From Distant Tales : Archaeology and Ethnohistory in the Highlands of Sumatra [1 ed.] 9781443807845, 9781443804974

Citation preview

From Distant Tales

From Distant Tales: Archaeology and Ethnohistory in the Highlands of Sumatra

Edited by

Dominik Bonatz, John Miksic, J. David Neidel, Mai Lin Tjoa-Bonatz

From Distant Tales: Archaeology and Ethnohistory in the Highlands of Sumatra, Edited by Dominik Bonatz, John Miksic, J. David Neidel, Mai Lin Tjoa-Bonatz This book first published 2009 Cambridge Scholars Publishing 12 Back Chapman Street, Newcastle upon Tyne, NE6 2XX, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2009 by Dominik Bonatz, John Miksic, J. David Neidel, Mai Lin Tjoa-Bonatz and contributors All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-0497-5, ISBN (13): 978-1-4438-0497-4

TABLE OF CONTENTS

List of Illustrations ................................................................................... viii List of Tables............................................................................................ xiv Introduction ................................................................................................. 1 Part I: General The Dawn of Humanity in Sumatra: Arrival and Dispersal from the Human Remains Perspective ...................................................... 28 Harry Widianto The Neolithic in the Highlands of Sumatra: Problems of Definition......... 43 Dominik Bonatz Highland-Lowland Connections in Jambi, South Sumatra, and West Sumatra, 11th to 14th Centuries .................................................................. 75 John Miksic Part II: Northern Sumatra Is there a Batak History? ......................................................................... 104 Anthony Reid Ceramics, Cloth, Iron and Salt: Coastal Hinterland Interaction in the Karo Region of Northeastern Sumatra........................................... 120 E. Edwards McKinnon Ethnicity and Colonization in Northeast Sumatra: Bataks and Malays ... 143 Daniel Perret The Role of Local Informants in the Making of the Image of “Cannibalism” in North Sumatra ........................................................ 169 Masashi Hirosue

vi

Table of Contents

Part III: Central Sumatra The Megaliths and the Pottery: Studying the Early Material Culture of Highland Jambi ................................................................................... 196 Mai Lin Tjoa-Bonatz Adityavarman’s Highland Kingdom........................................................ 229 Hermann Kulke Tambo Kerinci ......................................................................................... 253 C. W. Watson Piagam Serampas: Malay Documents from Highland Jambi.................. 272 Annabel Teh Gallop Settlement Histories of Serampas: Multiple Sources, Conflicting Data, and the Problem of Historical Reconstruction......................................... 323 J. David Neidel Social Structure and Mobility in Historical Perspective: Sungai Tenang in Highland Jambi.................................................................................... 347 Heinzpeter Znoj Kerinci’s Living Past: Stones, Tales, and Tigers..................................... 367 Jet Bakels Kerinci Traditional Architecture.............................................................. 383 Reimar Schefold The Meaning of Rainforest for the Existence of Suku Anak Dalam in Jambi ................................................................................................... 402 Retno Handini Part IV: Southern Sumatra Mounds, Tombs, and Tales: Archaeology and Oral Tradition in the South Sumatra Highlands .............................................................. 416 Dominique Guillaud, Hubert Forestier, Truman Simanjuntak

From Distant Tales

vii

Southeast Sumatra in Protohistoric and Srivijaya Times: UpstreamDownstream Relations and the Settlement of the Peneplain ................... 434 Pierre-Yves Manguin From Bukit Seguntang to Lahat: Challenges Facing Gumay Origin Ritual Practice in the Highlands of South Sumatra ................................. 485 Minako Sakai Contributors............................................................................................. 501 Index........................................................................................................ 504

LIST OF ILLUSTRATIONS

1-1: 2-1: 2-2: 2-3: 2-4: 3-1: 3-2: 3-3: 3-4: 3-5: 3-6: 3-7: 4-1: 4-2: 4-3: 4-4: 4-5: 4-6: 4-7: 4-8: 5-1:

Map of Sumatra (drawing by Sebastian Hageneuer) Upper part of a human skeleton from Pondok Selabe cave (Baruraja, South Sumatra) showing strong features of the Mongoloid race, dating from 2,700 years ago Another human skeleton and skull from Padang Sepan (Bengkulu) Map of paleo-drainage located now beneath the South China Sea (Hantoro 2006: 56) Possible migration route of the Homo erectus during the glacial periods of the Pleistocene. Sumatra was bypassed, being a highland region to the west (Sémah et al. 1990: 41) Late or post-Hoabinhian adze from Bukit Arat, Serampas Find assemblage including potsherds, obsidian blades and hammers, and stone tools from test pit 2 in Bukit Arat, Serampas Jar burials at Muara Payang, Lahat (printed with courtesy of Balai Arkeologi Palembang) Presumed jar burials at Renah Kemumu, Serampas Two polished stone adzes from Muara Payang, Lahat (printed with courtesy of Balai Arkeologi Palembang) Excavated area with postholes of a house next to the megalith at Pondok, Kerinci Megalith in the centre of the excavated area at Bukit Batu Larung, Serampas Hybrid model of Sumatran spatial hierarchy Sumatran highland divisions and lowland drainage Monumental sites and topography South Sumatran administrative hierarchy, 1900 Sites on the Batanghari and major tributaries Dharmasraya and Tanah Datar Tanah Datar sites Sites on the upper Batanghari Sites of Indian (pre-Islamic) influence in northern Sumatra, and modern ethnic subdivisions of the Batak homeland

From Distant Tales

5-2: 6-1: 6-2: 6-3: 6-4: 6-5: 6-6: 7-1: 7-2: 7-3: 7-4: 8-1: 9-1: 9-2: 9-3: 9-4:

9-5: 9-6: 9-7: 9-8: 9-9: 9-10: 9-11:

ix

Statue of Si Singamagaraja XII in Medan, inaugurated by President Sukarno in 1961. The Aceh-influenced flag of the dynasty is depicted on the pedestal Medan area: Deli (drawing by Sebastian Hageneuer after Veth 1875) Taneh Karo: Passes into the lowland dusun areas (Edwards McKinnon 1984: fig. 137) Karo Kampungs on the Lau Renun (drawing by Sebastian Hageneuer after van Vuuren 1910) Ship relief: the rock-cut gereten at Lau Garut, Kampung Tanjung Taditional Karo house from the Doha (unpublished: courtesy of Lim Chen Siang: 2005 0411 Sumut 001) The ravine at Doha: traditionally, for defence, Karo villages were constructed on the edges of ravines Northeast Sumatra (Ecole française d’Extrême-Orient) Exchanges and meeting spots (Ecole française d’Extrême-Orient) Circulation of native products, first half of the 19th century (Ecole française d’Extrême-Orient) Centres of messianic movements and ethnic associations (18901940) (Ecole française d’Extrême-Orient) Map of North Sumatra (Hirosue 2002: 23) Distribution of megaliths in the highland of Jambi/Kerinci (source: Highland Jambi Research Project) Megalith of Pulau Sangkar II Megalith of Dusun Tuo Megalith of Bukit Batu Larung with the figure in a bent position and a circle: The pebble stones are the foundation for the megalith which was flipped onto its side (photo by Dominik Bonatz) Megalith of Benik: “dancing figure” before its destruction (photo by C. W. Watson in 1973) Megalith of Benik: The standing figure (guardian Dvarapala?) Carved stone near Lolo Gedang (photo by Dominik Bonatz) Oldest preserved house from the 1930s in Sungai Tenang Wooden carving from Kerinci: A human figure standing on an animal (with courtesy of the Ethnological Museum Bale, inv. no.: IIc1295) Pottery of Pondok: rim shapes of open and closed vessels (drawings by Listiani) Pottery of Benik: rim shapes of open and closed vessels (drawings by Listiani)

x

9-12: 9-13: 9-14: 9-15: 10-1: 12-1: 12-2: 12-3:

12-4:

12-5:

12-6:

12-7:

12-8:

List of Illustrations

Decoration on the pottery of Benik (drawings by Listiani) Pottery of Bukit Batu Larung: rim shapes of open and closed vessels (drawings by Listiani) Open vessels from Sungai Hangat (drawings by Listiani) Closed vessels from Sungai Hangat (drawings by Listiani) Bhairava statue from Sungai Langsat, now at the National Museum of Jakarta (Bildarchiv “Lotos-Film”) Edict issued by Duli [Pangeran] Ratu (r. 1639-1670) in Merangin addressed to various dipati. Renah Kemumu MS A (photocopy in the possession of Bapak Abu Nawas of Renah Kemumu) Edict issued by Duli Pangiran Dipati, dated AH 1086 (AD 1675/1676). Renah Kemumu MS B [a] (photocopy in the possession of Bapak Abu Nawas of Renah Kemumu) Edict issued by Duli Pangiran Dipati,? confirming the authority of Dipati Pulang Jiwa and Dipati Penarang Bumi over the people of Teramang in Serapas. Renah Kemumu MS B [b] (photocopy in the possession of Bapak Abu Nawas of Renah Kemumu) Edict issued by Duli Sultan Ingalaga (i.e., Sultan Abdul Muhyi of Jambi, r. 1679-1687) to Dipati Paling Jiwa. Renah Kemumu MS C (photocopy in the possession of Bapak Abu Nawas of Renah Kemumu) Edict issued by Duli Pangiran [Suta] Wijaya, to the menterimenteri of Syiring, dated 23 Syawal 1120 (5 January 1709). Renah Kemumu MS D (photocopy in the possession of Bapak Abu Nawas of Renah Kemumu) Edict issued by Sultan [Astra Ingalaga] (r. 1719-1725, 17271743) and Pangiran Dipati Mangku Ningrat of Jambi. Renah Kemumu MS E (photocopy in the possession of Bapak Abu Nawas of Renah Kemumu) Edict granted by Sultan Anum Seri Ingalaga to six Dipati, dated 15 Safar 1173 (Monday 8 October 1759). Renah Kemumu MS F [a] (photocopy in the possession of Bapak Abu Nawas of Renah Kemumu) Addendum to the edict contained in Renah Kemumu MS F [a], followed by an agreement between Dipati Ranah Ulu of Serampas and Dipati Gantu Lulu and Dipati Agung on land boundaries, 28 February 1906, ratified by the Dutch Assistant Resident of Sanggaran Agung. Renah Kemumu MS F [b], [c] (photocopy in the possession of Bapak Abu Nawas of Renah Kemumu)

From Distant Tales

12-9:

12-10:

12-11:

12-12:

12-13:

12-14:

12-15: 13-1: 13-2: 14-1: 15-1: 15-2: 15-3: 16-1: 16-2:

xi

Edict issued by Duli Sultan Ratu Seri Ingalaga and his adinda Pangiran Ratu Cakra Negara to Dupati Anggu Baya and Dipati Suta Menggala, dated 20 Syaaban 1214 (17 January 1800). Renah Alai MS A (original manuscript in the possession of Bapak Ruslan of Renah Alai) A clearer impression of the seal (#640) of Pangeran Ratu of Jambi as found on Renah Kemumu MS A, from a letter to Governor-General Johan Maetsuijker in Batavia, 1669 (Leiden University Library, Acad. 98 (13)) A clearer impression of the seal (#899) of Sultan Abdul Muhyi as found on Renah Kemumu MS C, from a contract with the VOC, 21 August 1681 (Arsip Nasional Republik Indonesia, Riouw 68/2) A clearer impression of the seal (#901) of Pangeran Suta Wijaya as found on Renah Kemumu MS D, from a contract with the VOC, 16 October 1763 (Arsip Nasional Republik Indonesia, Riouw 68/7) A clearer impression of the seal (#900) of Sultan Ahmad Zainuddin Anum Seri Ingalaga as found on Renah Kemumu MS F, from a contract with the VOC, 16 October 1763 (Arsip Nasional Republik Indonesia, Riouw 68/7) Photograph taken by Petrus Voorhoeve in Kerinci in 1941 of a document bearing the same seal (#901) as found on Renah Kemumu MS D (Koninklijk Instituut voor Taal-, Land- en Volkenkunde, Leiden, Or.414, 48b) Piagam Lantak Sepadan, issued by Sultan Anum Seri Ingalaga of Jambi in 1756, from the village of Guguk in Kabupaten Merangin, Jambi (WARSI) Serampas villages and the Kerinci Seblat National Park Moats of Dusun Tinggi (with courtesy of Dominik Bonatz; drawing by Sri Wasisto) Dusun Kabu: umbrellas were hung in the branches of a big cordyline as offerings at the grave of the ancestor Trowong Lidah Celebration of the Kenduri Seko: the Depati reads the holy oath. In the old days it was said to be done at the pulu negeri Stones surrounding a Cambodia tree at Hiang. This place which was also called Pulu Negeri was made in honour of a deceased village founder and ancestor Round stone of Muak (photo by Dominik Bonatz) Megaliths in front of the chief’s house, South Nias Megalith in the rice-fields, Lolo Kecil

xii

16-3: 16-4: 16-5: 16-6: 16-7: 16-8: 16-9: 16-10: 16-11: 16-12: 16-13: 16-14: 16-15: 16-16: 16-17: 16-18: 16-19: 16-20: 16-21: 17-1: 18-1: 18-2: 18-3: 18-4: 18-5: 18-6:

List of Illustrations

Rumah kereta api (photo by Depati Alimin) Outward-slanting walls in a granary from Lempur Drum house in Sungai Penuh Drawing of a putative reconstruction of a former bale by Depati Alimin) Village plaza in Hiang Cross-section and ground-plan of a longhouse (scheme). Dapeu=kitchen; umoh dalea=rear house; umoh luwo=front house; panganjon=elevated platform (drawing by Gaudenz Domenig) Various ways of fencing the basement in Hiang Carved door in Tanah Kampung, Kota Baru Separating wall between front and rear section in Siulak Modern longhouse interior in Sungai Penuh Inclined upper part of the wall of a house in Koto Iman Front-room with raised ceiling in Tanah Kampung, Kota Baru Structural system of a section in a longhouse (scheme) (drawing by Gaudenz Domenig) Posts in a longhouse in Siulak Extremely long house in Abai, Solok, West Sumatra Village plaza in Bawömataluo, South Nias Houses and granaries in Tikala, Sa’dan Toraja, South Sulawesi (photo by U. Schulz-Dornburg) Drum-like end of a megalith from Lempur with central star (photo by Mai Lin Tjoa-Bonatz) Top of a Dongson drum, Heger 1, South China (Heger 1902: Taf. VIII) Map of Jambi Province with the distribution of Suku Anak Dalm (Retno Handini 2005: 32) Map of the Pasemah region and of the main localities cited in the text (Laurence Billault, IRD Orléans) The Batu Gajah (elephant stone) (Laurence Billault, IRD Orléans) The Batu Tatahan: a drum between two carriers (Laurence Billault, IRD Orléans) Some engravings from the Dongson drums (Laurence Billault, IRD Orléans after Goloubew 1940) Sketch of a benteng near Belumai (Laurence Billault, IRD Orléans) Sketch of the site of Benua Keling Lama (Laurence Billault, IRD Orléans)

From Distant Tales

18-7: 18-8: 18-9: 18-10: 19-1: 19-2: 19-3: 19-4: 19-5: 19-6: 19-7: 19-8: 19-9: 19-10: 19-11: 19-12: 19-13: 19-14: 19-15: 19-16:

xiii

The stratigraphy of the excavation in Benua Keling Lama (Laurence Billault, IRD Orléans) Grave of Atung Bungsu in Benua Keling Lama (Laurence Billault, IRD Orléans) Grave of Serunting Sakti near Pelang Kenidai (Laurence Billault, IRD Orléans after van der Hoop 1932: no. 91) Sketch of the benteng of Benua Keling Lama (Laurence Billault, IRD Orléans) Southeast Sumatra: general map with all sites mentioned in the text (Map: Ecole française d’Extrême-Orient) Candi Siva, Kota Kapur, plan and section, 1996 excavations (Drawing: P.-Y. Manguin/EFEO) Southeast Sumatra: schematic map, protohistory to 5th-6th century CE (Map: Ecole française d’Extrême-Orient) Buddha (from the city of Jambi?), photographed at Rumah Bari, Palembang, 1985 (Photo: EFEO) Buddha from Betung Bedara (Schnitger 1936), now at the Museum Jambi Buddha head from Muar Timpeh, 2007 (photo by Don Longuevan) Southeast Sumatra: schematic map, late 6th to mid-8th century CE (Map: Ecole française d’Extrême-Orient) Buddha from Tingkip, Museum Balaputradewa, Palembang (Photo: EFEO) Candi Tingkip, plan and profile, 1999 excavations (courtesy Balai Arkeologi Palembang) Sketch map of Bingin (Pierre-Yves Manguin, Ecole française d’Extrême-Orient; after Topographische Dienst and Balar Palembang plans) Southeast Sumatra: schematic map, 9th-14th centuries CE (Map: Ecole française d’Extrême-Orient) Candi 3 under restoration at Bumiayu (Photo: EFEO) Terra cotta with garland décor, Candi 3, Bumiayu (Photo: EFEO) Siva statue from Modong (near Bumiayu), now lost (Knaap 1904) Sketch map of Teluk Kijing (field notes, 1994, Pierre-Yves Manguin, Ecole française d’Extrême-Orient) Candi Nikan: earthen mound with visible brick structure, 1994 (Photo: EFEO)

LIST OF TABLES

4-1: 12-1: 12-2: 13-1: 17-1: 18-1:

Lower Batanghari sites of the 10th to 14th centuries above one ha Piagam Serampas Piagam Serampas: terms of self-reference, contents and dates Serampas village populations Typology of Suku Anak Dalam List of correspondences

INTRODUCTION

The chapters in this volume represent a selection of papers given at a workshop entitled “From Distant Tales - Archaeology and Ethnohistory in the Highlands of Sumatra”, which took place in late September 2006 at the Free University of Berlin.1 The meeting was a highly successful one, attracting 26 regional specialists from ten different countries who joined together for three days to discuss the historical dynamics of the interior region of this large and strategically important island. Representing a range of key academic disciplines, the participants drew upon a rich array of materials: anthropologists primarily relied on oral sources but also reassessed historical ethnographies, linguists and historians presented newly translated documents, while archaeologists and art historians highlighted the material culture and their excavated finds. This interdisciplinary approach provided multiple angles of illumination, allowing maximum light to be cast on the societies of the island’s interior, which have until now remained in the shadows of regional, historical scholarship. The conference organizers designed this event to examine the dynamics of the highland societies within the larger regional context, particularly in relation to their lowland neighbours. Until now the highland-lowland relationship has been viewed as one in which the lowlands were the centres of civilization, dominating the culturallyfragmented, more egalitarian settlements of the highlands. This development of “complex societies” and “tribal or segmented groups” have been interpreted according to a number of theoretical models. Unilinear evolutionary perspectives, which have long been present in Indonesian archaeology (Miksic 2004a: 235-236), explain these two types of social organization as reflecting different stages along the transition from primitive to more advanced civilizations with diffusion from the Indian Sub-Continent typically serving as the agent of change. More popular in anthropology and archaeology today are models of centre1

The workshop was staged with generous financial support from several institutions. We gratefully acknowledge the Volkswagen Stiftung, the Free University of Berlin, and the Asia Research Center of the National University of Singapore.

2

Introduction

periphery relations which draw upon World Systems Theory. The benefit of this interpretive approach is that different societies are linked through economic and political processes as parts of a single spatio-temporal system, with the characteristics of each society emerging as a result of that relationship. Recent comparative studies, however, have demonstrated that centreperiphery concepts are not necessarily consistent with forms of unequal exchange (cf. Rowlands; Larsen/Kristiansen 1987; Champion 1989; Kardulias 1999). Instead, they have shown that the nature of exchange is often complex and accompanied by the pursuit of profits and avoidance of risks by all actors in the system. It is therefore necessary not only to develop an appropriate analysis for the economy of the core area, but also to find ways of characterising the economy of the periphery and describing transformations that are largely internal to it. Because none of these studies touched upon the Indonesian archipelago, the conference organizers thought that the analysis of long-distance relationships in Sumatra, especially between highland and lowland societies, with their markedly different patterns of social or economic organisation, could prove to be an important contribution to the understanding of centreperiphery relations within the Southeast Asian regional context. Complicating these theoretical ambitions, however, is a more basic, underlying question of whether conventional wisdom about the nature of pre-colonial highland societies is indeed correct. To what extent were the lowland polities really politically and demographically dominant to the highlands as typically supposed? Or is this characterization in fact a historical misrepresentation, resulting from the biases and errors of earlier generations of scholars? There is significant geographical, archaeological, and ethnographic evidence to support this latter interpretation, but the jury remains out with scholars representing both sides of this debate present in the following chapters. In order to provide the context for exploring these broader theoretical and empirical issues, the introduction continues with a brief overview of the geography of Sumatra. It then turns to an overview of the history of scholarship, which enables us to see how research evolved, and how intellectual biases, which this volume strives to rectify, crept into the study of this large island. Finally, it provides summaries of the chapters in this book, all of which contribute to these central concerns. With such a wealth of new materials and analyses, the conference and these proceedings reveal a much more complex history of this region than was previously understood and point to topics in need of future research.

From Distant Tales

3

Geographical Context Sumatra is the sixth largest island in the world (fig. 1-1), measuring approximately 1800 km in length and a maximum of 450 km in width. The entire island, which is bisected by the equator, is tropical in nature. It is affected by a monsoonal wind system, which leads to wet and dry seasons, though rainfall remains high year round in most locations. The island supports a number of different ecosystem types, with extremely high levels of species richness - though much has been lost in recent years due to rampant, wide-scale exploitation. The island lies on the western edge of the Sunda Shelf, a large extension of continental Southeast Asia, consisting of Sumatra, Borneo, Java, and other islands of western Indonesia. During glacial periods, when ocean levels were lower, portions of the Sunda Shelf that are currently submerged under the South China Sea connected these islands to one another and the mainland, making one large land mass. At current times, Sumatra lies separated from the Malay Peninsula by the Straits of Malacca, Java by the Sunda Strait, and Borneo by the Karimata Strait. The island of Sumatra can be roughly divided into three major geographical regions: the central Barisan Mountains (Bukit Barisan), the western coastal strip, and the eastern plain. The Barisan Mountains run like a backbone along practically the entire length of Sumatra on the western side of the island. The range was lifted and folded by the subduction of the Indian Ocean plate under the Sunda Shelf, giving rise to much volcanism in the region. The Barisan Mountains actually consists of two parallel ranges (Verstappen 1973; Whitten et al. 2000). The western range is more or less continuous, with elevations averaging about 2000 m, while the eastern one is more broken with elevations ranging from 800 to 1500 m. Between the two ranges lies a series of rift valleys, which a number of highland groups consider their original homelands (Miksic 1985: 425; Scholz 1987: 460; Reid 1997; Neidel 2006). This region is the central concern of this book. The western coastal strip is a narrow shelf of land that lies to the west of the Barisan Mountains, disappearing completely in places where the mountains drop directly into the sea. Rivers that cross the western coastal strip tend to be relatively small and rapid, limiting their utility for transportation. This coast also has relatively large waves and poor harborage due to a deep ocean depression that lies between Sumatra and the series of small western islands, including Enggano, Mentawai, and Nias (Whitten et al. 2000). Nevertheless this coast has played an important role in interregional trade dating back to at least the 10th century, as

4

Introduction

evidenced by Tamil and Arab inscriptions (Drakard 1990; Guillot 1998). Exchange goods have historically made their way from the west coast to the highlands by way of a series of mountain paths, though the steepness and potential hostility of highland peoples have seriously limited access to the interior regions throughout history. The eastern side of the island consists of a broad plain which was formed over the millenia by alluvial sediments carried from the Barisan Mountains by the great Sumatran rivers, including the Musi and Batanghari. These rivers are navigable for hundreds of kilometers upstream. Much of the communications and transportation between the highlands and lowlands in the past happened by way of the rivers, which formed natural highways to the foothills of the Barisan range, where they connected with footpaths that headed to the highlands. Major conjunctions of these rivers were typically picked as settlement sites, including those associated with Srivijaya and the other kingdoms that began to form around the 7th century CE.

The History of Research in Highland Sumatra Sumatra has been known to outsiders for 2,000 years. The first-century AD Greek cosmographer Claudius Ptolemaeus (Ptolemy) included data on “Baroussae”, corresponding to the northern and northwestern coasts of the island. Chinese and Indians of the same period were aware of the rare and precious commodities found in Suvarnadvipa, the “Isle of Gold”. The first foreigner to document a period of residence there was Chinese Buddhist pilgrim Yijing who spent several years studying Buddhism and Sanskrit in Palembang and Malayu between 672 and 695. It was several hundred years later before medieval travellers such as Marco Polo and Ibn Battuta described their stays of several months in its ports. These writers are no doubt representative of all early visitors to Sumatra: they saw only swampy estuaries where they were quarantined in specially reserved enclaves. The first Europeans to establish themselves in Sumatra were similarly restricted to the tidal zone. The Dutch established a trading office in Jambi in 1616 and in Palembang during the following year. The English tried to establish a base on Pulau Lagundi, Lampung, in 1624, but abandoned it the next year. The Dutch built an office at Padang on the west coast, which was burned in 1669 (Lekkerkerker 1916: 319), and another in Siak on the east coast in 1755. During this long period they seldom ventured out of sight of the sails of their ships. The first eyewitness accounts of the Sumatran hinterland were set down in the late 18th century. In 1770 a British botanist, Charles Miller,

From Distant Tales

5

journeyed to the watershed between the British base in Bengkulu and Palembang (Marsden 1966: 363). Then in 1778 he travelled from the west coast of Tapanuli into the Batak region. During this second journey, his party reached a “very extensive plain, on the banks of a large river which empties itself into the straits of Malacca.” John Marsden, son of William, the East India Company employee and author of the first History of Sumatra (first published in 1783), later visited the same area, and mentions seeing “two old monuments in stone, one the figure of a man, the other of a man on an elephant” (Marsden 1966: 373). These observations suggest that Miller and Marsden reached Padang Lawas. In 1800 another British company servant, Charles Campbell, reached Lake Kerinci (Marsden 1966: 304-306). These were feats of exploration, but did not result in substantial publications. In 1818 Sir Thomas Stamford Raffles, British governor of Bencoolen (modern-day Bengkulu), made two expeditions into the hinterland. These were both justified by political concerns related to the British settlement on the coast, but he took the opportunity to collect information about the people and environment. His first journey took him to Pasemah, now known to have been the home of one of Sumatra’s most important prehistoric societies. Relations between the British and Pasemah inhabitants had been hostile, but Raffles felt that the fault lay with the British rather than the local population, and so he set out to repair them. This expedition resulted in the famous discovery of the gigantic parasitic flower now known by the botanic name of Rafflesia arnoldi, after Raffles and his botanist. Raffles formed a very favorable impression of Pasemah. We found ourselves in an immense amphitheatre surrounded by mountains ten and twelve thousand feet high; the soil on which we stood rich beyond description and vegetation luxuriant and brilliant in every direction. The people, too, seemed a new race, far superior to those on the coast - tall, stout and ingenuous. They received us most hospitably […] I was not a little gratified to find everything the reverse of what had been represented to me. I found them reasonable and industrious, an agricultural race more sinned against than sinning […]. (Wurtzburg 1984: 440-441)

Shortly after returning from Pasemah, Raffles sent Captain Salmon on another expedition to cross the island from Bencoolen, on the southwest coast, to Palembang, on the east, the first time that this journey had been documented. Even before the results of this expedition were reported, he himself went to make the first study of the Minangkabau highlands, the home of the kingdom regarded at that time as the most ancient and important in the island, to which other Sumatran kings sent envoys to

6

Introduction

request recognition. William Marsden had devoted much attention to this region, recording the presence of a large inscription at Priangan (Marsden 1966: 352), but all his information was based on secondary sources. Raffles set off from the port of Padang with 50 soldiers, 200 porters, all his servants, his wife Sophia, and his friend Thomas Horsfield, a medical doctor and naturalist. He was escorted by two Minangkabau princes and about 300 of their followers. In the Minang highlands, he found “quite classical ground”, and refers to “the ruins of an ancient city”, “the wreck of a great empire”, “traces of a former higher state of civilization”, etc. (Wurtzburg 1984: 445-446), but unfortunately does not give further details. The Minang highlands came under Dutch authority in 1833, but it would be decades before any information could be added to Raffles’ descriptions of Minangkabau history and culture. The north coast of Sumatra had attracted numerous foreigners with the camphor and gold found in the mountains, but, probably repelled by tales of cannibalism, the visitors never ventured into the highlands. No outsider saw the great lake of Toba in the North Sumatran highlands until Dr. H. Neubronner van der Tuuk found his way there in 1853. It was another 15 years before the next European, the controleur J. A. M. Baron Cats van de Raet, became the second. The mountain-climber and volcanologist Frans Junghuhn wrote an early account of the Batak in 1847. Two years later, an article appeared in English about the Bataks of the Padang Lawas region (Willer 1849). In 1865, the Dutch sent a military expedition to Asahan, where Lake Toba’s outlet runs to the Straits of Malacca. Then in 1872, they made contact with the Karo (Cats van de Raet 1875). A provisional outpost was formed at Tarutung in 1872, but only exerted real authority from 1890. The plateau south of Lake Toba was a stronghold of organized resistance to the Dutch. They established a representative at Portibi, Padang Lawas, during a brief period of influence in 1839-1841; then ensued a period of 40 years during which Dutch made only occasional military forays before a controleur was stationed there in 1879. The Dutch exerted nominal control of the Angkola-Sipirok-Mandailing region from 1832 to 1837, then evacuated it during the Padri Wars. A Christian mission was set up there in 1857. A rebellion erupted in the name of a local ruler with the title Sisingamangaraja from 1877 to 1889, against whom the Dutch mounted four military expeditions before finally subduing the area. Controleurs were appointed for the area northeast of Lake Toba in 1892, but their offices only became operational in the period between 1896 and 1906. The Dutch attacked the Dairi area inland from the

From Distant Tales

7

ancient port of Barus numerous times before finally establishing administrative control there in 1904. Other officials were stationed at the now-popular tourist resort on the island of Samosir in 1906; Sidikalang in 1906; and Habinsaran in 1908. The administrator Baron Cats van de Raet (1875) conducted some of the first journeys through the highlands north of Lake Toba in 1866 and 1867, but the Simelungun and Karo areas were only brought under Dutch administration in 1915 (Lekkerkerker 1916). In 1894, books with sensationalistic titles such as Besuch bei den Kannibalen Sumatras by Joachim Freiherr Brenner von Felsach were still being published in Europe, but by the early 20th century scholars such as M. Joustra were compiling systematic ethnographies and dictionaries of Batak groups (Joustra 1912, 1926). Frans Junghuhn (1847), H. von Rosenberg (1855), and J. B. Neumann (1855) provided the first detailed information about the very important Padang Lawas area of North Sumatra. This region of approximately 1,500 square kilometers lies at the source of the Panai River, which flows east to the Straits of Malacca. The easiest land route across Sumatra leads from here to the port of Sibolga on the west coast via Gunung Tua and Padang Sidempuan. This geographical situation made the Padang Lawas area economically strategic. The kingdom of Panai is mentioned in the Tanjor inscription of 1030 which commemorates the Chola empire’s conquest of Srivijaya, and the Desawarnana, a 14th-century Javanese text, claims Panai as a dependency of the Javanese kingdom Majapahit (Wolters 1967: 193). Dutch sources began to record the presence of antiquities here in the 1880s. Further south, the Rejang valley came under Dutch rule in 1860. Only thereafter was it discovered that Rejang and the nearby Lebong valley had for centuries been exploited by indigenous gold miners. The Pasemah plateau was annexed in the years 1864-1866. The Central Sumatra Expedition under the leadership of P. J. Veth of 1877-1879 collected much information about the people of Rejang-Lebong in Bengkulu, the Rawas area of the upper Musi, the Minangkabau, and the Batanghari in Jambi. The report of this expedition contains descriptions of clothing, houses, and language (Veth 1881). These were not based on long-term research, but they recorded the first data on a broad range of subjects relating to the Sumatran highlands. The important archaeological site of Muara Takus on the upper Kampar River was mentioned in the Notulen (minutes of the meetings of the Batavian Society) in 1876, and by several other scholars in the 1880s, including R. D. M. Verbeek, one of the major scholars who took an interest in Indonesian archaeology in general before the Batavian Society’s

8

Introduction

Committee in Netherlands India for Archaeological Research on Java and Madura was formed in 1900. After Raffles’ description of his journey to Pagarruyung, there was a lull before Dutch officials stationed in the Minangkabau highlands began submitting reports to the Batavian Society of Arts and Sciences (Bataviaasch Genootschap van Kusten en Wetenschappen) in the 1860s; reports appeared occasionally in the minutes of the Society’s meetings. G. P. Rouffaer, a well-known historian of ancient Indonesia, wrote about the West Sumatran antiquities in his article in the Encyclopedia van Nederlandsch-Indie on Sumatra (Rouffaer 1905). Another major expedition crossed central Sumatra in 1907, travelling from Padang to the Kampar, the upper Batanghari, and Siak (Maas 1910). The first systematic report on West Sumatran archaeology and ancient history appeared in the Oudheidkundig Verslag (“Archaeological Report” of 1912. Nicolaas J. Krom, the author of the report (Krom 1912), acknowledged that much of the data had been compiled by the Assistant Resident of Fort de Kock (modern-day Bukittinggi), Louis C. Westenenk (This official contributed numerous important publications on highland Sumatra; see for example Westenenk 1921, 1922, 1932). Krom concentrated on the epigraphy of the region. This subject had been pioneered by Dr. R. Friederich, but the first person to write extensively about the inscriptions of West Sumatra was Hendrik Kern, who published his first article on the Minangkabau region in 1872, and continued to contribute articles on the subject for 40 years (Kern 1916, 1917). Krom listed 24 inscriptions and 31 other types of sites, from findspots of single objects to brick ruins. His inventory covered a huge area, from Padang Roco in the south to Muara Takus in the northeast. This density of ruins and artefacts indicates intensive activity over a period of several centuries. Subsequent scholars, particularly members of the Indonesian government’s archaeological research center and the staff of the office of archaeological preservation in Batusangkar, have added many more sites to this list. Although the pace of discovery is accelerating and any attempt to compile a new inventory would become almost instantly obsolete, there is a great need for a synthesis of what is now known of West Sumatra’s archaeology and epigraphy. In the middle and late 19th century, several internationally-known scholars brought Sumatra to the forefront of scientific inquiry. The first to do so was Alfred Russel Wallace, who spent some time in Sumatra during his decade-long research in the “Malay Archipelago” in the 1850s. Inspired by Wallace’s speculations on the Southeast Asian origins of humans, Eugene Dubois decided to begin his search for early human

From Distant Tales

9

fossils by joining the Dutch colonial army and arranging to have himself sent to Sumatra. He spent several months exploring caves near Bukittinggi, but became discouraged by the lack of results and moved to East Java, with memorable results when he discovered the site of Trinil. The next major scholar to devote attention to Sumatra was the French historian George Coedès, who in 1918 published his identification of Srivijaya as kingdom rather than a king. This realization led to a long and fruitful exploration of the history and archaeology of this early trading kingdom, which played a major role in the economies of ports from the Red Sea to China for several centuries (Coedès 1918, 1930, 1964). In the 1920s, archaeologists devoted considerable effort to the study of major complexes of ancient ruins in both highland and lowland Sumatra, South Tapanuli and Muara Takus. In 1920 P. V. van Stein Callenfels travelled overland from Padang through the Minangkabau highlands, including the Padang Roco area, thence to Padang Lawas and then to Medan (Callenfels 1920). In 1930 Frederik D. K. Bosch sailed from Jakarta to Padang, then travelled by road to Fort de Kock, then to Kota Nopan, and thence northward to Padang Lawas, then south to Muara Takus, as well as Pagarruyung and Pagaralam. In the lowlands, his visits were confined to Benkulu and Palembang (Bosch 1930). He reported that “the Palembang Lowlands district belongs to the areas poorest in antiquities from Sumatra.” While this evaluation is ameliorated by his references to several sites where large brick complexes had probably existed but had been completely destroyed by recent construction, often at the instigation of Dutch administrators, the contrast between the rather impoverished lowlands and the highlands, relatively rich in remains, is clearly implied. The archaeological record of the highlands became richer still in the 1930s due to the explorations of Abraham van der Hoop at Pasemah and Kerinci (van der Hoop 1932, 1940), and Friedrich Martin Schnitger. Schnitger was not a member of the archaeological establishment; he was something of a free-lancer, an Austrian whom the Dutch tolerated but never admitted to the Batavian Society. He was allowed to become the first conservator of the museum in Palembang in 1935. He began his archaeological pursuits there, but moved progressively further inland: to Muara Takus, Padang Lawas, and Rambahan (Schnitger 1936a, 1936b, 1937, 1938, 1939, 1989; Miksic 1989) After 1939, it might be said that Sumatran studies became a peripheral concern for archaeologists and historians, though not for anthropologists. Sumatra became a source of comparative material, rather than a central focus for the development of theoretical perspectives to which other areas

10

Introduction

are compared, rather than vice versa. Since 1920 research on the history and archaeology of the Sumatran highlands has languished in a peripheral position. Most historical and archaeological research there has been conducted by scholars whose main geographical interest lay elsewhere. Ethnologists, though, continued to find many more interesting subjects to pursue among the major ethnic groups inhabiting the highlands than in the coastal areas. Major syntheses of Sumatra’s ancient history compiled in the last 40 years have concentrated on lowland polities, depicted as outposts of civilization at the edge of jungles inhabited by barbarians. This characterization was shaped by assumptions about the importance of external communication for the development of Southeast Asian civilization rather than by concrete data. The dissonance between the density of hinterland ruins and population, on one hand, and the attribution of inspiration to foreign contact mediated by the harbour polities of the lowlands, on the other hand, was not recognized. One may detect here the typical colonial desire to glorify the role of foreigners, but also perhaps the role of the lowland dwellers themselves, whose archival sources emphasized the political dominion which they pretended to exercise over the highland groups. Whether this dominion actually existed, and if it did, when and how it was exercised, are important subjects for future study. This balance of power in the past has been simply assumed to exist rather than demonstrated. There are of course exceptions to any generalization. A few excellent historical works deal with the relations between the Sumatran highland and lowland groups. These include Christine Dobbin’s fine study of the Minangkabau over a period of several centuries, but unfortunately she did not pursue the subject further after her excellent book appeared (Dobbin 1983). Jane Drakard’s excellent study of upland-lowland relations in the Barus region of North Sumatra (Drakard 1990) can be compared with the archaeological volumes edited and partly written by Claude Guillot (Guillot 1998, 2004). The archaeological research which he directed on the ancient settlement in Barus is a major contribution to the field, but one may suggest alternative interpretations of the relations between the ancient coastal port and the population in its hinterland (Miksic 2001, 2004b). William A. Collins’ work on oral history in Sumatra (Collins 1979, 1998) indicates another very important line of research. There is a surprising lack of stories in the Pasemah hinterland about Srivijaya, whereas those about Majapahit are more common. Why is this so? Is it because Majapahit was a more recent kingdom, or did Srivijaya in fact have less impact on the hinterlands than Majapahit or its predecessor,

From Distant Tales

11

Singasari? This is a critical area for research, because people able to recite the old legends are rapidly fading from the scene. No doubt much has already been irretrievably lost, but it may still be possible to preserve a few more fragments of this type of information. In a related vein is the linguistic research focused on indigenous manuscripts. P. J. Voorhoeve conducted important basic research on highland Sumatran languages in the 20th century, and collected a large number of documents, some produced locally and others originating in the lowlands, which had been retained as community heirlooms (Voorhoeve 1927, 1955, 1961, 1970). A team led by Dr. Uli Kozok has continued this work, having identified a manuscript in the Kerinci area which Voorhoeve listed in his inventory as the oldest surviving Malay text written on a form of paper (Kozok 2004). This type of research is of the utmost importance because these documents are highly susceptible to the elements and a thriving black market in antiquities. One factor which may have deterred additional scholars from giving the hinterland peoples their due is their division among several major and many smaller groups, as opposed to lowland people, the majority of whom had by the 19th century assumed a generalized Malay identity. The lowland societies of Sumatra over the past 1,500 years have tended to represent themselves as polities with long-term chronological continuity like kingdoms or states. It is extremely difficult to piece together a coherent narrative for the many loosely-organized polities, often based on the principle of egalitarianism, of the highlands. Conversely, even though the 14th century highland kingdom of Adityawarman issued a large number of lithic proclamations, there have been few attempts to use them to piece together a history of his reign, and no sustained efforts have been made to use archaeology to augment the written record of the Minangkabau zone. Another factor handicapping scholarship on Sumatra is that, whereas researchers from various disciplines who study Java, Borneo, and Maluku have long been affiliated in more or less formal organizations, no such organization currently exists for Sumatra. In the 1970s Mervyn A. Jaspan of Hull University attempted to form such an association for Sumatra. He edited the Berita Kajian Sumatera (“Sumatra Research News”) for several years in stenciled format. With his untimely death, however, this publication met its demise, and no subsequent forum or discussion group has emerged to facilitate communication among scholars working in Sumatra. This volume touches upon several new directions for research on highland culture in Sumatra which need to be further pursued. The nature of the relations between highlanders and lowlanders can be seen to have a

12

Introduction

common structure, a fact which is striking given the extreme variety of culture and history in this 2,000-kilometer-long island. This suggests that environment may play an important role in this dialogue. While an array of studies of human-environment interactions could be pursued, one important initiative which measured anthropogenic disturbance on the vegetation of highland Sumatra through pollen analysis was begun in the 1980s (Morley 1982; Flenley 1988), but has not been continued. Future studies of trade may also help us to achieve a clearer definition of the highland-lowland relationship, which is clearly quite different from that observed in mainland Southeast Asia (although many scholars seem to have reached erroneous conclusions about Sumatra on the basis of the assumption that they were in fact similar). A related area which has been much less examined concerns warfare and fortification. David Neidel’s research indicates the progress which further inquiry into this subject may yield (Neidel 2006). Another subject which still awaits proper examination is the Tamil connection with the Sumatran hinterland. This is documented by inscriptions in Padang Lawas and the Minangkabau heartland. The ancient Tamil inscriptions have long been unstudied despite acknowledgement of their great importance. This is due to the lack of old Tamil linguistic expertise in this region, and lack of interest on the part of scholars based in India (Christie 1998). The penetration of Tamils far into the hinterland, at both Padang Lawas and Minangkabau, is a possibility which cannot be discounted, despite the fact that it seems difficult to reconcile with the resistance faced by early European and other foreign visitors to Sumatra. Other than the Bandar Bapahat inscription, there are few traces of Tamil penetration into the Minang area, but the area from Medan to Portibi is rife with various kinds of traces of significant Tamil communication with Batak groups, including rock-cut chambers, lineage (marga) names, and vocabulary. One last topic worth mentioning here is the importance of roads versus rivers, and the related question of the applicability of dendritic versus more central-place-like models. Pierre-Yves Manguin, contributor to this volume, writes: Had land transport been a true alternative to river transport, rather than only a complement, sites would by now have been brought to light at crossroads in the lowlands, away from rivers. Why would almost all the nodes in the dendritic model described above at different periods, as evidenced by archaeological sites, be situated on river banks, at confluences of navigable rivers, or at transhipment […].”

From Distant Tales

13

One reason which may be postulated is that the rivers ran in one direction (downstream), whereas the roads ran in the other direction (along lines of similar altitude). Although this issue is discussed in several chapters of this volume, more evidence needs to be brought to the table before it can be completely resolved. A final word may be said regarding the quantity of archaeological research which has been conducted on highland Sumatra but never published. This includes much material in Indonesian language. After the memorable 1954 survey of South Sumatra by the Indonesian scholars Soekmono, Satyawati Suleiman, Boechari, et al., and J. G. de Casparis and L.-Ch. Damais (Soekmono 1955), a more extensive survey was conducted in 1975 (Bronson et al. n. d.). This survey was only printed in a graymatter format and is not widely available. In a similar vein, one may mention the important research by the late Indonesian scholar Teguh Asmar on painted cist graves in Pasemah paintings, which was never published, and cannot be replicated due to the deterioration of those murals. The list of unpublished reports of later excavations is voluminous; to cite only two examples there are Machi Suhadi and Soeroso (Laporan Penelitian Arkeologi Klasik di Situs Jepara, 20 Mei-2 Juni 1984. Jakarta: Puslit Arkenas); Rr. Triwurjani et al. (Survei Arkeologi di Situs Danau Ranau, Sumatra Selatan. Jakarta: Bidang Arkeometri, Puslit Arkenas, 1993). Manguin in his contribution to this volume refers to many unpublished reports compiled by Indonesian archaeologists. The Jambi branch of the government organization responsible for conservation and preservation of monuments has also compiled numerous reports. The possibility of posting these on the internet should be seriously considered. The huge amount of work put in by Indonesian archaeologists in recent years to recover the ancient history of highland Sumatra deserves much greater recognition than it has so far received.

Chapter Summaries The chapters in this book, which provide a broad overview of recent research on the history and cultural heritage of highland Sumatra, are divided on geographical grounds into four main parts: a general category and then sections devoted to northern Sumatra, central Sumatra, and southern Sumatra. Within each section, the chapters are then arranged roughly in chronological order on the basis of the time periods discussed. This division, which is reminiscent of the organization used during the conference, underscores the regional diversity that exists within this vast landscape, while the reoccurring topics and themes that result from the

14

Introduction

common geographical setting and shared prehistoric heritage help maintain an overall sense of unity throughout the book.

General Harry Widanto begins this section by examining the important role that Sumatra played in the human migrations of the Pleistocene. During the socalled glacial periods, when sea-level was 100 m lower than it is today, land bridges formed making migration possible from mainland Asia to Sumatra and the other islands of the Sunda Shelf. This chapter discusses the initial arrival of people on Sumatra, their habitation patterns during the Pleistocene and Holocene periods, and also the patterning of population dispersal up to the present time. Using a palaeoanthropological perspective, Widanto interprets human remains, including pre-neolithic skeletons that were recently discovered in South Sumatra and similar materials from kitchen middens in North Sumatra and Aceh. The possibility that the oldest human traces in Sumatra could be buried in the beds of streams that flowed across the temporarily emerged land during the glacial periods, as can now only be seen on naval charts, is discussed. The second chapter by Dominik Bonatz focuses on one of the main problems of Sumatran archaeology, the definition of the period commonly designated as the Neolithic. The advent of this period is typically characterized by such achievements as agriculture, domestication, and sedentarization, but Bonatz questions the validity of using common archaeological data for identifying human activities related to one or more of these Neolithic “criteria”. To underscore the problems of classification, Bonatz compares archaeological finds from the highlands of Sumatra with those from other locations on and off Sumatra. He also discusses the relationship between humans and their surrounding landscape during the Holocene as revealed through various types of environmental data. Bonatz concludes by attempting to outline the probable nature of the Neolithic period in the highlands, underlining interactions with the early polities in the lowlands as one of the key factors for economic, material, and social changes in the highlands. John Miksic finishes off this section by analyzing the highland-lowland connections in Jambi, South Sumatra, and West Sumatra during the 11th to 14th centuries. Whereas archaeological reports on the island’s early history concentrate on discussions of imported objects and the famous lowland kingdoms known from foreign sources, Miksic suggests that recent discoveries of Chinese ceramics and other archaeological evidence in the highlands indicate that this focus is unduly biased. If this is correct, early historic period settlements in the eastern lowlands, including Malayu and

From Distant Tales

15

Srivijaya, do not represent the oldest form of complex society in Sumatra, but rather are comparatively later offshoots of highland agrarian societies. In making this argument, Miksic reformulates an influential model proposed by Bennet Bronson, which was based on the assumption that routes of communication and transport in ancient Sumatra were restricted to rivers. Using spatial analysis and the results of ongoing archaeological research, he attaches greater importance to overland routes and argues that the relationship between highlands and lowlands was more equitable than previous theories have assumed.

Northern Sumatra Anthony Reid starts this section by asking “Is there a Batak History?” Collectively, the 6-8 million Bataks of North Sumatra are one of Indonesia’s most important and intriguing groups. They have been in Sumatra for a long time and have attracted a large number of studies of religion and missiology, and a few good ethnological and language studies as well. Yet their history remains almost completely unwritten. According to Reid, this has occurred because historians have been overly influenced by the nation-state project of the past century, neglecting peoples who did not organise themselves into states which made claims on the past. Meanwhile, Bataks themselves have sought a state-like simulacrum through exalting the shadowy Singamangaraja dynasty. This chapter discusses the dilemmas of writing a history of highland, stateless peoples such as the Batak, who despite a rich tradition of writing did not generate linear chronicles, and for whom most sources are external ones. In the following chapter, Edmund Edwards McKinnon explores the coastal-hinterland interactions in the Karo region of northeastern Sumatra. Edwards McKinnon examines these linkages by focusing on the flow of commodities, such as ceramics, cloth, iron, and salt which were undoubtedly in demand and exchanged between the coast and the interior by the 14th century, if not earlier. Passes from the Karo plateau provided access to the riverine ports on the east coast and the plateau itself was the key link between the gold bearing region of Alas as well as the rich forest resin producing areas in the Dairi Pakpak region west of Lake Toba. Edwards McKinnon contends that the archaeological evidence for an established south Indian mercantile presence, initially at Lobu Tua near Barus from the 9th century and latterly at Kota Cina until the late 13th of early 14th centuries, and their involvement in the gold and resin trade on both the west and east coasts of Sumatra may explain the strong “Indian” influences on Karo folk memories and the presence of Tamil words in the Karo vocabulary.

16

Introduction

Daniel Perret next examines the relationship between ethnicity and colonization in northeast Sumatra. On standard ethnic maps, the northeastern area of Sumatra is characterized by an ethnic dichotomy between a coastal Malay population and an inland Batak population. Perret’s examination of local and other precolonial sources, however, reveals close links that existed between interior and coastal populations for over a millennium, with the term “Batak” seeming to have been primarily a geographical indicator meaning the population living in the interior vís a vís the coastal “Malay”. Perret argues that this distinction was altered by the coastal Malay elites when, following the introduction of Islam, they began to use non Muslim people living in the hinterland as slaves, workers for their plantations, and as spouses. The progressive rupture of the traditional links between the coast and its hinterland then continued from the 1860s onwards as Western planters, missionaries, and later the first colonial officials relied on the socio-politic vision of the coastal Malay elite. These developments, Perret contends, paved the way for the emergence of communal feelings and an “ethnic” consciousness in the whole area at the turn of the 20th century. Completing the discussion of northern Sumatra, Masashi Hirosue examines the emergence of the image of “cannibalism” for which North Sumatra has long been noted. The practice of cannibalism was widely believed by Arabian, Chinese, and European travelers to have existed among the inland people, despite the fact that those foreign visitors did not usually travel to the inland locations where the inhabitants were suspected of being cannibals until the 19th century. Hirosue argues that these images of cannibalism were used by coastal rulers to exert their control over trade relations by frightening foreign merchants from making direct contact with inland people. After those coastal rulers were subjected to European colonial rule during the 19th century, Hirosue contends that inland chiefs continued circulating talk of cannibalism among their villagers, for the purpose of impressing foreigners with the importance of their role in mediating the relationship between foreigners and local cannibals.

Central Sumatra Mai Lin Tjoa-Bonatz starts this section by analyzing two elements of the material culture of highland Jambi: megaliths and earthenware pottery. The megaliths, 21 of which were mapped in the areas of Kerinci, Serampas, Sungai Tenang, and Pratin Tuo during investigations conducted between 2003 and 2008, are the most impressive archaeological remains scattered across the highlands. According to their iconography and form,

From Distant Tales

17

two different types can be distinguished: conical megaliths, which are decorated with floral, anthropomorphic, and geometric motifs represented in base relief, and cylindrical megaliths, which often show human images and which are rendered in full relief. Local earthenware found at these different megalithic sites reveals remarkable similarities in terms of materials, decoration, and form. Tjoa-Bonatz argues that investigations at these and other settlement places indicate that highland Jambi should be considered a single cultural region and point to a continuation of a potterymaking tradition from at least the 12th until the early 20th century. In the second chapter, Hermann Kulke examines what is known about Adityawarman (ca. 1347-1377), the founder of the Minangkabau kingdom. Having gained considerable fame at the Javanese court of Majapahit and become the ruler of Jambi/Malayu, Adityawarman is known to have retreated to the highlands of West Sumatra and founded a kingdom right in the heartland of the Minangkabau people. He thus embodies a strange and perhaps even unique case in Southeast Asian history of an exchange of a flourishing downstream “metropolitan centre” with a well-established direct access to the profitable Indian Ocean trade system, for a mountainous upstream region. Although more than a dozen inscriptions and an impressive number of archaeological sites associated with him exist in the Pagarruyung region, very little is known about his kingdom, its extent and administration and, in particular, its integration into the rather egalitarian Minangkabau society. This chapter tries to make use of whatever can be derived from the existing epigraphical and archaeological sources in order to come to some hypothetical conclusions about these matters. In the following chapter, C. W. Watson discusses the Tambo Kerinci, a title which Dr. Petrus Voorhoeve gave to the 300 Kerinci heirloom (pusaka) documents held by local chiefs which he and his team documented in Kerinci in 1941 and then transcribed. The transcriptions disappeared during World War II, only to resurface 34 years later. In an article published in Bijdragen tot de Taal, Land en Volkenkunde in 1970, Vooerhoeve discussed the variety of documents which he had seen and commented on how they had their provenance in different parts of Sumatra - the west coast, the Minangkabau highlands, and the Jambi courts and appanages. Anthropological work points to some of these loose affinities between Kerinci and neighboring regions as well. In this chapter, Watson explores these associations by taking what at this stage is a very cursory look at the rediscovered Tambo documents to see whether they can help us to draw up any hypotheses about relationship between Kerinci village clusters (mendapo) and these adjacent areas, looking in particular

18

Introduction

at issues of trade, religion, and political authority. Watson concludes with a strong argument that more research is called for which will combine archaeology, kinship studies, linguistics, genetics, and a study of texts. Complementing Watson’s work on the Tambo Kerinci, Annabel Teh Gallop examines eight piagam from Serampas, copies of which were obtained by David Neidel during field trips to Serampas and Sungai Tenang, in the present-day District (kabupaten) of Merangin, Jambi, in 2004 and 2005. These documents are written in Malay in Jawi script, and are royal edicts and grants of authority issued by the rulers and nobles of Jambi to the dipati (customary officials) of the areas concerned from the 17th to the 19th centuries. Four of the documents are dated (1675, 1709, 1759, and 1800), while the other four can be dated to the period ca.1650ca. 1880s from their seal impressions. The edicts set out land boundaries, lay down laws and associated punishments for non-compliance, and establish royal prerogatives over certain categories of forest produce. The survival of these paper documents in the interior of Sumatra for up to 350 years, in the possession of the descendants of those to whom they were issued, belies the commonly-held belief that written records in tropical areas do not survive for more than one or two centuries. The Piagam Serampas, together with the documents in the Voorhoeve collection, are an exceptional resource for the nature of Malay governance in Sumatra. In the following chapter, David Neidel takes a slightly different tack by examining the problem of multiple sources, conflicting data, and the problem of historical reconstruction. He notes that in recent years, there has been an array of initiatives by historians and archaeologists to study marginal peoples and places. One of the difficulties in conducting these studies is that conventional data sources are typically very limited, forcing scholars to draw their evidence from widely disparate sources of information, much of which may be inconsistent. The question then becomes one of how to reconcile these inconsistencies, while still respecting the divergent perspectives of the people who live in and interact with these peripheral locations. This chapter engages with this question through an examination of the settlement histories of Serampas. Comparing European documentary sources, archaeological remains, modern-day oral accounts, and Malay documents, this chapter works through the conflicting evidence that arises even in areas as fundamental as pre-colonial demography, village typology, defense, settlement mobility and trade, and territoriality. Having reached some preliminary conclusions about the settlement history of the region, Neidel concludes by returning to the larger methodological question, advocating for a representational

From Distant Tales

19

approach which embraces, rather than obfuscates, the ambiguity and complexity present in the data. Heinzpeter Znoj next examines the issues of social structure and mobility in Sungai Tenang. He asks: how did highland Jambi’s society evolve since the period, when the impressive megaliths were placed? Does its structure itself bear the imprint of such evolutions? In answering these questions, he argues that highland Jambi society’s enigmatic structure neither clearly matrilinear nor simply cognatic, and neither clearly kinbased nor truly territory-based - is the result of conflicting adaptation processes to unsteady economic, political, and religious conditions that shaped the region since the early modern period. He argues that they are best understood under the assumption that highland Jambi society has been involved, from a remote and subordinate position, in irregular but at times intensive exchange with lowland and coastal markets and political and religious centres. Due to its isolated location, it had been forced to rely during extended periods on its own for subsistence, security and legal and political order. The social structure that emerged is a compromise between kinship-based and political principles. This compromise expresses itself in two important realms of social organisation: in the mentioned mixture of matrilineal and cognatic reckoning of kinship and in ancestral ritual communities, membership to which was not only by descent but alternatively also by adoption of immigrants. Jet Bakels next uses oral traditions and current cultural practices in attempts to shed some light on the historic meaning and function of the archaeological remains of Kerinci material culture. Of particular interest are the megaliths and other stone structures (pulu negeri), holy community heirlooms (pusaka; including textiles and ceramics), and traditional houses. In some cases the pre-Islamic practice of giving ritual offerings in connection with these objects is very much still alive and Bakels sketches a pre-Islamic worldview in which the meaning and functioning of these objects is rooted. She also analyzes Kerinci tales that are locally believed to explain the dawn of Kerinci civilization, the foundation of certain houses and villages, and the existence of the megaliths, and that account for the magical forces attributed to the heritage objects. Bakels argues that it is in terms of a pre-Islamic worldview in which ancestor worship and an animated natural world are central, that the material remains of Kerinci are best understood. Reimar Schefold continues the discussion of central Sumatra with an analysis of the traditional houses of Kerinci. His contribution is based on a survey of old and contemporary forms of traditional architecture in Sumatra and Java that he conducted with Gaudenz Domenig in 1995 as

20

Introduction

part of the Royal Netherlands Academy of Sciences project devoted to the study of vernacular architecture in western Indonesia. The results from their short visit to Kerinci were complemented by the investigations of a team of local adat experts under the guidance of Depati Alimin. In this chapter, Schefold describes and illustrates some formal and structural properties of Kerinci houses that might be of interest in a (pre)historic perspective. He reviews certain features from a comparative viewpoint, as local transformations of characteristics of what he has called the Southeast Asian-type house, a house type which he argues is rooted in the neolithic and early metal age traditions of Southeast Asia. In the final chapter of this section, Retno Handini examines the meaning of the rainforest for the Suku Anak Dalam people of Jambi province. The Suku Anak Dalam have long been dependent on the rainforest environment for their subsistence, which historically relied on hunting and gathering tied to a nomadic lifestyle which allowed them to track various food resources. The widespread deforestation of the last several decades, however, has had a huge impact on the life of Suku Anak Dalam people, each group of which has had to adapt new survival strategies. These adaptive decisions have also been influenced by the peoples’ own desire to move into what they percieved to be a better lifestyle and government policy aimed at sedentarization. Given these changes, the Suku Anak Dalam groups can now be divided into three general categories based on their subsistence activities: foragers, semiforagers, and sedentary groups. This chapter describes various aspects of their daily lives.

Southern Sumatra The Franco-Indonesian team of Dominique Guillaud, Hubert Forestier, and Truman Simanjuntak start this section by examining the archaeology, geographic distribution, and oral traditions surrounding a series of mounds and tombs located in the Pasemah region of the South Sumatran highlands. The region of Pasemah has long been known for a series of carved megaliths, which have been tied to the Dongson period through their iconography, but many questions remain about the dating and other attributes of these and other less well known remains. Confronted with these uncertainties, Guillaud, Forestier, and Simanjuntak conducted an excavation of a burial mound at Benua Keling Lama, a haunted site containing the grave of a famous ancestor of the Pasemah people, tracing the evolution of the site from a Neolithic settlement to a Metal Age settlement and then to a burial site. This chapter combines the team’s

From Distant Tales

21

archaeological research with analysis of oral traditions and other presentday representation of the past in order to give us a richer picture of the regions historical dynamics. In the following chapter, Pierre-Yves Manguin analyzes the upstreamdownstream relations and the settlement of the peneplains during protohistoric and Srivijaya times. Manguin takes aim at the notion that the polity of Srivijaya was merely an entrepôt (i.e., a harbour-city geared mainly towards international maritime trade) with little or no relationship with the vast hinterland provided by the Musi and Batanghari River basins. To support his argument, Manguin presents archaeological and ecological evidence built up during fieldwork in southeast Sumatra in the 1980s and 1990s. It documents Srivijaya’s swift move to control the hinterland immediately after its foundation in the late 7th century CE and the progressive settlement in subsequent times of most of the peneplain upstream from the central places of the two rivers basins. It thus confirms that the Malay polity functioned as a true city-state, extending its sphere of influence far upstream into two of the largest river basins of Insular Southeast Asia. The chapter also shows how much the local world views, as expressed in downstream Palembang Sultanate texts or in upstream Pasemah oral literature, match the available archaeological evidence. In the final chapter, Minako Sakai examines the ritual practices of the Gumay, a people of the South Sumatran highlands who trace their origins to Bukit Seguntang near Palembang. In theory, each Gumay village has a founding ancestor or two to whom village residents connect themselves through genealogical links. This chapter summarises the tales of Gumay ancestors and explores the process of their journey. The tales highlight interactions with other ethnic groups and the colonial administration. The chapter then explains how the memory of such journeys is enacted through ritual practice and the spatial organisation of Gumay villages. Sakai argues that the practice of origin rituals remains a vital part of social life in the highlands of South Sumatra and the core of the ritual practice continues despite the eroding influences of modernisation, Islamisation, and globalisation.

22

Introduction

References Bosch, Frederik David Kam 1930 Verslag van een reis door Sumatra. Oudheidkundig Verslag: 133157. Bronson, Bennet, Machi Suhadi Basoeki, and Jan Wisseman n. d. Laporan Penelitian Arkeologi di Sumatra. Jakarta: LPPN University of Pennsylvania Museum. Callenfels, P. V. van Stein 1920 Rapport over een dienstreis door een deel van Sumatra. Oudheidkundig Verslag Bijlage G: 62-75. Cats, J. A. M. Baron de Raet van 1875 Reizen in de Battaklanden in December 1866 en Januarij 1867. Tijdschrift voor Indische Taal-, Land- en Volkenkunde 22: 164-219. Champion, Timothy C. 1989 Centre and Periphery. Comparative Studies in Archaeology. London: Unwin Hyman. Christie, Jan Wisseman 1998 Tamil language inscriptions in Southeast Asia and China. Journal of the Malayan Branch of the Royal Asiatic Society 29: 239-268. Coedès, George 1918 Le royaume de Çrivijaya. Bulletin de l’Ecole française d’ExtrêmeOrient 18(6): 1-36. 1930 Les inscriptions Malaises de Çrivijaya. Bulletin de l’Ecole française d’Extrême-Orient 30(1-2): 29-80. 1964 Les Etats hindouisés d’Indochine et d’Indonésie. Paris: de Boccard. Collins, William A. 1979 Besemah Concepts: A Study of the Culture of a People of South Sumatra. Ph.D. diss. University of California, Berkeley. 1998 The Guritan of Radin Suane. A Study of the Besemah Oral Epic from South Sumatra. Bibliotheca Indonesia 28. Leiden: KITLV Press. Dobbin, Christine 1983 Islamic Revivalism in a Changing Peasant Economy: Central Sumatra, 1784-1847. London: Curzon. Drakard, Jane 1990 A Malay Frontier: Unity and Duality in a Sumatran Kingdom. Ithaca: Cornell University. Flenley, John R. 1988 Palynological evidence for land use changes in South-East Asia. Journal of Biogeography 15: 185-197.

From Distant Tales

23

Guillot, Claude, ed. 1998 Histoire de Barus I. Paris: Association Archipel Cahiers d’Archipel. 2004 Histoire de Barus II. Paris: Association Archipel Cahiers d’Archipel. Hoop, Abraham N. J. Th. à Th. van der 1932 Megalithic Remains in South-Sumatra. Zutphen: Thieme. 1940 Prehistoric site near the Lake Kerinchi (Sumatra), in: Proceedings of the Third Congress of Prehistorians of the Far East: 200-204, eds. Frederick N. Chasen and Michael W. F. Tweedie. Singapore: Government Press. Joustra, M. 1912 De Bataks. Leiden: Uitgaven van het Bataksch Instituut 7. 1926 Batakspiegel. Leiden: Uitgaven van het Bataksch Instituut 21. Junghuhn, Frans W. 1847 Die Battaländer auf Sumatra. Berlin: G. Reimer. Kardulias, P. Nick 1999 World-System Theory in Practice. Leadership, Production, and Exchange. Oxford: Rowman & Littlefield Publishers. Kern, Hendrik 1916 Verspreide Geschriften Vol. 6: Inscripties van den Indischen Archipel. The Hague: Martinus Nijhoff. 1917 Verspreide Geschriften Vol. 7: Inscripties van den Indischen Archipel. The Hague: Martinus Nijhoff. Kozok, Uli 2004 The Tanjung Tanah Code of Law: The Oldest Extant Malay Manuscript. Cambridge: St Catharine’s College and the University Press. Krom, Nicolaas Johannes 1912 Inventaris der oudheden in de Padangsche Bovenland. Oudheidkundig Verslag, Bijlage G: 33-50. Lekkerkerker, C. 1916 Land en Volk van Sumatra. Leiden: E. J. Brill. Maas, Alfred 1910 Durch Zentral-Sumatra. Berlin/Leipzig: B. Behr’s. Marsden, William 1966 The History of Sumatra. [3rd edition reprint] Kuala Lumpur/New York: Oxford University Press. Miksic, John Norman 1985 Traditional Sumatran trade. Bulletin de l’Ecole française d’Extrême-Orient 74: 423-467. 1989 Introduction to Forgotten Kingdoms in Sumatra. Singapore: Oxford University Press: vi-xvii.

24

Introduction

2001 Review of “Histoire de Barus: Le site de Lobu Tua. I. Études et documents.” Bijdragen tot de Taal-, Land- en Volkenkunde 157(2): 430-432. 2004a The classical cultures of Indonesia, in Southeast Asia. From Prehistory to History: 234-256, eds. Ian Glover and Peter Bellwood. London/New York: Routledge Curzon. 2004b Review of “Histoire de Barus: Le site de Lobu Tua. II. Étude archéologique et documents.” Bijdragen tot de Taal-, Land- en Volkenkunde 160(4): 593-599. Morley, R. J. 1982 A palaeological interpretation of a 10,000 year pollen record from Danau Padang, central Sumatra, Indonesia. Journal of Biogeography 9: 151-190. Neidel, John David 2006 The Garden of forking Paths; History, its Erasure and Remembrance in Sumatra’s Kerinci Seblat National Park. PhD. diss. Yale University, New Haven. Neumann, J. B. 1855 Het Pane- en Bila-stroomgebied. Tijdschrift van het Koninklijk Nederlandsch Aardrijkskondig Genootschap 2,1 (second series): 1133. Reid, Anthony 1997 Inside out: The colonial displacement of Sumatra’s population, in Paper landscapes: Explorations in the environmental history of Indonesia: 61-87, eds. Peter Boomgaard, Freek Colombijn, and David Henley. Leiden: KITLV Press. Rosenberg, H. von 1855 Hindoebouwvallen in het landscap Padang Lawas. Tijdschrift voor Indische Taal-, Land- en Volkenkunde uitgegeven door het (Koninklijk) Bataviaasch Genootschap van Kunsten en Wettenschappen 3: 58-62. Rouffaer, G. P. 1905 Sumatra. Geschiedenis. I. Oudste periode, tot op de komst der Hollanders (1596), in Encyclopaedie van Nederlandsch-Indië 4: 199-210, ed. John F. Snelleman, Vol. 4. ‘s-Gravenhage: Martinus Nijhoff/E. J. Brill. Rowlands, Michel, Morgan Larsen, and Kristian Kristiansen, eds. 1987 Centre and Periphery in the Ancient World. Cambridge et al.: Cambridge University Press. Schnitger, Frederic M. 1936a Hindoe-Oudheden aan de Batang Hari. Leiden: E. J. Brill.

From Distant Tales

25

1936b Hindoe-Oudheden aan de Batang Hari. Bijlage A. Utrecht: J. van Boekhoven. 1937 The Archaeology of Hindoo Sumatra. Leiden: E. J. Brill [Internationales Archiv für Ethnographie, Supplement zu Band XXXV]. 1938 The Archaeology of Hindoo Sumatra. lnternationales Archiv für Ethnographie 35 Supplement: 1-44. Leiden: E. ]. Brill. 1989 Forgotten Kingdoms in Sumatra. Singapore: Oxford University Press [Originally published 1939. Leiden: E. ]. Brill]. Scholz, U. 1987 The impact of natural conditions on the structure and development of smallholders farming in Sumatra, in Cultures and Societies of North Sumatra: 459-466, ed. Rainer Carle. Veröffentlichungen des Seminars für Indonesische und Südseesprachen der Universität Hamburg 19. Berlin: Dietrich Reimer. Soekmono 1955 Kisah perdjalanan ke Sumatra Selatan dan Djambi. Amerta 3: 1-50. Verstappen, H. T. 1973 A Geomorphological Reconnaissance of Sumatra and Adjacent Islands (Indonesia). Groningen: Wolters-Noordhoff. Veth, P. J. 1881 Midden-Sumatra, Reizen en Onderzoekingen Der SumatraExpeditie, Uitgerust door het Aardrijkskitndig Genootschap, 18771879. Leiden: E. J. Brill. Voorhoeve, Petrus 1927 Overzicht van de Volksverhalen der Bataks. Vlissingen: van de Velde. 1955 Critical Survey of Studies on the Languages of Sumatra. ‘sGravenhage: M. Nijhoff. 1961 A Catalogue of the Batak Manuscripts in the Chester Beatty Library. Dublin: Hodges Figgis & Co. Ltd. 1970 Kerintji documents. Bijdragen tot de Taal-, Land- en Volkenkunde 126: 369-399. Westenenk, Louis Constant 1921 Uit het land van Bittertong (Zuid-Soematra). Djawa 1(1): 5-11. 1922 De Hindoe-oudheden in de Pasemah-hoogvlakte (Residentie Palembang). Oudheidkundig Verslag, bijlage D: 31-37. 1932 Het Rijk van Bittertong. ‘s-Gravenhage: Leopold. Whitten, Tony, Sengli J. Damanik, Jazanul Anwar, and Nazaruddin Hisyam 2000 The Ecology of Sumatra. Hong Kong: Periplus Editions.

26

Introduction

Willer, T. J. 1849 The Battas of Mandheling and Pertibi. Journal of the Indian Archipelago and Eastern Asia 3: 366-378. Wolters, Olivier W. 1967 Early Indonesian Commerce. A Study of the Origins of Srivijaya. Ithaca: Cornell University. Wurtzburg, C. E. 1984 Raffles of the Eastern Isles. Singapore: Oxford University Press.

Fig. 1-1: Map of Sumatra (drawing by Sebastian Hageneuer)

Fig. 9-1: Distribution of megaliths in the highland of Jambi/Kerinci (source: Highland Jambi Research Project)

PART I: GENERAL

THE DAWN OF HUMANITY IN SUMATRA: ARRIVAL AND DISPERSAL FROM A HUMAN REMAINS PERSPECTIVE HARRY WIDIANTO

Introduction As a part of the Old World, the Indonesian Archipelago has been one of a few leading locations for the study of human evolution. The history of humankind on these islands can be traced back at least to the Lower Pleistocene period, some 1.5 million years ago, when a leggy human ancestor - Homo erectus - began prowling the steamy swamps and uplands of Java. Although hominids apparently evolved in Africa,1 Indonesia is a Garden of Eden in its own right, with a wealth of Homo erectus fossils. The first finds were gathered in 1891 with Eugene Dubois’ discovery of a skull-cap from the well-known Pithecanthropus erectus from Middle Pleistocene deposits in the Kendeng Hills, near the village of Trinil, East Java (Dubois 1894). This spectacular find was complemented by a left femur showing the morphological features of bipedal locomotion, proving that its owner had walked erect. Following the Trinil finds, many more Homo erectus specimens have been unearthed from hominid fossil-bearing formations in Java including those at Sangiran, Ngandong, Sambungmacan, and Modjokerto. The more than 100 individuals dug up on the island show 1

There is still controversy over the nature of human evolution and the possible role of Asia, specifically Southeast Asia, in this process. The question has not been solved. According to the “Out of Africa”-theory, Homo erectus left Africa 1.8 million years ago (e.g. Tattersal 1997), migrated and settled in other parts of The Old World. However, another theory called “multi-regional evolution” (e.g. Wolpoff 1991), suggests there is a possibility that several regions of the world served as centers of human evolution simultaneously. According to this theory, humans evolved in Southeast Asia in parallel to developments in Africa over a period of 1.5 milion years.

Harry Widianto

29

three stages of evolution during more than one million years - from archaic to progressive forms. The startling discovery of “hobbits” - the diminutive Homo florensiensis on Flores Island in 2002 - added a controversial “new” hominid to the Indonesian menagerie (Teuku Jacob et al. 2006; Matthew et al. 2007). In Indonesia human fossils have as yet only been discovered on Java, while remaining notably absent on other islands. No hominid discoveries, for instance, have been reported from neighboring Sumatra or Kalimantan. Taking into account the position of Java during recent glaciations, Sumatra and Kalimantan should have had the same opportunity to be colonized by mammals migrating from the Southeast Asian mainland, including humans. During the so-called Ice Ages, seawater froze due to drastic cooling in the northern hemisphere, resulting in sea levels declining by as much as 100-120 m and the occurrence of vast coastal regressions. Such fluctuations formed intermittent land bridges connecting the islands of the Sunda Shelf (Java, Sumatra, Kalimantan) and opening a migration route from the Southeast Asian mainland to the south. Along such a route, it might appear that Sumatra and Kalimantan should have been colonized before Java (Lower Pleistocene), but so far such evidence is missing. As a contribution to the symposium and this volume, this chapter focuses specifically on the peopling of Sumatra, with particular emphasis on the human remains. While evidence of human prehistory is scarce in Sumatra compared to the extraordinarily rich record of Java, it is nonetheless significant.

Human Evidence in Sumatra Unlike on Java where human remains abundantly relate the ancient history of the island, Sumatran remains are poor in quality and very limited in quantity. Before Dubois reached Java and was honoured by his spectacular discovery - Pithecanthropus erectus - he landed first in West Sumatra, ventured into the Minangkabau highland village of Payakumbuh, and explored caves in the Bukit Barisan hills (Shipman 2001). Dubois spent almost three years in that region and made several excavations. Unfortunately, while seeking Charles Darwin’s “missing link” in the tropical region that he considered highly suitable for anthropoid evolution, he found no real fossils in West Sumatra, just a lot of recent bones. He turned to Java when he heard of the discovery of Wadjak Man in 1889. Not long afterward, he found the second skull of this species (Wadjak II), and at last, Pithecanthropus erectus in 1891.

30

The Dawn of Humanity in Sumatra

Early Research Following Dubois, the next prehistoric studies on Sumatra were conducted in 1913 by the Swiss geologist August Tobler who undertook excavations in the karst hills between the Merangin and Batang Tabir Rivers in Jambi (Bonatz, infra). A cave named Ulu Tijanko yielded a handful of human remains, represented by cranial fragments, small teeth, a fragment of a mandible with chin eminence, and fragments of humerus, tibia, and femur. The fossa olecrani of the humerus was artificially perforated. These remains indicated a population of rather small stature and were associated with flakes, borers, arrow heads, scrapers, shell mounds (kjokkenmodinger or kitchen middens) and cores, all made of obsidian (Bonatz, infra). Research on the island was next continued by P. V. van Stein Callenfels in 1920, who excavated shell mounds in North Sumatra and described for the first time the existence of Mesolithic pebble tools in Indonesia (van Stein Callenfels 1920). He found cranial bones at one site, and sent them to Josef Wastl, who concluded that the bones showed morphological features characteristic of Papua-Melanesoid stock (Wastl 1939). From the same site, Binjai Tamiang, located near Medan on Sumatra’s east coast, H. H. E Schürmann excavated human skeletal remains including a cranium with occipital and temporal bones, fragments of three other skulls, a maxilla with six worn teeth, and about 30 fragments of limb and other bones. The cranium had a slight supra-orbital torus, and these remains were interpreted as deriving from a short-statured individual, with a dolichocranic skull, belonging to the Melanesoid stock (Wastl 1939). The skeletal remains were associated with Hoabinhian artefacts (monofacial pebble tools of an oval shape as well as pestles and mortars) and faunal remains of elephant, rhinoceros, bear, deer, tortoise, crab, and fish (Schürmann 1931). Wastl’s determination of race for these skeletal remains conforms well with sites in Vietnam and in the Province of Wellesly in the Malay Peninsula, where most pre-neolithic human remains belong to the Melanesoid racial group. Early researchers also examined urn burial sites on Sumatra. From the Lesungbatu village, Muara Danau and Tebing Tinggi in West Sumatra, for example, J. C. Noorlander reported the discovery of a couple of urn burials in 1939, each containing human skeletal fragments (Oudheidkundig Verslag 1939/1940: 13). Not much information was obtained from them, however, especially with respect to human remains.

Harry Widianto

31

Recent Research Long after initial investigations at the beginning of the 20th century, most of Sumatra still remains untouched by archaeological research. Fortunately, the situation began improving in the 1990s with the establishment of archaeological research institutions (Balai Arkeologi) in Medan (North Sumatra) and Palembang (South Sumatra). Together with the National Research and Development Center of Archaeology (Puslitbang), these two institutions have had the opportunity to undertake archaeological fieldwork, both at open sites as well as caves. Several sites containing human remains were soon discovered, among others Pondok Selabe, Gua Putri, Muara Betung, Muara Payang, Padang Sepan, and recently, Gampang Pangkalan (e.g. Simanjuntak et al. 2006; Bonatz, infra). These sites barely scratch the surface of Sumatran prehistory, I believe, but as preliminary data sources, they offer intriguing glimpses of the island’s early human occupants. Pondok Selabe is a cave formerly used for human occupation, which is located in the karst hills of the Bukit Barisan, about 35 km northwest of Baturaja, South Sumatra. The Ogan River flows nearby, while the smaller Air Tawar River - full of palaeolithic implements - penetrates the base cave. In cooperation with the Institut de Recherche pour le Developpement (IRD), the Puslitbang excavated this cave and gathered some important information on the prehistoric occupants. They found three cultural layers dating from 5,000 to 1,100 years ago. This human occupation belonged to pre-neolithic and palaeometallic periods, represented in the deepest layer by an assemblage of reworked palaeolithic tools. The cultural chronology in this cave is very clear: obsidian flakes were found in the lower level, while earthenware -both in fragments and intact - was found in the upper layer together with iron objects. In this context the most important finds at the cave are remains of at least five humans. One skeleton is preserved as an almost complete anatomical structure, consisting of cranial and long bones in a primary extended burial position. The cranium, which is heavily damaged, is in brachycephalic form with small teeth. There is no robust posture of the Pondok Selabe individual. Collectively the cranium, teeth, and long bones indicate various features of the Mongoloid race. This interpretation fits well with dating results from the second layer containing the human skeleton: 2,730 r 170 BP to 2,680 r 170 BP (Simanjuntak et al. 2006) (fig. 2-1).

32

The Dawn of Humanity in Sumatra

Fig. 2-1: Upper part of a human skeleton from Pondok Selabe cave (Baruraja, South Sumatra) showing strong features of the Mongoloid race, dating from 2,700 years ago Only 500 m to the southwest of Pondok Selabe is another habitation cave, Gua Putri. This cave is much larger and its archaeological record is more complicated than Pondok Selabe, vertically representing three layers of human occupation. The first two floors are the richest, containing obsidian and silicified limestone flakes, hammer stones, massive stone tools, and animal bones, while finds in the uppermost level consists of pottery, both plain and decorated. Two excavations conducted by the Balai Arkeologi Palembang identified human remains in the terrace of the lowest level, near the entrance. The skeleton was in a primary burial, not far from the surface at a depth of 50 cm. The skull was missing and the bones, which were heavily weathered, consisted of long bones (femur and tibia) and the two patellae. In spite of its very poor condition, the racial aspect of this skeleton is rather clearly Mongoloid. No dating is available. While caves have contributed the bulk of human remains on Sumatras, some open sites have also proved valuable. The sites of Muara Betung and Muara Payang, both located some 60 km northwest of Pagar Alam (South Sumatra), and Padang Sepan, which is located in the neighboring province of Bengkulu, for example, have revealed remains from urn burials. These include at least five skeletons found in Muara Betung, one from Muara Payang, and four from Padang Sepan (Retno Purwanti 2002; Kristantina

Harry Widianto

33

Indriastuti 2003). Urn burials were found sporadically in the three sites. Skeletons were placed inside and outside the jars, and were associated with iron objects and beads. The ones buried outside the jars in general are in extended positionss, with face upward and both arms crossed on the chest and stomach (fig. 2-2).

Fig. 2-2: Another human skeleton and skull from Padang Sepan (Bengkulu) The anatomical position of the bones remains clear, despite the fact they have been heavily decayed. The skull, humerus, radius, ulna, femur, and tibia were still in good condition and well preserved. The cranial pattern is brachycephalic with a high and round contour in norma lateralis. No depression was found in the frontal, parietal or occipital of this norma. The occipital squama is very round with no depression on the lambdatic region. Its facial aspect shows a wide and flat face, not prognathous, with

34

The Dawn of Humanity in Sumatra

small teeth and incisors presenting a shovel shape. Based on these cranial characteristics, the urn burials indicated a Mongoloid community. A second open site is also worth mentioning. An excavation conducted by the Balai Arkeologi Medan in April 2007 on the shell mound site in Gampang Pangkalan (Aceh Tamiang) resulted in the discovery of three human skeletons associated with artefacts of Hoabinhian culture (Anonym 2007). At the time when this article was written, those finds had not yet been analyzed, but based on a comparison with other shell mound sites in the region, these human remains may represent an Australomelanesoid community.

Significance of Human Evidence in Sumatra: The Peopling Period As noted earlier, research to date has yielded relatively few human remains from Sumatra. The main evidence stems from shell mounds, caves, and urn burials. The shell mounds are associated with coastal areas, while habitation caves are found in karst hills, and urn burials are known from alluvial deposits. Skeletal remains from the shell mounds of northeastern Sumatra present the dolichocephalic skull, large teeth, and robust posture characteristic of the Australomelanesoid race. Remains from dwelling caves and urn burials in the southern Barisan mountains region reflect Mongoloids with an origin very close to our own, Homo sapiens sapiens. These human materials from prehistoric sites in Sumatra suggest at least three different population groups. First, Australomelanesoids occupied the coastal areas during the first half of the Holocene. They developed artefacts of Hoabinhian affiliation, called at present by various terms such as “Technocomplex”, comparable to the pre-neolithic culture developed by the same population in the prehistoric caves of Gunung Sewu in Java (Widianto 2002). Second, an early Mongoloid group colonized the caves in the karst hills of the Barisan Mountains, characterized by flake technology using obsidian and silicified limestone. In Pondok Selabe, they lived some 2,700 years ago during the Neolithic period. Third, a later Mongoloid group is known from urn burials at open sites in the highlands. Their remains are associated with palaeometallic culture and may correspond to similar remains from Plawangan (Rembang, Central Java) and Gilimanuk (Bali) that date to the end of the prehistoric period in the first centuries CE. Based on this evidence, the human remains of Sumatra constitute a very recent inhabitation process compared to Java. Radiometric dating prospects for Sumatran sites are

Harry Widianto

35

poor, with dates having been obtained from Pondok Selabe only. Even though Austromelanesoids arrived in Indonesia around 13,000 years ago (Simanjuntak 2002), the record of human prehistory in Sumatra based on skeletal remains encompasses no more than the Holocene period, at best.

Mechanism of Sumatra’s Peopling Did Homo erectus ever set foot in Sumatra? After discussing human occupation on Sumatra, having taken into account the possibility of human arrival during the Pleistocene, I will come to the most important questions: when did humans first arrive, and how did they disperse during the occupation of Sumatra? Based on cultural chronology, the distribution of palaeolithic tools is the “oldest” evidence of Sumatra’s settlement process. In general, these implements are associated with and regarded as indicative of the presence of Homo erectus. A huge number of such stone implements were discovered from Sumatra at least since 1940, in particular from the river beds of South Sumatra, Bengkulu, and Lampung (van Heekeren 1972), including choppers, chopping tools, hand-adzes, and flakes. Recently, the evidence of palaeolithic tools has been increasing with the discovery of the FrenchIndonesian joint team from the Air Tawar and Sumuhun Rivers in the karst region of Baturaja in South Sumatra (Forestier et al. 2005). This joint team identified an important number of hand-axes, considered to show the presence of an Acheulian type in Sumatra. But in light of the absence of any fossils of this species in Sumatra, the status of paleolithic tools remains an unsolved mystery. Did Homo erectus ever live in Sumatra? Who were the makers of the palaeolithic tools widely distributed along the Kikim, Ogan, Air Tawar, and Semuhun Rivers? It is believed that the first colonization of the archipelago during the Pleistocene occurred during the Glacial Period by way of migrations over the land bridges that connected the Southeast Asian mainland to the Sundaland. In this case, Homo erectus took a migration route located somewhere east of Sumatra and arrived in Java some 1,5 million years ago (Widianto 1993). During glaciations, the present-day rivers in Sumatra and Kalimantan, such as the Musi and Kapuas, were merely the upland edges of a giant palaeo-drainage (fig. 2-3) located beneath what is now the South China Sea (Hantoro 2006).

36

The Dawn of Humanity in Sumatra

Fig. 2-3: Map of paleo-drainage located now beneath the South China Sea (Hantoro 2006: 56)

Fig. 2-4: Possible migration route of Homo erectus during the glacial periods of the Pleistocene. Sumatra was bypassed, being a highland region to the west (Sémah et al. 1990: 41)

Harry Widianto

37

The assemblage of bathymetric and palaeo-geographic data shows that the space between Sumatra and Kalimantan was formerly dominated by large river valleys. These valleys would have been a “path of least resistance” for Homo erectus travelling from Africa to Java. Evidently, they headed from northwest to southeast, so that Central Java was the first place they reached before moving eastward. Sumatra, consequently, was bypassed as a highland to the west, far away from the Homo erectus route. If this scenario is correct, there will be no traces of this species in Sumatra. Their tracks would occupy the valleys between Sumatra and Kalimantan which are now flooded by the South China Sea (fig. 2-4). If this is the case, then who were the owners of palaeolithic tools on this island? This question still presents a complicated puzzle. If Homo erectus never set foot on Sumatra, then the nearest answer should be focused once again on the Australomelanesoids. In fact, these stone implements could have been made by the same Australomelanesoids who developed the Hoabinhian culture in Southeast Asia and were cave dwellers in the first half of the Holocene on Java and Kalimantan. There is ample evidence that lithic technology during the Pleistocene underwent a serious stagnation in this archipelago. No Middle and Upper Palaeolithic is evident here. Lower Palaeolithic preserved technology shows the same evolution from the oldest levels to the most recent. For example, there are no significant differences between the oldest flakes dated to 1,2 million years ago from the Pucangan Formation in Sangiran and those from the dwelling caves of the Holocene. The problem of Sumatra’s palaeolithic tools, thus, cannot be solved by the human remains gathered from the island.

They arrived in Sumatra at the beginning of the Holocene By excluding Homo erectus from the list of Sumatra’s colonizers, only Homo sapiens - i.e., Australomelanesoids and Mongoloids – would have left proof of their presence on this island. In this case like that elsewhere in Indonesia, the Australomalanesoids on Sumatra came earlier than the Mongoloids. The Australomelanesoid race was dominant in the archipelago throughout the first half of the Holocene, particularly in the west on Sumatra and Java, and as recently found, in South Kalimantan as well (Widianto/Retno Handini 2003). Radiometric dating shows that these people lived in this area from 13,000 until at least 5,000 BP (Simanjuntak 2002). The migration of Australomelanesoids from the mainland to the south is thought to have occurred by the end of the Pleistocene, around 15,000 years ago. Other than human remains, this occupation is underlined by

38

The Dawn of Humanity in Sumatra

cultural evidence in Vietnam, Thailand, and the western part of Indonesia. Following the distribution concept, Peter Bellwood (1997) differentiates two groups of Australomelanesoids in Indonesia. The first group occupied the prehistoric caves in Flores, considered the ancestor of recent populations of that region. Included in this group are the remains of Liang Toge and those from Tanjung Pura on Morotai Island. The second group is distributed in the western part of Indonesia (i.e., Sumatra, Java, and Kalimantan). Their vast distribution shows that they were cave dwellers during the first half of the Holocene. Later, a physically different population, the Mongoloids, replaced them. The latter who are assumed to have arrived in Indonesia by at least 4,000 BP, bore Austronesian culture from the north (Taiwan) and spread rapidly to the south (the Philippines and Indonesia), before reaching the Pacific region in about 2,000 BP (Bellwood 1997). Their migration route is conncected with the “Out of Taiwan” theory, which hypothesizes that southern China, maybe Fujian or Zhejiang, was their place of origin before they moved to Taiwan. Due to their very rapid crossing to Polynesia in just 4,000 years, the “Out of Taiwan” theory is also known as the “Express Train to Polynesia” theory (Oppenheimer 1998: 162). Cave habitation, which was common among Australomelanesoids in the early Holocene, was not the only model. Habitation on open sites was sometimes more interesting, as seen from the human remains of at least 37 individuals at Gua Kepah (Kepah Cave), a burial site considered to be preneolithic, which is more related to the mollusc shell mound culture on the west coast of the Malay peninsula. The dominant character of the human remains is a mixture of Mongolid and Australomelanesid features (Teucu Jacob 1967). Taking into account the co-existence of the Mongoloids, Melanesoids, and Australoids, we can see the influence of hybridization and evolution processes working together. At the time of the arrival of Mongoloids in Sumatra, around 4,000 years ago, they did not immediately inhabit open spaces, but rather used caves as their initial habitation sites. Evidence of this is shown not only by human remains with Mongolid characteristics but also indirectly by the existence of pottery in the upper layers of their caves. Some of them emerged completely from their caves and lived at open neolithic sites, which differ significantly from one site to another, while some continued to inhabit caves until as recently as 1,000-1,800 years ago, when the use of metal had become very common in Sumatra. Radiometric dating is not available for the open neolithic sites in Sumatra, due to insufficient material at the sites including human remains.

Harry Widianto

39

In addition to prehistoric caves, data on human remains from the second half of the Holocene in Sumatra were collected from urn burial sites. The leading urn burial sites in Sumatra are in highland regions like Muara Betung, Muara Payang, and Padang Sepan. This distribution is different than that known elsewhere in the archipelago where urns have been found in coastal areas like Anyer (Banten, West Java), Plawangan (Rembang, Central Java), Gilimanuk (Bali), and Melolo (Sumba, East Nusa Tenggara). Based on the existence of artefacts commonly used as burial gifts, such as lithics, pottery, and metal objects, these urn burials are thought to correspond to a neolithic or even palaeometallic period. Various data on human remains with Mongoloid characteristics at different types of prehistoric sites on Sumatra have yielded important insights into the mobility of the Mongoloid race since their first arrival on this island.

Conclusion Taking into account bathymetry and paleo-geography, present-day Sumatra represents the highest and westernmost part of the Sundal Shelf. During the Pleistocene glaciations, in particular during initial hominid migrations from the mainland to Java, Homo erectus did not move along the current landmass of Sumatra, but rather followed the valleys of ancient rivers, which are located to the east of Sumatra and now lie submerged under the South China Sea. The peopling of Sumatra began in the early Holocene, when land bridges disappeared along the Sunda Shelf around 11,000 years ago. During the first half of the Holocene, perhaps around 8,000-10,000 years ago, the first Australomelanesoids arrived on the island and occupied coastal areas along the east coast of Sumatra while developing the Hoabinhian culture. They came from the mainland in the north, and were part of long migrations to the south, before they arrived in Java and occupied caves at Gunung Sewu, in the southern mountains. Even though they differed from the point of view of their habitation patterns - shell mounds in the coastal area in Sumatra and caves in Java - the two Australomelanesoid groups were comparable. Around 3,000 years ago, the Mongoloids arrived in Sumatra, possibly as part of a migration wave, as hypothesized in the “Out of Taiwan” theory. It seems that they replaced or absorbed the Australomelanesoids around that time, and entered caves first before they occupied open sites. Some of them, like those known from remains in Pondok Selabe and Gua Putri, still lived in caves at least until 1,000 years ago, while others went outside and built settlements on open sites. However, not a single neolithic

40

The Dawn of Humanity in Sumatra

settlement site has been found so far in Sumatra; only a handful of urn burials in the highlands remain from this period. These may represent the burials of people who built the megalithic monuments in Pasemah (Bonatz, infra; Guillaud et al., infra).

Acknowledgment I want to express my sincere gratitude to Prof. Dr. Dominik Bonatz and Dr. Mai Lin Tjoa-Bonatz for providing the support that enabled me to attend the Sumatra conference in Berlin.

References Anonym 2007 Ditemukan fosil usia 7.000 tahun: manusia purba di bukit kerang. Kompas, page 13, column 1-2. Bellwood, Peter S. 1997 Prehistory of the Indo-Malaysian Archipelago. Honolulu, Hawai’i: University of Hawai’i Press. Dubois, Eugène 1894 Pithecanthropus erectus, eine menschenähnliche Übergangsform aus Java. Batavia. Jaarboek van het Mijnwezen 24: 5-77. Forestier, Hubert, Harry Truman Simanjuntak, and Dubel Driwantoro 2005 Les premiers indices d’un faciès acheuléen à Sumatra Sud. Dossies d’Archéologie 32: 16-17. Hantoro, Wahyu 2006 Austronesian prehistory from a paleoclimatological and palaeographical perspective: settlement and migration through time in maritime-islands, in Austronesia Diaspora and the Ethnogeneses of People in Indonesian Archipelago. Jakarta: Indonesian Institute of Science: 30-59. Heekeren, H. R. van 1972 The Stone Age of Indonesia. Verhandelingen van het koninklijk, Instituut voor Taal-, Land- en Volkenkunde 61. The Hague: Martinus Nijhoof. Kristantina Indriastuti 2003 Karakteristik budaya dan pemukiman situs Muara Payang, tinjauan ekologi dan keruangan. Berita Penelitian Arkeologi 8. Palembang: Balai Arkeologi Palembang.

Harry Widianto

41

Oppenheimer, Stephen 1998 Eden in the East. Drowned Continent of Southeast Asia. London: Weidenfels and Nicholson. Oudheidkundig Verslag 1939/1940 Oudheidkundige Dienst in Nederlandsch-Indie, ed. Koninklijk Bataviaasch Genootschap van Kusten en Wetenschappen. Batavia: Kon. Drukkerij de Unie. Retno Purwanti 2002 Penguburan masa prasejarah Situs Muara Betung, Kecamatan Ulu Musi, Kabupaten Lahat, Provinsi Sumatra Selatan. Berita Penelitian Arkeologi 7. Palembang: Balai Arkeologi Palembang. Schürmann, H. M. E. 1931 Kjökkenmöddinger und Paläolithicum in Nord Sumatra. Tijdschrift Aardrijkskundig Genootschap 48(5): 905-923. Sémah, François, Anne-Marie Sémah, and Tony Djubiantono 1990 Ils ont découvert Java. They discovered Java. Mereka Menemukan Pulau Jawa. Jakarta: Pusat Penelitian Arkeologi Nasional and Musée National d’Histoire Naturelle. Shipman, Pat 2001 The Man who found the missing Link, the Extraordinary Life of Eugene Dubois and his lifelong Quest to prove Darwin Right. New York: Simon und Schuster. Simanjuntak, Truman, ed. 2002 Gunung Sewu in Prehistoric Times. Yogyakarta: Gadjah Mada University Press. Simanjuntak, Truman, Hubert Forestier, Dubel Driwantoro, Jatmiko, and Darwin Siregar 2006 Daerah kaki gunung: Berbagai tahap zaman batu, in Menyusuri Sungai, Merunut Waktu: 23-24, ed. Dominique Guillaud. Jakarta: Puslitbang-IRD-EFEO. Stein Callenfels, P. V. van 1920 Rapport over een Dienstreis door een Deel van Sumatra. Oudheidkundienst Verslag. Tattersal, Ian 1997 Out of Africa: again...and again? Scientific American 276(4): 4653. Teuku Jacob 1967 Some problems pertaining to the racial history of Indonesian region: a study of human skeletal and dental remains from several prehistoric sites in Indonesia and Malaysia. Utrecht: Drukkerij Nederlandia.

42

The Dawn of Humanity in Sumatra

Teuku Jacob, E. Indriati, R. P. Soejono, K. Hsü, D. W. Frayer, R. B. Eckhardt, A. J. Kuperavage, A. Thorne, and M. Henneberg 2006 Pygmoid Australomelanesian Homo sapiens skeletal remains from Liang Bua, Flores: population affinities and pathological abnormalities. Proceedings of the National Academy of Science 103(36). The National Academy of Science of the USA. Tocheri, Matthew W., Caley M. Orr, Susan G. Larson, Thomas Sutikna, Jatmiko, E. Wahyu Saptomo, Rokus Awe Due, Tony Djubiantono, Michael J. Morwood, and William l. Jungers. 2007 The primitive wrist of Homo floresiensis and its implication for Hominin evolution. Science 317: 1743-1745. Wastl, Josef 1939 Prähistorische Menschenreste aus dem Muschelhügel von BindjaiTamiang in Nord-Sumatra, in Kultur und Rasse. Otto Reche zum 60. Geburtstag gewidmet von Schülern und Freunden: 237-243, ed. Martin Hesch. München et al.: Lehmann. Widianto, Harry 1993 Unité et Diversité des Hominidés Fossiles de Java: Présentation de Restes Humains Fossiles Inédits. Ph.D. diss. Paris: Institut de Paléontologie Humaine, Musée National d’Histoire Naturelle. 2002 Prehistoric inhabitants of Gunung Sewu, in Gunung Sewu in Prehistoric Times: 227-248, ed. Truman Simanjuntak. Yogyakarta: Gadjah Mada University Press. Widianto, Harry and Retno Handini 2003 Karakter budaya prasejarah di kawasan Gunung Batubuli, Kalimantan Selatan: Mekanisme hunian gua pasca Plestosen. Berita Penelitian Arkeologi 12. Banjarmasin: Balai Arkeologi Banjarmasin. Wolpoff, Milford H. 1991 “Homo erectus” et les origins de la diversité humaine, in Aux Origines d’Homo Sapiens: 97-155, eds. Jean-Jacques Hublin and Anne-Marie Tillier. Paris: Presses Universitaires de France.

THE NEOLITHIC IN THE HIGHLANDS OF SUMATRA: PROBLEMS OF DEFINITION DOMINIK BONATZ

Introduction Archaeologists use the terms Paleolithic and Neolithic to refer to the earlier and later epochs of human prehistory, which can be distinguished by the types of artefacts people left behind, especially the types of stone tools they used. These periods correspond with two geological phases: the Pleistocene and the Holocene. The earliest part of the Holocene, which began about 11,500 years ago and which marks the official end of the last ice age, coincides closely with the first human experiments in the domestication of plants. The transition to agriculture marks a radical change in the development of human society and is therefore often labelled as the Neolithic Revolution (Manzanilla 1987, Price/Gebauer 1995). For V. Gordon Childe, who coined the term in the 1920s, the Neolithic Revolution was mainly an economic transformation which resulted in a fundamental shift in orientation towards stable food production strategies and their accompanying techniques of food storage, in contrast to earlier subsistence-based economies that relied wholly on foraging. Since then, archaeologists have used the term “Neolithic” in a broader sense to refer to periods of human civilization, in which farming was the main method of subsistence, sedentism was more widespread than seasonal mobility, and people relied on pottery vessels for the preparation and consumption of food, as well as on ground stone tools. The phrase Neolithic Revolution is still widely used by archaeologists, even though most of them would nowadays argue that the development to agriculture and sedentary life should be seen as gradual evolutionary process rather than a sudden cataclysm. This process started independently in different parts of the world, and it entailed major changes in the structure

44

The Neolithic in the Highlands of Sumatra: Problems of Definition

and organization of societies as well as a totally new relationship with the environment. Attempts have been made to identify zones of “pristine domestication” in the Near East, northern Africa, Mesoamerica, North America, and northern China (Minnis 1985; Smith 1989; Harlan 1995). However, the question of the spread of agriculture and animal husbandry beyond these zones is important as well. In addressing the issue, how and why the formation of Neolithic societies took place in so many places and under such different conditions, a variety of premises must be made. First, it is important to appreciate that the complex features associated with Neolithic societies such as farming, pottery production, and ground stone tools varied by region and developed at different rates. Second, changing environmental conditions and increasing social interactions do not inevitably change the economic orientation of human populations. It is therefore reasonable to question the teleological character of agriculture and sedentism which is often taken for granted. Third and finally, archaeological data are not necessarily indisputable to our interpretation of Neolithic communities. They are just means to generate and to test our set of hypothesis and reconcile often contradictory evidence. The process by which specific Neolithic communities could have emerged in island Southeast Asia will now be modelled in the case of Sumatra. With special focus on the highlands, the available archaeological data will be examined in order to define the distinct nature of a Neolithic period. However, Sumatra neither belonged to the zones of “pristine domestication” nor was it geographically directly connected to one of them. It is therefore questionable if the development of sedentary agricultural communities was an inevitable step in the prehistory of the island. Since alternative options to agriculture and sedentism may have existed for a long time up to the present, the possible causes of food production must be discussed on a theoretical level. For this purpose, different models which seek to explain the transition to agriculture are presented in the following section. At the end of this paper, the model which most closely fits the archaeological data is taken to formulate a hypothesis on the spread of Neolithic communities in the highlands of Sumatra.

Theoretical Approaches on the Origin of Agriculture In her article on the “Origins of Food Production in the New World” Barbara Stark (1986) describes three main types of models which are used by archaeologists to explain the transition to food production: “push” models, “pull” models, and “social” models (see also Bogucki 1999: 847848, fig. 21.4). “Push” models were proposed as early as the late 1960s

Dominik Bonatz

45

(e.g. Binford 1968; Flannery 1969). They claim that stress situations caused by exceeding resources due to population growth or climatic changes have pushed people to adopt agriculture. Such models have been common to explain the origins of agriculture in the Near East. For K. V. Flannery (1969), for example, population pressure was the main cause for people to move from the optimal zones of the Fertile Crescent to adjacent marginal zones where they started to cultivate cereals for the first time, that is about 10,000 years ago. With new evidence at hand, such as pollen diagrams which clearly showed that the earliest cultivating communities were situated within the natural habitat of wild cereals (van Zeist/Bottema 1991), later developed “push” models tended to prefer multiple factors for the transition to food production including changing environments, alternating subsistence strategies, and social organization (e.g. Moore 1982, 1989). Most recent finds in the Upper Euphrates region of modern Turkey prove the outstanding complexity and economic success of early Neolithic hunter-gatherer societies which even managed the construction of monumental cultic building complexes (Schmidt 2006). This evidence once again raises the question why hunter-gatherers found that the advantages of an economic system involving food production outweighed the options available to them as foragers. Some archaeologist consequently started to argue that cultural and symbolic processes are not secondary or tertiary in relation to the economic and social spheres. In fact these processes are considered as one of the main reasons to push people into agriculture (e.g. Hodder 1987; Cauvin 1994). In contrast to “push” models “pull” models do not concentrate on single stress factors but emphasize the role of increasing reliance on specific resources. The continuous exploitation of specific plants and animals led to the dependence on them. Modifications of plant-humananimal relationships therefore pulled people into agriculture. Such models were first proposed for the Mesoamerican highlands (Flannery 1968). Later they were also used to explain the origins of agriculture in the Near East, especially the Levant (Henry 1989; Bar-Yosef 1995: 65-72). Here the emergence of farming communities is seen as a socio-economic response to the forcing effects of climatic changes. The first change occurred around 12,500 years ago, in which an increase in temperature promoted a complex foraging system which heavily relied on wild cereals. This system collapsed after a second climatic change about 2,000 years later. As an option to balance the lack of wild plant resources people began to cultivate cereals in vegetationally rich areas with high water tables such as the Jordan valley, the lakeshores or riverbanks of the so-called Levantine Corridor (Bar-Yosef 1995: 70, fig. 3.4).

46

The Neolithic in the Highlands of Sumatra: Problems of Definition

Finally, “social” models stress the importance of collective food consumption for the development of food production. As ethnographic examples from foraging and cultivating societies show, many aspects of status depend on the provision of food for feasting. Therefore, the demand for food to meet social needs may have strengthened efforts to maximize food supply. Approaches to identify the social power of feasting as one cause for food production are, however, rare since the archaeological evidence is often elusive. B. Hayden (1992) has proposed such a model which he terms the “competitive feasting” model. He argues that the need to generate large amounts of desirable foods in order to stage competitive feasts would have stimulated foragers to experiment on cultivation. This hypothesis has no specific geographic reference but it is in line, for example, with observation from rice cultivating societies of East and Southeast Asia where surplus production apparently led to status and wealth (Higham 1995: 147-148). A similar approach relates surplus production to trade. C. Runnel and Tj. van Andel (1988) propose that agriculture developed in the eastern Mediterranean because a surplus of food was needed for trade or to support craftsmen who made goods for trade. The link between agriculture and sedentism is implicit to most models which aim to explain the transition to food production. This is especially true for theories on the spread of Austronesian-speaking groups into the Indonesian archipelago and Oceania. These groups are believed to have been farmers who cultivated rice and other crops, and who brought with them material innovations such as pottery and new lithic industries. These early farming communities would then meet the characteristics of a period which is called the Neolithic. The conception of the term sedentism, however, needs special reflection, and the material objects which could trace Neolithic communities still need to be identified. Before summing up the archaeological data which relate to these critical points of interpretation, a short review on the research history will be given.

The Neolithic in the Southeast Asian Archipelago: The Question of Early Farming Communities In his pioneering work “The Stone Age of Indonesia” (1972), Hendrik Robert van Heekeren puts the beginning of Neolithic period after the Mesolithic or Sub-Neolithic stage some 10,000 years or more ago. He concluded that for most parts of Indonesia archaeological evidence for the Neolithic was insufficient (van Heekeren 1972: 203). Only two artefact groups, rectangular polished adzes and paddle-and-anvil pottery, were

Dominik Bonatz

47

considered by him as indicative of the Neolithic. However, only a few sites in the northern and eastern Indonesian archipelago had yielded pottery within a distinct Neolithic context, and most of the stone adzes (fig. 3-5) were recorded without any archaeological context at all.1 When new archaeological material became available in the 1970s and 1980s, the efforts to define a Neolithic period shifted to the question of the origins and dispersal of farming communities, and the archaeological data to which agricultural activities could be attributed. The archaeological assemblages which Peter Bellwood in his research on the expansion of Austronesian-speaking peoples associates with agriculture are pottery, polished adzes, ornaments of stone and shell, spindle whorls, and villagesized settlements (Bellwood 2004: 21, fig. 7.12; 2006: 106). The documentation of at least some of these assemblages at archaeological sites in the Philippines, Malaysian and Indonesian Borneo, Sulawesi, eastern Java, and the northern Moluccas, is taken as proof of the Austronesian colonization, and with it the spread of farming communities, during the second and early first millennium BCE. However, unequivocal proof of agricultural activities is rarely given. In one site, Gua Sireh cave in northwest Sarawak, rice phytoliths were recorded in a context of around 2000 BCE (Bellwood et al. 1992; Beavitt et al. 1996). In another, Kimanis Cave in East Kalimantan, some potsherds were found that seemed to contain rice impressions (Bellwood 2006: 111). A date of 2000 BCE has been proposed for rice grains from a Maros cave in South Sulawesi (Paz 2005). However, secure evidence for rice cultivation in South Sulawesi is not before approximately 500 AD when rice still was produced on a modest scale (Bulbeck/Caldwell 2008: 15). The absence of any fragments of processed or unprocessed cereal, such as Orzya spp. (rice), at sites where recent research, including archaeobiological studies, has taken place, however, is even more striking (Barker et al. 2002: 160-161).2 Bellwood, nevertheless, believes that rice cultivation spread from the Philippines through Borneo into western Indonesia after 2,500 BCE connecting it with the dispersal of paddle impressed pottery style (Bellwood 2006: 111). He distinguishes this movement from that which carried red-slipped pottery into the eastern 1

The illustrated examples on fig. 3-5 are surface finds found in the vicinity of the excavations at Muara Payang in Sumatra, see below. 2 In other cases the evidence is just too scarce. At the west mouth of Niah Caves in Sarawak, for example, a single rice grain found in a potsherd associated with a burial yielded a date of c. 3,000 BCE (Brooks et al. 1977). Since this remained a unique find, the present excavators do not believe that the people were already cultivating rice (Doherty et al. 2000).

48

The Neolithic in the Highlands of Sumatra: Problems of Definition

Indonesian archipelago and Oceania where the farmers apparently replaced rice cultivation by tubers and fruits.3 Bellwood’s Austronesian dispersal hypothesis has attracted much attention and, with some modifications, seems to be accepted by many researchers (Bellwood/Renfrew 2002; Forestier 2003). It mainly works however for the northern sphere of the Southeast Asian archipelago, but does not provide us with a model for the spread of agricultural communities into the western part of the Indo-Malaysian archipelago (e.g. Forestier 2003: 32, fig. 3). Due to the lack of archaeological information, the latter migration remains largely theoretical. The date for the assumed arrival of Austronesian-speaking populations into Sumatra is for example considered to be quite recent at around 500 BCE (Miksic 2004: 191), and it cannot therefore be reliably stated that the beginnings of food production in this area are related to the arrival of external language groups. The challenge for archaeologists, then, is to understand why and when the transition from foraging to farming, i.e., the process of Neolithisation, took place in Sumatra. Putting this question into a framework of current archaeological research, I will now review the evidence so far at hand, starting with Sumatra in general and then focussing specially on the highland’s region.4

The Neolithic in the Prehistory of Sumatra: Current Stage of Research Fundamental to the definition of a Neolithic period in the prehistory of Sumatra is recognition of the preceding period. So far, the only recognized post-Pleistocene, pre-Neolithic tradition in Southeast Asia is the Hoabinhian (Glover 1977). The term Hoabinhian, borrowed from excavations in the 3 Note, however, that red-slipped pottery has recently been found at Pondok Silabe in Sumatra (see below). 4 This chapter draws upon the results of several archaeological investigations carried out in Sumatra during the last 20 or so years including our own archaeological project in the highlands of Jambi. This project was initiated in 2003 as a joint research project between the National Research and Development Centre of Archaeology (Puslitbang Arkenas) in Jakarta and the Institute of Ancient Near Eastern Archaeology at the Free University Berlin and was sponsored by the Swiss-Liechtenstein Foundation for Archaeological Research Abroad in Zurich. Special credit is given to those team members who have continuously contributed to the progress of the project, especially J. David Neidel (National University of Singapore), Bagyo Prasetyo (Puslitbang Arkenas, Jakarta), Mai Lin Tjoa-Bonatz (Berlin), Tri Mahaini S. Budisantosa (Balai Arkeologi Palembang), and Agus Widiatmoko (Archaeological Preservation Office Suaka, Jambi).

Dominik Bonatz

49

North Vietnamese province of Hoa Binh, refers to a “technocomplex”, rather than a culture, characterized by the ample use of pebble artefacts, especially simple pebble tools, unifacially retouched pebble tools (“Sumatraliths”), edge ground tools and short axes (Moser 2001: 30-33). Hoabinhian tools start to appear c. 14,000 BCE in Vietnam and thereafter spread over mainland Southeast Asia to the Philippines, Borneo, and Sumatra. The end of the Hoabinhian period depends much on local developments and can therefore only be dated with much variation to between the 9th and 4th millennium BCE or even much later. Some scholars have compared the Hoabinhian with the idea of the Mesolithic in Europe (Narr 1966; Gorman 1969; van Heekeren 1972). This comparison, however, is inappropriate since neither the lithic tool kits nor the ecological-economic circumstances of the Hoabinhian can be paralleled with the Mesolithic (Moser 2001: 30). Others, therefore, have reasonably questioned the applicability of chronological phases such as Mesolithic and Neolithic for Southeast Asia (Callenfells 1936; Hutterer 1985). Efforts to redefine the archaeological chronology of Southeast Asia are nonetheless limited to the point that a common pattern for periodisation can not be found within an area of such geographic and biodynamic diversity. A good example for this dilemma is the relation between the Hoabinhian and the Neolithic in Sumatra. In Sumatra, Hoabinhian sites usually refer to large shell middens (or kitchen middens) known from coastal areas in the north between Aceh and Medan (Brandt 1976; Edwards McKinnon 1990; Moser 2001: 130-131). One of the rare excavations at such an open air Hoabinhian site in Sukajadi near Medan has yielded a non-calibrated radiocarbon date of 7,340 ± 360 BP (Bronson/Glover 1984). Only recently, two Hoabinhian cave occupations have been excavated by a French-Indonesian team at the site of Tögi Ndrawa on Nias Island (Forestier et al. 2005) and at Gua Pandang near Padang Bindu in South Sumatra (Forestier et al. 2006: 183185). In addition to the typical pebble artefacts, 11 radiocarbon dates from the shell accumulation inside the cave of Tögi Ndrawa provide us with one of the longest stratigraphic sequences in an Indonesian cave site dating from about 12,000 until 1,300 BP, and thus prove the persistence of the Hoabinhian tradition in such remote places as the island of Nias. Dates from Gua Pandang range between 9,270 until 6,590 BP. Findings from these site are especially important since they prove the expansion of the Hoabinhian technocomplex through Sumatra. Two single finds from our own survey in the highlands of Jambi can now be added to this information. One is a Hoabinhian pebble tool found in the Air Hitam valley south of Lake Kerinci (Bonatz 2006: 317,

50

The Neolithic in the Highlands of Sumatra: Problems of Definition

fig. 29.10), the other is a bifacially retouched tool, apparently an adze, collected from the surface at Bukit Arat in the Serampas valley (fig. 3-1 and see below). In comparison to similar tools, e.g. from the Niah Cave (Moser 2001: pl. liv), this tool can be considered to belong to the late or post-Hoabinhian technocomplex. Both finds, the pebble tool and the adze, raise the possibility that Hoabinhian technology also spread througout the highlands.

Fig. 3-1: Late or post-Hoabinhian adze from Bukit Arat, Serampas The peoples using Hoabinhian tools are believed to have been huntergatherers since no Hoabinhian site has provided undisputed evidence of domesticated plant remains or bones of domesticated animals. The site of Tögi Ndrawa, for example, shows a spectrum of food typical for the Holocene local natural environment: wild-boar, deer, monkey, fish, seafood, and shell (Thiaridae brotia, Veneridae venus) (Forestier et al. 2005). Thus one could propose that the transition to food production marks the end of the Hoabinhian tradition, and with that the beginning of the Neolithic period. Archaeological research in other parts of Southeast Asia, however, demonstrates that foraging traditions must not have necessarily been replaced by farming. T. N. Headland and L. A. Reid (1989), for example, propose that in northern Luzon in the Philippines rice-farming populations and Negrito hunters were living in close vicinity

Dominik Bonatz

51

for at least the last 3000 years. The interaction with the Austronesian speaking farmers would have been so intensive that the Negritos adopted the language as their own (Headland/Reid 1989: 46). But they did not take over agriculture. One should therefore be cautious in putting periodical terms such as Hoabinhian, Mesolithic or Neolithic in a strict unilinear evolutionary and chronological relation. As archaeological terms they rather serve to set up criteria for the distinction of the variety of socioeconomic orientations in human civilizations and the identification of their material correlates. The main problem of inquiry, then, is how to define a Neolithic assemblage which seems to have replaced the Hoabinhian technocomplex if, on the one hand, no criteria for its classification exist, and, on the other hand, no additional information stems from archaeobiological or archaeozoological studies. To focus on this problem, two sites at the border of the Sumatran highland region will briefly be compared, before moving on to examine the highlands themselves. The first example, Tianko Panjang Cave, is situated at the distal end of a limestone promontory that borders the valley of the Tianko River. Immediately south of Tianko is a wide break in the northeastern wall of the Barisan Mountains through which the Mesumai, Merangin, and Tembesi rivers drain the actively volcanic region around Kerinci. After a short test excavation by J. Zwierzycki in the 1920s, the site was extensively excavated by a joint American-Indonesian team under the direction of Bennet Bronson and Teguh Asmar in 1974 (Bronson/Teguh Asmar 1975). The soil in a sheltered area of the cave was excavated to 2 m deep. Pottery was only found in the upper 40 cm, and thus appears only later in the cave’s occupational period. The potsherds were found mixed with obsidian blades, and obsidian blades continued to be the major find group down to the bottom of the excavations. Among the 600 plus pieces collected, a very low percentage of intentionally shaped and retouched artefacts were recognized. As formal tool types such as shouldered scrapers, borers, gravers, and arrowheads did not occur, the excavators made the still valid conclusion that taxonomic procedures for lithic artefacts will not work in central Sumatra. Unfortunately, no radiocarbon dates are given for the upper strata in Tianko Panjang Cave, and therefore do not cover those layers that contained pottery. The three samples taken are all from layers deeper than 90 cm. They provided dates between c. 9,210 and 10,250 BP (Bronson/ Teguh Asmar 1975: 136), and are thus comparable to other middle Hoabinhian sites in Sumatra and on the Southeast Asian mainland

52

The Neolithic in the Highlands of Sumatra: Problems of Definition

(especially Thailand). The problem, however, is that during the apparently long period of occupation at Tianko Panjang, no change in the use and in the shape of lithic industries was observed, while the appearance of pottery seems to be abrupt and again without a secured date. No polished adzes or other lithic artefacts typical for a Neolithic site were found, and furthermore no faunal or plant remains are recorded which could lead us to surmise that the cave-dwellers had at some point experimented with food production. In contrast, they are believed to have been successful enough at exploiting wild forest produce to have resisted true agriculture until long after other peoples of the region had fully entered the Neolithic (Bronson/ Teguh Asmar 1975: 132). In conclusion, Tianko Panjang can be characterised as a “negative” example of a site which could for a very long time be an attractive dwelling place for foraging people who from a certain moment on incorporated pottery in their lifestyle but who never experienced the traditional Neolithic achievements of agriculture and animal husbandry. Following Tianko Panjang, the second example is Pondok Selabe I which is situated in the karst region of Batu Raja near Padang Bindu in South Sumatra. Recent excavations conducted by the IRD (Institut de Recherche pour le Développement) and the Puslitbang (National Research and Development Center of Archaeology) in Jakarta have unearthed a cave site in which three different phases of occupation were clearly documented and fortunately dated by means of radiocarbon analysis (Simanjuntak/Forestier 2004; Forestier et al. 2005; 2006: 185-187). The sequence starts with a “Metallic Phase” in the upper stratum of the cave around 1,800 BP, and ends with an “aceramic Phase” in the deepest stratum, dated between 3,000 and 4,500 BP. This earliest phase yielded no pottery but in common with Tiangko Panjang did contain lithic blade implements. Based on a single radiocarbon sample, the intermediate stratum is dated to c. 2,700 BP. This phase is labelled as Neolithic since it contained potsherds incorporating regional traits of Neolithic pottery that while not known in Sumatra are typical for Neolithic sites in the Philippines, northern Borneo, Sulawesi, the Moluccas, and East Java: black and red slipped pottery, a pot with a cord decoration, and potsherds with deep incisions.5 All of the pottery is thin-walled and high fired. This pottery is really quite remarkable and raises the question of Neolithic farming communities coming from the north or the east coast, and bringing with them new technologies such as pottery making. If these 5

Personal communication with Hubert Forestier and Bagyo Prasetyo; see also Truman Simanjuntak et al. (2005: 49) with illustrations of impressed pottery.

Dominik Bonatz

53

farmers are the same who brought Austronesian languages to Sumatra then a new and very important argument can be made to stress the theory of a western route of the Austronesian expansion. For the moment however all that is speculation and it would seem more appropriate to start with a discussion about the nature of the Pondok Selabe site. Here the excavators have made a convincing suggestion (Simanjuntak et al. 2005: 49). The invention of pottery must not necessarily reflect a change in population, or a change in living circumstances of those who once settled within the cave. The absence of other typical Neolithic artefacts like the polished stone adze combined with the presence of bones from wild boar, deer, and monkeys, instead of domesticates again, like Tianko Panjang, speaks in favour of a group of hunter-gatherers who made seasonal use of the cave. The pottery then could have been an exchange good received from early farmers or horticulturalists who settled in the open terrain not far from the caves. In general, the equation between pottery and agriculture, and thus with the Neolithic, must be questioned. In many cultural contexts it is well attested that both do not necessarily relate to each other (c.f. Rice 1999). It is, for example likely that pottery was invented in northern China prior to the beginning of plant cultivation (Higham 1995: 133; Guo/Li 2000). The complex long standing hunter-gatherer cultures of Jomon in Japan and Chulmun in Korea employed pottery for thousands of years before they started to cultivate plants (Underhill/Habu 2005: 134-141). Since they also have used polished stone axes and practiced sedentary life both have been equated with Neolithic cultures in Europe, China, and elsewhere (Underhill/Habu 2005: 135). In contrast, the early Neolithic communities of the Near East had experimented for about 3000 years with techniques of food production before they started to use pottery for storage and food consumption (e.g. Thissen 2007). The problem, then, for archaeologists is how to tease out early farming groups in prehistoric environments such as Sumatra, and how to find their settlement places given that the evidence we have is often ambiguous. We see that in places where pottery has been found life and subsistence continued on Sumatra without drastic changes. Therefore, first the context and the combination of various artefacts have to be investigated before any conclusions can be made about the residential nature of a given site. Keeping this difficult starting point in mind, I will now shift to the highlands and consider the three essential types of archaeological sites so far investigated there: open air sites, burial sites, and megalithic sites. After that a second important source for archaeological information will be considered: the natural environment.

54

The Neolithic in the Highlands of Sumatra: Problems of Definition

Archaeological Sites in the Highlands of Sumatra Open air sites Much of prehistoric research depends upon the visibility of sites to archaeologists. In the landscapes of island Southeast Asia especially, open air sites are hard to identify because environmental conditions and perishable building materials such as wood and bamboo did not allow for the accumulation of cultural deposits, and the formation of mounds which could be explored by means of stratigraphic excavations. The typical situation, instead, is that which we found when excavating the site of Sungai Hangat in the highlands of Jambi in 2003 (Bonatz 2004: 118-123; 2006: 313-316; Bonatz et al. 2006: 495-496). During excavations many potsherds including Chinese porcelain from the Song Dynasty (11271279) until the Qing Dynasty (1644-1912) were collected, but since virtually none of the pieces was found in a stratified context, no chronological relationship could be established between the datable imports and the local earthenware or other finds like metals and stone implements. Most of the other sites we surveyed in the highlands of Jambi have the same problem: Different artefacts of different periods appear within the same context, thus the criteria to classify a site and its find assemblage as typical for a certain period are lacking. An exception might be the sight of Bukit Arat which has been identified in 2006 in close vicinity to Renah Kemumu in Serampas.6 It is a small mound situated at the boarder of the rice fields from where it overlooks a good part of the valley. A survey of the place provided surface finds of the late Hoabinhian tool, already mentioned here (fig. 3-1), and pieces of obsidian. Due to the importance of these finds large-scale excavations were carried out in 2008. The excavated surface of about 250 m2 yielded a huge amount of potsherds and obsidians, some flints and a set of round grinding stones (fig. 3-2). The obsidians include totals of 1178 flakes, 220 blades, 18 roughly ovoid hammers for producing the flakes and blades, about 200 pieces of core preparation debris, and 1392 chips. Thus the site exhibits all characteristics of an obsidian workshop. But it provides also evidence for domestic activities. 1040 pieces of earthenware were collected among which 55 are rims. The fabric of the potsherds is almost exclusively coarse red with many thin-walled pieces. Some show traces of paddle-marked decorations. The rims mostly stem from cooking pots but also few bowls and mediumsized containers are attested. 6 For a map of the region see Neidel, infra: fig. 13-1.For a preliminary report on the excavations at Bukit Arat see Bonatz 2009.

Dominik Bonatz

55

Fig. 3-2: Find assemblage including potsherds, obsidian blades and hammers, and stone tools from test pit 2 in Bukit Arat, Serampas The finds from Bukit Arat form an exceptionally homogeneous assemblage which, in theory, shows all the characteristics of a Neolithic assemblage. One might therefore be tempted to speculate about the nature of the site as one which marks the beginning of farming and sedentism in the Serampas valley. In order to achieve an idea of the age of the site, potsherds from different find contexts were analysed by means of thermoluminescence. They lead to surprisingly high dates between c. 1650-800 BCE.7 This date is among the earliest ever since achieved for a Neolithic site in Sumatra. It can only be compared with another recently excavated open air site at Benua Keling Lama on the Pasemah plateau, which has a “Neolithic” level dating back to c. 1550 BCE (Simanjuntak et al. 2006; Guillaud et al., infra). This level contained the fragment of a 7

The samples were analysed by CUDaM, Laboratorio di Termoluminescenza, Università degli Studi di Milano-Bicocca, using the standard fine-grain technique: BA 08-57-2: 1205 ± 270 BCE (Lab. Code D2242), BA 08-57-4: 1285 ± 265 BCE (Lab. Code D2241), BA 08-107-1: 910 ± 200 BCE (Lab. Code D2243), BA 08123-4: 1340 ± 250 BCE (Lab. Code D2245), BA 08-157-2: 1390 ± 260 BCE (Lab. Code D2246), BA 08-157-8: 1040 ± 235 BCE (Lab. Code D2247).

56

The Neolithic in the Highlands of Sumatra: Problems of Definition

polished stone tool. Considering the fact that the dates for the Neolithic in Pasemah and now also in Serampas are older than any date so far obtained from a site at the foothills or in the lowlands, one might be tempted to locate the beginnings of the process of Neolithisation in the highland regions (c.f. Guillaud et al., infra).

Burial sites In other places of Southeast Asia burial sites are by far the most promising sites to identify Neolithic assemblages. One need only recall the cave burial sites of Gua Cha in the Malaysian Peninsular and the Niah caves in Sarawak, Borneo. The Gua Cha burials date between 2,500 and 1,000 BCE. Most of the human skeleton at that site were found in extended positions, and were buried with grave goods that included stone bracelets, quadrangular-sectioned adzes, and pottery of footed, round, and flat-based forms with a predominance of cord-marked decoration (Bellwood 1997: 260-265). In the Neolithic, between about 3,000 BCE and 800 CE, the Niah Cave complex was exclusively used as a cemetery. The broad sequence of burial types includes inhumations in extended and flexed position, cremations, and jar burials in lidded jars (Bellwood 1997: 238-241; Barker et al. 2002: 149, 159-160). Typical artefacts of the Neolithic assemblage at the Niah caves are polished quadrangular and trapezoidal stone adzes, paddle-decorated types of earthenware, and the double-spouted earthenware vessels, the so-called Niah Cave vessels. For the highlands of Sumatra the main problem in finding burial places stems from unfavorable conditions for the preservation of human skeleton material. In many parts of the region the soils are acidic, leading to poor preservation of calcium, so that all bone material would quickly have been destroyed. There is, however, more than one way to identify a burial place. One distinct practice of burial, for example, is the deposition of the human corpse or cremation ash into a jar (or urn), which is quite a common practice in prehistoric burial sites of the Indo-Malayan archipelago. Among the burial types attested in the Niah Caves, for example, are also lidded burial jars. While the Niah Cave complex is a distinct Neolithic burial site, the bulk of the jar burial sites in Indonesia, are believed to coincide with the Metal period (Bellwood 1997: 306-307).

Dominik Bonatz

57

Fig. 3-3: Jar burials at Muara Payang, Lahat (printed with courtesy of Balai Arkeologi Palembang)

Fig. 3-4: Presumed jar burials at Renah Kemumu, Serampas

58

The Neolithic in the Highlands of Sumatra: Problems of Definition

At the first glance this seems to be also true for jar burial sites which have more recently been excavated by Indonesian archaeologists in the mountainous region west of Lahat in South Sumatra. The burial grounds at Muara Payang (fig. 3-3; Kristantina Indriastuti 2002, 2003) and Muara Betung (Retno Purwanti 1997, 2002; Sri Mulyati et al. 1999) are marked by unworked stones (so called dolmens). They contain classical vertical mouth-to-mouth jar burials meaning that one pot has been used as a cover or lid for the other pot.8 Human skeleton material inside the jars is very rare but in Muara Betung the teeth and fragments of a human skull have been found in one of the jars. These interments are, however, remarkable for the appearance of skeletons in extended position which were placed beside the jars. Thus, we are dealing either with different burial rites or one rite which includes a primary and a secondary burial. Unfortunately, no analysis of bones or soil samples has yet been made. Absolute dates for these sites are therefore unavailable. The ceramic assemblage which was found in the area of the burial grounds is also not very helpful for precise dating. It includes Chinese porcelain which ranges from the Song until Qing period, i.e., from the 11th century onwards. Another jar burial complex was excavated during our 2005 field campaign in Renah Kemumu in Serampas (fig. 3-4; see also Bonatz et al. 2006: 500-502, fig. 10-11), which lies only a short distance from the Bukit Arat site. I have to admit that we cannot claim to be absolutely sure of our interpretation of Renah Kemumu as a burial ground, since no traces of human bones could be identified even with the help of chemical analysis. The jars, of which 35 have been mapped and four excavated, range from 0.60-1.20 m in diameter. They have no visible lids, thus the question how they were covered remains open. Among the body and rim sherds fallen into the jar there are also sherds of much smaller vessels which could be interpreted as funerary gifts. Conventional radiocarbon dating of a charcoal sample taken from the fill of one jar yielded a date of 1100 ± 120 BP.9 In addition, optically-stimulated luminescence (OSL) measurement of one potsherd taken from the body of a jar dated this piece to 813 ± 128 CE, and a second piece from one of the smaller vessels to 1039 ± 124 CE.10 Sufficient evidence exists, therefore, to conclude that the 8

For similar jar burial sites in the coastal area see Padang Sepan north of Bengkulu (Kristantina Indriastuti 2002a, 2003a, 2004) and Lebakbandung near Jambi (Eddy Sunarto et al. 1996). 9 Non-calibrated radiocarbon date, without Geolab-number Radiocarbon Dating Laboratory Bandung 10 The analysis of the potsherd samples was conducted by C. Goedicke of the Rathgen Laboratory at the Staatliche Museen in Berlin, using the single grain OSL system.

Dominik Bonatz

59

burial ground at Renah Kemumu was used between 800 and 1,100 CE. This, however, is apparently not the date which is commonly associated with a Neolithic burial site. Instead it falls within the range of other jar burial sites in Indonesia and those from South Sumatra which certainly do not date back to before the first millennium CE and which are commonly associated with the Metal period. One should, however, be cautious with one-sided conclusions based on absolute dates. Jar burial sites like Renah Kemumu are rooted in a tradition which outside of Sumatra, i.e., Niah in Sarawak, goes back to the late Neolithic period around 1500 BCE. In Muara Payang (fig. 3-5) and Padang Sepan (at the foothills north of Bengkulu) polished stone adzes have been found in close vicinity of burial sites, therefore the Neolithic tradition seems to be almost tangible at these places. Last but not least, the already mentioned excavation of the French-Indonesian team at Benua Keling in Pasemah has shown how a Neolithic site changed its appearance to a Paleometallic site before it was finally converted into a large burial mound complex (Guillaud et al., infra).

Fig. 3-5: Two polished stone adzes from Muara Payang, Lahat (printed with courtesy of Balai Arkeologi Palembang)

60

The Neolithic in the Highlands of Sumatra: Problems of Definition

What follows is that jar burial sites can be related to Neolithic traditions and that, concerning the nomenclature, it only depends on the point of view whether to give prevalence to the either Neolithic traditions or Metallic innovations at a particular site. The decision, however, is in itself problematic since it is often ideological. The terminology should not be completely based on the evaluation of material culture which in the Sumatran case seems to be a problematic instrument for the division into periods. Jar burial sites on Sumatra can not be classified as Neolithic assemblages, but they might have been still embedded in a context of Neolithisation which could have been a long lasting process.

Megalithic sites Megaliths are by far the most well known archaeological remains of Sumatra. In the highlands they are clustered over four main regions: The Batak-region, the Minangkabau-area, the highlands of Jambi, including the Kerinci region, and Pasemah.11 Each region has its own distinct type of megalithic setting. Thus, there is no reason to believe that one “megalithic culture” has directly influenced the other. Excavations at megalithic sites are still very rare. They prove, however, that none of the megalithic complexes dates to earlier than the first millennium CE, a fact which was formerly only indicated by the iconographic evidence of the Pasemah megaliths. Only one site in the Minangkabau region has yielded, though enigmatically, evidence for an earlier date. In Bawah Parit in the Mahat valley Indonesian archaeologist have dug up human skeletons buried under two kris-hilt shaped stones with carved decoration (Miksic 1986, 2004). The analysis of the bone material provided a 14C date of 2,0702,130 BP (Aziz/Siregar 1997: 20). This early date and the exceptional use of the megaliths as tombstones are difficult to explain. Also elusive is the explanation for several polished stone adzes which were found beside, not under the stones, apparently in conjunction with pottery of recent date (Miksic 2004: 196). Can the stone adzes, the burials, and the upright stones be related in a chronological and functional way, or do they appear next to each other coincidently? The bulk of the other megaliths in the Mahat and Sinamar valley date certainly much later and overlap with the time when king Adityawarman took control over the region in the 14th century (Miksic 1987, 2004). That 11

Concerning the megaliths in Pasemah, the highlands of Jambi and the Minangkabau area see also the contributions of Guillaud et al., infra and TjoaBonatz, infra and Bakels, infra.

Dominik Bonatz

61

means that connections with the historical period of Srivijaya-Malayu in the lowlands can be claimed.12 The same can be suggested for the megalithic sites in Pasemah and those documented in the highlands of Jambi.

Fig. 3-6: Excavated area with postholes of a house next to the megalith at Pondok, Kerinci Two excavated sites in Pondok in the Kerinci region (fig. 3-6) and Bukit Batu Larung near Renah Kemumu in the Serampas region (fig. 3-7) show that a single megalith (batu larung) was erected in the centre of a settlement place and in conjunction with a stilted house which could be interpreted either as a domestic building or a meeting hall (Bonatz 2006: 317-322, fig. 29.11-29.13; Bonatz et al. 2006: 495, figs. 1-2; 497-500, fig. 6-7). Both sites have yielded a rich material culture including imports of Chinese porcelain, Indo-Pacific glass beads, and iron tools (Bonatz 2006: 318-322, fig. 29.14-29.16; Bonatz et al. 2006: 499-500). These finds prove exchange contacts with the harbour sites on the east and probably also the west coast of Sumatra. Given the date of the megalithic sites, which has been confirmed by radiocarbon dating and OSL-analysis of potsherds, trade was occurring around the 12th century CE, and thus during the period 12 For Adityawarman’s rule over the Minangkabau highland see also the contribution of Kulke, infra.

62

The Neolithic in the Highlands of Sumatra: Problems of Definition

when the capital of the Srivijaya-Malayu emporium had been based at Muara Jambi. A similar connection with the lowland empires can be assumed for the Pasemah megalithic sites. Given to the Dongson-style kettledrums, metal ornaments, and weapons which are depicted on these stones, the Pasemah complex can roughly be dated between 200 BCE and 500 CE or even later.13 That means it could at least partly coincide with the raise of Srivijaya’s first city-state at Palembang in the 7th century CE (c.f. Manguin, infra).

Fig. 3-7: Megalith in the centre of the excavated area at Bukit Batu Larung, Serampas What does this connection of megalithic sites to historical kingdoms and thus to regions of “classical archaeology” in Indonesia - tells us about the nature of these highland sites? First, it explains why certain items such 13

Two of these stones (Guillaud et al., infra: fig. 18-2 and 18-3) depict scenes in which bronze kettledrums, certainly one of the most prestigious goods by that time, were brought to the highlands and given to the locals. These scenes apparently mirror the ritual of exchange between the highlands and the lowlands. The phenomenon of erecting stone monuments seems to be ultimately linked to this exchange.

Dominik Bonatz

63

as metals, including prestigious bronzes like the Dongson kettledrums and the bronze flask from Kerinci,14 Chinese porcelain, and glass beads appeared there for the first time. Second, it suggests that the erection of the megaliths could in some way have been related to the issues of trade and by extension the social-economic changes which trade caused in this region. And third, it demonstrates that at least some communities had advanced to such a prosperous level that they committed themselves to the erection of a megalith which may imply a high degree of territoriality. I would even like to go further with this statement and add that these are in fact the megaliths which give the first firm evidence for sedentism in the highlands, even if this conclusion does not necessarily imply agriculture. It is doubtful if the subsistence of the people who erected the megaliths was already fully based on farming, or if it was still in the process of transition from foraging to farming. I raise this question because I would like to stress that material culture is only the secondary trait of the phenomenon called the Neolithic while subsistence is the primary trait of this phenomenon. In the case of highland Sumatra one should not inevitably classify a site as post-Neolithic, i.e., Metallic or early Metallic, because some metal finds have been made there. In specific historical situations one should rather expect that different material objects which are considered to be typical for certain archaeological periods such as Paleolithic, Neolithic or Metallic may have coexisted. Obsidian flakes, for example appear as early as the late Paleolithic in Sumatra but they continued to be used there without greater typological variation until modern times. The designation of highland Neolithic communities, therefore, does not exclude the possibility that megaliths and metal belonged to their specific cultural traits as did pottery and stone implements. Once again it must be said that the definition of the Neolithic does not work on the basis of material traits but only in correspondence with the reconstruction of the subsistence strategies of past communities.

Prehistoric Subsistence Strategies in the Highlands Two palynological studies conducted in Kerinci suggest that forest clearance for swidden agriculture started around 4,000 BP (Morley 1982) or 7,000 BP (Flenley/Butler 2001), although these conclusions remain 14 A bronze fragment, what appears to have been part of a kettledrum, and the famous bronze flask from south of lake Kerinci were already recorded in the 1930s; see van der Hoop (1940: 2001, pl. 80.1) and Wolters (1967: 60-61) on the kettledrum fragment; Glover (1990: 218-220, 224, fig. 1, pl. 5) and Glover (2004, 2008) on the bronze flask.

64

The Neolithic in the Highlands of Sumatra: Problems of Definition

speculative. Palynological studies in northern Sumatra, i.e., in the Toba Lake area lead to the general conclusion that humans have been clearing forest in the upland areas for some 7,000 years or more (Flenley 1988). At first this was for shifting cultivation and at a level which allowed complete forest regeneration (Flenely 1988: 187). Only from about 2,000 BP, do permanent clearings begin to be made. A possible tree cultivation phase, attested to so far only by the pollen record of Arenga, the sugar palm, is in turn succeeded by today’s intensive rice cultivation. In my opinion the environmental studies, as limited as they are, prove that early forest disturbance was not necessarily caused by farming but by foraging, which included such activities as controlled burning of vegetation, gathering, and protective tending. Only the evidence for permanent forest clearance 2,000 years ago indicates the transition to a true agro-ecosystem, i.e., the process of Neolithisation.15 We are thus envisaging a date which is not so far from that indicated by the archaeological record. It is furthermore important to note that the beginnings of farming were apparently not marked by the introduction of rice or other cereals but by vegetatively reproduced root crops, especially yams (Discorea sp.) and taro (Colocasia and Alocasia), which unfortunately leaves no archaeological trace. This means that the transition from foraging to food production was characterized by what David R. Harris (1973) called the “palaeotechnic tropical agriculture” in which diverse assemblages of crops and animals are raised in structural and functional interdependence, mirroring the complex structure of the natural ecosystem.16 Shifting agriculture is the characteristic of this subsistence technique. It seems that in most parts of the Sumatran highlands this type of agriculture was practiced until very recently before fixed field agriculture, i.e., wet-rice cultivation, was introduced as an alternative source of food production.17 Given the possibility that sedentary communities might have practiced swidden agriculture, the status of “Neolithic” communities must be inferred from their settlement-subsistence strategies. The archaeological 15

For an elaborated description of the evolutionary sequence of plant cultivation see Harris (1989). 16 See also Glover (1977: 155-158). 17 In the Kerinci region, for example, evidence for wet-rice cultivation only goes back to Thomas Barnes’ expedition in 1818 (Kathirithamby-Wells 1986). Even today in some highland regions dry rice cultivation in the upland fields appears as a modern variation of shifting agriculture since it is part of a rotational agroforestry system, i.e., rice and annuals being replaced by coffee and other crops. See Neidel (2006: 140, fn. #130 “i” on p. 408-409).

Dominik Bonatz

65

settlement sites of the highlands fit in well to the idea of communities practicing shifting or swidden agriculture because they reflect either a mobile settlement pattern, or they form a sort of communal place for people living not in villages but on their fields. Megalithic sites are the most probable candidates for such sites, but other settlements fall into this category as well.18 A high residential mobility would then still have been an important factor for the economic orientation of the people who we designate farmers. Therefore, I propose to consider the Neolithic in the highlands of Sumatra as a period of long-term changes in economic and socio-politic complexity. Such changes include the adaptation of flexible cultivation strategies, an at least semi-sedentary lifestyle, and a prolific material culture. The circumstances under which the process of Neolithisation has been accomplished in the highlands still lay beyond our scopes. But it may be suggested that it well extends to the time when early state polities became established in the low- and midlands.

Conclusion Considering both the archaeological and the environmental evidence discussed so far, there is no reason to believe that the beginning of the Neolithic in the Sumatran highlands was marked by the introduction of fixed field agriculture. Instead, it seems that the different regions of the highlands independently developed shifting or swidden agriculture on the basis of indigenous plants. Sedentary life-style, then, was adopted up to the point that constant movements within a broader regional ecosystem still allowed flexible subsistence strategies. These strategies probably included early farming and horticulture as well as hunting and collecting. Based on a mobile settlement pattern, first complex settlement systems might not have started much earlier than the beginning of the first millennium BCE, and it even appears that in some regions like Serampas they only developed at the end of this millennium. Assuming that the transition to food production in combination with first steps towards sedentary life correlate with the process of Neolithisation, the main question is not why this process started so late but why it occurred at all. The answer to this question, as speculative as at it might be at the moment, lies in the archaeological evidence. All the regions in question share from a certain moment onwards, which varies within a time span of about 500 years, the same basic features of permanent settlement activities such as the marking of territories and settlement places by burial grounds and 18

See Neidel, infra with references.

66

The Neolithic in the Highlands of Sumatra: Problems of Definition

megaliths, the accumulation of local earthenwares and stone implements at certain places, and in some instances the appearance of metals, beads, and import ceramics at the same places. The latter group of finds, as it already has been said, put the highland settlements in a trade or at least economic exchange relation with the low- and midland polities of Srivijaya-Malayu. The “pull” model, which is one of the models discussed at the beginning of this paper, seems to me plausible enough to explain the way in which the archaeological evidence relates to the definition of the Neolithic and the question of its occurrence in the highlands. In contrast, no evidence exists for stress factors caused by exceeding resources or drastic climatic changes which would have pushed the people into agriculture. And no “competitive feasting” or surplus production can be inferred from the archaeological data as to explain the investments in food production. Instead, one may reasonably assume that the protection of indigenous plants which were continuously exploited not only contributed to their genetic manipulation but also stimulated humans to cultivate them. In addition to this basic argument of a “pull” model I would like to argue in favour of a situation in which economic exchanges with the lowlands increased the reliance of the highland populations on specific resources, for example their tradable forest products, and thus pulled them into agriculture. Since trade has always been something to do with territoriality, i.e., the claim on territories from which trade goods stem or through which they are traded, it can be proposed as one of the main causes for the orientation towards sedentary life. The “pull” model may therefore not only put emphasis on economic, material, and social changes caused by external relations, it could also help to accept that the formation of early agricultural societies and early state-organized societies could have been intertwined and that therefore, in the case of Sumatra, a Neolithic Period in the highlands could very well have coexisted with a Classical Period in the coastal areas. An unlinear approach to models of cultural evolution in Sumatra is for that reasons just as hazardous as for other regions in the Southeast Asian archipelago.

Dominik Bonatz

67

References Fadhila A. Aziz and Darwin A. Siregar 1997 Pertanggalan kronometrik sisa rangka manusia dari situs Bawah Parit, Mahat, Sumatera Barat. Siddhayatra 1(2): 12-22. Bar-Yosef, Offer and Richard H. Meadow 1995 The origins of agriculture in the Near East, in Last Hunters First Farmers: 39-94, eds. T. Douglas Price and Anne Birgitte Gebauer. Santa Fee: School of American Research Press. Barker, Graeme et al. 2002 Prehistoric foragers and farmers in South-East Asia: Renewed investigations at Niah Cave, Sarawak. Proceedings of the Prehistoric Society 68: 147-164. Beavitt, P., E. Kurui, and G. Thompson 1996 Confirmation of an early date for the presence of rice in Borneo: preliminary evidence for possible Bidayuh/Asian links. Borneo Research Bulletin 27: 29-38. Bellwood, Peter 1997 Prehistory of the Indo-Malaysian Archipelago. [2nd ed.] Honolulu: University of Hawai’i Press. 2004 The origins and dispersals of agricultural communities in Southeast Asia, in Southeast Asia. From Prehistory to History: 21-40, eds. Ian Glover and Peter Bellwood. London: Routledge/Curzon. 2006 Asian farming diasporas? Agriculture, languages, and genes in China and Southeast Asia, in Archaeology of Asia: 96-118, ed. Miriam T. Stark. Malden. MA/Oxford/Carlton: Blackwell Publishing. Bellwood, Peter, R. Gillespie, Gillian Thompson, I. Wayan Ardika, and Ipoi Datan 1992 New dates for prehistoric Asian rice. Asian Perspectives 31(2): 161-170. Bellwood, Peter and Collin Renfrew 2002 Examining the Farming Language-/Farming Dispersal Hypothesis. Oxford: Oxbow Books: McDonald Institute for Archaeological Research. Binford, Lewis 1968 Post-Pleistocene adaptations, in New Perspectives in Archaeology: 313-341, eds. Sally Binford and Lewis Binford. Chicago: Aldine. Bogucki, Peter 1999 Early agricultural societies, in Companion Encyclopedia of Archaeology, vol. 2: 839-869, ed. Graeme Barker. London/New York: Routledge.

68

The Neolithic in the Highlands of Sumatra: Problems of Definition

Bonatz, Dominik 2004 Archäologische Forschungen in Kerinci (Indonesien). Bericht der ersten Kampagne, 2003. SLSA Jahresbericht 2003: 117-136. 2006 Kerinci - Archaeological research in the highlands of Jambi on Sumatra, in Uncovering Southeast Asia’s Past. Selected Papers from the 10th International Conference of the European Association of Southeast Asian Archaeologists: 310-324, eds. Elizabeth A. Bacus, Ian C. Glover, and Vincent C. Pigott. Singapore: NUS Press. 2009 Archäologische Forschungen in Hochland von Jambi, Indonesien. Bericht der vierten Kampagne, 2008. Swiss-Liechtenstein Foundation for Archaeological Research Abroad (SLSA) Jahresbericht 2009 (in print). Bonatz, Dominik, J. David Neidel, and Mai Lin Tjoa-Bonatz 2006 The megalithic complex of highland Jambi. An archaeological perspective. Bijdragen tot de Taal-, Land- en Volkenkunde 162(4): 490-522. Brandt, R. W. 1976 The Hoabinhian of Sumatra: Some remarks. Modern Quaterly Research in Southeast Asia 2: 49-52. Bronson, Bennet and Teguh Asmar 1975 Prehistoric investigations at Tianko Panjang Cave, Sumatra. An interim report. Asian Perspectives 18: 128-145. Bronson, Bennet and Ian Glover 1984 Archaeological radiocarbon dates from Indonesia. Indonesia Circle 34. Brooks, S. T., R. Helgar, and R. H. Brookes 1977 Radiocarbon dating and palaeoserology of a selected burial series from the Great Cave of Niah, Sarawak. Asian Perspectives 20: 2131. Bulbeck, David and Ian Caldwell 2008 Oryza Sativa and the origins of kingdoms in South Sulawesi, Indonesia. Evidence from rice husk phytoliths. Indonesia and the Malay World 36 (104): 1-20. Callenfels, P. B. van Stein 1936 The melanesoid civilization of Eastern Asia. Bulletin of the Raffles Museum, Series B1, no. 1: 41-51. Cauvin, Jacques 1994 Naissance des Divinités, Naissance de l’Agriculture. La révolution des symboles au Néolithique. Paris: CNRS.

Dominik Bonatz

69

Doherty, C., P. Beavit, and E. Kuruia 2000 Recent observations of rice temper in pottery from Niah and other sites in Sarawak. Indo-Pacific Prehistory: the Melaka Papers 4: 147-152. Eddy Sunarto et al. 1996 Hasil Ekskavasi Penyelamatan Situs Lebakbandung, Kecamatan Jelutung, Kotamadia Jambi, Propinsi Jambi. Suaka Peninggalan Sejarah dan Purbakala Propinsi Jambi, Sumatera Selatan dan Bengkulu. Edwards McKinnon, Edmund 1990 The Hoabinhian in the Wampu/Lau Biang Valley of northeastern Sumatra: An update. Bulletin Indo-Pacific Prehistorian Assocociation 10: 132-142. Flannery, K. V. 1969 Origins and ecological effects of early domestication in Iran and the Near East, in The Domestication and Exploitation of Plants and Animals: 73-100, eds. Peter J. Ucko and Geoffrey W. Dimbleby. London: Duckworth. Flenley, J. R. 1988 Palynological evidence for land use changes in South-East Asia. Journal of Biogeography 15: 185-197. Flenley, J. R. and K. Butler 2001 Evidence for continued disturbance of upland rain forest in Sumatra for the last 7000 Years of an 11,000 year record. Palaeography, Palaeoclimatology, Palaeocology 171: 289-305. Forestier, Hubert 2003 Les butineurs d’îles, d’Asie en Océanie, in Îles Rêves: Territoires et Identités en Crise dans le Pacifique Insulaire: 27-53, Séminaire Affirmation Identitaire et Territoriale dans l’Aire Asie Pacifique 1999-2000, sous la dir. de Dominque Guillaud. Paris Sorbonne: Presses de l’université. Forstier, Hubert, Arnoult Seveau, Valéry Zeitoun, Dubel Driwantor, and Chinnawut Winayalai 2005 Prospections paléolithiques et perspectives technologiques pour redéfinir le hoabinhien du Nord de la Thaïlande (campagnes 20022005). Aséanie 15: 33-60. Forestier, Hubert, Dubel Driwantoro, Dominique Guillaud, Budiman, and Darwin Siregar 2006 New data for the prehistoric chronology of South Sumatra, in Archaeology: Indonesian Perspective, Festschrift R. P. Soejono: 177-192, eds. Truman Simanjuntak, M. Hisyam, Bagyo Prasetyo, T. Surti Nastiti. Jakarta: LIPI.

70

The Neolithic in the Highlands of Sumatra: Problems of Definition

Glover, Ian 1977 The Hoabinhian: Hunter-gatherers or early agriculturalists in South-East Asia?, in Hunters, Gatherers and first Farmers beyond Europe. An Archaeological Survey: 145-166, ed. John V. Megaw. Leicester: University Press. 1990 Large bronze urns in Southeast Asia. Some new finds and a reappraisal. Southeast Asian Archaeology: 209-225. 2004 Decorated bronze urns. Splendid enigmas from Southeast Asia later prehistory. Arts et Cultures: 173-193. 2008 Bronzes en marge de la culture de Dong So, in Art Ancien du Viet Nam - Bronzes et céramiques: 31-45, eds. M. Crick and H. Loveday. Geneva: Collections Baur. 1969 Hoabinhian: A pebble-tool complex with early plant associations in Southeast Asia. Science 163: 671-673. Guo, Ruihai and Li Jun 2000 Cong Nanzhuangtou Yizhi Kan Huabei Diqu Nongye He Taoqi De Qiyuan (The origins of farming and pottery in northern China from the perspective of the Nanzhuangtou site), in Dao Zuo Taoqi He Dushi de Qiyuan (The Origins of Rice Farming, Pottery, and Cities): 51-63, eds. Yan Wenming and Yasuda Yoshinori. Beijing: Wenwu Press. Harlan, Jack R. 1995 The Living Fields. Our Agricultural Heritage. Cambridge: Cambridge University Press. Harris, David R. 1973 The prehistory of tropical agriculture: An ethno-ecological model, in The Explanation of Culture Change. Models in Prehistory: 391417, ed. Colin Renfrew. Pittsburgh: University of Pittsburgh Press. 1989 An evolutionary continuum of people-plant interaction, in Foraging ad Farming. The Evolution of Plant Exploitation: 11-26, eds. David R. Harris and Gordon C. Hillman. London/Boston: Unwin Hyman. Hayden, B. 1992 Models of domestication, in Transition to Agriculture in Prehistory: 11-19, eds. Anne B. Gebauer and T. Douglas Price. Madison: Prehistory Press. Headland, T. N. and L. A. Reid 1985 Hunter-gatherers and their neighbors from prehistory to the present. Current Anthropology 30(1): 53-70. Heekeren, Hendrik Robert van 1972 The Stone Age of Indonesia. [2nd ed.] The Hague: Martinus Nijhoff.

Dominik Bonatz

71

Henry, D. 1989 From Foraging to Agriculture. The Levant at the End of the Ice Age. Philadelphia: University of Pennsylvania Press. Higham, Charles 1995 The transition to rice cultivation in Southeast Asia, in Last Hunters First Farmers: 127-155, eds. T. Douglas Price and Anne Birgitte Gebauer. Santa Fee: School of American Research Press. Hodder, Ian 1987 Contextual Archaeology: An Interpretation of Catal Hüyük and a Discussion of the Origins of Agriculture. University College London, Institute of Archaeology Bulletin 24: 43-56 Hoop, Abraham N. A. Th. à. Th. van der 1940 Prehistoric site near the Lake Kerinchi (Sumatra), in Proceedings of the Third Congress of Prehistorians of the Far East: 200-204, eds. Frederick N. Chasen and Michael W. F. Tweedie. Singapore: Government Press. Kristantina Indriastuti 2002a Tempayan kubur, budaya prasejarah situs Padang Sepan, kabupaten Benkulu Utara. Siddhayatra 7(2): 50-64. 2002b Laporan Penelitian Ekskavasi Kubur Tempayan Situs Muara Payan, Kecamatan Jarai, Kabupaten Lahat. Balai Arkeologi Palembang. 2003a Lapon Penelitian Arkeologi Ekskavasi Kubur Tempayan, Situs Padang Sepan, Kecamatan Air Besi, Kabubaten Bengkulu Utara. Balai Arkeologi Palembang. 2003b Karakteristik budaya dan pemukiman situs Muara Payang. Berita Penelitian Arkeologi 8. Palembang. 2004 Pemukiman megalitik di wilayah propinsi Bengkulu. Berita Penelitian Arkeologi 11. Palembang. Kathirithamby-Wells, Jeyamalar 1986 Report of a journey to the interior of the island of Sumatra, in Thomas Barnes’ Expedition to Kerinci in 1818: 32-71, ed. Jayamalar Kathirithamby-Wells, Occasional Paper no. 7. Canterbury: University of Kent at Canterbury, Centre of Southeast Asian Studies. Manzanilla, Linda, ed. 1987 Studies in Neolithic and Urban Revolutions. Oxford: BAR. Miksic, John 1986 A valley of megaliths in West Sumatra; Mahat (Schnitger’s Aoer Doeri) revisited. Journal of the Malaysian Branch of the Royal Asiatic Society 59: 27-32.

72

The Neolithic in the Highlands of Sumatra: Problems of Definition

1987 From Seri Vijaya to Melaka; Batu Tagak in historical and cultural context. Journal of the Malaysian Branch of the Royal Asiatic Society 60(2): 1-42. 2004 From megaliths to tombstones; the transition from prehistory to the early Islamic period in highland West Sumatra. Indonesia and the Malay World 32(93): 191-210. Minnis, P. 1985 Domesticating people and plants in the greater Southwest, in Prehistoric Food Production in North America: 309-339, ed. Richard I. Ford, Anthropological Papers no. 75. Ann Arbor: Museum of Anthropology, University of Michigan. Moore, A. M. T. 1982 Agricultural origins in the Near East: A model for the 1980s. World Archaeology 14: 224-236. 1989 The transition from foraging to farming in Southwest Asia: Present problems and future directions, in Foraging and Farming. The Evolution of Plant Exploitation: 620-631, eds. David R. Harris and Gordon C. Hillman. London: Unwin Hyman. Morley, R. J. 1982 A Palaeological interpretation of a 10,000 year pollen record from Danau Padang, Central Sumatra, Indonesia. Journal of Biogeography 9: 151-190. Moser, Johannes 2001 Hoabinhian. Geographie und Chronologie eines steinzeitlichen Technokomplexes in Südostasien. AVA-Forschungen 6, Köln: Linden-Soft. Narr, Karl J. 1966 Die frühe und mittlere Altsteinzeit Süd- und Ostasiens. Handbuch für Urgeschichte I. Bern: Francke. Neidel, J. David 2006 The Garden of forking Paths: History. Its Erasure and Remembrance in Sumatra’s Kerinci Seblat National Park. Ph.D. diss. Yale University. Paz, V. 2005 Rock shelters, caves, and archaeobotany in island Southeast Asia. Asian Perspectives 44: 107-118. Price, T. Douglas and Anne Birgitte Gebauer 1995 New perspectives on the transition to agriculture, in Last Hunters First Farmers: 3-19, eds. T. Douglas Price and Anne Birgitte Gebauer. Santa Fee: School of American Research Press.

Dominik Bonatz

73

Retno Purwanti 1997 Laporan Penelitian Arkeologi Ekskavasi Situs Kubur Tempayan Desa Muara Betung, Kecamatan Ulu Musi, Kabupaten Lahat, Propinsi Sumatera Selatan. Palembang: Departemen Pendidikan dan Kebudayaan Pusat Penelitian Arkeologi Nasional Balai Arkeologi Palembang. 2002 Penguburan Masa Prasejarah Situs Muara Betung, Kecamatan Kabupaten Lahat, Propinsi Sumatera Selatan. Berita Penelitian Arkeologi 7. Balai Arkeologi Palembang. Rice, P. 1999 On the origins of pottery. Journal of Archaeological Theory and Method 6: 1-54. Runnels, C. and Tj. van Andel 1988 Trade and the origins of agriculture in the Eastern Mediterranean. Journal of Mediterranean Archaeology 1: 83-109. Schmidt, Klaus 2006 Sie bauten die ersten Tempel. Das rätselhafte Heiligtum der Steinzeitjäger. München: C. H. Beck. Simanjuntak, Truman and Hubert Forestier 2004 Research progress on the neolithic in Indonesia: Special reference to the Pondok Silabe Cave, South Sumatra, in Southeast Asian Archaeology, Wilhelm G. Solheim II Festschrift: 104-119, ed. Victor Paz. Diliman: The University of the Philippines Press. Simanjuntak, Truman, Hubert Forestier, and Bagyo Prasetyo 2005 Gens des karsts au Néolithique à Sumatra, in Asie du Sud-Est, De l’homo erectus a l’homo sapiens. Dossiers d’Archéologie 302: 4649. Simanjuntak, Truman, Hubert Forestier, Dubel Driwantoro, Jatmiko, and Darwin Siregar 2006 Daerah kaki gunung: Berbagai tahap zaman batu, in Menyelusuri Sungai, Merunut Waktu. Penelitian Arkeologi di Sumatera Selatan: 23-35, ed. Dominique Guillaud. Jakarta: Puslitbang-IRD-EFEO. Smith, B. D. 1989 Origin of agriculture in eastern North America. Science 246: 15661571. Sri Mulyati et al. 1999 Ekskavasi Penyelamatan Situs Kubur Tempayan di Desa Kunduran dan Muara Betung, Kecamatan Ulu Musi, Kabupaten Lahat, Propinsi Sumatera Selatan. Suaka Peninggalan Sejarah dan Purbakala Propinsi Jambi, Sumatera Selatan dan Bengkulu.

74

The Neolithic in the Highlands of Sumatra: Problems of Definition

Stark, Barbara 1986 Origins of food production in the New World, in American Archaeology Past and Future: 277-321, eds. David J. Metzler, Don D. Dowler, and Jeremy Sabloff. Washington DC/London: Smithsonian Institution Press. Thissen, Laurens C. 2007 Die Anfänge der Keramikproduktion in der Türkei - ein Überblick, in Vor 12.000 Jahren in Anatolien - Die ältesten Monumente der Menschheit: 218-229, ed. Badisches Landesmuseum Karlsruhe. Stuttgart: Konrad Theiss. Underhill, P. Anne and Junko Habu 2005 Early communities in East Asia: Economic and sociopolitical organization at the local and regional levels, in Archaeology of Asia: 121-147, ed. Miriam T. Stark. Hong Kong: Blackwell. Zeist, Willem van and Sietse Bottema 1991 Late Quaternary vegetation of the Near East: Beihefte zum Tübinger Atlas des Vorderen Orients. Reihe A (Naturwissenschaften), no. 18. Wiesbaden: Reichert. Wolters, Oliver W. 1967 Early Indonesian Commerce. A Study of the Origins of Srivijaya. Ithaca, N. Y.: Cornell University Press.

HIGHLAND-LOWLAND CONNECTIONS IN JAMBI, SOUTH SUMATRA, AND WEST SUMATRA, 11TH TO 14TH CENTURIES JOHN MIKSIC1

Introduction In the late 7th century two regions in southeast Sumatra vied with each other to dominate trade in the Straits of Malacca: Malayu, on the Batanghari, and Srivijaya, on the Musi River. Yijing, a Chinese Buddhist pilgrim, implies that this contest ended when Srivijaya acquired hegemony over Malayu between 672 and 689. Srivijaya seems to have won the contest and controlled trade between India and China for the next 150 years. Yijing used the name Malayu to designate a zhou or geographical region. Whether it was also the name of a kingdom, we cannot say. The Batanghari reappeared in history in 840 when a Chinese text entitled You-yang za-zu mentioned a kingdom named Jambi (Zhanbei). Jambi sent missions to China in 852/853 and 871, implying that the Batanghari region enjoyed at least semi-autonomy from Srivijaya during that period. No missions from Srivijaya to China are recorded between 704 and 904; during those two centuries Jambi was the only east Sumatran kingdom to maintain diplomatic contact with China, though Srivijaya probably carried on commercial relations with China during that period.

1

This paper is partially based on data collected during a collaborative research project between the Asia Research Institute, National University of Singapore; the Balai Pelestarian Peninggalan Purbakala, Propinsi Jambi; the Balai Arkeologi Sumatra Selatan, Palembang; Dr. I Wayan Ardika and the Jurusan Arkeologi, Universitas Udayana, Bali; and the Pusat Penelitian dan Pengembangan Arkeologi Nasional, Indonesia. Their contributions, and those of others too numerous to mention here, are gratefully acknowledged. Funding for that project was generously provided by Orchard Maritime Services Pte. Ltd., Singapore.

76

Highland-Lowland Connections

In early times China seems to have either ignored or been unaware of the oscillating relationship between these two neighbouring Sumatran realms, which were so similar in culture and language as to be difficult to distinguish at a distance of several thousand kilometres. Only further archaeological research holds the possibility to reconstruct this competition. China found it unnecessarily troublesome to delve into the region’s shifting political currents, so long as luxury tropical items flowed to the Chinese court. In the latter half of the 11th century, Chinese references to Sumatra become more plentiful and possibly more accurate. This increased attention coincided with new trading activity: higher volumes, new commodities, and more Chinese merchants visiting Southeast Asia. According to these records, Jambi succeeded Srivijaya as the dominant polity in the Straits of Malacca around 1080 (Coedès 1968: 178-180). Jambi sent missions in 1082, 1084, 1088 (this mission may have come from either Jambi or San foqi, a term which could mean either Palembang or Jambi; Wolters 1966: 225, fn. 4), 1090, 1094, and 1095 (Wolters 1966: 235; 1983: 60). Increased trade coincided with a shift in the locus of political power in southeast Sumatra in the 11th century. Historians and most archaeologists attribute the evolution of complex societies in southeast Sumatra to stimuli provided by maritime trade. Archaeological evidence does not unequivocally confirm this view. The alternative hypothesis, that conditions in the Sumatran highlands exerted significant influence over the evolution of centres of power and commerce in the lowlands, deserves serious consideration. Examination of this hypothesis requires a study of commercial and cultural relations between the upstream and downstream areas along the rivers which flow from the West Sumatran highlands to the Straits of Malacca. Archaeological data could enable us to define the nature and intensity of interaction between highland and lowland Sumatra, which could then be compared with the effects of interaction between polities in the eastern lowlands with their trading partners overseas. Very little research has been conducted in highland Sumatra; archaeologists have concentrated on areas mentioned in foreign sources, all of which are in the lowlands. Uplandlowland connections, which can only be studied archaeologically, have received little attention. Archaeological reports concentrate on discussions of imported objects from China, India, or the Persian Gulf, subjects which admittedly are glamorous and of great interest to the wider public. The analysis of such artefacts as earthenware pottery, which might reveal details of contacts between highland and lowland Sumatra, has been almost entirely ignored.

John Miksic

77

At its current rate of progress, the study of earthenware will take a very long time to reach a stage at which it would become possible to trace Sumatran-made ceramics to their sources. Such an endeavour should be supplemented by research into the movement of forest products such as dammar, gold, and ivory from their highland sources to the eastern estuaries. At Fort Canning, Singapore, a significant quantity of damar was found in a 14th-century stratum. In the 1980s Rosemary Gianno conducted very useful research on resin use and exploitation in the Malay Peninsula, using Fourier transform infrared spectrometry, gas chromatography, and mass spectrometry to specify the genus and sometimes the species of resins found in archaeological sites (Gianno 1990a, b). She was able to identify resin used to form the core of gold leaf anklets found on Java as originating from Agathis borneensis, a tree found on Borneo but not on Java (Miksic 1990: 90). This is precisely the type of information needed to examine critically the existing model of early settlement and trade in Sumatra, but this important work has not been continued. An example of distortions consequent upon common assumptions found in analyses of archaeological data is the case of Lubok Tua, Barus, northwest Sumatra. Working under difficult conditions, an IndonesianFrench team acquired much valuable data and has published important reports on their discoveries. The editor of the reports, M. Claude Guillot, has postulated that Lubok Tua was founded in the 11th century by Indians, and inhabited largely by foreigners, including Arabs and Persians (Guillot 2003). According to the authors of the report, few artefacts can be connected with the presence of women on the site (Guillot 2003: 65). It is difficult to imagine an enclave inhabited only by foreign males without a substantial local presence. Although it is possible that the site was purely populated by men, another hypothesis is plausible: that locals, including women, were present, but that their presence has remained undetected. One artefact type which is highly correlated with female activity is earthenware pottery. Lubok Tua has yielded much earthenware, some of which may be imported. The excavators of Barus infer that resemblance between Indian and some Lubok Tua pottery indicates that earthenware in Barus was imported from India. Such types as oil lamps and pots with a grid pattern on the bases are very different from anything in Southeast Asia, and therefore can be identified with a high probability as imports. Some designs found at Lubok Tua, such as kendi (goglet) spouts with moulded flower petal décor, are common in the Kaveri valley (Guillot 2003: 78), and may have been imported from there. Another decorative motif, an incised sawtooth design, is common in Barus but is not found in South Asia, suggesting that it came from somewhere else. A local source is highly probable.

78

Highland-Lowland Connections

The authors of the study are willing to consider the possibility that such items as large coarse jars were locally made (Guillot 2003: 99). In chapter seven, “’Local’ ceramics”, pottery which is “Indianized” in shape, but of “inferior quality,” is considered to have been made locally (Guillot 2003: 210). Furthermore, it is stated that pottery among the North Sumatrans (often grouped together under the term Batak), if it ever really existed, never has been developed until today. During our stay in this town, very simple pottery sold in the market was imported from Padang, a port several hundred kilometres away, and the oldest inhabitants, recognizing the absence of local production, reported that it had also been known to come from the province of Aceh in the past. Such pottery has also been found in Kedah and Brunei. The presumption that local pottery must be primitive cannot be taken for granted, but should be verified or refuted by research in the Barus hinterland. Such research may challenge the assumption that Barus was an isolated foreign enclave maintaining little contact with its hinterland.

Traditional Sumatran Trade and Settlement It cannot be assumed without proof that settlement patterns in southeast Sumatra evolved mainly in response to maritime trade. Europeans became involved in southeast Sumatran trade in the 16th century, but had little impact on settlement patterns or the structure of trade and exchange for the next 300 years. Why should the situation have been different earlier, when foreign merchants were less common and less invasive? The earliest known concentrations of populations and centres of political power in east Sumatra formed not at river mouths, but 90-100 km inland, at the confluences of river tributaries which linked the highlands to the lowlands (Miksic 1985: 433). Local commodities transported from the highlands to the lowlands in the internal Sumatran networks included animal and forest products, metal, prestige items, and utilitarian and staple goods, only some of which were sought in overseas markets. Iron sources in the Sumatran highlands (van Bemmelen 1970; Guillaud 2006: 73-74) may also have been exploited by highlanders for export to the lowlands. Other metals found along the Barisan chain include gold, silver, copper, and lead. The soils of the eastern lowland Sumatra are not suitable for rice agriculture; rice was imported by lowland population centres from both upstream and overseas sources (e.g. Reid 1980: 247; Miksic 1985: 431). Rice was a prestige food; commoners probably had to be content with sago

John Miksic

79

as their main carbohydrate. Until 1900, population in Sumatra was mainly concentrated in the highlands, where rice and fresh water were plentiful. Although as linguists believe, Austronesian-speaking peoples may have entered Sumatra only in the last millennium BCE, archaeological evidence for their activity (such as ceremonial sites related to those of Polynesia) are limited to the highlands. This suggests that the first Austronesian settlers quickly found their way to the upland valleys. Contrary to common assumption, the settlements in the eastern lowlands do not represent the oldest form of complex society in Sumatra; they are comparatively later offshoots of highland agrarian societies. It is to the highland cultures that we must look to find the forebears of the populations of Jambi and Srivijaya.

Fig. 4-1: Hybrid model of Sumatran spatial hierarchy Bennet Bronson (1977), in an influential article cited by many subsequent scholars (e.g. Manguin 2002; Guillaud 2006: 62), asserted that routes of communication and transport in Sumatra have historically been restricted to rivers. Bronson’s diagram closely resembles Dunn’s rarelycited earlier diagram for the Malay Peninsula (Dunn 1975: 100, fig. 7.1). Although neither used the term, geographers and locational analysts have identified this as a common pattern labelled dendritic (for its resemblance to a tree’s root system). Dendritic networks are often associated with areas dominated by overseas metropoles, as in colonial systems of resource

80

Highland-Lowland Connections

extraction, or with simple economic systems in areas where population is dispersed and transportation routes are few. Dendritic networks are associated with “gateway communities” (Hirth 1978: 37-38) which are normally located at boundaries between areas of very different topography and resource distribution; river mouths are typical examples. In southeast Sumatra, the dendritic model implies that gateway communities should have been located at river mouths. This was not true. Palembang, the political centre of Srivijaya, is 90 km from the mouth of the Musi; Muara Jambi, the 11th-century capital of Jambi, is 75 km inland. These sites occupy superior positions in the internal communication networks, an attribute which river mouth sites lack (Miksic 1985: 437). Pierre-Yves Manguin modified Bronson’s diagram by adding an overland connection between two river systems (Manguin 2002: 78). This is an improvement on Bronson’s diagram, but networks recorded in historical sources record are even more complex than this. Manguin shows that anak sungai dan teluk rantau is a literary expression used to mean the frontiers of Malay harbour-cities, and glosses it as “bends and reaches”. An alternative rendering suggested here is “tributaries and estuaries”. This would convey the worldview of nobles who considered themselves to be the centres of their world. They would not choose to live in marginal zones at the very extremities of the hilir (downstream) regions, subject to floods, tidal waves, and sudden attacks from the sea; they established their courts, around which a unique form of urban life evolved, at the junctures of the brackish and the fresh water, where the first permanently dry land, the outliers of the cooler hulu (upstream), appeared. In the Malay Peninsula, hulu came to mean “remote, uncivilized”, and hilir stood for “cultured, sophisticated”, but Sumatra, where these terms originated, presents a far different picture. The location of centres at hinterland river junctions conforms to the pattern of kadatuan, samaryyƗda, and mandala sketched by Hermann Kulke (1993: 172). Although Manguin notes quite rightly that there is no reason to take this diagram as a physical expression of a symmetrical geographical distribution, since it is concerned with political relations, there is nevertheless a logical justification for arguing that mental and physical landscapes of Sumatra closely resembled one another. A historically accurate map of Sumatran centres and routes must focus equally on highland regions, where intensive agriculture was feasible, and lowland confluences (for a model of this system, see fig. 4-1 in which the zigzag lines represent mountains or other barriers to overland communication).

John Miksic

81

The lowlands of Sumatra have a few large population concentrations, but the highlands (above 100 m) have many more mid-sized towns (fig. 4-2, Sumatran highland divisions and lowland drainage). There are more monumental sites in highland Sumatra than in the lowlands (fig. 4-3), although this may be partly the result of recent lowland urbanization. Little research has been conducted on overland routes in precolonial Sumatra (Edwards McKinnon, infra), but it is probable that they linked the early monuments. The oldest monuments in Sumatra were created in the prehistoric era, in the Pasemah highlands, far in the hulu of the Musi drainage (Guillaud et al., infra). Remains here consist of stone sculpture and elaborately-painted tombs, probably for nobles. Imported items found in sites of the Pasemah culture include South Asian beads and bronze drums from North Vietnam. Two Dongson drums have also been found in south Bengkulu, on the west coast, indicating that one or more transinsular routes existed in prehistory. Other important prehistoric highland centres lay near Lake Kerinci (the surface of which is 783 m asl). The former tea estate called Danau Gadang, 9 km south of the lakeshore, was traditionally linked by footpaths to both the Jambi highlands, and to the Minangkabau region further north (van der Hoop 1938: 200). This area yielded a wide range of items, from stone adzes and obsidian blades and numerous fragments of earthenware pottery, pieces of bronze slag, and a bronze arm-protector, to imported items such as a fragment of a Dongson drum and carnelian beads of double-hexagonal shape. At Lolo Gadang were found a bronze urn of prehistoric style, and numerous megalithic sites (such as the villages of Muak, Lempur, and Benik, Kecamatan (Sub-District) Gunung Raya, Kumun Ilir, Kecamatan Sungai Penuh, and Bento, Kecamatan Gunung Kerinci, Kabupaten (District) Kerinci. The Balai Pelestarian Peninggalan Purbakala (BP3) team reported finding a wide range of other artefacts, from obsidian flakes and stone axes to Chinese porcelain of the SongYuan period (12th to 14th centuries), in superficial association with the megaliths (Bonatz et al. 2006; Bonatz 2006; Tjoa-Bonatz, infra). At Bangkinang, on the upper Kampar River, Riau province, the oldest evidence of human activity is a stone adze (van der Hoop 1938: 200). Bronze statuettes or pendants in the form of human dancers found here display a style as yet unparalleled elsewhere.

82

Highland-Lowland Connections

Fig. 4-2: Sumatran highland divisions and lowland drainage

John Miksic

83

Fig. 4-3: Monumental sites and topography In the same eastern foothill region of the Barisan Mountains lies the site of Muara Takus. The site is small but contains several important Buddhist monuments, and must have been occupied for a rather long period, for two principal structures were rebuilt subsequent to their original construction. One legend links the temples to Adityawarman, a 14th-century ruler, and suggests that he extended his influence to this area by arranging a marriage for his daughter to a local chief (Ijzerman 1893: 51; Schnitger 1940: 390; Kulke, infra). It is more likely that Muara Takus was linked to Padang Lawas, which lies only 100 km further north, via Pasir Pengarajan and Rokan. This conclusion is supported by small finds at Manggis and Kotanopan-Penarik (penarik is Malay for “portage”), and the legend that the site was once attacked by the Batak (who inhabited Padang Lawas; Schnitger 1936: 40). Tanjung Medan lies 75 km northwest of Muara Takus. A stupa-like structure stood here in 1876, but by 1930 this had been reduced to a heap of brick. Small finds here such as a kendi with gold neck and silver rim and a gold plate were found in the rubble. The plate was inscribed with

84

Highland-Lowland Connections

images of eight bodhisattvas and an inscription of Buddhist formulae in Nagari script, probably datable to the 12th century (Bosch 1930: 133). In North Sumatra, the fearsome reputation of the Batak is belied by numerous records of overland traffic from the highlands to both east and west coasts. Traditional trade routes converge in the Padang Sidempuan/Gunungtua/Portibi region (Neumann 1856; Frey-Wisseling 1933) where over 20 complexes of brick ruins, mainly of Buddhist shrines of the 11th through 14th centuries, have been found (Schnitger 1940; Miksic 1980). Small quantities of Chinese porcelain have been reported from the site; further exploration is likely to reveal one or more centres of population and economic activity in the region. The lower course of the Panai River is likely to yield evidence of a settlement which may have functioned as the lowland gateway to (or outpost of) an ancient centre of power and wealth in the Padang Lawas hinterland. In the 16th century, produce from Haru (probably centred at Deli Tua) was exported in several directions: to the east coast, to the northern ports of Pediri and Pasai, and to Barus (Cortesão 1944: I, 148). John Anderson in 1823 met Alas people from Singkel who carried gold, camphor, and damar to the east coast where they exchanged it for cloth, an activity which Anderson described as “very extensive” (Anderson 1971: 83, 250, 263; see also Edwards McKinnon, infra; Perret, infra). S. C. Crooke, in an appendix to Anderson’s report, mentions that people walked from Bengkulu to Jambi in 24 days (via Kerinci and the Batang Asei; Anderson 1971: 399). He added that The mode of communication between villages, as well as distant parts of the country, is almost exclusively by water, there being few habitations that are not situated on the rivers or near them; and such routes as do exist, are mere footpaths through the woods. They, however, extend to Padang, Bencoolen, and other places on the western coast of the island, with which they are the means of commercial intercourse.

Crooke was unable to obtain details of the routes to other locations on the west coast, but he was told that it was possible to go up the Batanghari for ten days to a point on the Tembesi River, then overland for one day, then downriver to Palembang in two days. He also met an “itinerant trader” who described another overland route between Jambi and Palembang which took seven days; and “It is positively and generally asserted by the natives, that these land routes are their only means of intercourse with Palembang” (Anderson 1971: 400). There are ten important rivers in east Sumatra. On none of them is the primary settlement at the river mouth. Besides Palembang and Jambi, Gunung Sugih, 100 km up the Wai Seputih, Lampung, is the largest port

John Miksic

85

on that river system; and Rengat, 130 km up the Indragiri, occupies a similar position in its drainage. Siak Sri Indrapura on the Siak River, and Rengat on the Kuantan, are similarly located at or near the upstream limit of the tidal zones of their valleys. Neither the Rokan nor the Kampar have any major towns along their courses; both rivers have large tidal bores which render navigation along them unsafe. On the Musi’s tributaries, the important riverheads are Baturaja on the Ogan, Muaraenim on the Lematang, Sunganianik on the Musi, and Kotatanjung on the Rawas, all located at another transitional point: the 100-meter contour. Probably important overland routes ran between rivers at that approximate elevation. While the river valleys of Sumatra have some general features in common, each also displays significant distinguishing characteristics. The rivers in northeast Sumatra from the Barumun to the Kuantan flow through weathered sandstone and limestone. These rocks yield sediments of low fertility. The Batanghari sits astride the boundary between sedimentary and volcanic formations. Kerinci volcano provides rich soil for the Jujuhan and Tebo tributaries, but the benefits of these do not reach the lowlands around Muara Jambi; as an agronomist wrote many years ago, “the lands of Lower Djambi are poor or bad” (Mohr 1944: 437). The Musi River’s alluvial plain is 200 km long, from the Straits of Malacca to Muara Kelingi. The Musi and its tributaries drain numerous active volcanoes, include Mount Kaba and Mount Dempo, and the Lake Ranau area. An area of flood recession agriculture called the lebak area covers about 30,000 ha along the Ogan and Komering; in the 1920s this region yielded a surplus of rice worth 30,000 gulden (Vaas et al. 1953). It is probably not a coincidence that Palembang, in the centre (not the seaward margin) of east Sumatra’s largest and most fertile drainage, has been the largest city in east Sumatra for the last 1,500 years. The Musi valley covers 61,000 km²; the Batanghari is the second largest at 40,000 km²; the Kampar is third (31,000 km²), followed by the Siak (16,350 km²), and Rokan (16,500 km²) (Bezermer 1921: 528-530). Oliver W. Wolters suggested that one of Palembang’s advantages was “the belt of swamp reaching a considerable way inland and giving protection against attack from the interior” (Wolters 1967: 225, 342, fn. 163). In fact Palembang is located on the ulu side of the swamps, not the hilir, and thus is protected from attack from downstream, not upstream. Its placement facilitates rather than impedes intercourse with the hinterland. Wolters, so percipient in most of his analysis of ancient Sumatra, was misled by the common assumption that in Southeast Asia “the sea unites, the land divides” into overlooking this fact.

86

Highland-Lowland Connections

Regional Analysis and the Location of Palembang The most important indicator of a settlement’s position in a regional hierarchy consists of the number of activities carried out in it. It is often but not always true that the number of activities is correlated with the number of people in the site, i.e., that the larger the population, the more numerous are the specialized economic activities carried out in it. It is possible to obtain population figures for Sumatran cities during the 20th century, and thus to establish the shape of hierarchies of settlement in various regions. This data can be compared with the dendritic pattern postulated by Bronson et al., and with the hexagonal lattice of conventional central place theory, to see which pattern fits the data better (Miksic 1984). Central place theory recognizes three main variations: the K3 (where a lower-level centre has equal access to three higher-level centres; this is called the marketing landscape), the K2 (the transport landscape), and K1 (administrative landscape) (Christaller 1966; Smith 1974, 1976). Another important dimension of analysis measures the availability of transport, communication, and marketing connections. One useful objective means of calculating connectivity among centres is a figure known as the Beta index; this measures the number of alternative connections between centres of human activity (Haggett 1966: 238-239). The highest possible Beta index is three, but values above one signify that each centre or apex of a communication network has more than two alternative access routes. Centres with beta indexes below one can be equated with an administrative landscape, because a lower-level centre will have only one connection with a higher-level centre. In developing countries, beta indexes below one are often found. Obviously, the higher the Beta index, the more developed and complex is the network of communication and transport in a region. Other indices useful for archaeology are centrality and network shape. The geographer K. J. Kansky developed a simple formula for calculating a number which is correlated with the basic shape of a communication network (Kansky 1963). Another formula is used to calculate the Konig number. This is calculated by counting the number of other centres in a network which someone must pass through in order to travel from one end of the network to the other. The location with the lowest Konig number is the centre of that communication network (Haggett 1966: 237-238). By applying these methods to South Sumatra, it is possible to obtain an objective indication of the influence of local transport on the development of complex culture there. According to archaeological theory, one

John Miksic

87

important factor in the evolution of complexity is the efficiency of communication between human beings. The communication factor may be more important than settlement, i.e., the transition from village life to urbanization, or the transition from the tribal to the chiefdom level in Elman R. Service’s (1975) terminology. Although we have some ancient writing from Sumatra, the texts are not useful for reconstructing the ancient communication system within the island, and thus for the time being we must rely on ethnographic analogy. If Bronson’s theory were correct, the Beta index of the communication system in Sumatra should be below one, because there would be only one connection between each pair of centres in the network. In Bronson’s model there are eight locations and only seven connections between them; thus the Beta index is seven divided by eight or 0.875. This would indicate that the system could be controlled from one strategically-located centre, so that it would favour a monopoly system. If however we take into account the reports from the 18th and 19th centuries that there were footpaths along the mountains between river systems, then there would be at least one other route in Bronson’s figure, between the four sites at level D. This would change the value of the index to 10/8, or 1.25. This value indicates that a complex communication system existed in South Sumatra in about 1800; it is impossible to determine the Beta index for South Sumatra in the 7th century. From the standpoint of centrality, or the tendency for activities to concentrate in certain places, we have to begin by counting the number of centres through which a person would have to pass to get from one end of a network to another. We then add the distance between all centres in the system, and then divide the total distance by the number of centres passed through in the shortest journey. This procedure yields a value which is correlated with the shape of a network. This value in advanced countries such as France is around 30, while in developing countries it approaches one. This value is considered a sensitive indicator of the economic system which a given transport system supports. In the case of South Sumatra, the riverine distance which can be covered by a ship above five tons deadweight in the dry season is about 1,010 km (calculated from Robert R. Nathan Associates, reproduced from Bonn University Report on South Sumatra Infrastructure 1976, exhibits 3 and 4). The land route from Pulau Panggung in Komering Hulu to Kotatanjung in Hulu Sungai Rawas is about 250 km; thus the entire distance included in the South Sumatra network is 1,260 km. The largest number of centres which have to be passed through in a journey across the network, from Kotatanjung to Cahayanegeri, is seven. There are two

88

Highland-Lowland Connections

routes from Kotatanjung to Cahayanegeri: by river (370 km) and overland (330 km). The average distance is thus 350 km. The figure 350 is then converted into standard miles (350 x 0.6=210). When divided by seven, this gives the number 30, the same as France. If we calculate the Konig number which indicates the most efficient location for a centre in a communication network, then we look for the number of centres which have to be passed through once again. But this time we calculate the lowest number rather than the highest. In the South Sumatra network, the highest figure is seven, calculated from the hulu, but the lowest number is five, which is found for two centres (Muaraenim and Palembang). This predicts that Palembang and Muaraenim would be the best locations for a centre for the network. In the Konig formula, Sungsang, at the mouth of the Musi River, has a higher value. This would indicate that its status as a third-level centre in the regional hierarchy is correlated with a local communication network, not because of communication with overseas centres. If these figures are compared with the administrative hierarchy in South Sumatra during the colonial period, we find a high degree of correlation (fig. 4-4). High-level colonial administrative centres have low Konig numbers; Dutch administrative centres were efficiently located from the standpoint of communication efficiency. The Dutch system reflects that of the earlier Palembang Sultanate, so that it can be assumed that this pattern reflects the traditional political system of earlier times. South Sumatra’s hinterland supplied some important resources which were in demand in international commerce in the Classic period, but further supplies of Sumatran commodities important for international trade would have had to come to Palembang from other river drainages. If maritime trade had been a major determinant of site location, it would not have been logical to locate export centres far from the production centres. The development of Palembang as the centre of Srivijaya makes more sense if we assume that internal communication and transport networks formed in the ancient past, and stimulated evolution of complex civilization. At a later time, when traffic between Indonesia, India, and China increased, the people of the Musi drainage would have been ready to exploit the new opportunities because they had already evolved expertise in administration and transportation. Pasemah is not far from Muaraenim, which is equal to Palembang in terms of its location near the centre of an internal communication network. Perhaps the centre of complexity shifted from Pasemah to Palembang in the early historic period as a result of increased communication between South Sumatra and the outside world. However, local factors were still primary determinants which shaped the regional system.

John Miksic

89

Fig. 4-4: South Sumatran administrative hierarchy, 1900 The level of confidence we can assign to an index of Palembang’s position in an early Classic period communication network cannot yet be determined, because we lack data from the hinterland of Sumatra. During the Islamic period, it is clear that local factors were still strong enough to give Palembang a dominant position in the Sumatran political and economic hierarchy. More data is needed to determine whether to accept or reject this hypothesis for earlier periods.

The Batanghari Spatial System The Musi River drainage has been explored in some detail. The other important river course in southeast Sumatra, the Batanghari, also originates in the central highlands: the Kerinci and Minang areas. A systematic survey was carried out in 2005 along the lower Batanghari, from Muara Jambi, to the Berhala Strait, including the Batang Kumpeh,

90

Highland-Lowland Connections

where surface finds of Chinese ceramics dating between the 10th and 14th century had earlier been reported at two locations (Edwards McKinnon 1985: 26). The Berbak, the Batanghari’s southern mouth, could not be explored, because of its status as a protected nature reserve. Tab. 4-1: Lower Batanghari sites of the 10th to 14th centuries above 1 ha Site

Area (ha)

Possible site function

Tier One: Muara Jambi

1100

Ceremonial centre

Tier Two: Kota Kandis

132.5 (50.3)

Habitation centre

Suak Kandis

92.1

Habitation centre

Tier Three: Lambur

54.7

Habitation centre

Candi Teluk Complex

27.6

Ceremonial centre

Jebus

12.6 (4.9)

Habitation centre

Tier Four: 13 sites