Virtual Sambor Prei Kuk, Cambodia

Daniel Michon, Ph.D.
1. Cambodia and the Indianization Debate
2. Śaivism and Kingship
3. The Organization of the State and Religion: The Maṇḍala as Model

Download PDF version of the full essay

1. Cambodia and the Indianization Debate

That the study of pre-Angkor and Angkor Cambodian civilization (c. 6^th c. ce – 13^th c. ce) has been India-centric is clear.[1] Early studies of Cambodian religion and polity posited the dominance of a state sponsored “Khmer Hinduism.” This Khmer Hinduism was reconstructed from two primary sources: one, the over one thousand inscriptions written in both Sanskrit verse and/or Khmer prose,[2] and two, the impressive material culture primarily in the form of temple complexes and their attendant architectural and artistic evidence.[3] Both sources, on their surface, suggest a strong Indic influence, and early studies naturally turned towards Indian notions of religion and civilization to decode them. Thus, a third primary source was introduced to understand Khmer Hinduism, South Asian texts from the Hindu tradition. The seminal work for this understanding of Khmer religion is George Coedés’ 1944 Les États Hindouisés d’Indochine et d’Indonésie.[4] Based on over thirty years of research—his first article on the subject was published in 1911—it became, and in some ways still remains, the standard textbook understanding of pre-Angkor and Angkor Cambodia. Coedés concludes his study of the Hindu nature of Southeast Asian societies by asserting that for the religious, literary, and legal domains, “underneath the diversity of the civilizations of Farther India, underneath their apparent uniqueness . . . lies the imprint of the Indian genius, which gives the countries studied in this volume a family likeness and produces a clear contrast between these countries and the lands that have been civilized by China.”[5] This notion of Southeast Asia as essentially a “Farther India” was followed by a series of studies that all took Indian civilization as their model for understanding the inscriptions and material culture of pre-Angkor and Angkorean Cambodia.[6]

However, this understanding of early Southeast Asian civilization did not go unchallenged. Immediately before World War II, Pal Mus,[7] J.C. van Leur,[8] and F.D.K. Bosch[9] looked to indigenous cultures for the impetus for the development of Southeast Asian civilization. Most forceful was J.C. van Leur’s thesis which suggested that the Indian civilizational overlay was “a ‘thin and flaking glaze’ under which the main form of an older indigenous culture continued to exist.”[10] Van Leur’s thesis was republished in English in 1955, and it served to re-invigorate the “Indianization” debate in Southeast Asia.[11] Van Leur’s argument was taken up in the following decades by O. W. Wolters[12] and I. W. Mabbett.[13] While both Wolters and Mabbett cautioned that the evidence coming from Southeast Asia itself was so thin that any historical reconstruction must be tentative,[14] they nevertheless argued for a more nuanced understanding of Indian influence. They suggested that Indian influence was not thrust upon Southeast Asian societies, but rather local elites chose certain Indic forms to gain economic and political advantages. These “external” forms were so integrated with Southeast Asian society that any notion of “Farther India” became irrelevant. However, at least in the study of early Cambodia, these arguments were overwhelmed by the majoritarian “Indianization” view as put forth by Coedés, and as Michael Vickery wrote in 1998, despite these alternative theses, “Cambodia scholarship has never entirely escaped the notion that the Cambodian state and civilization were transplanted from India.”[15]

Michael Vickery, then, is the most recent scholar who sets out to rectify this overemphasis on external sources and external influence. His 1998 book Society, Economics, and Politics in pre-Angkor Cambodia: 7^th and 8^th Centuries re-situates the debate firmly in the Mon-Khmer milieu and presents the most radical argument for the independent development of Khmer civilization. He argues that “details of early Cambodian history do not have to be studied with reference to Indic models, and that any interpretation of Cambodian data or institutions which relies for justification on what an Indian model would have been is to be rejected.”[16] His book relies exclusively on the over two hundred inscriptions from the seventh and eighth centuries to reconstruct the religion, society, and politics of pre-Angkor Cambodia.[17] As for religion, he is quite pessimistic that pre-Angkor religious thought and practice can be reconstructed at all, he writes, “[i]t must be emphasized that the intention of this chapter [Chapter 5: The Cult Component] is not to describe the ‘religion’ of pre-Angkor Cambodia, or Khmer religion, a task which may be impossible on the extant evidence.”[18] He claims that the inscriptions are not primarily of a religious nature, that the cult roles described in the inscriptions were “not their exclusive, or even principal, occupation,” and therefore the overwhelming majority of the inscriptions deal with practical matters, not religious ones.[19]

While Vickery’s “corrective” is a welcome continuation and refinement of the arguments put forth by Mus, van Leur, and Bosch to the study of pre-Angkor civilization, he often grossly overstates the case for an independent development of an indigenous pre-Angkorian religion and fails to see the profound influence, at least at the élite and state level, of Śaivism. It is Vickery’s commitment to a materialist interpretation of the data that clouds his analysis. There are many inscriptions that reflect the religious concerns of the Khmers, but because they are in Sanskrit he ignores them. For example, two inscriptions with clear religious import, K. 441 and K. 442—two inscrpitions that will be essential to the Sambor Prei Kuk Heritage project—are only mentioned two times in his 450 page book. Further, these two inscriptions, and many others, are in clearly religious locations. K. 440 and K. 441 are found on the pilasters of boundary wall gate enclosing a sacred area, but others are found on temple pilasters, image-pedestals within these temples, and stone stelae within the temple precincts. David Chandler, in discussing the pre-Angkorian inscriptions, argues that the very fact that inscriptions were written on stone indicates their religious importance. Chandler writes, “The reason they [the inscriptions] were carved at all may have been that writing on stone, the medium of gods, served a special purpose. Stone was not used in secular sites; these, including palaces and ordinary dwellings, were built of wood, bamboo, and other perishable materials . . . In incising stones, Cambodians were speaking, collectively, to their ancestors.”[20]

With this said, Vickery is certainly right that local forms of religion would have remained strong throughout the pre-Angkor and Angkor periods. But the type of religious influence coming from India would have encouraged such practices. Of the three major strands of Indian religion that emerged in this period—Śaivism, Pāñcarātrika Vaiṣṇavism of the Bhāgavatas, and Mahāyāna Buddhism—Śaivism had the greatest impact, especially on the pre-Angkor period. Evidence that Śaivism was the most influential of the three Indic traditions that made their way into Cambodia is born out from Vickery’s own analysis of the inscriptions:

Within the inscriptions used for this study [that is, the pre-Angkor corpus], there are over 130 instances of named Indic gods (vraḥ) and about 90 names of gods, thus rather few multiple occurrences of any single god name. Nearly 50 of the names, most ending in īśvara, seem to be Śivaite, 14 are Viṣṇuite; there are 8 occurrences of combined Śiva-Viṣṇu (harihara, śaṅkaranārāyaṇa), one apparent Sūrya, eight gods with non-Sanskritic names, and three with Sanskritized titles which may be regarded as local pre-Hindu deities, plus seven Buddhist inscriptions.[21]

Vickery argues this evidence away by claiming that this only reveals that local deities took on Śaiva and Sanskritic names, but underneath these names were the local gods and goddesses of Cambodia.

However, when the pre-Angkor inscriptions are coupled with the material evidence in the form of Khmer art and architecture, particularly that of temple art and architecture, it seems undeniable that Śaivism—and, as we shall see, its attendant form of subsidiary Brahmanism—played a large role at the élite level. Alexis Sanderson, the foremost scholar of early Śaivism, writes,

. . . we have the evidence of a great wealth of material culture in the form of the remains of religious edifices, images of their deities, ritual objects, and bas-reliefs showing scenes from the Indian epics and the life of the population. The sheer number of the Khmer’s temples, the vast scale of the greatest of them, and the inscriptions that detail their endowments, reveal that the creation and support of such foundations was central to the economic, cultural and political life of the whole society. They channeled and promoted agricultural production, engaging a very substantial proportion of the region’s human and material resources, they integrated the realm, and they legitimated the tenure of land and power.[22]

In light of all of this evidence, Sanderson concludes that “there can be no doubt that for several centuries after the sixth century it [Śaivism and its attendant subsidiary Brahmanism] was the principal faith of the élites in large parts of the Indian subcontinent and in both mainland and insular Southeast Asia.”[23]

2. Śaivism and Kingship

Alexis Sanderson attributes the success of Śaivism in India and Southeast Asia to a few key features in the way Śaivism was implemented and practiced. First, Śaivism posited the possibility of multiple layers of religious adherence and practice. Only a few élites were initiated into the inner-knowledge of the liberatory Śaiva path, but these élites also actively cultivated the support of the wider community of uninitiated, lay devotees.[24] They cultivated this support in two ways: one, they offered a ‘lesser Śaivite religion’ of merit-gathering that centered on the more mundane concerns of lived life to both the royal family and the non-élites. And two, they accommodated other religious traditions. That is, they did not take offense to the worship of non-Śaiva deities and to the implementation of non-Śaiva rituals. This was because Śaivas saw their religion as dealing with a different set of concerns. The non-Śaivite rituals, practices, and gods were efficacious in the limited domains they sought to control, but for a more ultimate understanding of the universe and for ultimate liberation, one had to be initiated into the Śaiva path. Thus, élite Śaivism existed in harmony with other forms of religion. Sanderson notes when Śaivism encountered other these other religious forms

. . . they did not attempt to reform worship by restricting it to the narrow pantheon that they propitiated as initiates. This they imposed on the worship of Śiva in the Liṅga at the heart of these foundations; but they also took over, preserved, and regulated in accordance with the expectations of the uninitiated laity a much wider range of ancillary deities, deities that have no place in the scriptures and ritual manuals of [Śaivism].[25]

Now, the above argument is situated in the Indian context with regards to the layering of Śaiva ritual on top of brahmanical ritual, but Sanderson shows how this would also hold true of the incorporation of non-Indic religions. Thus, while Śaiva adherents would like to project an unchanging tradition to the public, it is clear that

. . . in the domain of worship performed by professional priests for others, such as we encounter in Khmers’ inscriptions, the pressures to depart from the purists’ model would surely have been [great]. We should consider it very probable that the Paddhatis [manuals of ritual] that guided their ceremonies among the Khmers were freely modified over time to appeal to or satisfy the expectations of new clients.[26]

Sanderson demonstrates the fluidity of the relationship between Śaivite scripture and paddhati with three examples: Kashmiri Śaivism, Nepali Śaivism, and Balinese Śaivism.[27] With these three examples, he argues that it would be wise to assume the fluidity of the Khmer Śaivite tradition as well.

The most important factor for the success of Śaiva religion both in India and Southeast Asia was its links with the institution of kingship. Śaivism became attractive to royal patrons by “extending and adapting its repertoire to contain a body of rituals and normative prescriptions that legitimated, empowered, or promoted all the key elements of the social and political developments that characterize the early medieval period.”[28] These characteristics of the early medieval period that Sanderson refers to are similar to what Vickery suggests in his modified version, or better his fluid version, of the Asiatic Mode of Production (AMP).[29] Vickery has an ambiguous relationship with the classical formulation of the AMP:[30] while on one hand he recognizes the limitations of such a model—and in particular he wants to disassociate it from Karl Wittfogel’s notion of ‘Oriental Despotism’[31]—he also recognizes its power, as “those who would reject an AMP, or any Marxist category, are constrained to make use of something similar under a different, or no, name, unless they are satisfied to lapse into the one of the current fashions of new irrationalism.”[32] Vickery’s fluid notion of the AMP rests upon a list of the basic, common traits of the AMP, or as Sanderson would call them early medieval, societies. Each society would be its own unique mix of these traits, but no two societies would be exactly the same. Thus, Vickery concludes that the power structure described in these inscriptions would not be at all like the static, classic vision of the AMP, but rather a unique formation that would draw from the common traits which the AMP identifies. For a list of these traits, Vickery suggests starting with Lawrence Krader’s list of twenty-four characteristics of AMP societies.[33] The most important traits on this list are “state domination of agrarian production, little private property, and middle levels of society consisting of state functionaries rather than private landlords or commercial groups . . . [and although] large hydraulic works, in particular centrally controlled hydraulic works, are not an essential feature . . . public works of some type, in Angkorean Cambodia the temples, are [essential].”[34] Vickery also adds one more critical trait, not on Krader’s list, for Southeast Asia, “the appropriation by leaders, first at the village, and later the state, level of direct linkages to the spirit world, through which they justify their leadership.”[35] These last two echo Sanderson’s elements of the proliferation of land-owning temples and expansion of an agrarian base which is assimilated through cultural and religious symbolism.

3. The Organization of the State and Religion: The Maṇḍala as Model

This conclusion, that the élites used Śaivism as a way of solidifying their control of a fluid state apparatus in pre-Angkor Cambodia while allowing for local traditions to continue unabated, leads directly to the contentious issue of state development in Southeast Asia. How was this state apparatus organized and how did it function? What was the role of religion and ritual? The debate concerning the nature of the early Southeast Asian state is long and complicated, and the best discussion comes from Robert L. Brown’s study of Dvāravatī circular stone sculptures called dharmacakras.[36] While the primary focus of Brown’s study is the Dvāravatī culture of central Thailand, he argues that

. . . South East Asia (particularly pre-ninth century South East Asia) can be viewed in many ways as sharing cultural features, characteristics that are often discerned best in the Indian mirror . . . Indeed, already in the seventh century Khmer élites are developing patterns—the use of inscriptions to record genealogies, Hindu art to demarcate territorial victories—that will lead them to a political integration [with other regions of South East Asia].[37]

Brown adamantly rejects the notion that this “political integration” was achieved through western notions such as Marx’s and Wittfogel’s ‘Oriental Despotism’ or Western notions of a “market model.”[38] Rather, it is the analogy of the maṇḍala that most adequately conveys the nature of such polities.[39]

In both Hindu and Buddhist ritual, a maṇḍala is a complex geometric design, most often either a circle or a square, meant to harness the energy of the cosmos as it replicates, in miniature, the nature of the cosmos itself. The source energy of the maṇḍala, and thus the cosmos, emanates from the center, flowing outward and decreasing in power as it reaches the periphery. A key characteristic of a maṇḍala is its fluidity. Any particular position in a maṇḍala is not static, but depends on an ever changing network of relations. While in theory the center of the maṇḍala is stable, in practice the center itself is inhabited by different forms of the divine depending on the religious orientation of its users. This is even more true when the political structure is modeled after the maṇḍala with the king at the center and his tributaries flowing out from that center—inhabitants of the maṇḍala are constantly jockeying for a position closer to the center. Further, particular regions might fall in the realm of more than one maṇḍala, or certain leaders might try to create their own center, and their own maṇḍala, in turn. Thus, as Wolters argues for the state in Southeast Asia, the state as maṇḍala

. . . represented a particular and often unstable political situation in a vaguely definable geographical area without fixed boundaries and where smaller centres tended to look in all directions for security. Maṇḍalas would expand and contract in concertina-like fashion. Each one contained several tributary rulers, some of whom would repudiate their vassal status when the opportunity arose and try to build up their own networks of vassals.[40]

Brown posits three overlapping characteristics of Southeast Asian maṇḍala states: (1) a center occupied by the city/palace/king, and the authority of the king is religious and symbolic rather than political. (2) The ruler’s power was legitimated in proportion to his ability to convince others that he was at the center of the maṇḍala. The primary technology of legitimation was the ritual theater of symbols,[41] that is, the use of art/architecture and ritual to link the ruler to the cosmos. (3) A state apparatus which used this ritual theater to convince others to see itself as the legitimate center.[42] Therefore, art/architecture, ritual, and statecraft were not separate realms of civilizational development, but were fundamentally tied together in very real ways. While Robert Brown’s focus is on the art of Dvāravatī, his conclusions could be extended to include not only artistic expression, but also architectural and ritual expression as key elements to Southeast Asian polities,

If the ruler expresses his power primarily through a symbol system, rather than through military [or economic] coercion alone, art [and architecture, but also just as importantly ritual] could have enormous appeal for the ruling élite. There would also be an impetus to spread this art [and architecture and ritual] to the surrounding polities as an expression of the ruler’s control. In addition, if there was a concomitant local concern in copying the center, there would be an equal interest by the local ruling élite to mimic the art of the center.[43]

Thus, when Richard O’Connor identifies pre-Indian influenced Southeast Asian cultural features, such as bilateral-kinship and the “big-man” concept, that are similar to the notion of a model for state development based on a maṇḍala, it becomes clear that the Śaivite religious ideas would find fertile ground in which it could plant its seeds and grow.[44]