Khmer Empire

← Encyclopedia

The Khmer Empire was the dominant power of mainland Southeast Asia from the ninth to the fifteenth century, ruled from a succession of capitals in the Angkor region of present-day Cambodia. Its conventional span runs from 802, the date a much later inscription assigns to the consecration of Jayavarman II, to 1431, the year tradition fixes for the abandonment of Angkor after pressure from the Thai kingdom of Ayutthaya. Both dates are tidier than the history behind them, but they bracket the period in which Khmer kings governed a territory reaching across modern Cambodia into Thailand, Laos, and southern Vietnam, and raised the monuments for which the civilization is remembered.

Behind the round number 802 lies a document written two and a half centuries afterward. The Sdok Kok Thom inscription, carved in 1052 by a priestly family tracing its service back to the founder, supplies the standard account: that Jayavarman II returned from Java, reigned from several seats in turn, and at last ascended the hill of Mahendraparvata — Phnom Kulen, north of the later city — where a brahmin named Hiranyadama and the chief priest Sivakaivalya performed a rite consecrating him as cakravartin, a universal monarch who would owe no allegiance to any foreign overlord. The events the stone narrates appear almost nowhere else. The kingdom it describes as already unified was in fact assembled piecemeal over generations, out of the smaller polities of the pre-Angkorian period that scholarship calls Chenla, themselves heirs to the older trading state of Funan in the Mekong delta. The number 802 is the empire’s chosen origin, not a frontier visible in the contemporary record.

The god-king and the mountain

At the center of Khmer kingship stood the cult later scholars labeled the devarāja — a Sanskrit compound read most often as the god who is king, or the king of the gods, and sometimes as the god Indra, king of the deities, presiding over the realm. Exactly what the rite enacted is disputed; the inscriptions are sparse and have been read in opposed ways. What is clear is that the Khmer court drew its legitimacy from Indian religion, identifying the ruler’s power with a great deity, most often Shiva, and binding it to a sacred royal linga installed on a temple-mountain. The state temple was understood as a working image of the cosmos: a stepped pyramid standing for Mount Meru, the axis of the world in Hindu and Buddhist cosmology, ringed by enclosures and a moat that figured the encircling mountain-walls and oceans. To raise such a temple was to anchor the realm to the order of the universe — and to fix, at the summit, the point where royal power and divine power were made to coincide.

What the devarāja actually was remains genuinely contested, and the disagreement runs deep rather than at the margins. One reading takes the rite as an apotheosis: the living king becomes the god, or the god’s earthly body, so that to serve the throne is to serve Shiva. A second treats it as the consecration of a cult object — a linga or image, perhaps movable, that housed the protective essence of kingship and could be carried from capital to capital, a palladium of the realm rather than a deification of any one man. A third reads the compound the other way around: not the king made god, but the god installed as the true sovereign, the human ruler holding the throne as the deity’s viceroy. The term itself is concentrated in a single late inscription, written by the very lineage whose hereditary office was to maintain the cult, which has led some to suspect its centrality is partly an artifact of who held the chisel. The encyclopedia leaves the question where the evidence leaves it. What is not in doubt is the political work the architecture performed: a sandstone mountain at the heart of the capital, the king bound to the god at its summit, and a court whose right to rule was written into the structure of the cosmos.

The court’s changing god

The religion of the court shifted across the centuries, and the great temples are a stratigraphy of that change. Early Angkor was predominantly Shaivite, with Vaishnava and indigenous strands alongside it; the kings of the ninth and tenth centuries raised brick and laterite pyramids — at Hariharalaya, at the new capital of Yasodharapura, at Koh Ker during a brief northern interregnum — each crowned with a linga and dedicated, in the main, to Shiva. The eleventh century brought consolidation and the first stone temple-mountains of the classic type. Then, in the early twelfth century, Suryavarman II dedicated Angkor Wat to Vishnu and oriented it, unusually for a Khmer foundation, to the west — a turn read both as fitting to Vishnu and as marking the building for the king’s own funerary cult. It is the largest religious monument ever built, and the fullest surviving statement of the temple-as-cosmos that the empire had been constructing in stone for three hundred years.

A generation later the program broke. In 1177 the Cham, from the coastal kingdoms of central Vietnam, sailed up the Mekong and the Tonle Sap and sacked Angkor itself. The man who drove them out and refounded the capital, Jayavarman VII, reigned from 1181 into the second decade of the next century, and under him the state turned toward Mahayana Buddhism. He walled a new city, Angkor Thom, and set at its center the Bayon, whose towers carry great serene faces — most often identified with the bodhisattva Avalokiteshvara, the embodiment of compassion, though some read them as the king himself in bodhisattva form, or as other powers. The compassion was not only carved. The inscriptions of his reign record hospitals founded across the realm, rest-houses along the roads, and a rhetoric of the king’s suffering on behalf of his subjects that is recognizably Buddhist rather than the older idiom of divine sovereignty. The same court that had bound the throne to Shiva now bound it to the bodhisattva’s vow.

After Jayavarman VII the monumental building economy wound down, and a quieter, more lasting change moved through the population from below. Theravada Buddhism — the Pali-canon form transmitted through Sri Lanka and the Mon country, with its village monasteries and its renunciate sangha — spread among ordinary Khmer in the thirteenth and fourteenth centuries. It asked no temple-mountains and consecrated no god-kings; its merit economy ran through the offering bowl and the wooden monastery, not the sandstone pyramid. The cult of the devarāja and the great state foundations belonged to a theology the new faith did not need. Theravada outlasted the empire that the older cults had served, and it remains the religion of Cambodia today.

A Chinese envoy named Zhou Daguan, who lived at Angkor for nearly a year around 1296 and left the one substantial eyewitness account of the city in its late maturity, describes exactly this layered religious world: brahmins, a class he calls the Pan-ki, and saffron-robed Buddhist monks all present at once, the royal palace and its golden tower, the markets run by women, the customs of a court still magnificent but no longer in its building age. His report, written from outside the frame and preserved in Chinese, is the closest thing the period leaves to a visitor’s photograph.

City of water

How far these ideas were held by ordinary Khmer, as against the court and priesthood that commissioned the inscriptions, is among the things the record does not say. The temples and their dedications express the theology of those who paid for them. But the scale of the settlement they sat within has become, in the last two decades, the empire’s most astonishing single fact. Airborne laser survey flown over the Angkor plain in 2012 and again in 2015 stripped away the forest and the rice paddy and revealed a low-density urban landscape of house-mounds, ponds, field systems, and a gridded road network sprawling far beyond the walled temple-cities — together one of the largest, perhaps the largest, settlement complex of the entire pre-industrial world, covering on the order of a thousand square kilometers. Estimates built from that survey put the population of Greater Angkor at its twelfth- and thirteenth-century height somewhere near three-quarters of a million people.

A settlement of that size, on a seasonally flooded plain that swings between monsoon deluge and dry-season drought, lived or died by its control of water. The Khmer cut enormous rectangular reservoirs — the barays, the largest of them several kilometers long — and laced the plain with canals, channels, and embankments. Bernard-Philippe Groslier, excavating in the 1950s and writing in the 1970s, read this apparatus as the skeleton of a hydraulic city: a centralized state whose power rested on the management of irrigation water, the barays feeding the rice that fed the labor that raised the temples. The thesis was elegant and for a time dominant. It is now heavily qualified. Later scholars argued that the great barays were as much cosmological and ceremonial as agricultural — sheets of water mirroring the temple-mountains they surrounded — and that the water system grew by accretion and local initiative rather than from a single planning hand. The current picture, drawn largely from the Greater Angkor Project’s mapping, keeps the network’s vast extent while abandoning the tidy machine: a sprawling, partly improvised infrastructure that worked for centuries and then, when the climate turned, failed catastrophically.

The unmaking

The end, when it came, had no single cause. Tree-ring records from the region register two prolonged droughts in the fourteenth and early fifteenth centuries, the second including the driest year in seven hundred years of reconstruction, broken by years of violent monsoon — precisely the regime that a vast, aging water network is least able to absorb, scouring channels in the wet years and starving fields in the dry. The shift to Theravada drained the ideological energy from the temple-building that had organized so much of the realm’s labor. The center of economic gravity was moving toward the maritime trade of the lower Mekong, away from the inland rice plain. And to the west the Thai kingdom of Ayutthaya was rising, pressing on Angkor through the fourteenth century and, by tradition, taking the city in 1431.

After 1431 the kings moved their seat toward the south — to the region of Phnom Penh and, later, Longvek — closer to the river trade and farther from the Thai frontier. The great temples passed gradually into the keeping of Buddhist monks, never wholly abandoned: Angkor Wat in particular kept its worshippers through every century, so that the European travellers of the nineteenth century who reported it as a marvel half-lost to the forest had in fact arrived at a living shrine. The stone cosmos remained standing after the empire that built it was gone.

Reading the empire

Almost everything known of Angkor before Zhou Daguan comes from stone. The Khmer kept their records — genealogies, donations, foundations, the dedications of temples and the rosters of their slaves and lands — incised in Sanskrit and Old Khmer on the doorjambs and stelae of the temples themselves, and the modern history of the empire is, before it is anything else, a history of reading those inscriptions. The foundational synthesis is George Coedès, whose decades of epigraphic work underlie The Indianized States of Southeast Asia (English edition 1968), the book that fixed the standard chronology of the kings and gave the devarāja reading its long currency. Lawrence Palmer Briggs’s The Ancient Khmer Empire (1951) remained the principal English narrative for decades. Charles Higham’s The Civilization of Angkor (2001) set the dynastic story within the deep archaeological record of the lower Mekong, from prehistoric villages up through the rise and fall of the state.

The later twentieth century brought a sustained revision. Hermann Kulke’s monograph The Devaraja Cult (1978) recast the central rite as the cult of a movable royal palladium rather than the deification of the king, unsettling the reading Coedès had made canonical. Michael Vickery’s Society, Economics, and Politics in Pre-Angkor Cambodia (1998) went back to the seventh- and eighth-century inscriptions and argued that the Indianization model had read too much Indian theory into what were, on the ground, indigenous Khmer structures dressed in Sanskrit. The most dramatic recent additions come not from epigraphy but from the air: the airborne-laser results published by Damian Evans and colleagues, Uncovering archaeological landscapes at Angkor using lidar (PNAS, 2013), remapped the city as a low-density megalopolis, and the climate reconstruction of Brendan Buckley and colleagues, Climate as a contributing factor in the demise of Angkor (PNAS, 2010), gave the long-suspected drought of the empire’s final century a measured shape. The inscriptions remain the spine; the laser and the tree ring have given them a landscape and a weather.

→ Related: Angkor Wat · Shiva · Vishnu · Hinduism · Buddhism · Mahayana · Buddhism Theravada · Indra