Articles by alphabetic order
A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
 Ā Ī Ñ Ś Ū Ö Ō
1 2 3 4 5 6 7 8 9 0


Providing for the Buddha: Monastic Centres in Eastern India

From Tibetan Buddhist Encyclopedia
Jump to navigation Jump to search



Himanshu Prabha Ray

Providing for the Buddha: Monastic Centres in Eastern India


When he had been consecrated eight years the Beloved of the Gods, the king Piyadassi [[[Asoka]]] conquered Kalinga [identified with large parts of modern Orissa]. A hundred and fifty thousand people were deported, a hundred thousand were killed and many times that number perished. Afterwards, now that Kalinga was annexed, the Beloved of the Gods very earnestly practised Dhamma, desired Dhamma and taught Dhamma (XIII Major Rock Edict of Asoka).


Fig. 1. Map showing location of Asokan inscriptions


As with other regions of the subcontinent, the edicts of the Mauryan ruler Asoka (272-232 BC) provide the earliest evidence for the presence of Buddhism in Eastern India. These may be classified into Rock Edicts (Major and Minor), cave inscriptions, and Pillar Edicts (fig. 1). The Minor Rock Edicts are also said to form the earliest group of Asokan inscriptions, followed by the Major Rock Edicts. The Pillar Edicts were the last to be promulgated, but no example of these has so far been found in peninsular India (Salomon 1998: 136). The Major Rock Edicts consisting of a set of fourteen proclamations inscribed on rocks have been found across the Indian subcontinent from Shahbazgarhi in Peshawar district of modern Pakistan to sites in Andhra Pradesh in south India. Rock Edict XIII is the largest in the group and it is here that Asoka expresses remorse at the loss of life in Kalinga. This edict is however replaced in the edicts inscribed in Kalingan territory at Jaugada on the north bank of the Rishikulya river in Ganjam district and at Dhauli in Puri district, but there is little doubt regarding Asoka s’ mission in spreading the Buddhist dhamma. The Dhauli and Jaugada edicts indicate that they were to be proclaimed on the eighth day of the star Tisya and were to be read aloud.

Another significant feature is the association of Asokan edicts in Orissa with the elephant, as evident from the rock-cut elephant at Dhauli above the inscribed stone. Other examples of this are known, for example at Kaima in Jajpur district. Kaima is a small hill on the banks of the river Kimira about one kilometre from its confluence with the river Brahmani and is 1

Fig. 2. Rock-cut elephant at Kaima, Jajpur district.


located to the north of Langudi hill. A rockcut elephant is known from the site located within what was perhaps an enclosure, but the only remnants now are four khondalite pillars (fig. 2). Structural vestiges of the Mauryan period continue to be scarce and include only a few examples, such as a bell capital found in a tank near the Ramesvara temple at Bhubaneshwar. Railing pillars dug up near the Bhaskareshwar temple also at Bhubaneshwar indicate the existence of a stupa (Donaldson, 2001: 48). Thus it is evident that in the third century BC, the Buddhist religious landscape came to be defined through a series of monumental remains, such as pillars, edicts on rock and Buddhist stupas. Though there are references in the Pali Canon to donations of buildings and residences to the Buddhist Sangha, there is little archaeological evidence for these before the Mauryan period.


The issue that I address in this paper relates to the agency responsible for the spread of Buddhism across the Indian subcontinent and the setting up of monastic establishments. Was it the ruler and the state, as indicated by Asokan inscriptions or was it the Buddhist monks and nuns, as evident from records of donations and gifts to religious sites (Ray 1986: 104-6)? Another common assumption is that by the end of the reign of Asoka in the second century BC, Buddhist monks and nuns were established in monasteries throughout the Indian subcontinent and that these monasteries, located near cities relied on state support (Lopez, 1995: 4). How valid are these statements? The Buddhist texts, at the same time, credit merchants and trading groups with the setting up of religious shrines, as for example in the case of merchant Anathapindaka’s purchase of the park known as the Jetavana in order to present it to the Buddha and his monks for their use as a monastic establishment. This is a theme that finds representation at several early Buddhist sites such as Bharhut, Bogh Gaya, Sannathi,

Amaravati, etc. In the context of eastern India, the Vinaya (Mah•vagga, First Khanndaka, 4) records that two early disciples of the Buddha, the merchants Tapussa and Bhallika travelled from Ukkala, often identified with Orissa to where the Buddha was meditating under the Bodhi tree and presented him with rice cakes and honey. They are also credited with establishing caityas and monastic complexes on their return. In this paper I address the issue of patronage to Buddhist monasteries through the three strands mentioned above, viz. polity, religious functionaries and trading groups, my emphasis being on the last two. One of the related concerns is the question of identity of the lay Buddhist, i.e. was Buddhism a religion of the community at large or was it restricted to those who renounced the world, i.e. the monks and nuns . It has been suggested that the Buddha prescribed a path to salvation and was not interested in

expounding a ritual or communal religion. While this may by and large be valid, there is no denying the fact that the Buddhist canon makes a concerted attempt to inculcate a sense of moral and ethical values among the laity based on Buddhist ethics and loyalty to the Triratna, viz. Buddha, Dhamma and Sangha. In his Minor Rock Edicts, Asoka refers to himself as a Buddhist upasaka and expresses his desire that Buddhism percolate down to the lowest level to include elephant-trainers, charioteers, teachers and scribes (Sircar 1979: 12). It would seem that the southern versions all belong to a single recension, which was issued for engraving extensively on the hills at one specific time, as indicated by the reference to 256 nights . The dhamma of these records is largely ethical, with the injunction to: “Obey mother and father, obey the teachers; have mercy on living beings; speak the truth and propagate the dhamma”. Schopen (1997: 116) suggests close correspondence between Asokan inscriptions and the Mahæparinibbæna sutta in which the Buddha states: After I have passed away, monks, those making the pilgrimage to the shrines, honouring the shrines, will come [to these places], they will speak in this way: Here the Blessed One attained the highest most excellent awakening, etc.


This paper addresses the contribution of the lay community in its continuing support of monks and nuns and sponsorship of ritual activities, pilgrimage, as well as maintenance of monastic structures. The region of eastern India is here taken to encompass the coastal areas of the present states of Orissa and Andhra Pradesh.


As in Orissa, in the coastal region of Andhra also, the presence of Buddhism clearly coincides with the Mauryan period, indicators of contact including Asokan inscriptions in the Kurnool district and a fragmentary pillar inscription and stupa at Amaravati dated to the fourth-third centuries BC. Archaeological excavations conducted at the twin site of Dharanikota have provided evidence for five periods of occupation from fourth-third centuries BC to the eleventh century AD . It is significant that the sites of Dharanikota and Amaravati are located at the point until which the river Krishna is navigable and may be defined as landing places for the coastal traffic (fig. 3). The river Krishna takes a sharp turn at this spot and this link of Amaravati with the river is preserved in a pillar fragment discovered during clearance of the site in 1958-59 with an Asokan inscription on it (Epigraphia Indica XXXV, 1962: 40-3). Engraved on one of the faces is the legend: “the gosthi called Vanda at Dhanyakataka” together with the representation of waters (Ghosh & Sarkar 1964 & 1965: 175)4. The term gosthi has been translated as committee or association and occurs in the inscriptions of Sanchi in central India, Bhattiprolu on the Andhra coast and Mathura in north India.


The stele from Amaravati is significant for this discussion as it graphically depicts the creation and legitimisation of sacred spots in an area as far removed from the Buddha’s peregrinations as the south-east coast and the linking of these to sacred spots in north India. The three extant faces of the stele depict scenes relating to the last three months of the life of the Buddha from his stay at Vaisali (identified with modern Basarh in Bihar) to his passing away at Kusinara (or Kasia, 52 kilomteres from Gorakhpur in north India) – the scenes further substantiated by explanatory labels. The label on the first face mentions


4. A subsequent publication suggests a somewhat different reading as “Dhanakada Vandana Mago Ca” translated as “and the pathway to worship Dhanakada” (Roy, 1994: 193).

Fig. 3. Map showing monuments at Amaravati.


the Bahuputa-caitya and other caityas at Vaisali, while on the second face the legend Savathi or Sravasti (134 kilometres from Lucknow, the capital of the present state of Uttar Pradesh) occurs on one of the three stupa-like structures. Another scene represented is the purchase of Jetavana in Sravasti by the merchant Anathapindika for the residence of the Buddha and the monks. It is significant that these depictions match in detail accounts of the life of the Buddha as narrated in the Digha Nikaya (Mahaparinibbana Sutta, 5.8; Ghosh and Sarkar 1964-5). This stele dated to the Mauryan period thus provided justification for the association of the site of Amaravati with sacred spots in the north, such as Vaisali and Sravasti and perhaps also explains the prominent part played by monks and nuns in sponsoring religious architecture at the site. An early pillar inscription from Amaravati refers to a donation by a bhikhu or monk from Pataliputra.  


Fig. 4. Map of Andhra showing location of Buddhist sites.


Expansion of Buddhist monastic sites and early trading networks


Starting from modest beginnings dated to the Mauryan period, nearly 140 sites are listed in the Andhra region dating from the second century BC onwards and the distribution along the coast from Srikakulam in the north to Ramatirtham and Nandalur in the south is quite striking (fig. 4; Hanumantha Rao, et.al. 1998: figure 1). Of these, thirty Buddhist sites have yielded stone inscriptions recording donations made by a cross-section of the populace to the Sangha. Thus the pace of culture change from the third century BC onwards was brisk and to appreciate its impact, it is important to highlight the antecedents in the region, as also the reasons for mobility across peninsular India.


In the second-first millennium BC Andhra was home to iron-using megalithic communities and social integration was measured by the construction of large monuments of stones, often sepulchral in nature. Chronologically, the Iron Age megalithic sites span several centuries from 1200 BC to 300 AD and extend across all regions of peninsular India with the exception of the western Deccan encompassing parts of the present states of Maharashtra and Gujarat. There is evidence not only for trade and exchange, but also for an increase in conflict and aggression within the Iron Age communities of peninsular India in the first millennium BC, as compared to earlier levels. In addition, manpower required for the construction of megalithic monuments indicates mobilisation and organisation of labour by these societies. The variability in the scale of the megaliths in peninsular India ranging from simple urn burials at one end of the scale to large stone circles enclosing elaborate chambers at the other indicates several small, locally oriented systems of varying scale and geographic extent. Nevertheless within the larger Iron Age cultural sphere, regional specificities are clearly defined.


A survey of published literature reveals thirteen megalithic sites in the lower Krishna valley with the number increasing to 95 in the subsequent early historic period, with four sites showing continuity of settlement. One of these was Amaravati where seventeen urn burials were found beneath a small stupa (Annual Report-Archaeological Survey of India 1908-9: 88-99). Yelleswaram in Andhra is another site where an urn burial was excavated beneath a Buddhist stupa (Khan 1963: 8-9). Thus it is evident that at several sites in Andhra, there was continuity in the demarcation of sacred space between the Iron Age megalithic burials and the setting up of Buddhist stupas.


As compared to Andhra, a Megalithic substratum is not known from Orissa, though as mentioned earlier, Mauryan inscriptions and other structural remains attest to the presence of Buddhism in third-second centuries BC. A third century AD Prakrit inscription from Nagarjunakonda in Andhra refers to the fraternities of monks at Tosali, Palur and Puspagiri, identified as major Buddhist centres in Orissa (Epigraphia Indica XX: 23). Recent archaeological research has brought to light several Buddhist sites in Orissa, though the major expansion occurs from fifth to thirteenth centuries AD, when more than one hundred Buddhist sites are known in the region. Several sites have earlier beginnings, one of these being Langudi, on a low hill running from north to south, located in the plains of the Mahanadi Delta about ninety kilometres from Bhubaneswar in the Jajpur district. The river Kelua, a tributary of the Brahmani, the second largest river system in Orissa, meanders across the northeast and eastern parts of the Langudi Hill. Thirty-four rock-cut stupas on the northern spur of the hill Fig. 5. Rock-cut stupas at Langudi, Jajpur district. Fig. 6. Stupas at Langudi hill. Fig. 7. Seated Buddha at Langudi.


have been dated to second-third century AD (fig. 5 & 6), though most of the sculpture dated from seventh to ninth century AD (fig. 7). Archaeological excavations at Lalitagiri have provided evidence for a second-first century BC stupa, an apsidal shrine built over the remains of an earlier structure, inscriptions in KushanaBrahmi, Gupta and post-Gupta script, as also a Gupta coin. Relics in gold foil were found at the site (Bandyopadhyay, 2007: 9-10). A second century AD Brahmi inscription from the pavement of the apsidal shrine records the completion of a seat by several lay devotees and a nun. Also found from the site were dull red ware sherds with inscriptions dated to first century BC (Indian Archaeology – a Review 1987-88: 88).


An issue that requires consideration at this stage relates to the extent and vigour of the trading network linking eastern India with the Bay of Bengal and evidence for the spread of Buddhism across the Ocean. A cluster of fifth century inscriptions of unequivocal Buddhist affiliation was found in Kedah on the west coast of the Malay peninsula. This includes engraving of the Buddhist formula on stone – a feature that does not occur among contemporary records from the Indian subcontinent, though the formula is found on terracotta sealings. Three of these inscriptions are made of local stone and bear similar illustrations of Buddhist stupas. Texts very similar to these inscriptions have been found on the island of Borneo and on the coast of Brunei (Christie 1995: 256). The most interesting of these inscriptions in Sanskrit is that of Buddhagupta, which refers to the setting up of the stone by the mariner Buddhagupta, resident of Raktamrttika, identified with Rajbadidanga in Bengal, on the successful completion of his voyage (Chhabra 1965: 23-4). There was a shift in maritime networks around the middle of the first millennium AD and pilgrims visiting sites


associated with the life of the Buddha formed a major category of travellers. The Chinese pilgrim Faxian arrived overland in India in AD 399 and returned by sea to China in 413-414 AD from Sri Lanka heading towards the northwest tip of Sumatra. The ship was wrecked on the way and perhaps landed in the Andamans. The next phase took Faxian to the northwest of Borneo where he arrived in 414 after 90 days at sea. The pilgrim remained in Borneo for five months and then left for China in mid-414 heading towards Canton. The ninth century marked a turning point and several major coastal sites in peninsular Malaysia date to this period, such as Ko Kho Kao near Takuapa, Laem Pho near Chaiya and Kampong Sungai Mas in south Kedah known for large collections of Chinese as well as West Asia ceramics. Other characteristic features of these sites are the brick architectural remains that survive and have been unearthed in archaeological excavations as well as stone statuary, both Hindu and Buddhist (Jacq-Hergoualc’h, 2002: 134-160).


The most interesting information about circum-peninsular navigation is contained in Yijing’s accounts of the voyage of the Chinese pilgrims who travelled to India and returned during the second half of the seventh century AD. Of the sixty monks whose trip is briefly recounted at least thirty-seven travelled by sea on merchant ships and none seemed to have given any thought to crossing the Malay Peninsula, even if some of them stopped off there, notably in Langkasuka. Yijing also provides an account of his own journey from Canton in October-November with the northeast monsoon and reached Palembang a month later. He stayed there for six months, then went to Jambi near Palembang sometime around May. He stayed there for two months and then reembarked in order to profit from the winds of the southwest monsoon to reach Kedah (Jiecha). He did not leave this region for India until the beginning of the following year when the north-east monsoon was well established. He reached the Nicobar islands in ten days and fifteen days later arrived at Tamralipti in Bengal. This was clearly the most direct route to the holy places of historic Buddhism (Jacq-Hergoualc’h, 2002: 53). It is then evident that in addition to trade and trading networks, pilgrimage and religious travel provided impetus for mobility across South Asia as well as across the Bay of Bengal into Southeast Asia and China. It is also accepted that worship of the relics played a key role in the spread of Buddhism across Asia and relics have served to forge a distinctive Buddhist geography articulated through pilgrimage routes .


Division and worship of relics


wooden chariot and monks and lay devotees ‘from within the borders’ assembled to pay homage (Legge 1886/1965: 79). In addition to these annual festivals, monasteries also regularly displayed the relics possessed by them such as, for example, the alms-bowl of the Buddha. The importance of the relics that a monastery possessed determined the following that it commanded (Gunawardana 1979: 227). The Buddhist site of Bhattiprolu, a village on the Guntur – Repalle railway line in Krishna district of Andhra Pradesh is significant on several counts. It provides details of second century BC practice of relic worship in coastal Andhra, but more importantly patronage for the setting up of the stupa comes from the local chief or raja and members


After the Buddha’s death his ashes were divided into eight parts and distributed to the eight kings in whose kingdoms he had lived and stupas constructed over these. Asoka subsequently opened seven of these and redistributed their contents over a much larger expanse of the subcontinent, thereby adding to the creation of the sacred geography. This is substantiated by archaeological finds of relic caskets with a host of small objects (Willis 2000: 12-7). A graphic representation of the division of the relics is provided by the three architraves on the gateways to the main stupa at Sanchi. Recent scholarship asserts that the Mahaparinibbana sutta (Dighanikaya 16, 6: 23-8) indicates that the “authoritative Theravada tradition both affirmed the value of relic veneration and at the same time cautioned that it should not be the primary preoccupation of members of the Sangha” (Germano and Trainor, 2004: 10). In the beginning, the relics involved were corporeal relics of the Buddha himself. Gradually the scope was enlarged and by the beginning of the Christian Era not only the bone relics of the monks were included but also precious objects such as gold, silver, pearls, crystal, lapis lazuli and so on. A majority of the relic caskets from peninsular India have been found in the Andhra region (Ray 2003: 372-5). Descriptions provided by the Chinese monk Faxian who travelled through


India in the fifth century AD indicate that festivals were regularly organised for the worship of these relics and cult objects. For example, at Pataliputra he refers to a procession of images celebrated every year on the eighth day of the second month. The images were placed in a the qualities that animated and defined the living Buddha” (Schopen 1997: 160).

of the gosthi and nigama. Of the three mounds, the largest is 12 metres high and has been known from 1870 onwards. Archaeological work at the first mound in 1871 yielded a crystal reliquary with ash, gold leaf and pearls, etc, while the maha-stupa at the site yielded caskets of crystal, gold, copper, stone and pottery placed within three inscribed granite containers with lids. Beneath one of the granite containers, a group of twenty-four silver punch- narked coins were discovered arranged in the form of a svastika. It is suggested that the Bhattiprolu script represents a provincial offshoot of the early Brahmi script in the south and dates to the second century BC (Salomon, 1998: 35). A significant reference in the inscription is to members of the gosthi and the nigama with raja Khubiraka as their chief .


The Pali dictionary derives the meaning of the term nigama from the Sanskrit root gama with the prefix ni. The compound term thus has the sense of coming together or meeting. On the basis of early Buddhist texts, Wagle defines the nigama as a gama composed of more or less integrated members of various kin groups and occupational or professional groups. It is therefore a larger and more complex economic and social unit than the village or gama (Wagle 1995: 21). Another related, though distinct, term is the negama or assembly of persons connected with the nigama . The terms nigama and negama are found inscribed on unbaked clay sealings from several sites in north India. The earliest of these date from the  


Fig. 13. Step-well at Udayagiri with 10th-11th century inscription.


Mauryan period (Thaplyal 1972: 223) . Beyond the northern plains, there are references to the nigama in inscriptions from early Buddhist sites. The nigama of Karahakata in the Deccan is mentioned in the second century BC inscriptions of the Buddhist site of Bharhut in central India (Cunningham 1879: 131, no. 16). Similarly, there are references to the nigama of Dhanyakataka along the east coast in Andhra (Epigraphia Indica XV: 263). Both the gosthi and the nigama existed at Dhanyakataka.


The nigama continues to be associated with urban centres in the post-Mauryan period. Several seals found at Bhita near Allahabad bear the legend ‘nigamasa’ in Kushana Brahmi, perhaps representing the authority of some autonomous urban administration (Archaeological Survey of India — Annual Report 1911-12: 56). The term nigama also occurs in the region around Madurai in Tamilnadu where fifty-five inscriptions in Tamil-Brahmi Fig. 14. Apsidal shrine at Udayagiri.


have been dated between the second century BC and the second century AD. A record from Nasik in Maharashtra of the early centuries AD provides some clues regarding the use of the term. It mentions that donations of land were first proclaimed in a nigama-sabha; these were then written down on cloth or else on a copper-sheet by an officer and delivered to the done, which was in most cases the Buddhist Sangha (Senart 1902-3: 82-5).


It is then evident that in the secondfirst centuries BC it was members of the nigama and gosthi who provided patronage to the early Buddhist coastal sites of Andhra. Unfortunately archaeological details of the beginnings and development of monastic centres in peninsular India are not often available because the unsystematic clearance of many of the Buddhist sites in the nineteenth century by British military officials. The history of the stupa at Amaravati, a long-lasting monastic centre (termed Dhanyakataka in early epigraphs) in the lower Krishna


valley, is a case in point (Singh, 2004). While collections of sculptures from the site are in several museums in India and Britain, it is only recently that attempts have been made to unravel the architectural history of the stupa.


An important factor that needs to be taken cognisance of is the traditional practice of restoration of religious sites in Asia by rebuilding, repairs and renovation. Archaeological evidence from Amaravati indicates that the stupa continued to be rebuilt and added to during the period of its existence from third-second century BC to fourteenth century AD, though it is the early phase that has often formed the focus of scholarly attention. Many of the early sculptural pieces were crushed into small fragments and reused for the building of the later stupa. Several slabs were re-carved with newer sculptures on the back and reused (Shimada, 2006: 98). For example, a palimpsest drum slab (fig. 9 & 10; British Museum OA 1880. 7-9. 79) was carved in first century BC with a scene indicating the Enlightenment of the Buddha represented by an empty throne with his footprints at the base. In the third century AD it was re-carved on the other side with an elaborate and tightly decorated stupa showing


Fig. 15. Stupa at Lalitagiri. the Buddha standing in the gateway of the monument worshipped by snake kings and human attendants (Knox 1992: 119, 139). Similarly the railing around the stupa was built in at least three phases starting from circa AD 50 to 1 BC; 50 to 100 AD and the last phase dated to circa 200 to 250 AD (Shimada, 2006: 134). Of the two hundred inscriptions dated to the early period recovered from the site, 108 provide relevant data on issues of patronage to the site and fortytwo of these record donations by women. Often they identify themselves through the occupation of their husband, such as Somadatta, the wife of Bala, the royal scribe (Epigraphia Indica XV: 264) or Uta, the mother of Dhanamala (ibid.). Monks and nuns or bhikkhunis/ pavachitaya (twenty-four) form the single largest category of donors, followed by gahapatis and their wives (twenty-two) with trading groups and their relatives accounting for fifteen donations. There are references to female lay worshippers (uvasika, Epigraphia Indica XV: 267) and to navakammakas or overseers of works (Epigraphia Indica XV: 274). Neither money donations nor gifts of cisterns and tanks figure in the list; instead the emphasis is on gifts of pillars, footprints, lion seats, coping stones and sculpted slabs. In contrast to monastic sites in Andhra, in Orissa, railing inscriptions recording donations are scarce, though there is architectural evidence for early Buddhist presence in the region.


Tantric isolation or continuation of Mahayana Tradition?


Alexander Cunningham stressed that Buddhism had become decadent by the middle of the seventh century AD. Already by this time, he claimed, the monks had become an indolent and corrupt body content to spend their lives in the monotonous routine of monastic life (1966, reprint: 2). In a number of instances, the decay of Buddhism was seen as the direct result of the growth of idolatry. This trend continues to be repeated in historical writings. The degenerate form of Buddhism, it is argued, was reduced to psychosexual practices, with little or no resemblance to the ‘nibbanicBuddhism of Buddha’s time or the devotional boddhisattvahood of the Mahayana tradition. Vajrayana Buddhism survived in a few pockets of eastern India where the abstruse rituals were limited to monastic institutions without a popular base or distinct identity (Chakrabarti, 2001: 142). This view does not take into account the vibrant and dynamic artistic regional traditions that developed from eighth to thirteenth centuries at a time when manifesting divine presence played a central role in many Indian religious practices (Leoshko, 2003: 18). Particularly important is the worship of the stupa from the eighth-ninth centuries onwards, which helps explain the increase in the number of stupas dedicated throughout the Buddhist world. The kriy• and cary• tantras emphasised the maintenance and worship of stupas and identical funerary and meritorious practices were prevalent in Mahayana, Mantrayana and Vajrayana texts (Hock 1987: 109).


There was a corresponding increase in the resource base as evident from grants of land and villages to monastic establishments in Anhra. Thus the Patagandigudem copper plates of the Iksvaku ruler Siri Ehavala Cantamula record the grant of 32 nivartanas of ploughable land (halaksetra) inhabited by the kin of the chief monk and 64 nivartanas of land west of Pithunda to the monks (bhikkhubhoga) and these examples may be multiplied (Hanumantha Rao et. al., 1998: 191219). The nature of donations now included lamps, bullock carts, servants (dasi-dasa), brass cauldrons, tanks and money (e.g. Alluru grant, Epigraphia Indica 39: 9-10). This was in addition to gifts of items of ritual use such as dipadhupa-gandha-puspa-dhvaja-panabhojana-sayana-asana-glana-bhaisajya (lamps, incense, sandalwood, flowers, flags, drinks, food, beds, seats and medicines for the sick, Tummalagudem copper plates of the Visnukundins, Hanumantha Rao et. al., 1998: 197202). The Caitanyapuri inscription engraved on a boulder records gifts of Fig. 16. Apsidal caitya at Lalitagiri.


residential cells for use of persons incharge of incense and clothes attached to the Govindaraja vihara (Hanumantha Rao et. al., 1998: 196-7). Thus the ritual requirements of the monastic complexes not only required a wider range of donations, but also manpower to monitor and maintain these and though trading communities are no longer mentioned, yet the continuation of trade to supply the ritual needs of the religious establishments is beyond doubt. The fourth to fourteenth century period of Amaravati’s existence has perhaps been the most neglected, though there is no doubt that the site continued in existence. Fourth to sixth century bronze images of the standing Buddha from the site are now in the Government Museum in Chennai, Tamilnadu (Sarkar and Nainar, 1992: 17) and eighth-ninth century limestone sculptures of Boddhisattva Avalokitesvara, Manjusri, Vajrapani and Cunda are in the British Museum (Knox, 1992: 215-220). In


Fig. 17. Monolithic votive stupa with Buddha image at Lalitagiri.


the eighteenth century, Colin Mackenzie had recorded two major inscriptions from Amaravati, but one of these is now lost, while the other twenty-line seventheighth century epigraph in Sanskrit dealing with matters of Buddhist principles is now in the collection of the British Museum (Knox, 1992: 223). There is little information on the physical characteristics of the shrines that housed these images or the settlements around the site. Walter Elliot’s drawings and sketches made during his work at Amaravati in 1845 would suggest the presence of a shrine complex, though its date seems uncertain. The Gadaladeniya rock inscription of AD 1344 from Sri Lanka records the restoration of a two storey image house at Dhanyakataka in Andhra by a sthavira named Dharmakirti (Epigraphia Zeylanica 4: 90). The pillar inscription of AD 1182 during the rule of Keta II refers to the presence of a Saiva temple of Amaresvara at the site and adjacent to it a very lofty caitya of lord Buddha (Epigraphia Indica, 6, 1900-1901: 146-60). Another epigraph dated to AD 1234 mentions a grant to the Buddha ‘who is pleased to reside at Sri Dhanyakataka’.

In addition to the lower Krishna valley, Buddhism in its Vajrayana form continued to flourish at the sites of Sankaram, near Anekkapalle, 50 kms. from Visakhapatnam on the highway to Vijayawada and Ramatirtham, 15 kms north-east of Visakhapatnam. A ninth century inscription from the site records gifts to the Buddhist monastic establishment. Bojjanakonda, the eastern of the two hills at Sankaram is dotted with a series of rock-cut and occasionally brick built stupas at different terraces. These undoubtedly represent offerings of pious pilgrims to the site in the ancient period that came to worship at the apsidal shrine-cum-monastery to the east (Mitra, 1980: 219). Further north along the Andhra coast, the site of Salihundam is situated on a hill about eight kms. from the ancient port of Kalingapatnam located on the sea coast. Images of Marici, Bhrkuti, Tara and Manjusri have been found at the site indicating continuity of the Buddhist monastic establishment into the early medieval period and influence from the art and iconography of ancient Orissa (Mitra, 1980: 222).


It was in Orissa that the major expansion of Buddhism took place in the fifth to thirteenth century period and stupendous monastic complexes were constructed in the three hill ranges of Jajpur district, the Alti or Nalti, Assia and Mahavinayaka. Local tradition refers to the region being close to the sea in the past. Ratnagiri is a 25 metre high isolated hill of khondalite formation of the Assia range bounded on three sides by the rivers Brahmani, the Kimiria and the Birupa. In the vicinity of Ratnagiri, the extensive Buddhist site of Udayagiri is located in the easternmost part of the Assia hills in a horse-shoe formation, while Naltigiri or Lalitagiri is not very far on the south bank of the Birupa. The three monastic complexes form a triangle, with the distance between Ratnagiri and Udayagiri is 5.5 kilometres as the crow flies and 3.5 kilometres between Udayagiri and Lalitagiri. The latter two hills are much broader and higher than Ratnagiri (Mitra 1981: 7). Excavations at Ratnagiri unearthed an impressive stupa surrounded by large numbers of stupas of all dimensions (fig. 11), two quadrangular monasteries and remains of eight shrines with small temples around some of them. Varied deposits were found inside the structural stupas and included partly charred bones, as well as reliquaries in the form of earthen vases, at least one of which contained a coin identified as a Ganga fanam dated to thirteenth century AD (Mitra, 1981: 28, 53). Also associated with the main stupa were a large number of monolithic stupas, many of them with the Buddha or Buddhist deities carved in relief. The monastery at Ratnagiri comprising of a stone-paved central courtyard surrounded by a verandah and twentyfour cells with a shrine in the centre of one of the sides, is the finest in the Indian subcontinent (fig. 12; Mitra, 1981: 152). It had an upper storey and indicates two phases of repair and construction. Cell 17 in the monastery housed a large number of bronze images of the Buddha and other deities, as well as a metal seal and sealings reading ‘Sri-Ratnagiri-mah•vih•riy•rya-bhiksu-sanghasya’. Two stone stele fixed to the eastern wall of the monastery record donations made by a kayastha named Janananda from the Buddhist site of Nalanda in Bihar. These included the grant of two pravara measures of paddy from each hala measure of land to the monks residing in the complex at Ratnagiri (Mitra, 1981: 213-4).


Though Buddhist remains from Udayagiri were first recorded in 1870, it was only in 1985 that large-scale excavations were undertaken at this extensive site, which in many ways is different from that of Ratnagiri. At the base of the hill is a rock-cut well with thirtyone steps leading down to the water. A tenth-eleventh century inscription in


Fig. 18. Image of Gajalaksmi on lintel of doorway of the monastery at Ratnagiri. the rock records that it was dedicated by R•naka Vajran•ga (fig. 13; Chanda, 1930: 10). Excavations uncovered a stupa, an imposing monastery covering 30 metres square and a residential-cumshrine complex, with a central shrine chamber housing the image of a seated Buddha in bhºmispar‹amudr• as also a massive water reservoir. Seals found from the excavations refer to the site as Madhavapura mahavihara. The southern section of the valley contained an apsidal shrine (fig. 14), monolithic stupas, two-storeyed monastery and several other structures (Bandopadhyay, 2007). Near the top of the western spur of the hill overlooking the river Birupa are a series of rock-cut images of the Buddha and Boddhisattvas with a votive stupa in front. The first image is that of Avalokitesvara and bears two inscriptions. One is a popular Buddhist dhæran∞ and the other records that the image was a gift of Simyaka. Peculiar to Udayagiri is the alignment of the eight Boddhisattvas forming a mandala around a central figure (Donaldson, 2001: 64).


As at Udayagiri, at Lalitagiri also a gallery of rock-cut Boddhisattvas is in evidence. Chanda recorded this when he visited the site in 1930 (Chanda, 1930: 9). In 1985, the Archaeological Survey of India excavated the mound at the top of Landa hill and unearthed a stupa, which yielded three small relic caskets (fig. 15). In subsequent excavations an apsidal

Fig. 19. Seated Buddha on lintel of monastery at Udayagiri. caitya was discovered. Forming an outer ring around the caitya were small monolithic stupas, some with inscriptions of fourth-fifth century alternating with Buddha images (fig. 16 & 17). Four monasteries were unearthed, with monastery 4 providing evidence of a shrine chamber with the image of the seated Buddha. A seal found in monastery 4 had the Sarnath sign with a cakra flanked by deer on top and an inscription reading ‹r∞ candr•ditya vih•ra samagra •rya bhiksu sangha indicating that the name of the establishment was ‹r∞ candr•ditya vih•ra. Several pillars found from monastery 3 are similar to those from Brahmanical temples in Bhubaneshwar. Though the caitya at Lalitagiri dates to the second-third century, only a few images have been found dating from the

early period. Unlike Udayagiri, however, four-armed images are the exception at Lalitagiri (Donaldson 2001: 53-6).

How does the presence of these remarkable monastic complexes in the fertile coastal tracts of Orissa, often in close proximity to Hindu temples, impact the dominant historiography regarding the degenerate and isolationist nature of Buddhism? Over the last few decades, there have been attempts at identifying specific texts that may have been used at Buddhist monastic sites, both in India and elsewhere. For example, A. Ghosh published the text of an inscription found on a stone slab now in the museum at Cuttack and also presented similar texts found on terracotta tablets from Nalanda written in Nagari characters dating from sixth to ninth century AD, as well as those from Paharpur and Bodh Gaya. Though Ghosh was unable to identify the text, it is now evident that the

stone slab represented a short Sanskrit version of texts preserved in the Tibetan Kanjurs. Texts such as the Bodhigarbh•lank•ralaksa and Vimalosn∞sa form a group of texts that were widely known and used for inscribing dh•ran∞s on terracotta plaques not only at sites in Orissa, but across the Buddhist world in Tibet, China as well as Sri Lanka. This led Schopen to conclude: If by ‘tantric’ we mean that phase of Buddhist doctrinal development, which is characterized by an emphasis on the central function of the guru as religious preceptor; by sets – usually graded – of specific initiations; by esotericism of doctrine, language and organization; and by a strong emphasis on the realization of the goal through highly structured ritual and meditative techniques, then there is nothing at all ‘tantric’ about these texts. They are texts dealing with ritual forms open to all and religious problems common to all – monks, nuns, lay men and women. Second, they show a 

marked continuity in terms of religious concerns with the literature that preceded them. They are, like much of the canonical Mahayana sutra literature that came before them, preoccupied with death and the problems of rebirth (Schopen, 2005: 337).


This then raises the question of patronage and the support base of the Buddhist monastic complexes at a time when the ritual requirements of the monastic complexes had changed. The setting up of more than 700 monolithic stupas at the site of Ratnagiri itself indicates the nature of interaction between the community and the monastic complex. Linked to this is the change in the nature of grants that are now recorded. These largely record donations of land in peripheral areas and for the expansion of agriculture in coastal regions, though in some cases the beneficiaries are also religious institutions. Of a total of 340 inscriptions from Orissa dated from sixth to thirteenth centuries AD, there are references to grants made to Buddhist centres in at least fourteen cases. A copper plate charter from Jayrampur in district Balasore dated to the first half of the sixth century AD records the grant of a village Svetabalika to the

bhikshusangha for constructing a vihara at Bodhipadraka (identified with present Jayrampur) for •rya Avalokite‹vara. The grant provided for ritual requirements of worship such as bali, caru, gandha, puspa, prad∞pa, etc and for providing for necessities of the •rya Sangha like food, bed, food, medicine, etc (Sircar, 1965: 530-1). This brief overview indicates that in contrast to the conventional linear development from Buddhist shrine to Hindu temple, recent research establishes that both the Buddhist shrine and the Hindu temple were contemporaneous in third-first centuries BC and shared sacred space with a diverse array of domestic, local and regional cults. It is also significant that rather than being subsumed through a process of acculturation within the dominant Brahmanical tradition as generally proposed, many of these local and regional cults continued to maintain autonomous religious traditions (Ray 2004: 350-375). This also raises issues of inter-religious negotiation and interaction.


Religious change and diverse religious landscapes


A region, which provides crucial archaeological data on the religious landscape from the earliest Neolithic settlement in the third millennium BC to the sixteenth century, is the secluded Nagarjunakonda valley shut in on three sides by offshoots of the Nallamalai Hill Range. More than thirty Buddhist establishments, nineteen Hindu temples and a few medieval Jaina shrines were unearthed in several seasons of archaeological excavations conducted at the site after its discovery in 1920 until its submergence in 1960. Early Buddhist religious architecture occurs throughout the valley, while Hindu temples were located mainly along the banks of the river Krishna and around the citadel area. It is significant that religious architecture of both the Hindus and the Buddhists was built on diverse ground plans and both used the apsidal form.


At Nagarjunakonda, the apsidal shrine with a stupa in its apse is either associated with the Mahacaitya (sites 1 and 43) or forms a part of a residential enclosure for Buddhist monks (site 38) and in the latter case no stupa was found in the apse leading to the suggestion that it may have enshrined either a Buddha image or Buddhapada found in its vicinity (Sarkar 1993: 77-78). In another case (site 9) the apsidal shrine located near the citadel formed a part of the monastery and was meant for the image of the Buddha (Sarkar and Misra 1999: 34). Site 85 was also situated close to the southern rampart wall and the monastic complex contained two caityagrhas - one oblong and the other apsidal – both enshrining images of the Buddha (Sarkar and Misra 1999: 36). Sarkar and Misra suggest that apsidal shrines (sites 1 and 43) located in the vicinity of the main stupa or mahacaitya were later additions to the Buddhist establishment at Nagarjunakonda


(1999: 33). Inscriptions on the floor of the shrines record the history of their foundations. The lay devotee Bodhisiri built the apsidal caityagrha for the welfare of her husband, his family and her maternal family at the vihara of Culadhammagiri for the benefit of monks from several regions such as Sri Lanka, China, Kashmir, Gandhara and so on (Vogel, 1929-30: 22, no. F). The other apsidal shrine was also built for the benefit of the monks from the different countries, but the donor in this case was Chamtisiri, the sister of the Iksvaku ruler Vasithiputa Siri Chamtamula (Vogel, 1929-30: 21, no.E). Another apsidal shrine was found in the northeastern part of the valley (site 24) and in addition to its proximity to the monastery shows two new features: one, a circular image shrine and the other, a memorial pillar raised in honour of the king’s mother (Sarkar and Misra, 1999: 37). The sculpture on the pillar depicts a lady seated on a high stool with a lady attendant standing nearby. Thus variations in apsidal shrines connected to Buddhist monastic complexes are evident and no single pattern prevailed. Nagarjunakonda is unique in that it provides a large corpus of inscriptions associated both with Hindu as well as Buddhist architecture that allows insights into several aspects of interaction between the community and religious architecture. An analysis of these records of donations indicates that in a majority of the Buddhist inscriptions the motive as stated by the donor relates to attainment of nirvana by the donor and the welfare and happiness of the world.


The data from Nagarjunakonda clearly establishes the intersection of diverse religious traditions, some associated with communities settled earlier in the region and others drawn in by movements of trading and craft groups, Buddhist monks as well as internal dynamics associated with the emergence of the site as a centre of political power. Nagarjunakonda as is evident from the archaeological record participated in both coastal networks as well as overland trading activity. It is also unusual in the varied secular architecture preserved at the site, including a citadel, an elaborate water system, residential complex and what has been termed an amphitheatre.


The pattern of a poly-religious landscape is by no means unique to Andhra and was replicated in Orissa as well, especially from the fifth century AD onwards, which not only saw the expansion of Buddhist monastic sites, but also the emergence of the Kalinga style of Hindu temple. The district of Jajpur and the region around Bhubaneshwar in coastal Orissa were marked by the presence of temples dedicated to Siva, while Vaisnavism and Saktism also spread across the region. There are significant reciprocal influences between the Hindu and Buddhist art and architecture of Orissa, such as in door frame fragments from Ratnagiri, Udayagiri and Lalitagiri. While the image on the lintel at Ratnagiri is that of Gajalaksmi (fig. 18), at Udayagiri it is a seated Buddha flanked by

Avalokite‹vara and Vajrap•ni (fig. 19; Donaldson 2001: 376-7). One of the issues that the survey of religious architecture presented here raises is the extent to which Buddhist monasteries were a reflection of the practice of the religion within a regional cultural milieu. Thus the architectural and sculptural features of Buddhist monastic sites in Orissa were closer to those of Hindu religious architecture. Was this only a case of shared artistic traditions between the two religious communities in eastern India? Did this similarity extend to all monastic sites or are inter-site differences significant? We have referred to several variations between the monastic centres of early Andhra and Orissa, but do these divergences also relate to centres within a particular region? A comparison of the plans of the monastic complexes at Ratnagiri, Udayagiri and Lalitagiri provides evidence of inter-site variations.


Regional Traditions and Memorial Landscapes


A striking feature of Buddhism in Orissa is the lack of narrative sculptures, one of the few exceptions being the astamah•pr•tih•rya sculpture, now in the Raghunatha temple at Solampur (Donaldson 2001: 94). Tara was a popular deity in Orissa, especially at Ratnagiri, where she is found sculpted on 99 niches of monolithic stupas. She is represented both in her seated lalit•sana form as also in various other types described in the S•dhanam•l•. From seventh to twelfth centuries she is shown as the saviour from a variety of dangers including shipwreck in sculpture as also in epigraphs (fig. 20). An early representation occurs in cave 9 of Ellora in western India and two standing images are known from Ratnagiri. She is invoked in several inscriptions, such as the Nalanda record of Vipula Srimitra dated to the first half of the twelfth century, as also the Kalasan inscription from Java. It would then seem that though there were several similarities between contemporary sites, yet every site placed emphasis on certain images suggesting local preferences for cults and specific texts. The eleventh century manuscript of Astas•hasrik• Prajñ•p•ramit• in the Cambridge University Library (MSS no. Add 1643) illustrates Buddhist images and shrines from different parts of India and four of the illustrations relate to Orissa. It is not surprising that three of these refer to images of Tara, perhaps from Banpur, while the fourth refers to the monastic complex at Kuruma, near Konarak in Puri district (Donaldson, 2001: 25).


In addition to the regional variations in the sanctity of monastic sites, there is another important feature of the landscape, viz. the location of monuments to the monastic dead. As early as 1854, Cunningham published the results of his cursory excavations at Sanchi where it was clear that an extensive cemetery was associated with the Buddhist monastic site at Bhojpur and both here and at Sonari and Andher there was evidence for the elaborate housing and worship of the monastic dead. Similarly at Bodh Gaya hundreds and thousands of small stupas of various sizes were crowded in a jumbled mass around the central point of the site.


More than seven hundred portable monolithic stupas were exposed in the area around the monastery at Ratnagiri and by their sheer numbers these indicate that the site competed with Bodh Gaya (Mitra, 1981: 31). Five hundred and thirty-five of the portable stupas were found on the south-western side of the main stupa area and date from the ninth to thirteenth century AD. Of these 269 stupas have a single niche with the image of a deity enshrined within it, while one has two niches. The most popular deity was Tara followed by the Buddha, Avalokite‹vara, Mañju‹r∞, M•ric∞ and Vajrasattva. “These stupas were evidently raised by devotees who intended to earn merit by installing them near the main stupa” (Mitra, 1981: 27) and the deities enthroned in the niches of stupas may indicate personal devotional images chosen for the monk or layman whose remains were deposited in the reliquary (1987: 112). Hock argues for a change in style from the seventh-eighth century undecorated monolithic stupas found at Lalitagiri to the more ornate ones at Ratnagiri with a Buddhist figure in the niche. Though Cunda as a goddess does not find frequent representation in stone sculptures at Ratnagiri, she is depicted repeatedly on the monolithic stupas and seems to have been popular not only in Orissa, but also at other sites such as Ellora in western India and Candi Mendut in Java (Hock, 1987: 116).


Some of the portable monolithic stupas found at Ratnagiri were meant for enshrining bone relics, as evident from the provision of sockets at the base. This was also the case with a large number of the minor structural stupas, which yielded partially charred bones and reliquaries. The reliquaries were generally plain earthen vases or stone blocks with sockets. In some cases (for example stupas 23, 24, 25, 62 and 115) gold beads and gold coins were also found. A second category of deposits consisted of inscribed stone slabs, terracotta plaques and tablets. The inscriptions consisted of the Prat∞tyasamutpæda s tra along with the nirodha section, Buddhist creed and dhæran∞s (Mitra, 1987: 28-30). The dhæran∞s had funerary associations, as is evident not only from the texts themselves, but also by te fact that they were found inside small monolithic stupas (Schopen, 1997: 122). It is also evident from archaeological excavations at sites such as that of Sannathi located on the eastern bank of the river Bhima 60 kilometres south of Gulbarga that the practice of erecting memorials for the deceased re-appeared in a somewhat altered form in the early centuries AD. A chance discovery during renovation work in the Chandralamba temple at the site of Asokan Rock Edicts XII and XIV and Pillar Edicts I and II


Fig. 20. Standing stone image of astamahabhaya Tara from Ratnagiri.

clearly shifted its beginnings to the Mauryan period. Distinctive finds from the site include memorial slabs with one of the complete specimens recovered measuring 3 metres in height. These slabs are divided into panels; the top panel is generally arched and decorated with a series of tiered roofs and windows. The second panel carries the portrait of an individual or couple, while the third commonly shows either an unyoked bullock cart or a horse without a rider (Howell 1995: 69).


Thotlakonda is a Buddhist monastery located 16 km north of the modern city of Visakhapatnam and along with Bavikonda and Pavuralakonda forms a cluster of Buddhist monastic sites in north coastal Andhra Pradesh. Archaeological investigation based on a systematic survey revealed roughly 200 stone cairns located on the hill sides with a clear view of the monastery. These cairns were constructed of natural boulders, easily available in the fields dotting the hilltops and were contemporary to the several viharas, chaityas, stupas and other Buddhist remains at Thotlakonda, which the excavators have dated between the second and third centuries B.C. through the second or third centuries A.D. Most of these cairns range between one and three metres in diameter, but a few larger ones are more than 10 metres in diameter (Fogelin, 2004: 376-391).


Based on the Mºlasarvæstivæda Vinaya, known from Buddhaghosa’s fifth century commentary and later translations in Chinese and Tibetan Schopen asserts that the performance of death rituals for fellow monks, negotiating ritual privileges, control of sacred relics and economic resources characterised institutionalised Buddhism of the first to fifth century AD. Monastic funerary rituals involved removal of the body, honour of the body, including bathing before cremation and recitation of sacred texts. Another aspect of these rituals was lay participation and claim to monastic relics. At this point the text categorically states that lay persons can build stupas for the relics of the monastic dead, but all such stupas should be built within the monastic complex and hence remain under monastic control (Schopen 2004: 98-106). A comparison of the three major Buddhist sites in Orissa shows that while at Ratnagiri, monolithic votive stupas were found crowded around the stupa, at Udayagiri and Lalitagiri, the clustering occurs around apsidal caityas. In addition, the evidence from Sannathi would suggest that memorial slabs were set up within the precincts of the monastery for important lay disciples.


Several conclusions are valid at this juncture. First, that the monastic sites underwent irreversible change during the process of discovery in the colonial period and this process is crucial to an understanding of the archaeological context of the site. Second, the dynamism of the religious process associated with the setting up of the monastic centre can only be comprehended through an analysis of their historical trajectories firmly rooted as these were in regional and local cultural traditions. Finally, the stupa formed the nucleus of interaction between the monastics and the lay community and the diversity of this interaction is evident from the range of burials, cairns and memorial slabs associated with the stupa landscape.


In the final analysis, how does one place this paper within the larger historiographic context of studies on Buddhism, including on art and architecture? The objective of this paper was to unravel the archaeological context of the monument in an attempt to understand its distinctive personality, which in my view transcends issues of location, patronage and identification of the sculptural elements. The focus was on understanding religious architecture in its myriad manifestations across eastern India, with marked distinctions not only between Andhra and Orissa, but also between contemporary sites, such as those of Ratnagiri, Udayagiri and Lalitagiri in Orissa. Archaeological data encompasses not just religious structures and standing monuments, but more importantly it also includes an analysis of the location of religious architecture within the social domain. Religious architecture is thus an important indicator of interaction with diverse interest groups, such as worshippers, ritual specialists, patrons, artisans, etc. No religious architecture can survive without adequate maintenance and we do know that shrines and other sacred architecture far outlived their patrons. Of crucial importance to this paper then was the social history of the monument, rather than an emphasis on architectural style and form. All the photographs have been taken by the author and the maps have been drawn by Ms. Uma Bhattacharya.


References


Anguttara Nikaya, R. Morris & E. Hardy translated, London: Pali Text Society, 1885-1910, 5 volumes.
Bandyopadhyay, B., Udayagiri – 2, New Delhi: Archaeological Survey of India, 2007.
Chakrabarti, Kunal, The Religious Process: The Puranas and the Making of a Regional Tradition, New Delhi: Oxford University Press, 2001.
Chanda, Ramaprasad, Exploration in Orissa, Memoirs of the Archaeological Survey of India, no. 44, 1930: 1-23.
Chattopadhyaya, B. D., The Making of Early Medieval India, New Delhi: Oxford University Press, 1994.
Chhabra B. Ch., Expansion of the IndoAryan Culture during Pallava Rule, New Delhi: Munshiram Manoharlal, 1965.
Christie J. W., State Formation in Early Maritime Southeast Asia: A Consideration of the Theories and the Data, Bijdragen tot de Taal-, Land- en Volkenkunde, 151, 2, 1995: 235-88.
Cunningham, A., The Bhilsa Topes, Varanasi: Indological Book House, 1966 (reprint).
Desai, D., Social Dimensions of Art in Early India, In Proceedings of the Indian History Congress, Gorakhpur: Indian History Congress, 1989-90: 21-56.
Donaldson, T. E., Iconography of the
Buddhist sculpture of Orissa, New Delhi: Indira Gandhi National Centre for the Arts and Abhinav Publications, New Delhi, 2001.
Fogelin, Lars, Sacred Architecture, Sacred Landscape: Buddhist Architecture in north coastal Andhra Pradesh, Ray H. P. and Sinopoli C. edited, Archaeology as History in Early South Asia, New Delhi, ICHR & Aryan Books International, 2004: 376-391.
Germano, David and Trainor, Kevin, ed., Embodying the Dharma: Buddhist Relic Veneration in Asia, Albany: State University of New York Press, 2004.
Ghosh, A. and Sarkar, H. Beginnings of Sculptural Art in Southeast India: a stele from Amaravati, Ancient India, 20 & 21, 1964-65, pp. 168-177.
Ghosh, Mallar, Development of Buddhist Iconography in Eastern India, New Delhi, 1980.
Goody, J., The Power of the Written Tradition, Washington-London: Smithsonian Institution Press, 2001.
Gunawardana, R.A.L.H., Robe and the Plough, Tucson: University of Arizona Press, 1979.
Hanumantha Rao, B.S.L., Ramachandra Maurthy, N.S., Subrahmanyam, B. and Sivanagi Reddy, E. Buddhist Inscriptions of Andhradesa. Secunderabad: Ananda Buddha Vihara Trust, 1998.
Hock, Nancy, Buddhist Ideology and the Sculpture of Ratnagiri, seventh through thirteenth centuries, PhD thesis, University of California, Berkeley, 1987.
Howell, J. R., Excavations at Sannathi
1986-1989. New Delhi: Archaeological Survey of India, 1995.
Jacq-Hergoualc’h, Michel, The Malay Peninsula: Crossroads of the
Maritime Silk Road (100 BC – 1300 AD), Brill, Leiden, 2002.
Khan, M.A.W., Monograph on Yelleswaram Excavations, Hyderabad, 1963.
Knox, Robert, Amaravati: Buddhist Sculpture from the Great Stupa, London: The British Museum, 1992.
Lee, Kwangsu, Buddhist Ideas and Rituals in Early India and Korea, New Delhi: Manohar, 1998.
Legge, James, A Record of Buddhistic Kingdoms, London: Sacred Books of the East, 1886/1965.
Leoshko, Janice, Sacred Traces: British Explorations of Buddhism in South Asia, Aldershot: Ashgate, 2003.
Longhurst, A. H., The Buddhist Antiquities of Nagarjunakonda, Madras Presidency, Memoirs of the Archaeological Survey of India, no. 54, New Delhi: Archaeological Survey of India, 1938.
Lopez, Donald S. ed., Buddhism in Practice, Princeton: Princeton University Press, 1995.
Mahadevan, I., Pottery Inscriptions in Brahmi and Tamil-Brahmi, Begley V. et. al., 1996, The Ancient Port of Arikamedu: New Excavations and Researches 1989-1992, Pondicherry: EFEO, 1996: 287-315.
Mitra, Debala, Buddhist Monuments, Calcutta, 1980.
Mitra, Debala, Ratnagiri (1958-1961), Memoirs of the Archaeological Survey of India no. 80, 2 volumes, New Delhi: Archaeological Survey of India, 1981.
Nandi, R. N., Social Roots of Religion in Ancient India, Calcutta: K. P. Bagchi, 1986.
Nath, Vijay, Puranas and Acculturation – A Historico-Anthropological Perspective, New Delhi: Munshiram Manoharlal, 2001.
Ray, Himanshu Prabha, Monastery and Guild: Commerce under the Satavahanas, New Delhi: Oxford University Press, 1986.
Ray, Himanshu Prabha, Inscribed
Potsherds: A Study, Indica, 24, 1: 1-14, 1987.
Ray, Himanshu Prabha, The Archaeology of Seafaring in Ancient South Asia, Cambridge World Archaeology Series, Cambridge University Press, 2003.
Ray, Himanshu Prabha, The Archaeology of Sacred Space: Introduction. In Ray, H. P. and Sinopoli, Carla ed., Archaeology as History in Early South Asia. New Delhi: Indian Council of Historical Research & Aryan Books International, 2004: 350-375.
Roy, Anamika, Amaravati Stupa, Delhi: Agam Kala Prakashan, 1994.
Sahu, N.K., Buddhism in Orissa, Cuttack, 1958.
Salomon, R., Indian Epigraphy, Austin: Center for Asian Studies, 1998.
Sarkar, H. and Misra, B. N., Nagarjunakonda. New Delhi: Archaeological Survey of India, 1999.
Sarkar, H. And Nainar, S.P., Amaravati, New Delhi: Archaeological Survey of India, 1992.
Sarma, I.K., Studies in Early Buddhist

Monuments and Brahmi
Inscriptions of Andhradesa, Nagpur: Dattsons, 1988.
Schopen, Gregory, Bones, Stones and Buddhist Monks, Honolulu: University of Hawaii Press, 1997.
Schopen, Gregory, Buddhist Monks and Business Matters, University of Hawaii Press, 2004.
Schopen, Gregory, Figments and Fragments of Mahayana Buddhism in India, Honolulu: University of Hawaii Press, 2005.
Sastri, T.V.G., Kasturibai, M. and Veerender, M. Vaddamanu
Excavations. Hyderabad: Birla Archaeological & Cultural Research Institute, 1992.
Shimada, Akira, The Great Railing at Amaravati: An Architectural and Chronological reconstruction, Artibus Asiae, LXVI, no. 1, 2006: 89-141.
Singh, Upinder, Documentation and
Destruction: The Case of Amaravati, H. P. Ray and Carla Sinopoli, edited, Archaeology as History in Early South Asia, New Delhi: Indian Council for Historical Research & Aryan Books International, 2004: 34-50.
Sircar, D.C., Select Inscriptions Bearing on Indian History and Civilization, Calcutta, 1965.
Sircar, D.C., Inscriptions of Asoka, third edition, New Delhi: Government of India, 1979.
Spiro, Melford E., Buddhism and
Society, London, 1971.
Subrahmanyam, R., Salihundam: A
Buddhist Site in Andhra Pradesh. Hyderabad: Government of Andhra Pradesh, 1964.
Thaplyal, K.K., 1972, Studies in Ancient Indian Seals, Lucknow: Akhila Bharatiya Sanskrit Parishad.
Wagle, N., 1995, Society at the Time of the Buddha, Bombay: Popular Prakashan.
Willis, M., Buddhist Reliquaries from Ancient India, London: The British Museum, 2000.
Wink, A., Al-Hind, The Making of the Indo-Islamic World, New Delhi: Oxford University Press, vols. I & II, 1990 & 1999.



Source