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


The mystery of the face towers

From Tibetan Buddhist Encyclopedia
Jump to navigation Jump to search
1375 584 n.jpg




This is a pre-print version of a chapter published in Bayon, New Perspectives (2007:230-81) River Books, Bangkok ed. Joyce Clark; please refer to the final published version if quoting.


The mystery of the face towers* by Peter D. Sharrock

The towers built by Jayavarman VII with colossal faces staring out in four directions over Angkor are the most singularly striking architectural feature of the ancient capital of Cambodia. And yet, despite many years of scholarly attention, they remain the greatest enigma of the vast complex of temples. This paper considers most of the earlier interpretations, reconsiders the material facts of the temple art in the context of Jayavarman’s Angkor and proposes a solution to the enigma.

At the heart of the mystery is the identity of the god in the face towers, but the enigma also encompasses the fundamental idea, the ‘intellectual “model”’, as Bruno Dagens puts it, that the architects and sculptors of the Bàyon were projecting. What statement is made by the first state-level Khmer Buddhist temple? What Buddhism does it celebrate? What is its political context and purpose? Why are there so many towers and why are they raised in this unique formation? Many scholars see the Bàyon as Jayavarman’s pantheon of the gods, past and present, of ancient Cambodia. Our task is to determine how the layers of hierarchy in the pantheon were structured, to look for the overall religious/political strategy which informs this structure, and to identify its principal divinity.

Most guide books interpret the Bàyon as bearing giant images of the Mahāyānist Bodhisattva Lokeśvara ‘facing all ways’, carved in the likeness of the king – an expression of royal-divine power streaming out to connect with a network of local territorial deities. I retain the idea of a network of territorial deities but question both the identification of the central, linking god as Lokeśvara and whether the god is carved in the likeness of the king. I will propose instead that the image of the sixth supreme Tantric Buddha, Vajrasattva is what is carved with maximum prominence on the regnal temple. And that this image is the ultimate symbol of a programme to supplant Śaivism as the religion of state and to strengthen Buddhism in Cambodia at the moment it faced destruction by Islam in northern India.


Giant faces

It should be said from the outset that the Bàyon enigma arises partly from the fact that the Bàyon’s giant faces are, iconographically, recalcitrant objects: their iconography is minimal and non-specific. As a consequence, the art historians have not reached a consensus as to their identity – shifting, over many decades, from Brahmā, to Śiva, to the Buddha, to the Bodhisattva Lokeśvara and more recently to the Buddhist Brahma.

The Bàyon itself has long been the most disconcerting monument for Khmerologists. It was first thought to be a ninth century Śaiva foundation; then thought to be 11th century, until finally discovered to be Buddhist and placed in the 12th century. The god of the giant faces changed in line with these shifts. George Coedès, who played a key role in this slow evolving archaeology, shows signs of the frustration and embarrassment of the long interpretive struggle in his final words on the faces, as he ends up shifting the blame from the indecision of the interpreters to the indeterminacy of the object:

I can already hear our critics crying out: ‘Just look at those archaeologists of the French School! Faced with an image as distinctive as the faces of the Bàyon, they are not even capable of telling us whether we are dealing with a Brahmā, a Śiva or a Buddha!’ But the point is that between Brahmā the Creator of the Universe, Śiva who spreads his blessings to all regions of space, the Buddha who multiplies himself indefinitely in the Great Miracle, and Lokeśvara ‘facing all ways’, there are not for the Indian religions those distinctions founded on individualism that the gods of Olympus have inculcated in us. What is hiding under these Indian divinities, which the architect wished to represent, is not so much a real being, an individual; it is only an abstraction…‘royal power blessing the four quarters of the country’. Faces of Brahmā, faces of Śiva, faces of Lokeśvara equally fit with this abstraction, and if we have decided on these last, it is only because of the distinctly Buddhist character of other elements in the Bàyon.

Coedès, the primary epigrapher and dominant historian of ancient Cambodia, here asserts, and attempts to redress, the extended problems of dating and understanding the Bàyon temple, constructed by Cambodia’s first officially Buddhist king. Reflecting on an influential 1936 article by Paul Mus, he sees the tower deity and Khmer state icons in general not as real, individual gods but only as abstractionsrepresentations of royal power beaming blessings to the whole country. Coedès reluctantly concedes to a more precise identification of, say, the Bàyon tower deity as Lokeśvara only on consideration of circumstantial evidence. And yet the following objection to Coedès immediately springs to mind: if the great Khmer icons of state, with their deep Indian roots, were not potent with the individuality of the gods they embody, why would one of Jayavarman's (probably Śaiva) successors have felt the need to desecrate the Buddhist icons of the Bàyon and his other temples on a massive scale, in order to transform them into Hindu foundations? In this poassage, although Coedès is specifically addressing the Bàyon face towers, his claim is general and refers to all Indic, including Khmer, divinities. I reject Coedès’s theory that the gods of the Indians and the Khmers were mere abstract royal power symbols without any dense individual life and identity that could be swapped with impunity. For, although the Bayon faces – for reasons unknown to us – were not desecrated, untold thousands of other major Buddhist icons did not enjoy this clemency. The date of the desecration is naturally not recorded in any inscription, but the reworking into Śivali gas of many of the Buddha icons in the Bàyon is usually attributed to the reign of Jayavarman VIII. My own view is that a Hindu-Buddhist clash appears more likely to have occurred in the reign of Jayavarma Parameśvara (Skt. Jayamādiparameśvara), the last king to leave an inscription in Angkor before the return of Ang Chan in the mid-16th century. In 1 2 , in his first year on the throne, Jayavarma Parameśvara erected a Śivali ga in the (I would ask still Buddhist?) Bàyon.6 Installing a Śivalin ga in the Bàyon immediately upon enthronement looks like a politico-religious act of some significance, so did Jayavarma Parameśvara convert the whole Bàyon for Śaiva ritual and then order the systematic desecration of Jayavarman VII’s other temples in Angkor? The question cannot be answered with the data we have, but the question should be raised. Jayavarma Parameśvara’s 2 -line record of his Bàyon Śivapūjā, may quietly announce the arrival of the desecrator of the Buddhist Bàyon. On the evidence we have, there is no stronger candidate.7


Description of the face towers

The Bàyon faces are indeed a special case of recalcitrant, nondeictic iconography set in a culture loaded with readable signs. But as I hope to show in this chapter, even they, on close study and in conjunction with contextual evidence, can be shown to yield clues for a more precise identification of the god whom Jayavarman VII chose to make so extravagantly prominent in his newly fortified capital.

suggests, if anything, a climate of tolerance for Hinduism and Buddhism, like that clearly attested in the inscriptions of Jayavarman VII.


Coedès 1942:18 ‘Inscription du Bàyon K.4 0’ Inscriptions du Cambodge vol.II EFEO, Hanoi


Many of the Bàyon’s Buddha reliefs were at some date chiselled into Śivali gas, which were in turn removed by a chisel, presumably by Theravādin monks who later controlled the building. Jayavarman VII’s huge .6 m. Buddha on a āga was also pushed into the shaft under the Bàyon’s central sanctuary. There are numerous Śivali gas left today in the Bàyon and one of these was presumably anointed and covered with the bronze (lotus-shaped?) cover (padmavitāna) mentioned in K.470. We have no further information on Jayavarma Parameśvara, except that he commissioned Angkor’s last inscription in Sanskrit on a temple on the ancient Kapilapura mound just northeast of the moat of Angkor Wàt, which Coedès characterises as redolent ‘with Śivaite mysticism’. (Coedès 1968:228) Moreover, two bits of circumstantial evidence argue against Jayavarman VIII being the desecrator. The first is the fact that the 1295 Mańgalārtha shrine is the last monument in Angkor decorated in the Bàyon style, with false windows and lowered curtains (Cunin 2004:363), which speaks if anything for continuity with Jayavarman VII, as does their sharing the same Brahmanical purohita. (The first temples in Angkor in a post-Bàyon style are the Hīnayānist foundations of Pra Pallilay and Pra Pithu near the royal palace, which must therefore be dated to post-1300). The second is the inclusion of eight Hīnayānist icons among the 2 4 Buddhist objects recently uncovered by Sophia University in Bantéay Kdei temple. Eight of the buried icons are in Hīnayānist styles current in the Cambodia in the 14th century. This does not rule out their being made during Jayavarman VIII’s reign, because inscription K.241 of 1267, perhaps the 14th year of Jayavarman VIII’s long reign, marks the installation of an earth-touching, māravijaya (victory over Mara) Buddha, the icon which developed into the dominant image of the later Hīnayāna of all Indochina. K.241 demonstrates that a Buddhist community a few miles from the capital felt free to erect a new Buddha image. This bespeaks a climate of tolerance, rather than of persecution and hurried icon burial. The icons buried in Bantéay Kdei include four reliefs of standing Hīnayāna Buddhas with right hand held forward on the chest, as found on the false doors of monument 486, which Woodward considers ‘may date from the 14th century’, (Woodward 19 5:14 Studies in the art of central Siam 950-1350 Yale), three earth-touching Buddhas and a bronze Hīnayāna standing Buddha with belt and centre fold, which also became standard in the 14th century. The most likely date when all three Hīnayānist icon styles had developed is post-1300. Apart from Jayavarma Parameśvara, all Jayavarman VII’s other successors were Buddhists, so I find it difficult to imagine anyone else with the motivation for wholesale desecration of Buddhist icons. It is conceivable that later Theravādin Buddhists destroyed or desecrated some of the major icons of the earlier Mahāyāna temples, but as most of the icons chipped from the temple walls were simple meditating Buddhas in niches, with no clear sect affiliation, the attribution of the bulk of the large campaign to remove Buddha icons to Theravādins seems unlikely.


Let us start with a brief description of the face towers. They form the superstructure and outline of the massive temple. They are projected high into the sky and their dominant position over the city confers on them superhuman and hierarchical superiority. The 59 towers function both as icons and architecture – an unusual overlap which may have contributed to Coedès’ sense of their expressing a politically significant abstraction of royal power. The Bàyon is not only a work of architecture; it is also an icon. The massive bulk of the temple is a god with 200 faces . Thus, while the conventional ‘presiding deity’ and the focus of ritual activity remains the traditional Khmer Buddha protected by a āga erected in the central sanctuary, I suggest that the architectural superstructure above this image represents a higher vision of ultimate reality, a reality beyond human conception and known only by supreme Buddhas.

The 161 extant faces measuring up to 1.80m. , which originally numbered over 200 , are virtually identical, with only minor physical variations among them. [Plate 1 Face tower] Only the four (or for want of space, three or even two) back-to-back (addorsed) heads of the deity in the tower are visible. The heads are mounted with a high, structured jatā of piled up ascetic’s hair and, where space and design allow, this is surmounted by a large blooming lotus flower. The heads are not hermetically addorsed, as they are in bronze and smaller stone images of multi-headed Khmer icons, but are slightly separated by an ornamented architectural column partially echoed in higher, structural elements in the piled-up hair. The deity’s head is, in each case, regally dressed with a diadem, heavy earrings, a high ‘choker’ necklace (collier gorgerin) of floral or medallioned links and a foliate helmet or hair cover that reaches down towards the invisible shoulders. The facial features – square jaw, full cheeks and broad, pacific, full-lipped, flickering ‘Khmersmile – are typical of the sculptures of the ‘Bàyon style’. The eyes are wide open and stare horizontally forward. The forehead bears a raised diamond-shaped lozenge which is sometimes carved with the pupil of a third, ‘vajraeye. otably absent is any Amitābha figurine in the hair, the hallmark of Avalokiteśvara, and although some of the eyes have been closed shut by later deliberate erasure, none originally had the lowered lids of the great Bodhisattva or ’Lord who looks down’ that is distinctive of the earlier Bàyon style. The facial features are not very close to those of the presumed portrait statues of Jayavarman VII, although the portrait statues also fall within the norms of the Bàyon style. We will come back to this.

None of the iconographic details excludes the faces from either the Buddhist or Hindu pantheons, as Coedès would no doubt have readily agreed, and yet Jayavarman VII surely knew exactly which deity he had chosen to remodel the skyline of Angkor. Scrutiny of the constrained number of traits and details of ornament and dress should help us narrow down the options.

Iconographic details:


1. Open eyes

One of the creative breakthroughs of the royal workshops under Jayavarman VII was the capturing of an astonishing mix of quietism and power in the lowered, meditative eyes of Buddhas and Bodhisattvas. This is seen in the king’s triads of the Buddha, Lokeśvara and Prajñāpāramitā that dominated the first decade of his reign. Boisselier memorably described the rapt smiles and lowered lids: An impassive ideal of superhuman beauty [expressed in the earlier ‘Angkor Wàt style’] is replaced by a highly sensitive search for human truth. The Buddha, like the Mahāyāna or brahmanical divinities of the time, ceases to be supernatural. Overflowing with compassion, he reverts to the human state and appears in the form of a Khmer.

The lowered eyes continue to grace the huge output of robust, compassionate and human Lokeśvaras through the second decade of Jayavarman’s reign, [Plate 2 Lokeśvara at New York Metropolitan Museum] but in the Bàyon towers the eyes of the deities suddenly open into an unwavering yet still smiling stare. Although, as Dagens notes, some of the Bàyon faces have had their eyes smoothed shut in later rework, this was not their original state.15 The open eyes, charged with knowledge and confidence, become the hallmark of the king’s third decade in power. In this, they are almost a throwback to the staring deities of the two largest foundations built before the Bàyon: Angkor Wàt and Phimai.

In this chapter, I am following the three phases of the Bàyon style distinguished by Philippe Stern16, which coincide roughly with the first, second and final decades of the reign. the first phase under a Buddhist triad, the second dominated by Lokeśvara and the third, which includes the final work on the Bàyon, Bantéay Chmàr and Bantéay Kdei. Stern suspected a major religious shift in the third phase, because Lokeśvara sculptures were covered over in the Bàyon and his image all but disappeared from the sacred art, but he was unable to define the new phase: Was the great vogue…[of Lokeśvara] ended? Given that one of the hidden pediments was dedicated to him and that he is absent from all the pediments left visible, had a new religious change taken place? One can ask, but as the change is only observable on a single great temple, the Bàyon itself, it is, despite the importance of the monument, difficult to say.18 I will argue (see ‘The Angkorian context’ below) that the third phase, in the opening two decades of the 13th century, was focused on a Hevajra cult, which we can, in imitation of Stern’s name of 'Lokeśvarization' for phase two, be called ‘Yoginification’, after the Yoginis who dance around Hevajra in his mand ala.

The opening of the eyes may be a signal that the Buddhist conceptions with which Jayavarman remodelled the country’s spiritual defences had been adjusted. Staring eyes are usually indicators of Tantric deities (whether Hindu, as in icons in the ‘Angkor Wàt style’ or Buddhist, as at Phimai). The Khmers did not go in for horrific icons of wrathful deities, so an open-eyed stare, as in the Khmer icons of the supreme Tantric deity Hevajra, was used as an adequate indicator of such a fierce god of Tantric Buddhist meditation rituals.19

2. Lozenge/third eye

Another feature of the Bàyon deities that has caused puzzlement is the lozenge shape on the forehead of the faces. This is mostly raised in relief but sometimes incised into the stone. It sometimes contains a bulging iris and is sometimes only a border, rather like the mounting of a jewel. [Plate 3 Lozenge with pupil] In Khmer bronze figures any of these variants would be taken as the third ‘vajraeye, which Tantric Buddhism adopted from Śiva’s third eye, and which confers the omniscience of a Buddha’s vision. Similar raised lozenge eyes are sometimes carved on the forehead of the middle Bayon period Lokeśvaras; they are invariably present on the brow of Hevajra and the eight Yoginīs that dance around him in the icons of what is understood here as the last, more overtly Vajrayānist part of Jayavarman’s reign.

3. Choker necklace The necklace below the Bàyon faces is close in design to the high floral or medallion choker worn by the asuras of Angkor Thom’s causeways. In the latter case this appears to be part of military dress and the ornament is shared with Khmer temple door guardians. But the chokers on the towers may be no more than a suggestive device, used in the limited space available to the sculptors, to indicate that the unseen giant bodies standing or sitting in the Bàyon towers are adorned in full regalia.

4. Diadem & hair cover

allowed himself to be borne along by the interpretation of Lokeśvara samantamukha formulated by Mus and Coedès, even while referring to Boisselier’s 1951 interpretation of Vajrasattva. (Stern 1965:145, 163)

18 Stern (1965:164)

19 Woodward points to the Hevajra in the Bangkok Museum as demonstrating that in Khmer art ‘there need be no greater facial indication of Hevajra’s fierce aspect than open eyes.’ Woodward, H.

(1981:5 ) ‘Tantric Buddhism at Angkor Thom’ Ars Orientalis 12

The diadem of the face tower deities is in the form of a high coronet or tiara, with attached foliate ornaments falling behind the ears towards the shoulders, and possibly covering the hair. Regal attire in Buddhism points to the princely Bodhisattvas or the crowned Buddhas of the Tantras, and in Khmer sacred art to the crowned Buddhas of Phimai. Long hair and neck covers like these, according to Boisselier, are ‘normally reserved for terrifying persons: the dvarapālas of the right door post, Asuras and occasionally warriors.’ Hair covers could suggest a vigilant guardian aspect in the faces – what Woodward calls the conquering and guarding aspects of Tantric Buddha images. Coedès published in 1923 a Khmer bronze in the Bangkok Museum, which shows Vajrasattva, identifiable in his classical mudra with vajra in the right hand before the chest and the bell in the left on the left hip, wearing a lightweight helmetlike hair and neck cover. [Plate 4 VS with hair cover Coedès 1923 pl.28.3] If it is not a hair cover that is represented on the towers (which cannot be ascertained) the foliate ornament behind the ears, like the choker, may, again, simply denote full regalia.

Interpretations


We now survey the principal attempts at identification of the Bàyon faces – including the Buddha, Brahmā, various composite deities (many involving Lokeśvara), the king, territorial gods and deva/asura guardians.

1. Buddha


The earliest historical description of the face tower deities calls them Buddhas. Exceptionally in Khmer history, we have an almost contemporary eyewitness report dating from the same century in which the Bàyon was completed. In 1296, nearly 80 years after Jayavarman died, and just one year after his son’s successor Jayavarman VIII abdicated, an observant and numerate Chinese ambassador spent a year in Angkor. Zhou daguan described the giant four-faced towers over the gates of Angkor Thom as five-headed Buddhas:

Above each gate are grouped five large heads of Buddha in stone, whose faces are turned towards the four cardinal points; in the centre is placed one of the five heads which is ornamented with gold.

Paul Pelliot, who translated the Chinese text, suggests the carvings were not of Brahmā but were the five-headed Śiva pañcānana, that was much produced in images by the Khmers in this period. Pelliot speculated that the fifth, golden head had eventually fallen to the ground. No trace of added heads has been found, but the ambassador’s report has given rise to several theories about the fifth golden head, which have tended to strengthen the inculpation of Jayavarman VIII. Bruno Dagens, for example, proposes that the fifth golden head may have been added in stucco and gilded and was part of the reworking of Buddha images in the Bàyon to ‘Śiva-ize’ the city gates and the state temple ‘when the Bàyon became Śivite in the mid-13th century.’ However a new look at the evidence for Jayavarman VIII being the desecrator by Michael Vickery, as noted in footnote five, now prompts a fundamental reappraisal of when the Bàyon was ‘Śiva-ized’ and the evidence now points to the 1320s, some 30 years after the ambassador visited the city. Zhou made no mention of whether the Buddhists were perpetuating their rituals in the great stone temples built by Jayavarman VII. He wrote only of Buddhists in villages using simple, single room temples with clay icons, but he said the Hindu priests (pa-sseu-wei) were worse off. The ambassador’s mention of four stone and one golden head on the gates of Angkor Thom therefore remains unexplained.

2. The most disseminated interpretation: Bodhisattva Lokeśvara


For many decades, scholars viewed the Bàyon as a ninth century Hindu foundation until Parmentier in 1924 discovered a large, built-over pediment of Lokeśvara in a routine cleaning operation that proved the Bàyon's Buddhist vocation and changed the status quo overnight. In 1927 Philippe Stern questioned the ninth century date after analysing the evolution of faces in Khmer sculptures and of temple structures and in 1928, George Coedès argued conclusively that the Bayon was the 12th century state temple of the Buddhist king Jayavarman VII whose inscriptions it bears. The Bàyon’s Buddhist cult was further confirmed in 19 when Georges Trouvé recovered and reconstructed a large, smashed āga Buddha while cleaning a shaft under the central sanctuary tower, after a storm dislodged stones from the cella, obliging Trouvé to insert iron supports.

In two articles in the 1930s, Paul Mus then laid the groundwork for this quasiconsensus which sees the faces the size of a man in the towers as icons of the Bodhisattva Lokeśvara samantamukha (‘facing all ways’), in the likeness of the king.

Mus proposed to resolve the ‘enigma of Khmer archaeology’ by applying to the Bàyon the vision of Avalokiteśvara in the great early Mahāyāna classic Lotus o the ood La addharma-pu ar ka). He sees the Buddhist and Hindu deities whose names appear in short inscriptions in alcoves and on the doorjambs of the Bàyon salles-passages or vra ku i as widely varied icons that are all emanations of the presiding Lokeśvara in the towers. Indeed, Avalokiteśvara in the Lotus sūtra is capable of appearing in the form of any deity. Mus also sees the uniformity in the 200 giant faces as an outcome of this high level synthesis of the empire-wide pantheon of gods below: [The faces] signify that beneath them in the shadowy chapels, Buddhas, gods and local genii are merely so many masks worn by the great succouring bodhisattva whose portrait appears in broad daylight on the four sides of the towers.

For Mus, each face tower corresponds to a province or city, whose local gods are united in the icon of Lokeśvara in the tower, carved in the likeness of the king and generating ‘royal power blessing the four quarters of the country’.31 Several of the extant 41 short inscriptions include geographical references, but all attempts to formally map their random distribution to the kind of ‘mystical geography’ Mus seeks, have failed. Certainly Jayavarman invested resources in erecting a network of temples and deities throughout his kingdom, and then built a state temple that honoured the whole, expanding pantheon. And imperial synthesis was a priority, as seen in the 1191 Pra Khan stela, which commands the presence at an annual spring festival of the icons of 122 gods (including 2 ‘Jayabuddhamahānāthas’ distributed to major cities). Coedès presumed they were carried to the capital in bronze replicas. This annual capital festival tradition may have been extended in whatever activities took place in the 16 vra ku i built across the outer courtyard of the Bàyon. Maxwell, in this volume, notes a geographic aspect to images installed in the outer sections of the temple, so the vra ku i may have been reserved for deities of particular regions of the empire. Coedès was prepared to speculate that as positions for two of the

Jayabuddhamahānātha images are marked in the short inscriptions of the Bàyon all 23 might have had named places there. Dumarçay and Royère estimate the vra ku i were built as late as ‘1210?’ and dismantled perhaps five years later, but cite no evidence. Transporting replicas may also have continued, so that the short, carved Bayon inscriptions may in some cases have only marked the position reserved for each icon.

But the cult of Lokeśvara at the heart of the Pra Khan festival underwent a significant change. In 1191 there is indeed much evidence to link Pra Khan, Neak Pean and the king’s other temples with Lokeśvara. Philippe Stern defines a radical switch in the 1190s from the early Buddhism focused on the triad of āga Buddha, Lokeśvara and Prajñāpāramitā to a ubiquitous veneration of the compassionate Bodhisattva Lokeśvara, propagated in an empire-wide campaign he calls ‘Lokeçvarisation’. The core second level of the Bàyon around the central sanctuary was built by this time, suggesting it may first have been conceived as a temple dedicated to Lokeśvara. Cunin’s measurements of the magnetism held by the stone (0.7 ~ 1.3 x 10-3),which make the core Bàyon contemporary with, or at least built from the same quarry shipments as the third series of additions to Pra Khan and the core of Bantéay Kdei, are consistent with this. Moreover, Cunin points to the erection of scaffolding around the Bàyon’s ‘massif central’ and surrounding face towers as requiring that the towers were built before the galleries that later connected them. The same requirement for scaffolding space would also dictate that the faces were sculpted on the towers from the same scaffolding. The rest of the Bàyon appears to have been left undecorated while work on the king’s other temples went ahead. Stern remarks on the homogeneity of the decoration of the walls of the state temple and attributes this all to the third and final phase of the Bàyon style in the first 20 years of the 13th century. Cunin, concurring with Stern’s analysis of the late (fourth) devatā type and other decorative motifs in the Bàyon, concludes that after the central architectural phases and the carving of the first face towers, there must have been a significant time gap between the building and the decoration of the walls and galleries of the Bàyon. Cunin indeed considers this a unique characteristic of this temple.

The abrupt end of Lokeçvarisation directly impacted the Bàyon, for the first phase of the carving of the state temple also appears to have included large pediments of

Lokeśvaras in the second level facing the central sanctuary , which were later consigned to obscurity below the extension of the third level platform bearing some of the face towers. Indeed, as Stern remarks, Lokeśvara seems to have been deliberately ousted from the state temple: We notice the total or quasi total absence of ‘Lokeçvara’ in the easily visible pediments. His great vogue, responding to the reform of the beginning of the second period, is it ended? Given that one of the hidden pediments was dedicated to him and that he is absent from all the pediments left visible, had a new religious change taken place?

The hidden pediment was the one whose discovery in 1924, when stone slabs in the platform were raised in a cleaning operation, first signalled the Bàyon’s Buddhism. The Lokeśvara pediment was thus unceremoniously built over. The Mus and Coedès reading of the face towers would therefore leave us with the irreconcilable propositions that the temple builders were carving Lokeśvara as the monument’s great deity of the face towers, while at the same time consigning his image to ignominious obscurity in the courtyard below.

Furthermore, seeing Lokeśvara as absent from the face towers is supported by a series of significant iconographical anomalies. First, the deity of the face towers does not have Lokeśvara’s figurine of Amitābha above the coronet; second, icons of Avalokiteśvara with four faces are unknown. It is hard to believe that such a committed message propagator as Jayavarman VII would have tolerated the omission of the instantly recognisable Amitābha figurine, attribute of the great Bodhisattva, if the latter was indeed the god he was celebrating. The absence of Amitābha was alone enough to discourage Bernard Philippe Groslier, the last French curator of Angkor, from seeing Lokeśvara in the Bayon.43 A third iconographical problem was immediately raised by Pierre Dupont in response to Mus’ article: Khmer Bàyon-style Lokesvaras wear a cylindrical chignon and never wear a diadem.44 Yet a fourth iconographical difficulty is that unlike the Lokeśvaras of Jayavarman’s second decade in power, the Bàyon deities have their eyes open and not lowered, as in all the images of the powerful Bodhisattvas of the 1190s.

2.1 Buddharāja


We now turn to the other components of the composite deity that Mus and Coedès identify in the face towers. Coedès believed the God of Angkor was the king: We know that this god of Angkor was the king, the god-king, personified before the 12th century by a golden li ga and at the time of the Bàyon by the statue of Buddha, which was recovered from the bottom of an open pit under the central tower.45 Coedès coined the nameBuddharāja’ to suggest that Jayavarman’s royal power was based on a Buddhist restatement of the Śaiva Tantric devarāja cult created for his ninth century namesake, Jayavarman II:

Instead of the devarāja of previous reigns represented by a golden liòga, however, the central sanctuary sheltered an enormous stone statue of the Buddharāja. This statue was not only a substitute for the Śaivite devarāja but also a statue of apotheosis of the founder king, whose features are undoubtedly also to be seen on the upper parts of the towers in the form of the Bodhisattva Samantamukha, ‘who faces in all directions’.46 Coedès held to his interpretation of the devarāja cult as the paradigm of divine kingship legitimization throughout Southeast Asia until his death in 1969, despite growing dissent from Stern, Filliozat and others. Later scholars, notably Hermann

Paris. Boisselier points out that at Bantéay Chmàr, where Lokeśvara appears in all his Bàyon-style forms, he is never represented with four faces but with one, eight, 11 and 16. (Boisselier 1951:329). Mus is aware of the problem of associating Lokeśvara with a four-faced icon and offers the less than convincing explanation that it is one face continuously facing the devotees who walk in pradak ina around it, somewhat in the way Brahmā, on a longer timescale, faces the four ages of the world (Mus 1937:73).In late Bàyon style Khmer icons, Lokeśvara eventually appears in the KVS-inspired Tantric ahākaru ika form with 11 heads and (representations of) 1,000 arms, which became a major icon in Tibet, East Asia and Vietnam. The largest actual number of arms in Cambodia is 32. This Tantric form of Avalokiteśvara with 11 heads appeared first in the Kanheri cave of western India in the sixth century (Neville, Tove E. 1998:fig. 9 Eleven-headed Avalokiteśvara, Chenresigs, Kuan-yin or Kannon Bodhisattva: it’s origin and iconography Munshiram Manoharlal, Dehli). The best known Khmer icons of the 11-headed Mahākaru ika Lokeśvara are among those linked by Boisselier to the KVS on the outer gallery of Bantéay Chmàr. 43 ‘…Lokeśvara always has his Dhyāni-Buddha on his chignon. In fact nothing allows any explanation for its disappearance on the towers of the Bayon, where it could have been easily sculpted.’ Groslier B.P. (1973:305) Le Bàyon: inscriptions du Bàyon EFEO Paris (my translation). 44 Some iconographical objections to Mus’ article were immediately raised by Pierre Dupont: ’We must however observe that that no known Avalokiteśvara – and there are many in the Bàyon period repertoire – corresponds exactly with the head sculpted on the towers of Jayavarman VII. This latter always wears a gilt diadem, decorated with petals and prolonged behind the ears, regularly ornamented with pendants; the neck is encircled by a tightly adjusted necklace. Khmer Lokeśvaras, in as far as they are now known, wear a cylindrical chignon and sometimes jewels, but we never see them with such a diadem and necklace.’ Dupont, Pierre (19 6:6 0) ‘Chronique’ BEFEO 36 (my translation).

45 Coedes (1963:64)

46 Coedès 1968:175

Kulke, Claude Jacques, Michael Vickery, Saveros Pou and Hiram Woodward have rejected Coedès’ understanding of the devarāja cult and his ‘Buddharāja’. But the identification of Lokeśvara in the Bàyon faces, led by Mus’ invocation of the Lotus Sūtra, with Coedès’ assent, was the first attempt to formulate an intellectual model for how the Bàyon functioned, and certainly merits the broad influence it has had. But it does present problems.

2.2 Portrait of the king

Mus’ and Coedès’ vision of a living, royal-divine network of āga Buddha/face tower Lokeśvara and royal portrait makes the king’s face ubiquitous as the supreme expression of power and the protection of the empire. This is presumably what eventually led Coedès to see megalomania as a principal drive behind the huge temple-building programme. But were all these icons really made in the king’s image?

Coedès believed that a living king can appear as a human being, a Bodhisattva and a Buddha. Since the sculptors of the Bàyon style invented the veridical portrait, we would expect all three forms of the kinghuman portrait statue, Lokeśvara and ‘Buddharāja’ – to be true likenesses of the same man. And Coedès indeed holds that the king’s portrait was sent to the cities and his own features appear in the major icons of his state temple. We can therefore test Coedès' interpretation by comparing the portrait statues with the divine ones. The generally acknowledged portrait statues of the king found in Kompong Svāy and Phimai [Plate 5 Jayavarman, Phimai] show him as an uncrowned man with lowered eyes probably inclining in añjali mudrā before a god – but not himself as a god or Buddha. These statues do appear to be portraits of the king; for they depict a muscular, bull-necked, moustachioed, squareshouldered, middle-aged military type, with a slightly expanding girth, such as we might expect Jayavarman to have been after his long military career with the Čams, and his fight to secure the Khmer throne in his forties. [Plate 6 Guimet Jayavarman portrait]

Images of the gods in the Bàyon style have none of these features. The portrait statues are different in many details, for example, from the āga Buddha in the Bàyon, which has the conventional narrow waist and somewhat narrow shoulders of most Bàyonstyle Buddhas, and no moustache. The Bàyon Buddha also has a pronounced and pointed cranial protuberance or u a on the top of the head, while the portrait statues have a small, tight ball of hair secured by a ribbon on the back of the head. The features of the portrait statues echo neither the ascetic, withdrawn Buddha of the

Bàyon cella, nor the more youthful, regally-adorned, wide-eyed, assertive god of the Bàyon face towers. It can be readily granted that all three icons share the basic facial features of the Bàyon style – wide, flat mouth, high forehead, heavy eyebrows – but this applies, as Dagens points out, to ‘virtually all of the statuary of the Bàyon and it is therefore not really meaningful.’49 We may then conclude that only a handful of portrait statues were made of the king, while images of the gods were produced in thousands and were quite different.

The Coedès scheme of a personalised divine network uniting the empire weakens if we extract the portrait element from the presumed composite of deities in the Bàyon. And if the direct link to the king’s person is removed, the differences between the assertive face tower deities who wear heavy crowns and choker necklaces, and the withdrawn, uncrowned Bàyon Buddha sitting under a āga hood, with his earrings removed, seem more pronounced. Coedès’ scheme even erodes further if we accept Woodward’s argument – which I propose we do – that from the material record, it is much more likely that the 23 distributed Jayabuddhamahānāthas – whose name implies a deity not a man – were not the portrait statues of the king but the large eightarmed ‘radiating Bodhisattva’ icons found in much greater numbers than the portraits. Without the king’s portrait, Coedès’ network reposes solely on a centrifugal Lokeśvara and is therefore fully vulnerable to the objections listed in section 2.0 to seeing Lokeśvara in the faces.

2.3 neak ta

Bruno Dagens modifies the idea of a divine network articulated by Mus and Coedès by defining the Bàyon as the centre of a spiritual map guarded by protective, territorial deities, which he sees as the Khmer equivalent (neak ta) of India’s ks of the empire.etrapāla , ‘guardians o51 Dagens says the f the field/domain’, extending from the capital to the frontiers ks etrapāla of southern India, because they protect

boundaries or sacred places, ‘are generally considered as a form or even a double of Bhairava, the “terrifyingform of Śiva…’ Although the Bàyon faces do not have the fierce aspect of Indian ks etrapāla , he sees them as wearing ‘demonic’ diadems, which he believes would convey to everyone their dangerous nature. Dagens therefore proposes that the Bàyon constitutes ‘the divine assembly of the Protectors of the kingdom, a monumental expression of that, made up of the gods of the great temples of the kingdom grouped in the chapels of the [Bàyon] temple,’ and that the divinities in the face towers above this assembly are a kind of super, state-level neak ta, the traditional local Khmer ancestral spirits, protective deities of a hill or a tree, which come under this modern Khmer term. He acknowledges the iconographical reservations about seeing Lokeśvara in the face towers, but nevertheless concludes that the guardians in the face towers do link up with the ‘radiating’ Lokeśvara images distributed throughout the empire in a single divine web of ‘statues in the provinces and face towers in the heart of Angkor’. The way Dagens accounts for there being

200 faces carved on the Bàyon is by saying these territorial guardians are ‘numerous because they have come from everywhere and watch over the map of the kingdom: visible from far and looking in all directions, they make tangible for all the divine protection which, thanks to the sovereign, extends over the whole kingdom’. From Paul Mus’ last words on the āga Buddha of the Bàyon, it seems he was moving towards a position similar to that later articulated by Dagens (though retaining the king’s portrait assumption that Dagens has helped us to dislodge):

Jayavarman VII on the āga at the centre of the Bàyon is like a very great village genii at the level of the kingdom.

Mus’ attempt to imagine a village genii rising to the position of pre-eminent guardian deity of state raises the major problem entailed by identifying the face towers closely with neak ta – namely that the latter exert their influence over very limited territory – a tree, a hill, a river, a pathway or a home. Their power may be exerted over a neighbourhood, but not a city, let alone over one of the greatest empires on earth. Dagens gives us no idea of how Jayavarman, a learned man steeped in Sanskrit texts and Indic cosmology, could have raised such a homely and lowly group of spirits to be the driving force and the defence of a major empire; nor does Mus. India’s (and Central, Southeast and East Asia’s) Brahmans and Buddhists worked for centuries to impose over the myriad ancestral deities of Asia a superstructure of universal gods, which then had state level application in politics and war. Dagens’ proposal that we see neak ta presiding over Cambodia’s first state Buddhist temple at the high point of imperial power, albeit doubling with the form of Lokeśvara, implies that Jayavarman would have been reversing this long-established trend by ceding precedence to local deities in the new Buddhist temple of state. I will argue, against Dagens and Mus, that Jayavarman followed a strong historical trend and that although his mission was designed primarily for maximum impact among the Khmers, he also sought to counter the contemporary destruction of Buddhism in northern India and boost Buddhism and his own syncretic regal creed by imposing a Buddhist pantheon over all other deities in Angkor. The kings of Burma, Nepal and Tibet all showed their concern to support the institutions that generated the Buddhism that lent them international support and legitimacy. A geographical ma ala linking up the state level power with local territorial guardians still resonates with the evidence today, but we have still to identify the central deity thus invoked by the medieval Buddhist king.

3. Brahma

Before the Lokeśvara interpretation took hold, the Bàyon faces were thought by many to represent the four-headed Vedic and Hindu god Brahmā. Scholars first drew on the traditional Khmer identification as they believed a Hindu king built the temple in the ninth century. Modern Khmers call the Bayon faces 'Brahma' (without the long final ā of the Sanskrit name) or ‘Tà Prohm’ (‘ancestor’ or ‘grandfather’ in Khmer), the name which remains attached to the temple with huge fig trees growing on its walls. The Khmers are referring to the form of the Hindu god Buddhicized in early Buddhism and prominent in the later Theravādin cultures of Cambodia and Thailand, but when and how this identification took hold is unknown and calls for research. In Thai sacred art it would clearly help if we could establish a firm date for evidence such as the massive, four-faced stone in the Ayutthaya museum, which Woodward refers to as being part of a gate which was perhaps called brahmasukuta or brahmasugata.61

Scholarly opinion swung away from this identification towards Lokeśvara after the chance discovery of Lokeśvara pediments in a cleaning operation proved the Bàyon was Buddhist before it was converted into a Hindu temple. The articles of Mus charted the new direction. More recently, some scholars have been drawn back to seeing the Buddhist Brahma in the face towers. Boisselier changed his mind in his later writings, without referring back to his earlier view, and opted for a Brahma interpretation, which Woodward subsequently expanded on.

The state Śaivism dominant since the ninth century (though perhaps weakened by the arrival of the Mahīdhara dynasty in 1080) was presumably maintained in a calendar of Brahmanical rituals and ceremonies ensuring continuity in royal dedications, state liturgy and hagiography. Even Jayavarman VII’s predominantly Sanskrit Buddhist inscriptions are shot through with Brahmanical poetic allusions, borrowing constantly from the Hindu epics and mythology. Boisselier changed his interpretation of the Bàyon on the basis of such allusions. He took one section of a fragmentary inscription as yielding clear allegorical evidence that Jayavarman VII consciously built the Bàyon and Angkor Thom on the model of the city of Indra in the Trāyastri śa heaven of the 33 gods. But I find the evidence of the passage, in which the city is compared with a bride, is neither direct nor clear because it depends on translating sudharma (‘good order’, ‘justice’, ‘righteousness’) in its secondary sense of a name for the assembly hall of the gods in Indra’s celestial city. One would hardly suspect this to be the case from the forceful way Boisselier puts his view across:

…the Khmer inscriptions still establish clearly that the new capital is the City of Indra (with whom the king is identified) and Tavatimsa Heaven – the Heaven of the 33 Gods – (with whom the princes and provincial governors under the king’s authority are identified), with its Royal Palace, its pleasure gardens and the Assembly Hall of the Gods, which is none other than the Bàyon (it is not I who put forth this idea, but epigraphy which asserts such a notion, destroying a multitude of perilous or whimsical hypotheses).

Coedès offers the allegorical translation in parentheses as a second rendering of the Sanskrit; this is Boisselier’s version:

’...worthy of praise, containing the great [[[garden]]] andana, and having at her summit the

[hall of] Sudharma of the [city of] Sudarsana, her domain was comparable to heaven.’

But if we take Coedès’ preferred translation of the lines, the clarity of the supposedly key interpretative reference for the Bàyon evaporates before our eyes: Pure, thanks to the conduct of her master, possessing celestial power, worthy of praise, experiencing a great joy, versed in the righteousness of honest people, her land resembled heaven.

Boisselier made the allegorical allusion the basis for a dramatic change in his interpretation of the Bàyon, its face towers and Angkor Thom – a base which I find precarious, and surely not robust enough to destroy a ‘multitude of perilous or whimsical hypotheses’.

In the poetical inscription in which the Čam king Jaya Indravarman is called āvan ā for attacking Angkor, presumably sometime in the 1170s, Boisselier sees a parallel with the Asura attack on Indra’s city. The Asuras were defeated and Boisselier compares the four Great Kings set to guard the city with the giant faces above Angkor Thom’s gates -- where they indeed loom above an image of Indra seated on his threeheaded elephant Airavata. On the Bàyon itself, whose faces Boisselier calls, without further elaboration, ‘quite different’ from those on the gates, he sees Brahma as he appears on auspicious days in the guise of the ever-youthful chieftain Gandharva Pañca ikha who multiplies his image to honour each of the 33 gods simultaneously. Boisselier admits his reliance here on Theravādin sources, justifying this by claiming ‘there is no significant difference in cosmological matters between the Mahāyānist and Theravādin traditions.’67 As we will shortly see (3.1), Woodward makes a more sustained case for seeing the Hīnayāna Brahma in the Bàyon faces by taking head-on the issue of which Buddhism we are experiencing in the late Bàyon and by claiming that the king took some steps towards Hīnayāna.

Boisselier again appeals to pan-Buddhist doctrines when he finds the Bàyon faces as not exactly smiling but expressing the meditational “‘active state of mind’ which the scriptures call the four brahmavihāra, the ‘things pleasing to Brahma’, the ‘sublime state’ leading to charity, compassion, joy and tranquillity.’ These meditational stages are indeed, as Snellgrove says, ‘very old property of Indian Yoga and probably belong to the earliest Buddhist practice’ and they survived in all forms of Buddhism. They are preserved in mature Vajrayāna, where they feature in the hevajra-tantra and the tārā-sādhana of the sādhanamālā.71 Sanderson finds them in another Yoginītantra, a still untranslated Tantra for visualizing the deity Sam vara (the Heruka deity who is known to be in the Mahīdhara pantheon from his image on one of the principal interior lintels at the central sanctuary in Phimai): For example one carries out the second evocation of Sam vara as follows . After reciting the Mahāyānist formulae: confessing sins, delighting in the merit of others, transferring one’s own merit for the benefit of others, and taking the vow of the bodhisattva, one contemplates the four brahmavihāras: benevolence, compassion, joy and patience. One then meditates on the essential purity of all phenomena and oneself, sees oneself and all things as nothing but mind

(cittamātra), realizes their emptiness, and then out of this emptiness generates the deity’s icon. If we transform this picture of the meditator into a king generating the icon of his city or state, we arrive at a vision of the Bàyon and its faces, which is very close to that which will be endorsed later in this chapter.

But Boisselier’s drift seems to be towards Hīnayāna, for the caturbrahmavihāra are seen primarily as a meditational technique for advancing to the Theravādin Brahmaloka and their major canonical exposition is in Buddhaghosa’s fifth century Pāli visuddhimagga. Moreover Boisselier, in seeing Jayavarman’s southeast Prasat Chrung inscription as indicating the Bàyon was conceived as the assembly hall of the gods, adduces the support of the Pāli janavasabha-sutta for the visit to Tāvati sa heaven by Brahma Sanaòkumāra (‘ever-young’) in the form of adolescent Gandharva Pañca ikha (with five topknots) when he created an image of himself on the couch of each of the gods – a vision realised architecturally in the Bàyon face towers.

3.1 in r i eka

Hiram Woodward significantly buttressed Boisselier’s account by showing that a cosmological document compiled in Bangkok in 1802 states that when Sanaòkumara appears over the throne of devaputra he ‘feels great joy, as if a king had received a new abhi eka and were rejoicing in the prosperity of his kingdom’.75 Southeast Asian Theravādin cosmology had thus actually absorbed the vision of Brahma appearing simultaneously all over Mount Meru that Boisselier had divined in the Bàyon; furthermore it was linked to a second consecration of a king as an emperor. The subject of a second consecration is a contentious one because the evidence is thin, but it is worthy of exploration because it usefully raises the unanswered question of what state ceremony officially inaugurated Jayavarman’s vast state temple – was this a second consecration of the king as cakravartin?

Woodward responds affirmatively, suggesting that Angkor was the scene of one of the early absorptions of this Brahmanical material. He believes that all but the first 16 face towers with giant faces of Vajrasattva were being carved at the end of the Bàyon construction phase (which Cunin’s contribution to this volume does not support) and were converted into multiple images of Brahma to accommodate them in a Theravādin indrābhi eka ceremony. In this way he addresses the larger issue of when the Hīnayāna arrived in Angkor, which Boisselier had left aside. Coedès believed an indrābhi eka ‘took on a particular character in the Indic kingdoms of Indochina’ and proposed its origin could be found in the Vedic aitareyabrāhman a . The Vedic text in fact refers to a higher consecration of Indian emperors who achieved paramount power after a major war.77 It was called aindra mahābhi eka, recalling Indra’s consecration as the king of the gods, a name which seems to have survived in Southeast Asia as the contraction ‘indrābhi eka’. Such a ceremony is mentioned, as Coedès noted, in the Thai Palatine Law, but with no indication of when it was deemed appropriate. The 1557 indrābhi eka recorded in the Ayutthaya annals marked the construction of a new palace by king Mahācakravarti, but there is no direct evidence for linking this ceremony back to Angkor.

I find that a case for Jayavarman propagating a form of Hīnayāna, and for this being visible in the Bàyon, rests on a very delicate web of evidence. Jayavarman undoubtedly turned Cambodia to Buddhism, but his major extant inscriptions were in

and his smaller temples, which were in Khmer) and the first use of Hīnayāna Pāli in a Khmer inscription occurs only in 1308, almost a century after his death. Much less tenuous, however, is the amount of evidence, which we will examine shortly, to suggest that the king advanced ever further into Vajrayāna and that the final phase of his royal Buddhist cult emphasised the presence of Hevajra, Yoginīs, Vajrapā i, Vajradhara, and Vajrasattva -- evidence which Woodward was the first to draw together.

Furthermore, it is hard to imagine Jayavarman VII choosing as a celestial figurehead the Buddhist version of Brahma, who may have been accorded less status than Vi u in Cambodia’s Theravādin Middle-period , and who had long declined in his Hindu form in India. The king’s strategy after seizing power by force – in circumstances still far from clear – following a Čam attack that killed the previous king and possibly left the palace sacked in Angkor – was to effect a major shift to impose Buddhism as the dominant religion of state for the first time. Placing Brahma at the pinnacle of his pantheon at the ceremonial high point of the reign would just not have packed the requisite message of cosmological backing for the new vision for the state.

Moreover, post-Bàyon style sculpted reliefs and icons on the Angkor royal terraces suggest that Jayavarman’s kind of Buddhism was perpetuated for at least two decades after his death during the reign of his son Indravarman II (r. ca 1220-43). These terraces maintain the father’s focus on the kind of dancers in the ‘salles aux danseuses’, although their wings and their ornate, bejewelled and crowded style betrays a loss of intensity, compared with the Bàyon dancers. [Plate 7 crowded vajra-eyed post-Bàyon Yoginis] Stern called it an ‘internal, precious evolution’ of the Bàyon style. Indravarman’s primary Buddha image is known as the Buddha of

Commaille, after Angkor’s first – sadly murdered – Conservateur who found it. This Buddha is transitional in style in that it shows a distinct development towards the kind of tranquil, unadorned images in monk’s robes, with softened features and with eyes again lowered, that was to characterise the late 13th and early 14th century Khmer Hīnayāna. But Indravarman’s ‘Commaille Buddha’ still sits naked or in a diaphanous robe on the Mahāyānist āga and has not yet metamorphosed into the standing monk who was shortly to dominate.

The definitive Khmer swing to Hīnayāna is first seen clearly in the decoration of the Pra Pallilay and Pra Pithu temples, built close to the palace in Angkor, which, in footnote seven, I have suggested should be dated to post-1 00. The pediment of the Pra Pallilay gopura indeed holds a new style standing monk Buddha. [Plate 8 Preah Pallilay gopura standing Buddha] These temples are also carved with early Khmer Buddhas in earth-touching mudrā , of the kind first encountered in epigraphy in the inscription K.241 from Prasat Ta An near Kralanh, northwest of Angkor (footnote seven). In the no doubt later Pra Pallilay model, the earth-touching Buddha is accompanied by crowned adorants. Popular Hīnayānist scenes like the taming of the alagiri elephant are also present. Pra Pallilay and Pra Pithu are built on a significant scale in stone and are positioned strategically close to the palace – all of which suggests the return of royal patronage to the Sangha, perhaps immediately after Jayavarman VIII’s abdication in 1295.

A further reason for resisting the Brahma interpretation is that a second coronation or indrābhis eka could have been recorded with a Tantric Buddhist version of the Indic Ocean of Milk churning myth. In the next section I argue that the Indian monastery at ālandā had already supplied the northern Buddhist world with a Buddhicized version of the Hindu creation myth that could have supported the celebrations of a Tantric Buddhist indrābhi eka in Angkor. Indeed, I suggest there is direct evidence in Angkor for the ālandā version of the creation myth being followed by the Khmers.

3.2 A Tantric Buddhist indrieka

From what little we know of how a Southeast Asian indrābhi eka was undertaken, formal celebrations appear to have centred on re-enacting the Indic ocean churning myth. From an unusual source, we find an indrābhi eka recorded in the Bàyon – an informal inscription, which is possibly a roughly chiselled instruction by a master mason or a temple official on the northwest section of the outer gallery of the Bàyon: …the peace of the whole universe. Then the king retires into the forest at the time when he celebrates the holy indrābhi eka. This appears on the least finished section of the Bàyon outer gallery walls, which may explain why four such informal inscriptions are left here. The script is cursive and difficult to read or date, according to Coedès, but he was convinced they were ‘not simple graffiti without relation to the bas-reliefs.’ I am aware of attempts by some scholars to link the indrābhis eka inscription on the outer gallery to the Churning of the Ocean of Milk relief directly opposite on the inner gallery , but I would personally resist these on the grounds that Jayavarman VII seems to have developed a new, Buddhist version of the Churning myth, described below, whereas the Churning relief on the Bàyon inner gallery is a reversion, presumably ordered by Jayavarman VIII or another king who refurbished the Bàyon for Śaiva ritual, to the Hindu classical version involving Vis u , Indra and Śiva. The inner gallery relief is indeed fairly close in design to the version of the myth depicted in Angkor Wàt. Although the Devas and Asuras have changed sides, a flying Indra returns to the top of the spinning Mandara and the jar of amr ta carved in the inner Bàyon relief has just been sketched in an almost identical form on the unfinished Angkor Wàt relief just below the projecting fold of Visn u’s sampot .

Bernard Philippe Groslier saw the idea of an indrābhi eka linked to the ocean churning myth borne out in the causeways of giants outside the walls of Pra Khan, Angkor Thom and Bantéay Chmàr:

Let us say very summarily here that Neak Pean, associated with Preah Khan, marks for me the establishment of Jayavarman VII’s power at Angkor, his abhi eka, while the Bayon will be the sign of his Indrābh eka. This is manifested notably by the giants’ causeway, invented at Preah Khan as Stern determined, and which imitates the creationist churning of the new world.88

Groslier’s seeing the Bàyon as the sign of Jayavarman’s indrābhi eka is consistent with what architects Dumarçay and Royère describe as the festive decorative touches the master builders added for the occasion of the formal dedication, but also as the topping-out ceremony for a massive project that had been pulled through some major design changes: ‘Finally, to indicate that the objective was obtained, the moment when this culmination was reached is shown in the decoration of the monument.…In its final state, the monument appeared as though dressed overall for a festival, with garlands of leaves hanging from the cornice…’ The causeways of giants may also be associated with the festivities, but they cause puzzlement because they are seen as depicting an odd version of the Churning of the Ocean of Milk – one performed in the absence of the three key actors: Vis u , Śiva and Indra. The Khmers had for centuries reproduced an accurate, conventional version of the Churning myth on their temple walls, so the peculiarity in Jayavarman’s design must have been a deliberate move away from this Brahmanical version.

Jayavarman’s Churning has only two main actors, one of whom looks godly and the other demonic. This in fact provides a closer fit with a Tantric Buddhist version, which features a titanic battle between two antagonists – Vajrapān i and the demon leader āhu. Vajrapā i eventually wins the amr ta produced by the churning for the Buddhas and turns blue after being made to swallow the fluid contaminated by āhu.

āhu is punished by the Buddhas for defiling the am ta by being turned into a nineheaded monster -- and the leading causeway Asura indeed has nine heads. I propose therefore that we see the new Buddhist version of the myth as the one being staged on Jayavarman’s causeways, and go on from this to assume that the anointment and celebrations could have been led by Buddhists.

The Khmer version of the Indic aindra mahābhi eka most likely celebrated the 1203 subjugation of at least some of the neighbouring Čam kingdoms to direct Khmer rule, after many years of war. The re-establishment of Khmer control of at least some of the Čam kingdoms in the 120 may have been the trigger for Jayavarman’s indrābhi eka, particularly given the attachment he must have felt for the Čam city of Vijaya, where he apparently lived during his early adult years between the ages of about 25-40 15-35.


4. Devas, Asuras & Devatās

Continuing uncertainty about the deity of the towers is reflected in the work of a Japanese team which is at present stabilising the structure of the Bàyon. The team’s art historians measured many of the faces and noted minor variations in the shape of an ear or the width of a jaw. From these small differences they devised a scheme of three types of face: square-jawed Asuras, oval-faced Devas and slim-faced goddesses. The team put up a notice outside the temple stating that the face towers represent a broad coalition of Hindu and Buddhist demi-gods securing the city against the repetition of what they assumed to be a psychologically jarring invasion by the neighbouring Čams: ‘Devatā, Deva and Asuras are common to Buddhism and Hinduism. They are depicted on the Bàyon as symbols of all gods and goddesses in both celestial worlds.’ The team’s empirical approach leads them to propose the small differences in measurement betoken a range of deities of both sexes in a broad, pan-religious celestial warrior coalition in the city’s defence against another invasion.

The Bàyon faces do bear some resemblance to the faces of the Devas of the Angkor Thom causeways – though the causeway Devas have a more military, square-jawed and less regal aspect than the pacific tower deities – but they assuredly bear no resemblance to the heavily-jowled Asuras with angry, bulging eyes and deeply furrowed brows. Nor do the narrower faces and slightly more sloping eyes of a few of the tower deities display any femininity. But the more fundamental difficulty I have with this new taxonomy of the faces is that I cannot see three face types in the towers and find myself siding with Mus, who deems the degree of uniformity achieved in the sculpting of the faces as extraordinary, given the scale of the project and the fact that the faces were carved in situ with simple technology after the rectangular blocks of sandstone were hoisted into position.93 The Khmersmastery of the problems of carving sandstone on a gigantic scale is I think suitably acknowledged by Claude Jacques:

Given the difficulties associated with creating sculpture on a gigantic scale, it has to be acknowledged that the Khmers of this era created, at the Bayon temple and elsewhere, the most beautiful giant sculptures which, moreover, are imbued with a symbolic significance that makes them deeply spiritual. Moreover this uniformity of model and execution extends to the face towers of Bantéay Chmàr and Prea Khan of Kompong Svay and to the other temples of Jayavarman VII where face tower gopuras were inserted into the outer enclosure walls. There are minor differences in the execution of the faces, but they are indeed minor, accidental to the homogenous and carefully implemented single design idea. The intellectual model that the architects and sculptors of the Bàyon were projecting required just such visual homogeneity. It is the impressive sameness of the repeated faces that is powerful and eloquent, not the minor differences in physical execution.

The international context: the Buddhist crisis

A glance at the international context of Buddhism at the start of the 13th century may help us in our quest for the identity of the deity of the Bàyon towers. Buddhism in Northern India was in terminal crisis at this time, and the Buddhist court in Angkor must have been psychologically affected. Moslem general Muhammed ibn Bakhtiyar Ikhtiyar-ud-din Khilijii had stormed the monasteries of the Ganges valley in 1197 and butchered the monks. The theologians or mahāntas who were not killed on the spot, fled with texts and portable icons into the Himalayas. According to Tibet’s historian Lama Tāranātha (15 5-1634), a large group of the refugee Tantric masters from the monasteries managed to reach Southeast Asia. Tāranātha is looking back over four centuries here and some of the names of the Tantrikas’ destinations are unclear. But his knowledge of South and Southeast Asian geography had a solid base, acquired from his teacher Buddhagupta, who travelled in the 16th century through India and the southern seas and stayed in Arakan, Pagan and Haripunya.

In the 11th and 12th centuries, there had been raids by the Islamic armies based in the Sind, which had dismayed neighbouring Buddhist courts. Burmese king Kyanzittha’s last Prome inscription in 1105 expresses concern at destruction of the Vajrasāna temple, at the geographical heart of Buddhism. Tilman Frasch notes that ‘the considerable number of Indian names [in Pagan inscriptions] suggests an increasing human influx from India’ in the 12th century. He believes the stream of refugee monks from the Bihar monasteries may have extended the period during which Mahāyānist ideas flourished in Pagan. But the situation became grave in the late 1190s and Burma took actions to prop up Buddhism at this time: king Narapatisithu sent a mission to Bodhgayā. An early 1 th century text supports Tāranātha’s report of an exodus to other Buddhist centres. It records a local ruler of Bodhgayā supplying provisions for 1,000 monks from a Sinhalese monastery on the north side of the Bodhgayā compound, who were returning to Sri Lanka. Narapatisithu began building a replica of Bodhgayā’s Mahābodhi temple so that Pagan could fill the gap left by the loss of the enlightenment shrine and become the new centre of the Buddhist world. The project amounted to transferring the Buddhist holy land, and its representative temple, to Pagan whose own Mahābodhi ‘came to stand as a symbol for Pagan’s new role as a Buddhist world power.’104

But the stream of Buddhism heading north and northeast in the form of refugees from Vikramaśila and Odantapurī in 1200 was Tantric, and the Pagan court, which had by then begun founding a long-term Theravādin alliance with Sri Lanka, would have been circumspect about Tantric overtures from the refugee masters. In Angkor, however, the first Mahāyāna Khmer emperor may possibly have welcomed the support and authority of an influx of the leading exponents of Buddhist Tantrism. From the point of view of the survivors of the Moslem onslaught, a Mahāyānist ruler in Cambodia would have been a major magnet. ālandā, Vikramaśila and Odantapurī had been parts of the world centre of Tantric Buddhism for 600 years but were now laid waste and undefended by any earthly sponsor. More than anything, Buddhism needed wealthy patrons with armies. No shred of direct evidence has been found in Angkor for the arrival of Tantric mahāntas from India. I will shortly claim (‘The

Angkorian context’ below) that in the final years of Jayavarman’s reign, an overtly Tantric pantheon led by Hevajra, Vajrasattva, Vajradhara and Vajrapā i moved to centre stage, but Jayavarman had all the resources needed for this in the Phimai tradition from which his Mahīdhara dynasty hailed. evertheless, a learned, Sanskritist Tantric Buddhist court ruling over the largest empire of the region could have accommodated the mahāntas without difficulty. Moreover, there may be indirect evidence of their presence.

From Nepal to Angkor

The influence of the fugitive Bihar mahāntas or ajrācāryas in Angkor may be attested from afar in Kathmandu, where most scholars see huge pairs of eyes of a Tantric Ādi-Buddha peer out over the city from each side of the square harmikā of the Kathmandu’s vayambhū ahācaitya, and similar monuments in nearby Patan. B.P. Groslier, following Jean Filliozat, speculates that the painting of the eyes on the towers in Kathmandu and Patan was inspired by the refugee masters from Bengal, at exactly the time that giant four-fold faces were being carved on the towers of the Bàyon.

It has been shown recently105 that it [the new form of Buddhism of Jayavarman VII] very probably consisted of the doctrine elaborated in ālandā, then taught in Angkor – finally in Japan – by the doctors of the [[[Buddhist]]] law who had to flee before the Moslem invaders in the closing years of the 12th century. It is therefore to this school and its texts that we should turn for Jayavarman VII’s conceptions of Buddhism, and therefore for the sources of the Bàyon. This is just as much the case for the Bengali zealots who took refuge at the same time in Nepal and Kashmir, who were very probably the initiators of the stupas marked with four stylized faces, oriented to the four directions, which are the only exact parallels that can be found with the Khmer face towers. Lack of evidence makes it difficult to date the Nepalese stūpas, and more particularly the moment when the giant eyes were first painted on the harmikā; but there are reasons for proposing, as Groslier does, that an innovative rework of the stūpas occurred at the end of the 12th century, when the Newar Buddhists were reinforced by the Bihar masters and Arimalla (r. 1200-16) founded the first Malla Dynasty of Nepal. Because of the destruction of Buddhism in India, this period marked a Buddhist renaissance in the Kathmandu Valley, with a distinct focus on religious art innovation. This is how Ulrich von Schroeder imagines Kathmandu bustling under the influence of the immigrant ajrācāryas (‘vajra masters’) of 1200:

The annihilation of Indian Buddhism caused a great influx of refugees to Nepal who swarmed to the Buddhist monasteries of the Kathmandu Valley, mostly in Patan and Kathmandu…The arrival of these Buddhist refugees was beneficial to Nepal in many ways, one of them being that among these immigrants were many eminent Indian Buddhist scholars who had salvaged valuable manuscripts and probably also many cast images. There is every reason to believe that among these displaced Buddhists were also many skilled artists and craftsmen. At the same time the importance of the vihāras as centres of Buddhist studies increased and the Tibetan Buddhists shifted their focus from north-eastern India to Nepal.

Buddhist art in this period was in a unique phase in which one predominant style is diffused internationally – a phenomenon reinforced by Islam’s driving the Buddhists out of their Indian monastery-universities. Much of the northern Buddhist art of PālaSena India, Nepal, Tibet and Burma can hardly be differentiated at this time and scholars consequently define a ‘(Pāla) International Style’. Images of Tantric Buddhas, whether painted on cloth in Nepal and Tibet, or on temple walls in Pagan, have many iconographic and artistic features in common and convey a powerful, brooding conception of Buddhahood. These features include heavy, lowering eyes with the upper lid sinking over the iris and the whites fired below with a tinge of red. Above them a pronounced ūrnā swells between strong, arched eyebrows. The giant eyes of the vayambhū- ahācaitya (‘self-existent great monument’) fit within this convention. [Plate 9 Vajra ara of Svayam ū Ma acitya]

We will see in subsequent sections of this chapter that Khmer Buddhists in the 11th and 12th centuries made icons of the Hevajra mandala and were therefore influenced by the Vajrayānist practices radiating from Bihar just before the disastrous Islamic onslaught on the monasteries that developed them.. It may be worthwhile incurring the obvious risks of anachronism and cautiously enquiring whether there are practices that have survived into modern times among the Newar ajrācāryas, which throw light on what we know of the 12th century Buddhist cult of the Khmers. Recent research on the unique Nepali experience of Buddhism has illuminated the beliefs and rituals of the ajrācāryas in ways that may suggest traces, conserved through time, of features that the two geographically-separated cultures once shared. The Bàyon and the Nepalese caityas are the only religious monuments in the world that share the architectural innovation of four giant faces of a deity peering out in the four directions over a city. In epal, the Svayambhū Mahācaitya is seen by several scholars as a monumental representation of the Ādi-Buddha, whose eyes are painted on the sides of high, square, metal-covered harmikā tower above a stūpa-shaped dome. I will later suggest that the same ancient design idea of the body of a stūpa or temple being the living Buddha present in the community is built into the Bàyon. Maxwell and Snellgrove identify this epalese Ādi-Buddha as Vajradhara; Brian Hodgson inclined towards ‘Vajra Satwa’ and saw the eyes as representing omniscience. Hodgson, a long-term resident of Kathmandu, no doubt gleaned this from his Newar ajrācārya acquaintances, whose daily guru ma ala ritual is devoted to Vajrasattva. David Gellner more recently recorded Vajrasattva as ‘both the guru of the Vajrācārya priests and an exoteric deity who is a kind of representation of the absolute in

Vajrayāna Buddhism.’ The interchangeability of the two names for the ewar’s supreme Buddha is made plain in Gellner’s citation of Vajrācārya Hemraj Sakya, who says Vajrasattva is ‘Vajrayāna’s main guru, also known as Vajradhara, he is worshipped as Adibuddha Svayambhū.’ The five Buddhas of the Vajradhatu Pentad appear on high pentagonal metal panels which rise like a crown above the eyes of the supreme Buddha of the Svayambhū. On a massive scale, this seems to be similar in design to Vajrasattva icons with the Pentad of Jinas in the hair or crown that is known in eastern India from the seventh to 10th centuries. The closest parallel in world religious architecture for thus representing the supreme deity of the state in a giant face overlooking the capital is in the much larger stone face towers which may have been constructed simultaneously in the Bàyon, as Groslier remarked. In both cases the whole building becomes the god, an innovation that blurs the traditional Indic dividing line between architecture and sculpture.

During the 10th to 12th centuries Kathmandu, Angkor and Phimai (as we will shortly see) embraced fierce deities such as Hevajra and Sam 118 Gellner describes the ewar Buddhists’ daily vara, who appear in the Yoginī guru ma ala ritual class of Tantras. as focused on the invocation and visualization of Vajrasattva, the lustration of his ma dala and the recitation of his renowned 100-syllable mantra (śatāk ara). Gellner gives an example of a ma ala ritual using an icon in this way on day five of the secret initiation ceremony for ewar Vajrācāryas: ‘[Cakrasamvara’s] statue is placed in the middle where his mandala is to be made up, and the Vajrācāryas begin making it with coloured powder.’119 Later we will consider the place of the supreme deities Vajrasattva and Samvara among the Khmers.

Both of these Buddhist cultures shared liberal admixtures of Śaivism and apparently performed animal sacrifices. Gellner records that Newar Buddhist priests ( ajrācārya) oversee ‘Spirit-Offerings’ (bali) which include rites involving blood, alcohol and buffalo entrails120 and we find a buffalo tethered for sacrifice on the eastern outer gallery of the Bàyon (such a ritual is of course not only found in Nepal and Cambodia)121. Gellner’s ajrācārya informants indicate that buffalo sacrifice is allowed in Buddhist rites for fierce deities: ‘It is significant that while Spirit-Offerings are never made to Buddhas or other high, pure divinities, they are made to Tantric gods such a Cakrasam vara and Vajravārāhī. As fierce gods, they can receive such offerings, which are not acceptable to Buddhas.’122

Both cultures crowned their Buddhas from the 11th century. Paul Mus’ study of crowned Buddhas concluded that although separate crowns and jewels were perhaps first draped over Śākyamuni images in Bodhgayā in celebration of the Lotus Sutra123, the carving and casting of crowned Buddhas began in Pāla India and was driven to Nepal by the advance of Islam:

[The royal Buddha] appears to be a response to secret practices. We can be certain of this when we consider that we see, under the Pālas in the convents of Bihar, the simultaneous spread of the doctrine of the Vajrācāryas and of the images of the crowned Buddha. And indeed, when we find these Vajrācāryas later chased from India into epal and Tibet, we find

in Kathmandu: ‘The main bulk of the stūpa was then regarded as the symbol of the Ādi-Buddha. Along with this metaphysical elaboration, which transformed the “horizontal” construct of the Pentad into a “vertical’” hierarchy, went elaboration of the personifications of these two new super-additions to the Vajrayānist cosmic structure. The Ādi-Buddha concept was specially embodied in a masculine being named Vajradhara, Wielder-of-the-Thunderbolt…’ T.S. Maxwell (199 :164). Snellgrove records the supreme Buddha as represented in the whole body of the Svayambhū structure: ‘The fifth, allcomprehending Buddha is sometimes shown on the great stūpas of epal…but there is really no need to show him, for the whole stūpa is his symbol.’ Snellgrove (195 :96). Dagens says ‘the idea of the temple being the body of the god is quite frequent in Indian thought…’ (Dagens 2000:10 ) 118 Sa vara is known to be in the Mahīdhara pantheon from his image with the distinctive elephanthide cape, on one of the principal internal lintels of the central sanctuary in Phimai. Hevajra, another Heruka deity with eight heads and 16 arms, is known from a large series of Khmer bronze and stone icons.

119 Gellner (1992:274)

120 Gellner (1992:149)

121 Michael Vickery informs me that for example in Chiang Mai, offerings of blood, alcohol and raw buffalo meat are part of a ceremony for the Buddha. Buffalo sacrifices also continued into modern Cambodian times.

122 Gellner (1992:149)

123 As attested by Hsüan-tsang in his travels in the seventh century.

they have in their hands images of crowned Buddhas, which play an essential role in their beliefs, and are still today the object of a secret cult (Vajradhara-Vajrasattva).

The crowned Buddhas carried to Nepal were contemporary with the first crowned Buddhas made in the Khmer empire. Khmer standing crowned Buddhas and regal Buddhas enthroned on āgas first appear in the 11th century when Phimai was flourishing. Woodward too links the 11th century Khmer Buddhas in royal attire with the esoteric practices developed ‘[i]n northern India, in the decades around 1000, [in which] the crowned Buddha became very popular…’ Although von Schroeder records one ninth century Nepalese bronze Vajrasattva, crowned Buddhas otherwise appear in his account of epal’s material record from the 11th century and are coeval with those of Phimai and Angkorincluding a Khmer finial of Vajrasattva described thus by Mus:

Finally a crowned Buddha figure, identical to the Buddha of the triad and seated on the nāga, reappears above a beautiful Khmer image of Vajrasattva and can only here be a representation of the Ādi-Buddha from which Vajrasattva emanates…We are therefore led to reserve an interval, out of precaution, between the Khmer inscriptions which date roughly from the 10th century [Bàt Čum etc.] and the figure of the Ādi-Buddha as Vajrasattva, which is not dated but may well be from the 12th or 13th centuries. In effect, Vajrasattva ranks only among the very developed forms of the Mahāyāna.128 In 1910 Henry Parmentier tried to find a direct connection between the temple architecture of Bengal and Angkor when he claimed that Chinese pilgrim I-ching saw early prototypes of the giant face towers in ālandā in the late seventh century. Filliozat reopened the question after he translated Tāranātha’s Tibetan text that recorded the flight of Bihar theologians. Bruno Dagens later discounted

Parmentier’s attempt to find a direct architectural link between ālandā and Angkor by showing that the Chinese phrase ‘heads of a man of natural size’ had been mistranslated as ‘heads the size of a man’. Dagens concluded that I-ching was describing conventional Indian temple ku us with human-size faces looking down from false windows, not giant faces as tall as a man’s body. So there was no direct architectural influence. The only observation of value we can now make is that in two remote geographic locations, there arose conceptually close instances of a major Buddhist architectural innovation where Tantric theologians exited the Islamic occupation of Bihar to settle in Nepal, Tibet and Southeast Asia to build new bases for the preservation of Buddhism.

The Angkorian context


Against the background of the international crisis of Buddhism, we now turn to ask what theology and cosmology were being projected by the Bàyon. In the years 11971207, when all the Ganges Valley monasteries were being destroyed, the Khmers were carving face towers in the Buddhist Bàyon, Bantéay Chmàr and other temples and the royal workshops were producing stone and bronze icons of Tantric deities such as Hevajra, Vajradhara and Vajrapā i – all progeny of ālandā, Vikramaśila and the other monasteries. Indeed, from the number of surviving bronze icons , it could be said that Hevajra, the wrathful form of the supreme, formless Buddha Vajrasattva (‘Adamantine being’), may have reached his world apogee in Cambodia. The abundant iconic evidence of Hevajra suggests a close link between the king and this baroque deity with eight towering, addorsed heads, 16 arms and four legs. Groslier, indeed, exhumed a finely cast, 22 cm gilt bronze dancing Hevajra in the elaborate late Bàyon style at his excavation site in the royal palace in Angkor Thom in 1952-3. [Plate 10 Hevajra excavated from the royal palace, Angkor Thom] In Tantric thinking, a king’s personal meditation in discovering the Buddhas inside himself generates a mandala of deities, not only for himself but for the whole state.136 I will conclude this chapter by saying that the king’s meditated mandala is projected in the architecture of the Bàyon.

I have argued elsewhere that, following Stern’s characterisation of the previous phase as ‘Lokeçvarisation’, we could describe the third and final phase of Jayavarman’s Buddhism as ‘Yoginīfication’, reflecting a return to a broader pantheon resembling that of Phimai Tantrism, and further indebted to the Yoginī Tantras of the northern Indian monasteries. My contention is that the so-called ‘Apsaras’ or celestial dancers that cluster around the entrances to the Bàyon (originally, 6250 goddesses were carved on the pillars and gopuras, according to my estimate from their positioning in gopura friezes and on entrance pillars) are in fact Tantric: they are  

Yoginīs or the projections of a Yoginī-Hevajra cult; the same emanations of Hevajra that appear in late ‘Bàyon style’ bronzes, or accompany the Tantric deities of the central sanctuary at Phimai. The large friezes of dancing Yoginīs enwrapping the Bàyon and Jayavarman’s other temples are the public projection of the third phase of the king’s state Buddhism, rooted in the Yoginī Tantras.

This interpretation is based on a close analysis of the postures and attitudes of similar deities in the Khmer iconographic tradition, especially in those found in the Bàyon, Angkor Wàt and Phimai. The ardhaparyanka (half cross-legged) pose, with one knee bent and the other leg retracted to touch the thigh of the other, and the open-eyed, unsmiling stare of these challenging goddesses indicate kinship with the Yoginīs. Their iconography separates them clearly from the demure apsaras that fly above the Devas and Asuras of the Churning of the Ocean of Milk relief in Angkor Wàt. The emphatic use of this goddess motif in the Bàyon, where they dominate all the entrances, is unprecedented in Khmer temple decoration. Moreover, large halls, called simply ‘salles aux danseuses’ by French scholars, and adorned with similar goddesses, were inserted late into the king’s other temples (the insertion was on a particularly impressive scale at Bantéay Chmàr) in an update package that suggests a late cultic shift. A further sign of a Hevajra-centred cult is the large number of lustration conches bearing the embossed image of Hevajra in the museum collections. The goddesses found in such prominent proliferation on the temple walls do not bear the specific attributes of named Yoginīs, but they hold flower garlands, which Yoginīs give to initiates in Hevajra consecrations. Khmer ritual bronzes of Hevajra, in the late Bàyon style, include some with eight Yoginīs twirling in ardhaparyanka around the deity, exactly as they are described in the hevajra tantra. These ascetic Yoginīs play prominent roles as intercessors in the Tantra and an abundance of female deities would be appropriate to a mature Vajrayānist Hevajra cult. The hevajra-tantra was the first of a new class of 'Mother' Tantras that gave a strong female orientation to its mandalas. Apart from this remarkable change in temple decoration, there is one piece of contemporary evidence from a Chinese source which suggests Jayavarman’s temples were known for the special focus they gave to female officiants. The 1225 chronicle of Chau Ju-kua, the Chinese Superintendent of Maritime Trade in Canton, was based on hearsay rather than visits to the countries whose goods he taxed at the northern end of the maritime trade route, but his testimony contains a unique indication of what was taking place in Jayavarman’s temples:

‘[In Chen-la, i.e. Cambodia] the people are devout Buddhists. In the temples there are 300 foreign women; they dance and offer food to the Buddha. They are called a-nan [Skt. ānanda (bliss)].…The incantations of the Buddhist and Taoist [[[Śaiva]] yogin] priests have magical powers.’ Calling the women ‘blisses’ probably has an origin in ritual. The hevajra-tantra, for example, defines the four ‘blisses’ that an adept strives for in executing a four-stage system of meditational and yogic-sexual initiations, the highest of which is said to be ‘free of passion and nonpassion.’141 Yogic self-consecration (svādhi hānakrama) was the motor of the Tantric vehicle for achieving the highest realisation through the psychic power of erotic experience. This is Davidson’s account of the four blisses: With the [third] insight/gnosis consecration, the disciple copulated with the female partner under the master’s tutelage. This sexual act is for the purpose of obtaining the proper method of the ‘centers of the internal ma ala’ (ma alacakra) to realize the four joys (ānanda) during the four moments of ritual orgasm. Finally, the fourth consecration varied considerably, involving either a symbolic revelation in a highly-charged charismatic environment or, more frequently, lengthy instruction about the nature of reality in which the experiences of the previous two sexual consecrations were to be integrated into a larger Buddhist philosophical context.142

In the absence of any other account of the ritual or liturgy practised under Jayavarman’s Buddhism, Chau’s brief chronicled report can be taken to reinforce the evidence of the Hevajra-dedicated ritual paraphernalia as indicating that a classical mature Tantric system of four consecrations was deployed in Jayavarman’s temples. The larger iconic record of the period also points to a royal cult of Hevajra, the Yoginīs, and Vajrasattva. The Khmer ritual bronzes of the complete Hevajra ma ala in the Phnom Penh, Sydney and Guimet collections, with the central deity surrounded by eight dancing Yoginīs, leave no possible doubt that the highly influential hevajratantra reached Angkor.

Hevajra or Vajrasattva or Vajradhara?


It seems prudent, in trying to identify the deities of the Bàyon, to limit the descriptions of Vajrayānist deities to those contained in the two seminal Indic texts we know were present in Angkor. The STTS, the root Yoga Tantra, which I have argued elsewhere reached Angkor in the 10th century143, features Vajrasattva as a

141 Elizabeth English (2002:92) ajrayogin , her visualizations, rituals and orms Wisdom Publications, Boston.

142 Davidson 2002:198

143 I have argued in a forthcoming article in the Journal of the Sanskrit Studies Centre, Bangkok that the root Tantra of the Yoga class, the sarvatathāgatatattvasa graha (STTS) and its ninth century commentary by Śākyamitra are identified as being brought into Cambodia by Buddhist guru Kīrtipandita in the Wàt Sithor inscription (K.111). Coedès published the Romanized Sanskrit of the key stanza which establishes the presence of the Tantras in Cambodia as: B7-8 laksagran t ham abhiprajñam yo nves ya pararāst rata tattvasaògrahat īkādi- tantrañ cādhyāpayad yamī // and translated it as:

Ayant recherché en pays étranger une foule de livres philosophiques et les traités tels que le commentaire du Tattvasangraha, ce sage en répandit l’étude. [Having searched in a foreign country for a great number of philosophical books and treatises, such as the Tattvasangraha commentary, this sage then spread the study of them].

Whereas a new translation by Tadeusz Skorupski of SOAS, published in this article, gives: Having searched from/in a foreign kingdom for one hundred thousand books of higher wisdom, and for the tattvasangraha -t kā and Tantra(s), the self-restrained one [[[Wikipedia:Sage (sophos|sage]]] taught (them).

Coedès mistook Śākyamitra’s tattvasangraha -t kā commentary on the Tantra for Kamalaśīla’s tattvasangraha- pañjikā, a text book on classical Mahāyānist doctrines, and translated tantran as

Bodhisattva. The text opens with Vairocana as the presiding Buddha, but gradually shows Vajrapā i, the master of ceremonies throughout, as his equal or possibly more than equal.144 Vajrasattva enters as the first of the 16 Vajra Beings145 of the Vajradhātu ma ala, the pre-eminent ma ala of the Yoga Tantras, and the Bodhisattva who heads the Tathāgata family. Vajrapān i’s celebrated subjug ation of

Maheśvara in the STTS starts a process which Paul Williams calls the ‘vajra-isation of

Buddhism,’ a complex and determinant shift towards the mature Vajrayāna.146

Another seed for change can be seen in the way Vajrapā i is at times called Vajrasattva, ‘adamantine being’. Snellgrove suspects this heralds the germination of Vajrasattva as the later Vajrayāna’s ‘sixth, utterly supreme Buddha’: Vajrapā i’s personal triumph as a ‘mere yaksa’ who is recognized as a Bodhisattva , then as the most powerful of all Bodhisattvas in that he wields the vajra, and finally his identity as Vajra-Being (vajrasattva) when he becomes the expression of perfect enlightenment, as conceived in tantric tradition…147 In the hevajra-tantra, one of the principal Yogin -tantras of the mature Vajrayāna, the transformation is complete and ‘Vajrasattva’ has become the personification of absolute being in the vajradhātu or ‘adamantine sphere’ from which the universe emanates.148 Hevajra, the lord of the man ala defined in the Tantr a, is a fiery, wrathful form of this absolute being, as we hear in the opening lines: There the Lord pronounced these words: ‘Do thou listen to this which is named Hevajra, the essence of Vajrasattva, Mahāsattva and Mahāsamayasattva…It is indivisible and so known as Vajra. A Being which is unity of three; because of this device he is known as Vajrasattva.’149 The Buddha has become something fearful, supreme and cosmic, as Snellgrove notes:

Moreover the Lord (Bhagavan) is no longer Śākyamuni in any of his recognizable hypostases, but a fearful being with the name of Śambara, Vajra āka, Heruka, Hevajra or or Ca amahāro a a (‘Fierce and Greatly Wrathful’).150

‘treatises’ instead of ‘Tantras’. He therefore saw only classical Mahāyānist doctrinal works arriving, instead of what is clearly identified as the leading Tantra of the Yoga class and its major commentary. 144 ‘It is scarcely necessary to observe at this stage that Vajrapāni may appear as the equal of Vairocana (the various eulogies make this quite clear), as the chief divinity in his own right…’ Snellgrove, D. (1981:42) ‘Introduction’ sarva-tathāgata-tattva-sa graha: Facsimile reproduction of a Tenth Century Sanskrit Manuscript from Nepal edited by Lokesh Chandra and David L. Snellgrove pp. 5-67 New Delhi. This may be seen as presaging his evolution into the sixth supreme Buddha Vajradhara. 145 This is the form in which Vajrasattva appears in the texts of the Tantric Shingon sect of Japan and which we will see later is the deity that Boisselier and Woodward see being carved in the Bàyon face towers in its original Mahāyānist plan. Woodward’s case for seeing a reference to the 16 Vajra Beings in the original 16 Bàyon tower sanctuaries, by Parmentier’s count, is however indirectly countered by Cunin, in this volume [Cunin typescript p.16] who concludes that 24, not 16, face towers were simultaneously constructed in the first phase of building.

146 Williams, P. with Anthony Tribe (2000:218) Buddhist thought Routledge, London. Williams adds: ‘The role of the vajra as the core symbol in tantric Buddhism continues for the remainder of its history in India, vajra names being characteristic of both Mahāyoga and Yoginī Tantra deities.’ 147 Snellgrove, D. (1987:136) Indo-Tibetan Buddhism: Indian Buddhists and their Tibetan successors 2 vols Shambala, Boston; reprint 2004 Orchid Press, Bangkok

148 Snellgrove comments: ‘Such then is the complex mystery at the heart of nirvān a and sam sāra. It is this that is referred to as the Body, Speech and Mind of all the Buddhas, as the Three Adamantine Ones (trayo vajrina ), as the unity of three states of being (tribhavasyaikatā) and may well be indicated by any agreed name. It pervades all things for there is nothing other than it, and yet transcends all things for it is not involved in their accidental or purely unreal defilement. It can be experienced only by learning to associate oneself with its true nature, which is identical with one’s own nature, and so on.’ Snellgrove (1959:28)

149 Hevajra-tantra I.i.2-3

Rob Linrothe also remarks the near absence in the hevajra-tantra of a conventional Mahāyānist Buddha or Bodhisattva: Buddha and bodhisattva have minor roles in the Hevajra Tantra. The chief interlocutor is Vajrasattva, considered to be a kind of ‘Ādibuddha’ whose ‘form comprises all the Buddhas’ and who is ultimately identical with Hevajra himself. Vajrasattva answers the questions of the Vajra-being or bodhisattva, Vajragarbha, and of yoginī (female adepts).

The unifying Buddhist conception of Ādi-Buddha -- primordial Buddha or first mover -- was defined for western scholarship in Tibet in the early 19th century by the remarkable explorer and scholar Alexander Csoma de Körös. The term Ādi-Buddha is in fact not used to describe Vajrasattva in the hevajra-tantra, although the epithet ādi – ‘first, original, primordial’ – is added to the name of the fifth Buddha Vairocana as ‘ airocanādi’. For Vajrasattva to receive the primordial epithet, we have to await the late eighth century sa varodaya-tantra155, but most of the attributes, such as being ontologically omnipresent are already there in the hevajra-tantra, which calls Vajrasattva ‘the one unity of all that is’ and shows him as a sixth Buddha whose family embraces all those of the Vajradhātu Pentad – Vairocana and the four directional Jinas.

In his personified form in icons, Vajrasattva is seated in the lotus position with a vajra held in his right hand, as if issuing from his breast, and a gha tā in his left, resting on his hip. This mudrā is found in several bronzes of Phimai and Angkor. In the texts, Vajrasattva and Hevajra can be difficult to separate, but in icons they can never be confused.

In the text, Hevajra is described either in his ardhaparya ka (‘half cross-legged’) dance posture surrounded by eight twirling Yoginīs, or braced in yab yum sexual union with consort airātmyā, who represents non-duality or the doctrine of the nonexistence of self. Yet according to the text even in this maithuna posture of coition ‘the place of this union is the seat of Vajrasattva’.158 Their differences are, however, made clear. Hevajra is portrayed as brimming with life, colour and emotion and

promises that through his Yoginīs he will act as an intercessor between the adept or sādhaka and ultimate reality; Vajrasattva, on the other hand, is a more remote, pacific, perhaps androgynous being or even state, which manifests variously at key moments in the rituals. Making contact with ultimate reality, in order to experience reality as non-dual, is seen to require the blurring of some distinctions and the shattering of inculcated dichotomies, such as subject/object, pure/impure. This can be achieved when an invocation of one aspect turns into an invocation of the other: O Vajrasattva…He of all eyes, the vajra-eyed, makes manifest, see sublimity, Hevajra! Indeed, ‘He-vajra’ is a form of address (‘hail, vajra!’) both for what the wordvajra’ (‘adamantine’, etc.) represents and for the essence of vajra (Vajrasattva). Tucci describes this absolute being of pure abstraction as ‘the supreme Tathāgata, the germinal point outside time and space, from which issue the five directions of the cosmogram, symbolized by the five Buddhas…’ Etienne Lamotte selects this description of Vajrasattva from the Jñānasiddhi: ...the unique eye of knowledge, immaculate, knowledge incarnate, Tathāgata, undivided, omnipresent, immanent, subtle seed, exempt from impurities. From this theological and cosmological complex, two main possibilities present themselves as to what is being projected in the art of the Bàyon. One option is to see the temple as a unique and extravagant extension of the Hevajra ma ala with the dancing god represented in the face towers surrounded by Yoginīs of the gopuras and pillars.

From the large number of bronze statues, libation conches and other ritual paraphernalia in the material record, as well as large Hevajra images in stone, it would appear that Hevajra consecrations formed a significant part of the royal cult in this period. During the Lokeśvara phase, Woodward is surely right to see ‘Phimai Tantrism playing a secondary role.’ But if we see Jayavarman’s Buddhism and sacred art steadily evolving from an early triad, through ‘Lokeçvarisation’ to a late, royal cult derived from the Yoginī Tantras, there is no need to sideline the Hevajra ritual conches; they belong to the mainstream of the third phase of the new state religion, when Hevajra consecrations may have been conducted on a significant scale. A state programme of Hevajra consecrations may be further manifested in the architectural friezes of Yoginīs carved in and around the Bàyon gopuras and cut into the pillars standing before the outer gallery walls. These Yoginīs number 6,250 in the Bàyon alone, according to my calculation of the makeup of the entrances and surrounding pillars. If we take up Groslier’s intuition to seek a ālandā text to provide the key to unlocking the mysteries of the Bàyon, then from the material record, the hevajra-tantra has the strongest claim. There are indications of other Tantras being adapted to Khmer Buddhist use, but signs of the presence of the hevajra-tantra are manifold and incontrovertible.

The second option is to see a Hevajra cult expressed in the Bàyon architecture but with the unifying form of Vajrasattva appearing in architectural-iconic form in the face towers. The giant all-seeing faces which address the four cardinal directions high above the city would then depict the supreme, formlessadamantine being’ (Vajrasattva) of which Hevajra is a wrathful emanation. This option would be consistent with a royal cult in the Bàyon focused on Hevajra initiations. Which option is more compelling?

The question addresses the form of the deity because in essence the candidates are aspects of the same. Disentangling the claimants however requires careful navigation. Paul Wheatley for example states that Vajradhara (‘bearer of the vajra’) is in the Bàyon towers; yet he remains very close to Coedès’ solution, for he also sees a portrait of the king and Lokeśvara there, with the latter manifesting himself in the form of the Bodhisattva Vajradhara, an aspect of Vajrapā i assumed by Lokeśvara when expounding the Buddhist law.165 The complexity of this analysis reminds us of the crosscurrents of multiple interpretations we are navigating when we attempt to resolve the enigma of the Bayon faces. Boisselier, for example, before eventually turning to see Brahma in the faces, arrived at an even denser composite than Wheatley: he said we see the form of Vajrapā i in the faces, but this is really Lokeśvara, who is an aspect of Vajradhara and of Vajrasattva and of Vajradharma. Complex overlapping roles and definitions in a constantly enlarging pantheon were one of the outcomes of the bewildering creativity of the sages of Vajrayāna as it spread through many countries over many centuries and with no centralising hierarchy or authority.

However, thanks to the prolific evidence for a cult of Hevajra, we can justify circumscribing the list of candidates for the supreme Khmer deity portrayed in the Bàyon towers to the three who appear in the hevajra-tantra. Is it Hevajra, the fierce Heruka form and central figure of the Yoginī cult, whose image was also excavated from the ruins of Jayavarman’s palace? Is it Vajradhara, identified by one of the great medieval commentators on this Tantra, and as seen by several modern interpreters of the contemporary eye towers of Nepal; or is it Vajrasattva, named by the other medieval commentator as the overarching being of the Vajrayāna cosmos? Here are arguments for each.

Hevajra

1. The thousands of Yoginīs embellishing the entrances to the Bàyon make it conceivable that Hevajra is carved above them in the towers of the state temple. The presence of a large number of Hevajras cast in bronze in this period would tend to support this. Moreover, two large stone statues of Hevajra were found outside the east and west gates of Jayavarman VII’s fortified city, Angkor Thom.

2. Hevajra and his circle of eight Yoginīs are given the central role in a belief system recorded in several Khmer votive tablets. One in the Lopburi Museum, known as the ‘Trailokyavijaya’ tablet, in that it seems to show three linked worlds, has Hevajra and his Yoginis dancing within the pillars demarcating the principal level of the pavilion or palace ( āgāra). At the summit, above rows of Buddhas, is the āga Buddha with two devotees in the anjali posture. This tablet evinces the centrality of the Hevajra concept in Jayavarman’s Angkor. Hevajra also appears seated at the centre of a smaller votive tablet, now in the Bangkok Museum, where he again supports the āga Buddha.

3. In design, the individual faces in the face towers share many characteristics with the heads addorsed in triple tiers in Hevajra icons – piercing open eyes, third vajra eyes, and heavy, smiling lips. Hevajra’s regalia are also similar to that indicated on the face towers.

However, as noted earlier, Hevajra and Vajrasattva, although inseparable conceptually, are unmistakable in iconic form. From the perhaps 100 Khmer icons of Hevajra known in public and private collections, all have the same tower of eight addorsed heads, as specified in the hevajra tantra. There is some variation in the number of arms between16 and 20, but there is no variation in the number of heads. The iconography of Hevajra’s heads is so consistently deployed in the sacred art of the Khmers as to make it a sine qua non for portraying Hevajra. This feature alone is enough to eliminate an identification of Hevajra with the four-faced deity of the Bàyon towers. We are now left with the entwined Vajradhara/Vajrasattva option.

Vajrasattva/Vajradhara?

1. The most striking link I have found between the hevajra-tantra and the appearance of the Bàyon temple, presents itself when the architecture is compared with the insight and vision that results from ‘concentrated meditation’ (samāhitayoga) on Hevajra. The meditator first imagines a shiny black hook emerging from his breast to ‘draw in the Buddhas who are stationed throughout the threefold world’; having venerated them with Hevajra’s eight Yoginīs, the Buddhas are asked to consecrate the sādhaka. A shower of flowers rains down, drums beat and ‘vajra songs’ are sung as ‘Heruka will be revealed in you’. This, according to the Tantra, initiates a process of gradual fusion between the initiand and the deity: Performing morning, noon and night this meditation, which bestows such power, you should arise and at all times remain consubstantiated with the divinity. Later in the Tantra, the concluding moment of the empowerment ritual describes how the meditator conceives of utpattikrama, the ‘process of divine emanation’ which generates the universe and herself absorbs this back internally in utpannakrama (‘the process of perfection’). The apparent duality of existence implicit in two processes of emanation and absorption is then perceived by the meditator as an essential unity and a vision emerges of ‘the ma ala [which] appears from continuous application to the practice.’ This is presented as the climactic deity visualization of the text: The great bliss, such as one knows it in the consecrations of the Great Symbol, of that the man d ala is the full and efficacious expression, for nowhere else does it have its origin…This bliss is Wisdom, this bliss is Means, and likewise it is their union. It is existence, it is nonexistence, and it is Vajrasattva.

This hevajra-tantra vision of a ma ala of deities emerging three-dimensionally into space on a vast scale provides an ‘intellectual “model”’ of the kind we have been seeking in the design of the Bàyon. Indeed, Snellgrove’s account of the meditator’s final ma alic vision seems to uncannily echo the Bàyon’s architecture:

One may envisage the final stage of emanation of the ma ala of sixteen divinities, or one may envisage it as countless forms of Hevajra filling space in a regularized order in every direction. Then one must realize the identity of oneself with the whole process by associating one’s personality with the emanation, which is achieved by a strenuous effort of belief: OM HERUKA-svabhāvātmako ‘HA – ‘I am of the essence of Heruka.’ The imagined forms are conceived as sinking into one’s own heart, and from here the process may be repeated, so that

one becomes oneself the twofold process of emanation and absorption, of (My emphasis). sam sāra and nirvān a . We may indeed envisage the architect of the Bàyon as seeking architectural expression for the ma ala of 16 Vajra Beings or of the ‘countless forms of Hevajra filling space in a regularized order in every direction’ – in the form of the empowering state ma ala of Jayavarman VII. But as we have seen, Khmer iconography excludes Hevajra’s face from the towers, and as noted in footnote 1 8, the case for seeing the 16 Vajra Beings of the earlier STTS and the pantheon of

Japan’s Shingon sect (in which Vajrasattva heads the Tathāgata lineage) appears weakened by its dependence on the older Tantra and by Cunin’s determination (in this volume) that the original plan of the Bàyon included 24 sanctuary towers, not 16 as Parmentier had believed.

So what case can be made for seeing the hevajra-tantra’s version of supreme Buddhahood -- Vajrasattva or Vajradhara – being given plastic form in the Bàyon sanctuary towers? The multiple, pacific faces on the Bàyon towers and the sublime effect of their smile, suspended in stone above the city, have some natural affinity with the notion of a Vajradhara or Vajrasattva as the ultimate being of the Vajrayāna cosmos. Maxwell’s admirable analysis of the architectural concept in epal is that the vertical dimension brought by the tower with eyes represents the sixth Buddha, hierarchically and physically elevated above the ‘horizontal construct’ of the Vajradhātu Pentad, which encloses the original stūpa below. In this volume (Maxwell typescript page 28), Maxwell rightly sees the same concept at work in the Bàyon architecture: ‘In this interpretation, Jayavarman’s face-towers would have represented the Buddhist universe at the highest level, while lesser deities – dominated by the colossal meditating Buddha located on the central axis inside the central massif – were worshipped in the interior spaces below them, incorporated into the labyrinthine body of the architectural mass’.

Separating Vajrasattva from Vajradhara is also problematic. The names of the Vajrayāna’s sixth Buddha oscillate to the point of being fully interchangeable in Vajrayāna texts, as they are in the current usage of epal’s ewar Vajrācāryas. ‘Oscillations’ is Tucci’s term for the interchangeability of Vajradhara, Vajrasattva and Mahavairocana:

...oscillations are frequent and the supreme Tathāgata, the germinal point, outside time and space, from which issue the five directions of the cosmogram, symbolized by the five Buddhas, may assume the name of Mahāvairocana, Vajradhara, Vajrasattva, etc. Snellgrove acknowledges ‘the impossibility of making any final distinction between Vajradhara (Holder of the Vajra) and Vajrasattva (Vajra-Being), for both represent buddhahood in its adamantine aspect’: [The vajra-yāna’s] sphere of utmost potency is the vajra-dhātu-ma ala, and whoever rightly occupies this sphere is consubstantiated in Vajra-sattva. It is possible to force some distinction between Vajra-dhara, the idealised personification of the possessor of the power, who thereby becomes implicitly the supreme being, and Vajra-sattva, who is the properly consecrated being, but of course the two are never properly distinguishable, just because their selfidentification is the whole purpose of the rite. But he notes that the two deities are ‘distinguishable iconographically’, a point we will shortly take up.

2. Four faces are frequent in images of the fifth Buddha Vairocana, who turns the wheel of the Buddhist law, and are also known in Southeast Asia and China in images of the sixth Buddha Vajradhara or Vajrasattva.

3. The large open lotuses that crown the jatās of the heads in the Bàyon towers never appear above Hevajra’s tower of heads and so they too tend to argue for seeing Vajradhara or Vajrasattva. Jean Boisselier at one point thought the Bàyon represented the cosmic mountain with Vajrasattva manifested continuously at its summit. In this interpretation he saw the lotuses crowning the face towers as symbols of the eternal flame of the primordial Ādi-Buddha. In the ku alin yoga that plays a major role in Vajrayāna, the force driving towards enlightenment (bodhicitta) is conceived as a fluid in the body which rises during skilful meditation to engage directly with higher spheres of reality and achieve gradual transformation into Buddhahood. The fluid blazes up through nodes or cakras in imagined channels beside the spine. The nerve nodes are conceived as being in the perineum and navel, the heart, the throat and the crown of the head, which correspond with the spheres in which Buddhas exist and the sahasrārapadma or u sa-kamala (‘1,000-petalled lotus on top of the head’). The large lotus topped with a cone above the face tower deities may then be a rendering of the ultimate state aspired to by adepts of Vajrayāna in which they merge into the supreme Buddha. In Nepal, what Boisselier here identifies as a cone is called the bindu – ‘the emergent essence; a drop of radiance that flows upwards from the central yogic channel at the moment of enlightenment’ – and is another architectural feature shared by the Bàyon and the epalese stūpas. The geographically remote but possibly contemporary architectural parallels between the eye towers of Nepal and the face towers of Angkor, perhaps reflecting common notions brought by the Mahāntas fleeing the advance of the Moslem army into Magadha, may be synchronic circumstantial evidence that weighs on the side of seeing the Vajradhara in the Bàyon.

Fortunately the Khmer context makes a special local use of one of these deitiesVajradhara – which enables us to take a further step forward in determining the deity of the Bàyon towers.

4. Vajradhara


Vajradhara, identifiable in Khmer icons by the prajñālinganābhinaya or prajñāembrace mudrā, was assigned a special and important role in connection with Jayavarman’s network of hospitals. The Bàyon style Vajradhara usually has the vajra and ghantā not quite crossed, but held forward in front of the chest. His eyes are lowered, whereas the eyes of Khmer bronzes of Vajrasattva’s are wide open and staring forward. The differences are well illustrated in Coedès 1923 photograph of icons from Thai collections which shows, from left to right, Vajradhara, Vajrasattva and Vajrapā i, all three of them crowned and regally adorned. [Plate 11 Vajradhara, Vajrasattva, Vajrapani]

Vajradhara is also known in a significant series of large sandstone statues. In these images the hands are crossed at the wrists – more likely out of sculptural caution against breakage than from any iconographic imperative – with the vajra prongs of each instrument pointing forward. These deities are all seated in the half-lotus position, minimally dressed and unadorned; they have single heads with u sas covered with a lotus petals and their eyes are lowered. One of only two major icons found in the ruins of Jayavarman’s Bantéay Chmàr temple is one of the larger Vajradharas (80cm) from this series, which has damaged hands and the left arm missing. Victor Goloubew saw it in situ in 1921 before it was removed from the remote and unprotected site to the security of the National Museum in Phnom Penh. [Plate 12 Vajradhara of Bantéay Chmar] Although the hands are damaged, the remnants leave no doubt about the prajñā-embrace mudrā.

Goloubew also wrote in 1937 about a smaller headless Vajradhara which was set on the only vajra-decorated pedestal in Angkor in front of the East Chapel of the hospital complex just west of Takeo temple. [Plate 13 Vajra pedestal before East Chapel] (The image has now disappeared from the chapel and is not in the Conservation depot in Siem Reap, though the pedestal remains in place). Goloubew also published an image of a very similar Vajradhara with tiara which is in the depot marked with inventory number 308, but now sadly without its head. In 1951 Jean Boisselier said there were six other statues in the depot that were Vajradharas but mistakenly identified as Buddhas, but I was able to find only one other in the depot in 2005. On the art market in London, dealer John Eskenazi in 1995 published a 94cm Vajradhara almost identical to the one found in Bantéay Chmàr but in far better condition and with the mudrā perfectly clear. Sotheby’s auction catalogues offered two similar Khmer Vajradharas in the same year in London.

This series of large Vajradharas in sandstone implies a significant cult of this deity and the discovery of one of the icons on a vajra pedestal of the East Chapel suggests a link with medicine. The link is reinforced by the discovery of three more crowned Vajradharas with earrings, in a more squat provincial style, but wearing crowns like that of number 308 in Siem Reap and with bell and vajra held before the chest, in the sanctuaries of hospital complexes in Īśān.

4.1 Buddha of Medicine


The Buddha of medicine is known from many icons in the Bàyon and Angkor Wàt style, usually seated on a āga and holding an ointment pot (traditionally in lapis lazuli from the Badakhshan deposit in Afghanistan) or a myrobalan fruit (har tak ) or simply raising one finger of his right hand while in dhyāna or abhaya mudrā. The name Bhai ajyaguru appears no less than 10 times – more than any other, and sometimes with the name of his city of residence – in the small inscriptions identifying icons in the vra ku i and small sanctuaries of the Bàyon. The familiarity of the Khmer gurus with the bhai ajyaguru sūtra is attested in inscription K.293-24A in face tower 21, near the central tower of the Bàyon, which uses the full formal name of the master of healing and the names of the sun and moon Bodhisattvas who lead the assembly in Bhai ajyaguru’s eastern paradise. These icons, or their stations for replicas in the Bàyon, evidently anchored the royal network of hospitals, which had already grown to number 102 by the time the Tà Prohm stela was erected in 1186.

Vajradhara is called a Bodhisattva in the fourth century classical Mahāyāna version of the bhai ajyaguru-sūtra, where he is among several Bodhisattvas enjoined by Śākyamuni to help those who are ill. But in the later, Tantric version of the Sūtra translated by I-ching in 707 , Vajradhara is given prominence at the end of the text when, in an added section not present in Hsüan-tsang’s 650 translation, he utters a special, concluding protective dhāra – a sacred formula of sounds handed down by Buddhas from aeons past. Although both versions of the Sūtra are included in the esoteric section of the T’ang records (Taisho XIV, 450 and 451) Birnbaum considers that the addition of Vajradhara’s special dhāra underlines the Tantric credentials of the later text. The Khmer Bhai ajyaguru is identified with the Yogācārin version of the three ‘bodies’ (trikayā) or modes of being of Buddhas in the cosmos, in the opening lines of the dedication stelae of the hospitals, which addresses the Buddhas in their ‘selflessness’ (nirātmaka) and ‘nondual’ (advaya) existence in a kind of compressed creed: ‘Homage to the Buddha in the forms of compassionate incarnation, visionary bliss and the supreme awareness in emptiness, whose nonduality in selflessness is beyond the duality of being or not being …’192

The trikāya of dharmakāya, sa bhogakāya and nirmanakāya is also evoked in the in the opening stanzas of both the Tà Prohm and Pra Khan stelae. The dedication on the hospital stelae points to the Buddhistsbelief that although the monks were experts in herbal cures,193 many diseases had psychological or spiritual causes and were treated by guidance in meditation skills194 and with powerful mantras.195 The Khmer hospitals had a doctor and medical assistants but a key part of any cure was administered by monks who were adept in texts focused on healing the sick. The cult of Bodhisattva Vajradhara, the healer, should in my view be dated on stylistic grounds to the ‘Lokesçvarisation’ decade of Jayavarman’s reign (1190s),196 for the images have single heads and two arms and are dressed like monks. This sober dress is in contrast with the icons of both the early and late Bayon period, where there is a tendency towards full regal attire, as in the Angkor Wàt style, with crowns, earrings, heavy necklaces, armlets, anklets and sampots bedecked with jewellery.197 The Vajradharas, like the Lokeśvaras and Buddhas of the first and second Bàyon periods, also have their eyes lowered.


5. Vajrasattva

Khmer epigraphy generally offers limited help in determining the name the Khmers used for their supreme Buddha, but a recently discovered inscribed stone (K.1158) from Sab Bāk near Phimai in modern Thailand is exceptionally clear.198 Dated 1066 in the reign of Bàphûon builder Udayādityavarman, to whom it pays homage, the Sanskrit and Khmer language inscription marks the restoration of nine decaying Buddhist images originally erected probably early in the 10th century, like similar ones attributed to a common patron in inscription K.111 of Wàt Sithor. K.1158 calls the supreme BuddhaVajrasattva’, the sixth Buddha and lord of the five Buddhas


bhāvābhāvadvayātito dvayātma yo nirātmaka

193 The Khmer medical tradition seems to have been strong in herbal cures from as early as the seventh century, when Indian sage Punyodaya was sent from China to gather medicinal herbs in Chen-La. Lin Li-Kouang (1935:83-100) ‘Punyodaya ( ’ati), un propagateur du Tantrisme en Chine et au Cambodge à l’époque de Hsüan-TsangJournal Asiatique Juillet-Septembre 1935.

194 Dowman describes the healing process as: ‘The [[[Wikipedia:patient|patient]]] performs his sādhana and attains mahāmudrā-siddhi and in the process the original disease is cured.’ Keith Dowman (1985:1 ) Masters o ahāmudrā SUNY, New York

195 Gellner records how modern Tantric healers in Nepal rely primarily on mantras: ‘The source of Tantric healers’ power is the possession of powerful, and ipso facto dangerous, mantras. As indicated above, the practitioners often see themselves as providing a selfless service to others. They have no fixed fees. As with a priest’s dak i ā, they take whatever is offered (though they can specify prices if they prescribe a medicine).’ Gellner (1992: 29)

196 Woodward, noting a ‘horizontal emphasis to the facial features’, assigns them to 1186 in the first phase, the date the first 105 hospitals are announced in the Tà Prohm stela. Woodward (2003:208) However, the hospital building programme presumably continued throughout the reign, so more would have been built in the 1190s.

197 The heavier jewellery and heavier human body are visible in the late bronze Hevajras (for example the image within the ma ala frame, the gilt icon found at the Angkor Thom palace, and the bronzes in the Berlin and Guimet museums). The heavier jewellery is found in the stone reliefs of Bantéay Chmàr – the Hevajra lintel and the Lokeśvaras of the western gallery – as well as in the Leper King Terrace, which I have attributed to the following decades under Indravarman II (Woodward took the same view Woodward 1975:149). It is also in contrast with the unadorned portraits of the king and his wife and the Lokesvaras of the 1190s (the Jayabuddhamahānātha images are uniquely bedecked in tiny Buddha images, sometimes taking the form of jewellery).


Chirapat Prapandvidya (1990:11) ‘The Sab Bāk inscription: evidence of an early Vajrayāna


Buddhist presence in Thailand’ Journal of the Siam Society, Bangkok

(‘śr pañcasugata’) of the Vajradhātu Pentad. Woodward has recently made the interesting suggestion that the principal deity in the pentad of Buddhas in the northern lintel at Phimai appears akin to the form of Vajrasattva called Mañjuvajra in the first ma ala defined in the 11th century Ni pannayogāval compendium. The image has Mañjuvajra’s three faces and six arms, the principal ones in dhyāna mudrā and one of the left holding a combined bell with vajra. [Plate 14 Manjuvajra (Vajrasattva) of Phimai]

Chirapat Prapandvidya, the translator, renders the word ‘sugatādikādika’ in the Sab Bāk text as ‘Ādi-Buddha’ but the Sanskrit construction is unusual, suggesting a more likely translation as ‘images of the first Buddha and the others.’ This stone, found in the heartland of the Jayavarman VII’s Mahīdhara dynasty, offers unchallengeable evidence that the Khmer Buddhists were naming their supreme Buddha 'Vajrasattva' just a century before Jayavarman ascended the throne in Angkor.

The Khmer Vajrasattva tradition may also be traceable to the late 12th century Bantéay Chmàr temple. On the southern wall of the cruciform axial cella, marked ‘C’ by Lunet de Lajonquière and presumably not far from where Goloubew saw the Vajradhara stone image (he did not specify its exact placement), I came across the frieze of a Tantric trinity of deities carved in low relief which I later found had just been partially published, though not identified, by Christophe Pottier.203 The trinity is composed of two multi-armed, polycephalous deities dancing in Hevajra’s ardhaparyanka (‘half crossed-legged’) posture on each side of a seated, multiheaded, six-armed deity whose principal arms are in dharmacakra mudrā. [Plate 15 Vajrasattva and Herukas of Bantéay Chmàr] Four more arms rise in an arc behind the central figure, which has three or four faces with open eyes that in some respects resemble those of an unpublished Hevajra in a large doorway lintel still standing in the large neighbouring Hall of Yoginīs to the east. If we think of the Vajradhara statue and the central seated deity in relief in close proximity in Bantéay Chmàr, their differences are emphasised in a way helpful to our search. The single-headed, twoarmed stone Vajradharas with lowered eyes are monk-like in their lack of jewellery or regal dress, which diminishes the probability that they are the adorned deity in the four-faced Bàyon and Bantéay Chmàr face towers. But the six-armed, smiling polycephalous deity on the Bantéay Chmàr central sanctuary wall has the earrings, diadem and open eyes of the deity in the towers. The seated deity’s central position between two dancing Herukas, in a triptych similar to that in a votive casting in the Bangkok Museum, suggests the representation of an overarching, central force of the Buddhist cosmos. The regal deity staring into the Bantéay Chmàr cella is quite different from the nearby monk-like statue of Vajradhara in prajñā-embrace mudrā: I propose that here we have icons of the Khmer Vajrasattva and Vajradhara in close proximity, which clearly separate the two in the Khmer Buddhist world and eliminate Vajradhara as the name of the Bàyon and Bantéay Chmàr face tower deity.

The most complete record yet found of Jayavarman’s late pantheon is contained in the small casting from a votive tablet mould published by Woodward in 1981. The casting presents a three- or four-faced supreme Buddha with six arms at the centre of a ma ala that includes Hevajra, Sam vara (a Heruka variant with Vedic antecedents who appears prominently in the central sanctuary at Phimai and who was venerated at a temple dedicated to him at ālandā ), Vajrapā i, Lokeśvara, another Bodhisattva and the āga Buddha. Woodward’s analysis, as noted earlier, sees Vajrasattva at the centre of the tablet in the earlier form he takes in the STTS, MVS and the Shingon pantheon -- i.e. one of the 16 Vajra Beings and head of the Tathāgata lineage. This is the Bodhisattva Vajrasattva, before he was elevated to the status of supreme Buddha, which Boisselier first saw in the Bàyon faces. This identification would also bring the tablet close to the Čam version of Tantric Buddhism found in the 902 inscription from An Thai, Quảng Nam, where Vairocana presides over an assembly of the Buddhas Śākyamuni and Amitābha, who emit the Bodhisattvas Vajrasattva, Vajrapā i, and Lokeśvara in a cult, which appears based on the Yoga Tantras of the early Vajrayāna. Woodward also associates the Vajrasattva at the centre of the tablet, which is seen as emanated by the image of the āga Buddha of the central sanctuary, with the lingering martial or krodha aspect of the Bodhisattva Vajrasattva described in the ni pannayogāval (NSP), such as that shown in an account of him enforcing Śiva’s attendance at the ceremony for the Buddha’s entry into nirvā a. ‘

His drawing our attention to the Vajrasattva of the 16 Vajra Beings also appeared to be supported by Parmentier’s conclusion that the original plan of the Bàyon contained 16 sanctuary towers , but this detail is in conflict with Cunin’s review of the chronology of the Bàyon architecture in this volume, in which he concludes that 24 face towers were begun simultaneously in the first phase of construction and not 16.


Y . āga-protected Buddha

A. Hevajra

[7 visible faces, 16 arms]

Y2. Vajrasattva

[3 faces, 6 arms]

B. Sa vara

[3 visible faces, 12 arms]


X. Lokeśvara

[4 arms]

Y1. Bodhisattva?

[6-8 arms]

. Vajrapā i

[pratyālidha, feet on corpse]


The identification of the Khmer Vajrasattva with the 16 Vajra Beings also runs into an iconographic problem. All the Khmer bronze Buddhas in the mudrā of Vajrasattva sit in the lotus position of a Buddha and never in the ‘attitude noble’ (royal ease) with the right leg extended forwards. But Mallmann identifies these two leg positions as the consistent distinguishing mark between Vajrasattva as one of the 16 Vajra Beings as they are represented in the major ma ala collections, the NSP and sādhanamālā (SM):

The 16 Vajra [[[Beings]]] are depicted as humans, with peaceful expressions, and seated in the royal ease posture…[T]he white Vajrasattva Buddha can only be differentiated from the Vajrasattva of the 16 Vajra [[[Beings]]] by his seating position: the first is seated in the vajra position, the second in the royal ease position.

Khmer bronzes of Vajrasattva of the 11th and 12th centuries all represent the seated Buddha Vajrasattva rather than the Bodhisattva, as we would indeed anticipate from the contemporary 1066 Sab Bāk inscription (K.1158), which calls the supreme BuddhaVajrasattva’, the sixth Buddha and lord of the five Buddhas (‘śr pañcasugata’) of the Vajradhātu Pentad. Here Vajrasattva is hierarchically above Vairocana.

The casting can be read in various ways; its importance is in the supreme pantheon it contains. My inclination is to read it as a ma ala with the major deity in the centre -- and between Hevajra and Sa vara the crowned, seated polycephalous deity could only be the supreme Buddha Vajrasattva that we find in the hevajra-tantra and the sam varodaya-tantra, rather than his earlier manifestation as one of the 16 Vajra Beings in the STTS and the ahāvairocana-sūtra, the Yoga Tantras that remain core texts for the Shingon.

The casting may give other indications of a deliberate correlation of the temples of the empire, where the major icons of Jayavarman VII resided. The three lotus stalks snaking upwards may be seen as delineating the principal deities of the three great Buddhist foundations that straddled the Dangrek Mountains and formed the spiritual backbone of Jayavarman’s empire: the Bàyon in Angkor, Bantéay Chmàr and Phimai. Phimai was dedicated in the early 1100s but was later extended by Jayavarman, who built a chapel with his own portrait image kneeling and facing the central sanctuary -- which assures us that it was still a thriving foundation in 1200. The Bàyon and Bantéay Chmàr, halfway along the imperial road between Angkor and Phimai, were being completed simultaneously at the turn of the 13th century. This vertical reading therefore takes account of the prominence of both Sa vara and Vajrapā i in Phimai, and Hevajra and Lokeśvara in Bantéay Chmàr.

Bantéay Chmàr

Bàyon

Phimai


4. āga Buddha of the Bàyon

2. Hevajra of Bantéay Chmàr (8 faces, 16 arms, dancing on one corpse) 1.Vajrasattva (3/4 faces, 6 arms dharmacakra mudra) Bàyon faces

Sa vara of Phimai (elephant hide, 4 faces, 12 arms, prajñā-embrace, 2 corpses)

5. Mahākaru ikaLokeśvaras of Bantéay Chmàr West wall 6. Prajñāpāramitā (6 arms)

Vajrapā i of Phimai (dancing wildly on corpse)

A further possible interpretation is to read the whole casting as a representation of the Bàyon alone. Such a reading would bring the casting closer to the triptych on the wall of the Bantéay Chmàr cella. But this scheme can only survive if a major assumption is granted, namely that the Bàyon cella below the faces of the central sanctuary contained large icons of Hevajra and Sa vara, the wrathful emanations of Vajrasattva. We have a candidate for such a role in a giant stone Hevajra [Plate 16 New York Metropolitan Museum’s Hevajra] found dumped outside the Angkor Thom walls. But no similar icon of Sa vara has been uncovered. Certainly the size and quality of carving of the heads and torso of the Hevajra icon, now in ew York’s Metropolitan Museum, suggest a major royal purpose. From the photographs taken of the exhumation of Hevajra’s torso and legs in March 1925 in the EFEO archive, we can establish that the icon was nearly three metres high and was erected on a one metre pedestal. [Plate 17 Hevajra legs] I have attempted a ‘virtual’ reconstruction of this imposing sculpture from the photograph and the remnant pieces in the Siem Reap depot (the legs have now disappeared). [Plate 18 Hevajra reconstructed] We cannot know where this large stone Hevajra was originally erected before it was dumped in a mound of earth outside the east gate of the city walls. Its broken state may suggest it was either one of the thousands of icons smashed when the Buddhists were driven out of Jayavarman’s temples, or possibly later by Theravādin Buddhists. Chinese envoy Zhou daguan, who arrived at the end of the 13th century, makes no reference to whether Buddhists ceremonies were still being performed in Jayavarman VII’s large stone temples. He refers only to a widespread village network of modest, single room Buddhist temples with a tiled roof and a single icon made of painted clay. None of the surviving small Bàyon inscriptions refers to a Hevajra or Sa vara, but if this icon had been erected in one of the Bàyon tower sanctuaries beside the central āga Buddha an inscription would not have been required, because only the ancillary shrines in the central tower have short inscriptions. Sanctuary BY.2 beside the cella could have easily accommodated it. Cunin says this sanctuary ‘…constitutes one of the largest spaces under a corbelled roof in all Khmer art.’

Careful analysis of the facial features and regal dress of this Hevajra show them to be very close in style to those of the four metre high Bàyon style Vi u in the Angkor Wàt temple, which was presumably erected by Jayavarman VII. As Jayavarman VII was honouring Vi u and Śiva with images and ritual space in his new, predominantly Buddhist, temples it would be logical to assume that Angkor Wàt, the Vai ava shrine of his kinsman Sūryavarman II, continued as an active foundation throughout Jayavarman’s reign. And if the indrābhi eka also marked the restoration of Khmer power in Čampā, the Khmers had a Vi u called Cāmpeśvara (Lord of Čampā), who from pre-Angkorian times had been erected in thanks for victories against the Cams.

ājendravarman in Pre up inscription of 961 thanked Cāmpeśvara for his victories over the Cams. Japanese archaeologists recently uncovered a very similar, but smaller Vi u to the Angkor Wàt image (their arms broke off at exactly the same points) in the pond behind one of the Prasat Suor Prat towers, which I suggested (footnote 81) may also have played a role in the indrābhis eka celebrations.

The Angkor Wàt Vi u, whose broken arms have recently been replaced and whose original head has been returned from the Phnom Penh Museum, of course survived the Śaiva reaction. The striking similarities in the carving of the faces of this Vi u and the Hevajra now in New York suggest that these two exceptional, large statues were among the last great icons commissioned by Jayavarman VII in the final phase of the Bàyon style.

Whatever reading is preferred for the tiny Bangkok Museum casting, it seems to hold the key to the enigma of the Bàyon faces. And, whichever way it is construed, the casting, whose mould I have now located in the Phnom Penh Museum [Plate 19 Poipet bronze mould], establishes Vajrasattva at the centre of the Khmer state ma ala, as he is the lord of the guru ma ala of the ewar Vajrācāryas. For only the Vajrasattva of the Mahīdhara Khmers, could take a central position in a cosmic diagram bearing images of his own fierce emanations Hevajra and Sa vara. When we apply this hierarchy to the Bàyon and see the āga Buddha presiding in the cella but Vajrasattva as ‘filling space in a regularized order in every direction’ above him the face towers, no other identification of the deity in the towers – Brahma, Śiva, Lokeśvara, Siva-Lokeśvara, neak ta, Hevajra, Vajradhara – comes closer to fitting the facts of the material record, as well as the context of Khmer Buddhism responding to the dramatic crisis of international Buddhism.


Conclusion


The Bàyon’s towering circular sanctuary, enclosed within two square gallery walls, has the form of a ma ala in its ground plan and also in its projection upwards as Mount Meru. Seeing Vajrasattva as everywhere visible in the monumental structure resolves a key part of the Bàyon enigma. We have turned full circle to seeing, with Zhou daguan, the giant deity as a multi-headed Buddha; and we can now confidently add (what a 13th century Chinese ambassador probably assumed), that the Buddha was Tantric.

Jayavarman’s deliberately slow unveiling of his Tantrism climaxed in the final years of his reign with his inscribing the cosmogony of Vajrayāna into the skyline of Angkor. The sheer number, as well as the uniformity, of the 200 giant faces convey a synthesis of the Bàyon’s broad pantheon of Buddhist and Hindu deities in an overarching conception – captured in Boisselier’s earlier vision of the Bàyon as the magical Mount Sumeru on which Vajrasattva is visible at all times. The uniformity of the faces and their lack of individual detail can be seen as a deliberate attempt to render the qualities of the ultimatefourth state’ (caturtha tur[i]ya), defined in the

Śaiva Tantras and adopted by the late Tantric Buddhists, that Kamaleswar Bhattacharya lists as: invisible (ad a), impalpable (avyavahārya), ungraspable (agrāhya), indefinable (avyapadeśya), unthinkable (acintya), without any distinctive marks (alak ana). For the Tantric Buddhist Jayavarman VII, this was the ultimate projection of the ethereal, formless Vajrasattva, beyond time and space yet omnipresent and omniscient. Tantric Buddhism approached such large, cosmic, syntheses through constructing complex ma alas and the Bàyon is the ma ala of Jayavarman VII. In the hevajra-tantra we are told the ma ala appears of itself ‘as the full and efficacious expression’ of ‘concentrated meditation’ (samāhitayoga) on Hevajra, until the king’s realisation of emptiness generates the icon of the city and the state, envisaged in countless forms of the deity filling space:

‘This bliss is Wisdom, this bliss is Means, and likewise it is their union. It is existence, it is non-existence, and it is Vajrasattva.’217 It is a major new statement addressed to the Śaiva Khmer empire as well as the setting for the consecration of Cambodia’s first Buddhist king as cakravartin. At some moment late in the reign, we can assume that a major ceremony was held, beneath the entranced smile of Vajrasattva in the face towers, that was at once the dedication of Jayavarman’s regnal temple, the celebration of his indrābhi eka and an assertion of the future of Vajrayāna Buddhism in a hostile world.

Plates:

1. Face tower
2. Lokeśvara at ew York Metropolitan Museum
3. Lozenge with pupil
4. Vajrasattva with hair cover (after Coedès 1923 pl.28.3)
5. Jayavarman, Phimai
6. Guimet Jayavarman portrait
7. crowded vajra-eyed post-Bàyon Yoginis
8. Prah Pallilay gopura standing Buddha
9. Svayambhū Mahācaitya (after Held, Suzanne and Beguin, Gilles 199 :5 Nepal
Himer Verlag, München)
10. Hevajra excavated from the royal palace (courtesy Phnom Penh Museum)
11. Vajradhara, Vajrasattva, Vajrapāni (after Coedès 1923:pl.27)
12. Vajradhara of Bantéay Chmàr (courtesy Phnom Penh Museum)
13. Vajra pedestal before East Chapel
14. Mañjuvajra (Vajrasattva) of Phimai
15. Vajrasattva and Herukas of Bantéay Chmàr
16. ew York Metropolitan Museum’s Hevajra
17. Hevajra legs (courtesy EFEO archive)
18. Hevajra reconstructed
19. Poipet bronze mould  
Bibliography:
 
-- Plan and report of the Survey and Excavations of Ancient Monuments inn NorthEastern Thailand (1959, 1960-1 reprint 1979) Fine Arts Department, Bangkok.
 
Ang Chouléan 1986 Les êtres surnaturels dans la religion populaire khmère Cedoreck, Paris
Astley-Kristensen, Ian (1991) The Rishukyō: The ino-Japanese Tantric
Prajñāpāramitā in 150 erses Amoghavajra’s ersion) Institute of Buddhist Studies,
Tring
Bautze-Picron, Claudine (2003) The Buddhist murals of Pagan Orchid Press,
Bangkok
Bhattacharyya, Benoytosh (1924) The Indian Buddhist Iconography mainly based on the ādhanamālā and other cognate tāntric te ts o rituals OUP
Bhattacharya, Kamaleswar. (1961) Les religions brahmaniques dans l’ancien
Cambodge d’apres l’épigraphie et l’iconographie PEFEO
Bhattacharya, K. (1966:5-18) ‘Linga-kośa’ Essays offered to G.H. Luce by his colleagues and friends in honour of his seventy-fifth birthday Artibus Asiae, Ascona
Raoul Birnbaum (1980) The Healing Buddha Rider, London
Boeles, J.J. (1966:14-29) ‘Two Yoginis of Hevajra from Thailand’ Essays offered to
G.H. Luce Eds. Ba Shin and J. Boisselier, Artibus Asiae, Ascona
Boisselier, Jean (1951:324- 2) ‘Vajrapani dans l’art du Bàyon’ in Proceedings of the
22nd Congress of International Orientalists Istanbul
Boisselier, J. (1955) La statuaire khmere et son evolution EFEO, Saigon
Boisselier, J. (1967:275-334) 'Notes sur l'art du bronze dans l'ancien Cambodge' Artibus Asiae XXIX,4
Boisselier, J. (1978:323- 1, plates 24, 25:410) ‘Cambodia’ in The Image of the Buddha Ed. D. Snellgrove, UNESCO
Boisselier, J. (1988:2-5) ‘The symbolism of Angkor Thom’ The Siam Society Newsletter March 1988 vol.4.1
Boisselier, J. (1993:129- 8) ‘A Buddhist presence amidst the godsAngkor the serenity of Buddhism ed. M. Ribaud Thames and Hudson, London
Chirapat Prapandvidya (1990:11-14) ‘The Sab Bāk inscription: evidence of an early
Vajrayāna Buddhist presence in Thailand’ Journal of the Siam Society Bangkok
Clark, Walter Eugene (1937) Two Lamaistic pantheons I Harvard University Press
Coedès, George (1906:44-81) ‘La stèle de Ta-Prohm’ BEFEO 6
Coedès G. (1913) ‘La foundation de Phnom Penh au XVe siècle d’après la chronique cambodgienne’ BEFEO 13
Coedès, G. (1928:81-10 ) ‘La date du Bayon’ BEFEO 28
Coedès, G. (1932:71-81) ‘Etudes Cambodgiennes XXVIII. Quelques suggestions sur la methode à suivre pour interpréter les bas reliefs de Bantéay Čhmàr et de la galerie extérieure du Bàyon’ BEFEO 32
Coedès, G. (1941:26 ) ‘Le stèle du Prah Khan d’Angkor’ BEFEO 41
Coedès, G. (1942:187-9) ‘Inscription du Bàyon K.4 0’ Inscriptions du Cambodge vol.
II
Coedès, G. (1943) Pour mieux comprendre Angkor Imprimerie d’Extrême Orient,
Hanoi
Coedès, G.1952 Inscriptions du Cambodge IV
Coedès, G. (1960:179-98) ‘Le portrait dans l’art khmèr’ Arts asiatiques Musée Guimet, Paris
Coedès, G. (1968) The Indianized states of Southeast Asia, East-West Center Press
Commaille, Jean (1912) Le Bàyon d’Angkor Thom EFEO, Paris
Cunin, Olivier (2004) De Tà Prohm au Bàyon: contribution à l’histoire architecturale du temple de Bàyon Institut National Polytechnique de Lorraine.
Dagens, Bruno, ‘Les tours à visages du Bayon et leur signification’, (2000: 98-120), 5th annual Bayon Symposium Siem Reap
Dagens, Bruno (2000:98-119) ‘The Bàyon Face Towers and their Meaning’ 5th annual Bayon Symposium Siem Reap
Dagens, B. (2003) Les Khmers, Éditions les belles letters, Paris
Dalsheimer, Nadine (2001) Les collections du musée national de Phnom Penh EFEO Paris
Dasgupta, Shashi Bhusan (1950) An introduction to Tantric Buddhism University of Calcutta
Dhani, Prince (1961:275-8 ) ‘The gilt lacquer screen in the audience hall of Dusit’
Artibus Asiae 24, Ascona
Dowman, Keith (1985) asters o ahāmudrā SUNY, New York
Dowson, John (1869) The history of India as told by its own historians. The Muhammadan period. Edited from the posthumous papers of Sir H.M. Elliot KCB Trübner, London.
Davidson, Ronald (2002) Indian Esoteric Buddhism: a social history of the Tantric
movement Columbia University Press, New York Dupont, Pierre (1936:630-4) ‘Chronique’ BEFEO 36
English, Elizabeth (2002) ajrayogin , her visualizations, rituals and forms Wisdom Publications, Boston
Eskenazi, John (1996) Silk & Stone: The Art of Asia Hali London
Filliozat, Jean (1981:59-99) ‘Sur le çivaïsme et le bouddhisme du Cambodge, à propos de deux livres récents’ BEFEO 70
Filliozat, J. (1969:45-8) ‘Emigration of Indian Buddhists to Indo-China c. A.D. 1200’
Studies in Asian History: proceedings of the Asian History Congress 1961 Asia
Publishing House, New Dehli
Finot, Louis (1925:372-92) ‘Phimãnàkàs’ BEFEO 25
Finot, L. (1925:393-406) ‘Temple de Mangalārtha à Angkor Thom’ BEFEO 1925 Frasch, Tilman (1996) Pagan Stadt und Staat Steiner, Stuttgart’
Frasch, T. (2000:41-9) ‘A remark on the Mahābodhi temples at Pagan’ Proceedings of the 7th International Conference of the EASEAA Berlin 1998 Eds Wibke Lobo &
Stefanie Reimann
Gellner, David (1992) Monk, householder and Tantric priest: Newar Buddhism and its hierarchy of ritual Cambridge University Press
Giteau, Madeleine (1997) Khmer art: the civilisations of Angkor ASA editions, Paris
Goloubew, Victor (1937:97-104) ‘Sur quelques images khmèrs de Vajradhara’ Journal of the Indian Society of Oriental Art 5
Groslier, Bernard Philippe (1954:228-9) ‘Fouilles du Palais oyal d’Angkor Thom’
Proceedings of the 23rd International Congress of Orientalists Cambridge 1954 Royal Asiatic Society, London
Groslier, B. P. (1973) Le Bàyon: inscriptions du Bàyon EFEO Paris
Hirth, F. and Rockhill, W. (1911) Chau Ju-kua: his work on the Chinese and Arab trade in the 12th and 13th centuries, entitled Chu-fan-chi Imperial Academy of Sciences, St Petersburg
Hodgson, Brian H. (1874) Essays on the languages, literature, and religion of Nepal and Tibet: together with further papers on the geography, ethnology, and commerce of those countries Trübner & Co, London Huntington, John C. (2002:16-2 ) ‘The Iconography of Svayambhū Mahācaitya: the main mandalas’ Orientations vol.33:10
Jacques, Claude (1976:250-4)) ‘Bernard-Philippe Groslier.”Inscriptions du Bàyon”.
Le Bàyon.’ Artibus Asiae Vol. XXXVIII,i
Jacques, C (1999) Angkor Koenemann, Cologne
Kabese, Kanyū (1961:218-9) ‘Ādi-Buddha in Tibetan Buddhism/Japanese Buddhism
Encyclopaedia of Buddhism Fascicule: A-Aca Ed. G.P. Malalasekera, Government Press of Ceylon
Klimburg-Salter, Deborah E. (1998:1-1 ) ‘Is there an inner Asian international style
12th to 14th centuries? Definition of the problem and present state of research’ The
Inner Asian International Style 12th -14th Centuries, Verlag der Österreichischen
Akademie der Wissenschaften, Wien
Körös, Alexander Csoma de (1833:57-9) ‘ ote on the Origin of the Kāla-Chakra and the Ādi-Buddha Systems’ Journal of the Asiatic Society of Bengal Vol. 2
Lamotte, Étienne (1966:112-59) ‘Vajrapani en Inde’ Mélanges de sinologie offerts à
M. Paul Demiéville PUF Paris
La Vallée Poussin, L. de (1929) ijñaptimātratāsiddhi: la siddhi de Hiuan-tsang Geuthner, Paris
Lin Li-Kouang (1935:83-100) ‘Punyodaya ( ’ati), un propagateur du Tantrisme en
Chine et au Cambodge à l’époque de Hsüan-TsangJournal Asiatique JuilletSeptembre 1935
Linrothe, Rob (1999) Ruthless compassion Serindia, London
Lobo, Wibke (1994:113-2 ) ‘ eflections on the Tantric Buddhist deity Hevajra in
Cambodia’ Proceedings of the 5th International Conference of the European
Association of Southeast Asian Archaeologists Paris, 24-8 October 1994 Vol.II’ Lobo, W. (199 : 18) ’Catalogue101’ Angkor et di siècles d’art khmer Réunion des Musées Nationaux, Paris
Malalasekera, G. P. (1938) Dictionary o Pāli proper names John Murray, London Mallmann, Marie-Thérèse de (1986) Introduction à l’iconographie du tântrisme bouddique Librairie d’Amérique et d’Orient, Paris
Mannikka, Eleanor. (1996) Angkor Wat Time, Space, and Kingship, University of Hawaii Press
Maxwell, T.S. (1997) The Gods of Asia OUP, Oxford
Mus, Paul (1928:153-2 8) ‘Le Buddha paré. Son origine Indienne. Çakyamuni dans le
Mahāyānisme moyen’ BEFEO 28
Mus, P. (1936:57-68) ‘Le symbolisme à Angkor-Thom: Le «Grand Miracle» du
Bàyon’ Compte Rendus des éances de l’Année: Academie des Inscriptions & Belles-
Lettres, Paris;
Mus, P. (1937:65- 5) ‘Angkor in the time of Jayavarman VIIIndian Arts and Letters
XI’
Mus, P. (1962:527-9) ‘Angkor vu du Japon’ France-Asie XVIII Bhikkhu ā amoli (1991) The divine abidings: The path of purification visuddhimagga) by Bhadantācariya Buddhaghosa Buddhist Publication Society, Candy
Neville, Tove E. (1998) Eleven-headed Avalokiteśvara, Chenresigs, Kuan-yin or
Kannon Bodhisattva: it’s origin and iconography Munshiram Manoharlal, Dehli’ Orzech, Charles D. (1998) Politics and Transcendent Wisdom: The Scripture for
Humane Kings in the Creation of Chinese Buddhism Pennsylvania State University Press
Park, Hyounggook (2005:280- 10) ‘Appendix : Art History Survey eport of the
Bàyon Temple’ The Bàyon master plan JSA
Parmentier, Henri (1936:281-6) ‘Autres modifications subies par le Bayon au cours de son exécution’ BEFEO 36
Pelliot, P. (1951) Mémoires sur les coutumes du Cambodge de Tcheou Ta-Kouan:
version nouvelle Librairie d’Amérique et d’Orient, Paris
Pottier, Christophe (2004:131-50) ‘À propos du temple de Bantéay Chmàr’ Aséanie 13, juin 2004
Przyluski, Jean (1937: 01-18) ‘Heruka- Śa bara’ The Polish Bulletin of Oriental Studies I, Warsaw
Ray, Nihar-Rajan (1936) Sanskrit Buddhism in Burma H.J Paris, Amsterdam
Ray, Reginald (1973) ‘ andala’ symbolism in tantric Buddhism Chicago University
Samuel, Geoffrey (2003) Ritual Technologies and the State: The Mandala-Form
Buddhist Temples of Bangladesh. A paper for 5th international congress on Bengal Art,
Dhaka and Mainmati ‘
Sakai, Shinten (1961:214-8) ‘Ādi-Buddha: development’ Encyclopaedia of Buddhism Fascicule: A-Aca Ed. G.P. Malalasekera, Government Press of Ceylon
Sanderson, Alexis (1994:87-101) ‘Vajrayāna: Origin and Function’ Buddhism into the year 2000 Dhammakaya Foundation, Bangkok
Schlagintweit, Emil (1863) Buddhism in Tibet; illustrated by literary documents and objects of religious worship. Reprint 1988 Indological Book House Dehli Schroeder, Ulrich von (1981) Indo-Tibetan bronzes Visual Dharma Publications, Hong Kong
Sharrock, Peter D. (2004) ‘Yoginis of the Bayon’ Proceedings of the 10th
International Conference of the European Association of Southeast Asian
Slusser, Mary (1982) Nepal Mandala: a cultural study of the Kathmandu Valley, Princeton
Snellgrove, David (1957) Buddhist Himālaya: travels and studies in quest o the origins and nature of Tibetan Religion Cassirer, Oxford
Snellgrove, D. (1959:204-1 ) ‘The notion of divine kingship in Tantric
Snellgrove, D. (1981:5-6 ) ‘Introduction’ sarva-tathāgata-tattva-sa graha:
Facsimile reproduction of a Tenth Century Sanskrit Manuscript from Nepal edited by
Lokesh Chandra and David L. Snellgrove pp. 5-67 New Delhi
Snellgrove, D. (1987) Indo-Tibetan Buddhism: Indian Buddhists and their Tibetan successors 2 vols Shambala, Boston; reprint 2004 Orchid Press, Bangkok
Archaeologists (EurASEAA) London (forthcoming)
Buddhism’ La regalita sacra: E.J. Brill, Leiden
Stern, Philippe (1927) Le Bayon d’Angkor et l’évolution de l’art khmer: étude et discussion de la chronologie des monuments khmers Librairie orientaliste Paul
Geuther, Paris
Stern, P. (1965) Les monuments khmers du style du Bàyon et Jayavarman VII Universitaires de France, Paris
Tajima, yūjun (1959) Les deu grands ma alas et la doctrine de l’esoterisme Shingon Bulletin de la Maison Franco-Japonaise VI Tokyo/Paris
Thomas, F. W. (1908:20-2) ‘Abhi Encyclopaedia of Religion and Ethics Ed.
James Hastings, T & T Clark, Edinburgh
Thompson, Ashley (2000:122-14 ) ‘An oblique view of the Bàyon’ 5th international symposium on the preservation and restoration of Bàyon – Final report JSA, Siem Reap
Thompson, A. (2004:13-39) ‘The future of Cambodia’s past: a Messianic MiddlePeriod Cambodian oyal Cult’ History, Buddhism and new religious movements in
Cambodia Eds Marston, J. and Guthrie, E. University of Hawai’i press
Thurman, Robert A.F. (1958:119-48) ‘Vajra hermeneuticsBuddhist Hermeneutics ed.
Donald Lopez University of Hawaii Press
Trouvé, Georges (1933:1116- ) ‘Chronique: Bàyon’ BEFEO 33
Tsuda, Shinichi (1974) The a varodaya-tantra, selected chapters Hosuseido Press,
Tokyo
Tucci, Giuseppe (1949) Tibetan Painted Scrolls La Libreria dello Stato, Roma
U Lu Pe Win (1971:3-10) ‘Old Burma – Early Pagan, a radical review’ Journal of the
Burma Research Society LIV
Vickery, Michael (1977) Cambodia after Angkor, the chronicular evidence for the
14th to 16th centuries Yale
Wheatley, Paul (1971) The Pivot of the four Quarters Edinburgh
Williams, Paul with Tribe, Anthony (2000) Buddhist thought Routledge, London
Woodward, Hiram (1975) Studies in the art of central Siam 950-1350 Yale
Woodward, H. (1981:57-6 ) ‘Tantric Buddhism at Angkor Thom’ Ars Orientalis 12
Woodward, H. (1994:106-10) ‘The Jayabuddhamahānātha Images of Cambodia’ The
Journal of the Walters Art Gallery 52/3 (1994/5)
Woodward, H. (2003) The art and architecture of Thailand Brill, Leiden
Woodward, H. (2004:329-54) ‘Esoteric Buddhism in Southeast Asia in the Light of ecent Scholarship’ Journal of Southeast Asian studies 35.2
Wyatt, David K. (2001:3-66) Relics, oaths and politics in 13th century Siam JSEAS
32.1




Source