1

This is a pre-print version of an article published in in Materialising Southeast Asia’s Past (2013: 41-
55) eds Marijke J. Klokke & Veronique Degroot, University of Hawai’i Press, Hawai’i;
please refer to the final published version if quoting.

The tantric roots of the Buddhist pantheon of Jayavarman VII

By Dr Peter D. Sharrock SOAS, London University

ps56@soas.ac.uk
(7,300 words)

Abstract
The state Buddhism adopted by the ancient Khmers in the late 12th century under king
Jayavarman VII has generally been taken as a form of Bodhisattvayāna, focused on a cult of the
Avalokite vara samantamukha (‘facing-all-directions’) of the Lotus Sutra. George Cœdès saw
such a cult in large multi-armed icons of the ‘radiant’ Bodhisattva as well as in the giant faces
on the towers of the Bayon state temple. This paper challenges the consensus view that Khmer
Buddhists followed a mainstream Mahāyāna from the seventh century until the imposition of the
Buddhist state and argues for seeing their Buddhism as predominantly tantric throughout the
ancient period.

The Buddhism of Jayavarman VII’s father, king Dharanindravarman II, who probably
built Beng Mealea temple outside Angkor and erected the large four-armed
Avalokite vara icon now in the Siem Reap conservancy (plate 1), may have been
similar to many other cults of the great compassionate Bodhisattva that gained
adherence from many Buddhist communities of South, Southeast and East Asia from
the 10th century on. But Jayavarman’s Buddhism seems to have been more elaborate.

A critical test of Jayavarman’s

Buddhism lies in determining
who is the deity whose giant
faces loom over Angkor from
towers on the Bayon temple, the
monument that tipped the ancient
Khmers from 4 centuries of state
aivism into state Buddhism
(Plate 2).

Plate 1 Lokesvara of Beng Mealea, Siem Reap

Conservation
 2

Plate 2 Bayon face towers Plate 3 Bayon Loke vara lintel

Iconographical reasons for rejecting the current identification of the faces

Most guide books interpret the Bayon as bearing giant images of the Mahāyānist
Bodhisattva Loke vara ‘facing all ways’(samantamukha) of the Lotus Sutra, 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 support the idea of a network of
territorial deities still so important to the Khmers today, 1 but I question both the
identification of the deity at the hub 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 of
Jayavarman’s Angkor.

The first iconographical reason for questioning the Loke vara identification is the
absence of an Amitābha figurine in the hair of the face tower deities. Bernard Philippe
Groslier wrote: ‘…Loke vara 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.’2 Further reasons are that Loke vara is
unknown anywhere with four faces and in the Bayon period he is invariably
uncrowned, whereas the Bayon faces wear tiaras, hair covers and earrings. A further
consideration is that if the Bayon was indeed dedicated to Loke vara, it would be
highly unlikely that a major Loke vara pediment would be covered over in the final
Bayon build. Yet Plate 3 shows the pediment exposed by Henri Marchal in 1924
which overnight confirmed the Bayon’s Buddhist vocation. (Before this the Bayon
had been thought to be a 9th century Brahmanical temple). Moreover, Preah Khan, the
temple Jayavarman VII expressly dedicated to Loke vara and his father, is the one
temple he built in Angkor whose towers and gateways bear no giant faces. If
Loke vara were in the face towers, there would certainly have been face towers in
Preah Khan.

Seeing an image of the king in the face towers seems to be ruled out by the portrait
statues of the king. These statues do appear to be portraits for they depict a muscular,
bull-necked, moustachioed, square-shouldered, 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 Chams, and his fight to secure the Khmer throne in his
 3

forties. This is an uncrowned man with lowered eyes probably inclining in añjali
mudrā before his god – but not himself a god or Buddha as Cœdès proposed in his
divine king or ‘Buddharāja’ theory long ago rejected by scholars.

It can be readily granted that the

portrait statues and the Bayon faces
share the basic facial features of the
Bayon style – wide, flat mouth, high
forehead, heavy eyebrows – but this
applies, as Dagens points out, to
‘virtually all of the statuary of the
Bayon and it is therefore not really
meaningful.’3 In contrast to the five
known portrait statues of the king,
the giant renderings of deities in the
Bayon towers are heavily bejewelled
with earrings, high tiaras, long neck-
covers and choker necklace.

In abandoning the
‘Avalokite vara/king-in-the-faces’
hypothesis, I propose we look to
Khmer icons of supreme tantric
Buddhahood for a better account of
how the face towers project
Jayavarman’s supreme pantheon.
Plate 4 portrait statue of king Jayavarman VII, Can we establish a case for saying
Phnom Penh National Museum the rīghana (‘the glorious ones’) –
the mysterious epithet for the supreme Khmer Buddhas that Claude Jacques has found
in 9 inscriptions4 – were headed by Vajrasattva & Hevajra? (Plates 5 & 6)

Plate 5 Bayon face tower (Vajrasattva?) Plate 6 Hevajra bronze (PP Museum)
 4

If we now turn back to the earliest indications of Buddhism among the ancient
Khmers, we find clear traces of a distinguished and internationally-known cult of
Mahāyāna evolving into tantric Buddhism from the 7th century onwards. The Khmer
temple inscriptions show the elite of ancient Cambodia as highly literate and there
were monasteries, but these had limited independence and were all royal foundations.
The Buddhists’ lack of favour at court therefore kept their profile low and left only a
faint trace of their cult in the material record.

Seventh century
Nancy Dowling has noted strong resemblances to the H nayāna icons of Dvāravat in
the series of standing Buddhas in the Phnom Penh National Museum, which she
credibly dates to Zhenla in 610-65.5 The Buddhas come mostly from the region
around the old Funanese centre of Angkor Borei in southern Zhenla, west of the
Mekong, but similar wooden icons have been found further south in the Mekong
Delta. The Khmer experience of Buddhism apparently came to an abrupt end later in
the seventh century, for the Chinese monk Yijing travelled through the territory and
reported that although Buddhism had flourished there, it had been wiped out by ‘a
wicked king’.6 Cambodian craftsmen were to produce no more of the quiet,
contemplative Buddhas in flowing bhiksu’s robes until the arrival of Theravāda at the
end of the 13th century.

We also find evidence from Zhenla for the presence of early Mahāyāna Buddhism.
This tradition is attested in a large and impressive image of the crowned Bodhisattva
Avalokite vara, now in the Musée Guimet in Paris, but unearthed at Tan Long in
Rach Gia on the coast of what is now Vietnam. Mirielle Bénisti dates the Bodhisattva
from Rach Gia to 650-720 from the motif of a band of alternating round and
rectangular jewel settings at the base of the crown, which appears on temple
colonettes in these years. 7 Yijing’s report would indicate the early part of this date
range.

Direct influence in Zhenla from India

The seventh century Khmers of Zhenla seem to have been among the first people
outside India to feel the impact of Buddhism’s tantric or esoteric third wave, known
variously as Mantrayāna or Vajrayāna, which was to eventually spread from Bihar
through, Nepal, Tibet, Mongolia, China, Korea, Japan, Sri Lanka and much of
Southeast Asia. Chinese records first disclose that the Khmers of Zhenla were
exposed to the tantric Mantrayāna, or early Vajrayāna.8 Chinese annals contain a
revealing document, the official Tang biography of Puṇyodaya, one of the first Indian
tantric masters recorded as travelling abroad. Shortly after arriving in the Tang capital
and translating Sanskrit texts into Chinese, Puṇyodaya was sent on an imperial
mission to Zhenla. Lin Li-Kouang, who translated Puṇyodaya’s biography, traces the
emergence of tantric Buddhism out of secret monastery rituals to the period between
650 and 750. Puṇyodaya played a role in this evolution with his 663 translation of the
Astamandalaka-sūtra (T.486). His translation reflects the concept of the Mantrayāna
or vehicle of spells more than the later Yoga tantra evinced in the 770 translation of
the same text by Amoghavajra, the most eminent tantric Buddhist guru of the Tang
court.9
 5

Puṇyodaya was a contemporary and perhaps rival of the Yogācārin pilgrim and
translator Xuanzang. Like Amoghavajra a century later, Puṇyodaya was sent on a
Chinese imperial mission to gather tantras.10 He also studied, as Amoghavajra did, in
the tantric institutions of Sri Lanka11 before reaching China then being sent to Zhenla
to gather medicinal herbs in the mid-seventh century. The biography says ‘when he
arrived in the Southern Seas, the kings paid homage to him and built religious
foundations especially for him.’ His teachings and skills made such an impact that
after he returned to China a Zhenla delegation of Buddhists arrived in the Tang capital
in 663 to plead for his return. The emperor permitted his departure by decree and the
Indian master spent the rest of his days in Zhenla.12 However, his work to establish
early tantric Buddhism in Zhenla suffered a rapid reversal in the suppression usually
attributed to Jayavarman I (r.657-90+: K95 v.5), who married a daughter to an Indian
Brahmin Agastya of Aryade a.13

Eighth century
Outside Zhenla, the eighth century saw Vajrayāna mushroom across East Asia from
the northern Indian monasteries. George Cœdès indeed calls the arrival of Vajrayāna
on the world stage the dominant fact of the eighth century. 14 The advent of the Pāla
dynasty in Bengal in mid-century had a major impact, and through the ailendras
(c.750-850) there was a surge of Mahāyānist temple building in Java and the Malay
Peninsula (Chaiya).15

Amoghavajra narrowly survived shipwreck in 719 as he accompanied Nālandā’s

former abbot Vajrabodhi (670-741)16 to China to translate the tantras into Chinese for
the Tang emperor.17 After Vajrabodhi’s death in 741, Amoghavajra returned on an
imperial mission to Sri Lanka to fulfil his master’s dying wish that he bring to China
the Sarvatathāgatatattvasaṃgrahanamamahāyānasūtra (STTS) and other large works
they had been forced to throw overboard in the earlier storm. On his return to China
the emperor Hiuan-tsong had an altar erected in his palace where he received the
tantric consecration (a h a). ‘In consequence all classes of society adopted the cult.
This was the most prosperous era of esoteric Buddhism in China.’18 For a time it
moved Daoism to one side at court.

At the end of the eighth century, under the guidance of Amoghavajra’s pupil Houei-
kouo, Kōbō Daishi (774-835) arrived from Japan for two years instruction by this
master before returning in 806 to found the Shingon sect that held great sway at court
and that still prospers today. The two great texts of Shingon Buddhism are the
Da n ch yō (Mahāva rocana-sūtra)19 and the Kongōchōgyo or STTS, translated into
Chinese by Amoghavajra. Reviewing the prominence achieved by the tantric
Buddhists in the Tang court under the Tai-tsung emperor (762-779) and of the
Shingon sect in Japan, Ronald Davidson concludes:
There appears no exception to the rule that, when the Mantrayāna becomes culturally
important outside India, it is principally through the agency of official patronage, either
aristocratic or imperial. 20
This applied in Tibet, China, Mongolia and Japan. In Cambodia the rule also held
because the country’s Buddhists kept only a low profile until the Mah dhara seizure of
power in the 11th century.

However, the advantage of rapid top-down advancement through the power of royal
courts exposed the Buddhist royal cult leaders to equally sudden reversals of fortune.
 6

Several of these Buddhist communities suffered setbacks like that reported by Yijing
in seventh century Zhenla. In broad brush terms there is only space here to refer to the
Buddhist ailendras suddenly disappearance from Java in the ninth century; the
closing of 4,000 monasteries and burning of books and vestments burned in ninth
century China; the purging of Mahāyāna cults in Sri Lanka in the ninth century and
from Burma in the 12th century; while Buddhism was crushed in Tibet, before
returning in a second triumphal wave a century later.

Closer to Zhenla in the eighth and ninth centuries, indeed next door in ān, a hoard of
200 superbly modelled and cast Bodhisattvas found recently at Prakhon Chai attests to
a perhaps regal cult of Bodhisattvayāna, possibly in the shadowy state ‘ r Canā ā’
mentioned in inscriptions. Nothing more is yet known of this community that
suddenly re-entered history with the discovery of the buried bronze masterpieces.21 In
Cambodia proper, Jayavarman II left no inscriptions in this period but is credited by
succeeding kings with founding Cambodia and asserting its independence in a tantric
aiva ‘d varāja’ cult in 802. For the next 150 years the Khmer Buddhist community
left little sign of its existence.22 At this time the Vedantic theologian aṇkara was
travelling throughout India preaching against Buddhism and it cannot be excluded
that his influence was felt in Cambodia, for ivasoma, a cousin of Jayavarman II,
studied the āstras under him and returned to Cambodia, where he eventually became
chief minister to Indravarman I. 23

Ninth century
Ninth century Champa had an early Vajrayānist cult at a major new Buddhist complex
at Đông-dừờng. In ān and the peninsular communities, bronze tantric deities were
produced in the ninth century. Similar icons were produced in Pāla Bengal and Java at
this time and Hiram Woodward calls the peninsular bronzes ‘Bengali-influenced’.24 In
Cambodia, a ninth century Buddhist patron and guru named r Satyavarman is
identified in two tantric inscriptions as the creator of a foundation at Sak Bāk village
near Phimai, in modern Thailand, and one at Wat Sithor southeast of Angkor. This
pioneering Buddhist appears from the inscriptions to have been active in the ninth
century in the eastern and western wings of the area dominated by the Khmer
Buddhist ‘Mah dhara’ families, which eventually came to power in 1080 through the
usurpation of Jayavarman VI.

10th century
The Khmer Buddhists received a major boost in the 10th century at the Renaissance-
style court of king Rājendravarman, where Buddhism was tolerated as a secondary
creed. Rājendravarman was a powerful and cultivated aiva king 25 who presided over
a Renaissance-style court that contained active communities of aivas, Vai ṇavas and
Buddhists. The king allowed his senior minister, general and architect
Kav ndrārimathana to build the elegant, moated Buddhist temple of Bat Čum amid the
palaces and villas of the regenerated city. The small brick and stone temple is
dedicated to the Buddha, Vajrapāṇi and Praj āpāramitā.26 The icons of Bàt Čum are
lost but the temple probably housed a Buddha enthroned on a multi-headed giant
Nāga, as this is the principal Buddha icon of this period, which appears both in the
round and on dated 10th century caityas in the Phnom Penh National Museum and the
Musée Guimet, Paris. These early Khmer Nāga-enthroned Buddhas, of which a dozen
survive, bear striking similarities-- notably the long separated necks of the serpents
and their crested heads -- to the icons found in the Mahāyāna communities that dotted
 7

the Khmer foreign trade route to the pre-Thai peninsula. No known inscription
explains why the Khmer Buddha in dhyāna mudrā is enthroned on a Nāga but there
contexts suggest it may represent Vairocana, the supreme Buddha in this period in
China and Japan. No Khmer evidence I have found comforts the widely held view
that this image is connected with the Mucalinda myth known from the biographies of
the akyamuni and propagated broadly in Southeast Asia’s Theravāda in later
centuries. The cult of Bàt Čum is clarified in an inscription a few years later that
recounts the principal Buddhist scholar at court, K rtipaṇ ita, sending abroad for
tantric texts, which included the STTS, the root tantra of the Yoga class in the Tibetan
classification.

Coedès published the Romanized Sanskrit of the crucial stanza which establishes the
presence of the Tantras in Cambodia as:
B7-8 lak agraṇṭham abhiprajñaṃ yo nve ya pararā ṭrataḥ
tattvasaṅgrahaṭ kādi- tantra cādhyāpayad yam //

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 Tattvasaṅgraha commentary, this sage then spread the
study of them].

Whereas a more literal translation gives:

Having searched in a foreign kingdom for one hundred thousand27 book(s) of

higher wisdom, the self-restrained one [sage] taught the Tantra teachings
(tantram) of such texts such as the Tattvasaṅgraha and its commentary.28

The word ‘Tantra’ can be used to describe chapters in texts and Coedès translates the
word ‘Tantra(s)’ in the broad sense of ‘traités’ (treatises). Having rendered
abhiprajñaṃ (‘higher wisdom’) as ‘philosophiques’, he goes on to indicate the
doctrinal basis of lines B27-8 as simply ‘Le Mayāhāna’Ś29

B27-8 advayānuttaraṃ yānam anye ām svam ivārjjayan

yo di an munaye haimaṃ rājataṃ ivikādvayam //

Procuring for others as if for himself the nondual (advaya) and supreme
(anuttara) vehicle (yāna), he bestowed on the Sage (muni) a pair of golden and
silver palanquins ( ivikā).

[Coedès: Procurant aux autres, comme à lui-même le véhicule suprême et sans

second, il consacra au Muni deux litières en or et d’argent].

Coedès’ evaluation of this Buddhism has naturally been influential and his translation
has gone unchallenged for many decades. Yet the words of the opening Sanskrit
compound advaya-anuttara-yāna (nondual-supreme-vehicle) of this stanza suggest
the third Buddhist vehicle, the Vajrayāna, rather than the much broader and older term
‘Mahāyāna’ that the later vehicle enormously enhanced in terms of ritual, liturgy and
 8

text over many centuries. Furthermore, the Wàt Sithor text indicates that K rtipaṇ ita,
in teaching the Tattvasaṃgraha, a common abbreviation in Indian sources for the
Sarvatathāgatatattvasaṃgraha (STTS), had a preference for this the principal
scripture of the Yogatantras.30

Three major works are known by the short name Tattvasaṃgraha ‘compendium of
truth’.31 Kamala la’s commentary on āntaraksita’s manual of Mahāyānist doctrines
is called Tattvasaṃgraha-pañj ā, but the commentary brought in by K rtipaṇ ita,
according to Wat Sithor, was the ‘tattvasaṃgraha-ṭī ā’. Tibet’s 17th century historian
Tāranātha has the support of modern scholars in identifying the Tattvasaṃgraha-ṭī ā
as the short name for the Kosa ā am āra-tattvasaṃgraha-ṭī ā.32 This ṭī ā is not a
commentary on āntaraksita’s academic manual of Mahāyāna doctrines but rather the
leading commentary on the Tattvasaṃgraha-tantra itself. The STTS is the third work
which shares the same short name. The author of the ṭī ā was identified by Tāranātha
as ākyamitra, who lived in the late 9th century33 and composed the Kosa ā am āra-
tattvasaṃgraha-ṭī ā in his home town of Kosala during the reign of Devapāla, the
successor of Gopāla, the eighth century founder of the Pāla dynasty. 34 In
K rtipaṇ ita’s time, ākyamitra’s ṭī ā was a major work of recent scholarship –
indeed the current scholarly classic that sought to elucidate the dominant yogatantra
that was playing the pre-eminent role in advancing Vajrayāna throughout South, East
and Southeast Asia.

The upshot of this small textual clarification is significant. It means that Wat Sithor’s
Sanskrit compound ‘tattvasaṅgrahaṭī ād -tantram’ identifies the root tantra of the
Yoga class (in the Tibetan classification), and its major commentary, as the textual
platform for the Khmer Buddhist revival. If we think of K rtipaṇ ita bringing such
texts into Cambodia to re-launch Buddhism, we envisage a mission invested with the
vigour and ambition of the gurus Buddhaguhya and ākyamitra, who had already
contributed much to making tantric Buddhism a success in Tang China. This is far
removed from the kind of scholastic enterprise based on the textbooks of āntaraksita
and Kamala la that Cœdès and subsequent historians have assumed.

11th century
A century later, a new dynasty from the as yet unidentified Khmer city of Mah dhara
took power and was to sustain itself with only brief interruptions until the late 13th
century. Mah dharapura was a city presumably near Phimai, where a vast temple
complex was constructed by King Jayavarman VI in the last two decades of the
century. Although the state religious calendar in Angkor was presumably still
honoured at the pyramidal aiva Bàphûon temple in Angkor, the construction of the
Phimai tantric Buddhiust complex appears to indicate that the centre of political
power had moved at least in part from Angkor to Phimai, where a temple complex on
an unprecedented scale set the standard for Angkor’s greatest builders, kings
Sūryavarman II and Jayavarman VII. The central sanctuary at Phimai leaves no doubt
about the Tantrism of the Buddhist Mah dharas. Although the central icon is not
known, around the inner sanctum of the temple are deeply-carved reliefs of the tantric
Heruka Samvara, a crowned (Pāla-derived?) Buddha and what looks like a version of
Vajrasattva and the Vajradhātu Pentad.
 9

12th century
Phimai was built from c.1080 onwards and dedicated in 1108 and no doubt remained
a powerful religious foundation for many years. The enduring importance of the
Phimai Buddha to the dynasty is reflected in the prominent place it was afforded 80
years later in Jayavarman VII’s court spring festival. It was also singled out for
offerings by his chief queen Jayarājadev , when she set up a Buddhist foundation for
abandoned girls. Jayavarman himself enlarged Phimai and built a chapel for his own
portrait statue venerating the god of the central sanctuary.

But within a decade of the completion of the Phimai complex, Sūryavarman II came
to power in a coup against his own kinsmen and began the construction of the
extraordinary Vaisnava shrine we now know as Angkor Wat. Sūryavarman’s devotion
to Visṇu, whose role in Khmer religious art had been minor since pre-Angkorian
times, remains something of a mystery.35 It is worth noting that Vai ṇavism, boosted
by reforming guru Rāmānuja, bloomed in India in the early 12th century and the great
Jagannātha temple rose at Puri, Orissa, exactly when Angkor Wat was being built.

Although this survey attests the presence of Buddhism in Cambodia for seven
centuries before he came to power, Jayavarman VII made a major break with tradition
when he made Buddhism the state cult for the first time. The decision probably owed
something to the extraordinary political circumstances of his accession and
presumably reflected his personal convictions. He eventually chose the posthumous
title Mahāparamasaugatapada (‘who has attained the domain of those who are
supremely devoted to the Buddha’).36 The record is far from clear, but Angkor was
attacked by a Cham army and the usurper Tribhūvanādityavarman, who had
overthrown Yo avarman II in 1167, was killed. Never before had a foreign army
penetrated the military and spiritual defences of the Khmer capital. The revamp of the
city defences undertaken by the new general-king Jayavarman, who reached Angkor
in 1182, probably with another Cham army, included massive defensive earthworks,
moats and walled compounds. Even though Jayavarman’s cult was syncretic, with
large sections of his temple complexes dedicated to Brahmanical rites for iva and
Vi ṇu, the central sanctuaries were emphatically Buddhist. His move to assert his own
creed as supreme over the aivism of the long-entrenched Brahmanical aristocracy
must have entailed a blend of conviction, contingency and adventurism. He inherited
a mix of religious traditions and sought a balance of power among them that gave him
ultimate control. Syncretism was then a matter of realpolitik rather than ecumenism,
as Davidson observes of feudal India:
…the patronage of many kings towards multiple religious traditions could be viewed in part as
an attempt to keep them all beholden to the ruler – and divided from one another – rather than
a dedicated catholicity among Indian monarchs. 37

Socio-religious reengineering
In taking the significant risk of reducing the power of the court Brahmins and
challenging the status quo of the entrenched Brahmanical aristocracy, Jayavarman
appears to have gambled on winning popular support by defending the people in a
large socio-engineering plan. A network of walled complexes directed by monks and
officials with guaranteed supplies, water, transport and artisans was constructed in
Angkor, and across the empire, within a decade. Louis Finot describes these multi-
purpose communities as ‘complex establishments …which were at the same time
temples, convents, universities, and no doubt fortresses when needed, capable of
 10

protecting the population and withstanding a siege.’ 38 Later there are signs that he
sought to embed the Buddhist root through mass conversions and consecrations. The
largest community of all was presumably lodged around the new Bayon state temple,
within the new earth ramparts of Angkor Thom. Compared with Suryavarman’s
remote Vai ṇava rituals on the small, elevated, third-level platforms deep within
Angkor Wat, Jayavarman’s new Buddhist temples were a departure into open, city-
centre architecture. Paul Mus contrasts the remote and ‘ordered splendour of Angkor
Wat, evoking Versailles’, with the massive presence of the Bayon open to the heart of
the city, without moat or enclosure wall:
…the time for religious and artistic academism had passed. The most audacious forms from
the mystical experience and the sculptural and architectural expressionism appropriate to the
Great Vehicle of Buddhism had taken hold of Cambodia and were refashioning it in
announcing – fallaciously in fact – a worthy destiny…Now [the sovereign’s] portrait breaks
out everywhere. He throws himself before his people, he exposes himself to defend them,
risking all…hence this art with a difference, in the service of a new creed, because the politics
have changed, in a disastrous and revolutionary situation that had to be faced. 39
These large, temple-centred popular complexes were the base units of an ambitious
politico-religious re-engineering programme that helped turn Cambodia into a
Buddhist society.

Jayavarman VII’s icons

Jayavarman, on taking power, began erecting triads of the Nāga-enthroned Buddha,
Loke vara and Praj āpāramitā. In his second decade, with the national defences
secured against Cham adventures and internal rebellion, he committed himself to
foreign expansion. This went beyond restoring Khmer control over parts of Champa
and halted only at the borders with the Đai Vi t, Burma and r vijaya. During this
period the triads made way for a broadly-disseminated iconographic campaign in
which the temples and the empire underwent what Philippe Stern called
‘Lokeçvarisation’. Standing Loke varas were inserted into the central sanctuary of the
aiva healing spa of Neak Paen in Angkor and large eight-armed sandstone and
bronze images of Loke vara, along with portrait statues of the king, were distributed
to the religious power centres in the major cities. Preah Khan (‘sacred sword’) temple
in Angkor, which Jayavarman called Jaya r nagar , ‘city of victorious fortune’, was
built around a central shrine to ārya-ava o t a surrounded by 238 deities.40 Stern
went as far as to describe Lokeçvarisation as a ‘brutal blow, which would give
“Lokeçvara” first place at the transition to the second period.’41 The powerful yet
contemplative Khmer image of the ‘Lord who looks down’ became paramount in
religious art – reproduced in more icons than the Buddha. These Loke varas are an
artistic breakthrough in rendering a new naturalism and quietism in a vigorous human
frame. Boisselier memorably describes their lowered eyelids and soft, rapt smiles:
An impassive ideal of superhuman beauty 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.42
The lowered eyelids that give the images this arresting human presence became the
hallmark of ‘Lokeçvarisation’.

The final phase of the king’s reign is signalled in the art of the new state temple, the
Bayon, where the decoration indicates a move to public displays of the Vajrayāna of
the kind first seen at Phimai. Yogin dancers in Hevajra’s ritually charged tantric
 11

ardhaparyanka posture appear in their thousands carved in relief entablatures on the

gopura and columns that make up the approaches to the Bayon. Similar Yogin s whirl
around Hevajra in bronzes that are three-dimensional mandala based on the liturgy of
the Hevajra-tantra. One of the late Bayon style Hevajras was indeed excavated from
the ruins of the royal palace by Groslier in 1952, suggesting the wrathful emanation of
the supreme tantric deity had become the king’s personal deity or ṭad vatā.43
Moreover halls with the dancer emblem were added to the earlier temples like Preah
Khan, Tà Prohm and Bantéay Kdei and their collective impact on the ritual spaces of
the temples towards the end of the reign is such that we might extend Stern’s model
and call it ‘Yogin fication’. Were these halls designed for large scale Hevajra
consecrations? The large number of bronze lustration conch holders and other ritual
paraphernalia in the world’s museum collections, which bear Hevajra’s embossed
image, support this interpretation. Indeed from the material record left by this
movement, Hevajra may have reached his world apogee in Cambodia at this time.44
This wrathful emanation of Vajrasattva also appears in stone. Two large surviving
sandstone statues of Hevajra – one intact one with eight heads, 20 arms and four feet,
a larger (three metre) and finer but heavily damaged one, the bust of which is now in
the New York Metropolitan Museum -- were found, respectively, near the west and
east gates of Angkor Thom (Plates 7 & 8) .

Plate 7 Hevajra bust, NY Metropolitan. Plate 8 Hevajra, West Gate, Siem Reap
Conservatory
 12

I recently located the legs of the New York Hevajra bust in the forest outside Angkor
Thom and these have now been moved by the Khmer authority to the Siem Reap
conservatory. This has enabled us to reconstitute electronically how the New York
Hevajra may originally looked, using the bust and legs and blowups from Khmer
Hevajra bronzes. (Plate 9)

Plate 9 Hevajra reconstituted electronically

Vajrasattva and his Heruka emanations Hevajra and Saṃvara appear together in a
Phnom Penh museum bronze mould found at Poipet, near the modern Thai border.
They dance on corpses beside a seated, six-armed, three or 4-headed faced figure who,
at the centre of such a mandala, can only be Vajrasattva. (Plate 10) I propose we see
the same supreme deity in the face towers of Jayavarman’s temples.

Plate 10 Bronze mandala mould from Poipet, Phnom Penh National Museum
 13

A similar triad is carved in relief on the wall of a sanctuary in Jayavarman’s large

Bantéay Chmar temple in this region. I suggest that these elements coalesce into a
body of evidence that attests the existence of a royal Vajrayāna cult, focused on
Hevajra/Vajrasattva. This puts in question the current consensus of scholars,
according to which there are tantric elements in Jayavarman’s Vajrayāna, but they are
peripheral to his core creed. I would propose instead that his Hevajras, Yogin s,
Praj āpāramitās and Vajrasattvas as eventually occupying centre stage in a royal
Hevajra cult that enabled the king to master political turmoil and maintain empire,
like his tantric forebears in feudal India.

Conclusions
From c.600-1080 Khmer Buddhism survived as a secondary creed, despite repression
and only sporadic royal patronage. In 1080 the Mah dhara dynasty established
Buddhism as the power behind the throne and in 1182 Jayavarman VII made it the
state religion. The Buddhism of the Khmers was mostly tantric from 600-1300. At
some as yet undetermined point, probably in the late 13th century, the Khmers then
turned to a Thai- or Burmese-influenced Theravāda. In the 16th century an energetic
Khmer king retuned to Angkor and refitted it on a major scale as a regional
Theravādin Buddhist pilgrimage centre. Jayavarman’s strategic shift to Buddhism
thus proved permanent among the Khmers, though the nature of the cult rapidly
evolved far away from his own.

How the Theravāda arrived in Angkor remains obscure. Pali inscriptions and temples
with Theravādin iconography suddenly appear at the end of the 13th century,
strategically positioned beside the royal palace. And an eyewitness imperial Chinese
emissary in 1296 describes a village level Buddhism which sounds Theravādin in
spirit and ritual. He mentions no role for the huge Vajrayāna stone temples built by
Jayavarman. Jayavarman's form of Buddhism seems to have declined fast, though the
leaders of nascent Ayutthaya sought to capture its imperial sway by occupying
Angkor and taking many Angkorian Buddhist features into their Theravādin state
Buddhism.

1
There is abundant evidence today in Cambodia of central government routinely establishing formal
links with neak ta when officials are installed in the provinces. (Ang Chouléan 1986:217 Les êtres
surnaturels dans la religion populaire khmère Cedoreck, Paris). By ministerial decree, local
administrators must take off their hats every time they pass the ana tā spirit house. Certain neak ta
have even achieved national renown. In a story still widely related in Cambodia, king Ang Chan
(r.1516-66) was on the verge of defeat by a Siamese army, when General Moeuṅ proposed to commit
suicide in order to generate an army of ghosts to support the Khmers. Before the whole army, the
general threw himself into a ditch filled with stakes, followed by his wife and sons, and their violent
deaths were believed to have resulted in the Siamese army succumbing to cholera and being defeated.
The general was venerated as neak ta Ghlāmng Mioeng ‘and the popularity of this spirit extended
through the whole of Pursat province and even throughout Cambodia.’ (Chouléan 1986:207)
2
Groslier B.P. (1973:305) Le Bayon: inscriptions du Bayon EFEO Paris
3
Dagens, Bruno (2000:107) ‘The Bayon Face Towers and their Meaning’ 5th annual Bayon Symposium
Siem Reap
4
Jacques, C. (2007:192) The Khmer Empire River Books, Bangkok
5
Nancy Dowling (2000:122-155) ‘New light on early Cambodian Buddhism’ in The Journal of the
Siam Society vol. 88.
6
Yijing (1896:12) trans. J Takakusu A record of the Buddhist Religion as Practised in India and the
Malay Archipelago (A.D. 671-695) Clarendon, Oxford.
 14

7
Mirielle Bénisti (1969Ś109) ’Recherches sur le premier art khmerŚ«la bande a chatons», critère
chronologique?’ Arts asiatiques XX.
Jean Boisselier reached a similar chronology by analysing the centre-folded dhoti. Jean Boisselier
(1957Ś272) ‘A propos d’un bronze cham inédit d’Avalokiteçvara’ Arts asiatiques IV
8
Vajrayāna and tantric Buddhism are used as synonyms here, although the second is normally
considered broader because it includes the Mantrayāna or Mantranaya (‘path of mantras’), whose
practitioners before the seventh century did not see themselves as distinct from Mahāyāna.
9
Lin Li-Kouang (1935:83-100) ‘Puṇyodaya (N’ati), un propagateur du Tantrisme en Chine et au
Cambodge à l’époque de Xuanzang’ Journal asiatique Juillet-Septembre 1935.
10
Leon Wieger (1922:536) Histoire des croyances religieuses et des opinions philosophiques en Chine
Beijing
11
Lin Li-Kouang (1935:83)
12
Lin Li-Kouang (1935:89).
13
Sanderson (2005:401)
14
Cœdès, G (1968:96) The Indianized states of Southeast Asia trans. S Dowling University of Hawaii
15
Chaiya in the 8th century flourished on the maritime trade network taxed and policed by a federation
of r vijayan ports. Three brick temples were erected there by the King of r vijaya, according to the
‘Ligor’ inscription of 775 AD. Mahāyāna Buddhist masterpiece bronzes were excavated at Chaiya. As
there was no prior bronze-casting tradition on this isthmus tract the bronzes may have been imported
from Java, where a similar Bodhisattva was found from this period.
16
Vajrabodhi left Nālandā to study in south India for seven years at the feet of the tantric master
Nāgabodhi (Tajima 1936Ś22).
17
In 719 Vajrabodhi spent 5 months in Java waiting for the monsoon winds to take him to China. He
was received with great honour by the king and met a 14-year old monk Jnāngarbha(?), who took the
name Amoghavajra and who was to remain at his side until his death in 741.
18
Tajima, R (1936:23) Étude sur le Mahavairocan-sūtra (Da n ch yō) Librarie d’Amérique et
d’Orient, Paris
19
Mahāva rocanā h sam odh v urv tādh sthānava pu yasūtr ndrarājanāmadharmapararyāya.
20
Davidson, R (2002:115) Indian Esoteric Buddhism: a social history of the Tantric movement
Columbia University Press, New York
21
Hoshino thinks there could have been an overland link between this Mon-Khmer community of ān
and Siem Reap in northwest Cambodia where the first inscription to the erection of a Bodhisattva is
dated 791. It seems to me possible that the first Prakhon Chai Buddhists may have been among those
expelled from Sambor Prei Kuk in the late seventh century, as recorded by Yijing. Hoshino, Tatsuo
(1986 :33) Pour une histoire médiévale du moyen Mékong Editions Duang Kamol, Bangkok
22
Pierre Dupont commentsŚ ‘La première partie de l’époque angkorienne, qui débute en 802, est
marquée par le déclin, voire la disparition du bouddhisme. Tout au plus le roi Yoçovarman (889-900)
réglemente-t-il le fonctionnement d’un couvent dans la région d’Angkor...Nous n’avons aucune idée de
ce que pouvaient être en ce temps les images du Buddha. C’est un demi-siècle plus tard que le
bouddhisme prend in développement nouveau, sous les rois Rājendravarman (944-968) et Jayavarman
V (968-1001).’ ‘Les Buddha sur Nāga dans l’art khmer’ (1950Ś39) Artibus Asiae 13
23
Briggs, L.P. (1951BŚ234) ‘The syncretism of religions in Southeast Asia, especially in the Khmer
empire’ Journal of the American Oriental Society 71, Yale University Press, New Haven Connecticut
24
Woodward, Hiram (2003:92-93) The Art and Architecture of Thailand Brill, Leiden, Boston
25
Rājendravarman is not only credited with restoring the capital and restructuring the state into
provinces he is referred in inscription K.806 v.143 as being expert in the 18 aiva topics of legal
business (vyavahārah) (astāda apadajñ na). Alexis Sanderson (2003-4Ś381n.113) ‘The aiva religion
among the Khmersś Part I’ BEFEO Paris
26
In the Bàt Čum inscription she is also referred to as ‘Divyadevi’ (celestial goddess) and ‘Devi’.
27
Lak a or 100,000 was conventionally used for large indistinct numbers. In the Chinese canon the
Prajñāpāram tā sūtras are said to have consisted of 100,000 gāthās or lokas of 32 syllables (Kwon
2002:27).
28
This is from a new selective translation of the Wàt Sithor inscription, published for the first time at
the end of this paper, for which I am beholden to Dr Tadeusz Skorupski, Reader in the Study of
Religions, SOAS.
29
Footnote (3) p.206 IC VI.
30
In this preference for the Yoga Tantras, K rtipaṇ ita was in the footsteps of China’s great Buddhist
sage Amoghavajra, who is counted by Japan’s Shingon sect as the sixth patriarch of Sino-Japanese
esoteric Buddhism. Amoghavajra was born in India or Sri Lanka and followed his guru Vajrabodhi to
 15

China. After Vajrabodhi’s death he left China to acquire a copy of the STTS in Sri Lanka and
subsequently translated it into Chinese in the late eighth century for the Tang emperor. Amoghavajra
said in his final testamentŚ ‘The great doctrine, in its totality and in its particulars, is vast and deep!
Who can fathom the source of the Yoga-tantra?’ Orlando comments on thisŚ ‘The term in its most
particular sense refers to the Yoga-tantra, the esoteric texts regarding the Vajradhātu or “Diamond
Realm”…In a more general sense, the term yoga in Esoteric Buddhism means “to concentrate one’s
mind in order to harmonize with the supreme doctrine and to identify with the deity one worships.”
Hence all the rites performed by the monks in this sect, whether simple or complicated, are called yoga,
because these rites are the means to identify oneself with the deity…’ Orlando, Raffaello (1981:106) A
study of Chinese documents concerning the life of the tantric Buddhist patriarch Amoghavajra (A.D.
705-774) Princeton
31
Alexis Sanderson (2003-4Ś427) ‘The aiva religion among the Khmersś Part I’ BEFEO Paris
32
Chattopadhyaya, Lama Chimpa Alaka (1990:270) Tāranātha’s H story of Buddh sm n Ind a Motilal
Banarsidass, Dehli; Mkhas-grub-rje, Rgyud-sde-spy ’ -rnam-par-gzhag-pa-rgyas-par-brjod translated
by F.D. Lessing and A. Wayman (1968:25) as Introduction to the Buddhist Tantric Systems, The
Hague; Do-Kyun Kwon (2002:25) STTS a study of its origin, structure and teachings PhD thesis,
SOAS, who also cites Matsūnaga Yukei’s M yō R sh p. 68
33
Kwon 2002:25; Winternitz, M. (1932:396) A History of Indian Literature University of Calcutta
La Vallée Poussin agrees the mid-ninth century date and notes ākyamitra appears to have added a
(signed) chapter to the Pañcakrama attributed to Nāgārjuna. Louis de la Vallée Poussin (1896:IX)
Etudes et textes tantriques Pañcakrama Engelcke, Gand
34
Tāranātha (1608Ś274-283) History of Buddhism in India
35
Apart from Sūryavarman II, who took the posthumous title Paramavi ṇuloka (‘gone to the paradise
of Vi ṇu’), only one other Khmer king is known to have been a Vi ṇu devotee – Jayavarman III (r.835-
before 877) – who took the posthumous title Vi ṇuloka (K.256A v.6).
36
Sanderson (2004:429)
37
Davidson 2002:192
38
Finot, M.L. (1925Ś239) ‘Loke vara en Indochine’ Etud s As at qu s pu é s a ’occas on du 25eme
ann v rsa r d ’EFEO, EFEO, Paris
39
Mus, P. (1964:31-32) ‘Un cinéma solideŚ l’intégration du temps dans l’art de l’Inde at dans l’art
contemporainŚ pouquoi?’ Arts Asiatiques X, A. Maisonneuve, Paris
40
K.908 A69, 70 āryāva o t asya mahyamasya samantaḥ / atadvayan trayo īt s t na d vāḥ
prati ṭh tāḥ. (‘Around the holy Avalokite a the king erected 238 deities’).
41
Stern, Philippe (1965:134) Les monuments khmers du style du Bayon et Jayavarman VII
Universitaires de France, Paris
42
Boisselier, J (1978Ś327) ‘Cambodia’ in The Image of the Buddha ed. D. Snellgrove, UNESCO
43
Groslier, Bernard Philippe (1954Ś229) “Fouilles du Palais Royal d’Angkor Thom” in Proceedings of
the 23rd International Congress of Orientalists Cambridge 1954 Royal Asiatic Society London
44
This possibility is attested by Rob Linrothe (1999:274) Ruthless compassion: Wrathful Deities in
Early Indo-Tibetan Esoteric Buddhist Art Serindia, London.
_________________________________________________________________________________
 16

Bibliography
Ang Chouléan 1986 Les êtres surnaturels dans la religion populaire khmère Cedoreck, Paris
Bénisti, Mirielle (1969) ’Recherches sur le premier art khmerŚ«la bande a chatons», critère
chronologique?’ Arts asiatiques XX
Boisselier, Jean (1978) ‘Cambodia’ in The Image of the Buddha ed. D. Snellgrove, UNESCO
Briggs, L.P. (1951BŚ234) ‘The syncretism of religions in Southeast Asia, especially in the Khmer
empire’ Journal of the American Oriental Society 71, Yale University Press, New Haven Connecticut.
Chattopadhyaya, Lama Chimpa Alaka (1990) Tāranātha’s H story of Buddh sm n Ind a Motilal
Banarsidass, Dehli
Cœdès, George (1968) The Indianized states of Southeast Asia trans. S. Dowling University of Hawaii
Cœdès, G. (1954:195-211) Inscriptions du Cambodge VI EFEO Paris
Dagens, Bruno (2000) ‘The Bayon Face Towers and their Meaning’ 5th annual Bayon Symposium Siem
Reap
Davidson, Ronald (2002) Indian Esoteric Buddhism: a social history of the Tantric movement
Columbia University Press, New York
Dowling, Nancy (2000:122-155) ‘New light on early Cambodian Buddhism’ in The Journal of the
Siam Society vol. 88
Dupont, Pierre (1950:39-62) ‘Les Buddha sur Nāga dans l’art khmer’ ArtibusAsiae 13
Finot, M.L. (1925) ‘Loke vara en Indochine’ Etudes asiatiqu s pu é s a ’occas on du 25 m
ann v rsa r d ’EFEO EFEO, Paris
Groslier, Bernard Philippe (1954:228-9) ‘Fouilles du Palais Royal d’Angkor Thom’ in Proceedings of
the 23rd International Congress of Orientalists Cambridge 1954 Royal Asiatic Society London
Groslier B.P. (1973:305) Le Bayon: inscriptions du Bayon EFEO Paris
Hoshino, Tatsuo (1986) Pour une histoire médiévale du moyen Mékong Editions Duang Kamol,
Bangkok
Jacques, Claude (2007) The Khmer Empire River Books, Bangkok
Kwon, Do-Kyun (2002) STTS a study of its origin, structure and teachings PhD thesis, SOAS
la Vallée Poussin, Louis de (1896) Etudes et textes tantriques Pañcakrama Engelcke, Gand
Lin Li-Kouang (1935:83-100) ‘Puṇyodaya (N’ati), un propagateur du Tantrisme en Chine et au
Cambodge à l’époque de Xuanzang’ Journal asiatique Juillet-Septembre 1935
Linrothe, Rob (1999) Ruthless compassion: Wrathful Deities in Early Indo-Tibetan Esoteric Buddhist
Art Serindia, London
Mus, Paul (1964:21-33) ‘Un cinéma solideŚ l’intégration du temps dans l’art de l’Inde at dans l’art
contemporainŚ pouquoi?’ Arts asiatiques X, A. Maisonneuve, Paris
Orlando, Raffaello (1981) A study of Chinese documents concerning the life of the tantric Buddhist
patriarch Amoghavajra (A.D. 705-774) Princeton
Sanderson, Alexis (2003-4:349-462) ‘The aiva religion among the Khmersś Part I’ BEFEO Paris
Stern, Philippe (1965) Les monuments khmers du style du Bayon et Jayavarman VII Universitaires de
France, Paris
Tajima, R (1936) Étude sur le Mahavairocan-sūtra (Da n ch yō) Librarie d’Amérique et d’Orient,
Paris
Wieger, Leon (1922) Histoire des croyances religieuses et des opinions philosophiques en Chine
Beijing
Winternitz, M. (1932:396) A History of Indian Literature University of Calcutta
Woodward, Hiram (2003) The Art and Architecture of Thailand Brill, Leiden, Boston
Yijing (1896:12) trans. J Takakusu A record of the Buddhist Religion as Practised in India and the
Malay Archipelago (A.D. 671-695) Clarendon, Oxford.

_____________________________________________________________________________

List of plates (author’s own)

Plate 1 Lokesvara of Ben Mealea (Siem Reap Conservation)
Plate 2 Bayon face towers
Plate 3 Uncovered Bayon-style Loke vara lintel
Plate 4 Portrait statue of Jayavarman VII, Phnom Penh National Museum
 17

Plate 5 Bayon face tower (Vajrasattva?)

Plate 6 Hevajra bronze (Phnom Penh National Museum)
Plate 7 Hevajra bust, NY Metropolitan Museum of Art
Plate 8 Hevajra, West Gate Angkor Thom, Siem Reap Conservation
Plate 9 New York Metropolitan Hevajra reconstituted electronically
Plate 10 Bronze mandala mould from Poipet, Phnom Penh National Museum