The Language of Balinese Shadow Theater [Course Book ed.] 9781400858767

Table of contents :
CONTENTS
PREFACE
ACKNOWLEDGMENTS
ORTHOGRAPHY
PART ONE. THE SHAPE OF THE WORD IN BALI
PART TWO. THE DISCOURSE OF BALINESE SHADOW THEATER
APPENDIX
GLOSSARY
BIBLIOGRAPHY
INDEX

Citation preview

THE LANGUAGE OF BALINESE SHADOW THEATER

THE LANGUAGE OF BALINESE SHADOW THEATER

MARY SABINA

ZURBUCHEN

P R I N C E T O N U N I V E R S I T Y PRESS PRINCETON, NEW JERSEY

Copyright © 1987 by Princeton University Press Published by Princeton University Press, 41 William Street, Princeton, New Jersey 08540 In the United Kingdom Princeton University Press, Guildford, Surrey All Rights Reserved Library of Congress Cataloging in Publication Data will be found on the last printed page of this book ISBN 0-691-09428-4 Publication of this book has been aided by the Whitney Darrow Fund of Princeton University Press This book has been composed in Linotron Trump Clothbound editions of Princeton University Press books are printed on acid-free paper, and binding materials are chosen for strength and durability Paperbacks, although satisfactory for personal collections, are not usually suitable for library rebinding Printed in the United States of America by Princeton University Press, Princeton, New Jersey

To My Parents

CONTENTS

Preface, ix

Acknowledgments, xm

Orthography, xv

PART ONE T H E SHAPE OF T H E W O R D IN BALI

1 1. Language Patterns and the Linguist's View, 7 2. Language from Birth to Death, 41 3. Literate Traditions and Literary Acts, 82 PART TWO T H E D I S C O U R S E OF BALINESE S H A D O W T H E A T E R

113 4. Speaking the Play The Use of Kawi, 123 5. Speaking the Play· The Use of Bahnese, 183 6. Shaping, Selecting, and Setting the Play, 206 7. Cultural Change and Noetic Change, 255 Appendix, 267 Glossary, 269 Bibliography, 275 Index, 287

PREFACE Invocation for the Performance of Shadow Theater

JUST AS THE boundaries of awareness become perceptible, there is per­ fect tranquility, undisturbed by any threat, and even the utterances of the gods subside. It is none other which forms the beginning of my obeisance at the foot of the Divine. Greatly may I be forgiven, for my intention to call forth a story. And where dwells the story? There is a god unsupported by the divine mother earth, unsheltered by the sky, unillummated by the sun, moon, stars, or constellations. Yes, Lord, you dwell m the Void, and are situated thus. You reside in a golden jewel, regaled on a golden palanquin, shaded by a floating lotus. There, all the gods of the cardinal directions approach in audience, these being: Isora, Mahisora, Brahma, Ludra, Mahadewa, Sangkara, Wisnu, Sambu, Sadasiwa, not to mention the five seers, and these are: Kursika, Gargga, M6tn, Kurusya, Pratanjala Not far away are the seven seers, these being. Parasu, Janaka, Kanwa, Narada, Bhayu, Bajra, Erawana, and even the four world guardians, these being: Indra, Yama, Baruna, Kuw6ra. Not distant are the three great powers, namely: Siwa, Sadasiwa, Paramasiwa, not distant are the male and female celestial creatures. All are nearing the foot of the divine, exalted Supreme Teacher, in the Void. There are found the young palm leaves, the one lontai, which, when taken and split apart, carefully measured are the lengths and widths. It is this which is brought to life with hasta, gangga, uwiia, tanu. And what are the things so named? Hasta means 'hand', gangga means 'water', uwiia means 'writing-instrument', tanu means 'ink'. What is that which is called "ink"? That is the name for, and none other than, the soot of the oil-lamp, mixed with the ashes of the kepuh-tree, placed in a vessel of copper. It is these things which are gathered together and given shape on leaf. "Written symbol" is its name, of one substance and different sound­ ings. And thus there are the Ten Written Symbols, and the Five Writ­ ten Symbols, and the Three Written Symbols, the total count is eight­ een m number, the complete set of written symbols, in the form of vowel and consonant types. The symbols a, 1, u, έ, ο· the a symbol is attached with the taiung-, the ι symbol is attached with the hulu, the u IX

PREFACE

symbol is attached with the suku, the έ symbol is attached with the tahngj the ο symbol is attached with the tahng and the tamng. Ka-khaga-gha-nga, ca-cha-ja-jha-na, ta-tha-da-dha-na, pa-pha-ba-bha-ma, yara-la-wa, sa-sa-sa-ha. Of many forms is the greatness of the symbols, concealed, disguised, like ants all in a row. It is that which is spoken of by the exalted poet in his wisdom. And who is he called "exalted"? He is well born, fore­ most among poets, with knowledge of the smallest details in the field of letters. Rightly explained by him is the nature of the sounds of the four stops. There are anudanti, anunasika, alpaprana, and maha­ piana. What are these so named? Anudanti are the written symbols sounded on the ends of the teeth, anunasika are written symbols sounded through the nose ; alpaprana are written symbols sounded with fluttering of the lips, mahapiana are written symbols sounded from within the mind. These are what resound, allowing the shaping of a story. Thus, ap­ parent is the unity of the astadasoka, the Eighteen. How? As follows: asta means 'eight', dasoka means 'ten'. The total is eighteen, the num­ ber of the Paiwas, from the land of India. These are recounted by the divine Supreme Poet, and include, for example, Sramaparwa, Santikaparwa, Udyogaparwa, Swargarohanaparwa, Dirawanaparwa, Bhismaparwa, Dronaparwa, Krepaparwa, Salyaparwa, Karnaparwa, Wirataparwa, complete, with the Sanskrit phrases of the Adiparwa. And how does tonight's story begin? .. . THE INVOCATION translated above (see the Appendix for the Bahnese version) is an example of a traditional opening "apology" uttered by a Bahnese puppet master, dalang, before the first scene of a shadow pup­ pet play, or wayang kuht performance. Such invocations vary in con­ tent, in length, and in regional distribution throughout the island. In such texts, the use of symbolism, allusive language, and metaphor is densely complex, when uttered aloud, the dalang's vocal techniques of chanting, roaring, and guttural resonance combine a phonological po­ etics with the mysterious, incantatory atmosphere of the opening mo­ ments of the shadow play. It is indeed a too-ambitious undertaking to attempt to explain all the meanings of such invocations, says the pup­ pet master; their truth lies beyond words and analytic reason. The out­ sider can hope to catch only part of the ultimate sense, and even that only after much study and contemplation. In the thus imperfect understanding of the foreign visitor to Bah, then, one puzzles over the intent of the performer's invocations. The χ

PREFACE

example above seems to join cosmological vision with metahnguistics. It shares with other examples of the genre two particular features: first, a direct address to nonspecific divinity asking forgiveness, and second, the linking, in the last words, of the play about to be performed with the books, or parwa, of the originally Indian great saga of the Bharata heroes. The last verses of the Sanskrit Mahabhaiata compare the epic itself to the oceans and mountains, "mines of precious gems." The image of the literary work is that of a perpetual source of meaning, yielding up jewels of wisdom from boundless depths. In Southeast as well as South Asia, this particular epic has indeed been a recurrent inspiration and cultural force for many centuries. It is intriguing to find on the small island of Bali a continuing, exuberant expressiveness in which Mahabhaiata tales and characters are a vital presence. This presence becomes even more provocative when it is realized that direct contacts between Bah and the South Asian sources of Sanskrit literature ceased more than a millennium ago. The Bahnese people on their volcanic island, speaking Austronesian languages and retaining a prehistoric base of ancestor-and-animist beliefs even while fashioning a unique blend of Hindu and Buddhist conceptual systems, have also woven the Mahabhaiata into their characteristic cultural proliferation. The intriguing persistence of an originally Sanskrit epic in the form of Bahnese shadow theater is a spark that ignited my interest in the languages and verbal arts that are subjects of this study. Yet I am not concerned here with tracing Indian entities from historical sources to present forms. There is little that is exclusively Indian in the Bahnese expression of the struggles of Pandawas and Korawas, the antagonists of the epic, and the wayangparwa (shadow theater using Mahabhaiata stories) is a rich and enduring form because it shows important things to the Bahnese about themselves. One of the many things wayang parwa portrays is the role of language as the shaper of Bahnese experience. As a student of linguistics, I found the configuration of linguistic forms, sound, and written literature, together with beliefs about the cultural past and present as suggested by shadow theater, to be of primary importance in my research. I came to Bah after studying linguistics and Indonesian languages and literatures, and I was initially drawn to wayang because of its use of Old Javanese (Kawi), a language fundamental to Bali's history and social evolution. With my husband I took up residence in a village in south-central Bah where shadow theater is the dominant performing XI

PREFACE

art, and I soon found myself following performers as they carried their puppets, screen, and instruments from village to village, temple to temple, covering most districts and ceremonial occasions of the territorial and temporal layout. Watching wayang was an effective method of approaching modern Balmese language, another object of my studies, and as time went on I was increasingly impressed by the nature of wayang as a complex and perplexing form of language behavior. Of course, there is more to wayang than language—there are also light and shadow, movement, music, the sounds and smells of tropical nights, the setting of crowds and busy ritual—yet more than other forms of theater Balmese wayang is a linguistic tour de force, and the dalang perceives the task in this way I eventually became convinced that the linguistic shape of wayang parwa could only be explained after coming to terms with many other ways in which languages and texts are experienced in Bah. The interrelations of sound, verbal code, ritual use of language, beliefs about writing, manuscript tradition, textual structures, social force of certain linguistic forms, and coherence of the universe itself—all these have repercussions in terms of wayang that are neither trivial nor always obvious. For this reason, I offer in this study a perspective on Balmese verbal art which, emphasizing contemporary wayang practice, attempts to ground itself in a broader context—the context of the particular culture's language behavior, and also the wider scope of a linguistics that encompasses the patterns and purposes of all verbal communication.

Xll

ACKNOWLEDGMENTS

tan wismtti sangka mkang hayu η teka For never did he forget from whence the good came —Ar/unawiwaha XIII 1 THE LEARNING that led to this book could never have occurred with­ out the advice and assistance of many persons. In Bah from 1977 to 1979,1 was fortunate enough to be helped in my studies by Ida Bagus Aht Budhiana, I Mad6 Bandem, Ketut Ginarsa, Ida Bagus Maka, Mad6 Sanggra, I Gusti Bagus Sugriwa, Nyoman Sumandhi, and Wayan Warna. At Udayana University's Faculty of Letters, I Gusti Ngurah Ba­ gus, Gusti Putu Dharsana, Wayan Tondra, and Nengah Mddra were most kind. The staff of the Gedong Kirtya and its director Ketut Suwidja were generous with their assistance. I owe a lasting debt of grat­ itude to the late Ida Bagus Ged6 Gena of Sidemen for sharing his deep understanding of languages and literature with me. To the shadow theater performers of Bah who unselfishly shared their knowledge, time, meals, and seats near the screen, who willingly answered questions and explained performances both day and night, I express deepest gratitude. Particular thanks must go to Wayan Loc6ng and to dalangs Wayan Nartha, Ketut Brata, Nyoman Ganjring, Mad6 Sua, D6wa Rai M6si, and Ida Bagus Gria. As neighbors, artists, and friends, the members of Banjar Babakan and soioh dalang Sukawati were patient and giving. Among those who taught me about the shadow theater, there is one above all whose help and inspiration will most endure. Part Two of this study is dedicated to the memory of Dalang Sukawati I Ketut Madra Besides these many teachers, there are friends in Bah whose hospi­ tality and good cheer made the hard times easier and the good times even better In Sukawati, the family of Anak Agung Ged6 Putra, in Budakehng, the families of Mad6 Kuduk, Wayan Kamasan, and Ida Wayan Ged6; in Sidemen, Cokorda Ged6 Dangin and Dewa Ayu Putu, in Denpasar, the families of Putu Sukosti, Otong Wirawan, and Verra Darwiko—all deserve deep thanks. Outside of Bah, the advice and hos­ pitality of I Gusti Ngurah Oka in Malang, Gloria and Soepomo Poed}osoedarmo in Yogyakarta, and the Ikranagaras in Jakarta were much appreciated. xni

ACKNOWLEDGMENTS

The fieldwork leading to this thesis was carried out under a fellowship granted by the Social Science Research Council and the American Council of Learned Societies. In Indonesia, sponsorship was kindly provided by the Pusat Pembinaan dan Pengembangan Bahasa, under the directorship of Amran Hahm, and the Lembaga Ilmu Pengetahuan Indonesia. My graduate studies and dissertation write-up were assisted by grants from the U.S. Office of Education and the Horace Rackham School of Graduate Studies at the University of Michigan. I was helped in the preparation of the manuscript by the facilities and staff of the Ford Foundation's Southeast Asia Regional Office in Jakarta, especially Sophie Saerang. To all whose friendship and scholarship have inspired me, many thanks—especially to David Stuart-Fox, Francy Hays, Jeff and Gail Dreyfuss, Pinky and Jim Henry, Richard Walks, Gary Mcdonald, Susan and Ken Walton, and Alan Feinstein. Colleagues in Berkeley and Santa Cruz—Amin Sweeney, Bart van Nooten, Kathy Foley—have provided valuable consultation. The interest, wise counsel, and encouragement of my husband, Mark Poffenberger, have been constant through the years. To him goes credit for all photographic illustrations and to the Gedong Kirtya special thanks for permission to photograph illustrated manuscripts. The members of my doctoral committee at the University of Michigan (Judith O. Becker, Madhav Deshpande, and Paz B. Naylor) were important in bringing this text to fruition. Each presented viewpoints and offered opinions from within his or her range of experience that proved most valuable. To my chairman and friend, A. L Becker, a most profound gratitude is directed. His guidance, scholarship, criticism, humility, and good cheer have provided many lasting lessons. I am privileged to be his student.

xiv

ORTHOGRAPHY

FOR THE MOST PART, the spellings used in this book conform to the modernized Balmese orthography of the Ramus Bali-Indonesia (KBI). Thus, έ as in English hey and e as in English dirt are distinguished from each other, word-final a should also be pronounced as in English first, shirt, and so on Sanskrit and Old Javanese terms may not conform to their respective conventional orthographies unless they appear in the title of a text or are quoted from a manuscript or printed work. Indo­ nesian words employ the orthography in effect since 1972, for both In­ donesian and Balmese terms, / and y and c correspond to English /ump, yes, and ch&u, respectively. In the bibliography, archaic spellings in personal names (e.g., oe in­ stead of u) have been retained. The first time a non-English word occurs in the text it is italicized. Subsequent occurrences are not italicized, except when necessary for clarity. Frequently cited Balmese or Indonesian terms are listed m the accompanying Glossary.

xv

PART ONE THE SHAPE OF THE WORD IN BALI

IN THE LONG history of Western scholarly commentary on Bah, the phenomena of language and literary arts have never been overlooked, indeed, in some of the earliest general accounts, such as those by Crawfurd (1820) and Fnedench (1876-1878), the possible origins and observed transmission of literary languages and manuscripts were discussed at length Subsequent generations of researchers, including van der Tuuk, Gons, and Hooykaas among the most prominent, have continued to labor in the fields of lexicography, epigraphy, and textual studies, these and many other scholars have contributed the collections of manuscripts, translations, and critical editions that form the basis of our present knowledge and opinions of Balinese words and the works into which they have been shaped If one scrutinizes this wide range of valuable contributions to Balinese letters, several models for the general orientations of its authors come to mind Not every scholar has clearly manifested one of these orientations, but on the whole there is much evidence for the following outlines of thought First, some authors have approached Balinese letters as a route of access to the historical past, following the traces of a once great and powerful, civilizing Indian presence in the Indonesian archipelago According to this notion, Hindu and Buddhist culture and Sanskrit texts were brought, either in a thin trickle through centuries, or as part of a colonizing wave, to the maritime and agrarian kingdoms of ancient Southeast Asia, whence subsequent historical tempests caused them to be swept from their original island beachheads and to be stranded in the "living museum" of Bah By studying the Old Javanese and Sanskrit works preserved there, scholars felt, we could reconstruct the circumstances, persons, places and texts involved in the scarcely documented process of "indenization" in Southeast Asia The priority given to Indian-born theology, liturgy, epic themes, and Sanskntized language in the views of some scholars resulted from a methodological concern for tracing Balinese manuscripts and ritual to the hypothesized South Asian homeland It was assumed that "pure" Indian forms had been superimposed on the prehistoric animistic, "pagan," indigenous culture of the archipelago, and after many centuries of contamination and acculturation they would have to be carefully sifted out from the accretions, corruptions, and distortions of a confused transmission In this view, the Balinese could only dimly grasp the significance of the lost Indian origins of things they had come to call their own, and thus any Balinese usages, interpretations, and 3

THE SHAPE OF THE WORD IN BALI

changes of the heritage not only were of scarce interest, they indeed actually obscured the underlying original forms. Another general perspective common among students of Balmese letters emphasizes not the Indian so much as the Javanese cultural legacy on that island. For certain authors, the ancient Javanese world of Hindu-Buddhist kingdoms, shrines, arts, and literature is the object of study, since Java's more recent Islamization and colonization, many of the signs of this past vanished or were moribund, and so scholars would look to the smaller island neighbor for Hindu-Javanese reflections and continuities. In this view, Old Javanese "classical" literature, architecture, royal genealogies harking back to the Ma)apahit courts, and anything else that could be traced to ancient Java were of paramount importance. The processes of adaptation whereby Javanese entities were "bahmzed," or the contemporary meanings given to Old Javanese letters in Bah, were largely ignored. A third general view of languages and literary arts in Bah has taken both Sanskritic and Javanese origins and influences into account, yet has focused primarily on the material of the manuscript tradition. Many scholars have seen the wealth of works on palm-leaf strips as a massive archival and philological challenge. They have conceived their task to be the collection and "rescue" of as much material as possible, and the transferral of it into more durable form. These scholars have assumed that multiple versions of a text emanated from a single original and ideally should be grouped together, analyzed, and collated, with any inconsistencies, ambiguities, lacunas, or other undesirable features eliminated. From this process would emerge a "purer" text resembling that of the "original" author—a text free from corruptions and errors that are the inevitable result of repeated transmission through copying. This work would then stand as a "critical edition," a solid basis on which any further linguistic and textual research could be founded. Like the perspectives emphasizing Indian and Javanese origins, the philological1 viewpoint in Balmese studies tends to downplay any spe1 My use of the term philological refers to the text-focused study of language The original object of philology, the "unified study of language and texts," seems in many respects to have become obscured by a word-focused philology that has as its objects both etymological analysis according to the science of historical linguistics and the compilation of critical editions and translations of texts In a recent essay, A L Becker (1980) has suggested a "re-visiomng" of philology in order to reunify linguistics and textual studies This "modern philology" aims at deeper cross-linguistic and cross-cultural understanding through

4

THE SHAPE OF THE WORD IN BALI

cifically Balmese processes, interpretations, and contributions within the long evolution of the literary tradition. In the arduous task of manuscript scrutiny and attention to minute detail demanded by the production of critical editions, scholars have not always been concerned to analyze the relationships between a text and its immediate cultural surroundings. Manuscripts written in Javanese linguistic forms were considered part of "Javanese" or "Javano-Balmese" literature irrespective of their significance or possible origin m Bah; words borrowed from Sanskrit or modeled on Sanskrit analogy were to be "corrected" to resemble the "true" originals Texts were taken as self-sufficient entities suitable for study regardless of any real-world users or uses, any difficulties of form or meaning withm the text should thus be resolvable by comparison with other manuscripts Similarly, although great attention has been paid to issues of provenance, original form, and historical significance of manuscripts, and although a diachronic viewpoint is certainly necessary, there has been a notable lack of an interpretive thrust that seeks to explain Balinese letters primarily in terms of Balinese behavior and beliefs. Questions regarding how and why the Balmese maintained such an active and multiform literary tradition, the impact of Balinese and non-Balinese languages and texts on linguistic and non-linguistic cultural forms, and the synchronic configurations of languages and literature in terms of everyday life have long remained unanswered. The present study aims to bring forward such questions of interpretation as a context for understanding the linguistic forms and cultural role of one genre of verbal art, the Balinese shadow theater. Part One is thus devoted to a discussion of the meanings and uses of language and written texts in the Balinese world, under the assumption that these may exist according to wholly different patterns, and with a quite distinctive character, when compared with standard Western conceptions of the "universal" qualities of language and literature The study begins with a survey of important diachronic and synchronic patterns of interest to the linguistics-minded observer. I also try, in Chapter 1, to make explicit some of the theoretical perspectives and assumptions that have characterized my research and reporting of conclusions. Chapter 2 covers certain aspects of language in both spo"samphng the whole of the relations of the text to its various contexts " The philological expertise that has traditionally been applied by Western scholars to Balinese texts is not the same "linguistics of particularity" that Becker describes 5

THE SHAPE OF THE WORD IN BALI

ken and written manifestations, with reference to important social values and cultural patterns. In Chapter 3, the focus shifts from language in general to verbal art in particular, as we examine the media of literate expression and the characteristics of the Balinese "literary" entity and event. The subject matter of all three chapters is interrelated and can be seen as a setting within which both terminological and conceptual problems regarding language in Bah take shape, and as a background against which my later evaluations of the discourse of Balinese shadow theater must be viewed

6

1 LANGUAGE PATTERNS AND THE L I N G U I S T ' S VIEW This is our fabric, the language is, and it's very specific It's really like a warp and a woof The idea that there is an archetype, either deep in your psyche or in a metaphysical world, is nothing compared with the fact that language is flooded with archetypes as we write or read Powerhouses and residence these words are, and they're built into a vast fabric and they extend way back in time —Robert Duncan, "Warp and Woof Notes from a Talk" LANGUAGE PATTERNS IN BALI T H E L A N G U A G E world of Bah is one of remarkable richness and diversity in all its spoken, written, sung, and chanted manifestations. So various are the different linguistic forms employed, so complex the interweaving of vocal styles and literary genres, that both language and literature seem a tangled confusion that escapes characterization and conceals both sources and structures. Yet this linguistic proliferation is indeed like a richly woven fabric, no matter how complex the design, a discemable warp and weft underlie its form and provide essential unity. Two general descriptive dimensions have long been important in language studies· the diachronic, or historical and developmental aspect, and the synchronic, or timeless and systemic one Like the unjoined warp and weft in weaving, however, which unloomed have neither the art nor serviceability of textiles, the abstractions of diachronicity and synchronicity, if taken individually, do not resemble a language. Both these notions must be employed to highlight those features of the general Balmese linguistic setting that facilitate analysis in following chapters. Diachronic Considerations Balmese is an Austronesian language of the Hesperonesian subgroup, 1 and it is related to the languages of Bali's closest island 1

See the classification provided by Bellwood (1979 121-124)

7

THE SHAPE OF THE WORD IN BALI

neighbors, Lombok to the east and Java to the west. On the basis of 2 phonological correspondences, Esser groups Bahnese w i t h Sasak of Lombok and Sumbawan of western Sumbawa, making Bahnese part of an Eastern Indonesian subgrouping. Yet this assignment is in some ways unsatisfying because both Bali's cultural history and its presentday linguistic forms make us turn west toward Java (and even beyond Indonesia, to India) as well as east toward Lombok. While it is theoret­ ically useful to examine a stratum of Bahnese that might be called "in­ digenous" or "pre-Sanskrit, pre-Javanese influence," and thus that might be closer to some Eastern Indonesian languages in terms of pho­ nology, in reality such a stage of the language is unrecoverable, instead of being able to look at the language of a h u m a n group at that abstract level of investigation, one finds only a very limited list of relationships between abstract " s o u n d " categories. 3 T h e fact is that Bali from the time of the earliest written records had already come under the influence of non-Balmese cultural and linguis­ tic traditions. 4 We have n o way of knowing how Bahnese was spoken before the adoption on the island of certain Buddhist and Hindu ideas, along with t h e use of Sanskrit, some time before the ninth century A.D., the period of the oldest dated metal 5 and stone inscriptions. 6 Aside from a relatively small number of the oldest inscriptions that are entirely in Sanskrit, including the Buddhist clay seals that are sup­ posed to date from the eighth century A.D., 7 Bahnese inscriptions prior to t h e time of t h e first close ties with Java are in a mixture of Sanskrit and Old Bahnese (Coedes 1968:129, G o n s 1954). It appears that, just as in other Southeast Asian areas where Indian 2

See Map 9B of the Atlas van tiopisch Nederland (Esser 1938) For this dimension of comparative Austronesian linguistics, see especially Dempwolff's extensive vocabulary (1938), drawn from a large number of lan­ guages, as well as the work of Dyen (1965) and Dahl (1973) 4 Goris's (1954) collection of ancient edicts with notes, translation, and glos­ sary remains the most thorough work published containing data from this early period The name Bah, it should be noted, is thoroughly documented in both Javanese and Bahnese inscriptions as being synonymous with banten, meaning 'offering' (Sukarto Κ Atmod]0 1977 13) 5 Although the metal inscriptions are usually referred to as "copper plates," the actual material is bronze (Bernet Kempers 1977 97) 6 Since the inscriptional evidence from Java starts at an earlier date, from at least the fifth century, we can speculate that the Bahnese, too, may have begun hearing Sanskrit sometime in the early centuries A D 7 For the text and translation of the formula on these artifacts, see Gons and Dronkers(1955 190n 3-01) and Bernet Kempers (1977 101-103) 3

8

LANGUAGE PATTERNS AND THE LINGUIST

influence led to the use of Sanskrit, the language as it was adopted in Bah did not alter indigenous sound systems at all (Gonda 1973 579) Sanskrit seems to have functioned as a court-based medium for the teaching of literature and religion, and it was never spoken or even studied by the population at large Yet Indie words were used in great numbers in the Old Bahnese inscriptions, this fact, while shedding little light on the nature of ancient spoken Bahnese, does raise interesting issues about the communicative situation that existed between the kings authorizing the writing of inscriptions and the communities of presumably nonhterate villagers who received them We will return to this relationship between the "speakers" and "hearers" implied by the ancient Bahnese edicts The general consensus among historians regarding the spread of Indie notions of religion and statecraft in the Indonesian archipelago has shifted from a characterization in terms of "colonization" or "Indian expansion"8 to a view that gives native peoples a more active role in searching out and incorporating Hindu and Buddhist cultural features 9 This view proposes a syncretic process of acculturation whereby priests and scholars, not necessarily of Indian blood but possibly Indonesians who had studied in India or at one of the monasteries within the archipelago, contextuahzed indigenous kings and native religion m terms of highly congenial ideas expressed in the sacred texts and epic literature brought from India (Stutterheim 1935 7, Bernet Kempers 1977 40) Using writing systems whose origins were Indie, learned officials in the Old Bahnese courts created inscriptional monuments to the supremacy and royal decrees of the rulers They mainly employed a script derived from the Pallava writing systems of southern India, which had spread in ancient times through Ceylon as well as mainland, peninsular, and insular Southeast Asia The form of this script used in the Old Bahnese period is generally called "Early Kawi," emphasizing its close relationship with contemporaneous scripts of Java There is no definitive proof, however, that the scripts used in ancient Java and Bali might not have developed separately from a common Pallava source (Bernet Kempers 1977 43) 8

See, for instance, Gonda (1973 66) "This spread of Hinduism was neither sudden nor violent By marrying into leading local families, by displaying qualities as magician, practitioner or warrior, by the prestige of their higher civ lhzation some of them |the Indians) impressed or even subjected local populations " 9 See, for example, the essays m Hall and Whitmore (1976) 9

THE SHAPE OF THE WORD IN BALI

Indian influences on the Old Bahnese courts came from diverse sources, and their paths of transmission are difficult to trace—a fact that is borne out by the history of writing in Bali For in addition to the Pallava-denved "Kawi" syllabary, we also find the "Early (or Pre-) Nagari" script of North Indian origin, the ancestor of modern Indian Devanagari The use of this script in ancient Java and Bali is tied to the spread of Mahayana Buddhism I0 In one particular case, the form of the Early Nagari script used on a non-Buddhist stone pillar erected in southern Bah in the early tenth century suggests a possibly direct link with ongoing changes in the Early Nagari of the same period in India (de Caspans 1975 37), that link may, perhaps, have been a native Indian scribe engaged at court Whether directly or indirectly infused, however, Indie tradition emerges not only in the written shape but also in the language of the Old Bahnese inscriptions The priests and scholars responsible for directing the Old Bahnese kings' ritual and civic affairs were perhaps initially literate only in Sanskrit Yet by the time of the earliest (late ninth century) inscriptions, they were inscribing native formal discourse—what we know as Old Bahnese but which by now was mixed with Sanskrit words and phrases—in the Indian-derived script They used Indian concepts to organize increasingly centralized bureaucracies, implementing complex systems of hermitage and temple management, taxation, corvee, irrigation management, and justice The character of the early inscriptions leads one to the conclusion that even prior to the late ninth century the Bahnese had been hearing and using some Indie vocabulary in the context of their relations with religious centers and courts u Gonda (1973 180) notes that most of the Sanskrit terms in the Old Bahnese inscriptions are used in connection with the king, and that "many Sanskrit items occur only, or predominantly, in those passages of the inscriptions which contain the date or refer to religious ceremonies, festivals etc " In the very oldest inscriptions there are Sanskrit titles for court and religious officials (e g, senapati 'commander', bhiksu 'monk'), as well as Sanskrit roots serving 10 Coedes (1968 30, 89) mentions a "temporary vogue" for this script in Java and Cambodia, which seemingly indicates the influence of the West Bengal and Nalanda monasteries in India 1 ' One authonty who agrees with this conclusion is Gons (in Gons and Dron kers 1955 16) On the Javanese side, Robson (1983 294) suggests, in a similar vein, that "the main result of acculturation with India was that the institution of the ratu ['king') became Indiamzed "

10

LANGUAGE PATTERNS AND THE L I N G U I S T

as Bahnese word stems, for example, paitapanan 'hermitage', from Sanskrit tapa 'austerities, solitary life'.12 The sociohnguistic situation in the Old Bahnese courts was already complex. On the one hand the king dealt with his Sanskrit-titled rehgio-pohtical officials in formal speech heavily laced with Sanskrit and heard himself consecrated in religious ceremonies using Sanskrit prayers and chants, some of which were composed in Bah following Indian models 13 On the other hand the village leaders used a language that was some form of Old Bahnese—a language that represented the populace which formed the kingdom's economic base and retained the traditions of village autonomy, native religion, and indigenous social organization even as it embraced aspects of Hinduism and Buddhism expressed in Indie terms. For the common Bahnese, "Sanskntization" was profoundly a linguistic phenomenon, and the monks and scribes were agents of an ongoing cultural revolution the advent of literacy in the archipelago.14 To honor the king properly, Sanskrit terms were necessarily used by courtiers and village representatives, to avoid alienating his subjects, the king could not move too far away from the Bahnese linguistic framework. After reading the inscriptions, one imagines that a good deal of paraphrase and cross-linguistic explanation went on in the Old Bahnese courts as boundaries, taxes, ritual obligations, and inheritance settlements were debated by the king's council, which heard from village elders and scholar-priests alike.15 The sociohnguistic aspects of the Old Bahnese court, then, promoted the use of different languages or speech styles by different speakers in different settings. A crucial result of this "multiple code" communication was the need for translation and paraphrase The king, in recording his edicts, would wish to make certain that the Sanskrit words constituting as well as recording his authority and divinity were understood by the general populace. He was after all not addressing 12

For these and other examples, see edicts 001 and 002 in Gons (1954, vol 1) According to Goudnaan and Hooykaas (1971 5), the "great ma)onty" of these hymns were written in Java and Bah For an explanation of how some of them may have been composed, see Schoterman (1979] 14 To date there is no evidence that a non-Indie script was ever used in Indo nesia, so we continue to assume that writing and Sanskrit were imported hand in glove For a consideration of the processes of linguistic "Sanskntization" in India itself, see Deshpande (1978) 15 The most ancient edicts in Gons (1954) are interesting in this regard, see especially numbers 001-006, 101, 104-107, 108, 110 13

11

THE SHAPE OF THE WORD IN BALI

himself only to the gods or to an abstract posterity, but to a specific and contemporary Balinese audience, as we know from the speech context reflected m the inscriptions. A series of the oldest inscriptions is characterized by an opening for­ mula: Yumu pakatahu sarbwa [Xj, which may be glossed as 'Let it be known by you all', with [X] standing for various terms referring to the mtended audience. Yumu is a second-person pronoun probably related to the -mu of Old Malay and the kamu and kanyu of Old Javanese, the inscriptional context strongly suggests a non-honorific sense, such as fits with the notion of a ruler addressing his subjects. Pakatahu is a derivation composed of the root tahu 'know', plus the nominalizing prefix paka-, which has both causative function (emphasizing the re­ sult of an action) and imperative force.16 Sarbwa is a written variant of Sanskrit saiwwa 'all' The titles referring to the king's audience vary from one edict to an­ other, but many of them allude to age status, which seems to have al­ ways been an important social referent in Bah. Among the terms meaning 'elders' are kiha (reminiscent of the Javanese hononfics ki and kyai), kumpi (in Modern Balinese, 'great-grandfather'), and sanat (Sanskrit). For 'younger (persons)' we find dyah, kumara (Sanskrit), and addhi (cf. Modern Balinese adi 'younger sibling') The mixture of in­ digenous terms with Sanskrit borrowings is characteristic of the edicts as a whole. 17 From the point of view of the Balmese villagers, the inscriptions au­ thorized by the king were kept as relics, repositories of the power of the king's word.18 In certain cases their contents may have been explained to the general public, in colloquial Balinese, on holy days when the stone or bronze plate was ntually cleansed, given offerings, and read aloud. Readings of inscriptions do still take place in many Balinese vil16

Whereas paka- only occurs with tahu in the Old Balinese inscriptions, the related form maka- is frequent and still functions as a numerical prefix in mod­ ern Balinese See the discussion in Gonsj 1971 2-5) on paka17 For examples of the use of the yumu pakatahu opening formula, see the edicts cited in η 15 above 18 Bernet Kempers (1977 97) comments on the situation encountered by twentieth-century Dutch scholars collecting epigraphic material "Villagers considered these metal bearers of a mysterious script to be magical, sacred Be­ sides, the plates were connected with the ancestors living in the mountains and the jungle Best, they thought, to wrap the plates in white cloth to keep away evil influences, and to place them in closed temple shrines " 12

LANGUAGE PATTERNS AND THE LINGUIST

lages that store the sacred heirlooms, with archaeological specialists called in to decipher the old script.19 The Old Bahnese kings seem to have been aware of the impact of the introduction of new cultural attributes through the medium of a new language. On at least one occasion a Bahnese king inscribed his edict in two languages, and even in two different scripts, giving us a record in stone of functionally distinct linguistic codes in ancient Bah. This inscription is the Blanjong pillar of Sanur, dated A.D. 914 (Coedes 1968:129; Bernet Kempers 1977:103), which tells of the victories of a king called Sri Kesari Warmma (d6wa). One side of the pillar contains an Old Bahnese form of the message in Early Nagari script, the other side, in Sanskrit, is in the Pallava-denved Kawi scnpt. This unique and somewhat cryptic form was perhaps intended to blend and unify the two literate traditions then active in Bah, rendering the more indigenous sounds in the more remote script while elevating the "local" alphabet by using it to record the prestigious pure Sanskrit version of the text. The form of the Blanjong pillar edict is conservative in that it throws nothing away, making use of all available scripts and idioms; at the same time it is also quite radical in its crisscrossing of foreign and indigenous elements. Yet however it came about, the pillar is an early textual manifestation of a crucial aspect of Balmese language and literature from ancient times up to the present- the preservation of various "archaic" linguistic codes and their translation or interpretation via more "contemporary" ones. The many ways in which this principle has operated and, indeed, continues to operate will become clearer as this study proceeds.20 Moving through the Old Bahnese or "Hindu-Bahnese" historical frame, when the island was acculturating language and ideas that were originally Indian, we come to a period when Bah was closely tied with Java politically, and when Javanese-Bahnese cultural exchange no 19 During a series of visits made in 1977 and 1978 to Bahnese villages to gather data on local edicts, a team led by Putu Budiastra invariably was welcomed by residents eager to know the contents of the inscriptions (Budiastra 1977-78) 20 Another indication of the antiquity of the practice of "dual code" composition of texts comes from manuscripts in which interlinear Old Javanese glossing of Sanskrit occurs Regarding both poetic and didactic forms, Chandra (1978) argues that these lontars provide unique insights into the pedagogical context for the transmission of a variety of Sanskrit genres in ancient Indonesia

13

THE SHAPE OF THE WORD IN BALI

doubt intensified. This period was probably initiated by the marriage in the late tenth century of Udayana of Bah and the Javanese princess Mahendradatta, although there is no reason to completely rule out the possibility of dynastic intermarriage even before this time.21 The famous son of this king and queen of Bah was Airlangga, who reigned in East Java, and presumably also held some sway over Balinese affairs, from A.D. 1019-1049. It is possible that at least by this time the most ancient extant examples of Old Javanese literature, such as the kakawin Ramayana written in central Java during the ninth century, were brought to Bali, where they were studied and copied. This would also have held true for the growing body of Old Javanese poems, treatises, didactic works, and so forth written in East Java. These works in palmleaf manuscript form [lontai, wntal) gave impetus to the assimilation of new ideas from Java, leading to a "Javano-Balinese" cultural period following upon the age of the Old Balinese rulers. Javanese influence began to be felt in many spheres of Balinese life (all of which probably utilized the growing numbers of written manuscripts from Java), including literature, religion, philosophy, statecraft, and the arts and sciences. In the context of language history, an especially significant change took place in late tenth-century Bah. During the reign of Airlangga's parents, court officials began to write inscriptions in Old Javanese,22 indeed, after A.D. 1016 there are no more inscriptions in Old Balinese.23 While one can only speculate on the precise historical conditions that accompanied this change in the code of the inscriptions, it seems clear that the Balinese kings were again adopting and adapting a congenial model of written expression for their own needs. In the ninth century, priests trained in Sanskrit were an important institution at court because of their literacy; in the eleventh century, the highly developed Old Javanese court-based code became available and desirable as Bali's political and cultural ties with Java became stronger. Linguistically, the shift from Old Balinese to Old Javanese is clearly distinguishable. Even though the two languages show structural similarity, share a certain percentage oi vocabulary, and utilize Sanskrit 21

The inscriptional clues, primarily based on similarities in dating and titles, that indicate some previous ties between the courts of Bah and Java are discussed in Sumadio (1975) 22 Thefirstedict in Old Javanese is numbered 303 in Gons (1954) and is called Buwahan A 23 The last Old Balinese inscription is numbered 351 in Gons (1954) and is called Sembiran A III 14

L A N G U A G E P A T T E R N S AND THE L I N G U I S T

borrowings, they are "clearly different, in phonology, morphology, syntax and lexicon" (Teeuw 1965-273) For instance, analysis of pho­ nological correspondences based on Dempwolff's (1938) sound-change laws shows Old Balinese to be closer to Malay than to Old Javanese 24 In the area of morphology, the functional categories of the Old Balinese and Old Javanese verbal systems seem comparable, although the forms taken by parallel affixes show significant differences (Teeuw 1965 281-282)25 And from a lexical perspective, the addition of Javanese vo­ cabulary at the time of the change from Old Balinese to Old Javanese possibly helped to intensify the development of status- and mtimacyrelated registers in the spoken language, a process that had previously been in operation when Old Balinese rulers used specialized Sanskrit borrowings to refer to themselves 2 6 24

See Ward (1973 chap 4) for some discussion of phonological and lexical changes from Old t o Modern Balinese 25 Summarizing the information presented in Teeuw's paper, and at the same time altering some of the syntactic labels, we see the following correspond­ ences in Old Javanese and Old Balinese verbal systems It m u s t be noted, how ever, that conclusions about syntactic function aie provisional, since not all Ba­ linese mscnptional data have been analyzed Syntactic function

Old Javanese affix

Old Balinese affix

Nominahzer Transitive Intransitive Goal focus Agent focus Stative aspect Locative particle

pa aken (m)a m (m)aN um ι

pa ang ma y maN um in

26 In the changeover from Old Balinese to Old Javanese, several prominent elements of t h e inscriptions are formally altered without, however, their struc­ tural significance being eliminated The dating system in the Old Balinese edicts, for example, used Sanskrit m o n t h names [bulan being the Balinese term for 'month'), the number of the day according to the moon's waxing [uukla] or waning [kisna], the day according to the three-day market week {iggas), and fi­ nally, the number of the Indian Saka year When officials began to write the in­ scriptions in Old Javanese, they renamed the Sanskrit m o n t h s using the San­ skrit masa, and they noted the day of the week with abbreviations from the six-, five- , and seven day weeks of the Javanese waia (another Sanskrit term) sys­ tem, as well as the n a m e of the particular seven-day week according to the thirty-week wuku calendar Thus, the entire dating formula of the edicts was altered Similarly, the important mscnptional topic of territorial boundaries is re­ phrased in the edicts, using Javanese instead of Balinese directional terms

15

THE SHAPE OF THE WORD IN BALI

Since the Old Javanese of the later Bahnese edicts was also a specialized court and administrative code, the non-inscnptional, unattested Bahnese language of the period must not be thought of as entirely lost or outmoded, even though Old Bahnese no longer appears in mscnptional records. What may very likely have occurred in the area of spoken language was an extremely gradual elaboration and "complexification" of the linguistic component of ancient Bahnese culture A higher-status, more "refined," Javamzed idiom emerged among the social elite, for the arts, religion, literature, and government were all dependent on a written tradition dominated by the use of Old Javanese. For villagers geographically and socially distant from the courts, comm o n spoken Bahnese (related to the language of the earlier edicts) remained the usual mode of expression. When dealing with the elite, of course, or when involved in ritual or other activities encompassing a written tradition, some Bahnese would have to be able to operate using the more "formal" or "refined" code. This pattern, which I sketch only roughly here, accounts for the fact that almost all the lexical items from the Old Bahnese inscriptions that are still in use in modern spoken Bahnese are of the "lower," common register From the time of the first edicts in Old Javanese through at least the sixteenth century, features of Hindu-Javanese cultural influence become increasingly difficult to separate from possible indigenous or Hmdu-Balmese ones Both textual and archaeological evidence suggest that there was some cultural similarity between Bah and at least the eastern part of Java, but the origins and paths of transmission of certain phenomena, such as temple architecture, remain murky. And such is the case when we look at the continuing influence of Javanese language and written tradition in Bah. Other characteristic changes in the Old Javanese language edicts involve demonstrative pronouns, with Old Bahnese hento or tua 'that' becoming Old Javanese lka(ng), deictic "pointers," with Old Bahnese ditu 'there' becoming Old Javanese ngkana, and prepositions, with Old Bahnese di 'at' being replaced by Old Javanese (r)i Certain set phrases are also entirely substituted, as in Old Bahnese kabudhi kabudhi 'forever and ever', which becomes Old Javanese n(ng) dlaha rung dlaha, with the same meaning Old Bahnese kangm kaiuh klod kadya

Old Javanese 'east' 'west' 'seawaid' 'mountainward' 16

wetan kulwan/kulon hdul loi

L A N G U A G E PATTERNS AND THE L I N G U I S T

Like our knowledge of the language in pre-Hindu times, we have lit­ tle evidence of how Bahnese was spoken before the adoption of Old Javanese as the court medium We do know that for a millennium the Bahnese have performed, studied, and continuously recopied thou­ sands of manuscripts in Javanese idioms 2 7 The wealth of manuscripts in Bah, which began to be centrally collected and catalogued in the late nineteenth century, represents not only the "classical" Old Javanese (such as that of the eleventh- and twelfth-century East Javanese kakawin, poems in Indian-based meters with a heavily Sanskrit lexicon), but also more recent "dialects" (such as are found in the kidung, poems in indigenous meters written in Java and, primarily, in Bali from the sixteenth century on) Presumably, all this Javanese "literature" (in the strict sense of "written" language) could not but affect the native tongue of generations of Bahnese so intent on preserving texts that were geographically, temporally, and linguistically distant from their own idiom In fact, there is a great deal of apparently Javanese borrowing in both spoken and written Bahnese The Bahnese themselves, while not par­ ticularly interested in tracing historical threads of transmission, feel that "Kawi" (meaning ancient and medieval idioms, ι e , Old Javanese or Middle Javanese or combinations of these with Bahnese) as well as spoken Javanese were increasingly common in Bah beginning in A D 1343, when the island was officially brought under the Javanese king­ dom of Majapahit as a result of Gajahmada's conquest 2 8 At this point, say the Bahnese, aristocratic Javanese began to settle in Bah, culmi­ nating in a virtual flood of noble refugees when Majapahit disinte­ grated (in the face of growing Islamization of Javanese courts) in the late fifteenth century As a result of this influx of Javanese priests and princes who were "fleeing Islam" and who established Majapahit-style courts in Bali, a number of important things happened First, the uppercaste Bahnese lineages were founded through the intermarriage of prominent Bahnese families with Majapahit royalty Second, religion and the arts were influenced by another great wave of Javanese ideas, specifically the Majapahit traditions, many of which are embodied in Kawi literature And finally, the spoken language of the newly arrived Javanese affected spoken Bahnese While there are few specific ideas about how this latter process took 27

See the catalogue of Pigeaud (1967 70, 1980) for some idea of the extent of this manuscript tradition 28 This episode is well known from the fourteenth-century court chronicle of King Hayam Wuruk's Majapahit, the Nagarakrtagama |Pigeaud 1960 63)

17

THE SHAPE OF THE WORD IN BALI

place, the Balinese scholar Ketut Ginarsa has suggested one theory Ac­ cording to Ginarsa, in the most ancient period Old Balinese was influ­ enced first by Sanskrit, particularly in orthography and lexicon Fol­ lowing this era was a pre-Majapahit period of contact with Javanese culture, which resulted in the alteration of the Sanskrit-Bahnese mix­ ture by Old Javanese Under Majapahit and later influences, the pre-ex­ isting mixture of Old Bahnese/Sansknt/Old Javanese was increasingly used as the non-honorific code for daily life and informal situations Meanwhile, honorific speech style began to incorporate the spoken Javanese used in Majapahit-styled Balinese courts This code very likely was a less formal Javanese style (which would nowadays in Java be called ngoko), used by the heirs of Majapahit to familiars and to per­ sons of lower status That this less formal style should be integrated with an evolving elevated speech "register" fits with the sociological facts, Balinese individuals would wish to address the nobility of Java­ nese descent using Javanese-derived vocabulary associated with that group This explanation accounts for the noticeable presence in mod­ ern Balinese alus ("high," polite) vocabulary of many Javanese terms classified as "low" {ngoko in Javanese or kasai/kapaia in Balinese) The theory is graphically summarized in Figure 1 Western scholars have not yet addressed the topic of Javanese-Balinese linguistic interaction specifically Focusing almost entirely on written language from a traditional philological standpoint, their ap­ proach has involved comparing and editing collected versions of a sin­ gle work In this field the question of what is Javanese and what is Ba­ linese in the manuscripts is most often unanswerable, except for the oldest Kawi works which independent evidence (such as inscriptions,

TIME

LINGUISTIC BASE

A D 50» (first inscription A D 8 8 2 )

Old Balinese Iτ Sanskrit Balinese | -« Τ Kawi Bali

AD

1100

A D 1400 AD

1980

Kasar Balinese

Alus Balinese

SOURCE OF INFLUENCE (historical period)

Sanskrit (pre Javantzation) Old Javanese/Kawi (pre Majapahit) (Majapahit and after) Javanese

Dates are intended to reflect the general histonial per od and do not represent specihe events

FIGURE 1

Theory of the evolution of spoken Balinese 18

L A N G U A G E PATTERNS AND THE L I N G U I S T

place names, and so forth) definitely places on one island or the other We face a situation in which "archaic" linguistic codes were not only preserved but also used to create new works As Berg (1927) discovered with his work on the "Middle Javanese historical tradition," largely based on texts composed in Java-styled courts in Bah, and as Worsley somewhat ruefully comments (1972 94-95), the fact that a manuscript is in Old or Middle Javanese does not necessarily mean that it origi­ nated in Java, or that the surviving manuscript is not a Balinese version as well as a Balinese copy Zoetmulder (1974) repeatedly cautions against the dating of texts based only on internal linguistic evidence, yet it is this same evidence to which he and others must continually return in the attempt to sort out Javano-Bahnese literature according to authorship, time of composition, and place of origin It seems, after surveying all the difficulties involved in sorting out Old Javanese from Middle Javanese from supposedly later "Kawi-Bah" codes, that a chronology based on linguistic evidence is not only diffi­ cult to establish but in some cases irrelevant as well Similarly, after accepting "the anonymity of the literary traditions" (Worsley 1972 87), in which authors' names are only rarely mentioned and take pseudonymic form, and the confusion that attends the determination of a text's provenance as actually Javanese or as part of Javanese-idiom literature created in Bah, it seems that authorship, date, and place of origin are likewise not the only principles around which we should or­ ganize our studies of this major Southeast Asian literary tradition As this book proceeds, it is hoped that other, more contextually grounded notions regarding Balinese languages and literatures will emerge One such notion was mentioned earlier, m the discussion of the Blanjong pillar Old Balinese kings who established monuments and commissioned charters in Old Balinese and Sanskrit languages occa­ sionally must have found it necessary to gloss, paraphrase, or other­ wise translate the formal, ritual language into the local idiom The so­ cial utility of making commands and regulations understood would require the king or more likely his representative to reword parts of royal decrees in common Balinese The Blanjong pillar is a unique case in which an inscriptional message is encoded in two different lan­ guages 2 9 However, it does set a precedent for something that I wish to 29 The pillar is also unusual in its Indian or Asokan form, with a carved lotus seat atop a column Three such pillars have been found in Bah (Sukarto Κ Atmodjo 1967 23-25) In speculating about the purpose of the Sanskrit part of the edict, it is useful to note that in India itself the adoption of Sanskrit as an "of­ ficial" code probably reflects sociopolitical changes that enhanced the prestige

19

THE SHAPE OF THE WORD IN BALI

argue became a basic textual feature in all Balinese literature, namely, the translation or paraphrase of a more "distant" code in terms of a code "nearer" that of the hearers or audience. According to this principle, as Javanese literary idioms became known in Bali, they too would be recast in the common person's tongue. 30 The reading of literary forms—in the earliest periods, perhaps even inscriptions—in the context of oral performance establishes the basic structure of Balinese literature. The "older" code is accompanied by paraphrase into a more contemporary one. Both the "original" archaic idiom and the paraphrase form parts of one "text." In such a case, archaic styles, one might predict, would not disappear. Some specific ways in which this process has operated, along with the implications and results of this principle as it has worked in Bah for more than one thousand years, form part of the discussion immediately following. Synchronic Considerations With the advent of the electronic age, students of language could begin to take for granted the taped preservation of actual speech. In the past, such sounds were inevitably unrecoverable because they were unrecordable. Today in Bah it is no longer unusual to see the microphones and cameras of foreigners at work recording the particulars of the island's speech, theater, and music. Yet the Balinese themselves, without the techniques of electronic or even print reproduction, through many centuries have sustained a remarkable variety of linguistic codes and styles that are still alive and still uttered. This variety reflects the course of linguistic and cultural history discussed above. As an example of how the languages of the past are part of present experience, Sanskrit is still regularly heard in the ceremonial songs and invocations of the padanda, consecrated priests of Brahmana caste. Old Javanese prose and poetry from as early as the ninth century, as well as from later periods of the East Javanese kingdoms, the Majapahit empire, and post-Majapahit Balinese kingdoms, are still read and performed; these are the parwa and kakawin forms, primarily. Other of the ancient heritage The earliest dated Sanskrit inscriptions are later than the Asokan edicts, which were in Prakrit After the second century A D , mixed Sansknt-Prakrit inscriptions occur, and by the eighth and ninth centuries the language is entirely Sanskrit 30 In early Javanese literature, the concept of retelling or rewriting a text in another form was referred to using derivations of the Sanskrit loan piakrta, thus, pmrakrta could describe language or a story that was 'retold, translated', and amiakrta variously meant 'perform, (re)tell, gloss' (see Zoetmulder 1982 1389) 20

L A N G U A G E PATTERNS AND THE L I N G U I S T

Old Javanese texts are used in religious study, temple ceremonial, and a variety of traditional professions. Other forms of Javanese language, generally referred to as the "Middle Javanese" idiom, are preserved in the kidung poetry sung to accompany rites of passage and temple ceremonial, or in the courtly gambuh theater. This linguistic form is also extant in many prose dynastic genealogies, or babad, which are often brought to life through masked dance-drama (topong) performance.31 "Literary" Bahnese language, sometimes called "Kawi-Bah" style, is the code for poetry in the geguntan form, which is dramatized through the songs of the "operatic" arja and is also popular as entertainment and accompaniment for social and ceremonial occasions. These are only a few of the uses of archaic linguistic codes and traditional literary forms in present-day Bah. Archaic codes also appear m a vast number of chants, charms, and invocations sounded aloud, softly mumbled, or silently recited. The occasions for such language use are as diverse as the activities of daily life itself—the construction of a building, the start of a business enterprise, the onset of a stage in an agricultural cycle, embarking on or returning from a journey, falling in love, the treatment of illness, or the recovery of a lost object All these situations may call for the use of verbal formulas whose force in part derives from the antiquity of their origins and linguistic form, and which are generally referred to as mantra. What underlies and ties together all these uses of the language of the past, echoing in diversity its range of codes and styles, is a manuscript tradition whose palm-leaf lontars32 number in the thousands. For rather than being strictly a "prehterate" example of oral transmission, each style of uttered ancient language presupposes a genre of written text as its background. The complex and mutually supportive relationships between writing and sound in Bah will be a continuing focus of this study While an exhaustive typological description of Bahnese manuscripts has never been attempted,33 Pigeaud's Literature of Java (1967-1970, 1980) classifies and catalogues many of the genres extant in texts found 31

Recent treatments of topeng performance in Bah include Emigh's (1979] description of the tradition and Kakul's (1979) performance transcript, which is translated into English See also Young (1982), who provides transcribed Bahnese text and English translation 32 The etymology of lontai is traced to a metathesized form of zontal, from ron 'leaf plus tal 'borassus palm' Other terms used to refer to the manuscript type are ental, cakepan, kropak 33 Hooykaas (1979) provides an introductory overview of many genres and styles found in scribal form

21

THE SHAPE OF THE WORD IN BALI

in Bah. Unfortunately (from the Bahnese viewpoint), Pigeaud tends to regard the manuscripts from Bah as an extension, or as remnants, of Javanese culture and literature, rather than from the perspective of an ongoing Bahnese tradition. His summary of four categories of Javanese literature, four periods of Javanese literary activity, and five Javanese literary idioms (1967:45) tends to obscure the fact that many, perhaps the majority, of the non-Islamic texts of this "Javanese" literature were collected in Bah and in the Bahnese kingdoms of Lombok. We are thus left uninformed as to how a great body of manuscripts, orphaned from its "true" Javanese sources, came to be adopted and nourished by the Bahnese.34 Certain scholars (Robson 1972, Zoetmulder 1974) have more directly alluded to a Bahnese role in the preservation of so-called "classical" Old Javanese literature, particularly the kakawin (stylized poetry in syllable-quantified meters derived from the Sanskrit kavya tradition), the parwa (Old Javanese prose retellings of the Sanskrit Mahabharata), and the kidung (romantic and historical poetry whose content and metrical form are indigenous). It is clear that most of the extant texts of these three genres would not be known if Bahnese manuscripts did not exist: "It is to Bah that we owe the survival of Old Javanese literature to the present day" (Zoetmulder 1974:21). Thus when we talk of "Hindu-Buddhist" or "pre-Islamic influence" Javanese literature, we are also implying the sustained activity of generations of Bahnese readers and audiences, who both preserved the old works and created new texts based on older models. When we look for the "how" and "why" of this complex Bahnese manuscript tradition, we return to the synchronic pattern of ritual, theatrical, and daily-life usage of multiple languages sketched above. While I do not here presume to classify all the forms and functions of Bahnese written tradition, a brief, functionally oriented overview, presented in Table 1, gives some of the important categories, uses, and users, of manuscripts in Bah. There are several points that must be made before one is simply swept away by the mass and variety of Bahnese manuscript traditions. First, as I indicated earlier in this chapter, the usual notions of the "provenance" of a literary work (its author, date, and place of composition) fail to produce an adequate picture of the historical and socio34 Some Bahnese literati still prefer the general term "Kawi" for all archaic styles, asserting that the more specific Bahasa Jawa Kuno (Old Javanese), used to refer to the kakawin language, gives the impression that the texts do not really "belong to" the Bahnese

22

L A N G U A G E PATTERNS AND THE L I N G U I S T

cultural context of Balinese literature Likewise, a classification based on purely linguistic criteria is unsatisfactory, since even today a Ba­ linese may choose an archaic code, with little relation to everyday lan­ guage, as the vehicle for a literary work The serious emulation of Eliz­ abethan English by an American playwright today is felt to be an anachronism, however, the imitation of twelfth-century East Javanese poets by present-day Balinese is not considered so Even widely accepted Western criteria of genre as commonly deter­ mined by a text's "form" and "content" fail to yield a coherent typol­ ogy A single episode from the originally Indian epic Ramayana—for example, the advice of Rama to his brother Bharata—might appear in any number of metric or prose versions inscribed over a period of sev­ eral centuries in various linguistic styles Some commonly accepted generic names, such as kakawin or kidung, used in the field of Balinese literature actually specify metrical type and vocal realization without predicting content Yet even these categories are not airtight, since the meaning of a text depends so greatly on the setting of its performance The kidung Malat, when sung to accompany a tooth-filing ceremony, has a different signification than the singing of Malat as part of a the­ atrical character's stately evocation of the Majapahit courtly heritage Other generic terms, such as tutui 'teachings, revelation' and αμ 'sa­ cred text', refer to rhetorical structure, which often is that of a guru 'teacher' didactically addressing a pupil, rather than to form, which may include both prose and metric elements The difficulties of categorizing Balinese literature satisfactorily should provide us with clues as to the inadequacy of our own literary models Walks (1980) has shown that classification of much Balinese literature must be to a certain extent based on vocal musical criteria, since it is as oral performance, as singing rather than silent reading, that poetic forms exist for the Balinese Yet scholars have persistently chosen not to look at performance and contextual aspects of literature in Bali Many, both Western scholars and Javanese literati, prefer to view this literature as if it were an accidentally preserved relic from the classical Javanese past The regrettable thing about this view is that it makes the Balinese into passive archivists, mechanically copying palm-leaf books whose true significance they only dimly grasp, and that is a serious misconception The professional prejudices of some cataloguers and traditional phi­ lologists have tended to obscure other facets of literature as well The work of critical editing is based on Western, not Balinese, notions of what a text is While the present list of available editions of Balinese 23

TABLE 1 . SOME USERS AND USAGES OF TEXTS IN BALI

T H E S H A P E OF T H E W O R D IN B A L I

LANGUAGE PATTERNS AND THE L I N G U I S T

works in European languages includes extremely valuable contributions,35 it should not be forgotten that these editions are not just translations but complete transformations of literary acts and literary entities We need a clearer understanding of how the biases of Western literary studies have colored our understanding of "literariness" as it exists for the Balinese. It was pointed out earlier that in the linguistic history of Bah new languages repeatedly came on the political and cultural scene without wholly eliminating the pre-existing codes Instead, each language mode has become associated with certain events, settings, participants, and acoustic features. Sanskrit chants, ancient Kawi poetry, modern "literary Balinese" verses, and the changing nuances of everyday spoken Balinese with its chne of intimacy versus distance—all of these, when coupled with appropriate contextual features, have come to represent different "worlds" within and between which the Balinese are happy to move. Sanskrit, for example, is perceived as a ritual medium specific to the religious specialist addressing himself to the deities. Old Javanese evokes a heroic, partly Indie world conveyed through both the oral performance of literary texts and the theater. Middle Javanese has associations more specifically Javano-Balmese in nature, and it also serves to unite groups of people in communal worship of the ancestral past.36 The preservation of archaic language modes continues to be necessary in order to summon the past in the present, to recreate the various worlds making up the Balinese universe Simultaneously, the strategy of using the vernacular as a translation "bridge" between the past and the present has been maintained, and, in turn, it reinforces the viability of archaic ways of speaking. The oral performance of texts continues to be the primary path to enjoyment of the Kawi literary heritage, and such performance is marked by a structure in which each textual unit is accompanied by vernacular paraphrase37 Similarly, in the more than one dozen forms of theater currently popular, we also find the use of multiple languages and literary styles Thus the pattern of linguistic multiplicity pervades Bah, and after considering the linguistic history described earlier, we can detect as a 35

See Zoetmulder (1974 52) for a list of some of these textual editions, see also Hooykaas (1978), Drewes (1975), Supomo (1977), Santoso (1975), and Teeuw and Robson (1981) 36 See Walhs (1980 chap 6) for an analysis of kidung performance in Bali today 37 This is discussed in Chapter 3 25

THE SHAPE OF THE WORD IN BALI

cultural constant the presence and use of many codes among one rather small population.38 Variation in code and style is not limited to the realms of written literature and the theater, however. Spoken language also partakes of the principle of variable code and context. Just as different literary modes are appropriate for different rituals or theatrical situations, so there is a range of spoken forms that varies according to the relations between participants in the speech situation.39 Modern Balinese does, of course, predominate in daily speech, but even this linguistic setting is not so homogeneous as one might think. Each Balinese is also a member of the Indonesian nation, and the national language [Bahasa Indonesia) is regularly and widely used in education, government, the mass media, and other institutions that increasingly reach toward even the most isolated villages. Moreover, there are throughout Bah small enclaves of various minority ethnic groups whose languages are accepted as familiar sounds, especially around the markets of regency [kabupatan] centers. Thus one hears other Indonesian languages, such as Madurese, Sasak (from Lombok), Javanese, and Buginese (from southern Sulawesi), in the northern, eastern, western, and northeastern coastal areas, respectively. Many older Balinese, especially those educated during colonial times, can still speak Dutch, and they use certain Dutch expressions regularly. Other people remember phrases of Japanese, a vestige of their experience as an occupied territory during World War II. While most of the ethnic Chinese residents of Bah, who have for centuries been merchants and are located in larger towns, use Balinese and Indonesian, there are still some speakers of dialects of southern China. Other non-Austronesian languages are heard with increasing frequency, having become economically and socially important as Bali's place in the itineraries of international travelers has become established. English, French, Japanese, German, and Italian have all become potential channels of communication. Tour guides, hotel workers, airlines and taxi personnel, and owners of souvenir shops all try to gain some skill in one or more of these languages in order to better tap the resources of the tourist trade. The discussion in this section, while not exhausting all possible variations in spoken and written language in Bah, has nevertheless established the presence of such variation and its dimensions. Operating 38 Bali's population was estimated to be between 500,000 and 800,000 at the beginning of the nineteenth century Censusfiguresfrom later periods showed 1,101,400 in 1930, 1,782,500 in 1961, and 2,475,555 in 1980 39 This topic is treated in Chapter 2

26

L A N G U A G E PATTERNS AND THE L I N G U I S T

between a number of linguistic codes, involving a range of sociocultural settings, infusing the awareness of all island residents, "multiplicity" is the key to linguistics in Bah. The aim of my studies in Bali was to search for patterns in this variable linguistic landscape. The focal point within this sphere is literary expression—an area that is perhaps more fittingly termed "verbal art," since it includes both written and sounded forms. At this point it is useful to consider a number of approaches to the study of verbal art, by way of highlighting central issues for my argument as a whole. THE LINGUIST'S VIEW Having discussed the general historical context and present configuration of language systems and literary traditions in Bah, I now turn to certain conceptual issues that might be useful to the linguist trying to understand language and literature in a cross-cultural situation. More specifically, I will look at notions of "literary structuralism" and "discourse theory" which have been brought forward by scholars attempting to come to terms with literature as a linguistic phenomenon. Following that, I will consider ideas from the field of literary studies proper which are germane to the enterprise of understanding "discourse" or "text," in the sense of language units larger than the sentence. The purpose of this rather theoretical interlude, to use a metaphor, is not to advocate the universal application of a specific and limited set of tools in mining the ore of raw language data. Yet throughout the field experience on which this book is based, and in post-research organization and analysis, certain concepts have been heunstically useful. Even a faulty theory can be useful, inducing errors and misjudgments that eventually may direct us toward new insights and explanations. Textuahty and Discourse One of the important contemporary fields of study contributing to the understanding of literature has been that of structuralism. A major impetus in the development of the theory of literary structuralism has been the conviction that principles of linguistic science could be appropriately and fruitfully applied to poems, plays, and other works of verbal art that traditionally were considered to he outside the province of linguistics itself Accordingly, structuralists looked to the primary Saussunan postulate in linguistic studies: the distinction between 27

THE SHAPE OF THE WORD IN BALI

langue, the language system, and parole, the particular utterances manifesting that system. In an extension of the langue/parole notion to another field, students of literature asked, What are the formal laws and aesthetic norms that make up the "system" of literature manifested by particular works? How can we generally express the forces shaping the production of and response to literary texts? In other words, How can we characterize "literariness," that entity that distinguishes literary language from ordinary speech, forming the "langue" of which literary texts are the "parole"? In the more than fifty years since these questions began to be asked, structuralists have pursued the basic units and models of segmentation underlying literary texts. Some early efforts focused on the functions of elements in poetics; other approaches emphasized the search for universal laws of narrative.40 Students of myth attempted to generalize formal conceptual oppositions that revealed psychological and sociological truths.41 All these studies tended to see the "minimal unit" of literature operating in the same way as the phonemic contrast underlying speech production. Eventually this model changed, as sequences of logical figures began to be posited, in the work of "text grammarians," as the basis of narrative structure 42 The linguistic model of literary structuralism changed from that of phonology to syntax: if there is a grammar of the sentence, it was said, there must also be a grammar of the text. The "text grammar" concept developed elaborately, with universal laws of combination, finite categories, and closed-system structures m line with the generative theories of American linguistics.43 Other structuralists, emphasizing issues of code and the verbal message of a text, developed the general science of signs called semiotics, which carried the analysis of signifying systems far beyond the realm of verbal texts.44 It should be clear from a survey of prevailing structuralist 40 Jakobson (1960) set out many fundamental structuralist concerns in a classic essay, for examples of structuralist analysis of poetics, see his treatments of Baudelaire and Shakespeare (1970a, b) Propp's (1958) work on narrative "functions" was likewise widely influential 41 This more anthropological study of texts is epitomized by the work of L€viStrauss(1955) 42 Examples of such work are Greimas (1966) and Bremond (1973), the analysis of literary genres by Todorov (1969, 1977) is a related development 43 The "formalist" bent of the text grammarians is represented, for example, by van Dnk (1972) 44 Barthes (1968, 1975) has been the acknowledged leader in this field

28

LANGUAGE PATTERNS AND THE LINGUIST

approaches45 (which have dominated poetics and the linguistic study of literature) that many elements are generally left out in such theories. While trying to characterize the system that underlies and constrains the production of texts, structuralists seem to have lost sight of the text as an actual linguistic entity. Instead of considering real arrangements of sounds or words, scholars often work only with semantic abstractions from texts, such as "motifs," "movements," or "mythemes." Leaving line-by-line explication to literary critics, structuralists by and large have aimed toward universal laws that can then be applied to literary works. Yet once we are immersed in, say, some version of Propp's thirty-one functions, is it ever possible to go back to an actual text with its inevitable "residue" of elements that do not quite fit the analytic scheme? Another objection to the structuralist bent, along the same lines, comes from the field linguist's intuition that without understanding each word in a text, working only from translations, one is dangerously far from the sources of not only formal but also aesthetic coherence. If iakobson's "poetic function" has to do with the aesthetic response elicited by the verbal material of a text, how can scholars ever consider the "artistry" of literature in another language tradition without firsthand knowledge of that language? This argument is important in the case of Bah, where language and texts are very different from the IndoEuropean material that formed the basis of much structuralist analysis. Translations and abstractions, which are actually transformations, allow too many opportunities for unjustified concepts to slip into the interpretation; the analysis of the formal coherence of texts in foreign languages must be firmly anchored in the language of the text. A telling critique of much literary structuralism would point out a crucial flaw in the basic premise that formal linguistic principles (of phonology, morphology, and syntax) are also the substance of literary studies, and that "literariness" is somehow analogous to "wellformedness" in language. Yet perhaps this rather metaphorical application of linguistics to literature—an application that sees a narrative unit in terms of a phoneme, or hears in a novel's plot the echo of sentential syntax, or senses underlying all texts a "deep structure" that inexorably "generates" literary creations—has been carried too far. While the linguistics-based study of literature is necessary to an un45

More detailed examinations of the structuralist view are included in Culler (1975) and in Barthes's essay entitled "Introduction to the Structural Analysis of Narratives" (1977 79-124) 29

THE SHAPE OF THE WORD IN BALI

derstanding of the full significance of a work of verbal art, it still does not follow that all linguistic elements have literary analogues that behave in precisely the same fashion. The formal features operating within a sentence are of a different order than the textual and con-textual elements contributing to meaning in a language unit larger than a sentence. The analysis of texts, or of the larger-than-sentence entity which many linguists call "discourse," requires knowledge of the wellmapped territory covered by formal linguistics and other sorts of analysis as well. Certainly there can be no ultimate objection to the use of linguistic postulates as starting points in literary analysis. Important contributions have been made by linguists to the understanding of "discourse," an area that may be seen to include literary as well as oral "texts." Indeed, the much-needed corrective that linguistics can provide for literary structuralism is the revelation that texts, like sentences, are not made up merely of concatenated units like beads on a string, but are composed of hierarchies of different types of grammatical relations.46 Linguistics also can provide a great deal of insight into verbal art precisely because linguists do concentrate on actual verbal material, rather than (as in some endeavors in styhstics, poetics, and semiotics) using linguistic metaphors for problems of textuahty and aesthetics toward which the linguistic principles themselves were never directed.47 In addition to questions of literary form, another area of textual meaning where structuralism has often fallen short is the tradition within which any literary work is contextuahzed. Audiences do not approach plays, or readers books, without certain expectations regarding "type" or "genre." To each text the creator or author and the hearer or reader bring the accumulated experience of the texts they have encountered before. To be fair, it should be noted that structuralists have been spared the duty of pointing out exactly what is meant by "folktale" or "novel" because both the scholars and their readers share assumptions about and experience of those literary types. But when we look at Bali's unfamiliar and complex variation of literary material, we have to consider the place of the text in literary and cultural traditions, as this bears on both form and meaning. We are, then, beginning to 46

Pike's (1967) and Longacre's (1972) concept of the tagmeme is exemplary, linguists such as Grimes (1972) and Halhday and Hasan (1976) have pointed out other basic dimensions of discourse linguistics 47 Along these lines, see the critique of some structuralist perspectives in "Surfacing from the Deep" by Smith (1978 157-201) 30

LANGUAGE PATTERNS AND THE L I N G U I S T

look for ways of bringing text-external material into the analysis. From an ethnohnguistic viewpoint, the analysis of speech styles takes account not only of structural aspects (relations of forms), but also of "use" factors (meaning of verbal features in terms of non-linguistic contexts).48 In addition, sociohnguistic studies have enabled us to better understand how social differences influence stylistic norms in speech. The corresponding problem in literary studies, that of genre, must be addressed in an analysis of discourse. One way to understand recurring issues such as what is "normal" or "deviant," "traditional" or "innovative" in literature is in relation to the presuppositions and expectations based on experience of "prior text," a term that includes the awareness of genre. Related to the idea of genre or prior text is another notion that is often downplayed in structuralist analysis. This is "intentionality," the speaker or text-creator's purpose in conveying the text. This idea is also extra-textual and contextual, taking as its base the factors inherent in every communication act in which some speaker, hearer, and third-person "other" are participants. Intentionality really includes two types of information. One is the prepositional content of utterances, the logical and semantic component of discourse, the "meaning" that is relatively explicit in the linguistic forms of the utterances. The other is the relational or pragmatic content, having to do with the mutual attitudes and motives of speakers and hearers, which are often implicit and expressed through parahnguistic means.49 The notion of prepositional content, lending itself to logical and rhetorical considerations of whether an utterance is a "statement," "vow," and so forth, is nicely counterbalanced by pragmatics, in which paradox and unresolvable ambiguity are features of communication. In the case of Bali, the outsider's understanding of intentionality cannot always be so subtle, as we begin with the basic task of determining the linguistic forms that show prepositional and relational content in texts. Moreover, in written literature the "speaker" and "hearer" become complexes of different personae, and when an orally performed literature such as Bali's is taken into account, relational content becomes closely tied with text as performance event and text as channel of communication. There is one more type of information about texts that is pertinent 48

SeeHymes(1968, 1972) See Labov (1965), who showed that unconscious loyalty was a crucial notion for the understanding of speech styles in Martha's Vineyard, and R Lakoff's (1973] illustration of the functions of syntactic choice in conveying politeness 49

31

THE SHAPE OF THE WORD IN BALI

here, and like the preceding concepts of prior text and intentionality it concerns things extra-textual This is "reference," that component of discourse that points to the world outside the text by linguistic means Reference is linked to individual signs within a text and also to higherlevel networks of signs constituting culturally shared knowledge The matters in the extra-textual world to which a sign refers can be of varying complexity, just as the linguistic sign can be either denotative or poetic and metaphorical This area of textuahty takes account of the fact that semantic elements are not equivalently expressed across cultures, and that different languages imply different systems of naming or "referring " 5 0 The discussion in this section has presented four general types of information that give coherence and meaning to a text The first of these is internal structure the hierarchies of parts of speech, phrases, sentences, and larger-than-sentence elements, whether these are called paragraphs, act sequences, movements, or scenes The second type of information is prior text, or constraints on form and content that locate a text within a tradition and fulfill or violate expectations and assumptions about the text's "type" or genre The third type of information about discourse is intentionality—the identification of participants in the "speech" situation, the types and functions of things participants "say" (speech acts), and the relations between communicating participants evoked or presupposed by the discourse Finally there is reference, the way discourse uses language to point beyond itself to the world outside discourse Taken together, these four types of information provide a heuristic approach that is hopefully broad enough to avoid imposing categories derived from one language 50 The concept of reference and its importance to discourse have been well expounded by the language philosopher Paul Ricoeur (1976), who differentiates between two fundamental branches of the science of language semiology or semiotics, which concerns the closed system, virtual, immanent relations of signs within language, and semantics, which concerns the contextual, actual, event based aspects of language as discourse Within this latter branch, the function of reference is to point to the extra linguistic world, it is a nonstruc turalist function, since reference only reveals itself in the use of language, not in language's abstract, synchronic existence Reference is contrasted with sense in Ricoeur's thought The latter involves the "what is said" of an utterance, that is, the correlation of predication (application of universal categories) and identification (pointing out of a logical subject), the former involves the "about what" of an utterance and shows language "transcending" itself "[R]eference relates language to the world It is another name for discourse's claim to be true"(p 20)

32

LANGUAGE PATTERNS AND THE L I N G U I S T

onto the texts of a quite different one. Recognizing both intra- and extra-textual sources of information about discourse, we escape the lure of the "formal fallacy," whereby an analysis dispenses with the idea of a cultural system interacting with a cultural product such as a literary text. These four types of information are necessarily interrelated, forming part of a culture's "text-building" strategies. The types of textuahty brought forward here were not fabricated out of whole cloth. They reflect a desire felt by many students of language for models of discourse beyond the level of taxonomies, which are static structures, or generative devices, which are linear mechanical processes. A series of exchanges in conversation, or any literary text, cannot be adequately accounted for without the possibility of feedback, self-correction, and constraint, as are encompassed by cybernetic devices; notions of purpose and transformation, as are provided by the model of the biological cell, and perhaps even more complex systems, as are exemplified by animal organisms and communities of organisms.51 The explanation of discourse involves complicated investigation of humankind's communicative powers, and the exploration of this realm has just begun. Faced with such a complicated task, one may despair of ever coming to grips with textuahty; however, the four areas of discourse information I have been discussing have already been shown to be useful in describing the dramatic and literary tradition of Javanese shadow theater, or wayang. A. L. Becker (1979) has outlined the plot structure of wayang performances using a hierarchy of text-internal elements, and he has demonstrated the performer's attention to prior text in his linguistic and thematic choices. He has also described the constraints posed by speech situational factors and the strategies of "naming" that form the referential component in wayang plays. After considering this work on an Indonesian traditional genre, we find it easier to proceed in the area of Balinese literature. First, however, it is necessary to look at one more area of scholarship that has an important bearing on the present study. Noetic Economy Throughout the discussion up to this point I have used terms such as "literature," "literary tradition," "oral performance," and "text" without always specifying how they are to be understood. The present sec51

See Boulding (1964) for a well-made argument demonstrating various levels of systemic complexity 33

THE SHAPE OF THE WORD IN BALI

tion aims to clarify these terms against a background of general literary studies. The modern scholarship of "oral tradition" or "oral epic" was established with the work of Albert Lord and Milman Parry, who outlined the forms and processes of what they saw as the prehterate Homeric tradition of orally composed and performed epic poetry, specifically as this tradition was recorded by Parry m Yugoslavia during the 1930s. The fundamentals of their theory of oral tradition are expressed in the following extract. [Ojral epic song is narrative poetry composed in a manner evolved over many generations by singers of tales who did not know how to write; it consists of the building of metrical lines and half lines by means of formulas and formulaic expressions and of the building of songs by the use of themes.. . . By formula I mean "a group of words which is regularly employed under the same metrical conditions to express a given essential idea." . . . By formulaic expression I denote a line or a half line constructed on the pattern of the formulas. By theme I refer to the repeated incidents and descriptive passages in the songs. (Lord 1976:4) In the years since their work appeared, Lord and Parry's conclusions have been the starting point for much debate. While generally considered valid for most aspects of the Yugoslav guslar's art, some problems arose when this theory was applied to other traditions, such as Old English or African literatures. The notion of oral poetry as a distinct entity completely uncontaminated by any contact with writing is unrealistic in most parts of the world. Regarding Lord and Parry's most basic notion of the formula, it is in some cases ultimately impossible to distinguish instances of formulaic speech from the general phenomenon of poetry as "patterned" expression. Perhaps Lord and Parry's most valuable insights he in their appreciation of the Yugoslav singer's—and, by extension, Homer's—skilled art as radically different from "literature" as experienced and presupposed by the literate minds investigating oral traditions. That there need not necessarily be a fixed, pure textual form underlying the wide variations characteristic of oral performances of a story; that the singer of tales works in harmony with the situational context and the audience; that the poetics of orality is a complex interaction of traditional constraints and language alive in the here and now—all these ideas were indicated by Lord in passages like the one following. 34

L A N G U A G E PATTERNS AND THE L I N G U I S T

Each theme, small or large—one might even say each formula— has around it an aura of meaning which has been put there by all the contexts in which it has occurred in the past. It is the meaning that has been given it by the tradition in its creativeness. (Lord 1976:148) Continuing investigation of oral tradition suggests that it is indeed our own Western literate condition that needs to be examined if we are to adequately comprehend the meaning of "oral" verbal art. Eric Havelock undertook to show the roots of Western loss of a hypothetical "oral state of mind" in trying to understand why poets were outlawed in Plato's Republic Instead of the integrative, experiential, mimetic tradition of music and words shared by poet and audience in oral performance, Plato wanted to establish an analytic, abstract, reflective consciousness with a separation between "the personality which thinks and knows" and "a body of knowledge which is thought about and known" (Havelock 1963:201). Without the objectification of knowledge as written script, it would be difficult to conceive of separating knower from known, says Havelock, and thus Plato's enemy was orahty, defined as the prehterate transference of cultural knowledge through patterned speech performance, that is, poetry. The work of Walter }. Ong, from which I have taken the heading of this section, represents an important step in analyzing the implications of the oral literature concept. Ong (1977 44) uses the term noetics to refer to the specific processes of producing, remembering, recalling, and sharing through which different cultures manage their knowledge store.52 In terms of modern information science, noetics is thus the system of shaping, storing, retrieving, and communicating knowledge. Ong has treated in great detail the changing "noetic economy" of Western society from classical times onward, distinguishing characteristics of prehterate "primary orahty" from those of "residually" oral, manuscript, print, and electronic media stages (1967). In the state of primary orahty, cultures are obsessed with the transience of knowledge as embodied in the spoken word, and the mnemonic qualities of 52

Noetic is defined in the Oxford English Dictionary as "of or pertaining to the mind or intellect, characterized by, or consisting in, mental or intellectual activity " The original Greek root meant "to have mental perception or intelligence " The term has gradually come to be used in philosophy and aesthetics to refer to a science of the intellect, of speculation and investigation, to a training of the mind and, in Ong's terms, a conditioning of consciousness 35

THE SHAPE OF THE WORD IN BALI

patterned expression are vital if knowledge is to transfer between generations The noetic economy of an oral culture demands that knowledge be processed in more or less formulary style and that it be constantly recycled orally—otherwise it simply vanishes for good unless it be discovered anew (1977:151) With the introduction of writing, language acquired its first fixed form, visual retrieval became possible as cultures imposed new order, consistency, and control over language (196745). N e w modes of thought took root and were expressed in "learn6d languages" based on script. Yet ancient and medieval European societies still preserved a sense of the word as spoken event, as indicated by the oral/aural roots of the classical syllabus in rhetoric and dialectic and the tradition of reading manuscripts by mouthing the words aloud It was the technological innovation of print 53 that represented the final "lodging of speech in space" (1967'49), as knowledge became completely frozen and objectified in textual forms such as encyclopedias, and the truth about the world came to be perceived as a set of " 'facts' which tend to be regarded as physical objects available without any reference to verbalization" (197788). The Gutenberg revolution made possible a kind of "delusional systematization" leading the mind to imagine a superbly ordered world that could be analyzed by the individual with scientific objectivity, leaving behind the shared, experiential aspects of knowledge as oral communication (1967.136-137). In our present noetic economy, certain aspects of "oral/aural synthesis" have returned to Western culture with the electronic media (albeit in altered form), such as a sense of simultaneity and participation as part of an audience. 54 53 It is the achievement of Elizabeth Eisenstein (1980) to have elaborated the sociohistoncal complexities surrounding the advent of print technology in Europe, and to have rightly pointed out the problems involved in modern schol arship's observation of the passage from manuscripts to printed books, since our own material requisites constantly intrude "(T]he conditions of scribal culture can only be observed through a veil of print" (p 8) A stimulating, contrastive, more ahistoncal approach to Western stages of media awareness pervades Mar shall McLuhan's Gutenberg Galaxy (1962), in Eisenstein's view, McLuhan in his avant-garde preoccupation with the newer media forms, overlooks many of the effects of the Gutenberg communications revolution that are still with us (Eisenstein 1980 40-41, 129, 151) 54 As Ong and others have noted, however, the electronic media have brought

36

LANGUAGE PATTERNS AND THE LINGUIST

Ong's studies are useful in guiding us toward consideration of nonWestern groups whose models of language may be differently shaped by the "condition of the media in their own cultural background" (1967 65) One of his most important insights is that cultures do not move in neat dichotomous stages between prehteracy, chirographic (manuscript) awareness, typographic consciousness, and the "secondary orahty" brought about by electronic media influences The introduction of a new medium alters, but does not do away with, the noetic strategies that preceded it We are now prepared to move beyond a simple model of literature as either "oral" or "written" to the complex interrelations of various noetic patterns encompassed by a single culture's use of language55 In a more meta-theoretical mode, Ong points out ways in which our own grounding in modern literate information management may betray us when we look at cultures with different noetic economies Scholars have distorted strongly or residually oral literatures with their own typographic prejudice when they insist on fixed-form texts where none exist, the term "literature" etymologically implies "technologically processed speech" and thus, one could argue, should properly not be applied to primary orahty at all (1967 42) Warnings about noetically conditioned prejudices in research have also been heard in the work of Jacques Dernda (1976), who looks at the ways linguistic science, although ostensibly the science of speech, is fundamentally and inescapably enmeshed within principles of phonetic script and the written word Barthes (1975 49-51) has written of the preoccupation of linguistics with the sentence, leading to scholarly obliviousness to the open-ended, fragmentary locutions that make up so much of everyday discourse Elsewhere, Dennis Tedlock (1977) has strongly criticized folklorists who study oral tradition using only conventionally transcribed alphabetic texts, linguists who wish to posit closed-system "grammars" for the production of texts, and structuralists who want to "dismantle" the translation in search of grandiose abstractions Oral tradition, into being entirely new forms of performer-audience relations When we watch a television game show, for instance, we are actually part of a second "audience" which is observing the stage performers and studio audience, the talk show host who glibly tosses off one-liners in apparent "spontaneous" dialogue with the studio audience is, in reality, working from a carefully prepared monologue script 55 As A L Becker remarked to me, the "dirty joke" is an important genre in American society that is still basically oral in its learning and transmission 37

THE SHAPE OF THE WORD IN BALI

moreover, represents "patterned repetition on various scales" (p. 513), and the distinction between the so-called "oral poetry" of Homeric epic and "prose narrative" such as Eskimo storytelling is wholly an artifact of alphabetic and textually conditioned thought. Homer came to us already changed by manuscript form, from which we ferret out an "oral poetic" regularity, much of which possibly could not have existed prior to writing. At the same time we disregard "the dimensions of the speaking event that were screened out by alphabetic treatment" when we look at the Eskimo story, blinding ourselves to its poetic features as we focus on a transcript that gives the illusion of written prose. Other views of how noetic processes affect discourse are presented by the philosophers of language. For Ricoeur (1976:16), the essence of language lies in our motivation to overcome "the radical non-commumcabihty of the lived experience as lived"; this we can do only through the "being-together" imposed by communication. Languageas-used presupposes a speech situation with participants in a two-way or dialogic relation. Our existence as "speakers" includes the intention to be recognized and understood by some "hearer." From this intention comes the noetic tension of discourse, the making-known of one's interiority and existential solitude. Thus, for Ricoeur (1976:18), "the noetic is the soul of discourse as dialogue." For Kenneth Burke (1970) the noetic aspects of language are traced using the science of "logology," or "words about words." In his view, our basic nature as symbol-using creatures places us in a communicative paradox from the start. Our capacity for symbol-using is primarily manifested in verbal ability. Thus human beings are intuitively responsive to notions of "transcendency," since words "transcend," or are different from, the things they denote. Yet our basic store of referential tools is based on "everyday experience" in the empirical realm of nature and social life. Our only options for expressing notions of theology, or the transcendent "supernatural," then, are by analogy with the natural order, the political/social order, or the symbolic order, which is language itself. Whether or not there actually exists a "supernatural" order or (in other terms) an object of theological inquiry is not at issue. What interests Burke is that the language of theology is often analogically based on a culture's conception of logos, the uttered word. A model of logos is closely linked, furthermore, with the noetic patterns that shape, store, and transfer information in all cultures.56 56 See Burke's The Rhetoric of Religion (1970 14) for a discussion of the different realms to which words may refer His "science of logology" is based upon

38

LANGUAGE PATTERNS AND THE L I N G U I S T

A noetically balanced perspective on language behavior and verbal art seems to be required in the investigation of traditions distant from Western culture or university-educated experience. The fundamentals of our own attitudes toward language are at issue, since the analyst's noetic background inevitably colors the interpretation of quite different patterns of shaping culturally important information.57 Basic media features are important to the meaning a "text" bears, whether it is shaped in oral performance, on a palm leaf, or on an offset-printed page. Our understanding of the terms "literature" and "text" in the context of the present study, then, must extend beyond the margins of the typographically fixed page and permit the values of sound in verbal art, the oral/aural literary mode, to dwell alongside our dominant visual orientation toward literature. SUMMARY In the discussion thus far I have looked at some aspects of language history and contemporary language use in Bah. Since a variety of different languages and literary genres are common, I found it necessary to give an overview of certain historical and sociopolitical processes so that readers might better understand the functions and meanings assigned to different languages and literary types. Speech varieties and written language in Bah show remarkable variation along the axes of formal/informal, ritual/secular, epic/contemporary, mystic/mundane, and so on. To characterize this variety requires consideration of the setting and purpose of each work, as well as its audience. The continuing use of "archaic" codes in Bah is best approached as a combination of context, intention of the participants, and formal stylistic and acoustic clustering. Similarly, modern spoken Balinese language is a constantly altering pattern of multi-variant "shifting," as speakers make lexical and parahnguistic choices according to different meta-communicative needs As we try to assess the full range of verbal art m Bali, it further becomes apparent that literature as commonly presupposed in most the notion of "words about words," or all the disciplines (grammar, etymology, poetics, and so forth] that form part of the symbolic order He wishes to put forward a concern for terminology, by means of logology, that will counterbalance the concern for imagery characteristic of mythology 57 For a critique of the ways Western scholars have characteristically imposed their own noetic orientations on Malay literature, see Sweeney (1980), particularly chapters 1 and 2 39

THE SHAPE OF THE WORD IN BALI

structuralist and literary studies is not our object of study Literate activity in the form of a broad and complex manuscript tradition has continued m Bah for a millennium or more On the other hand, this same artistic verbal expression includes oral composition as a major component of textual form There is no question of calling the Balmese ethos "prehterate" or "radically oral," but neither is it one in which literariness and textuality are embodied exclusively by the printed page To understand the Balinese view of how language fits into human experience, we must investigate the meanings given to speech and to writing

40

2 LANGUAGE FROM BIRTH TO DEATH

Characters are written on leaves, the dead burned on a pile, adorned with gold, and with gold in their mouth(s), with all kinds of scents —Ancient Chinese records, quoted by A J Bernet Kempers, Monumental Bah L A N G U A G E , even though a universal h u m a n attribute, may not always possess universal significance across cultures. That is to say, different language groups may view the place of language in the world, and in h u m a n experience, in different ways Beliefs about language in turn reflect as well as shape the techniques a culture uses for processing and communicating knowledge. The Bahnese manifest complex beliefs about language consistent with their elaborate material and ritual culture, m surveying these beliefs we are brought into contact with an equally ornate system of metaphysical interpretations. These interpretations have important bearing on vocal traditions, literary forms, and speech functions Our scope thus includes both "the means of speech in h u m a n communities, and their meanings to those who use t h e m " (Hymes 1974:1471) LANGUAGE AS WRITING Seen from the perspective of the history of communication, the development of writing in h u m a n societies wrought a profound and stillresonating change in the relation between the individual and language. In Western culture, there seems to have been a certain ambivalence about the scribal transformation of experience, as evidenced in the following proverbial contrast The letter kills, but the spirit (breath) gives life. The word flies, but writing remains. 1 1 The first proverb is from 2 Cor 3 6 Litteia emm accidit, Spmtus autem vivificat, the second reads Verba volant, scnpta manent (My thanks to Stuart Robson )

41

THE SHAPE OF THE WORD IN BALI

The first saying refers to the "dead" nature of writing as soundless sub­ stitution for the shared awareness that accompanies speech. Yet the second proverb notes with relief the permanence of writing, its facili­ tation of the recovery of information, its easy preservation of what needs to be known. In the following pages, I will examine certain as­ pects of writing's impact and meanings in the non-Western setting of Bah. As discussed in the previous chapter, the introduction of writing in ancient Bah most likely was through the agency of priests and scribes attached to the Old Bahnese courts. For the general populace, initial contact with written script would have occurred as royal edicts were issued. Most of the ancient inscriptions embody agreements between kings and village localities regarding tax obligations, boundaries, up­ keep of religious sites, and so forth, these permanent metal and stone edicts were intended to perpetuate verbal acts through time, and they were often enshrined and preserved as sacred objects in local temples. This custom of endowing the written object with sacred meaning was pervasive and persistent, and it is evidenced today in edicts such as the prasasti Gunung Agung ('inscriptions of Mount Agung') which are rel­ ics of the Besakih temple complex. These inscriptions, dating from the fifteenth century and painted on wood, are still considered sacred ve­ hicles inhabited by some of the temple's highest gods on important oc­ casions. If the practice of ritually cleaning and sometimes reading the in­ scriptions on holy days can be taken as an indicator, we may assume that renewing the force of the written word through periodic ritual ac­ tivation was an important consideration for the communities holding inscriptional documents in ancient times. The most famous example of faithful commemoration of an inscription was recorded by Stutterheim (1935:28-29), who came upon the stone edict establishing the shrine at Tirta Empul (Tampaksiring). He learned that each year it was ceremonially revered on the very same date recorded in the inscription itself nearly one thousand years before. Even when the text itself was no longer understood, the sacred meanings of its nature as writing con­ tinued. Aside from epigraphic activity, we know that writing was used in Bah in the transmission of manuscripts of many varieties, a process that undoubtedly began in the earliest historical period. The continu­ ing dissemination of texts, many of them originating from Java, must also have been primarily a court-based tradition, especially in the ιηι42

LANGUAGE FROM BIRTH TO DEATH

tial stages. The writing materials used down to the present in this tradition are specially prepared leaves [ion] of the tal palm (hence the words rontal and lontai, describing certain manuscripts), into which letters are cut using a small, pointed blade [pengutik, pangrupak). There is evidence, however, that in the composition of original literary works in Java in the past other sorts of materials were used, including a flat rectangular board or slate, called karas, which was inscribed with a sharpened soft-stone pencil, called tanah. Zoetmulder (1974 127135) and Robson (1976) have clarified the significance of karas and tanah as writing materials, in Bali, the fact that these particular instruments were ever used is only evidenced by illustrations (of fairly recent date) in a few lontars.2 Also cited by Zoetmulder (1974 135-137) is the tradition, known from frequent mention m poetical literature, of writing on the petals of the pandanus flower, or pudak, using a thorn or other sharp instrument. It may be the case that less permanent materials requiring little preparation, such as karas, tanah, and pudak, were used in more spontaneous, original composition, and that only when a text was intended to last, or was prepared for wider transmission, was it recorded in a palm-leaf manuscript, which requires elaborate preparation.3 The history of writing in Bali itself reveals that the most prominent artifacts are of course the inscriptions on metal and stone and the lontar manuscripts, which have been recopied many times throughout the centuries, any materials that were of impermanent nature or unlikely to be copied would long since have disappeared It is clear from certain peculiarities of the archaeological remains, however, that the island very likely had some sort of writing activities even predating the earliest edicts. Early Buddhist clay seals4 bear a standard litany that indicates they were used as "amulets having an auspicious influence" (Bernet Kempers 1977 102) In addition, nearly all the early Bahnese inscriptions themselves are remarkable for being in the form of metal plates emulating the rectangular shape and the arrangement of lines of 2

One such manuscript is the Dampati Lelangon in the library of the Faculty of Letters (Fakultas Sastra) at Udayana University in Denpasar 3 Two descriptions of the process of making a lontar manuscript and of inscribing and illustrating it are published in Indonesian, see Museum Pusat (1975) and Suwidja (1979) See also the reference to the elements composing the literary manuscript in the Preface to this book Ginarsa (1975) gives the most detailed English account of contemporary Bahnese manuscript production 4 These are the same artifacts referred to in note 7, Chapter 1 43

THE SHAPE OF THE WORD IN BALI

script used in the lontar tradition, in contrast, contemporaneous edicts from other parts of Indonesia are on stone (de Caspans 1975 43) This evidence invites interesting speculation as to )ust how long lontar manuscripts have existed in Bah It is, however, even more difficult to know from archaeological re­ mains to what extent script may have been known or used outside of the courtly inscriptional and literary traditions Yet perhaps, in spec­ ulating on the spread of script in old Bali, we need not presuppose a rigid division between "court" and "village " At the end of the first millennium A D the island's population was not so large, or the court complex so elaborate, that anything resembling total separation of "royal" and "commoner" spheres should have developed The inscrip­ tions themselves testify to a continuous give-and-take between kings desirous of maintaining economic and political power and villages with semiautonomous, self-governing traditions, but which neverthe­ less needed an external political referent and arbiter for supra-village concerns (such as the maintenance of irrigation networks) Along this court-village axis5 whose "historical compromises" were perpetuated as inscriptions, there may have been individuals from the villages whose task it was to transmit and explain messages to and from the king, who were able to read the inscriptions and perhaps even help write, carve, or forge them Perhaps it was also necessary for certain village members to keep records of matters such as the king's taxation, corvee, temple maintenance, and the like In any case, it is not incon­ ceivable that a few commoners in any area near the court may have needed literacy in their dealings with the king, and that writing was also used for non-court-related activities such as the making of magic charms (as in the case of the Buddhist amulets), the recording of fines in village councils, marketing, and so forth While the above discussion is largely speculative, it does indicate an avenue of exploration that fits rather well with the role of writing in more recent times For, even though literacy was traditionally a con­ cern primarily in princely compounds [pun] and priests' homes [gna], there have long been groups outside the social elite who also have uti­ lized script Among these persons are the architects/carpenters [undagi) and metalworkers [ραηάέ) whose arts are expounded in lontar form throughout Bali Prominent as literate individuals in the "com­ moner" {jaba, literally 'outside') realm are the bahan, whose role as 5

The expression "court village axis" I have borrowed from J S Lansing (1977), who has used it in his work on Bahnese social and temple organization 44

LANGUAGE FROM BIRTH TO DEATH

healers and diagnosticians of magic activity requires the reading of lontars and the writing of charms. The bahan's medical science [usada] is found in lontars, and the more mystico-magic side of the profession is often gleaned from tutur texts. In treatments, the balians frequently include some written language in the form of sacred syllables or mantras This writing is done, often as an accompaniment to drawings, on a variety of materials ranging from palm leaf and the sheaths of bamboo stalks [klop6kan] to bits of cloth or wood to stone and hammered metal (panpih) of gold, silver, or copper 6 Other people whose work requires reading or writing skills are the bahan wanga, specialists in astrology and time-reckoning. Even in precolomal times, before Dutch-mandated education spread the knowledge of Bahnese script, some village leaders (e.g., khan, bendasa) had to be able to keep written records such as membership rosters and simple accounts. In some cases, the temple or "lay" priests [pamangku] were also literate, since their duties require the knowledge of mantra and kidung, as well as the mobilization of large congregations and their resources for ritual ends. In addition, there must have been many people, especially women, who were to some degree literate as a result of their involvement with marketing and commerce We know that even jaba women could learn to read for the purpose of orally performing literary texts, and those with especially pleasing voices were often "adopted" by royal households and trained to read and sing there. Similarly, prominent figures in the performing arts have long needed to read the literary works that provide stories for dance-drama, shadow puppet theater, and other performance genres That Bali's social organization is multiform, crosscutting, and complex is by now a truism of cultural studies of the island.7 The operation of dense networks of irrigation associations, temple congregations, work groups, performing arts troupes, and so on requires a certain amount of writing ability on the part of a few individuals. In the premdependence and especially pre-colonial days, before public education increased literacy, people with reading and writing skills were endowed with special prestige. The need to read and write in order to be a leader in community affairs was reason enough for a literate man 6

This information was provided by D J Stuart-Fox, whose unpublished manuscript on the art of the bahan I was kindly allowed to read On connections between the bahan and the realm of written texts see also Hooykaas (1978) 7 The works of Hildred and Clifford Geertz contain the most enlightening material on this subject, see Geertz and Geertz (1975) and Geertz (1959, 1964, 1973) 45

THE SHAPE OF THE WORD IN BALI

from one neighborhood or hamlet [banjar] to be asked to serve as khan in another. Individuals who could read and write were considered bet­ ter endowed as leaders of their communities because they were also within reach of the vast knowledge contained in the lontar tradition. Control in the realm of script, then, meant special powers in political life as well. Yet it does not follow that the written word has no significance for an individual who has never had cause to read and write. On the con­ trary, there are certain times in each Bahnese life when written lan­ guage is essential for passing from one stage to another, when the power of the inscribed word surpasses the conventional functions of writing defined in terms of literacy At the birth of each Bahnese child a number of ritual actions are nec­ essary to ensure the safe entrance into human life of a soul recently separated from the deified ancestors {kawitan, from wit 'tree' or 'ori­ gin'). Among these actions is the burial in the house compound of the an-an (placenta), this organ being venerated as a representation of one of the four metaphysical siblings who accompany a baby after concep­ tion, through birth, and on into mature life.8 Along with the an-an is buried a pipil, a palm leaf on which have been written magic syllables, mantras, or occasionally the name of the newborn.9 The pipil is writ­ ten by someone in the family, usually the child's father, but it can also be taken to an outside scribe such as a pamangku or bahan. Although pipils vary in content, it is not the message itself that is considered most important. Many Bahnese feel the pipil represents the divine pa­ tron of language and learning, Sang Hyang Αμ Saraswati, expressing a wish that the child be empowered with knowledge of writing later in life. Another rite of passage is significantly connected to the use of writ­ ten language. During cremation ritual, the human corpse or exhumed bones are accompanied by a pangawak ('embodiment', from awak 'body'), which can stand for the physical body even when no remains are present, and whose elements include a strip of sandalwood on which a human-like image and letters are inscribed. In addition, along with the holy water vessel are small palm-leaf "notes" written by the padanda, these are called racadana (Hooykaas 1976a:39-40, 1976b:47; Ramseyer 1977177). The white shroud covering the body to be cre­ mated is called a kajang. Although the style and motifs used by the 8

The four birth-siblings, or Kanda Empat, are described in many Bahnese manuscripts (see Hooykaas 1974) 9 Pipil is also a term used for traditional land deeds 46

LANGUAGE FROM BIRTH TO DEATH

Brahmana households making the kajang vary from area to area and according to the caste of the deceased, some arrangement of aksaia (letters) always seems to be included.10 On some shrouds, for example, there may be drawn an image of the three worlds [Tnloka], at the bottom is the great world tortoise Bedawang, bearing the intertwined serpents Anantabhoga and Basuki, atop of whom rests the cosmic mountain Mahameru, itself an icon of the human body. On the kajang the human form replaces the mountain image, with a series of place names listed on either side of the entire drawing. On one side is a list of places in the macrocosmos, the Buana Agung 'Great World', on the opposite side are detailed the places within the human body, the microcosmic Buana Aht 'Small World'. This dualism on the illustrated funeral shroud—a dualism so important in Bahnese thought—represents the unification of the individual being with the greater cosmological order at the point of liberation through cremation. Yet it is also true that the content of what is drawn on the kajang is transcended by its meaning as an inscribed ritual element. The kajang can be compared to a bus ticket, one member of a priestly household explamed, it is something with writing on it that allows us to go from one point to another. And indeed, another meaning of the term kaiang in Bahnese is 'to move, to transport'. The aim of Bahnese death ritual is to safely transport the soul of the deceased back to its point of origin in the ancestral world, in the course of burial, cremation, and post-cremation ceremonies, the physical form becomes less and less an object of focus as the immaterial self, the atma, is purified and released. Part of the "de-matenahzmg" process is a stage in which each body part is marked with a written symbol from the syllabary, this ritual, carried out by the padanda and called the Dasabayu 'Ten Breaths' or 'Ten Forces',11 is yet another example of how written language literally "embodies" the human self whose physical body has been vacated. The origin in Bahnese ritual of devices such as the pipil and kajang is lost in time, but their recurrence among the tokens with which human life is begun and ended gives special meaning to individual awareness of writing. Inscribed language accompanies each Bahnese as she or he arrives in and departs from madyapada, the 'middle world' where humans dwell. Far from being a secondary, derived result of language, 10 The Babad Pasek, the genealogical epic of the large Pasek descent group, includes drawings of the type of kajang its members are entitled to use (Sugnwa 1976 153] 11 For information on the Dasabayu concept, see Gons (1926 61), Week (1976 75-76), and Hooykaas (1978 63)

47

THE SHAPE OF THE WORD IN BALI

an artifact merely representing the human capacity for speech, Balinese script manifests language's otherworldly powers and functions at crucial boundaries in the life cycle At these times the communica­ tive function of writing is the dovetailing of human experience with other more metaphysical orders It is not within the intention or range of most Balmese people, how­ ever, to plumb the esoteric depths suggested by the meanings and uses of the pipil and kajang Only Brahmana priests, and others who desire access to specialized and extensive uses of the written word, such as in literature, the healing arts, and so forth, need to be fully aware of writ­ ing's mystic implications For the nonspeciahst majority, however, there is still an occasion at which one celebrates in a general sense the existence of language and the written shape of learning This occasion is Saraswati Day [Odalan Sang HyangAji Saraswati) As mentioned above, Sang Hyang Αμ Saraswati is considered the pa­ tron goddess of language and learning, this 'Great God of Teachings' shares some characteristics with the Indian divinity after whom she is modeled Yet it is only m the last few decades that the four-armed im­ ages of her holding conch, vina, lotus, and lontar have become wide­ spread, and it seems unlikely that the day set aside for her special wor­ ship was as widely observed in the days before schooling encouraged literacy as it is at present Saraswati Day occurs once every 210 days according to the Javano-Balinese calendar, when Saniscaia of the 7-day week and Umanis of the 5-day week coincide during the week Watugunung of the 30-week wuku cycle 12 On this day all written materials (and even printed aids to knowledge such as schoolbooks) are given special offerings and sanctified with incense and holy water [tirta] Lontars are especially honored, often kept on a high shelf in a northeast corner of the house, on this day they may be placed in the houseyard temple itself There are a number of lontar sources on the practice of Odalan Sa­ raswati, including the Tutur An Saraswati13 and the Sundan Gama 14 The latter advises devotees that they may not open books, read, write, or sing for the entire holiday period, while the former states that some form of literary performance is necessary to honor the goddess, but that 12

A detailed treatment of the Balmese calendncal system is available in Gons(1960) 13 Hooykaas (1964 22) lists the different versions of this text available in the library of the Gedong Kirtya in Singaraja, Bah 14 For information on this manuscript, see Ginarsa's (1970) discussion of Sa­ raswati worship 48

LANGUAGE FROM BIRTH TO DEATH

this may only be carried out after the sun has passed its zenith and the required offerings have been made. Both these textual traditions have their adherents in present-day Bah Yet it is not enough just to honor one's own books on Saraswati Day In most areas, this is also the time to bring offerings to the homes of one's teachers, be they priests, bahans, or artists. In this way, even families not owning books can pay homage to the nearest sources of knowledge which on this day are imbued with the presence of Saraswati. Furthermore, persons learned m religion and literature may on Odalan Saraswati perform special fasting or meditation, brata Saraswati, to worship the goddess as the female counterpart, the iakti, of the god Brahma. According to some lontars,15 when humans still lived as animals in the most ancient times, it was she who, on Brahma's command, descended to bring language, culture, and civilization to human groups. The role of Saraswati in Bahnese culture is broader than it first appears. The goddess is mentioned in numerous rituals as the divine essence of language.16 She is also worshipped (on Soma Ribek, during wuku Smta) as one of the triad representing the source of fertility and prosperity connected with rice cultivation.17 Popular belief holds that the goddess is heard in the voice of the familiar house lizard, the cekcek, whose chirps, if heard during nighttime conversation, are Saraswati's corroboration of what is being said. Hooykaas (1964:23) mentions the association of Saraswati with water, rivers, and irrigation, Sri gives rice, and Saraswati "the indispensable water for its cultivation." Some Bahnese explain that different regional accents develop because people drink water from different sources. The meanings given to the goddess's name are, like the sounds of speech, linked with water 18 15

For example, the manuscript Purwagama in the collection of Gria Pidada, Sidemen, includes this story 16 In his collection of texts and rituals performed by lay temple priests (pamangku), Hooykaas (1977 75| includes the following among "Thoughts of a Pamangku whilst Making Offerings" "Conduct the syllable OM to the tip of your tongue, when OM has arrived there, imagine that the goddess Saraswati stands on the tip of your tongue " 17 The other two members of this triad, known collectively as Sang Hyang Tnpramana or Sang Hyang Trimurti-Amerta, are Sang Hyang Sri, the rice Goddess, and Sang Hyang Sadana, the god of wealth and Sn's male partner in fertil lty ritual (Ginarsa 1970 18) 18 In the invocation entitled "Homage to the Gods of the Sun and of the Gangga" (Hooykaas 1977 52-53], Saraswati is named as one of the Seven Sacred Rivers 49

THE SHAPE OF THE WORD IN BALI

Saras is interpreted as 'speech', 'river', or 'that which flows', while wati, in some accounts, is glossed as 'honored', 'exalted', or 'that which possesses' " The commemoration of the written word performed on Saraswati Day, as well as the use of writing in key life cycle ceremonies, helps us understand the special significance of inscribed language in Bahnese experience The written word in this culture is not a reduction or reflection of oral speech processes, but extends beyond ordinary sensory awareness of the sakala, or 'manifest' world, into metaphysical, 'unmanifest' reality, the mskala realm In ancient times, the king's word became a permanent relic to be enshrined in temples, today, the Bahnese still honor books and manuscripts by keeping them in special places and periodically celebrating the deified origin of language activity The human capacity for verbal expression, in their view, bears withm itself the pattern manifested by all creation Both written and sounded language embody ultimate truths and potentially reveal the inner natures of the human and the cosmos, as we will see in the next section In the ordinary person's awareness of language, as m the skilled person's specialized knowledge, writing is a potent icon in the individual cycle of life and death LANGUAGE AS SPECIALIZED STUDY While the average Bahnese traditionally may have had limited direct contact with the written word, experiencing a rich manuscript heritage primarily through the sounds of ritual and the performing arts, there have always been persons who choose to study lontars and become more closely involved with the written word It is in the nature of the Bahnese conception of written tradition, however, that initially there be no question as to whether one studies the lontars for purely religious, professional, or aesthetic reasons No matter what one's status, whether that of a future padanda, a bahan, or just someone who enjoys singing, the activity of reading brings one close to that interface between ordinary sensation and metaphysical existence represented by the role of writing in the rituals of birth and death Thus, anyone who wants to read manuscripts, even a student from a foreign country, should come to understand certain ordering principles of the universe It is common in cultures where literacy is limited and much cultur19 See the working papers by I Ged£ Sura and Ida Pranda Putra Telaga included in a Bahnese seminar on religious aspects of Saraswati Day (Pembahasan Pra saran 1979)

50

LANGUAGE FROM BIRTH TO DEATH

m ψ0Φ'-$& «δ ΐτ> m m f ^ ¾ ^ ¾ ¾ ¾ gap ΐΓ{ τΛ)ΐηΛ

$M SSii 6 ¾ 31 '-

„•»

"~~·

Balinese script [aksaia). Close-up of a lontar leaf, showing four-line format. From a recent copy of the Saiasamuccaya. ally valued information must be orally transmitted that individuals have a visually conceived map, schema, or other design that functions as a mnemonic device for the storing and communication of knowledge. The "residually" oral societies of medieval Europe, for example, used the basic architectural outline of cathedral-type structures as a mental organizing principle; the elements or terms within a field of knowledge could be keyed to the various parts of the "building" according to their proportions and place in the hierarchy. 20 Present-day Bali still widely employs such a visual aid to memory; it is called the nawasanga. Often glossed as 'rose of the winds', this is a schema representing the four cardinal directions, the four intermediate directional points, and a central " h u b , " as for a wheel or compass. These nine elements are the basis of the design's name, both terms, nawa (Sanskrit) and sanga (Old Javanese, Balinese), meaning 'nine'. The role of the nawasanga as a holistic organizing principle in Balinese thought involves the assignment of different values to the directional points, each of which is associated with a particular deity, a 20

See Yates's (1966) discussion of the art of memory from its Greek origins onward; the architectural analogues of memory are treated in chaps. 4, 5, 6, 7, 15, and 16, particularly. 51

THE SHAPE OB THE WORD IN BALI

color, a material element, a body part, a weapon, a written or sounded syllable, and so forth. The design has both horizontal and vertical poles, and when fully elaborated it can have eleven directions in all— the eight points plus center, nadir, and zenith. The nawasanga's origins as a directional index may go back to early historic times when the Bahnese, primarily agriculturalists rather than seafarers, revered the volcanic heights as the source of fertility-bearing irrigation waters and the sea as the home of more dangerous, chthonic forces. These principles of spatial orientation still permeate basic cultural patterns, such as the layout of the Bahnese house, the positioning of the human body both in the physical world and in social settings, and the arrangement and design of components of religious ritual. The esoteric force of this cosmological map is fully realized only by high priests and experts in meditation, for it is a complex and far-reaching subject (Hooykaas 1974, 1976a, b, Soebadio 1971). The padanda's deeply intoned Sanskrit invocations are aimed toward unification of the self with the supreme divinity (Siwa, or Sang Hyang Tunggal/Acintya/Widi21), who in turn is the coalescence of all the dimensions and forces represented by the nawasanga. In the experience of most Bahnese, however, the nawasanga functions more as a guide to the organization and recall of many classes of things. For example, in laying out the circular caiu (an offering to chthonic or demonic forces placed on the ground), the Bahnese knows that kelod is associated with scarlet flowers, fiery elements, "red" rice grains, the seaward side of the offering area, and it is under the aegis of the god Brahma. Under the god Wisnu, on the mountainward side, are the black emblems and watery elements symbolizing ka]a22 The nawasanga also is a classifying principle for movements in dance, for temperaments, illnesses, and physical "humors", and for time itself, as each day is dominated by one or another of the directional deities. While the nawasanga (or its more succinct fivefold version with just the cardinal points plus center) may be used to classify many different 21

Tunggal means 'one, single, unified', acmtya can be glossed as 'ineffable, inconceivable', widi invariably signifies the highest god to the Bahnese For additional explanation of these concepts, see Soebadio (1971) 22 Regarding the Bahnese conception of directionality, see Belo (1970 90-93), see also the pamangku's invocations dealing with this topic in Hooykaas (1977 46-47, 66, 88) The Preface to the present volume also includes a reference to the Godhead surrounded by the directional deities—a nawasanga image—as part of the shadow puppeteer's "Invocation for the Performance of Shadow Theater " 52

L A N G U A G E FROM BIRTH TO DEATH

types of information, it is its connection with beliefs about language that we wish to consider here, for the place of both written and sounded language on the Bahnese cosmological map associates lin­ guistic form and verbal activity with other fundamental elements in the conceptual system. The nine directional points may be divided into two groups of four-plus-center (the cardinal and intermediate points are grouped separately and the center is counted twice), giving a total of ten elements When a written symbol is assigned to each point, we find the cornerstone of Bahnese linguistic mysticism, the dasak­ sara (from dasa 'ten' plus aksaia 'letter'). 23 While the high priest uses them as just one element in a convoluted esoteric ritual, many ordinary readers of lontars are also familiar with the implications of the dasaksara. They are based on a mantra praising five of Siwa's many names, the "Ten Letters" and the full mantra are shown below (see also Figure 2). Dasaksara. SA ΒΑΤΑ AINA MA SI WA YA Full Mantra: Sadyojata, Bamadawa, Tatpuiusa, Aghoia, Isana, namah Siwa ya (The five names), homage to Siwa. In reading the mantra from its display in Figure 2, one makes two clockwise rounds [pradaksma), ending in the center each time In ac23

See Gons (1926 62) for information on this topic The "Ten Letters" are sometimes referred to in theological treatises as the catui-dasa pianawa, the 'Fourteen Syllables', when the supreme sound OM is included, along with its three components (A, U, M) See also the reference to the "Ten Letters" in the Preface of this book (4)

A (8)

Wa

Si

(9)

(5)(10) I/Ya

(3)Ta Ma

Sad) Na

(7)

(6)

Ba (2)

FIGURE 2

Place of the dasaksara on the nawasanga (numbers in parentheses indicate sequence of sounds) 53

THE SHAPE OF THE WORD IN BALI

cordance with the main philosophical thrust of Bahnese Hinduism (which was heavily influenced in ancient times by the Saiva-Siddhanta school of South Indian thought), Siwa is the supreme principle from which all creation emanates and the void into which, eventually, all will dissolve. Thus Siwa (also called Sadasiwa) embodies the central region of the nawasanga. The other eight deities (Iswara or Isora, Mahoswara or Mahisora, Brahma, Rudra or Ludra, Mahaddwa, Sangkara, Wisnu, Sambu) are in the final analysis different aspects of Siwa's oneness. In uttering the dasaksara, one m effect recreates the dimensions of the universe with all its ordered variation, at the same time naming the higher-order principle that subsumes it all. When this whole system is in motion, revolving, the ordered differences can begin their condensation into fewer and fewer forms until the final, single distillation is achieved—the goal of the mystic adept's meditation. 2 4 For example, when the padanda performs his daily ritual [surya sawana),25 the making of holy water is achieved through an inner activation of the dasaksara, and a mental "revolving" of the directions, such that the ten letters are progressively condensed and transformed. 26 First, they become five, the pancaksara (Na-Ma-Si-Wa-Ya), then three, or tnaksara (Ang-Ung-Mang) This triad in turn becomes a duality, rwa bhmn&da 'the two distinguished' [Ang-Ah], a principle referring to the bipartite-yet-indivisible nature of universal oppositions such as male-female, life-death, and macrocosmos-microcosmos 27 Finally, there is the all-encompassing ONG, or ongkaia, the representation of essential sound and ultimate reality all m one 2 8 14

The starting point in this process is the realization of the macrocosmic-microcosmic correspondence, the mahapadma nng sania 'great lotus within the self (Gons 1926 63) 25 The ma)or work treating this ritual is Hooykaas (1966) 26 This process is commonly practiced by seekers of spiritual insight other than the padanda The shadow puppeteer performs a similar manipulation of the directions (see Chapter 4) Lontars such as the Rahasyopadtsa 'Secrets of the Teachings' (subtitled 'the way of knowledge for persons striving after liberation') also may include elaborate, speculative, and revelatory instructions for revolving and condensing the directional and linguistic symbols In the Bahnese-language poem Basui, which includes allusions to formulas for magic and curing, there is a description of a "sorcerer's" rites involving homage to the nine directions, midei buana (Hooykaas 1978 49, verse 53) 27 This dualism has its Indian roots in both Upamsadic and Tantric traditions (Gons 1926 59) 28 The term ongkaia is variously explained by the Bahnese Some say that the individual {wong) can realize his or her ultimate nature as the holy syllable [kara] Others interpret the word as a derivation of wong 'person' plus ka-rwa 54

LANGUAGE FROM BIRTH TO DEATH

The meaning and use of the dasaksara point to an entire field of elab­ orate mystico-hnguistic speculation, all the details of which are be­ yond the scope of the present study. I would, however, like to mention three relevant aspects of this science that are phonological, ortho­ graphic, and textual in implication, and that are related to the dasak­ sara principle First, there is the relation of the dasaksara to sound The nawasanga as an information-storing device applies not only to the macrocosmic world-at-large, but also to intrinsic, microcosmic reality. The organi­ zation of space through the mapping out of different orders of being ap­ plies just as much to the inner, individual self as it does to the outer, natural world. It is in the nature of human vocal activity that sound be experienced "inwardly," through its physiological components, at the same time that it is outwardly uttered. 29 Thus, the Bahnese logically ascribe to the dasaksara a simultaneous internal and external reality. Accordingly, they are seen to have their own domains within the hu­ man body Kay&ki hngganya, Sanghyang Dasaksara ring laganta, sa kara nngpapusuhan, ba kara ring ati, ta kara ring ungsilan, a kara ring ampru, ι kara ring witmg hati, na kara ring paparu, ma kara ring urung-urung gadmg, si kara ring hmpa, wa kara ring ineban, ya kara ring tungtungmg hati, dasaksara ngaran, kayaki genah6 ring ]ero These are their seats, the Great God Ten-Syllables in your body, the 'sa' syllable is in the heart, the 'ba' syllable is in the liver, the 'ta' syllable is in the kidneys, the 'a' syllable is in the bile, the Ί ' syllable is in the center of the liver, the 'na' syllable is in the lungs, the 'ma' syllable is in the intestines, the 'si' syllable is in the spleen, the 'wa' syllable is in the diaphragm, the 'ya' syllable is in the tip of the liver, they are called the Ten Syllables, and these are their places inside.30 'both, two together', emphasizing the unity of opposites that precedes the final realization of indivisibility See also Hooykaas (1973 42) and Gons (1926 6263] 29 For this reason, claims Walter Ong, sound has primacy in communication between human beings, and it is a major factor in socialization and shared com­ munal sense "Because the spoken word moves from interior to interior, en­ counter between man and man is achieved largely through voice" (Ong 1967 122-125) 30 From the lontar Tegesmg Warah 'Explanation of Speech', in the Gedong Kirtya collection, number 1426 55

THE SHAPE OF THE WORD IN BALI

In all verbal activity, then, inner "seats" of sound are involved in outer expressive realization.31 In the practice of meditation, or in the utterance of mantras and other verbal formulas, the dasaksara are "brought to life" in speech with the addition of the velar nasal, and thus Sa becomes Sang, Ba becomes Bang, and so forth This nasal component {anuswaia) of the ten uttered syllables is associated with the cerebral, low-pitched, inwardly reverberating qualities of meditative sound as found in the most powerful of sacred syllables, ONG. When one writes out the dasaksara as sacred syllables, it becomes clear that their significance derives from orthographic shaping as well as from their sound associations. The three parts of the fully written velar nasal in the Bahnese writing system are considered symbols of the Tnsakti 'Three Powers', namely, Iswara, Wisnu, and Brahma, whose manifestations as air, water, and fire are essential to all life 32 These three written elements are the vertical or tear-shaped nada, the circular windu, and the crescent-moon aidacandra (see Figure 3) With the addition of the nasal symbol, the dasaksara derive fundamental sig31 The relation between sounds and the body is not always expressed in the same way in the metalinguistic texts that deal with this subject In the lontar Bramokta Widhi Sastra (Institut Hindu Dharma 1979 39-461, we find a description of how the first-created humans were made from aksara Using the consonant and vowel symbols, the text connects each phonetically related group of aksara with a different part of the body The "Ka" group—ka, kha, ga, gha, nga—became skin and hair The "Ca" group—ca, cha, )a, jha, nya—became flesh and membranes The "Ta" group—ta, tha, da, dha, na—became blood, sweat, and other body fluids The "Ta" group—ta, tha, da, dha, na—became sinews and joints The "Pa" group—pa, pha, ba, bha, ma—became the stomach and other internal organs The text continues with the assignment of the vowels and semivowels to limbs, orifices, and organs of sense, and it states that the nervous system, vocal apparatus, and motor faculties were created from various sacred syllables While this text uses a basically Sanskntic model, in which the Devanagari syllabary is divided into five groups of stops of five symbols each [pancaksara waiga), another lontar, the Tutur Anacaraka, employs the Bahnese variant of Javanese alphabetical order, that is, ha-na-ca-ra-ka-da-ta-sa-wa-la-ma-ga-batha-nga-pa-dha-ja-ya-nya, to assign each aksara to its place in the body, not only in one-to-one relation, but also in groups [hanacaraka, datasawala, etc ) The point is that any one of several "schemata"—whether derived from an Indian or Javanese model—is available when giving form to a mystic tradition 32 The anuswara is usually written in abbreviated form in syllable-final position In certain esoteric texts, such as the Anggaksaia, the tripartite nature of the nasal as found in the ongkara is the basis for a highly refined yogic practice

56

LANGUAGE FROM BIRTH TO DEATH

nificance through shaping as well as sounding. The proposition about language revealed here is that the "hidden" meaning of a sound can be explored through its written shape. This proposition has led to the for­ mulation of a typology of the written forms of language, from those that are basic, "neutral" elements of words to those that are powerful and highly charged and have more specialized usage. This typology is explained in a number of lontars that include esoteric teachings on the nature of language.33 While precise definitions vary from one text to another, in general the following four classes of written symbols are of primary importance 1. Swara-Wianiana 'vowels and consonants', the basic uncombined shapes of the Bahnese syllabary. 2. Wiyastia (or Wiuastia), letters as they are found in words, with all the modifications, addition of vowel symbols, vocalic merging [sandhi), shortening, and so on that occur when letters are conjoined. 3. Swalahta, letters from the first two groups that are combined or abbreviated to make another symbol that may or may not have mystic import In texts, swalahta appear as panten, "shorthand" abbrevia­ tions for commonly used words and phrases, for example, pu is an ab­ breviation oipumama 'full moon', and ti stands for tilem 'new moon', both terms being regularly found in the calendncal literature [wanga). A more complicated type of swalahta is exemplified by the Dasabayu (discussed in the first section of this chapter), in which abbreviations of the words for natural phenomena and sacred syllables are combined in an elegant design. The swalahta introduces us to the type of written symbol used by the Bahnese for the visual display of hidden meanings and secret knowledge M 4. Μοάιέ, a term used for the mystically powerful signs whose sounding can call forth gods, subhumans, and other unseen orders, and that are important in the practice of both magic and meditation 3 5 33 34 35

For example, in the Καήα3γοραάέ$α and the TuturAji Saraswati See t h e discussion of sacred syllables in Sugriwa (1978 14-15) These elements of language are also mentioned by Hooykaas (1964 37,

or ng

6

-

ο

- ardacandra

nada

= Iswara

= air

windu

= Wisnu

= water

= Brahma

= fire

FIGURE 3

A written anuswaia

symbol and its cosmic referents

57

THE SHAPE OF THE WORD IN BALI

Modro frequently include nasals and a final glottal fricative {-h), since these sounds "originate" with Ang (inward breath) and Ah (outward breath), the sacred paired syllables, the rwa-bhmn6da. Ultimately the rwa-bhinneda can be dissolved into the ongkara, the O N G symbol, which is itself the most powerful of modr6. The distinguishing char­ acteristics of modre are, first, their nature as syllables, such that no matter how convoluted and unpronounceable they may appear in writ­ ing, each modro is considered a unit of sound, and second, their exist­ ence as utterances. For, while swalahta symbols may be elucidated as words and even sentences, and are a sort of lexical shorthand device, the lexical content, if any, of modr6 is peripheral, for these signs exist ultimately only in their own oral realization, their sounding. The ut­ terance of modro is the province of priests, balians, and practitioners of yogic disciplines and magic. Modr6 take many hundreds of forms and are listed in special Ion tars called Kiakah.36 In practice, modro are often used w h e n the writing of sacred syllables is called for, for exam­ ple, in cremation ritual They are also included in the many types of protective charms written on cloth {rajah, tumbal, ulap-ulap, and others) that are hung above doorways, gates, new buildings, and so forth, carried about on one's person, or used in other efficacious ways. So far we have considered aspects of Bahnese mystico-hnguistic studies that are related to beliefs about sound and writing. The textual implications of this tradition of metalinguistic speculation are also important The dasaksara, for instance, are not a " u n i t " or "level" of language according to any generally agreed-upon definitions, and yet they are found as a common constituent of formulaic chants and h y m n s 3 7 Many sorts of swalahta and modrd appear at the beginnings and endings of manuscripts, and their function is not a purely decora­ tive one. Moreover, sacred syllables invariably occur in the thousands of mantras in daily use, and they are considered the galvanizing force

1973 100) and Week (1976 78-79), the latter includes drawings of a number of interesting modre 36 The term kiakah may come from the Sanskrit kaiaka, which can mean 'making, doing, acting, who or what does or produces or creates', 'intending to act or do', 'instrumental in bringing about the action denoted by a verb', 'the notion of a case (as in expressing a relation between a noun and a verb)' It seems that a logological explanation (in Kenneth Burke's terms) is relevant here, as a term originally used to express a feature of language (ι e, case relations) has been applied to the process of supernatural control using language 37 See Hooykaas (1977 44), in which OM and the dasaksara form the ending section of one of the pamangku's seho, or octosyllabic line hymns

58

LANGUAGE FROM BIRTH TO DEATH

that activates the mantras' potency. In this role, they are called bijaksaia 'seed-letters' or 'seed-syllables'.38 The tnaksara (Ang-Ung-Mang) and the ongkara |ONG) are frequently used as bijaksara, for example. What emerges from an examination of the specialized study of lan­ guage in Bah is the recognition that language is a non-trivial force in human experience. Underlying all verbal activity is an inherent order that echoes the dimensions of the nawasanga. The written shapes of language are not on the order of neutral codification of speech, neither are sounds arbitrary signifiers of the entities they denote. 39 There is, furthermore, a certain noetic foundation for the Balinese reverence for writing, facilitating as it does the preservation of utterances beyond the speech event that produced them, and allowing the silent symbol to carry the potential force of utterance (as in the modre). The aksaia, a term that in Sanskrit meant 'indestructible' (Hooykaas 1978.76), rep­ resents the "eternal" quality of language recognized in cultures under the influence of technological processing of speech, which perpetuates verbal acts through time It follows, then, in perfect concord with general Balinese opinion on matters of mystic importance, that it can be quite dangerous to ap­ proach the aksara and the knowledge they contain without basic prep­ aration in the mystic aspects of linguistic meaning. The lontars, like other sakti 'powerful, magically endowed' objects in Bah, are pos­ sessed of an ambivalent power that is beneficial or harmful depending on the conditions under which that power is activated The following passage from a text on healing arts used by the bahan gives an impres­ sion of the dynamics of literate awareness. Yika yan sua arep wruh ring kejatianing sangiang α)ΐηέ tan wan­ ting muang wenang ta sua angambekang ring awak d£w6k Yan durung weiuh ta sua ring kejatianmg raga muang otot pemuhhpemuhhamng awak, aywa ta sira anyemped dasaksaia, mapuaia kado ta sua, muang baya yan lwang έάαη palan ta Pati jdamut, seksekang angsengan, kebus bang siiah, matemahan bisu If one wants to know the truth of the noble teachings, [there are] 38 The Balinese term bi]a is also used to denote special grains of rice that are added to holy water during certain purificatory rituals 39 Other cultural traditions manifest their own varieties of this principle The Judaic tradition of the Khabbala, for example, outlines a mystic doctrine of lan­ guage transmitted orally from teacher to pupil, whereby the twenty-two letters of the Hebrew alphabet are each considered to bear a concealed metaphysical truth

59

THE SHAPE OF THE WORD IN BALI

some do's and don't's [regarding] contemplation of one's body. If one does not yet know the truth about physical form and sinews, the joinings within the body, one must not take the dasaksara lightly. [This] causes failure, and danger: if [there be] mistakes, in­ sanity results. Confusion, short-windedness, a red and feverish head, [and] finally one becomes dumb. 40 Thus, persons who have prepared themselves to study written texts are also getting ready for contact with extraordinary forces. Special protection becomes necessary in the face of intensive contact with the written word, which is under the divine aegis of Saraswati.41 Special­ ists in the goddess's field are expected to undergo rituals of purification called pawintenan, pangidep hati 'awakening the mind', or panggalang hati 'enlightening the mind' (Hooykaas 1977:30, Ensink n.d.). There are a number of types of these ceremonies, and the particular ones an individual participates in depends on whether he or she is be­ ginning a course of study, or has completed an endeavor in letters and sciences, or is taking on a new, spiritually engaging role, such as be­ coming a lay temple priest (pamangku). In the case of the pawintenan, a padanda performs an extensive purification that includes bathing the initiate's torso in holy water and uttering long chants requesting that his or her thoughts, words, and deeds (idep, sabda, and bayu) be aided by Saraswati. Often, sacred syllables are written on various parts of the initiate's body, such as the forehead, tongue, chest, and shoulders. The idea that protection and initiation are important in the study of written literature reinforces belief m the special potency and mystic power of certain aspects of language activity, particularly those involv­ ing the use of lontars. Even commonplace activities connected with the use of manuscripts take on special significance in the Balmese world of supercharged linguistic meanings. The Tutur An Saraswati lists special mantras to be used when (1) inscribing [anyurat] a lontar; (2) destroying (literally "killing," mamatiang) old, no longer usable lontar leaves (these must be burned specially, not just "thrown away"), (3) opening a lontar and preparing to read {amaca), and (4) putting away [nyimpen] a lontar after reading. Related to special proscriptions about handling the written word is a doctrine of revelation, haywa ννέια or a/α w£ia, 'do not reveal/divulge'. Although this doctrine has often been interpreted as forbidding people 40

From the lontar Usada San, collected by D J. Stuart-Fox from the materials used by Mangku Liyir, Pangosikan, Ubud 41 See the discussion of Saraswati in thefirstsection of this chapter 60

LANGUAGE FROM BIRTH TO DEATH

of common caste access to the lontars, many Balinese literati nowa­ days concur that aja wera-type injunctions commonly found in the manuscripts are meant to restrict certain types of knowledge, invaria­ bly mystic, magic, or esoteric in nature, to those persons who have studied enough to receive them, regardless of caste. One's teacher or guide in esoteric inquiry is responsible for deciding when the pupil should be given access to certain types of information, discretion in revelation is an issue when it is believed that formulas, chants, and se­ crets of meditation can be turned to undesirable ends by those of ma­ levolent intent. The following excerpt from the lontar Anggaksaia is a variation on the aja wera theme. Nihan upadasa warahakena ring sisia, aywa kaienga d&mng wong han, Ιιιιτέ Sang Hyang Niskala, tan paiupa, tan pawaina, tan pad6sa This is the lesson to be taught to students, do not [let it] be heard by other people, [as it] concerns the Great God Unmamfest, with­ out form, without color, without place. When we consider the above notions about the potency of the writ­ ten word, it comes as no surprise that the Balinese revere manuscripts and place constraints on the study of certain languages and texts Where and how lontars are kept, the ceremonial activity that accom­ panies their handling, the process of "initiation" into the study of sastia 'written texts', the special times set aside to honor books and teach­ ers, the power of language to aid the human being at critical points in life's cycle—all these cultural facets build into the Balinese world a rich symbolic component of language meanings. While for the linguist an "utterance" may exist as a "given" bit of information and stand at the beginning of the analytical procedure, for the Balinese an utterance may instead be viewed as the end of a long process of bringing language from its seat of power within the body into outer, active, verbalized re­ alization. This further implies that while the study of language and written texts is not exactly the same as the study of mystic truth in Bah, the two are very closely related, and each shares elements of the other. One does not become an expert in classical Old Javanese poetry without being aware of the relation between the written sign ONG and the Balinese concept of an absolute divinity. As human beings, our capacity for speech heightens our awareness of what is within us—in spiritual terms, the "inner" self—if only be­ cause, physiologically, spoken sound emanates from and resounds 61

THE SHAPE OP THE WORD IN BALI

within the the central torso area. Inherent in verbalization is a force associated with our vital organs The Bahnese concept of the dasaksara links the physiological fact with a multitude of linguistic symbols, each of which reiterates the relation between language and various lev­ els of universal truth. In Burke's (1970) terms, there is a "logological" relation between Bahnese metaphysics and linguistics, in that the na­ ture of transcendent reality is expressed as an analogue of written sym­ bols. These symbols stand for sounds that are successively " b o r n " and "dying" in speech, while themselves remaining in "imperishable" [aksaia 'not-perishable') testimony to the eternal quality of divinity, cosmic order, truth, and so forth. The completion of the logological process is the re-borrowing into beliefs about language of the same con­ cepts of transcendency that were originally modeled after language, and t h u s we find the heightened connection of deities and metaphysi­ cal wisdom with writing 4 2 The student of traditional letters in Bah is not always in a priestly or esoteric role, however. One can enjoy literature and use manuscripts for practical ends without pursuing mystico-hnguistic awareness Yet at t h e same time, and because of the general cultural context in which language and writing are not taken for granted as mere h u m a n attri­ butes, there is the realization that studying texts may be the beginning of a spiritual journey and transformation. Extensive exploration of lit­ erary and esoteric works is considered particularly appropriate for per­ sons whose social position, age, or experience have prepared t h e m for encountering the metaphysical side of language. In fact, many people only take up t h e study of Ion tars, join reading groups, or become in­ volved in professions based on the use of manuscripts when they have reached an age at which earning a living, raising a family, and other de­ manding "worldly" affairs may be partially set aside Others may turn to sastia after a serious illness, a psychological change such as pro­ longed depression, or some mystic experience, any of which might be 42

Other possible logological patterns can be seen in the meanings of words like nyutia (from Sanskrit sutra 'silk') and sandhi 'merging, becoming one' Both are used extensively as metaphors for mystic transformation and are based on linguistic terms for processes of sound assimilation Another example of a logological figure much favored by Bahnese literati comes from the "Bhagavadgita" section of the Bhismapaiwa, where Kresna reveals to Arjuna that Kresna himself is the life-force behind language, the sound that is part of each syllable "Among all letters my name is Ά,' " he says, referring to the basic form of each element of the syllabary, a consonant shape plus a sound (see Gonda 1935 75) 62

LANGUAGE FROM BIRTH TO DEATH

interpreted as a sign from the gods that the individual should become more involved with spiritual concerns—by serving as pamangku for a village temple or by taking up the bahan's training, for example. A recent case in point, which attracted the notice of the provincial newspaper, was that of a thirty-one-year-old man who, after a protracted illness that neither traditional arts nor modern medicine was able to cure, as a last resort took to sitting in temples and meditating at night, finally emerging both cured and with the announcement that he had been instructed by the gods to become a dalang (shadow-puppet master), a profession that also requires the study of lontars 43 In Bah, as we have seen, close contact with literature and manuscripts reaffirms the place of the individual within an ever-revolving symbolic order—the nawasanga—which is highly valued throughout the culture. The significance of language in its written form is tied to the role of writing as the shaper of that profound relation between the self and eternal cosmic order, and also to its role as the representation of the past. Revivable utterances are those recorded in painstaking inscription on palm leaf, and the archaic linguistic forms of old messages are another token of the value of recorded, repeatable language. Because of the existence of lontars, many generations of Balmese, literate or not, directly or indirectly, have had access to a succession of utterances in manuscript form, each of which, now separated from its historical origins, represents a complex of both diachronic and synchronic perspectives and ways of knowing, by virtue of its different formal and contextual features The language specialist is closest to the language of the past, and is responsible for the way that past meaning is transmitted in the present—an mterpreter of valuable communications from vague predecessors, serving society so that recorded experience and cultural truth may not be lost to unwritten silence. LANGUAGE AS SOCIAL DYNAMIC Thus far this chapter has dealt primarily with the role of written language in Balmese experience The present section, therefore, will treat some of the cultural and social aspects of speech. The Balmese vocabulary includes a great variety of words that refer to spoken language, from those generally denoting verbal activity to those specifying particular sorts of speech acts. The terms raos or baos 'speech', basa lan43 As reported in the Bah Post (24 January 1979) under the headline " 'Dalang Gaib' dan Penarungan "

63

THE SHAPE OF THE WORD IN BALI

guage', suara 'voice', omong, munyi 'speak', orah 'say', andika 'a ruler's speech to subordinates', and tutur 'tell, advise' are only a few of the words having to do with spoken language In Chapter 1, I noted that in modern spoken Balinese there exists synchronic formal variation specifically tied to social structure—the feature well known under the rubric "language levels " I also alluded to certain diachronic processes that have contributed to the presentday shape of spoken Balinese, including periods of influence from Sanskrit, Old Javanese, and the Majapahit-era language of Java (related to, but not identical with, the "Middle Javanese" of the kidung literature) The present discussion is aimed toward clarifying some of the forms and functions of language levels as these levels relate to Balinese beliefs about language in general and also to the concerns of "textuahty" that are a focus of this book The as yet untraced development of language levels or speech registers is a large and complex issue, yet we do have several facts at hand to guide our inquiries 1 Balinese speakers have been familiar with multiple-code practices from the earliest historical period, when Sanskrit terms were already being used in royal and religious spheres 44 2 Indonesian languages in general tend to place importance on the "chne of person," that is, on those linguistic forms that show variable degrees of status and intimacy among participants in a speech event 45 3 Among the Indonesian languages, elaborate systems of language levels are found only in Sundanese, Madurese, Balinese, Javanese, and Sasak, which suggests that Javanese or Javano-Bahnese political and cultural influences have been central to the development of this linguistic feature 46 4 In Bah, Javanese linguistic influence has been important, at least as regards texts, since the tenth century 5 The much later settling in Bah of Javanese priests and nobles during the Majapahit era catalyzed a period of increased influence of Javanese language, such that spoken Balinese developed a more complex range of variation along the "chne of person," as the social order was modified along the lines of Hindu-Javanese models 44

See the first section of Chapter 1 for discussion of this process The linguistic implications of the "chne of person" have been explained for the case of Old Javanese by Becker and Oka (1977) 46 For analysis of the language levels of Central Java, see Poedjosoedarmo (1968), Poedjosoedarmo and Koendjono (1977), and Wolff and Poedjosoedarmo (1982) For Madurese, see Stevens (1965) 45

64

LANGUAGE FROM BIRTH TO DEATH

6. Present-day language level usage, while varying from one area of Bah to another, is a fundamental feature of spoken language in those regions (primarily the lowlands and coasts) having prolonged interaction with the Bahnese courts of the post-Majapahit period, with the most elaborate language level systems traditionally being found in closest proximity to those courts. 7. At the same time, more outlying areas (usually in mountainous regions), which are associated with a sociocultural type less influenced by Hindu-Javanese Majapahit, do not evidence the language level phenomenon 47 In the absence of a thorough developmental analysis of Bahnese,48 the facts listed above can still broaden our perspective on today's language levels Synchronic descriptions of the language level phenomenon are several, beginning with Friederich's (1876-1878, first published 1849) statement that Bahnese language "is divided into a High and a Low, the first being spoken by the lower to the higher orders, and the last by the higher to the lower" (p. 2) Both Fnedench and Kern (18701871) noted that the Bahnese "high" language had many Javanese terms, but that their relation to the "high" [kiomo] and "low" [ngoko] Javanese levels did not indicate a straightforward borrowing. The next description that focused on the issue of language levels was by de Vroom (1872), who discussed the "dialects" involved in masoismggih, a Bahnese term meaning 'to lower and elevate [using language]'. De Vroom talked about formal variation (noting, for instance, that pairs of "high" and "low" words often showed minimal phonological contrast), as well as social functions (e g., the difference in usage between the three highest castes, tnwangsa or anak manak, and the common caste, sudra or jaba, majority). When van Eck (1874) produced the earliest grammatical description of Bahnese, he included sections discussing the "high" versus "low" vocabularies, and he related these to syntactic categories such as the personal pronouns In a later grammar, which has become a standard reference for Bahnese, Kersten (1970, first published 1948) pointed out the distinction between "speaking to" (the second-person interlocutor) and "speaking about" (the third-person referent) as reflected in formal classes of words. Kersten noted that different vocabulary items had different numbers of level variants, one set of variants might have two forms, another three or more, and most words had none. He also 47 48

See Jendra's (1976a) survey of dialectal variation Some discussion of diachronic change in Bahnese is given in Ward (1973)

65

THE SHAPE OF THE WORD IN BALI

suggested that speakers of Balinese often mix forms from different levels in one utterance. Most of this early work on Balinese made use of the lexicography of van der Tuuk (1897-1912), who included notes on regional and social variation in his massive dictionary. In the last decade, a number of works addressing the issue of Balinese language levels have appeared. Ward (1973) formalized the notion of a semantic set that comprised all the level variants of one concept, and he also made the distinction between "structural" variation (involving primarily function words, "functors") and "lexical" variation (involving content words). He also provided a fairly complete list of terms demonstrating language level variation within one area of Bah. Keeler (1975) took a more ethnohnguistic approach in his analysis, describing the overall cultural setting that determined formal variation. Shadeg (1977) produced a list of 1,000 Balinese words, giving level variants along with example sentences for each set. Most recently, Bagus (1979) described a shift in the use of "polite forms" concomitant with sociocultural change, continuing his work on the notion that "polite form" usage is changing to override traditional social categories and to highlight newer distinctions and definitions of the elite (Bagus 1975). While the contributions of these authors have gradually built a composite picture of Balinese language levels that is largely accurate and consistent, there nevertheless remain several points insufficiently clarified and a number of basic assumptions that require further examination. At this point I think it is necessary to review a few of the explicit conclusions or implicit assumptions of previous analyses and to present my own evaluations and suggestions. I will first treat conclusions regarding formal aspects of language levels, then consider functional concerns. A good way to begin talking about Balinese speech register variation is to reconsider the appropriateness of the term "levels" itself. This label is perhaps accurate in implying a hierarchy of more or less honorific forms, yet it wrongly gives the impression that every element of Balinese speech can be clearly identified as belonging to a fixed set of airtight categories, such as alus or kasai. These terms, which can respectively be translated as 'refined' or 'high' and 'coarse, common' or 'low', have been the basis of conceptualization of spoken Balmese in terms of high versus low levels, and they have given rise to the fallacy that a binary opposition is the foundation of level variation. This binary opposition of alus/kasar is partially accurate in reflecting a duahstic hierarchy of caste status (tnwangsa versus jaba) that is intimately tied to formal variation in speech; yet level choices are made even between 66

LANGUAGE FROM BIRTH TO DEATH

jaba speakers, and the lexical oppositions are not always so binary, or the status issues so bilateral, as the alus/kasar terminology suggests. To illustrate the points I have just stated, when we look at the Balinese lexicon, we see that most items do not show level variation at all. Of those that do (between 600 and 1,000 items, depending on the estimate), the dimensions of contrast (that is, all the level-marked var­ iants of one concept) clearly differ from one semantic set to another. A small cake/cookie (noun)

SET Β eat (verb)

sanganan ]aja

marayunan ngajeng nunas madaar neda ngamah

SET

SETC

not yet (aspect marker) durung doiang ίοηάέη

Each of the semantic sets above has multiple lexical referents for one concept. Only in Set A, however, is there the two-way contrast con­ forming to the high/low or alus/kasar paradigm, sanganan is the "po­ lite" term for 'cookie' and would be used in the context of speech di­ rected toward some higher-status auditor, while jaja is the "common" form to be used with intimates and inferiors. A large proportion of the semantic sets in spoken Bahnese are of the type exemplified by Set A. There is a significant group, however, that conforms to the pattern of Set B. Here, referents for the concept 'eat' all have specific and limited uses and crosscut the alus/kasar categoriza­ tion in various ways. Marayunan is an honorific used only for persons of the highest status, such as priests and princes; ngajeng is a more gen­ eral high term that is also used in polite speech. Nunas is used in polite speech too, but its function is to deprecate, rather than elevate, the firstor third-person referent in relation to the second person. Madaar is the low or kasar form used reciprocally among intimates or status equals, and used by superiors to inferiors Both neda and ngamah are terms used for animals, and when applied to humans they can be derogatory, pejoratively kasar in the sense of 'coarse, crude'. The semantic sets with contrasts like those of Set Β have varying numbers of terms per set. Most of them seem to share the feature of being linked to the con­ ceptual realm of "personhood," or, as Ward (1973:105) has stated, they are used for "speaking of actions, attitudes, thoughts, feelings, posses67

THE SHAPE OF THE WORD IN BALI

sions, body parts, social relationships, or other cultural concepts, val­ ues, or objects which serve as a means of self-identification and -es­ teem " A third pattern of lexical contrast is represented by Set C, which in contrast to Sets A and Β falls into the category of "functional" rather than "contentive" vocabulary Duiung is the highest form in this set, while daiang is also polite, and ΐοηάέη is low Other "functors" in Bahnese may have more or fewer members per lexical set, but functional vocabulary almost always contrasts in some fashion Thus, while every word in a Bahnese utterance does not necessarily show "level" contrast, almost every utterance must evidence some choices among lexical items, "since very few sentences are totally free of functors" (Ward 1973 104) The three sets of lexical contrasts given above do not exhaust the types of such oppositions in Bahnese language There are other pat­ terns of opposition, such as the "double reference" verbs (e g, 'give', 'order', 'speak to', and so forth), in which the vertical status trajectory between both actor and recipient is gauged vis-a-vis the speaker and hearer for each speech level (Ward 1973 110) Another example is the direct mention of person (first, second, or third), which involves a com­ plex of pronouns, relational terms, status titles, birth-order names, and personal names Bahnese speakers thus must control a profusion of se­ mantic sets whose conditions of contrast are variable Not all speakers have equal command over an identical range of vocabulary, however Geographical area, social position, age, and experience combine to condition each speaker's competence Not all Bahnese may agree, fur­ thermore, as to the lexical contents of or level distribution within any one semantic set, regional variations are many Despite the obvious difficulties in developing a single paradigm to encompass the varieties of contrast among semantic sets, Bahnese metalinguistic terminology indicates the important distinctions to be made Among native speakers, the terms basa Bah 'Bahnese language' and omong Bah 'Bahnese speech' are frequently heard referring to the language as a whole, with omong alus or basa alus being used for the "polite" component 4 9 The alus area of language can be further subdi­ vided into the following categories (see Bagus 1979 197) 49

Less frequently used terms for alus speech include basa smggih, basa manak, and basa bakti Archaic, rarely encountered terms include natia (alus), sogol (kasar), mabasa 'to speak', used in the sense of 'to speak using polite vo cabulary', and jumalada 'middle range speech' More specialized terms for cer tain speech styles include asah basa 'symmetrical use of kasar', maseh basa 68

LANGUAGE FROM BIRTH TO DEATH

alus mider ('all-around'), "basic" alus vocabulary, used in polite speech alus madia ('middle'), middle-range alus vocabulary, slightly less po­ lite than mider forms alus smggih ('affirmation' or 'respect'), special honorific vocabulary used for second- and third-person reference, elevating in relation to the speaker alus sor ('beneath'), special deprecatory vocabulary used for first- and third-person reference, lowering in relation to the hearer The non-alus area of Balinese, including those lexical items that have level contrast and those that do not, is often termed basa kasar Some speakers, however, use the terms basa kapara, basa ketah, or basa biasa—all of which mean 'general/usual language'—in order to avoid the negative connotations of "kasar," namely, 'crude, coarse, un­ refined' 5 0 We are beginning to see that Fnedench's convenient and symmetri­ cal definition of language levels glosses over a number of significant is­ sues Likewise, the statement in one of the most widely read accounts of Balinese culture that "the high and low tongues are two distinct, unrelated languages" is totally misleading in terms of the nature of so­ cially determined variation in spoken Balinese (Covarrubias 1937 51) Rather than conceiving of each utterance in Balinese as belonging to one of a fixed number of discrete "levels," it may be better to consider levels as primarily consisting of criteria for ordering and selecting items from within the semantic sets themselves This point becomes more apparent when we look at contrastive sentences formed from the lexical items included in sets A, B, and C above Sentence 1 Icang PI (k)

ίοηάέη not yet (k)

naai eat (k)

jaja cake (k)

ηέ baang τηέτηέ rel give mother (k) (k) (k)

I haven't eaten the cake which mother gave [me]

'code-switching from alus to kasar', nebah basa 'using kasar with the interloc utor', degag 'using the wrong level', )abag 'using kasar speech insultingly' There are many other regional terms for speech styles in Balinese, many of the above are found in Bagus (1979 chap 3) 50 The issue of applying the term kasai to lower-level speech remains a sen­ sitive one in some quarters, since the high-caste and low caste Balinese com­ mentators on the subject do not always share the same perspective on the con69

T H E S H A P E OF T H E W O R D I N B A L I

2.

Icang PI (k)

tondénnot yet (k)

naar eat (k)

jaja cake (k)

rel (k)

icénin give (a si)

ida P3 (a si)

ratu ruler/lord (a si)

biang. mother (a) I haven't eaten the cake which she [i.e., the queen mother] gave me. 3.

Tiang PI (a)

déréng not yet (a)

naar eat (k)

sanganan cake (a)

ané rel (a ma)

icénin give (a si)

ida. P3 (a si)

I haven't eaten the cake which s/he gave me. 4.

Titiang PI (a sr)

duiung not yet (a)

nunas eat/ask for (a sr)

sanganan cake (a)

san6 rel (a)

wdhin give (a sr)

ipun P3 (a sr)

I haven't eaten the cake which s/he gave me. 5.

Ratu ruler/lord (used as P2) (a si)

duiung not yet (a)

ngajengang eat (a si)

sanganan cake (a)

sané rel (a)

aturang give (a si)

titiang! PI (a sr) My lord hasn't eaten the cake which I gave [you]? 6.

Ida P3 (a si) atuiin give (a si)

padanda Brahmana priest (used as title)

tondén not yet (k)

ngiayunang eat (a si)

jaja né cake rel (k) (k)

icang. PI (k)

The priest hasn't eaten the cake which I gave [him[. Symbols used:

(k) (a)

kasar or 'ordinary'-register words alus mider

notations of kasar, which in non-linguistic contexts signifies roughness and inferiority.

70

L A N G U A G E FROM BIRTH TO DEATH

(a ma) (a si) (a sr) Ρ rel

alus madia alus singgih alus sor pronoun—1, 2, or 3 relative pronoun

When explaining the meanings of the sentences above, one focuses on the continuous alteration of social relations indicated by personreferencing words and "shifters" such as pronouns. Sentence 1 indi­ cates a symmetrical relationship between speaker and hearer. If the first-person pronoun were altered, say, to beh 'older brother', while maintaining the first-person reference, the situation would include slight asymmetry, as from an older male relative to a younger, albeit within the same caste and "intimacy" range. Sentence 2 again presupposes a speaker and hearer symmetrically exchanging low-level speech, except that the reference to the third per­ son employs a high-level verb [ιοέη, for someone of higher status giving to someone of lower) and an honorific title [ida ratu biang, a term on the order of 'queen mother' for royalty). In sentence 3, the functors, an aspect marker and relative pronoun, have changed and are of the alus madia category. This utterance could be used between persons not fa­ miliar with each other's position, or perhaps by a lower- to a highercaste person when the relationship is one of long familiarity. The third-person referent is still indicated by hononfics and is of higher sta­ tus than either speaker or hearer. Sentence 4 is spoken by a lower-status speaker to an interlocutor of much higher status; the utterance might also be used in a situation of extreme formality between high-caste persons of differing rank. The third-person referent here is of lower status than the hearer, as indi­ cated by the deprecating verb and pronoun [w6h, ipun). The speaker uses a deprecating form {nunas) for his own "eating." In sentence 5, the action of "eating" is directed toward the elevated second person, who is also the recipient of the action of the verb atui, used for 'give' when the giver is lower than the (second-person) recipi­ ent. In sentence 6, the speaker addresses the hearer with kasar speech, except for references involving the high-status third person, for whom the highest honorific "eat" is used. The speaker also employs a form of the verb atur that is used when the third-person recipient is higher than the giver. As the above examples show, Bahnese speech cannot be compared to a train that switches between one or another of various parallel 71

THE SHAPE OF THE WORD IN BALI

"tracks." It is, rather, a continuous movement along a continuum of more-or-less, showing reciprocity, deprecation, or respect at different points in an utterance. The term "language levels," implying absolute and discrete oppositions in utterances, rather than just within the lex­ icon, is in this sense misconceived The Bahnese clearly feel that speech can be directed " u p " and "down" at the same time. Their image of language evokes the waves and contours of utterances that delineate all the dimensions of Bahnese social space, terms such as anggah-ungguhmg basa, tegeh-6nd6k basa, and sor-smggih basa, which can be translated as "the highs-and-lows of language," do not imply that "high" and "low" are mutually exclusive at the level of discourse Honorific as well as deprecating vocabularies, although theoretically subdivisions of alus speech, occur along with kasar vocabulary when the context is appropriate (sentences 2 and 6 above] Yet there remain several important properties of formal variation in Bahnese speech that are not highlighted in the six sentences given above. Those examples exhibit extreme formal symmetry, they are il­ lustrations of maximal contrast in that each morpheme m their paral­ lel syntax has correspondence with another morpheme of the same class in all other examples. Actually, many Bahnese utterances that are semantically equivalent (i.e., have the same content) and yet prag­ matically opposed (must be used in different contexts) do not share for­ mal structure. A possible contrast between alus, more honorific speech, and utterances in the biasa or "everyday" mode is shown in the following sentences. 7.

Titiang PI (a sr)

nunas ask for (a sr)

wau just now (a)

rauh come (a)

lugia, ratupadanda, permission/ lord priest pardon (used for P2) sakwg from (a)

padanda P2 (a si)

napii what/where (a)

I beg pardon, my lord, where is my lord coming from? 8

'Κα come (k)

'h from (k)

'jai where (k)

Where are you coming from' The above two examples demonstrate in extreme form the tendency for higher, more polite, and formal speech to have proportionately 72

LANGUAGE FROM BIRTH TO DEATH

more verbal material than lower styles. Sentences 7 and 8 are both per­ fectly typical greetings, on the order of "How are you?" The addressing of any such remarks to a Brahmana priest (padanda) is, however, of ne­ cessity accompanied by an apologetic phrase [titiang nunas lugia) and honorific lexical choices, indicating the speaker's respect and self-ef­ facement, which are also important parts of utterance meaning. In con­ trast, sentence 8 would probably be spoken between low-caste inti­ mates. Furthermore, each of the morphemes in sentence 8 is actually a shortened form of a two-syllable word,- the full lexical representation of the utterance would be Teka uli dija. Such contraction of forms in fast-speech style is extremely common in everyday intimate language, particularly in the southern and south-central parts of the island. In po­ lite styles, fast-speech rules cease to operate, and particles, affixation, and reduplication become more elaborate. Thus, the Bahnese exploit a principle of "more sound higher, less sound lower" as a non-lexical strategy for marking social distinctions in speech Most equivalent morphemes across levels demonstrate this principle. For example, the relative pronouns ηέ, αηέ, and sana in ex­ amples 1 through 6 above represent the kasar/alus madia/alus mider hierarchy, and each has progressively more phonological material. The information carried by the redundant sound is not prepositional but pragmatic, having to do with the relationship between speaker, hearer, and other. In conveying propositional content, then, Bahnese speech utilizes a phonological redundancy principle to indicate degrees of re­ spect and formality on the one hand and informal reciprocity on the other. The considerations discussed above help us to realize that speech "registers" in Bahnese are not just a lexical phenomenon within the language There are indeed complex and standardized parahnguistic features that go hand in hand with lexical opposition in determining utterance register. Such features include speech rhythms and tonal contours, position of torso and hands, head and arm gestures, direction of gaze, relative vertical location of interlocutors, and relative place­ ment of interlocutors according to the cardinal directions.51 The emphasis in Bahnese speech registers on status-referencing forms of address, on the use of lexical hierarchies for objects and ac­ tions closely associated with the individual, and on norms of parahn­ guistic behavior geared to the range of formal variation all lead to the 51

Correct parahnguistic behavior can compensate for imperfect use of vocab­ ulary, and "Injot using correct intonation and bodily expression in status rela­ tions is just as serious as choosing inappropriate words" (Walks 1980 71-72)

73

THE SHAPE OF THE WORD IN BALI

conclusion that the speech register system is in fact centered around issues oi person. By "person," I refer not to an abstract grammatical category, but to a dynamic concern for adjusting speech so that in each unique speech act individual identities are appropriately marked according to their mutual relationships.52 The methods used to differentially mark relationships involve, as I have suggested, many levels of the linguistic system. We have noted that on the lexical level not all items show level variation, it is also abundantly clear that not all terms that do vary have the same weight m determining the register of an utterance. Those lexical items that are most powerful in marking speech as higher or lower are those that have the greatest number of level variants, it is these forms that speakers must be most careful to choose correctly, and that are most dangerous to misapply.53 It is even possible to be forgiven errors in the less sensitive areas of vocabulary as long as the crucial choices are correctly made. If we look at the range of parts of speech in Balmese, it becomes possible to sort out a "hierarchy of force" that ranks classes of words according to the relative strength with which they mark utterance register. At the top of the hierarchy are certain "person-references," namely, the group of pronouns, relational terms, titles, names—all of which serve to distinguish first, second, and third persons. Slightly less powerful than this category are certain functors—negators, demonstratives, existentials, and deictic pointers—which also have a great deal of force in marking register. Interrogatives, verbals, nominals, and so forth can also be evaluated for the force with which they mark register. One procedure for testing the relative strength of various classes is to evaluate "middle-level" speech, the utterances of which include both alus and kasar/biasa vocabulary. In order to raise a lower-register utterance, one notes those morphemes that must be altered. To take a 52 So important is the notion of person that there are actually constraints on leaving out from utterances this highly marked category of personal reference For instance, the term basa gabeng is used to characterize an utterance lacking specific mention of the hearer, gabeng literally means 'a grain of rice without a kernel' Similarly, speech addressed to persons of high caste often includes the particle tu, from latu 'king' as a way of continuously marking the speaker's polite intentions 53 In the past, severe penalties could be claimed from persons using insulting or misapplied vocabulary, at the very least, the offender would be required to furnish a purifying ceremony with offerings called piayascita

74

LANGUAGE FROM BIRTH TO DEATH

simple example, the utterance "There isn't/aren't any" in kasar/biasa speech is 9.

Sing (negative)

ada. (existential)

In a more polite or alus madia form, one would say 10. Ten (negative)

ada,

and in honorific speech there are two higher forms, 11. Ten

winten (existential)

and, higher still, 12. Νέηίβη (negative)

w6nten.

From these examples it is possible to conclude that negative particles are higher on the scale of register force than existentials, since the neg­ ative ten (a shortening of ηέηίβη) is sufficient to mark utterance 10 as alus madia. Furthermore, the utterance Sing wonten, which combines a low negative and a higher existential, is not found. The "hierarchy of force" principle allows us to see the dynamism of registers as they are actually applied in Bahnese speech. Speakers are aware, and make differential use, of those areas of language that are most significant in determining the hearer's evaluation of an utter­ ance. The manner in which individuals make decisions about which forms to use when and where belongs to the area of the functional roles of speech registers. It has often been stated that the hierarchy of caste is the primary de­ terminant of who speaks what to whom (see, e.g., Keeler 1975:1 ΙΟ­ Ι 11; Ward 1973:102-103). Caste is a matter of assigned status—it can­ not be acquired other than through birth or, for some women, marriage 54 (and then in a somewhat altered form) —and in some cases caste op­ erates independent of any other variable in determining the linguistic register used in speaking. Many Bahnese, when asked to account for 54 A woman who marries a man of higher caste, especially one from a ruling princely or Brahmana family, is usually given a special title or new name to sig­ nify her changed, but not entirely equal, status

75

THE SHAPE OF THE WORD IN BALI

the presence of lexical variation in spoken language, ascribe the varia­ tion to the social fact of the four caste groupings, known as the catur janma or catur wangsa, or simply as the anak jero or triwangsa 'insid­ ers, three high castes' versus the anak jaba 'outsiders'. While there appears to be relatively little information about lan­ guage registers handed down in written form in Bali, a few treatises do bring up the subject. For example, the A]i Purwa Basita Krama,55 the 'Teaching on the Origin of Order in Language', renders in some detail the manner in which members of different castes should address each other. Kunang kramamng sabda basita, y&n tan tunggal kulawangsa, luirnya, γέη Sudra abasita ring Wasia: Smggih si Tankawur, tiang; ring Satna andirv Smggih dowa, titiang; yan ring Brahmana ulaka· Smggih paduka sang mahampu, titiang; yan ring sang Brahmanapandita• Smggih pukulun paduka sangmahapandita, titiang; yan ring sangprabu, ab^ka jumenengratu Smggih pukulun paduka sang mahabupati, atur titiang sahasembah Muah γέη W6sia abasita ring Sudra· kita, wargamngsun-, yan ring Satna andm: Smggih d6wa, titiang; yan ring Brahmana ulaka Smggih paduka sang pangempuan, titiang; yan ring Brahmana pandita: Smggih pukulun paduka sang mahapandita, atur ti­ tiang, atur ulun, wenang; yan ring sang ]umeneng abisoka ratu: Smggih pukulun paduka sang sn maharaja, atur ulun These are the forms for spoken language, when [speaker and hearer are] not of the same group. The situation is, if a Sudra speaks to a W6sia: "Smggih mister-not-confused" [to the hearer], "tiang" [re­ ferring to the speaker]. If [he speaks] to nobility descended from a king: "Singgih lord," and "titiang." If to a non-ordained Brah­ mana: "Smggih sandal of the great master," and "titiang." If to a Brahmana priest: "Singgih the slave of the sandal of the great priest," and "titiang." If to the king, consecrated to govern as ruler: "Singgih the slave of the sandal of the great ruler," and "I [titiang] offer obeisance." And if a W6sia speaks to a Sudra: "You," and "my group." If to nobility descended from a king: "Singgih lord," and "titiang." If to a non-ordained Brahmana: "Singgih san55

The text is in the form of a dialogue between Sang Hyang Brahma and Sang Empu Mahaddwa, who requests information on the essence of knowledge Sang Hyang Brahma summons Sang Hyang Tigajanyana, who provides the revelation on language 76

LANGUAGE FROM BIRTH TO DEATH

dal of the master," and "titiang." If to a Brahmana priest: "Singgih slave of the sandal of the great priest," and "I [titiang] speak," or "the slave speaks," will do. If to the one consecrated to rule as king· "Singgih the slave of the sandal of the honored great king," and "the slave speaks." The manuscript continues in the same fashion, covering all the combinations of forms between one caste and each of the others. Then the text takes up the forms of address between members of the same caste {tunggal kulawangsa) whose religious status [gama) and sociopolitical role [lungguh wibawa) are also parallel. In such a case, "their language forms are also the same" {kiama basitama tunggal ]uga). When the speaker and hearer are not of the same caste yet both have received religious consecration, special forms of address may be in order, a sudra who is consecrated as a sengguhu (a special priesthood for certain lower-caste persons) must be addressed as jew sang empu 'noble master' or jew gedo 'great nobility'. The speech norms enumerated in the An Puiwa Basita Kiama axe certainly not all in currency in modern Bali; neither can we assume that all the forms in this prescriptive treatise were ever in common use in one region of the island. The text combines Kawi (Old Javanese) forms with Bahnese ones, and at one point it glosses Kawi status titles with better-known Bahnese equivalents. Yet the types of distinctions that are made in the text, and the points of contrast illustrated, are quite typical of Bahnese speech today. One of the facts demonstrated by this manuscript is that, although caste is a primary determinant of speech variation, it is definitely not the only one. A ruling noble [piabu) must distinguish, say, other members of the Satria group according to whether they are of his own rank [andm) or subordinate (mawak kaula) or consecrated [wus magama). When speaking to a subordinate Satria, for example, he must determine whether the relation is most appropriately couched in terms of "child-father" ties [cemng-bapa) or those of "younger brother-older brother" [adi-beh). And he must bear in mind that low-caste (sudra) persons can have higher-than-usual status in the religious sphere (e.g., as a resi or sengguhu) or in public life (e.g., with the special titles pasek, pand6). Thus, even a ruling prince must be careful to differentiate between speech forms according to the relative status of his subordinates. In the dynamic social context of contemporary Bah, non-caste determinants of status have become increasingly important. Notions of ac77

THE SHAPE OF THE WORD IN BALI

quired status (related to education, wealth, employment, and urban residence) are competing with traditional inherited-status values 56 In the urban setting of the provincial capital, low-caste persons with jobs in the civil service or the professions may no longer regard caste differences as foremost in determining speech styles In place of speech in which degrees of social distance are precisely delineated by traditional status titles and honorific and deprecating vocabularies, a more neutral "language of politeness" using primarily middle-range [madia] polite vocabulary is gaining currency This style is used reciprocally in nontraditional contexts (e g , in a government office), between members of different castes, and even in the homes of urban, upwardly mobile groups The changes in the use of traditional speech registers among certain age and occupational groups in modern Bah have been researched and analyzed in detail by I Gusti Ngurah Bagus (1979 212317) Among other findings, Bagus has determined that rural, agriculturally employed groups have more control over the speech register range, and employ traditional register styles more often, than do employees of the civil service, persons in business, or students at university As a result of both the persistence of traditional speech norms and more recent shifts in register usage, it has become even more inaccurate to base a functional analysis of register variation on caste membership alone The degree of intimacy, age, sex, marital status, physical setting, event, features of the speech act, presence and identity of audience, even the types of clothes participants are wearing—all these variables differentially interact to control register choice In a situation in which register-determining factors are in conflict, we find speakers making special choices, such as the neutral middle-range polite speech of urban groups Partial or entire code-switching may also occur In this regard, the increasing use of Bahasa Indonesia, the national language, is especially significant 57 56 It is important to mention here that status rankings based on caste are not at all a static framework and may never have been As Boon (1977 chaps 7 and 8) illustrates, competition and jockeying between descent groups in the mythic, ritual, social, and political spheres are characteristic features, which both the Dutch colonial officers and the "Indiamzation" view of Bahnese society tended to ignore 57 Warnings about the "decline" of the regional language are not new in Bah As early as the 1930s, one commentator noted that Malay had already become a second mother tongue for many inhabitants, he also urged Bahnese who left the island for schooling to continue to use Bahnese upon their return, and he

78

L A N G U A G E FROM BIRTH TO DEATH

The complex of lexical, phonological, and paralmguistic features that make up speech registers is governed by a principle of variation that the Bahnese express as d&sa, kala, patra. This phrase, which is widely used when talking about Bahnese cultural variation in general, literally means 'village, time, leaf, [or name]', and it can be understood in the sense of "place, time, and custom." The linguistic operation of the principle provides that speech behavior is subject to constant fluc­ tuation, as participants find themselves in a variety of contexts and speech events and engaged with a range of different interlocutors. One must always be attentive to the different norms for changing situa­ tions and new localities. As an illustration of the desa-kala-patra principle, two Bahnese speakers not strangers to each other will not always use the same forms of address. For instance, two noble acquaintances, one a Brahmana and one a ruling prince,58 may use various forms depending on the circumstances in which they find themselves According to two such individuals I interviewed, when the Brahmana would visit the pun of the prince he would refer to himself as titiang, when he at­ tended ritual functions at the pun in a more official capacity, however, he would use as a first-person reference the more formal Brahmanan cokor ι dawa ('your majesty's Brahmana', literally 'Brahmana [at] the foot of the divine') Another instance of dosa-kala-patra as a functional norm is provided by the institution of the banjar, the sub-village residential unit or "hamlet" that holds monthly council meetings, called pauman or sangkepan. The members of the banjar council [kiama banjar) are. married heads of households who have known each other all their lives, yet even in banjars with no high-caste membership, the language of the council is always elevated in style 5 9 Anyone who addresses the meeting must use alus mider and alus madia vocabulary and refer to the members of the banjar with the polite forms ida and άαηέ. Persons who out of anger or carelessness use the more intimate, informal code of everyday exchange are subject to fines levied by the banjar. The principle of context dependency represented in d6sa-kala-patra lamented the tendency to substitute some Dutch words (especially pronouns, it seems) for Bahnese vocabulary (Ranoeh 1938) 58 Being a member of the ruling palace household of an area elevates one's sta­ tus considerably above that of other, "non-ruling" Satna groups Terms used to denote this highest Satna standing are abisika jumeneng ratu (see the excerpt from the lontar A]i Puiwa Basita Kiama above) and satna ning xat 59 See Hobart (1975 73-76) on the rhetoric of banjar meetings 79

THE SHAPE OF THE WORD IN BALI

applies not just within communities but also between them. Variation in the lexical membership of semantic sets and normative judgments as to what counts as an utterance within a certain register lead to distinct regional styles and tastes in speaking For example, the citizens of eastern Bah (Klungkung or Karangasem regencies) are sometimes regarded by the residents of southern Bah as "traditional" or "proud" because in polite speech with strangers they may use titles or relational terms (e.g., bapa 'father'), as first-person reference, rather than the more self-effacing titiang Communities count such formal stylistic variation in register usage among their unique identifying characteristics, thus, the use of maiayunan versus ngiayunang as an alus singgih form of 'to eat' may signal the speaker's membership in or exclusion from certain localized groups. The operation of the d6sa-kala-patra principle provides that Balmese speech registers involve only a limited amount of predictability. We cannot determine with accuracy what two persons will say to each other merely by knowing their relative caste status, a great deal of other information about the particular speech setting may be necessary. An example of this circumstance occurred when I heard several young Balinese of different castes speaking colloquial Bahasa Indonesia with each other. I was puzzled, for they were all from two neighboring villages and I was accustomed to hearing them use Balinese. Finally, I realized that the setting—a "modern"-style wedding reception complete with rock music and guests in Western-style clothes—determined that they would feel more comfortable if they masked their status differences by using a different code than if they adhered to the everyday, caste-indicating speech forms of the village setting. A corollary of the principle that speech registers involve limited predictability is the axiom that registers cannot be accounted for merely by diagrams with vertical arrows indicating the social hierarchy of the interlocutors. While speech registers are closely tied to the vertical hierarchy of status, they are also geared to the horizontal clme of intimacy that gauges relative degrees of familiarity and formality. Thus, when two Balinese speakers exchange alus forms reciprocally, they may be indicating a relatively formal "distance" from each other, not the egalitarianism of familiars that might be suggested by the use of such symmetrical forms Likewise, the asymmetrical exchange of kasar and alus can be coded for extreme intimacy, as between an older court servant and a younger member of the aristocracy. This is the special function of the speech style called (in some areas) basa kuma,60 60

The word kuma is said to come from ka uma 'to the fields', suggesting sit80

LANGUAGE FROM BIRTH TO DEATH

which uses shortened status titles, among other forms, to encode intimacy while maintaining status difference. Another indication of the dynamism underlying the surface categories and oppositions of speech registers is provided by the true function of alus madia, or "middle-level" speech. There are relatively few morphemes regularly labeled as alus madia forms, Bagus (1979 200) counts only 28 items out of his total of 755 forms that show variation.61 The traditional motivation for using alus madia forms involved two criteria: first, they were used by a high-caste individual when speaking to a sudra of special status, such as a village headman or lay priest, second, and more generally, they were and are used between interlocutors who do not know each other. The "neutrality" of the alus madia vocabulary ensures that two individuals meeting in a marketplace or on the road will largely avoid any major register errors Until the identities of the individuals are better known, alus madia speech facilitates feedback and self-correction as speakers adjust their language to the social situation at hand. The combined effect of the functional criteria I have been noting should be convincing evidence that speech registers belong to a special linguistic realm in which, as with other sociohnguistic indicators, the data we study do not always combine to form elegant paradigms governed by precise rules. The intentions and motivations of Bahnese speakers, both as individuals and as members of society as a whole, are subject to fluctuation So great is the determining force of context on speech register that lexical levels are difficult to test with the same reliability as syntactic intuitions. This is the case because registers are a discourse strategy in Bahnese, used to show shifting emotions, relationships, and points of view In the Bahnese world of language, speech registers are yet another index of the culture's particularly elaborated view of personhood defined in both social and metaphysical, political and metalinguistic terms. uations of informal intimacy in which language distinctions would likely be muted 61 The majority of alus madia forms, it should be noted, belong to the categories of vocabulary that are strongest according to the "hierarchy of force" principle These terms represent additional subdivisions, colorations, and gradations of the register range, by using alus madia forms, a speaker indicates polite intentions even though most of the morphemes in the utterance are in the kasar/biasa range

81

3 LITERATE T R A D I T I O N S AND LITERARY ACTS

The word reduced to an inscribed surface, silenced, then resurrected, has a potential, new fecundities —Walter J Ong, Interfaces of the Word T H E PREVIOUS chapter has concentrated on certain cultural aspects of language in relation to both the average Balinese and the more specialized person adept in the manuscript tradition An important symbolic dimension evoked by the nature of language as script was noted in the discussion on mystic interpretations of writing Yet in Chapter 1, it was pointed out that manuscripts coming under the heading of "verbal art" are regularly performed orally and rarely read silently by individuals ' Traditional literature in Bah, then, presents us with a situation in which texts in the form of written manuscripts are nevertheless experienced as vocal sound The particular conjoining of noetic features, oral and written, that shape the textual tradition are the subject of the present chapter As noted earlier, literary texts have existed in Bah in palm-leaf manuscripts since at least, and perhaps even before, the period of the earliest historically dated ties with Java in the early eleventh century The remarkable preservation of thousands of both metrical and non-metrical texts through repeated copying has been frequently cited Less clear has been the role that Balinese reading and studying have played in the history of verbal art, that is, in "classical" or "belles-lettres" genres 2 Yet a manuscript tradition that depends for its existence on regular 1 An early researcher of Balinese Sanskrit, Sylvain Levi(1933 xvi|, found that when he asked Brahmana priests to show him their liturgies, the padandas could not help but "sing-song" the texts as they were going over them 2 For example, Pigeaud (1967-70, 1980), in his catalogue of extant manuscripts in major collections, almost ignores the role of Balinese traditions of creation and performance in Javanese-idiom literature Even after a fairly early article by Bhadra (1937) dealing with the practice of literary performance in Bah, scholars generally have not seemed to take seriously this aspect of textuahty in the archipelago

82

LITERATE TRADITIONS, LITERARY ACTS

and painstaking copying is necessarily equally dependent on thorough knowledge and sustained interest on the part of its literati. A text that is unused or unstudied is less likely to receive scribal attention; an unread text in a non-colloquial code later becomes an unreadable one, and then ceases to exist in manuscript form as lontar leaves fall victim to weathering and insects.3 Thus, in the case of Old and Middle Javanese works particularly, we need to ask, How was it that a large body of archaic language texts remained readable for centuries m Bah? In the absence of dictionaries, manuals, grammars, and critical editions, how did the Bahnese (and, up to a certain point, the Javanese| preserve and transmit the techniques and knowledge necessary for understanding a complex prosodic endeavor such as the kakawin? The answers about the cultural context of classical lontar literature in Bali are beginning to appear. In the work of Walhs (1980) we see how certain genres comprise distinct sets of interdependent formal, contextual, and acoustic features. To analyze the noetic components of literature in Bah, we need to look at the activity of group reading and studying, called pepaosan in Bahnese. This is certainly true in the case of, say, paiwa (so-called "prose") works, ancient forms par excellence that have continued to influence the structure and content of various sorts of texts up to the present and have had an important role in the evolution of theater traditions. At the same time, moving toward an even more general understanding of the traditional noetic economy of Bali, we need to examine the nature of lontar manuscripts as shapers and bearers of information. Our investigation, then, concerns both the formal and material means of language preservation, as well as the rhetorical ends that inevitably give meaning to verbal art. THE LONTAR AS A LITERARY M E D I U M

The inscribed surface of the Boiassus flabelhfoimis palm is the standard form of manuscript in Bah. Detailed descriptions of the lontar-making process are found in Zoetmulder (1974:37-38, 127-128) and Ginarsa (1975), while the wide variety in the content of what has traditionally been recorded in lontar form is largely (though, in the case of Bah, not completely) covered by Pigeaud (1967-1970, 1980). From a noetic point of view, the presence of lontars in Bah signals 3

Some Bahnese readers of lontars maintain that the more a manuscript is read, the longer it will last, and frequent handling does seem to give a certain patina andflexibilityto the leaves, forestalling the drying and cracking of deterioration 83

THE SHAPE OF THE WORD IN BALI

the absence of the kind of "radically oral" management of information such as that indicated by the ancient Homeric and Vedic traditions, in which patterned oral repetition was the only method available for perpetuating knowledge In a culture that knows writing, the word has status as an object, and its material, fixed, non-sound qualities are one source of its significance With any sort of "technologically-processed speech" (Ong 1977 42) is born "literature," which can stand as a separate "voice" apart from any specific person's vocalized participation Even when all members of the culture do not control the scribal medium, or are simply illiterate, the notion of writing is still crucial Anderson (1972 47) has commented on the sociopolitical value of writing in classical Javanese culture The literacy of the ruling class was a symbol of Power largely because it presupposed the ability to make the qualitative leap out of illiteracy Their power derived not from their ability to disseminate new concepts through society, but from their ability to penetrate to and conserve old and secret knowledge The notion that manuscripts embody old truths and ancient learning is common to scribal cultures generally and to Balinese lontar tradition specifically One reason for this is that, with lontars as with other sorts of manuscripts, the linguistic code tends to be of the archaic, "learned language" variety, in Bah, this would be Old Javanese or "Kawi," with all the diachronic and stylistic variety that term implies As characterized by Ong (1977 27-29, 1982 112-114), learned languages developed throughout the world in manuscript-using cultures before the appearance of print, and they have a number of features in common, including the following 1 2 3 4

They are never learned as mother tongues They are always learned with the help of writing They are used primarily by men They tend to disappear or become mother tongues as print technology and vernacular literature become dominant

Ong's generalizations, largely valid, must be clarified in relation to specific cultures' uses of learnod languages, for instance, the orally transmitted metalanguage of Pamni's Sanskrit stands in special relation to written Sanskrit in India For the Balinese, the term "Kawi" can evoke associations of specialized male-dominated learning that is acquired by means of the written word, a situation that is typical for learnod languages This is true of Kawi because it is no one's mother 84

LITERATE TRADITIONS, LITERARY ACTS

tongue in Bali, because it is often learned with the aid of writing, because it is used within specialized realms such as esoteric philosophy or the dramatic arts, and because those individuals who through predilection or necessity come to use Kawi are, by and large, men Yet to many Bahnese, "Kawi" is a notion linked as much to a temporally distant, semi-divine, epic world as it is to a specific linguistic code in which ideas about that world are expressed. Although the preservation and vitality of Kawi literary modes depends on lontars, it is not always, and not even primarily, through direct contact with lontars that most Bahnese become familiar with the "Kawi ethos" of, say, the great war of the Bharatas. These ideas are conveyed in certain social, ritual, and theatrical contexts through specific patterns of participant behavior and vocal sound. Thus it is true that for most Bahnese the lontar represents not so much a direct, historical link to the past as an empowering force that enables the re-creation of distant worlds in the present. It is this attribute of continual recurrence of the "past," linked to basic cultural beliefs about cyclic time,4 that contributes to the continued vitality of the Kawi tradition side by side with vernacular spoken and written modes Kawi, both as a language and as a realm of experience, is not disappearing from the Bahnese scene, as the learn6d language theory might predict. Even in the face of the media revolution, which is making printed books, radio, television, and film common experiences, there are specific times when and contexts in which Kawi verbal and behavioral "styles" are necessary and appropriate. The discipline of philology was a science born after typography had transformed European information management by making possible fixed-form texts and frozen arrangements of language in space.5 Attitudes based on experience with books often underlie scholarly works on Bahnese lontars—works in which noetic issues are not considered Many aspects of lontars as manuscripts clearly show that it is misleading to equate their form and function with those of printed books For instance, printing makes possible multiple, identical copies of one work, whereas two manuscript copies are never quite the same. The human intervention that accompanies the copying of even one lontar is extensive, the size of the leaf, the style and size of the writing, the 4 For an interpretation of Bahnese notions of time and how these interact with ideals of individual and social behavior, see Geertz (1973) The related Javanese concept of cyclic time, and its important connection with traditional music, is treated in J Becker (1979) 5 See Eisenstein (1979 200), who links the emergence of a "rationalized" view of antiquity to the rise of print

85

THE SHAPE OF THE WORD IN BALI

spelling and punctuation conventions followed, the amount of text per leaf, the colophons and prologues—all these elements can vary, even though Bahnese scribes have for centuries been diligent and scrupulous in the perpetuation of a fairly standard manuscript format. Furthermore, each new copy of a lontar is rather like a new "edition" from the point of view of the copyist, who may make alterations, substitutions, and deletions as his own knowledge and experience dictate. Bahnese scholars will sometimes look over a number of manuscripts of a work and engage in "horizontal borrowing" in the creation of a new copy (Robson 1972:311); this borrowing, sometimes called "contamination" by critical editors, can affect all levels of the text from spelling and word grouping to large sections of content. In typographic tradition, books are standardized containers of knowledge; the verbal entities within them must be identical between copies and capable of being indexed, cited, and referenced when quoted. In the case of Bahnese lontars, elements can be reshaped, excerpted, expanded, condensed, and otherwise altered more or less freely. In many cases, a lontar seems less a replica of one work in its entirety and more a collection or compendium of one person's knowledge and interests. Several works, such as a number of esoteric teachings (tutur), may be concatenated in one manuscript. Among the large group of metalinguistic manuscripts, the lontar An Canda (which can be glossed as 'Rules of Metrics') contains at least two discourses on esoteric topics related to the rules of syllabic quantification (the canda of guiu-laghu). Similarly, the lontar called Guiu-laghu6 is a compilation of teachings, some of which are also found in lontars with different titles, such as the Kandan Sastia 'Elements of Writings' and the Aji Canda. Another way in which lontars differ from books as knowledge-conveying media concerns titles. Books, as we know, have titles that assume their contents; the title of a book is supposed to imply the content or parts of that book, and thus it is logically "prior" to the book itself. With lontars, however, titles are often taken from the first words or opening sentences of the manuscript, earning their place as titles by virtue of "temporal" rather than logical priority.7 In didactic texts, for example, the first word following the opening formula is often a term meaning 'this', such as nihan (Old Javanese), lti (Sanskrit), lki (Javanese), or puniki (Javanese, Bahnese). The words following the formula are taken as the title of the text, as in In tegesmg waiah 'This is the 6 Guru refers to the metrically "long" syllables of Old-Javanese kakawin prosody, while laghu refers to the "short" syllables 7 For a more detailed explanation of this distinction, see Burke (1970 25-26)

86

LITERATE TRADITIONS, LITERARY ACTS

explanation of speech'. While this method of titling permits reference to particular manuscripts, it does not provide that two manuscripts with the same title have exactly the same contents. This is not a problem for the Bahnese, however, who are not unnerved by great variety in form and content between similar written works. If, as the diagrams of philologists suggest, works are tree-like, producing many branches from the original trunk/form, the Bahnese are content with the profusion of foliage that results, and they do not feel the need to cut away the growth concealing the shared base and ancestry. While the preceding discussion considers features of lontars that might seem rather obvious, it is necessary to go over this ground if our notions about how the material substance of Bahnese knowledge is stored are to be relatively free of print-conditioned prejudice. There is another important point, however, that suggests just how different lontars are from books. Works of verbal art have voices that m u s t be sounded and heard, and the text represented by the manuscript is only one element of the text as experienced

PEPAOSAN: THE LITERARY EVENT

In the history of Western scholarship in the Javanese-Bahnese literary heritage, it is only recently that the performative context of literature in Bah has emerged as a necessary and relevant feature that requires investigation. Although a small amount of published commentary on pepaosan or mabasan8 does exist (Bhadra 1937), the mtro8

The most general term for the activity of group reading and studying is pepaosan (a partially reduplicated form of the root paos 'read' with nominalizing suffix -an, or, alternately, a nominal derivation with pa- and -an affixes, meaning 'a reading of something') The term refers to the group study and performance of literature Mabebasan (a reduplicated form of the root basa 'language', with suffix -an) and mabasan are sometimes used synonymously with pepaosan, but these terms actually focus on the process of 'giving meaning through translation or paraphrase', which is the central aim of the literary gathering Within the reading group (called seka pepaosan or seka bebasan, and in some areas prasantian), two major "expert" roles are distinguished that of reader and that of translator The former is the JUIU tnaca (stative verbal form of paca 'read'), in more honorific language, this is sang ngwacon 'the one reading' [sang is an honorific definite article, while ngwacin is an active verbal form of the root waca, an honorific form of 'to read', plus m object focus suffix) The translator, /uni mabasan, is also called juru teges or, in honorific style, sang ne gesm [teges translating as 'meaning', 'order', with a secondary sense 'to cut') The translation function is also referred to simply as masin 'to "language" something', from the active form of basa plus m, or melutm from pelut 'to 87

THE SHAPE OF THE WORD IN BALI

auctions to critical editions only intermittently and with caution mention that kakawin prosody perhaps is linked to the ongoing oral performance of kakawin in Bah (Teeuw et al 1969 7, Worsley 1972 91) While most aspects of the pepaosan process have been scarcely discussed, Walhs (1980) has analyzed the vocal musical structure of makakawin or mawirama/ngwirama, the singing of kakawin poetry in wirama meters (based on the Indian kavya prosody called vaina vitta) From his work, we learn of the interlocked textual and musical features of the kakawin, with the translator dependent upon certain musical cues for correct verbal interpretation, and the singer needing information from the written text in order to give the appropriate musical rendering (Walhs 1980 140-155) Setting The setting of pepaosan activity reveals the cultural value of group literary enjoyment in Bah Basically, pepaosan occur (1) in the context of rites of passage and temple ceremonies, (2) in semi-regular meetings of a group at the home of one of its members, or (3) in formally arranged "invitational" meetings of literati from different villages or regions (Overlapping of functions between the first and third of these settings is also common ) Pepaosan of the first type are part of festival "atmosphere" and provide aesthetic enrichment on the occasion of weddings, cremations, temple odalan (anniversaries), and so forth As part of the larger ceremonial event, the pepaosan is just one of a number of other verbal and musical activities, and it may receive only sporadic attention from persons not directly involved in the reading group When a rite of passage is held in a home, the "host" family head may sit in on the pepaosan, temporarily in a state of ntually endowed "closeness" to the divine world, a participant in all ceremonially significant moments Although the pepaosan is not absolutely required, it is highly valued as a mark of the ceremony's importance and association with traditional values The texts that are read on these occasions vary widely, and the specific selection depends on group members' expertise, one text may be chosen as particularly suited to the occasion, or parts of several different works may be read peel' These terms convey the image of giving verbal form to, and uncovering the sense of, texts whose performance is a musical art requiring elaboration and interpretation 88

LITERATE TRADITIONS, LITERARY ACTS

The second type of pepaosan involves a fairly stable group of serious students, often under the guidance of a respected literary expert The members of the group meet, usually at night, to read and develop their skills. These situations are less formal than ceremonial settings, and texts are read with particular emphasis on pedagogical considerations. The correct musical phrasing of a line from a kakawin may need to be repeated several times in such a study group, whereas in a ceremonial context small errors in singing technique most often go unchallenged. The acknowledged guru, or teacher, of the study-group serves as the model to be emulated both in vocal style and in knowledge of the literary language. In contrast to more informal study group or ceremonial pepaosan, the third type, "invitational" group readings, involves primarily only acknowledged experts, and sometimes it has a subtly competitive atmosphere. This variety of pepaosan can have a mixture of religious and scholarly aims, for example, near the time of Odalan Saraswati, Brahmana residences of the Buda sect may host a reading of sections from the Sutasoma, a kakawin that elaborates the Buddhist philosophical perspective of fourteenth-century Java. At two such occasions that I attended literary authorities from a number of villages in the district were present, and any participant who wished to sing or translate had to have a degree of expertise Furthermore, the interpretations of the text that day dealt exclusively with philosophical and mystical points and were intended to enhance the spirituality of the occasion. Issues of esoteric meaning were treated at length, this was in contrast to another occasion, a local temple odalan, at which the same experts performed the Sutasoma and emphasized the adventure and drama of the hero's confrontation with man-eatmg beasts before a mixed audience of local residents. This third, "invitational" type of pepaosan seems traditionally to have been very popular,9 and it is still common today Local and provincial government authorities, for instance, may sponsor pepaosan competitions and "festivals."10 Such readings are often held on auspi9 See Bhadra's account (1937) of pepaosan competitions in northern Ball Another occasion calling for the reading of a specific text is the night before the new moon of the seventh month {Magha, sasih kapitu), called Siwaratn, when the kakawin Lubdhaka is the object of study and the god Siwa the ob)ect of devotion (see Teeuw et al 1969) 10 Recently there have been efforts to encourage pepaosan skills, particularly among younger Bahnese, with government subsidies available for establishing literary study groups

89

THE SHAPE OF THE WORD IN BALI

Literary reading [pepaosan). Older and younger generations use both lontars and mimeographed texts during a village temple celebration (note offerings on the table).

cious days (such as the full moon, purnama), and representatives from different areas may read far into the night, comparing each other's vocal techniques, translations and interpretations. Although differences of opinion tend to be tolerantly muted in the interest of group harmony and cooperation, there are occasions when conflict in interpretation arises and literary prestige is won and lost in the debate. The process involved with convening any type of literary reading event is fairly standard. Generally, participants are seated cross-legged in roughly circular fashion in an open pavilion or bali. The persons who will begin the reading and translating may be somewhat nearer the center of the group; the translator is often turned slightly, so as to face the reader. In front of the reader is a round, raised tray [dulang) such as is used for arrangements of offerings; on this tray are the lontars to be performed. There are usually small offerings and burning incense next to the lontars; reading group participants are provided at intervals with betel, cigarettes, coffee, and snacks. When the text performance is to begin, the first member to read picks up the lontar and, having decided where in the text he is to start, 90

LITERATE T R A D I T I O N S , LITERARY ACTS

sings the first line or stanza by way of "warm-up." He then repeats the initial line, and this time the translator follows with his Bahnese-language rendering. These two continue to perform, often pausing for questions and commentary from other members of the group. After a period of time, which may be anywhere from fifteen minutes to one hour, the reader stops, and (perhaps after an "intermission") another reader and translator take their places near the lontars. The roles of reader and translator do not have to switch at the same point; individual participants decide how much they are willing and able to perform. During performance sessions that continue for many hours, there are lengthy breaks for meals, rest, and informal conversation. Structure In considering the different settings in which the pepaosan can occur, we must not lose sight of the fundamental features that define a pepaosan as a literary event and are present in all of the settings described above. A primary structural principle, for example, is the alternation between the reading (i.e., singing) of a written text and the translation/glossing/paraphrase that follows. No matter what the prosodic requirements of the text at hand, a segment of verse or prose is always succeeded by its paraphrase, at which point discussion of the text's meaning is optional. It often happens that the end of a singer's phrase will be picked up by the translator, who then elides directly into his paraphase, an unbroken stream of vocal sound thus reinforces the notion of unity of written text and oral translation. Only in the more pedagogical setting of the pepaosan, where practicing correct vocal technique is the central aim, will any verbalization intervene between the maca 'reading' and mabasan 'translating' components of this basic textual unit. The fundamental "distinctive unit" of pepaosan, then, is made up of two integrated yet different and complex activities, one a highly constrained musical performance of a fixed written text, the other an extemporaneous rendering of the text's meaning in stylized vernacular.11 Thus, for the Bahnese, "literature" implies group activity to the extent that a minimum of two persons is engaged in mutually dependent ver11 This pattern holds true even for texts written in Balinese vernacular, such as geguritan poems Tape recordings of geguritan readings, with the paraphrase component, are popular items played for entertainment at home or to add to the "busy" acoustic texture of a temple festival It is possible to rephrase Balinese poetry in Balinese, using the style of elaborate, formal expression called utama nmg basa 'best of speech' favored by pepaosan translators

91

THE SHAPE OF THE WORD IN BALI

bal interchange. Although this conclusion seems unstarthng at first, it masks an important principle. The relation between the individual and literature in Bah is not that of the one-to-one, passive, visual consumption of a fixed, silent text such as has become usual (and even obsolete) in some Western societies. Rather, the Balmese experience of literature is multi-sensory, participatory, communal, and socially integrative in nature. For the participant in pepaosan, literature is more a process than an object. The literary performer experiences the interweaving of reading, sounding, listening, speaking, and contemplation, while for other persons present there is an immediacy of both musical and verbal meanings that is close to the model of preliterate oral-aural synthesis For all, the enjoyment of verbal art does not suppose a cleavage between sight and sound, the learn6d language, rather than cutting people off from the spoken word through the distance imposed by writing, promotes the unfolding of an aesthetic experience that goes beyond the formal bounds of the medium.12 Strictly speaking, there is generally little obvious distinction between performing and nonperforming pepaosan participants. All are seated, in the same area, according to a socially hierarchical allocation of space Performers themselves become audience as reading and translating roles are rotated. Anyone who wishes, moreover, can be an aesthetically participating auditor and enthusiast The oral-tradition aspects of the pepaosan are underscored by the fact that audiences need not be literate in order to be familiar with the sounds and meanings of texts.13 Another important facet of pepaosan is also linked to process and to the fluidity of a literary tradition in which so many "voices," of both the past and the present, can be heard Just as there is no single individual who alone spins out the poetic threads, so also there is no single, permanent source of interpretation for texts that are a millennium old. The information needed to paraphase Old Javanese literature appropriately is available primarily in the dynamic of pepaosan itself, for traditional Bah made only limited use of dictionaries, grammars, or works of literary criticism.14 Instead, the process of oral performance 12

"Sound situated man in the middle of actuality and in simultaneity, whereas vision situates man in front of things and in sequentiahty" (Ong 1967 128) 13 This point is made by Sweeney (1980) in regard to traditional Malay literature, its production and consumption 14 There do exist lontars in the form of glossaries, which are ordered according to universal hierarchical categories such as "gods," "natural elements," "ani91

LITERATE TRADITIONS, LITERARY ACTS

presents interdependent musical and verbal keys to meaning. The component of paraphrase, meanwhile, provides for an ever-changing translation encompassing both the formulaic and the spontaneous poles in language. With sounding comes the beginning of textual interpretation, and through paraphase the text is renewed each time it is performed. In this atmosphere of shared experience of both fixed and fluid forms of literary texts, pepaosan experts with the most age and experience are naturally looked upon as guides and teachers Such individuals, while they may also dictate norms of vocal style and technique, most often turn their attention to translation rather than singing The most eminent literati may eventually cease to perform, reserving their authority for critiques of the translations—critiques they deliver in judicious and eloquent commentary drawing on the full range of their literary experience While it is possible (though not highly valued) for an individual to become a translator largely through memorization, the skill required for commentary and debate is dependent on analytic thought and knowledge of many texts. Even today, prestige in pepaosan circles is gained in the traditional manner of long years of imitation, observation, and experience in the actual event, a university degree in Old Javanese does not necessarily lend an individual's comments more weight. Yet because texts can ultimately be perpetuated only through reading and translating, the pepaosan cannot simply lapse into polished demonstrations given exclusively by experts. Students and beginners have their place, especially in the role of juru maca 'reader'. The commentary that follows paraphrase is an opportunity for investigation, questioning, and exploration, when all voices are willingly heard. The pepaosan contains a learning process reminiscent of the "oral rhetorical" format, in which knowledge is transmitted and skills developed mals," and so forth These texts, often called Dasanama, cannot of course be used for quick dictionary-style reference Other Sansknt-Bahnese or SanskritOld Javanese glossaries traditionally used were perhaps composed from memory Chandra (1978) describes a Bahnese lontar called Svarasamhita, an interlinear Sanskrit and Kawi "vocabulary," that actually is a word-for-word paraphrasing of the Sanskrit kavya called Janakihaiana There clearly are other works dealing with metrics, lexicography, grammar, and poetics We need a full analysis of the Bahnese texts that have been used for the study of language and literature, and an explication of their relationship to the classical Sanskrit literary arts, in order to complete our understanding of this branch of Bahnese learning

93

THE SHAPE OF THE WORD IN BALI

through debate and dialectic, argument and refutation. In marked contrast to European classical rhetoric, however, Bahnese cultural stress on harmony and consensus overrides the "agonistic" polemic of debating "opponents." The statement of opposing views is muted through indirectness and acceptance; sapunika dados taler 'that is also acceptable' is murmured with nods of assent even when the opinions expressed are apparently diametrically opposed.15 If the pepaosan were set entirely in the code of classical Javano-Bahnese literature, however, it would seem unlikely that such vigorous study of archaic language should have persisted A key point here is that the "distinctive units" of pepaosan are based not only on two vocal functions—the singing-reading of the fixed text and the translation that is not fixed—but also in two linguistic codes. The use of vernacular Bahnese, albeit mostly a formal and literary style, causes the learned language to be recast in a verbal medium nearer the immediate experience of the participants. Through the pepaosan, the language of lontars, "old" both temporally and because it is being repeated, is phrase by phrase and verse by verse woven with "new" (because it is extemporaneously expressed) translation. This linguistic multiplicity reflects a larger cultural pattern that values the maintenance of widely varied types of vocal expression. Another way in which the pepaosan tradition echoes general cultural patterns is in the careful attention paid to the hidden, mystic import underlying language both as speech and as writing. Lontars are handled with respect, certain days are considered auspicious for literary readings and others are not. Persons in a state of ritual pollution [sebel) because of a recent birth or death in the family may avoid handling manuscripts. In a variety of formal and contextual ways, then, the pepaosan as an event and a process defines "literature" as the Bahnese have known it. Neither the linguistic and material media of pepaosan nor its rhetorical structure much resemble "literature" as experienced in cultures where typographic and electronic organization of knowledge is dominant. The preceding discussion has pointed out the indivisibility of the basic structural unit, the reading of a text plus its translation. Yet the reading component itself is an indivisible combination of a written message (the manuscript] and its oral realization (the singing or reci15 Bhadra (1937 8) also mentions this point It is even permissible to consult a scholarly edition of a kakawin text as a source of interpretation, but it is not considered a "higher" authority even though it is a printed book and the work of foreigners'

94

LITERATE T R A D I T I O N S , LITERARY ACTS

tation). Thus the pepaosan is an integrated hierarchy of elements having to do with writing and reading as well as memory and oral composition. Throughout this hierarchy is felt the dialectical resonance of musicahty and textuahty, archaic language and vernacular, and both fixed and changing forms. In this section we have looked at the settings of pepaosan activity, and also at the central features that define this particular Balmese "literary event": 1. An invariant structure whose basic unit is composed of two elements· the sounding and interpretation of one segment of written text. 2. Group participation and sharing of knowledge and experience; communal, multi-sensory enjoyment of verbal art. 3 Literature as a process rather than an object, with interpretation available only from the ongoing dynamic of the event. 4. Prestige gained through experience and age, often with the oldest figures as commentators. 5. Openness to and tolerance of inquiry and commentary. 6. Linguistic multiplicity. 7. Attention to extra-textual values of speech and writing. 8. Interweaving of a variety of noetic techniques for storing and transmitting knowledge. Together, all these aspects of the literary event of pepaosan further characterize the nature of "the word" within Bahnese culture. Along with the values of written and spoken language discussed in Chapter 2, they form a background against which specific verbal elements of literary discourse must be viewed. The linguistic, material, and rhetorical aspects of pepaosan constitute a body of constraints on the shaping of texts. Neither the traditional philologist's perspective, influenced by the nature of texts when typographically fixed, nor the folklonst's tendency to oppose "oral" and "literate" traditions, much avail us here. The verbal art of Bah, and especially as we have viewed it here in terms of the pepaosan, intertwines elements from the realms of oral-aural patterned speech and manuscript literacy. In the next sections, we will examine more closely some features of Bahnese verbal art that reveal and highlight this unique "noetic economy." ORAL-AURAL CONSTRAINTS ON TEXTUAL FORM In Chapter 2 of this study, it was stated that language, according to Bahnese traditions of metalinguistic commentary, is a representation 95

THE SHAPE OF THE WORD IN BALI

of the congruence between the structure of the inner self and external reality (micro- and macrocosms). As elucidated in esoteric texts dealing with this subject, mystic speculation and language converge in lessons on controlling the breath [prana] and on concentrating upon the elements of the syllabary (divided into "small" and "great" aspirants, alpaprana and mahaprana) within one's body The culmination of the tantnc-type meditation, referred to in treatises that are called tutur and tatwa,16 is the transmitting of power, often expressed in terms of heat and fire [gem], upward through the body until it reaches the area of the forehead where the power syllable ongkara is written and then passing out through the ends of the hair {tungtungmg rambut) or the fontanelle [wunwunan] Sacred syllables (modig) uttered by the priest, bahan, or other practitioner are also a vehicle of power, and they can affect the outer world because they are a distillation of the "written language within the body" [sastra ning sarira), a literal exhalation or "ex-pression" (forcing out) of inner power toward some external goal, which may be either positive, "right" [tengen) or negative, "left" [kiwa] in its result Uttering, then, can indeed be direct action Our appreciation of this view of the power of language would fall short, however, if we thought of it merely as superstition or sophisticated dabbling in sorcery. For the Bahnese use of language as the connector between micro- and macrocosmos, or inner and outer worlds, also reveals an orientation toward the word as ongoing activity and action in the present. This orientation may be fundamental to primarily oral-aural noetic processes. Cultures which do not reduce words to space but know them only as oral-aural phenomena, in actuality or in the imagination, naturally regard words as more powerful than do literate cultures.. . Words in an oral-aural culture are inseparable from action for they are always sounds. (Ong 1967. 112-113)17 There are several ways in which beliefs and processes grounded in a basically oral appreciation of the word influence the form and function of written texts in Bah. The beginnings of texts are characterized by dense textual boundaries, for example, that acknowledge the power of 16

Tutur is an Old Javanese word meaning 'thought, memory, account, teaching', tatwa is a Sanskrit borrowing and means 'truth, reality, essence' 17 Ong's notion here of the "power" of sound for oral-aural cultures is different from concepts of the sociopolitical power associated with literacy 96

LITERATE TRADITIONS, LITERARY ACTS

uttered language as well as the significance of literary beginnings in outlining the dimensions of the particular "speech" act. These bound­ aries, seen in their full form only in lontar manuscripts, may have three components, each with its own role in the discourse. The first component, on the verso of the first leaf of a lontar (the recto is left bank to prevent wear from exposure), are the initial written signs, which invariably consist of one or more non-morphemic config­ urations of letters (aksara). Two common figures used in lontar man­ uscripts (there are many variations) are the papadan-windu-papadan and the panti-windu-panti, as illustrated in Figure 4. Although these aksara configurations are not sounded during reading, they are not merely decorative for the Bahnese, who classify them as swalahta (ab­ breviations that may have mystic import) or as modro (power sylla­ bles). They also are used to mark the ends of stanzas and changes in meter in kakawin literature. Although several different interpreta­ tions of the letter shapes are available, and there might be disagree­ ment on just how a papadan should be written, it is clear that the open­ ing aksara of the lontar enhance the text as well as clearly signal its beginning (Sugriwa 1978:15). They suggest a certain role of language in meditation, and the force of magic utterance, as they direct attention toward the opening sounds of the text proper.18 And although they are not reproduced in the transliteration of printed editions, they are an in­ tegral part of the text in its lontar form. 18

According to some literati, the papadan actually conceals the symbols for a, u, and ma, the three components of the ultimate sound O N G

m\ ο b papadan c panti-windu-panti

a papadan-windu-papadan

d papadan FIGURE 4

Types of initial aksaia in Bahnese manuscripts

97

THE SHAPE OF THE WORD IN BALI

T h e second component of the textual boundary is a word, phrase, or group of phrases that serves as an invocation. The phrase most com­ monly used in this way begins with the modre O N G and proceeds: ONG awighnam

astu nama or or

siddam siwa ya swaha ya

The spelling of the mantra varies widely, and often just the first three words [ONG awighnam astu) are used This originally Sanskrit for­ m u l a can be glossed 'May there be no hindrance in the realization', and it is found not only at the beginnings of works of verbal art but also in almost every other sort of manuscript as well: didactic, esoteric, tech­ nical, genealogical, and so forth. This brief mantra seems to have as its purpose t h e invoking of a divine source that can somehow foster the successful writing or use of the text. T h e most fully elaborated expression of the literary invocation is found in the kakawin, whose introductory stanzas (manggala) are di­ rected toward a royal or divine source of patronage and aesthetic inspi­ ration. 1 9 In Zoetmulder's (1974:173) words, " a manggala is anything, any word, act or person, which by its salutary power is able to assure the success of the work that is ahout to be undertaken." In the mang­ gala of the kakawin, we see the elaboration of a tradition of yogathrough-hterary-creation, a tradition that had as its supreme goal the union of t h e poetic imagination with the d i v i n e 2 0 The achievement of such union with a god [dewaSraya] is a momentary pinnacle of aes­ thetic rapture in which transcendent awareness unfolds in selfless contemplation of surpassing beauty. 2 1 In Old Javanese, the term used for both the sensation and its cause—beauty—is the same: lango. The power of the kakawin to stimulate the desired experience of profound aesthetic rapture [kalangwan), at least for present-day Balmese readers 19 There is a difference of opinion regarding the sense of the term manggala, Robson (1983 310) argues that it refers to the source of "the power to bring the work to a successful conclusion," and not to the introductory stanzas of kaka­ win 20 See the manggalas of the kakawins Hanwangia (Zoetmulder 1974 473, 162) and Nagaiakitagama (Pigeaud 1960-63, vol 1, ρ 3, vol 3, ρ 3) for exam­ ples of this poetic figure 21 See Zoetmulder (1974 172-185) for a comprehensive treatment of what he terms the "rehgio poetae " As he explains it, "This union is of a transitory na­ ture, lasting no longer than the brief moment of ecstatic rapture experienced in surrendering oneself to the overwhelming power of the aesthetic experience" (P 185)

98

LITERATE T R A D I T I O N S , LITERARY ACTS

and perhaps for the creators and audiences of kakawin in ancient times as well, is through its activation as vocal sound. The godhead imma­ nent in the work of verbal art (and the kawi 'poets' often liken their works to candi, temples which enshrine deities) becomes perceptible through the temporal unfolding of the poem as performed. The transi­ tory character of kalangwan is analogous to the experience of uttered language as it flowers into vocal sound and fades away again into si­ lence. The manggala thus encloses a powerful linguistic icon for mys­ tic awareness, based on the momentary and experiential nature of lan­ guage seen from an oral-aural standpoint. The third component of the textual boundary is the apology, or as the Balinese call it, pangaksama (from Sanskrit ksam 'to forgive, be pa­ tient'). Although it occurs generally only in poetic works and may form part of the ending rather than the beginning of a text, the apology is an important formal and rhetorical device in many works of verbal art. In the kakawin Lubdhaka, it is interwoven with the manggala and in­ cludes a typical expression of the poet's self-deprecation. But I myself am far from being endowed with the talent of com­ posing sweet-sounding words in verse, For how could I achieve all that a poet longs for, exceedingly dull as I am? Thus indeed I cannot but incur the utmost derision of others, My one hope, however, is that this may succeed in being an aid in my search for the Absolute. (Teeuw et al. 1969:69, translation of canto 1, stanza 1) As is clear from this example, a pangaksama can be an apology for the speaker or writer's own shortcomings, it may also express a plea for tolerance on the part of the intended audience, as in the following, more recent composition in Balinese language and pupuh poetic form (indigenous meters influenced by Javanese macapat, the so called sekar aht meters). Pmunas titiangpiiengang, antuk ida άαηέ sami, mmakadi nng Ida Sang Kawiswaia miaga lewih, ampurayang titiang gusti, sampun banget ngangen kahyun, Udangang ugi matutang,

I respectfully ask to be heard by my honored audience all, and to the Great God Kawi-Voice exalted, forgive me, lords, for my great longings, and be truly pleased to approve 99

THE SHAPE OF THE WORD IN BALI

pupuh katah tuna luwih, these verses greatly lacking, pasang ιριιη, their written form saksat sami εαηέ lobah 21 altogether inconsistent. (Regeg 1975· 1, canto 1, stanza 3, my translation) The rhetorical significance of pangaksama-type devices is a tacitly understood convention in Bahnese society, and one that seems to be a feature of the politeness systems of other Southeast Asian languages as well. In Bah, formal occasions such as a visit to a high-caste household (gria or pun) or a meeting of the hamlet council (banjar) require special speech styles that include an apologetic phrase as a rhetorical device initiating a person's turn in speaking.23 Moreover, when lontars are published in romamzed versions in Bah today, special prose or verse pangaksama are often added in the introductions [pamahbah) of the printed works 2 4 These apologies are added even when the text to be published already includes an introductory pangaksama This contem­ porary trend resembles the practice, attributed to some lontar copyists of former times, of adding apologies to the texts of certain kakawin, such as canto 39 of the Lubdhaka (Teeuw et al. 1969" 144-145) and the two colophons of the Nagarakrtagama (Pigeaud 1960-1963). The pangaksama is generally important in textual and natural speech "beginnings," orienting the participants m the discourse to the special speech styles or literary aesthetic required by the situation Be­ cause a poet aspires to mystic transcendence, he must claim that he is not worthy of it; in contexts in which enhanced speech is most clearly demanded, the speaker must excuse his inarticulate coarseness. The literary convention of the "apologia" is thus a figure of oral rhetoric as well. This correspondence represents only one of the many areas of close relationship in Bah between written texts and the realm of spo­ ken language. Thus far we have considered three different aspects of the formal be­ ginnings of written texts that have been directly or indirectly lnflu22

Poems in sekar aht meters, called geguntan, are still frequently composed in Bali 23 1 refer to Bahnese expressions such as nunas lugra 'begging your leave' See Hobart (1975) for information on speech styles used in addressing the banjar Regarding traditional greetings using apologetic figures, Bagus (1979 168-169) remarks on a tendency among more modernized groups to replace these apolo­ gies (which are associated with caste-determined forms of deference) with a general greeting of Om Swastiastu 'May you go in peace' 24 See, for example, the texts Bhima Swaiga (Ananda Kusuma 1977) and Ge­ guntan Usaba Dangsil (Yasa 1973) 100

LITERATE TRADITIONS, LITERARY ACTS

enced by cultural values surrounding the oral or spoken word. The final point to be considered here concerns a Bahnese interpretation of the very first sound of the manggala section of the kakawin. Because this is the beginning of the initial poetic line, the sound chosen for this slot, it is felt, should be one that represents the essence [tatwa] of all language. The ongkara symbol (OM or ONG) is ideal because it is the sound of supreme, indivisible unity. Any syllable with the -a- sound (i.e., a consonant, wianjana, without any change in inherent vowel quality) is also auspicious, since -a- is the basic living "soul" of all aksara. And finally, an s- at the beginning of the first word stands for sandhi, the Sanskrit term for sound assimilation, which in Bahnese metaphysics refers to mystic union and the dissolving of opposites. This unrecorded interpretation, handed down among Bahnese literati and, it would seem, virtually unknown outside the island, can of course prove nothing about the intent of the poets who first wrote down the kakawin that we know today. It is interesting to note, however, that among the fifty-three different kakawin whose manggala and epilogues are included by Zoetmulder (1974:473-505), no fewer than forty-seven begin with either the ongkara, the -a- syllable, or the sconsonant. Turning from the "border" between silence and sound (or between ordinary language and the specialized aims of verbal art) which characterizes the beginnings of texts, we will now look at other structural patterns in written literature that show conditioning related to oral processes of handling knowledge. Certain features of the lontar, for example, seem to function as memory prompters, mnemonic aids, that help the reader to store and recall information. We have already seen (in Chapter 2) how categorization according to a compass-like model of the universe (the nawasanga) is an aid to the recall of many different types of information, and that classifications based on the circular nawasanga shape make their way into the tradition of metalinguistic analysis. It is clear from the many texts having to do with linguistic matters that Bahnese scholars have long been concerned about preserving a heritage of Javanese and Sanskrit learning for future generations. Thus, we find manuals on prosody (e.g., Aji Canda, Kandan Sastra, Guru-laghu), glossaries of Sanskrit vocabulary (e.g., Bhasa Ekalawya, Prayoga nmg Bhasa),15 and treatises including both summaries of stories and materials on language {Cantakapaiwa).16 Some of these works make 25 26

Cf note 14 above Ensink (1967) has discussed the contents of the Cantakaparwa

101

THE SHAPE OF THE WORD IN BALI

the recall of important information easier by grouping terms according to aurally perceived qualities such as alliteration and assonance For ex­ ample, the glossary in the Bhasa Ekalawya includes sets of three terms that differ minimally from each other according to vowel quality:

LARU, lkangpanas.

LARA, LARI, LARU LARA, that is 'sorrow'. LARI, that is 'to move forward'. LARU, that is 'heat'.

BHAYA, BHAYI, BHAYUBHAYA, lkang koweh. BHAYI, anwam. BHAYU, unp.

BHAYA, BHAYI, BHAYU: BHAYA, that is 'trouble'. BHAYI, 'youth'. BHAYU,'life force'.

LARA, LARI, LARU LARA, lkang papa LARI, lkang laku

In another text, the Swara Wianjana 'Vowels and Consonants', rules of syllabic quantification (difficult to remember since they represent a Sanskrit prosody arbitrarily applied to indigenous words) are illus­ trated and explained entirely in verse, using a widely known and easily learned poetic form (in this case pupuh smom).27 The Tutur Anacaraka gives philosophical meanings for each written consonant based on a sound correspondence with a mnemonic phrase Ka ngaian kamulanmg wong

'Ka' means the origin of people

Wa ngaian awaking sudi. La ngaian laksananmg kasiayan

'Wa' means the pure body . . . 'La' means the act of protection . .

.

Futhermore, the originally Javanese ordering of the syllabary (ha-na-cara-ka, etc ), often discussed in metalinguistic texts, has long been a basic mnemonic tool in both Java and Bali, with the syllables grouping into the words of a story from the legend of Αμ Saka, the king who is 28 said to have brought the original syllabary to Java Another way in which written information has been made more memorable is through the grouping under one heading of synonymous 27 This meter is particularly widely known in Bah, and its use fits with the poet's stated intention to teach his children some of the laws of poetics 28 Hana caraka, data sawala, maga batanga, pada jayanya This can be roughly glossed as 'There were once two servants,fightingwith each other, who both died, and thus were equally successful'

102

LITERATE T R A D I T I O N S , LITERARY ACTS

or near-synonymous words and phrases, whether these originate from Bahnese, Javanese, or Sanskrit languages. A text entitled Piayoga ning Bhasa 'Use of Language', for example, gives the following list of terms meaning 'mountain' mremuh, naka, waiadri, anala, adn, prawata, selapa, wukir, bukit, paibwata, GUNUNG The exposition of such "language groups" [seka ning basa) in metalinguistic texts often follows a hierarchical pattern, beginning with names of gods and cosmic elements, followed by kings, priests, demigods, ordinary humans, natural phenomena, animals, and so forth. Such accounts or classifications, also called dasanama 'ten names', are found in, among other lontars, the Cantakapaiwa and Basa Dwijodah 'The Language of Two Meetings'. The concept of dasanama or seka ning basa is an important Bahnese notion relating disparate bits of knowledge. The term seka in ordinary Bahnese means 'group, voluntary association', and it is widely applied to almost any collection of persons engaged in a single activity, whether it is coconut picking, pepaosan study, roof thatching, or gamelan playing. An individual may belong to several different seka at once, and his or her membership is ultimately a matter of voluntary choice rather than inherited duty (such as clan or temple membership} The linguistic idea of seka ning basa reveals a strategy for associating names in order to demonstrate shared characteristics. Each name is considered another aspect of the nature of the entity that is named. These collections of names, then, are actually outlines of the identity of the named object, culled from all available contexts. They closely resemble in many ways the "commonplace" tradition of the West, as described by Ong (196780-82, 1977 151, 188), which also brought cumulative collections of specific items under one heading The notion of the commonplace topic or saying is related both to written tradition and to oral formulaic expression. Commonplaces originated in order to help speakers manage all the various bits of information that society used in regard to a single topic. Our word "topic" comes from topoi or loci 'places'. Commonplaces, by extension, are commonly known items of information, gathered together under one topic, or place, in a written work The characteristic feature of Bali's commonplace tradition, it would seem, is its emphasis on names. Through the grouping of all the names associated with something, all attributes of that entity are recorded 103

THE SHAPE OF THE WORD IN BALI

and available. With their many lists of names and synonyms, however, the texts are still not asserting that all the terms collected under one heading are virtually identical. They emphasize, rather, the resemblances between the various contexts, codes, etymologies, and usages from which the names are gathered The commonplace tradition of Bah maps out in written form the connections between different moments in language-as-used. The way things "are" can be approached through how they are named 29 Among the features of traditional Bahnese written literature that show connections to the realm of oral expression, there is one formal principle that particularly stands out. This is the fact that word boundaries in lontar manuscripts are not indicated, written symbols of an entire prosodic unit (whether poetic line, stanza, phrase, or sentence) are concatenated with no divisions between individual words. In one sense, the scnptio contmua format is necessitated by both the writing system and certain phonological processes- the former determines the joining of, say, two consonants across a word boundary, while the latter dictates the vowel mergers of sandhi. But from a discourse perspective, it is the nature of the text as it is performed orally that allows the poetic line or sentence of prose to be conceived of as one continuous, unbroken stream of sound. The implication here is that the experience of verbal art, for readers as well as audiences, is based more on oral-aural than visual perception of language.'0 The psychological reality of the printed word with its white-space boundaries is absent from the experience of speaking and hearing. It is this condition, in which language is sounded and heard rather than silently read, that is paralleled in the lontar form without word divisions. The performer, juru maca, sounds a written line according to prosodic conventions that allow variation (e.g., to emphasize the meaning of a particular element) within a framework of identifiable constraints of pitch, tone, and rhythm (Walks 1980:144-149). The translator, juru mabasan, uses the sounded information to conceptualize smaller textual units, the morphemes and syntactic groups on which is based the Bahnese-language paraphrase. The working out of each paraphrase unit of text, in other words, requires the use of oral29

Epistemologically as well as etymologically, perhaps, "name" (Old Javanese aran, Bahnese adan) is close to "existence" (Bahnese ada) Goris (1971) also noted this correspondence from his study of the language of the Old Bahnese inscriptions 30 As Lord (1976 25) noted, "Man without writing thinks in terms of sound groups and not in words, and the two do not necessarily coincide "

104

LITERATE TRADITIONS, LITERARY ACTS

aural faculties to interpret a sounded entity that itself is an oral interpretation of written information. The complexities of the relationship between written tradition and orahty in Bah here become clearly perceptible. There are further implications of the written format of Balinese lontar literature. The prosodic units of poetry, such as lines, stanzas, and cantos, are signaled by punctuation or other notation in the written text. In prose works, we find caesura marking units of text that cohere semantically and also follow the natural rhythm of breath. Sometimes in prose texts, the unit between caesura tends to be octosyllabic. Yet linguistic units, for example, morphemes and paragraphs, that denote only semantic breaks are not graphically depicted in the written substance of the text. These emerge through the verbal art of the translator. He must produce an oral version of semantic content within the prosodic framework of the particular text—poetic line by poetic line, or phrase by phrase. Whereas the philologist tends to move word by word in establishing a "correct" version of an Old Javanese text, first breaking up the alphabetized transliteration into words, then working according to etymological and lexical criteria (note the morpheme-centered perspective), the Balinese performer is concerned with utterance-by-utterance coherence and continuity. The individual morphs or terms of the written text are not as central a concern as the musical, rhetorical, and semantic unity of the prosodic unit as a whole. The Balinese take the utterance as the starting point for an understanding of verbal art in the same way Ricoeur (1976 7) takes the sentence as the starting point for an understanding of discourse. "Only the sentence is actual as the very event of speaking . . . A sentence is a whole irreducible to the sum of its parts." We have examined some of the ways in which an oral-aural conceptualization of language shapes Balinese manuscript tradition. The specific ways in which the "inscribed" form of the text affects its oral rendering reveal still other aspects of the noetic patterns within Balinese literature. C H I R O G R A P H I C CONSTRAINTS ON ORAL PERFORMANCE

As we analyze the process whereby written texts are sounded in patterns of verbal art, it becomes clear that two separate yet interdependent roles must be considered: that of the reader-singer and that of the translator. For both of these participants in the pepaosan, the fact that 105

THE SHAPE OF THE WORD IN BALI

they are performing a written text, rather than working strictly from memory or orally composing, affects the oral-aural aspects of their performance. The singer working with the text in manuscript form has a generalized notion of the vocal style that accompanies each different metrical type. Between the boundaries of the prosodic unit signaled by punctuation, the singer applies a certain musical shape, or contour, according to remembered features of pitch, voice quality, rhythm, and so forth. In the case of kakawin performance (mawirama), musical realization is directly controlled by patterns of syllabic quantification [guru-laghu— see note 6, above). The syllable values, the guru-laghu pattern for each particular meter, are encoded by the writing system and reflect a borrowed Sanskrit prosody.31 Based on musical information indicated visually by the guru-laghu pattern, the singer can apply a generally similar vocal contour to each line of a given wirama meter, but both the duration of each contour and the breakdown of the poetic line into shorter word groups vary according to the text at hand. Depending on the metrical length of each poetic line (i.e., the number of syllables) as well as its content, the singer creates shorter phrases, each of which is given a Bahnese gloss by the translator. The singer of wirama poetry must to some extent understand the syntactic and semantic components of the poetic line—undivided in written form—in order to render appropriate musical phrasing, since the process of translation is more difficult if the written text is not delivered in cohesive segments. The conceptualization of the poetic line broken down into constituent phrases is common in the pepaosan tradition, and it is called guru basa (from guru 'teacher' and basa 'language')32 Yet even if the meaning of the kakawin line is unclear to the singer, he can usually follow the prosodic guidelines and produce an acceptable rendering; it is then up to the translator to clarify semantic issues through paraphrase. This individual applies the techniques of pasang kokoran (rules about fixed versus conditioned syllable quantity) and kiuna nyutia ('word merging', the rules of sandhi) to correctly interpret the frequent ambiguities that arise in the strictly controlled prosodic structure of the kakawin. 31

See Walhs( 1980 chap 5) for a detailed account of wirama vocal technique The notion of guru or 'governing principle' is applied to poetics in several ways. Besides guru basa 'phrasing' and guru-laghu 'syllable quantity', there is guru suaia 'final vowel [in the poetic line]' and guru wilangan kecap 'number of syllables [in the poetic line]' 32

106

LITERATE TRADITIONS, LITERARY ACTS

In another genre of orally performed literature—the body of "prose" works, or paiwa, often read in pepaosan groups—the situation is somewhat different. There is no codified notation that corresponds to gurulaghu, and that closely governs vocal realization. There is, however, a widely recognized style of vocal recitation used for reading parwa (just as there is a style for other genres), termed palawakya.33 While it is not within the scope of this study to present an exhaustive analysis of the poetics of palawakya vocal style, a brief discussion of certain features should give some notion of how the Balmese paraphrase of an Old Javanese text is an exercise in written tradition as well as oral composition. To call parwa works "prose," in fact, is only to acknowledge that there is in these texts an absence of a rigid metrical scheme, and that in romamzed printed editions parwa texts look like "regular" prose with clearly marked words, phrases, sentences, and paragraphs In Bahnese literary circles and in the theater, however, palawakya is a commonly identifiable reciting (as opposed to singing) technique characterizing any sounded non-metric text.34 Thus, we find in parwa performance vocal aspects—pitch, rhythm, tonal contour, dynamics— that form a distinctive oral poetics With specific regard to the written text, the prosodic structure of palawakya recitation presupposes the ability of the reader-singer to scan a unit of text, usually marked with a cank [ I —the same caesura used at the ends of poetic lines in kakawin texts), and to make judgments about syntactic structure as well as the role of that textual element in the larger discourse. Basically, parwa readers seem to identify for special vocal emphasis a focal point or "topic"35 within each textual unit. Clearly, much depends on the performer's knowledge of Old Javanese and his ability to scan rapidly the lontar lines in palawakya recitation. Pepaosan experts remark that parwa texts, though linguistically easier to interpret than kakawin, are much more difficult to 33

The Sanskrit term phalavakya means 'a promise of reward' In Old Javanese, the compound is made up of phala 'fruit, result, judgment' and wakya 'voice, utterance' A synonym that is sometimes mentioned for palawakya is paca pa(i)mng, which literally means 'accompanied reading' but is used in the sense of 'rhythmic speech' Another term occasionally encountered is gegancaian 34 The masked dance-drama toping, which dramatizes stories from the prose chronicles (babad), uses palawakya-style recitation in its dialogue, for example 35 For a discussion of the nature of syntactic "topic" in Austronesian languages, see Naylor (1975) on Tagalog 107

THE SHAPE OF THE WORD IN BALI

sound correctly. In the absence of strict metrics, the performer must rely more on meaning in delivering an acceptable reading. Within the range of palawakya, too, are sometimes abrupt style changes. For example, direct discourse may be read differently from descriptive narration, and a certain low chanting tone is adopted for the fragmentary &loka, or Sanskrit verses, that are scattered throughout parwa texts.36 In some manuscripts, such rhetorical switches in the text are marked with punctuation devices that give visual cues to the reader. Besides the cank (caesura), a double carik (//) may occur, rather like "quotation marks," at the beginnings of sections of reported speech, a papadan (see Figure 4) at the end of an utterance sometimes marks the beginning of a Sloka interlude. Depending on the passage being read, moreover, the length of each performed parwa utterance varies widely, from a few words (e.g., Mo]ar ta sang prabhu 'Said the king') to long sentences that include several phrases The performer must make decisions based not only on form and content but also on the role of the particular textual element in the discourse hierarchy. Mojai ta sang prabhu may be a separate utterance not because of any syntactic or semantic complexity but because it marks a discourse boundary where description switches to reported speech Furthermore, in the context of the pepaosan performance, the reader is signaling such a boundary not )ust as a matter of form but also because his vocal variation is information for the translator, who is thus cued to the boundary between discourse units The reader of parwa, then, has a complex )ob because the translator formulates his utterances based on the sounded version of a written text, rather than translating directly from written Old Javanese into modern Balmese. The translator is guided by the rhetorical interpretation that occurs as the reader shapes syntactic, semantic, and discourse elements vocally. The artistry of translators in pepaosan events, regardless of whether the text performed is poetry or parwa, is a kind of oral composition They receive information from the written text aurally and formulate a paraphrase that they deliver orally. Translators learn to do these things through many years of practice at listening and participating in pepaosan groups, moreover, whereas the best singers many have relatively little experience in translating, the best translators also have a firm command of reading-singing techniques, though this is not true 36

For further mention of these Sanskrit fragments, see the fourth section of Chapter 4, below See also Zurbuchen (1976b 37-39) and Zoetmulder (1974 8994) 108

LITERATE T R A D I T I O N S , LITERARY ACTS

in every case. Of course, the more times translators have been involved in the performance of a specific text (whether as singer, translator, or simply auditor), the smoother and more confident their translations will be. Yet they do not work only from verbatim memory—although there are individuals who have virtually memorized popular passages from the better-known works of Kawi literature and their paraphrases—and indeed such a technique is disparaged by experts. Rather, translators compose their paraphrases in concert with the linguistic and vocal musical information transmitted to them bit by bit by the singers. It is difficult to describe in words the extent to which, in performance, Bahnese paraphrase mirrors the vocal dimensions of the text as sung. Even though the translator uses a measured speaking voice rather than a singing or recitative style, the rhythm, the tonal contour, and, particularly, the characteristic rise and fall of the emphatic caesura given to the final syllable of the utterance all resemble the technique of the reader. While the length and content of the paraphrase in relation to the Kawi text are variable, rarely is the paraphrase much longer or shorter than the text. A kind of proportioned, back-and-forth relationship obtains between the singing and translation components of the literary performance, as the text is first musically realized by the singer and then given altered linguistic shape by the translator. When laying out the meaning of an ancient manuscript before his audience, the translator is at an interface between languages and cultural systems. Linguistically, he preserves correspondences between two separate yet related systems, Old Javanese and Bahnese, through structural parallels with the written text. At the same time, he is concerned to render meaning in terms that are within the experience of an audience far removed from the society that originally gave shape to the written text. The meaning that he speaks is both stable and transformed, a matter of both resemblance and difference. Its "sense," in Ricoeur's (1976) terms the "what" that is being spoken, is fashioned so as to remain stable between the text and translation. Yet m the process, the translator may relate the "reference," or the "about what" of the utterance, to the immediate Bahnese surroundings rather than to the historical past. The verbal art of the pepaosan is thus complexly and paradoxically both a remnant of the past and a creation of the present, a matter of both conservation and change. Mythic events and archaic culture are incorporated into the present moment, when literature is a communally shared experience. 109

THE SHAPE OF THE WORD IN BALI

The noetic characteristics of Balinese literature, with both written and sounded components, are mutually constraining. The fixed form of the written manuscript is always limited by the conditions of oral performance and variable transmission, and the potentially wide variation of the oral-aural text is always checked by the shape of the manuscript as well as by the line-by-line performance format. The literary discourse of the pepaosan is thus both a repetition of something uttered in the past and a new creation uttered only now, it is both fixed and fleeting, somehow "the same" even as it is changed. It is a rich mix of tradition and spontaneity, encompassing both the past and present in linguistic form and cultural meaning. CONCLUSION The discussion of manuscripts and performance in this chapter has focused primarily on just one area within the large realm of Balinese literate activity, that of verbal art or belles-lettres. What has emerged is a clear picture of how necessary to the continuing vitality and "readability" of Kawi linguistic varieties has been the pepaosan practice of sounding and translation. Within a larger framework, the group sharing that takes place in literary performance undoubtedly supports, and is reinforced by, values about how the cultural past can be important to the present. For most Balinese, "literature" is not equivalent to the typographically fixed and silently experienced books of the West. The literary entity is a matter of emerging form and meaning, as the relatively fixed, silent manuscript is combined with techniques of sounding and spontaneous paraphrase. In this view, the compilation of critical editions of Javanese-Bahnese manuscripts, and the substitution of printed "works" for the lontar and the pepaosan, would represent fundamental transformation and displacement of Balinese verbal art. This is not to say that critical editions are misconceived, but rather that we cannot understand what literary discourse in Bah means solely through such forms. What now needs to be considered is the extent to which other, nonbelles-lettres texts can be understood in light of the findings of this chapter It seems, for example, that just as the kakawin and parwa need to be sounded for their literary meaning to be fully transmitted, so also do certain other types of manuscripts need a real-world context of action or expression for their full meaning to emerge. A case in point are the medical manuscripts of the bahan, the usada, 110

LITERATE T R A D I T I O N S , LITERARY ACTS

which contain lore, healing practice, and ritual instruction. Important in the texts are lists of ingredients, series of ritual actions, and magic formulas to be uttered. These are signaled in the texts not by explicit statement but by a series of abbreviations, or panten (see Chapter 2), such as ma for mantra 'magic formula', ca for caru 'sacrifice to chthonic forces', and sa for saiana 'elements for offerings'. The user of these manuscripts has a practitioner's knowledge of the actual pronunciation of mantras and the combination of arrangement of elements in offerings; without such extra-textual experience, he or she would not be able to interpret the manuscripts. A similar situation obtains with regard to other types of manuscripts—for example, those used to perform elaborate ceremonies such as the huge centenary Eka Dasa Rudra of 1979. These texts state the names of dozens of types of offerings and how many of each are required, without specifying the composition and ingredients of each type. Yet another genre includes mystic, speculative treatises, tutur, in which indirect, allusive reference is made to yogic practices and altered states of consciousness brought about through esoteric discipline For both the offering texts and the tutur, the essential key to textual meanings is in the person of the living practitioner—the (generally) female offering expert {tukang banten) and the spiritual guide (guru), respectively—who is in charge of the reenactment of the past (the written word) in the present. Textual transmission is thus not )ust a matter of copying faded lontar leaves, it is also the active performance of texts whose meanings must be periodically rephrased in Bahnese speech or Balmese behavior to be meaningful at all. We are here, I believe, edging close to one of the intriguing paradoxes of this culture. Why is the island so full of the signs of its past, so dense with expressive forms preserving various historically prominent layers of belief? Bah has been cited as a "living museum" of syncretized and interspersed cultural patterns. At the same time, the culture's dynamic character, along with the lush growth and decay of its tropical ecology, means that its material remains from the distant past are rare, and most of what is cited as traditional or "old" is in fact relatively recent expression of patterns originating in the past that are still relevant in the present. Some differences between Western and Bahnese perceptions of the past were in 1937 the subject of an article by Gregory Bateson. While studying the renovation of a long-unused temple, he came to wonder why the villagers had not used the lengthy sequence of ceremonies and trance sessions leading up to the renovation to elaborate some version 111

THE SHAPE OF THE WORD IN BALI

of the temple's origins and history and the stories of the gods residing there, instead, he found, the only information about "the past" to come out of the events was that the god of the unused temple was the "daughter" of the god of the village temple. In fact, wrote Bateson, even this sparse account was as much about the present as the past, to have "talked" through the entranced priests, the gods would have had to be placed in some sort of relationship, so as to know how to behave and what sort of honorific or deprecating language to use with each other. Bateson concluded that the past is only necessary to the Bahnese as it gives form to immediate experience. "The past provides not the cause of the present but the pattern on which the present should be modeled" (Bateson 1970-135). While one may argue with Bateson's evaluation that in Bali things are not intrinsically valuable merely because they are "old," and while one must be careful in drawing analogies between old texts and old temple stones, there is a sense in which neither manuscripts nor overgrown carvings are of much significance to the Bahnese if they lack an essential feature: interpretability. In other words, unless there is some agent or event whose explication or action will bring the relic of the past into present focus, the importance of that relic is lost. In the case of a temple, past relations between deities and social groups are reiterated and reexpenenced through periodic ceremonial. In the case of lontar writings, the pepaosan circle, where vocal sound and the language of the present are vehicles of interpretation, is where the language of the past receives the necessary completion of its meaning. In both cases, the temple ritual and the literary event, there takes place a "reading" of the past; in both cases, the object (or text) from the past is incomplete without its recapitulation (or paraphrase) in the language or behavior of the present. If Bah is full of signs of its past, it is also teeming with the interpretations and paraphrases that must accompany them. We have seen through examining literary discourse that neither archaic written codes not oral interpretations can exist alone, the ancient language lives as it is uttered now, and the fleeting interpretation is significant when tied to a written form.

112

PART TWO THE DISCOURSE OF BALINESE SHADOW THEATER

that, among all of Bali's varied dramatic arts, the shadow theater (wayang) has received relatively little detailed attention in the works of scholars The standard treatise on dance and drama of the island (de Zoete and Spies 1938, new edition 1973) mentions wayang only in passing ' In the most popular introduction to Bahnese culture in general, Covarrubias (1937 243) gives some description of the setting and technique of wayang, noting that even though foreign visitors might not see much of interest in the drama, IT IS SURPRISING

[wjith its elaborate magic, religious significance, its undiminished popularity, and as the probable ancestor of the Balinese theater, the wayang remains the most important form of Bahnese entertainment While the actual origin and evolution of shadow theater in Southeast Asia remains obscure and much-debated, a consensus prevails that some sort of dramatic ritual connected with exorcist magic or ancestor beliefs very likely had its place in prehistoric Indonesian societies How and when this ritual became imbued with Hindu-Buddhist religious significance, came to use the Indian epic literature as dramatic material, or took the form of shadows from carved cowhide puppets on a screen as (perhaps) opposed to paintings on cloth scrolls or unilluminated puppet manipulation, we can never know with precision In the absence of historical documentation, we have generally tended to associate wayang in Indonesia with Java, where shadow theater remains an extremely important, widespread cultural focal point Accordingly, when scholars looked at wayang in Bah, they tended to compare it to Javanese shadow theater—a not unreasonable approach, since so many things Javanese have found their way to Bah and, through adaptation, survived there 2 Nevertheless, some important aspects of Bahnese shadow theater tend to be overlooked when it is regarded merely as another example of Javanese cultural impact One must be cautious, too, in interpreting Balinese wayang in terms of the "classical" forms of the eighteenth- and nineteenth-century Central Javanese courts, for the two islands certainly had independent shadow theater traditions long before then 1

"The magically powerful shadow play is not only the first ancestor of the Bahnese theatre, it is also the first school of style, accustoming the Bahnese from babyhood to see reality in the symbol" (de Zoete and Spies 1973 163) 2 McPhee (1970 148) hypothesizes that wayang arrived in Bali in the fifteenth century at the time of the decline of Majapahit, clearly this date is much too recent 115

DISCOURSE OF BALINESE SHADOW THEATER

Gons (1955 81) hypothesizes that wayang was a Javanese import to Bah, although performances are mentioned in an Old Balinese inscrip­ tion that predates the earliest period of intensive Javanese influence. In a charter of A.D. 896 (#002 in Gons 1954), we find the term parbwayang 'one who performs wayang'. Later documents in Old Javanese in­ clude the terms awayang ι haji 'performing wayang for the king' and annggit 'performing wayang', from nnggit 'puppet'. More recently, Wayan Simpen (1975 3-4) has discussed the origin of Balinese wayang in terms of the etymology of nnggit, which originally meant 'serrated, finely cut' and as such was applied in an early period to Chinese coin­ age and cut-out palm-leaf decorations that were used for offerings. It may also have been used to refer to carved and cut-out figures whose shapes were silhouetted on walls and, later, on screens. Yet however wayang came to be known in Bah, there is little doubt that Hindu-Javanese culture had great influence on both the form and content of wayang during the periods in which Java and Bah were closely linked politically and culturally This is obvious from features such as staging techniques, puppet shape, and dramatic content. 3 Still, the differences between the two forms are such that analysis of Ba­ linese wayang in its own right can contribute to a better understanding of the development of both Javanese and Balinese traditions. As Worsley (1972 93) suggested, Some impression of the extent of the linguistic integration of Jav­ anese and Balinese elements within Balinese literature can be gained from the lampahan of the Balinese wayang where both ele­ ments are to be found. A brief introduction to Balinese wayang in its cultural setting is Colin McPhee's "The Balinese wayang koeht and its music" (1936, re­ published 1970). C. Hooykaas (1973) has done extensive work on cer­ tain texts used by puppet masters (dalang), and he has studied as well the function of wayang from the point of view of Balinese religion (1957, 1960). Ensink (1967-1968) has compared some important fea­ tures of Balinese shadow theater with their Javanese counterparts. Hinzler (1975) has published an illustrated Dutch account of wayang 3

It is well known that the familiar thirteenth- to fifteenth-century East Jav­ anese "wayang" style of carved relief figures, as represented at Candi Jago and Candi Panataran, for example, bears remarkable resemblance to the Balinese wayang iconography of today See Holt (1967 82), van Stein Callenfels (1925), and Suleiman (1978, 1980) 116

D I S C O U R S E OF BALINESE SHADOW THEATER

performance and iconography native to both northern and southern parts of the island, more recently (1981), she has presented a broad study, based on fieldwork, of the themes, composition, and rehgio-cultural significance of the Bima Swarga story in Bahnese wayang. Among the Indonesian-language publications on Bahnese wayang are two collections of essays and two volumes that include transcriptions of plays and literary works relating to wayang.4 With the exception of Hinzler (1981), however, none of the sources on wayang mentioned thus far provide much insight into the linguistic and literary importance of shadow plays as they are actually devised and performed. Hooykaas, while unrivaled in his treatment of some of the textual sources used by puppeteers, never treats the issues of how a particular play might be chosen, composed, performed, and evaluated in a certain setting. His Kama and Kala (1973) includes a collation of sixteen manuscripts of the Dhaima Pawayangan, a "manual" emphasizing mystic and religious aspects of wayang widely known by Bahnese dalangs. Yet because of its lack of interpretive viewpoint, Kama and Kala fails to provide insights regarding the relation of the texts to performance. Having read the book, we are still no closer to understanding how the artist operates, or how audiences perceive plays. We are provided with text but deprived of context, and thus the discourse of the Dhaima Pawayangan remains confusing and impenetrable. It is my intention in the second part of this study to look at the discourse of Bahnese shadow theater—its meaning as language and as text—in order to move closer to an understanding of wayang's cultural, historical, and social significance I also hope to show how, in terms of management of necessarily shared cultural knowledge, in an environment of both oral-aural verbal art and manuscript literacy, wayang plays a key role in Bahnese verbal communication. In what follows, I will be examining the linguistic material of shadow theater in terms of the textuahty of the lampahan (literally 'movement, step'), a term used to refer to both the abstract notion of the structure of plays and the plot of a particular story. I hope to demonstrate that, behind the strange lamp-lit flicker of intricate shadows on a screen, there exists a unified expression of myriad linguistic preoccupations—an expression both reflecting and shaping the complex of Bali's language beliefs and language behavior Before beginning our examination of actual performance, it may be 4 See Proyek Pencetakan/Penerbitan Naskah Naskah Seni Budaya dan Pembelian Benda Benda Seni Budaya (1975a, b), Yayasan Pewayangan Daerah Bali (1978a, b), and Bandemetal (1981-1982)

117

D I S C O U R S E OF BALINESE SHADOW THEATER

helpful to consider the general situation of shadow theater in presentday Bali, and to clarify in which area of this broad tradition we are working. There are several types of shadow plays, all using puppets carved from thin cowhide, that can come under the heading of wayang kuht {wayang originally meant 'shadow', kuht is 'skin' or 'hide'), and that are distinguished according to repertoire, puppet shape, and stag­ ing technique. Of these, the wayang parwa and wayang lamayana, which portray stories based on the originally Indian epics Mahabharata and Ramayana, respectively, are clearly the most important. In­ deed, the wayang genres Cupak, Calonarang, Gambuh, and Sasak are only intermittently performed.5 Of the two prominent genres, wayang parwa is the more frequently performed, both because of its perceived larger repertoire of stories and the smaller size of its troupe.6 It is the best known, most widespread form of shadow theater, to the point that its huge body of stories (the "eighteen books"—Astadasa Paiwa—of the Indian written epic give the genre its name) is almost synonymous with the word wayang A basic feature of all wayang performances is their connection, in greatly varying degrees, with religious concerns and the "unmanifest" (niskala) world of gods and ancestors. Yet while offerings are always made, the drama's entertainment function may obscure its ceremonial aims for many members of the audience. There are certain wayang gen­ res, however, whose link with a limited number of specific ritual mo­ ments is more important than their role as entertainment for a human audience. These forms are distinguished by the event context, by their role in the ceremony, rather than by their repertoire or literary sources only. Wayang Sudamala, the least strictly defined genre in this group of ritual shadow plays, is in the broadest sense understood to refer to a wayang performance whose aim is purification of a specific person or 5

Cupak stories are indigenous Bahnese tales about fraternal rivalry and greed Calonarang is the name of an originally East Javanese tale of magic and absolution from evil Gambuh stories are taken from the Ραημ realm of origi­ nally Javanese tales about an idealized princely figure For information on the content of these genres, see de Zoete and Spies (1973 chaps 3 and 4) Wayang sasak originated in Lombok and has been popular among Sasak-influenced com­ munities in eastern Bali, its repertory includes stories of Islamic heroes and saints 6 Wayang parwa is accompanied by two (in northern areas) or four (in south ern areas) gendai musicians (see McPhee 1970 for a discussion of this genre of Bahnese music) Wayang ramayana uses ten to fifteen musicians in its bat&l en­ semble of gongs, drums, cymbals, etc 118

DISCOURSE OF BALINESE SHADOW THEATER

soul (from Sanskrit iuddha 'pure, cleansed' and mala 'stain, impur­ ity'). Yet the situation is nonetheless confusing, because Sudamala is also the name of a specific lampahan (also called KuntiSraya) used in purificatory wayangs (van Stein Callenfels 1925) In practice, wayangs for the purpose of nyudamala, ngruwat, or malukat (Balmese terms meaning 'to cleanse, purify') are of two types one, for the benefit of a human soul during death rituals {pitrayadnya),7 and two, for purifying a living person, either during rites of passage [manusayadnya] or while ridding someone of the misfortune of having been born during the week called wayang (one of the thirty seven-day weeks of the origi­ nally Javanese wuku calendncal system). This latter type of wayang, called sapuligai? is only performed during the period of wuku wayang, and it features stories of the origin of the dalang's power to free human souls from the ravenous, all-consuming Betara Kala 9 An important as­ pect of both the sudamala and sapul£g£r genres is the distillation, through their being performed, of the dalang's holy water, which is known variously as toy a panglukatan or wangsuh pada 10 There is one more genre of ritual wayang that can likewise result in the creation of holy water. This is the wayang lemah, which is most frequently (but not exclusively) played during temple anniversaries (odalan, which are a type of dawayadnya, or temporary sanctification of a temple or shrine by the presence of the deity to whom it is dedi­ cated). Wayang lemah must be performed to coincide precisely with the making of una 'holy water' by an officiating padanda. The word lemah 'daytime' ostensibly refers to the fact that this kind of wayang is performed during the day, although it can actually occur at whatever hour, day or night, that the priest begins his maw&da ritual The unique feature of wayang lemah is, rather, its screenless and lampless staging. Although the puppets are stuck into a banana trunk [gede7

In this context, besides the Sudamala tale of Betari Durga's exorcism, are performed stories relating adventures of souls in the nether worlds, for example, Bima Swaiga (see Hinzler 1981) Occasionally, wayangs held on the night be­ fore cremation are said to be charged with a seance-like atmosphere, as dalangs' voices might mysteriously alter to resemble those of the deceased 8 The term's origin is obscure, but it may be a compound formed from sapu or sapuh 'broom' and 1έ$έτ, which in Javanese can mean 'naked' 9 See Hooykaas's (1973) extensive treatment of textual sources for this lam­ pahan 10 Toya is 'water,' panglukatan is 'purification', wangsuh is 'to wash a sacred object', and pada is 'foot' The dalang's making of holy water involves dipping, first into the oil of the wayang lamp and then into the container of water to be transformed, the central supporting handle or "foot" of certain key puppets 119

DISCOURSE OF BALINESE SHADOW THEATER

bong) as usual, instead of having a screen [kehi] behind them they are leaned against a skein of cotton thread stretched between two upright branches of the sacred dapdap tree, which have been affixed in place of the poles that ordinarily support the wayang screen. The fact that it is very frequently performed on the ground of the inner temple courtyard gives another sense to the genre's name, for lemah also means 'earth' in Bahnese. To affirm human ties with the "mother earth" {Hyang Peitiwi) from which life springs, the interpretation goes, wayang lemah stories must concern beginnings and the origins of both the world and the social order. Thus, the episodes that are performed are often chosen from the Adiparwa, the first part of the Mahabharata, or are stories from origin-tale collections such as the Usana Jawa, Usana Bah, or Puiwagama.11 Although in many cases wayang lemah (also called wayang gedog 'box' or 'to knock, pound') is compulsory if a ceremony is to be regarded as complete, and although it is commonly found throughout the island, its performance never captures the visual or aural attention of the persons present In the midst of temple ceremonies, it is yet another of the many voices of ritual, as the bright and syncopated resonance of the gendur accompaniment and the dalang's varied intonations merge with the padanda's deep chanting and steady bell ringing and with many other musical and speech activities I have spent some time discussing the major recognized types of ritual shadow plays (sudamala, sapul6g6r, and lemah) as a way of introducing the wayang parwa It is this form that constitutes the basis of my discussion, although many of my conclusions regarding wayang parwa are equally valid for other Bahnese wayang genres. It is necessary to consider the associations of wayang with Bahnese ritual if we are to understand the dalang's art, his12 audiences' evaluations, and the broad cultural meanings of the tradition as a whole. Furthermore, as will be shown, the links of shadow theater with ritual practice and beII See Ramseyer (1977 200-210) for an account of wayang lemah performance, the differences between performing with and without a screen are mentioned in Hinzler 11975 32-33) 12 Throughout my discussion of wayang I use the male pronoun to refer to the dalang, and indeed, until very recently in Bah, there were no known female performers Yet in 1978-1979, at least two women dalangs held wayang performances, and female musicians were sometimes seen in their troupes Thus, although the profession is still overwhelmingly male, the small number of women performers must be noted

120

DISCOURSE OF BALINESE SHADOW THEATER

liefs about the structure of the cosmos are clearly made manifest in some of its patterns of language use. Accordingly, we will turn in the following chapters to the language of the dalang, and we will consider the play he speaks to be the "text" under scrutiny. In Chapters 4 and 5, we will look at the ways utterances and episodes of a wayang play are linguistically defined and conjoined. The discourse of Balinese wayang is more than the structural relations of verbal entities, however. In Chapter 6, therefore, we will consider the contextual factors that complement the linguistic features discussed in Chapters 4 and 5. These factors include the shaping of a wayang parwa play according to literary and dramatic traditions, the selection of the play in regard to a specific event and audience, and the setting of the play in terms of the world of the audiences who interpret its meanings. The "conditions of possibility" of wayang parwa performance given in Part Two are conceived along the lines of the types of information about discourse that were presented in Chapter 1 of this study, namely, the structural, generic, intentional, and referential relations that underlie all discourse Throughout the analysis, the special and significant noetic economy of Bah (as discussed in Chapters 2 and 3) is a continual point of reference

121

4 SPEAKING THE PLAY: THE USE OF K A W I

Samangkana sangkanya ngaian dalang, sua ta wenangmangwasaken kata Thus the origin of the name dalang it is one who is empowered to command speech —C Hooykaas, Kama and Kola (my translation] THE PURPOSE of the present chapter is to analyze the way a wayang play is spoken, examining the linguistic forms employed by the dalang. Wayang parwa is without doubt one of the most remarkable of Bali's many modes of discourse, and each performance constitutes a unique text that cannot be adequately captured by any one medium of reproduction. Thus, at the same time that we concentrate on the purely verbal material of the shadow play in this discussion, we must keep in mind the continuous interweaving of musical accompaniment, the movement and visually oriented staging techniques, and the social and event contexts that inform the drama In approaching a wayang performance as a text, we want to consider all the speech forms and literary genres that the puppet master, the dalang, uses in speaking the play As is well known, wayang lampahans are not written down,1 and dalangs do not memorize set "lines" of dialogue. In the manner of the verbal artist of oral tradition, they have no "script," no fixed inscribed text. Yet at the same time, the dalang can be a specialist in written tradition, and he may have experience with many manuscripts—of various genres—that are used in conceptualizing and sounding the lampahan. Further, certain religious meanings traditionally ascribed to wayang performance, and the metaphysical significance of his verbal art, place the dalang in the realm of the ritual practitioner, which itself requires specific kinds of language use. Instead of being denned as a particular language form with its own stylistic features, then, Balmese wayang is best conceived of as a combi1

The Gedong Kirtya, Bali's major repository of manuscript collections, does house a number of "plays" written down by dalangs in the 1940s, as part of a project to commission such works However, the practice of writing down dramatic material "as if" it were dialogue being performed is not typical among dalangs 123

DISCOURSE OF BALINESE SHADOW THEATER

nation of codes, styles, and genres that are shaped according to norms and constraints familiar to all performers. The profession of dalang is a highly disciplined combination of talents that in the West are generally considered independent instrumental music, singing, poetry, comedy, puppetry, dance movement, dramaturgy, and religious ritual Yet none of these abilities is more important than verbal talents, no dalang is considered accomplished without distinctive language skill, since "swaranya sakawuwuswuwus 'His voice says all that can be said' "2 This statement from the Dharma Pawayangan reflects the fact that, even in the multiform linguistic surroundings of Bah, special prestige accrues to the dalang as verbal artist. The dalang involves the imaginations of his audience largely (though not entirely) through vocal means. However, his prestige can depend just as much on his knowledge and experience of the written word. Nearly every discussion of the dalang's duties and training mentions familiarity with the manuscript tradition. Ingsun aguguru Sastia, mgsun aguguiu anuhs, ingsun aguguiu amaca, ingsun aguguiu aiaiasan, aguguiu putus, amutus mg atmku 1 have studied the manuscripts, I have studied writing, I have studied reading, I have studied the art of composition, [and] study is completed, finalized in my heart 3 It seems that in the dalang's art we encounter another area where integration of oral-aural and scribal awareness and techniques is important Literacy is a legitimate component of the play, just as is oral composition. THE USE OF KAWI In speaking his play, the dalang may have recourse to many languages, including Sanskrit, Old Javanese, Middle Javanese, Bahnese, Indonesian, and even (occasionally) "tourist" languages such as French or English. In general, however, the languages of wayang are (1) Bahnese and (2) what is referred to as "Kawi," that is, styles of Old Javanese combining features familiar from the classical works, such as kakawin and parwa, as well as from certain later compositions. It is 2 3

This quotation is from Hooykaas (1973 18), with my translation Hooykaas (1973 34), my own translation 124

T H E U S E OF K A W I

important to set forth the range of each of these languages in performance, for they are never used randomly, but rather determine structure both within and between sections of the lampahan While realizing that performance practice varies not only between individual dalangs but also between the different areas of Bah, I will nevertheless attempt, in the present discussion, to give a general idea of the use of Kawi in Balinese wayang 4 The use of spoken Kawi by Balinese dalangs is in marked distinction to Javanese wayang, in which the "elevated" or "literary" style of modern Javanese is common, but Old Javanese itself is mainly limited to the suluk incantations, whose literal meanings may or may not be known by the dalang.5 The dramatic usage of Kawi/Old Javanese is not strictly limited to wayang performance in Bali, but also occurs in, for instance, the arja, topeng, and wayang wong theaters. Yet it is my experience that the dalang's Kawi is more representative of the Old Javanese literary language, as well as more varied in its application, more productive, and less dependent on memorization, than the Kawi of the dance-drama genres Mantras To understand the scope of the dalang's Kawi, we must look at four distinct uses of this language in actual practice, starting with the very beginning of the performance. Watching a performance of wayang parwa in Bah, however, the Western novice may have occasion to wonder at which precise moment the play does begin Because of our experience in front of the proscenium stage, we are keyed up for the start of the action from the time the puppet master and musicians take their places behind the screen Yet during the lengthy musical overture by the genddr, as the dalang dedicates the offerings for the play and then sits reflectively chewing a plug of betel or busies himself adjusting the height of the coconut-oil lamp [damai), there is not yet a focus for audience attention. Anticipation heightens again when the kwpak (also called gedog), the puppet box, is opened after three sharp raps with the wooden knocker {cempala) Yet even at this point, following the fluttering dance of the stylized "tree" puppet {Kayonan), there is no 4 Thus, throughout the following discussion, which is based primarily on South Bah performance data, there may be details at variance with certain local traditions Hinzler's (1981 298-303) description of Kawi usage is based on her own widely collected data 5 See A L Becker (1979) for discussion of the role of suluk in the Javanese shadow theater

125

D I S C O U R S E OP BALINESE SHADOW THEATER

"story" action, but rather a long interlude while all the puppets are removed from the box and placed upright on either side of the screen or flat on the performance platform where the dalang is seated—all this to the meandering accompaniment of the gend£r quartet. Finally, after another dance by the Kayonan, the dalang, employing a mixture of singing and chanting techniques, intones what the inexperienced viewer would expect to be the play's first "lines." At this point there are a few puppets displayed in the central area of the screen, however, the dalang's speech seems to come from no specific character, and it certainly is not directed toward the encircling crowd, for they pay only scant attention. It is still a matter of waiting and wondering before puppets are articulated and given voices on screen, the gendor music subsides, and "dramatic dialogue" becomes an obvious focus. From the viewpoint of Western theatergoers, all this might be just an elaborate and puzzling way of diverting and diffusing audience interest in the performance. But the Bahnese watchers are not perplexed by the absence of a precise boundary between "real life" and the beginning of the play such as is represented by the rising curtain on the Western stage. They are accustomed to an elaborate transition encompassing offerings, music, contemplation, and various stages of puppet manipulation—a transition important for both the success of the entire performance and the safety of the dalang's person. Furthermore, the audience knows that the language material of the play commenced long before the dalang's first audible speech, in the form of prayers, charms, and invocations that the dalang recites, for the most part silently, beginning before he leaves his home for the site of the performance. This series of mantras, as they are generally termed, represents one of the distinct styles of Kawi employed by the Bahnese dalang. A dalang acquires his knowledge of these mantras, or ritual verbal formulas, both directly from another dalang and from written texts (particularly the Dhaima Pawayangan already mentioned, see Hooykaas 1973) that concern mystic aspects of the puppet master's art. The exact phrasing of the mantras may vary between dalangs, but they are generally uttered at the same clearly defined moments. Before leaving his home, for example, the dalang invokes inspiration from the taksu shrine in his house temple.6 He also pauses in his courtyard gateway, and again when he arrives at the home of the client commissioning the 6

Taksu is a Bahnese term regularly used for a variety of connections with otherworldly forces To become katakson during a performance—translated as the state of being 'entered' or 'endowed' by unseen powers—manifests itself in a particular vocal vigor, graceful and compelling movement of puppets, and clarity and extraordinary aesthetic achievement in portraying the lampahan 126

THE USE OF KAWI

performance, to silently recite brief mantras. Similarly, before he eats or drinks at the client's home, as he sits down before the wayang screen, dedicates offerings, chews the betel quid [canang], opens the puppet box, takes out the first puppet, anoints himself with oil or smoke from the lamp, and as he holds the Kayonan close to his face prior to its first appearance on the screen—at each of these moments he commonly offers a mantra. 7 And yet because of the silent recitation and the fact that his mantras constitute a ritual whose only other participants are from the "unmanifest" world, there may be no outwardly 7 See Hinzler (1981:60-70) for details on mantra recitation, offerings, and the preparation of holy water.

Unspoken language. The puppet master [dalang], silently reciting a mantia, holds a betel quid at the beginning of the performance (note the unopened puppet box). 127

D I S C O U R S E OF BALINESE SHADOW THEATER

visible or audible signs of this particular use of language on the part of the dalang In their textual form, the dalang's mantras are clearly similar to the formulas and invocations used by other ritual practitioners in Bah They are partially composed of Sanskrit or Sanskrit-like elements, many of which are familiar from texts used by padanda, pamangku, and bahan alike, for example, ya nama swaha 'let there be homage' They make use of power syllables (modr6, bijaksara) such as ang, ung, mang, and they close with familiar incantatory phrases such as poma, poma, poma 'so be it' and teka asih 'come love' In terms of content, we find the mantras pointing to the performer's immediate artistic aims The pang&g&r (or pangg&g&r 'stimulator') for­ mulas are magic charms to compel the attention and affection of au­ diences both human and nonhuman In their character as pangasih 'love-charms', these invocations often deify the performer as the god of love Kamajaya/Kamaraja Other formulas, some of them clearly stem­ ming from the realm of pangiwa (literally 'left-handed', ι e , "black" magic), include panolak 'repellers', panungkul 'means of subduing, overcoming', pangalup 'means of drawing attention', and pangembak swaia 'raisers of sound' A clear theme throughout many of these charms is the power of the dalang's voice over both human and other­ worldly spectators, his vocalizing is compared to the raging of the ele­ ments, to the sound of a gamelan, to birdsong, to heart-melting lament The performer conjures special verbal powers so that "Teka resep αίιηέ wong kab6h, angiungu sabda swaranku 'Delight comes to the hearts 8 of everyone listening to the sound of my voice' " As used by the different ritual practitioners of Bah, mantras are a type of "speech magic," that is, an arrangement of words and sounds, the force of whose uttering affects the state of things in the world Some of the dalang's mantras provide indications of the particular sources of his powerful voice For example, the Pangiaksa Jiwa 'Soul Protector' requests divine strengthening Om pukulun Sang Hyang Paiama Wis6sa mangraksa nwa, Sang Hyang Taya mangraksa bayu, Sang Hyang Purusa Wisosa mang­ raksa sabda, Sang Hyang Acmtya mangraksa idep, sapa warn apaksa αίαηέ, bhuta layak nembah, sarwa wisasa nembah, janma manusa nembah, Όέν/α Bhatara asih, apan aku Sang 8 The quotation and translation are from my notes Hooykaas (1973 34 43, 62-63) gives many examples of pangeger and pangalup charms See also the texts in Sugnwa (1963 23) and Proyek Pencetakan/Penerbitan (1975a 40, 108)

128

T H E U S E OF K A W I

Hyang Tiga Wisasa, luhui mg Sang Hyang Ongkaia Mula, malungguh nng nada hem, tan kasoian bhawa ring wisasa, sing ala paksano uiung, teka wedi, teka asih, patuh, patuh, patuh Hail Great God Supreme Power protect soul, Great God Taya protect breath, Great God Masculine Power protect speech, Great God Unimaginable protect thoughts, whoever dares to side with evil, demon and witch bow down, all powers bow down, humans bow down, divinities and gods show affection, for I am the Great God Three Powers, above the Great God Original Ongkara, residing in the slippery sound, unvanquished by the doings of power, every evil shall fail, come fear, come love, even so, even so, even so 9 The dalang's bayu, sabda, and idep (breath, speech, and thought) are a focus in this mantra, which also conceives the dalang's spiritual supremacy in terms of the transcendent power syllable and "slippery sound" ONG, the ongkara The three terms, collectively the Tnpiamana 'Three Powers' or 'Three Ways of Knowing', are commonly used in Bahnese philosophizing, and in a general sense they refer to action or results (bayu), the form in which these occur (sabda), and the motives or meanings (idep) that underlie them In the context of wayang, the triad of terms can refer specifically to linguistic entities or, more generally, to aesthetic aspects of the performance as a whole Figure 5 is the schematic representation of how the role of bayu, sabda, and idep might appear in wayang performance 9

For versions of this mantra, see Sugnwa (1963 23) and Hooykaas (1973 44), the translation is my own

Sang Hyang Tigajnana Kawiswara

Guru Reka

I BAYU

In language = In performance

the act of uttering movement action

verbal/vocal artistry

semantic and symbolic force conceptualizing the lampahan (plot)

FIGURE 5

Linguistic and aesthetic aspects of bayu, sabda, idep 129

DISCOURSE OF BALINESE SHADOW THEATER

The divine source of these three elements of verbal art is said to be a trio of deities that is variously called Sang Hyang Suksma (Ineffable) or Sang Hyang Tigajnana (Triad of Knowledge), that is, Guru R6ka, Saraswati, and Kawiswara Guru Rika is the guardian of thought, or idep, inspiring the story line that the shadows will enact, this deity's name could be glossed as 'Master of Creation/ Inspiration' 10 Saraswati (as discussed in Chapter 2) represents the origin of verbal form, especially written texts, the sabda of wayang Kawiswara (Kawi-Voice) empowers the bayu, the strength and vocal variation required to bring to life both story and language material u The Dharma Pawayangan emphasizes the relation between the two triads in the following admonition Mangkana tmgkah kiama dalang utama, wekasira Sang Hyang Guru R&ka, Hyang Saiaswati, Hyang Kawiswaia, angwawa nng Bayu-Sabda-Idep, ya ta sadanna angregepana, kadi ling mg aksaialki

Such is the way of the great dalang he commands the divine Guru R6ka, Saraswati, and Kawiswara, ruling over Deed-WordThought, that is what one must strive for, as these writings say 12 Another mantra used by the dalang, often just before opening the puppet box, invokes the assistance of "four dalangs" who "guard me as I perform wayang "13 This quartet is said to be located in the "four colors" of the performer's heart, whence it is "brought out" to the area of the vocal apparatus (tongue and velum), following which the performer places each of the four dalangs in protective array around his body According to some performers, the Four Dalangs are special transformations of the well-known four guardians of the cardinal directions (Sang Hyang Caturlokapala) Both quartets are assigned the colors (white, red, yellow, and black) fitting the nawasanga14 imagery evoked by comparison of the Four Dalangs with Iswara, Brahma, Mahadowa, and Wisnu (one of the sets of names used for the Caturlokapala) The contextualization in terms of the nawasanga is also made 10

Cf Sanskrit lekha 'line, sketch, figure, drawing' For more information on this triad of knowledge, see Hooykaas (1964 26, 38-39) 12 This quotation is taken from Hooykaas (1973 56), the translation is my own 13 Mantras that refer to these figures can be found in Hooykaas (1973 20-21) and Sugnwa (1963 23) 14 The nawasanga was discussed above in Chapter 2 11

130

T H E U S E OF K A W I

clear by the guardians' being placed on four sides of the performing dalang. In their role as reinforcers of the dalang's strength, with their location in the vocal tract, the Four Dalangs especially assist the performer's verbal power. The origin of these guardians is not clear, although there is a folktale concerning one of them (Hooykaas 1973:146-155).15 The Four Dalangs are logical counterparts of the Balmese Kanda Empat, four birth-siblmgs who accompany each human soul from conception through life and on after death. It seems as though the Four Dalangs are a special form of the protective and strengthening Kanda Empat, who are also, in turn, the microcosmic version of the macrocosmic four directional guardians.16 The layers of symbolism based on the nawasanga image here become a resounding theme in the dalang's preparation for performance. One further aspect of the Pangawak Dalang Pat 'Four Dalangs Embodied', as they are sometimes called (Bhasma 1975), is yet another association with a well-known quartet: the panasar (or pandasai from dasar 'base')—the important clown-commentator puppets, or 'base' figures, as their name implies. According to the Dharma Pawayangan, the Four Dalangs and the panasar group all have seats within the performer's body (Hooykaas 197321-22). Further, the words with which the dalang strengthens himself by figuratively "placing" both quartets (the panasars and the Four Dalangs) around his body ('on the right hand, the left hand, the front, the rear') are almost identical (Hooykaas 1973:21, 37). In sum, the meaning of the Four Dalangs derives from a complex representation of power and protection-bestowing elements, on which the dalang concentrates in order to form an external array of spiritual strength and guidance during the performance. This element of the dalang's pre-performance, unheard verbal activity is based upon the spatial organization of knowledge represented by the nawasanga. By properly activating micro/macrocosmic reality within himself, the dalang's performance, and especially his verbal powers, are enhanced The Dharma Pawayangan makes clear that the basis of the dalang's vocal technique and variation is in his awareness of the nawasanga's dimensions. Yan ring bhuwana aht, Iswara pupusuh, Wisnu ampru, Maha15

One derives a sense of lost ancient associations from their names Samirana, 'gentle wind', Anteban, 'hand motion, as of waving toward oneself, Jaruman, 'companion', Sapurna, 'full, complete, perfected' 16 The Kanda Empat are extensively treated in Hooykaas (1974) 131

DISCOURSE OF BALINESE SHADOW THEATER

d6wa babuwahan, Brahma ati, aneius η kuwung mg ati pakumpulanya, ya sangkanya dadi angrak anguwuhi In the microcosmos Iswara is the heart, Wisnu is the bile, Mahadewa is the kidneys, Brahma is the liver, in due time they come together in the hollow m the liver These are the causes of scream­ ing and moaning (Hooykaas 1973 16-17) The image distilled from the dalang's mantras, and the spiritual and professional aphorisms of the Dhaima Pawayangan, place the dalang in the center of many symbolic orders As he prepares himself for per­ formance, the artist adds layer upon layer of meanings constructed ac­ cording to the fourfold arrangement typical of the nawasanga As illus­ trated in Figure 6, by the time the dalang has begun the play he has distributed around himself a dense array of symbols categorizing val­ ued perceptions of how the world is ordered Nowhere is this function of directing cosmic forces toward imme­ diate dramatic and vocal ends clearer than in the dalang's Kayonan mantra This is one of the last of the series of "unheard" language acts, occurring just prior to the first appearance on the screen of the Ka­ yonan, the 'world tree', one of the most important of the shadow pup­ pets 17 Holding the elaborately carved tree-figure before his face, the dalang begins a counterclockwise naming of the nawasanga deities, which are said to one by one merge together, ultimately becoming part of the Ka­ yonan, which itself "returns to," or is subsumed by, a transcendent cognitive and creative force The last deity named, Iswara (standing in the East), is called the d6wa mng nnggit (God of Puppets) in Dhaima Pawayangan texts In this context, also, Iswara is said to have the na­ ture of swaia (or suaia) 'voice' The Kayonan mantra is thus a distilla­ tion of the various directional forces into one vocal power in the serv­ ice of wayang performance Furthermore, by invoking the "divine Kawi" near the end of his mantra, the dalang indicates that for him Old Javanese is not just the language of vaguely historical cultural prede17

Kayon is from a common form in Indonesian languages, kayu 'tree', plus nominalizing suffix -an Among the Bahnese, kayon is often associated with kayun 'thought' (cf Old Javanese hyun 'affection, intent') For wayang perform ers, the tree-figure is especially important as a symbol of creative and imagina­ tive force The old Javanese term wwit had various meanings including 'origin', 'ancestry', 'tree', and so on In modern Bahnese, wit means 'tree', as well as re­ ferring to beginnings and ancestors Trees, ancestor beliefs, and beginnings of performances all come together here 132

THE USE OF KAWI KAJA (NORTH) a) Wisnu b) black (ireng) c) form of the world (rupa rung bhuwana) d) bile (ampru) e) Sa (m) purna - black of the heart f) throat g) Sangut h) on the left (kivia) KAVH (WEST) a) Mahadewa b) yellow (kunmg) c) earth (ksiti-dharam) d) kidneys (babuwahan) e) Jaruman - yellow of the heart f) back of the tongue g) Werdah h) in back (un) /

s

KANGIN (EAST) a) Iswara b) white fputih) \ c) sky (akasa) 1 d) heart (pupusuh) / e) Samirana - white of the heart f) tip of the tongue g) Twalen ·. h) in front (arep) /

1 1

\

TENGAH D A L A Ν G

(CENTER)

KELOD (SOUTH) a) Brahma b) red (bang) c) radiance of the world (teja nmg bhuwana) d) liver (an) e) Anteban - red of the heart f) middle of the tongue g) D£lem h) on the right (tengen)

>,

KEY'» a) directional deities (Sang Hyang Caturlokapala) b) colors of Sang Hyang Caturlokapala c) manifestation of Hyang Caturlokapala in macrocosmos (bhuwana agung) d) manifestation of Hyang Caturlokapala in microcosmos (bhuwana alit) e) Four Dalangs (Dalang Catur/Pat) and location in performer's body 0 place of Four Dalangs in vocal tract g) Panasar ('base'-figures) h) place of Four Dalangs and Panasar surrounding performer FIGURE 6 A poetics of space for Balmese wayang 18

This diagram follows, for the most part, the collation of Dharma Pawayangan manuscripts presented by Hooykaas (19 73) The texts are not consistent on several points, for example, some of t h e m mention Delem " i n front," Twalen " o n the left," and so forth, which would imply that the performance area is oriented toward the sea [kelod) Many dalangs, however, feel that kangm is the most favorable direction toward which to orient the screen and thus the per­ former's body, so I have here depicted the dalang to be facing eastward

133

D I S C O U R S E OF BALINESE SHADOW THEATER

cessors but also an active power in the present moment of perform­ ance. As he moves from silent to uttered speech, the dalang, through concentration and ritual, enters an idealized frame in which "kawi" is no longer only the name of a language but also a realm of intermingled cultural heritage, literary tradition, aesthetic aims, and metaphysical presences. Sambu muhh manng Wisnu, Wisnu muhh manng Sangkara, Sangkaia mulih manng Mahadowa, Μαηαάέν/α muhh manng Ludia, Ludia muhh manng Biahma, Brahma muhh manng Mahaswaia, Mahaswaia muhh manng Iswara, Iswara muhh manng Kakayonan, Kakayonan muhh manng manah mahasakti mahenmg, manah mahenmg muhh manng Sang Hyang Kawi, kawi kawikanan sakawuwus wuwus tmgkaha ngaianya. Sambu returns to Wisnu, Wisnu returns to Sangkara, Sangkara re­ turns to Mahad6wa, Mahaddwa returns to Ludra, Ludra returns to Brahma, Brahma returns to Mah6swara, Maheswara returns to Is­ wara, Iswara returns to the Kayonan, the Kayonan returns to the Pure All-Powerful Mind, the Pure Mind returns to the Divine Kawi, "Kawi" is the skill of speaking all that can be expressed by words.19 What the dalang's mantras show us about the linguistic aspect of wayang parwa is significant. First, use of mantras in Old Javanese di­ rectly links the dalang to other Bahnese ritual specialists who also have access to the "unmanifest" world by virtue of formulaic utter­ ances in this code, and whose verbal activities are a form of speech magic. Second, the use of Old Javanese to both prepare the performer and preface the play is an integral and necessary part of the entire per­ formance. The audible, publicly directed speech of the play is consid­ ered the natural outcome of the inaudible utterance of mantras. The performer's inner state is linked to the qualities of his performance, the sounds of the play presuppose the period of silent utterance. In the dalang's case, the first audible speech is a culmination, not a begin­ ning, of his verbal display and artistic involvement. Dialogue As the dalang begins to manipulate the puppets and use his voice for theirs, he does not stop using Kawi. However, the use of Kawi in dra19

This text is based on Hooykaas (1973 32) and Sugriwa (1963 23), the trans­ lation is my own

134

T H E U S E OF K A W I

matic dialogue has a different form and function from that in the mantras. First, from the viewpoint of drama, speaking in Kawi is a defining attribute of a certain group of characters. These include the duwa or betaia 'gods', resi 'sages', and all members of the satna 'noble' class, as well as the myriad demons, monsters, and nonhuman orders. Men, women, children, armies, and animals within this range all use Kawi. What ties the various orders together is that they are considered to have a common source in the epic realm. In fact, many characters (such as the demon-god Kala)20 originate outside the Mahabhaiata tradition, from other literary works or from within the creative realm belonging to many generations of dalangs. Yet in wayang parwa, all the characters who share the screen with the Pandawas and Korawas (the antagonists in the Mahabhaiata), and who belong to the classes of betara, resi, ratu, or laksasa (gods, sages, kings, or demons, respectively), must speak Kawi. The class of exceptions to this rule of Kawi-speaking is small but of critical importance. It includes the panasar ('base'-figures) or parekan 'servants', whose status as court retainers privileges them to explain their masters' and mistresses' minds in the vernacular.21 The qualities of Kawi as sounded in Bah claim particular attention, because nowhere else in that part of the world where this language has been important do we find Kawi commonly uttered. Indeed, most scholarly description gives the impression that Kawi has survived only through manuscript transmission, as an archaic, "leam6d" language impenetrable to all save a specialist elite. While Old Javanese texts are read by a limited number of people, there are at the same time many settings that require the use of Kawi, thereby ensuring its continued familiarity as part of the aural world of most Bahnese.22 So numerous are the ritual and theatrical usages of the language by priests and performers, in fact, that any adequate consideration must take into account not just the extant body of manuscripts, but also a complex and multiform range of oral-aural experience of Kawi. The dalang's dialogue represents the most extensive Bahnese use of 20 This is the same Kala whose powers the dalang is able to overcome in the sapuleger ritual shadow theater (see Hooykaas 1973], the character appears in many other stories as well 21 The parekan's linguistic functions will be discussed in detail in Chapter 5 22 Again, see Walhs (1980) for an analysis of the acoustic culture of Bah, that is, the ways different codes and vocal styles are used to communicate in specific contexts

135

D I S C O U R S E OF BALINESE SHADOW THEATER

spoken Kawi. Just how this dialogue is composed in performance and how the skill for its production is acquired are questions basic to the understanding of a tradition of verbal art that at present occupies perhaps several hundred Bahnese in the profession of dalang.23 For unlike the use of the language by other Kawi-speaking practitioners, such as the pamangku or balian, the dalang's Kawi is not limited to memorized chants or prayers. Even when compared to other forms of theater that use Kawi dialogue, the Old Javanese of wayang is more elaborate and more central to the performance as a whole. There are several aspects to the poetics of the dalang's orally composed dialogue. These include the sources for the dalang's Kawi—the "speech community" with which he conforms; the characteristics of the dialogue measured against criteria of oral poetry and oral composition, and the features of this type of Kawi as compared to other, primarily written, forms of the language. Sources. The skill of the Bahnese dalang is multifaceted and elusive, because the performer is master of so many arts, it may be difficult to pin down just which aspects of a wayang have either pleased an audience or left it unimpressed. There are few absolute artistic canons in the discipline and no lists of what a "real" dalang unquestionably must or must not do in performance. Yet there is a general consensus among performers and other Bahnese commentators on padalangan (the art of puppetry) that the ability to compose dialogue in Old Javanese and knowledge of literary texts are preeminent among the dalang's qualifications (Sugnwa 1963:19, Bhasma 1975:63-64). And indeed, the use of Kawi in dramatic dialogue is intimately linked with written textual tradition. In answer to the question "How did you learn Kawi?" the dalangs frequently give one of two replies. First, it is agreed, the skill can be acquired through "reading the lontars." But there is also the vaguely defined, unmeasurable influence of "hearing Kawi a lot." This second reply is readily understandable when one notes that many Bahnese dalangs come from families of wayang performers, and thus from early childhood they are continuously exposed to and involved with the profession. Such families or descent groups, moreover, may be located 23 After conducting a survey of 36 out of the 50 subdistncts in Bali, Arthanegara (1975 113-114) listed 207 persons as "active" dalangs and 51 as "inactive " Hinzler (1981 45-46) puts the total number of dalangs at 108, 72 of whom are from the commoner (jaba) descent groups

136

T H E U S E OF K A W I

in communities where many types of performing arts and literature are important activities.24 In such an atmosphere of supportive familiarity and constant performance activity, the small child's ears are as accustomed to hearing father, uncle, or cousin speak the dalang's Kawi at night as they are to hearing the everyday voices of the same persons. The guttural roar of the opening invocation, the intonation contours of dialogue, and the phonetic range that separates standard vocal types such as alus 'refined' or lehleh 'seductive' from kasai 'rough' or keras 'strong'—all these features the youngster knows intimately and can readily imitate.25 A child also has opportunities to practice handling puppets, often by playing with paper cutouts or worn-out figures from a performer's set, and he may even have a "practice screen" for imitating movement and language that he has absorbed through exposure to performances. As he grows up, the young member of the performing family acquires other skills important to the professional wayang troupe. He learns to play the gend6r accompaniment (one of the most arduous and intricate instrumental feats in Bah), to carve and paint puppets, and to serve during the play as one of the dalang's two assistants, called tututan 'that which follows along', who are responsible for setting up the performance area and handling the puppets during performance when the dalang himself is not using them As a dalang's assistant, the novice acquires expert knowledge of the wayang repertoire and the composition of individual stories. The assistant must ideally anticipate the dalang's scene and character changes, which can occur rapidly. At the same time, the novice's knowledge of Kawi broadens, not just through participation in performance, but also through the stable, close-knit, self-perpetuating traditions of a performing family. This is demonstrated in moments when, relaxing or working at home, a dalang "quizzes" his young children by asking for the Kawi equivalents of Bahnese words. Not all young boys in dalang families become dalangs, of course, and 24

In the southern Bah village where I lived, Sukawati, one banjar (hamlet) with a population of around one thousand persons included more than a dozen professional dalangs of whom no fewer than six performed from three to twenty times a month The banjar, moreover, consisted primarily of one interrelated descent group, or soroh, which was often referred to as soroh dalang 'the family of dalangs' 25 The Dharma Pawayangan texts include descriptions of the types of wayang "voices", see Hooykaas (1973 26-27, 70)

137

DISCOURSE OF BALINESE SHADOW THEATER

many puppeteers do not decide on the profession until they are well into manhood. The decision to seriously prepare a novice for actual performance may involve older clan members—the long-practiced professionals—in a discussion of the potential of the candidate for dalang. The same relatives will work closely with him, perfecting his speaking and singing voices, puppet manipulation, and cempala technique (the cempala, as previously noted, is a wooden knocker held in either the hand or foot). The serious candidate who undertakes the dalang's profession begins at some point to study Kawi texts. In former times, he might have learned to read especially for this purpose, under today's public education system, he might learn the rudiments of Balinese script as part of elementary education. While some dalangs participate in village literary performance groups (seka pepaosan), others may read lontars only at home with a few family members or sound the lines alone in their spare time Often, too, a person seriously committed to study will move into the house compound of a practicing dalang, and there will go over manuscripts with his teacher The materials studied by dalangs are primarily Kawi texts relating the events of the Mahabharata. Important among these are the Old Javanese parwa texts and the kakawins Bharatayuddha and Arjunawiwaha. The kakawin Ramayana is also considered fundamental, not only because it is "background" to the parwa stories proper, but also because of its moral/ethical content and much-revered poetic excellence. The kakawin Sutasoma is important because it expounds the Balinese religious viewpoint of Siwa-Buda synthesis. Aside from these basic works, there is a range of written material, including didactic treatises, more recent poetic or prose renditions of Mahabharata stories, and Dharma Pawayangan manuals, upon which dalangs draw It must be emphasized, however, that there is no specific "syllabus" that candidates must complete before "graduating" to become practicing dalangs, and therefore the involvement of performers with written works vanes greatly. In the actual composition and delivery of Kawi dialogue in performance, the parwa texts are undoubtedly closer than other literary styles to the dalang's language, in both syntax and lexical choice This is the case for several reasons First, the dalang's aim is to portray realistically the dialogue between puppet characters, and the parwa texts, which alternate direct discourse with straightforward narrative, provide a closer analogue to dialogic exchange than does the intricate poetic language of the kakawin texts. Second, since the speeches in the parwa texts are unconstrained by the elaborate prosody and resulting 138

T H E U S E OF K A W I

syntactic complexity of those in the kakawin texts, they are a better model for the construction of new utterances. The range of the Old Jav­ anese lexicon is narrower in the parwas than in the kakawins, there are fewer elaborate Sanskrit-styled compounds, and poetic figures and de­ vices Ialamkara) do not come into play. And finally, the composers of the Old Javanese parwas were concerned primarily to render a chain of tales succinctly, the result was greater accessibility of these texts as general "reference" works for stylistic and plot-creating purposes. It is possible to get an idea of the echoing of the parwas in actual wayang practice when the written texts are compared with performance style. In the following examples, the first quotation in each pair is taken from a tape-recorded wayang parwa performance26 and the sec­ ond from the Adipaiwa (Widyatmanta 1958) Sentence1(a)

Ahum (excl)

kita P2

pawekas advice

sang Yudistna/ (art) Yudistira

mnghulun of PI

rengenan be heard

yayatAki in this way,

mangkZI now

Ah, you, Yudistira, listen to this advice of mine now. 1(b).

Ai Paitha kita (excl) Partha P2

sang Dhananjaya/ (art) Dhananjaya

Rengonta be heard P2

canta mnghulun/ story, account of PI Hey, Partha, you, Dhananjaya, let my story be heard by you. [Adipaiwa, vol. 2:17, p. 104) 2(a).

Ibu/ mother hana be

yan if

lwii like

mangkana thus

kenoh pleasing

satuian miang in accord his

dahat very

nnasa/ be felt

apanl (conj)

K&sawal Kdsawa

Mother, in that case it is good, moreover, I, Kesawa, am in agree­ ment 26 Except where otherwise noted, quoted excerpts are from cassette record ings of complete performances collected in 1977-1979 The following perform­ ers, listed with village of residence, gave permission to record or permitted ac­ cess to prior recordings of their performances I Ketut Madra (Sukawati), I Wayan Nartha (Sukawati), I Nyoman Ganjnng (Sukawati), Ida Bagus Gria (Su­ kawati), I Madi Sija (Bona), I Nyoman Rajeg (Tunjuk), I Nyoman Sumandhi (Tunjuk/Denpasar), Dewa Rai Μέβι (Bangli), and Ida Bagus Buduk (Bongkasa)

139

DISCOURSE OF BALINESE SHADOW THEATER

2(b). Azya (title)

Widura/ Widura

Atyanta exceedingly

azya Bhlsma (title) Bhisma ι (art)

kenoh pleasing

kummkin paying attention to

putu grandchild

dahat very

άέ έζϊ by (title)

kahaywan well-being

nira kalih/ his both

Noble Widura, how exceedingly good that noble Bhisma is consid­ ering the welfare of both his grandchildren. [Adipaiwa, vol. 2:16, p. 95) Symbols used

(/) (excl) (art) (conj) (top)

caesura (caesura occurring in performance or in manuscripts are indicated similarly) exclamation article conjunction topic marker

The preceding quotations illustrate correspondence on a number of levels between the dalang's Kawi and the Old Javanese parwa style. The use of zengen/rengon 'hear' and the phrase kenoh dahat 'very good' are obvious and common examples of lexical similarities. The choice of pronouns [kita 'you', nmghulun 'mine', third-person possess­ ive miang/nira) as well as personal articles [sang] and titles [lbu, azya) also demonstrates shared features. The following excerpts indicate some syntactic parallels, in this case the use of a non-actor focus pred­ icate construction (the so-called "passive") employing the infix -m(goal focus marker). 3(a).

Gmawayaken be built

umah house

A house was built (for him) 3(b).

.. . gmav/έ be built

tikang (top) that

kadatwan palace

mkang at that

Khandawaprastha Khandawaprastha The palace was built at Khandawaprastha . {Adipazwa, vol. 2 16, p. 96) The dalang also uses the predicate prefix indicating non-mtentionahty as a feature of an action, just as in the parwa, with the ka- form 140

THE USE OP KAWI

4(a). Karenga d&nta sua be heard byP2 P3

sang Pandawa (title) Pandawa

ngwangun karya/ do/perform work

It was heard by you that the Pandawas are preparing a ceremony. 4(b). Kaiengo pwa be heard (top)

άέ nuanangis by P3 cry

άέ sang Arjuna/ by (title) Arjuna

Heard by Arjuna was that person lamenting. {Aaiparwa, vol. 2:17, p. 101) The use of the agentive preposition άέ is another similarity between the two examples in the last pair. Although there are also a number of areas of significant contrast between the Kawi of the dalang and parwa texts (see the discussion below), we can nevertheless observe that parwa manuscript tradition is one model for the performer's style. Another dimension of the relationship between the two varieties of Kawi emerges when we compare the dalang's delivery with the vocal realization of the parwa text in the context of group performance (i.e., the seka pepaosan). The dalang manages a gamut of phonetic tech­ nique in producing the guttural roars of demons, the high, whining ca­ dence of refined nobility, and the energetic, peppery style of rough he­ roes such as Bima. Yet according to dalangs themselves, there is a principle that underlies the production of the various voices, which is called malawakya. This term is a verbal form of palawakya, which in Chapter 3 was described as the vocal style used for sounding non-met­ ric ("prose") written texts such as the parwas. Again, palawakya vocal style is characterized by the intonation contours of phrases of unequal length sounded by the reader according to caesura marked in the man­ uscript before him. An examination of the acoustic features of the dalang's dialogue re­ veals major links with reading-group palawakya style, including the same parameters of intonation and, especially, phrase marking. In pe­ paosan sessions, each phrase unit is given an abrupt, heavy, falling con­ tour on its ending syllable, by way of vocally signaling the caesura. In wayang dialogue, caesura are instead marked by sharp raps of the footheld hammer (cempala). A Bahnese explanation of this process is that the dalang must manut angsel nyank αίιιτέ 'follow the pauses punc­ tuating the speaking' (Proyek Pencetakan 1975a43). The term nyank is a verbal form of cank 'written caesura', the knocking of the cempala is thus an audible cank in the dialogue of the dalang.27 27 Sentence 1(a) above shows a typical pattern—an utterance is composed of two constituent phrases marked by the cempala, one phrase identifies the hearer, while the second contains the proposition to be communicated 141

D I S C O U R S E OF BALINESE SHADOW THEATER

In the following example, the cempala is heard six times and indi­ cates the constituent phrases of this single "sentence" of dialogue. 5. Lakia I pwa kita nanak I tan Ιέη kagmucaia άέ lbunta I η tatkala I ang&makena kunang jagat lamakana sida I amangguhakena kertaraharja I Boys / my children / it is none other your mother speaks of / than a time / of great striving in the world in order to / establish peace and well-being. It thus appears that palawakya, both in oral performance of parwa texts and in the realm of the dalang's dialogue, refers to a phrase-focused prosodic model that the reader or performer applies to his material. Why, however, should the Kawi dialogue of wayang be described in terms of a nondramatic reading/recitation technique? After all, the dalang does not reproduce the language of the parwa in verbatim fash­ ion, since he does not memorize speeches from a written text. He uses no script at all, in fact, and two performances of the same story may include totally different Kawi speeches. Part of the answer undoubt­ edly lies in the fact that the dalang is familiar with the parwas, since these texts include most of the "core" tales of the repertoire. More­ over, many dalangs have experienced the pepaosan as auditors or par­ ticipants, for them, palawakya reading style is inherently lmked to the dramatic reproduction, m shadow theater performance, of the speech of the same epic characters. Long interdependence between literary and dramatic realms has in various Bahnese theaters given rise to spe­ cial "stage" styles for the poetic forms of the texts upon which the drama is based: in aija dance-drama, for example, tembang vocal mu­ sical style is a standardized, elaborated variant of the technique nor­ mally used to sing the same texts in nondramatic contexts. The relationship between palawakya vocal style in the two perform­ ance contexts, reading group and wayang play, can thus be viewed in terms of the written textual background for dramatic usage of Kawi. To adopt this view, however, is to overlook a key aspect of the dalang's verbal artistry, namely, its character as oral composition. For, sur­ rounded by dozens of puppets in the midpoint of a story, directing the interchange of music, movement, and voice, bound by the tenets of his profession to maintain the flow of spoken text, the dalang creates dia­ logue in a moment-by-moment application of composition technique. He continuously weds verbal convention and dramatic necessity. He has no time to scan his memory of a written text, looking for appropn142

THE USE OF KAWI

ate utterances to imitate, instead, he must have available a set of verbal patterns that are immediately combinable, and that allow him to compose the next speech even as he is pronouncing the present one. The dalang's Kawi dialogue, we can hypothesize, should possess features that identify it as belonging to the realm of oral composition. Composition m Performance. To analyze the dalang's Kawi dialogue in terms of its composition in performance, we do not have to presuppose the existence of the oral poet's "formula" as Lord and Parry defined it: "[A] group of words which is regularly employed under the same metrical conditions to express a given essential idea" (Lord 1976:4). The wayang performer does not sing a fixed melodic and metric line, but uses a highly stylized speaking voice to declaim non-metric utterances. The fact that the dalang is not speaking "poetry" need not blind us, however, to the prosodic entities that are present in the dialogue. The dalang's formulas exist, but not in accordance with Lord and Parry's definition.28 As will be seen, patterns of rhythm, tonal contour, syntactic units, and rhetorical shifts all combine to form a setting for formulaic linguistic displays As discussed above, palawakya is a term used in both wayang theater and pepaosan literary performance to refer to the delivery of nonpoetic verbal material. The pepaosan reader is guided by caesura marked in the manuscript, when there is too long a "phrase" between such caesura, he must break it into two or more syntactically cohesive constituents. This is necessary if the translator, the )uru mabasan, who does not have a written text before him, is to be able to compose coherent paraphrases smoothly. In wayang performance, the dalang begins with an idea that needs to be conveyed through dialogue, rather than with a fixed-form text. He then proceeds to compose the Kawi utterance, which in most cases is immediately followed by a Bahnese paraphrase. The Kawi speeches themselves vary in length, example 5, above, is slightly longer than average, while example 3(a) is quite brief. Typically, speeches at the beginning of the play are longer, since the first scene is invariably a meeting at court and has a stately, measured pace emphasizing formal, flowery dialogue over action or comedy. As the play progresses, Kawi speeches become more succinct, particularly in the fast-paced scenes emphasizing dramatic or comic action, battle, and rapid movement. 28 The types of formulaic expression employed by the Malay dalang of Wayang Siam are covered in Sweeney (1980)

143

D I S C O U R S E OF BALINESE SHADOW THEATER

Normally, a single Kawi speech will be punctuated at least once by a single rap of the cempala, which indicates phrase boundaries. The tendency for the dalang to compose in syntactically defined word groups or phrases, between which he may pause briefly to collect his thoughts and compose the succeeding phrase, establishes an irregular rhythmic progression of verbal material and the rapping of the cempala. When he is composing longer speeches, and in scenes emphasizing dialogue over action, the cempala is more regularly and frequently heard. In fast-moving battle scenes, by contrast, speeches are succinct, often amounting to no more than a short phrase or exclamation, and the cempala marks movement rather than dialogue. The wooden knocker, hammered against the puppet box by hand or foot, functions further as an aid to the dalang's thought and composition processes by supporting the continuous alternation of sound textures characteristic of wayang performance. The cempala is sounded not only to punctuate dialogue but also to accompany certain puppet movements and to tell the musicians when to start or stop playing, as well as which type of piece to play. By varying the force and frequency of the knocking, the dalang can heighten the mood of a sad or angry speech; he can emphasize puppet movement such as the joining of hands in respectful salute [sembah) or the parrying of blows in battle. During interludes of puppet dance and song, the cempala functions like the Bahnese gamelan's drumbeat, and the dalang's display of distinct, intricate rhythms rapped with the foot-held hammer, while he simultaneously manages the movements of a dance duet, requires great experience and skill. The cempala thus assists the entire range of the play's dramatic components, including its Kawi and Bahnese dialogue. Returning to our examination of the sorts of purely verbal patterning marked by the phrasing of the cempala, we move closer to an understanding of the phrasal entities upon which the performer relies in dialogue composition. For even though the "lines" of two wayang plays are never the same, there are regular and observable similarities that structure the oral composition of dialogue. The dalang depends on a certain amount of this "stock" verbal material for the desired fluidity of a performance during which he works without break. As indicated earlier in this section, the raps of the cempala delimit phrases of dialogue that are usually syntactically coherent, that is, word-group boundaries do not normally intervene between nuclear syntactic bits such as a transitive verb and its object. Moreover, in many typical speeches, contrasting types of phrases seem to occur in a fixed order. 144

T H E U S E OF K A W I

6. Singgih / inganika yes/indeed honored nimittanyan reason the

dateng come

ibu / dénikanang mother by that marikanang to that

kadiang how

jagat territory

punapa / what

Astina / Astina

Yes/ honored mother/ on what account/ for what reason have [you] come hence to the Kingdom of Astina? 7.

Singgih/ yes kadi like

inganika honored

pangaitika speech

sanghyang (title for deity)

Surapati/ Surapati

men6 just now

yan if

setepung pinaiiyuta tan ana singsal / one group come million not be difficult

Yes/ honored Great God Surapati/ if it be as just spoken, [if they come] in groups of millions, there's no problem. 8.

Uduh / ceraka oh servant wus finish

kerta do

Twal6n Twal6n

kita / 6nak pawistaning P2 yes, good speech of

ulun PI

lugraha / granted

Oh/ you, Twal6n, servant/ yes, it is granted you should proceed [to translate] my speech. 9.

Uduh / yayi oh younger brother lumaris departing

kaya such as

piawiia hero

sadaya one heart

adan pada (conj) all

mangki, / now

Oh/ brothers, dear heroes/ let us all depart at this time. 10.

Boh boh (excl)

kita P2

Arjuna / ngong Arjuna PI

tan not

alah defeat

dtnikang by that

astranta / arrow your Ho, ho, you Arjuna/1 can't be beaten by your arrows. These five examples of dialogue, along with example 5 above, illustrate a number of points. First, each utterance begins with an expletive or exclamatory particle. Each of these elements tends to characterize the relationship between speaker and addressee and thus the dramatic "mood" of the utterance as a whole. For example, singgih indicates obedience and respect and is used from a subordinate to a superior, as when the Pandawa brothers address their mother, D6wi Kunti (6). Al145

D I S C O U R S E OF BALINESE SHADOW THEATER

ternating with smggih, and pronounced with the same prolonged rising tone, are the forms namya snapakulun. Uduh, by contrast, is directed toward a lower-status character from a higher and connotes friendly in­ timacy, as when Yudistira addresses his servants (8) or his younger brothers (9). Filling the same slot as uduh are ahum, which is most often used by sages and teachers, and lakia, which is used to address male hearers, for example, by Dewi Kunti to her sons. A third category of initial expletives might be termed kasar 'coarse'; the terms in this group show familiarity (sometimes presumptively so), aggression, an­ ger, or contempt. Ih at the beginning of a speech usually shows impa­ tience or rage on the part of a warrior addressing an enemy, the plainspeaking Bima uses ah in many of his speeches, pwah/bah/yah express surprise or dislike, and demons address human beings with a rumbling boh (10). There is also a class of expletives that is more neutral in char­ acter, serving to move the action along rather than pinpointing char­ acter relationships. Included in this group are agehs, agya, and τηβηέ, which can all be glossed as 'quickly', and which can also be heard in the dalang's narration. The four groups of utterance-initial expletives (respectful, intimate, coarse, and neutral) are general types, and their typicality varies from one performer to another. While the expletives often comprise one ut­ terance phrase, followed by a rap of the cempala, they also can be con­ joined to the following phrase (10). Although not all Kawi speeches begin with an expletive, the pattern I have outlined is extremely com­ mon. The second characteristic phrase of a Kawi utterance in wayang fol­ lows the expletive opening and might be termed the "addressee iden­ tification" phrase. In examples 6 through 10 above, direct references to the hearer constitute distinct phrases within the utterance. The most important constraint on the use of addressee identification phrases is the signaling of relations between speech act participants. This con­ straint, like the operation of speech registers in modern Bahnese, has several dimensions. It represents a hierarchy of status conceived in vertical terms of "higher" versus "lower", it also has a horizontal di­ mension, a "chne of person" 29 indicating degrees of intimacy or dis­ tance between speakers and hearers. All Bahnese dalangs have in mind this multidimensional image of participant relations as they compose Kawi dialogue (as do, for that matter, performers in other theater forms, e.g., topeng, wayang wong, 19

See Becker and Oka (1977)

146

T H E U S E OF K A W I

and arja, in which Kawi dialogue is used). Within the addressee iden­ tification phrase, different classes of pronouns and titles are combined to indicate various degrees of status relationship. Kinship terms used for older relatives [lbu 'mother', bapa 'father', kakak 'older sibling', paman 'uncle') all show respect, without necessarily implying a literal km relation, when used in wayang. Conversely, kinship terms used for younger relatives {yayi 'younger sibling', putu 'grandchild', nanak or ταιέ 'child') indicate familiarity and the speaker's higher status when used in wayang. In the standard, tenderly romantic scenes between lovers, when a hero is seducing a princess or leaving his wife for the battlefield, the male character always addresses his wife or lover as yayi, while she uses kakak to address him. The effect of such language is to combine emotion with the inherent asymmetry of the siblingstyled bond. Status titles are another component of the addressee identification phrase. As in other Indonesian languages, titles for second person ad­ dress are more polite than references using only pronouns. The hon­ orific mgamka/ingandika prefaced to kinship terms (1) is more defer­ ential than the use of kinship terms alone. To show extreme deference, a character will use rahadyan 'noble highness' to a king and paduka betara 'the foot of the divine' to a god. Yudistira, for instance, uses ταhadyan when speaking to his older relative Kresna and receives the slightly less deferential but still honorific ηαιέηάια 'king' from him, since Yudistira is a king in his own right. Two noblemen who have )ust met will use the even less deferential d&wa 'lord', which is polite but not excessively honorific. When addressing a personage of the ratu group it is always impolite to use his or her name without accompa­ nying title, for this purpose, natha and niepa (from Sanskrit words for 'king') become titles. 11. Uduh I nanak niepa sn Padmanaba I 6nak mendek kita rumuhun I hana patakwon mrang bapanta I Oh/ child, noble king Padmanaba/ be pleased to bow down a mo­ ment/ your father has a question. In this speech the highest deity, Siwa (Betara Guru), addresses Kresna as a subordinate "child" (using uduh and nanak) but also gives him his due as "king." Moving from speech that "elevates" (in Bahnese nymggihang) the hearer to that which "lowers" (nebah), we find wayang characters ad­ dressing their courtiers as prawira 'hero', their servants with the titles 147

DISCOURSE OF BALINESE SHADOW THEATER

of ceraka or kalaganta (8), and their armies as bala p6ka or wadwa p6ka 'warrior servants'. In this range are encountered the second-person pronouns of Old Javanese kita, which is not polite when used without titles (10), and ko, used in insults and when speaking to subhuman orders such as demons and animals. Another grammatical category, personal articles, also signals statusand-intimacy variation. For divinities, sang hyang or hyang is used before a proper name (7), for royalty, sang or sang natha are used, the latter being polite, the former more familiar. The combination of different classes of terms of reference in the expletive and addressee identification phrases creates fine distinctions in the quality of the utterance 12. Ah/kita Arjuna/ Ha/ you [there], Arjuna. 13 Uduh / kita nanak sang Ariuna / Oh/ you, child, Arjuna. These two examples are both non-honorific speech directed to Arjuna. Any ear accustomed to the language of the dalang, however, immediately recognizes that while fragment 13 is an affectionately intimate salute from an older friend or relative, fragment 12 either is spoken in anger or is the aggressively familiar challenge of an enemy. Thus far we have considered two standard types of phrases that typically are components of a single utterance of Kawi dialogue in wayang. While the initial expletive and addressee identification phrases do not always occur in the dialogue, they are nonetheless characteristic, particularly of those scenes that emphasize dialogue over action, that is, the initial meeting scene that normally opens the play and any meeting scene that occurs later in the performance. The opening scene, called pagun(d)eman, pagmeman, paparuman, pagostian, or patangkilan, is crucial to the introduction of major characters, to laying out the central plot-related issues, to expounding religious or ethical philosophy, and to securing the audience's interest in the outcome of the story. During this scene, moreover, the performer may still be improvising the plot outline as he performs, and he needs to be especially attentive to audience reactions to the style and content of the performance A great deal depends on the language skills of the dalang, who composes his dialogue with concentration as characters are first introduced and then become involved with discussion, deliberation, and debate. 148

THE USE OF KAWI

It is in this context that the phrase-focused structure of the dialogue, and the two standard types of phrases discussed above, come into their own. By measuring his dialogue with pauses during which the cempala sounds, the dalang paces his thought and speech processes. By opening a dialogic utterance with the stereotypic sets of expletives and identi­ fiers described above, he can provide expected information about the characters on stage even as he mentally composes the next phrases containing the core propositions of the utterance. Both the prosodic units within the structure of the dialogue and the formulaic nature of the linguistic entities filling those units support the smooth, sure im­ provisation of dialogue. These verbal patterns bear striking resemblance to some of the char­ acteristic features of an oral culture's verbal art, as outlined in the work of Walter }. Ong.30 The most generally defined and inclusive of these features—stereoptypic or formulaic expression—has appeared in the phrasal units discussed above. A special case of formulaic expres­ sion is epithetic identification, which serves to distinguish between different classes of characters and stock types. The forms of secondperson address illustrated earlier are epithetic, as are the first-person references in the following sentences. 14. Ahum kita Ιαιέ sang Dmyodana I apa kita dateng maiek lawan sua τάπια ya bapa I anak sigia wistaiyakena I Ahum, child, you Duryodana / why is it you have come to meet with me (literally 'he the elder, he father') / quickly, let it be told. 15. Ih kita sang danuja I syapa aranta I lah pada waiah lawan gatimkanang Paitha I Ih you, demon / what's your name / come on, tell it to me (literally 'that Partha'). 16. Yan mangkana namya naiaiya Kasawa I In that case, I (literally 'noble Kesawa', i.e., Kresna) make my obeisance. Epithetic identification in speeches like the ones above involves a set of references for each major wayang character. Kresna, for example, can be called naiaiya K6sawa, sn Padmanaba, sang nathang Yadu, and sn Hanmurti, among other terms. The extensive use of epithetic 30

See Ong (1977 102-117), from whom I have borrowed for this discussion

149

D I S C O U R S E OF BALINESE SHADOW THEATER

identification in wayang is related to another feature characteristic of oral-aural noetics: the establishing of prominent, "heavy" or "ceremonial" characters or character types. The larger-than-life, epic dimensions of such characters are a pivot around which both old stones and newly created plots revolve. The importance of "heavy" characters in the super-epic "text" that includes all wayang parwa plays should not be underestimated. When Kresna or a raksasa (demon) appears on screen, his speech and actions are interpreted by the audience against a background of expectations resulting from countless previous appearances—in other places, in past generations, through many other performers—of the same individuals or character types. Thus the performer, like the "oral" poet, need not recreate the personalities and past exploits of his characters in each performance. Another feature of oral cultures' verbal art depends upon the presence of such "heavy" characters, for it is the actions of these figures that become wayang "history" as they are later formulated and retold. "Ceremonial" characters, then, condition a "ceremonial" construing of the past A fifth important element in oral tradition is the tendency toward thematic standardization, the organization of subject matter in terms of a regularized range of motifs. This characteristic appears in the language of wayang in the form of typical speech acts that are commonly woven into the Kawi dialogue. For instance, examples (14) and (15) above both involve requests for information, which in the dalang's language involves the terms wistaryakena, warah(akena), and weruh(akena), all meaning 'tell about' or 'inform'. Wayang characters frequently ask about the reasons for each others' actions and the causes of particular effects, thus, the terms kadiang punapa 'how is it that. . .' (6) and nimittanyan 'the reason t h a t . . .' (6) are regularly heard. Other actions appearing in formulary style are greetings, taking the form of obeisance, as in (16), and the asking of forgiveness, as in the following utterance. 17 Smggih / pakulun paduka Betaia pasang tabapatik Betaia I Yes / at the foot of the divine the slave of the divine asks forgiveness. Similarly conventional are the speeches requesting and giving permission for departure, in nearly every performance there is at least one exchange of mamwit 'request permission to leave' and lumakwa or lumans 'go, depart' (9). Agreement and concurrence are also expressed in 150

T H E U S E OF K A W I

standard forms, such as 6nak 'fine', mabener 'right, truly', saturan 'agreement', or tuhu kadi hngta 'just as you say'. All the examples above involve the verbal formulation of standard thematic material, however, these sorts of themes result more from the formal necessities of drama—the requirement that characters must meet and then separate, agree or disagree, ask questions and an­ swer them—than from wayang's substantive characteristics. Yet when we look at the subject matter of the play, we also encounter standard thematic expression. Right-minded heroes and heroines, for instance, always have as their goal the general well-being of creation at large 18. . . . lamakanA amangguhakena kunang kerta nikang bhuwana/ . . . in order to achieve peace in the world. 19. Pira I kunang lawas angeitiaken watekmg suralaya I lamakαηέ amangguh subiksa I How long / have the heavenly ones strived / to attain prosperity? One of the primary themes of the wayang parwa cycle of stories is recurring and inherent conflict within a single descent group, and, as their teacher Drona reminds the young Duryodana, 20. Koiawa bangsa tunggal lawan Pandawa I sama-sama Kuru wangsaI The Korawas are one race with the Pandawas / they are both of the Kuru race. This particular theme provides the material that attests to the pres­ ence in wayang language of another typically oral-cultural feature, the predominance of "agonistic" expression, what Ong calls the "cultiva­ tion of praise and vituperation." Within this range lie a host of formu­ lary threats and challenges, what scholars of oral tradition call "flitmg". 21. Kita wadwa p&ka Pandawa I mati tan uiung kita I kita mati I You, Pandawa army I you can't but die / you shall die! 22. Kiesna I Tonton kadang warganta mati I Kresna / watch your relatives die! 151

D I S C O U R S E OF BALINESE SHADOW THEATER

23. Ngong tan alah dining gem I ngong tan alah dining senjata I ngong tan alah dining sarwa astianta I You can't beat me with fire / can't beat me with weapons / can't beat me with all your arrows1 Characters accost each other in battle with demands for identification (15) and exhortations to "show your courage" [cangkah-cumangkah). Equally profuse are expressions of praise, for instance when strangers visit the Pandawas at their court of Amarta. 24. Yan mmgkini da wa I

nnasa tan hana mada kawibawamrang Pan-

And so one feels there exists no one to equal the greatness of the Pandawas. 25. Narindra sampun smanggah wruhing tatwing an I The king is considered to be wise in the teachings of great books. One final example of the typically oral-cultural aspects of the dalang's composition in Kawi is its copiousness. As Ong (1977:114) notes: Oral performance demands flow: though pauses may be used effectively in public speaking, hesitancy is always a vice.... An orator must have at his immediate command an abundance of things to say and of ways to say them, copia in Latin rhetoric, 'flow', 'abundance'. When he composes dialogue in Kawi, the Bahnese dalang is not in precisely the same position in relation to his audience as Lord's classical bard is to his. The spectators/auditors will have a chance to capture the meaning of the dialogue when it is paraphrased in Bahnese by the parekan. Thus, copiousness in the form of synonyms, parallelisms, repetition, syntactic redundancy, and so forth is not necessarily for the benefit of an audience that may have missed part of the utterance content. Yet copiousness, redundancy, and "flowery" speech are essential for the performer himself, to provide time to collect his thoughts and to compensate for the hesitancy and confusion that can occur during the improvisation of dialogue and action. The performer's aim is never to be entirely "at a loss for words." Copiousness in the dalang's Kawi can take several forms. The phrases gati nikanang, lwir mkang, and maka sadnyamng are used be152

T H E U S E OF K A W I

fore personal pronouns in the sense of "situation," "condition," or "opinion." 26... . mmitta antyan suka maka sadnyanmg hulun I ... for greatly pleased am I. A number of other expressions serve as "fillers" when the dalang is marking time while thinking up the next speech or scene within the play. At such times, a puppet's speech will be laced with 6nak 'fine', sakadi mganika 'as was said', yayatiki and kaya mangk& 'in these circumstances', yan mangkana 'in that case', and wyatara 'more or less'. These and other fillers are heard contributing "flow," but not necessarily content, to the dialogue. In some cases, the performer must compose a whole utterance with filler phrases, either because he cannot recall the exact Kawi vocabulary for what he wants to say or because he needs time to consider some other aspect of the performance. The phrases wus keita lugra 'go ahead and speak' or 'it is done', yan tan hana salah 'if there be no mistake', lamakan& sigia 'with all due speed', haywa sangsaya 'do not be concerned', and haywa tan wiwaka 'do not be inattentive/careless' are thus used in a wide range of situations and with various meanings. The dalang who is insufficiently versed in Kawi, it is said, will rely too heavily on such "empty" expressions instead of formulating utterances that more literally express what he is trying to say. As mentioned earlier, the dalang's tendency for copiousness in the Kawi dialogue is most obvious in the first part of the play (the meeting scene, pagun[d]eman). The introductory speeches in this scene are invariably ornate and indeed are given scant attention, since the audience is likely to be still settling itself around the screen and the performer still "warming up" to his tale. The following is an example of "copiousness" in a Kawi speech, in which each phrase is carefully and slowly pronounced, but the sum of the parts is little more than the requisite politeness and formality. 27. Smggih I mganika lbu / domkang kadiang punapa I katunan totya kayong dangu Imenawa /hana wetekamang anta weidaya I kunang wistaryakena I lawan y6ki Natha Kusawa I mangkana saturan Sn Janaidana I Yes / honored mother / on what account / [what] trouble as in the past / the reason I [if] there be some matter in [your] heart / so let it be told / to King K6sawa here I thus the words of noble Janardana. 153

D I S C O U R S E OF BALINESE SHADOW THEATER

The "content" of an utterance such as the one above has nothing to do with the particular story being played, it is extremely important, however, in setting the formal mood of the court scene and fulfilling expectations regarding the gentility of the characters portrayed (in this case, Kresna and D6wi Kunti). By examining several types of verbal patterns in the dalang's Kawi dialogue, we have established the presence of certain features characteristic of the art of the oral poet m largely nonhterate societies. Generalizing about the use of irregular prosodic structure with the cempala, and about composition using stereotyped or formulaic language, we see that the dalang is affected by many of the concerns affecting any oral poet- the need to measure verbal flow, to find strategic pauses, to compose while still speaking, to fulfill expectations about typical characters and their relationships, to build a text with standard themes, to have many ways of saying the same thing, and, no matter what exigencies prevail, to be able to keep on speaking. The sorts of constraints outlined here apply equally to all dalangs and are among the crucial aspects of his art as a whole. If the performer used a written script, or if he in fact memorized lines of dialogue from manuscripts, none of these factors would come into play. In the dalang's oral verbal art, there is a noetic unity that underlies the variable, transitory differences between the performances of countless individuals through time. Form and Variation. After studying the sources for the dalang's dialogue and examining the patterns underlying its composition in performance, one realizes the fundamental variability of Kawi in this tradition. Earlier in this chapter I described the influences on and training of a boy growing up in a family of wayang performers Yet not all dalangs, by any means, accede to the title through hereditary-occupational training. Some puppet masters begin to perform relatively late in life, after illness, misfortune, or spiritual exploration convinces them to take up the profession. Others study padalangan after careers in other, perhaps related, fields, such as dance-drama. One successful dalang in Bah today was once a roving seller of medicines in town squares and at festivals connected with temple ceremonies; so remarkable were his vocal powers and ability to "think on his feet" that his wayang performances quickly became popular. Just as there is no one way to become a dalang, then, so the method of learning Old Javanese language differs from performer to performer The texts the performer might read in the course of training, the style 154

THE USE OF KAWI

of dramatic Kawi prevalent in the area where he lives, and the language used by certain performers he takes as his models31—all these factors influence the dalang's particular use of Kawi. Thus, Kawi forms and styles vary geographically, temporally, and individually, and it is not always possible to talk about the Kawi of wayang or reasonable to expect to locate a stable group of forms used by all performers, or even consistently by a single performer. The Old Javanese used in wayang, then, is not identical with the language of any one Old Javanese text that has come down to us. While most dalangs use a style that is markedly similar to the language of the parwas and, as stated earlier, while these texts can be considered the closest written model for the dialogue of the dalang, there are nevertheless many differences between the forms of these tenth-century works and the performance of twentieth-century dalangs. Some of these differences can be traced to the influence of later forms of Javanese language and literature. For example, some dalangs employ forms that are generally associated with "Middle" Javanese, the language of the kidung, the use of mgsun and manna as first-person pronouns is characteristic of this style, and these forms are commonly heard in the northern Bah area. Another fairly standard Middle Javanese form, shown in sentences 10 and 23 above, is ngong (parwa texts have ngwang), also a first-person pronoun. Certain Middle Javanese forms that Zoetmulder (1974:441-443) cites as common in kidung language are not used by Bahnese dalangs, for instance, one does not hear ayo for Old Javanese haywa 'do not', or noiana for Old Javanese tan hana 'there is not'. The influence of Middle Javanese forms, it seems, is limited and selective. Grammatical categories well attested in the language of the parwas are not always fully represented in the dalang's Kawi. The elaborate deictic system described by Becker and Oka (1977) is, in general, greatly reduced: instead of the full range of six demonstrative pronouns that locate objects and events along the chne of person,32 we normally find in the dalang's usage only lki and ikana(ng). The deictic paradigm has collapsed to a single opposition. 31 Among the Yugoslav bards, Lord (1976 24) noted, "the singer imitates the techniques of composition of his master or masters rather than particular songs " 32 These six demonstratives are lki 'this close to me', lku 'that close to you', ika 'that yonder', ika 'that relating to speaker which happened in the past', iko 'that relating to hearer which happened in the past', and lka/ikana 'that relating to neither speaker nor hearer which happened in the past'

155

D I S C O U R S E OF BALINESE SHADOW THEATER

As one might predict, some of the syntactic variation between the language of the Old Javanese parwa and the dalang's Kawi is condi­ tioned by the structure of Bahnese itself. This is certainly true in the case of the "areahs" suffixes -a and -en, which indicate the potential­ ity, desirability, or future status of an action [-a is an actor-focus form, while -en is goal-focused, the so-called "passive" form). Some (but by no means all) dalangs use the -a form like the Bahnese verbal -a, which is a third-person agentive marker (reduced from ία), in utterances that describe an action that is clearly in the past. 28.

. . . rumuhun formerly

kmkmakena attempt P3

swijakarya ceremony

η in

tengahmg middle the

alas forest . . . formerly when they strove to conduct a ceremony in the mid­ dle of the forest. . . Similarly, in narrative such as the following, some performers use wainana instead of the grammatically precise wamanen. 29

Warnana (wamanen) be told

η at

wijil come out

ua P3

sang (title)

maha great

prabhu king To be told now, the appearance of the great king . . . Variation like that in the two examples above is sometimes pointed out and criticized by artists with the greatest knowledge of the classi­ cal Old Javanese texts such as the parwas. Yet these same performers may themselves use contrasting forms of, say, active verbal prefixes. In parwa language, these include ma-, mang-, a-, and ang-, but in modern Bahnese only a homorganic nasal initial (e.g., jemak 'take' becomes the transitive verb nyemak) is involved, while ma- functions as a stative prefix. Thus, in the language of the dalang are heard verbal forms such as ngmng 'accompany', with affixation Bahnese-style, instead of the angmng of the parwa texts. In a converse grammatical condition­ ing, verb roots that are not part of the Old Javanese lexicon can be "Kawified" when they are combined with appropriate affixes. Thus, kaplak, a Bahnese verb meaning 'strike, slap', is made into a transitive Kawi verb with -aken suffix. 156

T H E U S E OF K A W I

30. Smggih I angaplakaken ki&ta swahaia I Yes I striking up the chariot on the attack. Another example of the same phenomenon is mcep, in Bahnese a verb root meaning 'put in order, make understood', which is given Kawi form as demonstrated in the following utterance. 31. Sigia naiaiya K6sawa angmcepaken a]i tenung I Immediately noble Kresna interpreted the wisdom of the Tenung book. In outlining some of the variation between the oldest known forms of archaic Javanese and the Kawi used by the Bahnese dalang today, we must remember that not all performers attest to identical usages. While extremely knowledgeable Bahnese literati may be quick to point out "aberrations" from the language of parwa and kakawin texts, among many wayang performers there may be no absolute distinction between the language extant in written form and the language inherited through centuries of oral performance. In general, the Bahnese dalangs do not worry about compartmentalizing "Old Javanese," "Middle Javanese," and "Kawi-Bah" forms, preferring to consider all commonly accepted usages as part of the great Kawi realm. One reason the Bahnese dalang is not bothered by philological inconsistency is that he regards Kawi language as a performance strategy rather than as an object, as a mode of communication rather than as a fixed-form, silent text. His job is to evoke the world of the past through speaking it; his puppets engage in dialogue, not quotation. Like the oral poets described by Lord (1976:29), the dalang's aim is the "preservation of tradition by the constant re-creation of it." To achieve this, he cannot afford to dispense with the well-tested styles of composition in performance, or to disregard the forms and techniques of utterance composition handed down from one generation of dalangs to another. To the extent that the parwa manuscripts continue to be a primary resource in the dalang's training, the most ancient textual forms will continue to be one parameter of acceptable performance practice. Yet the ultimate criteria of acceptability are the norms of oral composition in performance, which can be conditioned by locational, individual, and non-Kawi linguistic variation. That the Bahnese dalang speaks in Kawi is unusual; that the form of his Kawi is so close to that of texts written a millennium ago is remarkable. The most interesting aspect 157

D I S C O U R S E OF BALINESE SHADOW THEATER

of all, however, is the noetic complexity, both in transmission and in form, with both scribal and oral/aural underpinnings, of the vigorous Kawi voices that continue to be heard in wayang performances. Narration The narration of the Bahnese dalang is also m Kawi, and it has both structural and dramatic aspects. Dramatic narration is linked to the particular story (lampahan) being played, and its content is specifically related to characters and events in that story. Structural narration is more predictably applied at standard points in all stories, and it is linked to the abstract notion of plot (also called lampahan) with its structural divisions. Whether the narration is structural or dramatic, it is expressed in Kawi and has distinct acoustic features. The style used by the dalang for narration is audibly different from that of any of the Kawi-speaking voices of the puppets; in general, the voice comes from deep within the chest, has a throaty yet constricted resonance, is projected with full force and volume, and uses a wider scale of tones than do any of the puppet voices. According to the Dhaima Pawayangan, the dalang's voice, which can "say all that can be said" (sakawuwus-wuwus), results from a "meeting in the heart" of a fivefold deity representing the four cardinal directions plus the center Matemu ring pupusuh, metu sabda nga, Mahadowa wit mg sabda, Wisnu pukuh mg sabda, Brahma madhya mng sabda, Iswara tungtung mg sabda, Bhatara Guru mula mng sabda. As a result of the union in the heart, speech comes forth. Mahadewa is the stem of speech, Wisnu is the base of speech, Brahma is the middle of speech, Iswara is the culmination of speech, Bhatara Guru is the root of speech. (Hooykaas 1973 18-19) As described in the discussion of mantras above, the preparation for performance using mantras concentrates the dalang's mind on a similar condensation of cosmic dimensions into one forceful, empowering distillation. In the dalang's case, this essential unity is called Sanghyang Kawi, in the Kayonan mantra, the dalang attains this deified essence of linguistic activity just before he begins the first audibly spoken material of the performance. Many dalangs state that it is Sanghyang Kawi who is speaking through them in narrative interludes, and 158

THE USE OF KAWI

that the narrative voice belongs to the Kayonan figure. 33 Throughout pre-performance preparation, then, the dalang is "working up t o " the production of this astoundingly strong voice, which gasps and groans in drawn-out, contorted tones—what the texts call "screaming and moaning" (see Hooykaas 1973:16-17) Structural Narration. The period of mantra recitation while the dalang is seated before the screen, making offerings and opening his puppet box, may last an hour or more before the first audible speech occurs. Eventually, as the dalang sings and then begins to chant in Kawi, he arrays on the screen those puppets with roles in the first scene of the play. While the silently uttered mantras are part of a gradual transition from the ordinary world and ordinary awareness to the dramatic world and a state of enhancement and inspiration, the initial audible vocalizing marks a qualitative shift in the progress of the performance. Following an elaborate, sung introduction in the tetandakan style (which will be discussed in more detail below) is the first and extremely important example of structural narration, the pangaksama or apology. This is signaled by a change in the gender music from that used for the sung introduction, and also by the quality of the dalang's voice, which begins "screaming and moaning" instead of singing. It is this voice that is characteristically narrational. The pangaksama commonly consists of one phrase, Dadia ta pmta or Dadia ta pira ta pmten 'May the request be fulfilled', 'May it be as requested', 34 which alludes to the granting of forgiveness and the performer's request that his play be unhindered in its enactment. 3 5 The pangaksama is followed immediately, in the same vigorous style of almost trance-like groaning, by the second form of structural narration, 33

That the dalang's narration is identified with the "voice" of Sanghyang Kawi, and symbolized by the Kayonan puppet, is also indicated in the few written records of lampahans collected by the Gedong Kirtya and in transcriptions of tape recordings that I commissioned from dalangs in 1977-1978 In these texts, the narrative speeches are labeled "Kayonan" or "Kawi Suara" (Kawivoice) instead of "dalang," as one might have expected 34 See the third section of Chapter 3 for discussion of apologies in the manuscript tradition The "dadia ta" figure would seem to have a parallel function to the "awighnam astu" phrase that is generally found in lontars 35 The longer pangaksama traditionally characteristic of Bahnese wayang is represented in the Preface to this book, this apology also includes a request for forgiveness 159

D I S C O U R S E OF BALINESE SHADOW THEATER

the panyacahparwa 'enumeration of the parwa' 3 6 This entity is a for­ mulaic invocation of Sang Hyang Astadasa Parwa 'Divine Eighteen Parwa', the books of the original written Mahabharata It includes ref­ erence to Byasa (Kresna Dwipayana), the sage "compiler" of the epic, and hails the emergence of two special divine principles a deified "puppet" that controls movement and action and a "void" that in­ spires individual plots from withm the entire epic sweep The next narrative episode is the pamahbah 'introduction, explana­ tion', a summary of the point at which the lampahan to be portrayed begins The performer, throughout the panyacahparwa and pamahbah, is narrowing down to a single moment of the epic span, moving from the invocation of abstract, remote forces underlying the performance to the specifics of time, place, and character While there is no overall term that all performers use for the three components of narrative introduction (pangaksama, panyacahparwa, pamahbah), they are sometimes referred to as the pangelengkara, which can be glossed as 'place of embellishments' or 'area designating a border' 3 7 In the past (and in some areas as recently as one generation ago), this complex of structural narrative forms was much longer and included the enumeration of the names of all the parwas as part of the panyacahparwa 3S A representative version of the introductory narrative elements of wayang parwa, taken from a recorded performance, is given in example 32 below It includes the pangaksama, the panyacahparwa, and part of the pamahbah 32 Dadia tapmta/ Cantanen kunangpwa samangk6/ya taripalangkepanira Sang Hyang Astadasa Parwa/ Ya ta nniket άέ resi Kresna Dwipayana/ Μηιΐ Sang Hyang Rmggit amolah coral kadi gelap sumarasah anusup mg randu parajamanala/ Ya ta rumijil Sang Hyang Sunyantara amunggel kunang canta/ Warnanen η wijd ira [pamahbah begins] Sang Kunti Bho]aputraka/ tan 36

This component of the dalang's opening invocations is also found in the last section of the text that is included in the Preface to this study Cacah is a root meaning 'to separate, chop up' 37 The root lengkara may originate from the Sanskrit alamkara 'poetic em­ bellishments', for a brief, useful survey of Indian poetics, see chapter 3 of Di mocketal (1974 115-144) 38 An interesting textual correlate of this enumeration is found in the San sknt stanzas that open the Adiparwa Here, both the name of Kresna Dwipa yana and the names of the parwas are mentioned 160

T H E U S E OF KAWI

Ιέη aiepat lawan nararya sn Padmanaba/ tan sah nararya Paitha/ hana η samipamra/ saw6tnmg lawas sua ngaitiyaken jagatraya May the request be granted. Now let it be told, that entirety of the Divine Eighteen Parwas, which were composed by the sage Kresna Dwipayana. The Divine Puppet moves and acts, like lightning everywhere illuminating the great world. The Divine Void comes forth, dividing up the story. We tell of the appearance of [here be­ gins the narration of the story to be played] Queen Kunti, together with noble Lord Kresna, and also with noble Arjuna, close at hand, since they had long been striving in the three worlds . . . Following the pangelengkara section, and before the beginning of character dialogue, there is often another narrative element that both literally and rhetorically signals a "border." This is the pangalang, a verbatim-repeated song or speech with which the dalang prefaces the opening scene. The pangalang is a loosely designated category, some­ times referring to any of the introductory elements and sometimes to a particular standard opening song or speech that varies from one per­ former to another. In South Bah, many dalangs use a pattern that be­ gins 33. Sunduk samita Indatan katon Ι αηίέη Imaweidayang

ati

Words and gestures do not reveal the inner heart. . . This pangalang example is a preface to the dialogue of noble characters (ratu) and suggests their enlightened and refined qualities.39 Together, all the introductory structural narrative elements form an elaborate verbal network establishing the setting of the drama. The pangelengkara has to do with the relation between the performer and a supra-textual realm where specifically literary entities are seen as di­ vine and inspirational. This "heavy" boundary (in that it has several parts and is lengthy) is not directed toward the human audience, but does bring forth the performer in a demonstration of verbal skill, which is also part of the meaning of lengkara. He is no longer involved with unheard, personal preparation in the realm of the "pure mind" at the center of cosmic space, yet he is still under the sway of a deified and mythic force that can guide him in the selection of the play. His voice 39 Panasar (clown-servant) characters have their own pangalang in the form of special songs and instrumental pieces. See the pangalang given in Hooykaas (1973 192-193)

161

D I S C O U R S E OF BALINESE SHADOW THEATER

is strange and otherworldly, as if it were the channel for unseen speakers. The words are repeated formulas (as are the mantras), part of the meta-text to which all wayang parwa plays belong, part of the larger epic context in which the present drama is but one of many episodes. While recitation of the mantras involves the dalang in purely personal communication with divine auditors, however, the pangelengkara, in enlivening the puppets, sets the stage for those presences to move and act {amolah caia) before the audience. The epic events and characters thus set forth [lampah] to play the story (lampahan). Another noteworthy point about the pangelengkara and pangalang as "borders" is that they signify the progression from exact verbal repetition to freely composed puppet speech. This is an elaborate process with several stages, during which the dalang always heralds the "original" text (the Mahabharata) and its compiler (Byasa|, bringing these powerful symbols of the literary and cultural past to bear on his own dramatic creation. He is aware of a written textual predecessor being linked to the orally displayed text that he composes. The introductory narration thus echoes some of the noetic tensions that shape the interdependence in Bali of written and sounded literary discourse. Dramatic Narration. The second type of narration dalangs use can be summed up as "dramatic," since when it occurs depends on the particularities of the story being performed. Some of this narration takes place within scenes, serving to emphasize a character, heighten a mood, introduce a speech, or simply give the dalang a chance to collect his thoughts or search for puppets. Some dramatic narration occurs between scenes, and as such it will be discussed in terms of constraints on plot later in this study (Chapter 6). No matter what the function of the dramatic narrative, however, it is in Kawi, and it tends to be uttered in the same "screaming and moaning" voice as the pangelengkara sections. The manner in which the dalang composes dramatic narrative is basically the same as the manner in which he composes dialogue. The narration is made up of formulaic entities, is measured by cempala rapping, and revolves around a standard core of thematic material. The theme is generally descriptive of either the psychological states of characters (such as anger, sorrow, agreement, surprise) or actions (arrival, departure, preparation, listening, replying, and so forth). The statement of the theme in narrative utterances is prefaced by a set of terms that relate the theme to its dramatic context. The following list of such terms is representative, though not all-inclusive162

THE USE OF KAWI

bawisiati an wawu agehs an tedun an wijil cantanen warnanen byatita saksana kancit

next, then, following upon just then, next, just as immediately, quickly upon the descent of upon the appearance/coming forth of let it be told let it be described formerly, in the past in the wink of an eye straightaway

As is clear from the terms in the list above, the dramatic narrative thus includes logical connectives linking one action or element of the plot to another. These introductory words are combined with formulas expressing thematic material and inserted at various points in the dia­ logue or action. 34. An wawu I kaienga de mrang natha K6sawa I And just as/ this was heard by king Kresna. 35. Bawisiati η lampah ira sang Dananjaya I Next, the setting forth of Arjuna. 36. Kancit prapta sang maharaja Astma ring pasraman I samangkana Straightaway the king of Astina arrived at the hermitage/ thus. 37. Saksana pejah I lkanang wadwa danawa I In the wink of an eye, dead/ was the demon army. 38 Agehs I gatimrang sn Dharmaputra I Quickly/ thus the noble son of Dharma. As the examples above suggest, the Kawi narration is often fragmen­ tary or peripheral in terms of its relation to context. It is important, however, because of its dramatic function within a scene, and because it allows the dalang to continue his vocal activity while he is busy put­ ting down or taking up puppets or collecting his thoughts before pro­ ceeding with the performance. The narrative interludes do not all follow the pattern indicated above. They may be either succinct or elaborate, depending on the da163

DISCOURSE OF BALINESE SHADOW THEATER

lang's inclination, the necessity of drawing out the narration while characters are shown "traveling," or other contextual factors. Between episodes, narration is generally protracted, while the Kayonan figure dances on the screen to indicate a major scene shift. 39. Warnanen I an tedun I η mangkana pwa ya I gatmikanang I sang Pandawa I aiep ataki-taki ngwangun kaiya I ya mangki cantanen I hana natha sang Mantarabhuwana I ngaian sang Darmadowa I Now let us describe/ the descent/ at that time it was/ the circum­ stances/ the Pandawas/ were preparing to hold a ceremony/ and now let it be told/ there was a king of Mantarabhuwana/ named Darmadowa. This longer narrative speech gives an idea of the dalang's phrase-byphrase composition process, as he switches his orientation from one group of puppets to another and bridges the locational and episodic breaks in the story. In formal terms, it seems, the composition of narration parallels dia­ logue creation. From the standpoint of acoustic features, however, the narrational and dialogic styles are easily distinguishable. In rhetorical terms too, narrative differs from dialogue. The "narrator" of wayang is an anonymous figure commenting from outside the immediate dra­ matic action, and also quite apart from the point of view of the audi­ ence. This Kawi-speaking figure is in fact the dalang's dramatic per­ sona, under the influence and inspiration of the "divine Kawi-voice." In its symbolic association with the Kayonan figure, and with its unique vocal style, the dalang's narration is the third distinct type of Kawi "voice" (along with mantras and dialogue) created by the dalang in wayang performance. Returning to the Text Among the mantras that the dalang recites silently before the play begins may occur one "soul-protector" (pangraksa jiwa) of particular interest when trying to understand the dalang's use of Kawi. Pukulun Sang Hyang Panca Pandawa, umor mg akasa, Nakula Saduwa ring takepan kahh, Arjuna ring lontar, Dharma-tanaya ring lontar sastra, Bhima ring belat nmg lontar, 164

T H E U S E OF K A W I

Unp apageh, hla nnwighna parama sakti, OM AM MAM UM OM My lords the divine Five Pandawas, aloft in the heavens, Nakula and Sad6wa in the two wooden covers [of the lontar], Arjuna in the palm leaf, Yudistira in the written word, Bhima in the cord stringing the leaves together, Unending life, unhindered joy, greatest power, OM AM MAM UM OM.40 This invocation is cited by wayang performers as an illustration of the profound and inseparable bond between shadow theater and written literary tradition No wayang would be complete without "quotations" from the great poems and stories handed down in lontar form Just as the five Pandawa brothers are a perpetual, inspiring, and powerful unity, so is the lontar—with its wooden end pieces, palm leaf strips, writing, and center hole with connecting cord—a unity that assures life for the words of the past, provides aesthetic satisfaction, and bestows special power In this section, I will consider how the dalang directly employs the Kawi texts in his play, constituting the fourth type of Kawi in Bahnese wayang parwa. The two major ways to be discussed can both be characterized as "singing" (since vocal music, it must be recalled, is one of the primary components of Bahnese literary expression), yet each has its own definitive vocal style The bebatuian are songs that have melodies peculiar to the dalang and are sung irrespective of instrumental accompaniment, while tetandakan axe. sung in conjunction with specific instrumental pieces played by the gend6r quartet. It must be emphasized here that not all performers use the same terms for these vocal musical interludes, neither do they necessarily use identical fragments for the same function. Furthermore, there is distinct regional variation in the way vocal music is applied in performance, and different localities tend to use different melodic patterns. However, the usage I describe below reflects widely found practice, and it is part of the wayang parwa tradition handed down between generations of performers in the Sukawati community of dalangs I studied. Bebatuian This class of vocal music includes excerpts from texts that are sung with melodies peculiar to the dalang The term itself 40

The text is from Hooykaas (1973 60), the translation is my own 165

D I S C O U R S E OF BALINESE SHADOW THEATER

comes from an Old Javanese root batur, meaning 'elevated flat surface' or 'stairway', in Bah the word is associated with "pure" [sua] objects and places, such as offerings, temples, mountains, hermitages, and ascetic practice. Bebatman has been glossed as 'a sacred building made of stone with a level top surface' (Panitia Penyusan Ramus Bali 1979:76) and 'certain ways a mantra is pronounced' (van der Tuuk 1897-1912, vol. 4, p. 940).41 While in theory any kakawin, kidung, or parwa text could be a source for bebaturan, in actual practice the kakawins constitute the most significant textual source of the excerpts used. In the major South Bah tradition discussed here, only the parekan (attendant) figures quote from the kakawins using the standard vocal realization for the wirama meters involved. This practice is not part of the bebaturan category, but will be discussed as part of the parekan's linguistic roles (see Chapter 5). While it is beyond the scope of this study to consider the strictly musical aspects of bebaturan usage (melodic and rhythmic contour, transmission, regional styles, and so forth), it is nonetheless important to stress that bebaturan are a musical, and not just a literary or dramatic, element of wayang. The talents required of the dalang include a pleasant singing voice for entertaining the audience and bringing characters to life, bebaturan are obligatory vocal complements to dramatic dialogue and action. One dalang has defined bebaturan as "vocal art which is free, not bound to the genddr tunes, using the Kawi language, and generally taken from kakawin verses."42 The dalang may sing the bebaturan either solo fashion or while the musicians are playing an unrelated melody in the background. The solo bebaturan occur at various points within scenes, but perhaps with greatest frequency as sung prologues to a character's speech. 40. [sung] Kunang sn Anmurti, bupati.. . [spoken] Smggih I mgamka lbu I dAmkanang kadiangpunapa I menawa I hana wetekanyan anta werdaya I kunang wistaryakena I lawan yaki natha KSsawa I mangkana satuian sn Janardana I Thus lord Anmurti, king.. . . 'Yes, your majesty the queen mother, in what manner, and by what cause, do you appear so preoccupied? 41

The second gloss may have to do with the usages of sacred or literary texts in the theater Zoetmulder (1982 225-226) illustrates other meanings of batui 'help, servant, companion' and 'contribution' 42 From a collection of notes on wayang parwa prepared by Dalang I Ketut Madra, Sukawati, in 1977 166

THE USE OF KAWI

Let the matter be related to me, King K6sawa', thus the words of Lord Janardana. Following the first three words of the bebaturan in the example above is a pause, and the fourth word [bupati) is sung on a lengthy falling me­ lodic contour. The dalang from whose performance the example is taken used this musical pattern for many such prologues, no matter what the meaning or syntactic structure of the bebaturan fragment in­ volved. This particular fragment is from the beginning of a play, and it is a standard opening to a patangkilan (court meeting) scene. It is ut­ tered by Kresna, who moves one arm as he talks, while the other pup­ pet characters remain stationary. This type of bebaturan can be called paselah, which roughly means 'something occurring during an unoc­ cupied interval'. The patangkilan scenes, with their more static and formal atmosphere, are prime opportunities for the dalang to embellish dialogue with brief literary excerpts, thereby exhibiting his vocal skill and, simultaneously, his knowledge of classical texts. It is considered desirable that the bebaturan fragment chosen men­ tion the character who is speaking. In the example above, Anmurti is a reference to Kresna, the other names used m example 40, K6sawa and Janardana, also refer to Kresna. Balinese audiences are, through such patterns, accustomed to a variety of names for the most familiar wayang characters Every dalang has a range of bebaturan in his repertoire, each gauged to specific character types and situations. When a noble female char­ acter such as Kunti speaks, the following fragment is appropriate. 41. Ling άένη may at angiegep, cunga . Said the queen, holding and pointing the dagger . . . The next verse fragment can be used to preface the speech of Yudistira, the oldest of the five Pandawa brothers 42. Yatna sn Dharmaputra puhh, agaiawala . .. Alert, the son of Dharma hurried back to the fray . . There are also bebaturan applied to stock character types, such as a vis­ iting king or hermit. The following is an example of such an all-pur­ pose fragment. 43. Duh yogya tan hana salah Ah, correct, there's no mistake . . . 167

D I S C O U R S E OF BALINESE SHADOW THEATER

Paselah-style bebaturan also occur within scenes to establish var­ ious dramatic moods, illustrate character emotion, justify a philosoph­ ical or ethical argument, and intensify activities such as mourning, meditation, or fighting. Before Arjuna releases an arrow in battle, for example, he may sing the following quotation. 44. Dwiyastra Paiupati tang nnegep ginuhya . . The powerful arrow Pasupati he held, consecrating it with a man­ tra . .. [Bharatayuddha 31:23)43 When he is overcome with sorrow or longing, he may sing the follow­ ing verse fragment. 45. Mulat maia sang Arjunasemu kamanusan kasrepan . . . As he gazed, Arjuna appeared to be touched with pity . . . [Bharatayuddha 10:12) These examples (44 and 45) each represent one line [acank) of a fourline kakawin verse; they are also heard in more fragmentary form, for example, mulat maia sang Arjuna. This latter form might occur by way of a signal to the dalang's assistant [ι tututan), if Arjuna were not actually on screen while the fragment was being sung, the assistant would be expected to have the right puppet ready when the dalang reached for it. In contrast to the paselah type of bebaturan, which are normally un­ accompanied and relatively brief, pangkat (traveling) bebaturan are lengthier quotations usually sung while gend6r music is played in the background. The term pangkat is the name of gendor pieces played during breaks between scenes while characters are moving from one place to another. Here, too, the literary excerpts are often keyed to spe­ cific characters. 46. Sang Madreyalara syuh rathamra kena de Kama Ιέη sang Yuyutsu/ apwanghh sang Sikandi dhwajamra kawilet Satyaki jrih kang&lanl sang Dhrestadyumna morang malajeng asasran mwang watek wira sanggha/ akwah sang Sura mating rana ratu kadadak έιιηα άέ Suryaputra/ The sons of Madn were distressed, their chariot destroyed by Kama, and similarly Yuyutsu; 43

All the quotations from the Kakawin Bharatayuddha in this discussion are from the edition of Wirjosuparto (1968) 168

T H E U S E OF K A W I

Sikandi was in disorder, his flag entangled, and Satyaki was in fear, Dreshtadyumna, in confusion, fled, dispersing with a group of warriors, Many were the heroes who died on the field, and kings annihilated by the sun's child, Kama (Bharatayuddha 30 3) 47 Bhidang Saratiiaya katara hanja-hanja/ tendas kawandha bhuia pada pupu mwang angga/ bhutadi raksasa sahayudha kalarupa/ muntab geseng ndah amangan kunapangmum rah/ Other forms which emerged from that horrifying arrow were goblins, Heads, headless bodies, hands, feet and other body parts, And also demons and monsters holding weapons, in the shape of dread Kala, Spitting fire, eating corpses, and drinking blood [Bharatayuddha 31 3) Example 46 given here was used for the traveling intervals of the Pandawa twins, Nakula and Sadewa, Madri's sons, while example 47 is "Delem's bebaturan," sung to introduce the chief clown who accompanies the Korawas or other antihero characters The dalang sings these bebaturan while the genders play melodically differing tunes Each example is a complete four-line kakawin verse [pada], however, one line or part of a line may occur elsewhere as part of a paselah-style bebaturan As the examples presented thus far have suggested, many bebaturan are to be found in the kakawin Bharatayuddha, which is a primary text for all wayang performers The core stories of the original written epic underlie all the later, orally transmitted creations of the productive wayang tradition Other texts from which bebaturan are taken include the kakawins Ramayana, Ar]unawiwaha, Bhomakavya, and Sutasoma, and the didactic prose moral/ethical manual, the Sarasamuccaya 44 In the rich, multiform texture of voices and sounds that is the aurally experienced shadow play, bebaturan are heard as echoes of the past, textual confirmation that the dalang is playing what is known and what is true, according to preserved written authority Although the 44 The manuscripts that dalangs are likely to study and quote from in their plays often are those most popular in literary study groups See Robson (1972) for a discussion of the current popularity of certain texts among Bahnese literati

169

D I S C O U R S E OF BALINESE SHADOW THEATER

audience for the most part does not know the sources or literal mean­ ings of the individual fragments, their aesthetic role as part of wayang's musical texture, and their collective significance as one of the requi­ site Kawi "voices," is clear Any dalang who did not apply bebaturan vocal embellishments would be open to criticism as unread or shallow, and his play would sound insufficiently wayah—a Bahnese term meaning 'ripe' or 'old', which is also used for things remote, elevated, and ancient, and which can also indicate learned and literary qualities From the standpoint of the dalang, bebaturan are among the most traditional and revered components of his art Melodic contours, as well as the verses themselves, are handed down from experienced to novice performers Depending on the skill, experience, and literary knowledge of the performer, of course, the way these "quotations" are used varies Some performers use a limited number of bebaturan and apply them similarly in all their stories, others prefer to dip into the lontars looking for new and appropriate excerpts that are then memo­ rized In general, however, dalangs are not concerned to cite "chapter and verse" for the poetry they use and cannot always identify which kakawin the excerpt has been taken from Since the quotations may be part of an orally transmitted technique, and not acquired through ac­ tual visual perusal of manuscripts, there are inevitable differences be­ tween the textual sources and performance practice In the following example, Kresna has sought a meeting with the god Siwa, and he greets the deity with a two-line bebaturan, here compared with the manu­ script source 48 Sn Nilakanta ya ta gelengmg η wekangku Yat kiwala ngmngi sigia jugan gegenta (performance) Sn Nilakanta myatagelengo wekangku/ Yat k&walangmnga cidxa ]uga η gegonta/ (Bharatayuddha 31 9) The blue-throated Siwa will surely be angry with you, my children, If you continue this deceit—take this advice The differences in the two versions are actually minor, having to do primarily with prepositions and the change of cidia 'ruse, trick' to sigia 'quickly', as well as the absence of non-phonemic, orthographically marked features such as vowel quantity It seems that the diver­ gence from written sources that might occur through purely oral trans170

THE USE OF KAWI

mission may be limited by the fact that many performers do return to the texts and, thus, continue to retrieve original forms. One reason dalangs might be motivated to look for new literary excerpts to use as bebaturan is the belief shared by the majority of performers that there should be a connection between the literary quotation and the dramatic moment being portrayed. The "relevance" of bebaturan to the lampahan is, however, more thematic than dramatic. The fragment cited in example 41—"Said the queen, holding and pointing the dagger"—comes from a highly charged scene in the Bharatayuddha when Satyawati tries to kill herself rather than bid farewell to her husband, Salya, who must leave for hopeless battle against the Pandawa forces. As a bebaturan, however, it can be used before any noble female speech, without necessarily implying that the character concerned is about to stab herself. Similarly, "Ddlem's bebaturan," in example 47, can be used to signal the appearance of Delem without necessitating an onslaught of apparitions, the connection here is Delem's association with the puppets of the "left" (i.e., lined up on the dalang's left-hand side), who include many demonic types. Thus, there is flexibility in the bebaturan choices a dalang can make. He can stick to the well-known excerpts or search in the texts for special citations he considers appropriate to a particular lampahan. A constraint on bebaturan choice that performers frequently mention, however, is the chronological order of the two epics—that the heroes of the Ramayana are the ancestors of, and thus temporally prior to, the Pandawas—such that quotations from the Ramayana may be used in wayang parwa stories, but Mahabharata texts should not be cited in wayang ramayana lampahans. This conception of time-ordering between the epics adds another dimension to the patterning of past and present so characteristic of wayang language and stories. These two constraints on bebaturan choice, thematic relevance and epic ordering, reflect the feeling of wayang performers that the traditional texts are points of contact with an epic world activated in performance. As one performer put it, when explaining that the bebaturan can also be calledpanititala' Paniti is from titi, which means 'bridge', or a true path, and that is the same as mti, which means 'wisdom' or the true way, tola means 'the earth we walk on', but can also mean the lontar we read, thus pamtitala means 'following along the path of truth according to the wisdom of the lontar'. Bebaturan are the voices of textual authority, not only literary but also 171

D I S C O U R S E OF BALINESE SHADOW THEATER

inspirational in effect It is even said that dalangs who become 6, katakson* during a performance—specially endowed by unseen mag­ ical forces—will sing bebaturan that they have never studied and do not know Tetandakan This group of vocal illustrations has been defined as "poetry which follows or is tied to the gend6r pieces " The term comes from a Bahnese root tandak, used m many contexts for singing, danc­ ing, and narrative activities 4 6 In wayang parwa, tetandakan (also re­ ferred to as ses6ndon 'songs') are linked to specific mood-setting in­ strumental pieces The dalang's melodic contour always parallels that of the gend6r, although the vocal and instrumental lines may be inter­ woven or offset, one moving slightly ahead of the other The texts used for tetandakan interludes vary greatly, indeed, the textual origins of many commonly used tetandakan are obscure How­ ever, the dalang may employ kakawin, kidung, or parwa excerpts, just as he may for the bebaturan Sometimes the dalang uses the same quo­ tation for both bebaturan and tetandakan functions 49 Kunangprawara Pandawa anggelai ι bajra tiksnalungid Dhananjaya lawan Wrekodara tumut Sikandi haiep I

I

And thus the Pandawa warriors were arrayed m the Sharp Thun­ derbolt formation, Arjuna and Bima, accompanied by Sikandi, were in the fore [Bhaiatayuddha 10 11) As a tetandakan, the example above was performed at the beginning of the play's opening scene (patangkilan) It followed the dalang's intro­ ductory panyacahparwa and was sung according to the gendmg (tune) Pangalang Ratu ('King's Border'), music which begins patangkilan scenes On another occasion, the same kakawin half-stanza was per­ formed as a bebaturan and was sung at the beginning of a verbal con45 The Bahnese notion of taksu 'mysterious or magic power' (see note 6 above) is related within the culture to many altered states, such as trance In the con text of wayang, katakson 'being entered' was explained to me as having three aspects wiguna, such that whatever is performed [guna] is believed, wibawa, such that the audience is silent, taking as authority [wibawa) whatever is said, and wiadnyana, such that the play is full of learning and deep insight [ad nyana) 46 One of the earliest nineteenth century commentators on Bahnese htera ture and culture, Friedench (1876-78), describes an offering dance called tan dak

172

THE USE OF KAWI

frontation between representatives of the Pandawa and Korawa families. Perhaps the most familiar tetandakan of all is the verse of the kakawin Bhaiatayuddha sung as the dalang's first audible voicing, at the very beginning of the play before the pangelengkara. This extremely well-known verse, which also describes a military scene before the start of battle, is called Alas Harum 'Fragrant Forest' after the accompanying gendir tune. It is also termed the Rahma 'Sunrise', after its first words: Rahma tatas kemantyan 'Thereupon day breaks'.47 Just why or how this particular literary quotation became formulaic in the opening of wayang parwa plays is not known, but, as with example 49 above, it is resonant with one of the defining themes of the entire wayang parwa repertoire: the fraternal antagonism of the "great war of the Bharatas." The mood-setting gendor pieces to which tetandakan songs are linked are part of the set instrumental repertoire Alas Harum and Pangalang Ratu have been mentioned, others include Tangis, for sadness; Tuniang, for the appearance of demons; R6bong, for romance and the appearance of a beautiful woman.48 The majority of tetandakan texts used for these pieces have been orally transmitted, and the original written sources, if any existed, have been forgotten or lost. A few dalangs are in possession of inherited lontars that may include bebaturan fragments as well as mantras and short summaries of didactic fables that the dalang can use as "exempla" in his lampahan. Such written sources are not known for many tetandakan, however, and, as the quotations below show, the lyrics most commonly used in southern Bah reveal wide variation between different performers' renderings, as well as distinct similarities that may indicate shared origins no longer traceable. 50. Bujangga anom kamahngan, kailangan genta mwang pustaka, sang kan manangis, ana m&TU mas tumpang sanga-welas mauncal-uncal mas mten komaladi The young student of religion has been robbed, he who has been 47 For text and translation of the complete verse (canto 26, stanza 1 of the Bharatayuddha), see McPhee (1970 195) or Winosuparto (1968 125, 277) 48 This list is not exhaustive but includes the most frequently heard tetandakan pieces Each title, moreover, represents a category of tunes, and each local wayang tradition may play the repertoire in distinctively different ways (McPhee 1970 163) Some dalangs use the same lyrics each time they perform a particular tetandakan piece. In southern Bah, the phrase muknya ng luh 'the

173

D I S C O U R S E OF B A L I N E S E S H A D O W

THEATER

robbed of bell and book weeps. There is a tower with nineteen tiers, one above the other, made of gold, pearls, and diamonds.49 51. Bujangga anom kamalingan, apan kailangan dining pustaka, sang kaii manangis, ana wiku anulup tangpaksi d&nulupi, paksi anglayang an6 tawang, sekar sida d6nu yuti, jambotpanjangpininggiring pinggir mas. The young student of religion has been robbed, [and] since deprived of his books, continuously weeps; there is a hermit shooting a bird with blowpipe, the bird flies in the heavens; flowers by the millions; a long beard edged with gold trim. 52. Bujangga anom k6langaning pustaka, mwang genta, apan masaning amuja kaiya manangis, toya milir sakSng watu, jambot panjang pinihng mas, totopong bang ronron mapindan jadma. The young priest, who has lost book and bell, because it is time for worship and work, cries. Water flows from a stone; a long pigtail of hair, edged with gold, red headgear whose leafy ornaments have a human design. 53. Ganda kasih asih, apan baya pegat sih i aii manangis, apan gegangsalan tunjung mas crincing mirik salar kumenyai tetalia d6n tangisin, ana wiku paksi d6n tulupin ronronn6 mapinda jadma, ana m&iu matumpang solas mauncaluncal mas kumala gempuk gempang tiba ring segaia madu. Fragrance [and] pity, for the fear of separation from the beloved she cries, holding a golden lotus, beautiful voice, perfume, radiance, bound by weeping; there is a hermit, a bird shot with a blowpipe, leaves human-shaped; there is a pagoda with eleven tiers and a golden peak, a lotus disappearing [as it] falls into a sea of honey. fragrance of a woman' is the opening of a popular RCbong song; just these words alone are sufficient to conjure up a romantically playful and sensuous atmosphere, whether in theater or daily life. 4 9 The translations of all the poetic fragments here are only approximate. The sources for the texts quoted are: (50) written lampahan given by Hooykaas (1973:192-193), used with gending Rundah; (51) written lampahan given in Proyek Pencetakan/Penerbitan (1975a:42), used with gending Rundah; (52) performance tape, used with gending Rundah, (53) performance tape, used with gending Tangis and gending M6sem ; (54) Yayasan Pewayangan (1978a:73).

174

THE USE OF KAWI

54. Ganda kasih asih, gunung gempal tiba ring segaia madu, bujangga anom kamahngan mwang pustakanu hilang sang kan manangis mangasih asih . . Fragrance and pity, a mountain breaks [as it] falls into a sea of honey, the young student of religion robbed and his books gone, continuously weeping piteously... Although the five tetandakan examples above originate from differ­ ent plays and performers, occur at different points in each perform­ ance, and are accompanied by different mood-setting pieces, their po­ etry nevertheless shows striking correspondences. Many of the images portrayed are standard motifs in a body of esoteric kidungs (poetry us­ ing sekai madya form) composed in the courts of sixteenth- and sev­ enteenth-century G6lgel (the old seat of power for Balmese rulers with Majapahit ancestry; it is located near modern-day Klungkung). The poems' language and metric structure are both obscure for most Bahnese, as are the contents of this particular style, called lelungid (from a word meaning 'sharp, pointed'), "poems of didactic, mystical or erotic nature, often disguised as evocations of Nature's beauties" (Walhs 1980:185). The poems were composed more for musical than for textual exposition; their metric and melodic structure is such that the meaning of the text is obscured, rather than highlighted, by its method of being sounded (Walhs 1980:204). In all but a few areas of Bah, these kidungs are not known, and it is probable that many lontars recording the genre have been lost.50 Among the images that suggest that some of the dalang's tetandakan may have their source in lelungid-type kidungs are those involving weeping, priests, bird-and-ammal similes, precious stones and metals, temple towers, and exotic Nature (e.g., lotus, sea of honey). In the highly allusive kidung lexicon, such metaphors are often veiled refer­ ences to mystical or erotic ecstasy. Even seemingly straightforward images can have multiple layers of meaning. In example 52 above, for instance, toy a mihr sakέng watu is literally 'water flows from a stone', but by poetic convention it invariably refers to weeping. The next phrase, jambot panjang 'a long swatch of hair/whiskers', can be under­ stood as jambat panjang 'long and unbroken' or lambat panjang 'slow and long'—references to prolonged, sorrowful sobbing. And the follow­ ing phrase, pinning mas, is literally 'edged with gold', yet it can be read 50

The historical kidungs treated extensively by Berg (1927) constitute a dif­ ferent genre from that of the mystico-erotic texts referred to here 175

D I S C O U R S E OF BALINESE SHADOW THEATER

as 'the beloved is hinted at/teased' (from pinning 'hint, tease, allude' and mas 'term of endearment') or even as 'the beloved turned round his/her head and looked back' (since pimring can also mean 'to address someone while turning the head back'). The texts of the dalang's tetandakan are thus often densely lmagistic, obliquely metaphorical, and lyrical All these qualities could suggest archaic origins in the esoteric and speculative kidung tradition. If the songs can be traced to written sources, we may begin to understand more about how wayang parwa performers came to use them in standard vocal musical interludes.51 Even in the absence of direct textual confirmation, though, there is other material that can guide us toward a better understanding—in historical, cultural, and noetic terms—of the shaping of wayang tradition through contact with literary sources. The Literary Shape of Wayang Parwa. Our attempt to gain a historical perspective on the use of literary quotations and manuscriptstyled language in the shadow theater is furthered by interesting information provided by the kidung Wangbang Wideya, which was probably composed in the late sixteenth century in G6lg6l and has been edited and translated by Robson (19 71). The poem describes three performances of wayang "prawa" (a variant of parwa), each based on a different identifiable kakawin text. The lampahan of at least one of these performances seems to have paralleled the kakawin rather closely (Robson 1971 36), and the singing of kakawin stanzas was clearly a feature of the performances. The dalang combines the singing of kakawin verses {akakawin) with interpretation in the ordinary language [amrakreta, homprakreta). The phrase sumelang amanjang sindonira ('alternately he would sing poetry'—Robson 1971188-189) is also said to refer to the interspersing, in performance, of kakawin verses Similarly, in the prose chronicle Tantu Panggelaran, which originated in East Java in Majapahit times, a description of how the "first" wayang was played by the gods states that "beautiful kakawin verses were sung" (kmudungan/kinidungan panjang langonlangon— Pigeaud 1924· 103), if, indeed, panjang is the "declamation of poetic excerpts" (Mardiwarsito 1978:221, Hooykaas 1973 50). While Bahnese dalangs today do not refer to the singing of kakawin excerpts in performance as amanjang, there is still a clear connection 51 Not all tetandakan lyrics are of the romantic/esoteric type The songs for Tunfang music, which heralds demons, witches, and graveyard scenes, describe the singing and dancing of ghouls Even this imagery is allusive, however, and is frequently interspersed with obscure phrases

176

T H E U S E OF K A W I

between this term, kakawin verses, and wayang. The wayang paman]ang is an increasingly rare form of shadow theater associated with cremation ritual, in its full form, it includes a screenless stage, musicians, special puppet forms,52 offerings, and a dalang, all placed on the cremation tower [wadah] and carried in procession to the graveyard. The dalang's role was to play a lampahan,53 to sing kakawin verses, and at certain places along the way to chant mantras while tossing offerings54 onto the ground from atop the wadah. Today in Bah, genddr musicians, an officiating family member or pamangku, and offerings are carried on the wadah, but the full wayang pamanjang complement is hardly ever found.55 Knowledgeable persons (particularly Brahmanas) still are asked to ride along and "mamanjang," however, and the verses performed are some of the same kakawin excerpts used as bebaturan in contemporary wayang parwa.56 Returning to the kidung Wangbang Wideya, then, we can see that amanjang represents a traditional association between wayang performance and kakawin singing that goes back at least to Majapahit times in East Java and is not forgotten even today. It is worthwhile recalling, in this regard, that the courtly environment and themes depicted in Wangbang Wideya were created and enjoyed by an aristocratic elite who were centered in Gelgel and Klungkung, and who consciously styled both life and art after "ancestral" Javanese patterns 52

These puppets differ greatly from the usual forms, for surrounding each puppet figure is a rectangular "frame" carved from the same piece of hide 53 One informant who had seen many wayang pamanjang claimed to remember a particular lampahan played in this context, the "Prastanika Parwa," the story of the Pandawas' )ourney to heaven (see summary in Zoetmulder 1974 81) 54 Certain locations on the road to cemetery are considered magically dangerous for the soul about to be liberated through cremation, and offerings must be made at all such points These locations include crossroads, gulleys, streams and rivers, forks and turns in the road, the entrance to the graveyard, and each point at which the wadah is picked up or put down The Dharma Pawayangan (see Hooykaas 1973 50-53) lists the mantras to be performed while the dalang carries out the ritual 55 There is a general opinion that wayang pamanjang was particularly important in the courts of Klungkung but would only be seen today in the most lavish cremations It is still not unusual, though, for wayang figures, either cardboard cutouts or actual puppets, to adorn the cremation regalia 56 Examples are Bharatayuddha, canto 12, beginning Atha sedengira man tuk, and Arjunawiwaha, canto 13, beginning Mamwit narendratmaja In wayang lampahans, these vocal illustrations can be used to highlight moments of great sorrow, such as the death of a heroic character 177

D I S C O U R S E OF BALINESE SHADOW THEATER

Thus, the Javanese-modeled court culture that sparked the creation of the Bahnese kidung genre is the same aristocratic audience depicted in Wangbang Wideya, watching wayang parwa performances whose dalangs would "intermittently sing kakawin poetry " It is quite likely, then, that if dalangs were influenced by courtly styles of inserting kakawin verses in plays (i.e., amafijang, which may have influenced the development of bebaturan], they would also have been familiar with other court-centered vocal art—and the dominant genre was the kidung—that they might have used to beautify and embellish their performances.57 In such a way, certain kidungs or kidung-styled texts would become part of standard padalangan practice In the wayang "prawa" played for the courtiers in the Wangbang Wideya, it is clear that the dalang followed a story based directly on a kakawin episode and sang kakawin verses What is less certain is the extent to which the performer playing the story "Supraba Duta" used the same verses that tell the story in kakawin AT;unawiwaha (cantos 15 and 16). Robson (1971-36) states that "passages . [from the kakawin] alternated with an explanation, in which the story was developed " The kidung text does not entirely bear out this assumption, what we have is a dalang who "declaims the wayang" [angecapaken wayang), makes "interpretation in ordinary language" (amrakreta), "sings kakawin," and "inserts kakawin verses m his singing" [akakawin, sumelang aman]ang smdanira) The text nowhere clearly indicates that the play itself was an elaboration of certain kakawin passages. Rather, we have a pattern remarkably similar to present-day wayang parwa, in which spoken dialogue is combined with vernacular exposition and sung literary excerpts. In current practice, kakawin verses are thematically linked to characters and dramatic action, but even when dalangs perform lampahans directly based on kakawin literature (as opposed to stories composed by themselves or other dalangs), they use a variety of textual material and are in no way constrained to follow only those verses containing the lampahan being performed. It thus seems that the description of wayang performance in kidung Wangbang Wideya is congruent with, and indeed clarified by, the present-day Bahnese dalang's use of literary quotations, which was described earlier. Another explanation for the performance style described in Wangbang Wideya might propose that in ancient times dalangs were story57 The kidung Malat, another product of the Java-styled Bahnese courts, mentions a kidung sung during a wayang performance (cited in Zoetmulder 1974 154)

178

T H E U S E OP K A W I

tellers who sang their plays, modeling the lampahan on written literature; they did not, as is the custom today, speak the dialogue. Allusions to wayang performance in the oldest inscriptions and poems from Java often occur in conjunction with references to widu 'singers', and the kakawin Ramayana mentions a widu mawayang 'a wayangperforming singer'.58 Yet if the Balmese dalang of former times structured his play around the singing of kakawins, it would be difficult to explain how the oral composition of Kawi dialogue, "declaimed" in palawakya style rather than sung, eventually evolved. Perhaps only the Javanese shadow theater emphasized (kakawin) singing over spoken Kawi dialogue, it may, then, be a Javanese-derived style, popular in Majapahit times and preserved in the G6lg6l court, which is depicted in the Wangbang Wideya. However, neither in modern Java nor in Bah do we find any shadow theater practice that suggests that singing was once more central to the performance than spoken or declaimed dialogue. Vocal music is prominent in and essential to wayang in a number of ways, but in Bah it is certainly spoken Kawi dialogue, along with vernacular interpretation, that provides the fundamental shape around which the various styles of vocal musical embellishment are entwined. The literary genre that provides the closest written analogue to the verbal shape of Balmese wayang is, to my mind, none other than the Old Javanese parwa. The fact that the sounding of parwa texts in performance and the vocal technique used in the dalang's dialogue are in the same category—palawakya—is one clue to the correspondences between these manuscript and theater traditions The shared content of parwa texts and wayang parwa performance is another close link, one that has been reinforced for more than a millennium as dalangs have continually returned to the parwa manuscripts in search of dramatic style and substance. Another shared feature of parwa texts and wayang parwa provides perhaps the most revealing information of all. The written epic, it is generally agreed, was composed as a basically straightforward rendering of the Indian Mahabhaiata prototype.59 Yet the presence of many 58 From Sarga 24, verse 12, these and other references to wayang in Old Javanese literature from Java are discussed in Zoetmulder (1974 207-212) 59 The author of the Wirataparwa states his desire to mangiawaken Byasamata 'render the work conceived in Byasa's mind into plain Javanese' (Zoetmulder 1974 87) Still, no one has determined a specific Indian recension used as the source for the parwa texts, neither, for that matter, have exact prototypes been located for any Old Javanese work other than the kakawin Ramayana

179

DISCOURSE OF BALINESE SHADOW THEATER

interspersed Sanskrit "fragments," in the form of full or partial sloka, has remained puzzling to commentators. Zoetmulder (1974:87-92) provides the best recent summary of the character of these Sanskrit fragments, concluding that the fragments themselves do not form a skeletal narrative framework such as might be used for reciting the Old Javanese text; neither do they unequivocally point to any particular Sanskrit recension as the source used by the Javanese compiler)s). Pigeaud (1967, vol. 1, p. 15) suggests a relationship between the fragments and significant points or pauses in the narrative style of recitation exemplified by the dalang. Going further than either Zoetmulder or Pigeaud might approve, I would suggest that the structural and functional features of the Sanskrit words and phrases in parwa texts have much in common with the bebaturan of the Bahnese dalang. Zoetmulder (1974:90-91) notes that the Sanskrit fragments (1) occur in the beginnings of characters' speeches, as a preface containing the "opening words" of the speech; (2) contain moral, ethical, and didactic advice, such as on right conduct or politics; (3) take the form of curses, mantras, and prayers, and (4) embody metaphors and stock similes that are recurring themes in the text. All of these features of the Sanskrit scattered throughout the parwas are found in the bebaturan of the Bahnese dalang; the "Sanskrit quotations" used by the late tenth-century parwa authors, and the various textual quotations employed as bebaturan by Bahnese dalangs, all "served to enhance the value of the text" (Zoetmulder 197492). The specific value of the Sanskrit quotations and the bebaturan reflects the meaning attached to the use of archaic language as a vehicle of the potent messages that have been considered important enough to be transmitted through time. Also important is the rhetorical paradigm evoked by the process of quotation itself—as the signal from speaker to hearer of external authority and verification that supersedes the single, transient speech situation. Related to the connections between the Sanskrit quotations and the structure of the Bahnese dalang's linguistic performance is the observation that the words following a Sanskrit sloka in the text are usually a "translation or paraphrase" of the quotation in Old Javanese (Zoetmulder 1974:89). The Sanskrit fragments, in other words, were given the same recasting into contemporary idiom (at that time, Old Javanese) that is so intrinsic to Bahnese verbal art, whether literary or dramatic, up to the present day. The dalang echoes ancient written discourse, with its clear resonance of text and translation, in the immediate, spontaneous weaving-together of the language of the past 180

T H E U S E OF K A W I

with vernacular recapitulation. It is almost as though these ancient parwa manuscripts are not just versions of the epic, but also hold the key to how the epic was once performed. We can never know for certain to what extent the Old Javanese parwa texts were formulated along the lines of an ancient style of dramatic recitation, or even theater, that involved the Indian epics.60 That the mutual influence of manuscripts and theatrical forms on each other has a deep historical past in Java and Bah seems unquestionable when the structural similarities of the linguistic components of literature and the theater are examined. In Western terms, literature and the theater are contrasting forms of art that are clearly separated by a gulf of noetic and rhetorical implications; literature is a silent and solitary communing with print, while the theater is group experience of sound and spectacle.61 In Bahnese awareness, however, "performance" is equally important as a basic principle in both literary and dramatic verbal art. The force of oral-aural experience shapes the text in the lontar and wayang alike. The pattern evoked by the dalang's use of literary quotations, as well as text-with-translation form, harks back to prevailing noetic conditions as they certainly existed and influenced the creation of the Old Javanese parwa texts. SUMMARY In this chapter, we have considered the four major ways Kawi is used in Bahnese wayang parwa: in mantras, dialogue, narration, and sung embellishment. Each of these constitutes a different sort of "voice" within the wayang world. The four Kawi voices contrast in several dimensions: linguistic form, structural or dramatic function, and vocal sound quality. In rhetorical terms, too, they are distinctly different. The mantras emanate from a mystic realm where macro- and microcosmos are seen to correspond, Kawi dialogue belongs to the epic world inhabited by the puppet characters, the narrative voice is the 60 In this regard, it is interesting to note Levi's (1933 xxxm) observation that of the three Sanskrit stanzas that open the Adiparwa, the second is "the wellknown benedictory verse of Bhatta Narayana's Venisamhara," a text having to do with Sanskrit theater 61 Another aspect of noetic conditioning makes manifest the differences between world theatrical traditions in the Western theater we tend to take for granted the existence of afixed-formtext underlying every performance of a play, a specific written object followed by performers and anticipated by the audience In Bahnese theater, as we have seen, this is never the case

181

D I S C O U R S E OF BALINESE SHADOW THEATER

dalang's persona enhanced by and imbued with supernatural forces; and the sung quotations are textual voices, strengthening an oral performance through contact with the words of the past. For the dalang, "Kawi" is all these voices with their various sources, structures, and purposes. Together they are an inheritance encompassing the timehonored and traditional meanings of wayang parwa, as well as a noetically complex model for new performances. The Kawi voices store the cultural and theatrical past and simultaneously facilitate the continued recall of this heritage in present performance. Yet the Kawi voices are not the only verbal patterns important to the play. Perhaps the most fundamental linguistic shape of all emerges when dramatic Kawi is linked with vernacular—that is, Bahnese—explanation and commentary. In the following chapter, we will examine the uses of Bahnese in wayang parwa, m order to complete our understanding of how the dalang speaks his play.

182

5 SPEAKING THE PLAY: THE USE OF B A L I N E S E

Dalem Sangut Merdah TwaUn Medetn bangun ngamah dogtn Delem, Sangut, Merdah, Twalon, They're sleeping-waking eating men —Ketut Ginarsa, Panbhasa Bah1 BALINESE LANGUAGE as used in wayang parwa is a separate realm from that of Kawi, both linguistically and rhetorically It is the particular province of the four parekan 'court retainers', characters so wellloved by Balinese audiences, so crucial to the play, and so important as general cultural symbols that their origins and role bear careful scrutiny. The different terms used to identify this group of characters reveal various aspects of its identity and function as perceived by the Balinese The term parekan is from a Balinese root parek ('to approach', a related form, pack, means 'near') According to notions of power and its distribution, court servants are "those nearer" to the sources of political and spiritual authority embodied by the ruler Proximity to power is literally "inside" [jero] a glorious and protective atmosphere ranfied by the ruler's presence The area "outside" this influence [jaba], on the far side of the palace walls, lacks the luster radiated by the monarch's charisma Thus to be lowly at court is at least to have a certain prestige outside of it by virtue of participation in the "inside" world The four parekans in wayang are socially concrete roles for the Balinese audience, which views them as participants in a social order similar to its own Because they speak in Balinese, they are even within the same "universe of discourse" as the audience, in contrast with the epic world inhabited by the Kawi-speakmg figures2 1 Ginarsa (1971 9) classifies this two-line rime as a wewangsalan, the translation is my own 2 Not only in the shadow theater do Bahnese-speaking attendants "mediate" between universes of discourse Nearly all dramatic forms employ the dual-language device in Bali Genres based on shadow theater (wayang wong, parwa) em-

183

D I S C O U R S E OF BALINESE SHADOW THEATER

Among performers within the padalangan tradition and in the Dhaima Pawayangan texts, there is another name for the parekan: panasai or pandasar, from a root [dasai] meaning 'base, support'. In what particular sense this quartet is basic to the shadow play is a question worthy of consideration. Do the "base"-figures represent shadow theater's prehistoric origins, when the dalang made contact with ancestral spirits using puppet mediums? 3 In that case, the panasars may be the original "base" upon which the edifice of Indian epic was later erected; they are, after all, indigenous Indonesian characters with no precise counterparts in ancient epic written tradition. 4 Yet the panasars are also important in the dalang's mental and spiritual preparation for performance, which is associated in the mantras with other quartets of macro- and microcosmic significance (see the first section of Chapter 4 above). In this way, they are "basic" to the niskala (unmanlfest) side of wayang performance as well as its sakala (manifest) sounds and shapes. Another sense in which the panasars are fundamental to wayang is their omnipresence—the fact that they constantly appear before and speak to the audience. Their continual commentary is, furthermore, essential to the dalang himself, contributing to the process of oral composition and the play's overall coherence. In the remainder of this chapter, we will examine, through the panasar figures, the formal and structural, contextual, and noetic dimensions of the use of Bahnese language in contemporary performances. The four puppets under consideration here form two generally antagonistic pairs. The two characters allied with the Pandawas or other ploy identical parekan figures, while other genres (e g, topeng) have different but analogous clown-servant roles (see de Zoete and Spies 1938 and Bandem and de Boer 1981, for extensive treatment of dance-drama forms) 3 The Bahnese term dasaian, meaning 'persons often entered by spirits during ceremonies honoring the ancestors' (Panitia Penyusun Kamus Bah 1979 146) suggests this association 4 There are two ancient texts that tend to bear out the suggestion that shadow plays or other theatrical forms were having an influence on the shape of literary works in East Java One is the kakawin Ghatoikaca&raya, written in the early thirteenth century, in which are mentioned some "companion" characters of the hero Abimanyu These characters might be related to parekan figures (van Stein Callenfels 1925, Zoetmulder 1974 547-548) The second text is the more recent kidung Sudamala, probably written down in the late Majapahit era in East Java In this text, the Javanese counterpart to Twalen, named Semar, appears with that name as a parekan-style companion to the hero Sadewa (Zoetmulder 1974 435). 184

THE USE OF BALINESE

heroes of the "right" 5 are the father-and-son pair of Twalen and Werdah.6 Twal6n (called Maldn by his familiars) is the obese; blackskinned, snub-nosed, and fat-lipped audience favorite. Beneath his good-natured drollery, fumbling, and fussiness is the grumbling yet af­ fectionate wisdom of old age, hidden deeper is his special sakti (magic power) stemming from his origin as Sang Hyang Ismaya, younger brother of the deity Sang Hyang Manik Maya (Betara Guru, or Siwa). He often bickers with his son Werdah (Merdah, 'Dah, or Wana in some areas), whom he accuses of being a foolish upstart, and he is assailed in reply as a backward bumpkin. Twalen's voice is deep, slow, throaty, and rumbling, while Werdah speaks in high-pitched energetic bursts. Their opposite numbers, D6lem and Sangut, attend the heroes of the "left " Although there is not always a presupposed blood relationship, Delem imperiously expects to be called beh 'older brother' as a sign of his dominance. Loud, confident, boastful, and pushy, Delem (or Μέlem) glosses over his own physical and character blemishes, such as his goiter and his arrogance He is gullible and easily scared, yet always ea­ ger to start a fight. Sangut (or 'Ngut), on the other hand, is the cynic of wayang, he looks for the truth, even if it makes the "left" appear un­ favorably, yet he shies away from any philosophical absolutes. He is simple, honest, not easily fooled, yet compromising and flexible when it comes to avoiding unnecessary conflicts. He must humor Dolem, taking his postunngs with a gram of salt, and haplessly follow the Korawa camp into certain humiliation and defeat. These four characters/ ostensibly antagonistic, assaulting each 5

Puppets of the "right" are those lined up next to the screen on the dalang's right-hand side, those of the "left," dominated by the Korawas, are placed to the dalang's left Each wayang puppet belongs by tradition to one of these cate­ gories, and although the "right" side tends to be favored as morally superior and more alus (refined), there is no absolute distinction between "good" and "evil " 6 In some localities, Twalen and Werdah are considered to be older and younger brothers 7 Although the four panasars already mentioned dominate this puppet cate­ gory, there occasionally appear other "court servant" types who speak in Bahnese and provide bebanyolan In North Bali, Twalen and Werdah are joined on the "right" by Τοίέ and/or Dawala, also sons of Twalen There are also special clown-attendants for Gatotkaca and Hanoman (Hinzler 1975 49) In the south, there is a brash, outspoken female attendant [condong) who may accompany royal females Certain dalangs, specializing in humor, inject other physically bizarre, Bahnese-speaking characters into the play from time to time More­ over, in other genres of wayang, such as Cupak and Calonarang, special panasar characters are used For additional information on clown-servants in wayang, see Hinzler (1981 54,303-304) 185

D I S C O U R S E OF BALINESE SHADOW THEATER

other in the name of Korawa or Pandawa, yet members of the same social group and serving the same dramatic function, are for the wayang parwa audience a focal point whose importance can scarcely be overstated. Although they have been called "clowns" and are the source of the play's jokes and slapstick humor, their significance extends beyond the realm of bebanyolan 'humor'. As we will demonstrate, the panasars speak with Bahnese hearts and voices and represent a "commoner" Balinese heritage that is as much a part of wayang parwa's textual form and meaning as the courtly world of epic tradition is. FORMAL AND STRUCTURAL ASPECTS

A primary linguistic function of the four panasars is the translation into Bahnese of each speech of Kawi dialogue as it occurs. They must accurately convey the utterances of royal masters who, aloof and silent, wait for their words to be paraphrased. The panasars do not translate narration, use Kawi (except for quotation purposes), or speak for any persona other than their own except as commentator or translator. Because the wayang audience is not so familiar with Kawi that dialogue is transparently understandable, the panasars are crucial to the audience's understanding and enjoyment of the play, yet they maintain their unique roles and individual qualities throughout. During a great part of any wayang performance, an alternating pattern of languages prevails, as Kawi speeches are followed by their Bahnese paraphrase. (In the extracts that follow, the Kawi speeches are italicized, while the Bahnese paraphrases are not.) 1. Drona (in Kawi): Kadiang apa kita hana I hyun mangkana angerug kaiyan nang Pandawa I

Why do you have this desire to destroy the Pandawas' work?

Sangut (in Bahnese): Dadi ada ja keneh ceningo lakar ngrusak gaon sang Pandawa I

So you've got this notion, child, of destroying the Pandawas' work.

2. Bhisma (in Kawi)· Antyan satwa budi gmelar I Twalen (in Bahnese): Dulur ngelarang kadarman san6 utama /

How great the righteousness [you] have demonstrated. Along with demonstrating principles of the loftiest sort, 186

THE USE OF BALINESE

The standard, straightforward translation pattern is nearly word for word in these examples, even though the Kawi and Balinese versions may gloss somewhat differently when separately analyzed. The acoustic material is roughly balanced between the two versions, although certain syntactic contrasts are evident. For instance, utterance 1 begins with an interrogative phrase [kadiang apa 'why, how come'), while the Balinese gloss begins with a conjunction [dadi 'so, such that'). In utterance 2, predicate case roles are differently distinguished: the Kawi is a goal-focus construction, while the Balinese uses an agent-focus form of the transitive verb ngelarang, from gelai 'regulate, arrange, demonstrate'. Using the aesthetic ideal of formal balance between dialogue and Balinese translation, the dalang works with discourse units that are long enough to be complete utterances in themselves, but not so lengthy that he has forgotten what he stated in Kawi by the time he is ready to paraphrase in Balinese. While stylistic variation between performers precludes absolute definition, most dalangs would agree that an overlong Kawi speech, in contrast with one that is broken down into several utterance-with-translation pairs, is likely to forfeit audience interest in the performance. After a series of utterances by one character, before another character begins to speak or following a single, lengthier speech-with-translation, many dalangs use a device that "punctuates" the discourse unit. The form of this device is: [Kawi-speaking character]: [PanasarJ:

Yogya Patut

Correct. Right.

While the exact words vary among performers,8 the function of this "coda" is generally the same: it marks the end of a dialogue sequence that is longer than a single brief utterance. The dalang can also rely on this device, as he can on his cempala or Kawi narration, to "fill in" moments of hesitation or confusion in the course of a scene. It is not always the case that the Kawi and Balinese speeches correspond in relatively literal fashion. The dalang regularly uses the Kawi utterance as the "kernel" [buhh, binih) of what he wants to say, one performer explained, the panasar elaborates on a "skeletal" (pokok) figure provided by the Kawi utterance, said another dalang. The follow 8

Mangkana and mabenei are also used in the sense of "yogya" while asapumka, akanten, and akito can be equivalents for "patut " 187

D I S C O U R S E OF BALINESE S H A D O W THEATER

ing exchange illustrates this pattern; it is excerpted from a heated de­ bate during which Duryodana (eldest of the Korawas) tries to convince Drona that he will do evil no matter what instruction his teacher gives him. Duryodana's utterances are translated by Ddlem. 3. Duryodana: Nangmg tan paguna patutui mganika lawan Duryodana I

But it's no use your lecturing tome.

Ddlem: Yan titiang manahin / )eg ten w6nten gunan6 ten ι ratu nuturin kenten ring

What I think, is, well, there's no use your lordship lecturing me like that.

titiang /

Duryodana: Αρα kawenganyan I

What's the reason?

Delem: Napi k6 and mawinan I

What's the cause?

Duryodana: Lvnr αάέ I simpang lawan sadnyan nang Duryodana I

[Your] way is different, branching off from mine.

Dolem: Dwaning sampun ldmpas tek6n tetujon titiang/ titiang makeneh kangin ι ratu nujuin kauh / suba bes duegan ι ratu nuturin anak / anak bedak / entungin ι ratu galeng / anak kiap entungin ι ratu y6h / anak seduk entungin ι ratu kamben / dija ada unduk ta / unduk sing adanin ento / titiang makeneh lakar nguugang gaen Pandawa / tonden ι ratu mapahayu / yan ten ι ratu bes lebihan dueg I

Because you're opposed to my inclination. I'm thinking east and you're heading west. You try too hard to give advice—a guy's thirsty and you throw him a pillow, he's tired and you give him water; a fellow's hungry and you toss him clothes, what's the point? There isn't any. I intend to ruin the Pandawas' work and you haven't done anything to help. You're simply too clever.

The first two utterances by Duryodana have fairly literal glosses, while the third occasions a long elaboration by the panasar D6lem. The dalang is thus not bound to a "one-to-one" correspondence as the basis for Balinese translation of Kawi dialogue but, rather, can use additions and expansions of the material. This added material fills out the utter­ ance, incorporating detail, argument, and examples frequently taken 188

THE USE OF BALINESE

from the "life-world" of the audience—the everyday objects and expe­ riences known to all Bahnese. The fact that the panasars share the universe of discourse of the Ba­ hnese audience is further manifested in the formal aspects of their lan­ guage. They make full use of Bahnese lexical resources for the expres­ sion of intimacy versus social distance, of politeness, honorifics, and deprecation. This system outlines both a vertical dimension (hier­ archy) and a horizontal one (deixis, or the ordering of linguistic forms according to their distance from the speaker) as it obligatorily marks the relations between speech participants.9 While Kawi makes use of some hierarchically and deictically determined forms, it does not pos­ sess the pervasive language registers of modern Bahnese. In the context of wayang, an utterance in Kawi, relatively unmarked for the sort of pragmatic intent signified by language registers, when translated by the panasar explicitly outlines the dimensions of Ba­ hnese social space. 4. Darmadewa: Narandra sampun smanggah wruh mg tattwSng aji I

The king is famous for his knowledge of the essence of the venerated texts.

Sanguf Wantah cokor ι dowa san6 kasinanggahan maraga uning / ring tatwa-tatwa / dagmg kasusastrano sami wantah pradnyane" ring jagate" wantah cokor ι d6wa /

Only your highness is famous as the embodiment of learning, of the essences, the contents of all the written works, such is your wisdom in the world, just such, your highness.

5. Kresna: Yan anruwoyaken kunang aitha, yadnya kmantiniia I

If [you] possess wealth, make a friend of ceremony.

Werdah: Yan palungguh biang madruwe' mas pipis / patut anggdn mayadnya / kocap punika sampunang anggena mabiutakarma, kenten ida ι anak/

If your maternal majesty possess riches, it is fitting to carry out ritual, it is said, do not use it in mere noisy commotion. Such [says) your child.

In both the examples above, the panasars use respectful and honor9

See the third section of Chapter 2 for discussion of modern Bahnese lan­ guage register variation 189

DISCOURSE OF BALINESE SHADOW THEATER

lfic (alus singgih) Balinese style; in fact, every word in these "transla­ tions" by Sangut and Werdah presupposes selections from lexical sets of semantically equivalent but pragmatically opposed terms. Accord­ ing to padalangan (the art of wayangl tradition, all characters have sta­ tus relative to each other that must be reflected in the Balinese glosses. Indeed, it would be impossible for the panasars to speak in Balinese at all if such status-and-intimacy differences were not presupposed. Thus, Kresna addresses his aunt (Kunti) deferentially, and the phrase "your maternal majesty" [palungguh biang), not present in the Kawi form of the speech, is a crucial register component of the Balinese ut­ terance in example 5. Similarly, the repetition of cokon dawa (literally 'the foot of god') in example 4 emphasizes the politeness of Darmadawa's manner upon his arrival at the Pandawas' court—a politeness in dramatic contrast to his subsequent insults (see example 9 below). The use of varying registers in Balinese paraphrase also clearly iden­ tifies epic characters in terms of Balinese social and kinship roles. For example, while Kresna has several Kawi names that are borrowed from Sanskrit tradition, m Balinese his attributes and social status are given a more "native" hue: Ida he

Όέν/α god

Agung great

Rai younger brother

Sang Nathong the king of

Yadu Yadu people

This elaborate titling is similar to the naming patterns that distinguish high-caste nobility from commoners in Balinese society. The panasars are supposed to distinguish their own social position as well as those of their noble superiors. When they "speak for" an­ other character, they must use appropriate registers, whether high or low. If they switch to reported speech, they speak from a "distance," incorporating their own perspective into the indirect discourse. (In the examples below, boldface type is used to indicate honorific [alus] Ba­ linese.) 6. Hyang Indra: Twal6n I

Twalen!

Twalen: Aduh titiang, titiang parekan singgih betara ι tua ratu /

Aduh, I am here, the servant of the highest god, this old one, lord.

Indra: Enak pawistanmg ulun wus keita lugraha Ipamuhh

Proceed to deliver my command, and offer your 190

THE USE OF BALINESE

sembah ta π hyang-hyanging smembah I

reverence to the deities who are revered.

Twalen· Aduh / cai cai parekan gelah cai Mal£n, muhh sembah came ring sang patut kasembah I yan asapunika pangeledang singgih betara /

Aduh, you, my servant Malon, offer reverence to those who are fittingly revered. In such manner do [I] bow before the highest god.

7. Kunti: Kadiang apa nnasa dinta nanak I

What is your opinion, children?

Twalen: Kenken karasa baan cemng, kenten ida betari ratu /

"How d'ya feel about it, kids?"—that's what her divine majesty said.

Example 6 begins with a brief exchange, as the god Indra calls to Twalen and is answered in elaborate, formal Balmese. Then as the pa­ nasar translates Indra's reply he initially employs kasar forms, a deity would naturally use cai, a "low" second person pronoun, to a servant. But as he refers to deities and reports his own action, Twalon switches to high speech [nng sang patut. ..). In example 7, the first part of Twaldn's speech is intimate, everyday Balmese, he is translating for Kunti, who as a mother appropriately "speaks down" to her sons. Then Twa­ len changes to indirect discourse and also switches registers, using the hononfics befitting his own status as a subject referring to a royal su­ perior. The switch to alus speech ("that's what her divine majesty said") also functions as oral-aural quotation marks, setting off the translated dialogue from the panasar's own remarks. The panasars often preface or conclude their translations with this sort of device, clarifying not only the relationships among noble characters but also those between the elite and commoners. Another rhetorical device the panasar uses to point to his own status is particularly common in the patangkilan scene opening the play. Here, two or more ornate Kawi speeches may occur before the panasar humbly introduces himself, requesting permission to "imitate" the Kawi dialogue. 8. Twalon. Singgih / aratu sasuunan ι tua / pidaging mamitang lugra / kenginang ι tua kadi mangkin I antuk

Yes, king who is venerated by this old one, the intent is to beg leave; what is desired by the old one now, is to dare to 191

D I S C O U R S E OF B A L I N E S E S H A D O W

purun jagi ngojah / munggwing wecanan palungguh i ratu / mantukd ring ida i anak sareng kalih, boya asapunika /

THEATER

imitate according to the words of your kingly majesty which are directed toward these two youngsters; isn't that so?

While at this point in a performance many in the audience may not be attentive to the dialogue—since its content is predictable, and everyone knows the nature of the formal greetings being exchanged—it is a set feature of the opening scene, nonetheless, for the panasar to ask permission to paraphrase. The speech of the panasar figures continually makes distinctions by way of the person-related system of Balinese language registers. When Kawi dialogue is paraphrased in Balinese, then, this speech register information works to place the characters of the epic world, and the shifting relations between them, squarely within the bounds of Balinese social truth. The meanings of speech register variation in Balinese lives—social, political, religious, emotional—become strategies for text-building in the dalang's art, articulated through the voices of the panasars. The implications of misuse of language registers are not trivial to the Balinese. In the following excerpt, Kresna responds to Darmaddwa, a stranger who has come to the Pandawas' court of Amarta and boldly challenged Yudistira's qualifications as king. In his previous Kawi utterance, Darmad6wa has used ko, translated into Balinese as cai-, both are kasar second-person pronouns and in this case highly insulting. 9. Kresna: Ah / Darmadiwa meneng juga kita rumuhun, y&ki Kresna ngentiana kita / apa ngaran ikanang wang pradnyan, mingk6n£ kapradnyananta/

Ah, Darmad6wa, just be quiet first—this is Kresna who stops you. What manner of man is called wise? And is this your sort of wisdom?

Werdah: D6wa sang Darmaddwa nathdng Mantarabuwana i d6wa kad6n matitel ratu, mula k6n6-k6n6 katatasusilan i d6wan6 dadi ratu, karuan tiang ngelah sesuunan sang Darma dan6 suuna baan gumi Amartand, k6nk6n baos idan6, "ratu sang

Lord Darmad6wa king of Mantarabuwana: you think yourself titled king, and this is the ethic you have as king. Certainly I have the great Darma, venerated in the land of Amarta [as king]; and how did he address you? As "your majesty the king," since that's 192

THE USE OF BALINESE

prabu," kaden keto ida dong ι dewa jeg teka mai macaicai dong kone buktin kapradnyanan ι dewan61

what he thought you were, ha! Then here you come saying "cai"; ha, so this is the evidence of your wisdom.

For the Bahnese audience of wayang, although the non-panasar char­ acters speak in Kawi, they mean largely through the Bahnese of the panasar. An important implication here is that if the Kawi dialogue were not paraphrased, much of the play would be unintelligible to its auditors. The extent to which this generalization holds true, however, needs clarification. Like the epic characters themselves, the Kawi dialogue is distant from the here and now, and the panasar serves to bridge the gulf be­ tween both cultural and linguistic past and present. Yet there are many elements of a wayang parwa performance that need no explanation to any audience at all familiar with the genre: the standard character forms, names, and traits; the fundamental plot shapes, episodes, and dramatic moods; the underlying and endless conflict between Pandawa and Korawa. Much information is conveyed through the music, movement, and local socio-ntual context of the play as well. Thus, there are moments in the performance when the dalang can be assured that the meanings of the dramatic action are getting across to the au­ dience even without the panasar voices. Such a moment might occur during a battle scene, as heroes and de­ mons hurl Kawi challenges at each other without benefit of transla­ tion. 10. Raksasa (demon): Kadiang apa

How's it going'

Satyaki: Ih

I h . . . lh . . . (gasping)

.ih . .

Raksasa: Satyaki I cangkah

Satyaki—steel yourself1

cumangkah help

1

Satyaki: Pakulun . . tulung

My lord,

Raksasa: Kresna, tonton kadang waiganta mati . eeaau . . . hS, ho, h6. . .

Hey Kresna, see your family and friends die . . . (roaring and laughter).

During scenes typified by example 10, the action depicted on screen is more central than dialogue, m many cases the audience is familiar enough with the Kawi phrases that translation is superfluous in any 193

D I S C O U R S E OF BALINESE SHADOW THEATER

case. The Old Javanese lexicon is not so remote from Bahnese that there are no cognate forms to aid comprehension. In the example above, the only morphemes not shared by Kawi and Bahnese are kadiang (Bahnese has kadi 'like, as') and the second-person enclitic -ta, which many Bahnese accurately comprehend as a shortened form of kita. The verbal elements of the play that definitely do require translation are those containing details of the frequently intricate and circumstantial plots, as well as any philosophical or ethical points; these meanings all emerge through the Bahnese language component. As examined thus far, this component has an utterance-by-utterance structure, as Bahnese paraphrase is interwoven with Kawi dialogue. There is, however, another dimension to the use of Bahnese by the panasars. Both during and at the end of episodes, the panasars are free to discuss among themselves the content and implications of each scene as the plot unfolds. In these interludes, they can go over the conflicts and conversations of the various characters, in the process filling in for the audience any points it perhaps missed, highlighting material the dalang feels was insufficiently explained, and elaborating the key plot elements so that everyone in the audience continues to follow the story. These panasar exchanges are a second chance to improve the presentation of the lampahan (on the part of the performer) and to clarify understanding (on the part of the audience). Such exchanges are, in addition, extremely important for the incorporation of Bahnese beliefs and attitudes into the epic-based lampahan (see example 12 below). The panasar dialogue is also useful for commenting on actions that do not involve dialogue by other characters—as when, for instance, someone is meditating and a panasar, standing by, describes the intensity and supernatural events accompanying the exercise of spiritual power. The panasars also explain events that occur "off-screen" but are necessary for dramatic or thematic purposes. In the following excerpt, Twalen and Werdah are alone on screen, watching the busy activities of many citizens of Amarta who are preparing for a large ceremony. 11. Twalon: M6me ratu, nydn nyen to 'Dah, selidan )a suba ada anak hngsir rauh, ada padanda ba rauh, 'Dah / Werdah: Ida begawan Dhomya, nang

Oh my heavens, who is that, 'Dah, it's early and already some elderly personage has arrived, a priest has arrived, 'Dah' That's the sage Dhomya, Dad. 194

THE USE OF BALINESE

Twalen: Begawan Dhomya Werdah: Beneh / ida maraga yajamana dand lakar muputang karyand /

The sage Dhomya? Right! He is the officiant who'll carry out the ceremony.

Twalon: Beh, ratu, mggih, mangda masadana rahayu karyand / Werdah· Aduh nang, luh muani, cenik kelih, bajang tua / dand bisa ngigel panga ngigel / dan6 bisa magambel apang ngayah magambel / dane bisa lakar matetandingan ngurusang banten / nga6 banten /

Well, lord, yes, so that the ceremony has all the proper requirements.

Twal6n: Beh, makejang Werdah: Aduh, dewa ratu, nang. Nang, nang / Twal6n- Beh, an6 luh-luh mekejang jegeg-jegeg / Werdah: Mapayas patuh / aduh/

Well! Everyone [is involved].

Oh Dad, male and female, big and little, young and old; those who can dance, dance, those who can play the gamelan play, those who can arrange them are taking care of the offerings.

Oh, good heavens, Dad. Dad, Dad. Well, the women are sure pretty, all of 'em. Dressed up all alike, oh . . .

The Balmese terms for these short scenes of panasar dialogue are anda-anda 'chaining, linking' and telekang 'intensify, persevere'. The terms reflect the performer's awareness of the structural role of the panasar, in between either utterances or scenes, in the smooth flow and comprehensibihty of the performance. If the function of the panasar is to paraphrase, to recapitulate meaning, then this function is indeed carried out on at least two textual levels—that of the utterance, the basic unit of discourse, and that of plot structure, the coherence of sequences of utterances. CONTEXTUAL ASPECTS When the language of the past is transformed into the meanings of the present, many things happen. The duty of translation is for the panasar/parekan figure also an opportunity for audacious, pungent, moralizing, or critical commentary on the world and on the exploits of his 195

D I S C O U R S E OF BALINESE SHADOW THEATER

noble lords. Because other characters do not address each other without intermediaries (for paraphrasing the Kawi), one or more panasars are always present on screen, either participating directly in the action or eavesdropping from the periphery.10 Although there are sometimes brief scenes—travel, a fight, the descent from heaven of a deity—in which the panasars are absent, they always reappear to summarize and explain the action that has taken place. Every character's speech is thus in some way, whether through direct paraphrase or general commentary, expressed to the audience in Bahnese idiom. One result of the structural necessity for the continuous presence of the panasars, then, is the injection into every episode of their own "world view." This is partly a matter of time-honored individual traits—each clown's different voice, mannerisms, characteristic expressions, philosophical disposition, and so forth. At the same time the panasars are thoroughly Bahnese in attitude, expressing political, social, and cultural concerns—both traditional and contemporary— which the dalang can exploit in capturing the favor of the audience. We have seen in example 11 how the panasars can evoke a typically Bahnese atmosphere, the hustle and bustle of a temple congregation performing a ceremony, by means of a clever bit of dialogue describing offstage action. The spiritual, ethical, and practical aspects of Bahnese religion are contextual elements regularly brought into the panasar commentary. In the following example, Twalen and Werdah discuss the laudatory activities of the Pandawas in terms of ideal Bahnese behavior. 12. Werdah: Yasani ngahh dan6 madan rahayu . . . waluya tetujon naanang wak6 / nyujurang Ida sang Pandawa uh kenk6n baan naanang apang Ida kapucukan / r6h Ida lakar ngwangun karya rajasuya apang pragat / uhng ditu Ida lakar nyinang dan6 madan "Tnpaiamartha" Twalon: "Asih"

Their endeavor is to pursue that which is called Good . . . just like you, Dad, upholding the Pandawas m whatever way, so that they are revered, Because they are about to perform in entirety the kinginauguration ceremony; in that, they will show what's called the "Three Virtues": "Affection"—

10 This is in contrast to the Javanese wayang, in which puppets speak using various styles of modern Javanese, and thepunakawan (clown-servants) are discontinuously present, having their own extended scenes of action, commentary, and humor

196

THE USE OF BALINESE

Werdah: Lakar nywocain panjak

Blessing the people,

Twalen· "Punya"

"Endowment"—

Werdah: Lakar ngemaang panghargaan tekon anak dane" kaucap guru di gumind Twaldn: Apa adann6 to, "Baku" Werdah: Ngukuhang agamano r£h saadani di gumma wit sakong pakaryan Ida Sanghyang Widi Wasa

Directing esteem toward people acknowledged as teachers/leaders in the world, What's it called— "Devotion"— Strengthening religion and the order of things as they were originally established by the One Great God. That's it, the way of the Pandawas. Aduh, however will they handle such a busy ceremony? Hey, Dad, never mind, that's not our concern; the gods have already taken care of all that.

Twalen- K6to y6n gegelaran Ida Sang Pandawa. Aduh, k6nk6n 1a lakar ram6n karya Werdah. B6h, naanang, da, suba raosanga to, mirib sing )a naanang wak6 ngelarang to, watek dewata suba kal ngelarang to

In the following translation of a warning from Drona to his wayward pupil Duryodana, Sangut discourses on religious notions of reward and retribution [Karmapala). 13. Drona (in Kawi): Pwah I apa I yan mangkana I smgsal kita nanak / mmitta hana tnloka mankanang ]agatiaya, pernah paran lka I

Well1 What? If that's the way [you feel], you're lost, child, for there are three realms in the cosmos, what sort of places be they?

Sangut (in Bahnese): Aduh, ndon malu, cenmg, adengad6ng malu ngraos I yan k£toang cening / nang tegarang pinehin raos bapa I awinan ada kojaranga tnloka di jagati / lakar anggon gena ento, apa gunano ento / madyaloka,

Oh, hold on, child, and think about what you're saying. If you [behave] like you've said, you're lost. Try to think about what I'm saying, the reason it's said there are three places in the cosmos: earth, hell, and heaven. Earth, that's here, the 197

D I S C O U R S E OF BALINESE SHADOW THEATER

nerakaloka, swargaloka I Yan madyaloka dim kon£, di madyaloka tongos anako lakar malaksana / Yan malaksana jelok dim di madyaloka / di nerakaloka lakar ngalap buah pagainno dani jelok masih / Ada madan swargaloka dim kone" tongos anako ngalap buah pagao dan6 melah / [sings kakawin) "Penh benei ring swamarga, palamng anemua suka mageng" I Ta ngudiang cening jeg mamuruh / ldupe mai ke mercapada anak lakar kautus nebus dosa / ngudiang cening jeg mabalik dadi ngawe dosa / Lan darman6 kinkinang patuto gelarang apang sida kayang π wekas cening nepukin dan6 madan rahayu /

place where people carry out actions. If you act badly here on earth, you'll reap the fruits of it in hell. Heaven's the place for the rewards for right action. "If you truly strive in your way, you'll be rewarded with great )oy." So why are you being difficult, child? We live on this earth to redeem sin, why are you reversing this and creating sin? Get yourself together, act appropriately, so that later on you'll achieve happiness and well-being

The world that encloses and contextuahzes the "epic" universe of discourse in wayang parwa is a Bahnese world, where actions and con­ flicts are explained in terms of familiar beliefs and behavior. Not only are social divisions in the play world parallel to those of contemporary society; heroes are said to worship in Bahnese-style temples [puia], perform the same five types of ceremonies [pancayadnya], and hold to the same syncretic, Hindu-Buddhist "religion of holy water" [agama Siwa-Buda, agama tnta). In the days before widespread elementary ed­ ucation conveyed religious instruction, and even now m some areas of Bah, wayang performances are a major channel for moral and spiritual teachings such as are found in examples 12 and 13. The panasars further draw upon the variety of the culture's entire store of knowledge in evoking the Bahnese meaning of the drama. They may sing kakawin excerpts in standard wirama forms and ex­ plain the contents (as in example 13). The traditional orally transmit­ ted folktales and fables (dong6ng, satua) are found in panasar dialogue as examples drawn from popular, not necessarily written, wisdom. The Bahnese possess a large body of knowledge accumulated in the form of 198

T H E U S E OF B A L I N E S E

proverbs,11 which dalangs regularly exploit. For instance, a thoughtless character who boasts of his own merits might be reminded: Buka padinS, an6 puyung nyeleg, an6 misi nguntul Like an ear of rice, when empty it faces upward; when it has contents [i.e., is wise] it bows down. A hero who begins to quail in a moment of self-doubt is told of his noble parentage and warned not to act as though Ada singa manakang

bikul

There was a lion who fathered a mouse. When the Pandawa warrior Bima loses his temper and castigates his older brother Yudistira's judgment, he is said to be Mejek tai di gidatn6 Smearing shit on his own forehead. Panasar speech thus communicates using many of the forms in which culturally valued information is stored. Talking between themselves, Twal6n and Werdah cajole, jibe, and moralize in highly characteristic fashion, at once entertaining and educating the audience; in the following scene, Werdah is being reprimanded for criticizing old age. 14. Twal6n: K6wala k6n6 ny6n / Diapin jani gigin cain6 ngretek, buin pidan cai nak kal pawah cara bapa / diapin jani cai matan cain6 cedang / buin pidan cai lakar masih lamur cara bapa / ento mawinan ada rawos: "Lar6 ginanti ning anwam". . .

It's only this: even though your teeth can chomp now, one day you'll be toothless like your dad; your eyes may be sharp now, but someday they'll be blurry like mine. That's why there's the old saying: "Infancy is replaced by youth"

Werdah: Cenik ngantusang bajang... /

"Childhood is linked to youth"

11 Two collections of Balinese proverbs are van Eck (1868-1872, 1875) and Ginarsa (1971).

199

D I S C O U R S E OF BALINESE SHADOW THEATER

Twaldn: "Anwam gmanti nmg tua" ...I Werdah: Bajang ngantusang tua . . .

"Youth is replaced by old age"

Twalen: "Tua gmanti nmg pati" I Werdah: Tua ngantusang mati Twalen: Koto nyen llehang pang cai nawang /

"Old age is followed by death"

"Youth is linked to old age"

"Old age is linked to death" So there's the cycle, for your information.

Werdah: Aduh, naanang, saja nang I Aduh, nang / mara wak6 nawang, nah yen koto )a ulmg jam wak6 / lakar malajah sing ja bani lakar wak6 teken anak tua, nang

Oh Dad, you're right dad, oh dad, now I get it, and if it's like that, from now on I'll try not to be brazen with old people, dad.

Twalen: Uhng jam lakar cai sing bani, bum mam kala lakar bani k£to I Werdah: Sing ja k£to, nang / Twalon: Dong I kadin cai kenkdn tuyuh icang£ nga6 cai, nak kanti bubul entud icango

From now on you'll not be brazen, but come tomorrow you'll be brazen, so. It's not like that, dad.

Werdah: Aduh, da ngorahang kdto, k6nk6n /

Oh! don't talk like that, good grief.

Haven't you ever thought of my toil in making you? [I worked] till my knees were calloused!

This exchange highlights the way a skilled dalang can blend didactic and humorous content without violating the characteristic personalities of the panasars. Moreover, the panasars' quotation of written texts often takes the form of text-and-translation exchange; in example 14, the dalang draws upon the Sarasamuccaya, a collection of edifying sloka from Indian epic and Puranic tradition, which is in Sanskrit with Old Javanese glosses (Sudharta 1968, vol. 1, p. 18). The panasars are a channel for displaying the dalang's verbal and rhetorical resources. By playing off each pair's personalities, he demonstrates the kind of verbal sparring, with all its teasing, pointed rejoinders, and silencing comebacks, of which the Bahnese are inordinately fond. The dialogues of Sangut and Decern are particularly full of siat 200

T H E U S E OF B A L I N E S E

layah (fighting with tongues), thrilling audiences by constantly going beyond the ordinary bounds of civility. 15. Duryodana (to Drona): Yan suka rumekwaken kunang tutur Duryodana ngaturaken mangké /

Tf you'd like to listen, Duryodana will give you some advice now.

D6lem (translating): Sampunang i ratu akw6h nutur/ . . . ddwa suba ngldwa/ agama suba gamana/ widi suba magedi/ darma suba daar m6ng

Don't you talk too much . . . the gods are all mods, of religion there's hardly a smidgen, the godhead has already fled, and righteous law is in the cat's craw.

Sangut: Aduh ratu, aduh, aduh. Nguda ia k6to ida d6wa agung gedd, Mdlem /

Oh lord, oh, oh! Why is our great divine Majesty talking like that, Mdlem?

D6lem: Bedikan cai ngomong Ngut, buin cai masih milu/ cara nawang unduk, b6h, cara nawang swargan, cara nawang d6wa takeh caind /

Shut up, you, Ngut; there you go, butting in as if you know what's going on—like you personally know heaven and the gods, that's how you act.

Sangut: Lagut layah cai ten patulang da k6to ngraos

Since your tongue seems to be loose, you'd better not talk like that.

D6lem: K6nk