths Reference Griffiths and Perret2014c: 230),Footnote 37 as well as the fourteenth-century inscription of Saruaso II, issued by crown prince Anaṅgavarman, son of Ādityavarman, which mentions his “daily meditation on Hevajra”, or his “ever keeping in mind the Hevajra(tantra)” (hevajranityāsmṛtiḥ; see Hunter Reference Hunter and Kozok2015: 324–327). It would thus seem that Ādityavarman was following the same ideology and ritual technology adopted earlier by Kublai Khan and Kṛtanagara, who equated themselves to the central deity of the maṇḍalas of Buddhist tantras such as the Guhyasamāja and the Hevajra. The wild, ferocious character of the Esoteric Buddhist (and tantric Śaiva) iconography that developed at the east Javanese courts of Kaḍiri-Siṅhasāri and Majapahit shares features with the Sumatran iconography, as attested in Biaro Bahal and Muara Takus, and in the Mahākāla statue of Padang Roco, attributed to the fourteenth-century Ādityavarman. As suggested by Bautze-Picron (Reference Bautze-Picron and Perret2014: 107), images found at the Sumatran sites of Padang Lawas are part of a network that connects them to east Java, South Asia, Cambodia and Campā in the eleventh and thirteenth centuries; overall, the sites show a kind of Buddhism that belongs to the same phase of Vajrayāna as that which was present in the Khmer and Cham domains between the twelfth and thirteenth centuries, and in east Java and China in the thirteenth century.
The contribution of Śrīvijaya to Vajrayāna Buddhism in Tibet, which can also be seen through the handful of texts composed in Suvarṇadvīpa that were introduced into the Tibetan canon, is acknowledged by the Tibetan tradition from the eleventh century, and confirmed by modern scholarship (see Chattopadhyaya Reference Chattopadhyaya and Chimpa1981: 93–94; Schoterman Reference Schoterman and Acri2016 [1986]). Atiśa, a native of Bengal, is said to have transmitted the Durbodhāloka (a Sanskrit commentary on the Abhisamayālaṅkāra) to Tibet. This was composed in Southeast Asia by Atiśa's teacher *Sauvarṇadvīpī-Dharmakīrti under King Cūḍāmaṇivarman, who may have founded a Buddhist temple in Nāgapaṭṭinam around AD 1019 (Skilling Reference Skilling1997, Reference Skilling2007).Footnote 38 This Dharmakīrti may have been the master who imparted to Atiśa the teachings of the Kālacakratantra, and has been identified by van der Kuijp (Reference van der Kuijp2003: 420, n. 6) as the author of the Netravibhaṅga, a commentary to the Hevajratantra. Both Atiśa and *Dharmakīrti were fervent devotees of Tārā, whose cult was widespread in insular Southeast Asia, and which may have been popularized in Tibet by Atiśa after his stay in Suvarṇadvīpa (Schoterman Reference Schoterman and Acri2016 [1986]: 119). Wilkinson (Reference Wilkinson and Brown1991: 236–239) presents evidence for a connection between the cult of Gaṇeśa in Tibet and *Dharmakīrti. The transmission of Buddhist ideas from Sumatra and/or Java to the Himalayan region has been suggested on the basis of artistic and architectural similarities between the Tabo monastery in Himachal Pradesh, which Atiśa visited in AD 1042, and Borobudur.Footnote 39 Recently, Sinclair (Reference Sinclair2016b: 164–166) has also presented a case that the eight-armed form of Amoghapāśa in the Thaṃ Bahī monastery in Nepal (founded by Atiśa) was introduced from the Śrīvijayan domains.
The Malay Peninsula
Given its strategic geographical location, the Malay Peninsula, which was included in the domains of the staunchly Buddhist Śailendra-Śrīvijayan rulers from the seventh to the thirteenth century, acted as an important intersection in the traffic of merchants, monks, and pilgrims plying the maritime routes. The middle part of the Peninsula was dominated by the city-state of Panpan from approximately the fifth to the eighth century, and has yielded Buddhist remains consisting primarily of statues, tablets, and votive stūpas. The city-state of Laṅkasukha extended over the region of present-day Pattani on the east coast, and has yielded the same type of material evidence plus Buddhist brick sanctuaries. According to Jacq-Hergoualc'h (Reference Jacq-Hergoualc'h2002: 183), this legacy reflects the dominance of the Mūlasarvāstivāda school up to the end of the seventh century, and the concomitant rise of Mahāyāna Buddhist tantra from the seventh century onwards.
The earliest inscription associated with Buddhism in Southeast Asia, dated to the fifth century, has been found in Kedah. Having been commissioned by sea captain Buddhagupta from Raktamṛttikā (probably Raktamṛttikā Mahāvihāra in what is now Rajbadidanga in West Bengal, or another unknown location in mainland Southeast Asia), it highlights the link between Buddhism and trade in the peninsula prior to the establishment of monastic centres and the rise of institutional support in the Śrīvijayan period (Sen Reference Sen, Wong and Heldt2014c: 46–48). Another important testimony to the translocal nature of Buddhism in the area is the existence of many votive tablets found at multiple sites in the period from the sixth to the twelfth century. Two of them, recovered from sites in Kedah and tentatively dated to the seventh century, quote a passage from the Mahāyāna Sūtra Sāgaramatiparipṛcchā, whose Sanskrit original has survived through fragmentary quotations but has been integrally transmitted in Chinese and Tibetan translations (Lammerts and Griffiths Reference Lammerts, Griffiths, Silk and Eltschinger2015: 994). Several clay tablets recently found in the Khao Nui cave in Trang province bear an abbreviated inscription of the Four Truths of Buddhism in Sanskrit (Revire Reference Revire and Tan2015: 301–303, figs. 26.4–26.6), while other specimens display a Mahāyāna/early Mantranaya iconography (including, for example, the Eight Bodhisattvas). The specimens inscribed in northeastern Indian scripts, as well as those displaying a twelve-armed Avalokiteśvara from the Perlis caves, could have belonged to pilgrims transiting from the Subcontinent (unlike the tablets recovered from difficult-to-reach caves, which rather suggest a local context of religious practice: see Jacq-Hergoualc'h Reference Jacq-Hergoualc'h2002: 47).
Besides controlling the trade routes, the Śrīvijaya thalassocracy extended at its height over Sumatra, the Malay Peninsula, and Java, and was actively engaged in Buddhist diplomacy with India and China. In addition to the donation by Bālaputradeva at Nālandā, other gifts and diplomatic exchanges with South India and Sri Lanka are documented in inscriptions dating from the eleventh and twelfth centuries (Jacq-Hergoualc'h Reference Jacq-Hergoualc'h2002: 274–275, 346–347, 400). Under Śailendra rule, the Malay Peninsula, Sumatra, and Java shared several elements of Sanskritic Buddhism. The early triad of bodhisattvakulas of the Kriyātantras that we find at Mendut is mentioned in the Chaya/Vieng Sa (‘Ligor A’) inscription of AD 775 as Kajakara (Padmapāṇi-Avalokiteśvara), Māranisūdhana (Śākyamuni), and Vajrin (Vajrapāṇi), where a brick temple was dedicated by the royal chaplain (rājasthavira) Jayanta and his disciple Adhimukti at the instigation of a Śailendra king (Long 2014: 25–27). The exquisitely crafted late eighth-century bronze Avalokiteśvaras found in the Chaiya district of modern Thailand and in Bidor (Perak, Malaysia) show close similarities with the Avalokiteśvara found at Wonogiri in central Java (Fontein Reference Fontein and Fontein1990: 210–211). These remains suggest a link between the Malay Peninsula and Java under the Śailendras, perhaps also via *Vajrabuddhi (Sharrock and Bunker Reference Sharrock, Emma and Acri2016), Amoghavajra, *Prajña, and other monks.
Links between Nusantara and Nepal can be evinced from the AD 1015 Manuscript CUL Add. 1643 (f. 120r) of the Aṣṭasāhasrikā Prajñāpāramitā, which mentions a Lokanātha on Mount Valavatī in Kedah/Kaṭahadvīpa (Kim Reference Kim2014: 49, 63, 65).Footnote 40 De Mallman (Reference de Mallman1951) and Soper and Chapin (Reference Soper and Helen B.1970) proposed that the so-called ‘art of Śrīvijaya’ was influential on the Avalokiteśvara images produced in Yunnan around the eleventh to thirteenth century. De Mallman (Reference de Mallman1951: 573–574) also noted a similarity between certain stylistic features of these statues that, while being absent from Pāla art, are shared by specimens found in places located along the maritime networks, including the western Deccan caves, South India, Sri Lanka, Java, Sumatra, the Malay Peninsula, Campā, and Japan. This state of affairs led her to conclude that these motifs were of South Indian origins, yet they were diffused through the intermediary of Śrīvijaya.Footnote 41
Acknowledgements
I am grateful to Iain Sinclair for his most valuable comments and insights on a draft of this article; any mistakes are exclusively my own. I acknowledge the support of the National Research Foundation of Korea Grant [NRF-362-2008-1-B00018], thanks to which I was able to present this paper at the conference “Maritime Silk Road in Southeast Asia: Crossroad of Culture”, held at the National Museum of Korea, Seoul, in December 2017.
