Temples as Economic Centers in Early Cambodia

In chapter 3 it was proposed that Southeast Asia’s early history can be understood as being characterized by a series of economic and social transformations. In tribal societies, a reciprocal sharing of economic resources among family, community, and religious groups maintained the social unit. The development of entrepreneurial activities brought social imbalances, resulting in the transformation of the indigenous economy and the emergence of political entities based on redistributive exchange. Several early Southeast Asian societies went beyond these pri­mary levels of integration, developing organizational mechanisms for the acquisition, control, and disposal of resources in pursuit of collective goals, most of which were political. 1

The Khmer state that came to be based in Angkor—at its height from the ninth through the thirteenth centuries—demonstrates this transfor­mation thesis. Especially noteworthy is the development of Khmer temples as centers of redistribution and the “continuous and massive movement of products” from villages toward temples. 2 Khmer temples collected and ultimately returned to the countryside a portion of the local output, goods that represented a significant share of the society’s total economic production. These goods were redistributed according to the wealth, power, or prestige of the recipients. With the temple network as its base, Khmer society went beyond primary redistributive integration and reached a higher level of centralized economic control. The massive public works projects directed by Khmer monarchs—highlighted by the construction of Angkor Wat by Sūryavarman II in the early twelfth cen­tury and the even more impressive Angkor Thorn by Jayavarman VII at the end of the same century—required a degree of economic and social integration high enough to provide the economic resources necessary to fund and carry out these projects. They could not be supplied by a simple redistributive economy. To achieve these ends, goods and services from the Khmer agrarian system had to be channelled into the hands of those 137representatives of the state who were responsible for the achievement of broadly political goals. They created an integrated state political system.

Although separate mechanisms of administration were developed in each temple, Khmer temples were never autonomous from Khmer society and its stratified political order. At the primary level, Khmer temples were subject to the authority of a landed elite and were an instrument by which this elite reinforced its economic and political control. Temple staffs were often members of elite families, and the staff of a local tem­ple was frequently supervised by members of the local landholding elite who had entered the clergy or by priests who owed their positions and their prosperity to continued patronage by that elite. Normally, the con­solidation into an estate of an elite’s land rights in regional agrarian com­munities was legitimized by the construction of a temple, with income from specified estate lands designated to support the temple’s activities. Collections, gifts, and offerings flowed to the temple as part of a cycle of economic redistribution. They flowed outward in the form of support for the performances of rituals and construction and in so doing rein­forced the prowess of the elite as patrons of the temple deity and as the source of prosperity in the eyes of the local inhabitants.

By the height of the Angkor era, local temples and their cults had been integrated into a statewide network of temples that was ultimately tied to the king’s central temple at the royal capital. The priests from local tem­ples participated in rituals at major royal temples that were constructed at strategic points throughout the state. They as well as their aristocratic patrons derived legitimacy from this participation. In return local tem­ples helped to finance the activities of the central temples by assigning to them a portion of the annual collections. Under the guise of religion, Khmer monarchs, who dominated the central temples and their staffs, could draw part of the wealth of their realm to the royal capital without creating an elaborate secular bureaucracy to collect revenues in the state’s name.

This chapter summarizes research on early Cambodian history and through Khmer temple inscriptions examines the political and economic transformation of the Khmer state. From these inscriptions we know that temples consistently served as centers for the collection of economic resources and that the development of the temple network paralleled the development of the Khmer state. 3 Yet temples were not centers for the redistribution of economic capital alone; they were equally important as centers for the redistribution of “symbolic capital.” The temple network came to integrate the Khmer realm in two ways. On the one hand, tem­ples linked disparate agricultural regions horizontally into an ever-expanding economic network whose wealth fueled the Khmer state. On 138the other hand, temples were the locus for the manipulation of cultural symbols to vertically integrate the higher and lower levels of the Khmer socioeconomy. As institutions imbedded within the traditional socio­economy, temples assumed the leading role of “limiting and disguising the play of economic interests and calculations.” 4

Khmer Religion

In early Cambodia Hindu forms of religion were introduced in temples and temple ceremony to provide the basis for legitimizing political (and by implication economic) integration. As noted in chapter 1, local elites used the Śaivite religion to reinforce indigenous symbols of authority relationships. Since Śiva was referred to in Indian philosophy as the “Lord of the Mountain,” proprietors of early Khmer temples were able to connect the worship of Śiva to indigenous beliefs in the sanctity of mountains, which were believed to be the abode of ancestor spirits responsible for the prosperity of the living. Equally responsible for the success or failure of the living were various local spirits who bestowed fertility on the land. Such fertility spirits were represented in early Cam­bodia by a liṅga, a stone or metal phallus usually inserted upright in a cir­cular “vulva” (yoni). Since Indian tradition associated the liṅga with Śiva, the god of fertility, it was possible to link Śiva with these traditional spirits as well. Thus worship of Śiva “fit” with indigenous traditions in which worship of mountains and phallic symbols was necessary for the society’s success.

Worship of Śiva in Cambodia became formalized in the devarāja (“god-king”) cult of Jayavarman II (770–834), which was based on a mountaintop that became the site of his realm’s principal temple at the center of the royal capital. 5 In the Brahmanical concept of the universe, a circular central continent, Jambudvīpa, was surrounded by seven oceans and continents. In the center of this continent was Mount Meru, around which the sun, the moon, and the stars revolved. On its summit was the city of the gods, where Indra, the “Lord of Heaven,” reigned. 6 On the slopes of Mount Meru was the lowest of paradises (Lokapāla) where the “Guardians of the World” resided. Jayavarman II accordingly centered the cult of the devarāja at his capital city of Hariharālaya, south of Angkor on the edge of the Great Lake (Tonle Sap) in central Cambodia, on the summit of “Mount Mahendra,” the Khmer equivalent of Mount Meru.

While building a Khmer state through a combination of conquest and the formation of a network of personal alliances, Jayavarman II also 139consolidated the worship of regional deities in his royal devarāja cult. 7 He incorporated the veneration for mountains, subordinated local ances­tor spirits to the worship of Śiva, and then proclaimed himself Śiva’s rep­resentative on earth. By associating himself with Śiva and the royal mountain, Mount Mahendra, he symbolized his ability to guarantee the flow of life-power from the spirit realm to his subjects. The establish­ment of the devarāja cult tied Jayavarman spiritually to his supporters, and the cult became an emblem of the unification of the Khmer realm. Henceforth, although there were struggles for power within the royal domain, struggles for local independence were denied religious sanction. 8 The Khmer monarch from Jayavarman II’s time on could monopolize temporal power within the realm, power that was justified by a royal cult in which he alone could represent Śiva or any other Indian deity.

The royal capital and its realm were thus under the protection of the “Lord of the Heavens,” and the king, his representative on earth, was the “Lord of the Mountains,” the guardian of law and order, the protector of religion, the defender of his land against external foes, and the sum of all authority on earth. From the summit of Mount Mahendra (Mount Meru, the center of the universe), the devarāja entered into a relationship with the divine world. The royal temple was dedicated to the living monarch—the god-king—and it became his mausoleum when he died.

The blending of the indigenous cult of ancestor worship with the Indic religious forms gave the monarch magic properties that conferred im­mortality upon him. Statues and liṅgas of gods placed in the central and subordinate temples of the Khmer realm were portraits and symbols of kings, their names a fusion of monarchs’ personal titles with the names of the gods. Jayavarman’s successor, Indravarman I (877–889), con­structed a stone temple to shelter the royal Śiva-liṅga Indreśvara in which the devarāja resided, and he thus became associated with Indreśvara. 9 Although early Khmer kings such as Jayavarman II and Indravarman I were identified with Śiva, in the early twelfth century Sūryavarman II assumed an intimate relationship with Viṣṇu and built Angkor Wat to honor this union. He became the Viṣṇurāja rather than the devarāja. At the end of the same century, Jayavarman VII was the Bodhisattva Loke­śvara who decorated the faces of Angkor Thom’s Bayon temple complex (see page 144) and was honored in a new Buddharāja royal cult. Whether the devarāja, Viṣṇurāja, or Buddharāja, the Khmer king was the inter­mediary between man and the divine powers, the upholder of the estab­lished order (dharma) handed down by his ancestors, the intercessor with the spirit world for the fertility and prosperity of his realm. His capital city, with his royal temple at its center, was constructed in the image of the universe interpreted in terms of the Indic Mount Meru. 10

140The key religious elite of the Khmer realm was not composed of legal­istic Brahmans but devotional (bhakti) ascetics—“wayward Brahmans” who were not interested in Vedic ritual sacrifices and preferred to stress Śiva’s grace rather than the laws of orthodox Hinduism. 11 Gifts to Śiva or to any other “Hindu” or local deity linked to him brought one merit and thus hope of a superior death status. 12 A political overlord—a landed aristocrat—not a Brahman priest as in Vedic India, was viewed as having the foremost spiritual influence on his followers’ lives and their hopes for salvation. The secular landed elites performed customary ritual roles, including providing those of lesser spiritual prowess (śakti) oppor­tunities for enhancing their prospects after death. Political allegiance, expressed by personal loyalty, was based on indigenous attitudes about death and spiritual prowess. As acts of homage, and in theory to secure the favor of their overlord, subordinates offered presents to temples or erected statues or liṅgas in honor of the god-king. 13 Landed elites saw subordination to an overlord as providing further means of earning merit and satisfying their death wishes. 14 Gifts to ritual priests (Brahmans) were rare. Merit was earned by personal achievement rather than by honoring “Brahmans.” Temple priests were not honored because of the rituals they performed; they benefited instead through their superintend­ing of temples.

Jayavarman II’s Devarāja Cult

Jayavarman’s devarāja cult was a watershed in the history of Khmer reli­gion. Even though Khmer kings after Jayavarman II began to formulate their own royal cults, they continued to venerate his devar āja cult. Although his devar āja cult was like other personal liṅga cults that pro­moted a chief’s prowess, Jayavarman’s spectacular achievements guaran­teed his status as an ancestor of note among his Khmer “kin” who fol­lowed him. Early Khmer epigraphy glorified rulers in terms of their personal achievements and spiritual prowess, which rewarded their sub­jects with prosperity in this life and the next. Jayavarman’s reign estab­lished new criteria for accomplishment; his military victories were so impressive that simply honoring his spiritual capabilities was insufficient to praise his powers as king. Jayavarman’s religious rites not only consol­idated Khmer ritual in a statewide cult, they also used a new sacred vocabulary to proclaim the king’s glory and abilities. 15 After Jayavar­man’s reign, succession disputes ended with the victor reaffirming the ancestral claim to lead the Khmer people in his generation in return for the loyalty of his subjects. 16

141The most important aspect of Jayavarman’s devarāja cult was his incorporation of local deities—and in particular the indigenous ancestor worship—into the state’s religious ceremony. While earlier Khmer mon­archs had to some extent brought together the local and “Hindu” gods, they had always been worshiped separately, and indigenous deities were never placed in “Hindu” temples. In the indigenous belief system of Cambodia, death was held to represent the passage of the spirit into the realm of the ancestors. Death rites ushered the spirit into this world of the dead. In return for faithful worship of the dead, the ancestors granted a certain life-power to the living. Additionally, indigenous deities were traditionally important as “protectors” of the local realm; a chief was the repository of their prowess and could activate their powers on his subjects’ behalf. To proclaim their legitimacy, Jayavarman II and subse­quent Khmer monarchs merged the worship of Indic deities such as Śiva, Viṣṇu, and the Buddha with the indigenous deities and ancestor spirits, and even installed the local deities in the king’s “Hindu” temples to rep­resent his control over the traditional protective forces of his realm. True to this tradition, Angkor Thom’s Bayon focused on the central shrine, which was dedicated to the Buddhist Bodhisattva Lokeśvara, but the Bodhisattva (i.e., Jayavarman VII) was surrounded by images or Śiva and Viṣṇu, as well as those of important indigenous protector deities. Beginning with the reign of Jayavarman II, Khmer monarchs did not build temples or invoke the cults of past monarchs to honor their ances­tors but to install the traditional protectors of their ancestors and past kings in their central temple compound. 17 The devarāja cult and the Khmer royal cults that were initiated in its image exalted but did not deify the reigning monarch. At their deaths, however, Khmer monarchs were elevated to divine status so that their superhuman powers might be drawn upon by the living. Jayavarman II is the first example of the deifi­cation of a Khmer monarch at his death. 18 He was known posthumously as Parameśvara.

The significance of this synthesis is also shown in the oath of loyalty that was sworn to Sūryavarman I in 1011 by his administrative corps (tamrvāc). 19 In this instance these subordinates of the state requested that, in return for their faithful service, the king first maintain their fam­ily ancestral cults and then, secondly, that he perpetuate their families. Through proper observance of the rites calling forth the protective forces of the ancestors, their kin would be guaranteed the resources necessary for a prosperious life. 20 The oath ends: “May we obtain the recompense of people devoted to our masters in this and the other world.” The emphasis of this oath was not on Sūryavarman’s genealogical legitimacy, nor his divinity, but on the concerns of the present: that the monarch guarantee the prosperity of the Khmer people and secure for them a superior death status.142

Central tower, Angkor Wat. Photo by Charles Mark.143

Angkor Wat, Cambodia. Photo by Charles Mark.

Ornamental apsara, divine dancers who lavish the joys of paradise upon the elect. Angkor Wat; photo by Charles Mark.144

Angkor Thom, the Bayon. Photo by Charles Mark.

Relief of naval battle, Angkor Wat. Photo by Charles Mark.145

The Bayon, face of Lokeśvara. Photo by Charles Mark.

146This emphasis on the prosperity of the present is also the theme of the bas relief of Angkor Wat, which shows the departure from the “present” and deteriorating age—the kali yuga in Indian cosmology—and a return to the first or golden age (kṛta yuga) that was represented by Sūryavar­man II’s reign (1113–1150). 21 According to this depiction, the new age began when Sūryavarman II seized the throne from two “kings” (one of whom was an elderly uncle), put down a civil war, and went on to regen­erate a maṇḍala (i.e., he restored order in the world). True to the literary battle scenes of the Kurukṣetra, the episode of the Indian Mahābhārata epic on which the Angkor Wat relief is based, the king’s dead enemies and their supporters descended into the underworld to be judged by Sūr­yavarman, who assumed the role of Yama (Viṣṇu), the judge of the dead. 22 After the dust of the battle had settled, Viṣṇu churned the cre­ative ambrosia of the universe to sustain the golden age.

Indian symbols were used in this bas relief to make a Khmer statement; the use of the “Hindu” conventions of Viṣṇu and the “golden age” drew attention to Khmer tradition. Sūryavarman’s identification with Yama, the judge of the dead, was consistent with Khmer views of a sovereign’s powers, where obedience or disobedience to one’s overlord had conse­quences in the afterlife. While in Indian tradition the Kuruk ṣetra battle initiated the last age of the world (the kali yuga age of the present), at Angkor Wat the battle episode was used instead to introduce the first, or golden age (the kṛta yuga), which was still said to exist. The message of the Angkor Wat bas relief was that it was a privilege to live in Sūryavar­man’s generation when internal peace was restored and the Khmer mili­tary marched to the edge of the Southeast Asian mainland. 23 As in the oath of loyalty to Sūryavarman I, the emphasis of Angkor Wat’s bas relief was not on royal genealogy but on the present golden age that had been initiated by Sūryavarman II’s personal accomplishments, which had fulfilled the Khmer people’s expectations of one who claimed to be their monarch.

The depiction of Viṣṇu churning the creative ambrosia also supported the traditional conviction that the Khmer monarch was the source of his subjects’ economic welfare. In the Angkor era the king was the creator and director of public works that were designed to ensure prosperity—in particular the Khmer hydraulic system, a network of waterworks that irrigated some twelve and a half million acres, constructed around the Khmer capital at Angkor and in a number of regional domains under Khmer authority. Without this hydraulic system, the water supply was irregular and thus limited agricultural productivity. The nāga, the water spirit, was widely portrayed in Khmer art, and was a central figure of 147popular religion. 24 Chou Ta-kuan, Mongol envoy to Angkor in 1296, reported that the Khmer people believed that their ruler slept with a nāga princess, and that the result of their union was the country’s prosperity. 25 This report implies that the Khmer monarch enjoyed a ritual relationship with the spirit of the soil that released the fertility that guaranteed the earth’s productivity. In this same tradition Yaśovarman I constructed an artificial lake (baray), Yaśodharataṭāka, northeast of his new capital city of Yaśodharapura (Angkor) at the end of the ninth century. According to the inscription reporting this event, the king wished to “facilitate an out­let for his abundant glory in the direction of the underworld.” 26 This underworld, also depicted as the place from which Khmer monarchs judged the dead, was the abode of the nāgas, the source of fertility. Another inscription notes that Yaśovarman, “resplendent with glory,” made the lake “beautiful as the moon to refresh human beings.” 27 Yet another inscription equates Yaśovarman’s lake with the moon, which in Indian tradition provided life-sustaining ambrosia. 28

A 1980 examination of the Angkor era water management system demonstrates that Yaśovarman’s lake was not a critical source of water for the Angkor region’s agricultural production in a technical sense, though as the focus of Khmer religion it was important symbolically in the Khmer system of “theocratic hydraulics.” 29 Archaeologists have assumed that water seeped through the dike base of Yaśovarman’s lake (which measured four miles long by one mile wide) into collector chan­nels outside the dike, which subsequently carried the water to surround­ing fields. 30 But studies conducted in the late 1970s found that Angkor-era agriculture was based instead on bunded-field transplanted wet-rice cultivation that allowed the planting of approximately fifty million fields. 31 In the Angkor region flood waters would slowly rise from the lake to its tributaries, but would rapidly recede after the rainy season. A network of dams and bunds diverted and retained the receding floodwa­ters of the Great Lake, the Tonle Sap, after the rainy season. The Khmer lacked the technology to build large-scale dams that could have allowed an integrated regionwide hydraulic system; instead they depended on a network of small, simple earthworks on minor streams to retard and spread floodwaters into clay-based ponds, which stored the water for later use. 32 Archaeological evidence demonstrates that Angkor itself was not a major center of this water management network but rather the hill Phnom Kulen, which was located upriver from Yaśovarman’s lake some fifty kilometers northwest of Angkor. Phnom Kulen was near the headwaters of the Siem Reap (river), which flowed from that area through Angkor to the Tonle Sap. A network of small earth dams regu­lated the flow of water downstream from Phnom Kulen to Angkor.

One striking feature of the water management network at Phnom 148Kulen is that its dams, in addition to their effectiveness in managing water, all were constructed running east-west and north-south. Similarly, throughout the Angkor region Khmer temples were constructed at the intersection of east-west and north-south oriented moats and roads, which was done purposely to project the image of the heaven on earth (maṇḍala) that had been initiated by Khmer monarchs. 33 In addition to being consistent with Indian and Khmer cosmological focus on east-west and north-south, the water management network was consecrated by the traditional symbols of fertility. A number of liṅgas were carved in the rocky riverbed between Phnom Kulen and Angkor, suggesting the sanc­tity of the water that flowed from this mountain region to Angkor. 34 It is significant that Jayavarman II consecrated his devarāja cult at Phnom Kulen, which became the original “Mount Mahendra,” prior to the establishment of his new capital downriver at Hariharālaya and the con­secration of a new Mount Mahendra there. 35 The original Mount Mahen­dra at Phnom Kulen was thus not only a source of legitimacy for later monarchs who drew upon the protective powers of Jayavarman’s deva­ rāja cult but it was also seen quite correctly as the source of the Angkor region’s water supply, a fact that enhanced the possibilities for success for the monarch’s subjects.

In these examples and others, Khmer inscriptions and archaeological evidence well reflect the growing religious sophistication of Khmer soci­ety in the Angkor era. The sacred language and symbols of “Hindu” reli­gious philosophy proclaimed the king’s glory and abilities. The king was filled with life-sustaining energy derived from “Hindu” and indigenous deities, as divinity flowed from the heavens or from the earth and per­meated him, endowing him with the power to dispense “purifying ambrosia” or other less abstract forms of prosperity upon his subjects. 36 While local deities and spirits protected the monarch and his subjects, “Hindu” gods suffused Khmer monarchs with their superior creative and purifying energy, enhancing further the prospects for prosperity in this and other worlds. Having now an overview of the Khmer religious sys­tem, let us step back and examine the economic and political implications of this system’s evolution from pre-Angkorian to Angkorian times.

Temples and Khmer Statecraft in Pre-Angkorian Khmer Society

The principal concern of the leaders of early Khmer society, as reflected in their epigraphy, was the establishment and endowment of local tem­ples, for which they accrued religious merit and economic return. Key figures in the foundation of these early temples were the consentual 149leaders of local populations, rather than persons claiming royal author­ity. Inscriptions recording such activities emphasize the religious prowess rather than the physical might of the local elites who were establishing the temples. Regional leaders held official titles; the power of the landed elites was seemingly recognized by those claiming the authority to rule over the Khmer people by the bestowal of titles on these preexisting leaders, giving them “new” authority as district officers in the state administration. In such a way the landholders, the regional economic, social, and political leadership, were integrated into an emerging state system.

The actions taken by landed elites reported in nonroyal inscriptions involved the worship of local and “state” divinities to “acquire merit” and to “exhibit devotion.” 37 Inscriptions celebrated the presentation of gifts made to the temple by local leaders as part of their worship, and the wealth of those making the gifts was stressed and donations were care­fully calculated. The landed elites who were responsible for this epigra­phy emphasized the giving of gifts to temples as the foremost means to ensure the prosperity of society. The literacy of the temples’ patrons was also acclaimed in the inscriptions recording gifts, as if this literacy, in addition to the merit they acquired as temple benefactors, legitimized the donors’ status as the leaders of society. Gifts were normally provided not by single “officials” but by several “officials,” members of the landed elites, who made endowments as a group rather than as individuals, denoting the existence of kinship-type bonds among the Khmer aristoc­racy who may or may not have been blood relatives. The landed elites’ patronage of temples as a group, thus, may be seen as in some way for­malizing their political alliances. Maybe one man could not afford to make a gift by himself and took up a collection from those who would benefit by the gift.

The Khmer aristocracy concentrated economic resources under a tem­ple’s administration, whether to acquire the merit associated with such donations, to allow for a more efficient management of an elite’s resources, or to avoid the revenue demands of those political elites claim­ing rights to a share of the local landed elite’s possessions. The foremost method of accomplishing this goal was to donate land to temples. Boun­daries of donated lands were clearly defined in the inscriptions, usually associated with place names—perhaps a village, an estate, a pond, or the riceland of another landholder. Past and present holders of the assigned lands were enumerated, along with the mode of the property’s acquisi­tion and its price if acquired by purchase; 38 the land’s productivity (rice yield) was even estimated. Inscriptions reporting assignments of popu­lated land gave the parcel’s current occupants and spelled out what por­tion 150of the occupants’ production was assigned to the temple. If the land was unpopulated a labor force was assigned the task of working the newly donated lands. These laborers were counted—males, females, females with children—and their ethnic identity (e.g., Mon or Khmer) was recorded. 39

It is not uncommon to discover that the relatives of the donors of such land endowments were members of the priesthood servicing the local temple, and who became managers of the assigned property. 40 In many instances the donor family rather than the temple staff managed family land assigned to a temple, the temple receiving only a designated share of the income. What was transferred by donors was not “ownership” of land but the right to income from land. As explained in chapter 1, con­trol over manpower and production rather than ownership of land was critical for early Southeast Asian state development. To Khmer elites “landholding” meant rights to the production and labor service of the inhabitants of a parcel of land rather than absolute possession of it. In land donations to temples, only certain rights over the land were trans­ferred; while inhabitants of the assigned land normally continued to farm the land, the recipient temple collected much or all of their produc­tion. Donated property was usually subject to a combination of claims, those of the temple receiving the donation as well as those of the donor’s family, who retained certain personal rights to the property—for exam­ple, the right to a share of the land’s production as well as administrative or political rights over the inhabitants.

The economic diversity of local temples is reflected in the variety of donations “for the service of the property” assigned: domesticated ani­mals, goats, buffaloes, cattle, coconut palms, fruit trees, areca nuts, betel leaves, clothing, a threshing floor, plus numerous individual objects are examples. 41 The type and size of gifts to temples not only indicate economic specialization within early Khmer society in order to create such wealth but also the developing institutional capacity to utilize and administer this production. An economic system was emerging centered in the temple. The assignments of land and its production by Khmer aris­tocrats turned temples into local storage centers; goods deposited in tem­ple storehouses were a source of social and economic power, reinforcing the prestige of the temples’ primary benefactors, the local landed elites, who influenced the redistribution of the temples’ stores in support of their followers (including those working for the temple). Income from the temple was utilized by a local elite to maintain a variety of subordi­nates and was redistributed to expand the family’s social and political power.

The concentration in temples of the authority to manage local re­sources 151had a significant impact on the process of local political integra­tion. The implication of such centralization for the economic control sys­tem is reflected in epigraphic references to the “joining together for the enjoyment of the gods,” whereby a single or several landholders shifted a share of the income (goods and services) destined to one god or temple to that of another or amalgamated the administration of one temple’s lands with that of another. 42 This joining together or joint usufruct is termed saṃparibhoga or miśrabhoga. 43 Through these actions a pattern of sub­ordination of one local deity to another as well as one local temple to another began to emerge in Khmer society. Regional temple networks came to be sustained and controlled by a secular, landed elite who trans­ferred income from their lands or donated material goods and services to temples in the network.

During the Angkor era miśrabhoga arrangements were normally not undertaken without the approval of the king himself, but in pre-Angkor society private landholders and not a royal authority dominated the con­centration of economic resources and the amalgamation of temple administration. Prior to the ninth century Khmer monarchs were con­cerned only with recruiting as their allies the local landed elites who had initiated such transfers and consolidations, confirming the transactions rather than challenging them. This is illustrated in the records of the reign of Jayavarman I (657–681) who was based in northern Cambodia and northeastern Thailand (a region the Chinese knew as “Chen-la”). Jayavarman actively extended his authority over local communities and an important aspect of this process was his attempt to bestow legitimacy upon local temples. 44 His recognition of land income transfers is the topic of seven of his eight surviving royal edicts. 45 Five involve copartici­pation and the consolidation of land management under the authority of regional temples. 46 However, during the reign of Jayavarman’s predeces­sor Īśānavarman I (616–ca. 635), such shifts of income rights and land management consolidations had been undertaken with neither the mon­arch’s involvement or his favor; Īśānavarman was merely mentioned as having reigned at the time they took place. 47 Īśānavarman and other early Khmer monarchs seem not to have had sufficient authority to manipulate the temples controlled by dominant regional political elites, which were growing into centers of regional religious and economic power. On the other hand, Jayavarman I’s inscriptions show an attempt by the monarch to dominate lands and temples, even though he clearly did not have authority in land transactions. His royal edicts, although expressing his concern for land, never claim that the monarchy’s author­ity over land superseded that of the local landed elites.

An inscription of the Angkor-era ruler Jayavarman IV (928–941) also 152speaks of a Khmer monarch’s concern for land, forbidding the careless grazing of buffaloes that might destroy good riceland. 48 But in express­ing this concern Jayavarman IV still had to take an indirect approach and respect the landholding rights of the local elites. In this instance Jayavarman ordered a royal official to notify the region’s political leader (a khloñ viṣaya—a district chief) of his wishes; the local leader in turn decided to acknowledge the edict by permitting it to be published in his district. Clearly it was the regional chief’s option to acknowledge Jaya­varman IV’s concern for the local lands; it would appear that the chief equally had the option of ignoring this expression of royal interest. Thus while Angkor-era kings had a greater involvement in land transfers and assignments, early Khmer rulers could only express interest in the trans­fers of income and management rights of lands that were part of a regional elite’s domain. Even in later times the nonroyal private sector, dominated by landed regional elites, was still quite strong and func­tional.

Jayavarman II, whose reign (770–834) marked the transition from pre-Angkorian kingship, was particularly adept at forging alliances with powerful regional families and their networks of supporters. 49 The “commander of the royal army,” Mratāñ Śrī Pṛthivīnarendra, was a regional chief from the eastern Khmer realm who was allied to Jayavar­man by marriage. 50 This kinship linkage was critical to Jayavarman’s establishment of a territorial base along the Mekong and drew to Jaya­varman’s cause two brothers, members of the Mratāñ’s personal alliance network, whose ancestral home was Angkor Borei near the old Funan capital of Vyādhapura at the edge of the Mekong Delta. 51 At the insis­tence of the Mratāñ these brothers and their family were rewarded for their service; Jayavarman assigned the family significant land grants in territories in an area west of the Tonle Sap following its conquest by Mratāñ Śrī Pṛthivīnarendra, the two brothers, and their allies for Jaya­varman. 52 Not only did Jayavarman bestow income rights to this land, but seemingly to perpetuate his alliance with this family Jayavarman married one of the brothers’ sisters. 53 Mratāñ Śrī Pṛthivīnarendra made requests of Jayavarman for redistributions to other of his followers and their families, who were given land assignments, titles of honor, and posts at court. 54 It was Mratāñ Śrī Pṛthivīnarendra, a commander of troops and the leader of an extensive regionally based alliance network, rather than a temple priest who performed the critical Mount Mahendra ceremony consecrating Jayavarman’s rule and initiating the royal deva­ rāja cult. 55

Within the Khmer realm religious foundations (i.e., temples) of pow­erful local families, who also held official titles in the Khmer state’s 153administration, became a means of integrating the land and its produc­tion into the structure of the state. 56 Family temples and their properties were subordinated to central temples placed strategically throughout the realm. A portion of the production collected by private temples was channelled to the state temples. In return the priests of local family tem­ples received validation through periodic participation in the rituals of the central temples. Local family cults also became legitimized via their worship by Khmer monarchs and their subordination to royal cults. In these times temples were not just religious centers but were important links in the state’s economic and political network. Religion supplied an ideology and a structure that could organize the populace to produce, tap this production, and secure a region’s political subordination, without the aid of separate, secular economic or political institutions.

Temples and Temple Networks in the Angkor Era

A regular order of royal priorities becomes discernable in the activities of four great Angkor kings: Indravarman I (877–889), Yaśovarman I (889–900), Rājendravarman II (944–968), and Jayavarman VII (1181–1220). They first built public works, usually large water reservoirs, or rehabili­tated the capital, including the hydraulic works. They also built ancestral temples honoring their immediate predecessors. However, similar pat­terns were not followed in the period between the reigns of Rājendravar­man II and Jayavarman VII (968–1181), leading historians to describe the reigns that occurred during this time as a “zone of imprecision.” 57 Yet within this zone of imprecision the reigns of Jayavarman V (968–1001) and Sūryavarman I (1002–1050) left a great abundance of inscriptions and along with the reigns of Udayādityavarman II (1050–1066) and Sūryavarman II (1113–1150) were eras of great royal construction. Dur­ing these four reigns the seeming “imprecision” may well reflect the rapid growth of population in the Angkor area and the resulting tensions in property relations as powerful families at the court competed for eco­nomic resources—especially land and labor—turmoil that ultimately benefited the kings. 58

In the reign of Sūryavarman I, who is viewed in Khmer history as a usurper, a process can be observed by which royal favor enabled some elite families to prosper at the expense of others. 59 Those families who especially prospered were those who had willingly subordinated their own interests to the king’s and were incorporated into the royal “bureau­cratic” order. At least sixteen major inscriptions recording the histories of “bureaucratic” families were issued between 1002, the beginning of 154Sūryavarman’s reign, and 1080. In two inscriptions from Sdok Kak Thom, for example, Sūryavarman I is assigning land to officials of high rank for the purpose of increasing production that could be shared by the officials and the state. 60 Royal inspectors were sent to the site of the land to be transferred to validate the transaction, especially to determine if the land for development was “unowned.” This did not mean that the land was unoccupied but that another aristocratic family did not hold rights over the land. Indeed one of the inscriptions notes that villagers were already occupying the land and that the king turned over rights to the production and labor of these well-established agriculturalists to the family receiving official recognition. One noticeable feature of this trans­action was the role of the local temple in the proposed development scheme, for the villagers came under the administration (cāt camn āt) of a family temple.

From this evidence we can see that Khmer monarchs in the pre-Angkor era were clearly incapable of projecting this degree of royal interest. Pre-Angkor monarchs could only recognize the possession of land rights, rather than having wide powers of land assignment. In tenth-century Cambodian epigraphy, on the other hand, although the extent of the king’s authority over religious establishments (temples) is unclear, the king became more actively involved when the resources of two or more temples were merged. The religious network forged by Angkor’s kings insured that the landholding rights of Khmer elites would remain frag­mented; only those in royal favor could consolidate or extend their prop­erty rights. This was done by forbidding miśrabhoga, the joining of the rights to lands of more than one local estate’s temple, without the approval of the king himself. 61 As noted earlier, miśrabhoga consolida­tions, initiated by regional elites, were common in pre-Angkorian society as a means by which a regional elite established its power base. Seemingly in the Angkor era miśrabhoga consolidations were closely supervised in order to limit the growth of a regional elite’s resources. Land-right trans­fers to temples not involving the king still took place, but Angkor-era monarchs were as a matter of course asked to sanction the transactions. The Khmer monarch became the spiritual overlord of private property rights, acting as a judge in disputes, but in most instances only inter­fering when his participation (or that of his officials) was first solicited.

The apparent key to the Khmer capital’s control over manpower was its ability to form “lord-subordinate” alliances with local leaders. Kings, acting from a center of authority, fragmented the power of potential ene­mies by formulating agreements in which these prospective opponents became subordinates of the state. Eleventh-century Khmer inscriptions show that Sūryavarman I, whose rule marked a high point in Khmer 155administrative development, was particularly adept at incorporating the chiefs of peoples on the outer edges of his personal domain. After “con­quering” a territory with his armies, he recruited the local chiefs (elite) of the “defeated” peoples as district chiefs (khloñ viṣaya) and their power status changed little. 62

It was during Sūryavarman I’s reign in 1011 that the king’s administra­tive corps (tamrvāc) pledged an oath of allegiance, swearing to become the local “eyes” of the king and pledging to feed the capital with infor­mation about local activities. 63 Tamrvāc who pledged their loyalty to Sūryavarman were most active in the districts of those appointed khloñ viṣaya and were likely intended to serve as a reminder of the king’s authority in the khloñ viṣayas’ regions and as a source of information in regard to their loyalty. However, in later inscriptions tamrvāc are specifi­cally listed as being among those royal officials whose rights in a temple domain were voided. This suggests either a royal response to local pres­sure to remove royal officials, or it may also suggest that tamrvāc became too closely allied to the local khloñ viṣaya. The elimination of the tamrvāc’s local authority—which in such instances could have been util­ized to reinforce the local elite’s authority at royal expense—would thus have had a negative impact on the power of the khloñ viṣaya.

Early historiography on classical Southeast Asia depicted the Khmer state as a social pyramid with the king and his elite sitting on top and lit­tle contact between them and the people below. 64 In Sūryavarman I’s state, however, there seems to have been a more intense relationship between the royal capital and the local populations. The Khmer state protected regional interests by incorporating the local status quo into its formal structure instead of replacing the local landed elites with officials sent out from the capital. It further benefited local communities by mak­ing its army available to guard the entire state, maintaining order in the regions under state control and protecting the domain from invasion. Unlike pre-Angkorian inscriptions, which denote that the presence of court officials at the local level was small, Sūryavarman’s inscriptions indicate the physical presence of royal officials to administer the transfer of land and to directly collect revenues due to the king, as well as local visits made by the royal retinue (kaṃsteṅ). The kaṃsteṅ was a mobile body of state administrators who traveled from place to place within the realm, acting on disputes that could not be solved locally or on affairs that were considered to be within the state’s area of interest. 65 The kaṃ­ steṅ was thus a periodic visible symbol of the king’s administration, while the tamrvāc represented royal interest locally in the king’s absence.

Sūryavarman’s reign thus represents a critical phase in the develop­ment of an integrated Khmer economic and political order, with the 156Khmer temple network assuming a major role in the developmental pro­cess. The prosperity of families in the Khmer realm—a prosperity based upon control over the production of land and manpower—came to depend more and more upon royal favor. Interestingly, in the period immediately prior to Sūryavarman’s initiatives, as the Khmer realm moved into the theoretical “era of imprecision,” the percentage of royal inscriptions actually decreased during Rājendravarman’s and Jayavar­man V’s reigns. The most impressive new temple construction of that period was being attributed to named officials and not viewed as being due to the king’s direct initiative. Although the number of royal inscrip­tions relative to those of his subordinates increased during Sūryavar­man’s reign, temple construction inaugurated by families subordinate to the king continued to greatly outnumber that of the Khmer monarch. Only in Jayavarman VII’s reign did the old pattern of royal institution of temple construction again overshadow the efforts of Khmer “aristo­crats.” 66 This new pattern does not demonstrate weakness at the center; instead it came about through the intensified integration of the regional aristocracy into the Khmer state as “bureaucrats.” They were given offi­cial bureaucratic titles; their authority over land and manpower was rec­ognized; and they were charged with responsibility for the expansion of the Khmer state’s economic base. Along with this recognition, however, went the responsibility of sharing their land’s production with the state. Land transferred to aristocratic families was assigned specifically for the benefit of family temples, whose staffs assumed responsibility for super­vising its development. Because they were subordinated to royal temples, these family temples had to share the local production received with the central temples and thus with the Khmer kings who had made the origi­nal assignment.

What developed through this pattern of land assignment and develop­ment was a network of private and temple landholding rights that was subject to the supervision of the monarchy. Angkor’s society was elite-dominated. The king held the power to maintain elites with patronage and at the same time needed to prevent or neutralize the emergence of rival power centers. Elites were linked to the royal court through the Khmer temple network, if not through the bestowal of royal favor. Angkor’s rulers were capable of reducing local power centers to subordi­nate provinces of their government, but in the process of awarding terri­torial grants or property right transfers they were not able to dispense landholding rights at will. The landed elites subject to the Khmer state did have independent rights and were not subject to the constant demands of the state. In the Angkor era, although local autonomy was assaulted—sometimes with a good deal of energy—the landed aristoc­racy 157retained power. The Khmer state system was not highly centralized or “bureaucratic,” nor was it a “feudal” order in which the king assigned bureaucratic duties to a landed elite who derived their landholding rights and status as a consequence of the king’s favor. Records of land assign­ments to families and their temples during the Angkor era reflect con­stant friction between the center and its periphery and provide evidence about the nature of the relationship between the Khmer monarch and his regional elites. 67

“Ownership” of land in the Khmer realm, as noted above, was embedded in a system of rights held by related people. Exclusive prop­erty rights imply the exclusion of various official claims that would oth­erwise have force. The king in theory held the final authority to validate landholding rights, although this authority was not normally exercised—Sūryavarman I’s initiatives were unusual; the Khmer monarch thus assumed the role of patron placing land under the “exclusive” control of favored families and their temples. The aristocracy, drawing their liveli­hood from the land, theoretically owed their continued prosperity direct­ly or indirectly to royal favor.

Exercising the Khmer monarch’s theoretical land rights, Sūryavarman I thus dealt directly with the lineage of Pas Khmau, which he viewed as being constantly violent to his interest, by dictating that henceforth their land was to be controlled by the Khmer kings and the lineage’s resources merged with those of the king’s temples. 68 However, this was not a total seizure of land as aristocratic “rights” over the land were not reassigned to another family. In the Sdok Kak Thom inscription cited above, a fam­ily of officials claimed land and rights over land that was administered by a family temple. Rights over this land had been granted by a series of kings—new settlements (sruk), provisions, and manpower had been endowed to equip the temple and to guarantee the growth of the commu­nity around it. In the turmoil associated with Sūryavarman I’s rise to the throne the landholding family had apparently sided against Sūryavar­man. Sūryavarman gained authority over areas where the family held land but did not reassign all this family’s land rights to his direct sup­porters. Instead, Sūryavarman forced a temple priest related to the fam­ily—who was responsible for the property’s assessment—to leave the temple staff and to marry a younger sister of Sūryavarman’s first queen. Some of the original family’s land rights were reassigned, and the family was provided new rights over land in an area beyond the royal core where it would seemingly be less a threat to royal interests. The temple’s liveli­hood and as well the family’s prosperity were said to have been devas­tated in the disturbances of the era. Thus Udayādityavarman II renewed Sūryavarman’s endowment, allowing the new family of officials to clear 158land overgrown by forest, consecrate temple statues, and build its eco­nomic well-being. 69

In other instances Khmer monarchs imposed their authority locally by forbidding temple staffs to employ people living on and cultivating assigned land to also work temple lands, withdrawing certain rights of local officials over land assigned to temples, or favoring families loyal to the capital by exempting their temple lands from the demand of the king’s officials. 70 In these inscriptions Khmer monarchs reveal two pur­poses in their dealings with the land endowments of local temples: on the one hand they intervened where possible to limit the power of potential rivals, and on the other hand they desired to enhance the economic strength of their supporters. Avoiding possible conflict, Khmer kings rarely intervened in local temple affairs unless there was a direct threat to royal interests but adeptly employed their right to assign their sup­porters’ income rights to unsettled lands; rights to previously settled land left vacant could be solicited from the king by a family, the land rights of extinct lineages could be reassigned, and landholding rights to unpopu­lated and overgrown land could be assigned. 71 Families with existing estates were sometimes encouraged to resettle in new territories. 72

Several Khmer monarchs expressed concern that too much of a fam­ily’s resources was committed to the financing of temple affairs. In such cases they intervened to maintain the family’s economic well-being. A family in one such instance had transferred so many of their income rights to support temple activities that the king, upon finding their resources insufficient to maintain the family’s everyday requirements, ordered a halving of royal assessments on the unassigned land they had remaining. 73 In another instance the rights on land transferred to a tem­ple were excluded from the revenue claims of various officials but income rights not subject to revenue claims were explicitly distinguished from other landholding rights of the donor’s family, which provided the family’s subsistence and allowed them to carry out government require­ments. 74 Yet the substantial transfers of land rights to family temples by Khmer aristocrats and kings continued to be made, where family mem­bers who were members of the temple staffs were assigned the exclusive right of administering this land for the temples’ (and their families’) ben­efit. 75

If land assigned to a temple was not to be administered by the temple staff, the heads of the family or other members of the lineage might act as property managers on the temple’s behalf, utilizing income from the land to erect buildings, construct hydraulic projects, secure additional manpower to work the land, or in general ensure that the land’s produc­tivity 159would increase. The dominant landholding families thus also bene­fited from the development of temple land and the accumulation of wealth by the family temple. This income was tapped for redistribution by the family to its supporters in various forms. Family-temple relation­ships were rarely questioned by Khmer monarchs, who instead attempted to obtain a share of the wealth accumulated locally by the right their cen­tral temples had to a percentage of the local temples’ income. However, as will be discussed below, income from local temples covered only a small percentage of the royal temples’ expenses and this revenue sharing was more symbolic than critical to the financial well-being of the state temples. In this light the revenue demands of central temples upon local family temples would appear to have had more political and social than economic importance. Royal interest in local temples, aside from guar­anteeing the financial well-being of allies and limiting the economic resources of potential rivals, was more concerned about a local temple’s ritual being in harmony with that of royal temples than about making a local temple economically subordinate to a royal temple.

The existence of a hierarchical religious network gave the monarchy control over ritual. Moreover, the elaborate royal cults developed by Khmer monarchs helped to integrate subordinates with the center. 76 The king’s powers were generated by ceremony. The royal court, its activities, and its style recreated a world of the gods—in theory a heaven on earth, in which all greatness and glory were concentrated. By successfully ful­filling his role as the hypothetical focus of all sanctity and power, the king maintained the orderliness of the world. Subordinate centers of power in the Khmer realm sought to imitate the ritual of the royal court. 77 This ritual unity was more important than administrative control in maintaining the state’s dominance over areas outside its core. Territo­rial unification was not sufficient to sustain the realm. This came about through the integration of indigenous folk traditions, symbols, and reli­gious beliefs into a cult that was visibly concentrated in the center.

The participation in the state’s rituals by those who worshiped local spirits emphasized the subordination of local deities and their temples to those of the state, thus enhancing the king’s image as the holder of supreme spiritual power and divine prowess among the living. There were thus political implications to the institution of a centralized Khmer religious network system. As temple collections flowed from local to cen­tral temples and the integration of local and central cults became concen­trated in central temples, the Khmer temple network facilitated the Khmer monarchy’s manipulation of the regional landed aristocracy—themselves integrated into the royal “bureaucratic” order by having offi­cial 160religious and secular titles bestowed upon them—via the subordina­tion of these aristocrats’ local temples and cults to the monarchy’s tem­ples and cults.

Endowments to temples represented the mobilization, organization, and pooling of economic resources (capital, land, labor, and so forth) to support portions of the overall ritual process of the temple—for exam­ple, financing a single event in the temple’s religious calendar of ritual, the construction of a building in the temple compound, clothing for a temple image, or a subsidy for a temple priest. While this redistribution of economic capital was central to a temple’s existence, the mobilization of “symbolic capital” was also critical, as temple endowments generated one or more ritual contexts in which honors rather than material returns were distributed to and received by donors. In this way economic capital was converted to culturally symbolic capital, honors that enhanced the stature of the donor in the minds of his kin and clients. 78

An endowment permitted the entry and incorporation of Khmer cor­porate units (e.g., families and kings) into a temple as temple servants (priests, assistants, and so on) or as donors. The donor represented a social, economic, or political unit; the gift was a means by which the group or its leader could formally and publicly receive recognition. While an endowment supported the deity, and often returned some mate­rial advantage to the donor, perhaps more important were the symbolic returns of the “donation.” 79 Rulers—regional elites or Khmer monarchs—were patrons and protectors of temples, ensuring the continuance of a temple’s services, resources, and rules. They were not “rulers” of tem­ples, however, but were servants of the temple’s deity, human agents of the lord of the temple—a stone image that could not arbitrate in the real world on its own behalf—who protected and served the deity.

In the Indian tradition a ruler’s relationship with a temple represented a symbolic division of sovereignty, whereby the ruler became the greatest servant of the temple’s lord, his patronage of the temple’s deity sustain­ing and displaying his rule over men. 80 Yet kings and others who claimed political authority were subject to challenges by those who perceived their shares and rights—a consequence of other individual or group endowments to a temple—to be independently derived from the sover­eign deity. An example of the competition to claim the shared sovereignty of the secular (the ruler) and the sacred (the temple) is provided in the Khmer realm, where local elites and kings each patronized temple deities. The issue of who was the ultimate servant of the temple’s lord had politi­cal significance and explains the attempts by Angkor’s monarchs to sub­ordinate the deities of local temples to those of royal temples or to inte­grate these local deities into royal cults. 81

161The returns from temple donations thus had both economic as well as political implications and explain the significant flow of economic resources from rulers at various levels to Khmer temples. If an identifica­tion with a deity was as essential to legitimize rule as seems to have been the case in the Khmer realm, then instead of material returns upon one’s investment symbolic returns must have been equally desirable, especially to Khmer kings who never reaped substantial material benefit from their assignment of local income rights to local temples. The temple’s redis­tributive role was thus critical to issues of sovereignty in both the secular and sacred Khmer world order and raises questions of how to equate eco­nomic and symbolic capital, if it is really possible to quantify this conver­sion process, and whether there was any attempt to achieve “equity”—topics beyond the scope of the available historical records. 82

Family and Central Temples as Economic Centers

The extension of agriculture into previously uncultivated lands and the construction of hydraulic networks to facilitate the production of rice surpluses were central to the development schemes of Khmer kings and their subordinates. 83 Khmer temples at the state and regional levels ful­filled three economic functions in the agricultural development process. First, they were centers of investment (“banks”), the source of invest­ment capital and management; donors’ gifts were redistributed to indi­viduals or groups of peasant and bondsman cultivators as capital invest­ments (e.g., seeds, livestock, and land to be cultivated), which stimulated the agrarian sector. Second, temples were repositories of technological information and knowledge, directly or indirectly supporting scholars, astrologers, and artisans whose expertise and literacy could be drawn upon by cultivators. Third, Khmer temples were supervisory agencies that involved agricultural laborers in the development process, offering sufficient returns to encourage them to remain on the land. As noted ear­lier, lands assigned to temples for development were often unpopulated, requiring the assignment of a labor force with no previous claim to the land’s production. This work force might be acquired by moving a popu­lation, possibly war captives, from an area peripheral to the state’s core domain to the lands to be developed. 84 Laborers assigned to develop new lands were incorporated into the local economic and social system by temples. Lone peasant cultivators could not likely have borne the eco­nomic burdens of shortfalls in production as they brought new land into production or implemented agricultural technology associated with the construction of irrigation projects. 85 Temples, however, could mobilize 162their storage and redistributive mechanisms to meet the subsistence needs of the laborers in such an event, drawing from material resources assigned for this purpose by kings or regional elites. Temples were also in charge of agricultural development, engaging diggers, scribes, managers, and other specialists, combining the technical expertise and human resources necessary for the extension of agriculture. 86 Furthermore, tem­ples offered laborers emotional security; workers sought not only per­sonal economic profit but also worked for the spiritual gain derived from service to a temple’s deity. 87

A twelfth-century inscription from Trapaeng Don On illustrates the economic function of a “personal” or family temple. This inscription is especially valuable because it contains parallel Old Khmer and Sanskrit texts that are identical in context and permit full translation of both texts. 88 The founder of this temple was an individual claiming to be a Brahman, who had been the head herdsman for several kings. He had saved the treasures paid to him for his distinguished service and pur­chased lands, adding these lands to those he had acquired in other ways, and then had built a temple to complete the foundation of his estate:

Certain lands from this estate were assigned to support the temple’s activities (see fig. 5). This assigned land was divided into (1) a “common field of the cult” from which the harvest went into the common granaries of the temple, (2) “the field of the servants of the cult,” (3) “the field of the chief priest” (purohita), and (4) the “sustenance fields” of the temple kñuṁ (“bondsmen”), 89 agricultural laborers bound to serve the temple. There were nineteen adult kñuṁ, nine men and ten women, who worked the “common field of the cult” and the “field of sustenance”; other kñuṁ are likely to have farmed the additional fields that provided income for individual members of the temple’s staff. 90 The nineteen kñuṁ were divided into two groups; each labor team worked on the “field of the cult” half of a lunar month. This division of the lunar month was a feature of the Indian dating system adopted into the Khmer system. 91 Inscriptions recording the assignment of kñuṁ to temples usu­ally gave two lists, equal or approximately equal in length, for the bright and dark fortnights respectively. Temple treasuries, repositories for food 163produced by the kñuṁ on the temple’s behalf, were normally constructed in pairs, one for the deposit of each fortnight’s collected production. 92 Temple kñuṁ worked for their own subsistence or for other masters with coexisting claims on the land (e.g., the donor’s family) during periods free of temple duties. 93

The persons in charge of the temple were relatives of the founder:

The property, however, was not free of limitations, for the inscription continues: “Let him [i.e., the purohita] not sell or give away the slaves and the lands of these slaves….” Thus the land assigned to support the temple and its staff was at the personal disposal of the relatives of the temple’s founder, the Khmer aristocrat and “bureaucrat,” but with cer­tain restrictions.

Lands assigned to family temples were in some cases divided among the temple personnel or in other instances were considered together as “fields of the cult”; the temple personnel received shares of the “sacrifi­cial rice” from the “fields of the cult.” The temple kñuṁ “bondsmen” who worked the land had “fields of sustenance” apportioned to them by the temple staff for cultivation. They were allowed to retain a portion of their production from their assigned land or were given a share of the harvest from the produce collections belonging to the entire temple. Family temple lands also had specific obligations to the state, except 164when there was a royal decree relieving the temple of these responsibili­ties. 94 Land assigned to temples was free of direct revenue obligations to the state; especially forbidden were collections by royal officials who normally derived a portion of their income from making assessments for revenue demands. 95

In an inscription from Sūryavarman I’s reign (1028) a military leader bearing a royal title received the land rights and property of a royal enemy. 96 In developing his estate this individual consecrated a Śiva-liṅga and built a temple around it. Thereupon the king required the payment of miśrabhoga duties to a royal temple, although the inscription does not state the amount and type of duty that was imposed. Thus the “family” temple of this state “bureaucrat” was linked, as a criterion of its estab­lishment, to a central temple. The duties paid to the royal temples in miśrabhoga linkages of smaller local temples to larger royal temples were relatively small and were more symbolic than economically critical to the central temple’s existence. For instance, a typical local temple in Sūrya­varman I’s reign received an income of 25,000 kilograms of hulled rice annually but shared only 90 kilograms of it with a designated central temple. 97

That only limited demands were placed on local family temples is fur­ther conveyed in the accounts of the Khmer royal temples. The Ta Prohm temple’s rice needs during the reign of Jayavarman VII were said to be 6,589 kilograms daily for cooking and 2,512,406 kilograms annually. 98 This rice fed the temple’s personnel, who included 18 high priests, 2,740 officiants, and 2,632 assistants—among whom were 615 female dancers, 439 learned hermits who lived in the temple monastery (aśrama), and 970 students. A total of 12,640 people lived within the walls of the temple compound. The sum of rice consumed annually was thus 2,512,406 kilo­grams, of which only 366,800 kilograms were delivered by villages assigned to the temple and 42,157 kilograms from royal storehouses, together covering less than one-fifth of the temple’s rice consumption. 99 Villages assigned to the central temple supplied rice through their local family temples, but in comparison to the total amount of production annually drawn by local temples from their assigned lands the annual payment of roughly 90 kilograms of rice that these family temples were obliged to pay to a central temple was insignificant. The central temple’s primary source of income was the temple’s own assigned lands worked by kñuṁ “bondsmen” at the behest of Khmer monarchs and state elites. State-level central temples thus functioned economically in a fashion similar to the local family temples, but on a much larger scale.

Following the twelfth-century inscription of Jayavarman VII in his Prah Khan temple complex describing the Khmer realm as being com­posed 165of 306,372 subjects inhabiting 13,500 villages, one historian calcu­lated that Khmer subjects were producing roughly 38,000,000 kilograms of hulled rice annually for 20,000 gold, silver, bronze, and stone gods. 100 Each worker supplied an average of 120 kilograms of hulled rice, or 60 percent of his productivity. 101 The potential of the flow of production to and the concentration of economic resources in Khmer temples is demon­strated in the Ta Prohm inscription’s enumeration of the temple’s stores, which contained a set of golden dishes weighing more than 500 kilo­grams; a silver service of equal size; 35 diamonds; 40,620 pearls; 4,540 gems; 523 parasols; 512 sets of silk bedding; 876 Chinese veils; a huge quantity of rice, molasses, oil, seeds, wax, sandalwood, and camphor; and 2,387 changes of clothing for the adornment of temple statues.

The Khmer Temple Network in Southeast Asian Perspective

The Burmese state of Pagan, at its height from the eleventh to the four­teenth centuries, also drew economic support from a temple network. The economic basis of both the Khmer and Pagan states was irrigated rice agriculture, and the ensuing wealth from this production was con­centrated in temples. In the Khmer state, the temples were integrated into the state structure—temple officiants were state functionaries subordi­nate to secular powers, and conflict between the state and religious insti­tutions did not arise. In Pagan, however, rights to produce from the land shifted to the Buddhist monastic order (saṅgha) and away from secular political authority. Socioeconomic power shifted from the Pagan monar­chy to a rival, hierarchical order dominated by monks. With an increas­ingly diminished control over the economic resources in their domain, Burmese kings were forced to redefine the structure of the Pagan state; a new state system was organized around a new Theravāda Buddhist elite. 102

In contemporary Java, there was a pattern of royal temples and local temples—the latter built under the patronage of local rulers—similar to that in the Khmer realm. Although hundreds of small, local temples existed, during the eighth through the tenth centuries only four major temples—Caṇḍi Sewu, Caṇḍi Plaosan, Caṇḍi Loro Jonggrang, and Caṇḍi Borobudur—as well as the burial temples of kings were built. 103 The central temples were built under the authority of Javanese rulers claiming mahārāja status but with the cooperation and support of their state authorities and regional chiefs. The temples helped integrate the state; three levels of state offices and corresponding temple networks emerged at the state, regional, and village levels.

166Javanese temples operated within a royal ritual policy that was intend­ed to consolidate and extend the state’s political authority. 104 The pri­mary level of competition for political power, which was based on how much economic power one or the other could control, with economic power being linked to temples, was at the regional and state levels. The regional elites (rakrayān) controlled regions (watĕk) and subordinate vil­lage clusters (wanua), while mahārājas attempted to draw the rakrayān and wanua under royal authority. One strategy employed by the kings was to urge rakrayāns to stage great festivals. For example a ruler would encourage the establishment of a local family temple, usually assigning the land to the temple as sīma, land free of royal revenue demands for taxes and services, but would require the temple’s benefactor (a rakra­ yān) to stage a great public festival in honor of the temple’s founda­tion. 105 Such ceremonies and festivals were a means of redistributing wealth that might otherwise have been used in ways disruptive to society—and to the king’s sovereignty.

In the Singhasari-Majapahit era (eleventh through fifteenth centuries) eastern Javanese monarchs generously donated land to religious institu­tions throughout the countryside, freed temples from numerous royal taxes and dues (dĕrwaya haji), and also prohibited royal administrators and soldiers from entering religious domains to make demands on their own behalf. 106 The latter prohibition prevented collections by royal tax collectors and officials who were also working for local magnates and village authorities. Transfer of dĕrwaya haji thus often was of more con­sequence in a negative sense to local officials, especially rakrayān regional chiefs, than the king. 107 The transfer of the rights and property of local leaders to religious institutions subtracted a portion of the wealth directly available to a local magnate. A temple’s establishment under royal patronage also provided an institutional foothold from which Javanese rulers could attempt to extend their political and economic power in the hinterland beyond the centers of royal power.

The redistributive functions of Khmer and Javanese temples are sum­marized in figure 6. In this diagram the state’s resources flow as dona­tions or dues (gold, land, livestock, food, and manpower) from villages and local and regional temples to the state’s central temples. Returns may have been in kind, utilizing the temple as a “bank.” In such instances the donor received a material (i.e., “economic capital”) return upon his “investment,” either directly via a prescribed rate of return or indirectly through the greater productivity achieved regionally as a consequence of the temple’s efficient management of land. Returns might also have been “symbolic capital,” which contributed to the legitimacy of various regional and state aristocrats through the performance of temple ceremo­nies 167that emphasized the donor’s superior spiritual prowess or through the act of recording the gift in an inscription that perpetually honored the donor’s piety.

Figure 6

Temple Hierarchy in Angkor and Java (Classical Era)

Thus in Java, as in the Khmer realm, temples never became indepen­dent of those in political authority, whether at the local or state level. Rivalries and tensions developed, rather, among local and state authori­ties whose power was based on their control over manpower, landhold­ings, and temple administration. There was never a shift of socioeco­nomic power away from the secular authorities to a religious order. However, both the Javanese and Khmer monarchs were limited by the very nature of their policy of utilizing temples and temple networks as a means of integrating their domains economically and politically. Neither developed a centralized bureaucratic order, depending instead upon assignments of land rights and impressive titles to those in royal favor to 168elicit the loyalty of semiautonomous regionally based landed elites. Yet neither appears to have been lacking in the ability to finance major royal projects. In both realms, in the absence of a bureaucratic system for col­lecting large amounts of income for the state’s treasury, temples were viewed as important centers of economic accumulation that could be tapped to finance the king’s patronage of religion. Most conspicuous were temple construction and elaborate temple ceremonies and festivals, which provided a pretext under which the state’s economic and social resources could be mobilized to achieve the state’s political goals. These goals mainly focused on the construction of a state-dominated political system replacing the previous system, which had been built on a series of personal alliances. In both systems temples assumed major roles in the process of political integration.