THE DEVARAJA CULT THE CORNELL UNIVERSITY SOUTHEAST ASIA PROGRAM The Southeast Asia Program was organized at Cornell.nUniversity in the Department of Far Eastern Studies in 1950. It is a teaching and research program of interdisciplinary studies in the humanities, social sciences, and some natural sciences. It deals with Southeast Asia as a region, and with the individual countries of the area: Brunei, Burma, Indonesia, Kampuchea, Laos, Malaysia, the Philippines, Singapore, Thailand, and Vietnam. The activities of the Program are carried on both at Cornell and in Southeast Asia. They include an undergraduate and graduate curriculum at Cornell which provides instruction by specialists in Southeast Asiap cultural history and present-day affairs and offers intensive training in each of the major languages of the area. The Program sponsors group research projects on Thailand, on Indonesia, on the Philippines, and on linguistic studies of the languages of the area. At the same time, individual staff-and students of the Program have done field research in every Southeast Asian country. A list of publications relating to Southeast Asia which may be obtained on prepaid order directly from the Program is given at the end of this volume. Information on Program staff, fellowships, requirements for degrees, and current course offerings will be found in .an Announcement of the Department of Asian Studies, obtainab.le from the Director, Southeast Asia Program, 120 Uris Hall, Cornell University, Ithaca, New York 14853. . . ll THE DEVARAJA CULT Hermann Kulke Translated from the German by I. W. Mabbett with an Introduction by the Author and Notes on the Translation of Khmer Terms by J.tM. Jacob Originally published as "Der Devaraja-Kult," Saeau"lum, XXV, 1 (1974), P.tP· 24-55, by the Verlag Karl Alber, Freiburg. Data Paper: Number 108 Southeast Asia Program Department of Asian Studies Cornell University, Ithaca, New York January 1978 $4.50 iii CORNELL UNIVERSITY SOUTHEAST ASIA PROGRAM 1978 International Standard Book Number 0-88727-108-9 lV TABLE OF CONTENTS Translator's Preface • • • • • • • • • • • • • • • • • • • • • • • • vii Introduction • • • • • • • • • • • • • • • • • • • • • • • • • • • • • xv The Devaraja Cult: Legitimation and Apotheosis of the Ruler in the Kingdom of Angkor . . . . . . . . . . • • • • • • • l I. Preliminary Remarks • • • • • • • • • • • • • • • • • • • 1 II. Previous Theories • • • • • • • • • • • • • • • • • • • • 4 III. The Sdok Kak Thom• Inscription and the Devaraja Cult • • • 14 ,, IV. Devaraja--a "CaZanti Pratima" of the God Siva? • • • • • • 24 V. Were the Kings of Angkor "Participants" in Divine Rule? • 29 VI. Conclusion . . . • • • • • • • • • • • • • • • • • • • • • 36 Abbreviations • • • • • • • • • • • • • • • • • • • • • • • • • • • 41 Bibliographyt. • • • • • • • • • • • • • • • • • • • • • • • • • • • 42 Appendix: �otes on the Translation of the Khmer Text . . • • • • • • 48 V vi TRANSLATOR'S PREFACE Dr. Kulke should need little introductiont.to Indologists who do not read German, for his studies on Orissan history have yielded a number of articles in English which offer interesting parallels to the themes dis­ cussed in the present work on Angkor.1 This study on the devaraja or so­ called "god king" cult in Cambodia is among his most challenging contribu­ tions; the propositions it advances are found at least to provoke a general and earnest reexamination of accepted ideas about the significance of the cult. The implications of this are far-reaching, for the discussion to which this study is a contribution is not merely of an isolated ritual in a single kingdom: it bears on our understanding of the role and limits of royal authority in ancient Southeast Asi.a generally; the term devaraja has been (perhaps .illegitimately) taken from its original Cambodian context. and, interpreted as "god-king," applied to various other places which appear to have had similar institutions.. Indeed such institutions may have been similar, but if Dr. Kulke is .right they do not amount to the sort of "god-kingship," or to the p"lenitudo potestatis which may be glibly read into the term, that are so often inferred by those who write in any way about the "Indianized" kingdoms of ancient Southeast Asia. Evidence of the devaraja cult is one of the main props for the widely accepted notion that Indochinese monarchs were ."god-kings"; if the prop is removed the notion must be questioned; thus the desirability of an English trans­ lation of this study is too self-evident to require detailed justification. Nor is it necessary to explore here the ramifications of the cult across Asia or the historiographical background to it, because they are compendiously discussed and documented below by the author. He examines in some detail the previously advanced theories· about the cult, several of them addressed to the questio,,n whether the term devaPaja refers to a central cult object, such as a Siva linga on one of Angkor's numerous pyramidal monuments, or to the ritual as such, and argues that the views adv�nced by such authorities as G. Coedes are commonly unsatisfactory or inconsistent. The chief difficulty in identifying the devaraja with the ¾f. Kulke, "Some Remarks about the Jagannatha Trinity," Indologen­ Tagung, eds. Herbert Hartel and Volker Moe. .ller (Berlin, 1971), pp. 126-39; "'K�atriyaization' and Social Change: A Study in Orissa Setting," Changing India: Studies in Honou:ra of Ghu:raye, ed. S. D. Pillai (Bombay, 1976), pp. 398-409; "Kings without a Kingdom: The Rajas of Khurda and the Jagannatha Cult," South Asia, 4 (October 1974), pp. 60-77. vii ., monumental Siva lingas is 'that, whereas each ruler wished to re-create his kingdom symbolically upon a new shrine of his own with its own lihga or statue, the epigraphic references indicate that there was one devaraja, which could follow kings about. The chief difficulty in identifying it with an abstract ritual is that, on the author's interpretation, it had a physical personality. At this point the argument turns partly on the meaning of certain terms in the Old Khmer part of the Sdok Kak Thom• stele inscription, and the notes on Old Khmer terms contributed by Mrs. J. Jacob in the appendix, as well as the author's further discussion in his Intro­ duction, are particularly valuable on this account. The author's novel solution is to propose that the devaraja was a sort of,palladium of royal­ ty, a portable substitute for an original "fixed" temple cult object established by Jayavarman II, who founded the cult in the ninth century. Integral to the author's method, and a major contribution to the methodology of the subject whatever conclusions one may reach, is his rigorous separation of the two categories of evidence which may otherwise be misleadingly confused: the inscriptions mentioning the devaraja cult, and the monuments which used to house lingas as the centers of royal cults. We are not obliged a priori to infer that the two sorts of evi­ dence refer to the same thing, and the reasoning noticed above leads the author to the conclusion that in fact they do not. When they are sepa­ rated, it becomes easier to recognize that the epigraphically attested devaraja does not have to mean "god-king," and may well, on the indica­ tions provided by J. Filliozat's recent article using Tamil material,n2 ., refer to Siva as "king of the gods.n" Similarly, in the monuments, we can observe the importance of royal cults without being obliged to treat them as the cults of "god-kings." Dr. Kulke's Orissan researches, like J. Filliozat's article just mentioned, point to the value of Indian analogies in reaching an under­ standing of Indochinese material; and indeed the progress of research tends to emphasize the similarities rather than the differences which are sometimes postulated. These similarities stem in part of course from the direct influence of Indian culture upon Southeast Asia. In part also they stem from a common ancestry in the web of territorial cults that were dotted all over monsoon Asia in prehistoric times, influences or vestiges of which survive in folk religion still. Since the devaraja cult may well have in it echoes of such archaic religious sensibilities, this is one dimension of the subject, easily overlooked by historians, which de­ serves brief notice here. In enquiring into the motifs in prehistoric religion discussed by such authorities as R. von Heine-Geldern3 and H. G. Quaritch 2J. Filliozat, "New Researches on the Relations between India and Cambodia," Indiaa, III (1966), pp. 95-106. See also the author's Intro­ duction, below, where further evidence is cited. 3In English, see his "Conceptions of State and Kingship in South viii Wales,t4 certainly, we enter dark and speculative realms. What should not be particularly controversial is the general observation that in prehis­ toric China, India, and Southeast Asia there was a nexus between earth gods, phallic cul ts, and ancestor worship. For example·, the original ele­ ment 11 in the Chinese character tsu meaning "ancestor" is derived from a phallic pictograph, while t 'u, .±. "earth" appears to have been a represen­ tation of a pole or tree branch stuck in the ground as a focus of a fer­ tility ritual.t5 Links between gods, kings, ancestors and a sacrificial ritual appear in the varying meanings attached to the character ti,'* "emperor" in its more modern meaning, according· t_o Fu Ssu-nien' s analysis of sixty-four instances on oracle bones: sometimes the.character refers to a sacrifice to a god, sometimes to royal ancestors as people, sometimes to a cult object representing an ancestor in heaven, and sometimes to a high god.t6 Karlgren's researches point to megalithic cults in China link­ ing an earth god with phallicism conducted on an earth or stone platform, and in later times perhaps on ·a stepped pyramid.t7 Lin Shun-sheng has pointed to the originally phallic significance of the rock in which the eartb-god is focused.t8 W. Eberhard refers .to a phallic god -called tsu (ancestor) in the Me Ti and called "field ancestor" in the Shih Ching.t9 Southeast Asian prehistoric cults in which such associations were made, and correspondences with India and China, are discussed in some depth by P. Mus in.his seminal article which was ·originally published in 1933. 1 O He attempts to reconstr_ uct the structure and religious thought East Asia," Far Eastern.Quarterly (Nov. 1942), pp. 15-30, also published as Data Paper No. 18, Southeast Asia Program, Cornell University (Ithaca, 1956; fourth printing, 1972). . 4H. G. Q. Walets, The Making of Greater India (London, 1961); Prehis­ tory and Religion of Southeast Asia (London, 1957). 5see B. Karlgren, "Some Fecundity Symbols in Ancient China," Bulle­ tin of the Museum of Far Eastern Antiquities, 2 (1930), pp. 1�66. 6 .In Chinese; citedtby D. H. Smith, "Divine Kingship in Ancient China," Numen, IV (1957), p. 189. 7B. Karlgren, "Some Fecundity Symbols," pp. 1-66. 8 .Lin Shun-sheng, "Ancestor Temple and Earth Altar among the Formosan Aborigines," Bullet.in of the Institute of Ethnology, Academia Sinica, 6 (1958), pp. 1-57. 9t w. Eberhard, The Loaal Cultures of South and East China (Leiden, 1968), p. 188. lOP. . Mus, "Cultes.Indiens et Indigenes au Champa," BEFEO, XXXIII (1933), pp. 367-410. English translation as India Seen from the East, Centrefor Southeast Asian Studies, Monash University, Monash Papers on Southeast Asia, No. 3 (Clayton, Vic. , 1975). lX of earth-god cults across monsoon Asia, in which an originally inacces­ sible abstract divinity identified with the earth was made concrete and accessible in a cult object such as a stone: community c'hiefs as priests provided a bridge between cormnunity and divinity, becoming, during each ceremony, further embodiments of the divinity themselves, and th.eir ances­ tors, literally absorbed by the tutelary earth, were divine embodiments also. The stone slabs known as kut in Champa, the focus of Mus's study, represent the survival of the ancient cult stones into historical times under a n,ew aspect;n 11 and, with the importation of Indian religions, it was the Siva Zinga which lent itself especially to ·the service of the old ideas; the anointing of a Zihga, in which sacral fluids pouredn.down into the stone cistern representing the earth, evoked for the Southeast Asians the fertility ritual of their cults of stones or mounds of earth.n12 Dr. Kulke cites below the literature on royal cults in various other parts of Southeast Asia;n13 there are interesting parallels in Java, where rituals were conducted to "open" the eyes of the statues of dead kings, vestiges of ancestor cults are.implied in the practices of placing the statue of a dead king over a receptacle containing his remains, and of placing remains of royal bones in model stone houses on posts at Hindu and Buddhist burial sites. 14 As for India, it is likely that the Zihga similarly served the pur­ poses of prehistoric cults, attested for example in the Harappan civiliza­ tion;n15 the yakEJas, "demons," of later Hindu myth are likely to represent territorial gods of indigenous communities absorbed into the new composite culture,16 and reverence for ancestors, though more thoroughly submerged than in other parts of Asia by the celestially oriented religion of the Aryans, may be detected in the rich complex of myths associated with the pitr:yana, "way of the fathers," in which the underworld of the dead, the idea of reincarnation, and cult s of moon, waters, •Snakes and so forth were bound together.n17 11Ib1·d. , E ng. trans. , pp. 36-43 • 12Ibid. , p. 31. 13see pp. 1-2. 14see W. F. Stutterheim, "Iets over Prae-Hindu:i:stische bijzettings­ gebruiken op Java," MededeeZingen der KoninkZijke NederZandse Akademie van Wetensehappen, n. s. 2, 5 (1939), pp. 105-40. 15see Sir Mortimer Wheeler, The Indus CiviZization, ·3rd ed. ( Cam­ bridge, 1968), p. 10n_ 9. 16see P. Mus, "Cultes Indiens," Eng. trans. , pp. 28 f. 17see S. Bhattacharji, The Indian Theogony (Cambridge, 1970), p. 76. X The prehistoric background may thus shed a little light on the pur­ poses and origins of the cults of lihgas on stepped pyramids at Angkor. What is particularly significant is that they embody the idea of bi­ presence: in itself, a divinity was abstract and unapproachable, but itt. could through ritual acquire also a concrete form in which it could com­ municate with its devotees. Since it was essentially abstract, there was no contradiction if there were two or more concrete forms: the-sacred cult object, the community chief who was thus temporarily and ritually divin- . ized, and his ancestors. Each of these was a partial manifestation of a divinity that was everywhere and nowhere. As with the prehistoric cults, so, perhaps, with the pyramid-studded religious landscape of Angkor. "These considerations apply to the pieces of archaeological evidence, the monuments with their lihgas, that are not central to Dr. Kulke's study: he identifies them in order to distinguish them from the epigraphic material that bears on the devaraja cult properly so called� But the idea of bi-presence without contradiction is relevant to one of his main sug­ gestions--that the devaraja wast·a secondary, portable, image distinct from the various fixed images in permanent shrines. If this suggestion is cor­ rect it finds parallels in the practice of reduplicating embodiments of the primal divinity, as if by mirrors, without implying that the divinity itself was anyt�ing but a single eternal.unity. In China, the god of the soil could be embodied in two tablets, one fixed,· the other transported in procession or carried into war by the commander-in-chief; while in India fixed lihias in temples could be duplicated by processional images, utsavavigraha. 1 · · Such a processional image is involved in the nexust·tbetween ruler, ancestors, soil, and fertility, and means as much and as little for the question of "god-kingship" as the nexus as a whole. Perhaps, if we wish to refine this question, we must adjust ourselves to a vocabulary of ideas different from our own. What can be asserted, negatively, is that there is no warrant· in the archaeological evidence for the conclusion that.the kings of Angkor, individual human beings, were made into gods in their own right. For more specific conclusions, it is to the epigraphic evidence that we must turn, and it is to the epigraphic evidence that the bulk of Dr. Kulke's study is addressed. There is only one really informative inscription, and thatt.tis the Sdok Kak Thom• stele inscription; but as it is a long one, and as it con- tains a Sanskrit section and an Old Khmer section paraphrasing it, it might be thought that the meani�g of the devaraja, both as a term and as an institution, could be derived from the inscription without ambiguity. However, this is not so. The Sanskrit devar,aja can mean either "god-king" or "king of the god,s.t" The Old Khmer paraphrase·is kamrateh jagat ta· raja. The meaning of this in the context is open ·to discussion, a discus­ sion which must be left to those proficient in Old Khmer; my own under­ standing, on the basis of-the opinions �f others, is that the expression 18These examples are given by P. Mus, "Cul tes Indiens," Eng. trans. , p. 47. Xl suffers from much the same ambiguity as devaraja and that the problem can only be solved, if at all, by a thorough study of the contexts.n19 This being the case, there seems to be good warrant f,,.or Dr. Kulke to accept the indications of J. Filliozat's research that Siva was wor- shipped as king of the gods. Further, he points to the distinction be­ tween the expressions kamrateh phdai karom., "lord of the lower plane/ earth," often used to designate a human king, and kamrateh jagat ta rcija., "lord of the world, who is king," arguably applied to an opposite, divine, pole. The earthly ruler was ''protected" by the latter. Dr. Kulke con­ cludesn. unambiguously that "the Sdok Kak Thom inscription offers no warrant at all for the theory that the kings of Angkor were 'god-kings. '"20 If this is so, it is iron.ic that there is less evidence of the iden­ tification of kings with gods in Angkor, the home of the devaraja., than in India, where a division of powers between royal and br-ahmanical orders may in a sense fairly be seen, but where nevertheless authoritative texts tell us that the king is a god in human form,21 that he is a god on earth,22 that the coronation of a king is really the coronation of In­ dra,23 and that even an infant king is not to be despised as he is a god. 24 (As Dr. Kulke remarks in his Introduction, however, such references embody literary theory, not epigraphically attested fact.) These remarks must on no account be taken as an attempt to gild the lily; they are intended only to introduce then.reader to certain aspects of the background to the cults discussed by Dr. Kulke, and to point to the significance of his study against this background. The study is an impor­ tant one, and in bringing a translation of it before the public I must first acknowledge with considerable gratitude the kind interest which the author himself has taken in the project and the great care he has taken in going through drafts of the manuscript. I am also happy to acknowledge the encouragement of Professor 0. W. Wolters, who inspired the project and brought the translation to the Southeast Asia Program for publication, �nd my wife, who has looked at the English style of the translation with fresh eyes and suggested improvements. Further, the present publication benefits considerably from the contribution of Mrs. J. Jacob who, in discussion with the author, has 191 am indebted in this connection to S. Pou, personal communication. 20See below, p. 21Mahabha.rata., 12.68.40. 22Ibid. , 12.59.128 ff. 23Ibid. , 12.67.4. 24M anusmr .• t1,,., 8 . 8 • xii suggested a number of changes to bring the English here more in line with the Khmer terms translated (which is why the reader may notice a few dis­ crepancies between the English and the original German in such contexts), and who has supplied some notes on Khiner terms in the Sdok Kak Thom• stele inscription which appear as an appendix. I.W.M. xiii . XlV INTRODUCTION This study began as part of a broader project on the cult of divine royalty in Southeast Asia, with an emphasis on the ·vais. n. ava apotheosis of Hindu rulers of that region, a rather late development. This idea is well known, for example, by the idealized "portrait sculpture" from Belahan· showing the Javanese king Airlangga (1016"'.'"49) as the god Vi�I)_ U, and by Angkor Vat itself which is supposed to have been dedicated to king Surya­ varman II (1113-ca. 1145) deified as Vis• n• u. The study on monarchy ·.and divinity was extended to Orissa,n. India, where the author as a member of the Orissa Research Project worked on the Vaisnava Jaganna.tha cult and its influence on Hindu kingship ideology in o·rissci.: 1 The medieval kings of Orissa, too, are famous for their alleged divinization as "Moving Vi�IJ.u" (aalant'i Vi�1JU) and as son and viceroy (putra, rauta) of the god Vi�I)U­ Jagannatha. There was an even more direct correspondence between this · part of Eastern India and early twelfthn.century Cambodia. In the same generation, in the early twelfth century, the two largest temples of India and Southeast Asia were constructed and dedicated to Vi��u by rulers whose forefathers and predecessors until then had been strong adherents of the Hindu god Siva (i.e., the Jagannatha-temple in Puri, ca. 1135, by King Co�aganga, and Angkor Vat by Suryavarman II). Although the leaning of kingship ideology towards Vai��avism during these centuries was evident on both sides of the Gulf of Bengal, a criti­ cal analysis of Orissa's rich epigraphical sourees completely disproved the theory that there ever existed anything like a genuine deificati,on of living kings in this part of India, irrespective of Vai�IJ,avism and Saiva- ism. A further and mor.e general study of the epigraphical sources of India showed that this result is quite representative for the whole of. medieval Hindu India. The discussion· of the Hindu kingship ideology as a whole so far seems to have been unintentionally overloaded by interpreta­ tions of texts of legendary origin. which described the ideal kings of the hoary past, neglecting the rich epigraphical evidence of India's histori­ cal kings. With this experience in my mind I returned to my Southeast­ Asian studies, still believing that at least in the Indianized States of Southeast Asia "men were worshipped as gods" (G. Coedes). But the reas­ sessment of the evidence from Angkor whose Devaraja has become a synonym for the divinized "god-king" soon resulted in very similar conclusions. It wa,s not the king of Angkor who was praised as the devaraja, but the god Siva himself as "king of the gods.n" And the .famous temple mountains 1Jagannatha-Kult und Gajapati-Konigtum. Ein Beitrag zur Gesahiahte religioser Legitimation hinduistisaher Konige (Universitat Heidelberg, Schriftenreihe des Siidasien-Instituts, Bd. 23), in press. xv and the royal lingas (named after their founder-kings) were dedicated not to "the king who was the great god of ancient Cambodia" as G. Coedes ,, assumed, but to Siva. Further comparisons with Hindu rituals in India, especially with the well-known institution of worshipping "moving images" (calanti pratima) of sometimes far-away original cult images (mula bera) made it highly probable that the Devaraja of Angkor was nothing else but a bronze image of Siva which was worshipped as a substitute for the original linga which had been established in 802 A. D . when Jayavarman II founded the Angkorian empire. As one of its palladia it was carried by the kings of Cambodia to their successive capitals. Besides this "impersonal" state cult there existed the "personal" royal lingas which were again and again newly established by the various kings on top of their famous temple mountains of Angkor. It was their cult which dominated the ideology of kingship in Angkor from the tenth century. Because the idea of the devaraja cult as an all-pervading state cult of Angkor was due mainly to the absence of any distinction between these two separate cults by modern scholars, it is the consequence of this interpretation that the devaraja cult as the cult of the deified living "god-king" never existed. This result corre­ sponds to the findings in India. But in the course of my further research it became evident that there is at least one distinction between the Hindu kingship ideology of India and Cambodia. In their inscriptions, the living Indian Hindu rajas often were venerated in a rather functional comparison with the gods and their cosmic duties and only very rarely were they praised as their earth­ ly representatives. In Cambodia, however, the kings of Angkor went a step further than the Indian kings. From the time of Jayavarman IV (921/28- ,,,,. 941) onwards they claimed in their inscriptions to be part (amsa) of Siva . Their "subtle inner self" (Suks ,,,,. . ma-antara-atman) met with the "royal self" (niyoktra-atman) of the god Siva in the personal royal lihgas on top of the temple mountains erected by the respective kings. Although the idea of the Hindu king's being an am• sa of the god is known in the sacred texts of the Hindus (e.g., in the Bhagavata Puran• a) in connection with the ideal kings of the past, the application of this idea to living kings of Angkor certainly meant a further step towards royal divinization. It is not un­ likely that this more accentuated development of the idea of a communion between kings and gods on the sacred temple mountains of Angkor is the heritage of an autochthonous Southeast Asian mountain cult with a Hindu blending, known since the days of Funan. But again, this idea also proves that even in Angkor Hindu kings thus claimed to be part of the god, not the god himself. As "the subject of monarchy and divinity in Angkor is not new" and "the field of Angkor studies is an arena trodden by giants of scholarship" as I. W. Mabbett rightly puts it, 2 I gratefully acknowledge the support 2I. W. Mabbett, "Devaraja," Journal of Southeast Asian History., X (1969), p. 203. XVl which I found in J. Fil.liozat's and I. W. Mabbett's articles,n 3 both of which had already questioned Coedes' famous interpretations of the devaraja cult (see pp. 4 ff . and notes 17 and 51). Furthermore my thanks are due to Professor H. von Stietencron, Tu.bingen, for his criticisms and numerous comments after he had thoroughly read the .first draft of the article. -After its publication in 1974, I r·eceived several encouraging com­ munications by friends and colleagues and further suggestions. Professor E . H. S. Simmonds traced back aspects of the Thai shadow play to Cambodian influence of the Ayudhya period and concluded that "it is not impossible that there is reference to the concept of the devara.ja.which, as 'far as Thailand is concerned, is derived from the Khmers,"4 an� Professor H. D. Evers recently suggested an·ninfluence of the devara.ja cult on "the idea of a combined Bodhisattva-God-King Role"5 in Ceylon. A very important supplementary.reference I owe to Dr . Friedhelm Hardy, London, who drew my attention to the fact that in South India devaraja was al-so a rare but well-known name of Vi9�u (e . g. , of Varadaraja-pperumai in Kanchipuram). It was precisely in the eleventh century, when th.e name devara.ja was used for the first time in Cambodia in the famous Sdok Kak.Thom• inscription, that Tirukacci Nampi composed a hymn on Varadaraja under the title of "Devaraja9takam."6 Hardy thus fully corroborates J. Filliozat's arguments pointing out that (in the context of the ninth century's South Indian §aivaism) "Devaraja, meaning 'king of gods,' designates Siva himself" and not the ki· ngs of Cambodia. ( See below, pp. 13-14.) I am very much obliged to the Centre of-South East Asian Studies of the School of Oriental and African Studies and its members for giving me, in February 1977, the opportunity·nof a thorough discussion of the problems of the devara.ja cult. · I am particularly grateful to Mrs. J.nM. Jacob. As my former teacher in Old Kroner she took the trouble to read most thor­ oughly the German version and its English translation. In connection with the Kluner text of the Sdok Kak Thom• inscription she suggested a few slight alterations to its translation and wrote some notes on key Khmer expres- sions (see Appendix). These notes represent a few points discussed in this past February by J.M. J. and H.K. at the stage of preparation of the English translation by Dr. I. W. Mabbett. Perhaps the two most important key expressions referred to by her notes are man and daiy. I have argued (pp. 20 ff. ) that the transfer of 3J. Filliozat, "New Researches on the Relations between India and Cambodia," Indiaa, III (1966), pp. 95-106; I. W. Mabbett, "Devaraja.n" 4E. H. S. Simmonds, "New Evidence on Thai Shadow-play Invocations, " BuZZetin of the SahooZ of OrientaZ and African Studies, XXIV (1961), p. 555 (see below, n. 13). . 5H. D. Evers, Monks, FPiests and Peasants: A Study of Buddhism and SoaiaZ Struat'Ur'e in Central Ceylon (Leiden, 1972), p. 66. 6Friedhelm E. Hardy {London), personal communication. • • XV11 the devaraja from Roluos to the newly established capital at Angkor and the consecration of the "personal" royal lihga on the central temple moun­ tain (Bakhen) were two distinct actions by Yasovarman (899-900), which are clearly separated in the Skok Kak Tho1!1 inscription (D, 12) by the word man. I translated it with "dann" in the sense o.f "darauf" (then), which indi­ cates a time interval between the two actions. Mrs. J.n M. Jacob's inter­ pretation of man, however, does not exclude a synchronism. But still there is the fact that the consecration (sthap ana) of the "holy linga at the center" (vrat lihga ay kantal) is described in a separate sentence (D, 13; see below, notes 77 and 81). The expression daiy "various (ly)," "different (ly)" is one of the most crucial of the Khmer portion of the Sdok Kak Tho1!1 inscription. To some extent it depends on its translation in C, 81 whether there existed various devarajas, an interpretation which would lead again to Coedes's theory of an identity between the devaraja and the various "personal" lingas on the temple mountains. But it has been shown that Coedes in his most famous article on the devaraja cult (1952) brought the word daiy in relation with kamraten jagat ta raja (= devaraja) . Therefore he spoke in this article of "les divers rois-dieu:x;," whereas in the translation of the relevant portion of the Sdok Kak Thom• inscription together with P . Dupont, he had written in 1943/46: ·"Le dieu-roi changea de residence suivant les capitales ou le monarque le conduisit" (see below, p.n11). In her note in the Appendix Mrs. J.nM. Jacob shows it to be highly probable that daiy has to be understood as some sort of pluralizer of the capitals (nagara) in which the kings variously stayed. Consequently the kamrateh jagat ta raja should thus be understood as an "idole unique. " Finally, a statement of the article needs some correction. I have written that "after 1052 A.D. , the date of the Sdok Kak Tho1!1 inscription, we hear nothing more of the devaraja cult" (p. 39). However, one inscrip­ tion of the late twelfth century at the Bayon mentions the kamrateh jagat ta raja. 7 In his article on the devaraja, I. W. Mabbett, following G. Coedes, has already mentioned these inscriptions which list obviously the various sculptures which once stood in the chambers of the Bayon temple.n8 G. Coedes concluded from this inscription that the devaraja seems to have been relegated to a chapel in the outside galleries, although the Bayon itself must have been associated with the cult of the king.n9 It is a pity that I had missed this inscription when I wrote the German version of the present article in 1973 and I owe my knowledge of it to I. W. Mabbett's article. This inscription, I am sure, verifies beyond doubt the hypothe­ sis that the famous devaraja of the Angkor period was a single sculpture which was the calanti pratima or substitute of the lihga which had been 7G. Coedes, "La date du Bayon. Appendice. Les inscriptions du Ba.yon," BEFEO, XXVIII (1928), pp. 81-112 (inscriptions, pp . 104-12) .n. 81. W. Mabbett, "Devaraja," p. 209. 9G. Coedes, "La date du Bayon," p. 100. xviii installed by Jayava.rman II on the Mahendragiri in 802t. This_ is corrobo­ rated by the fact that the same inscription which mentions the devaraja lists also the names of other deities whose temples are outside Angkor: e. g., kamratenjagat ahok gargyar and the kamratenj agat sambhupura. 10 These names can refer to nothing else but.-to cal.anti pratimas or substi­ tutes for the deities of Koh Ker and Sambor which had been brought with misc_ellaneous other sculptur�s to the Bayon, which had become the new state temple of Angkor in the late twelfth- century.t· I am most grateful to Dr. I. W. Mabbett who has given so generously of his time and his own great knowledge irithe field of Cambodian studies to translatet-the article. And I am highly indebted to Professor Wolters for his great interest in the article and_ for his consent to publish it in the Cornell University Southeast Asia Data Papers.· H. ·tKulke Heidelberg, May 1977 lOibid. , p. 105, inscription No . 2. XlX xx -· THE DEVARAJA CULT Legitimation and Apotheosis of the Ruler in the Kingdom of Angkort1 I. Preliminary Remarks We would probably. not go far wrong.if we described the devaraaja cult as the most widely discussed .state cult of the Hindu middle ages in South­ east Asia. In Cambodia, the country where it acquired its distinctive hallmark, the "devaraja" has become a synonym for the divinized "god-king.t" In the cult of this divinized kingship, it is believed that we have found the essential "source of ·inspiration for the great monuments of Angkor, "2 and that we see "the fundamentalt·tunifying element in the ancient Khmer society"3 of the ninth to thirteenth centuries A.D. The quest for the origins and ramifications of this cult reveals a web of connections spreading out widely across south, southeast and east Asia. In the voluminous literature concerning the devaraaja cult we find indications of influences stemming from ·rndia,t4 China,t5 Yun�n,6 Chenla,t7 Champa, 8 and Indonesia, 9 and ·tfrom the megalithic culture of th.e whole 1The author wishes to thank the South Asia Institute of Heidelberg University for the provision of a grant for a field trip to Southeast Asia in 1971, and to express his gratitude to his teachers in Cambodian, F. Huffman. (Yale University) and J. Jacob (S.O.tA.S. London), and to Professor H. van Stietencron (Tilbingen), who read a draft of the article and made valuable suggestions. 2coedes, 1952a, 1952b. 3 · o' Sullivan, 1962, p .t 88. 4sastri, 1957; Majumdar, 1963, p. 189; Chatterji, 1964, p. 226; Filliozat, 1966. 5coedes, 1952a, pp. 17 ff. ;· Coedes, 1952b, pp. 52 f. 6Briggs, 1951�·, p. 25; Coedes, 1968, p. 36. 7coedes, 1970, p. 58. 8Bhattacharya, 1961, p. 21·' Briggs, 1951a, pp. 15, 38, 44. 9aroslier, 1960, p. 109; Coedes, 1968, p. 100; Stierlin, 1970, p. 14. 1 2 region as weu.n1 0 Furthermore, with respect to the world mountain motif, the cult has been linked to the ziggurat in the Near East.n1 1 The ramifi­ cations of the devaraja cult in Indonesian1 2 and (especially in the Ayudhya period) Thailandn1 3 have been pointed out. The influences of this origi- nally Saivite cult have been discovered in Cambodia, no less in the Vaishnavite temple of Angkor Wat than in the Mahayana Buddhist temple of the Bayon, where we see the S.,.aivite devaraja cult reincarnated variously in a Visnuraja and a Budciharaja.a14 We also have numerous evidences of the continued efficacy of the influences of this cult down to the present time. Thus, the life-size statue of a standing Buddha in the central (Theravada Buddhist) temple of Wat Preah Koh in Pnom Penh could represent "in reality" the . idealized statue of the Cambodian king Norodom (1859- 1904 A.D. ) , a statue which Coedes considers to be closely associated with the statues of the "Lord of the world" (Kamrateri Jagat) in the temples of Angkor. l s Even Sihanouk's "Buddhist socialism" exhibits belated Buddhist vestiges of the devaraja cult.n1 6 For all that the devaraja cult is accorded a dominant importance in numerous studies of the history of Cambodia and Southeast Asia, this role stands in striking contrast to the meager number of assertionsn1 7 that can be made about the cult with any confidence. For example, whereas it had commonly been confidently assumed that the devaraja-lihga embodied the divine essence of the kings of Angkor, or even represented the kings them­ selves, it was possible for J. Filliozat a few years ago to demonst.,.rate convincingly that it is not the kings of Angkor but the Hindu god Siva (as 10Wales, 1953, p. 167; Wales, 1957, p . 128, note 1. 11Heine-Geldern, 1930, pp. 75 ff. ; Bhattacharya, 1961, pp. 23 f. 12coedes, 1934; Coedes, 1968, pp. 88 f.; Groslier, 1960, pp. 96 f. 13wales, 1931, p. 60 ; Dhani, 1954, p . 171; Coedes,n. 1967, pp. 146 f. ; Bechert, 1967, p. 222 . 14Briggs, 1951b, pp. 237, 246. 15coedes, 1966, p. 33. 16Osborne, 1966, pp. 4 ff.; Bechert, 1967, pp. 253 ff. ; Sarkisyanz, 1969, p. l; Kulke, 1970, p. 336. 17Mabbett, 1969, p. 204: "The devaraja cult is not synonymous with the general practice of king-worship. It is, on the contrary, a specific practice known only from a handful of references -in the inscriptions_. The meanings of these references have to be conjectured from their contexts, and the association that we may make between the cult so named and other apparent evidences of a cult surrounding the king is not explicit in the sources so much as reasoned out by modern scholars." 3 "king of the gods") that is to be understood by the designation devaraja . 1 8 Filliozat's propositions are of great importance for the whole complex of questions about the deification of the ruler in the realm of Indian cul­ ture, because the devaraja serves as a paradigm for the "god-king" in Southeast Asia. However , if the considerations advanced by Filliozat are correct , it follows that one of the propositions essential t.o the previ­ ously accepted theory of the divinization of Angkorian kings as devaraja is disproved .t1 9 It is therefore all the more curious that, so far as I know , no serious attempt has yet been made to ta.ke up Filliozat's "chal­ lenge" and either refute his claim or else drawtthe logical implications for the theory of the dev.araja cult. However, it is not only Filliozat's research data that call for a . renewed examination of earlier theories about the devaraja .cult. In what follows it will become.t apparent that there is a whole series of further propositions which do not stand up to critical examination , or at least must be considered as.not yet established.t2 0 There are Angkorian in­ scriptions from the ninth to the eleventh centuries which have , previously , been adduced as testimony to the significance of the devaraja cult. How­ ever , an examination of these has the effect , on the contrary, of raising questions which finally compel us to forsake the picture we had previously formed of the devaraja cultt. If the propositions to be advanced below should eventually. be confirmed , then indeed the devaraja rite played a central role in the legitimation of thet·newly established regime at the beginning of the ninth century, dt1il"ing the sacral foundation ceremony of the Angkorian kingdomt. However, in ·the course of the following centuries (the tenth and e�eventh) , the significance of the devaraja cult as Cambo­ dian state cult was eclipsed by. the cult of the royal lingas (which is to be distinguishedt.from that of the devaraja ) . These were the lingas which were erected by all great kings up to .tthe eleventh century on the monumen­ tal "stepped pyramids" or "temple mountains" that are so characteristic 18Filliozat , 1966 , p. 103; Mabbett , 1969 , p. 207. 19This conclusion is also important for India in that there too we have no historical proof , in the sense .tof inscriptional evidence, of a genui�e divinization of the king going beyond an apotheosis of the ruler , for example, in respect of his functional proximity to particular gods. "Historically speaking we have , in ancient India , no evidence at all of the actual worship of the. king or of the theoretical advocacy of it" (Varma, 1959, p. 237 ) a. In this connection see also B . B. Mishrat, Polity in the Agni Pura1J,G, (Calcutta , 1965), p. 33 : . "The divinity. of the king is vigorously criticized by Bana [Kadambari , Purvabhaga, para. 107J. Though a brahmana, he calls it an invention of unscrupulous flatterers of the king. t . The minister §ukana.sa informs the prince ChandrapiQ.a that the king is made by �he flatterers to consider himself to be four-armed Vi9I}u·tand three-eyed Siva, and in trying to act ·t1ike them he becomes an object of public ridiculet. " (See also note 50 . ) 20rn this connection seet·also Mabbett , 1969 , p . 204. 4 of the architecture of Angkor. It is this Saivite cult of the royal Zihgas on the temple mountains of Angkor, and not the devaraja, which we reencounter later in the twelfth and thirteenth centuries, in the form of a Vais• navite and Buddhist syncretism, in Angkor Watn- and in the Bayon. • The reason why the striking decline in the importance of the devaraja has not previously been recognized may, perhaps, be attributed to the fact that, hitherto, the royal Zihgas on the temple mountains of Angkor have invariably been seen as "devarajas.a" Hence it must be one of the pre­ eminent concerns of the following discussion to demonstrate that this conception of the identity of the devaraja with the royal Zihgas is not tenable. But this could have far-r�aching implications for the whole legiti­ mating system of Angkor. In relation to the content and ideology of this system, it will in future no longer be possible to treat the devaraja as the kingpin of the royal Zihga cult that was crucial to the legitimation of Angkorian rulers. In the same way, it could no longer be appropriate to characterize the devaraja cult with the help of the inscriptions which describe the disposition of these same Zihgas on the temple mountains. The most important consequence for the question of legitimation in the Angkorian kingdom could, however, be the recognition that there was no cult for the divinization of the king as "god-king," a recognition which could certainly be of significance for the whole complex of problems con­ cerning the Indian influence in Southeast Asian. In connection with the social aspects of this legitimating system it will be necessary to give new thought to the function and significance of .,,. the family of the founda- tion priest Sivakaivalya. This family, among other things on account of its centuries-long monopolistic status in the devaraja cult, has been labeled as the most important priestly family in Southeast Asia an.,,.d as Angkor's "shadow dynasty." This hig. h valuation of the family of Sivakai- valya is derived in part from their position in the devaraja cult and from the supposition that this cult was at the same time the important cult of the royl;U Zihgas on the temple mountains of Angkor. However, if it should turn out that, once the inauguration of the empire had been enacted, the devaraja was only one of various re�alia of royal might in the Angkorian kingdom, can this assessment of the Sivakaivalya family be sustained in its entirety?n2 1 II. Previous Theories Before we turn to the examination of the epigraphic sources, some of the previous theories about the devaraja cult should be briefly out­ lined. Deliberately, no attempt at exhaustiveness is made here, as a 21The answer to this question must wait on a later study. Here it may simply be remarked that the Sdok Kak Thom• inscription reports in only one instance that a member of this family also consecrated a royal Zihga. ( See pp. 30-31 . ) 5 number of recently published·tworks have reviewed the discussion of the deva:raja cult in its several ph. ases. 22 It is George Coedes, without any question, who is to be credited with the most thorough investigation of. the devarcija cult and who produced the contributions which·tweret_ essential for ·the working out of the theory.t23 Since his "Note sur l'a potheose en c·ambodge," which appeared in 1911, the problem of interpreting the ·aevwaja cult runs like a red thread through his monumental work. Even after. his death, there appeared in the R. C. Majumdar Felicitation volume of-1970 his important contribution on Jaya­ varman IV (921/28-941 A.D.) as "le veritable fondateur . du cuite de la royaute divine au Cambodge," in which he substantially enlarged our knowl­ edge of the devarcija cult·. Probably .the best known synthesis of his re­ search findings is to be found in his masterly introduction to the world of Angkor, "Pour mieux. comprendre Angkor."24 In the section on "personal cults"t25 Coedes brings together the results of the researches which he had presented in various articles, ·tand which in the following sections were to be investigated in even greater detail. In the Cambodian cult,,. of the devaraja or "god-king," according to Coedes, one propensity of Saivism came to be more conspicuous.tthan in India,.its country of origin,_ _and developed into the cult of the kingdomt. The "essence" of the reigning king, his "subtle, inner self" (suks• manta- ratman), was located in a Zinga on a temple mountain which stood in the exact center of the capital of the empire and thus in the center of the cosmos. The supernatural Zinga, the phallic symbol of the god Siva, was 22Filliozat, 1966; Mabbett, 1969; de Bourg, 1968/69. 23The best known andt_probably· also the most comprehensive descrip­ tion of the devarcija cult in German is t_o be found in R. Heine-Geldern's "Weltbild und Bauformt· in Siidostasien" (1930), pp. 33-45. Integral to his analysis is the role oft·cosmography as the determinant of planningt. It is to be regretted· that Heine-Geldern completed this work before Stern, Coedes, etc., in the early thirties, made possible the definitive dating of·the most important central temple mountains of Angkor such as the Bakhen and the Bayon.t· There is an excellent summary in German of the theories of Coedes in "Angkor. Die Hauptstadt des alten Kambodscha--ein Abbild des Kosmos," SaeauZwn, VI (1955), pp. 154-165. See also Werner Muller, Die heiZige Stadt· (Stuttgart, 1961), pp. 135-45. For comprehen­ sive accounts in German of the culture of Angkor see B . P. Groslier, 1958 and 1960, and also Sti.erlin, 1970t. 24The version cited below is the English translation (Coedes, 1966). [This is the authort' s note. The English translation is cited in the original. ] On the organization of the devarcija cult see Coedest/Dupont, 1943/1946a, - pp . 65-68. 25coedes, 1966, pp. 22 f. 6 ,,,,. generated directly by Siva and was received by a brahman who passed it on to the first king of the Angkor dynasty. This solemn foundat·ion of the devaraja cult2 6 took place at the beginning of the ninth century under Jayavarman II on Mount Mahendraparvata, 27 the present-day Phnol!l Kulen to the northeast of Angkor. According to Coedes, it is furthe- r known that Indravarman, a successor of Jayavarman, established the lihga Indresvara on the Bakon at Roluos (= Hariharalaya) for the cult of the devaraja , and that shortly before 900 A.D. Yasovarm.an, the founder of Angkor, conse­ crated the lihga Yasodharesvara for the devaraja cult on the first central mount of Angkor, the Bakhen. Then, in the year 921 A.D., King Jayavarman IV moved the devaraja to the temporary new capital Koh Ker, where it was honored on the Prasat Thol!l under the name Kamrateh jagat ta rajya as "Lord of the world who is royalty." After the reversion of the capi.tal to Angkor the devaraja was, at the beginning of the eleventh century, conse­ crated on the �emple mountain Phimeanakas that was built specially for the purpose in the Angkorian palace precincts. Towards the end of the eleventh century, Udayadityavarman constructed the Baphuon in order to embody the "inner self" of the king in a golden lihga. Then, towards the end of the twelfth century, after the victory of Buddhism, the devaraja "forsook" the Saivite lihga and was worshipped at the Bayon asn- "Jayabud­ d.ha," in a statue of the Buddha--a statue in which Jayavarman VII, the builder of the Bayon, was divinized.28 "From all this evidence it is safe 26 · · This ceremony on Mahendraparvata seems to have been preceded, in the injunction of Jayavarman II, byn-an "auspicious magic rite" (kalyana­ siddhi) in the south of Cambodia, at Ba Phnom, "desti�ee a empecher que le Kambujadesa ne put etre pris par Java" ["intended to prevent Cambodia from being annexed by Java"J. Inscription of Vat Samron (K.956), lines 15-16 ( IC, VII, p. 133). In this connection see also Jacques, 1972, p. 212, and Wolters, 1973, p. 22. 27The chronology of Jayavarman II's reign most generally accepted at the time, 802-850 A.D., has been a subject of continuing keen dispute (on this see Majumdar, 1943, pp. 52 ff.; Briggs, 1951a, p. 81 ; Dupont,n_ 1952, pp. 157 ff.). Very recently there have appeared almost at the same time two more articles which review the whole range of problems once again (Jacques, 1972; Wolters, 1973). The most difficult problem is the ques­ tion whether the year 802 A.D. represents the date of Jayavarman's coro­ nation or that of a later consecration of the devaraja cult on Mahendra­ parvata. 28Elsewhere, Coedes (1952b, p. 51) lists altogether thirteen royal temple mountains: Prasat Ak Yol!l, Krus Prah Aram Ron Chen, Bakori, Bakhen, Prasat Thol!l in Kol}nn_ Ker, Baksei Cha.mkron, the Eastern · Mebon, Pre Rup, Ta Keo, Phimeanakas, Baphuon, Angkor Wat, Bayon: "Thus of the thirteen tem­ ples enumerated six were certainly dedicated to the royal lihga between the ninth and eleventh century, a seventh, Angkor Vat, became a mausoleum of its founder, and, finally, the last contained a Buddhist image of which the name recalled that of Mayavarman VII. The association of this par­ ticular architectural type, the pyramid, with royalty is therefore cer­ tain.n" 7 say that it was the king who was the great god of ancient Cambodia, the one to whom the - biggest groups of monuments and all the temples in thet· form of mountains were dedicated.t"29 Thus, according to Coedes, the cult of the devar.aja was founded. by Jayavarman II on Mount Mahendra and continued by h.is successors through the consecration of new Zingas,a3 0 which commonly bore the names of their respective founding kings. So for Coedes the devaraja, the imperial pal­ ladium of Angkor, is identical with the current : personal Zingas on thet· temple mountains. On the other hand, as Coedes says in other places,t3 1 it remains unclear whether the devar.aja-Zihga of these temple mountains was regularly one and the same cult object or whether on the accession of a new king a new cult object was regularly consecrated. According to Coedes , moreover , the "inner self" of the kings of Angkor was located in the deva:Paja-Zinga on the temple mountains of Angkor. t . To these kings,·tas "god-kings" (deva-raja) were the great temples of Angkor dedicated. In the abovementioned, posthumously published, article on Jayavarman IV (921/28-941 A.D .) as "the veritable founder of the cult of divine king­ ship in Angkor ," Coedes brings together a series of arguments which he. had already advanced in various places.t32 They represent an important 29coedes, 1966, p. 31 (emphasis added). Here , .Coedes sees the kings as the actual gods of Angkor, whereas in other contexts his argument is more qualified. ·rt is th·erefore regrettable that this particular passage, whose thesis (as we are attempting to show) is untenaole , is all· too fre­ quently the authority followed by general works on Southeast Asian his­ tory. In this category can perhaps be included Le Thanh Koi (1967, p. 36)t, "Letroi etait le Dieu a qui de son vivant le temple etait dedie" [The king was the god to whom in his lifetime the temple was dedicated"]; and Hall (1966, p. 99)t, "He himself [the .king as devarajci] was the god to whom in his own lifetime the temple was ·dedicated." Stierlin (1970 ,. . p. · 22) even goes a step fyrther when he writes, "As he [the· ruler o-f Angkor] was the king of the gods, he was also the king of men.t" 30coedes, 1952b , p. 52: 11Thetroyal essence was identified for each reign with the subtle ego of the reignin_ g ki_ng , and . . . the devaraja, unique when considered as a philosophical and religious conception imply� ing the existence of an image of the abstract king, was in reality multi­ ple, each reign having its own . '' . 31coedes, 1968 , p . 119: "We do not know whether this Cdevar.aja'J Zinga that contained the 'royal essence ,' the 'moi subtilt' of the king remained the same iinga throughout the successive reigns or whether, on the other hand, each·tof the various Zingas consecrated by the kings upon their accession and bearing their names (Indresvara, Yasodharesvara , Rajendresvara) was in turn the Devaraja �" 32coedes , 1931a; 1952a/b; 1961; and IC, I (1937), pp ; 7o ·f. 8 supplement to the earlier discussion. Starting from the fact that the Zihgas on the central temple mountains of the ninth century are known only by the names of their builders (Indresvara on the Bakon at Roluos and Yasod.haresvara on the Bakhen), Coedes expresses legitimate doubt whether the royal Zihgas of the ninth century were yet called devaraja (or, in Khmer, kamrateh jagat ta raja ) . All that is so far certain is that both expressions occur for the first time in the famous inscription of Sdok Kak Thom• in the year 1052 A.D. However, as the kings Bhavavarman, Isana- varman and Pu�karak�a in the pr�-Angkorian period of Chenla (sixth to ninth centuries) had previously erected Zihgas under their own names (Bhaves. vara, Isanesvara and Pus. karesvara), this would necessarily leave as the sole demonstrable innovation of the Cambodian rulers of the ninth century the idea of consecrating the royal Zihga on a stepped pyramid in the center of their capitals. According to Coedes, the decisive step to the divinization of king and the kingship in the devaraja cult presumably did not happen until Jayavarman IV. As a usurper, he transferred his capital to Chok Gargyar, the present-day Ko� Ker, about eighty kilometers to the northeast of Angkor. There he constructed the Prasat Thom• , the highest Cambodian tern- ple mountain built to that time. However, he did not name the Zihga of this temple after himself as his predecessors had done, but instead he dedicated it,, to Tribhuvanesvara, the "lord of the three worlds," a famil- iar name of Siva. The Khmer inscriptions at KoQ Ker praise the god, first of all, as "the divine lord who is lordship" (vra'JJ. kamrateh jagat ta rajya).a3 3 In his discussions Coedes accords special significance to the Sanskrit expression rajya (kingdom, royalty, lordship) in the Khmer name of the Zihga. In contrast to the Khmer name of the devaraja in the later Sdok Kak Thom inscription, kamrateh jagat ta raja ("the lord of the world, who is king" }, here at KolJ Ker it is called kamrateh jagat ta rajya ("lord of the world, who is lordship"). According to Coedes, the rule of Jaya­ varman IV is thus divinized, with the name Tribhuvanesvara, as "lord of the three worlds. " While Jayavarman II, as cak.ravartin at the beginning of the ninth century, aspired after unlimited dominion over the earth, Jayavarman IV a century later went a step further: "Sa fondation du linga Tribhuvanesvara, 'Seigneur des trois mondes' qui est le rajya , doit lui assurer une sorte de souverainete cosmique, identifiant la royaute khmere a la maitrise des trois mondes qui constituent l ' univers. "34 Coedes then sets out which successors of Jayavarman IV followed his innovation in their inscriptions, and, similarly, which kings again resumed the tradi­ tion of his predecessors and, in the Zihgas of their temple mountains, conjoined their own names to that of Siva (Isvara). 33This name appears with several variations :n. vra'JJ. kamrateh an ta rajya; vrah kamrateh an jagat ta rajya and vra� kamrateh jagat ta rajya ; vrah is here equivalent to "holy, exalted," and as substantive signifi­ cantly "temple," "king ," and "god" ; kamrateh = Lord ; jagat Sanskrit)n= "world"; afta= "my"; taa= relative pronoun; rajya (Sanskrit)n= "kingship," "kingdom," "sovereignty.n" 34 Coede,s, 1970, p. 60. 9 In consideration of the fact that the name kamraten jagat ta rajya, which was to reappear in the Sdok Kak Thom• inscription in the form kamraten jagat ta raja, is first documented in the inscriptions of Jayavarman IV, we must agree with Coedes that Jayavarman IV is the veritable founder of the divinization of kingship in Cambodia.n35 · It will be shown below, how­ ever, that the transformation under Jayavarman IV could have taken a basically different course, as Coedes appreciated.n36 Coedes developed his views. about the devaraja cult over the course of almost sixty years. During this time, the rerlections of his col- · leagues in the Ecole Francaise d'Extreme-Orient also drew him ihto ear­ nest debate.n37 In this connection the studies of L. Finot and Ph. Stern should particularly be mentioned; they advanced, against Coedes, the con­ tention that the devaraja was a· "'roi abstrait" (Finot), or, again, "plus un rituel qu'un symbole materialise" ( Stern) . The discussion was pursued by Coedes with almost the same involvement as his famous debate with J. Przyluski over the question whether Angkor Wat was a vai��avite temple or a mausoleum for its royal builder Suryavarman II (1113-ca. 1150 A.D.).n3 8 However, whereas the latter discussion was more or less settled by an agreement that Angkor Wat was a vai�I)avite mausoleum for Suryavarman II deified as Paramavi�l).uloka, the discussion between Coedes, Stern and Finot led to no definitive explanation of the nature of the devaraja cult. The reason for this may lie, as will be shown in the following sections, in the conflicting interpretations of the epigraphic data, especially those 35The expression devarajya appears, certainly ; as early as in an inscription of King Indravarman ( see note 134). The discussion of the divinization of the king is here deliberately confined to the epigraphi­ cally attested royal cult in Hindu Angkor. The question of royal cults in Southeast Asia in megalithic times ; in which for example Quaritch Wales sees the essential foundations of the devaraja cult, is not taken into consideration here. "It is when we compare the peculiarities of the Khmer religion with the Older Megalithic beliefs in South-East Asia, that the similarity of concepts is so striking that there seems little doubt as to the direction in which to look for the cause of the distinctive Khmer traits'.' ( H. G. Quaritch Wales, 1961, p. 128). 36see below, pp. 38 ff. 37summaries of this discussion are . to be found in Coedes, 1952a, and de Bourg, 1968/69. 38 J . Przyluskin, "Pradak�ir:a et. Prasavya en Indochine,n" in Winternitz­ Festschrift ( Leipzig, 1933), pp. 326-32; idem, "Is Angkor Vat· a Temple or a Tomb,n" JISOA, V ( 1937), pp. 131-44; G. Coedes, "Angkor Vat, temple ou tombe,n" BEFEO, XXXIII (1933), pp. 303-9; idem, "Etudes cambodgiennes XXXIII : La destination funeraire des grands monuments Khmers,n" BEFEO, XL (1940), pp. 315-43; idem, "Le grands monuments d'Angkor, sont-ils des temples ou des tombeaux? " Cahiers de l 'Ecole Francaise d 'Extreme-Orient, XXVI (1941), pp. 26-29. 10 of the Sdok Kak Thom inscription. Today, Coedes's contentions about the devaraja cult are generally accepted, and those of his opponents are men- tioned almost only in footnotes,19 but the reputation which Coedes's life work has justly earned should not be allowed to close the door to a re­ examination of his arguments. Finot writes of the devaraja cult : ffCe roi, ainsi eleve aux honneurs de l'apotheose, n'est pas le roi fondateur, car dans ce cas, l'idole eut reGu, selon la regle, son nom, et elle fut restee dans son temple sans suivre le roi regnant dans toutes ses residences. Le devaraja est le roi abstrait, dans sa nature surhumaine, l'essencen_nroyale confondue avec l'essence divine sous l'apparence du Zihga. C'est pourquoi il accompagne partout le roi regnant qui est comme l'emanation changeante de sa sub­ stance imm.uable et re saurait se separer de lui."40 Hence, Finot seems to distinguish between the Zihga of the king (such as Indresvara) and the devaraja. The former, in obedience to the rules of Hinduism, was a fix­ ture in a temple and thus could not follow the king to his various resi­ dences. The devaraja, on the other hand, was the "abstract king," which was united, as the ruler's supernatural essence, with the divine in a Zihr;p. . In this immaterial form the devaraja followed the kings of Cambo­ dia to their various capitalsn. According to Coedes, Finot was thus defending the idea of an "idole unique," against which he maintained that previously the search in Angkor for a temple which could have been the shrine of a "permanent"·n devaraja sculpture had been in vain. As evidence in refutation of the idea of the "idole unique" C9edes then adduces the Tribhuvanesvara-Zinga, which for him was the devaraja of king Jayavarman IV in Kol]. Ker. It would be highly improbable that the potent Tribhuvanesvara Zihga should have been "carried off" by Jayavarman IV from Angkor to Ko� Ker, only to be subsequently transported back to Angkor again in the reigns of his successors. On the contrary, according to Coedes, the Zihga of Kol]. Ker is the pre-eminent example of a royal Zihga that was consecrated for a particular king �nd for a particular templen. So, if inscriptions mention kings taking the devaraja with them when they change capitals, this must in Coedes's view be understood as referring to the cult and ritual, but not as denoting the devaraja sculpture.n4 1 Coedes then seeks. to trace the whole idea of 39Exceptions in this respect are de Bourg, 1968/69, and Sahai, 1970, pp. 41 ff. 40cited by Coedes, 1952a, p. 12. 41coedes, 1952a, p. 13: ·"Et quand les texte s nous dis ent que tel roi changeant de capitale emmena avec lui le devax>aja il faut evidemment comprendre qu'il en transfera le culte, le rituel, et non pas l'dole elle­ meme.n" [ "And when the texts tell us that such�and-suc·h a king, in chang­ ing capital, took with him the devaraja, it is evidently necessary to understand that he transferred the cult, the ritual, and not the idol it­ self. "] Here Coedes comes very close in his reasoning to Stern ; 1934.n· 11 the "idole unique" back to a misinterpretation bf the Sdok Kak Thom• in- scription, "qui, loin de parler comme on l'a cru d'une ·nseule ·et meme image 'residant dans toute capitale ou les rois le conduisirent, ' dit textue+le­ ment: 'Quell- e que soi t la capitale OU les rois sont all es resider, les ' divers rois-dieux y ont .ente emmenes aussi.'" 42 We must agree with Coedes that this interpretation rules out the notion of an "idole unique." Still, the question remains open whether the interpretation of the text which Coedes offers here in his debate with Finot is tenable. To begin with, it is enough here to point to the striking �act that Coedes ( in collaboration with Dupont) translates the corresponding part of the Sdok Kak Thoip inscription in exactly the sense understood by Finot: "Le dieu-roi changea de residence suivant les· capitales ou le monarque le conduisit. . • . "4 3 . In 1934 Philippe Stern put together his views on the devaraja cult in his important article, "Le temple-montagne khmern. Le culte du linga et le devarajaa. " After enumerating the great temples of Angkor and ar­ ranging them chronologically, Ste_rn countsn· up the Angkorian temple lihgas whose names are known to us. We have the Indresvara of the Bakon, the Yasodharesvara of the Bakhen, the Rajendresvara of the East Mebon, and the Rajendrabhadresvara of Pre Rup. Further, the lihga of Baphuon is .wor­ shipped as the "golden lihga" (suVarrJ;a�Zihga). Stern then poses the deci­ sive question of the relationships between.these Zihgas known by_ name and the devaraja , "ce dieu-roi sous forme du Zihgaa, symbole de la royaute, qui residait dans cha.que capitale?" He ·seeks to supply the answer in the example of the Yasodharesvara-Zihga on the Bakhen: Stern believes it can be established from the Sdok Kak Tho:qi inscription that, when Yasodharapura ( = Angkor) was founded, the Yasodharesvara-Zihga was consecrated on the Bakhen in the center of Angkor as the devaraja of Yasovarman.44 However, he warns against the "simple and logical hypothesis" that the erection of every temple mountain in Angkor coincides- with the accession of a new king and the dedication of a new lihga as devaraja. "The. reality is,· on the contrary, more flexible and less logical . " As evidence against t- his same "simple and logical hypothesis," Stern adduces the Eastn-Mebon and Pre Rup, both of which were built by Rajendravarman II ( 944-968). These examples 42coedes, 1952a, p. 14 ( emphasis added). ["Which, far from speak­ ing, as has been thought, about a single unique image 'residing in each capital to which the kings conducted it," says, in the context, 'tq what­ ever capital the kings went to take up residence, the various king-gods were taken thither also.n'"J 43coedes et Pupont, 1943/46a, p. 110 ( emphasis added). ["The god­ king changed its residence according as the monarch conducted it to dif­ ferent c- apitals. . . ." J 44stern, 1934, p. 6�3. It will be shown below ( p. 33) that this identification of the Yasodharesvara with the devaraja in the Sdok Kak Tho:qi inscription is, specifically, not admissible. 12 demonstrate that a king could have built, one after the other, tu.Jo temple mountains for two lingas named after him, Rajendresvara (952 A.D.) and Rajendrabhadresvara (961 A . D.). These two temple mountains of one and the same king contradict the proposition that every king built one temple for his royal devaraja-li-hga.a4 5 The difficulties which are bound to arise if one attempts to demonstrate the connections, and also the distinctions, between the devaraja and the li-hgas of the Angkorian temple mountains can only be obviated, according to Stern, by the proposition that the devaraja was a "movable cult.n" This cult, says Stern, could be celebrated as a rule (among other occasions) when a new royal temple was consecrated. Hence Stern is inclined to see in the devaraja rather "a rite than a mate­ rial symbol . "4 6 As confirmation of this proposition, Stern adduces the data of the Sdok Kak Tho� inscription, which indeed bring the ritual (siddhividya, vidhi, etc. ) of the devaraja into the foreground. Hence Stern comes to the conclusion that, in the central sanctuaries on the tem­ ple mountains of Angkor, there was invariably (or commonly) a linga which invariably (or commonly) united the name of the reigning monarch with Isvara, the name of Siva. In contrast to these royal lingas, however, the devaraja seems to have been the ritual of a cult rather than the cult object of a particular cult in a particular temple. This devaraja cult could have been celebrated "around" the royal linga. In Stern ' s view this interpretatiJn would explain how and why the cult was celebrated in vari­ ous temples.n47 On the views of Finot and Stern it may be said by way of summary that both distinguish at least partially between the devaraja and the per­ sonal king-lihgasa. Finot sees in the devaraja the abstract royal essence ("essence royal"), which existed as it were independently of the currently reigning king and independently of his personal Zingas established on the temple mountains. The suprapersonal royal essence was united with the god Siva in a Zinga that was passed down from king to king through the agency of brahmansn. Stern, on the other hand, appears to have in mind a devaraja ritual independent of the personal lihgas of kings and without any definite temple. This ritual was celebrated in various places, and this could include the temple mountains of the personal lingas belonging to the kings of Angkor. But Finot and Stern both studiously refrain from 45coedes (1952a, p. 51) included the Mebon and the Pre Rup in a list of the royal temple mountains of Cambodia. It is not clear why, in a sim­ ilar list (1955, p . 158) , he did not mention the Mebon temple mountain. 46Stern, 1934, p. 614. 47stern, 1934, p. 615: "Quant au devaraja, il parait etre un rituel celebre autour d'un Zihga au nom royal, plutot qu'un Zinga determine ayant son temple particulier. S'il en est ainsi, ce culte a pu etre celebre dans des temples differents." ["As for the devaraja, it seems to have been a ritual celebrated in the royal name before a lihga, rather than a specific linga with its own unique temple. If such is the case, this cult could have been celebrated in different temples. "] 13 indicating , even by so much as a hint, what this nabstract" cult--a cult without definite cult object and without permanent temple---could have looked like. This uncertainty impelled both scholars to revert , in par­ ticular· cases , to the identification of the devaraja with the personal Zihgas of the kings; in so doing they fell back into the circular argument from which they had originally wished to escape. The most important contribution of recent years , whose implications reveal new avenues . to the explanation of the devaraja cult, comes, as was mentioned above , from J. Filliozat. On the basis of a series of data from the Tiruvasagam of the South Indian Saivite sacred Ma1:ikka-vasagar (which Filliozat dates to the ninth century), he is enabled to demonstrate � that in South India Siva was worshipped as "king of the gods" on Mount Mahendra: 48 "So it is unnecessary to suppose pe. culiar Khmer ideas of the king as god to understand how the devaraja was established on the Mahen­ draparvata ;n the shape of a Zinga. Devaraja, meaning 'king of gods ', designates Siva himself, normally represented by a ZiYJ,ga and established on the verr mountain referred to by Mal:}ikkavacakar."49 After taking account of the custom, familiar throughout India, of naming the Zihga of a king with his name in combination with Isvara (= Siva) ,n5 ° Filliozat 48"rf we now refer to the Tamil Saivite literature of the time when Jayavarman II first established the devaraja on the Mahendraparvata . . . we meet with a very simple explanation of all the facts" (Filliozat , 1966 , p. 101). As an example , Filliozat cites fromnthe Tiruvasagam, inter alia, "o God of the gods themselves" (XXVIII , 9 ) , or "OnKing of those who are above" (XXVII , 7 ) . "Taking as his abode the great mountain Mahendra where he established his seat and his glory, there he granted the grace of mani­ festing the tradition (agamam)" ( II_, 8-10 ) . The historically most famil- . . � � iar example in India of the worship of Siva on a Mount Mahendra is Siva Gokar1:asva.min (or, Gokar1:esvara) on the Mahendraparvata (Mahendracala) as state divinity of the eastern Gangas of the Kalinga kingdom, from about the sixth to the twelfth centuries (see, for example, the Ponnunturu in­ scription of Samantavarman from the year 64 of the Ganga era, whose begin­ ning is usually dated to 496 A.D. S. N. Rajaguru , Inscriptions of Orissa, II [Bhubaneswar 1960 ] , p. 10 ) . 49Filliozat, 1966, p . 102 (emphasis added). 50"so in Cambodia, an Indresvara, for example, was the Lord of King � Indravarman , that is, simply Siva, the Lord of the Universe, as worshipped by this king in a peculiar place." Filliozat, 1966, p. 102 . In this con­ nection see also J. Duncan M . Derrett, The Hoysafas, a Medieval Royal Family (Madras, 1957 ) , p. 223 : . "Without a word of�explanation it might be supposed that if the Hoysal. esvara , or Linga of Siva dedicated by the . Hoys94a, or in the name of the Hoysa1a, were the product of a figment of the king's own brain , it would be tantamount to an admission that the medieval India was a 'divine king ' a. This, of course, was not the case . The custom of naming a Linga either after the person who had the temple built and the consecration performed or a nominee of his . . . was indeed 14 comes to the conclusion that "the Khmer expression kamrateh jagat ta raja means simply 'the lord of the World (jagadisvara ) who is King,' not 'who is the [Khmer] king,' having performed the Zihgasthapanaa. "5a1 This brief review of various theories about the devaraja cult has served not only to demonstrate in outline the development of the royal cults in Angkor, but also to point to a few of the controversial questions about the nature of the devaraja cult, to which we shall now address our­ selvesn. The most important problem is certainly the ,,question raised by Filliozat: who was venerated in Angkor as devaraja , Siva or the Cambodian kings? On the basis of South Indian sources, Fil,,liozat was able to show convincingly that in the devaraja the Hindu god Siva was venerated ; but he stopped short of going on to make a more thorough examination of the Cambodian epigraphic sources. We shall revert more than once to the ques­ tion how far Filliozat's thesis is valid for Angkor as well. III. The Sdok Kak Tho� Inscription and the Devaraja Cult The most important source for the devaraja cult is rightly consid­ ered to be the famous inscription of Sdok Kak Tho� from the year 1052 A.D. In it the brahman Sadasiva recounts the history of his family, which pos­ sessed an uncontested monopoly in the discharge of priestly office per­ taining to the devaraja cult. The part played· by the family of �ivakai­ valya in the devaraja cult parallels the role which the Sdok Kak Tho� inscription plays in our knowledge· of the devaraja cult. Without the data from this inscription about the foundation of the cult, and the precise enumeration of the various priestly functions of the seven forebears of Sadasiva, it would not be possible to fit such isolated references to the devaraja cult as occur in other inscriptions into any broader pattern. The inscription, which has been edited and translated several times,n52 consists of a Sanskrit part and a Khmer part. About the devaraja ctilt, not unconnected with vanity, since by this means the donor achieved a kind of immortality." It is obviously one of the gravest errors in the discus­ sion of the devaraja cult, and the question of divinization of rulers in general, to perceive in the attempt to achieve salvation in a particular god (perhaps by the erection of a statue of oneself in the likeness of the god, a "portrait sculpture"n) a process of divinization of the one striving after salvation . 51Filliozat, 1966, p . 103 (emphasis added ) ; Mabbett, 1969, p. 105: "Neither this passage, nor any other, says that Jayavarman is a devaraja. The term appears as the name of a rite, and should be thought of as suchn. The occurrence of the term alone is not enough to show that anybody iden­ tified a king with a god .11 52F. Aymonier, 1901; Finot, 191 5 ; Coedes/Dupont, 1943/46a. Where not otherwise specified, the edition of Coedes and Dupont is citedn. 15 · the Kluner version is substantially more informative. In it is recounted how King Jayavarman II came from ,Java 5 3 to Indrapura in Cambodia. · There Jayavarman nominated the priest Sivakaivalya as his teacher (guru) and court chaplain (ro.japur_ohi ta) . After two shifts of capital., to Harihara­ laya and Amarendrapura, it was established upon Mahendraparvata (C , 70). "At that time54 there came a brahman named HiraQyadama from Janapada, a savant versed in magical science (siddhi vidyo.)a. He was invited by His Highness, Paramesvara [posthumous name· of JayavarmanJ, ih order to conduct a ceremony (vidhi) which should prevent.this land of Kambuja from ever being dependent (o.yatta) on Java, and to bring abo.ut CinsteadJ 5 5 that there should be only one single 'Lord of the lower earth' [= King ; Khmer: kamraten phdai karo�J, who would be Cakravartin [universal lo�dJ. This brahman conducted the ceremony in accordance with the Vinasikha. He con­ secrated (prati[!t;ha.) the lordn·of the World, who is king Ckamraten jagat ta raja = Skt : DevarajaJ (C, 71-74 ) . This brahman taught the holy Vina­ sikha, Nayottara, SaJilIIlOha and Sirascheda [Tantras? J,n56 all of which5 7 he recited fromn- beginning , to end, in order t- o have them written down and . , to teach them to Sten afi Sivakaivalya. He gave instruction to Sten afi Siva- kaivalya, so that the latter could perform the ritual (vidhi) in the presence of (na) the kamraten jagat· ta raja (C, 74-76 ) . His Royal High­ ness Paramesvara and the brahman HiraQyadama granted a conces�ion and swore an oath, ordaining that the family. line of the Sten afi Sivakaivalya should officiate in the presence of5 8 the· kamraten,jagat ta raja, forbid­ ding that other people _should offic iat�. Sten aff Si vakai valya, the pUPohita, appointed his whole family to the service of the ritual (C, 76-78 ) . Then- His Royal Highness Para.mesvara, the king, went back again to be ruler59 in_ the royal city of Hariharalaya. His.Highness the 53 .manavra� pada Paramesvara mok an;ivi Java pi kurun ni �u nagara Indrapura (C, 61 ) . According to the recent researches of Professor Boe­ chari, cited by O. W. Wolters as an authority, ,it is not to be ruled out that the expression "bhumi Java" in then·nearly Srivijaya inscription from Kota Kapur ( 686 A.D.) r�fers to the extreme south of Sumatra. Wolters, 1973 , p. 22, note 8 . 54or "In that a brahman came. . . . " See Appendix under man. 55see Appendix under le�. 56vrahmar:a not thve vidhi toy vrat Vinasikha. prati�t;ha kamraten 4agat ta raja. vrahmar:a not paryyan vra� Vinasikha. Nayottara. SaJnmoha. Siraecheda (C, 73-74 ) . The substantive prati�pha , like sthapana. , is used wit·hout a verb. On these texts, see note 74 below. 57see Appendix under man and syan. 58see Appendix under na. 59see Appendix unde·r kurun. 16 kamraten an ta raja was conducted ( nam 60 • ) back also. Sivakaivalya, to- gether with his whole family, ,, officiated according to the rules ( C, 78-79). Sten afi Sivakaivalya died during this reign [of JayavarrnanJ. His Royal Highness Para.mesvara went to heaven while [residing] in the city of Hariharalaya (C, 80). The Kamraten jagat ta raja moved from place to place,6 1 accompanying the king to the respective capital cities, in order to protect ( ccvrz ) the rule ( rajya) of future kings ( kamrateh phdai karorri ) (C, 80-82). Here follows a short description of the activity of the priest Suk�mavindu, who was the purohita of the devarajaa, and successor to Siva­ kaivalya, under King Jayavarman III. With regard to the devaraja cult we are told merely: "During the �eign of His Royal Highness Vi��uloka [Jayavarman IIIJ, the kamraten jagat ta raja resided in Hariharalaya." Sten afi Su.ktJmavindu "was purohita of the kamrateh jagat ta raja.a· The whole family officiated in the presence of the kamraten jagat ta raja." (C, 82-84) The following sixty-four lines of the inscription's Khmer text re­ c,,ount, in practically the same terms, the careers of the successors of Sivakaivalya who celebrated the official ritual of the devaraja under the various kings of Angkor, up to the priest Sadasiva in Udayadityavarrnan II's reign ( 1050-1066 A.D. ).n62 Additional information about the devaraja in the Sdok Kak Tho� inscription is to be obta�ned only from the time of Angkor ' s foundation during Yasovarman I's reign ( 889-ca. 910 A.D. ) , and from the time of Jayavarman IV ( 921/28-941), who temporarily moved the capital to Chok Gargyar (Ko� Ker), north of Angkor. Under Yasovarman, Vamasiva was the tutor of the king, and his "whole family officiated in accordance with the rules in the presence of the kamrateh jagat ta raja. _nWhen6 3 His Royal Highnes s Paramasivaloka [Yaso­ varman IJ founded the city of Sri Yasodharapura he transferred the kamrateh jagat ta raja from Hariharalaya to place it in this city6 4 (D, 11-12) . 6 5 His Royal Highness Paramasivaloka then constructed ( sthapana ) the central 60see Appendix under nam• . 61see Appendix under daiy , and Introduction, p. xviii. 62Even if Sadasiva's religious functions were reduced under Surya­ varrnan ( Briggs, 1951b, pp. 242 f. and 1952, pp. 180 ff. ), there is still no question of any weakening of Sadasiva and his family. The service of the deva.raja was carried out by his family both under Suryavarrnan and under the latter's successor Udayadityavarman (D, 44; D, 64)n� 63see Appendix under man. 64see Appendix under na'[/ • 65Here I follow the translation of Finot ( 1915, p. 89, lines 12 f. ) . 17 ,,,. (temple- ) mountain (vnam kanta l ) . a66 The lord of the Sivasrama [ ::; Va.masivaJ erected the sacred lihga in the center.t" (D, 12-13)6 7 Concerning Jayavarman Iv,t·twe learn of "the reign of His Royal High­ �ess Para.masivapada 'c_Jay_avarman IVJ; then His Highness left the city of Sri Yasodharapura in order totbe ruler ·a(kur-uh) in Chok Gargyar. He took with him also the kamrateh jagat ta rajaa· [thither]." (D, 31-32) Under Har�avarman II ( 941-944 A.D.t), the successor of Jayavarman, the kamrateh jagat ta raja continued to reside in Chok Gargyar. Not until the time of his successor, Rajendravarman II ( 944-968 A,,,..D. ), did the capital move . back to Angkor. ''Then His Royal Highness Sivaloka (Rajendravarman II) returned again to be ruler in the city of Sri Yasodharapura, and he con­ ducted the kamrateh jagat ta ·raja back with him. The whole family offi­ ciated in the presence of the k(1Jrlrateh jagat ta raja in accordance with the rules.t" ( D, 36-37) Let us now glance briefly at the Sanskrit part of the inscription, which recounts the solemn inauguration of the d,,,.evaraja cult. Here it is said that Hira�yadama taught the sastra texts Sirascheda, Vinasikha, Scuzrmoha, and Nayottara, "the· four faces of Tumburut." "When this brahman, full of zeal, employing his knowledge and experience in occult science, had brought together the essence of the sastras, then, for the increase of the well-being of the earth, he performed the success-ensuring (ritual ) called devaraja.a"6 8 Besides the importantt- designation of the devaraja ritual, the Sanskrit version of the Sdok Kak Tho:rp. inscription contains little concrete evidence. The name devaraja crops up only twice more at the end of this family chroniclet. There it is said that Sivacarya, the . ,,,. penultimate member of the l.ine of priests from the family of Sivakaivalya, "offered worship (area ) ·tdaily, _tfull of zeal and excluding other priests" to the devaraja (B, 34 ),t. and that Sada.siva, the· la. st chief of the Siva·­t kai valyas, honored the devaraja ( B ,' 37) . It is significant for our further consideration that, in connection with the founding of Yasodharapura, the consecration of the lihga upon the Yasodharagiri (Bakheri ) is especially stressed, as it.is also in the Khiner 66Bakhen = Yasodharagirit. 67In later inscriptions this lihga of Yasovarman is called "Sri Yasodharesvara"; for example the Bakhen inscriptions K. 464 and K. 558, line 6 (Jacquest, 1970, p .t.65). . .68s,,,.a-stram• s,,,. ·i ras" ch eda . -v1.. na-s"t1.·.tk.h a-k. hyam• . sa.rpmohana.mapi nayottarakhyam (5 5 ) tat tumvuror vaktra-catu�kr� asya siddhyeva vipras samadarsayat sa� (56) XXVIII divijas sa.mmuddhrtya sastra-sara.m. rahasya-kaustalyadhiya sayatna� ( 57 ) siddhir vvahantih• kila devaraja bhikhy8.J!l vidadhre bhuvana-rddhi-v�ddhyai ( 58 ) XXIX 18 version of the inscription : Va.masiva, the guru of Yasovarman (887-ca. 910 A.D.), "erected a Siva-linga on the king's request upon the Sri Yasodhara­ giri, equal in splendour to the King of the Mountain [MeruJ" (B, 15). Further, it is said of Isanamurti, the hotar of Jayavarman IV (921/28-940 A.D . ) , that, full of devotion (bhaktia) , he honored Tribhuvanesvara (B, 27) --the linga that Jayavarman IV. caused to be erected on the Prasat Thom• in Koh Ker . In attempting to solve the problems posed for us by the devaraja cult, scholars have sometimes been all to readily tempted to see.ncertain enigmatic allusions to the devaraja cult in the Sdok Kak Thom inscription . as having an immediate bearing on the archaeological problems of Angkor. Sometimes it seems that a number of the difficulties attending research into the devaraja cult can be reduced to a methodological problem. We have been attempting to elucidate the nature of a cult whose existence is known, ultimately, only from an inscription, by using archaeological means, and we have even been trying to solve further problems by the same means, before the possibilities of clarifying the problem through the in­ scription have been completely exhausted .69 However, as the Sdok Kak Tho� inscription is still, as it always has been, the only source which makes possible any s9rt of coherent statement about the devaraja cult, we shall here deliberately follow the opposite path--that is, we shall in the first instance assess the data of the Sdok Kak Tho:rri inscription in. their own rightn. It will be apparent that the ·analysis of the Sdok Kak ThoI!1 inscription ·nyields a relatively clear picture of the devaraja in .at least three particulars : (1) the devaraja is Siva; (2) it is worshipped in a movable cult image, which (3) is not identical with the lingas of the large temple mountains of Angkor. Hence the temple mountains of Ang­ kor, rich as they are in abundant archaeological and epigraphic indica­ tions of their cosmographic significance, are ruled out as sources for the devaraja cult in the strict sense . They are, however� of crucial significance for the problem of royal apotheosis in the kingdom of Angkor in the wider sense.70 69certainly, when we think of the size and number of the at first apparently almost insoluble archaeological problems confronted by archae­ ologists in Angkor, this methodical approach becomes entirely comprehensi­ blen. It seemed capable not only of relating the large number of undated Angkorian temple buildings to a chronological sequence, but also of pro­ viding a key to the question why this prodigious building activity had been undertaken in Angkor. The constantly repeated references of the Sdok Kak Tho� inscription to the priests who carried out the kamrateh jagat ta raja rite under the various kings immediately presented themselves as a key to the solution of this problem. It followed from the theory that after their death, the temple mountains of Angkor became the mausolea of the men who built them, that the new kings were virtually·compelled to erect new temples for "their" devarajasa. 70Heine-Geldern, 1930, pp. 33-45; Coedes, 1955 ; Filliozat, 1954 ; Stierlin, 1970, pp . 81 ff . See also Mus, 1933, pp. 696-710 and M. Eliade, 19 If we read the Sdok Kak Thom• inscription bearing in mind the famil- iar theory that the devaraja was consecrated by - the kings from time to time on the temple mountains built by them, it instantly strikes us that, on the contrary, the kamraten jagat ta raja was consecrated, or even solemnly inaugurated (pratif!pha)a; only once, namely under Jayavarman II, on Mahendraparvata. 7 1 Thereafter, r�·peating i tself. almost mechanically, it is said only that, under the supervision of a particular successor � . Sivakaivalya, the latter's "whole family officiated in the presence of the kamraten jagat ta raJa in accordance with the rules."n72 · From this, to my understanding, it can only be concluded that a particular cult ob­ ject, perhaps·na sculpture, was consecrated once, 7 3 and that the further celebration of the ritual pertaining to this same cult object ·was observed by the successors of the foundation priest. This signification is con­ firmed in the inscription where it is said that the brahman Hiran• yada.ma, as a savant of occult magic, consecrated the kamrateh jagat ta raja ac� cording to the Vinasikha ritual. Then he recited the four presumably tantric textsn74 from beginning to end, and wrote them down in order to instruct the court priest Sivakaivalya. "He taught Steri afi Sivakaivalya to celebrate the rite (vidhi ) performed in the presence of the kamrateh aft ta raja . "75 One of the consequences of this prominence accorded in the Sdok Kak Thom• inscription to the texts and to the rite (vidhi ) based on them has be.en that Philippe Stern presumed the devaraja to be a rite "Centre du monde, temple, maison," in Le symboZi.sme cosmique des monuments religieux ( Roma, ·1957), pp. 57-82 ( Serie Orientale Roma, XIV). 71Before the passage translated above ( C, 73-74), in which the brah­ man Hiral}yadama performed the consecration, the assertion was made right at the beginning of the Khmer text that "then His Royal Highness Parames- � vara c·nonsecrated the kamrateh jagat ta raj a in the city of Sri Mahendra- parvata: (man vra� pada Paramesvara prati�tha kamrateh jagat ta raja anau nagara Sri Mahendraparvata, C, 56). See also Coedes/Dupont; 1945/46_, p. 103, note 2. 72gi kule phon sin na kamrateh jagat ta raja ru ta tapra ¾au ( D, 11-12). This explanation is repeated almost word for word, nine times altogether, for the reigns of all the Angkor kings up to Udayadityavarman II. 73Briggs advanced most�inphatic�lly the opposite interpretation ( 1951b, p . 233): "Hiran.yadama taught Siv akaiv alya the magi.c ritual to en-. able the purohita or other members of the family to create a new devaraja on the accession of a new king." · · 74As early a� 1915, Finot supposed . ( 1915, p. 57) that Tantric texts . were involved here. See P. C. Bagchi, "On Some Tantrik Texts Studied in Ancient Kambuja," Indian Historical Quarterly, V ( 1929), pp. 754-69, and VI ( 1930), pp. 97-101. Dupont and Coedes ( 1943/46, p. 64), on the other hand, referred cautiously to ''sastra siv ai tes." 75gi ta thve vidhi na kamraten jagat ta raja ( C, 75-76). 20 rather than a particular cult objectn. But, if we start from the premise that the kamrateh jagat ta raja or the devaraja was a particular sculp­ ture , then it is only to be expected that, once the cult object had been consecrated , th.,,.e rite should occupy the foreground in the view of the successors of Sivakaivalya ; for the celebration of it , and , even more, the knowledge of the secret texts on which the whole cult rested , were the most important possession of this priestly family. As we have already observed , there are pre-eminently two objections to the theory of the devaraja as an "idole unique." Coedes adduces , against it, above everything else, the grandeur of the Tribhuvanesvara lihga of King Jayavarman IV, and Stern sees the lihga of Yasovarman on the Bakhen temple mountain as clear evidence that on occasion the devaraja as lihga was erected for a particular ruler on the temple mountain built by him. Bearing in mind these objections , let us first examine the text of the inscription in relation to the lihga which Yasovarman erected upon the foundation of Angkor.76 Yasovarman's procedure at the founding of Angkor is very clearly described in the Sdok Kak Tho� inscription. In the relevant passages of the text in the Khmer version, four successive transactions are protrayed : ( 1) Yasovarman founded the city of Yasodhara­ pura (Angkor) ; ( 2) he transferred the kamrateh jagat ta raja from Hariha­ ralaya, his previous seat of government, to this city; ( 3) thereupon Yasovarman founded (sthapana ) the "central mountain" of Angkor, the Bakheri temple mountain; and (4) the guru Vamasiva founded (sthapana ) the sacred lihga in the center--quite obviously referring to the lihga on the central temple mountain , the Bakheri .n77 The inscription makes it clear that first the kamrateh jagat ta raja was brought to Yasodharapura and only later was a lihga consecrated on the central temple mountain built in the meantime. Even Coedes , who argues vehemently for the identity of the kamrateh jagat ta raja I devaraja with the lihgas on the Angkorian temple mountains ,n78 comes , together with 76on the tribhuvanesvara lihga of Jayavarman , see below, pp. 38 ff. 77man vraJJ, pada Paramsivaloka cat nagara Sri-yasodharapura ncurz kam­ rateh jagat ta raja Cll[lVi Hariharalaya yok duk ta,,,,nagara no� man vra� pada Paramasi valoka sthapana Vncurz KantaZ. kamrateh Sivasrama sthapana vra� lihga ay kantaZ ( D , 12-13). 781n 1932 Finot (BEFEO, XXXII [1932 ] , p. 3, note 1) raised the ques­ tion of why then.Bakhen , which was consecrated in 889 A.D. [allegedly ! ] to the devarajaa, should in a later inscription, in 968 A.D.n, be eulogized under the name of Yasodharesvara. Coedes answered ( communication cited by Finot, BEFEO, XXXII [1932] , p . 3 , note 1) with the following suggested alternative explanations : ( 1) In the beginning, the Bakhen was possibly consecrated not to the devaraja but to the Yasodharesvara. This supposi­ tion is , however, certainly not confirmed by the Sdok Kak Tho� inscription 21 Dupont, to the same conclusion in their translation of the inscription: "Le developpement du recit montre clairement que le transfert du devaraja a Yasodharapura constitue ·un episode defini, tandis que la fondation du Mont Central, OU un Zihga est place, en constitue un autre .n. Chaque epi­ sode est·nd'ailleurs introduit par man, 'alors,n' qui indique une reprise dans la narration.n"n79 So we can take it as certain that, in the Khmer version of the Sdok Kak Tho� inscription , for one thing, the transfer ( nairz) of the kamrateh jagat ta raja is recounted, and, for another, it is reported that thereafter80 ( man). a Zihga was consecrated on the central temple mountain. It is not legitimate to infer the identity of both icons from the inscription, as Stern did.n8 1 So we may di'spose of a major argu­ ment that has in the past been adduced for the identity of the devaraja with the lingas of _the temple mountains- of Angkor. (w hich is at least two hundred years later). (2) After h·is return from Kocy Ker (944 A.D.) Rajendravarman did not reconsecrate the Bakhen, alien­ ated as it had been from its original purpose by Jayavarman IV, to the devaraja, but instead he erected a new temple mountain to the devaraja and consecrated the Bakhen to the Yasodharesvaran. In his article of 1970 about Jayavarman IV, Coedes raises the question whether the expression devaraja was known at all as early as the ninth century. 79coedes and Dupont, 1943/46a, p. 113, note 6. 80On man see Appendix, and Introduction, p . xviii where the author makes the qualification that man is consistent with the transfer and the consecration occurring at the same time. 81Stern (1934, p . 613), on the basis of the text, reasons from a presumed simultaneity of the above-mentioned sequence of events to an identity of both icons: "la partie khmere ·du texte indique, avec preci­ sion, que le Kamrateh jagat ta raja, au moment de la fondation d 'Angkor (Yasodharapura) fut erige dans cette ville sur le mont central ; or la par­ tie correspondante sanskrite se. bornen· a signaler (BEFEO, XV [1915], p. 80 , verse 43) que, d'apres les ordres du roi, au moment de la fondation d 'Ang- ,,. kor, il (le Guru) erigea un Zihga sur le mont Srf Yasodharagiri, egal en beaute au roi des Montsn" (emphasis added). tThe Khmer section of the text indicates precisely that the kamrateh jagat ta raja, at the moment when Angkor (Yasodharapura) was founded, was erected in this town upon the cen­ tral mountain ; but the corresponding Sanskrit passage confines itself to the information (BEFEO, XV [1915], p. 80, verse 43) that, following the orders . of the king, at th,,.e moment when Angkor was founded, he (the guru) erected a Zinga on mount · sri _nYasodharagiri, equal in beauty to the king _ of the Mountains."] This synchronism of the erection of a linga and the shift of the kamraten jagat· ta raja must, however, rest on an error: the expression "at the moment of the foundation of Angkor'.' does not occur .,,. .. either in the text or the translation of Finot cited: "D 'napres les ordres du roi ' il erigea un linga sur. le mont Sri Yasodharagiri, egal en beaute --�- . au Roi des monts il'Himalaya). " [Following the king 's orders, he erecte� a linga .on Mount Sri Yasodharagiri, equal in beauty to .the King of the � Mountains (the Himalaya). "J Finot, BEFEO, · XV ( 1915), p. 80, at verse XLIII � -- - -._· · ..._ . . .-· ·--.. ' - · · • . \ 22 A further feature of the Sdok Kak Tho:rri inscription counts, to my understanding, just as unambiguously against the notion that from time to time the devaxaaja was consecrated afresh for a new ruler in the tihga of a new temple. In the Sdok Kak Thom. inscription it is certainly explaine� what "personal tihgasa" the variousnsuccessors of Sivakaivalya caused to be consecrated in their respective villages . In the same way , we are told , in chronological sequence , under which famjly chiefs of this priest­ ly dynasty the "whole family officiated in accordance with the prescrip­ tions in the presence of the kamrateh jagat ta rajaa. " On the other hand , we hear nothing about the devarajas alleged to be rededicated constantly. If the devaxaajas always had to be consecrated afresh in the royal tingas of the reigning kings , then it becomes a problem for us to see why these lihgas and their consecration are not named in the inscription in associa­ tion with their constantly repeated references to the kamraten jagat ta raja , especially since we know that the names of the tingas of these tem­ ples were in no way forgotten or altered with the death of their founders. As a particularly good example in this connection , the Indresvara on the Bakon in Hariharalaya (Roluos) may be noticed. If this lihga , which was still known by its founder's name generations after its founder Indravar­ man, 82 and famed for the magnificence of its t,,,,.emple , had been consecrated as the devaxaaja by a member of the family of Sivakaivalya, then it would be incomprehensible why this consecration is not mentioned in the inscrip­ tion. 8 3 The only answer to this question that can properly be given is that the In,ti,pesvara was not the devaraja during the reign of Indravarman. 84 A number of further controversial problems which have arisen in the course of debaten_ about the devaraja cult can be largely resolved by refer­ ence to the Sdok Kak Thom• inscription. Thus , the Khmer rendering of it leaves no doubt that a cult image and not an abstract rite occupies the central position in the devaraja cultn. For it is said constantly, as had already been mentioned more than once, that the- members of the �ivakai­ valya family officiated in the presence of85 the kamrateh ·jagat ta raja. 82 ,,,,.In an inscription of 960 A.D. it is named vrah• kamrateh an Sri- Indresvara ( IC, IV , p. 103 CK . 265J A , 14-15). 83Instead, however, we find only the laconic statement that "under the reign of His Royal Highness Isvaraloka (Indravarman I)· the kamrateh jagat ta raja was in Hariharalaya, (and) the whole family officiated ac­ cording to the rules in the presence of the kamrateh jagat ta raja" (D , 4-5). As further examplesn, the tihgas of Rajendravarman in the Eastern ,,,,. Mebon and Pre Rup may be cit�d. Moreover, there is the lihga of Siva erected around 1000 A.D. by Sivacarya on the Hemas�nga (Ta Keo) ; cp . the Stele of Tuol Ta Peen, sloka XCI ( IC, V , p. 256). 84 ,,,,. In the foundation inscription of the Bakon is clearly written: Sri-Indresvaxaa iti lihga { IC, I , p. 32, sloka XXIII) . · See also Baksei Chamkrori , sloka XXVI ( IC, IV, pp. 88 ff.). 85or "in connection with." See Appendix under na. 23 This "in the presence of" is invariably represented i. n Khmer by . na, which has an unambiguously locative significance, and here it can only be said that a rite was performed in front of a cult image.a8 6 The word na1z1,a8 7 which is always used when the devaraja follows the king on a change of capital, points in the same directionn. Nam, w.hich is equivalent in mean­ ing to the expression "to escort a person,n'' is applicable rather to a cult image than to an abstract rite. So also with area or arcana in the San'­ skri t version of the inscription, whic·h is the equivalent of "veneration, worship." In a context such as, "they offered worship to the devaraja" (devaraja cakrur arccam, B, 34 ) , devaraja can only represent a cult ob­ ject, but not its ritua1.n88 A further question, critical for th_e problem of the di viniza.tion of rulers in Cambodia ; is likewise clearly answered in the Sdok Kak Thom. in-. scripti on. · That is, in complete accordanc·e with Filliozat' s interpreta- tion, we find absolutely no indication in the Sdok Kak Thom inscription . . that the kings of Angkor were worshipped as devaraja or kamrateh jagat ta raja. On the contrary, it is stated unambiguously that the rite on Mahen­ draparvata was performed by Hiranyada.ma at the request of Jayavarman II so that there should be "onlyn·on� king who was cakrav tin'.' ar ( C , _ 73). 89 A further indication of the "earthly" character of the rule wielded by the kings of Angkor may be seen in the expression kamrateh phdai karorri which is used many times in the inscription for "king. ." Phdai karom. means "lower plane, the earth," so kamrateh phdai karorri is equivalent to "lord of the lower plane, lord of the earth."9 0 While the conceptions kamrateh jagat ta raja ( "lord of the world, who is king") and kamrateh phdai karon:z ( "the lord of the lower plane/earth") seem indeed to be parallel ideas, in the Sdok Kak Thom. inscription they .in fact designate two poles of 86Mabbett, 1969, p. 206. 87see Appendix under nam. . 88 Sahai, 1970, p. 41 : ·"Notre texte indique clairement que le deva- raja n'etait pas un simple rituel, mais une divinite adoree avec des rites precis." [ Our text shows clearly that the dev�raja was not a mere ritual but a divinity worshipped with specific. rites."] 89It is not in the least disputed here that the idea -of a "universal ruler" (cakravartin) contained in itse·lf a species ·nof divinization of the ruler. The concern here is merely to establish whether the conception of the devaraja assumed proportions exceeding those of its Indian·nmodels-­ perhaps in then_nsense of an outright identification : the king is God. 901ewitz, 1971, p. 98: "Kamrateh phdai karorri , Lit. 'seigneur de la surface inferieure,n' c'est-a-dire 'le roi' regnant sur 'la terre', la­ quelle s 'oppose au ciel qui estn· 'la surface superieure' (phdai Ze) •" C "Kamrateh phdai karon:z, lit. 'lord of the lower surface', that is, 'the king' reigning over 'the earth', as opposed to the sky which is '- the upper surface' (phdai Ze) •"J 24 ,, lordship. The god Siva is the "lord of the world, who is king" (one might say : the highest king) , while here below the king rules as "lord of the earth. "9 1 This antithesis becomes even clearer when we read that the "lord of the world , who is king" protects (cam• ) the "lord of the earth" (C, 81-82). One must therefore ask how the king of Angkor can be the kamrateh jagat ta raja himself when he is actually protected by the lat­ ter. We may take it as established , then , that the Sdok Kak Thol!1 inscrip­ tion offers no warrant at all for the theory that the kings of Angkor were "god-kings" ( devaraja ) : on the c.ontrary , it is stated unambiguously that the "lord of the earth" is protected by the "lord of the world." ,, IV. Devaraja--a "Galanti Pratima" of the God Siva? In the foregoing discussion we were able to establish that an analy­ sis of the Sdok Kak Thom• inscription shows the devaraja to have been of necessity a cult object and that this cult object was not identical with the Zihgas which the kings of Angkor had erected on the monumental temple mountains of their capitalsn. But the Sdok Kak Thol!l inscription does not help us any further with the question what sculpture or what cult object was worshipped as devaraja , as it contains no relevant datan. In any event, the inscription allows the legitimate supposition that1 in the do-,, ,, main of royal Saivism , a Siva Zihga could have been involved.� 2 It is likewise unclear in which temple the conjectural devaraja sculpture was worshipped. Indeed the fact that to date it has not been possible to assign this function to an Angkorian temple induced Coedes to reject Finot's theory of the "idole unique."9 3 Since neither epigraphic nor 91A verse of the Prah• Bat inscription cited by Filliozat (1954 , p • 550) seems to point in a similar direction. Here it is said that the two worlds have two protectors: "this world" has the king (Yasovarman) and the heavens have the Great Indra. 92coedes , 1968 , ,,p. 23: "Most of the kingdoms founded in Farther In- dia soon adopted the Saivite conception of royalty , based on the Brahman- Kshatriya pairing and expressed in the cult of the royal Zihga.a" 93coedes , 1952a, p. 13 : "L' hypothese d 'un devaraja unique a travers les siecles souleve de grandes difficultesn, car on a vainement cherche son sanctuaire parmi les grands monuments de la capitale." ["The hypothesis of a single devaraja enduring through centuries raises great difficulties , because its sanctuary has been sought in vain among the great monuments of the capital."] See further Coedes , 1952b, p . 52. Possibly it is Dupont who is responsible for the suggestions in the Preface to Coedesn' s and Dupont ' s edition of the Sdok Kak Thom inscription where it is said, characteristically, that "Ln' object, un Zihga sans doute, auquel s ' adres- sait le culte , etait unique et pouvait etre deplace.n· L ' inscription de Sdok Kak Thol!1 mentionne a plusieurs reprises que le kamrateh jagat ta raja a ete transporte a la suite de tel roi , installe dans telle capitale. Il archaeological sources seem to afford us any information showing in which sculpture and in which temple the devaraja was worshipped, and since we have now had to dispense with the theory of identity between devaraja and the temple pyramid lihgas� only the avenue of hypothesis remains open to us at present if we are to approach the solution to this problem. It appears that in the past, whenever discussion has turned on the devaraja as "idole unique," it is only the central god statue, or the lihga in the central sanctuary of the relevant state temple, that has been considered. But there is a further form of the murti of a god that still plays an extremely important role today in the Hindu temples of India. This further form, found in very many of the larger temples of India, is the calanti pratima, which takes the form of a ( usually) bronze "mobile image" of the chief divinity. These movable god-images are an important constituent of the cult, especially during the major temple festivals, when they are conveyed through the streets of the temple city. At that time, they are carried on temple carts or litters as utsava murti ("festi­ val image") of the god--whosen_nprimary image remains standin� _ninn·nthe tem­ ple. A calanti pratima was and is especially impo.rtant in Saivi te tem­ ples, in which a lihga is worshipped. As indeed Finot has remarked , this lihga that stands fast in the temple naturally cannot be carried out on the occasion of festivals, and certainly it cannot be taken to the new capitals. 94 ,,. ' ' One of the most famous of the Siva temples of India is. the Lingara.ja temple in Bhubaneswa.r, Orissa ( in eastern India).9 5 This temple provides us with a good example. At the center point of the Lingara.ja cult stands a svayambhu-lihga, a "self-existent" manifestation of the god t3iva, in mas,,.s ive stone. It goes without saying that this "self-generated" ,,. image of Siva cannot be removed from the spot where Siva originally manifested himself. So, during the numerous L:i.ngaraja festivals, the function of avait done une personnalite .physique et n'etaitn.nvraisemblablement pas rem­ pla�able" ( 1943/46a, p. 64). Cf'The object to which the cult was directed, doubtless a lihga, was unique and could be moved. The Sdok Kak Tho:qi in­ scription mentions repeatedly th�t the kamrateh jagat ta raja was trans­ ported in the retinue of this or that king, installed in this or that capital. Thus it had a physical personality and probably. was not replace­ able." ( Emphasis added.) J This explanation of the devaraja cult comes closest to the one offered here, but clearly contradicts earlier and later statements by Coedes. 94Thus, it is stated unambiguously : ."A Sivalihga may not be moved" (sivalihgcurt na calayet)a. 95This, the largest Siva temple in eastern India, is evidently a "contemporary" of the largest ·saivite temple mountain of Angkor, the Baphuon, which was built at the beginning of the second half of the eleventh century. K . C. Panigrahi, Archeological Remains at Bhubaneswar ( Calcutta, 1961), p. 166. 26 "deputy" for Siva outside the temple is discharged by a four-armed bronze sculpture, about 45 cm. in height, which represents �iva as Candrasekhara.n9 6 During the festivals, this sculpture is the focus,, of all those royal rituals that are directed to the Lingaraj a as " ( Si va-)Linga which is the. king."9 7 In Orissa, just as in other parts of India, calanti pratima sculp­ tures played an important role which could be of very considerable signi­ ficance for the meaning of the devaraja. There are numerous examples of the elevation of powerful regional deities to the status of imperial divinities ( ras• t• r adevata) of the ruling dynasties. Especially on occa- sions when the holy places of these divinities were remote from capitals established by new dynasties, it seems to. have been only a matter of time before bronze sculptures were made of the original god-images ( mula bera) and worshipped in the imperial capitals, usually in the immediate vicinity of the palace. Aniconic effigies of fhakura�is or SvayCll[lbhu-Zihgas espe­ cially are thus embodied as Hindu divinities with particular regularity. A familiar example from Orissa would be perhaps the Thakurani Bhattarika. While she is worshipped in her original site on the River Mahanadi in an uncarved stone, she is at the same time in the palace of Bara.mba as Durga­ Mahi�asuramardini, the imperial divinity (ra�tradevata) of the former princely state of Baramba.n9 8 The rite of this "deputy" sculpture was per­ formed in the palace of the king ' s court chaplain "with the exclusion of other people"--just as the Sdok Kak Tho� inscription says with regard to the devarajaa. Just as the devarci.ja as "sourcen·of the treasure of power" seems to have been held most precious in the sight of the king, so in Orissa it was permitted only to the king and his court priest to set foot in the sanctuary of the ra�pradevata in the palace precincts . Naturally 96Rajendralala Mitra, The Antiquities of Orissa (Calcutta, 1875, new impression Calcutta, 1963), II, pp. 133 ff. [ Indian Studies, Past and Present]. On Siva's Candrasekharamurti, see J. N . Banerjea, The Develop­ ment of Hindu Iconography , 2nd ed . (Calcutta, 1956), p . 463. 97The name Lingaraja is not without interest for our discussion of the devaraja. ,,The two names are constructed in parallel waysn. If we re- late lihga to Sivan= Deva, then the name Zihgaraja comes very close to the name devaraja in meaning as well. In the case of Zihgaraja, indeed,.nwe have to reckon with a double meaning: "the king of the Zihgas" ( in regard to the rest of the lihgas in India) and "the king who is a Zihga. " In ritual, Zihgaraja receives royal honors and attributes, possibly in emu­ lation of the state cult of the vais• n• a vite Jagannatha in Orissa and in competition with it. 98For an aetiological explanation of this "metamorphosis" see Pilahka BadZunb• a Itihasa , 1940, p. 26. The example best known throughout Orissa of the promotion of an originally tribal, regional divinity to the rank of a "Hinduized" state divinity is the Vi��u-Jagannatha in Puri ( see Kulke, 1973, pp. 129 f . ). On the "promotion" of Durga see Goetz, 1974, pp. 70 ff. 27 enough, the observance of this cult, just like that of the devaraja , was made hereditary in the family of the 'court priest. Like the ·adevaraja of Angkor, the sculpture as ra�tradevata was always transferred on a change of capital, while the primal divinity remained as ever in its original �bode. It seems that sometimes these bronze ras• t• radevatas were even taken into battle to protect "their" kings. It , is not difficult to surmise the reason why a calanti, pratima was brought into the vicinity of the palace. It was held to make the power of the state divinity effective for the legitimation of rulership. In the later middle ages, the kings of - the central Orissan dynasty were reigning as rauta, regents of their imperial tutelary divinity.n99 Almost simultaneously in tenth and eleventh century Angkor, the notion prevailed that the king was a "part" (am. sa) of the � state god Siva (see below, pp. ·29 ff. ). When we examine the evidence in the Angkorian inscriptions where· references to the devaraja cult are made, we find one inscription which could perhaps offer confirmation of the hypothesis that the devaraja was also worshipped in the form of a calanti, pratima. An inscription of Kok Rosei enumerates the endowments made by the priest Sivacarya, who was pre­ sumably the famous purohita of the devaraja under kings Jayavarman V and Suryavarman I at the end of the tenth and the beginning of the eleventh centuries. In this inscription it is reported that he .nhad made three presentations to Bhadresvara (the state divinity of the pre-Angkorian Chenla state), "my divine l. ord, who is· lordship" (vrah' . kamrateh an ta rajya ) , and to the "Lord of the world"n. in Lingapura ( Kol]. Ker).. Now, the immediate context in which "the divine lord who is -lordship" is named should strike us_: ta vrat kamrateh an ta rajya sru vra (tJ vleh pratidina Zit mvay--"for the div ine lord who is _lordship, _ daily one Zi"fj, of rice for the holy fire."n100 The holy fire (vrah• vleh ) of Angkor here stands in a close relationship with the kamrateh an ta rajya, or_ rather the two to- gether seem act·ually to compose a unity. The holy fire must have played � prominent role in the state cult of Angkor. Thus in 1001 A.D. .the dig­ nitaries of the Cambodian empire swore their famous oath of allegiance to 99Kulke, 1974. See also G. Sontheimer, "Religious Endowments in India: The Juristic Personality of Hindu Deities," Zeitschrift fur vergleichende Rechtswissenschaft, LXVlI (. 1964), p. 76 : 0The king's being an agent of the god proved to be a very effective assurance of the loyal­ ty of the subjects and prevented any further disturbance- s from the side of the people, which shows that the deity can be a very potent force." 100IC , VI, pp. 173-80 (K. 175). Stele of Kok Rosei, A, 16-17 : "au Dieu- royal _ (° V• . K.A. ta rajya) quotidiennement : 1 Zit de paddy pour le Feu sacre." ["Daily, to the royal God (vraYJ, kamrateh an ta rajya) one li� of paddy for the Sacred Fire."J .The damaged west side of the stele mentions the family (of �ivacarya?) in connection with the vraYJ, kamrateh an ta rajya (IC, VI, p. 176, lines 5-6) ! 28 Suryavarman I in front of the holy fire,n1 0 1 and in the magnificent "his­ torical gallery" of Angkor Wat, in the scene showing Suryavarman II's military parade, the holy fire is clearly visible on a litter carried before the rajapurohita and the brahmans .n1 0 2 Significant for our hypothe­ sis is the fact that the holy fire here appears in the form of a calanti pratima. Its outward, cylindrical, form, from whose upper curve fan­ shaped flames seem to spring forth, was explained by Groslier as a minia­ ture stupa and by Bosch as a lihga.a1 0 3 Bosch bases his explanation chiefly on the mythologically derived and epigraphically attested relationship between the lihga and fire. . None of this should be allowed to give the impression that the deva-ra-Jaa is identical with the holy_ fire of Angkor . The intention here is merely to point to the possible "functional" contiguity of both cult ob­ jectsn. If, as has commonly been assumed in the past, 1 0 4 the "lord who is lordship" endowed by Si vacarya should be identical with t·he devaraja, then this "functional" contiguity emerges from Sivacarya's inscriptionn. Vra'IJ, kamrateh an ta rajya and vra1 vleh obviously belong closely �ogether in the cult realm, in that the holy fire seems to have been almost a consti­ tuent of the devaraja cult. In this connection, a conclusion established by Bhattacharya has definite significance for our discussion: "Certains textes permettent meme de conclure que le Feu etait garde· en permanence dans le palais royal. Les inscriptions de Phimanakas et de Prasat Tor (regne de Jayavarman VII) parlent de la 'salle du Feu' (agnigrha, vahnya­ gara) . " 1 0 5 With the help of parallels from eastern India ( where calanti pra­ timas of imperial divinities were kept in palaces of their precincts), and information gained from analysis of both the Sdok Kak Thom. inscription � and that of Sivacarya, the following hypothetical explanation may be ad- 101c oede... . s, 1913a, pp. 12 ff . Likewise the author of the Sdok Kak Thom inscription, Sadasiva, .was married to the sister-in-law of Suryavar- • man I before the Sacred Firen. 102 Coede...s, 1932, pl. 549. 103Bosch, 1932, p . 11 : "l'objet porte en procession solonnelle est, soit un lihga venere comrne le feu sacre, soit un receptacle du feu sacre en forme de lihga.a" ["The object carried in solemn procession is either a lihga venerated as the Sacred Fire or a receptable of the Sacred Fire in the form of a lihga.a"J 104IC, VI, pp. 173-80; Sahai, 1970, p. 43 ; Coedes, 1970, p. 59. 105Bhattacharya, 1961, p. 148 (emphasis added) .n. [ uCertain texts even allow the conclusion that the Fire was kept permanently in the royal palacea. The inscriptions of the Phimanakas and at Prasat Tor (reign of Jayavarman VII) mention the 'chamber of the Fire' (agnigr• ha, vahnyagara)a" . J 29 vanced: the devara;ja was a calanti pratima in the form of a bronze sculp­ ture representing Siva in one of his divine embodiments. This sculpture was worshipped in successive capitals ., either in the palace or within the palace grounds in a special building , which, in the style of the royal palaces of Angkor, was built for the "god who is the king" of perishable but valuable materialn. When kings removed to other palaces or new capi­ tals, this calanti pratima of the devarada was escorted there (nal[l) a.s palladium of the empiren. It had to remain close to the king in order to watch over ( cam• ) the "lord of the earth"nas "the lord of the world ."n1 0 6 V. Were the Kings of Angkor "Participants" in Divine Rule? With these hypotheses envisaging the devaraja as a calanti pratima of an original lihga on Mahendraparvata, this discussion' ·s immediate task , in the strict sensen., is now concluded . But the devaraja cult belongs also to a broader context--the glorification and divinization that characterize the proceedings of kingship (rajya) and kings (raja) as bearers of domin­ ion in India and the regions· influenced by India. Though we cannot attend here in any greater detail to- the various forms the Cambodian rulers' apotheoses have taken, still there is one reason that requires us to establish more specifically how we should regard some aspects of the Ang­ korian kings' divinization. The inscriptions of Angkor contain allusions to the divine essence of the kings as a "portion" (amsa) of Siva that is located, as the king's "subtle inner self" (suksmantaratman) , in a Siva linga·, and these allusions regularly refer to the lihgas which the kings caused to be consecrated on the temple mountains built by them. Since these lingas have previously been seen as the devaraja of successive kings, the inscriptional references concerning these lih . g as have commonly been applied directly to the interpretation of the devaraja cult as a wholen. This is true in large measure for the central lihgas of the Bakhen, the Prasat Tho� in Ko9 Ker, and the Baphuon; which have previously been seen as the devarajas of their founders, kings Yasovarman I, Jayavarman IVa, and Udayadityavarman II , and which therefore played a crucial role in the discussion of the devaraja cult. However, as the opposite theory is advanced here, it is therefore necessary to make clear at the outset that these lihgas are not identical with the devaraja. After clarifying this matter, we may briefly notice the significance of these royal· lihgas on the temple mountains for the royal cult in Angkor . These observations are so to speak extraneous to our discussion of the devaraja cult as such . Having cited above evidence refuting the identification of the Bakhen lihga with the devaraja, we may now turn to the Tribhuvanesvaran· lihga of the usurper Jayavarman IV. At the very outset of his reign in his new capital of Chok Gargyar (Ko9 Ker) to the northeast of Angkor, Jay.avarman appears to have begun with the construction of what was then 106Majumdar ( 1963, p. 210)n-even designates the devaraja as the "tutelary deity of the kingdom." 30 the largest temple mountain of Cambodia, the Prasat Thom• . On it he con- secrated the Tribhuvanesvara Zihga , which was to play a prominent role in the legitimation of his rule. In the inscriptions of Ko� Ker this lihga is called kamrateh jagat ta rajya as well as tribhuvanesvara ("lord of the three worlds" ) . Now as Coedes sees both the kamrateh jagat ta raja in the Ko� Ker inscriptions and the kamrateh jagat ta raja familiar to us from the Sdok Kak Thom• inscription, similarly as names of the devaraja, it follows that the Tribhuvanesvara Zihga represents for him the demon- stration par excellence that the devaraja was erected and consecrated afresh by successive reigning monarchs on their new temple mountains.n1 0 7 To the arguments previously adduced against this theory, touching on the devara.ja cult as a whole, may be added two more important ones which in connection with Koh Ker militate against Coedes 's theory. Through • them it is possible to establish that, on the one hand, the Tribhuvanesvara Zihga cannot have been the devaraja of Jayavarman, and that, on the other, · in the first years of his reign Jayavarman ruled in Ko� Ker without the legi�imation of the devara.ja cult. In an inscription of Kocy Ker dated 923 Saka (1001 A.D.n) it is reported that high officials in the service of King Udayadityavarman I (ca . 1001-1002 A.D.n) bestowed endowments upon the kamrateh an jagat ta raja in Chok Gargyar (Ko� Ker ) . n1 0 8 So there can be no doubt that in 1001 A.D. the kamrateh an jagat ta raja = tribhuvanesvara was still in Ko� Kern. At this time the original kamrateh jagat ta raja = devaraja had already been back in Angkor for about half a century, since Rajendravarman II (944-68 A.D. ) bad returned the capital to Angkor and "taken back the kamrateh jagat ta ro:,ja at the same time."n1 0 9 It should therefore be apparent that the name kamrateh jagat ta rajya was applied to the tribhuvanesvara and not to the devaraja. In another inscription of Koh Ker it is mentioned that as early as • 921 A.D. Jayavarman IV had the Tribhuvanesvara lihga consecrated in Ko� Ker. 1 1 0 Elsewhere, however, in the Sdok Kak Thom• inscription, we learn 107coedes, 1970, pp. 58 ff., and 1968, p. 314, note 86. See also Sahai, 1970, p. 42. 108Koh Ker, Prasat Thom (K. 682 ) , lines 3-6 (IC, I, p. 50). • • 109sdok Kak Thom inscription, D. 36-37. • llOThe vrah kamrateh an jagat ta rajya is named in other inscrip­ 0 tions of Koh Ker as early as 921 A.D. (Coedes, 1931a, p. 15 ). On the chronology �f Jayavarman IV see Jacques, . 1971, pp. 168 ff . He contradicts Coedes's contention that by 921 A.D. Jayavarman had already assumed power as usurpern. Jacques bases himself on two inscriptions (Prasat Nan Khmau, IC, II, p. 32, verse III, and Tuol Ta Pee, ,,IC, V, p. 256, verse LXXXVI ), which both date Jayavarman's accession to Saka 850 = 928 A.D. For Jacques, the royal titles of Jayavarman and the mention of rajya in his early inscriptions of 921 and 922 A.D. apply to Jayavarman as 1'roi de Ko� Ker, et non pas roi des Khmer" (p. 169n) ["king of Kol} Ker, and not king 31 that the priest Kumarsva.min and his whole family "officiated before the kamrateh jagat ta raja" under kings Har�avarman I and Isanavarman _ II (D, 29 ). Yet until about 922 A.D. or 925/928 A.D. these kings were ruling in Angkor.n1 1 1 From this, to my understanding, it can only be concluded that in the first years of his reign as opposition king in Koh Ker Jaya­ varman IV was not in posse�sion of the devaraja.a1 1 2 In the Sdok Kak Thom• inscription it is said that Jayavarman IV "went forth from the city of Sri Yasodharapura to be ruler in Chok Gargyar, ( and) took with him the kamrateh jagat ta raja ( D, 31-32). Accordingly the second part of this statement, concerning the devaraja, cari only apply to the perio. d after . the end of the rule of -I�s anavarman, when Jayavarman I.V was sole ruler of Cambodia, and when he had already consecrated the Tribhuvanesvara Zihga in Ko� Ker and installed it at the center of his religio-legitimating system. The history of political ideas as a whble, in Cambodia and, in a measure, in Southeast Asia, has been stro.ngly influenced by the fact that Jayavarman IV must have reigned as a usurper without the sanction of legal succession, and, for several years without legitimation by the devaraja cult. To use Max Weber's terms,n1 1 3 Jayavarman must have striven to estab­ lish a new legitimacy of his own stemming from the "charisma of office"n1 14 in opposi"tion to the legitimacy �temming. from· the "charisma of succession" of the Khmers"J. · But it still remains unclear how, in Jacques's view, it can have been possible for the kings of Angkor to grant Jayavarman "volun­ tarily" the liberty to build himself such a "bastion" in Kol_). Ker, north of Angkor, equipped with all the insignia of royal dominio_n, including then· tallest temple mountain in Cambodia--while at ·the same time the construc­ tion of the Bak�ei Cha.mkron temple was manifestiy a far less pretentious undertaking. 111 Coede' s, 196 8, p. 114n; Jacques, 19 7 1, pp. 168 ff. 112Briggs ( 1951a, p .- 116) attempts to resolve the difficulties with the idea of a schism in the devaraja cult : ''The existence of two God-Kings would not seem so strange if, as Stern thinks, the kamrateh jagat ta raja was a ritual, which could transform any Zihga into a God-King." It is remarkable that, _nas far as I know, Coedes did not explore this obvious contradiction to his theory in the epigraphic evidence. 113Max Weber, Wirtsahaft und GeseZZsahaft. Studienausgabe, ed. J. Winkelman ( Koln/Berlin, 1964), pp. 184 ff., 865. 114Another alternative to legitimacy based on the "charisma of suc­ cession," less familiar in Angkor, was constituted by charismatic rule on the strength of divine election. This method of legitimizing a coup d 'etat retrospectively, which was common especially in India and Indone­ sia, was usually mediated by·nthe "court mythographers" who concocted · legends recounting how the usurper became the elect of _the national divin­ ity. On this, see for -example C . C. Berg, "Javanische Geschichtsschrei­ bung," SaeauZum, VII ( 1956), pp. 168-181; VIII ( 1957), pp. 249-66 ; also Kulke, 1969. 32 possessed by the kings rightfully enthroned in Angkor. This occurred when Jayavarman caused the then-biggest temple mountain erected in Cambodia to be built, the architectural austerity of which generated a stark monumen­ tality that was not exceeded even by later buildings . However, instead of combining the name of the linga of this temple with his own name ("Jayesvara"), he dedicated it to Siva Tribhuvanesvara, the "Lord of the three worlds."n1 1 5 To this divinity , the lord of his "own" state temple, Jayavarman gave veneration as the supreme lord of the world, and thereby attributed sovereignty to it so that in future he would rule as "part" of (am• sa) or participant in Siva. So, as "deputy" for Siva he was account- able only to the god . Thus Jayavarman derived the legitimacy of his rule directly from §iva. An attack on the king would be tantamount to an attack on the rule of Siva.n1 1 6 Hence Jayavarman IV , the usurper, pos­ sessed a higher legitimation for his rule than the kings of Angkor whose legitimacy was derived from the "charisma of succession."n1 1 7 This development can easily be discerned in the inscriptions of Jayavarman. As early as 921 B. C., we read, "In his humility, this vie- 115What Fillinozat says of the devaraja in general is especially true of the Tribhuvanesvara linqa--it is not the kings of Angkor that are war- ';, shipped in it but the god Siva (as king) . Thus the name of the linga in ,,,,, Ko0 Ker is not that of the king who endowed it, but t�at of the same Siva Tribhuvanesvara, and its designation as the linga of Siva is made explicit in many ways--for example ugrasya lingam, lihgam idam siva�ya ( Koq Ker, Prasat Damrei, K. 677, verses rv, XIX, IC, I, pp. 58 f.), Sambhor lingam ( KoQ Ker, Prasat Andet, K. 675n, verse XX:VIII, IC, I. p. 64). The name Tribhuvanesvara, for Coedes an index of Jayavarman 's claim to "cosmic ,,,,, dominion," is also another name of Siva current in Cambodia. Thus the inscription of Palhal from the year 1069 A.D. records the consecration of a Tribhuvanesvaradeva ( Coedes, 1913c, p. 28). Jacques ( 1971, p. 169) fol­ lows the views of Filliozat concerning the Tribhuvanensvara. 116From Orissa in India we have a convincing parallel to this. King Kapilendra, who usurped the throne in 1435 A.D. and founded the SuryavaJ!lsa dynasty, had himself extolled as the elect of the Orissa state divinity Jagannatha , transferred sovereignty (sco�rajya) to this divinity, and reigned as its servant (sevaka)a. As such he threatened in numerous in­ scriptions that opposition to his oT,m orders would be treason (droha) towards the state divinity Jagannatha. On Kapilendra's inscriptions see K . B. Tripathi, The Evolution of Oriya Language and Script (Cuttack, 1962), pp. 251-73; see also Kulke, 1974. 117"rf the legitimacy of the ruler is itself not secured by the charisma of succession in accordance with unambiguous rules, he then re­ quires legitimation by another form of charismatic power, and in the nor­ mal way this can only be the hierocratic form. This is also, and espe­ cially, true for the ruler who embodies a divine incarnation and thus possesses the highest 'personal charisma.n'" Max Weber, Wirtschaft und Gesellschaft, p. 865. 33 torious Sri Jayavarrnan caused to b·e remitted to Tribhuvanesvara the entiret. glory and the power of sovE?reignty. " l l 8. In the same year, we find the Tribhuvanetsvara lihga eulogized as vraf} kamrateh an jagat ta rajya--"My Divine Lord of the World Who is Lordship."t1 1 9 In the undated inscription �f Jayavarman from Prasat D8.l!lrei in Kot Ker , it is further announced that Siva.came down to eartht.in order to protect the steadfast [King Indravar­ manJ who was part ( <17!1Sa ) of his own self. 1 2 O · In the inscriptions of the two most ·tpowerful rulers of Angkor in the eleventh century, we came· across the influence of this idea that the king is a part of the god who wields _supreme sovereignty as tutelary state divinity--the idea which began to produce marked effects on the legitima­ tion ideology of Cambodia only under Jayavarman IV.t1 2 1 Suryavarman I ( 1002-1052 A . D. ), who is generally considered to have been a Buddhist , endowed "his" temple mcrunta,,in in Angkor ; Sri Suryaparvata, · with a Siva­ linga under the name of Sawbhu .t1 22 In doing so he was ,,following the exam-. ple of Jayavarman IV in· · bestowing one of the names of Siva, and not his own name, upon the linga endowed-by him on the royal temple mountain .t. J,, ust as Jayavarman, in his inscription, i·s lauded as "fixed .p ortion" of . Siva, for whose protectio. n ,,S iva descended to earth, so a hundred years later we find King Suryavarman I. eulogized as an "enduring image of Siva upon earth."t1 2 3 There is a�other relevant inscription which is essentially 118"cte J na csrijayaJ varman.,a vijayina rajyasya saradbhutan;z bhaktya sarvam adiyata tri . · . . . " ( Prasat Tholp inscription, in A. Barth and A . Bergaigne, Inscriptions Sanskrites duaChampa et du Cambodge:, LXIV, verse III ), following the construction of C.oedes, 1931a, ·pp. 13 f, . He relates . sarvam . to the endowment recorde·d by _t he inscription: "ce [Sri JayaJ varman victorieux a donne avec devotion. tout ceci,.merveille et essence de la royaute." ("This victoritous Jayavarman has in devotion granted all this, the wonder and the essence of royalty. "J · 119 Coedes, 1931a, p. 15· .t. Inscription of ·prasat ThoJ!l, eastern gopura, second prakara-wa.11. 120 (K. 677t) Kol'J. Ker, Prasat DSJ!lI'ei , sloka XVII (IC� · r , p . 58 ). This very likely represents the influenc·e of the _vai�r:avi te Avatara doctrinet. The amsa idea points in the same direction,. for this too seems to belong in the realm of Vai�r:avism rather thant.�aivism. Thus, in the Bhagavata Pura.1:a, the Ur-kingt·tP�tnu is praised as aJ?1Sa of Vi�i:iu (Harer an;zsa1J, XV, 6 , or again Vis• n• oh• kala, XV, 3 ) ·• · . .121 For Rajendre.·v ar· man also the,, claim was made in an inscription of 948 A .D. that he was a "portion·" of Siva (ayam mama-arrzso bhumisaf} ) . In- · scription of Prasat Pram, A XIX (Coedes, 1913b, p. 19). 1221ovek inscriptioni of. Har�avarman III, sloka XXXVII ; Majumdar, 1953, p. 423. See also Ta Keo inscription, A 5-6 ( C�edes � 1931b, p . 18 ). 1231.,.. s,, . v araE?ya ksitau vaddha-murti . . Stele of Tuol Ta Pee ( K. 834 ) t, • sloka XCII (IC, V, p . 256 ) . 34 concerned with the accession of Suryavarman even though it dates from the period of Udayadityavarman ' s ;eign il050-l068 A.D . ) : here too Suryavarman is eulogized as a portion of Siva (Siva-arrzs a)a.a1 2 4 This testimony is im­ �ortant for the reason that the inscription begins with an invocation to Siva in which reference is made to the portions (arrzsa) which must be familiar to those who strive for salvation. Included among these portions of Siva is, surprisingly, the "ruler's self" (niyoktratman) . What is the relations hip between the amsa , the "ruler' s self" of ,;# • the god Siva, with King Suryavarman, who for his part is eulogized in the same inscription as an amsa of Siva?n1 25 • It is very probable that in the tenth and eleventh centuries the kings of Angkor were extolled as amsa of ,;# • Siva, and that this amsa of the god was seen in direct association with ,;# • the curzsa of Siva which was eulogized as his "ruler's self." The reign of Udayadityavarman II (1050-1066 A.D . ) also plays an im­ portant part in our knowledge of the devaraja cult and of the ritual sig­ nificance of the royal temple mountains of Angkor. For one thing, the Sdok Kak Thom inscription dates from the beginning of his reign (1052 • A. D .)--the only source which makes possible any coherent statement about the devaraja cult. For another, we find for the first time in inscrip- . tions concerning Udayadityavarman clear epigraphic evidence about the function of the royal lingas in the state cult of Angkorn. Thus, the famous Lovek inscription of the priestly Saptadevakula family, dating from the reign of Har�avarman III (1066-1080 A.D.), gives a clear account of the cosmographical import of Udayadityavarman's Baphuon temple mountain. From it we learn how Udayadityavarman erected a golden mountain (the Baphuon) in his own city, vying with the abode of the gods, the golden Mount Meru standing in the middle of Jambudvipan. On the summit of this golden mountain in a temple resplendent with divine radiance, he conse­ crated a Siva-lihga with the name of the "Golden Liriga. " 1 2 6 In the Prat1 124Inscription of Prasat Khna (K. 661), B, sloka LXI ( IC, I, p . 202) . While in Cambodia the king is exalted as the "fixed image" of the god on earth, in Orissa even to modern times he bears the name Calanti Vi�QU ("Record of Rights" of 1955) . As the "Moving Vi�9u" he is the deputy for the "fixed" Vi�r:u-Jagannatha in the templen. The name Calanti Vi�QU in Orissa does not in any way--as one might at first suppose-·-embody a di­ vinization of the king as the great god Vi�QUn. Here, too, it is simply a question of a functional similarity between the state divinity and the king of the staten. As such he functions as the farmer ' s rauta (deputy) and is as bhakta its "first servant" (sebaka)a. 125It is not to be ruled out that we see here the influence of the Buddhist bodhisattva ideal, as it is the supreme .ntask of the bodhisattva to obtain the salvation of all men--an idea which is alien to the saivite conception of royalty in the eleventh centuryn. 126Lovek inscription of Har�avarman III, verses 40-43 ; Majumdar, 1953, p . 44. Similarly, Inscription of Ph1:101!1 Sandak and Pratt Vihar, line 9 (Coedes and Dupont, 1943/46b , p. 141)n. 35 Nok inscription, General S8Jllgrama sought permission to endow this golden Zihga, which harbored within itself the "subtle inner self" of Udayaditya­ varman, with his spoils of war.n1 27 Here, in a few lines, we find in epitome the apotheosis of the ruler in eleventh century Angkor. The "subtle inner self" of the king ("le roi abstrait" ascording to Finot) dwells in a lihga, the phallic manifestation of the god Siva, which a king has consecrated in the course of his reign in a temple mountain, and which is a guarantee of fecundity and strength. �en the kings of Angkor are exalted as a "portion" (a,zzsa ) of the god Siva, it appears that this portion and the "subtle inner self" of the king are one and the same.n1 2 8 Hence the god Siva and the king of Angkor are united in a Zihga upon the topmost step of a temple pyramid which consti­ tutes a microcosmic replica of Mount Meru, the .above of the gods, and represents the center of the Angkorian Kingdom. Now, as "fixed image" of ; Siva on earth, the king of Angkor in his capacity as "lord of the earth" (kamrateh phdai karorri) wields dominion on behalf of the "lord of the world who is [the possessor ofJ sovereignty" (kamrateh jagat ta rajya, and deva­ rajya). . Af ter his death the king enters the region (pada) of the highest god. Whereas these inscriptional data concerning the divinization of the kings of Angkor have in the past been regularly applied directly to the devaraja cult, it shoul_d now be apparent from the testimony of the pas­ sages cited above that they refer exclusively to the cult of -the royal Zihgas on the temple pyramids of Angkor .n 1 2 9 The only conclusion that can be drawn is that it was these Zihgas and their temples, and not. the deva­ raja, which were·nthe focus of the state cult. These Zihgas· on the temple mountains were the visible manifestation of the god Siva .n1 3 0 Here 127Pra� Nok Stele inscription, verse 159 (Majumdar, 1953, p. 398)n. 128In the Pre Rup foundation inscription of 961 A. D. , it is said of the Zihga in the southeast of. the upper platform of the temple that King Rajendravarman "erected the Lord (Isvaran) Rajendravarmesvara for his own salvation, just as if it were itself imbued with the royal essence" (IC, I, p . 102, verse CCLVIII). . 129It is apparent also in the case of the Zihga of the Baphuon that it was not, as Coedes among others supposed (Coedes, 1931b, p. 22), the devaraja. In the Khiner section of the inscription of Lovek mentioned above, the reference is clearly made to the kamrateh jagat survar�alihga (IC, VI, p. 285, line 17), and the inscriptions of Phno� Sandak and Pra� Vihar speak of a kamrateh an survar�alihga which King Udayadityavarman caused to be erected (Coedes and Dupont, 1943/46, p. 141, line 10). 130 .There can be no doubt that, from the time when the idea became ; current that the king was an Cll[lSa of Siva, the kings of Angkor increasing- ly preferred to name their royal Zingas explicitly as the Zihga of the ; god Siva, and so they forbore to associate their own names with the 36 "dwelled" the "ruler ' s self" belonging to the god, which was lodged in the reigning king of Angkor as his "subtle inner self . " Through this connec­ tion , the king of Angkor became "participant" in the divine rule of Siva . VI . Conclusion It has been the object of the foregoing discussion to call into question the previous explanation of the devaraja cult as a cult of the divinized ruler, and thus to work out with greater clarity the role of the devaraja cult in the legitimating system governing Angkorian kingship . At the same time , a fresh examination of the Sdok Kak Tho� inscription has successfully shown that the devaraja cult can in no way be identified with the cult of the royal lihgas upon the great temple mountains of Angkor . The two should therefore be examined in isolation from each other in order to avoid confusing their peculiar characteristics at the outset . For the interpretation of the devaraja cult , we should appeal only to inscriptions which refer unambiguously to it and not to the cult of the royal lihgas. In the first instance this requirement confines us almost exclusively to the Sdok Kak Tho� inscription . Most of the other sources which have pre­ viously been seen as sources for the devaraja cult 1 3 1 concern the Tribhu­ vanesvara lihga on the Prasat Tho� temple mountain in Ko� Ker , and so do not quali fy as evidence about the devaraja cult . From the Sdok Kak Tho� inscription we learn that the magic ritual enactment of the presumably tantric consecration took place on Mount lihgas. The royal dominion of the "lord of the world" had to be "un­ bounded" if the legitimation of the king as am• sa of Siva was to be seen to be efficacious . 131See Coedes/Dupont , 1943/46a, p . 64 . Still unclear i s the expla- nation of verse VI of the inscription, dating from the years 877 and 879 A .D . , commemorating the foundation of the Pra� Ko temple built by Indra­ varman . In it there is reference to the divine rule (devarajya ) of Mahendra which was established by Svaya.J!lbhu ( Brahma ) . This verse is often seen as an allusion to the foundation of the devaraja cult on Mahendra­ parvata ( Coedes , IC, I , p . 25 , note l ; Dupont , 1952, pp . 171, 175 ; Bhattacharya , 1964 , p . 45 ) . In spite of an ambiguity which is undeniably inherent in the text , the term devarajya is more likely to refer to the "Great Indra" ( Mahendra) whose "dominion over the gods " had been founded by Brahma . By this same rite , the inscription says , King Indravarman ( "he whose protection is Indra" ) was consecrated . After this consecra­ tion Indravarman acquired [ earthly ! ] dominion ( labdha-rajya, verse VII ) . Certainly, there is here in addit ion an allusion to Siva, the "true" king of the gods , and his dominion, in the sense demonstrated by Filliozat . A direct allusion to the foundation of the devaraja cult would on the other hand be difficult to prove . 37 Mahendraparvata. The object of this solemn ceremony was , firstly, to free Camt,odia from dependence on Java and to consecrate King Jayavarman II as sole earthly universal ruler (cakravartin) of Cambodia. Secondly , at the same time that the kingdom - was founded by this solemn rite, the deva­ raja was consecrated . The inscription now leaves no doubt that the king of Cambodia was not , as previously believed, consecrated as god-king . On The contrary , t.his consecration took place around an image of the god Siva, who , as "god, who is king" and as "lord of the world" protected the kings of Cambodia ( as "lords of .the earth" ) . After making an examination of the Sdok Kak Tho� inscription , and appealing to Indian parallels , we may draw the initially hypothetical conclusion that Jayavarman caused a Zihga to be consecrated -on Mahendraparvata, and that a movable image (calanti pratima ) of this was m_ade and venerated as the devaraja in the successive capitals of the kingdom at least up to the eleventh century . The state ceremony of Mahendraparvata at the beginning of the ninth century pre-eminently facilitated th-e achievement of a domestic policy obj ective . According to the recent rese arches of Jacques and Wolters , the solemn consecration was preceded by · decades of struggle by Jayavarman for mastery in Cambodia. By· the "political sacrifice" 1 3 2 on Mount Mahen­ dra Jayavarman elevated himself, after subs:tantial territorial gains , · to the status of sole cakravartin ruler of Cambodia . This explanation of the devaraja cult as a unique poZiticaZ act on the occasion of the terri­ torial unification of Cambodia would make sense of the relatively trivial · significance of the devarcija cult in the following centuries , 1 3 3 when the principle of the unity of the Angkorian kingdom was no longer in ques­ tion . 1 34 There is much to. be said for the supposition that a century later the opening regnal years of Jayavarman IV between 921 and 925 /928 A .D . contributed to a weakening of the devaraja cult . It was the usurper Jayavarman IV who was able to- s-ucceed in setting himself up, without the legitimation of the devarcija , against the kings of Angkor , whose rule the devarcija was supposed to protect . Jayavarman' s challenge must have be_en all the. graver when, in the interim, he publicly and deliberately caused to be consecrated a "rival '.' . s tate divinity in the Tribhuvanesvara lihga 132"The political sacrifices were the concrete mechanisms not only for obtaining the blessings of the gods but also for asserting the polit- ical power of the king . " V . P . Varma, 1959 , p . 222 . 133This necessarily hypothetical contention is based on the striking fact that royaZ inscriptions scarcely mention the devaraja cult at all . The Sdok Kak Thom. inscription and presumably that of Sivacarya, on the. � . . contrary, originate from the family of Sivakaivalya, to whose benefice the devaraja cult is specifically assigned : "Aucun autre texte royal n ' en parle . " [ "No other royal text mentions it . " J Coedes /Dupont , 1943/46a, p . 64 . . · · 134 .Wolters , 1973, p . 30 : "Local independence was no longer the ac­ ceptable objection as it had been in the eighth century. The integrity of the Angkorian kingdom was no longer in question . " 38 upon the then-highest temple mountain of Ca.mbodia.n1 3 5 To this Zihga' s service he committed his sovereignty in order to constitute himself a portion (ClJ?lsa) of this divinity and thus to possess a higher legitimation than the legal kings of Angkor. The idea of participating as an GJ!ISa of the god in divine lordship ( devarajya) decisively affected the apotheo­ sis of the ruler in the Angkorian kingdom from the time of Jayavarman IV on. Thereby the king became a participant in divine lordship, without himself being a god. A further reason for,,. the decline in significance of the esoteric- tantric devaraja cult of Siva, as against the cult of the royal lihgas on central temple mountains, may lie in the growth in the personal power of the kings of Angkor. At the center of the royal temple cult stood lihgas which sometimes bore the names of the kings who endowed them, and which were venerated as abodes of their "subtle inner selves." These indeed constituted the essential royal cult of Angkor. Even in the late period, when Buddhism was already dominant in the Cambodian state cult, the temple mountains were still at the heart of the "state�sustaining" cult. This is what the famous Chinese traveler Chou Ta-kuan reported at the end of the thirteenth century about a ritual union of the king with a snake prin­ cess, the "snake mistress of the country," which lasted through the night.n1 3 6 This union took place within the palace precincts on a "golden tower," in Phimeanakasn. By contrast, the cult potentialities of the calanti pratima of a remote foundation lihgaa, consecrated generations be­ fore, must have paled to insignificancen. Indeed, the devaraja cult was preserved as one of the sources of magic power, but we may presume that its significance was very markedly diminished in the course of time, and came to approximate that of the holy fire ( vra� vleh ) or the holy sword (prae� khan ) in the state cult of Angkor. The fortunate circumstance that the devaraja cult is in its broad outline familiar, through the Sdok Kak Tho� inscription, we owe presumably to the biography of Sadasiva, the last purohita of the devaraja known to us. Suryavarman I (1002-1050 A.D.) married Sadasiva to his sister-in-law, withdrew him from religious functions, named him as his royal chaplain ( rajapurohita ) , and invested him with one of the highest state offices (D, 44-45). Under Udayadityavarman II he became the king's guru and even received the highest royal title dhuli jen vra� kamrateh an and the name §ri Jayendravarmann. In spite of this dizzy career of Sadasiva there is no mistaking that his family had forfeited its monopolistic position as the most important priestly family of the kingdom. 1 3 7 In its stead, the Saptadevakula priestly family, connected to Suryavarman I by bonds of kinship, advanced further and further into the foreground under Suryavarman 135From this point of view one could presumably speak of a schism in the devaraja cult in the sense employed by Briggs (see note 112 above). 136Heine-Geldern, 1930, pp . 37 ff.; Pelliot, 1902, pp . 144 f. 137Briggs, 1951a, p . 150; 1952, p. 178 . 39 .,,. and his successors. The head of this family, the famous San.karapa�iita, became sacrificial priest (hotar) and teacher (guru) of Suryavarman. Under the latter's successor Udayadityavarman he also became the sacrifi­ cial riest (yajaka) of the "golden· Zinga" on the Baphuon temple moun- . tain, f3 8 which (similarly to the Tribhuvanesvara Zihga under Jayavarman IV) stood right at the center of the state cult under Udayadityavarman. The final passing over of the cult in the staten·sanct�ary of the Baphuon seems to have induced Sadasiva, at the peak of his (no longer temporal) power, to establish, in the Sdok Kak Tho� inscription, a lasting monument to the dominant role which his family had played right from the beginning in the state cult of Angkor. However, it was not· through the inscription alone that Sadasiva sought to announce the greatness of his family to future generationsn� In the small temple .in Bhadraniketaria (= Sdok Kak Tho�) in which the inscription was found, Sadasiva endowed a Brahma sculp­ ture for �he priest Hira�yadama and a Harihara sculpture jointly for his .,,.a ncest-or Sivakaivalya and the priest Siva.,,.srama, the co-founder of the . Siva-Asra.ma that was so important to the Sivakaivalya family (C, CXXVII). We must agree with Briggs when he describes the temple of Bhadraniketana as a sort of "family pantheon" and· the inscription as the "swan-song" of the Sivakaivalya family.n1 39 In our present state of knowledge , it seems probable that we should see the endowments of Sadasiva as the swan-song of the devaraja cult as well . After 1052 A.D. , the date of the Sdok Kak Tho� inscription, we hear nothing more of the devaraja cult. Certainly, it was not "abolished" in the twelfth century ; but it was more and more downgraded to the status of a stage prope!tY in the royal apparatus of magic·power. The fate of the caZanti pratima of thea_adevaraja , the sometime state palladium of Ang­ kor, may. have been finally sealed soon after the overthrow of Angkor by the Cha.ms (1177 A.D ; ), when Jayavarman VII came to power and, as a con­ vinced Buddhist, had Angkor completely rebuilt. - It was no longer the old §aivite central temple mountain (VYLGJ!l kantal) of the Bakhen that .stood at the midpoint of the new city of Angkor Thom,·nbut the Buddhist temple com-. . . . . plex of the Bayon. I.,,.n- his inscription Jayavarman declared allegorically that the mountain of Siva in the Himalayas was uprooted (unmuZita)a, so that new kings had recourse to him, Jayavarman, in order to obtain secur­ i.,,. ty. The god Brahma had fashioned Jayavarman out of a half each from .,,. Siva and Vi��u, so that the king thereby surpass ed Siva in majesty and virility. Even more unmistakably, it is said elsewhere that Brahma took from the god Siva the power (sakti) which which he had formerly conquered the Tripura demons and handed it over to King Jayavarman.n1 4 0 .,,. Indeed it .nwas not only in Cambodia that the Saivite ideology of leg'itimation had lost its sakti in the twelfth century. In all the states 138Lo vek 1. n scr1. p ti· o n, s.,,.1o k.a s 37 -43 ; MaJ.umd ar, 195· 3, pp. 423 f. 139Briggs, 1951a, p. 170. 140Inscri. pt i· ons of Prasat Chrun, Angkor Tho!!)- (K. 281, K. 288), IC� IV, pp. 207-50. 4o .,,. of southern and eastern India as well, Saivism as an ideology of state experienced a crisis which was generally connected directly or indirectly with the activities of the great vai��avite reformer Ramanuja. Thus , at the beginning of the twelfth century , King Suryavarman II (ca. 1113-50 A.D.) in Cambodia submitted to the "allure" of Vais• n• avism at the same time as Anantavarman Coga�anga, king of 0rissa (ca. 1112-46)n, in eastern India. Both gave up the Saivite state religion of their forefathers and built their gigantic new state temples in honor of the god Vi��u. So , on either side of the Bay of Bengal, huge vai��avite temples appeared simul­ taneously--the Jagannatha temple of Puri , in 0rissa , and Angkor Wat. In 0rissa and large parts of India , however, Vai��avism increasingly suc­ ceeded during the following centuries in adapting itself to the popular bhakti cults, and thus resolved_ the legitimation crisis in the Hindu realm � whereas Cambodia was no longer affected by this last mediaeval re­ form movement in Indian Hinduism. After the devastatin.,,.g defeat of Angkor in 1177 , Jayavarman VII in- herited a realm in which Siva's mountain of the gods , the Hindu symbol of the temporal dominion of Angkor's earlier kings , was uprooted.n1 4 1 Mis­ construing the "Indian" signs of the times , he adopted the path of the esoteric Mahayana Buddhist Lokesvara cult , and saddled the population of his state with a form of royal apotheosis that was so far unknown. Having covered his kingdom with a network of temples , statues of gods , and hospi­ tals in a frenzy of missionary zeal , he expressed his compassion for suf­ fering humanity in the moving language of his inscriptions . But his words could no longer reach a people afflicted by wars and compulsory labor. The people , exhausted by the burden which the Hindu apotheosis of the ruler and its later- Buddhist form placed upon them , turned to the Cey­ lonese Theravada Buddhism which from the end of the twelfth century began to spread across Burma to the rest of Southeast Asia. 141The reason why the night-long ritual union between the snake princess and the king upon the Phimeanakas , described by Chou Ta-kuan at the end of the thirteenth century , had such a long continuous history may lie in the "pre-Hindu" origin of the snake mythology which can be traced back through the Funan empire to the foundation legend of Cambodia; see Heine-Geldern, 1930 , pp. 37 f. ·ABBREVIATION$ BEFEO Bulletin de Z 'Ecole F'ranr;aise d 'Extreme-Orient IC Inscript.ipns du Cambodge · ( s.ee Coedes , 1937-64 ) JISOA Journal of the Indian Society of Oriental Art JRAS Journal . of the Royal Asiatic Society of Great Britain · :and· Ire Zand JSS · Journal · of the Siam Society · 41 BIBLIOGRAPHY Aymonier, E. 1901 "La stele de Sdok Kak Tho!!l," Journal Asiatique, Ser. 9 , XVII, pp. 5-52. Bechert, H . 1967 Buddhismus, Staat und Gesellschaft in den Landern des Theravada Buddhismusa. Voln. II, BiPma, Kambodscha, Laos, Thai·Zand. Schriften des Institutes fur Asienkunde in Hamburg, voln. XVII, 2 . Wiesbaden . 1969 "Zum Ursprung der Geschichtsschreibung im indischen Kulturbereich," Nachrichten der Akademie der Wissenschaften in Gottingen. I. Philologisch-Historische Klasse (1969), pp. 35-38n. Bhattacharya , K . 1961 Les religions brahmaniques dans l 'ancien Cambodge . Publications de l'Ecole Fran�aise d'Extreme-Orient, vol. XLIX. Paris. 1964 "Recherches sur le vocabulaire des inscriptions sanskrites du Ca.mbodge ," BEFEO, LII, pp . 1-72. 1964 "Hari-Kambujendra," Artibus Asiae, XXVII, 1/2, pp . 72-78. Bosch, F . D . K . 1932 "Notes archeologiques VI .--Le temple d ' Angkor Vatn. a) La proces­ sion du Feu Sacre ," BEFEO, XXXII , pp . 7-11n. Bourg, H. de Mestier du . 1968 "Anpropos du culte du Dieu-Roi ( devaraja) au Cambodgen. En hommage au professeur George Coedes," Cahiers d 'Histoire Mondiale, XI, pp . 499-516. Briggs , L . P . 1951a The Ancient Khmer Empire. Transactions of the American Philo­ sophical Society, New Series, vol. 41 , pt. 1 . Philadelphia. 1951b "The Syncretism of Religions in Southeast Asia, Especially in the Khmer Empire ," Journal of the American Oriental Society, LXXI , pp . 230-49. ,,. 1952 "The Genealogy and Successors of Sivacharya . Suppression of the Great Sacerdotal Families by Suryavarman I , " BEFEO, XLVI, pp. 177-85 . 42 43 Chatterji, B. R. 1964 Indian Cultural Influence in Cambodia. 2nd rev. ed. Calcutta. Coedes, G. 1911 "Note sur l'apotheose au Cambodge," Bulletin de la Corronission Archeologique de l 'Indochinea, pp. 38-49 . 1913a "Etudes cambodgiennes. --IX. Le serment des fonctionnaires de Suryavarman I," BEFEO, XIII, 6, pp. 11-17. 1913b "Etudes cambodgiennes.n--X. Inscription de Prasat Pram," BEFEO, XIII, 6, pp . 17-26. 1913c "Etudes cambodgiennes.--XI. La stele de Palhal,n" BEFEO, XIII, 6, pp. 27-36n. 1931a "Etudes Cambodgiennes XXVI. La date de Kor: Ker,n" BEFEOa, XXXI, pp. 12-18. 1931b "Etudes Cambodgiennes XXVII. La date du Baphuon," BEFEO, XXXI, pp . 18-23. 1932 Le temple d 'Angkor Vat. Avec une introduction de G . Coedes, 3 vols. La galerie de bas-reliefs, troisieme partie. Memoires archeologiques, publies par l'Ecole FranGaise d'Extreme-Orient, vol. II. Paris. 1934 "On the Origin of the Sailendras of Indonesia," Journal of the Greater India Society, I, pp. 61-10. 1937-64. Inscriptions du Cambodge. Vols. I-VII. Collection de textes et documents ·sur l'Indochine. Ecole FranGaise d'Extreme�Orient. Hanoi, Paris. 1951 "La divinisation de la royaute dans l'ancien royaume khmer a l'epoque d'Angkor," Proc. ?th Congress History of Religions, Amsterdam, pp. 141-42. 1952a "Le culte de la royaute divinisee, source d'inspiration des grands monuments du Cambodge ancien,"n·nSeri e Orientale Roma, V, 1, pp. 1-23. Institute Italiano pernil Medio ed Estremo Oriente. Roman. 1952b "The Cult of the Deified Royalty. Source of Inspiration of the Great Monuments of Angkor," A rt and Letters: India, Pakistan and Ceylon, XXVI, 1, pp. 51-53. 1955 "Angkor. Die Hauptstadt des alten Kambodscha--ein Abbild des Kosmos," Saeculwn, VI·, pp. 154-65. . 1961 "Les expression vraJ:i kamraten afi et kamraten jagat en vieux­ khmer," The Adyar Library Bulletin, XXV, pp. 447-60. 44 Coedes, G. 1966 Angkor. An Introduction. Translation of Pour mieux comprendre Angkor. 2nd impression. Hong Kong : Oxford University Press . 1967 The Making of South East Asia. Translation of Les peuples de la peninsule indochinoise. 2nd ed. London. 1968 The Indianized States of Southeast Asia. Edited by W. F. Vella. Translation of Les Etats hindouises d 'Indochine et d 'Indonesie . Honolulu. 1970 "Le veritable fondateur du culte de la royaute divine au Cambodge," R. C. Majumdar Felicitation Volume, pp. 56-66. Edited by H.nB. Sarkar. Calcutta. Coedes, G., and Dupont, P. 1943/46a "L'inscription de Sdok Kak Thom• ," BEFEO, XLIII, pp . 57-134. 1943/46b "Les steles de Phnom Sandak et de Prah• Vihar," BEFEO, XLIII,• pp. 134-54n. Dhani, P. 1954 "The Old Siamese Conception of the Monarchy," The Siam Society Fiftieth Anniversary, II, pp. 160-75. Reprinted from JSS, XXXV, 2. Dupont, P. 1952 "Les debuts de la royaute angkorienne," BEFEO, XLVI, 1 , pp . 119-76. Filliozat, J . 1954 "Le symbolisme du monument du Phnom• Bakheri," BEFEO, XLIV, pp. 527-54. 1966 "New Researches on the Relations between India and Cambodia," Indica ( Heras Institute, Bombay ) , III, pp. 95-106n. Finot, L. 1912 "Notes d'Epigraphie, XIII: L'inscription de Ban That," BEFEO, XII, 2, pp. 1-28n. 1915 "L'inscription de Sdok Kak Tho!!l," BEFEO, XV, pp. 53-106. 1932 "Le Yasodharagiri et lanPhimanakas," Journal Asiatique, CCXX, pp. 47-59- Goetz, H. 1974 Studies in History, Religion and Art of Classic and Medieval India. Edited by H. Kulke. Schriftenreihe des Siidasien-Instituts der Universitat Heidelberg, vol. XVI. Wiesbaden. Gonda, J . 1966 Ancient Indian Kingship from the Religious Point of Viewa. Leiden. Griswold , A. B. 1970 "The Real Foundern.of the Cult of Divine Royalty in Cambodia" ( English summary of Coedes , 1970) ; in R. C. Majumdar Felicitation Volume, pp. 63-64n� Edited by H.nB . Sarkar. Calcutta. Groslier, B. P . 1958 Angkor. Eine versunkene Kultur im indochinesischen Dschungel. Koln·. • 1960 Hinterindien. Kunst im Schmelzatiegel der Rassen. Kunst der Welt. Baden-Baden. Hall, D. G. E. 1966 A History of South-East Asia. 2nd ed .n. London. Heine-Geldern , R. 1930 "Weltbild und Bauform in Siidostasien,". Wiener Beitrage zur Kunst und Kulturgeschichte Asiens, IV , pp. 28.:...78. Hoffmann , H. 1953 "Die Begriffe ' Konig' und 'Herrschaft' im indischen Kulturkreis ," Saeculum, IV , pp. 334-39. Jacques, C. 1970 "Etudes d'epigraphie cambodigenne.--IV. Deux inscriptions du Phnom Bakhen ," BEFEO, LVII, pp. 5 7-67. 1971 "Etudes d'epigraphie cambodgienne.--VI. Surnles donnees chronolo­ giques de la stele de Tuol ta Pee ( K. 834) ," BEFEO, LVIII, pp. 163-76. 1972 "Etudes d'epigraphie cambodgienne.--VIII. La carriere de Jayavar­ . man II," BEFEO, LIX, pp. 205-20. Kroef , J . nM. van der 1968 "Kingship and Political Legitimacy in Southeast Asia : Patterns of an Enduring Tradition ," Asiatische Studien, XXII, pp. 68-87. Kulke , H. 1969 "Funktionale Erklarung eines sudindischen Mahatmyas. Die Legende Hiranyavarmans und das Leben des Co�a-Konigs Kulottunga I ," Saeculum, XX, pp. _412-22. 1970 Cidambaramahatmya. Eine Untersuchung der religionsgeschichtlichen und historischen Hintergrunde fur die Entstehung dera·Tradition einer sudindischen TempeZ.stadt.a· Freiburger Beitrage zur Indologie, vol. III. Wiesbaden. 1970 "Der Sturz Sihanouks--Angkors zweiter Tod?" Internationales Asienforum, I , pp. 325-40. 46 Kulke, H. 1973 "Some Remarks about the Jagannatha Trinity, " Indologen-Tagung 19?1, pp. 126-39. Edited by Herbert Hartel and Volker Moeller. Berlin. 1976 " ' Knsatriyaization' and Social Change in Post-Medieval Orissa,n" Changing India. Studies in Honoura of Ghurye. Edited by S. D. Pillai . Bombay. Le Thanh Khoi 1967 Histoirae de l 'Asie du sud-est. "Que sais-je?" No. 804. Paris. Lewitz, S. 1971 "L ' inscription de Phimeanakas ( K. 484) ( Etude linguistique), " BEFEO, LVI I I , pp. . 91-103. Mabbett, I. W. 1969 "Devaraja,n" Jouranal of Southeast Asian History, X, pp. 202-23. Majumdar, R. C. 1943 "The Date of Accession of Jayavarman II , " Jouranal of the Graeatera India Society, X, pp. 52�55. 1953 Inscraiptions of Kambuja. The ·Asiatic Society Monograph Series, vol. VIII . Calcutta. 1963 Hindu Colonies in the Fara East. 2nd ed. Calcutta. Mus, P. 1932-34 "Barabuq.urn. Les or1.g1.nes du stupa et la transmigration, essai d'archeologie religieuse comparee,n" BEFEO, XXXII ,. pp. 269-439; XXXIII, pp. 577-980 ; XXXIV, pp. 175-400. . · Osborne, M. E. 1966 "History and Kingship in Contemporary Cambodia,n" Jouranal of Southeast Asian Historay, VII , pp. 1-1.4. O ' Sullivan, K. 1962 "Concentric Conformity in Ancient Khmer Kinship Organisation,n" Bulletin of the Institute of Ethnology (Academia Sinica, Taipei), 13, pp. 87-96. Pelliot, P. 1902 "Memoires sur les coutumes du Cambodge, par Tcheou Ta-�uan,n" BEFEO, II,. pp. 123-77 : - Sahai, S. 1970 Les institutions politiques et l 'organisation ailministrative du Cambodge aneien. Publications de l ' Ecole Fran�aise d'Extreme­ Orient, vol. LXXV. Paris. Sarkisyanz, E. 1965 Buddhist Backgrounds of the Burmese Revolution. .The· - Haguet. "Die Neutralit at Kabodschas- -ihre Geschichte und ihr Wes en " Verfassung und Recht in Ubersee, Heft I , pp. 1-12t. Sastrit, K. A. N. 1957/58 The Cult· of Devaraja in Kambuja. Transactions of the Archaeo­ logical Society of South India. Sedov, L. 1969 · "La societe angkorienne et le probleme du mode de pro�uction asiatique,t" Sur le "mode de production asiatique. " Foreword by R. Garaudy. Editions socialest. Paris. Stern, Ph. 1934 "Le temple-montagne kluner. Le culte du linga et le devaraja, " BEFEO, XXXIV, . pp . .611-16 . Stierlin, H. 1970 Angkor. Weltkulturen und Bauform. Munchen. Varma, V. P. 195. 9 Studies in Hindu Political Thoughta. and Its Metaphysical Founda­ tions . ·2nd ed . Delhi • Wales, H. G. Q. 1931 Siamese State Ceremonies. Their History and Function. Londont. 1953 The Mountain of God. London. 1957 Prehistory and Religion of Southeast Asia. London. 1961 The Making of Greater India. 2nd enlarged ed. London. Wolters, 0. W. 1973 "Jayavarman II's Military Power: The Territorial Foundation of the Angkorian Empire,t" JRAS, I, pp. 21-30. APPENDIX NOTES ON THE TRANSLATION OF THE KHMER TEXT By J. Jacob man is not really equivalent to an adverb or adverbial phrase such as "alors , " "damals , " "at that time." It needs to be understood (though not necessarily translated) as a conjunction· introducing a subordinate clause, "whereas, inasmuch as, in that, which, when." On C . 70 "man vrahmaJJ,a . . . mok . . . pi . . . " may be literally translated as "In-that a Brahman . . . c a.me, it-was-because . . .n ." Similarly on D . 12 "man vrah• pada cat . . . , nam• . . . " may be understood as "When His Majesty founded . . . , he took . . n. . " On C. 74 it must be translated as "which.n" leh• C . 7 2 . This word may be understood by reference to Modern Khiner mleh• "so, like that.n" syan C . 74 . In spite of the Modern Kraner meaning "almost," in Old and Middle Kraner this word implies plurality, often linking a remote plural subject to a following verbn. Here it gathers up the list of texts, "all of" which the Brahman recited. kurun, C. 78, occurs as a verb "to govern ." na C . 77. "at, ( place or time), with reference to, in respect of, when, where ." Note that on D . 5 na occurs in combination with anau "at, remain." In occurrences where "kamrateh jagat ta raja" fol lows, it might be translated "in connection with" just as well as by the more precise "aupres dea, " "var" or "in the presence of ." daiy C . 81 . "various(ly), different (ly) ." In this sentence, "daiya" seems to need to be construed as followsn: "the kamrateh jagat ta raja stayed variously/in different places according-to the cities to­ which the lord of the world below, he (particle) took (him) there also ." Note that tadaiy on D . 17 and 25 _is the form which means "other." ncvrz D . 12 . "nctl[l • • • y ak duk . " YlCU!1 "take" usually has a person as object. yak "take" usually has a thing as objectn. "duk" means "put in the proper place, put away, keep." 48