Tantric Buddhism and the Pala rulers of Eastern India

C Tantric Buddhism and the Pala rulers of Eastern India

The Pala-style: regional variations, national claims and international legacies

I.I. The Pala rulers as patrons of Tantric Buddhism in medieval Indian historiography

The chapter begins with introduction to the Pala dynasty-one of the three apices of the Kanauj triangle after the fall of Harshavardhan’s empire, thereby interrogating the role of the imperial Palas in establishment and patronage to monasteries like Nalanda, Odantapuri, Vikramshila, Somapuri, Jagjivanpur, Jagaddala etc. This sub-chapter primarily focuses on copper plates, sculptures and palm-leaf manuscripts and the role of these objects of further art-historical scrutiny, in determining the Pala genealogy which glorifies itself through such inscriptions.

With no historical record of the Pala rulers in contemporaneous literature except the late Ramacarita to discuss the genealogy of the kings, the closest account is of the 16th century chronicler Lama Taranatha, whose History of Buddhism in India, is written in a distinct Buddhist faith, believing the supernatural to be true and real. For Taranatha, the dynasty was established by the grace of a miraculous siddhi obtained by a boy born in Pundravardhana who got the initiation from an acarya to propitiate the goddess Cunda. (Chattopadhyaya A. , Lama Taranath's Hisory of Buddhism, 1980, pp. 257-264) It was the blessings of the Goddess Cunda, that enabled the devotee to overpower and kill a nagini, who had been devouring the next deemed king since the time of

Gobicandra or Lalitcandra for several years. On selecting him their king for consecutive seven days and experiencing his pious nature, the people selected him their king, naming him Gopala, literally meaning– the leader of the herd. Gopala is also credited by Taranatha to have founded the Odantapuri or Nalendra Vihara as an act of generosity, close to Nalanda in the present-day town of Bihar Sharif on his victory over Magadha. However dubious with no distinction whatsoever between history and legend; Lama Taranatha’s account of the times still remain crucial as his interest in the lineage of the mahasiddhas forms the core of the chronology. Gopala I, for example is mentioned to have been enthroned that year following the death of mahasiddha Krsnacarya by Pandit Indradutta as cited by Taranatha; Krsnacarya or Kanha, a very important figure in this thesis thus appears to be essentially related to imperial genealogy of the Palas– the patrons of Tantric Buddhism.

Lama Taranatha’s own biography of Kanha, however places him towards late-ninth–earlytenth century to have been the abbot at the recently established Somapura Mahavihara by Devapala, the most active period for the mahasiddha tradition.

As per Rakhaldas Banerjee, one of the first historiographers to ascertain the genealogy of the rulers based on extensive epigraphic evidences, the Pala dynasty was founded by the result of an election by the subjects to select their king, after a spell of foreign invasions and authority and anarchy in Bengal after the fall of Harsavardhana’s empire. Verse 4 from the Khalimpur copper-plate grant of Dharmapala testifies to this by stating– Matsya-nyayamapohitum prakrtibhir-laksmyah karan-grahitah Sri-Gopala iti kshitisa-sirisam chudamanistatsutah. Banerjee mentions that the translator K.P. Jaisawal’s translation “to escape from anarchy” summarizes the specific cause of the election embodied by the phrase– Matsyanyayam-apohihmi (Banerjee R. , 1915, p. 45), where Matsya-nyaya usually

means the anarchical state of political affairs where one kingdom is taken over by a larger kingdom, like a minnow is gulped down by a gigantic fish. On the basis of inscriptions of various copper plates, Rakhaldas says that Gopala, the son of a successful military warrior named Vapyata, came to the throne around 750 CE, thereby refuting the earlier assumptions of V.A. Smith of the premature chronology of the dynasty beginning from 732 CE. According to him, Gopala was an aged man when he came to the throne and was succeeded soon after by his son Dharmapala from the marriage with princess Daddadevi from the Bhadra country.

Dharmapala is credited by Banerjee to be the real founder of the dynasty, re-gaining lost control over the territories annexed by the Gurjara king Vatsaraja, who in-turn was defeated by the Rashtrakutas– Dhruva and Govinda III, creating a lack of imperial authority and hence social unrest in Northern India. Dharmapala, a contemporary of the Gurjara king, Nagabhatta II, and the Rashktrakuta king, Govinda III, is believed to have ruled in the latter years of the

eighth century and the early years of the ninth century. The most illustrious event of his rule was the northern conquests– the displacement of a certain king Indraraja from the imperial capital of Kanauj, and the instatement of a puppet king, Cakrayudha at his behest, probably hailing from Pancala. This however was not conducive to the Gurjara Nagabhatta II whose momentary seizure of the throne, mentioned in the Gwalior copper-plate inscription of Bhoja I , was ousted by the coalition of the Palas and the Rashtrakuta king, Govinda III, as recorded in the Amogavarsha I grant and poetically eulogized in the Radhanpur grant of Govinda III stating that the Gurjara king fled just as monsoon does with the coming of autumn.

Rakhaldas, the scientific native historian, as noted by Tapati Guha Thakurta (GuhaThakurta, 2004, p. 112), says that Dharmapala must have ruled at-least for thirty-two years as the Khalimpur grant mentioned above comes from this regnal year, but negates any possibility whatsoever of Lama Taranatha’s hagiographic account of the ruler, having ruled for glorious sixty-four years. Another inscription from a sculpture of a four-faced Mahesh erected by a certain Kesava, son of a sculptor named Ujjvala, found at Bodhgaya and later moved to the Indian Museum, Calcutta where the inscription was deciphered by Pandit Nilmohan Chakravartti, is dated to be in the twenty-sixth regnal year. Banerjee by citing Dr. Kielhorn mentions that Dharmapala was married to Rannadevi, daughter of Parabala, a powerful Rashtrakuta chief. According to R.C Majumdar, the Tibetan tradition holds Dharmapala to be the founder of the Vikramshila Mahavihara with 114 teachers in various subjects and a central temple surrounded by 107 smaller shrines (Majumdar, 1971, p. 111).

Dharmapala’s eldest son Tribhuvanpala can be assumed to have died during the former’s life-time and thus he was succeeded by Devapala, the second son, who according to Taranatha would go on to capture Varendra in the east and consecrate the Somapuri vihara; capture Odivisa from the control of tirthikas and also rule over Rara in Bengal. Dharmapala had yet another third son named Vakpala, whose son was called Jayapala. The Somapuri vihara and the Vikramshila vihara on excavation has yielded a high-degree of similarity in groundplan, elevation and programming; suggesting that they were initiated in the rule of the same king Dharmapala, Somapuri being completed under the reign of Devapala in all probability (Images 1 and 2). The Badal pillar inscription also mentions Darbhapani,

the minister to the king on whose counsel Devapala could expand his kingdom far and wide. Kedaramisra, the grandson of the minister holds the position in the reign of Vigrahapala I or Surapala I, who both Rakhaldas Banerjee and his mentor at the Presidency College, Haraprasad Sastri held erroneously to be the same person from confusion arising in the genealogy mentioned in the pillar inscription and the Bhagalpur copper-plate. However, to his merit Banerjee notes that Guravamisra, from the same family of Darbhapani and Kedaramisra, was the minister to Narayanapala from the Bangarh grant, thus establishing a familial continuum within the imperial court.

Taranatha’s historical legend of Devapala’s conquests of Odivisa is also testified by the phrase “Utkilit Otkala kulam” in the second verse of the pillar inscription, translating to bhayat=tatha na samaram svapne = pi pasyed = yatha. Verse 15; Ibid. “eradication of the race of the Utkalas”; besides curbing the pride of the Hunas, the haughtiness of the Dravidas and the Gurjaras. (Majumdar, 1971, p. 112) In this copper-plate, one can discern a strong hold of the ministerial family on political affairs of the state and what R.C.

Majumdar says “extravagant pretensions” on their behalf. Majumdar also deduces from other epigraphic evidences that the Gurjara king whose pride was subdued was none other than Bhoja I, the Utkala invasion during or being immediately after king Sivakara of the Kara dynasty. The reference to the king of Kamarupa or Pragyotisha, accepting the sovereignty of Devapala is also cited in regards to the Bhagalpur copper plate of Jayapala, while the Dravida ruler mentioned, is supposed to be the Pandya ruler Sri Mara Sri Vallabha.

Devapala’s rule which extended for almost thirty-five years, the Monghyr grant coming in the thirty second of his reign, marked a pan-Indian expansion of an empire rooted in Bengal.

The copper plate in the fifteenth verse records Devapala’s vast kingdom to be bound by Himalayas in the north and the Ramsevar Setubandha in the South, which though possibly a poetic exaggeration, seems true to a large extent. Devapala’s name and fame was not only restricted to India but went far in the South-east so much so that king Balaputra Deva of the Sailendra Dynasty ruling over Sumatra, Java and the Malay Peninsula sent an ambassador to the Pala court for endowment of a grant of five villages to build a monastery at Nalanda, the seat of international Buddhist learning and culture with the seal of Pala patronage.

(Majumdar, 1971, p. 115) Nalanda Mahavihara, with its present foundations led by Kumaragupta and enjoying the patronage of Harshavardhana in the seventh century; was yet another trope for inheriting the classical antiquity and claiming national and international sovereignty with emphasis on their distinct Buddhist religious affiliation. Sylvian Levi also identified Devapala as the ruler of “Wu-cha” who does “what is pure” and sent an ambassador to the king of China in 795 CE (Majumdar, 1971). In an opposite vein to suzerainty of the Palas, the Chronicles of Ladakh suggest that Bengal had to face some difficulty from the Tibetan side, a relationship which historians cannot be exactly determine in lack of other sources (Chowdhury, 1967).

In the section of the The History of Image Makers, Lama Taranatha mentions the names of two artists, Dhiman and his son Bitopalo, active during the reigns of Dharmapala and Devapala, whose techniques of metal-casting, engraving and painting earned them great repute in Varendra and Vamgala. The followers of the son are said to have formed the Madhya-desa tradition because they were mainly widespread in Magadha, while those of the father formed the eastern tradition centred around Somapura Mahavihara, near present-day Rajshahi, Bangladesh. While Taranatha attempted in his encyclopaedic treatise to touch upon the regional styles of Buddhist art during the eighth and the ninth century CE, mentioning the styles of Kashmir Art, Western Indian (Gandharan) Art and the Pala-styles influencing the art of Nepal; the early twentieth century scholarship of Haraprasad Sastri and Rakhaldas Banerjee, hinged on the history and antiquity of the Bengali script aimed at evoking a glorious past of Bengal.

(Guha-Thakurta, 2004, p. 112) Rakhaldas, trained in the “arena of museums” as Guha-Thakurta mentions, attempted at a systematic chronology of Pala-Sena sculptures by deciphering the inscriptions, with the assistance of Pandit Nilmani Chakravartty. He also for a brief period headed the Eastern wing of the Archaeological Survey in the excavation of the Somapura Vihara, after the tenure of D.R. Bhandarkar was over; a joint-venture by the Varendra Research Institute, Rajshahi and the Department of Ancient History, Calcutta University. In 1911, even while appointed as the archaeological assistant in the Indian Museum, he played a key-role in founding the Bangiya-Sahitya Parishat, an institution to hone vernacular scholarship focused on deciphering the Bengali script.

With a territorial focus on Bengal to invoke the glory of the pre-Islamic past of the region, he displayed the objects collected by the society in 1912-’13 to coincide with the annual session of the Indian National Congress. (Guha-Thakurta, 2004) By an analogic analysis of sculptural styles with his previous scholarship on the development of the “proto-Bengali” script, distinct from the northern Nagari script, he argued for a regional school of sculpture extending from eastern part of Magadha to Varendra active for about four centuries– suggesting that the regional schools of Dhiman and Bitopala were influential till about the twelfth century. During the Islamic invasions, he believes the “proto-Bengali” script remains frozen and static, and the regional school of visual arts dies a pre-mature death.

After Dharmapala and Devapala, the next king mentioned in the inscriptions– Surapala I and/or Vigrahapala I had created a prolonged confusion among historians in determining the individuals. D.K. Ganguly argues that since the Bhagalpur copper-plate mentions Vigrahapala I after describing the conquests of Jayapala, he is the son of the latter; while Surapala I is the son of Devapala with his wife Mahata as affirmed by excavation of a copper-plate at Mirzapur, Uttarpradesh. (Ganguly, 1994, p. 29) The excavation of yet another copper-plate grant excavated from Nandadirghi Vihara, Jagjivanpur Malda helps Ganguly for the first time

introduce in the dynastic lineage– Mahendrapala, son of Devapala and Mahata as an intermediate ruler who died childless before Surapala I came to the throne. (Image 4) The copper-plate helped clear the earlier confused identity of the Pala king Mahendrapala with that of the Pratihara king, who was mentioned in several inscriptions. The re-constructed model of the vihara, panels of terracotta reliefs and the metal icons with the copper-plate in the State Archaeological Museum, Behala, Kolkata; also marks the presence of a vihara in Malda between Vikramshila and Somapuri Mahaviharas, between Magadha and Varendra; besides reinstating Mahendrapala in the Pala-genealogy.

The Badal Pillar inscription and another inscription of a Buddhist statue erected by an old Buddhist monk, named Purnadasa, at Odantapuri from Surapala’s second regnal year (Banerjee R. , 1915, p. 56) re-affirms his claim to power, before Vigrahapala I, son of Jayapala and grandson of Vakpala, could rule for a very short while. Subsequently, Narayanapala took the reins of the empire in troubled times against the onslaught of the antagonistic GurjaraPratiharas who took control over Magadha as mentioned in the Gwalior Prasasti of Bhoja I.

1 Vikramshila Mahavihara, Antichak, near Kahalgaon, Bihar– Photograph by author

2 Somapura Mahavihara, Paharphur, Bangladesh– Image from Internet

3 Copper-plate grant of Mahendrapala with metal sculptures– Photographs by author

Narayanapala’s long reign of fifty-four years is marked by several inscriptions– the first coming from his seventh regnal year at Gaya temple, one in his ninth regnal year found from Bihar and brought to the Indian Museum and the Bhagalpur copper-plate issued from Munger in his seventeenth regnal year. However, from the seventeenth till the fifty-fourth regnal year there is no epigraphic evidence of the ruler from Magadha, suggesting that he had lost control of the northern areas, testified by some inscriptions of the Gurjara king Mahendrapala I– specifically one mentioned by A.M. Chowdhury at Paharpur in the fifth regnal year of Mahendrapala I. (Chowdhury, 1967, p. 55) An inscription from Bihar dated to the fifty-fourth regnal year of Narayanapala hints at the reclamation of the territory in the last days of his reign from the after the death of Mahendrapala I and devastating attacks of the Rashtrakuta king,

Krishna II, which rattled the Pratihara empire. Another interesting facet of Narayanapala’s inscriptional record suggests him as a patron to Pasupata temples, an exceptional boast of having built a thousand Siva temples from the ruler of a distinctly pro-Buddhist dynastic lineage. (Chowdhury, 1967, p. 56) Rajyapala, son of Narayanapala, is referred to in inscriptions from his twenty-fourth, twenty-eighth, thirty-first and thirty-second regnal year; asserting that he ruled for the latter number of years. He got married to a Rashtrakuta princess. During the reigns of his successors Gopala II and Vigrahapala II, the

Palas had to face the wrath of aggression of the Candelas and the Kalacuris, both dynasties emerging with the wane of Pratihara strong-hold in Madhya-desa. Gopala II’s reign of about seventeen years is marked by epigraphic evidences from the first (Image 4) and second regnal years, a colophon of an Astasahasrika Prajnaparamita manuscript, dated to his fifteenth regnal year from Vikramshila Vihara attesting his control over Magadha while a copper-plate from his sixth regnal year found in Comilla asserts his power in the South-Eastern part of Bengal, a strong footing of the Candra dynasty. The flattering inscriptions attempt to repress the reverses suffered by the empire. Colophon of a Pancaraksha preserved in the British Library helps assign a reign of twenty-six years with inscriptions from Kurkihar, near Gaya, coming from the third and nineteenth regnal years of Vigrahapala II.

The first Candela conquest of Bengal by Yasovarman is dated in a Khajuraho inscription dated to VS 1011 or 954 CE. Another inscription dated to 1002 CE mentions Yasovarman and his son Dhanga. Two Kalacuri rulers, Yuvaraja I and Laksmanarja, as well as the Kamboja and the Chedi rulers sacked Bengal during this the combined reigns of Gopala I and his son Vigrahapala II, limiting the kingdom only to Anga and Magadha; when Mahipala

I ascended the throne to establish what Rakhaldas Mukherjee calls the “second empire.” (Banerjee R. , 1915, p. 68) While the Pala prowess was waning under Gopala II and his son, other dynasties assumed self-honorific titles such as Gaudapati assumed by the Kamboja rulers (or the Mongolian rulers as postulated by Rakhaldas Banerjee from interpreting the Dinajpur pillar inscription) and Paramsevara, Paramasaugata and Maharajadhiraja by Bhavadeva

(Majumdar, 1971, p. 129). The Eastern and Southern parts of Bengal saw the rise of regional powers such as of Bhavadeva known from an inscription in a Chittagong temple and his son, Kantideva known from an inscription of a grant in Vardhamanpura, present-day Bardhhaman in West Bengal (Majumdar, 1971, p. 130) and the Candra dynasty with Trailokya Candra and Sri Candra being the most significant rulers after the death of Narayanapala. Banerjee also mentioned another independent dynasty in Eastern India known as the Khadgas, known from the copper-plate of Devakhadga from the first half of the tenth century.

4 Vagesvari (1st regnal year of Gopala II), Indian Museum, Kolkata– Photograph by author

Anadhakrta vilupta or “lost owing to non-occupation” (Majumdar, 1971, p. 131) was the state of the paternal kingdom of the Palas (Janakabhuh as mentioned in the Ramacharita) and pitr-rajya as mentioned in the Verse 12 of Bangarh grant (Majumdar, 1971), that Mahipala I had to recover on ascension. In the third and the fourth regnal years respectively, the Baghaura inscription on a Visnu image and the Narayanpur inscription on a Ganesa image (both in Tippera district) reclaims the authority of Mahipala I over Samatata or Easter Bengal. The Belwa grant issued on the fifth regnal, Majumdar argues, helps one recognize that

Mahipala I recovered Eastern and Northern Bengal in the very early years of his long reign which saw hostility with a new set of nemesis– the Cedis and the Calukyas and the Colas, besides the Vakatakas under Gangayadeva. The most celebrated of these raids was by Rajendra Cola as testified in the Tirumalai Prasasti which also discusses the occupation of West Bengal by three political powers– Dharmapala of Dandabkuti near Utkala, Ranasura of Southern Radha and Mahipala of the Northern Radha; all three of whom had to flee the battlefield against the might of Rajendra Cola who was thus able to fulfil his ambition of reaching the Ganga. Rajenda

Cola’s expedition of Northern India which lasted somewhere between 1024 and 1026 CE, though hardly altered the regional politics and claim of subsequent power, but is to be considered significant in relation to his naval expeditions and imperialistic ambitions in the South-East Asia, However, it is important to note that there is not a single reference of the Colas crossing the Ganga; and drawing from a Sanskrit manuscript of a play titled Candakausika by Ksemisvara found by Haraprasad Sastri, Rakhaldas tried to put forth the theory that the Cola king had to reverse from the bank of the Ganga. A section of the play compared Mahipala’s victory over the Karnatas to the victory of Candragupta Maurya over the Nandas. The identifications by Sastri and Banerjee with Mahipala I and the Colas, have been objected by scholars who identified Mahipala in the play with the Pratihara king and the Karnatas as the Kalachuri or even the Calukyans from Kalyani. (Sinha, 1954, p. 416)

Despite these reverses suffered by Mahipala I, he earned the repute of resurrecting the Pala kingdom from almost ruins, so much so that he is remembered in the popular Bengali proverb “Dhan bhangte Mahipal er geet”’; DC Sen in his Eastern Indian Ballads also argues for numerous ballads remembering Mahipala’s peaceful pursuits till about five centuries later when Brindaban Das wrote Caitanya Bhagabat in 1572 CE (Chowdhury, 1967, p. 88) . The Amagaci and the Manahali plates also mention Nyayapala’s birth as result of merits earned by Mahipala I. The rise of the power of the Palas is also verified by religious inscriptions from Sarnath near Benaras and for the restoration of the Nalanda Mahavihara in the eleventh regnal year, parts of which were destroyed in fire. Another palm-leaf manuscript of the Astasahasrika Prajnaparamita (Asiatic Society G 4713) was copied in the 6th year of his reign at Nalanda with twelve illustrations. (Image 5). Another illustrated copy of the Astasaharika manuscript (Cambridge MSS Add.1464) with two brilliantly painted wooden covers mentions the name of the donor– a woman named Ladoka in the colophon, helping ground the urgency of this dissertation to relocate the position of women as patrons. (Image 6)

The Sarnath inscription of Vikram Samvat 1083 (=1026 CE) spoke of repairs and construction of temples under the Mahipala, the king of Gauda (Gaudesvara), while the actual work was entrusted to his brothers, Sthirapatra and Vasantapala. Historians from R.C Majumdar onwards, have however noted that the inscription doesn’t prove the possession of Northern India by the Palas, but a customary grant from the royalty for a religious place like Benaras in order to earn merits

or punya. Though it has been difficult to ascertain the dynastic authority of the king ruling over Mithila or Tirhut around 1020 CE, with scholars debating the lineage of Gangayadeva, from multiple interpretations of a Ramayana copied in Samvat 1076 of an unknown era (V.S 1076= 1020 CE– Gangayadeva of Kalacuri/ Saka 1076= 1153 CE– Gandayadeva, son of Naynadeva of Mithila); on the sole basis of the Sarnath inscription; Mahipala I cannot be established as the ruler of the region. The Imadpur inscription from the forty-eight regnal year of Mahipala I also matches up quite closely to the estimate of Lama

Taranath of his reign to be of fifty-two years. Mahipala’s son and successor Nyayapala who had to face persisting trouble from the Kalacuris– Karna, the son of Gangayadeva. The Tibetan tradition also holds Atisa Dipamkara to be responsible for mediating a treatise between the rulers; while Atisa’s date of leaving for Tibet at the age of around sixty nears 1041 CE, Karna’s ascension to throne is fixed by an inscription equating to 1039 CE (Majumdar, 1971, p. 138). Weaving together these textual sources, it seems probable that Mahiapala’s reign came to an end around circa 1027 CE as per D.K. Ganguly who agrees with D.C.

Sircar; with Nyayapala ascending the patrilinear throne and ruling for fifteen years; whose only inscriptional evidences come from the fifteenth regnal year at Gaya Krsna-Dvarika temple and from a Pancaraksa manuscript in the Cambridge Library (Banerjee R. , 1915, p. 79) from the fourteenth year (Images 7 and 8), from a woman donor Daddaka, who Bendall thinks to be the same donor of MSS 1464, i.e. Laddaka. (1883:175)

5 Astasaharika Prajnaparimta (G 4713, Asiatic Society) – Buddhist goddess and Buddha’s birth (details)

6 Astasaharika Prajnaparimta (Cambridge MSS 1464) painted covers and Khadiravani Tara detail (226 v)

7 Pancarksa (Cambridge Mss Add 1688– 14th regnal year of Nyayapala) Top to bottom– 45 recto; 66 verso; 67 recto

8 Details: Sahasrapramardini and a wrathful yogini (from 45 r and 66 v respectively)

Nyayapala’s son, Vigrahapala III, also ruled for a short-while assigned to about seventeen years; their combined reigns Chowdhury notes were not long enough attested by the fact that the son of the craftsman who incised the Bangarh inscrpiton of the ninth regnal year of Mahipala I, inscribed the Amacgachi plate from the twelfth regnal year of Vigrahapala III.

Although there a number of epigraphic evidences of the latter’s reign, it seems to have been marked by the return of the Kalacuri king Karna whose advancement as far as to West Bengal crushing the last of the Candra kings can be verified by epigraphic evidences such as a pillar at Paikot in the Birbhum district (Chowdhury, 1967, p. 96). Karna’s might in the latter years of his life was considerably crushed by the monarchs of the neighbouring kingdoms– Kirtivarmans of the Candelas, Udayaditya of Malwa and Bhimadeva I of Anihilavad who is eulogized by the grammarian Hemacandra (Banerjee R. , 1915, p. 79). Bilhana, the court-

poet of the Western Calukyans, describes the career of Vikraditya’s (VI) conquests up to Gauda and Kamarupa (Chowdhury, 1967, p. 97). The manuscript of Ramacharita discovered by Haraprasad Sastri mentions the defeat of Karna at the hands of Vigrahapala III who as a gesture of a peace-treaty had to give his daughter, Yuvanasri to Vigrahapala III in marriage. It seems possible that South-Eastern Bengal was occupied temporarily by the Palas before the rise of the Varmans, who rose power on the aftermath of the Kaivarta revolution during Mahipala II. Also, in the middle of the eleventh century, there were attacks from the Somavamsi rulers of Odisha and of Ratnapala, the king of Kamarupa, as mentioned in his Bargaon grant.

Vigrahapala had three sons– Mahipala II, Surapala II and Ramapala; Mahipala’s tumultuous short reign– the revolutions of a specific Kaivarta caste and the subsequent seizure of power by Ramapala is the narrative of Ramacarita, written by Sandhyakaranandi, a partisan of the Ramalapala’s court and hence cannot be regarded as unbiased testimony. Mahipala II on the basis of the narrative seems to have prisoned the younger brothers Surapala II and Ramapala and gone to war with the rebellion led by the Kaivarta chief Divya, disregarding the counsel of the ministers. Divya, by the author of the text is portrayed as dasyu or

villain who on killing Mahipala II in the battle dared to usurp the throne, a patriarchal symbol of the claim to temporal medieval power. Divya was succeeded by his younger brother Rudoka who in turn was followed by his son, Bhima– thus creating a hiatus in the Pala rule. Surapala II seems to have been at the helm of the Pala power after Mahipala II for about only two years, as recorded in the list of Pala rulers in the Manahali grant of Madanapala, a historical fact unmentioned in the Ramacarita (Banerjee R. , 1915, p. 85).

On ascension, Ramapala remained inactive for a while unable to decide the course of actions, but soon had to take a resolution pacified by sons and ministers to prevent the complete collapse of dynastic power where the barbaric Kaivartas attacked Ramapala again. By humbling seeking help from local chieftains with well-equipped forces, all whose names are enlisted in the Ramacarita, Ramapala was able to defend his inherited kingdom from impending disaster; thereby also mobilizing the political dismemberment of Bengal with the decline of the imperial authority of the Palas (Majumdar, 1971, p. 147). Allied with Ramapala’s forces where his maternal uncle Mathana or Mahana, the Rashtrakuta prince, with his two sons

Kanharadeva and Suvarnadeva. Having gained the support of fourteen Samantas, Ramapala’s troops proceeded towards Varendra under his cousin Sivaraja who succeeded in breaking up the frontier guards making way for the main army which entered after crossing the Ganges. It was but by an evil fate of destiny that Bhima was frightened out of his life and got captured alive in the battle as mentioned in the commentary to the Verse 16 of the Ramacarita. The victory was followed by the customary sack of the enemy’s capital city Damara and establishment of a new capital Ramavati where he laid the foundation of the Jagaddala Vihara. Ramavati continued to be the capital of the Palas as the Manahli grant of Madanapala discussed earlier was also issued from here (Banerjee R. , 1915, p. 91).

The narrative also mentions the Varman kings of Vikramapura sending his own elephant as a gift to accept Pala suzerainty. Ramapala also set forth on his ambition to recapture lost territories, taking control over Kamarupa and indulging in long-standing hostilities with the late Ganga dynasty of Utkala and Kalinga, the Gahadavalas of Benaras and Kanauj.

Ramapala is considered to have lived a long age almost notching up to Taranatha’s estimate of sixty-four years; colophon of a manuscript mentioning his fifty-third regnal year and the Chandimau image inscription from his forty-second year. Among four sons of Ramapala, no traces are found of Vittapala and Rajyapala after their father’s death in the Ramacarita; though they had played significant roles as military commanders in Ramapala’s military career.

Kumarapala succeeded the Pala throne; the Kamauli grant of Kumarapala’s minister, Vaidyadeva records a naval victory probably over the Varmans and subjugation of Timgyadeva, a vassal ruler from Kamarupa. Kumarapala was followed by his son Gopala III, who scholars have come to believe died as an infant or was even possibly killed by Madanapala

(Banerjee R. , 1915, p. 102); the events of Gopala’s death have been shrouded with mystery with only single verse dedicated in the Ramacarita alluding to some unknown enemy. However, inscription from a Sadasiva image from Rajibpur, Dinajpur district negates the theory of an early death of Gopala III, ascertaining that he ruled at least for fourteen years (Chowdhury, 1967, p. 131).

Madanapala came to the throne and ruled for at least eighteen years as testified by the Valdugar inscription dated to Saka 1083, 11th Jyaysistha corresponding to 4th May, 1161 (Majumdar, 1971, p. 156). But inscriptional evidences suggest a sharp decline in Pala territorial control as a part of Bihar was taken up by the Gahadavala king, Govindacandra and the advance of the Calukyans under Somesvara III and finally the loss of Gauda or northen Bengal to the Senas sometime after the eighth regnal year of Madanapala. Omission of specificities about the death of Gopala III and praises showered on Madanapala in majority of a section in Canto IV have made scholars come to conclusion that Sandhyakara Nandi completed the composition of the epic during

Madanapala’s rule, in the courtly culture of eulogy or prasasti. Madanapala can be said to be the last king from the lineage of the Palas though Taranatha’s estimate of Yaksapala, a contemporary of Ramapala and inscriptional evidences from seven manuscripts and two inscriptions of Govindapala suggest the

continuation of the imperial title related to some smaller dynasties with no established links to the imperial lineage, ruling around Bodhgaya. Another king called Palapala has been mentioned in a Jaynagar inscription dated to his thirty-fifth regnal year; but no evidential connection can be established; the same holds true for another Indradyunmapala. A dynastic tree drawn by D.K. Ganguly along with approximate regnal periods garnered from a huge plethora of inscriptions, are being given here along-with.

9 Map of Ancient and Medieval Bengal with territorial distinctions

Summarizing this section with an image of a map of ancient and medieval Bengal with sub-territorial divisions and sites mentioned in the inscriptional evidences (Image 9), the author would like to re-invoke Lama Taranath’s phrase ‘left no marks of their hands’ (i.e. did not construct any new monastery etc.) as a parameter to be counted among the Seven Palas. For the devout Buddhist Taranatha, the Palas, living many centuries before him, were the last patrons of the Law in India before the Islamic invasions of Bakhtiyar Khilji; the Seven Palas were the glorious patrons of Buddhist Siddhas, Panditas and Mahaviharas, which helped the final phase of flourishment of Mahayana and Tantrayana; the migration of monks from Vikramshila to Tibet from Santaraksita onwards, possible through diplomatic political relations. The relationship between the monastery and the state, Ronald Davidson mentions “was based on a system of mutual identity and on an ideology of magical performance.” (Davidson, 2002, p. 111)

Besides, the Kaivarta Revolt– mentioned passingly in the Ramacarita by Sandhyakarananandi, who was doing his best but to recast his patron Ramapala as the reincarnation of Rama from the classical epic as the re-enforcer of law and order, needs to be probed further in connection with the rebellious attitude of the siddhas as can be seen in many caryas which are not simply reflections of the inner microcosmic world of tantra within the body leading to Nirvana, but also on the outward world of the mundane Samsara. The pertinent question raised by Debiprasad Chattopadhyaya in the post-script to the introduction to Alaka Chattopadhyaya’s Caturasi Siddhir Kautha about the class and caste tensions of the times in relation to the caste-inclusiveness of the mahasiddha tradition (attested by the castes and social positions of each mahasiddha in the list) is significant and needs further enquiry (Chattopadhyaya D. , Post-script to the Introduction, 1960, pp. 37-55). However, with the lack of historical evidences it becomes difficult to theorize the revolt as a planned peasant-rebellion;

Debiprasad though with a Marxist stance allegorically compares Divya, Kaivarta chief, as the leader amongst the masses to Lenin, the leader of the Russian Revolution who infused the peasants and the labourers against a feudal Tzar rule in Russia. Further, Debiprasad postulates that the grounds for a counter-ideology against the dominant state ideology was created through the discourses of the eighty-four mahasiddhas (often coming from the lowest rungs of the society doing lowly occupational jobs) who chose to compose the caryas in a vernacular dialect to stay true to the politics of inclusiveness of the

marginalized, thereby infusing the representatives of the subaltern with a rebellious spirit and the ambitious claim to power– temporal through politics and eternal through tantric praxis. The Kaivarta Rebellion, therefore, needs to be taken into serious consideration while understanding the social dynamics at least in the waning phase of the Pala empire, which became instrumental for the final collapse of the kingdom, the last imperial strongholds of Buddhism in India. It is ironic to notice that the “Subaltern agency” becomes the instrument firstly for decentralization of power– marked by the rise of the Varmans in Samatata and Harikela and then for re-centralization under the Brahmanical Sena rulers who would vanquish the Palas and re-unite Bengal with a more firmly established caste hierarchy and sectarianism even among the Vaisnavas and the Saktas.

The specific caste of fishermen or boatmen– the Kaivartas also come to the foreground along with other Sudras with occupational or behavioural details in the mahasiddha tradition related to specific occupations like weaving, shoe-making to even burglary and massdeception. So, both within the political and the metaphysical texts of the times, there is a strong dissentious tendency against the dominant institutional ideology. In Divya and his nephew, Bhima who managed to flee after getting caught by the army and his friend Hari who reorganized the troops for a second attack on Ramapala, one therefore finds the early embodiments of political dissent amongst representatives of the governed masses and their gallant attempt to seize power when the monarchy was caught up in

familial disputes of succession. Debiprasad quite rightly mentions the high possibility of regional folk-tales and ballads about the bravery of these chieftains which unfortunately have not survived to go down in written history like the songs of Mahipala I as remembered in the proverbial saying: Dhan bhangte Mahipal er geet. For Debiprasad, the historian of the Lokayata movement(s) in ancient and medieval Indian history which emerged from teachings of the Carvakas to the Ajivikas, the Kaivarta Revolt becomes a seminal case for registering probable connections between the heterodox tradition of the mahasiddhas who at multiple instances were caught up in confrontations with the Samgha and were even expelled, and the aware political organization amongst leaders of suppressed castes and classes, the chief of who– Divya, was described by Sandhyakaranandi, the court-poet as demonic and evil with no true right to royal authority, which belonged to the Palas exclusively by birth.

Debiprasad through a citational matrix of Shashibhushan Dasgupta’s Obscure

Religious Cults and thereby of Gunaratna’s estimate of the cultic ‘carnivalesque’ practices of the Lokayatas (or the Carvakas or the Brhaspatyas) argues for the promiscuity of Tantra as a discourse of Indian materialism (Chattopadhyaya D. , 1959). Gunaratna’s mention in the commentary on Sat Darshan Sumccaya of the Lokayatas as Kapalika Yogis who smear their body with ashes and indulge in eroticism, enable Debiprasad to claim the common Lokayata source of kama-sadhan (pursuit of sexual pleasure through deha-vada) and artha-sadhana

(pursuit of wealth). Debiprasad also weaves in the ‘demonic’ Asuras and the female Asuri who was the first one to hear about Samkhya Darshana from sage Kapila in his ‘bottom-up’ narrative. For the Marxist historian, even ‘Original Samkhya’, free from the later spiritual overwritings of Advaita and Yoga (which itself was over-written with a theistic faith), was a strictly materialist philosophy where Purusha meant the male body (and not the metaphysical Soul or Atman) and Prakriti meant the female body (not only immanent matter). Thus, in trying to ally the orthodox Samkhya and Yoga Schools, the heretic Tantrika orders and the heterodox Lokayata philosophies at their materialistic origins, Debiprasad also reflects on the Vamadevya Samana of Chandogya Upanishad from the Vedanta tradition. The inter-relatedness of speech, magic, meter and sexual rites become seminal for framing the notion of “Tantric Art and Aesthetics” within a Buddhological framework.

Equally important as recognizing the class and caste politics of the Pala reign, is to reposition the critical agency of women within the Tantric Buddhist tradition in matters of patronage to art and culture, in Tantric rituals and rites as the Significant Other of the male mahasiddha, in initiating lineages of Tantra, in jointly composing or performing some of the caryas, in originating visualizations and hence iconographies of female goddesses like Nairatmya and Vajrayogini in union with the male Buddha as well as independent Female Buddhas (Shaw, 1995) with the possibility to attain Buddhahood in a female body. Citing Gerda Lerner. Elisabeth Schissler Floernza and John Scott, Miranda Shaw quite passionately argues throughout her book Passionate Enlightenment: Women in

Tantric Buddhism to reclaim “the historical agency of women, that is to concentrate upon how women acted, rather than how they were acted upon” (Shaw, 1995) to reconsider how they acted and viewed events rather than how they were viewed or described in textual sources predominantly written by male scribes in Sanskrit or Tibetan where the syntax of the language former use a neuter gender verb for a collective group of authors while the latter for simplicity’s sake often

effaces the female signature suffix “mo”, thereby re-gendering the historical agency, Taranatha’s identification of the lives of the Siddhas with particular reigns of Pala emperors citing the Sanskrit Pandits, combined with inscriptional evidences and the chronology discussed, may be useful for interrogating the biography of the Mahasiddhas in the milieu of Tantric Buddhist theory and praxis. With a critical lens of gender and subaltern are studies, the next chapter would also take into folds the dohas and caryas of the Mahasiddhas, a tradition which according to the classicist Taranatha is “the corrupt history of the doha” ; thereby grounding the divide between the imperial and the vernacular, which is to be dealt in detail subsequently.

I.II. The Pala style- regional variation(s), national claim and international legacy

Despite considerable regional variations of the Pala-style, it can be argued to be an inheritor of the classical Gupta idiom of the Sarnath Buddha with the tendency of mannerist dynamism and ornamentation, seeking to leave behind an international legacy at colonial outposts such as Java and Sumatra from the eighth to the eleventh century and to Tibet and the Trans-Himalayan belt via Santarakshita and Padmasambhava in the seventh century CE and via Naropa, Tilopa, Atisa and the translators from the eleventh century CE onwards. The Pala-style termed after the last imperial dynastic patrons of Buddhism in India becomes subject to critical enquiry with a dialectical engagement with regional dialects of Magadhi Apabhramsa and the notion of The Cosmopolitan Vernacular (Pollock, 1998).

The Pala style and its relation with Gupta Classicism: From a national visual language to regional dialects in plastic arts “Classical” in the artistic or literary contexts is generally defined as something pertaining to a Greek or Roman writer or work; any work possessing a similar quality or relationship to later art-forms is said to be classical thus. In the context of Indian culture, literature and arts (sculpture in particular), the age of the Imperial Guptas gained the label of “classical” in Indian historiography. Gupta Classicism became the ancient ideal of tradition for scholars like Havell, Coomaraswamy, Kramrisch as well as Sri Aurobindo and Sister Nivedita, who through their writings on art wanted Indians: the educated elite and the

middle classes, to look back to this glorious phase of central authority under the Guptas which would make the colonized nation culturally united against the foreign oppressors. The pan-Indian (and through colonization a Pan-Asian) claim of Gupta classicism was foregrounded through a proliferation of standardized iconography of the Brahmanical pantheon, in the arena of structural temple activity and in the Buddhist context through an array of Buddha and Bodhisattva sculptures whose signature style attribute them to single school of artists. The unity in pictorial style in the Golden Age of the Guptas in the words of Joanna Williams was exceptional as “Nothing in Indian art of the two centuries preceding AD 370 shows the self-conscious aesthetic harmony that distinguishes the Gupta oeuvre” (Williams, 1982). The historiographical bias becomes evident in quoting Williams at length:

Our knowledge of the political history of India is sketchy until the fourth century. At this point the Guptas appear, clearly Indians, with their capital in the north-east along the Ganges. The archival impulse in their records in their records as well as the praises of Indian playwrights have given a rounded picture of their accomplishments, rising above the mist that shrouds much of the Indian past. The Gupta emperors have become identified with the Gupta Age that pre-ceded that darker present.

Gupta Art, thus with the notion of ‘national art’ or national style since Asoka’s Mauryan pillar edicts deeply engaged the European eye trained in the classical valorisation of Greco-Roman aesthetics. The emergence of a singular Gupta style was the direct implication of the vast empire and provinces with the central authority of the Guptas; the artistic centre of Sarnath achieved the title of “classical” with its transcendental images of Buddhas and Bodhisattvas which for scholars like Kramrisch celebrated the true Indian spirit unlike the more immanent Gandharan Art. The Sarnath Buddha (Image 10) for the early twentieth century art-historians embodied the qualities of radiant inner calm and stillness (Abhinavagupta’s ninth Rasa Santa) the products of supreme wisdom. The Buddha robed in the simple, uncut monastic clothes (executed by the sculptor as a transparent drapery revealing the Buddha’s ever youthful male body) and

his religiosity exhorted by a large halo and auspicious markings or lakshanas became the bench-mark for the Buddha image throughout Asia from the fifth to about the eighth century CE. The Gupta artists brought together a range of standing and seated Buddha images in distinct mudras and marked characteristic features like the stylized ushnisa, the prominent urna and the arched eyebrows. The resemblances between the Gupta Sarnath model and the Sultanganj Buddha (arguably from the seventh century from Bihar) were so striking that they were classified as offshoot of the “classical national language” The Gupta period thus went down in historiography as a the “golden age” when the quintessential iconic Buddha image was created and later disseminated throughout the Asian Buddhist world. Gupta style thus stood at the cross-roads in historical development of the iconic meditative Buddha image in South Asia, creating a stable prototype for a multitude of later images of the expansive Vajrayana pantheon as verified by analogous standing sculptures of Tara from as far as Swat Valley in present-day Pakistan and Kurkihar near Gaya (Image 11).

The classical Gupta style persisted in lingering traces in Eastern India more than anywhere else for nearly three centuries after the fall of the Imperial Gupta dynasty till about the ninth century when Lama Taranatha announces the emergence of a new school. Dhiman and his son Bitopalo were the forerunners of the Pala style during Devapala and Dharmapala who for Taranatha also formed the regional schools of Magadha (Madhyadesa) and Vamgala (Pracyadesa) respectively. Frederick Asher in The Art of Eastern India, 300-800 tried to bring rupture to this classical understanding of artistic styles with dynastic affiliation through an elaborate survey of reliefs, sculptures and inscriptions, interrogating the bridge from Gupta to

Pala Art. (Asher, 1980) Broadly, Asher’ study covers the art and iconography from the presentday states of Bihar, (West) Bengal and Bangladesh, along with stylistic connections with Odisha and Assam. Eastern India for Asher could have a unified identity based on the stronghold of Buddhism, beginning with Kumargupta patronizing the establishment of Nalanda Mahavihara and the linguistic commonality of the Magadhi dialects of Apabhramsa, which enabled the migration of artists’ workshops. Taranatha himself with the access to previous sources speaks of an artist named Bimbasara under the ruler Buddhapaksa (perhaps identical with the Gupta ruler Buddhagupta) followers of whose school was known as that of Madhyadesa irrespective of their place of birth. Asher also brings in a significant discussion on the materiality that unifies Pala sculptures particularly in stone: the use of greyish black basalt stone of Rajmahal hills. Besides, architecturally the regional peculiarity of Eastern India was to build temples in brick or stucco and not in stone as in neighbouring Odisha. Magadha or Southern Bihar was an important cultural centre under the Guptas and

Bengal itself was divided further into four provinces: Samatata, Varendra or Pundravardhana, Gaud and Vanga. Pundravardhana was mentioned in the Allahabad Prasasti of Samudragupta, marking the ancient city of North Bengal within its political territory. In Magadha, the most important religious centres were the Nalanda-Rajgir area and Bodhgaya, the very historical site of Buddha’s Enlightenment; the Sarnath style having a strong bearing on both the centres. Rajgir especially was an important centre for the serpent

cult of Manasa attested by the exceptional brick temple and stucco statues of the sixth century Maniyar Math (Image 12) about 500 yards southeast to the Jain complex of Sonabhandar Caves carved in the fifth century. Asher argues that the local cult named after the Mani Naga was appropriated by the Saivites with the cylindrical pancayatana Linga structure. Also, Asher writes that the finely joined yellow bricks of the Maniyar Math temple was analogous to the ones used in the first phase of construction of the Nalanda Site under the Guptas, later to be replaced by coarser

reddish-brown bricks during the Pala phase. The stucco sculptures of Maniyar Math modelled on the Sarnath idiom and replacing the earlier Kusana Mathura style, post-dates the Gupta style and anticipates the seventh century Nalanda style. The foundations of the Nalanda Mahavihara in the village of Sariputra as mentioned by Huan Tsang seems to have been patronized by Kumaragupta and Buddhagupta in the late fifth century. However sculptural representations from the site as preserved in the museum date as late as the eighth century: sensuously carved stone sculptures of Avalokitesvara and Samantabhadra, which also mark the transition between late Mahayana and Tantrayana phases of [[Buddhist

Iconography]] (Image 13). Asher’s hypothesis of a migration of artists and their workshop-based styles from Sarnath via the pilgrimage site of Bodhgaya to Nalanda with the establishment of the Mahavihara finds yet another node in the town of Bhagalpur, close to the Vikramshila monastery later built by Devapala. The artistic activity of the Varendra region reached its climactic phase with the Somapura Mahavihara at Paharpur with the same ground plan and design as Vikramshila; the style of building and sculptural decoration of the Somapura Mahavihara was closely related to the seventh century Shalban Vihara at Maynamati, Comilla.

The highly acclaimed bronze statuette in the Birmingham Museum known as the Sultanganj Buddha (Image 10) after its findspot ignited a heated debate since the time of Rajendralal Mitra and Alexander Cunningham with both these archaeological historians claiming it from a Gupta workshop citing the similarity of the parallelly incised thin drapery lines. Douglas Barrett on the other hand credited it an early Pala provenance which however is refuted by Asher who places this very important piece of

metal sculpture in the seventh century CE. The seventh and the eighth centuries CE in eastern India were marked by the persistence of the old Gupta style (till the patronage of the last Magadha Gupta ruler, Jivitagupta II) until the emergence of the new Pala style in the ninth century (from when onwards we find an enormous corpus of Tantric Buddhist icons in the form of stone and metal sculptures and palmleaf folios).

This transitional phase between the dynastic imperial styles of the ‘classical’ Guptas and the ‘neo-classical’ Palas also saw the emergence of regional dynasties like the Khadgas, the Devas and the Candras with a diverse patronage to Brahmanical (both Saivite and Vaisnavite), Buddhist as well as Jain religious orders. The decentralized political authorities (samanta-feudalism) as well as the simultaneous co-existence of religious sects and orders are reflected in the Tantrayana pantheon which was as inclusive of the tribal cultic substratum as aspiring of the Brahmanical icons and the ideals they represented. Bodhgaya, the perennial Buddhist centre (with Theravada, Mahayana and Tantrayana centres like Vajrasana Vihara) with its close proximity to the ancient tirtha centre at Gaya shows this religious and hence iconographic hybridity in the Kurkihar hoard of magnificent bronzes, the earliest dated sculpture of which is that of a seventh century Balarama, found within the precincts of the Buddhist monastery at Kurkihar, 14 km from Gaya on the way to Rajgir-Nalanda dated to the ninth regnal year of Dharmapala. Asher postulates that by the time of Dharmapala there was considerable flux of Brahmanical iconography into the Tantrayana pantheon, which was

both preceded and succeeded by Tantric Buddhist textual sources which mention a host of Brahmanical deities, subjugated by the Buddhist faith. His study also likens the style of Singhbhum district to the style of Vaitala deula at Bhuvanesvara, and that of the Dinajpur district to the styles of ancient centres at Mahasthangarh and Tamralipti. In the field of metalsculptures of Eastern India particularly, Nihar Ranjan Ray pronounces the North Indian denominator against which the later Pala styles gained currency (Ray, 1986). The late eighthninth revamped patronage to

Buddhist plastic arts also coincides with the increase in importance of the feminine icons of Prajnaparamita and Tara, attested by large number of bronze statuettes in the Nalanda site museum, the National Museum in New Delhi and the Indian Museum in Kolkata. S.K. Saraswati in a monograph (Tantrayana Art: An Album, 1977) and a Bengali publication on the art of painting during the Palas (Saraswati, Pal Juger Chitrakala, 1978) elaborates on the technical aspects of the craft besides engaging with the regional variations of the style with a distinct preference to a shorter physiognomy, stout and fleshy with broad hips and shoulders, oval faces with broad elongated eyes: more immanent

and humane than transcendental and divine. Also, in the context of the regional variations of the Pala style, I would like to invoke the associative linkage of dialects within a language, the Pala visual language analogous to the vernacular Magadhi Apabhramsa (which also made its presence felt in the hybrid Sanskrit Buddhist Tantras), gaining currency at the moment of rupture of the Sanskrit Cosmopolis. However, on a parallel note it is crucial to recognize the “poetry of politics that gave presence to the Sanskrit

Cosmopolis” (Pollock, 1998). By ‘poetry of politics’, Pollock refers to the persistence of classical Sanskrit in the inscriptions to objects and monuments with royal or at least elite patronage. Just as in the literary domain, the localized dialects of

Apabhramsa had set up a dialogue with the Gaudiya and Pancali variations or reetis of classical Sanskrit poetry, the regional dialects of the Pala visual idiom get surfaced in the formalistic sense in relation to the classical Gupta ideal of Sarnath, which it seeks to displace and replace. Against Pollock’s conception of regional territorialisation of the vernacular, the next section tries to uncover the travels of the Tantric Buddhist siddhas and monastic centres far outside Eastern India, arguing for a pan-Indian geography of Tantrayana praxis as encoded in textual sources and visual iconography. However, many such centres ceased to be active post eight century CE, especially the caves of Western India; in many instances such

Tantric Buddhist centres later got appropriated by other religious cults and sects. Analogous to the spread of the literary Sanskrit cosmopolis from Bactria to Java, the spread of the visual idiom or style of Tantric Buddhism is subject to enquiry. The connection of schools of Tibetan arts to the Pala style with a Newari interlude in Nepal (of bronze-casting and painting) has been

pointed out by several scholars starting with Lama Taranatha’s hierarchical dictum that in the traditions of plastic arts (since after the Buddhist migrations from the tenth century onwards): first came the skilled artists from Bengal and Magadha, then the artists of Nepal and still lower in rungs: the artists of Tibet (Chattopadhyaya A. , 1980). The Buddhist art of Nepal from the eleventh to the

thirteenth century CE is modelled highly after the Pala idiom which in turn has the bearing of Sarnath Buddha; thus, in a way being twice removed from the classical Gupta ideal. Different schools of art can be seen emerging in Tibet from the thirteenth century onwards; the art activity in Tibet becomes prolific particularly with the patronage of the fifth Dalai Lama, who incidentally claimed both the spiritual and the political powers, gaining imperial authority over the plateau of Tibet. The imperial Lama wanted to establish his political authority and monopoly over the highly profitable trade of pashmina silk, engaging in aggressive military campaigns, curiously allied with the Mongol forces against the Buddhist kingdom of Ladakh, which had to seek help from the Sunni Mughal emperor Aurangzeb.

Again, in the South-eastern colonial outposts: Java, Indonesia, Sumatra and Cambodia late Mahayana and Vajrayana iconography thrive from the eighth to the tenth centuries, monastic missions being patronized by the rulers of Eastern India. For these South-East Asian outposts of Tantric Buddhism and likewise true for the spread of Mantrayana and Tantrayana to Tibet and China: small votive icons, ritualistic objects and manuscripts played a seminal role for dissemination of sutras such as the Karandavyuha Sutra (visual manifestations of which is evident in 10th century Buddhist Art of Cambodia (Woodward, 2007)), the

Mahavairocana Sutra and the Gandavyuha Sutra (both marking their presence in China from the seventh-eighth centuries CE, a manuscript of the latter being sent to the Tang emperor by Subhakardeva of the Bhaumakara Dynasty from Odisha). The hoards of Tantric Buddhist sculptures from Achyutarajpur near Bangarh in Odisha and from different sites in Southern Bangladesh, published as twin monographs By

Debala Mitra from the Agam Kala Prakashan, New Delhi also needs to be paid close attention to (discussed at length in Chapter III), in the context of dissemination of Vajrayana iconography to Northern or South-Eastern centres. The Achyutarajpur bronzes in particular are exceptional as the artist’s skill and conception realize the shift from the early post-Gupta style (8th century images

of Buddha and Avalokitesvara) to a much more dynamic mannerism in the 11th century images of wrathful gods and goddesses like Heruka, Bhrkuti and Vajrahumkara (Image 14). Though beyond the scope of this dissertation, the text-flow and iconographic syncretism in any of such Buddhist centres: China, Tibet, Nepal, Mongolia, Myanmar, Thailand, Java, Sumatra, Indonesia, Cambodia, Bali present distinct individual cases of hybridity of Tantric Buddhist discourses, mottled with indigenous cultic esoteric practices and beliefs (Taoism in China, Bon in Tibet so on and so forth), and often rivalled by Tantric Saivism, especially from Kashmir. From the late-tenth century onwards, the Chola suzerainty over the South-eastern colonies also suggest later Saivite appropriations of Tantrayana texts, iconography and sites. The Pala style named after the imperial dynastic patrons is thus to be located within such a complex and widespread Buddhist network of dissemination, across restricted boundaries of present-day nation-states.

10 Gupta Buddha from Sarnath, 5th century CE, sandstone, British Museum, 1880.6 Sultanganj Buddha, 7th-8th century CE, copper, Birmingham Museum, IS. 171-1947

11 Standing Tara, 9th century CE, Swat Valley, bronze, Metropolitan Museum of Art, 1984.484.2 Tara, 11th century CE, Kurkihar, bronze, Bihar Museum: Photograph by author

12 Maniyar Math, 6th century CE, brick architecture with stucco decoration Rajgir: Photograph by author

13 Samanthabhadra, 7th century CE, basalt, Nalanda site museum

Eleven-armed Avalokitesvara, 8th century CE, sandstone, Nalanda site museum: Photographs by author

14 Achyutarajapur bronzes: Vajrahumkara, Heruka and Bhrkuti (Left to Right), 10th- 11th century, Orissa State Museum, Bhuvaneswara Tantric Buddhist Art in India outside the Eastern Provinces (Pracyadesa)

This section is an attempt to argue for a pan-Indian geography of Tantric Buddhism, particularly in Western India and the continuum of the Gupta style till almost the twelfth century. This pan-Indian or national claim for Tantrayana contests the more conventional eastIndian regionality of the cultic ritualistic practices post 7th century CE. From the Konkan and Maharashtra to the Swat Valley and to Odisha and Assam, Tantric Buddhism through the travels of the mahasiddhas seems to have travelled the entire sub-continent from the 8th to the 12th centuries, often parallel to the rise and spread of the Natha order of Shaivism and their hatha yogic practices.

The Ajanta cave-complex with the earliest phase of Theravada activity began in the second-first century BCE with Cave 9, Cave 10, Cave 11 and Cave 12. Cave 10, one of the earliest and perhaps the biggest Caityagriha of the 1st century BCE bear stylistic similarities to other Western Indian Theravada cave sites like Bhaja, Pitalkhora and Kanheri (all these Buddhist cave complex activities beginning in the second-first centuries CE as Theravada or Hinayana centres and later refurbished with Mahayana and Tantrayana icons, painted and sculpted intrusively). While the much-discussed site of Ajanta (since its rediscovery in 1819 by John Smith, a British officer on shooting expedition) witnessed a very brief ‘proto-Tantric’ activity in its last intrusive phase in the late fifth century CE

before an abrupt end to the largescale artistic activity; the nearby site of Pitalkhora: an elaborate Theravada cave complex was refurbished in the Mantrayana phase as can be seen in the painted images of manushi buddhas on the pillars in Cave 3: an earlier caityagrha (Image 14). In the celebrated monastic complex of Ajanta realized under the principal patronage of the Vakatakas (vassals of the classical

Guptas) and their vassals: the Asmakas and the Risikas, manushi Buddhas or the mortal Buddhas become the visual trope for painterly refurbishing of earlier caves: Cave 9 and Cave 11, besides appearing extensively in the antarala (threshold to the inner sanctum) of Cave 2 as the visual manifestation of the Miracle at Sravasti. The painted image of Dipamkara Buddha on pillars in the same area of Cave 2, as well as the sculpted icons of Dipamkara on the façade of Cave 19 has been suggested by scholars to be a result of migration of

Gandharan artists at some point during the last phase of artistic activity. This last phase according to Walter Spink’s short chronology of the fifth century Mahayana reclamation of the site, centred around the Vakataka emperor Harisena (Spink, 2005), is marked by a rather hastiness in completion of the project. The unfinished caves: Cave 24 and Cave 21 in this last phase of activity (before the Samgha probably declared the site non-sacred or as Spink romanticised the death of the project with that of emperor Harisena) was contemporary to

the intrusive refurbishment of earlier caves through insertion of icons (Spink, 2007). These later icons for refurbishment and embellishment were mostly votive statues which suggest the popularity enjoyed by Ajanta for a brief while as a centre of Buddhist pilgrimage attracting merchants and artists from far and wide. That the complex of Ajanta is not mentioned in the Western Indian travels of the seventh century Chinese Hiuen Tsang hint that this site was abandoned by the Samgha.

Paintings of Ajanta with its inherent notion of classicism mostly take on the themes of the Jataka narratives, delighting in the sensuousness of the female body, particularly the female dancer as evident in Cave 1, identified as the Palace Scene perhaps from the Mahajanaka Jataka (Image 15) (and many other similar scenes in Cave 1, Cave 2, Cave 16 and Cave 17: the most elaborately painted caves.) The seductive enchanting dancers and courtesans, objects of disdain and refrain for the celibate

monks become the delightful subjects for artistic and aesthetic engagement. Cave 26, certainly one of the latest caves to be carved also takes on the sensuality of female dancers: Desire, Pleasure and Lust, the three daughters of Mara sent by the evil king to seductively withdraw Sakyamuni Buddha from his pursuit of Nirvana (Image 16) (Cohen, 2006, p. 117). The metaphor of the dancers for a modernist Orientalist like W.B. Yeats becomes the hybrid Buddhist notion (from Theravada to Mulasarvastivada to Zen with elements of Brahmanical Shakti) for hypothesising the ‘Unity of Beings’. Yeats, an avid

collector of plates of Ajanta reproduced by Heringham, Laurence Binyon and Mukul Dey, had a quasi-religious and quasi-romantic modernist idea of the erotically charged female dancers in a sacred monastic space (Kampily, 2017). Also, interestingly an Ajanta inscription of Buddhagupta, the chief minister of emperor Harisena deifies Sakyamuni Buddha as the bhagvana or God superior to Krsna and Siva. (Cohen, 2006) This ambivalence of the sacred and the profane is crystallised in the symbolism of erotic female dancers in Ajanta, a trope that transforms quite radically in what one can term the early phase of Tantrayana in the caves of Aurangabad, in the same

district in Maharashtra. ‘Taming of Nalagiri’ relief of Cave 26 with the seductive enchantresses: the three daughters of Mara are subordinated by the earth-touching gesture or bhumisparsha mudra of Sakyamuni in a baroque mannerist carving with deeper undercuts, suggesting the emerging out of sculpted icons from the volcanic rock surface. The upper section of Cave 6 has been conjectured based on the iconographical analysis to to be a nunnery, a significant later Mahayana institution within the patriarchal Samgha. Also, the Litany scene of Avalokitsvara or Astamahabhaya Avalokitesvara

showing the Mahayana male deity dispelling eight fears of a traveller as seen in Cave 4 (in the last intrusive phase of “arrival of the uninvited”) becomes a popular icon in Western Indian cave sites from the fifth century CE (Image 17). The Litany scene carved analogously at Aurangabad, Kanheri and Ellora makes way for the female counterpart by late seventh century CE possessing the same power of dispelling the eight fears: the icon of Astamahabhayatarini Tara.

The activity at Aurangabad Caves like Ajanta and many other Western Indian Buddhist cave complexes began in the first-second centuries BCE and was revived in the late Mahayana phase in the late fifth century CE. The newly carved caves of Cave 2, Cave 5 and additions to Cave 1, Cave 3, Cave 4a were patronized by the regional powers claiming power during and soon after the Vakatakas were exiting the volatile medieval political arena. Cave 3, for Pia

Brancaccio with its “devotees and exuberant style, is a testament to the rising power of the periphery versus the centre” (Brancaccio, 2011). The later additions, particularly the votive reliefs of the iconic bodhisattvas of Tantrayana: both male and female and the dancing Vidya in Cave 7 mark a watershed moment in the esoteric development of Buddhism in Western India and a peripheral encounter with the Saiva Pasupatas attested by the Saptamatrika relief with Ganesa (also evident in Ellora and Panhale Kaji).

While in Cave 2 and Cave 6 at Aurangabad, the mature Mantrayana iconography of Avalokitesvara Maitreya and Vajrapani represent the transitional phase, Cave 7 stands as the climactic fulfilment of iconographic complexity with sculptures of Avalokitesvaras and Buddhas with female attendants, female Buddhist goddesses (in the antarala) and porches in the mandapa dedicated to Pancika, Hariti on one end and to six female figures holding lotus flowers, bracketed by a Buddha in varada mudra and a standing Avalokitesvara besides representations of Gajalakshmi (Images 18-21). The Litany scene or Astamahabhaya Avalokitesvara is an

imposing image from this cave with the effigy of Amitabha in a typical hairdo of dreadlocks or jatamukuta. Cave 9 has a four-armed image of Avalokitesvara (partially damaged) below the Parinirvana relief of Buddha (which has a similarity to the same theme Cave 26 Ajanta) also bearing an effigy of Amitabha pronounce the Vajrayanist formation of the concept of Buddha families or kulas with the five Dhyani Buddhas as the spiritual ancestors of the lineages (Image 22). The diversification of the Buddhist pantheon goes in hand with the rise of the samanta model in administrative politics. Brancaccio, following his mentor Spink, postulates the ‘The Aurangabad Renaissance in the Fifth Century’ to be patronized by the Asmakas, the feudatories of Harisena. (Brancaccio, 2011)

The Aurangabad sculptures likewise outdo the Gupta-Vakataka model, breaking new grounds for improvisations and iconological interpretations contained in the new class of Buddhist literature: the hybrid Sanskrit sutras like Saddharmapundarika Sutra and Sarvatathagatatattvasamgraha embodying complex methods of visualization. One such carving on the walls of Cave 2 is the delineation of a squatting female

alike to the Goddess Lajjagauri. At Aurangabad thus, there is also the strong resurgence of the problematic of the historical agency of women in Tantric Buddhist praxis. Also, the iconic representation of Vajrapani (as exemplified Ajanta Cave 1 painting) persisted in Cave 2 and Cave 6 of Aurangabad which posit the shift to Vajrayana from late Mahayana or Mantrayana. The

sculptures of female goddesses in the antarala or vestibule of Cave 7 identified as manifestations of Goddess Tara finds a textual parallel in a seventh century text: Tarakalpa, closely associated to the more widely known Manjusrimulakalpa, which mention Tara as a dharini goddess who can cast a spell. Another dwarfish stout female goddess identified as Ekajata, like Tara assume independent status and verses dedicated to her in the ninth century Sadhanamala. Basic iconohraphical features of the Goddesseses Mahamayuri and Marici are also introduced in Cave 7 at Aurangabad as a part of the retinue of Sita Tara; all these

iconographical details find their codified form available to us as the manuscript of Sadhanamala. The sculptural representation of the female dancing goddess flanked by female musicians in the same cave, furthers the urge towards experimentation and innovation in visualization techniques. John C. Huntington proposes through analysis of iconographic programming that Cave 6 was meant for the realization of the Mahakarunagarbhadhatu mandala mentioned in sources like the Mahavairocana Sutra and the

Sarvatathagatatattvasamgraha. Huntington postulates a connection of Aurangabad to the north-western spread of Buddhism along the Silk Route. While Subhakardeva’s (from Odisha) embassy with a copy of the Karandavyuha Sutra was received with great pleasure by the Tang emperor in the seventh century, the mandalas of Dunhuang (or closer home: those in the Tabo Monastery in Spiti Valley or complexes of Alchi-Mangyu-Sumda in Ladakh) and those is the Shingon sect of Buddhism in Japan has perhaps its earliest archaeological prototype in Western India, especially in Aurangabad: Cave 6 and Cave 7 from sixth century CE (Huntington, 1981).

The diffusion of Mahayana at Kanheri, Brancaccio argues to be a coastal phenomenon of connection between the Konkan and the North-western provinces of Sindh and Gandhara (Brancaccio, 2011, pp. 120-123). The twin colossal statues of Buddha carved outside Cave 3, standing in varada mudra are closer to the Gandharan typology rather than the Sarnath one (Image); also discernible is the analogy to colossal Buddhas in Gandhara, for instance the site of Bamiyan. Kanheri or Krsnagiri as the cave complex is known from ancient and medieval sources, remained an exclusively Buddhist site with inscription mentiong the Satavahana ruler, Vashisthaputra Satakarni from the second century CE (Cave 5) to Silahara inscriptions as late as 12th century CE. A feudatory of the Rashtrakuta king Amoghavarsha I (817-878 CE), Pullasakti’s inscription mentions a grant donated towards the monastic cave-complex, a part of which was spent in buying books

(Gokhale, 1987). Another inscription dated to 854 CE mentions the permanent endowment of 100 drammas for construction on meditation cells for monks living in Maharaja Mahavihara (Durbar Hall or Cave 11) by a donor from Gauda (in Eastern India) during the reign of Kapardin II. Kanheri, thus with more than a millennium of Buddhist monastic activity patronized by the maritime traders saw all the stages of development of Buddhism: from the huge early Theravada complex of caityas and viharas to later Mahayana votive appropriations from the 5th century CE (Cave 2, Cave 3) to what we may say the earliest and only representation of Eleven-Headed Avalokitesvara, a Mantrayana deity of great importance on the right niche to the entrance of Cave 41 (Image 23). This image as Shobhana Gokhale elaborates on in article is “…the singular sculpture of its kind which shows a superb illustration of the imagination of learned Buddhist scholars and the exquisite skill (and visualization: my stress) of the then artists.” (Gokhale, 1987, pp. 371-376)

Gokhale, starting from precise scientific measurements of the Eleven-headed Avalokitesvara image datable to fifth century (height: 145 cm, breadth: 37 cm, height of eleven heads: 32 cm, breadth of eleven heads: 21 cm and each individual head: 8x8 cm), moves on to complex issues of iconology explaining the curious iconography for the first and last time in Western Indian Buddhist Caves. In the left hands, Avalokitesvara holds a bottle (replacement for kamanadalu) and a lotus bud with long stalk and the right hands in Abhaya mudra and holding a rosary. Gokhale’s premise of study is the cultural environment of Kanheri:

inscriptional records, relation to Saivite caves of Elephanta, Jogesvari, Ellora and Parel and moreover cultural relations with Central Asia and China. Mentions of Avalokitesvara as the spiritual son of Amitabha and as Eleven-headed or Ekadasasirsa are found in the Sukhavativyuha Sutra and the Gunakarandavyuha Sutra. Gokhale wonders about the possible relation to the Ekadasa Rudra, the eleven violent Vedic Gods before arriving at a very interesting discussion on the concept of Dasabali Buddha, earliest reference of construction of caityas dedicated to whom can be found in Nagarjunakonda. Mention of Dasabali is found also in Devanimori stone casket inscription mentioning charity in the name of Dasabali (bringing into fold the idea of punya or merit in the Mahayanist context) and from an inscription dated to 500 CE from Punjab, West Pakistan. To quote Gokhale at length:

The concept of building Stupa in the name of Dasabali Buddha travelled along the trade route but the eleven-headed image of Kanheri with Dasabali in the headdress is a masterpiece of Buddhist iconography. Highly intellectualised spiritual experience of Dasabali Buddha is beautifully transformed into sculpture.

The unique image of Eleven-Headed Avalokitesvara is interestingly located in the western corner of the complex (in the extreme western naval outpost of Indian subcontinent): the western cardinal paradise of Sukhavati Vyua, the paradise of the Dhyani Buddha, Amitabha. Colossal image of Eleven-Headed Avalokitesvara or Kuan-yin from late seventh century CE is found from the Tang court of China and strengthens the argument for trade relations between Western India and China. The ten surmounting faces perhaps symbolizing the ten bhumis or territories in the Kanheri image is arranged in four tiers: Sakyamuni’s face on the top, and then three frontal faces depicting the Bodhisattvas, three angry faces on the left and three on right with canine teeth upwards. In extension to this

terrorization of the Mandala, it is also important to make note of the only painted fragment of a mural: the only extant figure of Aksobhya at the centre of the five Dhyani Buddhas, located in the eastern corner of the cave complex in the ceiling of Cave 34 (Image 24). The style of painting as evident from this fragmentary mural, culls from the post-Ajanta iconic tradition with the artist emphasizing on linear delineation of the iconic deities and is closer to the Pala-style of painting in palm-leaf folios from the tenth century onwards. The Cave 34 ceiling painting of Aksobhya, thus seeks a date of circa 8th century CE on stylistic grounds, perhaps executed from the donations made by Amoghavarsha’s feudatory or the grant from the Brahmin from Gauda.

The iconic visualization of Mahayana deities began in Kanheri, like Ajanta and Aurangabad with the Litany Scene or Astamahabhaya Avalokitesvara, intrusively carved in Cave 2 and more programmatically in Cave 90. In Cave 67, the Litany is carved exactly opposite to a colossal image of Avalokitesvara with Amitabha in the effigy flanked by female divinities: Tara and

Bhrkuti. In Cave 90, the extensively carved interiors with iconic variations of Buddha images fits into the transitional phase between proto-Tantric understandings of the Miracle at Sravasti and later codified mandala arrangements of the Mahavairocana Sutra, painted murals of which can be seen at Mangyu, Ladakh (painted originally in the tenth century CE with later repairs and re-painting). Kanheri or Krsnagiri, as a flourishing centre of

Tantrayana is also attested by the biography of the Bengali monk later known as Atisa Dipamkara. Atisa, born as Candragarbha, the son of the Eastern-Indian Candra king from Samatata, received his first Tantric initiation at Krsnagiri or Kanheri under guru Rahulagupta getting his name Guhyajnanavajra (Chattopadhyaya A. , 1967). Atisa, however could not reach the final stages of Tantric initiation, unable to participate in the sexo-yogic practices in a ganacakra or magical ritualistic feast and later received his Mahayana initiation in Sumatra from Sri Dharmakirti, getting his Mahayanist name: Dipamkara Srijnana. In a Newari palmleaf manuscript of Illustrated Astasahasrika Prajnaparamita in the Cambridge Library Collection, one finds mention of the monastic complex of Krsnagiri accompanied by a painted image of caityagrha, which scholars tend to identify as a caitya from within the cave complex with more than 100

rock-cut caves: carved, embellished and inhabited for over a millennium from the third century BCE (with Asokan rock-edicts found at Sopara, close to Kanheri) to the thirteenth century CE (with the last Silahara king Somesvara losing power of the region to the Yadav prime Mahadeo). Also, a unique wooden image of the Buddhist Goddess Tara found from Kanheri (Cave 31 by Moreshwar Dikshit) published in an article in Marg

(present location: unknown) attests the importance of this coastal site of Vajrayana and stylistically appears to be an import from Eastern India in the ninth or the tenth century CE. (Sankalia, 1984) The wooden image of Tara (Image 25) found from Kanheri is both formally and stylistically strikingly similar to images of Lokanatha and Sthriacakra at Bangladesh National Museum, Dhaka (Gupta, 2016). Caves 84-87 considered the Kanheri Necropolis with a number of brick stupas and terracotta sealings with characters of the Buddhist creed from circa tenth century is believed to be the burial site or cemetery of the large cave-complex, which probably played significance role in esoteric rituals of Tantrayana.

The centre-periphery political dialogue becomes the most evident in the Buddhist caves at Ellora, excavation of which possibly started with the Kalacuris or the Calukyans of Vatapi (Badami) and later taken over by the Rashtrakutas. Ellora’s uniqueness lies in being a site where sects and orders of all institutionalized religions meet and interact: Vaisnavism, Saivism, Buddhism and Jainism. The occult or esoteric aspects of Buddhism, which goes on to form the tenets of Vajrayana can be seen reaching a mature stage in Ellora with R.S. Gupte’s scholarship rectifying the ignorance of earlier scholars on Tantric Buddhist iconography. Gupte enumerates a large number of Avalokitesvara images from Ellora with interesting forms like Rakatananda, Sadakshari and the Litany, number of Manjusri images with forms like Siddhakavira and Sthiracakra, 8 images of Jambhala: the Buddhist god of wealth, 27 images of Cunda (Accompanying Buddha or Bodhisattva images), 12 sculptures of Bhrkuti and 25 of Tara in the Buddhist caves enumerated from 1 to 12 in the eastern corner of the complex (Gupte, 1964,

p. 7) This full-flowering of Mantrayana and Tantrayana iconography at Ellora needs to be seen in continuation of the eclectic developments at Aurangabad and Kanheri in the preceding centuries and to a lesser extent: the ‘classical’ tradition of Ajanta (with marital alliances between the Vakatakas and the imperial Guptas). Also, Gupte mentions the Crowned Buddha icon at Ellora (Cave 11): one of its type in the

Western Indian Buddhist caves, an icon which becomes extremely popular for the Pala artist of the Eastern Provinces from the ninth century onwards. The artistic activity (sculptural iconography as well as architectural elements) at Ellora also needs to be seen as a hybrid product of the transactions between Mantrayana Buddhism and Saivism; the three floored

monastery and chapel hosting a large number of iconic deities and mandalas (Cave 12: for instance a sculpted icon of Vajrasattva in the antechamber on the second floor) paving the way to Dasavatara Cave (Cave 15) and the celebrated colossal monolithic temple of Kailashnatha (Cave 16) bearing the seal of imperial Rashtrakuta patronage of Dantidurga. Cave 15 with images of Buddha on the top part of the wall suggest that the cave started being excavated as a Buddhist shrine but was later converted into a Saivite structure with the introduction of the Nandimandapa architecturally and through sculpted icons of Brahmanical deities giving its name: Dasavatara.

The three-storeyed rock-cut viharas of Ellora (Cave 11 and Cave 12) are the richest depositories of Tantrayana iconography in Western India. While the two-storeyed Cave 11 has icons of Avalokitesvara (with Amitabha in the hair-dress), Samantabhadra, Manjusri, Jnanaketu, Vajrapani, Cunda and Tara (accompanying Avalokitesvara or Buddha images); the three storeyed Cave 12, excavated in the eighth century has elaborate Mandalas or magic circles: for instance in the ground floor with nine icons of Rakta-Lokesvara,

Sthiracakra, Maitreya, Budddha, Padmapani with Amitabha in effigy, Vajrapani, Jnanaketu, Lokanatha and Manjusri. Cave 12 has representations of the five Dhyani Buddhas and arguably the Dhyani Buddha Saktis: twelve female figures carved in the antechamber to the shrine on the third storey. Extensive representation of Female Goddesses: Vasudhara (Cave 12), Mahamayuri (3 images in Caves 6, 10 and 12), Janguli (Cave 12), Cunda (27 images: Caves 10, 11 and 12), Tara (25 images: Caves 2, 4, 6, 9, 10, 11 and 12), Bhrkuti (Caves 4, 5, 8, 10, 11 and 12), Sarasvati (3 images: Caves 10 and 12) as well as Hariti and Pancika (on the façade of Cave 8) are telling of the sculptural realization of the Yogacara ideal of Mahasukhavada in Ellora (Gupte, 1964). (Images 26-28)

Though Gupte’s thorough iconographic research and attempts of tracing Vajrayana icons to the textual sources of Sadhanamala and the Nispannoyogavali, revises and rectifies the earlier scholars, he still holds the belief that the ritualistic art of Vajrayana was the phase of decline and deterioration of the ‘original’ Buddhist ideal. Against this general contempt of scholars for Tantric Buddhism in

India (especially in Western India), this research aspires to subvert this assumption and re-position Tantric Buddhism in India as an advanced mature phase of nervous excitement and hybrid experimentation, proliferating through the esoteric lineage of the siddhas. Also, through the praxis of the siddhas, it becomes significant to throw some light on Kalacakrayana and

Sahajayana, two distinct trajectories of the last phase of development of Tantrayana in India. While Kalacakrayana rituals were primarily based on visualization techniques deriving from textual sources in Sanskrit (and later Tibetan translations); Sahajayana thrived on oral performativity and can be traced through the living performative traditions of the Sahajiya sects and cults in Bengal with various religious affiliations. Sahajayana also becomes a significant moment of the return of Buddhist discourses to orality and a pronounced emphasis on kaya-sadhana or the rigorous sexo-yogic practices to train the body and the mind. Tantric Buddhist literature and art, emerging from the performative discourses of the mahasiddhas through the carya songs, further evoke the

problematics of gender and agency, elite and vernacular aesthetics and ethics. Alternatively, the form of Kalacakra that a sadhaka should would meditate upon to attain a siddhi would be prescribed as, to quote Biswanath Banerjee:

In the variegated mandala, on the great lotus, with the Sun above, and shedding lustre, is seated with the left leg stretched (the Lord) with the bhutanatha trampled under the feet, causing horror in the minds of gods, demons, and snakes, and who is jnanasattva and located in the smasana, should be meditated upon for a month and (thus) ghosts should be controlled. (Banerjee B. , 2005)

Panhale Kaji: a cave-complex in Southern Konkan (in the present state of Maharashtra but close to the border with Goa) known in the medieval times by the name of ‘Pranalaka’ is a site of singular importance in the area. The caves are marked with a distinct iconography of an advanced phase of Tantrayana till 12th century CE and still later appropriation of the site by the Natha siddhas (Deshpande, 1986). The Tantric Buddhist siddhas (Kanhupa, his guru: Jalandharipa, along with Matsyendranath, Gorakshanatha and Cauranginatha appearing in the Buddhist list of eighty-four as well the original nine Nathas) with their prominent

centres in Eastern India travelled to as far as Andhra-desa and Konkan, spreading the discourses of Sahajayana and Kalacakrayana. At Panhale Kaji, images of Dhyani Buddha Aksobhya appear at two cave shrines (Cave 6 and Cave 8) besides a number of loose images found from the site. Another loose image of Siddhakavira from the site has Aksobhya in his effigy. Image of Candamaharosana found in

Cave 10 is carved on a stupa which is telling of its connection to Ratnagiri in Odisha (Image 30). Candamaharosana Cave 29 has the image of Tripurasundari surrounded by the eighty-four siddhas: an image which is syncretic of the cult of Dattatreya and that of the Nathas along with the Tantric Buddhists (Image 29). Cave 14 also has the carved relief of Kaulajnana, an advanced initiation stage of the esoteric ritualistic practice of Tantric sects and orders. Further south, in Tulu Nadu near Mangalore the temple to Manjunatha has been speculated to be preceded by a centre dedicated to the Buddhist Lokanatha. Or further West in Gujarat at the Jain centre of Taranga near Mehsana, which has the icons of Tara popularly known as Taran Mata and a male icon worshipped, dressed as a female goddess Dharan Mata in cave-shrines behind the Jain pilgrimage complex. In Andhra and Telengana, besides the ancient sites of Amravati and Nagarjunakonda:

where Madhyamaka Buddhist philosophy flourished with the Satavahana and the Ikshvaku patronages, a number of Buddhist sites witnessed the later Vajrayana phase such as Bavikonda, Salihundam, Bojjanakonda and Lingalametta. Further south, female deities appear in a celebrated Buddhist Tamil Epic: Manimekalai written by Cattanar (c. 450-550 CE), named after the protagonist who evolves from the heroine to the female Bodhisattva. The author, essentially a Buddhist of the Sautantrika-Yogacara school through the fascinating epic narrative attempted to disseminate vital principles of Yogacara and aspects of Tantric Buddhism in the south.

(Kandaswamy, 2005) Manimekalai tevyam, the rescuer of people from shipwrecks in the myths is a possible allegory to denote salvation from the sea of birth, a miraculous power of the Buddhist Goddess Tara whose cult grew in popularity steadily from the sixth century CE. In South India, the staunchly Saivite Chola kings (and later the Brahmanical Hoysalas and Kakatiyas) may be held responsible for the overpowering and appropriation of the Buddhist monastic centres by the Saivite orders. The bronze-casting workshop at Nagapattinam, however produces some excellent bronzes of calm and peaceful Buddhas and Bodhisattvas, predating the terrific yet elegant dance of Nataraja.

In the north-western frontiers of the Tantric Buddhist Cosmopolis, Tantric Buddhism flourished in Oddiyana or Uddiyana located in the Swat Valey, mentioned in Tibetan texts as Orgyan or Urgyan. Oddiyana is particularly known in the Buddhist Tantras as one of the prime esoteric centres of dakini-vidya for the ritual-feast or ganacakra and also the very site of revelation of many of the Tantras. Oddiyana

with its obscure geographic locale has been memorized in the name of Oddiyana Marici (a wrathful Tantric Buddhist Goddess) and Uddiyanakrama, an advanced stage in Yogini-tantras of Buddhism. While the Kalacakra Tantra suggests Oddiyana to be the birth-place of Guru Padmasambhava, a Tibetan scholar Buddhagupta from the sixteenth century calls it the city of Mahmud Ghazni, the Islamic destroyer of Buddhism in the Swat-Valley. (Tucci, 1940) The Italian translator of the travels of Urgyan-pa (named so for visiting the sacred site of Buddhist Tantra sometime in the first quarter of the thirteenth century) delineates the extensive travels of this Tibetan monk from

Nalanda and Vikramshila (in the last days of existence after Bakhtiyar Khilji’s invasion) to Northern India: Bodhgaya, Sarnath to Punjab: Jalandhara (another sacred pitha mentioned in the Tantras and from where Kanhu’s guru Jalandharipa hailed from) to Chamba to Kashmir to the Mongolian lands and finally to Urgyan, somewhere in the Swat Valley, recently captured by the Turks, before turning back. Tucci found the manuscript of the travelogue of the Tibetan pilgrim at the collection of the Hemis Monastic Sanctuary in 1930 and subsequently his biography at Spiti. (Tucci, 1940, p. 10)

Urgyanpa seems to recognize the importance of the sacred pitha at Uddiyana, frequented by flesh-eating and blood-drinking dakinis, tamed under the spell of Adi-Buddha by the siddhas: Kambalapa and Indrabhuti (Indrabhuti and her sister mentioned in the siddha list to hail from Uddiyana and been initiated by Kambalapa). Despite the Turko-Mongol aggressions, Urgyanpa witnessed obscure traces of occult

Tantric practices still thriving in the forests and cemeteries of the north-western frontiers of Tantric Buddhism. The founder saint of the Hemis monastery was also acclaimed to have travelled the same itinerary in the sixteenth century with little success to trace down the sacred places after three centuries of Islamic activity. Urgyanpa, back in the thirteenth century

however experienced the enchanting songs and dance of the dakinis in cryptic hybrid sandhyo-bhasha. Also, the lingual Apabhramsa Cosmopolis (intertwined with the Tantric Buddhist Cosmopolis) is evident in the dialogue of the Tibetan pilgrim and two Brahmins who joined the caravan at some point of travel in Central Asia: the cross-roads of civilizations. Both the Tibetan monk and the Brahmins speak in a variation of Apabhramsa which also bears early indexes of Persianization (or the Persian Cosmopolis) in utterances like ‘Hami bhotanti dsogi huva’ (I am a yogi from Tibet) and ‘Tumi abo eham bhesa, roti vela khahi ki nahi? (Come, sit here; you eat bread or not?) (Tucci, 1940, p. 38) The Italian translator of the Tibetan text however mis-interprets the dialogue in local

dialect as later Hindi interpolations; however, Suniti Kumar Chatterjee rightly suggested that this was distinctly an archaic medieval dialect which was essentially hybrid in nature. In his History of Buddhism in India, in the section on Image-makers Lama Taranatha likewise mentions the Kashmiri style of Buddhist Art which was known widely coming from the northwestern provinces. The Kashmiri bronzes of Buddhas and Avalokitesvaras, treasured in the Hemis Museum, dated from the seventh to the twelfth centuries, speak volumes about the correctness of Lama Taranatha’s account (which unfortunately could not be photographed).

In the collaborative skill and imagination between Kashmiri and Central-Asian artists and probably artists from Eastern part of India as well, the Buddhist mahasiddhas with their significant female partners (and thus shareholders of Mahasukha) come to life at Alchi in the painted dhoti of the colossal terracotta image of Manjusri in the three-storeyed Sumtseg (Images 31 and 32) (Alchi: Ladakh's Hidden Buddhist Sanctuary, 1996).

  
15 Manushi Buddhas, Cave 3, Pitalkhora, Post-Ajanta paintings
  
  
16 Mahajanaka Jataka: renunciation scene, Cave 1, Ajanta
 
17 Dance of Mara's Daughters, Cave 26, Ajanta
  
18 Litany Scene or Astamahabhaya Avalokitesvara, Cave 4, Ajanta
 
  
19 Female Deities, Cave 7, Aurangabad
 
20 Avalokitesvara and Tara, Cave 7, Aurangabad
 
 
  
21 Main shrine, Cave 7, Aurangabad
 
  
22 Dance of Vidya, Inner Shrine, Cave 7, Aurangabad
 
  
23 Four-armed Avalokitesvara, Tara with Attendant female deities, Cave 9, Aurangabad
  
24 Eleven-headed Avalokitesvara, Cave 41, Kanheri
 
  
25 Dhyani Buddha Aksobhya, Cave 37, Kanheri
  
26 Wooden Image of Tara found from Kanheri, published in Marg Vol 36, No.1 (1984) and lost since
 
  
27 Teen-tala vihara (Cave 12), Ellora
  
28 Avalokitesvara with female attendants and Amitabha at the top, Cave 12, Ellora
 
  
29 Avalokitesvara with Tara and Bhrkuti and other Gods of the Tantrayana pantheon, Cave 12, Ellora
  
30 Tripurasundari with Siddhas, Cave 29, Panhale Kaji
 
  
31 Candamaharosana, Cave 12, Panhale Kaji
 
  
32 Manjusri Shrine, Sumtseg, Alchi, eleventh century CE
 
  
33 Mahasiddhas, details of Manjusri's dhoti, Alchi


Bibliography
 
1. Asher, F. M. (1980). The Art of Eastern India, 300-800. New Delhi: Oxford University Press by arrangement with the University of Minnesota Press.
2. Banerjee, B. (2005). The Kalacakra School: The Latest Phase of Buddhism. In N. Bhattacharyya, Tantric Buddhism (pp. 263-267). New Delhi: Manohar.
3. Banerjee, R. (1915). Palas of Bengal. Kolkata: Asiatic Society of Bengal.
4. Brancaccio, P. (2011). The Buddhist Caves at Aurangabad: Transformations in Art and Religion. Leiden: Brill.
5. Chattopadhyaya, A. (1967). Atisa and Tibet. Calcutta: Firma K.L Mukhopadhyay.
6. Chattopadhyaya, A. (1980). Lama Taranath's Hisory of Buddhism. Kolkata: KP Bagchi and Company.
7. Chattopadhyaya, D. (1959). Lokayata: A Study on Indian Materialism. Calcutta: People's Publishing House.
8. Chattopadhyaya, D. (1960). Post-script to the Introduction. In A. Chattopadhyaya, Caturasi Siddher Katha (pp. 37-55). Kolkata: Papyus.
9. Chowdhury, A. (1967). Dynastic History of Bengal. Dacca: Asiatic Society of Pakistan.
10. Cohen, R. S. (2006). Beyond Enlightenment: Buddhism, Religion and Modernity. New York: Routledge.
11. Davidson, R. M. (2002). Indian Esoteric Buddhism- A Social History of the Tantric Movement. New York: Columbia University Press.
12. Deshpande, M. (1986). The Caves of Panhale-Kaji. New Delhi: Archaeological Survey of India.
13. Ganguly, D. (1994). Ancient India: History and Archaeology. New Delhi: Abhinav.
14. Gokhale, S. (1987). Eleven-Headed Avalokitesvara from Kanheri. Annals of the Bhandarkar Oriental Research Institute, 371-376.
15. Guha-Thakurta, T. (2004). Monuments, Objects and Histories. New York: Columbia University Press.
16. Gupta, S. S. (2016). Wooden Idols of India: The Antiquity of a traditional experience. Chitrolekha International Magazine on Art and Design, Vol.6, No.1 , 76-82.
17. Gupte, R. (1964). The Iconography of the Buddhist Sculptures (Caves) of Ellora. Aurangabad: Shri M.B. Chitnis, Registrar, Marathwada University.
18. Huntington, J. C. (1981). Cave 6 at Aurangabad: A Tantrayana Monument? In J. G.
Williams, Kaladarsana: American Studies in the Art of India (pp. 47-55). New Delhi: Oxford.
19. Kampily, M. (2017). The Ajanta Mountain Caves Hindu/Buddhist Tantric Dancers and W.B. Yeats' Later Eastern focused concept of Unity of Being. Yeats Journal of Korea, Vol. 52, 143-183.
20. Kandaswamy, S. (2005). Female Deites in a Buddhist Tamil Epic. In N. Bhattacharyya, Tantric Buddhism (pp. 268-276). New Delhi: Manohar.
21. Majumdar, R. (1971). History of Ancient Bengal. Kolkata: G. Bharadwaj and co.
22. Pollock, S. (1998). The Cosmopolitan Vernacular. The Journal of Asian Studies, Vol 57, No 1, 6-37.
23. Poncar), R. G. (1996). Alchi: Ladakh's Hidden Buddhist Sanctuary. London : Serindia
.
24. Ray, N. R. (1986). Part I: History of Art Form- On the Common North Indian Denominator. In K. K. Nihar Ranjan Ray, Eastern Indian Bronzes (pp. 17-24). New Delhi: Lalit Kala Academy.
25. Sankalia, H. (1984). A Unique Image of Goddess Tara from the Kanheri Caves. Marg, Vol 36, No.1, 84.
26. Saraswati, S. (1977). Tantrayana Art: An Album. Kolkata: The Asiatic Society.
27. Saraswati, S. (1978). Pal Juger Chitrakala. Kolkata: Ananda Publishers.
28. Shaw, M. (1995). Passionate Enlightenment: Women in Tantric Budhhism. New Jersey: Princeton University Press.
29. Sinha, B. (1954). The Decline of the Kingdom of Magadha. Patna: Motilal and Banarsidass.
30. Spink, W. (2005). Ajanta: History and Development; Volume I: The end of the Golden Age. Leiden: Brill.
31. Spink, W. (2007). Ajanta: History and Development; Volume V: Cave by Cave. Leiden: Brill.
32. Tucci, G. (1940). Travels of Tibetan Pilgrims in the Swat Valley. Calcutta: The Greater India Society.
33. Williams, J. G. (1982). The Art of Gupta India: Empire and Province. New Jersey: Princeton University Press.
34. Woodward, H. (2007). The Karandavyuha Sutra and Buddhist Art in 10th century Cambodia. In e. b. Pala, Buddhist Art: Form and Meaning (pp. 70-83). Mumbai: Marg Publications.