Hundreds of thousands of people bound to the service of temples: 79,365 for Ta Prohm in 1186 and 306,372 for all Jayavarman VII's foundations in 1191, war captives among them; household slavery fed by the purchase of hill peoples, with runaways branded and fettered; the first Theravāda monastery endowed with slaves in 1308; and thousands of Buddha images chiselled out or recut as liṅgas in the Śaiva reaction that followed Jayavarman VII.
CONNECTIONS · 802–1432 · RELIGION · From Early medieval Indic religious traditions → Angkorian Khmer

Indian gods at Angkor (c. 900) — and the Khmer turn to Theravāda

In 802 a brahmin consecrated a Khmer king on Phnom Kulen with a rite taken from four Indian tantras. For four centuries Angkor's kings built for Śiva, Viṣṇu and the Buddha, and assigned hundreds of thousands of people to the gods' service. Then a monk's religion came from Sri Lanka by way of Burma, and Angkor Wat became a Buddhist shrine.

In 802, on the Phnom Kulen plateau, a brahmin named Hiraṇyadāma consecrated Jayavarman II with a rite drawn from four Indian Śaiva tantras, and founded the kingdom of Angkor. Around 900 Yaśovarman I built a new capital around the Bakheng hill for the liṅga of Śiva. For four centuries Khmer kings raised temples to Śiva, Viṣṇu and finally the Buddha, and assigned people to them on a vast scale: one stela of 1186 counts 79,365 in the service of a single temple. Then Theravāda monks, ordained in a Sri Lankan lineage and carried through Burma and the Mon country, won the villages. The first Pali inscription dates to 1308, the last Sanskrit one to 1327, and the monks' first monastery was endowed with slaves.

Several enormous carved stone faces with calm, faintly smiling expressions and diadems, built from stacked sandstone blocks on adjacent temple towers, against a clear blue sky.
Face towers of the Bayon, the state temple Jayavarman VII raised at the centre of Angkor Thom in the late twelfth and early thirteenth centuries, when Mahāyāna Buddhism briefly became the religion of the Khmer state. The serene faces are usually read as the bodhisattva Lokeśvara or as the king in his guise. A century later Buddhist images in the same temple were recut as Śiva-liṅgas.
Anandajoti Bhikkhu (Photo Dharma). Face towers of the Bayon, Angkor Thom, late 12th–early 13th century (photographed 2010). In situ, Angkor, Cambodia. CC BY 2.0 via Wikimedia Commons. · CC BY 2.0

Before: spirits of the soil and many small kings

Around 800, the Khmer-speaking country of the lower Mekong and the Tonle Sap basin had no single capital and no single king. It had a patchwork of rice-growing principalities, and it had already lived for several centuries with gods from India. Since at least the sixth century its rulers had dedicated temples to Śiva and Viṣṇu, written their foundation charters in Sanskrit and called themselves by Sanskrit names; how that first layer arrived, with the Brahmi script and the merchants of the Bay of Bengal, is the subject of a separate record in this atlas. Alexis Sanderson, the Oxford scholar of Śaiva scripture, describes the religion of the Khmer principalities from the fifth century to the fourteenth as three faiths of Indian origin, Śaivism, Pāñcarātrika Vaiṣṇavism and Mahāyāna Buddhism, and observes that "the three coexisted harmoniously for the most part but with Šaivism predominant" 2. Theravāda Buddhism, the religion of Cambodia today, was not yet among them.

Beneath and beside those Indian cults lay an older religion that never needed a temple charter. Ian Harris, whose Cambodian Buddhism (2005) is the standard history, describes the neak ta, the tutelary spirits of Khmer villages, as beings who "lack any explicit association with Buddhism from a purely doctrinal perspective" and yet hold "important and complementary functions at the level of popular ritual" 4. Following the colonial-era ethnographer Souyris-Rolland, he sorts them into three kinds: spirits of nature (of the hill, the water, the field, the forest), ancestral spirits, and spirits that began as Brahmanical gods and heroes and were absorbed into the village pantheon 4. Ang Choulean's Les êtres surnaturels dans la religion populaire khmère (1986) remains the fundamental French study of this world 5. Its shrines were stones, termite mounds, trees and posts; its priests were local; its gods belonged to a place rather than to a book.

The liṅga as heir to the stone

French scholarship of the 1930s proposed that the Indian cults had not displaced this religion so much as dressed it. Paul Mus, writing on the neighbouring Cham kingdoms, argued that the royal liṅga, the stone emblem of Śiva at the centre of a Sanskrit temple, had been grafted onto the cult of a soil-genius held in a raw stone or post: "il a fallu que le liṅga prît la suite des emblèmes bruts, poteau ou pierre, d'où le génie sortait quand on l'appelait selon les rites" 3. The liṅga, he concluded, would remain "l'équivalent du poteau, de l'arbre, de la pierre ou du tertre d'un dieu du sol" 3. Mus's evidence was chiefly Cham, and applying it to the Khmer extends his argument rather than proving it. Daniel Michon has since set out the two models that still divide the field: an "India-centric, 'Khmer Hinduism' model" and a "Khmer-centric, indigenous model", which he reconciles by distinguishing the religion of the elite from the practice of everyone else 6.

Some features of that older order survived inside the Sanskrit cults themselves. The best documented is matrilineal succession. When the founders of the Angkorian royal cult fixed the rules of its priesthood, they ruled that only "male hermits or women born in his matrilineal family" could serve as its priests, and the office passed in the female line, beginning with Śivakaivalya's sister's son 79. A goddess called Bhagavatī received images and festivals: the Sdok Kak Thom stela records a priest who set up "an image of Śaṅkaranārāyaṇa and an image of Bhagavatī" and gave them slaves, and the great stela of Ta Prohm in 1186 still celebrates the spring festival "suivant la tradition de Bhagavatī" 813.

What the Khmer country held around 800

The religious landscape on the eve of Angkor can be summarised in five layers, each attested in the sources:

  • Village spirits (neak ta) of hill, water, field and ancestor, with local shrines and no scripture 45.
  • Sanskrit cults of Śiva and Viṣṇu, patronised by princes since the sixth century, with brahmin priests and land endowments 12.
  • Mahāyāna Buddhism, present at court and in monasteries, a minority among the three Indian faiths 2.
  • Matrilineal priestly descent, which the Sanskrit cults absorbed rather than abolished 79.
  • Bonded labour attached to temples, the Old Khmer khñuṃ, already documented in the oldest inscriptions in the Khmer language 1.

What the country did not have was a single centre. Coedès's The Indianized States of Southeast Asia describes the eighth century as a time when the old kingdom of Chenla had broken into rival principalities 1. The man who would put them back together was a prince the later inscriptions call Jayavarman II.

The transmission: a rite on the mountain, a library of tantras

Mahendraparvata, 802

Jayavarman II left no inscription of his own. Everything known about his consecration comes from a stela set up two and a half centuries later, in 1052, at Sdok Kak Thom, today in Thailand's Sa Kaeo province. It is a family history: a bilingual Sanskrit and Old Khmer record of the priestly lineage of Śivakaivalya, which had served the kings of Angkor for some 250 years 7. George Coedès and Pierre Dupont edited it in 1943; Kamaleswar Bhattacharya retranslated the Sanskrit in 2009 and Chhany Sak-Humphry the Khmer in 2005 789.

According to the Khmer text, the king, whom it calls by his posthumous name Parameśvara, "came from Java to reign in the royal city of Indrapura", with Śivakaivalya as his guru 8. He moved his court several times and finally settled on Mahendraparvata, the "mountain of the great Indra", on the plateau of Phnom Kulen north-east of the later Angkor. There a brahmin performed the rite that founded the state:

"When the brāhmaṇa Hiraṇyadāma, proficient in the lore of magic power, came from Janapada in response to His Majesty's having invited him to perform a sublime rite which would release Kambujadeśa from being any longer subject to Java and which would enable the sole High Lords of Earth to be cakravartin, this Hiraṇyadāma celebrated a rite from the Vināśikha and established the Sovereign High Lord of the World." 8

Coedès dated the event: "we have abundant evidence that the effective beginning of the reign was 802" 1. Where Janapada lay, and what "Java" meant (the island, the Cham coast, or simply a foreign overlord), are still argued. What is not in doubt is what the brahmin brought. He taught Śivakaivalya four texts, "the holy Vināśikha, the Nayottara, the Saṃmoha and the Śiraścheda" 8. The Sanskrit verses of the same stela call them "the four faces of Tumburu", and say that from them the brahmin "contrived the magic rites bearing the name of Devarāja" 9. These are Indian Śaiva tantras of the Vāma stream, whose cult centred on Tumburu, a four-faced form of Śiva. The founding act of the Angkorian state was the performance, by an imported specialist, of a rite drawn from imported scriptures.

The mountain city itself was lost to sight for a millennium. In 2012 an airborne lidar survey covering 370 square kilometres "revealed an entire, previously undocumented cityscape" on Phnom Kulen, which Damian Evans and his colleagues identified with "the 8th- to 9th-century city named Mahendraparvata" 11. The same survey showed that greater Angkor's formal urban core covered at least 35 square kilometres, against the 9 inside the walls of Angkor Thom 11.

The god who is king?

The Sanskrit name of the rite, devarāja, produced the most durable argument in Khmer studies. Coedès read it as a cult of the "god-king": a royal liṅga, consecrated at the centre of each capital, in which the sacred essence of kingship resided, so that the king "appears less as an administrator than as a god on earth" 1. On this reading each of the great temple-mountains, from the Bakong to the Baphuon, was the shrine of a reign's royal liṅga 1.

Hermann Kulke attacked the reading in 1974 and, in English translation, in 1978. The inscription, he argued, "now leaves no doubt that the king of Cambodia was not, as previously believed, consecrated as god-king. On the contrary, this consecration took place around an image of the god Siva, who, as 'god, who is king' and as 'lord of the world' protected the kings of Cambodia" 10. The devarāja was a movable image of Śiva that travelled with the court, not the temple-mountain liṅgas; "the devaraja cult as the cult of the deified living 'god-king'", Kulke concluded, "never existed" 10. The Khmer text supports him on the point of translation: Sak-Humphry renders its kamrateṅ jagat ta rājya not as god-king but as "Sovereign High Lord of the World" 8. And the stela's own account of the cult's later life describes an object that moved: "Le dieu-roi changea de résidence suivant les capitales où le monarque le conduisit afin de veiller sur le pouvoir royal" 7.

The debate matters for this record because it changes what was transmitted. On Coedès's reading, India supplied an ideology of divine kingship. On Kulke's, it supplied something more specific and more practical: a protective tantric cult, the texts that prescribed it, and a priestly family licensed to perform it. The second is what the Sdok Kak Thom stela actually describes.

Brahmin families and their books

The rite of 802 created an institution that outlived every king who used it. The stela fixes the priesthood in Śivakaivalya's maternal line, and records its succession through his sister's son Sūkṣmavindu and on for generations 79. The same family history shows how texts travelled with teachers. Śivasoma, the guru of King Indravarman I, was himself a grandson of a Khmer king, and an inscription at Prasat Kandol Dom says he "a appris les śāstra (de la bouche) de celui qui est nommé Bhagavat Śaṅkara" 12. Whether this Śaṅkara was the Advaita philosopher Śaṅkarācārya, as has been proposed, cannot be proven from the inscription, but the claim itself, that a Khmer royal teacher had studied at the feet of an Indian master, is the point the inscription was making.

The library grew. Sanderson's study of Khmer Śaivism, the most detailed analysis of the Sanskrit corpus, traces the Saiddhāntika scriptures, the mainstream of Indian Śaiva ritual, in the Cambodian inscriptions; the oldest of them, the Niśvāsa, is already named in tenth-century Khmer epigraphy 2. Royal charters show the pull of Indian forms even in their script: the steles of Yaśovarman I's monasteries repeat the same text twice, once in the ordinary Khmer script and once in a north-Indian one 1. The named priests run through the whole period:

  • Hiraṇyadāma, the brahmin from "Janapada" who performed the rite of 802 8.
  • Śivakaivalya, royal chaplain to Jayavarman II, founder of the hereditary priesthood 7.
  • Śivasoma, guru of Indravarman I, who learned the śāstra from "Bhagavat Śaṅkara" 12.
  • Yajñavarāha, a royal grandson trained in grammar and Śaiva yoga, founder of Banteay Srei in 967 1.
  • Divākarapaṇḍita, "the indispensable Brahman Divākara", who crowned Sūryavarman II in 1113 1.

Contact with India itself was intermittent but real. Coedès cites the copper plates of the Chola emperor Rājendra I for a royal gift: "It may be that in 1012 or shortly afterwards, Suryavarman I, feeling himself threatened, solicited the aid of Rajendrachola I by making him a present of a chariot" 1.

Yaśodharapura, around 900

The Angkor of the maps begins with Yaśovarman I. His predecessor Indravarman I (877–889) had ruled from Hariharālaya, near modern Roluos, where in 881 he inaugurated the Bakong for the royal liṅga Indreśvara, "the first great monument in stone" in Coedès's phrase 1. Yaśovarman, who acceded in 889, founded in his first year "about a hundred monasteries (āśrama)", each with a royal pavilion and a stela of rules, built the Lolei temple in 893 in the middle of his father's reservoir, and then founded a new capital, Yaśodharapura, around the hill of Phnom Bakheng 1. The Sdok Kak Thom stela records that the family priest Vāmaśiva "érigea un liṅga sur le mont Śrī Yaśodharagiri" 7. The liṅga on the Bakheng was Yaśodhareśvara, the lord of Yaśodhara's mountain, and the city it anchored remained the Khmer capital, with interruptions, for some six centuries 1.

The temple-mountains of the next three centuries map the reception of Indian religion in stone:

Date Monument King Principal cult
881 Bakong, Hariharālaya Indravarman I Śiva, royal liṅga Indreśvara
c. 900 Phnom Bakheng, Yaśodharapura Yaśovarman I Śiva, liṅga Yaśodhareśvara
961 Pre Rup Rājendravarman II Śiva
967 Banteay Srei founded by the brahmin Yajñavarāha Śiva as Tribhuvanamaheśvara
c. 1113–1150 Angkor Wat Sūryavarman II Viṣṇu; funerary temple of the king as Paramaviṣṇuloka
1186 Ta Prohm Jayavarman VII Prajñāpāramitā, for the queen mother
1191 Preah Khan Jayavarman VII Lokeśvara, for the king's father
c. 1190s–1200s Bayon, Angkor Thom Jayavarman VII Mahāyāna Buddha and Lokeśvara

Viṣṇu at Angkor Wat, the Buddha at the Bayon

The cult was never fixed. Sūryavarman II, crowned in 1113, built the largest religious monument in the world for Viṣṇu; Coedès read Angkor Wat as a temple "constructed during the lifetime of the king for whom it was to serve as a funerary temple", in which the king would be deified under the posthumous name Paramaviṣṇuloka 1. Seventy years later Jayavarman VII (1181–c. 1220), who had driven a Cham army out of Angkor, made Mahāyāna Buddhism the religion of the state. Peter Sharrock describes this as the moment Angkor "turned definitively to state Buddhism ... after four centuries of state Śaivism", accompanied by "a very large, politically motivated campaign of tantric Buddhist initiations" 15.

A small green-patinated bronze of a crowned Buddha seated in meditation on three scaly coils of a serpent, with a fan of seven serpent heads rising behind him.
Buddha sheltered by a nāga, Khmer bronze of the twelfth century, Angkor period. The Buddha sits in meditation on the coiled body of the serpent, whose seven hoods spread behind his head: an Indian legend of the Buddha's enlightenment joined to the water spirits of Khmer religion, and the most characteristic Buddhist image of Angkor. The Metropolitan Museum of Art, 1987.424.19a, b.
Unknown Khmer artist. Buddha Sheltered by a Naga, 12th century, Angkor period. The Metropolitan Museum of Art, Gift of Cynthia Hazen Polsky, 1987 (1987.424.19a, b). Public Domain via The Met Open Access. · Public Domain

Jayavarman VII's foundations were Buddhist but still Indian in their categories. Ta Prohm was consecrated in 1186 for his mother in the form of Prajñāpāramitā, the Perfection of Wisdom; Preah Khan in 1191 for his father as the bodhisattva Lokeśvara 11314. The great face towers of the Bayon, at the centre of his new walled city of Angkor Thom, are usually read as Lokeśvara or as the king in the bodhisattva's guise. The protective nāga, the serpent of Khmer water cults, coiled beneath the meditating Buddha and spread its hoods above him: the most characteristic Buddhist image of Angkor joined an Indian legend to a local spirit.

What changed and what was replaced

A state organised around its gods

The Indian cults gave Angkor a way of turning land and people into offerings, and the inscriptions record the result with an accountant's precision. The Sdok Kak Thom stela shows each new liṅga or settlement endowed with rice land and slaves, "200 esclaves" in one grant under Yaśovarman, "mille esclaves hommes et femmes" in another under Sūryavarman I 7. Under Jayavarman VII the scale changed. The Ta Prohm stela of 1186, edited by Coedès in 1906, counts the temple's dependants verse by verse 13:

Ta Prohm (K.273), 1186 Number
Villages assigned to the temple 3,140
Chief officiants 18
Officiants 2,740
Assistants, "parmi lesquels 615 femmes (qui sont) danseuses" 2,232
"Au total ... personnes, y compris ceux qui ont droit au logement" 12,640
Men and women who "font là le service des Dieux" 66,620
"Au total ... avec les Birmans, les Cams, etc." 79,365

Five years later the stela of Preah Khan gave the totals for all the king's foundations: 306,372 men and women "y compris les Cāmpa et les Yavana avec les Pukāṁ et les Rvañ", and 13,500 villages 14. Preah Khan alone had 97,840 people and 5,324 villages assigned to it, and its own staff included "1.000 danseuses" 14. The temples were also the state's welfare system. The Ta Prohm stela records that "one hundred and two hospitals (ārogyaśālāḥ) are established in every district", supplied with rice yearly "for the sake of the gods, residents and patients", and supported by 838 villages and 81,640 men and women 13. Preah Khan lists 121 rest houses "avec du feu" along the royal roads, 57 of them on the road to Champa 14.

This was what the Indian religions changed most in Khmer life. Rice, labour, dance, medicine and travel were organised through foundations dedicated to Indian gods, staffed by people whose status was defined by their service to those gods. Coedès's verdict on the system's political face was that "the sovereign, for whom reigning consists of 'devouring the kingdom' (as a governor 'devours' his province), appears less as an administrator than as a god on earth" 1.

The monk from Kamboja

The religion that would replace the Indian cults came by a different road. Theravāda Buddhism, preserved in Pali by the Mahāvihāra monastery of Anurādhapura since the third century BCE, had been reformed in Sri Lanka under Parākramabāhu I, and its purified ordination lineage began to travel to the mainland in the late twelfth century. The Kalyāṇī inscriptions, set up at Pegu in lower Burma by King Dhammacetī in 1476–79 and translated by Taw Sein Ko in 1892, record the mission that brought it to Pagan. The Burmese monk Chapaṭa returned from Lanka with four companions:

"Sivalithera, a native of Tamalitthi, Tamalindathera, the son of the Raja of Kamboja, Anandathera, a native of Kincipura, and Rahulathera, a native of Lankadipa" 16

The party included a monk from Kāñcīpuram in south India and a prince from Cambodia. The inscriptions date the founding of the Sinhalese lineage at Pagan to Sakkaraj 543, 1181 CE 16. The Burmese Glass Palace Chronicle makes Tāmalinda "one of Jayavarman's sons", which would place a son of Jayavarman VII in the founding generation of mainland Theravāda; the Kalyāṇī text says only "the son of the Raja of Kamboja" 416. Harris notes that Tāmalinda was ordained into "the lineage of the Theravada Mahāvihāra, which had been recently purified by Parakkamabāhu I", and that he "seems to have died in Burma, never returning to his homeland" 4. For Harris the episode shows that "Angkor now felt the need to reach out, both diplomatically and spiritually, to other influential regions of the Theravada world" 4. The Theravāda that reached Cambodia was Indian in origin, Sinhalese in lineage, and carried by Mon, Burmese and Tai networks.

Yellow robes in 1296

A century later, a Chinese observer saw the new religion established in the streets of Angkor. Zhou Daguan arrived with an embassy of the Yuan court in 1296 and stayed about a year; his Zhenla fengtu ji (真臘風土記, "Customs of Cambodia") is the only eyewitness description of Angkor while it was a living city 18. He sorted Khmer religion into three teachings:

為儒者呼為班詰,為僧者呼為苧姑,為道者呼為八思惟。……苧姑削髪穿黄,偏袒右肩,其下則繫黄布裙,跣足。18

In a literal rendering: literati are called banjie, monks zhugu, and Daoist-like adepts basiwei; the zhugu shave their heads, wear yellow, bare the right shoulder, tie a yellow cloth skirt below and go barefoot 18. The banjie were the brahmins, recognisable by the white cord they wore at the neck for life; the basiwei were Śaiva ascetics who worshipped "a single stone", the liṅga 18. The yellow-robed zhugu, eating one meal a day, reading scriptures written on palm leaf, were Theravāda monks. Zhou also recorded two facts that show how far they had already penetrated Khmer life: boys who went to school were taught first by the monks, and when the king had important affairs of state he consulted them 18.

Pali replaces Sanskrit

The epigraphy confirms the shift and dates it. In 1933 Georges Trouvé found at Wat Kok Khpos near Siem Reap a stela inscribed in Pali on one face and Khmer on the other. Coedès published it in 1936 as "de beaucoup le plus ancien texte pâli" of Cambodia, dated 1230 of the Śaka era, 1308 CE 19. Its donor, a king of the Śrīndra line (Harris names Śrīndrajayavarman, Ludivine Provost-Roche Śrīndravarman), gave a village to a senior monk, together with "an additional four villages and slaves of both sexes for the funding of a monastery" 417. Sharrock is categorical: "it is certain that first use of Pali in a Khmer inscription occurs in 1308 CE" 15. Sanskrit ended within two decades. Harris dates the last Sanskrit inscription at Angkor to 1327 4.

The changeover was not peaceful at the top. Provost-Roche has shown that the Prasat Kombot in Kompong Thom was rebuilt as a Theravāda prang under Śrīndrajayavarman (1307–1327), two centuries earlier than usually assumed, and that his successor Jayavarmaparameśvara answered with a final Śaiva revival: his inscription calls for a "purification de la Terre", and in his first year, 1327, he set up a Śiva-liṅga inside the Bayon itself 17. She reads it as "une opposition et/ou une réponse au theravāda" 17. It was the last such gesture.

Angkor Wat becomes a Buddhist shrine

The capital moved south in the fifteenth century. The conventional marker is an Ayutthayan attack on Angkor in 1431, which Provost-Roche calls the "date conventionnelle de ce que l'on nomme « la chute d'Angkor »" 17. The evidence now points to a slower end. Dan Penny and his colleagues found that land use in the core of the city "began to decline more than one century before the Ayutthayan invasion that conventionally marks the end of the Angkor Period" 22; Brendan Buckley's tree-ring record shows that Angkor "experienced decades-long drought interspersed with intense monsoons in the fourteenth and fifteenth centuries that, in combination with other factors, contributed to its eventual demise" 21. Michael Vickery linked the shift of the court towards the Phnom Penh region to the growth of maritime trade with China and the decline of the old agrarian elite, and stressed how thin the fifteenth-century record is: almost no inscriptions, little architecture, and chronicles largely composed later 23.

When the court came back to Angkor in the sixteenth century, it came as a Theravāda court. King Ang Chan, who built at Udong in the 1530s, added two late bas-reliefs to the galleries of Angkor Wat, which Coedès dated to 1546 and 1564 24. They show Kṛṣṇa and Viṣṇu; the Indian epics remained the court's stories even after its religion had changed 4. Inscriptions of 1577 and 1579 record that King Satha piously restored Angkor Wat, rebuilt its central tower and covered it with gold 4. The Bakheng received a giant seated Buddha, and the western face of the Baphuon, the eleventh-century Śaiva temple-mountain, was rebuilt from its own stones as an enormous reclining Buddha 4. Japanese pilgrims took Angkor Wat for the Jetavana, the Buddha's own monastery in India; one of them, Morimoto Ukondayu Kazufusa, came in January 1632 "to make offerings to four buddha images" and left his inscription in ink in the temple's galleries, one of fourteen Japanese inscriptions that survive there 4.

The Cambodian settlement

The religion that resulted is neither the Indian religion of 900 nor the Theravāda of Sri Lanka. It has three parts, each with an Angkorian history:

  • The sangha: Theravāda monks, Pali scripture and the village wat, which by the time of Zhou Daguan already schooled boys 184.
  • The court brahmins, the baku: the early nineteenth-century Rapāl ksatr chronicle ranks them with the heads of the sangha immediately after the royal family 4.
  • The neak ta: the spirits of place, some of whom now live in Angkorian images; the neak ta Toeuk Lic on Phnom Kulen inhabits an Angkorian Buddha statue 4.

The sangha itself was reorganised from Bangkok. In 1854 King Ang Duang asked the Siamese court for a pure Pali canon, and the Khmer monk Mahā Pan brought the reformed Thommayut order to Udong with eight monks and some eighty bundles of texts; the Cambodian Thommayut was founded "probably around 1855", beside the older Mohanikay 4. Anne Hansen has traced how the colonial-era "modern Dhamma" movement of the 1920s grew out of those reforms 25.

What the Indian cults left behind can be listed as precisely as what they built:

  • Sanskrit as the language of record, replaced by Pali for religion and Khmer for everything else after 1327 415.
  • The royal liṅga cult, whose last recorded act was Jayavarmaparameśvara's liṅga in the Bayon in 1327 117.
  • The hereditary brahmin priesthood's monopoly of royal ritual, reduced to the baku of the court 4.
  • Mahāyāna Buddhism, which survived Jayavarman VII's death briefly at the Bayon and then disappeared from Cambodia 17.
  • The temples as owners of land and people, a role the Buddhist monasteries inherited in reduced form 4.

What the cost was

The transmission of Indian religion to the Khmer was not a conquest. No Indian army ever stood at Angkor, and the Khmer kings chose, paid for and reshaped every cult they received. Its costs were paid inside Khmer society, by the people the new gods were given, and in the violence that the competition between cults produced.

Bound to the gods

The largest cost was labour. Bonded temple dependants, the khñuṃ, are older than Angkor; the oldest Khmer-language inscriptions already list them 1. What the Indian cults changed was their number. The Sdok Kak Thom stela shows slaves given with each foundation; the Ta Prohm stela counts 79,365 people in the service of one temple; the Preah Khan stela counts 306,372 for all of Jayavarman VII's foundations 71314. Scholars still dispute how to translate khñuṃ: Coedès rendered it esclaves and Sak-Humphry "slaves", while others prefer "servants" or "dependants", because the category ran from owned captives to free villagers whose rice dues were assigned to a god 78. The inscriptions themselves are explicit that some were foreigners. The Ta Prohm total is "avec les Birmans, les Cams, etc."; the Preah Khan total includes "les Cāmpa et les Yavana avec les Pukāṁ et les Rvañ", Chams, Vietnamese, people of Pagan and highlanders 1314. These were, in large part, the peoples Angkor fought.

No inscription records how many people built Angkor Wat, and figures such as "300,000 workers" that circulate online have no traceable source. What is documented is the quarrying. Etsuo Uchida and Ichita Shimoda located more than fifty sandstone quarries at the south-eastern foot of Phnom Kulen, linked to the temples by a canal route of some 34 kilometres; every block of Angkor's sandstone temples was cut there and moved by human and animal labour 27.

Slaves bought from the hills

Zhou Daguan, the only outsider who saw the system working, described the trade that fed it:

人家奴婢,皆買野人以充其役,多者百餘,少者亦有一二十枚,除至貧之家則無之。18

That is: households buy "wild people", hill peoples, as slaves; the richest have more than a hundred, even those with few have ten or twenty, and only the very poorest have none 18. A young, strong slave cost a hundred lengths of cloth, an old or weak one thirty or forty. Slaves could not enter the upper part of the house except kneeling, and runaways were tattooed blue on the face or fettered with iron at the neck, arms or legs 18.

The Theravāda turn did not end this. The first Pali inscription in Cambodia, the founding charter of the new religion in 1308, endowed its monastery with "slaves of both sexes" 4. Harris records a monastery of the late sixteenth or seventeenth century with "the reputation of having more slaves than any other monastery in the kingdom" 4. The new religion inherited the labour regime the old cults had built.

Chiselled Buddhas

The competition between the cults left a second bill, this one in stone. After Jayavarman VII's death, a Śaiva reaction systematically destroyed the Buddhist images of his temples. Sharrock describes the "reworking of Buddha icons into Sivalingas in the Bayon, and the chiselling off of thousands of Buddha images in Jayavarman VII's other temples", a desecration "quite unprecedented in Khmer history" 15. Provost-Roche describes the method: meditating Buddhas recut into liṅgas, Avalokiteśvara turned into Śiva by removing the upper arms and the Buddha in the headdress, images of Prajñāpāramitā defaced 17. "Cette réaction šivaïte," she writes, "s'apparente à une entreprise iconoclaste ayant transformé les monuments mahâyâniques attribués à Jayavarman VII en monuments šivaïtes" 17.

A moss-stained sandstone wall with a row of seven pointed arched niches under carved friezes; the niches are empty, their flat chiselled surfaces showing where seated figures were cut away.
Emptied Buddha niches on the library of Preah Khan, the temple Jayavarman VII dedicated in 1191. The seated Buddhas that filled the niches were chiselled away in the Śaiva reaction after his death, while the carved arches and friezes around them were left intact. The reaction is conventionally attributed to Jayavarman VIII (r. 1270–1295); some scholars place it later or question its religious motive.
Greg Willis. Destroyed Buddha niches, library of Preah Khan, Angkor, 1191 (photographed 2004). In situ, Angkor, Cambodia. CC BY-SA 2.0 via Wikimedia Commons. · CC BY-SA 2.0

Who ordered it is disputed. The conventional attribution is to Jayavarman VIII, whose reign a 2007 re-reading of the inscription K.567 now dates to 1270–1295 rather than from 1243 1517. Claude Jacques holds to that attribution 17. Sharrock points out that Jayavarman VIII left no inscriptions and that Zhou Daguan in 1296 still saw five great stone heads of the Buddha over each city gate, and proposes Jayavarmaparameśvara, who placed his liṅga in the Bayon in 1327 1518. Martin Polkinghorne, Christophe Pottier and Christian Fischer have questioned whether the destruction was religious at all, suggesting it may have targeted images of the king and his officials, and found sixteenth-century tool marks on the Bayon's Buddha pedestal 20. Harris thinks the iconoclasm "rather more selective than has sometimes been thought": Banteay Chmar escaped entirely, and the target may have been the syncretic tantric imagery rather than Buddhism as such 4.

The material record is not in dispute. In 1933 Trouvé recovered the colossal Buddha of the Bayon's central sanctuary, in pieces, from the shaft beneath it; the reassembled statue now stands near the Bayon 20. In March 2001 archaeologists from Sophia University found 274 Buddhist sculptures, "mostly broken Buddhas on naga", carefully buried in sand at Banteay Kdei 15.

War on the frontiers

The cults also travelled with armies, and so did their plunder. Around 950 Rājendravarman II's expedition into Champa "removed the gold statue from the temple of Po Nagar at Nhatrang", the Cham goddess of the country 1. The foreigners counted among the gods' dependants in 1186 and 1191 were in large part the product of Jayavarman VII's wars with Champa and his other neighbours 1314. By Zhou Daguan's time the war had turned: villages near the Siamese frontier "因屢與暹人交兵,遂至皆成曠地", had become waste through repeated fighting with the Siamese 18. The Ayutthayan attacks of the fifteenth century, and the Siamese wars against Longvek under King Satha at the end of the sixteenth, belong to the same long conflict 4. No source counts the dead or the captives of 1431.

A coda: the sangha in 1975–1979

The settlement this record traces was nearly destroyed in the twentieth century, by a regime that set out to end religion altogether. This is not a cost of the transmission, and is set down here only because it shows how much the settlement had come to carry. Cambodia's monastic population had risen from 37,531 in 1955 to 61,014 in 1967 4. Under Democratic Kampuchea monks were forcibly disrobed, sent to labour and in many cases killed; official sources later gave a figure of only "12 monks to have remained in robes until the end of the Democratic Kampuchea period" 4. "A year later," Harris writes, "Cambodian officials computed a total of 2,311 monks in 1,821 monasteries ... an overall decrease of around 60,000 monks since the Khmer Rouge seized power" 4. He treats the larger claims of monks murdered "with considerable caution", and concludes in his later study that "although Buddhism was in mortal danger during Democratic Kampuchea, it was never totally extinguished" 26.

Weighing the bill

Cost When Evidence
Temple bondage at state scale: 79,365 people for Ta Prohm, 306,372 for all Jayavarman VII's foundations 1186–1191 Ta Prohm and Preah Khan stelae 1314
War captives (Chams, Vietnamese, Burmese, highlanders) counted among the gods' dependants 12th century Same stelae 1314
Household slavery fed by purchase of hill peoples; branding and fettering of runaways 1296–1297 Zhou Daguan 18
First Theravāda monastery endowed with slaves 1308 K.754 419
Thousands of Buddha images chiselled out or recut as liṅgas late 13th–14th century Sharrock; Provost-Roche 1517
Plunder of Po Nagar's gold image c. 950 Coedès 1

The bill is real and it is not small. Hundreds of thousands of people were assigned to the service of gods whose names they may never have spoken, and some of them were captives of war. The cults' quarrel destroyed thousands of sacred images. But it is the bill of a society that chose its gods: the brahmins came because Khmer kings invited and paid them, the tantras were read because Khmer priests copied them, and the monks' religion spread because Khmer villagers fed the monks. Cambodia's religion today, Buddhist monks, court brahmins and village spirits under the five towers of Angkor Wat on the national flag, is the settlement that came out of six centuries of that choosing.

What followed

Where this lives today

Theravāda Buddhism as the religion of most Cambodians Angkor Wat on the Cambodian national flag The baku, court brahmins of the Cambodian royal household The neak ta spirit shrines of Cambodian villages The Mohanikay and Thommayut monastic orders Angkor, a UNESCO World Heritage Site since 1992

References

  1. Cœdès, George. The Indianized States of Southeast Asia. Edited by Walter F. Vella, translated by Susan Brown Cowing. Honolulu: East-West Center Press, 1968. en
  2. Sanderson, Alexis. "The Śaiva Religion among the Khmers (Part I)." Bulletin de l'École française d'Extrême-Orient 90–91 (2003–2004): 349–462. en
  3. Mus, Paul. "Cultes indiens et indigènes au Champa." Bulletin de l'École française d'Extrême-Orient 33 (1933): 367–410. fr
  4. Harris, Ian. Cambodian Buddhism: History and Practice. Honolulu: University of Hawai'i Press, 2005. en
  5. Ang Chouléan. Les êtres surnaturels dans la religion populaire khmère. Préface de Georges Condominas. Paris: Cedoreck, 1986. fr
  6. Michon, Daniel. "Pre-Angkor Khmer Religion and State." Religion Compass 5, no. 1 (2011): 28–36. en
  7. Cœdès, George, and Pierre Dupont. "Les stèles de Sdŏk Kăk Thoṃ, Phnoṃ Sandak et Práḥ Vihār." Bulletin de l'École française d'Extrême-Orient 43 (1943): 56–154 (K.235, 1052 CE). Digital edition by Dominic Goodall and Kunthea Chhom, DHARMA project. fr primary
  8. Sak-Humphry, Chhany. The Sdok Kak Thom Inscription (K. 235), with a Grammatical Analysis of the Old Khmer Text. Phnom Penh: Buddhist Institute, 2005. en primary
  9. Bhattacharya, Kamaleswar. A Selection of Sanskrit Inscriptions from Cambodia. Phnom Penh: Center for Khmer Studies, 2009 (K.235 at 139–187). en primary
  10. Kulke, Hermann. The Devarāja Cult. Translated by I. W. Mabbett. Data Paper 108. Ithaca: Southeast Asia Program, Cornell University, 1978. Originally "Der Devarāja-Kult," Saeculum 25 (1974): 24–55. en
  11. Evans, Damian H., Roland J. Fletcher, Christophe Pottier, Jean-Baptiste Chevance, Dominique Soutif, et al. "Uncovering Archaeological Landscapes at Angkor Using Lidar." Proceedings of the National Academy of Sciences 110, no. 31 (2013): 12595–12600. en
  12. Cœdès, George. Inscriptions du Cambodge, vol. I. Hanoi: École française d'Extrême-Orient, 1937 (K.809, Prasat Kandol Dom, 37–46). fr primary
  13. Cœdès, George. "La stèle de Ta-Prohm." Bulletin de l'École française d'Extrême-Orient 6 (1906): 44–86 (K.273, 1186 CE). Digital edition by Kunthea Chhom, Dominic Goodall and Nina Mirnig Sarma, DHARMA project. fr primary
  14. Cœdès, George. "La stèle du Práḥ Khằn d'Ankor." Bulletin de l'École française d'Extrême-Orient 41 (1941): 255–302 (K.908, 1191 CE). fr primary
  15. Sharrock, Peter D. "Garuḍa, Vajrapāṇi and Religious Change in Jayavarman VII's Angkor." Journal of Southeast Asian Studies 40, no. 1 (2009): 111–151. en
  16. Taw Sein Ko, ed. and trans. The Kalyani Inscriptions Erected by King Dhammaceti at Pegu in 1476 A.D.: Text and Translation. Rangoon: Superintendent, Government Printing, Burma, 1892. en primary
  17. Provost-Roche, Ludivine. "Nouvelles perspectives pour la connaissance de la fin de la période angkorienne." Arts Asiatiques 65 (2010): 43–58. fr
  18. Zhou Daguan 周達觀. Zhenla fengtu ji 真臘風土記 [c. 1297]. Chinese text at Wikisource; English translation: A Record of Cambodia: The Land and Its People, translated by Peter Harris. Chiang Mai: Silkworm Books, 2007. zh primary
  19. Cœdès, George. "Études cambodgiennes XXXII: La plus ancienne inscription en pāli du Cambodge." Bulletin de l'École française d'Extrême-Orient 36 (1936): 14–21 (K.754, 1308 CE). fr primary
  20. Polkinghorne, Martin, Christophe Pottier, and Christian Fischer. "One Buddha Can Hide Another." Journal Asiatique 301, no. 2 (2013): 575–624. en
  21. Buckley, Brendan M., Kevin J. Anchukaitis, Daniel Penny, Roland Fletcher, et al. "Climate as a Contributing Factor in the Demise of Angkor, Cambodia." Proceedings of the National Academy of Sciences 107, no. 15 (2010): 6748–6752. en
  22. Penny, Dan, Tegan Hall, Damian Evans, and Martin Polkinghorne. "Geoarchaeological Evidence from Angkor, Cambodia, Reveals a Gradual Decline rather than a Catastrophic 15th-Century Collapse." Proceedings of the National Academy of Sciences 116, no. 11 (2019): 4871–4876. en
  23. Vickery, Michael. "Cambodia and Its Neighbors in the 15th Century." Asia Research Institute Working Paper 27. Singapore: National University of Singapore, 2004. See also Vickery, "Cambodia after Angkor: The Chronicular Evidence for the Fourteenth to Sixteenth Centuries," PhD diss., Yale University, 1977. en
  24. Cœdès, George. "La date d'exécution des deux bas-reliefs tardifs d'Angkor Vat." Journal Asiatique 250 (1962): 235–243. fr
  25. Hansen, Anne Ruth. How to Behave: Buddhism and Modernity in Colonial Cambodia, 1860–1930. Honolulu: University of Hawai'i Press, 2007. en
  26. Harris, Ian. Buddhism in a Dark Age: Cambodian Monks under Pol Pot. Honolulu: University of Hawai'i Press, 2013. en
  27. Uchida, Etsuo, and Ichita Shimoda. "Quarries and Transportation Routes of Angkor Monument Sandstone Blocks." Journal of Archaeological Science 40, no. 2 (2013): 1158–1164. en

Further reading

Cite this article
OsakaWire Atlas. 2026. "Indian gods at Angkor (c. 900) — and the Khmer turn to Theravāda" [Hidden Threads record]. https://osakawire.com/en/atlas/khmer_hindu_buddhist_900/