From Its Origins to the Fall of Angkor
The great pioneering works of French scholarship on ancient Cambodia were primarily concerned with the construction ofroyal chronologies and with the problem of how Hinduism had been transplanted in an alien setting.1 The towering figure in the field, George Cœdès, had been trained as a Sanskritist and regarded Southeast Asia as a tabula rasa for the reception of Indic religious and cultural ideas and practices, which, rather astonishingly, appeared to bear exactly the same meanings as they had in their land of origin. This is odd, for old Khmer inscriptions are actually slightly more frequent in Cambodia than those in Sanskrit, even from the earliest period. One of the earliest, found at Angkor Borei [K. 600] and dated 611 ce, is written entirely in Khmer, for example.2 For the joint Sanskrit/Khmer inscriptions, it is a rule of
thumb that “the Sanskrit text is concerned only with the spiritual benefits acquired by some pious act or other, while the Khmer text is to some extent a notary’s deed placed under divine protection” (Jacques 1970, 24). The Khmer portion, then, is addressed to living men and women; the Sanskrit, to the gods. Given this fact, Vickery (1998, 139) has argued that the ancient Khmer were materialists “overwhelmingly concerned with practical, not religious, aªairs.” Jacques (1982, 40; 1990, 15), on the other hand, regards the inscriptions as “exclusively religious documents.” Here, then, are some of the quandaries facing the investigator exposed to divergent and incompatible readings ofthe most basic kind. What can we know about religion in ancient Cambodia? Were the old Khmer eªectively “Indianized,” or did autochthonous deities survive their exposure to foreign influences by putting on a suit of Indic clothing? There are no clear answers here, although Mabbett (1997, 350) has persuasively suggested that the storehouse of imported mythologies would have provided a rich resource for the forging of a unified identity in the expanding polity that emerged in the lower Mekong region in the early centuries of the Christian era. Little is known about the domestic life of ancient Cambodia. The inscriptions and other material remains are the work of elites and were often associated with the king himself. The eªorts of several generations of archaeologists have quite naturally been directed at the clearance, reconstruction, and interpretation ofgrand temples. The spaces between these impressive structures and much of the larger Khmer region—spaces in which ordinary people can be assumed to have lived, worked, and worshipped—have been largely ignored.3Any reconstruction of this enormous period will have many lacunae, and, in consequence, it would be unwise to claim that one could oªer a coherent account. As Mus (1975, 7) once sagely observed, “One sometimes risks confusing a library with a country.”
thumb that “the Sanskrit text is concerned only with the spiritual benefits acquired by some pious act or other, while the Khmer text is to some extent a notary’s deed placed under divine protection” (Jacques 1970, 24). The Khmer portion, then, is addressed to living men and women; the Sanskrit, to the gods. Given this fact, Vickery (1998, 139) has argued that the ancient Khmer were materialists “overwhelmingly concerned with practical, not religious, aªairs.” Jacques (1982, 40; 1990, 15), on the other hand, regards the inscriptions as “exclusively religious documents.” Here, then, are some of the quandaries facing the investigator exposed to divergent and incompatible readings ofthe most basic kind. What can we know about religion in ancient Cambodia? Were the old Khmer eªectively “Indianized,” or did autochthonous deities survive their exposure to foreign influences by putting on a suit of Indic clothing? There are no clear answers here, although Mabbett (1997, 350) has persuasively suggested that the storehouse of imported mythologies would have provided a rich resource for the forging of a unified identity in the expanding polity that emerged in the lower Mekong region in the early centuries of the Christian era. Little is known about the domestic life of ancient Cambodia. The inscriptions and other material remains are the work of elites and were often associated with the king himself. The eªorts of several generations of archaeologists have quite naturally been directed at the clearance, reconstruction, and interpretation ofgrand temples. The spaces between these impressive structures and much of the larger Khmer region—spaces in which ordinary people can be assumed to have lived, worked, and worshipped—have been largely ignored.3Any reconstruction of this enormous period will have many lacunae, and, in consequence, it would be unwise to claim that one could oªer a coherent account. As Mus (1975, 7) once sagely observed, “One sometimes risks confusing a library with a country.”
- Funan
The earliest known settlement in Cambodia, dating from c. 4200 bce, is Laang Spean, near Battambang. The site appears to have been occupied down to the ninth century ce (Mourer 1977), yet the first significant polity in the region is generally known as Funan. The word is thought to be the Chinese equivalent of the Khmer “bnar,”meaning “mountain.” Evidence relating to Funan is restricted to a small body of Chinese writings from diªering periods of history, some roughly contemporary epigraphic materials in Khmer and Sanskrit,4and the growing data assembled through recent archaeological investigation. The paucity of sources has meant that considerable imagination has been used in the reconstruction of Funan’s history. Even Cœdès fell into the trap when, with little evidence at his disposal, he argued that the kings of the region who employed the Khmer title “kuru¡ bnar” (= Sanskrit 8ailaràja, or “king of the mountain”) ruled from somewhere near Ba Phnom. This identification is now known to be untenable. The most likely candidate for the capital, called Vyàdhapura, appears to be Angkor Borei (Vickery 1998, 19). At its height, Funan’s influence probably extended throughout the Mekong Delta and into much ofpresent southern Vietnam, as well as the central Mekong region. It may also have included some ofthe Chao Phraya valley and the Malay Peninsula (Cœdès 1968a, 36–37). Archaeological investigations conducted by Malleret at Oc Eo and related coastal sites during the 1940s demonstrated that this important Funanese port was a link in the Indo-Roman trade during the early centuries of the Christian period. Certainly, Funan seems to have been a socially stratified society in which rice production and hydraulic works had both reached a significant level of sophistication. The ethnic character of the people is unknown, although the existence of inscriptions from the early seventh century points to the Khmer identity of at least some of the inhabitants, probably for many centuries prior to this period. Vickery (1998, 19–20) has argued that, possibly for reasons of prestige, some of the village chiefs of the region, previously known under the Mon-Khmer title of poñ,began to take the Indic su‹x “-varman,”5 from perhaps as far back as the fifth century ce. These individuals were probably ritual and clan chieftains who claimed descent from predominantly female, pre-Sanskritic deities (kpoñ). These, in turn, slowly transmogrified into higher-status divinities designated by the honorific term “our lord” (vraƒkamratà¡añ) (Vickery 1998, 153). Present levels of knowledge make it impossible to be certain about when Buddhism first arrived in Cambodia. The evidence is “at best sketchy” (Skilling 1997, 93), but around forty carved buddha images from the Mekong Delta region and from areas of Thailand associated with Funan6 have been found in a variety of materials, including stone, wood, glass, clay, bone, and metal. Their iconography is varied.7 Given the concentration of finds around Wat Romlok, near Phnom Da, this area has been regarded as a focus of Buddhist cult activity, although both Angkor Borei and Tra Vinh Province in Vietnam must also have been important (Bhattacharya 1961, 18). Dating must remain tentative,8 but other significant discoveries include a fine standing buddha in varamudrà, probably dating from the seventh century, discovered at Tuol Tahoy, Kompong Speu Province; two images of buddhas in parinirvà¢a, dating from the fifth– sixth centuries, one from Oc Eo and the other from Angkor Borei; and a roughly contemporary image ofthe Buddha seated under a multiheaded nàga,also from Angkor Borei. According to Boisselier (1967), the Tuol Tahoy buddha shows many similarities to Dvàravatìstatues, a fact that could be used to support the generally accepted thesis that Cambodian Theravada Buddhism was originally influenced by Mon culture. However, the most characteristic element of Dvàravatìart, the wheel of the law (dharmacakra),is not attested in ancient Cambodia. This need not imply that there were no links between Funan and Dvàravatì, but the evidence is slender. Nevertheless, a few pieces imported from India and China have been found, indicating Funan’s position as the hub of complex trading connections. Of particular note is a Gandhara-style buddha head of probable Indian origin, dated to the fifth–sixth centuries and discovered at Wat Kompong Luong, Angkor Borei (Tranet 1998, 439). Scholars of an earlier generation tended to regard these finds as evidence of Indianization in the sense that Brahmanical and Buddhist artifacts were exported from their place of origin by powerful individuals intent on establishing familiar religious and cultural ideas in alien lands. Cœdès (1968a, 55–56), for instance, postulates waves of emigration of southern Indians to Funan in the first half of the fifth century, perhaps as a result of the campaigns of Samudragupta (r. c. 335–375) and the resulting submission of the Pallava dynasty. But we know that the people of coastal Indochina were skilled seafarers from early times, and it is perfectly possible to envisage mariners bringing back familiar images as souvenirs from their trading activities around maritime Asia. This theory has a number of attractions. It obviates the need to establish why Indians might have been motivated to engage in Hindu-Buddhist missionary activity, and it fits with what we know about the seafaring traditions of early Southeast Asia.9 But, most significantly, there is little evidence that imported Buddhist objects had any great influence on the evolution ofCambodian sculptural styles (Boisselier 1966, 266). Certainly much ofthe Buddhist imagery from Funan reflects, though not slavishly, the schools ofBuddhist art that were flourishing within greater India at the time, but native aesthetics were not compromised. Had such images been introduced with specific missionary intent, one would surely have expected them to have exercised a more substantial influence on the indigenous artistic tradition. Scattered Chinese sources illuminate the scene a little further.10A fifth-century account given by an individual named Che suggests that “two hundredFodu (probably Buddhists)...from India” were living in Tuen-siun, a vassal state of Funan (Pelliot 1903, 279n5; italics mine). History of the Southern Qi Dynasty (Nan Qi shu), written between 479 and 502 ce, preserves a poetical account of the religion of Funan by a Buddhist monk Nàgasena (Na-kia-sien). It seems that in 484 he had been sent as an envoy to the Chinese emperor by a king of Funan, perhaps called Kau¢qìnya-Jayavarman, in an unsuccessful attempt to induce the Chinese to provide military aid. Nàgasena is reported to have told his hosts that Brahmanism and Buddhism both flourished in Funan. Although he emphasizes that “the custom of this country was to worship the god Mahe8vara (9iva),” he uses a number oftechnical terms when speaking ofBuddhism, such as “pàramità” and “bhùmi,” which strongly suggest a Mahayana presence (Cœdès 1968a, 58–59). A further point worthy of note is that, in both this embassy and its successor in 503, the Funanese king is said to have sent Buddhist presents to the emperor, including two ivory stùpas and a coral buddha image (Majumdar 1953, 32), a sound indication that Buddhist ideas and practice had, in some form, infiltrated royal circles. Another two Buddhist monks from Funan, named Sa¡ghapàla (460–524) and Mandra,11arrived at the court of Liang in Nanjing in 506 and 503, respectively. The Sanskritic forms of their names give no indication of their ethnic identity, and there is little justification in assuming, as some scholars have done, that they were Indians. Having said that, both had competences in Sanskrit, for translations into Chinese ofIndic canonical texts are ascribed to them, and they appear to have stayed at the Bureau of Funan (Funan-guan) while at Nanjing. As was the case with Nàgasena, it is di‹cult to derive a feel for the doctrinal a‹liations of these important early intermediaries. Of the two, Sa¡ghapàla appears to have been the more able scholar, yet both tended to work on Mahayana sources. This is, however, insu‹cient evidence to conclude that they were Mahayanist monks. Sa¡ghapàla’s biography tells us that the majority of his translations concerned Mahayana doctrine, although he was also interested in the Abhidhamma.12 One of the texts he brought from Funan and subsequently translated, although from what original language it is di‹cult to be certain, was the Vimuttimagga,a work probably composed by a certain Upatissa in the second century ce in northern India. This work will be of interest to us when we start to examine the much later features of traditional Theravada Buddhism in Cambodia. The Vimuttimaggahas widely, though not universally, been thought of as a text associated with the heterodox Abhayagirivihàra of medieval Sri Lanka, which may have played some role in shaping the forms ofBuddhist practice in Cambodia following the fall of Angkor in the fifteenth century.13 The presence of the Vimuttimagga might suggest the existence of both Mahayana and 9ràvakayàna styles of Buddhism in Funan. One other Funanese monk is known to the Chinese sources. His name was Subhuti, but his activities are not recorded (Yang Baoyun 1998, 141). We also hear of the Chinese monk Yunbao, who made a reverse journey from the court ofLiang to Funan in response to a series ofsix embassies sent by King Rudravarman (r. c. 514–c. 550) to China between 517 and 539 (Mabbett 1986, 295).14 He returned to China bearing the gift of a Buddhist relic, a twelve-foot-long hair of the Buddha. However, the most famous name to be associated with Buddhist contacts between Funan and China is undoubtedly the celebrated Buddhist translator Paramàrtha (c. 499–569). According to the second volume of Biography of Eminent Monks(Xu gaoseng zhuan),composed in the late seventh century, Paramàrtha was a native of Ujjain in western India but spent a period of time in Funan before arriving in China in 546 ce. It appears that the Chinese emperor had requested—some sources suggest a higher level ofcoercion— that the king of Funan, probably the same Rudravarman, might wish to supply his country with Buddhist texts and eminent teachers of the Buddhist dharma.15 Paramàrtha was subsequently chosen, arriving in China with 240 bundles of texts. It is unclear whether he had brought them all the way from his native India or had collected them in Funan itself. Paramàrtha never returned to Funan; indeed we have no idea how long he had been there when he received the call from China. But this evidence, particularly when taken together with the translation interests of Sa¡ghapàla and Mandra, points to the strong likelihood that the rulers of Funan were somehow involved in the patronage ofMahayana Buddhism. It is di‹cult to be more precise, although it is certainly tempting, given Paramàrtha’s own doctrinal propensities and the much later practice-oriented and quasi-tantric trajectory of Cambodian Buddhism, to wonder whether the Yogàcàra school of thought, as represented in India by the Valabhìmonastic center, might have found a toehold in Funan. A Sanskrit inscription at Ta Prohm of Bati [K. 40], probably from just before the middle of the sixth century, mentions both Jayavarman and Rudravarman and begins with an invocation addressed to the Buddha. This is followed by mention of holy relics, the Buddhist triple jewel (triratna), and details of how a Brahmin court o‹cial became a Buddhist lay-disciple (upàsaka)(Cœdès 1931, 8; Majumdar 1953, 41). A roughly contemporary inscription from Prasat Pram Loveng, Thap Muoi, in the Plaine des Joncs [K. 7],16 on the other hand, commemorates the installation by Prince Gu¢avarman17 of VaiU¢ava cult objects at the temple of Cakratìrthasvàmin in a ceremony presided over by Brahmins well versed in Indic sacred writings and traditions. The text also mentions the existence of Hindu priests of the Bhàgavata sect (Majumdar 1953, 40).18 From this evidence it seems likely that both Buddhism and Brahmanism were supported by influential figures. However, the presence of Buddhist artifacts and inscriptions from Funan is greatly outweighed by those influenced by Brahmanism. Indeed, the finest works of art associated with the region are statues of KwU¢a Govardhana, ViU¢u Balaràma, ViU¢u Para8uràma, and LakUmì, probably dating from the early to mid-sixth century (B.-P. Groslier 1962, 61).19 For most ofits history, then, Funan had some ofthe characteristics ofa Hindu state in which an atmosphere of religious tolerance seems to have operated. Bhattacharya’s view (1961, 16) that the kings of Funan sponsored Buddhist activity for “purely political imperatives” seems a little harsh and di‹cult to justify on the basis of available evidence. Both VaiU¢ava and 9aiva cults certainly operated at the level of the court, but Mahayana and 9ràvakayàna styles of Buddhism also oscillated in importance. It goes without saying that the majority of the population continued to practice ancestral cults.
2.Zhenla
According to Chinese sources, Funan was eventually subdued by a neighboring vassal state called Zhenla. Like Funan before it, there is no mention of such a place in Cambodian epigraphy, yet it must certainly have existed, given that an enormous growth in the inscriptional record in the seventh century confirms the existence of a polity very like that described by the Chinese. In Funan the poñhad not claimed ownership of the land, which remained communal property, but this type oflocal authority figure was eventually supplanted by a diªerent form of ruler, the mratàñ-varman, who became a distinctive element in the early Zhenla political landscape. These changes appear to be associated with transformations in the religious cult. In the words of Russian scholars, “[Earlier temples developed] from the sacred place of lineage god ancestors of the commune into a focus of wealth separated from the commune, and held by a privileged commune upper class. This was conducive to the replacement ofpreIndic local beliefs by Hinduism and the formation of a priesthood.”20 A fair quantity of statuary has survived from the early Zhenla period. It retains earlier features but is characterized by a certain stiªness and “frontality.” This mild degeneration into the so-called Phnom Da style B is nicely illustrated by an early eighth-century image of Avalokite8vara from Rach-Gia, near the Mekong Delta.21A number of buddha images, mostly from the south, also date from this transitional period. The most important of these is a Buddha image from Tra-vinh, the hair of which is rendered by large, flat spiral curls with an almost absent uU¢ìUa (B.-P. Groslier 1962, 64, 71). The architecture, on the other hand, usually consists oflone or grouped brick towers with stone-framed doors. The earliest of these exclusively Brahmanical sites are the brick tower of Preah Theat Touch in Kompong Thom Province and the sandstone structure of Asram Maha Rosei on the slope of Phnom Da, near Angkor Borei.22 These structures have been regarded as the first examples of a truly Khmer art, the initial phase ofwhich is sometimes termed the Sambor style after Ì8ànapura (modern Sambor Prei Kuk, to the north of Kompong Thom), an early capital ofZhenla that was founded by Ì8ànavarman I (r. c. 616–635). The southern group of temple buildings at Ì8ànapura certainly seems to date from this reign. Although in a ruinous state today, the brick core of its central shrine is supposed to have housed a gold li¡ga, while the eastern enclosure tower acted as a stable for 9iva’s mount, Nandin. Both were oriented to catch the rays of the rising sun.23 According to History of the Sui Dynasty (Sui shu), the original capital of Zhenla was on the middle Mekong at Bassac. Cœdès (1968a, 66) has identified the site with Wat Ph’u. At a sacred mountain nearby, human sacrifices were made to the spirit P’o-to-li (probably Bhadre8vara, a form of 9iva); the king visited the temple once a year to perform the rite at night. History of the Sui Dynasty also confirms the existence of Buddhist monks and nuns in Zhenla. They appear to have participated in funerary rites during the reign ofÌ8ànavarman I (Cœdès 1968a, 74–75), but by the time that Bhavavarman II came to power around 639, the veneration of 9iva had begun to eclipse both Buddhism and all other Brahmanical cults. This seems to fit the observations of the Chinese pilgrim Yijing in the late seventh century regarding the region. Although he never actually visited Cambodia, Yijing’s Southeast Asian informants had told him that in Funan “the people...venerated many gods.” Then “the law ofBuddha prospered and spread about” until an unidentified “wicked” king “expelled and exterminated them all, and there are no members of the Buddhist Brotherhood at all” (I Tsing 1896/1982, 12). Cœdès (1968a, 67) has oªered the opinion that this king may have been Bhavavarman II. Actually, there is some evidence that devotion to ViU¢u and Harihara did continue, although the inscriptions of the next king, Jayavarman I (r. c. 655–681), point to an increased emphasis on the li¡ga. Vickery’s survey (1998, 140–141) of extant Khmer inscriptions dating from Zhenla lists ninety diªerently named Indic gods (vraƒ). More than half, given their -ì8vara su‹x, are probably references to 9iva. Of those remaining, fourteen concern ViU¢u, eight mention 9iva-ViU¢u combinations (i.e., Harihara), and there is one reference to the sun god, Sùrya. But there are also seven specifically Buddhist inscriptions: K. 828, K. 49, K. 505, K. 755, K. 163, K. 244, and K. 132. To these we should add a few scraps of epigraphical material written in Sanskrit. K. 828 need not detain us, for it is only a small piece ofgra‹ti, but in K. 49— a dual Sanskrit/Khmer inscription from Wat Prei Val, Kompong Trabek, Prey Veng Province, dated 664 ce—two monks (bhikUu), Ratnabhànu and Ratnasirha, are named and described as brothers. In the Sanskrit portion we discover that they have donated slaves, animals, and land toward the foundation of a religious property.24 In addition, King Jayavarman will guarantee that a certain 9ubhakìrti is to have hereditary rights over the establishment. The Khmer portion of the text ascribes the title “pu caƒañ”(= sthavira), or “elder,” to both monks. Bhattacharya (1961, 16) has suggested that this means they must have belonged to the Theravada. This is possible, but the term, although monastic, really implies monastic seniority and is not convincing evidence of sectarian a‹liation. Of rather more significance as evidence of possible Theravada presence in Cambodia is a portion of Pali text [K. 820] engraved on the back of a seventh-century buddha figure from Tuol Preah Theat, Prey Veng Province.25Pali is, ofcourse, the canonical language ofthe Theravada. Furthermore, Dupont (1955, 190–221) believes that the figure shows some Dvàravatì influences. Another Khmer inscription from Prachinburi Province, Thailand, dated 761 ce, is not listed by Vickery, because it was discovered fairly recently (Rohanadeera 1988). It contains three Pali stanzas in homage to the triple jewel that appear to come from the TelakaVàha-gàthà, a poetical text believed to have its origin in Sri Lanka. As such, it represents the strongest evidence of a Theravada presence at this period of time. To this one might add a final, though not conclusive, piece of support from a dual Sanskrit/Khmer inscription [K. 388] from Hin K’on, Nakhon Ratchasima, in the Korat region of Thailand.26The Sanskrit portion mentions the donation of ten vihàras, four stone boundary (sìmà) markers, and some caityasby a royal monk (ràjabhikUu)to “provide for the body of Sugata [= Buddha].” Also mentioned is the donation of two sets of monastic robes (cìvara) in a kaVhina ceremony. Filliozat (1981, 84) has, dubiously in my opinion, interpreted this as a reference to Theravada practice. Two more inscriptions refer to the long-standing practice ofmonastic slavery.27 One—in joint Khmer and Sanskrit from Khlau Rang, Prachinburi Province [K. 505], dated 639 ce—enumerates the gift of pagoda slaves (kñur vihàra), plantations, and treasure to a vihàra by a certain Sinàhv.28 The roughly contemporary K. 163 from Prasat Ampil Rolum, Kompong Thom, is more specific. Poñ Prajñàcandra is recorded as having donated slaves to a trinity of Buddhist deities: vraƒkamratàñ añ 8àstà(= Buddha), vraƒkamratàñ añ Maitreya, and vraƒkamratàñ añ 8rìAvalokite8vara.The presence of Avalokite8vara as one of the triad has led some scholars to regard this inscription as the earliest explicit evidence for the existence of the Mahayana in the region. BernardPhilippe Groslier (1962, 77) has gone somewhat further, suggesting that the gradual increase in production ofMahayanist images throughout the Zhenla period could represent a gradual broadening ofBuddhist practice beyond the confines of the court into the mass of the people. Cœdès (1968a, 94), on the other hand, has chosen to characterize K. 244 from Prasat Ta Keam, Siem Reap Province, as the first explicit piece of evidence for the existence of Mahayana Buddhism in Cambodia; it is specifically dated 791 ceand mentions the erection of an image of Loke8vara, consecrated with the title jagadì8vara.Both theories are problematic, most notably because the veneration of Avalokite8vara is attested in Theravada settings and cannot be taken as a decisive indication of Mahayanist activity.29 Although these specific inferences may not be correct, there does appear to have been a considerable expansion of the Mahayana throughout the Southeast Asian region from the mid-eighth century, perhaps as a result of the sponsorship of the Pàla kings of northeast India and the growing influence of Nàlandàuniversity. In some ways it would be surprising for there not to have been a resurgence in Buddhist contacts with Southeast Asia during the Pàla period, for these royal patrons of Buddhism controlled and made accessible the major Buddhist pilgrimage sites. In consequence, Nàlandàbecame a mecca for Buddhist scholars, so much so that Bàlaputra, king of 9rìvijaya, built a monastery for monks from his realm in the precincts of the great monastic center around 860 ce (Kulke and Rothermund 1990, 119).30 It seems that a combination of tantric ideas and symbols contained within a Hindu-Buddhist syncretism, common to Bengal and surrounding regions, began to make its presence felt in Cambodia from around this time.31 The erection of an image of 9rì Vidyàdhàra¢ì (= Prajñàpàramità) by a physician—mentioned in K. 132, from Sambor Prei Kuk and dated 708 ce—may conceivably fit this context, while a Sanskrit inscription from the same location [K. 604], dated 627 ce, tells us that a Brahmanical teacher of the 9aiva Pà8upata sect, Vidyàvi8e8a by name, had studied Buddhism, although it is impossible to say whether this study took place in Cambodia or in India.
Angkor before Jayavarman VII
The founding figure of the Angkorian period, Jayavarman II (r. c. 802–850), was connected by blood to earlier rulers. He arrived in the region from a place named Javà32 around 800 ce, setting up a power base at Indrapura, a not completely identified location but probably Banteay Prei Nokor, to the east ofKompong Cham. He then gradually extended his influence across much of Zhenla, subsequently moving his capital to Hariharàlaya, fifteen kilometers southeast of modern-day Siem Reap, and finally to Mahendraparvata (Phnom Kulen), where a Brahmin named Hira¢yadàma “learned in the magical science (siddhi vidyà)” ordained 9ivakaivalya as royal chaplain (purohita)33 so that he might perform rites associated with the cult of devaràja.34 Earlier scholars, such as Cœdès and Dupont, understood the devaràjato be either the deified king himself or a singular image of 9iva standing in the king’s stead. The matter has not been adequately resolved, but it now seems more likely that the devaràja was a special mobile image (calantì pratimà) of a protective deity (Kulke 1993, 355) or, perhaps, some sort of sacred fire (Woodward 2001, 257–258). It was on Kulen that Hira¢yadàma also performed a ritual that sacralized Jayavarman’s claim to be a universal king (cakravartin)35and symbolically broke his dependency on Javà. According to an important and very long Sanskrit and Khmer inscription from Sdok Kak Thom [K. 235], dated 1052, this ritual derived from four Indic texts: Vinà8ikha, Nayottara, Sarmoha, and the 9ira8cheda.36 These were probably tantric, with a sixth- to seventh-century northern Indian provenance. We do know that many Angkorian inscriptions, from as early as the beginning of the ninth century, mention texts of the 9aiva Àgama corpus (Bagchi 1930). These describe the proper construction of the various classes of 8ivali¡ga, and this expansion of 9aivism fits in well with what we know about religious developments toward the end ofthe Zhenla period. The Àgamas make it clear that the priests conducting such rites must be drawn from Brahmanical families of northern Indian origin, and it looks likely that there would have been significant intercourse with India itself at this stage of Cambodia’s history. Indeed, the Phum Ta Tru inscription [K. 538], dated 978 ce, underlines this, for another royal chaplain, BhaVVa Divàkara, is said to have come from the banks of the Yamuna River in northern India. Such an arrangement operated, however sporadically, for a considerable time; even as late as the last Sanskrit inscription in Cambodia [K. 300], dating from the fourteenth century, the priest Sarvajñàmuni is said to have come from Àryade8a, that is, India. Having said all this, we should avoid assuming that either 9ivakaivalya or Hira¢yadàma was definitively Indian. Only the ruins ofJayavarman’s Krus Preach Aram Rong Chen—the templemountain on Phnom Kulen, which housed the devaràja—remain today. However, the king did not reside there. It was too sacred and, on a more pragmatic level, rather inhospitable. He died at Hariharàlaya and was deified with the posthumous name “Parame8vara.” His funerary temple,37 Preah Ko, was erected at his death, though not completed until 879. Preah Ko comprised six towers in two rows on a low platform. The three towers at the front contained images of9iva, while the back three housed representations of the god’s female consort Gaurì. In addition, the central tower of the front row was dedicated to Jayavarman, and the one behind to his wife, who had also been given a posthumous name, “Dhara¢ìndradevì.” In this way the royal couple were incorporated into a ma¢qala of divinities, a characteristic feature of Angkorian religion, both Brahmanical and Buddhist, that was to become increasingly complex as the period unfolded. The next important king, Indravarman (r. c. 877–889), had a chaplain (purohita),9ivasoma, who was not only a member of the royal lineage but also appears to have been a disciple of 9ankara (c. 788–820), the great Indian 9aiva founder of the Advaita Vedànta school of orthodox Hindu philosophy.38 Like Jayavarman II before him, Indravarman clearly favored 9aiva cults, dedicating a temple to 9iva at the already sacred site of Phnom Bayang in the south of the country. But his most important monument is the Bakong, dedicated at Hariharàlaya in 881.39 Although also a 9aiva shrine, the Bakong resembles the Buddhist temple-mountain of Borobudur on the island of Java, which had been constructed by the 9ailendra kings around the middle of the ninth century (B.-P. Groslier 1962, 99). Inscriptional evidence from the reign certainly demonstrates that Buddhism had not entirely disappeared as an element in the religious life of the region. A Sanskrit inscription from Ban Bung Ke—near Ubon in the Mun valley [K. 495], dated 886 ce, and erected by one Somàditya— mentions Indravarman as the reigning king and describes the donation offields, gardens, slaves, and buªalo to make merit for Somàditya’s deceased father. The text also notes the installation of a stone image of “the master of all the munis, Trailokyanàtha.” As Majumdar (1953, 74) notes, Trailokyanàtha is “evidently a Buddhist god.” Indravarman was also the originator of the great hydraulic schemes that so graphically typify the material and symbolic aspects of Angkorian civilization. The technical and managerial resources that he must have commanded in order to construct the Lolei Baray40 (IndrataVàka)—a rectangular tank of three hundred hectares containing ten million cubic meters of water, aligned on an east-west axis to the north of Hariharàlaya—remain quite remarkable even from the contemporary perspective. This artificial body of water was not simply a reservoir. It had a clear symbolic significance, given that some of its water was diverted along canals to nearby sanctuaries,41the whole design evoking the oceans surrounding Mount Meru in traditional Hindu and Buddhist cosmology. Indravarman’s successor, Ya8ovarman (r. 889–900), was associated with another enormous hydraulic project: the Eastern Baray (Ya8odharataVàka), to the northeast of his new capital, Ya8odharapura. One of 9ankara’s great practical achievements had been to establish monasteries (maVha)in the four corners of India, and it was perhaps in imitation of this that, in the year that he acceded to the throne, Ya8ovarman commissioned one hundred hermitages (à8rama), some of which have been discovered by recent archaeological investigation.42 It seems that these à8rama catered to a variety of religious sects. The inscriptional record makes it clear that the southern shore of the Ya8odharataVàka was the site for a number of à8rama, including one for each of the assorted 9aiva and VaiU¢ava groupings that flourished in the kingdom. However, a Sanskrit stele found at Tep Pranam, Angkor Thom [K. 290], dated from the end of the ninth century, suggests that one of the structures in this area was a Buddhist monastery (Saugatà8rama).43 Its remains have not yet been identified—they were probably replaced when Tep Pranam was constructed—but the existence of organized Buddhism so close to the symbolic heart of the state points to o‹cial toleration of the religion. It is di‹cult to be certain of the factors that contributed to this situation, but a close study of the text of K. 290, alongside the very similar and roughly contemporary inscription of nearby Prasat Komnap [K. 701], the site of a VaiU¢avà8rama, suggests that the religious landscape of the period had syncretic features. It seems that the establishment of both VaiU¢ava and Buddhist structures is preceded by an invocation to 9iva, but it would be incorrect to push the notion of interreligious tolerance too far. Residents at the VaiU¢avà8rama seem to have held some Buddhists in low regard, for K. 701 warns that ignorant Buddhists with bad morals, possibly with regard to celibacy, are not welcome in the hermitage. Bhattacharya (1955a, 112–113) notes that the same stipulation is not made for 9aivas. The construction of the 9aiva 9ikharì8vara temple at Preah Vihear and the sanctuary of Lolei dedicated to his parents apotheosized as 9iva and Umàwas begun during the reign of Ya8ovarman, but his finest achievement was the construction of the Bakheng in 893 ce. With some 108 towers arranged around the central shrine, this seven-level temple-mountain, surmounted by five sanctuaries arranged in quincunx, appears to be overflowing with complex Indian cosmological significances. This is certainly the view of Filliozat (1954),44 although Vickery (1998, 59, 172) has suggested that Indic cosmological norms may have been imperfectly known in Cambodia at the time of the construction but have been anachronistically transposed onto the structure without consideration of other possible and more culturally compatible factors. The squared arrangement of the Bakheng does seem problematic from an Indic perspective, although it makes perfect sense in the context of what we know about traditional Southeast Asian spatial concepts.45 Like many of the important architecture structures at Angkor, the Bakheng had a dynamic history, in that it has been significantly remodeled on more than one occasion. Indeed, we know that by the mid-sixteenth century it had become a Buddhist temple.46 During the reign of Ràjendravarman (r. c. 944–968) a high-ranking Brahmin named Kavìndràrimathana, the only named architect in the epigraphical record, was charged with the construction of the Eastern Mebon temple on an island in the middle ofthe Ya8odharataVàka. Kavìndràrimathana is described as an emissary (chàra), but it is also clear that he was a devotee of the Buddha.47 In this capacity he was involved in the establishment of the previously mentioned Saugatà8rama, to which he donated slaves. He also founded the temple of Bat Chum in 953 ce. According to Sanskrit and Khmer inscriptions on its three towers [K. 266–268], Bat Chum enshrined a triad of deities: the Buddha, Avalokite8vara-Vajrapà¢i, and Prajñàpàramità.48A mystical diagram (yantra) of forty-eight Sanskrit syllables arranged on a lotus blossom was engraved on one of the foundation stones of the temple, appearing to emphasize further that the cult practiced at the site was Mahayanist and tantric (Cœdès 1952). However, some caution is required in this context, for we know that later forms of Theravada Buddhism practiced in the region49 employed similar imagery. Whether Bat Chum represents an early manifestation of this latter tradition or was fully Mahayanist must, given our present state of knowledge, remain an open question. Whatever the answer, there can be little doubt that Buddhism was patronized at the highest level, for one of Kavìndràrimathana’s relations, Vìrendravikhyàta, was also associated with the Buddhist cult. Furthermore, the foundational inscription of Kdei Car, Kompong Thom Province [K. 157], dated 973, tells us that the vihàra was constructed to house pure bronze images of Loke8vara and Devì(= Prajñàpàramità) donated by Ràjendravarman himself.50 A number of very significant Brahmanical structures were also completed in Ràjendravarman’s reign. The best-preserved of these today, Banteay Srei (Citadel of Women), is a 9aiva temple in miniature that was dedicated in 968 by two Brahmins, Yajñavaràha and his younger brother ViU¢ukumàra. An undoubted architectural gem, it is richly decorated with numerous scenes, including many derived from the Mahàbhàrata and Ràmàya¢a. Ràjendravarman, on the other hand, constructed Baksei Chamkrong and Pre Rup. The latter’s inscription [K. 806], dated 961, is of particular interest. Bhattacharya (1971, 99) has described its author as probably the most erudite figure of the Sanskrit epigraphy of ancient Cambodia. An adherent of the Brahmanical philosophical tradition, he demonstrates a broad understanding of the great range of Indian literary and philosophical thought, even referring to two important concepts of Mahayana Buddhist philosophy, representation-only (vijñaptimàtra) and emptiness (8ùnyatà).51 It is clear that quite sophisticated understandings of Buddhist intellectual categories flourished among the religious elites of the period, even though the evidence indicates that Buddhism occupied a subordinate position in the religious imagination of the early Angkorian period. The situation begins to change from the second half of the tenth century, when we witness a steady increase in the production of Buddhist statuary and, for the first time, we discover Buddhist episodes inscribed on temple lintels. Little of this material has been discovered with supporting epigraphy, and it has consequently been rather neglected. Nevertheless, much of the material exhibits stylistic parallels with Brahmanical statuary, most ofwhich has been more fully studied and stratified, so it is possible to arrive at some fairly accurate datings (Boisselier 1966, 271).52 On this basis we can say that sculpted images of buddhas and bodhisattvas were quite common in the late tenth century. Good examples are the seated Buddha from Peam Cheang, encircled and protected by the nàga-king Mucalinda, now in the Phnom Penh National Museum, and a bronze Maitreya from Wat Ampil Tuk (Girard-Geslan 1997, 182). Both show Indic influences, particularly in the way that the conical uU¢ìUa is composed of many fine braids of hair in the style of a tiara (mukuVa). During the reign of Ràjendravarman’s son, Jayavarman V (r. 968–1001), 9aivism was still in the ascendant. However, a detailed Sanskrit inscription from Wat Sithor [K. 111], dated 968, indicates the extent to which Buddhism was percolating the region. The first section (vv. 1–9) starts with an invocation to the Buddha, dharma, and bodhisattvas, followed by verses expressing adoration for the Buddha’s triple body (trikàya). The statement that “this world is nothing but mind (cittamàtra)” (v. 8) indicates that we are in Mahayanist territory. The author of the text, Kìrtipa¢qita, is a servant of the king. Thanks to the latter’s eªorts, “the law of Buddha reappeared from the darkness, as autumn brings out the moon that up to a short time before had been veiled by the clouds of the rainy season.” The servant continues, “In his [the king’s] person the doctrines of emptiness (nairàtmya)and subjectivity (cittamàtra) . . . reappear like the sun that brings back the day. He reignited the torch of the true law, the 8àstra Madhyantavibhàga, and the other, that the destructive gusts of sin had extinguished. He brought in from foreign lands, in order to spread their study, many philosophical books and treatises, such as the Tattvasargrahacommentary” (vv. 26–29). There is a strong suggestion here of an attempt to reestablish the intellectual credentials of Buddhism in Cambodia after a period of persecution. It is di‹cult to be precise about the context, although the occurrence of Mahayanist concepts in K. 806 only six years before could suggest that any harassment of Buddhism may have occurred before 961 ce. The missionary strategy outlined by Kìrtipa¢qita appears to be based around the importation, possibly from India, ofrelatively sophisticated philosophical concepts and supporting texts. Specific mention of the commentary to the Tattvasargrahaand the Madhyantavibhàga, a key work of the Mahayanist Yogàcàra school, is particularly intriguing, given that Paramàrtha, a great champion ofthe same school, had a connection with the region some four centuries before. Could it be that Kìrtipa¢qita was trying to revivify this earlier tradition? K. 111 concludes with a series ofroyal ordinances relating to various aspects ofthe Buddhist cult, such as the correct performance of monthly rituals to the constellations [vv. 52–55], procedures for the dedication of a monastery [vv. 58–65], daily rites [v. 67], ceremonial ablutions of buddha images at the full moon and other occasions [vv. 70, 71, 74], copying sacred writings [v. 82], and much else besides. The inscription of Phnom Banteay Nan [K. 214], dated 982, is also connected with Jayavarman V’s reign.53The text’s author, Tribhuvanavajra, is “celebrated for his discipline (vinaya).” In Sanskrit he praises the Buddha’s triple body (trikàya)before invoking Loke8vara and Prajñàpàramità, the “mother of the buddhas.” The Khmer portion also mentions his many donations to Trailokyavijayàgì8varì (= Prajñàpàramità). Another purely Buddhist inscription from Thmar Puok [K. 225], dated seven years later, invokes an even more complex set of six Buddhist deities—the Buddha, Prajñàpàramità, Loke8vara, Maitreya, Vajrin (= Vajrapà¢i), and Indra—said to have been enshrined by the sage Padmavairocana.54 A preoccupation with Mahayanist pantheons, then, is a distinct feature of the period. Sùryavarman I (r. 1002–1050) was accorded the posthumous title “Nirvà¢apada,” clearly Buddhist in inspiration. Both epigraphy and the northeastern Thai Pali chronicles of the fifteenth and sixteenth centuries, such as the Càmadevìvarsa, suggest that Khmer power began to extend to the Chao Phraya basin during his reign. Despite the evidence of K. 701, relations between various religious sects fell into a more settled pattern. Sùryavarman I’s Lopburi inscription [K. 410] tells us that Mahayana and Sthavira Buddhist monks lived in proximity to Brahmanical ascetics. They are all ordered to oªer the merit of their austerities to the king, and anyone hindering their holy retreat is to be expelled and punished heavily.55 A more markedly syncretic outlook is revealed around this period. An early piece of evidence from Preah Khan of Kompong Svay [K. 161], dated 1002, starts with an invocation to 9iva NaVaràja and is followed by one to the Buddha. It contains a mélange of Buddhist and 9aiva ideas (Bhattacharya 1961, 35–36). Similarly, the inscription from 1041 ceat Phimai [K. 953] has a Sanskrit invocation to 9iva on one side of the stele, while a Khmer verse honors the Buddha on the reverse. That this state ofaªairs persisted is evidenced by the Trapan Don On inscription [K. 254] of 1129, which lists oªerings to 9iva, ViU¢u, and the Jina (victor) of Va¡8àràma—an epithet of the Buddha (v. 30). A probable explanation is that it is related to the growth of tantrism in the region. Indeed, a recently discovered bilingual inscription from Sab Bàk near Nakhon Ratchasima, present-day Thailand, dated 1066,56 provides clear evidence of the presence of tantric Buddhist concepts. The Sanskrit part of the text refers to the five buddhas (pañcasugata), together with Vajrasattva in his capacity as their chaplain (purohita). We also read ofa powerful individual, possibly Sùryavarman I, who has eliminated a threat to Buddhism and consolidated the religion in the country. The Khmer portion informs us that nine images of Buddha-Loke8vara, formerly erected on a mountaintop to forestall a threat from Java, have been renovated. This all appears quite consistent with what we know about the protective and exorcistic character of many tantric rites. A final interesting feature of the inscription is a reference to a secret tree (guhyavwk8a) from which the fruit of truth may be eaten (Chirapat Prapandvidya 1990, 12). There is, perhaps, some connection between this and other magical trees mentioned in the esoteric Buddhist texts of a much later period.57 We know virtually nothing about the reign ofJayavarman VI (r. 1080–1107), but later inscriptions link him both to 9aiva constructions, particularly at Preah Vihear and Wat Ph’u, and to Buddhist structures at Phimai and Preah Pithu.58 Phimai (Vimàya) is the most important and imposing Khmer monument to survive in present-day Thailand. Started around 1000 ce and architecturally representing an early transition between the Baphuon and the Angkor Wat styles, the site was probably associated with Buddhism before this date, as attested by the incorporation of a stone inscribed in Sanskrit [K. 1000], dating from the eighth century and mentioning a statue ofthe Buddha (muniràj)in the masonry ofthe first gallery (Dagens 1988, 18). Phimai once contained a large amount of Buddhist tantric imagery,59 most significantly a figure of Vajrasattva surrounded by a profusion of Brahmanical iconography and an image of Trailokyavijaya (“conqueror of the three worlds”), who, in a Khmer inscription [K. 397] of 1110 ce, is described as the general (senàpati) of the kamrate¡ jagat of Vimàya.60 This use of military terminology fits the tantric and exorcistic context quite nicely. The temple still boasts fine Buddhist scenes on the lintels of the sanctuary, including the first known Khmer Buddha in bhùmispar8amudrà(Boisselier 1966, 274). Banteay Samre and Beng Mealea, both Buddhist foundations dating from the early to mid-twelfth century, are similar to Phimai in structure and iconographical arrangement. Erected during the reign of Sùryavarman II, at about the same time that the great VaiU¢ava temple of Angkor Wat was being constructed, both have a central tower adorned with Buddhist imagery surrounded by VaiU¢ava structures and devices. These caused problems for earlier investigators who tended to misidentify the small number of Buddha figures on the lintels of Beng Mealea, concluding that the temple must have been exclusively Brahmanical.61 As Boisselier (1952–1954, 218) sagely notes, the iconography of Beng Mealea and Banteay Samre is su‹ciently problematic to make it di‹cult to form any clear hypothesis about the type of cult practiced there, but, given the obvious parallels with Phimai, the most likely explanation is that they were all conceived as three-dimensional ma¢qalas,instantiating elaborate Mahayana Buddhist-dominated pantheons arranged to provide a religio-military protective function in accord with the prevailing tantric mood. When we turn to the reign of the greatest Angkorian king, Jayavarman VII (r. 1181–c. 1220), we shall
have more evidence at our disposal to indicate how Mahayana Buddhism performed this apotropaic role.
Jayavarman VII: Symbolism of Temple and State
The various capitals of the Angkorian kings may be read symbolically as miniature images ofthe universe. The rivers and baraysrepresented the cosmic ocean; the enclosing walls, the iron-mountain chain (cakravàla) at the limit of the world’s golden disk; and the temples, the central world mountain, Mount Meru. The various hydraulic projects at Angkor may all be reduced to a simple form of a baray surrounding a temple-mountain. In essence the mountain collects heavenly water and is fructified by the encounter. The water then runs down into a baray,from where it fertilizes the surrounding soils. We find this arrangement first at Kulen, but it may also be observed at Angkor Wat and the great Buddhist temple complexes ofJayavarman VII. From his cosmic center the king ruled; he both “consumed” his domains and, by the proper performance of the royal cult, radiated back quasi-divine power. In this way, order and prosperity were sustained. At death he was absorbed back into the immaterial realms above the central world mountain and replaced by a new king. The great cosmic cycles of origination and dissolution so typical of the Khmero-Indic worldview were recapitulated as one reign followed another. This basic pattern worked whether the cult performed at the center was Brahmanical or Buddhist, the system reaching its apogee during the reign of Jayavarman VII. The temple of Ta Prohm,62 one of Jayavarman’s first great Buddhist temple projects, was constructed in 1186 to house an image of his mother, Jayaràjacùqàma¢i, in the form of Prajñàpàramità, the mother of the buddhas. Its foundational inscription [K. 273] includes invocations to the triple jewel, Loke8vara, and Prajñàpàramità. It also tells us that Ta Prohm was supported by enormous royal donations, including 3,140 villages and their 80,000 residents, plus more than 600 female dancers. The king and various landed proprietors (gràmavant)also supplied more than 5,000 kilos ofgold plate, 4,500 jewels, and much else besides. However, the cosmology of temple design is best illustrated at Preah Khan of Angkor, founded several years later on the basis of the astonishing donation of 13,500 villages and 306,372 male and female slaves from various foreign regions.63 In its foundational inscription [K. 908], dated 1191 ce, we read that the great temple housed an image of Loke8vara, rendered in the form of Jayavarman’s father, as its central divinity. Its cosmological significance is that it lies in the neighborhood ofthree barays, or sacred tanks (tìrtha), representing Buddha, 9iva, and ViU¢u [K. 908, v. 33]. In one of these, the Preah Khan baray,or JayataVaka, Jayavarman built another temple, Neak Pean.64 The structure is a reference to the mythological Lake Anavatàpta, whose waters were deemed so pure and inaccessible that anything coming into contact with them must be thoroughly cleansed. Only those with advanced magical powers, like the Buddha, who supernaturally transported himselfto Anavatàpta for his daily ablutions, can journey there, and it is said that gods brought sixteen pots of its holy water for the daily use of A8oka, the archetypal Buddhist king (Sp. i.42). Kingship, then, is strongly linked to the mythology of the lake.65 According to Boisselier (1970, 94–95), the construction of an artificial version of Anavatàpta was de rigueur for any self-respecting Buddhist monarch. The Chinese pilgrim Xuanzang, for instance, is supposed to have seen such a structure fed by both hot and cold springs on the outskirts of Ràjagwha when he visited India in the seventh century, and the Sinhalese tradition also describes something similar built by Parakkamabàhu I (r. c. 1153–1186), who was an almost exact contemporary ofJayavarman VII (Cv.lxxxix.49). According to Buddhist traditions, the four great river systems ofthe world issue from Anavatàpta, so canals oriented in the four cardinal directions also naturally flowed from Neak Pean. Its sanctuary incorporated an image—Balàha, the horse avatar of Avalokite8vara, with human figures clinging to him for dear life—to the east of a central lotus-throned Buddha. These figures appear to be the shipwrecked merchants mentioned by the Kàra¢qavyùha sùtra, one of the principal Indic sources of the cult of Avalokite8vara. To the west is a reclining male figure identified as ViU¢u; to the north is a block holding sculptures comparable to the thousand li¡gas found at Phnom Kulen and clearly related to 9iva.66 Unfortunately, the southern image is too eroded to be identified, but Boisselier (1970, 99–100) has concluded that this architectural ma¢qalais a political statement of the universal power of the kingdom committed to the ideology of Mahayana Buddhism founded on, and not antagonistic to, the veneration of both 9iva and ViU¢u. The arrangements at Phimai, Banteay Samre, and Beng Mealea, although not so explicit, were probably inspired by similar considerations. We have already had cause to note that the cult of Avalokite8vara had been around in Cambodia for some considerable time. Indeed, the two Sanskrit inscriptions K. 214 (dated 982, from Banteay Nan, Battambang Province) and K. 417 (dated 970, from Prasat Chikreng, Siem Reap) mention the bodhisattva specifically.67 K. 908 also mentions that Jayavarman VII erected twenty-three images, named Jayabuddhamahànàtha, around the principal towns of his domains. Woodward (1994–1995, 106–107) has suggested that these were the “radiating” Avalokite8vara figures covered with small buddhas and a Prajñàpàramità on the chest, many of which are still extant. The iconography, with buddhas emerging from every pore of the bodhisattva, is also based on the Kàra¢qavyùha sùtra.68 The likelihood, then, is that this or similar texts were well known in Cambodia, a likelihood reinforced by the iconographic evidence of Banteay Chmar, a late temple of Jayavarman’s reign in the far northwest of the kingdom. Banteay Chmar once possessed eight unique two-meter-high figures of Avalokite8vara, carved in low relief on the southern side of the western gallery.69 Iconographically, they are in surprising conformity with the Kàra¢qavyùha. Jayavarman’s great temple-mountain, the Bayon,70 was constructed quite late in the reign. It housed a colossal stone image of the Buddha protected by the nàga Mucalinda. But its most distinctive external features are its towers, each with four faces.71 Unfortunately, there is little consensus over their identity. It has customarily been believed that they are the faces of Loke8vara. However, to Bernard-Philippe Groslier (1962, 183), and to Mus before him, the 54 towers with their total of 192 faces are an architectural expression of the Buddha’s famous miracle at 9ràvastì, where he manifested multicolored beams of light from every pore. To these scholars the faces are those ofthe historical Buddha, whereas Thompson (2000, 261), if I understand her correctly, suggests a link with Maitreya. Woodward (1981, 62) has argued that the faces were originally conceived as representations ofVajrasattva, though, when sometime later the Bayon was redesigned under the influence of resurgent Brahmanism, they were reinterpreted as images of Brahmà. It is di‹cult to decide among the various identities advanced in scholarly circles, yet Filliozat (1969, 47) has noted one striking parallel. The appearance of face-towers in both Cambodia and the Himalayan region, where eyes facing the four directions are painted on temple towers, occurs at around the same time. It may be that during the Muslim conquest of Bengal at the end of the twelfth century Buddhist monks fled not just to Nepal and Tibet, as is widely known, but also to Indochina.72Dominant Buddhist philosophical and aesthetic notions, many ofthem connected to the tantric current that had run strong in northern India for many centuries, would inevitably have spread and adapted to local conditions in the process. There is certainly clear evidence that the tantric divinity Hevajra and his elaborate pantheon were venerated during the Bayon period (Boeles 1966). Contemporary Khmer bronze vajras and other ritual implements of the type associated with tantric cult activity are also well attested. It is possible that these were additions to the current of Buddhist tantrism that had already been active in Cambodia since at least the reign of Ràjendravarman. The Loke8vara-Buddha-Prajñàpàramità triad is found in earlier strata of the inscriptional record, but, as we have already noted, Jayavarman VII specifically associated Loke8vara and Prajñàpàramitàwith his father and mother. What we seem to have here is a Buddha family in which the central figure in the triad, the Buddha, equates to the king himself. To put it another way, father/Loke8vara and mother/Prajñàpàramitàact as the bodhisattva lieutenants of the king/Buddha. If Woodward is correct about the identification of the Jayabuddhamahànàtha images, and if the Kàra¢qavyùhaor related texts really do support the iconography of extant Avalokite8varas of the period, we may well be looking at a symbolic representation of kingship extending from its Angkorian center to the farthest extremities of the realm. This notion is picked up and extended by Hawixbrock (1998, 76–77), who argues that the total religious symbolism of the monuments constructed by Jayavarman VII expresses the king’s status as a righteous ruler (cakravartin)in the classical Buddhist mold. The cakravartinideal is given added resonance by the fact that Jayavarman VII also constructed 121 “houses of fire” (probably rest houses) along the principal roads of his kingdom and 102 hospitals, perhaps in imitation of A8oka.73 Hawixbrock believes that there is good reason to regard some of Jayavarman’s temples—Ta Prohm and Preah Khan at Angkor, Banteay Chmar, and Wat Nokor are good examples—as sites of important victories over the Cham. They each act as three-dimensional cosmograms symbolizing the restoration of order through the power of a righteous king, but the ensemble of religious foundations can also be read as an overarching ma¢qala by which the realm was restored and purified following its violation during the Cham invasions. Through their construction and accompanying rites, the kingdom was transformed into an ideal realm. At the center of the great city Angkor Thom stands Jayavarman’s pantheon, the Bayon, representing Mount Meru; the nearby Royal Palace is homologized to the residence of Indra; and Neak Pean, to the sacred lake Anavatàpta. We do know that the Siem Reap River, which flows through the city, had been identified with the Ganges sometime previously.74 To complete the vision, the great provincial temples ofthe reign, like Ta Prohm ofKompong Svay and Banteay Chmar, represent the various islands surrounding the world mountain of traditional Buddhist cosmology. Symbolically, then, the Angkorian state has become coextensive with the entire world. Jayavarman VII’s cult ofstate Buddhism was undoubtedly the high-water mark for the Mahayana in Cambodia.
