Book Reviews

BOOK REVIEWS

Copper and Bronze Metallurgy in Late Prehistoric Xinjiang: Its Cultural Context and Relationship with Neighboring Regions. Jianjun Mei. BAR International Series 865. Oxford: Archaeopress. 2000. 187 pp, 31 tables, 12 maps, 155 figures, bibliogra phy, £30.00. ISBN 1-84171-068-7. Reviewed by Vincent C. Pigott, Institute ofArchaeology, University College, London

The excavation by Chinese archaeologists cal interaction existed between Xinjiang of naturally mummified Caucasoid indivi and its neighboring regions during the duals dating as early as the second millen Bronze (c. 2000-1000 B.C.) and Iron Ages nium B.C. in the Tarim Basin in Xinjiang, (c. 1000-300 B.C.). His overarching goal China's westernmost province, are of un is to furnish an enhanced understanding questionable import in discussions of the of the archaeological and cultural contexts movements of people across Eurasia in later of late prehistoric Xinjiang, while at the prehistory. The cemeteries in which the same time offering a new perspective on Tarim mummies were found mark what how metallurgy spread into the province, is currently the easternmost presence of and how this technology may have reached ancient Europoid peoples, representatives eastwards into the Chinese heartland. of the Eurasian steppe culture (see Barber In his introductory chapter Mei reviews 1999; Mair 1998; Mair and Mallory 2000). both the background of the development Metal artifacts in copper and its alloys stand of archaeology and the foci of pertinent among the most important possible archae prior research in Xinjiang. Crucial here is ological markers of these wide-ranging the role of external cultural influences in movements. Jianjun Mei, in this publica the development of settlement occupation. tion of his doctoral thesis at the University Moreover, it is clear from Mei's discussions of Cambridge, has opened the door on a that the current wave of archaeological re wealth of hitherto uncirculated archaeo search, much of it from Eurasia during the logical data both on the archaeology of 1990s, has altered traditional thinking (even Xinjiang province and on the coming of among the Chinese) about the develop copperjbronze to this geographically and ment of Chinese civilization as an exclu culturally pivotal region ofdesert and oases. sively indigenous process, especially where Mei seeks answers to four research ques metallurgy is concerned. This theme under tions: (1) when, where, and how copper pins discussion throughout the volume. and its alloys began to be used, (2) what Mei divides the remainder of his publi metallurgical technologies were employed, cation into three major components. In (3) what the cultural context was for the the first component (Chapters 2 and 3) he beginning and early use of metals, and (4) reviews the archaeological evidence for at what cultural connections and technologi- least fourteen Bronze Age cultures and a

similar number of Iron Age cultures from socketed dagger-axes, and handled mirrors the various regions ofXinjiang (Chapter 2). as well as items in precious metals, silk, and He undertakes next a substantive typol lacquer strongly suggest that Central Asia ogical investigation of six major categories was linked by trade routes through Xin of metal artifacts from these periods, e.g., jiang with northwest China, i.e., the Gansu implements, weapons, harness and chariot corridor. It is in Gansu that some of Chi fittings, vessels, toilet articles, and orna na's earliest copper-base artifacts have been ments (Chapter 3). excavated (see Linduff et al. 2000). The second component (Chapters 4 and In Mei's detailed discussions of the ty 5) consists of a review ofprevious analytical pology of copper-base artifacts, there are research on Xinjiang metal finds as well repeated references to the similarities ap as an analytical program focused on metal parent between artifacts excavated in Xin samples supplied to him by local archae jiang and those of known Andronovo type ologists. Chapter 4 presents the results from from neighboring Eurasian locales. These the battery of analyses he performed to ob categories include shaft-hole axes, sickles, serve microstructure and determine com flanged adzes, and socketed celts. Through position of 58 metal samples with a goal of these artifact and site-specific discussions, comparing the technologies of the various Mei offers persuasive evidence for the im cultural groups he has identified. In Chap pact of Eurasian steppe culture on Xin ter 5 his investigation becomes site specific jiang. Nor can we ignore indications, and focuses on the important finds from though currently based on less substantive the mining and smelting site of Nulasai in evidence, of contact between Xinjiang and Nileke which, on current evidence, appears bronze-using cultures to the east. to date to the first millennium B.C., but Mei, who is trained in metallurgy as well may well be earlier. This site, with its un as archaeology, conducted his own labora common finds of mines and associated pro tory analyses. They give us the first glimpse duction debris, is one of but a handful of of the levels of sophistication attained in such documented sites currently known the metalworker's craft as well as the mul across Eurasia. tiple, alloying traditions being employed. In his final component (Chapters 6 and Copper and tin-bronze artifacts are present 7), Mei turns to a synthetic overview of in Xinjiang from the early second mil cultural interrelationships between Xin lennium B.C. while, interestingly, arsenical jiang and regions to the east, west, and copper doesn't seem to appear until the north. He concludes with a discussion of later centuries of this period. In the early the development of copper and bronze first millennium B.C. artifacts in copper ap metallurgy in the region. Significant new pear with more frequency due to what Mei archaeological data, much of it from Chi suggests is the exploitation of local copper nese sources, is presented in this volume, in deposits near Urumchi. Thus, when com particular that concerned with widespread pared to western Asia, copper and its alloys contact between Xinjiang and neighboring appear relatively late in Xinjiang and ap regions. The initial occurrence of artifacts parently not in the more time-honored in copper (at Gumugou) and tin-bronze (at sequence of copper, then arsenical copper Tianshanbeilu) in Xinjiang takes place in followed by bronze as seen in the Near the early centuries of the second millen East, Central Asia, and Eurasia. The some nium B.C. what jumbled Xinjiang sequence, in Mei's Turning to the Iron Age, this period is estimation, reflects the introduction of tin marked by major changes including not bronze and perhaps even arsenical copper only the coming of iron, but also the artifacts and/or metallurgy from outside the increasing use of gold and silver, and the region followed by attempts to produce practice of horse nomadism. However, as metal locally. On the more technical side, indicated by the unique evidence from Mei argues that the presence of sulfide Nulasi, copper mining and smelting con inclusions in the microstructure of artifacts tinued unabated. Bronze artifacts, including from the Tacheng region suggests that BOOK REVIEWS

copper sulfide ores were being smelted. and into northern Xinjiang. Increasing He adopts the traditional 'matte' smelting contact with the Chinese heartland was model involving the roasting of sulfide ores occurring in the late first millennium B.C. prior to smelting to explain the production as marked by the presence of silk, lacquer, of the Nulasi ingots. In future research he and mirrors. Cast iron and its technology might also consider the possibility of the reaches Xinjiang from the central plains of co-smelting of sulfide and oxidic ores China at this time as well, brought perhaps directly to copper in a one-step produc by the Saka. tion process without roasting. Research by What Jianjun Mei has achieved in this William Rostoker and colleagues (1989; volume is a unique synthesis, from a variety Rostaker and Dvorak 1991) and that by of mostly new sources, of the critical infor Heather Lechtman and Sabine Klein (1999) mation concerning copper-base metals and has introduced co-smelting as a highly fea metallurgy in Xinjiang from the point of sible alternative to the matte process and initial appearance shortly after c. 2000 B.C. one which can yield arsenical copper. down into the Iron Age. But this volume When Mei turns his attention to east is much more than a study in a single of Xinjiang, and to Bronze Age cultural technology, it is a harbinger of continuing contact with the Gansu-Qinghai region, he revelations concerning the complex later sees not only an influx of painted pottery prehistory of Eurasia. Mei's research con into Xinjiang from this region, but also cretizes the crucial role played by Xinjiang looks at the presently modest evidence for in the transmission of cultural and tech the spread of copperjbronze metallurgy nological traditions both East and West. into Gansu-Qinghai from external sources. He states rather decisively that "one thing One potential source is the possible in appears quite clear: Andronovo expansion teraction between the Machang (Gansu played a vital role in the transmission of Qinghai) and Afanasievo (southern Siberia) copper and bronze technologies in Eurasia cultures in eastern Xinjiang. Arsenical cop during the second millennium B.C." (p. per appears late in both regions, but it is 74). Furthermore, Mei's study does noth not clear ifit has any direct links to Central ing to dispel the suggestion that this rapid Asian-Eurasian traditions. cultural expansion may well have had an Iron appears c. 1000 B.C. in Xinjiang, influential role in introducing metal and but the tradition of copperjbronze metal perhaps metallurgical technology into north lurgy continues with new forms being west China and ultimately to the Chinese introduced during the Iron Age. It is from civilization ofthe central plains. this period that the one major documented In the Foreword, Colin Renfrew, Mei's production site (Nulasai) comes with its academic advisor at Cambridge, praises this rare find of smelting remains including five volume as a "pioneering work," "the first plano-convex ingots (an unusual copper coherent study of later prehistory in this arsenic-lead alloy). Mei suggests that the region," and the "first in any language" to high arsenic content may have been an detail the region's metallurgical evidence. intentional addition and not the result of Moreover, this newly opened window on smelting arsenical copper ores. what was transpiring technologically in It would be of particular interest if, as Xinjiang, a geographical and cultural "shat suggested by scholars, Nulasai was a site ter zone" between East and West, is crucial which lay within the Saka people's sphere to understanding issues of fundamental ofinfluence, but on current evidence this is interest to that substantial archaeological a difficult attribution to make. The Saka community whose interests lie in the Asian may have been responsible for the produc Old World. tion of a certain cauldron type, which Mei This volume is readable, rich in informa identifies and, given that all Xinjiang caul tion and has numerous tables, maps, and drons are made with Chinese-style piece figures. If it is read in concert with several molds, this technology may have spread recent studies of Asian metallurgy, e.g., from China through the Mongolian steppe Linduff et al. (2000) on China, Chernykh 170 ASIAN PERSPECTIVES 41 (2) . FALL 2001

(1991) on Eurasia, Pigott (1999) on South LECHTMAN, HEATHER, AND SABINE KLEIN west Asia, Agrawal (2000) on South Asia, 1999 The production of copper-arsenic and Higham (1996) on Southeast Asia and alloys (arsenic bronze) by cosmelting: Modern experiment, ancient prac the relevant papers in Mair (1998), a far tice. Journal of Archaeological Science clearer understanding of the development 26: 497-526. and spread of metallurgy across the vast LINDUFF, KATHERYN M., HAN RUBIN, AND SUN geographical expanse that is Asia can now SHUYUN be achieved. Its title understates somewhat 2000 The Beginnings of Metallurgy in China. the wealth of archaeological information it Lampeter: The Edwin Mellen Press, contains as this volume is more than a study Ltd. ofmetals and metallurgy-it is an insightful MAIR, VICTOR H. view of technology in cultural and histori 1998 The Bronze and Early Iron Age Peoples cal context-one which will certainly be of Eastern Central Asia. 2 vols. Wash ington, D.C.: Institute for the Study widely read. of Man, and Philadelphia: University of Pennsylvania Museum. MAIR, VICTOR, AND J. P. MALLORY REFERENCES CITED 2000 The Tarim Mummies. London: Thames AGRAWAL, D. P. and Hudson. 2000 Ancient Metal Technology and the Ar PIGOTT, VINCENT c., ed. chaeology cif South Asia. Delhi: Mun 1999 The Archaeometallurgy of the Asian Old shiram Manhorilal Press. World. MASCA Research Papers in BARBER, ELIZABETH WAYLAND Science and Archaeology No. 16. 1999 The Mummies of Urumchi. London: Philadelphia: The University Mu Macmillan Publishers Ltd. seum, University of Pennsylvania. CHERNYKH, EVGENY ROSTOKER, WILLIAM, AND JAMES R. DVORAK 1991 Ancient Metallurgy in the USSR: The 1991 Some experiments with co-smelting Early Metal Age. Cambridge, UK: to copper alloy. Archeomaterials 5 (1) : Cambridge University Press. 5-20. HIGHAM, CHARLES ROSTOKER, WILLIAM, VINCENT C. PIGOTT, AND 1996 The Bronze Age ofSoutheast Asia. Cam JAMES R. DVORAK bridge, UK: Cambridge University 1989 Direct reduction to copper metal Press. by oxide/sulfide mineral interaction. Archeomaterials 3 :69-87.

The Mons. A Civilization oj Southeast Asia. Emmanuel Guillon. Bangkok: The Siam Society, 1999. 349 pp. ISBN 974-8298-44-2. Reviewed by Tilman Frasch, South Asia Institute, Heidelberg University

It does not seem to be too far from truth gap by writing the history of the Mons. to state that the Mons have received rela In fact, both the author's reputation as a tively little scholarly attention so far, if leading scholar of Mon language, and the compared to other Southeast Asian peo format of the book (349 pages, foolscap ples. For whatever reason that this may size, with many colored illustrations) sug be, the present study of the French philol gest that we have a reference work in hand. ogist Emmanuel Guillon represents the first The book consists of two parts. In the large-scale attempt to make good for this first part, Guillon approaches the phenom-

enon "Mon" by describing the components Research on early Burma published after that shape their identity: language and 1980 is completely ignored; even though script, ethnicity and belief system. The sec studies by Michael Aung-Thwin, Janice ond part, entitled "A long history," covers Stargardt, and the present reviewer have, the period from the formation of the first each one in his (or her) own way, con polities in the third millennium B.C. to the tributed to a balanced view of the Pyus and present, including the first blossoming of Mons and their respective influence upon Mon culture in central Thailand (the king Pagan, by making thorough use of the dom of Dvaravati), the "classical Mon pe available epigraphical, historiographical, and riod" during the Pagan times, and the Mon archaeological data. Instead, Guillon main states of Lower Burma between 1281 and ly quotes from Luce who-with full re 1754; not to forget the Mon renaissance spect to his pioneering research on early in seventeenth-century northern Thailand. Burma-seems to have overestimated the Five appendices provide additional infor contribution of the Mons to the Pagan mation on palaeography, chronology, a kingdom. Quite paradoxically, Luce is short history of Mon studies, and two glos treated critically only on occasions when it saries on Buddhist texts and Mon words. appears to be unnecessary or even wrong. The author has used a wide range of writ Thus, Sudhammapura is translated as "Great ten texts including Mon historiographies Assembly Hall of the Gods of the Heaven and inscriptions, the oldest of which date of the Thirty-three" (p. 104) against Luce's back to the sixth century A.D. version "City of the Good Law" merely The great expectations that are roused by because of the long a. If the lengthening the outer appearance of the book are how of vowels had indeed to be taken that ser ever hardly met with by its contents. Basi iously, a good number ofministers in Pagan cally, the book has three major weaknesses. would have been female, as their names First, it is, by and large, a compilation of end with long a (Satya, Asankhya, etc.). dates and facts on Mon language, art, and Obviously, the long vowel is euphonic, culture without any theoretical reflection and Thaton/Sudhammapura can still be or methodological concept. This is cer considered as the "City of the Good Law." tainly not enough for a study that under Third, the author displays a noteable lack takes to provide a general history of the of language skills whenever he ventures Mon people. into languages other than Mon. This refers Second, the book is quite outdated. The mostly to Sanskrit/Pali and Burmese. To original French version of the manuscript give a few examples: the suffix "-dev!" in was finalized in 1969, but apart from those royal titles is always translated as "divinity" few areas which were among Guillon's instead of "queen," which is very common main research interest (i.e., Mon epigraphy in Sanskrit and Pali. A complete perturba and language), it has hardly been updated tion occurs on page 32 where ci1khi and since. On page 170, for example, it is stated cariy are described as "two kinds ofscribes." that the Buddha reached parinirvana in the The Burmese word cakhi indeed denotes year 543 according to the tradition of the the position of a scribe or clerk, but cariy southern Buddhists, in contrast to Sanskrit (lit. ca "letter" and riy "to write") usually sources which put it between the years 478 occurs in formulas such as ci1 riy so sa ka "he and 486 B.C., a statement that completely who wrote these letters." It is a mere de ignores the relevant writings of Bechert on scription and has nothing to do with the this topic (apart from being wrong insofar office or position of a scribe. In the foot as the 486 era was a creation of Indologists note, moreover, cakhi is linked to Pali sak who tried to reconcile the 544 era with the khi ("witness"), which also occurs in early date of the Indian king Ashoka). Or look at Burmese inscriptions, but as saksiy, which what is said about the kingdom Pagan, the is a direct derivation from Sanskrit, saksin. early phase of which is earmarked nothing Equally untenable is the derivation of the less than the "classical age" of the Mons. Burmese word mim ("king, ruler") from ASIAN PERSPECTIVES 41 (2) • FALL 2001

Mon smin (p. 160). This misinterpretation emerges after reading the book is for what is obviously based on the pronunciation of reason or what purpose it may have been man as "min." While man is a pure Burmese published. Full of mistakes, outdated, and word again, smin seems to be related to barely original as it is, it can hardly be con Sanskrit samin ("lord, master"). Similarly, sidered as a standard work on a "Southeast the remarks on Burmese epigraphy and Asian civilization," as stated in the title. historiography (p. 96) cannot be substan Nor can we regard it as a political mani tiated at all. This list could be continued for festo, as the relevance of Mons for the a while, and moreover become extended modern states is not discussed, be it their by a list of typing errors such as arafifiasi in contribution to the early Thai states or stead of arafifiavasi, the forest monks (p. 79 their situation in modern Burma. The an and passim), and in the glossary of Mon noyance about a book that is expensively words in Appendix A. made but faulty, poorly written (and trans Finally, it has to be noted that the book lated?), and unnecessary at the time, soon was arranged in a rather thoughtless way: gives way to the simple insight that the expressions such as "I shall come back to history of the Mons remains to be written. this point," "more about this later," "I shall Or, in Guillon's words: More about this have to say more about this," and even the later. announcement to tackle a certain problem Editor's note: A German language ver in a separate monograph (p. 164) occur on sion of this review was published in Peri several occasions-it may be 30 or more plus: Jahrbuch fur aussereuropaische Geschichte, and run like an unbroken thread through volume 10,2000, pp. 219-221. the text. The question that immediately

Land of Iron. The Historical Archaeology of Luwu and the Cenrana Valley. David Bulbeck and Ian Caldwell. 2000. Hull: Centre for South-East Asian Studies, University of Hull. 141 pp, 7 figures, 8 maps, 9 tables. ISBN 0-903122-08-1. Reviewed by Ian C. Glover, Institute ofArchaeology, University College, London

The authors describe this as the second, research findings in the light of the his but still preliminary, report on the joint torical traditions with suggestions for a new Australian-British-Indonesian OXIS proj chronology based on radiocarbon dates and ect on the Origin of Complex Society in an analysis of the imported, largely Chinese South Sulawesi, and it includes short re ceramics found at many of the sites; (4) ports on 29 excavated sites, and 56 sites Four appendixes listing the ceramic finds which were surveyed in the ancient king and forty-two new radiocarbon dates; and dom of Luwu at the head of the Gulf of (5) A six-page bibliography, useful in its Bone, Sulawesi, in central Indonesia. own right since the abundant literature on The slim volume contains five sections: the cultures and history of South Sulawesi (1) A historical account of the kingdom of is not always easy to track down as it occurs Luwu based on oral traditions, especially in Dutch, Indonesian, French, English, and the epic 'La Galigo' cycle, and the histori German, not to mention sixteenth- and cal lontara texts for which South Sulawesi seventeenth-century descriptions in Portu is justifiably famous; (2) A summary of the guese. field research program in different parts of The volume is perhaps best read III ancient Luwu; (3) An interpretation of the company with the book, The Bugis, by

the French anthropologist Christian Pelras 3. That another, more southerly king (1996) on which in some ways it is a dom of Cina located near the mouth of the commentary. Pelras advanced a proposi Cenrana River was either a trade-based or tion, developed on the basis of a reading of agrarian kingdom (or both?), later absorbed the lontara and analyses of the La Galigo for some time by Luwu. epic cycle by many people including him Fieldwork was started in 1992 with an self and Ian Caldwell, that Luwu was the extended survey and trek across Sulawesi earliest of the historical Bugis kingdoms by Caldwell when he recognized many ar and its court, from the fifteenth century, chaeological sites, which were identified by was the source of much of Bugis elite cul local informants with places known in the ture and traditions. Luwu and its succes lontara or from oral traditions. Further sur sors were essentially agrarian kingdoms in veys between 1994 and 1997 were fol which external trade was important but not lowed by an excavation season in 1998, the the main source of the wealth and power scope of which was curtailed by unusually of the rulers. In contrast, the La Galigo bad weather. cycle tells of an earlier society-perhaps Excavation consisted essentially of small, between the twelfth and fourteenth cen 1-m-sq test pits, sometimes enlarged to 2 tury A.D.-with divinely descended rulers by-1 m, and sometimes several to a single centered in the northern part of what is site. Deposits were sieved and soil samples now southwestern Sulawesi, in which the retained for more careful examination, and political economy depended on maritime charcoal samples were taken where avail trade with other parts of the Indonesian able. It could be argued that many small archipelago. The La Galigo society came pits-the 'telephone box' strategy-are to an end in the early fourteenth century quite inadequate to give a reasonable sam and after a period of anarchy, the 'Age of ple of surviving remains at heterogeneous Chaos' mentioned in later Bugis chronicles, living sites. This is probably true, but the a number of new, sago and rice-based abundance of local and imported ceramics agrarian kingdoms arose, of which Luwu together with some iron, bronze, and gold was the first and for long predominant. artifacts and ornaments, iron slag, chert In the aXIS project Bulbeck and Cald flakes, glass and stone beads, has enabled well are attempting to test this reconstruc the authors to develop rough chronologies tion through excavating and dating ancient for most sites using radiocarbon dates and settlement sites in and around Luwu-an ceramic histograms based on the frequency, ambitious attempt and perhaps the first sus layer by layer, of identifiable and dateable tained such piece of research in Southeast imported sherds. Asia aiming to integrate historical and ar Drawing on the wealth of archaeological chaeological data within a closely defined material, the authors feel able to comment region. Additionally, they have examined on, and sometimes modify earlier historical coastal-hinterland relationships around the reconstructions of South Sulawesi. Gulf of Bone within the framework of First they suggest that as an ethnic and Bronson's well-known (1977) model. To linguistic group the Bugis are relatively late this end they set out (pp. 14-15) a number arrivals, not earlier than A.D. 1300 in the of hypotheses, which can be summarized region north of the Gulf of Bone. Further thus: they find no archaeological evidence for 1. That the Bugis kingdom of Luwu was the fourteenth-century 'Age of Chaos,' founded at Malangkene by the fourteenth which figures in Bugis oral traditions, but century A.D. on the basis of earlier, but not quite the opposite; that the period saw an perhaps Bugis occupation of the region. increase of overseas trade and the rise of 2. That the presence of high-grade iron powerful kingdoms of which Luwu was ore, rich soils, and dammar (fossil resin) de one of many, but not the earliest. One re posits in the region were sufficient to en markable finding was of buried brick struc sure the prosperity of the kingdom. tures at Malangkhe, which seem once (now 174 ASIAN PERSPECTIVES FALL 2001 looted) to have been associated with Maja in the text related to the maps, which pahit-Javanese-style gold arm bands, brace themselves are clear enough. Some pointers lets, earrings, and eye covers, and a great in the text would have helped here. variety of ceramics. It is quite possible that this was a classic Javanese religious and burial site of the fourteenth century; and REFERENCES CITED if so, it is the first to be recognized in Sulawesi. BRONSON, B. There are many other points of interest 1977 Exchange at the upstream and down stream ends: notes towards a func in this brief, but information-packed re tional model of the coastal state in port, which is surely a model "preliminary Southeast Asia, in Economic Exchange publication," lacking only in photographs and Social Interaction in Southeast Asia, of the archaeological sites and material. ed. K. 1. Hutterer. Ann Arbor: Uni Looking for something to criticize I found versity ofMichigan. that it was often quite difficult to find early PELRAS, C. sites and geographical features mentioned 1996 The Bugis. Oxford: Blackwell.

Prambanan: Sculpture and Dance in Ancient Java. A Study in Dance Iconography. Alessandra Iyer. Bangkok: White Lotus Press, 1998. xii, 211 pp. ISBN 974-8434 12-5. Reviewed by Astri Wright, Department ofHistory in Art, University if Victoria, British Columbia, Canada

Alessandra Iyer's Prambanan: Sculpture and ter describing the Prambanan temple com Dance brings together under one cover ela plex, its main temple Chandi Shiva with its borations on several of the author's articles Ramayana and dance reliefs, interspersed on dance in Java published in academic with reliefs of deities and auspicious sym journals since the mid-1990s. This small, bols, and a discussion of the history of compact, and attractively packaged book restorations; a third chapter discussing the organizes the results of Iyer's doctoral and dance reliefs and their dance iconography post-doctoral research on the history, ico in greater detail, giving a more detailed nography, and practice of dance in Java introduction to the Indian Natyasastra's into a consecutive narrative. Drawing on a karana dance sequences as tentatively re combination of South and Southeast Asian stored by Indian scholars; a fourth chapter art history, dance history, and recent theory discussing the Natyasastra's (sanskrit dance in dance studies in general, Iyer presents treatise) possible presence in ancient Java, her material here to a broader public, with textually or dance-practice-wise, and a short interests (both academic and applied) in fifth chapter with concluding remarks. dance research, anthropology, and art his This text, interspersed with photographic tory in Southeast Asia and beyond. and line-drawn images, is followed by Part In just over 160 pages, divided into 2, which begins with a sixth chapter that Part 1 and 2, Iyer offers the following. is really a catalog of the 62 dance reliefs. Part 1 includes an introductory chapter These are located on the outside of the of disciplinary reflections and theoretical Chandi Shiva balustrade (on the inside of methodological navigations; a second chap- which is carved the Ramayana series of

Asia" Perspectives, Vol. 41, No.1 © 2002 by University of Hawai'j Press. BOOK REVIEWS 175 reliefs). In this section, Iyer gives us relief of the presence of Indian dance styles in by-relief photographs and descriptions, ancient Java rather than for its own intrinsic each matched up with its corresponding value" (p. 7). While this seems a somewhat karana number and line-drawing taken reductionist view of (among others cited) from Subrahmanyam's 1978 reconstruction Claire Holt's varied and pioneering inter of the Natyasastra's sections on karana. This disciplinary contribution to dance studies, is followed by an Appendix in which Iyer's Iyer then proceeds to discuss what to this analysis is summarized in note form in reader still remains unresolved in her own parallel columns; footnotes to the text; a book: the theoretical-methodological de glossary of dance terminology; and the bate around "Indianization" versus a model bibliography. of local assimilation of foreign (including Iyer's analysis begins with the riddle Indian) influences. (Iyer's footnotes here, noted by nearly everyone who has written as throughout, give a good orientation to about the Prambanan over the last sixty newcomers to the field on the different years: what might have been the sources of stages in these debates.) inspiration and practice depicted in the re While Iyer distances herself from the lief series of dancers and musicians on the 'Indianization' view of Southeast Asia as outer balustrade of Chandi Shiva (Shiva passive recipient to Indian influences in temple), most of them suggesting the difli statecraft and cultural forms (and no stu cult-to-substantiate relationship between dent-scholar in this age of post-colonial the Natyasastra (an ancient Indian body of analysis could fail to make such a gesture), teachings about dance techniques, postures she also critiques the "localisation of Indian and iconography put· into writing some influence" view as too passive and polar time between the first century B.C.E. and ized a model. Good, so far as the theory the fourth century C.E.). Iyer dismisses ear is concerned. But what happens in the rest lier scholars' positings of this very connec of her analysis, as it rests on her choices tion as intuitive but unsubstantiated by of methodology? Iyer's discussion is based quantifiable analysis (p. 99), even while almost entirely on recent results of the last noting that Suhamir, in his 1948 report on thirty years of Indian dance scholarship the completed Dutch-initiated restorations (and in particular that of the author's of the Prambanan, knew enough about this own Indian mentor, Padma Subrahmanyan, Indian dance tradition to identify the reliefs with whom she traveled to Prambanan in as depicting angahara, sequences involving 1994). Leaning on Subrahmanyan's pub more than one karana (p. 13). lications of the late 1970s, Iyer creates a Then, through an intermittently acro closer marriage between ancient Indian batic analysis that rests on a painstaking dance traditions and Javanese sculptural adaptation of Indian dance scholar Padma reliefs than this branch of scholarship has Subrahmanyan's reorganization of the Nat ever encountered before. At the same time, yasastra to the reliefs in Java, Iyer concludes she also presents data that allow the reader that these Chandi Shiva reliefs (their origi to study the full series of dance reliefs and nal sequence also no longer identifiable), to compare these with the line drawings can "unerringly be identified as sculptures of Indian reconstructions of that tradition. showing karana, the units of dance move Hence, while Iyer's book offers new in ment described in the Sanskrit text on traregionally comparative possibilities, her dance and drama known from India as analysis maintains the old India-Java axis, Natyasastra" (p. 11). demonstrating mainly a one-way flow of In her introduction, Iyer writes how, influence. The one possibility she cites as upon undertaking her research, she realized perhaps pointing to influence flowing the that she was "studying an obsolete non other way, is not developed and remains Western dance form which, if at all, had airily hypothetical (see below). only elicited interest in past times in terms Within Indonesian dance studies, most of finding out whether it was an indication of the early writers on dance in Indonesia ASIAN PERSPECTIVES 41 (2) . FALL 2001 focused on dance in Bali. Claire Holt was there are no references to individual draw among the first to study dance in Java and ings in the text. Indonesia and to investigate sociohistori This confusion is due to a major editing cal dimensions that link indigenous and and layout oversight. After much cross imported traditions, and sculpture, dance, checking, the drawings are, in fact, revealed and aesthetics. While Holt has a fairly to be illustrations of the karana discussed lengthy discussion of the reliefs of dancers in text section 3.4.1 (Karana descriptions); and dance-scenes at the Borobudur (1967), however, the existence of the line drawings she does not touch on the Prambanan that follow twenty pages later is not sig dance reliefs in detail. This textual focus naled in the written text. Furthermore, the is illustrated by the inclusion of only three small 'catalog' of line drawings have no photographs of dance reliefs from the Pram subheading of their own either in the banan: a reproduction of a dance scene of a table of contents or on the page where woman performing a sword dance in a they begin. Finally, the drawings are also court scene, as part of the Ramayana series numbered according to a second system, (Holt 1967, pI. 102, p. 120), a second pho according to their karana number, all of tograph showing a drummer dancing with which makes for a disorganized reading ex two other dancers (frontispiece), and a perience. third photograph from the outer balus Since dance studies of non-Western trade's dance reliefs (Holt 1967, pI. 41, traditions constitute the lesser part of that p. 57). These last two reproductions show literature, historiographically Iyer's book the reliefs numbered by Iyer as P6 and P55, represents a welcome contribution to a respectively (Iyer, pp. 111, 159). However, small field. Through her detailed studies one of the two publications has reversed and analysis, Iyer is able to tie the Chandi the reproductions of these reliefs-and Shiva dance reliefs more closely to the In the existence of the Holt reproductions is dian sanskritic text tradition than anyone strangely not mentioned in Iyer (the fron has before and perhaps this reflects more tispiece photograph in Holt belonged to closely a historical reality than what we the Dinas Purbakala, the Indonesian Ar have known in this subject area und now. chaeological Service; the one on page 57 However, this analysis has been done was photographed by Holt herself). Fur without the author adding any perspectives thermore, this comparison higWights how from what one might expect to be an ab poor the quality of the black and white solutely necessary component in her re reproductions of Iyer's own photographs of search-reconstructions (in texts, image, or the dance reliefs are, perhaps a signal to contemporary choreography) of indigenous White Lotus to improve their reproduction (here, Javanese) dance traditions and ico technology. nography. Similarly, it is stated clearly that Another issue, which relates both to the the Glossary of dance terminology (pp. author and the editors at White Lotus 193-198) includes only those Sanskrit dance Press, is the occasionally confusing organi terms in the Natyasastra that Iyer has iden zation of the book. The most blatant ex tified as relevant to the Prambanan dance ample is in Chapter 3. While the text is reliefs. Rather than an informed challenge organized into subsections, numbered 3.1 from a dance historian, the following is a to 3.5, a ten-page section of the line draw sincere question from someone rooted in ings of dancers illustrating different karana an interdisciplinarily informed, culture (taken from Subrahmanyan's publications, specific art history: Is there, then, no Java pp. 73-82) are also numbered with the nese dance terminology, as distinct from same numbering system (starting with Fig those in Sanskrit, used in Java in historical ure 3.1). This is done without there being texts or even today (which are linguistically any apparent correspondence between text archaic) to describe any movements identi and image bearing matching numbers, and fiable in the reliefs? BOOK REVIEWS 177

Perhaps the latter kind of material does pany it. The drawings on pages 73-82 do not exist. But any sense of a social history provide a lively and communicative break of dance in Java is missing. Hence, I cannot from the static nature of print text. Since help but wonder how Iyer's study would Iyer has been experimenting with digital have differed had she traveled to Pramba video-imaging, perhaps this is a next stage nan with other mentors at other times, in her research. mentors from Java, Bali, perhaps Cambodia While Iyer claims dance history as a and Thailand, perhaps even from parts of branch of art history, she herself displays the Oceanic world. I cannot help wonder limited art historical tools at times, such as what her conclusions would have been had when interpreting the famous early four she undertaken comparative local studies teenth-century statue ofKing Krtanagara as of tribal-primal dance forms, as far as they a fusion of Siwa and Buddha (p. 69) when still exist, alongside the studies of the more it is well established as a hari-hara (Siwa and obviously 'internationalized court cultures' Vishnu) image and simultaneously an ard dance traditions. The fact that many karana hanarisvara (Lord who is Half Man Half imitate animal movements or are inspired Woman) image. In addition, when chal by elements of nature (p. 89) would seem lenging the old theory that the three struc to point to a link, at a deep historical (in tures facing the three Prambanan temples deed, archaeological) stratum, between a once housed the vahana (vehicles) of each primal way of thought and ritual and the temple's main deity, Iyer cites the find of high-Sanskrit brahmin literary culture that 'a solar disk' opposite the Vishnu temple produced the written versions of the Nat as more likely evidence of Surya than of yasastra and other classic texts. Mter read Vishnu. Here, she (and the unnamed staff ing this book, I cannot help wonder at member at the Prambanan Archaeological how text-based many historians, including office she cites) misses the fairly basic point art historians (which might or might not that a circular disk (a weapon) looks very include dance historians), still are. similar (and at times identical) to the sym Adding first-person experience of cul bol of the sun, and such a disk is one of turally relevant kinetic movement to the Vishnu's mam iconographic attributes. art historical study of sculpture and space, Vishnu is classified a deity of solar lineage, urban and rural, interior and exterior, and as such complements his theological would greatly enrich insights into writings rival Shiva who is of lunar lineage and about a culture like Java's where both powers. dance and an appealing personal body This is not to say that dance would not language for both men and women is so be a major part of a culturally contextual central. The way the discipline of art his ized Indonesian art history more free of tory is currently constructed and practiced, European biases and outmoded disciplinary however, dance history, dance icono limitations than it is today (or that a solid graphy, and movement analysis occupies no knowledge of sculptural-mythological ico more than a tangential relationship to it. nography shouldn't be part of a dance Indeed, it would appear that dance history scholar's tools). Holt's Art in Indonesia: stands on its own methodological feet as Continuities and Change (Ithaca, NY: Cor much as does the history of music, theatre, nell University Press, 1967), divided into literature, and other areas of human en three parts (The Heritage, Living Tradi deavor that together constitute the arts of tions, and Modern Art, followed by rich humanity but are generally not included appendixes of translations of inscriptions, under the rubric of fine arts. In the same scenes from epics and shadow puppet plays) way that music needs the aural text to dis still stands out as a pioneering (even pre play its data, I would argue that this book scient) work of post-colonial art history needs a video of dancers offering various that has not been matched or recognized possible interpretations of karana to accom- for its contribution to theories and meth- ASIAN PERSPECTIVES 41 (2) . FALL 2001

odologies developed (and still developing) could arrive at a meaningful hypothesis of three-and-a-half decades after its publica one likely scenario. tion. Among national or regional histories The other interesting hypothesis of Iyer's ofart and cultural expression within South to follow up on is the fact that, if the dance east Asia, Holt's remains a lone monument reliefs indeed are dated to around the same to a culturally embedded, multimedia and date as the construction of the temple, they multivocal art history. are one to two hundred years older than Like Iyer, Holt was trained as a dancer the earliest representations of a karana series and knew the various aesthetics of different found in India today, the earliest known Indonesian (and other Asian and European) series being at the Chola Brhadisvara tem dance traditions through her body. How ple in Tanjore (Iyer, p. 37). Iyer posits that, ever, most of us in Southeast Asian art his while neither the idea of karana nor the tory have not been able to include more Natyasastra text originated in Java, perhaps than a smattering of dance studies (if any) the idea of carving a series of karana as part in our training, which makes the work by of a temple's sculptural program may have Alessandra Iyer, and her colleagues Felicia originated in Java and been transmitted Freeland-Hughes and Clara Brakel all the back to India, and as such, constitute a case more valuable. of cultural recycling-or the flow of influ To me, the two most interesting hy ence going the other way. potheses offered by Iyer are the following: However, before research like what Iyer (1) the possible presence of Shiva, Lord of outlines above can be taken any further, a Dance, on the Chandi Shiva and (2) the well-overdue translation seems an absolute linking of all the individual karana into a necessity. In Chapter 4, Iyer points out specific tandava (dance associated with that a balanced analysis of the dance reliefs Shiva). While Holt writes that no image of carved at Chandi Shiva or elsewhere in Shiva himself as Lord of the Dance appears classical Java is not possible until the Old at Prambanan or at any other of Java's Javanese text Nawanatya (late thirteenth sanctuaries (Holt 1967: 61), Iyer posits that fourteenth centuries), is fully translated and some of the groups of three dancers may analyzed. The fragments of the Nawanatya depict the Lord Shiva himself. While translated by Pigeaud in an appendix in his establishing the many variations in Shiva 1963 translation and study of the Nagara iconography from Indian to Cambodian Kertagama, include terms also found in the and Cham art and how definite Shiva attri Natyasastra. Iyer points out that this sug butes may have gone missing, Iyer does not gests prior knowledge of the Natyasastra in in the end point to any specific reliefs as Java (p. 92). It would seem clear, however, candidates for Lord of the Dance status. that the Javanese text would have to be This would be an interesting line ofanalysis read for all of its references, not only for all to continue. Combining close scrutiny of of its potential narratives of recycling but the dance reliefs by eyes very familiar with also for possible information of local in Shiva representations throughout South vention, whether more hybrid or less so. and Southeast Asia (to me, based on the These, then, are some of the studies we poor reproductions, reliefs P27, p. 132 and can hope for in the next round of scholar P31, p. 136 stand out as strong candidates ship pertaining to dance and cultural trans for Shiva status here) with the idea that mission in historical Java, and toward Javanese sculptors may have worked locally which Iyer's present book provides a step from imported workbooks with drawings ping stone. ofimages not entirely familiar to them, one