The Archaeology of Knowledge Traditions of the Indian Ocean World 9780367335465, 9780367337810, 0367335468

Citation preview

THE ARCHAEOLOGY OF KNOWLEDGE TRADITIONS OF THE INDIAN OCEAN WORLD

This book examines knowledge traditions that held together the fluid and overlapping maritime worlds of the Indian Ocean in the premodern period, as evident in the material and archaeological record. It breaks new ground by shifting the focus from studying cross-pollination of ideas from textual sources to identifying this exchange of ideas in archaeological and historical documentation. The themes covered in the book include conceptualization of the seas and maritime landscapes in Sanskrit, Arabic and Chinese narratives; materiality of knowledge production as indicated in the archaeological evidence of communities where writing on stone first appears; and anchoring the coasts, not only through an understanding of littoral shrines and ritual landscapes, but also by an analysis of religious imagery on coins, more so at the time of the introduction of new religions such as Islam in the Indian Ocean around the eighth century. This volume will be of great interest to researchers and scholars of archaeology, anthropology, museum and heritage studies, Indian Ocean studies, maritime studies, South and Southeast Asian studies, religious studies and cultural studies. Himanshu Prabha Ray is Research Fellow, Oxford Centre for Hindu Studies, Oxford. From 2014 to 2019 she held the prestigious Anneliese Maier research award of Humboldt Foundation. She was the first Chairperson of the National Monuments Authority, Ministry of Culture in New Delhi, India from 2012 to 2015 and former Professor, Centre for Historical Studies, Jawaharlal Nehru University, New Delhi, India. Her research interests include Maritime History and Archaeology of the Indian Ocean and the Archaeology of Religion in Asia. Her recent books include Decolonising

Heritage in South Asia: The Global, the National and the Transnational (ed. 2019), A ­ rchaeology and Buddhism in South Asia (2018), Buddhism and Gandhara: An Archaeology of Museum Collections (ed. 2018), The Archaeology of Sacred Spaces: The Temple in Western India, 2nd Century BCE to 8th Century CE (with Susan Verma Mishra, 2017), The Return of the Buddha: Ancient Symbols for a New Nation (2014) and The Archaeology of Seafaring in Ancient South Asia (2003).

THE ARCHAEOLOGY OF KNOWLEDGE TRADITIONS OF THE INDIAN OCEAN WORLD

Edited by Himanshu Prabha Ray

First published 2021 by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN and by Routledge 52 Vanderbilt Avenue, New York, NY 10017 Routledge is an imprint of the Taylor & Francis Group, an informa business © 2021 selection and editorial matter, Himanshu Prabha Ray; individual chapters, the contributors The right of Himanshu Prabha Ray to be identified as the author of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library Library of Congress Cataloging-in-Publication Data A catalog record has been requested for this book ISBN: 978-0-367-33546-5 (hbk) ISBN: 978-0-367-33781-0 (pbk) ISBN: 978-0-429-32185-6 (ebk) Typeset in Times New Roman by codeMantra

CONTENTS

List of figures vii List of tables xi List of contributors xiii Preface xvii

PART 1

Conceptualising the seas: Introduction Himanshu Prabha Ray





13

vi Contents



PART 2

The materiality of knowledge production: Introduction Himanshu Prabha Ray

97







PART 3

Anchoring the coasts: Introduction Himanshu Prabha Ray

169







Index 263

FIGURES

­ ­

­ ­ ­ ​­ ­ ­ 6 ­ 8.1 ­Palm-leaf ​­ manuscripts and rolled legal documents, eighteenth and nineteenth centuries 151 ­ ​­

viii Figures

­ 8.3 Ivory container carved for the gold letter written by King Alaungphaya of Burma to King George II of Great Britain (letter dated 7 May 1756). Dimensions: h: 132 mm; diameter: 50 mm  8.4 Document holder, brass inlaid with gold and silver, Mamluk Egypt or Syria, early fourteenth century 9.1 Map showing coastal shrines of ancient South Arabia (drawn by Uma Bhattacharya) 10.1 Nautical iconography in Angkor contains a great wealth of information, not only in terms of nautical technology, but also on boat usage and location. In this scene from the Bayon temple a noble lady picking a lotus flower appears. The action takes place in a pond, as marked by the stepped shoreline 10.2 Decorated boats are used by the elite in a variety of settings; the scene located on the west wall of the pond lying just outside the northeastern wall of the Phimeanakas complex, in Angkor Thom, represents seven boats, five of which are carrying noble people toward the stairs in the middle of the pond’s wall. Containers are seen onboard the vessels, and in one instance a nobleman is reaching into a chest held by one of his servants 10.3 The figureheads of the two barges depicted in Angkor Wat represent an idea of how the boats used during the Dvaravati Water Festival would have looked. The boat from the lower registry has a heavily ornate hamsa (goose) 10.4 One of the carvings of Vishnu Anantasayin bathed by the waters of Kbal Spean in the Kulen Mountains. Seeing the water touch the body of the snake, it is easy to imagine the serpent Ananta as a zoomorphic boat 11.1 A row of musicians – three oboes, a two-stringed lute, cymbals, a big-barrel drum with stand, two bossed gongs carried in a pole by four persons, two oboes, an eight-gong circle, a conch-shell, a small cylindrical drum hanging from the neck, and an oboe 11.2 A row of musicians – two long horns, a pair of cymbals, a small drum hanging from the neck, a conch-shell, a short horn, a big barrel drum – accompanying the procession of Suryavarman II 11.3 A big-bossed gong carried on a pole, with the person at the back beating it 11.4 A big-bossed gong carried on a pole and a small barrel-shaped drum in front of it 11.5 Two bossed gongs, one small, one bigger. The player is at the back, his left hand holding a beater with cloth wrapped

154 161 176

193

197

199

201

211

211 213 214

Figures  ix

at the tip to give off a loud, booming, nonmetallic sound when beaten, as practiced today in Cambodia, Thailand, Java and Bali, Borneo and the Philippines 11.6 A gong circle with nine bossed gongs, the musician using two hand beaters 11.7 A gong circle with nine graduated bossed gongs, the musician uses two beaters with cloth wrapped around the tip, another person carries the instrument 11.8 A gong circle with eight graduated bossed gongs, another person holding the frame 11.9 A gong circle with eight bossed gongs, with two beaters without the tip cloth wrapper. A conch-shell and a small cylindrical drum also are shown 12.1 Umayyad gold Dinar, dated AH77 12.2 Gold coin with legend ‘Shubhatunga’, British Museum 12.3 Gold coin with legend ‘Shubhatunga’ (Classical Numismatic Gallery, Auction 2, 18 December 2010, Lot 38) 12.4 Gold coin with legend ‘Shubhatunga’ (Todywalla Auctions, Auction 118, 1 September 2018, Lot 70) 12.5 Gold coin with horse rider facing left 12.6 Gold coin with horse rider facing right 12.7 Indian imitation of an Islamic Dinar, ex-Walter Elliot, British Museum (accession number BM 1876, 6.2.2) 12.8 Indian imitation of an Islamic Dinar, ex-Walter Elliot, British ­Museum (accession number BM 1876, 6.2.3) 12.9 Gold ‘Dinar’ of Offa, king of Mercia (CE 756–796) 12.10 Gold coin with seated goddess on obverse 12.11 ‘Boar’-type gold coin with pseudo-Arabic inscriptions in the ­margins 12.12 ‘Boar’-type gold coin with stylised pseudo-Arabic inscriptions in the margins

214 215 216 216 217 243 247 248 248 249 249 251 252 253 253 254 254

Taylor & Francis Taylor & Francis Group http://taylora ndfra ncis.com

TABLES

11.1 Distribution of bas-reliefs of musical instruments in the eight galleries of Angkor Wat based on field research in 2009 and 2011

218

Taylor & Francis Taylor & Francis Group http://taylora ndfra ncis.com

CONTRIBUTORS

Dionisius A. Agius  is a Fellow of the British Academy and Al-Qasimi

Professor of Arabic Studies and Islamic Material Culture at the University of Exeter and affiliated to King Abdulaziz University, Jeddah. Educated at the Jesuit Université St-Joseph, Beirut and University of Toronto, he is an Arabist, ethnographer and linguist working on the maritime cultural landscapes of the Western Indian Ocean. He is author of Seafaring in the Arabian Gulf and Oman: People of the Dhow (2005), winner of The Abdullah AL-Mubarak Al-Sabah Foundation and British-Kuwait Friendship Society major Book Prize; Classic Ships of Islam: From Mesopotamia to the ­Indian Ocean (2014) and The Life of the Red Sea Dhow: A Cultural History of Seaborne Exploration in the Islamic World (2019). Andrew M. Bauer is Assistant Professor in the Department of Anthropology

at Stanford University, where his research and teaching interests broadly intersect archaeological method and theory and environmental anthropology, with particular emphasis on the politics of environmental production in precolonial South India. He has published several books, including Climate Without Nature: A Critical Anthropology of the Anthropocene (2018, with Mona Bhan), Before Vijayanagara: Prehistoric Landscapes and Politics in the Tungabhadra Basin (2015) and The Archaeology of Politics: The Materiality of Political Practice and Action in the Past (2011, coedited with Peter G. Johansen), as well as numerous articles and essays. Shailendra Bhandare  is Assistant Keeper, Numismatics, Ashmolean Museum, Oxford University, Great Britain. His publications include “Not Just a Pretty Face: Interpretations of Alexander’s Numismatic Imagery in the Hellenic East,” in Memory as History: The Legacy of Alexander in Asia, 2007; “Numismatics and History: The Maurya-Gupta Interlude in the Gangetic

xiv Contributors

Plain,” in Between the Empires: Society in India 300 BCE to 400 CE.; “From Kautilya to Kosambi and Beyond: The Quest for a Mauryan/Asokan Coinage,” in Reimagining Aśoka: Memory and History, 2012. His research interests include Indian Coinage: its historical context, utility and evolution with specific attention to the early historic, medieval (Mughal) and modern periods. Also, its significance for undertaking studies of broad historical interests such as cross-cultural syncretism, urbanisation and colonialism, as well as the applicability of Indian Numismatics for various Indological themes such as art, iconography, epigraphy and archaeology. Matthew A. Cobb received his PhD from Swansea University in 2012 and currently lectures in ancient history at the University of Wales Trinity Saint David. His research focuses on Graeco-Roman participation in the Indian Ocean trade, as well as cross-cultural engagement between West and East in antiquity. He is the author of Rome and the Indian Ocean Trade from Augustus to the Early Third Century CE (2018) and the edited book The Indian Ocean Trade in Antiquity: Political, Cultural and Economic Impacts (2019). Peter G. Johansen is an Assistant Professor in the Department of Anthropology at McGill University. His teaching and research interests focus on archaeological approaches to investigating the politics of everyday life, techno-politics, landscape production, ritual and monumental practices and metallurgical craft production in southern India. He is the author of several articles, and book chapters, and co-editor of The Archaeology of Politics: The Materiality of Political Practice and Action in the Past (2011, with Andrew Bauer). Salila Kulshreshtha  teaches South Asian history at New York University Abu Dhabi. She is the author of From Temple to Museum: Colonial Collections and Uma Mahesvara Icons in the Middle Ganga Valley (2017). Salila has a PhD in History from the Centre for Historical Studies, Jawaharlal Nehru University, New Delhi. She has worked with the Indian National Trust for Art and Cultural Heritage on heritage education and as the Assistant Keeper for Dr Bhau Daji Lad Mumbai City Museum, Mumbai. She has also held teaching positions at Rizvi College of Architecture, Mumbai and at the Old Dominion University and Virginia Wesleyan University, Virginia, USA. Her research interests include archaeology of temples, colonial archaeology, history of museums and religion in the Indian Ocean world. Elizabeth Lambourn is a historian of South Asia and the Indian Ocean world

and committed to the interdisciplinary and cross-cultural study of medieval history. Originally trained in Art History, she now spends a lot of her time reading, and talking to, anthropologists, archaeologists and textual scholars and her work engages equally with texts and ‘things’, and with texts as material ‘things’. She has published widely on varied aspects of the circulation

Contributors  xv

of artefacts, animals, people, and ideas around the Indian Ocean area. Her monograph Abraham’s Luggage. A Social Life of Things in the Medieval Indian Ocean World was published in 2018. She is a Reader (Associate Professor) in South Asian and Indian Ocean Studies at De Montfort University in Leicester (UK) where she teaches specialist modules on the Indian Ocean in addition to more general courses in design history. Arsenio Nicolas  currently teaches MA and PhD courses in Asian music,

ethnomusicology, music history and archaeology and academic writing at the College of Music, Mahasarakham University, Thailand. His current project is on Musical Exchanges between India, China and Southeast Asia. After completing studies at the University of the Philippines with Professor Jose Maceda, he pursued studies in Javanese music and dance at the Akademi Seni Karawitan Indonesia Surakarta (ASKI), now Institute of Indonesian Arts (ISI, 1979–1983), funded by the Indonesian Department of Education and Culture and the Ford Foundation. He has conducted music research in cooperation with the Indonesian Institute of Sciences (LIPI), the Akademi Seni Tari Indonesia (ASTI) Denpasar, Institute of Southeast Asian Studies Singapore, National University of Malaysia, National Museum of Ethnology (Osaka, Japan), Kyoto University Center for Southeast Asian Studies, Nippon Foundation, National Centre for Archaeological Research, Indonesia (ARKENAS), Southeast Asian Ministries of Education Organization, Special Projects in Archaeology and Fine Arts (SEAMEO-SPAFA) and Bangkok and Mahasarakham University, Thailand. Srinivas G. Reddy  is a scholar, translator and musician. He trained in classical South Asian languages and literatures at Brown University and UC Berkeley. Srinivas is Visiting Professor of Religious Studies and Contemplative Studies at Brown University. He lives in Rhode Island and spends his time performing, teaching and conducting research around the world. www.srinivasreddy.org ­ Tara Sheemar Malhan teaches in the Department of History, Janki Devi Memorial College, University of Delhi and completed her PhD from Centre for Historical Studies, Jawaharlal Nehru University. Her book titled Plunging the Ocean: Courts, Castes and Courtesans in the Kathāsaritsāgara was published in 2017 by Primus Books and was based on her doctoral research. She has contributed a paper on elite sexuality in Beyond the Woman Question: Reconstructing Gendered Identities in Early India edited by Kumkum Roy. Her research relates to courts as well as everyday lives, power along with subversion, space, hierarchies and mobility, social production and reproduction, religion as practice and discourse and Sanskrit literature and narratives. Her research interests include cultural history, social identities, gender and sexuality, religion and ritual practices, performing arts and ­Sanskrit aesthetics.

xvi Contributors

Uthara Suvrathan  is an archaeologist who works at the intersection of

archaeology and history to examine the organization of polities and places on the margins of larger states and empires in pre-modern South Asia. She is also interested in archaeological approaches to the landscape and in examining premodern trade and contact across the Indian Ocean. Suvrathan is Assistant Professor in the Jyoti Dalal School of Liberal Arts, NMIMS University, Mumbai. Veronica Walker Vadillo  is a Core Research Fellow at the Helsinki Collegium for Advanced Studies, University of Helsinki, where she is currently studying the role of nomadic fishing communities in state development in the lower Mekong River. She first obtained an MA in Maritime Archaeology from University College London, conducting her dissertation on the nautical iconography of Angkor. She later obtained a D.Phil. in Archaeology from the University of Oxford with a research topic that focused on the riverine cultural landscape of Angkor. She is particularly interested in methods that bring together archaeology, iconography, ethnography and environmental studies for the study of past maritime societies, a framework that she will continue to develop and expand in her upcoming project on the ports and harbours of Southeast Asia in the early modern period.

PREFACE

The chapters in the book draw on two meetings that were held in 2018 on topics related to sharing of new research on the Indian Ocean. These were attended by contributors to the volume with perhaps one exception. The first was a panel discussion jointly coordinated by Tara Sheemar Malhan and me on ‘Ocean Imagineries and Cultural Heritage’ at the Association for Asian Studies in Asia Conference, which was held from 5 to 8 July 2018 in New Delhi. We would like to thank Murari Kumar Jha for participating in the discussion. The second conference was held at the Ashmolean Museum, Oxford on 29 and 30 November 2018 on ‘Knowledge Traditions of the Indian Ocean World’ jointly coordinated by Shailendra Bhandare and me. It was sponsored by the Anneliese Maier Research Award of Humboldt Foundation, Ashmolean Museum and Oxford Centre for Hindu Studies, Oxford. We are thankful to Ingo Strauch, Vincent Tournier, Nesrin El-Galy, Anna M. Kotarba Morley, Robert Bracey and Richard Gombrich for presentations at the conference. Joe Cribb, Damian Robinson, Michael Feener and Diwakar Acharya helped in the smooth running of the sessions. Manish Thakrar, Nandana Nagraj, Shaunaka Rishi Das and Rembert Lutjeharms of the Oxford Centre for Hindu Studies helped in important ways to make the conference a success. Above all it is the hard work and patience of the contributors to this volume that need to be applauded. We are also grateful to the anonymous reviewers for their insightful comments that helped shape the book. Finally, it is to Shashank Shekhar Sinha and his team at Routledge to whom we owe a debt for bringing our discussions to a larger audience.

Taylor & Francis Taylor & Francis Group http://taylora ndfra ncis.com

1 INTRODUCTION Himanshu Prabha Ray

The Danish underwater archaeologist Thijs J. Maarleveld raised the crucial issue of the extent to which maritime cultural heritage and maritime archaeology can be labelled as ‘international’, even though practitioners often regard themselves as a part of an international academic community.1 Heritage issues are part of the international dialogue and cooperation projects entailing ‘culture and development’ are stimulated worldwide and for most laudable reasons (UNESCO 1995). Politically, such projects are a means rather than a goal…. In other words, both in national and in international cooperative situations, it is local, regional or national politics that define the relevance of any attention at all, as quite rightly professional advice is embedded in or made relevant through competent authorities.2 How is this gulf between national histories and maritime cultural heritage to be negotiated keeping in mind trans-national mobility and connectivity in the ancient past? Histories of new nation states written in the post- colonial period have focused on present political boundaries and traced the history of these political entities overlooking not only the larger context within which historical developments took place, but also connectivity across the seas.3 In the context of the history of South and Southeast Asia, Farish Noor has shown that the colonial period marked an epistemic and politicoeconomic break between South and Southeast Asia, thus severing the long cultural, ethnic and commercial links between the Indian subcontinent and maritime Southeast Asia. This saw the demise of the fluid and overlapping world of an interconnected Asia that had hitherto been defined and controlled by Asians

2  Himanshu Prabha Ray

themselves.4 How is this overlapping Indian Ocean world to be identified in the archaeological record and what knowledge traditions held it together? Recent research indicates that intellectual advances in many fields were the outcome of cross-pollination of ideas resulting from travels across the Indian Ocean world in the premodern period. The interchange of ideas across societies and regions created the dynamism necessary for the emergence and sustenance of extensive civilizations and the movement of scholars and students. The challenge is to identify this exchange of ideas in the archaeological and historical record and to initiate interdisciplinary dialogue across several disciplines on the shared culture across the seas. The chapters in this edited book address this challenge through three themes: conceptualizing the seas (Reddy, Malhan, Cobb and Agius), the materiality of knowledge production (Bauer and Johansen, Suvrathan, and Lambourn) and anchoring the coasts (Kulshreshtha, Vadillo, Nicolas and Bhandare). Traditionally historians have used accounts in Greek, Arabic, Persian, Sanskrit and so on to emphasize the predominance of certain language groups in maritime networks, such as ‘Roman trade’, ‘Arab trade’, the ‘Sanskrit Cosmopolis’ or the ‘Buddhist Cosmopolis’. In contrast, conceptualization of the seas and maritime landscapes has been neglected. Four scholars address the subject from different vantage points: Srinivas Reddy from a reading of the Sanskrit Puranas and the narrative of the ‘Churning of the Ocean’; Tara Sheemar Malhan from an analysis of the narrative Sanskrit text, the Kathāsaritsāgara; Dionisius Agius based on Al-Maqdisī’s (d. after 378/988) description of the coastal landscape of the Persian Gulf and the Red Sea; and Matthew Cobb with reference to the kingdom of Da Qin and its people, dating from roughly the mid-third to seventh centuries CE as described in Chinese sources. Another theme of interest is the spread of writing across the ocean, especially scripts such as Brahmi used for inscribing on pottery from Berenike on the Red Sea coast to Vietnam on the east. The use of a diverse variety of scripts for inscriptions is also evident in the Hoq cave on the island of Socotra, as discussed in the next section. The spread of scripts has generally been associated either with trade or kings, such as Ashoka of the third century BCE Mauryan dynasty. The discussion in secondary literature has tended to focus on decontextualized inscriptions, such as the Ashokan writings on rocks and pillars found across the Indian subcontinent. An important site in this regard is that of Maski in Raichur District of Karnataka. Ashokan edicts inscribed on rocks were first discovered at Maski in 1915, and for the first time, the name of the king was deciphered. Earlier inscriptions had referred to the ruler as ‘devanampiya’ or beloved of the gods. Despite the importance of the site, little attention has been given to its archaeological context. This is an issue taken up in the chapter by Bauer and Johansen in the second section on ‘Materiality of Knowledge Production’. The emphasis on material evidence continues in the chapter by Suvrathan who uses archaeological data from the three sites of Berenike, Egypt, Tissamaharama, Sri Lanka and Arikamedu, India to analyse the nature of connectedness in the premodern Indian Ocean. The final chapter in the

Introduction  3

section by Lambourn centres on the materiality of writing in exchanges between the Middle East and South Asia before 1500. More specifically, it focuses on the writing supports and marking technologies used in the correspondence of Middle Eastern sojourners in South Asia, as documented in twelfth- century material from the Cairo Genizah. The shrine was a crucial component of the Asian landscape around which various social, economic and political systems and institutions evolved, as discussed in the third section of the book. Kulshreshtha highlights the coastal shrine at Khor Rori on the Arabian coast, while Veronica Walker Vadillo draws the temples at Angkor in Cambodia through an analysis of the nautical iconography. Epistemic pluralism and the transmission of knowledge through means other than through textual transmission such as recitation, performance, temples and medical therapies were raised in a 2002 Presidential Address to the American Academy of Religion. Relevant to this chapter is the fact that the first play to be performed for the gods, as narrated in the Nāṭyaśāstra of Bharata dated from 200 BCE to 200 CE, was the churning of the ocean of milk by the gods and the demons, which resulted in releasing the jewels from the ocean. Not only is this story prominently sculpted in several temples in India and Southeast Asia including in a huge 49-m-wide panel at Angkor Wat in Cambodia, but it is also a popular theme for dance performances.5 Nicolas addresses the theme from the perspective of Sanskrit musical terms used in inscriptions. He concludes that in Old Javanese inscriptions, there was extensive use of these terms during the Central Javanese period from the ninth to the tenth centuries, but then it virtually disappeared during the East Javanese period starting from the mid-tenth century onwards. Coins have generally been studied as instruments of trade and acculturation used by Empires in antiquity, such as the secondary writings on finds of Roman coins in India and the presence of Roman colonies in the subcontinent. The chapter by Shailendra Bhandare shifts the goal posts and examines how a coin’s appearance is determined and why some ideas develop wide currency, while others remain firmly local. Thus, the book raises new questions and opens several avenues for further research on the diverse means of communication across the Indian Ocean and ways in which this has encouraged dialogue between different communities and shaped the languages and knowledge traditions as well as impacting various religious developments in the area. In the remaining part of this chapter, I would like to place this edited study within the larger context of secondary writings on the theme.

Sacred Geography of the Coast: The Archaeological Evidence An important aspect of fourth- to seventh- c entury maritime networks is the development of multi-religious sites along the Bay of Bengal littoral.6 Nagarjunakonda is located about 150 km inland from the coastal site of Machilipatnam, identified with Maisolos mentioned by the Greek geographer

4  Himanshu Prabha Ray

Ptolemy. More than 30 Buddhist establishments, 19 Hindu temples and a few medieval Jain shrines were unearthed in several seasons of archaeological excavations conducted at the site after its discovery in 1920 until its submergence in 1960. The data from Nagarjunakonda is an indicator of the multi-religious networks that extended across the seas by the middle of the first millennium CE and evident at several sites across Southeast Asia. U Thong is located 45 km from the Mekong in central Thailand and was previously more accessible from the coast than at present. The fourthcentury BCE burial site of Ban Don Ta Phet, which has yielded a rich haul of goods obtained through maritime networks, is located 25 km south of U Thong.7 A copper-plate inscription in Sanskrit was found at the site in 1957 and published by Çoedes in 1958. In the inscription, king Harsavarman donates to two Siva lingas, an issue discussed at length in secondary writings, since U Thong has been considered ‘Buddhist’.8 Archaeological excavations have unearthed Siva lingas and Visnu images now housed in the museum at the site. This multi-religious nature of early sites has been considered the norm for both U Thong and Nakhon Pathom on the west bank of the Ta Chin river, which is closer to the Gulf of Thailand and several other sites in other parts of Southeast Asia. The Go Thap archaeological site in Dong Thap province of central Vietnam was declared a national relic by the Vietnamese Government in 1998 and a Special National Relic in 2012, known both for its rich archaeological heritage and as an important revolutionary base during Vietnam's resistance wars against France and the United States. The archaeological data shows continuity and links with the Oc-Eo culture of the Mekong delta and provides valuable insights into early religious practices in the region. Twenty- eight wooden Buddha images were found in the Go Thap area, and many of these are now in museums in Hanoi and Ho Chi Minh city. Wooden images of the Buddha have also survived in the Mekong Delta, including at sites such as Thap Muoi, Phong Me, Binh Hoa and Ma Noli. Some have been recovered during excavations.9 Radiocarbon dates suggest a date of sixth century CE for the images.10 Tingley in the catalogue to the museum at Ho Chi Minh city describes and illustrates standing Buddha images ranging in height from 30 cm to almost 3 m, usually in double abhaya-mudrā, ­ ​­ with smiling face and downcast eyelids. In addition to the wooden images, excavations have also revealed brick-built shrines to a range of Hindu deities. In addition to Hinduism, sites with affiliation to Buddhism are also wellrepresented along the coasts. Two themes provide unity to Buddhism across Asia: one is the image of the Buddha that has been venerated by communities in all regions through time, and the other is the worship of the relics.11 Both these practices brought the monastics and the laity together in celebration of the eightfold path shown by the Buddha, which was founded on wisdom, morality and concentration. The challenge is to trace the historical trajectory for the expansion of Buddhism across Asia, as donations were made and patronage provided by both the laity and the monastics in the building of stupas and relic shrines. Huge images of the Buddha were

Introduction  5

transported across South Asia and installed as evident from inscriptions recording details of the patron and the date of installation. This underscores the active participation and mobility of learned monks and nuns in the stupas and relic cults across Asia as new monastic complexes developed and the histories of these complexes came to be recorded. Ancient Buddhist texts functioned both as sources of knowledge and as objects of veneration. Wherever Buddhism spread, written works served to transmit and reinforce the religion’s doctrines, rituals, and institutions in new locations. The importance of written texts to Buddhist culture can hardly be exaggerated. At the same time, it must be acknowledged that the oral transmission of Buddhist texts continued for many centuries even after monks began putting sutras into writing.12 Another important element in the spread of Buddhism is the language used to preserve the Buddha’s teachings. In Suvarnabhumi or Southeast Asia, the classical Pali tradition was cherished and preserved by copying, reciting, studying and writing in Pali even after literatures in local languages had been developed.13 In the 1920s, digging at Khin Ba mound in central Myanmar led to the recovery of a relic chamber sealed by a stone slab. The relic chamber contained several sacred finds, the most extraordinary being a book of 20 gold leaves bound by a gold thread. A perfect replica of a palm-leaf manuscript, the ‘golden book’, contains a series of Pali texts and is one of the earliest examples of inscriptions on gold leaves.14 Both at Sukhothai in Thailand and at Pagan in Myanmar, Pali continued to be used for citations from texts and also for compositions, though it was written in the Khom script at the former site. In contrast, Cambodia and Laos have few Pali inscriptions. Thus, it is evident that the choice of language to be used was deliberate and taken after much reflection. The discussion on language cannot be delinked from the sacred geography of religious architecture along the coasts, with ancillary shrines, tanks, ritual networks and routes of pilgrimage connecting the littoral with the interior.15 It draws into dialogue riverine routes leading inland from the coast as well as to the ‘foreland’ across the seas. This brings into the conversation not only the diverse usage of sites and shrines as centres of social integration but also incorporates data from ‘movable’ or ‘portable’ ritual objects, such as bronze icons and clay votive tablets. Burials and shrines along the coasts were by no means restricted to South and Southeast Asia. There has been increasing interest among archaeologists working across the Indian Ocean in identifying coastal shrines and in formulating their wider networks. Eivind Seland refers to Judaist presence in the Southern Red Sea and Southern Arabia in late antiquity, including a synagogue in the Hadramawt port city of Qana probably dated to the fifth century CE.16 Christian communities are also attested in Southern Arabia, at the island of Socotra and in the Aksumite kingdom. One of the roles of the originally

6  Himanshu Prabha Ray

Egyptian goddess Isis in the Roman period is as a patroness of seafaring, and she seems to be one of the deities worshipped in the harbour temple excavated by the American–Polish team at the Egyptian Red Sea port of Berenike.17 Water and ritual bathing were important aspects of the cult of Isis.18 However, religious expansion continues to be linked to trade, and the Hoq cave at Socotra is often discussed in terms of trading networks. As emphasised in an earlier paper, there is a need to decouple trade as a causal factor from religious activity if the history of the Indian Ocean is to be studied in terms of oceanic space rather than as an extension of land-based routes.19 The sea then becomes the site of cultural encounters that can be traced through routes and island archaeologies. In the first centuries CE, Hoq cave on the northern face of the island of Socotra was known in a wide region by seafarers and was an important halting point for travel from western India to the east African coast. Explorations in the 2,000-m-deep cave in 2001 led to the discovery of 200 graffiti, drawings and small offerings dated from the late second to the fourth centuries CE. Of these, 192 are in the Brahmi script, 1 in Kharosthi, 1 in Bactrian, 3 in Greek, 1 in Palmyrene Aramaic and 20-odd in Axumite or ancient Ethiopian.20 The overwhelming presence of inscriptions in Brahmi indicating self-representation by travellers from the Indian subcontinent as compared to those in other scripts needs to be highlighted since many of these bear comparison with those from the Buddhist cave site of Kanheri on Salsette island off the west coast of India. The graffiti and inscriptions are not randomly engraved in the cave; instead, they occur in clusters at specific places, most likely representing deliberate choice of location. The second of the two Greek inscriptions found at Hoq expresses the sacred nature of the cave: ‘Septimius Paniskos, the ship-owner, kneeled before the gods and before (that/those) of the cave’.21 As documented by an Oral History Project of the University of Warwick, sailors from Kachchh in Gujarat on their way to East Africa often stop at Socotra to pay homage to Goddess Sikotar Mata, offering her ship models and seeking her blessings.22 The temple of Sikotar Mata (padia) situated on the Sikotar Hill coast is maintained by a Hindu priest known as the pujari, and a visit to the temple is an integral part of the rituals the seafarers have followed for centuries to protect them from the perils of the sea. Thus, several traditions of the past live on, and there is an urgent need to include them within the academic discourse not only for their protection and preservation, but more so for a holistic understanding of sea spaces. Lynne Newton has drawn attention to at least 50 shrines in the Salalah area of Dhofar, which are distinctive and different from those in Hadramawt.23 Many of the shrines have been modernized and are often not easy to detect, as they are invariably comprised of open-air graves of important people such as prophets and saints with no formal roofing. ‘There is a long tradition of shrines in the Hadramawt serving as places of mediation for tribal disputes and for trade’.24 There are at least five known shrines in the vicinity of Al Baleed, and these date to the early fourteenth century CE.

Introduction  7

In the premodern world, religious shrines were not merely centres of worship and devotion but were also vital institutions for providing learning and knowledge to members of society. There are references in early inscriptions from the Indian subcontinent to the presence of pathsala or schools attached to temples, as well as medical centres and arrangements for distribution of medicinal herbs. The beginnings of these networks may be identified in the archaeological record by specimens of writing on pottery, seals and sealings and by inscriptions on stone and copper plates.25 Horden and Purcell refer to the rural shrine in the Mediterranean region and the web of connectivity that links the microregional landscape, which blurs the distinctions that ‘historians of religion have been inclined to focus’.26 They also argue for self- conscious revival of religious practices rather than continuity and survival.

Languages of Knowledge Transmission The use of Arabic in the Indian Ocean world is often seen as marking a new beginning in seafaring activity, though there is evidence for continuity in terms of knowledge traditions, especially those related to astronomy and medicine. In the course of the eighth century, Arabic became the new language of prestige and interregional communication from North Africa to the edges of India. This also initiated a long period of translations, in which the parts of older literatures useful to the new elite were converted into Arabic from c. 750 to 1000 CE. Most of the texts that were translated were from Greek but also included Sanskrit texts on astronomy, medicine and political maxims. Van Bladel discusses the role of the Barmakid family from Tokharistan in the process of translation at the Abbasid court. At the time of Arab conquest in the seventh and eighth centuries, Buddhism flourished in Tokharistan roughly coterminous with the region of Bactria of the Greeks and Sanskrit studies were firmly established there. The main city of Tokharistan was Balkh, which was known for its Buddhist monasteries and Sanskrit learning.27 Fragmentary medical manuscripts in Sanskrit written in the Brahmi script have been found at Shahr–i Zohak, 15 km east of Bamiyan. In addition, Russian archaeologists have discovered Buddhist stupas and monasteries north of the Oxus. The great vizier Yahya al-Barmaki was a patron of scholars and poets. Yahya had sent a mission to India for two purposes: to collect Indian medicinal plants and to learn about the religions there. The surviving portion of the report mentions the way to India from Balkh and discusses the giant Buddha statues at Bamiyan. It is suggested that many of the accounts of India in Arabic were based on the now lost books written for Yahya. The Barmakids had a hospital in Baghdad at which Yahya employed Indian physicians who translated Sanskrit medical works into Arabic. These included Suśruta, the Aṣṭāṅgahridaya samhitā of Vagbhatta and the Siddhasāra of Ravigupta. In addition to medical treatises, Sanskrit texts on astronomy were also known in Baghdad. David Pingree has argued that an

8  Himanshu Prabha Ray

astronomical handbook was translated from Sanskrit into Arabic in Sind, in India, as early as ‘shortly after 735’.28 The tenth- c entury compilation of Balkhī School geographies was limited to what was known to Muslim authors as the Dār al-Islām ­ ​­ (Terrain of Islam).29 An exception at this time was al-Biruni who had learnt Sanskrit and was able to access the human geography of India. Islamic world maps represented the Indian Ocean in the form of birds and animals. The Indian Ocean was referred to by some ‘Muslim geographers as the “Great Sea” or the “Green Sea” both name tags were borrowed from Persian and Indian conceptual thinking’.30 The Sea of China was seen as an open sea, which was both frightening and dangerous for Muslim sailors.

Knowledge Traditions: Historiography In recent years, knowledge networks have drawn the attention of researchers, as evident from a spurt of publications on the theme. These edited books highlight different aspects of the flow of ideas with an emphasis on certain parts of the Indian Ocean world. The contributions of western India, for example, are examined in a book edited by Sara Keller, through intangible exchanges focusing on the spread of culture and knowledge with reference to ceramics, nautical charts, travelling scholars, Sufis, healing practices and crafts traditions. The chronological emphasis is on the post­fifteenth-century period.31 In contrast, another edited volume largely based on archaeological data from eastern India draws attention to ‘knowledge creation’ and ‘knowledge networks’, such as ceramics, beads and moulded clay tablets; institutions like monastic Buddhism; communities of seafarers; and individuals in the context of ‘social space’.32 These issues have a long prehistory in the context of the Bay of Bengal and the archaeology of the Indian subcontinent,33 and the edited book does not explicate the ‘revisionist’ structures it proposes within which the knowledge networks were situated in India, limiting the discussion of ‘revisionist’ trends to the period from 1000 to 1500 CE Southeast Asia. Unlike the above-mentioned studies, which largely deal with India, the two volumes edited by Angela Schottenhammer devote much attention to China. The study takes up interconnectivity across the Indian Ocean world through commercial exchanges as well as exchange of ideas, religions and technologies. The objective of the study is to stress the socio- economic foundations of the Indian Ocean world and to establish that the major religions played a very positive role in the development of commerce across the macro-region, though the process is not spelt out in the chapters.34 Buddhism and Hinduism are dealt with in sequential order, despite the fact that both South and Southeast Asia were multi-religious with Buddhism and Hinduism coexisting in the historical period and the emphasis continues to be on the centrality of the state.35 An edited volume that breaks away from the centrality of the state and socio- economic structures that have popularized the association of

Introduction  9

Buddhism with trade and that of brahmanas as a legitimizing agency of the state focuses on ‘religious orders’ conceptualized as a comparative category for discussion of Sufi brotherhoods and Buddhist monastic and ritual lineages.36 The region under discussion extends from India, Sri Lanka and the Himalayas through mainland Southeast Asia and across the Indonesian archipelago. The overlap between Buddhist and Sufi orders has been an under-researched theme, and this edited volume provides insights into a theme of research that opens several new avenues. The edited volume by Burkhardt Schnepel and Edward Alpers refers to the ‘island factor’ or the significance of islands in the history of the Indian Ocean.37 This significance draws on the location of islands as hubs of commercial exchanges or in terms of military strategy, but of equal importance is the role of islands for pilgrimage, e.g. the Shinto shrine of Kojima on Maekojima, an uninhabited islet about 60 m wide situated off the southeast edge of much larger Iki Island in Japan. The past six to seven decades have coincided with an expansion of archaeological techniques and methods, and a vast corpus of archaeological data is now available for understanding cross- cultural contacts. These writings date the beginnings of maritime activities to the prehistoric period and emphasize that these networks dealt with food stuffs and staples, such as cloth and medicinal plants.38 An analysis of the contents of pots from archaeological sites such as Khor Rori in South Arabia, Mleiha (c. third century BCE – m id-third century CE), Ed-Dur (c. first century BCE/CE–third century CE) and Dibba (c. first century CE–mid-third century CE) has led Anjana Reddy to suggest culinary change and the adoption of new foodstuffs and new forms of food preparation/consumption in the Arabian context.39 Seafaring was handled by mobile small-scale seafaring communities travelling across the seas, rather than by elite merchant groups at the behest of States or Empires.40 It is within these new research parameters that the chapters in this book provide fresh insights into hitherto under-researched themes.

Notes 1 Thijs J. Maarleveld, The Maritime Paradox: Does International Heritage Exist? International Journal of Heritage Studies, 18(4), 2012: 418–431. 2 Maarleveld, The Maritime Paradox, p. 420 [418–431]. 3 Himanshu Prabha Ray, Writings on the Maritime History of Ancient India, in Sabyasachi Bhattacharya (ed.), Approaches to History: Essays in Indian Historiography, New Delhi: Indian Council of Historical Research and Primus Books, 2011, pp. 27–54. ­ 4 Farish (Badrol Hisham) A hmad-Noor, Divorced by Empire: Colonial-Era Epistemology and the Division of South and Southeast Asia, in Shyam Saran (ed.), Cultural and Civilisational Links between India and Southeast Asia: Historical and Contemporary Dimension, New Delhi: Palgrave Macmillan, 2018, pp. 301–323. ­ 5 Vasudha Narayanan (September 2003), Embodied Cosmologies: Sites of Piety, Sites of Power, Journal of the American Academy of Religion, 71(3): 495–520.

10  Himanshu Prabha Ray

­

­

​­

­ ­

­ ­



­





​­



­

­

Introduction  11

25 Ray, Coastal Shrines and Transnational Maritime Networks, pp. 25–53. ­ 26 Peregrine Horden and Nicholas Purcell, The Corrupting Sea: A Study of Mediterranean History, Oxford: Blackwell Publishers, 2000, p. 406. 27 Kevin van Bladel, The Bactrian Background of the Barmakids, in Anna Akasoy, Charles Burnett and Ronit Yoeli Tlalim (eds), Islam and Tibet  – Interactions along the Musk Routes, Surrey: Ashgate, 2011, pp. 43–88. 28 van Bladel, Islam and Tibet, ­p. 75. 29 Dionisius Agius, Sea Zones: The Balkhi School’s Conceptual Mapping of the Indian Ocean, Topoi (Suppl.) 15, 2017, pp. 225–243. 30 Agius, Sea Zones, p. 234. 31 Sara Keller (ed.), Knowledge and the Indian Ocean: Intangible Networks of Western India and Beyond, Cham: Palgrave Macmillan, 2019. 32 Kenneth Hall, Preface, Kenneth R. Hall, Suchandra Ghosh, Kaushik Gangopadhyay and Rila Mukherjee (eds), Cross- Cultural Networking in the Eastern Indian Ocean Realm, c.100–1800, New Delhi: Primus Books, 2019, p. xv. 33 Himanshu Prabha Ray, The Winds of Change: Buddhism and the Maritime Links of Early South Asia, New Delhi: Oxford University Press, 1994; Himanshu Prabha Ray and Jean-François Salles (eds), Tradition and Archaeology: Early Maritime Contacts in the Indian Ocean, New Delhi: Manohar Publishers, 1996; Himanshu Prabha Ray, The Archaeology of Seafaring in Ancient South Asia, Cambridge: Cambridge University Press, 2003; P.Y. Manguin, A. Mani and G. Wade (eds), Early Interactions between South and Southeast Asia. Reflections Cross Cultural Exchange, Singapore: Institute of South-East Asian Studies, 2011; John Guy (ed.), Lost Kingdoms: Hindu-Buddhist Sculpture of Early Southeast Asia, London and New Haven: Yale University Press, 2014. 34 Angela Schottenhammer (ed.), Early Global Interconnectivity across the Indian Ocean World, Volume II: Exchange of Ideas, Religions, and Technologies, Cham: Palgrave Macmillan, 2019. 35 Himanshu Prabha Ray, Multi-Religious Maritime Linkages across the Bay of Bengal During the First Millennium CE, in Nicolas Revire and Stephen A. Murphy (eds), Before Siam: Essays in Art and Archaeology, River Books/The Siam Society, Bangkok, 2014, pp. 132–151. 36 R. Michael Feener and Anne M. Blackburn (eds), Buddhist and Islamic Orders in Southern Asia: Comparative Perspectives, Honolulu: University of Hawaii Press, 2019. 37 Burkhard Schnepel and Edward A. Alpers (eds), Connectivity in Motion: Island Hubs in the Indian Ocean World, Basingstoke: Palgrave Macmillan, 2018. 38 Himanshu Prabha Ray, Writings on Maritime History of Ancient India, in Sabyasachi Bhattacharya (ed.), Approaches to History: Essays in Indian Historiography, New Delhi: ICHR and Primus Books, 2011, pp. 27–54. 39 Anjana Reddy, Archaeology of Indo- Gulf Relations in the Early Historic Period: The Ceramic Evidence, in Himanshu Prabha Ray (ed.), Bridging the Gulf: Maritime Cultural Heritage of the Western Indian Ocean, New Delhi: India International Centre and Manohar Publication, 2016, pp. 53–78. 40 Satish Chandra and Himanshu Prabha Ray (eds), The Sea, Identity and History: From the Bay of Bengal to the South China Sea, Singapore/New Delhi: ISEAS/ Manohar, 2013; Himanshu Prabha Ray (ed.), Bridging the Gulf: Maritime Cultural Heritage of the Western Indian Ocean, New Delhi: India International Centre and Manohar, 2016.

Taylor & Francis Taylor & Francis Group http://taylora ndfra ncis.com

PART 1

Conceptualising the seas Introduction Himanshu Prabha Ray

The Indian Ocean is the third largest of the world’s seas and has been known by its present name since at least 1515, when the Latin form Oceanus Orientalis Indicus (Indian Eastern Ocean) is attested. Scholars have highlighted transformations that ensued as a result of European colonization which began in the sixteenth and seventeenth centuries through commercial imperialism.1 The mapping and territorializing of the Ocean emerged as a colonial project through scientific developments such as marine surveys and nautical cartography, establishment of port towns and the introduction of mechanised navigation. The chapters in this book deal with a period prior to these radical interventions when it was the sailing ship that guided mobility and connectivity across the ocean. The monsoon wind system provided unity to the Indian Ocean and determined not only the patterns and rhythms of voyaging, but also agricultural productivity across the region. The two main monsoon systems are significant: one blowing from the northeast in the winter and the other, southwest during the summer with a variable weather season in between, though there are certain regions such as the Red Sea which remain outside its fold. An ethnographic study of the Kharava, one of the traditional sailing communities of Gujarat2, shows the continued impact of the monsoon system even at present. The sailing calendar continues to be based on the concept of Naroj or ‘the new day to ply in the sea by the blessing of the sea god Varun’. The calendar commences around July 28 after the cessation of the monsoon and continues for around 300 days until the beginning

14  Himanshu Prabha Ray

of the monsoon towards the end of May. There is no sailing at the height of monsoon for around 60–65 days.3 The dependence on monsoon winds meant that sailing in the Indian Ocean took place along certain routes, though there were no doubt changes over time. For example, Mandvi on the north coast of the Gulf of Kachchh is a centre for dhows or traditional sailing vessels travelling to the Makran coast and east Africa, while vessels from Jamnagar on the southern side of the Gulf connect with coastal centres in Iran and Iraq.4 In contrast, Veraval on the Saurashtra coast is known to have linked with centres in Myanmar, though much of this sailing route is no longer possible. Porbandar was another important centre on the Saurashtra coast with ships travelling both to South Arabia and Africa as well to Burma, Singapore and other parts of Southeast Asia. In an earlier paper,5 I stressed the continued significance of the cultural route between islands such as Salsette, which now forms a part of the city of Mumbai and Socotra at the mouth of the Red Sea. Teakwood for the construction of the hawārī boats on Socotra is still imported from India6 and sailors from Kachchh in Gujarat on their way to East Africa often stop at Socotra to pay homage to Goddess Sikotar Mata, offering her ship models and seeking Her blessings. Researchers agree that the ‘island factor’ has not received the attention that it deserves in the writing of maritime history of the Indian Ocean, notwithstanding their geopolitical and strategic importance. India has a total of 1,208 islands (including uninhabited ones). While some of these lie off the Indian coastline, such as the Lakshadweep group including 39 islands of which 11 are inhabited, others such as the archipelago of Andaman and Nicobar group comprising 572 islands of which 34 are permanently inhabited are located at the juncture of the Bay of Bengal and the Andaman Sea. Jambudvīpa or island of the Jambu tree is a term that figures in the third-century BCE edicts of the Mauryan ruler Ashoka, especially the Minor Rock Edict at Brahmagiri in the Chitradurga district of Karnataka. Descriptions of Jambudvīpa and Uttarakuru occur in early Buddhist literature as well and are occasionally juxtaposed with Sihaladīpa referring to Sri Lanka, while Jambudvīpa stands for India.7 Ashoka uses the term in the Minor Rock Edict at Brahmagiri to mean the whole of his empire,8 which covered nearly the entire Indian subcontinent excluding the far southern part of its peninsula and the same term occurs in the Mahabharata.9 In Buddhist texts, Jambudvīpa or island of the jambu tree is the southernmost of the four islands/ continents where human beings reside in this world. At the centre of Jambudvīpa is found the Vajrāsana (‘diamond seat’), the spot where the Buddhas realise their enlightenment; hence, Jambudvīpa is always the continent where Buddhas spend their final lifetimes establishing their dispensations and is therefore the most auspicious site for beings to take rebirth. Alternatively, in Puranic cosmology, Jambudvīpa is said to be at the centre of seven islands that surround Mount Meru. The entire cosmos is divided into seven concentric island/continents (sapta-dvīpa vasumati)

Conceptualising the seas  15

separated by the seven encircling oceans, each double the size of the preceding one (going out from within). How does this perception of the Indian subcontinent as an island compare with archaeological evidence? The data from paleobotany indicates maritime travel in the Indian Ocean from at least the third-second millennium BCE and sustained contacts throughout the historical periods.10 There is consensus among archaeologists that the corridor extending from the Persian Gulf to the west coast of India was an important element in the cultural circuit of the third millennium BCE. There is increasing evidence for the possible presence of communities from the Harappan cultural region in the interior of central Oman during the second half of the third millennium BCE.11 At the other end of the Indian Ocean, archaeological research in several countries of Southeast Asia has provided insights into the maritime network in the first and second millennium BCE. The role of small-scale societies in these contacts is as yet inadequately understood. A significant factor facilitating the integrative potential of these communities was their large cargo-carrying vessels, which not only facilitated transformation of the local settlements into centres of commerce and production, but also linked local groups into regional and transregional networks. Underwater archaeology has contributed to an understanding of the boat-building traditions of the Indian Ocean, which was further supplemented by ethnographic studies of contemporary boat-building communities. By the beginning of the Common Era, these connections had expanded to include almost the entire stretch of the Ocean from the Red Sea all the way east to the coast of Vietnam, as evident from the distribution of certain characteristic ceramics. For sailing activity in the western Indian Ocean, the Periplus Maris Erythraei by an anonymous Greek writer from Alexandria dated to the first century CE is an invaluable source, while increasing archaeological data from several coastal sites in Sri Lanka and Southeast has helped flesh out the picture across the Bay of Bengal. However, the Persian Gulf is one region that continues to be somewhat neglected and a blind spot of early Greek writing. An important hub at the mouth of the Red Sea was located at Aden, often termed a ‘maritime city’ from the eleventh to thirteenth century and also the centre of Jewish activity in the Indian Ocean.12 The Aden to Mangalore route linking the mouth of the Red Sea to the west coast of India is prominently referred to in the India letters of the Cairo Geniza dated to the eleventh-twelfth centuries CE. These Judeo-Arabic texts recovered from the Geniza chamber of the Ben Ezra synagogue in Fustat–Old Cairo provide fascinating insights into the maritime world of the Indian Ocean and the recently translated 459 documents (comprising 523 shelf marks of the India Letters from the Cairo Geniza dated from CE 1080 to 1160) are crucial to this discussion.13 New sectors were added to the networks over time, as evident from accounts of sailing and travel in Arabic, such as the ninth-century Book of

16  Himanshu Prabha Ray

Routes of Ibn Khurradādhbih, which describes several stages in the voyage to China. Adventures of sailors across the Indian Ocean have been immortalised in stories in different languages, many of which survive in later Arabic versions, such as in the Ajaib al-Hind or Marvels of India and narratives of Sindbad the Sailor. One of the popular tales is that of Ishaq the Jew who left Oman with 200 dinars but returned after 30 years from China with a ‘couple of million dollars’ worth of merchandise’.14 Several clusters of Tamil inscriptions have been found on the eastern fringes of the Indian Ocean from Burma (Myanmar) to Sumatra. Of the seven mid-ninth to late-thirteenth century Tamil or part-Tamil language inscriptions so far found in Southeast Asia, one has been found near Pagan in Burma, two just south of the Isthmus of Kra in the Malay peninsula and four in north and west Java. Perhaps the easternmost record is the bilingual Tamil and Chinese language inscription found associated with the remains of one of the two Siva temples at Quanzhou in south China.15 These inscriptions connect groups operating out of south India with the founding or the endowing of temples or other structures for the use of the resident Indian community. Relevant to this discussion are the voyages of Zheng He, who is regarded within the Chinese tradition as a Ming envoy sent abroad at the head of naval armadas by the Yong-le Emperor (1403–1425) on seven occasions to develop relations of peace and friendship with rulers in the Indian Ocean region. The voyages sent by the Yong-le Emperor were different from earlier ones in that they were commanded by eunuchs. It is obvious that these fleets were crewed by a wide range of peoples. Many of the eunuch commanders were Muslims, the navigators were often non-Chinese, and it is possible that descendants of Fu-jian Arabs were also included in the crew.16 Tibbett’s suggests that navigation techniques and texts of the Indian Ocean shared several common methods that were probably used by all those bordering on the shores of that ocean and were to some extent representative of all Indian Ocean sailing.17 The use of compass card and stellar navigation in the Indian Ocean is well-documented in secondary literature, though it has almost disappeared in contemporary coastal communities. The Muallim nī pothī or captain’s manual now preserved in the National Museum, New Delhi provides fascinating insight into the sailing world of the Indian Ocean and changes over time. A part of this material was studied by B. Arunachalam, especially one of the manuals or pothīs,18 which contains five maps of parts of the coastline of peninsular India and Sri Lanka accompanied by sailing directions. This pothī bears the date VS 1710, or 1664 CE, which makes its contents the oldest known Indian coastline maps. These broad generalizations are important to understand cultural routes and patterns in maritime networks that defined the Indian Ocean in antiquity. These ancient routes have generally not formed a part of discussions of connectivity and mobility across the seas. Also marginalised are the several island sites known for their biodiversity and unique marine environment. This also raises questions about definitions of sea spaces and the knowledge

Conceptualising the seas  17

of the oceans. In a 1995 article Australian geographer Sue Jackson argued that the dominance of western views and concepts of nature and landscape predicated on western knowledge and values marginalised aboriginal idea of ‘caring for country’, which also encompassed sea space. This was an aspect neglected by a legal apparatus originated in the European occupation of Australia.19 A recent trend has been to emphasise transnational histories as a means to overcome the boundedness of the nation state, which runs the risk of retaining the centrality of the nation state and viewing the seas merely as extensions of it.20 Oceans undoubtedly are critical spaces for diverse projects of the imagination, governance and material exchange. In secondary writings, Sanskrit inscriptions and other texts, Arabic travelogues and European records have been used to indicate the development and complexity of different mercantile networks across the ocean. How far does the maritime network extend— only to littoral societies and connecting hinterlands? Another way of examining the issue is through various modes of conceptualization of the ocean that different societies adopt.21 It is these textual representations that form the theme of the first section of the book.

Notes 1 Sunil Amrith, Unruly Waters: How Mountain Rivers and Monsoons Have Shaped South Asia’s History, London: Penguin, 2018, p. 12. 2 The Kharavas are the premier sailing community with a long history of navigational skills and expertise on sea. Twelve villages in Kutch and Saurashtra are designated to them, starting from Koteshwar in Kutch to Mandvi, Beyt Island, Porbandar, Veraval and so on. However, ship-building is by no means restricted to these 12 villages and instead occurs at several small villages all along the coastline, Nancy Pinto Orton, Sea-Going Trade in Early Historic Gujarat (c. 100 BC to AD 500), unpublished PhD thesis, University of Pennsylvania, 2001, pp. 59–74. 3 Orton, Sea-Going Trade, p. 86. 4 Orton, Sea-Going Trade, pp. 95–96. 5 Himanshu Prabha Ray, From Salsette to Socotra: Islands across the Seas and Implications for Heritage, in Burhard Schnepel and Edward A. Alpers (eds), Connectivity in Motion: Island Hubs in the Indian Ocean World, Palgrave Series in Indian Ocean World Studies, Basingstoke: Palgrave Macmillan, 2018, pp. 347–368. 6 Julian Jansen van Rensburg, The Hawārī of Socotra, Yemen, The International Journal of Nautical Archaeology, 39(1), 2010: 99–109. 7 Mahavamsa, V.13; Culavamsa, XXXVII.216, 246; G. P. Malalasekera, Dictionary of Pali Proper Names, II, pp. 941–942, sv. Jambudipa. 8 D.C. Sircar, Select Inscriptions Bearing on Indian History and Civilization, ­Calcutta: University of Calcutta, 1965, I, no. 2, line 2. 9 Mahābhārata VI 12,1; VI 8,19. 10 Dorian Q. Fuller, Nicole Boivin, Tom Hoogervorst and Robin Allaby, Across the Indian Ocean: The Prehistoric Movement of Plants and Animals, Antiquity, 85, 2011: 544–558. 11 Dennys Frenez, Michele Degli Esposti, Sophie Méry and Jonathan Mark Kenoyer, Bronze Age Salūt (ST1) and the Indus Civilization: Recent Discoveries

18  Himanshu Prabha Ray

12 13 14 15 16 17 18

19 20

21

and New Insights on Regional Interaction, Proceedings of the Seminar for Arabian Studies 46, 2016: 107–124. Roxani Eleni Margariti, Aden and the Indian Ocean Trade, Chapel Hill: The University of North Carolina, 2007. S.D. Goitein and Mordechai Akiva Friedman, India Traders of the Middle Ages, Leiden and Boston: Brill, 2008. Abdul Sheriff, Dhow Cultures of the Indian Ocean: Cosmopolitanism, Commerce and Islam, London: Hurst & Company and Zanzibar: Zanzibar Indian Ocean Research Institute, 2010. H. Kulke, K. Kesavapany and V. Sakhuja, Nagapattinam to Suvarnadwipa: Reflections on the Chola Naval Expeditions to Southeast Asia, Singapore: Institute of Southeast Asian Studies, 2009. Geoff Wade, The Zheng He Voyages: A Reassessment, Asia Research Institute Working Paper Series, 31, 2004. Gerald R. Tibbetts, The Role of Charts in Islamic Navigation in the Indian Ocean, in J.B. Harley and David Woodward (eds), The History of Cartography, Volume II, Book I, Chicago: The University of Chicago Press, 1992, p. 257. B. Arunachalam , Traditional Sea and Sky Wisdom of Indian Seamen and their Practical Applications, in Himanshu Prabha Ray and Jean-François Salles (eds), Tradition and Archaeology: Early Maritime Contacts in the Indian Ocean, New Delhi: Manohar, 2012 reprint, pp. 261–282. Sue Jackson, The Water Is Not Empty: Cross-Cultural Issues in Conceptualising Sea Space, The Australian Geographer, 26, 1995: 87–96. Rila Mukherjee, Escape from Terracentrism: Writing a Water History, Indian Historical Review, 41, 2014: 87–101. Isabel Hofmeyr, The Complicating Sea: The Indian Ocean as Method, Comparative Studies of South Asia, Africa and the Middle East, 32, 2012: 584–590. S.R. Prange, Scholars and the Sea: A Historiography of the Indian Ocean, History Compass, 6(5) 2008: 1382–1393.

2 SEVEN SEAS AND AN OCEAN OF WISDOM An Indian episteme for the Indian Ocean Srinivas G. Reddy

The seven great insular continents are concentrically surrounded by seven great seas: the sea of saltwater, of sugar-cane juice, of wine, of clarified butter, of curds, of milk, and of fresh water. Jambudvīpa is in the centre of all these, and in the centre of this continent is the golden mountain Meru.1

With some minor modifications here and there, this short passage from the Viṣṇu Purāṇa summarises one of the most enduring formulations of ancient Indian cosmology. In a puranic worldview, we humans live on the great continental island of Jambudvīpa that is surrounded by a sea of saltwater. Emanating from this epicentre are six progressively larger concentric rings of land, each bound by six more oceans of various constitutions.2 And at the very heart of this cosmic design is the golden mountain Meru, the axis mundi, the centre of the whole universe. What does the idea of an ‘Indian Ocean’ mean for someone coming from such a cosmological framework? For one thing, the puranic worldview did not posit an Indian Ocean per se, but rather one unifying ocean for the whole world. From an ancient Indian perspective, there is but one ocean for humanity; one vast and dynamic watery expanse that surrounds the land and nurtures it. The ocean provides for us, just as it defines us. From this cosmological perspective we may interpret the ancient Indian notion of an ocean in variety of productive ways. As Sri Aurobindo believed, ‘the Vedic Rishis used the image of water, a river, or an ocean, in a figurative sense and as psychological symbol’.3 If we indeed interpret Indian mythologies of the ocean with such a hermeneutic, I believe we can reinvigorate critical academic debates regarding the nature of oceans and their relation to human history and culture. In particular I will draw on a popular mythic cycle within Hindu cosmology/mythology, namely the samudra-manthanam or Churning of the (Milky) Ocean. With this core myth as a foundation,

20  Srinivas G. Reddy

I hope to tease out generative ideas for a new vision of the ocean, and more specifically, the Indian Ocean as we commonly understand it today. As Sri Aurobindo adds, ‘let us see how far it takes us…’.4

I A return to the mythic archive of India allows us to engage with an alternative way of thinking: a new cosmology, a radically different ontology and a premodern epistemology that challenges many of the fundamental principles of our current episteme. From a Foucauldian perspective, our modern, post-Enlightenment episteme is a cognitive framework: in which knowledge, envisaged apart from all criteria having reference to its rational value or to its objective forms, grounds its positivity and thereby maintains a history which is not that of its growing perfection, but rather that of its conditions of possibility.5 Two important claims are put forth here. One is the objectification of knowledge, the envisaging of it as separate from its ontological reality, its form and value. The second is a flexibility built into the model that posits possibilities rather than boundaries. This second claim is critical, and we will return to it later. The first proposition, however, lies at the crux of our investigation. Can we so easily and conveniently disjoin an ontology from an epistemology? Doesn’t the nature of a thing influence how we perceive it? And isn’t our perception guided by what a thing is? An epistemology, a way of knowing, is a concept that necessitates a clear demarcation between the knower and the known, between a subjective perceiver and an objective perceived. This fundamental bifurcation is the backbone of the entire post-Enlightenment Western episteme. To put it another way, our current episteme privileges the cognitive regime of rational, evidence-based science, a worldview that has led to great technological advancements but also a zealous attachment to scientism.6 In such an episteme of science, the outside world (i.e. nature) is bracketed out, separated and distanced from us so that it may be (ab)used to meet the needs of humanity’s rampant avarice. Furthermore, nature is the object upon which, and through which, humans build their subjective reality. This ontologically limiting model becomes a prerequisite for the development of a robust Western epistemology. This is not to say that India, or other non-Western cultures, did or do not have epistemologies, for they certainly did and do. In India epistemology was a well-established philosophical discipline in which several Indian schools/traditions engaged in rigorous explorations of the pramāṇas or means of knowing.7 Some schools posited six pramāṇas, others three, others nine, but one mode of knowing remained common to all Indian schools of philosophy: pratyakṣa, or perception. No matter how much these rival schools disagreed and/or debated amongst themselves, they all concurred that the ultimate source of knowledge was the self, the

Seven seas and an ocean of wisdom   21

individual, vis-à-vis personal experience of both the internal and external world. Figures like Buddha and Mahavira were ideal embodiments of such experiential wisdom. All Indian schools of thought, as distinct from European/ ‘scientific’ thought, treated experience as the chief raw material of philosophical reflection. Moreover, they all valued revelation/insight in such states of heightened consciousness as the ultimate sources of valid knowledge.8 This may seem like utter nonsense from a ‘rational’, ‘objective’ and ‘scientific’ perspective, but that is the whole point. The critical issue is that there exists an epistemically bound hierarchy of knowledge forms. What happens to these regimes of truth when disparate cultures meet? Which ways of knowing are capable of being deemed true, valid and/or valuable? At some level the Indian episteme and the Western episteme are incommensurable. They are two different games, with different playing fields, rules and players. And yet, many of the rules overlap, and some players certainly learn both games. It is thus important to mention here that the real conceptual split here is not between East and West but rather what we have understood as modern9 as opposed to premodern. As post-Enlightenment Europe embraced modernity, insight was divorced from discovery; wisdom was relegated to the realm of the spirit while the domain of science claimed right over knowledge. In other words, knowledge had to be scientific. As Foucault would later write: I would define the episteme retrospectively as the strategic apparatus which permits of separating out from among all the statements which are possible those that will be acceptable within, I won’t say a scientific theory, but a field of scientificity, and which it is possible to say are true or false. The episteme is the ‘apparatus’ which makes possible the separation, not of the true from the false, but of what may from what may not be characterised as scientific.10 The inevitable bifurcation of faith-based religion from rational science, the manufactured split between the inner world and the outer world, is the conceptual underpinning of modern Western epistemology; it is also the fundamental principle that distinguishes it from Indian worldviews. This is not to say that Indians did not experience an objective, sensory world, they did, but they also embraced a subjective reality of self and cosmos. In a peculiar way, India never developed an epistemology in the modern sense of the world; for if knowledge itself is conceived of as an embodied reality, then knower and known collapse into singular phenomenon, thereby dissolving the essential subject–object duality of an epistemology. In general, Indian theories of knowledge were more concerned with questions of ontology, and more importantly phenomenology, rather than epistemology. The common underlying premise of all Indian schools of thought is that the knower and the known are ultimately one. It is a patently nondualist worldview. Indeed, the fundamental equivalency of humanity, nature

22  Srinivas G. Reddy

and knowledge had deep implications for how ancient Indians conceived of themselves, their cosmos and their relationship with the natural world. Similarly, the West was deeply transformed by the artificial cleavage of science and religion. As Conze remarks: there began in Europe that estrangement from reality which is the starting-point of most modern European philosophy. Epistemology took the place of ontology. Where ontology was concerned with the difference between reality and appearance, epistemology concentrated in that between valid and invalid knowledge.11 This shift away from ontology and towards epistemology had serious consequences. ‘The tone for all later phases of modern philosophy asserted that things by themselves have no relations to one another, and that a mind external and unrelated to them establishes all relations between them’.12 And as Conze rightly concludes, ‘ontology as a rational discipline lost its object’.13 This larger discussion of epistemologies and ontologies is invoked here, not in an attempt to resolve systemic cultural differences between East and West or modern and primitive, but rather to focus on the radical, revolutionary nature of reformulating a worldview. As Foucault would argue,14 we are scarcely even aware of the epistemic structures that govern our thoughts and capacities for analytic reflection. How then do such epistemic transformations take place? One possible answer comes from Thomas Kuhn in his classic The Structure of Scientific Revolutions in which he puts forth the idea of a paradigm, somewhat akin to Foucault’s episteme.15 These all-encompassing paradigms are hegemonic in nature but they are also subject to revolutionary changes from fringe elements that are ‘necessarily subversive’ to a given paradigm’s ‘basic commitments’.16 Over time these radical influences lead to a scientific revolution, a shift in worldview and the eventual establishment of a new paradigm and a ‘revolutionary transformation of vision’.17 Besides the impact of fringe elements, epistemic revolutions are often precipitated by the conflict of disparate epistemes, the coming together of different cultures, or the proverbial ‘the clash of civilizations’.18 Rather than reading this as an act of violence or conquest, can we see this friction of ideas as generative potential? As an opportunity for the development of a new way of seeing the world? Indeed, the Churning of the Milky Ocean myth cycle is an exemplar of the Indian episteme. It posits a productive/cooperative model of cultural interaction; it provides a cosmic blueprint for the evolution of a dynamic Indian way of seeing and interacting with the world.

II …the purpose of myth is to provide a logical model capable of overcoming a contradiction…thus, the myth grows spiral-wise until the intellectual impulse which has produced it is exhausted.19

Seven seas and an ocean of wisdom   23

Before preceding to the narrative details of the Milky Ocean myth, it is important to reflect on how Indian stories are propagated more generally. The fact that stories are essentially oral/performative in conception, transmission and reception is readily demonstrable, but India is also a land of great literacy and textuality. Hundreds of texts and thousands of manuscripts record versions of the Milky Ocean myth. One text itself has multiple recensions and variations. It is messy and imprecise, the antithesis of scholarship’s push towards clarity and fixity. One way the modern philologist manages this textual jumble is to create a critical edition of a text (i.e. an artificially constructed text that aims to capture the common core contents of a given work’s genealogy of transmission). The notion of a critical edition is an example of how the scientific mode permeates humanistic scholarship; it trims margins, deletes anomalies and sterilises the text of inconvenient variants, all of which are essential to an Indic text’s ontology, for it is meant to be variable, unfixed and fluid. Like an ocean with countless waves, India’s stories have washed upon her shores for centuries. No wonder the Great Story (Bṛhad-kathā) comes down through the ages as the Ocean of Flowing Stories (Kathā-sarit-sāgara). The deep origins of the Churning of the Milky Ocean myth cycle can be found in the images and tropes of the Vedas, from the soma-pressing rites to Indra’s slaying of the serpent Vṛtra.20 Later puranic versions abound, each one taking on particular text-specific variations. The Viṣṇu Purāṇa and the Bhāgavata Purāṇa both contain popular tellings and belie a heavy Vaiṣṇava bias in content, circulation and reception. Śiva of course plays an important role in the story and we will return to this point later. Indeed, the Rāmāyaṇa and the Mahābhārata both recount the story, and for our purposes here, we will follow the ādi-kavi Vālmiki’s version when the great myth is told to Rāma and Lakṣmaṇa by wise Viśvāmitra during their journey to Mithilā (Bāla-kāṇḍa 44.13–27). Even choosing a singular textual source does not relieve us of dealing with textual anomalies, inconsistencies and variations. The Indian myth does not lend itself to such stale analysis. The Indian myth is an unbound narrative, and although it is ostensibly couched within a stable text, that text too is variable and manifold. Is this sense, both Indian texts and the myths they convey are liquid: their contours are dynamic, their primary contents remain largely static yet malleable. In short, we have an amorphous myth with multiple variations embedded in a profusion of texts that are themselves unfixed.21 Given this high degree of variability, there still remains a core myth with key structural elements that almost every version incorporates. We turn now to the common generative motifs embedded within the Churning of the Milky Ocean myth cycle as recorded in the critical edition of the Vālmiki Rāmāyaṇa. The cosmogenic myth begins when the gods (devas) and antigods/demons (asuras) start to ponder how they might achieve immortality. Soon they collectively resolve to obtain the elixir of life (amṛta) by churning the great ocean of milk. With Mount Mandara as a churning rod, Viṣṇu as a tortoise base

24  Srinivas G. Reddy

and the serpent lord Vāsuki as the rope, the devas and asuras work together, agitating the great waters of the ocean and unleashing a veritable stream of cosmic/earthly entities. Depending on the telling, the list and order of evolutes are variable, but the great churning of the Milky Ocean produced the celestial physician Dhanvantari, the water-nymph apsaras, the goddess Vāruṇī/Surā, the mighty steed Uccaiḥśravas, the white elephant Airavata, the Kaustubha gem, the moon, the goddess Sri, the world-destroying hālāhala poison and of course amṛta, the elixir of immortality.22 As Viśvāmitra begins to narrate the story, he says ‘you shall hear what happened in this region just as it took place. Long ago, Rama, in the Golden Age…’.23 The key phrases here are asmin deśe (in this land), tattvena (in reality) and pūrvam kṛta yuge (in the earlier Kṛta Yuga). All of these time–space descriptors intimate that we are entering the world of myth: A realm encoded with an alternate conception of time and space, albeit one that is intimately connected to our own lived reality. In fact, the physical space seems to be the same24 —we are learning what happened in this very spot of Viśālā, not in the distant past, but in a whole different time cycle (yuga). And as temporally removed as the ancient myth may be, it still invests the present with primal principles for organizing the world. One of these core principles is the disintegration of oppositional polarities, or as Levi-Strauss would put it, ‘the purpose of myth is to provide a logical model capable of overcoming a contradiction…’.25 And furthermore, that ‘mythical thought always progresses from the awareness of oppositions toward their resolution’.26 This formulation means that myths are composed of: (1) elements that oppose or contradict each other and (2) other elements that mediate or resolve such oppositions. The Churning of the Milky Ocean myth cycle neatly captures both of these essentials in a way that establishes a unique ontology and epistemology; here the gods and demons are not enemies but coconspirators in the production of health, wealth and ultimately knowledge itself. As Coomaraswamy succinctly puts it, ‘the Devas and Asuras, Angels and Titans, powers of Light and powers of darkness…although distinct and opposite in operation, are in essence consubstantial, their distinction being a matter not of essence but of orientation, revolution, or transformation…’. 27 This underlying theory of consubstantiality posits a serious ontological claim, one that positions an Indian episteme in contradistinction to the Cartesian dualism of the Western episteme.28 Here the tried dualities of dark vs. light, man vs. animal and nature vs. culture melt away into a holistic reckoning of complementary elements. J. Bruce Long’s assessment of this epistemic model is noteworthy and I quote it at length because it clearly lays out some of the productive strategies deployed by Indian mythographers: One of the most remarkable characteristics of the religious vision of the Vedic sages and Hindu mythographers is this puzzling juxtaposition of apparently contradictory ideas (i.e. multiplicity from unity, life through death, order out of conflict, etc.). Sometimes the oppositions are juxtaposed and then reversed in order to demonstrate that they

Seven seas and an ocean of wisdom   25

are nothing more than interchangeable terms of the same reality…. The sage and the mythographers seem determined both to assert the unavoidability of the sets of polarities within the realm of mystical thought and, at the same time, to press beyond the level of polar opposition into a sphere of a larger and higher meaning.29 The primary and most ostensible polarity developed in the Churning of the Milky Ocean myth is the oppositional play between the demons and the gods, the Daityas and the Ādityas. In fact, these two factions are not so dissimilar; they are in fact half-brothers, the respective sons of the sage Kāśyapa and his cowives Ditī and Aditī.30 Although the devas and asuras ‘appear to exist in a state of polar opposition, they do not… the practice of interchanging and transposing their respective roles and characters’31 is one way in which the mythic register accommodates this rich complexity. In the story, for example, when all the celestials grab Vāsuki as a churning rod, the asuras quickly take up the head as a place of honor and the devas are relegated to holding on to the serpent’s tail. Later in the story when Vāsuki emits a thick and toxic smoke from his mouth, it is the asuras who suffer and the devas who rejoice. Over and over again, polarities are flipped, switched and complicated. The notion of oppositional positionalities is constantly being undermined and destabilised. Things are not this way or that way; they are always in a state of becoming, perpetually churning about in a dynamic sea. One narrative trope used to negotiate these shifting polarities is the mediator archetype. This figure is often a trickster, or in our case, the clever god Viṣṇu who plays a critical role in the story as both mediator between the devas and asuras, as well as the central support for the whole endeavour. As the churning progresses, the great rod Mount Mandara begins to plunge into the ocean, at which point the benevolent Lord Viṣṇu takes the form of a giant tortoise (kūrma) to prop up the sinking mountain. The kūrma avatāra thus comes into existence as the foundational plinth upon which this cosmic drama plays out. In another episode that further instantiates how things are not as they seem, Lord Viṣṇu takes on the form of a beautiful enchantress named Mohini. Later in the story when the amṛta finally appears, the asuras are the first to claim it. Then Viṣṇu in his Mohinī avatāra tempts the asuras into giving up the immortal nectar, which he/she then returns to the devas. Certainly, there is an element of deception here, and the trickster archetype is befitting. And so as a mediator, Viṣṇu’s ‘function occupies a position halfway between two polar terms, he must retain something of that duality—namely an ambiguous and equivocal character’.32 By cleverly straddling the normative gender dichotomy, Viṣṇu’s man–woman incarnation acts as a narrative fulcrum between the two consubstantial groups, just as his earlier kūrma avatāra had done in a more substantive sense. Thus, it is through the repetitive (almost cyclical) establishment and breakdown of polarities that the holistic nature of the Indian episteme develops.

26  Srinivas G. Reddy

One final example of this dynamism can be seen in the inclusion of the destroyer god Śiva within this heavily Vaiṣṇava myth cycle. In most accounts of the myth cycle, Viṣṇu compassionately embodies his role as the preserver god and is praised as the world-saviour. But even in this heavily sectarian context, Śiva plays a pivotal role. While churning the ocean in search of the life-giving nectar of immortality, the gods and demons also stir up halahala, a world-destroying poison that emerges to threaten the entire cosmos. This narrative twist encodes a key ontological claim: that good and evil, life and death are complimentary forces with a common origin.33 Now, in order to save the world, Śiva, the ash-smeared god of destruction, swallows the foul hālāhala poison. He takes all the dark evil within himself and neutralises it out of compassion for the world. The destroyer becomes the preserver; life is sustained by the ostensible bringer of death. Indeed, the world is a dynamic place, and although it is relatively stable, there exists a fine balance, a precarious equilibrium between polarities that gives life its vitality. Myths like the Churning of the Milky Ocean address these issues in reiterative narrative modalities;34 indeed, they go beyond the Levi–Straussian breakdown of dyads. These myths are not so much a resolution of the inescapable dialogic of the human mind; they embrace duality through a nonterminal, cyclical progression of diversification and unification. It is the postcolonial Nehruvian call for unity in diversity, or as Long puts it, ‘united polarity and a polarised unity’.35

III …in our pursuit of what we might call ‘oceanic knowledge’, a knowledge that is neither neatly divided into sciences and humanities, nor into geography and history, nor into Vedic verses and Cartesian cogitos. Like the ocean it will cross cultures and paradigms.36 The Churning of the Milky Ocean myth offers productive metaphors for conceptualizing both the creation and circulation of knowledge. In India, water, the seas and the great ocean itself are all symbols of knowledge. And knowledge, like water, is unbound, undifferentiated and imbued with great vitality and generative potential. The ocean (samudra) is vast and shoreless (pārāvāra), filled with all the riches and gems of the world (ratnākara); it is the dynamic (arṇava) source of water/life (payonidhi), the great container (abdhi) and the ultimate metaphor for the expansive and unlimited (akūpāra) nature of knowledge. The trope of vidyā-sāgara or oceanic wisdom encapsulates three key thematics in the understanding of knowledge: (1) fluidity, (2) potentiality and (3) circulation. Fluidity, wateriness, the liquid state, they all imply notions of flux, change and expansion. Water does not conform to a shape; it is variable and flexible, it spreads outward and yet is containable. Like our planetary seas that flow into each other and crash upon the shores of continents, the ocean is a powerful metaphor for that which is and is not (simultaneously). As a symbol

Seven seas and an ocean of wisdom   27

of human cognition, this fluidity represents the flexibility and openness of Indian epistemologies and ontologies. Rather than formal epsitemes or paradigms with recognizable (and in that sense historicizable) contours and boundaries that are periodically overturned or revolutionised, the Indian epistemic model, based as it is on a fluid ontology, offers a robust cognitive framework capable of being adapted to the contours of historical change. Foucault does mention an episteme having ‘conditions of possibility’ but the Indian episteme offers something grander, it embraces the possibilities of conditions. Here it may be useful to draw on A.K. Ramanujan’s famous paper, ‘Is There an Indian Way of Thinking?’ The somewhat controversial query was critiqued for its essentializing tone; perhaps it was posed only as a rhetorical device, but through some clever argumentation based on linguistic theory, Ramanujan does put forth an answer. His claim is that Indian modes of being and knowing (i.e. Indian ontologies and epistemologies) are primarily context-sensitive, in contrast to the heavily context-free modality of Western epistemes.37 The Indian model of being/knowing is flexible and dynamic, it does not posit universals but rather patterns: forms that are adaptable to an ever-evolving array of cultural patterns. The myths of India perfectly encapsulate this context-sensitive functionality. Let us return to the scene in Vālmiki’s Rāmāyaṇa when Rāma and Lakṣmaṇa hear the tale of the Milky Ocean. Upon seeing the present city of Viśālā, Viśvāmitra recounts what happened here in the distant past. But even this tangible physical reality is something transportable. Viśālā can exist anywhere, as can Ayodhyā, the Gaṅgā and other geomythical spaces. They come into existence whenever and wherever these sites are invoked by mythographers and experienced by listeners. There is but one Kāśi on the banks of the holy Gaṅgā, but this model is replicated in other parts of the subcontinent, in Dakṣiṇa-Kāśi, Uttara-Kāśi and countless others. I remember hearing of three distinct Chitrakūṭas in Andhra alone, each one claiming to be the real spot where Sītā had bathed. Indeed, the myths of India are atemporal and ageographical. Mount Meru is not a physical location in time or space; it is a transposable axis, a cosmic blueprint capable of being invoked and realised anywhere on earth. This context-sensitive way of thinking is the heart of the Indian episteme for it allows us to embrace both the universal and the particular, seamlessly and without conflict. A context-sensitive society requires a mythic paradigm that must naturally be generative and full of potential. Here again the ocean provides a fitting metaphor, for it is the great source of life, death, health and wealth. From the abundant ocean comes, amṛta, viśama, āyurveda and śrī. The ocean represents both a variegated array of resources as well as dissolution (pralaya) into an undifferentiated potentiality. Even in precosmogonic times, before the cosmos had assumed its present form, before time itself, all things existed in a state of perfect unity and absolute stasis,38 when the whole of existence was unmanifest water (apraketaṃ salilaṃ sarvam ā idam) and there was nothing but the waters of unfathomable depths (ambhaḥ

28  Srinivas G. Reddy

gahanaṃ gabhīram).39 The ocean thus represents the dynamic energy of the universe as simultaneously potential and kinetic. Hidden within these deep waters is knowledge itself,40 but it is latent only. It does not simply rise to the surface; but like butter from milk must be churned from within. The notion of churning, of circulation, of flow and of friction is the third and most critical trope that the Milky Ocean provides to us. In order to reap the vast riches of the ocean, it needs to be stirred up, for conflict becomes the precursor to production.41 And so cosmic rivals, the devas and asuras, resolve to enter into and alliance to churn the great waters. Here is how Viśvāmitra describes the scene in Vālmiki’s telling: Long ago, Rama, in the Golden Age, the sons of Diti were very powerful. Now the sons of Aditi were also mighty; but in addition, they were illustrious and extremely righteous. Once, best of men, a thought occurred to all of those great beings, ‘How may we become immortal—free from old age and sickness?’ As those wise beings were pondering this, an idea occurred to them, ‘If we should churn the ocean of milk, we would obtain from it the elixir of life’. Their might knew no bounds. So once they had resolved to undertake such a churning, they took the great serpent Vasuki for their rope and Mount Mandara as their churn and began to churn.42 Some key points become clear from the passage above. First, the Daityas and Ādityas are half-brothers, both children of the ancient sage Kāśyapa, but they are also first cousins, sons of the sisters Ditī and Aditī. These two daughters of the mythic progenitor Dakṣa represent, along with their respective offspring, the complimentary polarities of division/splitting (diti) and unlimitedness/freedom (aditi). Out of their joint desire to achieve immortality, the devas and asuras, who have unlimited creative vitality (amita ojasaḥ), resolve to join forces and churn the Milky Ocean in search of amṛta. Both sides are powerful, they are equally matched but different: The Daityas are superstrong (mahā-balāḥ) while the Ādityas are valorous (vīryavantaḥ) and truly righteous (su-dhārmikāḥ). This quality of being separate but united, different but the same, is essential because it is what creates generative tension. This friction of life is the catalyst that produces all the good, bad, ugly and beautiful things in the world. Thus, the basic mythic structure is simple: Covalent, consubstantial polarities join forces, activate the generative power of circulation and go on to produce multiple polyvalent entities, not least of which is knowledge itself. In the critical edition of the Rāmāyaṇa, the celestial physician Dhavantari is first to appear; he arises from the depths bearing the elixir of immortality, but also the science of life, the knowledge of āyurveda. From a structuralist perspective, the universal design of myths or ‘morphos’ is not about narrative details, but rather the network of formal relationships built up among the constituent parts of a myth.43 This network of

Seven seas and an ocean of wisdom   29

relations offers multiple readings and so even within the Indian tradition, the Churning of the Milky Ocean myth cycle is interpreted in a variety of ways. One recurrent thematic is the soteriological nature of the great churning. The quest for immortality represents a journey towards transcendence, and a return to the eternal oneness of the undifferentiated cosmic waters. As Wilson remarks: The account [of the Churning of the Milky Ocean] in the Hari Vamsha is brief and obscure, and is explained by the commentator as an allegory, in which the churning of the ocean typifies ascetic penance, and the ambrosia is final liberation.44 According to a more recent interpretation of the myth, the seven seas mentioned at the beginning of this article represent the constituent fluids of the body: salt water (urine), sugarcane juice (perspiration), wine (the senses), clarified butter (semen), curds (mucus), milk (saliva) and fresh water (tears).45 This kind of internal anthropomorphizing of natural phenomena is common in the Indian tradition; there is a sense that there is great swirling ocean within each of us. Thus, the power of yoga, represented by the rising up of kuṇḍalinī energy in the body, can also be mapped onto the mythic template: The ocean is the body, the base of Mount Mandara is the seat of yogic power at the root chakra (mūlādhāra), the churning rod is the spinal cord, the churning strings are the two nerve channels of iḍā and piṅgala and the churners are the air elements of vital forces on either side.46 Suffice it to say that grand mythic cycles like the Churning of the Milky Ocean provide rich epistemic frameworks capable of adapting to a variety of scenarios, agendas and perspectives. The question now is: What can a robust analysis stemming from an Indian episteme mean for our current understanding of the Indian Ocean in particular, and thalassology or ocean studies as an academic discipline more broadly?

IV Empty your mind. Be formless, shapeless, like water. Now you put water into a cup, it becomes the cup. You put water into a bottle, it becomes the bottle. You put it in a teapot, it becomes the teapot. Now water can flow, or it can crash. Be water my friend.47 (Bruce Lee) One of the most fundamental themes underlying the present discussion is the dyadic reckoning of nature and culture, the domain of the material world set in contrast to the life of man. Extending the parallels a bit further nature is the solid, stable land while culture is the liquid, dynamic ocean; nature is the world as it is, it is about an ontology; culture, on the other hand, is how we see the world, culture is an epistemology. But nature and culture are inseparable, just as ontologies and epistemologies are inseparable. Things

30  Srinivas G. Reddy

are the way they are because of how we perceive them, and how we perceive things defines the ways things are. We can never escape the subjectivity of our own ontologies and epistemologies; ‘epistemicide’ is untenable.48 Even the world of science has posited this interconnectedness in its own way with notions like the anthropocene, the current geological age of the planet defined specifically by the (deleterious) influence of humanity on the environment. Sadly, and ironically, the violence of our industrialised modernity and its effects upon the planet have only highlighted the ever-present symbiosis of nature and culture. This chapter grew out a conference entitled ‘Knowledge Traditions of the Indian Ocean World’, and part of the inspiration was to revisit how we conceive of the Indian Ocean as an area of study. Early on scholars like Michael Pearson could delineate an area of considerable geocultural unity and continuity, integrated not only by a physical seascape but a vibrant circulation of people, products and ideas.49 Pearson and others like Chaudhuri and McPherson would go on to view the Indian Ocean not as a unitary cultural area, but rather multiple cultural zones connected by maritime (trade) networks that exhibited certain cultural commonalities. They were wrestling with the tried/ tired nature/culture dyad, much as we still are today. Markus P.M. Vink’s detailed survey of Indian Ocean studies as a discipline looks at the field’s ‘rich historiography’, from ‘its humble beginnings in the 1950s and 1960s, and the cross-fertilization between the “Annales” school and world-systems analysis in the 1980s, to its—admittedly incomplete—institutionalization today’.50 That incompleteness speaks to our inability to fully accommodate a holistic nature–culture discourse in academia. This unfinished project of integration is clearly reflected in the way the Indian Ocean has been understood, particularly in regard to its status as heritage. In November of 2011 the French Ministry of Cultural Affairs organised The First Conference on Indian Ocean Heritage entitled ‘Indian Ocean Heritage, an Emerging Concept’. To conceive of an ocean as heritage was a new idea, but how was a global organization like UNESCO, the United Nations Educational, Scientific and Cultural Organization, categorizing the Indian Ocean as world heritage? According to UNESCO’s 1972 Convention Concerning the Protection of the World Cultural and Natural Heritage, heritage was to be bifurcated into two distinct categories: cultural and natural. Article 1 defines cultural heritage as buildings, monuments, sculptures, archaeological ruins, inscriptions, etc., while Article 2 defines natural heritage as geophysical and biological formations. All of these entities must be of ‘outstanding universal value from the historical, aesthetic, ethnological or anthropological point of view’.51 Clearly this was a staunchly dualistic approach to the idea of heritage. And though it may seem like UNESCO is just a strawman in this discourse, it only goes to show how deeply dialogic our conception of the ocean truly is. We are bound to see it in a dualistic way, as nature or culture. But hadn’t this forced dichotomization been recognised by Western philosophers? Did non-Western thinkers have anything to say about the matter?

Seven seas and an ocean of wisdom   31

Returning to the pioneering work of A.K. Ramanujan, a scholar who straddled (in both theory and practice) the proverbial East–West divide, we read: neither…does the Levi-Straussian opposition of nature-culture make sense; we see that the opposition itself is culture-bound. There is an alternative to a culture vs. nature view…culture is enclosed in nature, nature is reworked in culture, so that we cannot tell the difference. We have a nature-culture continuum that cancels the terms, confuses them even if we begin with them.52 It is the last statement that is the crux of the matter, the catch-22 of modern scholarly analysis. From the West, contemporary philosophers like Bruno Latour have opined that the conventional object/subject, nature/culture and reason/belief binaries are simply untenable as analytical tools in a productive understanding of knowledge. When Latour claims that ‘we have never been modern’, he is insisting that we humans falsely believe that we have transcended nature. And though he does not call for a return to traditional/ primitive premodern epistemologies, his imaginary Middle Kingdom tries to find space for a new discourse, one in which ‘all of culture and all of nature get churned up again every day’.53 This new space, this ‘new thalassology’54 must in many senses be aquatic, or in the words of Bruce Lee, it must become shapeless and formless like water. Anthropologist Stefan Helmreich argues that ‘water oscillates between natural and cultural substance, its putative materiality masking the fact that its fluidity is a rhetorical effect of how we think about “nature” and “culture” in the first place’.55 And furthermore that ‘water functions not so much as theory machine but as an other to theory: as description’.56 And here we come full circle. We don’t require more theories or fancier equations to model the cosmos; we don’t even necessarily need to reformulate epistemologies. We need a return to description, a shift back towards more ontological concerns, to think again deeply about the way things really are. But again, all ontological concerns are intimately related to one’s epistemic positionality. The suggestion here is that we simply dip into certain ontological assertions that have fallen under the surface so to speak; that we drink from the deep truths that still live in premodern societies, and listen to their stories, their myths.57 As Vink and Andre Gunder Frank have argued, the Indian Ocean as a distinct geocultural zone of interaction had a long and continuous history of some 6,000 years, a mere 200 of which were dominated by European powers.58 Unfortunately those two centuries have left us with a deep and persistent Eurocentric conceptualization of the Indian Ocean. If we recontextualise and reimagine the Indian Ocean in its long history, then we must address the gross imbalance of perspectives that have informed the bulk of modern scholarship on the subject. In other words, we need local perspectives and indigenous voices: an Indian episteme for the Indian Ocean.

32  Srinivas G. Reddy

A listener of the Churning of the Milky Ocean myth is never presented with a ‘resolution of dyads’,59 that mythic (for lack of a better word) episteme churns out a world view that is comfortable with holding contradiction in way that eliminates oppositional polarities from the get-go. As Long suggests, the ‘Indian Mind’ is somewhere ‘betwixt and between’ the two prototypical mentalities of the mythological mind and the scientific mind.60 But are these two minds so different really? Is this not the real myth of modernity? To assume that science is beyond reproach, to be convinced that science is rational truth and to trust that it is not dogmatic belief is the true indicator of modernity. Our faith in science blinds us to the meanings of myth. For many a modern subject, myths are mere stories, a form of entertainment and at best, an artifact of cultural heritage. But Kuhn, arguably a very scientific thinker, had something insightful to say about this: If these out-of-date beliefs are to be called myths, then myths can be produced by the same sorts of methods and held for the same sorts of reasons that now lead to scientific knowledge. If, on the other hand, they are to be called science, then science has included bodies of belief quite incompatible with the ones we hold today.61 It is this incompatibility that needs to be transcended, and the tools for such a transformation lie in the watery epistemologies of premodern cultures. But as Kuhn would argue, these ‘revolutions’ of thought are not placid affairs; they are ‘the tradition-shattering complements to the tradition-bound activity of normal science’. These extraordinary episodes witness a shift in ‘professional commitments’.62 For the professional academician, various ‘turns’ in the humanities and social science have certainly yielded new scholarship and fresh perspectives. The great ‘cultural turn’ from the 1980s onward pushed the scholarly pendulum towards meaning and affect rather than empiricism and positivism, but such theory-heavy research essentially re-entrenched the post-Enlightenment epistemic order. In other words, scholarship may have developed epistemologically, but not so much ontologically. It is no easy endeavour to abandon one’s ontological assumptions, but scholars like Eduardo Vivieros de Castro have made a spirited call for/towards indigenous histories and cosmological perspectivisms.63 This is an important and worthy course of inquiry, but before we rush too quickly towards (over) privileging native epistemologies, another vein of scholars assert that ‘the argument for indigenous critical traditions can be attacked as the newest species of liberal pluralism…of reading critical theory into gnosis itself… which effectively silences the very people—their voices—whose condition it purports to demystify’.64 How can we avoid the pitfall of propagating yet another form of appropriation and misrepresentation? Can we, as de Castro suggests, engage in a new discourse, ‘not to condemn it as colonialist, exorcise its exoticism, or landmine its intellectual field, but rather to turn it into something else?’65

Seven seas and an ocean of wisdom   33

This ‘something else’ I argue will have to be like the ocean: vast and varied in scope, constantly in flux and ultimately untenable. The myth of the Milky Ocean and how it was churned by the gods and demons to produce the wealth of knowledge is but one paradigm that allows us to see the seas, the continents, our planet and ultimately our universe as an unfathomable ocean of being. In our current episteme this kind of ‘being’ is epistemologically unavailable to us; it is scientifically unverifiable and thereby rendered obsolete. And yet ancient traditions everywhere have spoken of this kind of ontophenomenological unity. It is found in the depths of personal experience. The Bhagavad Gītā contains a beautiful verse that summarises this way of being with a water metaphor. Not the desirer of desires, but one who lets desires enter from all sides obtains peace, for she remains steady and full, like rivers flowing to the sea.66 What would Indian Ocean studies look like if we approached knowledge from such a free-flowing perspective? In this episteme, knowledge is organic, always filling up but never overflowing. The knower is not desirous of knowledge; it simply flows into her and through her, like waves upon the ocean.

Notes 1 H.H. Wilson, The Viṣṇu Purana: A System of Hindu Mythology and Tradition, London: Trubner & Co., 1864, Book II, Chapter 2, https://www.wisdomlib.org/ hinduism/book/vishnu-purana-wilson/d/doc115961.html (accessed on 3 July 2019). 2 Some traditions reckon 18 dvīpas but 7 appears to be the usual number. 3 Sri Aurobindo, The Secret of the Veda, Pondicherry: Sri Aurobindo Ashram Trust, 1998, p. 102. 4 Ibid. 5 Michel Foucault, The Order of Things: An Archaeology of the Human Sciences, London: Routledge, 1960, pp. xxiii–iv. 6 For a deeper discussion of science, scientism and non-Western epistemologies, see B. Allan Wallace, The Taboo of Subjectivity: Towards a New Science of Consciousness, New York: Oxford University Press, 2000, in particular Parts I and III. 7 The six most common prāmāṇas as discussed by various Hindu, Buddhist and Jain philosophers are: insight (pratyakṣa), inference (anumāna), analogy (upamāna), circumstantial implication (arthāpatti), negative proof (anupalabhdi) and verbal testimony (śabda). 8 Edward Conze, Buddhist Thought in India, Ann Arbor: University of Michigan Press, 1967, p. 17. 9 Here I am drawing on Bruno Latour, We Have Never Been Modern, Cambridge: Harvard University Press, 1993. More discussion on the idea of a false modernity follows later in the chapter. 10 Michel Foucault, Power/Knowledge: Selected Interviews and Other Writings 1972–1977, New York: Pantheon Books, 1980, p. 197. 11 Conze, Buddhist Thought in India, p. 25. 12 Ibid.

34  Srinivas G. Reddy

3 Ibid. 1 14 Take, for example, that What modern thought is to throw fundamentally into question is the relation of meaning with the form of truth and the form of being: in the firmament of our reflection there reigns a discourse—a perhaps inaccessible discourse—which would at the same time be an ontology and a semantics. Structuralism is not a new method; it is the awakened and troubled consciousness of modern thought. (Foucault, The Order of Things, p. 226). 15 Various scholars from both the sciences and humanities have compared and contrasted the Foucauldian episteme and the Kuhnian paradigm. According to Foucault’s early definitions, any given society in time is dominated by a singular, albeit multidisciplinary, episteme that is ‘silently invested in practice’ and experientially ‘invisible’ to individuals. For Kuhn, multiple paradigms can function simultaneously but his analysis is restricted to the domain of the sciences in which scientists are aware of their paradigmatic constraints. For a more sustained comparison of these two perspectives, see Jean Piaget, Structuralism, New York: Basic Books, 1970, pp. 128–135. 16 Thomas S. Kuhn, The Structure of Scientific Revolutions, Chicago: University of Chicago, 1970, p. 6. 17 Ibid., p. 112. 18 See Samuel P. Huntington, The Clash of Civilizations and the Remaking of World Order, New York: Touchstone, 1997. The phrase was used a half-century earlier by Albert Camus. 19 Claude Lévi-Strauss, Structural Anthropology, New York: Basic Book, 1963, p. 229. 20 See Ananda K. Coomaraswamy, Angel and Titan: An Essay in Vedic Ontology, Journal of the American Oriental Society, 55(4), Dec. 1935: 373–419. 21 Nothing can be said now of performance traditions, nor of the circulation of myths and texts, for that is beyond our present discussion. 22 There is some variety in the order and number of articles produced from the ocean. As I have observed elsewhere…the popular enumeration is fourteen; but the Rāmāyaṇa specifies but nine; the Mahābhārata, nine; the Bhagavata, ten; the Padma, nine; the Vayu, twelve; the Matsya, perhaps, gives the whole number. (Wilson, Viṣṇu Purana, http://www.sacred-texts.com/hin/vp/vp044.htm (accessed on 4 July 2019)). 23 Robert P. Goldman, The Rāmāyaṇa of Vālmīki: Volume 1 Bāla-kāṇḍa, Princeton: Princeton University Press, 1990, pp. 209–210. 24 We will return to the atemporal and ageographical nature of Indian mythic structures. 25 Lévi-Strauss, Structural Anthropology, p. 229. 26 Ibid., p. 224. 27 Coomaraswamy, Angel and Titan, pp. 373–374. 28 The two polar terms Lévi-Strauss begins with are “life” and “death.” But one must observe at once that in primitive culture “life” and “death”—which the secular mind conceives as irreconcilable—may not be perceived as polar at all, but rather as aspects of a single condition, the condition of existence. (Stanley Diamond, In Search of the Primitive: A Critique of Civilization, New Brunswick: Transaction Publishers, 1974, p. 308). 29 J. Bruce Long, Life Out of Death: A Structural Analysis of the Myth of the ‘Churning of the Milky Ocean’, in Bardwell Smith (ed.), Hinduism: New Essays in the History of Religions, Leiden: E. J. Brill, 1976, p. 173. 30 The Indian etymologists derive asura from the story of Surā, the goddess of wine, who emerges from the churning of the ocean. When she manifests the

Seven seas and an ocean of wisdom   35

devas (sura) accept her while the Daityas (asura, i.e. without wine) do not. This is of course a constructed derivation since asura is cognate with the Persian ‘god’ Ahura Mazda, thus adding another layer of reversed polarities to the equation. 31 Long, Hinduism, p. 204. 32 Lévi-Strauss, Structural Anthropology, pg. 226. 33 AM II.67: Flowers flourish in the severe summer heat and decorate the blackness of women’s hair, buds split braids like the parched cracked earth while trumpet flowers appear like a rain of honey. But how can this be? Well just look and you’ll see! Ambrosia and poison are from the same place.

34 35 36 37 38 39 40

41 42 43 44

45 46

As Long comments: ‘it is this conflict between opposed but interrelated forces that, at one and the same time, established the world in existence and threatens it with extinction’. Long, Hinduism, p. 172. For more on reiterative narrative modalities in India see A.K. Ramanujan, Repetition in the Mahābhārata, in Arvind Sharma (ed.), Essays on the Mahābhārata, Leiden: E. J. Brill, 1991, pp. 419–443. Long, Hinduism, p. 192. Devleena Ghosh and Stephen Muecke, Natural Logics of the Indian Ocean, Cultural Studies Review, 12(1), March 2006: 120. A.K. Ramanujan, Is There an Indian Way of Thinking? An Informal Essay, Contributions to Indian Sociology, 23(1), January 1989: 47–48 and 54–55. Long, Hinduism, pg 171. Nāsadīya sūkta, Ṛg Veda X.129. See multiple variations on this theme (e.g. the demon Somaka and the horseheaded Hayagrīva both steal the Vedas and hide them in the watery depths of the ocean). Viṣṇu recovers the sacred texts in both stories. See Wendy Doniger O’Flaherty, The Origins of Evil in Hindu Mythology, Delhi: Motilal Banarsidass, 1976, p. 101. cf. AM IV.17: ‘Taking the form of a giant fish, You ripped the Vedas from the monster Somaka, dreaded demon of the gods!’ Indeed the Churning of the Milky Ocean can be read as a cosmogonic myth, cf. Jeannine Auboyer in Long, Hinduism, p. 184, note 23. Goldman, The Rāmāyaṇa, p. 210, verses I.44.14–17. See Claude Lévi-Strauss, The Savage Mind, London: Weidenfeld and Nicolson, 1966. cf. Long, Hinduism, p. 174. It is interesting to note that Wilson, writing in the mid-nineteenth century, and coming from a very different epistemic foundation, finds this Indian interpretation to be ‘mere mystification’. http://www.sacred-texts.com/hin/vp/vp044.htm (accessed on 9 July 2019). http://bhagavata.org/canto5/chapter20.html (accessed on 9 July 2019). The metaphor is further extended upon: This total system is held upright by Viṣṇu…when kundalini power is incited the first results are always negative like, hālāha, ‘the lethal poison…’ and all all-ruinous ills befall on us. If this negativity is surrendered to someone who can handle it effortlessly, like Śiva, then the positivity results like Amrita/ Ambrosia. Even if this elixir of life is readily available, there will be still better lures, lusts, and longings like Mohiṇī and if enchanted by those instantaneous quirks, which are yet another kind of negativity, we will be lost completely, like the demons. Hence ‘churn the correct way…’ is the postulate of yoga. (http://www.valmikiramayan.net/bala/sarga45/bala_45_frame.htm (accessed on 9 July 2019)).

47 Bruce Lee, Pierre Berton Show, Bruce Lee: Mandarin Superstar, Hong Kong, December 9, 1971. Bruce Lee, Be Water My Friend, https://youtu.be/y5WJOtw2fDs.

36  Srinivas G. Reddy





­







3 GENDERED SPATIALISATION OF THE OCEAN IN THE KATHAˉSARITSAˉGARA Tara Sheemar Malhan

Text and context The chapter is an attempt to locate the depiction of the sea as a gendered space in the Kathāsaritsāgara (KSS), a Sanskrit text from eleventh century CE Kashmir, a vast compilation of timeless narratives. We see the manner in which fantasy, fact and fiction merge in the stories to create an impression of the sea, seafarers and of peoples across the sea. The literary gaze on the sāgara (ocean) reveals that the mentalité of the literary culture was not as landlocked as the emphasis on the graded hierarchy of the urban, rural and wild would have us think. In the spatialisation of the KSS the forest is the ‘other’ of the ideal brahmanical settlement, the city (nagara).1 The sea is envisioned as being beyond the forest; it is further on and far, but not inaccessible to the resolute of spirit. Determination is certainly required for crossing the villages and settlements and forests to reach the shore. Further, crossing the sea, battling its dangers and reaching another shore and landscape and returning home is a matter of great heroism and an empowering trope for the struggling protagonist. The fundamental landscaping of the text conceives of the essential spaces of the ocean and forests. Upon the basic natural spaces, like forests and oceans, human action and power structures created what Lefevbre termed as the ‘absolute space,’ as the space produced in pre-modern societies.2 Bordering the sea are the coasts and coastal spaces including the villages, cities and forests. Spaces across the sea include cities and ‘islands’ and lands in general, which are linked via the sea routes to the shore localities of the ‘mainland’. In this chapter I will explore two main questions: Who is depicted as crossing the sea and with what intent and in what manner are the areas, societies and individuals beyond the sea, or connected with it, portrayed in

38  Tara Sheemar Malhan

the text?3 These questions are investigated from the dominant perspective present in the text, the gaze, which is of those who traverse the seas, and they have been termed as ‘mainlanders’ in the chapter. In this process the sea is constructed as a textual space, and literary space, and individuals, objects and societies are located in that space; and these I term as the ‘islanders’, as the permanent inhabitants of the oceanic space, noticeably not as a homogenous ensemble. This is a gendered reading since gender is a running concern in the analysis throughout the chapter. The use of these terms is qualified since the concept of the sea or sāgara or samudra was a varied one in literature. Any large body of water could be termed as a sea. One could be going from a point on the coast to another and that could also be described as crossing a sea. The Puranic conception of the sāgara included the island—the entire land being in effect dvīpas (islands) bordered by the sea, there being seven overlapping islands, and from that perspectivethe ‘mainland’ would also be a dvīpa.4 Nonetheless, for purpose of analysis, the original area of the protagonists has been termed as the ‘mainland’. The areas that they seek are the ‘areas beyond the sea’, or the ‘islands’. The islands mentioned in the narratives include Simhala, Malayapura, Kataha, Suvarnadvipa/Suvarnabhumi, Hamsadvipa, Muktipura, Narikela, Karpura, Citrakuta, Tarapura, Ratnakuta, Karpurasambhava, Karkotaka, Kanakapuri and Utasthala. Some of these have been identified as islands or regions located in the South East Asian seas.5 The text reflects a sense of geography. For example, one island is mentioned as being on the western sea and approachable by Shashankapura, while most are on the eastern sea and approachable by shore cities, the main being Tamralipta, and also Jalapura and Vitankapura.6 In one case the eastern city connecting to Suvarnadvipa is not named but the unnamed city is placed in the area called Paundra, corresponding to parts of Bengal–Bihar.7 The western direction stories are few and refer to the island called Hamsadvipa, supposedly in the southwest.8 There are only a few cities located on the islands themselves, namely Malayapura, Muktipura, Tarapura, Kanakapuri and Kalashapura. Thus, a clear distinction is made between the eastern and the western shores and the areas they connect with while the eastern cities and islands are more frequent locations of stories. The depiction of the sea follows a distinct pattern in the narratives of the KSS. The template includes the shore, the sea itself and islands located in the middle of the sea. The notion that the earth contains many seas is present in the text, as seen in the instance when while riding a splendid air chariot Naravahanadatta, the princely protagonist is said to have looked down from heaven and saw the earth of the size of a mound and the seas small as ditches.9 The ocean occurs in numerous tales throughout the text and is used in various instances to explain the additional quality of vastness in any description of beauty, emotion, area, bravery, etc. For example, when the army of Vikramaditya assembled it is said to resemble ‘a broad sea’; king Simhabala is described as a sea of courage; another king is mentioned as ruling the earth with supreme authority ‘as far as the sea’.10 Sea as

Gendered spatialisation of the ocean  39

an analogy of vastness in beauty, is frequently mentioned—for example, a dancing girl of the name of Sundari is described as dancing as a wave of the sea of beauty tossed up by the wind of youth.11 The sea is essentially constructed as a passage, a doorway, a temporary dangerous space full of monsters and sudden storms, traversed by a variety of individuals to create interconnections and transferences. There are prominent recurring motifs in the depiction of the ocean that emphasize the following themes: role of monarchy in maritime contacts, flow of women and wealth from islands to mainland, role of merchants and their sea craft, depiction of the fishing community as interconnected with the sea and the shore, rituals and customs of ‘islanders’, transfer of ideas and rituals, masculine fantasy in the depiction of superhuman women and the depiction of the sea as beneficent as well as maleficent. The matter of this transference is essentially wealth, women and ideas. All these form significant tropes in the depiction of the sea as a space in the KSS.

Shaktideva, Kanakapuri and the Vidyadharis The story of a brahmana protagonist named Shaktideva and his search for the ‘Golden City’ or Kanakapuri, given below in some detail, sets down the ‘mainland-island’ interactions intricately before the audience, demonstrating all the aspects mentioned above.12 We can observe most of the themes in this lengthy winding narrative that recur in the KSS while depicting the oceanic space: • • • • • • • • •

kings and members of the ruling lineages, of the ‘mainland’ as well as the ‘islands’, are prominent in the stories; the sea is projected to be markedly far away, approached after crossing many landscapes and overcoming various hardships; the fear of the sea monsters and storms is a prominent trope used to highlight the dangers of the sea voyages; the forest and the hermitage appear as bridges connecting the spaces of cities, villages, coasts, ports, and islands; the merchant community is the visible connect between ‘mainland and ‘islands’ via the coastal spaces and their ship craft, transferring wealth from the latter to the former; the sea faring community of fishermen is given a crucial role in the islands, and in shipping and fishing activities; monasteries, temples, and brahmans are strategic ideologue transplants from the ‘mainland’; there is depiction of the alternate customs and rituals of ‘islanders’; the representation of empowered superhuman women, vidyādharīs, as symbolic of wealth, femininity and power is noticeable.

Fantasy functions as a remarkable narrative device— in depiction of monsters and other creatures like giant birds, giant whirlpools and trees, magical

40  Tara Sheemar Malhan

gateways, undersea or hidden cities, super-human women, etc— that transforms the socioeconomic and political significations into a symbolic and phantasmal delight for the audience. The story is as follows:13

The princess and the kaivartapati A brahman named Shaktideva decided to search for Kanakapuri because the princess Kanakarekha of Vardhmana had announced that she would marry the man who had seen the city. He went on a long search, crossed the Vindhya forest and reached an ashrama/hermitage. There he met a hermit who instructed him to contact the rich king of the nishadas on the island called Utsthala. According to the hermit the nishada king roamed between islands and would have information about the distant foreign city. Shaktideva set out and reached the city called Vitankapura on the shore, where he made friends with a merchant (vaṇijam) who took him onboard his ship (yānapātram) to Utsthala. On the way a fierce hurricane with great waves broke the ship. The merchant clung to a plank and was rescued by another ship. Shaktideva was swallowed by a big fish and captured by the servants (bhṛtya) of king Satyavrata, the king of nishadas, the same king he was trying to locate. The king interrogated him and upon finding out about his quest decided to help him.

The monastery (matha) and the temple (devakula) · For taking rest, Shaktideva was sent to a monastery of brahmanas (vipramaṭha) located on that island. He was supplied food by Vishnudatta, a brahmana resident of the maṭha. After conversing they discovered that they were cousins. Vishnudatta told him about the stream of merchants (vaṇika) and pilots (karṇadhāra) that came to Utasthala from other islands that might be able to assist him in his search. The next day, when the king (dāśendra) came to meet him, he spoke about an island in the middle of the sea called Ratnakuta as the location of a temple of god Hari, said to be created by the ocean. The temple organised a festival on a specific day along with a procession (yātrā) and people attended it diligently from all the islands (sarvadvīpa) to offer worship (puja). His view was that someone there might have heard of the ‘Golden City’.

The whirlpool and the city of vidyadharas Shaktideva took provisions furnished by Vishnudatta and went along with king Satyavrata in a ship arranged and steered by the king himself. They observed a massive banyan tree in the middle of the ocean, ‘home of marvels in which the monsters resembled islands’ under which was a whirlpool that began consuming the ship. Shaktideva escaped by catching a branch of the tree upon the advice of the king. The king, however, chose to go down with the ship since he considered the failure of the mission as his responsibility.

Gendered spatialisation of the ocean  41

Giant birds nested on the tree and Shaktideva overheard them speaking in human tongue of the landscapes they saw. He cleverly hid in the feathers of one of the birds that spoke of the ‘Golden city’. He was carried by the bird to Kanakapuri glittering with walls of gold and pillars of precious gems; it was a city of vidyādharas.

Vidyaˉdharˉı princesses and the magical doorway Shaktideva met the vidyādharī princess named Candraprabha in the city and she welcomed him as a guest and offered lavish hospitality. Later she briefed him about her family history and told him that she had three sisters—Candrarekha, Shashirekha and Shashiprabha—who had been cursed by a hermit to be born as mortals. The curse would end when they remembered their former existence. Candraprabha had been living in Kanakapuri waiting for the curse to end. She chose Shaktideva as her husband and went to ask permission from her father, the king of vidyādharas. Meanwhile, Shaktideva was entertained as a guest of the princess, but when he went sightseeing he was magically transported back to his city, Vardhmana, after plunging into a lake. There he told the princess Kanakarekha the way to the Golden City. When she heard his story she vanished, as she was in fact the vidyādharī Candrarekha, one of the three cursed sisters. Her curse was destined to end when a man would discover the way to the city. Subsequently, Shaktideva went a second time to the Golden City, this time in pursuit of the vidyādharī.

The repeat voyage, wealth and human sacrifice Once again, our protagonist reached Vitankapura on the shore of the sea to reunite with the merchant Samudradatta who had survived by floating on a plank for three days. He had been rescued by the ship of his own father who was on the return journey after making vast profit, that filled his ship with gold, from a distant land. Samudradatta’s father forbade him not to undertake any further risks (involved in sailing) and they settled down in the shore town. This time the merchant arranged for some of his agents to take Shaktideva to the island of Utasthala. On the way the sons of the deceased king Satyavrata saw him by chance and inquired about their father. They were angry and suspicious upon being told the events regarding the whirlpool and drowning of the ship because Shaktideva was himself alive, but their father had died. They bound him and took him to the temple of Candika to be sacrificed where the only choice the hero had was to pray to the goddess to save him.

The daˉs´endrakanyaˉ and food practices The morning after Shaktideva had been imprisoned, the beautiful daughter of the late king of fishermen emerged out of the inner cell of the temple.

42  Tara Sheemar Malhan

She was the sister of his captives,Vindumati, who had fallen in love with him at first sight, and planned to save him by publicly claiming him as a husband before he could be sacrificed. He agreed to marry her, and any complications connected with her social status are resolved in the story by stating that she did not belong to the fishing community but was in fact a celestial maiden (divya strī) degraded due to a curse. A observable incident connected with his stay on the island is when he glimpsed a candala coming along with a load of cow’s flesh and expressed shock to his wife inquiring that how was it possible that the cow, which was worshipped in all three worlds, be eaten? Vindumati agreed that the wickedness of that act was inconceivable but did not offer any real insight. She told him that in her former birth she had been a vidyādharī who had bit some strings made of dry sinews of cows with her teeth and fastened them to lyres. Due to the degradation she had been born in the family of fishermen (dāsakula), even for that small offence, implying that the punishment would be greater if the cow’s flesh was eaten.

Magico-ritual practices The story proceeds further and a daitya, in the form of a boar abducted a princess named Vindurekha, daughter of king Candavikrama, mentioned as lord of the southern region. Shaktideva saved her, they got married and she was soon enceinte. During an earlier discussion princess Vindumati had predicted that on that very island he would soon marry another wife who would become pregnant. In the eighth month he was to cut her open and take out the child and feel no moral doubt about it. Horrified, he had asked Vindumati the objective of the act, but she had been vague and had said it was for a certain reason. The eighth month of Vindurekha’s pregnancy arrived, but Shaktideva was still hesitant to cut her open. Finally, princess Vindurekha herself told him not to hesitate as she was also a cursed vidyādharī. In addition, to support the cause, she narrated a story in support of embryo and female sacrifice. It was about a hero who had obtained sovereignty over vidyādharas by sacrificing the child conceived by him with a yakshi (also a cursed vidyādharī).

Conferring of sovereignty Thus, Shaktideva finally agreed and cut open Vindurekha who disappeared as as soon as he did so; the child became a sword as he held it by the throat and he himself quickly became a vidyādhara. Vindumati then informed him that she and Vindurekha were two of the three cursed vidyādhara sisters from Kanakapuri mentioned earlier in the story and took him through the air to Kanakapuri with the help of his magic sword to unite with the other three princesses. She returned to her vidyādharī body and he became a king of the vidyādharas. Thus, he obtained four wives and sovereignty over the Golden City: Candraprabha (the princess who had not been cursed);

Gendered spatialisation of the ocean  43

Candrarekha (Kanakarekha, the princess who set him in search of Kanakapuri); Shashirekha (Vindurekha, daughter of the lord of southern regions); and Shashiprabha (Vindumati, daughter of the king of fishermen).

Monarchs and maidens The role of monarchy in maritime contacts is emphasized in the narrative; in other tales the royal connection is much more direct. Kings travel directly in search of maidens, and royal voyages are depicted as being for marital connections, which translate into political connections. The royal connection can be direct, as in the case of King Yashahketu of Anga, who discovered the adventures of his minister encountered during his journey to Suvarnadvipa and fell in love with a heavenly maiden described by the minister.14 Disguised as an ascetic, he journeyed in search for the woman, crossing many countries, rivers and mountains to finally reach the seashore. A merchant recognized the signs of royal birth on him and gave him passage on his ship. He found the maiden, a cursed vidyādharī, freed her from the curse by slaying a rakshasa, and returned to his city via a magic gateway—a tank in the gardens of the underwater city where she lived. The sea is significant in the various tales involving king Vikramaditya— he is said to rule the earth with all its dvīpas, and is also referred to as lord of the seven islands (saptadvīpeśvara); the story regarding his marriage to the princess of Simhala is also prominent.15 Within the islands there is the depiction of cities and a monarchy that is based exactly on the ‘mainland’ brahmanical pattern. The Sanskritised kingdoms of South East Asia, identified with the area termed as Suvarnadvipa, namely Java and Sumatra, Srivijaya, Sailendra and Mataram of the seventh to tenth centuries are well known. Though we cannot make any direct identifications, we can infer that the fictional symbolism of the narratives was not far from the actual processes taking place in the oceanic space.16 The connections are established at the political level through marriages between royal houses; in most cases the ‘mainland’ kings obtain ‘island’ maidens. Thus, carefully constructed political relations hinge upon the exchange of women, power flows in the same direction the princesses relocate. The ‘object’ exchanged must also cross the harrowing expanse of the sea, and so we have princesses duly put onboard ships destined for the land of their bridegroom along with much wealth and implied prosperity. Facing the sea storm is hardly an empowering trope in the case of princesses, and women in general, and nor is their depiction such that they are placed as coming in contact with any kind of superhuman beings that become the medium of their empowerment. Take the instance when king Gunasagara of the island of Kataha had a daughter born to him named Gunavati and it was announced that she would have the lord of the seven dvīpas as her husband.17 The king deliberated with his advisors and came to the conclusion that Vikramaditya was a suitable husband for her so he had her embark on a ship on the sea, with her retinue and wealth.18

44  Tara Sheemar Malhan

The exchange of royal women also involved the exchange of royal wealth, which flowed from the island to the mainland, and was further redistributed to supporters. When the marriage of king Prthvirupa of Prathisthana took place with the princess Rupalata of the isle of Muktipura, king Rupadhara gave many jewels (ratna) to the couple. The island king, being compared to a mine of jewels (ratnākara), honoured the painter and Buddhist mendicants (śramaṇa), who had been instrumental in bringing about the alliance, with dresses and ornaments, and bestowed gifts on all the others.19 King Prithvirupa remained in that city along with his attendants enjoying the best meat, drink and other luxuries arranged by his father-in-law. Afterwards he set out with his bride and retinue for his city and was escorted till the shore (samudrataṭa) by his father-in-law where he embarked on ships and crossed the sea in eight days and rejoined his army encamped on the shore of the other side. Finally, he reached his city to organize a festival (utsava) and gave villages and wealth to the painter (citrakrita), wealth to mendicants (śramaṇa) and complimentary presents to the chiefs (sāmanta), ministers (saciva) and rajputras. Undoubtedly the display of royal wealth and magnificence is a significant aspect of the narratives, as a marker of royalty that appears of enhanced import if the courtly context of the KSS is recalled, and possibly as a self-projection of the monarchy. The islands are rich in women as well as jewels/riches, the focus of monarchical aspirations.

Oceanic traffic In addition to the individuals belonging to royal lineages, there are various others that make the traffic across the seas ranging from royal messengers, bards, ministers, male and female mendicants, painters, fishermen, brahmanas and superhuman women, even children. Many such individuals were commissioned by the monarchy to undertake specific tasks—usually establishment of marital alliances. A few examples demonstrate the manner in which the narratives reflect the ‘easy’ traffic of peoples over the ocean, including those through monarchical intervention. A bard (vandi) named Shambarasiddhi states that while he was wandering the world he crossed the sea and saw the great city Malayapura, which king Vikramaditya was searching for.20 King Candasimha of Tamralipti, a city on the shore (taṭa) of the eastern sea, sent his dependent (kārpaṭika), named Sattvashila, a rajputra of the Deccan, who had been serving him for many years, to the island of Lanka to demand for him the hand of the king’s daughter.21 Sattvashila worshipped his patron deity and went onboard a ship (vāhana) along with the brahmanas whom the king appointed to accompany him. King Mahasena of Alaka sent an ambassador (dūta) named Surathadeva to Hansadvipa with a portrait of prince Sundarasena.22 King Mandaradeva of the island sent off his counter ambassador (pratidūta), who was a brahmana named Kumaradatta, to king Mahasena.

Gendered spatialisation of the ocean  45

There were also those who travelled privately, hinting at a broader social group for whom the oceanic route was accessible. A brahmana named Candrasvamin wandered in search of his lost son and daughter from island to island.23 There is the mention of a chaste (pativratā) female servant (karmakārī), named Shilavati, in the train of a merchant (vaṇika) based in Tamralipti who travels to an island called Ratnakuta with him.24 At other places there are references to pilots and steersmen of ships. See the following story of Dirghdarsin, where mention is made of the steersmen (karṇadhāra) in conversation with the minister.25 Also, see below the story of Devasmita who goes with her maids (ceṭikā) to Kataha, and the story of a painter (citrakāra) sent on a royal mission, along with two hermits. The oceanic route is indeed depicted as being busy with traffic.

Merchants and ships Significantly, the community most visible in the narratives frequenting the overseas routes and islands are the merchants and their private trading vessels. The sea and the merchant ships are inseparable. A man shipwrecked and lost in the sea and discovering a merchant ship as his rescue is a frequently appearing motif. Besides the merchant ships, the other ships plying the seas are the monarchical vessels and ships of the fishing community. Merchants are invariably mentioned in the ocean-related narratives. There are merchants who come from far away cities and return home with their profits or losses. There are also the merchants who live on the islands and engage in trade there. There are a few instances of wives of merchants travelling on ships  demonstrating that the ocean routes were not entirely prohibited for women. Numerous voyages are mentioned in the stories that were undertaken to acquire wealth and riches from the islands. In these the merchants take centre stage. Many stories also illustrate the dangers involved in sea voyages and loss of wealth might occur due to shipwrecks. Voyages were viewed as dangerous as well as profitable; similarly the ocean itself is connected with jewels (ratna), but also dangerous storms and monsters. The typical depiction of the seafaring merchants is visible in the incident when the minister of Anga, Dirghdarshan, left his country to visit holy places and at last reached the land of Paundra where in a certain city, not far from the sea, he met a merchant (vaṇika) named Nidhidatta who provided him hospitality because he saw him as a ‘distinguished’ brahmana.26 The merchant invited him to live in his home while he planned to go on a trading expedition to Suvarṇadvipa. However, the minister chose to go with him to the island so they embarked on a ship laden with merchandise. He remained some time in Suvarnadvipa with Nidhidatta (sārthavāha, vaṇika-pati), who was engaged in buying and selling. On the way back the steersman and the rest of the crew (karṇadhāra) are mentioned to be in conversation with the minister. The ship arrived at the coast and the merchant disembarked his goods, making his servants happy, and

46  Tara Sheemar Malhan

the minister went in high spirits with him to his house, which was full of joy at his arrival. Likewise, there are many voyages mentioned in the stories that were undertaken to acquire wealth and riches from the islands. Many stories also illustrate the dangers involved in sea voyages. For example, in one story it is mentioned that in the city of Dhavala there was a merchant’s son named Cakra who went on a trading voyage to Suvarnadvipa.27 There he gained great wealth in five years and in order to return embarked on the sea in a ship laden with jewels. However, he was caught in a terrible storm, barely escaped with his life and lost all his wealth. The same happened (i.e. loss of wealth) in the case of merchant Rudra.28 The narratives depict with great ease the presence of merchants on the seas and islands as is apparent in the story of brahmana Candrasvamin who went in search of his lost children. He was informed in the shore city called Jalapura that a merchant named Kanakavarman had come there a few days before and had found in the forest a brahmana boy with his sister and he had gone off with them to the great island to Narikela.29 Candrasvamin made acquaintance with a merchant named Vishnuvarman who was about to go to the isle of Narikela and went there in his ship (yānapātra). Upon enquiring, the merchants who lived there told him that the merchant had gone with the two children to the island of Kataha. Subsequently, the brahmana went in a ship with the merchant Danavarman to Kataha, then on to Karpura, Suvarna and Simhala with other merchants, but did not find the merchant he was searching for. Finally, he went with a merchant named Kotishvara to Citrakuta, crossing the sea in his ship. There he found the merchant Kanakavarman but the children he had rescued from the forest were not his. He found them later after much travelling in a city called Tarapura living in the house of the minister of the city. Take the example of prince Sundarasena and princess Mandaravati and the accompanying minister Dridhabuddhi, who left the hermitage of Matanga to go to the shore in order to find a way back home. When they reached the shore, they saw a lightship (laghu vāhanam) under the command of a young merchant coming near them. The prince asked through his minister for passage in the ship, hailing the merchant from a distance. The merchant saw the princess and was at once distracted with love, consented and brought his ship near the shore. Sundarasena first put the princess onboard and was preparing to get onboard from the bank where he stood when the merchant made a sign to the steersman (karṇadhāra) and so set the ship in motion to kidnap the princess (who was later rescued).30 From the gendered perspective the story serves to demonstrate the often depicted helplessness of women and the additional danger to them during voyages. For merchants, while wealth appears as the chief motive for sea voyages, ‘acquisition’ of women is a near second. This indeed reflects a very thriving trading and mercantile landscape. Noticeably, in most instances, the protagonists are kings, ministers and brahmanas. The merchant is depicted as the principal character in a few

Gendered spatialisation of the ocean  47

stories that speak of the voyage as a quest for wealth. The sea is made accessible to the wives of merchants too, though such instances are few. There seems to be some amount of negativity or additional fear associated with the movement of women, as we have seen above and also during the transfer of women as part of political alliances and/or romantic liaisons. We can observe this in the next story which depicts how a merchant’s wife came to accompany him on his trading voyage. The wealth of a merchant named Hiranyagupta began to diminish due to the expenses of his wife, Anangaprabha, who controlled him.31 To recover he got together some wares (bhānḍakaḥ) and went off to an island called Suvarnabhumi to trade taking his wife with him out of fear of being separated from her. In another story a merchant, stranded on a plank in the sea, seeks shelter on a ship (yānapātra) of a merchant named Krodhavarman, who makes him his companion.32 But on the ship he is discovered by the merchant secretly associating with his wife (patni). Thus, for women the standard trope is that they are either prone to getting abducted by other men, or adultery, so issues of sexuality become the focus. There is a single instance when a woman is depicted as crossing the sea without a male companion; this is in the guise of a man to save her husband from trouble. This is when Devasmita, the protagonist of a story, chaste wife of a merchant of Tamralipta, goes to Kataha to save her husband from some trouble with local merchants.33 She informed her mother-in-law in secret, was accompanied by her maids (ceṭikā), made the pretext of a mercantile expedition and took a ship (pravahaṇa) to the island of Kataha in the guise of a merchant (vaṇigaveṣa). This is in keeping with the general trend in narratives that bestow unnatural power and rights upon the chaste women, which are denied to other ‘ordinary’ women. We find depiction of a thriving merchant society on the islands. There is also some glimpse of the social customs of settlers and locals, which may be at variance with the normative Sanskritic pattern. This is briefly reflected in the story of a merchant named Candrasara of the city of Lampa who reached an island for the purpose of earning wealth.34 He began his business there with the help of a local merchant who knew his family. He asked to marry the daughter of a local merchant but was refused on some vague grounds. The father of the girl stated that he would send her to her grandfather on her mother’s side on the island of Lanka and the merchant was to go there and ask to marry her. Later in the story it is revealed that when the girl had been born a mendicant named Jinarakshita, a friend of her father’s, had told her father that he must not give the maiden away himself as she had another mother. This story betrays possible matrilineality, traces of alternate customs in narrative memory and the association with the heterodox religious traditions. Mostly mainland merchants go to the islands to acquire riches, or figure in the story as rescuers of those caught in a storm, or aides of those wishing to cross the ocean. The merchants are nearly as representative of the oceanic space as its dangers and riches.

48  Tara Sheemar Malhan

Fishermen and alternative practices The society located in the space of the ocean is not exactly the main subject matter of the narratives since the focus is on the mainland protagonists and their adventures. Associated with the shore, the sea and islands are the community of fishermen (kaivartas) and their ships. The kings of the fishermen are given some prominence in the stories as we have seen above in the story of Shaktideva the king of fishermen, Satyavrata, (niṣādādhipati, dāśendra, dāśapati and kaivartapati) is instrumental in providing assistance to the protagonist. There is an amount of ambiguity in representing relations with the fishing community due to putatively lower social status. Yet, any possible irregularity of characters, behaviour, and actions is resolved by use of narrative techniques like karma and rebirth, as when in the story of Shaktideva, the princess of the kaivartas is said to be a ‘fallen’ vidyādharī.35 The depiction of beef eating is possibly portrayal of alternate food habits of the community. While some hostility is also reflected in the story, as when the hero is imprisoned to be executed, the marital alliance serves the purpose of relieving the hostility. Use of the sea as a resource is briefly mentioned in one place, and the community of fishermen are described as catching fish a few times; the mundane sea is not the focus of the narratives, which are more concerned with fantasizing and entertaining and with subtle social rhetoric.36 For example, in the tale of Shaktideva, at one point, it is described that the servants of the king of fishermen who were engaged in the pursuit of small fish caught the big fish that had swallowed Shaktideva whole. The fishermen, proud of their catch, dragged it along to show to their king but the hero emerged from it when cut open. The fishing community appears in many other cases; for example, when a merchant took his wife along to Suvarnadvipa.37 Journeying they reached the city of Sagarapura near the sea, where he got acquainted with a chief of fishermen, called Sagaravira. The merchant and his wife went with him to the shore of the sea to travel via his ship. They got caught in a terrible storm and the merchant deserted his wife, jumped into the sea and oared himself to another merchant ship. Sagaravira saved her by making a raft of planks tied with a cord and took her to his city. These are extremely subversive depictions since they upturn the usual depiction of the brave elite urban protagonist. The cultic traditions relate to worship of the fierce goddess, magico-ritual practices similar to tantric beliefs and prominent depiction of the royal lineage and women of the community in independent roles. This is most prominent in the story of Shaktideva when he is imprisoned in the temple of Candika to be sacrificed.38 Usually in the KSS this is a motif that is connected with the depiction of the fierce tribes of the forest, who are placed in a supportive as well as antagonistic relationship with the urban protagonists.39 The magico-ritual practices that are depicted relating to the

Gendered spatialisation of the ocean  49

offerings of human sacrifice, embryo sacrifice, etc., to gain power, in this case symbolized through the magical sword, are also motifs that occur commonly in the KSS in the depiction of tantric ritual in urban, rural or wild spaces.40 Javanese inscriptions offer insights into local practices, which were clearly magico-ritual; tenth-century inscriptions mention communal feasts celebrating temple construction, rituals, donations of land and wealth.41 To an extent depiction of the fishermen follows a pattern similar to the depiction of forest tribes in the KSS. This is the pattern of composition of Sanskrit literature that standardized through repeated representation, while, as human lifestyle adaptations the difference between maritime and terrestrial hunter-gatherers has been well pointed out.42

The sacred connect I have pointed out above how mendicants, ascetics, Buddhist hermits and brahmanas are depicted as frequenting islands, and their role in the spread of ideas is well established. Of the institutions located on islands we find the brahmanical temples, monasteries and hermitages. These have long been considered bastions of brahmanical culture and once again demonstrate how the social processes are assimilated in memory through narratives. Take the case of Vidushaka, the protagonist of a story who in search of his wife, a vidyādharī named Bhadra, reached Karkotaka, a city located on the shore of the eastern sea.43 There he entered a monastery (maṭha) inhabited by many brahmanas from various lands, who were noted for their hospitality. It was a wealthy foundation of the king of the place, Aryavarman, who had annexed to it beautiful temples, all made of gold. We have also seen in the story of Shaktideva that he took shelter in a monastery of brahmans on instructions of the nishada king. While the king greatly supported and assisted the hero, he appears to follow taboos of commensality, and kept food packed by the brahmana inmate of that monastery for his journey with the king. The hermitage of sages, is also placed in spaces across the sea, near the shore. Their main purpose in the narratives is to provide shelter to stranded or shipwrecked individuals of ‘mainland’ origins, especially of royal lineages. We can take the example of princess Mandaravati of Hansadvipa who was the sole survivor of a shipwreck; the sea having lifted her up, ‘like a wave, as it were with an arm, and flung her up alive in a forest on the shore near the scene of the shipwreck’.44 She was rescued by a hermit named Matanga and his daughter Yamuna who gave her shelter in their hermitage (ashrama). Later prince Sundarasena and his minister were also rescued from the shore by two hermits who had come there to bathe and took them to the same hermitage and the prince was reunited with his beloved Mandaravati.45 Matanga sent them off to Alaka with blessings. Hermit Matanga and his daughter appear in another story as rescuers of a shipwrecked maiden, whom they named Vela, since she had been found on the shore (velā) of the sea.46 She came to be loved by all the hermits (munijanapriyā) in the hermitage (ashrama). While these hermitages are depicted as located on the shore

50  Tara Sheemar Malhan

or islands, their configuration is similar to the usual depiction of hermitages often appearing in Sanskrit literature. We see the placement of mainland brahmanical cults as part of the religious tradition of islands. Most island kings, as we have seen, are constructed on the pattern of the mainland monarchy. King Ratnadhipati ruling a great island in the middle of the sea, called Ratnakuta is called a great worshipper of Vishnu.47 Location of worship of Vishnu on the islands is noticeable in the stories. In the story of Vidushaka, king Aryavarman is said to have annexed beautiful temples of gold to a monastery.48 Above in the story of Shaktideva is the depiction of an island in the middle of the sea where the temple of Vishnu, ‘founded by the Ocean’, forms the location of a festival visited by people from many islands. The presence of male and female mendicants (pravrājakas) and Buddhist hermits (śramaṇa) can be seen in the narratives. A mendicant (pravrājikā) of mature age named Katyayani informed a prince about the princess of an island called Haṁsadvīpa, as his prospective wife.49 The ascetic (tāpasī) apprised him that she was in the habit of wandering the earth and the islands for the sake of visiting sacred bathing places and other holy spots (tīrtha). In Haṁsadvīpa she noticed the beauty of the local princess and painted her picture which she showed to the prince and he fell in love. Likewise, Naravahanadatta was informed by a female ascetic (tāpasi) while out for a hunt with his minister Gomukha about a city on the other side of the sea called Karpurasambhava.50 The king of the city, Karpuraka, had a lovely daughter named Karpurika who was destined for him. Buddhist monks are also mentioned as visiting the islands, as in the case of two Buddhist hermits (śramaṇa, bhikṣu) who went to king Prithvirupa of Prathisthana and seeing that he was very handsome told him that they had travelled through the world and had nowhere seen a man or woman equal to him in beauty except princess Rupalata of the island of Muktipura.51 Finally, when the marriage of the king took place and he reached his city with his bride he gave wealth to the mendicants (and others). The reflection of certain rituals that were meant to ‘appease’ the sea in some stories is very interesting. These reflect some form of animism, a rendition of the sea into a spirit. The material used to appease the sea is jewels, which seems to be a reversal of what the ‘mainlanders’ usually attempted to gain from the sea, jewels (ratna). This can be seen in the story of prince Sundarasena who sailed with his retinue and faced a great hurricane. The ranges of the forest on the shores of the sea are described as shaking to and fro and waters of the sea being inverted due to the winds.52 An offering of jewels was made to the sea accompanied by a loud cry of anguish.53 The pilots (karṇadhāra) let loose the sail, and flung out very heavy stones on all sides fastened by chains. After the voyagers had flung themselves in the sea and swam in different directions, the sea became calm again ‘and bore the semblance of a good man whose wrath is appeased’.54 Likewise, Vidushaka, the hero of a story and the merchant Skandhadasa were suddenly stranded on the merchant’s ship in the middle of the ocean on the way to

Gendered spatialisation of the ocean  51

a distant eastern island, and the merchant propitiated (arcita) the sea with much jewels.55 When the ship still did not move the merchant was grieved and said that whoever would release the ship would get half his wealth and his daughter. Thus, the sacred connects, at various levels, the ‘mainland’ with the ‘islands’ and the sea, hinting at a process of migration of ideas and culture.

Superhuman women and fantasy Narratives collocate superhuman women with the dangers of the sea and their emergence from inside the sea is reminiscent of the emergence of jewels at the churning of the ocean; Laxmi, the goddess of wealth and prosperity being one of the ‘jewels’. Just as overcoming the dangers of the sea is empowering for the hero, contact with these women is an empowering narrative device. They rescue shipwrecked kings and heroes, confer kingship of the superhuman domain and are an explanatory narrative device. Marital relations between mainland men and island women are validated by presenting them as vidyādharīs, as we saw in the story of Shaktideva and his marriage with the princess of fishermen. This seems to be a projection into the phantasmal realm of actual processes of interaction that took place in vast regions of the Indian Ocean.56 These heavenly women are not made available to everyone as can be seen in the story about the minister of the king of Anga who discovered a maiden (kanyā) coming out of the sea in the course of his travels and described her to his king upon return. The king fell in love with her and followed the route of travel taken by the minister to encounter the woman emerging from the sea on the way to the ‘Gold-Island’.57 The maiden disappeared in the sea and the king plunged after her and he saw a beautiful city in front of him. There he found his lady love, who happened to be a cursed vidyādharī, freed her from the curse and returned to his city from the path of a magic tank in its gardens, a magic gate to the world of men. It would be an alternative motif if the minister had been given access to the superhuman lady; that would have resulted in him becoming a king.

Narratives and history We must keep this broad context of commercial and cultural contacts between various regions of the Indian Ocean in perspective when we examine narratives of the KSS. It is established that shipping and maritime trade in early India and early medieval India was thriving with the emphasis on mainland and maritime South East Asia and that Indian Ocean trade connected diverse regions and communities. Coastal areas and ports were connected with local inland networks and with networks across seas and long-distance land networks, harbours on both sides having reaped considerable advantage due to the thriving trade as stopovers, transhipment points and movement of goods from within the subcontinent.58

52  Tara Sheemar Malhan

The two geographical areas prominently figuring in the stories are the areas that were the focus of trade from the subcontinent in the early historical and early medieval periods. One is Bengal with the port of Tamralipta on the eastern coast. The other is Suvarnadvipa/bhumi, identified with South East Asia and its islands, to which the eastern coast was the important connection point from the subcontinent. The fact that vastly separated and landlocked areas were connected by trade to the sea is also attested by inscriptional evidence. For example, in the country of Vanga inscriptions in Kharoshti and Kharoshti-Brahmi in North-Western Prakrit (first century CE to fifth century CE) have been noted and are used as proof of the suggestion that a community or communities from the north-west settled in lower Bengal in the period.59 Contacts of the region of Pundra with Gandhara has been suggested on basis of material evidence of the post-Mauryan period.60 The main point being that the northwest, was very connected with the eastern seaboard from a very early period. Elsewhere, I have asserted that the KSS has multiple layers that make possible its timelessness, as well as location in the early historical as well as early medieval context.61 Thus, I can suggest that processes connected with maritime socioeconomic connections are assimilated in memory throughout the narratives. The stories demonstrate how collective memory works in narratives and inculcates facts into fantasy. There is no interdiction of voyage; except a sense of its danger, and the projection of the profits to be made in terms of acquisition of women and wealth. There is a vibrancy of communications and depiction of transport of goods and peoples. The gaze of the mainlander dominates the stories; alternative customs are shown as being repressed through denigration. The role of fantasy is as a significant interspersal device in the stories. It connects the realms separated by form, space and time and makes possible the embedding of history in the narration. The wide range of core issues related to gender in early India including sexuality, access to oceanic routes and areas, identity, kinship, marriage and resources are all brought out in the stories.62 The ocean is constructed as a passage or a doorway, in itself a temporary dangerous space, filled with monsters as well as superhuman women and magical underground cities. Transference of wealth, women and ideas happens across the ocean. Monarchs and ministers, princes and princesses, brahmans and merchants and fishermen are the principal characters and social groups depicted in the story as participants in seafaring. Merchants and their ships offer the chief means of transport, for themselves and others. While temples, monasteries and hermitages are shown as crucial institutional links with the subcontinent; the rituals connected to the sea and of the sea dependent communities are relevant depictions. Juxtaposed with the dangerous ocean is the allegory of the sea as repository of riches and women accessible to the persons of origin in the ‘mainland’.

Gendered spatialisation of the ocean  53

Notes 1 For space and literature see Tara Sheemar Malhan, Plunging the Ocean: Courts, Castes and Courtesans in the Kathasaritsagara, New Delhi: Primus, 2017, pp. 44–46, 107. 2 Henri Lefebvre, The Production of Space, trans. Donald Nicholson-Smith, Oxford: Blackwell, 1991, pp. 48–49. 3 The necessity of studying communities living off the ocean resources has been pointed out. See Satish Chandra and Himanshu Prabha Ray (eds), The Sea, Identity and History, New Delhi: Manohar, 2013, p. 18. 4 The Vayu and other Puraṇas state that Bharatvarṣha has nine divisions called dvipas all of which are separated from each other by the sea and are not easily accessible. Jambudvipa (India proper) is the ninth. See P.V. Kane, History of Dharmaśāstras, Vol. III, Poona: Bhandarkar Oriental Research Institute, 1973 (Second Edition), p. 935. 5 R.C. Majumdar, Ancient Indian Colonies, Vol. II, Dacca: A.K. Majumdar, 1937, pp. 4, 52. Suvarnadvipa is used for the entire Malay Peninsula and Malay Archipelago, the main islands being Java, Sumatra, Bali and Borneo. He refers to the story of Candrasvamin in the Bṛhatkathā (Taranga 56, vv. 54) who is mentioned as visiting Narikela-dvipa (identified with Nicobar island), Kataha-dvipa (Keddah), Karpura-dvipa (either Borneo or north, especially the northwest side of Sumatra), Suvarnadvipa, and Simhala-dvipa (Ceylon). 6 Pandit Durgaprasad (ed.), Kathasaritsagarah, Bombay: The Nirnayasagara Press, 1889. 12. 34. 165 (Shashankapura); 9. 6. 51 (Jalapura); 5. 2. 35 (Vitankapura). 7 Ibid., 12. 19. 29–56. 8 Ibid., 12. 34. 174 (dakṣiṇapaścim). It is mentioned in a story that the prince Sundarasena sailed towards the south-west quarter to reach Hamsadvipa 9 Ibid., 9. 2. 7. 10 Ibid., 18. 1. 84 (samudra); 10. 2. 115 (sagara); 10. 3. 59. 11 Ibid., 10. 1. 75. 12 Ibid, 5. 1. 19–81, 5. 2. 1–73, 5. 3. 1–114, 255–272. It is in the book/lambaka (swell) five of the text titled Caturdarika (the four wives referring to the four vidyadhari sisters the hero gains as his wives). 13 I have created separate titled segments of the story to accentuate the themes which are discussed under various headings after the retelling of the narrative. 14 Ibid., 12. 19. 1–163. 15 Ibid., 18 3. 11, 16, 107. 16 Atsunobu Tomomatsu, Traditional Dietary Culture of South Asia: Its Formation and Pedigree, trans. Akira Matsuyama Routledge: New York, 2009. Srivijaya was an ocean-ruled kingdom during the seventh to eighth centuries and managed the trading network of the archipelago region in Java, Sumatra and Malay peninsula after the eighth century. Shailendra dynasty arose as a ruler of the port cities (disappeared after the ninth centuries) and constructed the Mahayana Buddhist temple of Borobodur. Former Hindu Mataram kingdom (constructed the Prambanan temple complex) emerged as the Muang state in the inland basin of Java having basis in the surplus rice production and trade of aromatic trees like Indian sandalwood, spices like pepper. 17 KSS, 18. 4. 105–117. 18 But when the ship came near Suvarṇadvīpa it was swallowed by a large fish with the princess and the people onboard. The monstrous fish was carried by the current of the sea to a coast near that island and stranded there. The people around ran to the fish with many weapons and cut it open and the ship came out. King Candrashekhara of the island heard of it and he was brother-in-law of Gunasagara. He took the princess home and the next day put his own daughter,

54  Tara Sheemar Malhan

19 20 21 22 23 24

25 26 27 28

29 30

31 32 33 34 35 36

37 38

39 40

Candravati, also onboard another ship with Gunavati along with magnificence and gifts to be married to the king. Ibid., 9. 1. 175–198. This analogy is so great that when after marriage the king enters the bed chamber of the princess it is said to be decorated with jewelled couches, jewelled pavement, jewelled pillars and jewelled lamps! Ibid., 18. 3. 79–80. Ibid., 12. 14. 31–33. Ibid., 12. 34. 94, 111, 121–123. Ibid., 9. 6. 1–74, 104–135. This story has been discussed below in the section on merchants and ships. Ibid.,7. 2. 9–50. The servant had a sister in Tamralitpti, named Rajadatta, and the king Ratnadhipati of Ratnakuta island married Rajadatta, the sister of the chaste servant (who cured his elephant with the touch of her hand but no other woman, even of his own household could), and took her on his prized elephant, which was his vehicle for inter-island travels, through the air and placed her in an uninhabited island in the middle of the sea in a palace and surrounded it with a guard of women so that she could never become unchaste as she would not meet a man. Ibid., 12. 19. 49. Ibid., 12. 19. 29–56. Ibid., 9. 6. 140–151. The merchant lost his wealth in this story because he had gone on the voyage against the wishes of his parents. Ibid., 9. 4. 86–88. Harishikha, general of Naravahanadatta, informed the prince, who enquired after hearing a great noise of drums that there was a merchant (vaṇika) named Rudra in the town who had gone to the island of Suvarṇadvīpa on a mercantile (vāṇijya) expedition. As he was returning, the hoard of wealth that he had acquired was lost and being sunk in the sea by his ship. He had escaped alive himself and had arrived at the town in misery. Ibid., 9. 6. 40–65. Later in the story the merchant was captured by shavaras while crossing a forest along with his wealth, followers and the abducted lady. The king of the foresters, who was a friend of the prince, returned his lady love and punished the merchant; Ibid., 12. 34. 264–269. Ibid., 9. 2. 316–319. Ibid., 7. 2. 62–102. Ibid., 2. 5. 179–180. The king of the island put the four erring merchants to justice—they had to pay a large sum to her and a fine to the king’s treasury since they tried to seduce her—and she returned home with her husband. Ibid., 11. 1. 48–53. These instances reflect the flexibility inherent in the Brahmanical system, wherein the normative philosophical concepts are in effect utilized to validate subversive acts. Another instance is a brief one when in a great tract of land assigned to brahmanas in the country of Anga, called Vrikshaghaṭa, a brahmana is mentioned as having sent his three sons to the sea to fetch a turtle for a sacrifice. See KSS, 12. 15. 1–10. Ibid., 9. 2. 318–338. This story is about the string of marriages of Anangaprabha before she finally becomes a queen. It has been noted that the Bengal coastal areas were known for worship of goddess Chandi as seen in the Chandi Mangal Kavyas. See H.P. Ray (ed.), Archaeology of Seafaring: The Indian Ocean in the Ancient Period, ICHR Monograph Series 1, Delhi: Pragati Publications, 1999, p. 23. See Malhan, Plunging the Ocean, pp. 91–97. Ibid., pp. 205–210.

Gendered spatialisation of the ocean  55



­











4 CONCEPTUALISING THE FAR WEST Early Chinese notions of Da Qin and the Indian Ocean trade Matthew A. Cobb

In the Liangshu 梁書 (Book of Liang), a seventh-century Chinese history (primarily covering the period 502–557 CE), a curious incident is reported about the arrival of a merchant called Qin Lun 秦論 on the coast of Jiaozhi 交趾 in 226 CE – a region equating to the Red River delta in Vietnam.1 This merchant, who is said to have been from Da Qin (literally ‘Great Qin’), a term often interpreted by many modern scholars as referring to the Roman Empire, was sent by the prefect of Jiaozhi to the ruler Sun Quan 孫權 (r. 222–252 CE) of the state of Wu, who asked him to provide a report on his native country and its peoples. After being gifted ‘men and women/boys and girls, ten of each’ Qin Lun returned home. It is stated in the passage directly preceding this anecdote that merchants from Da Qin frequently visit Funan 扶南, a Prefecture in the Mekong delta, as well as Rinan 日南 (Annam) and Jiaozhi, but few of the inhabitants of these southern frontier states have travelled to Da Qin. If it is accepted that Da Qin does in fact refer to the Roman Empire in this instance, and that this comment is directly connected to the anecdote about Qin Lun, then we would, on the face it, appear to have attestation of supposed frequent travel by merchants from the Roman Empire to areas along the coast of what is today Vietnam during the early-third century CE (possibly also from the end of the second century CE).2 However, this seems an unlikely proposition given our understanding of the levels of direct participation by Roman merchants in the Indian Ocean trade at this time, as is clear from the intensity of occupation at the key Egyptian Red Sea ports of Myos Hormos and Berenike, as well as the comparatively more limited distribution of Mediterranean finds (ceramics, bronze-wares, glass-wares, etc.) at sites in East Africa, Southern Arabia and India in this period.3 Additionally, as will be outlined subsequently, most merchants from the Roman Empire confined their activities to the northwestern Indian Ocean

Early Chinese notions of Da Qin  57

(Red Sea, Gulf of Aden and Arabian Sea). Regular travel to the regions of Funan, Rinan and Jiaozhi for merchants setting out from the Red Sea would probably take up at least two trading seasons – entailing travel from the Arabian Sea region, into the Bay of Bengal, and then subsequently into the South China Sea (at each stage following the rhythms of the monsoon winds). Given the availability of goods from this latter region in western Indian ports this would seem to make little economic sense.4 This raises the conundrum of how this passage should be interpreted. The problem posed by this passage also feeds into a broader debate on how to interpret and utilise claims made about Da Qin in early Chinese texts of the first millennium CE. Since at least the nineteenth century, scholars like Friedrich Hirth, who produced his seminal China and the Roman Orient, have attempted to match the claims made in these texts with what is known of the history and geography of the Roman Empire or parts of the eastern Roman Empire.5 This broadly credulous approach to the early Chinese narratives about Da Qin has continued in both sinology and classical scholarship up to the present day; although there are certainly scholars, like Shiratori Kurakichi, who have argued that we need to be more cognizant of imaginative elements in these narratives.6 Raoul McLaughlin, for example, has sought to utilise these narratives in his discussion of Roman participation in the Indian Ocean trade. He is willing to accept the claims made in these texts at face value, going so far as to describe them as ‘credible and compelling’. He assumes that information derived from travellers coming from the Roman Empire who reached the Chinese court was the source of the claims that appear in these Chinese texts.7 Similarly others like Sitwell, Li Feng and Christopoulos have assumed that many of the claims about Da Qin derive from what the envoy Gan Ying 甘英 heard while in Parthian territory in 97 CE.8 The consequence of this set of assumptions is that the claim made in the Liangshu about the frequent arrival of merchants from Da Qin on the coast of Vietnam would be taken at face value.9 It is the aim of this chapter to consider these early Chinese narratives about Da Qin and whether they can be used as uncomplicated evidence for the development of Indian Ocean trading activity in the early-mid first millennium CE as scholars like McLaughlin have done. In order to do this a brief contextual outline will be given of the nature on the texts themselves, the origin and meaning of the term Da Qin, what claims these texts make about direct contact between the Chinese world and people from Da Qin and finally a consideration of Indian Ocean trading networks in a wider context. It is argued here that very few people from the Mediterranean world made it as far as China, and vice versa, though this does not preclude the possibility of interaction at intermediary ports. However, it seems more likely that if information did reach China about the Roman world, it would very often be disseminated by intermediary cultures. Moreover, it is difficult to definitively prove the origins of claims made about Da Qin in texts like the Weilüe and Hou Hanshu, and each needs to be assessed on a case

58  Matthew A. Cobb

by case basis. In fact, it is apparent that many of the claims made in these texts can be better understood within Chinese mythical, religious, cultural and geographical frameworks for viewing the world – especially the more fantastical and utopian claims. Finally, it is argued that the term Da Qin had become sufficiently nebulous in meaning by the time the Liangshu was written (seventh century CE) that the claim made about merchants from Da Qin frequently visiting Vietnam may reflect conflation with other groups, if not outright fabrication on the part of the author.

Early Chinese texts, ethnography and the meaning of Da Qin It is necessary to start our contextual discussion with an outline of some of the key early Chinese texts that discuss Da Qin. The most important of these texts from the perspective of (purported) Romano–Chinese contact are the Weilüe 魏略, complied by Yu Huan 魚豢 in the third century CE, and the Hou Hanshu, compiled by Fan Ye 范曄 in the fifth century.10 As noted earlier there are also histories of additional interest, including the aforementioned Liangshu. The reason for focusing on the Weilüe and Hou Hanshu is that these are the earliest texts to describe Da Qin, and that later texts are often very derivative. Other poetic and noncourt texts also make brief allusions to Da Qin from the early centuries CE, such as the botanical text the Nanfang caomu zhuang 南方草木狀 (Plants of the Southern Regions).

Standard histories and ethnography Works like the Hou Hanshu and Liangshu were included in the cannon of 24 Histories, a grouping retrospectively created in the Ming period (1368–1644 CE). The Weilüe is not included in this list, but the section surviving to us was incorporated into the commentary of one of these standard histories, namely the Sanguo zhi 三國志 (Records of the Three Kingdoms), and was certainly a work compiled from official records and earlier histories.11 The process of compilation means that later works often repeat many of the claims made in earlier histories – only rarely do we see attempts to verify or critique earlier material, although this is not to say new claims and embellishments were never added.12 The authors of the Weilüe, Hou Hanshu and Liangshu were writing several generations or centuries after the purported incidents of contact (or attempted contact) and it is not always clear which earlier accounts they drew upon.13 But as noted above, it is often inferred that they are based on reports from men like Gan Ying who travelled as far west as Southern Mesopotamia. Leslie and Gardiner argue that the Hou Hanshu reflects knowledge originally derived from the first to second centuries CE, while the Weilüe reflects the early third century CE.14 However, this is entirely based on their own internal reading of the information contained within the texts and is far from conclusive. A fairly consistent feature of these histories is that they contain one or more chapters dealing with peoples living beyond the imperial bounds of

Early Chinese notions of Da Qin  59

China – such ‘ethnographic’ and geographic discussions have been a feature since at least the Shiji 史記 of Sima Qian 司馬遷 (c. 145–86 BCE).15 In the case of the Weilüe and Hou Hanshu they conveniently collate their discussion of Da Qin in the ‘Western Regions’ sections of their work. Unsurprisingly they tend to present a Sinocentric viewpoint and make little use of ‘external’ sources. As a result, any detail provided needs to be understood in relation to the interests of the Chinese compilers and scholar-officials (who were the primary audience for these texts) – which in the case of standard histories often tends to relate to the realities of imperial administration.16 One fairly common feature of what might be termed ancient ethnography is the attempt to explain other cultures and people within reference to one’s own. Scholars in previous decades have tended to focus on antithesis, extreme opposites, as a feature of these types of ethnographic study. It has often been assumed that part of the reason for this antithesis was that it provided a means of reflecting on one’s own cultural identity in opposition to the ‘Other’, rather than as a means of understanding the Other itself, consequently making redundant any attempts to use these accounts to establish clear comprehension of different cultures. However, recently various scholars have sought to nuance this perspective. Not all ethnographic narratives are focused solely on the polar – ‘us’ versus ‘them’ – some appear to implicitly and explicitly explore cultural relativism and commonality (or universality) – common, shared or similar values, customs and lifestyles perceived as being universal to all cultures.17 Of course, the latter may also be subject to authorial construction (i.e. they are not necessarily, unmediated factual reports). These caveats should be taken into consideration when analysing Chinese ‘ethnographic’ and geographic ideas about Da Qin. What is striking about the narratives concerning Da Qin in the Weilüe and Hou Hanshu is that this land is not presented as an ‘Other’ in terms of polar opposition. Indeed, many ideas about Da Qin are fantastical and utopian, with these peoples being presented as very much like the Chinese, as we shall see.

The meaning of Da Qin As previously noted, the term Da Qin has frequently been assumed by many scholars to refer to the Roman Empire or at least parts of the eastern Roman Empire (like Syria) – though some have sought to connect it variously with Macedonia, Egypt and Arabia. The term literally means ‘Great Qin’.18 This should most likely be understood as a reference to the Chinese notion of a significant civilisation (like itself) that existed in the extreme West. The term seems to be entirely Chinese in origin and does not have any derivation from another language. Its earliest use appears to connect to the explorations of the envoy Gan Ying in 97 CE.19 The debate over exactly which region the term Da Qin refers to has tended to focus on scholarly attempts to connect topographical and geographical descriptions to what is known about the Roman Empire. These arguments, of course, rest on the assumption that

60  Matthew A. Cobb

the information in the Chinese sources is based on some original factual report, rather than being the product of literary imagination. Indeed, Shiratori Kurakichi has suggested that the term Da Qin might actually be a physical reference to the size of the people (see below).20 The term Da Qin was not the only one in use. The Weilüe and Hou Hanshu also mention Lijian 犂鞬 as an alternative, as well as the term Haixi 海西, meaning ‘West of the Sea’. This latter designation is used both as a description and as an alternative name. The meaning and origin of these alternative names has generated a fair amount of contention among scholars. Debate surrounds whether they represent confusion or conflation in the sources, with their potential original denotation of another area. For example, the name Lijian has been variously interpreted as originally referring to Petra, Alexandria, the Seleucid Empire or possibly the region of Hyrcania.21 Likewise, Christopoulos has suggested that Haixi in fact refers to Egypt.22 Unlike the term Da Qin, these terms appear to have been transliterated foreign words, and attempts to identify them have particularly relied on phonology, an approach not free of difficulties given the nature of the source material.23 In light of these issues, it is probably best to accept that the Chinese sense of what Da Qin was remained quite vague and nebulous, in the same way that Roman notions of the Seres (‘Silk People’) continued to reflect a hazy conception of the distant East.

Embassies and reaching China via Indian Ocean networks of exchange There are a variety of means, both direct and indirect, by which information about particular cultures might be transmitted in the ancient world. The movement of peoples is one of the most obvious factors behind the transmission of ideas – this movement potentially being spurred by trade, migration, proselytization or other drives. The ‘quality’ of this information could vary, ranging from personal observations and direct contacts to secondhand reports and hearsay. More incidentally, material goods might also inform cultural perceptions, although the origins of this material (and its original significance) might not always be fully understood by the recipient culture.

Embassies Evidence for direct contact between peoples from the Roman world and China appears to be very limited. This is further complicated by the fact that the ethnographic terminology employed in both cultures is far from precise. In the Graeco-Roman literary tradition there are references to the Seres (‘Silk-People’), Sērikē and Sinai/Thinai, which are located by the author of the Periplus and Claudius Ptolemy at the distant eastern edge of the οἰκουμένη.24 However, the identification of these people has proved quite contentious, with suggestions ranging from an origin in Central Asia, East Asia or China. However, such ideas were probably quite

Early Chinese notions of Da Qin  61

nebulous and attempting to fix them in a specific geographic and cultural sense is almost certainly misguided.25 In any case, the epitome of Florus and the Historia Augusta are the only western sources to speak of emissaries from the Seres coming to the Roman Empire. Florus purports that one such embassy came during the reign of Augustus (along with ambassadors from the Scythians, Sarmatians and Indians), while the author of the Life of Aurelian makes reference to the Seres visiting during his triumph in 274 CE.26 There are no accounts of ‘ambassadors’ being sent from the Roman Empire to the land of the Seres, Sērikē or Sinai/Thinai. By contrast, no Chinese historical tradition makes reference to any of their ambassadors reaching Da Qin; although we are told of an attempt by Gan Ying, who was sent by the General Ban Chao 班超 (32–102 CE) to explore as far as the western sea. In the course of his journey it is claimed that he reached what appears to be the mouth of the Persian Gulf, before apparently being told by those in the region that if he wanted to reach Da Qin he would need to take a sea voyage that would entail a round trip of anywhere between two months to two years. Baulking at this prospect Gan Ying decided to return home. Most scholars tend to assume that this refers to a sea voyage round the Arabian Peninsula, in which rounding it and entering the Red Sea it would be possible to reach Egypt.27 We do not hear of any subsequent attempts to send embassies to Da Qin, but it is claimed in some of the Chinese dynastic histories that envoys from Da Qin were sent by the king Andun 安敦 (usually identified as Marcus Aurelius Antonius) in 166 CE.28 The envoys mentioned in this passage are often considered to be merchants rather than genuine representatives of the Roman Empire, and the precious but not spectacular goods – ivory, rhinoceros horn and turtle shell – were probably picked up in the course of their voyage.29 The next chronological date we have for contact is that of Qin Lun in 226 CE, in the aforementioned passage of the Liangshu. After this the Yiwen leiju 藝文類聚,a seventh-century reference encyclopaedia, records the claim that a few people from Da Qin came via southern China (Guangxi – i.e. via the Sea route) in 282 CE, bringing with them ‘tribute’, including fire-washed cloth (asbestos cloth).30 The Nanfang caomu zhuang, an early fourth-century botanical work, claims that in 284 CE Da Qin sent 30,000 roles of honey fragrance paper to China.31 McLaughlin goes so far as to suggest that the incidents in 282 and 284 CE represent attempts by private merchants from Alexandria to establish trade connections with the Chinese court. However, this is purely speculation on his part.32 Indeed, there may be some confusion in these sources, with this being a single ‘embassy’ connected to the year 284/285 CE.33 As a side note, it is worth stating that these few reported incidents of contact all relate to individuals arriving via the southern sea route (i.e. Indian Ocean and South China Sea) and not overland through Central Asia. Indeed, there is little reason to believe that people from the Roman Empire ever travelled the whole way along what is known as the Silk Road, or more accurately Silk Routes. Goods conveyed along these routes would have been

62  Matthew A. Cobb

exchanged by various groups at different stages. In fact, the only evidence we have for such an extended land journey is the expedition sent by Maes Titianus, a Macedonian merchant – the account of which was recorded by Marinos of Tyre (and survives in Claudius Ptolemy’s Geography). This expedition got as far as a place known as the Stone Tower, which is usually located in the region of the Pamir Mountains, the area merging with the Hindu Kush, the Himalayas and the Tianshan mountain range.34

Connecting West and East: wider Indian Ocean trade networks in the first millennium CE These few instances in which we hear of ‘envoys’ or merchants arriving in China appear to be quite exceptional and do not suggest regular direct contact. This is perhaps unsurprising, especially with regards to regular merchant activity. Not only because of the vast distances involved, but also from what we know about Indian Ocean trading patterns. The first millennium CE did see substantial (if not always even) growth in the level of interconnections between different parts of the wider Indian Ocean world. In the western Indian Ocean (especially the Arabian Sea region) networks of exchange developed between the Mediterranean world, East Africa, the Arabian Peninsula and western India. Similarly, in this same period, the eastern Indian Ocean also saw the growth of various trading links between eastern India and Southeast Asia.35 However, many traders operating in the northern Indian Ocean are likely to have confined their activities to a particular sphere like the Arabian Sea or the Bay of Bengal, likewise for those operating in the South China Sea in the case of Sino-Southeast Asian commercial relations.36 The regional patterns of the monsoons winds, as well as more localised conditions, meant that travel from the Red Sea all the way to the South China Sea was probably not achievable within a single sailing season, at least prior to the development of steam-powered ships.37 Of course, the existence of subsystems in the wider Indian Ocean did not preclude the wide-ranging movement of goods across the Afro-Eurasian world. Chinese silks were brought to the Mediterranean, often changing hands a number of times, while Roman coins and glassware appear to have found their way to Southeast Asia, most probably via intermediaries. Certain regions or entrepots acted as conduits between the different trading spheres: India (and later Sri Lanka) as a conduit between the Arabian Sea and Bay of Bengal, and the Thai-Malay Peninsula as a conduit between the Bay of Bengal and the South China Sea.38 It is not inconceivable that a small group of Roman merchants were occasionally adventurous enough to sail as far as Southeast Asia (and travel on even as far as China), but this should be seen as the exception rather than the rule.39 Likewise, travel from southern China to southeast India would have been fairly lengthy (at least two years for a return journey) and probably irregular.40 There is some evidence (Arabic and Chinese literary testimony) for a brief period during the last few centuries of the first millennium CE (although

Early Chinese notions of Da Qin  63

possibly going back to the late Sassanian period) of Persian (Bosi 波斯) and Arab (Dashi 大食) merchants sailing as far as Guangzhou (Canton), in southern China.41 The Belitung shipwreck has been assumed to be representative of this activity.42 However, it should be noted the nature of the excavations has proved controversial and the types of wood (both African and Indian in origin) that were used in its construction, as well as the wadding material (possibly Southeast Asian), complicates attempts at definitively confirming an Arabian, Indian or other origin of construction.43 In any case, by the beginning of the early second millennium CE this more direct trade activity came to an end, while places like Southern India and the Malay Peninsula continued to function as key intermediary points of exchange. There is little evidence to suggest that around the early third century CE we should see merchants from the Roman Empire as forerunners to this later (temporary) direct trade undertaken by Persian and Arab merchants. The vast majority of material culture from the Roman Empire found across sites in the wider Indian Ocean world clearly connect to Arabian Sea networks of exchange (integrated into overland and riverine networks in East Africa, Arabia and Central and South Asia). Moreover, it has been reasonably suggested that much Graeco-Roman information about Southeast Asia seems to have derived from Indian informants, since the former contain less detail than the Indian accounts but list similar place-names.44 Given the issues just outlined, and the fact that the literary sources suggest very limited contact with peoples from Da Qin, it is unlikely that direct contact was a significant means by which knowledge about the Far West was disseminated to China. Nevertheless, the possibility remains that some information about political structures, trading patterns and social and commercial networks could have been disseminated via intermediaries or at intermediary ports. There are certainly parallel historical examples pertaining to the Indian Ocean that indicate that information about other cultures could be indirectly acquired. A case in point is the Periplus Maris Erythraei – a mid-first century CE merchant’s guide to the Indian Ocean. The account was written by a Greek-speaking merchant operating from Egypt, who clearly had personal experience of the trade.45 Nevertheless, De Romanis has convincingly argued that the author’s descriptions of coasts, sea routes and commodities reflect the accumulation of information from a variety of different sources (including, but not limited to, personal observations).46 The author of the text aimed to provide key details about the major ports of the northwestern Indian Ocean littoral, including giving a description of the coast of East Africa below the Horn. The detail in the account suggests a reliance on secondhand information (it is notable that Claudius Ptolemy later mentions only two named Roman mariners who had explored the coast as far as Rhapta, one of whom had been blown off course by accident).47 Most likely the author of the Periplus heard reports from Arab skippers and agents from the Southern Arabian Peninsula who frequently engaged with and intermarried into the local population of Rhapta (probably near Dar es Salaam), and, as a result, had knowledge of the area and the language of its people.48

64  Matthew A. Cobb

Chinese notions of Da Qin Having outlined the nature of the texts, the debates over the meaning of Da Qin and the evidence (or lack thereof) for Romano-Chinese contact within the context of the wider Indian Ocean trade, it is worth turning to the claims made in the texts themselves. The authors of the Weilüe and Hou Hanshu make various statements about the kingdom of Da Qin, including its supposed political structures, wealth, the dress and appearance of the people and its proximity to the realm of Xi Wangmu 西王母 (Queen Mother of the West). Many of these claims are quite fantastical and are not easily married to what is known of Imperial Roman history. These details tend to elicit two broad responses. The first, adopted by scholars like McLaughlin and Christopoulos, we have already noted. This position holds that the information contained within these texts represents genuine transfer of (at least some) knowledge about this distant western region, albeit with a great deal of misconception resulting from its transmission and its filtering through the lens of Chinese cultural conceptions. However, there is a second broad approach that can be taken to the narrative in these texts. Namely that they largely reflect a construct of the distant West based on (rather than simply being moulded by) Chinese mythical, religious, cultural and geographical frameworks for viewing the world. If the former position is adopted, the Indian Ocean might be seen as a clear conduit through which knowledge of the Roman Empire (however, limited) filtered into the Chinese world. Whereas the latter position suggests little meaningful transfers of knowledge beyond the basic awareness of a powerful kingdom existing in the West. It is not necessary for us to take an absolutist view with regards to either position. It is not unreasonable to suppose that some reports or hearsay gleaned about this distant land in the west could easily have been woven into a narrative that is heavily influenced by Chinese conceptions of the universe in which they lived. Attempting to disentangle this is no easy task. Any assessment about the claims needs to be done on a case by case basis. I would, however, argue that a great many of the claims can be best explained in reference to Chinese mythical, religious, cultural and geographical frameworks for viewing the world. Clearly a great deal of the narrative we get from these texts is a construct, which has obvious implications for any scholar wishing to utilise statements from these narratives to further our understanding of the ancient Indian Ocean trade.

The people of Da Qin The constructive elements of these texts are immediately apparent from the physical descriptions given of the people of Da Qin. They are said to be tall and honest, like the Chinese, with the Weilüe even stating that they themselves claim to have originally come from China.49 It seems most logical to assume that the authors of these texts, who relied on earlier reports/account (or at least claim to) took the term Da Qin (Great China) – which may simply

Early Chinese notions of Da Qin  65

have been used in origin as an unqualified designation of a significant kingdom in the west – and subsequently embellished their descriptions by assuming that the term implied some kind of ethnic or cultural similarities.50 One other (complementary) possibility is that in making Da Qin in some respects like an idealised China, its use as an ethnographic ‘mirror’ became more palatable. That is to say a useful means of highlighting perceived issues or failings within one’s own culture by referencing another.

Location and environment The geography of the kingdom of Da Qin also fits Chinese conceptions of the lands to the West in both a geopolitical and in a mythic-religious sense. It is said to be located to the west of Anxi 安息 (usually understood as referring to Parthia) and Tiaozhi 條支 (sometimes taken to refer to Characene and Susiana). Da Qin’s western borders stray into the realm of the mythic-religious. It is described as the furthest land one reaches before arriving at the realm of Xi Wangmu, the Shifting Sands and Weak Water.51 Beyond this is the region where the sun sets. Xi Wangmu was associated, particularly in Daoism, with prosperity, bliss and immortality. Other facets of the accounts of Da Qin are likely coloured by this association.52 For example, some of the products associated with this kingdom are not ones we would identify with the Roman Empire, such as jade.53 If we consider that the Jade Mountain was said to be one of the homes of Xi Wangmu, then the Chinese association of jade with Da Qin is not too surprising. More broadly the kingdom of Da Qin is represented as being very prosperous – the kings lived in palaces with pillars made of crystal or glass.54 While the Roman Empire was known for its very fine glassware, it would be a bit of a stretch to take the ‘credulous’ view and suggest that this formed the underpinning of these descriptions. Perhaps more worthy of note is the fact that glass is frequently associated with fantasy worlds in the literature of the Han and Wei periods, such as in the Shizhou ji 十洲記 (An Account of the Ten Continents [pre-265 CE]), which mentions a glass palace in Fangzhang Mountain.55

Governance and administration The governance of the kingdom of Da Qin, as outlined in the Weilüe and the Hou Hanshu, also shows utopian elements and the influence of Chinese notions about ruler-ship. Most evident is the claim that the king is not permanent and can be replaced, particularly if unusual natural phenomena occur. Additionally, it is reported that the king has a council of 36 leaders/advisors who must be consulted, and that the common people can put their cases to the king, literally in sacks, to be addressed subsequently.56 The former notion of the resignation of the king at signs of the displeasure of heaven seems plausibly connected to the notion of the Mandate of Heaven.57 While the responsiveness of the king to his advisors and the concerns of the people also

66  Matthew A. Cobb

seem to broadly fit into an idealising trope, though why exactly the number of advisors was 36 is difficult to say.58 Some scholars have sought to offer explanations for these claims that suggest a misconstrued (rather than fabricated) sense of Roman history on the part of these Chinese authors. Theories include the notion that this is an idealised interpretation of the swift succession of emperors during the third century or an allusion to consular elections or the Nervan-Antonine practice of adopting successors.59 Similarly the practice of going around and collecting/listening to petitions has been connected with the practice of Roman provincial governors giving assizes (it was quite common for governors to tour their province, so as to hear legal cases and attend to related business), while some suggest that the replacement of kings may refer to the replacement of governors (like that of Egypt, who resided in the old Ptolemaic palace in Alexandria).60 Further attempts to rationalise the information from these texts can be seen with regards to the 36 advisors/generals of Da Qin. One such idea has been to suggest that they represent the consilium principis, the small council of select senators and magistrates that consulted with the princeps on matters of state.61 It is worth noting, however, that this was not a fixed body – either in terms of membership or operation under particular emperors (who could in any case consult as freely as they wished).62 A point that makes this group even less obviously based on anything tangible from Roman political history. It appears easier to place this claim into a broader idealising narrative (the ethnographic ‘mirror’). The chief city of the king of Da Qin is said to be more than 100 li around (41.5 km) and located near the mouth of a river. Both the Hou Hanshu and Weilüe describe the chief city of Da Qin as having five palaces each located 10 li apart (4.15 km), with the king travelling to one of these palaces each day.63 By the late first to second century CE the Palatine had become the main location for the wider palatial complex of the emperors, though they had property in other parts of the city as well, such as the Gardens of Sallust.64 There is no obvious topographic connection with what is claimed in these texts. Attempts could be made to suggest that since Lijian (possibly originally a designation for Egypt, but later equated with Da Qin; see below) and Haixi (‘West of the Sea’) are equated/conflated with Da Qin, then the authors may have been describing another city like Alexandria or Antioch-on-the-Orontes. For example, Hirth interprets the descriptions of the topography and regions in the chief city of Da Qin as more likely representing Antioch rather than Rome.65 Hoppál restates this case by suggesting the possibility that the authors may have conflated a prosperous administrative centre in the Roman East with the capital, that by the Late Antique period the western emperors were often away from Rome, that aspects of local administration of the city may suit the description of our sources and that they misconstrued some of the public buildings in the city as five palaces.66 The problem with this theory is that it requires sufficient knowledge of the topography of this city to have been transmitted to China (of which we have no obvious indication of by what mechanism this might have happened),

Early Chinese notions of Da Qin  67

with a simultaneous lack of knowledge of the political structures of the ­Roman Empire.67 The suggestion that information about the chief city of Da Qin was based on an actual city (or conflation of a number of cities) cannot be ruled out, but the limited details make any attempt to assert a specific site a doubtful prospect. An alternative explanation, which to my mind better fits, is that the five palaces accord with contemporary Chinese notions of the Wuxing 五行 (encompassing the Five Elements, Five Phases, Five Agents, Five Movements, Five Processes, Five Steps and Five Planets) and the cardinal direction, which had come to be associated with the Five Elements (west – metal, east – wood, south – fire, north – water, centre – earth).68 Indeed, Shiratori Kurakichi, suggests this idea may connect to the notion of five celestial palaces and the legend of the Emperor Shun’s periodic tour of inspection of the four sacred mountains.69 Numerology was significant in Chinese culture, meaning it is very likely, though not absolutely certain, that some of the numbers given in the Hou Hanshu and Weilüe have a symbolic significance, although this can only be speculation.70 In any case, there appears to be little obvious historical basis for the claim about the five palaces.

The products of Da Qin Discussion of the products of Da Qin perhaps has most pertinence with regards to Indian Ocean trade. McLaughlin, working on the assumption that the Weilüe and Hou Hanshu faithfully reproduce information appearing in ‘original reports’, regards these texts as being able to provide information about Roman products reaching the Far East (the former text associating Da Qin with some 59 products).71 He even suggests that many of these goods, which are clearly from other regions such as Arabia and India, were reprocessed and reexported items from Roman workshops.72 On the face of it, it might be tempting to accept this premise, at least in part, since some of the products listed in our Chinese texts do in fact derive from the Roman Empire (although not exclusively). These include gold, silver, various base metals, glass, red coral, asbestos cloth, realgar, storax, orpiment and saffron.73 A few of these products like red coral were in fact highly regarded across many cultures.74 While storax and glassware appear to have been particularly appreciated in China.75 It might even be tempting to give some credence to the claim made in the Weilüe about the people of Da Qin reworking imported Chinese silk and reselling it at a profit.76 This is because what seems to be corroborating evidence appears in a few Graeco-Roman sources. Pliny mentions women of the Empire dividing threads of silk and them reweaving them to make a lighter more gauze-like fabric; a practice also briefly alluded to in Lucan’s Pharsalia.77 Furthermore, there is a reference in the Edict of Maximum Prices to a variety of reworked silks including dyed and embroidered products (some with gold).78 However, there are reasons to be more circumspect. In the case of the notion of reworking woven Chinese silk products, Hildebrandt observes

68  Matthew A. Cobb

that any such practice (pulling it out and stretching) would damage the fabric. She argues that instead ‘Lucan may have alluded to a diaphanous, purple-dyed silken gown that was formerly a plissé-fabric, but stretched to make it transparent’ and that Pliny is talking about the unravelling of the cocoons not woven fabric.79 Further reason for doubt also derives from the fact that Da Qin need not automatically designate peoples from the Roman world. The existence of a reasonably developed silk industry in India in the early historic period should at least give us pause about whether the claim made in the Weilüe necessarily reflects accurate Chinese knowledge of Roman practices regarding silk processing.80 With regards to the range of items listed as products of Da Qin, most of these, as noted, clearly derive from regions such as East Africa (ivory, rhinoceros horn); Southern Arabia (frankincense and myrrh); the Persian Gulf region (pearls); India (dhūṇa, rock crystal, pearls, ivory, carnelian); and Southeast Asia (sea turtle shells).81 Rather than attempt to claim that all of these items were reprocessed and reexported from the Roman Empire, it seems more plausible to suggest that they were conflated with Da Qin, as a powerful semiutopian kingdom in the distant West, and that these authors were simply not in a position to clearly distinguish the origins of these products.82 This confusion and lack of clarity about the origin of particular imported products is not an unusual phenomenon. For example, there was a persistent and erroneous belief repeated in Graeco-Roman literature that cinnamon, cassia and nard came from Arabia.83 It may also be possible to draw certain parallels with the list of products from Da Qin given in the Weilüe and Hou Hanshu with Buddhist ideas that made their way into China. In Mahayana Buddhism there is the concept of the ‘Seven Treasures’. In the Great Event (an important Mahayana text) these seven consist of crystal or quartz, pearl, lapis lazuli, gold, silver, red coral and agate or coral (later texts substituted rubies or diamonds). These items became features of Buddhist heavens and also came to constitute elements of the immortal trees of Daoist paradises.84 All of these items are listed in the Weilüe with the exception of lapis lazuli, but these could be understood as being subsumed within the various semiprecious gems category.85 This appears similarly to be the case with the slightly less extensive list in the Hou Hanshu, which actually mentions blueish-green (qingbi 青碧) gems, quite possibly referring to lapis lazuli.86 While Da Qin was understood as a terrestrial empire, the fact that it was in the Far West near the location where paradise was situated (at least in certain religious-philosophical creeds like variants of Buddhism and Daoism) may explain why these products were associated with this empire. This suggestion is, of course, speculative.87

Da Qin as a nebulous concept? Given the contested debate in modern scholarship regarding exactly what Da Qin refers to, and the fact that many mythic-religious concepts are associated with it, we should consider questioning whether our Chinese authors

Early Chinese notions of Da Qin  69

had a precise, consistent and coherent sense of this land. In consequence this has obvious implications for the way in which we understand the reference to merchants from Da Qin frequently sailing to Vietnam, mentioned in the Liangshu. The notion that a term that designates an ‘other’ ethnic-cultural group might be vague, imprecise and fluid (i.e. open to being understood differently from context to context) should not be surprising. This type of problem has a number of historical parallels. For example, the term Yavana, in origin, appears to have been used in parts of India to designate Greek speaking peoples (it is generally assumed that the Persian term for Ionians, Yauna, was transmuted into the Prakrit Yona, and Sanskrit Yavana). However, by the early to mid-centuries of the first millennium CE it appears to have become much broader in meaning (not necessarily just designating Greek speakers), and could be understood in different terms (as appears to be the case in the Sangam literature of the Tamil cultures, as well as in inscriptional evidence from the Deccan and the island of Socotra).88 Similarly the Graeco-Roman designation of the Aethiopians, a people that are broadly associated with northeast Africa, could, depending on the period and context, be associated with the Meroitic kingdom, or, by the third century CE (if not earlier), with the Axumite kingdom. Similarly some classical authors conflate India and Ethiopia and, at least by the Late Antique period, the term India (sometimes distinguished between India Major and India Minor) could include regions like East Africa.89 This potential ambiguity should be borne in mind when considering the description of Da Qin and its people in the textual source material. We also need to be cognizant of the fact that Chinese historical texts, as has been discussed, usually followed their predecessors when describing foreign countries and situations. Indeed, it is of interest that texts dating several centuries after the initial composition of the Weilüe and Hou Hanshu appear to transpose a number of ideas connected to Da Qin with Fulin 拂菻 (the latter usually being interpreted as referring to Byzantium).90 Meaning some of the descriptions and designations repeated in later texts may represent fossilised ideas, which can become distorted in the process of conflation of earlier (sometimes more accurate) predecessors.91 This adds a further layer of complication when thinking about ‘knowledge transfer’, and how much can genuinely be taken as reports that resulted from trade contacts.

Conclusion In reviewing early Chinese narratives about Da Qin in texts like the Weilüe and Hou Hanshu, it appears that a lot of the claims are quite fantastical and utopian in tone. This is not to say that the genuine transmission (received either directly or via intermediaries) of information (however garbled) about the distant west never took place. However, many of the claims appear to fit into Chinese mythological, religious, philosophical, numerological and

70  Matthew A. Cobb

geographical conceptions of the world. It is also worth emphasising the point that the term Da Qin was probably quite vague in conception, and that the fossilisation of particular claims (and their subsequent transferal to descriptions of Fulin) as a result of the compiling by later authors adds a further complicated dimension to our assessment of the origins of these narratives. These factors have undeniable implications when it comes to assessing the usefulness of these texts for understanding facets of wider Indian Ocean trade in the first millennium CE. The conflation, particularly in the Weilüe, of many products from the west with Da Qin obviously suggests we cannot use this information as a clear-cut list of ‘Roman exports’ noted by the Chinese. More tentatively it could at least be used in a looser sense to vaguely tell us about the types of products making their way to China via seaborne and overland trade networks, but even here caution is required since it is not clear whether the compliers envisage all of these items as imports to China or merely items to be found in Da Qin.92 Furthermore, I would argue in light of the issues outlined in this chapter surrounding Chinese conceptions of Da Qin that the claim in the Liangshu (referring to the period of the third century CE) about merchants from Da Qin frequently coming to the coast of what is present-day Vietnam should not simply be taken at face value – it does not represent an unproblematic attestation of third-century Roman trading activity with the Far East. Ultimately, it is not my intention to claim that these early Chinese texts have no use in the study of Indian Ocean and South China Sea networks of exchange (commercial and cultural) of the early to mid-first millennium CE. However, I hope to have demonstrated that these reports need to be treated with a great deal of circumspection.

Notes 1 Liangshu 54.798. Compiled by Yao Silian 姚思廉 (557–637) – an official of the Sui and later Tang dynasty. 2 Preceding the reference to the frequent travels by merchants from Da Qin to what is now present-day Vietnam is a brief recounting of an earlier embassy from Da Qin that arrived in China in 166 CE; the fact that it precedes this statement adds a further degree of ambiguity. Does the claim allude to the very end of the second century as well as the early-third century CE? It should be noted that the term Roman merchants is used here as a convenient shorthand for various peoples of differing cultural and linguistic background (who seems to have frequently employed Greek as a lingua franca) who were subjects/citizens of the Roman state. 3 Most scholars regard the third century CE to be a period of low-level direct participation by merchants from the Roman Empire, with events such as the Antonine plague (from 165 CE) and subsequent social, economic and political problems in the Roman Empire seeming to exacerbate this decline (a modest downturn may have already set in from the early mid-second century CE). For a summary of this issue see Matthew Adam Cobb, ‘The Chronology of Roman Trade in the Indian Ocean from Augustus to the Early Third Century CE’, Journal of the Economic and Social History of the Orient, 2015, 58(3): 362–418.

Early Chinese notions of Da Qin  71

4 5 6

7 8

9

10 11

12

13

The tax rate on Indian Ocean imports was lowered from 25% to 12.5% at some point during the Severan period (193–235 CE – the octava was clearly in place by 227 CE – Codex Justinianus 4.65.7), which seems to both reflect a lower level of trade in this period and perhaps an attempt by the Roman state to restimulate it – Andrew Wilson, ‘Red Sea Trade and the State’, in Federico De Romanis and Marco Maiuro (eds), Across the Ocean: Nine Essays on Indo-Mediterranean Trade, Leiden: Brill, 2015, pp. 13–32; Steven E. Sidebotham, Review of Across the Ocean, by Federico de Romanis and Marco Maiuro, Journal of Roman ­Archaeology, 2015, 29: 915–919. Periplus 56. Friedrich Hirth, China and the Roman Orient: Researches into their Ancient and Medieval Relations as Represented in Old Chinese Records, Leipzig & Munich: Georg Hirth, 1885. See Shiratori Kurakichi, ‘Chinese Ideas Reflected in the Ta-ch’in Accounts’, MTB, 15: 25–72; Shiratori Kurakichi, ‘The Geography of the Western Region Studied on the Basis of the Ta-ch’in Accounts’, MTB, 15: 73–163. By contrast, Leslie and Gardiner, although accepting there are some imaginative elements in the texts, critique Shiratori and argue that while exaggerated their authors were ‘aiming at a real country, however dimly seen’. In particular, arguing that the earlier texts the Weilüe and Hou Hanshu are the most convincing – Donald Daniel Leslie and Kenneth Herbert James Gardiner, The Roman Empire in Chinese Sources, Rome: Bardi, 1996, p. xxv. Raoul McLaughlin, Rome and the Distant East: Trade Routes to the Ancient Lands of Arabia, India and China, London: Continuum, 2010, pp. 21, 107. Nigel H. H. Sitwell, Outside the Empire: The World the Romans Knew, London: Paladin, 1984, p. 176; Li Feng, Early China: A Social and Cultural History, Cambridge: Cambridge University Press, 2013, p.  280; Lucas Christopoulos, ‘Hellenes and Romans in Ancient China (240 BC–1398 AD)’, Sino-Platonic Papers, 2012 230: 1–88, pp. 45–46. See also Leslie and Gardiner, ibid., pp. 24–26. That the reports in the Liangshu are not always the product of critical scrutiny and careful filtering of information is perhaps a generous way of characterising this text. For example, Shiratori Kurakichi notes how a spurious account by a charlatan (effecting the role of a Buddhist monk) claiming to have visited the fantasy land of Fu-Sang 扶桑 in the eastern ocean was accepted into the text (as in other dynastic annals) – Shiratori Kurakichi, ‘Chinese Ideas’, p. 47. The surviving portions of the Weilüe come from a later fifth-century commentary by Pei Songzhi 裴松之 – Leslie and Gardiner, The Roman Empire in Chinese Sources, pp. 5, 20. Later travellers and geographers ‘add little to the descriptions’ in the Weilüe and Hou Hanshu. Another text, the Hou Hanji 後漢紀, was compiled the in fourth century, but it is generally not regarded as an independent source, and at best, in rare cases can supplement the other two texts – ibid., pp. 11, 21–22, 41, 57. Hence, the focus on the former two texts in this chapter. Michael Loewe, ‘Knowledge of Other Cultures in China’s Early Empires’, in K. A. Raaflaub and Richard J. A. Talbert (eds), Geography and Ethnography: Perceptions of the World in Pre-Modern Societies, Chichester: Wiley-Blackwell, 2010, p. 75; Krisztina Hoppál, ‘The Roman Empire According to the Ancient Chinese Sources’, Acta Ant. Hung., 2011, 51: 263–305, p. 269. Whether the ‘Western Regions’ (‘Xiyu Zhuan’) section of the Hou Hanshu (ch. 88) drew upon the ‘Xirong Zhuan’ of the Weilüe is a contested issue – Leslie and Gardiner, The Roman Empire in Chinese Sources, pp. 20–21; Yu Taishan, ‘China and the Ancient Mediterranean World: A Survey of Ancient Chinese Sources’, Sino-Platonic Papers, 2013, 242: 1–268, pp. 15–18. Yu Huan, the compiler of the Weilüe is not explicit about his sources of information – Hoppál, ibid., p. 268.

72  Matthew A. Cobb











­



Early Chinese notions of Da Qin  73

27 Hou Hanshu 88.10. A later Daoist text (Taiqing jinye shendan jing 太清金液神丹 經) presents Da Qin in even more utopian terms, reporting that someone from China was accidentally blown off course for 60 days ending up in the land of Da Qin. Pretending to be an ambassador from Funan, he was taken to see the king of Da Qin, was given a bunch of gifts and subsequently took four years to reach Funan on his return journey. See Yu Taishan, ‘China and the Ancient Mediterranean World’, p. 195. 28 Hou Hanshu 88.12. 29 For a summary of scholars see Matthew Adam Cobb, Rome and the Indian Ocean Trade from Augustus to the Early Third Century CE, Leiden: Brill, 2018, p. 121, n. 139. 30 Yiwen leiju 76. 31 Nanfang caomu zhuang 9. 32 McLaughlin, Rome and the Distant East, p. 139. 33 On the possibility that the Yiwen leiju and Nanfang caomu zhuang are alluding to the same incident see Leslie and Gardiner, The Roman Empire in Chinese Sources, p. 160. 34 For discussion of the silk route see Craig Benjamin, Empires of Ancient Eurasia: The First Silk Road Era, 100 BCE–250 CE, Cambridge: Cambridge University Press, 2018; David F. Graf, ‘The Silk Road between Syria and China’, in Andrew Wilson and Alan Bowman (eds), Trade, Commerce and the State in the Roman World, Oxford: Oxford University Press, 2017, pp. 443–529. 35 For a wider discussion of this trade activity see Himanshu Prabha Ray, The Winds of Change: Buddhism and the Maritime Links of Early South Asia, Oxford: OUP India, 1994; Federico De Romanis, Cassia, cinnamomo, ossidiana: uomini e merci tra Oceano indiano e Mediterraneo, Rome: L’Erma Di Bretschneider, 1996; Roberta Tomber, Indo-Roman Trade: From Pots to Pepper, Bristol: Bristol Classical Press, 2008; Eivind Heldaas Seland, ‘Archaeology of Trade in the Western Indian Ocean, 300 BC–AD 700’, Journal of Archaeological Research, 2014, 22: 367–402; Steven E. Sidebotham, Berenike and the Ancient Maritime Spice Route, London: University of California Press, 2011; Tom Hoogervorst, Southeast Asia in the Ancient Indian Ocean World, Oxford: Archaeopress, 2013, pp. 13–15; Sing Chew, ‘The Southeast Asian Connection in the First Eurasian World Economy, 200 BCE–CE 500’, in Michael Pearson (ed.), Trade, Circulation and Flow in the Indian Ocean World, London: Palgrave Macmillan, 2015, pp. 27–54. For an examination of not only economic but also cultural and intellectual interconnections in the ancient Indian Ocean see Matthew Adam Cobb (ed.), The Indian Ocean Trade in Antiquity: Political, Cultural and Economic Impacts, London: Routledge, 2019. 36 Philippe Beaujard, ‘The Indian Ocean in Eurasian and African World-Systems before the Sixteenth Century’, Journal of World History, 2005, 16(4): 411–465, p. 413. 37 Notionally the breadth of the Indian Ocean could be traversed in its southern region – peoples from Austronesia used winds and currents to reach Madagascar – Edward A. Alpers, The Indian Ocean in World History, Oxford: Oxford University Press, 2014, p. 9. Much later the Dutch sailed some 4,416 nautical miles from the Cape to reach Southeast Asia – Kirti N. Chaudhuri, Trade and Civilisation in the Indian Ocean: An Economic History from the Rise of Islam to 1750, Cambridge: Cambridge University Press, 1985, p. 131. However, the notion that regular trade networks (as opposed to periodic migrations) between Southeast Asia and Eastern Africa existed in the early centuries of the first millennium CE remains controversial. Miller argued that Pliny’s account of raftmen was evidence for a regular ‘cinnamon trade route’ connecting Southeast Asia and East Africa – James I. Miller, The Spice Trade of the Roman Empire 29 B.C. to A.D. 641, London: Clarendon Press, 1969. This idea, however, has

74  Matthew A. Cobb

38

39

40

41

42

43

44

since been heavily critiqued by Manfred G. Raschke, ‘New Studies in Roman Commerce with the East’, Aufstieg und Niedergang der Römischen Welt, 1978, 9(2): 604–1378, pp.  652–654. Although more recently Lytle has attempted to modify Millar’s argument, suggesting that cinnamon and cassia was brought (perhaps  directly) from Southeast Asia to Somalia or Southern Arabia (not Madagascar and Rhapta) – Ephraim Lytle, ‘Early Greek and Latin Sources on the Indian Ocean and Eastern Africa’, in Gwyn Campbell (ed.), Early Exchange between Africa and the Wider Indian Ocean World, Cham: Palgrave Macmillan, 2016, pp. 113–134. Reference in the Periplus (36, 60) to the very large Indian vessels known as kolandiophonta operating in the Bay of Bengal testify to the existence of Indian and Southeast Asian trade links. The Hanshu 漢書 (‘Book of Han’ – first century CE) describes maritime links between Guangzhou and Malay Peninsula, including the exchange of gold and silks for pearls and glass – Lin Ying, ‘Ruler of the Treasure Country’, p.  329. On the role of the Straits of Malacca, Malay Peninsula and Sumatra as intermediary points where merchants met to exchange goods see Sunil S. Amrith, Crossing the Bay of Bengal: The Furies of Nature and the Fortunes of Migrants, London: Harvard University Press, 2013, p. 14. Even travelling to the Ganges region of northeast India appears to have been uncommon for Roman merchants, at least initially. Strabo (15.1.4) notes that only a small number of merchants had reached the Ganges, although by the Flavian period Pliny (NH 6.24.82) does given an estimate for the amount of time it took for ‘Roman’ vessels to sail from Sri Lanka to the Ganges (7 days compared to a journey of 20 days made by local reed boats). On the limited Roman material found in this region see Cobb, Rome and the Indian Ocean Trade, p. 173. The Han Shu (28B, 1671) does refer to some ships travelling from Xuwen and Hepu (Guangxi and Guangdong) to Huangzhi (perhaps near modern Chennai) in the time of Wudi (r. 141–187 BCE). This took place in a series of stages. For the outbound journey there was a 5-month sea journey, 4-month sea journey, a 20-day sea journey, followed by a 10-day land journey and 2 further months at sea (i.e. one whole year for an outbound journey) – Loewe, ‘Knowledge of Other Cultures’, pp. 81–82. On the wind and current patterns impacting on the Malay Peninsula (both to the west – Bay of Bengal – and to the east – South China Sea) see Michel Jacq-Hergoualc'h, The Malay Peninsula: Crossroads of the Maritime Silk Road (100 BC–1300 AD), Leiden: Brill, 2002, pp. 17–21. For this evidence see George F. Hourani (revised and expanded by J. Carswell), Arab Seafaring in the Indian Ocean in Ancient and Early Medieval Times, expanded edition, Princeton: Princeton University Press, 1995; Dionisius A. Agius, Seafaring in the Arabian Gulf and Oman: The People of the Dhow, London: Routledge, 2005; Touraj Daryaee, ‘The Persian Gulf Trade in Late Antiquity’, Journal of World History, 2003, 14(1): 1–16. See, for example, Michael Flecker, ‘A Ninth-Century AD Arab or Indian Shipwreck in Indonesia: First Evidence for Direct Trade with China’, World Archaeology, 2001, 32(3): 335–354; Michael Flecker, ‘The Origin of the Tang Shipwreck: A Look at Its Archaeology’, in Alan Chong and Stephen A. Murphy (eds), The Tang Shipwreck: Art and Exchange in the 9th Century, Singapore: Asian Civilisations Museum, 2017, pp. 22–39; Jessica Hallett, ‘Pearl Cups Like the Moon: The Abbasid Reception of Chinese Ceramics and the Belitung Shipwreck’, in Venetia Porter and Mariam Rosser-Owen (eds), Metalwork and Material Culture in the Islamic World, London: I.B. Tauris, 2012, p. 350. For an overview of the controversies see Meg Lambert, ‘Belitung Shipwreck’, Trafficking Culture: Researching the Global Traffic in Looted Cultural Objects, 2012, https://traffickingculture.org/encyclopedia/case-studies/biletung-shipwreck/ (accessed on 11 January 2019). Ray (1994) cited by Hoogervorst, Southeast Asia, p. 12.

Early Chinese notions of Da Qin  75

45 Periplus 20.7.14–15. For a summary of the debate over issues such as the background of the author, and the dating and nature of the text see Cobb, Rome and the Indian Ocean Trade, pp. 22–24. 46 Federico De Romanis, ‘An Exceptional Survivor and Its Submerged Background: The Periplus Maris Erythraei and the Indian Ocean Travelogue Tradition’, in Giulio Colesanti and Laura Lulli (eds), Submerged Literature in Ancient Greek Culture, Vol. II, Berlin: De Gruyter, 2016, pp. 97–110. 47 Eivind Heldass Seland, Ports and Political Power in the Periplus: Complex Societies and Maritime Trade on the Indian Ocean in the First Century AD, Oxford: BAR, 2010, p. 15; De Romanis, ibid., pp. 101–03; Claudius Ptolemy Geography 1.9. 48 Periplus 16. 49 Hou Hanshu 88.11; Weilüe 11. Roman notions of the Seres are likewise quite fantastical and, at times, utopian. For example, it is claimed that they are fantastically long-lived (up to 200 years), just, unwarlike and shy about engaging with foreign peoples – Pomponius Mela Chorography 1.11, 3.60; Pliny NH 6.24.88. 50 Yu Taishan, ‘China and the Ancient Mediterranean World’, pp. 29–30; Hoppál, ‘Roman Empire’, p. 284. 51 Xi Wangmu is frequently associated with the distant West in Chinese texts like the Ganquan fu 甘泉賦 and Mutianzi zhuan 穆天子傳 – Yu Taishan, ibid., p. 47. A deity highly popular in Han times, she presided over the yin realm of the West – Paul S. Ropp, China in World History, Oxford: Oxford University Press, 2010, p. 33. The ‘Weak Water’ is water that is so weak not even a feather could float on it – Lin Ying, ‘Ruler of the Treasure Country’, p. 338. 52 The Far West was associated with being the abode of the immortals – if a traveller was able to gain access to the near impenetrable fortress of the Queen Mother of the West they would receive the elixir of immortality – Loewe, ‘Knowledge of Other Cultures’, p. 85; Yu Taishan, ibid., p. 30. During the Northern and Southern Dynasties period the Xi Wangmu was ‘elevated to the position of highest goddess in the Shangqing tradition of Daoism.’ – Mark E. Lewis, China between Empires: The Northern and Southern Dynasties, Harvard: Harvard University Press, 2009, p. 194. In Pure Land Buddhism the lay person could appeal to Amitabha Buddha (the Buddha of infinite light) to (eventually) be reborn into the Pure Land of the Western Paradise where Amitabha reigned – Ropp, ibid., p. 56. 53 Weilüe 10, 22; Hou Hanshu 88.12. 54 Hou Hanshu 88.12. 55 Lin Ying, ‘Ruler of the Treasure Country’, p. 338. 56 Hou Hanshu 88.11. 57 On the influence of Confucian principles on these ideas see Shiratori Kurakichi, ‘Chinese Ideas’, pp. 50–52. The means by which the legitimacy of a ruler’s power might be expressed underwent a series of complex developments throughout Chinese history. By the time of the Eastern (Later) Han and Northern and Southern Dynasties, a series of developments had taken place. It was not simply a case of the ‘Mandate’ coming directly from heaven (a mandate that was changeable), but omens needed to be recognised and interpreted (by diviners and experts). The ruler did not transcend everything (and hence control it), but was defined as being the centre of a web of relations between Heaven, Earth, the people and the ancestors. For a more sophisticated discussion of court legitimation see Howard L. Goodman, Ts’ao P’i Transcendent: The Political Culture of Dynasty-Founding in China at the End of the Han, Richmond, Surrey: Routledge, 1998, pp. 15–44; Ropp, China in World History, p. 30; and Michael Puett, ‘Ghosts, Gods and the Coming Apocalypse: Empire and Religion in Early China and Ancient Rome’, in Walter Scheidel (ed.), State Power in Ancient China & Rome, Oxford: Oxford University Press, 2015, pp.  230–259. On the development by Dong Zhongshu 董仲舒 (c. 175–105 BCE) of the concept of the emperor as Son of Heaven and an intermediary between Heaven and Earth (merging aspects of yin-yang theory,

76  Matthew A. Cobb

58

59

60 61

62 63 64

65 66 67 68

69 70 71 72 73 74

Five Phase theory, Taoism and Legalism, within a Confucian framework) see John King Fairbank and Merle Goldman, China: A New History, 2nd edition, Cambridge, MA: Harvard University Press, 2006, pp. 66–67, and more generally on ‘correlative cosmology’ (the mutual influence of correspondences between man, earth and heaven), pp. 64–66. It may possibly parallel the claim made elsewhere in the Hou Hanshu that the western region (the territories controlled by China that comprised the eastern end of the Silk Road) consisted of 36 kingdoms, though this could be just pure coincidence rather than signifying anything more meaningful. Alternatively, some have suggested that this could have been a pseudo-number, which simply meant many or countless – Yang Lien-sheng, ‘Numbers and Units in Chinese Economic History’, Harvard Journal of Asiatic Studies, 1949, 12(1/2): 216–225, p. 218; Hoppál, ‘Roman Empire’, p. 280. Anne Kolb and Michael A. Speidel, ‘Perceptions form Beyond: Some Observations on Non-Roman Assessments of the Roman Empire from the Great Eastern Trade Routes’, Journal of Ancient Civilizations, 2015, 30: 117–149, pp.  138–139. See also Leslie and Gardiner, The Roman Empire in Chinese Sources, p. 233. Ibid. Hill, Through the Jade Gate to Rome, Vol. I, p. 270. However, no references to various consilia specifically indicate a membership of 36 members – Augustus’ consilium princips seems to have comprised about 20 men, though it was subsequently supplemented by members of the equestrian order. More peculiarly, Christopoulos, ‘Hellenes and Romans in Ancient China’, p. 46, speculates that the 36 leaders reflect a ‘Hellenistic type of political kingship, indicating a form of Koinon, or general-councillor assembly’. However, the only evidence he cites in support of this claim is the mid-second century CE Lycian Koinon which consisted of 36 strategoi. On the consilium principis see John Crook, Consilium Principis, Imperial Councils and Counsellors from Augustus to Diocletian, Cambridge: Cambridge University Press, 1955. Weilüe 11; Hou Hanshu 88.11. While a number of scholars like Leslie and Gardiner, The Roman Empire in ­Chinese Sources, p. 102, and Hill Through the Jade Gate to Rome, Vol. I, p. 269, accept the interpretation that the city is Rome, this is not a universally shared view. On the Imperial Domus see Shelley Hales, The Roman House and Social Identity, Cambridge: Cambridge University Press, 2003, pp. 61–93. Hirth, China and the Roman Orient, p. 276. Hoppál, ‘Roman Empire’, pp. 276–281. For further critique of this theory see Shiratori Kurakichi, ‘Chinese Ideas’, p. 53. Fairbank and Goldman, China, p.  64; Li Feng, Early China, p.  304; Yu Taishan, ‘China and the Ancient Mediterranean World’, p.  69. On the developing association of qi with the notion of yin-yang and wu xing during Qin and Han periods see Michael Nylan, ‘Yin-yang, Five Phases, and qi’, in Michael Nylan and Michael Loewe (eds), China’s Early Empires: A Re-appraisal, Cambridge: Cambridge University Press, 2010. Shiratori Kurakichi, ‘Chinese Ideas’, pp. 55–59. On the significance of numerology see Fairbank and Goldman, ibid., pp. 64–65. McLaughlin, Rome and the Distant East, p. 21. In this he follows Leslie and Gardiner, The Roman Empire in Chinese Sources, pp. xviii, 218. McLaughlin, Ibid., p. 107. For a discussion of products exported from the Roman Empire as part of the Indian Ocean trade see Cobb, Rome and the Indian Ocean Trade, pp. 216–271. On the export of red coral from the Roman Empire in the Indian Ocean trade see Periplus 28, 39, 49, 56; Pliny NH 32.11.21–23; Kautiliya Arthasastra 2.11.42; Garuda Puranam 68, 80; Ratnapariksa 250–52.

Early Chinese notions of Da Qin  77













­

78  Matthew A. Cobb

Eastern and Southern Edges of the Inhabited World from the Greco-Roman Perspective’, in Serena Bianchetti, Michele R. Cataudella and Hans-Joachim Gehrke (eds), Brill’s Companion to Ancient Geography: The Inhabited World in Greek and Roman Tradition, Leiden: Brill, 2016, pp. 186, 191. 90 Yu Taishan, ‘China and the Ancient Mediterranean World’, pp. 14, 19. We see the word Fulin (often connected to the Byzantine Empire) appear for the first time in the fourth century CE in the Qiangliang (‘Annals of the Anterior Liang’). In a Tang Dynasty text – the Xintangshu (chap. 221) – there is reference to Fulin being another name for Da Qin. On this see Christopoulos, ‘Hellenes and Romans in Ancient China’, pp. 67–76. The Xintangshu (eleventh century CE) shares quite a lot of details with the earlier Weilüe and Hou Hanshu, albeit with some elaboration of the various stories. Also the Tang period text, the Tang huiyao, describes Fulin (99, 2019) in terms that are quite similar to our earlier depictions of Da Qin, including references to messages who carried a bag when they followed the king, the removal of the king in the event of unfavourable natural portents and the claim that there are said to be 400 settlements (Weilüe 11; Hou Hanshu 88.13). On this see Loewe, ‘Knowledge of Other Cultures’, pp. 85–86. 91 On the problem of later Chinese historians indiscriminately mixing up material from their predecessors see Leslie and Gardiner, The Roman Empire in Chinese Sources, pp. xx, 6, 22. Ibid., p. 200. 92

5 AL-MUQADDASˉI’S TENTH-CENTURY MARITIME LANDSCAPES OF THE ARABIAN RED SEA Dionisius A. Agius

Introduction The Arabian Red Sea coast, lying at a crossroads between Africa and Asia, was known as a commercial hub for the trade of frankincense, gold and slaves in antiquity, after which it emerged under the Romans and the Ptolemies (third c BCE to third century CE) as an important focal point for the India trade; it continued to do so in classical and medieval Islam and for many centuries thereafter, serving as a pilgrim–trade route until the last days of sail. Historical Greek, Latin and Arabic texts are our primary sources, the underlying structure for our understanding of the past; however, other benefits are resourced here: from archaeological finds and oral history I describe as the surface structure for interpreting the texts, enabling a wider view of what was described by classical and medieval authors. Social scientists today may be cognizant of the Greek and Latin texts dealing with the Red Sea and the Western Indian Ocean but less familiar with Islamic texts and so largely unaware of their importance in understanding late antiquity and what they offer to researchers studying the maritime landscapes of early medieval Islam. As a model for studying these maritime landscapes in coordination with archaeological investigation and oral history, I have chosen al-Muqaddasī (d. after 378/988), a geographer-traveller whose work is considered to be the foundation of physical and human geography, not only for its content but for the systematic inquiry he applied when recording details of the land and sea.

A new genre of geographical treatise The aim of ninth- and tenth-century classic geographical works was to survey routes and settlements, the lands and seas known to the Islamic caliphate

80  Dionisius A. Agius

and governors of the time.1 These works laid the foundation for the geography of Muqaddasī’s (d. after 378/988)2 Aḥsan al-taqāsīm fī macrifat al-aqālīm (The Best Divisions for Knowledge of the Regions), the most significant of early medieval geographical treatises, both in terms of content and methodology. It is a single work that in the words of Basil Collins, the translator, ‘bespeaks a product of Islamic civilization at its finest’.3 Al-Muqaddasī’s birth and death dates are not precisely known but he flourished towards the second part of the fourth/tenth century. He was a merchant and traveller by sea and land. Not much is known about his life, except what he tells us in his travel narrative (riḥla) written over a period of 20 years. He endeavoured to describe every region ‘with the deserts and the sea in them’; their routes, the commodities of towns and cities, the traditions and customs of various communities and their material culture.4 He had a special interest in seamanship and his descriptions afford us tantalizing glimpses of navigation and sailing techniques. In general, it may be said that his descriptions are readily understood and ‘accessible to the specialist and layman alike’.5 In view of the above comments, it is the aim here to examine al-Muqaddasī’s physical description of the Red Sea, the Arabian landscape and seascape in particular and his methodological approach to tenth-century scientific investigation, noting his attention to the coastal towns, their forts, wells and markets and the sea hazards encountered in navigating this perilous sea. An attempt is made in this chapter to answer the following questions: What is the scientific basis for al-Muqaddasī’s inquiry? How far does his treatise contribute to our understanding of the coastal landscape and seascape of the Red Sea and traditional seamanship? Finally, how significant is his information for the modern researcher? There are three parts to this study: the first is an assessment of the author’s scientific knowledge, the second is a discussion on climatic conditions and seamanship of the Arabian Red Sea and the third is an evaluation on the facilities and services offered to merchants and pilgrims when they landed.

A scientific approach In all the geographies noted earlier, we see the beginnings of the social sciences, the emerging need to show the day-to-day life, the practicalities, of Islamic culture in towns and villages, whether inland or on the coast. In these early texts we find descriptions of the life of individuals and their connections with the Islamic communities at large through a number of articles of commerce they produce in every region. But when we compare al-Muqaddasī to his predecessors, we find his approach much broader, encompassing an interest in all aspects of the material culture,6 addressing both the merchant and traveller. Water is the most essential ingredient of life and the subject of water management and urban development are topics he comments on. With participant observation, he provides his findings on different aspects of life from climate, vegetation and language to religion and culture ‘with objectivity and neutrality’.7

Maritime landscapes of the Arabian Red Sea  81

Al-Muqaddasˉı’s methodology I first demonstrated in an earlier work the uṣūl method, or the fundamentals that set the general criteria for fact-finding whatever the line of inquiry was in classical and medieval Islam.8 These research criteria were not all adhered to by travel scholars but essentially, seeing and hearing, were the chief fundamentals. What set al-Muqaddasī’s methodology apart from his predecessors, was his systematic approach, with the application of mainly three criteria, not always necessarily in the following order: seeing or observation (ciyān), hearing or oral communication (samāc or mushāfaha) and the consultation of manuscripts and maps (muṭālaca) that appears to be consistent throughout. He strove, in his own words, for accuracy, verifying information through a chain of reliable authorities (isnād); and if the chain had been broken he would reject the information given.9

The auditory and ocular cultures In Touati’s analysis of the seeing and the hearing experiences of geographertravellers and writers, he claims that both are given equal credit by medieval scholars; however, this may not have always been the case, as rivalry gradually increased then, leading them to claim the superiority of one method over the other. In order to understand the position that geographers took to fact-finding we have to examine the relationship between man and the environment and man in the environment of his work. Thus, Touati follows what the prose writer, al-Jāḥiẓ (d. 254/868) in his Kitāb al-Ḥayawān (Book of Animals) sketched out: ‘a framework of interpretation between man, work, and the ecological milieu’.10 These three factors cannot be separated from each other. Hence, the importance of describing the lands and the community’s activities in such lands as first-hand observers. The question is what judgement criteria did travel geographers follow? This is partly answered by the contemporary historian-geographer, al-Mascūdī (d. 345/956). In one of his narratives about Indian Ocean seafarers, he claims a sailor’s trustworthiness is judged by the criteria of experience, skill and age,11 the essence of true perception. In the collection of sea stories, the Kitāb cajā’ib al-Hind (The Book of the Marvels of India), alleged to be collected by Captain Buzurg Ibn Shahriyār (fl. mid-to-late fourth/tenth c.), we encounter navigators of great nautical experience who voyaged long distances, some as far as the dangerous seas of China.12 There is one story in particular narrated to him about a certain Captain Abhara who voyaged to China seven times commenting that ‘only adventurous men had made this voyage before’.13 If it is a true story he must have had enormous skill, knowledge of winds, currents and tides and experience to have returned from such a long and perilous journey alive and sound. The main issue here is whether we can rely on the informants. Not all the collected stories are endorsed with the names of people who provided information, though some do have details of their occupation and status among other things.14

82  Dionisius A. Agius

Of course, both auditory and ocular methods could lead to doubts but one way scholars checked the source was to apply the isnād method of the chain of authorities as mentioned earlier. Buzurg, al-Mascūdī and al-Muqaddasī were in contact with seafarers, merchants and pilgrims, who, with them, voyaged the seas of cUmᾱn (Oman), cAdan (Aden), East Africa and the China Sea, recounting stories of sea hazards, lost lives and cargoes thrown overboard.15 Thus, al-Jāḥiẓ’s principle that the ‘human mind is perfected only by voyages’16 is the acme of ocular and auditory experience: seeing lands, sensing the environment, talking to people of different social strata and observing them at work.

Al-Muqaddasˉı’s scientific knowledge of the Arabian Peninsula The traveller presents the facts about the Arabian Peninsula by first describing the region through maps he consulted at libraries; some giving details of things not familiar to him. He personally voyaged the ‘Peninsula of the Arabs’ and his information on winds, anchorages and islands was confirmed by ‘shipmasters, cargo masters, coastguards, commercial agents and merchants’.17 Not only were these people questioned on matters of landing places, winds and islands but some, he maintains, had nautical guides. Early Islamic maps like the ones that al-Muqaddasī had seen were roughly in shapes such as a ṭaylasān (headdress)18 or a bird. These forms perplexed him until one day he approached a learned Shaykh, Abū cAlī Bin Ḥāzim, head of merchants and a ship owner of several ships, who was sitting on cAdan (Aden)’s seashore, and asked him to draw the lines of the Arabian Peninsula. The shaykh drew on the sand a rough map containing ‘gulfs, tongues, and numerous bays’, sketching ‘two arms’ representing the Red Sea and the Persian Gulf.19 However, when al-Muqaddasī came to draw his own map of the Indian Ocean, referring to it as the ‘Sea of China’, he did not include bays and gulfs. His conceptualization of the whole Indian Ocean seems to have been abstract.

Climatic conditions and seamanship Al-Muqaddasī embarked on a ship at Al-Qulzum (today Suez) that sailed down the Arabian Red Sea coast reaching al-Yaman (Yemen), then round the southern Arabian coast to the Gulf of Oman and from there, the final leg, to the Persian Gulf, reaching the Persian coast of cAbbadān (East of Baṣra).20 The Red Sea is often described as the most difficult of seas, because of the frequent changes in the direction of the wind. In my recent fieldwork in the Red Sea, I recognized this description as being totally accurate and further corroboration comes from Early Modern travellers and colonial surveyors.21 His ship, most likely, was plank stitched, a construction technique found in most Western Indian Ocean ships for many centuries, something I

Maritime landscapes of the Arabian Red Sea  83

can personally attest as I found evidence of it on an abandoned fishing boat on the Southern Arabian coast in 1996 and in Kerala on the West Indian coast in 2013. This would have given his vessel greater flexibility to withstand the erratic behaviour of the Red Sea winds and waves.22 It needs to be mentioned that, then as now, the skilled mariners would have, as I observed in many places of the Red Sea, to watch the shallowness of the water because of the rocks and coral reefs found more frequently in the south. It is possible that the first leg of his journey was from Al-Qulzum to Judda (Jeddah). This voyage, according to written and oral reports, would have probably taken 20 days in fair winds allowing for stops at one or two other main trading ports such as Yanbuc and Al-Jār (see Figure 5.1 below), but the return journey would have been longer, about two months,23 counting the time spent in the harbour waiting for favourable winds and often struggling against the north westerly winds, covering only a short distance in a day.24 The data in Table 5.1, extracted from al-Muqaddasī’s observations of the climatic conditions of the Arabian Red Sea, points out the likely sea hazards navigators encounter in eight zones of the Arabian Red Sea. His terse comments for each sea zone in Table 5.1 appear rather general but when his notes are taken together, comparing one zone with another, the reader starts to form a holistic picture of the maritime landscapes. Fārān, in TABLE 5.1 Sea hazards in eight zones of the Arabian Red Sea according to

al-Muqaddasī25

Sea zones Arabian Red Sea

Seascape

Sea hazards

Al-Muqaddasī’s comments

Jubaylān

Fathomless deep

Fārān (Tārān)

Utter destruction of ships

Winds from Egypt and Syria

Al-Ḥawrā’

Ships taken by surprise attempting to enter the port Hazardous to ships; a source of apprehension Many ships have been wrecked

Al-Silāb

Strait

Jābir (Jā’iz)

Shallow waters

[Islands of] Kamarān [Island of] Al-Mandam (Mandab)

A difficult strait

A place not to be negotiated except with the force of a freshening wind [Island of] Usqūtra Like a tower Marks the limit of the (Socotra) in the Indian pirates; the terror dark sea of ships

Truly, a place of calamity Fearfulness and calamity

84  Dionisius A. Agius

the north, is exposed to the North Westerly winds that ‘blow from the direction of Egypt and Syria’. Here, for roughly nine months a year (May–January) prevailing northerly winds blow for what can be irregular periods of time. Of the next anchorage, Al-Ḥawrā’, near Al-Jār (north of Jeddah, see map below), he reports that ‘ships are taken by surprise’ because the anchorage has no protection. Al-Silāb is described as a dangerous zone where ships needed to take extra care. Mandam (Mandab) Island, also known in European texts as Perim Island, the southern gate to the Red Sea, was known as being a difficult strait between the African and Arabian littorals; with favourable winds, larger ships steer to the right and smaller vessels to the left. Shipwrecks were many: in the Fārān zone, he notes ‘utter destruction’ for ships and warns mariners of fearfulness and calamity around Jābir and the Islands of Kamarān. The warnings were needed as the Red Sea is notoriously exposed to many such winds that could take vessels by surprise. But gales and storms were not the only way to capsize ships: sometimes they were loaded with excessive cargo, passengers, cattle or camels; or water could seep through the planks’ seams tipping the balance of the ship. Human error could also play its part, of course; handling the sails and rigging incorrectly could well lead to a ship’s destruction.26 There was, however, much greater hazard caused by human beings: Pirates in the medieval period infested the Western Indian Ocean. They harboured their vessels in the Island of Usqūtra (Socotra) and pounced on passing ships entering the Horn of Africa into the Gulf of Aden. He described it as ‘the limit of the Indian pirates, the terror of ships hereabouts’.27 Reports of pirates go back at least to late antiquity where we are told they looted large vessels in the northern Nabataean sea zones.28 In the southern kingdom of the Sabaeans (c. 500 BCE),29 where there was ‘no central authority’, the Periplus of the Erythraean Sea records that herdsmen ‘eked out their meagre livelihood with profitable returns from piracy’.30 Classical Greek, Roman and Islamic texts mention anchorages that were positioned latitudinally opposite each other. This ‘twin geography’, in Ralph Pedersen’s view, is not accidental; he says: ‘[It] perhaps supported east-west sailing from the earliest times’.31 Consider Yanbuc opposite the Roman Berenike, while Jeddah is in line with the African cAydhāb (Aidhab). Strategically located as it was, this cAydhāb (see Table 5.2 below) became a pilgrimage port town in medieval Islam.32 However, at Al-Wajh (El Wejh), in the Arabian north, which lies opposite Myos Hormos, later known in the Mamlūk Islamic period (seventh to eighth/thirteenth to fourteenth c) as Al-Quṣayr (north of modern Egyptian town Quseir), sailing was possible in a west–east direction but the east–west direction could only be achieved by sailing north to Rᾱs Muḥammad on the south point of the Sinai Peninsula and then catching the North Easterly winds down to Al-Quṣayr (Quseir), as my informants reported to me during my fieldwork in Quseir between 2002 and 2004.33 Nautical documentation in early medieval Islam is patchy, but we have the occasional snippets of information al-Muqaddasī provides that enable

Maritime landscapes of the Arabian Red Sea  85 TABLE 5.2 Al-Muqaddasī’s data on the coastal landscape facilities and services34

Red Sea anchorages

Coastal landscape

Availability of water

Markets

Al-Qulzum

Seaport to Ḥijāz

Thriving markets

*Al- cUshayra

Drinkable water; poor quality

Incomparable caravanserai

Badr

Small town on the North-East African coast Pilgrimage port town on the North-East African coast Large sea town; Abundant water well-fortified c Ayn al-Nabī Small town

Al-Jār

On the seacoast; fortified Water from Badr

Al-Ḥawrā’

Port of Khaybar; fortress; suburbs densely inhabited On the seashore; fortified Water; many reservoirs Port of Sancā’ and Sac da Water from distance Coastal town; fort Rain reservoir Port of Zabīd Wells; fresh water

Thriving market; granary of Madīna Market

*cAydhāb (Aidhab) Yanbuc

Judda (Jeddah) c Aththar Ḥalī Ghalāfiqa Al-Sharja; al-Ḥirda; c Aṭana Mukhā (Mocha) c Adan (Aden)

Coastal towns

Water from distance

Flourishing sea town

Drinking water

Populous sea town; fortified

Salty water; number of cisterns

Markets bustling

Granary of Makka; entrepôt Fine market Source of grain Granaries

Entrepôt of al-Maghrib

us to discern the theory of navigation and the skills held by mariners in sailing the seas at the time. The winds are notoriously unpredictable in the Red Sea, as we have seen, and al-Muqaddasī explains how these winds are tackled by the mariners; he reports: A practice of theirs is to send a party of men to observe the wind; if the winds abate, or if that one prevails which blows from their side, they proceed; otherwise they have to stay a long time until the arrival of the hour of relief.35

86  Dionisius A. Agius

The information is unique, and has not been recorded by other geographers, nor mentioned centuries later by the master navigator Aḥmad Ibn Mājid (d. after 906/1500). The coastal route may have been safer but could be dangerous in strong winds because of reefs and rocks, in which case, al-Muqaddasī tells us that: ‘Ships changed course to islands and depths of the sea’.36 Island hopping offered an alternative to sailing in the middle of the sea. The midsea course was noted in later centuries by Ibn Mājid and Sulaymān al-Mahrī (d. 917/1511)37, but there is no mention of it by al-Muqaddasī. Al-Muqaddasī does mention a nautical technique when steering the ship in and around rocks or navigating through the reef channels. He observes that the captain climbs to the crow’s nest and on seeing a rock or shoal he cries out: ‘To the right!’ or, ‘To the left’. The same cry is repeated by two cabin boys presumably because the noise of the winds and waves could muffle the original cry of the captain. The author adds that the helmsman at the stern controls the direction of the ship by handling two ropes attached to the rudder, thus ‘pull[ing] right or left, according to the directions’.38 This rope-steering technique, survival of which lasted until fairly recent times, is a technological progression from using a quarter rudder.39 An alternative to climbing the crow’s nest is to stand at the foredeck and shout instructions from there; this is still practised today, as I myself saw in 2018 on Mafia Islands in Tanzania, East Africa. Not only did al-Muqaddasī inquire of shipmasters as to the theory and practice of navigation but, on observing them, he remarked that: ‘They study [the instructions] carefully together and on which they rely completely proceeding according to what is in them’.40 What does this imply? Are we talking here of oral or written instructions? Later assertions state that sailing instructions were memorized,41 often in the form of poetry, handed down through the generations. Ibn Mājid and Sulaymān al-Mahrī’s own treatises were perhaps written in verse form to aid memorization. Navigators needed no navigational instructions in coastal sailing, Ibn Mājid said, as they were guided by landmarks42 but needed to know, as al-Muqaddasī noted, the danger zones. I need to stress here that in my recent field trips when I asked Egyptian, Hijazi and Eritrean captains of small- and medium-size vessels about sailing instructions, they all seemed to agree that they depended on the oral and the visual. They claimed that they have never used any maritime guide or map to find their way all along the coast or across the African or Arabian coasts.43 However, Arabian, East African and Persian mariners in the tenth century could well have possessed portolan charts and nautical guides, not in Arabic but most likely in Persian, which was, at the time, the dominant maritime repertoire. It is likely that the geographer-traveller, al-Muqaddasī, would have been conversant in Persian as many scholars were in medieval Islam. A late Persian source of the twelfth century mentions a manual guide after the Persian rahnāma (manual of sailing instructions) model44 containing information on star positions,45 bearings and physical descriptions of coastal landscape and islands. The point here is that al-Muqaddasī’s

Maritime landscapes of the Arabian Red Sea  87

observation of mariners discussing nautical instructions before putting them into practice is noteworthy. Historically, this information predates any knowledge we have of nautical manuals in possession of mariners in the Red Sea and possibly in the Western Indian Ocean.

Anchorages and facilities Once al-Muqaddasī lays down the basis for our understanding of the winds the ships faced both at sea and when approaching the coast, he turns to supplying specific information on the anchorages (Table 5.2), describing each port town, its fort, if one is found, and comments on wells, markets and granaries. Sixteen landing places are recorded here: With the exception of two anchorages on the African coast, cUshayra and cAydhāb (marked with an asterisk), the rest are all on the Arabian coast, which could indicate that the latter is safer to sail along and offers better anchorages in spite of sea hazards noted earlier (see Figure 5.1). But the absence of data on the African ports suggests that the author had little access to information because he did not travel in that region and, even if he did receive reports from people he met in his travels along the Arabian coast, he would have preferred to rely on his observation. Table 5.2 highlights the trade links by sea and land for each sea town and region covering the topographic and urban space, with the understanding that the sea connects the land, the people and their material culture. The first on the list is Al-Qulzum (Ancient Clysma) that connects the land–lake–river route to Fusṭāṭ (Old Cairo) and the sea route to the Ḥijāz.46 It is at a cross-roads, gateway to the Mediterranean, with ‘thriving markets’, while cAdan (Aden), lying at the entrance of the Red Sea on the Horn of Africa, is well-populated in the tenth century, and classified as the ‘entrepôt of al-Maghrib (North Africa)’; it is commercially strategic, as it brings the India trade to North Africa and Europe via Fusṭāṭ.47 The India trade included the commodities that came from the West Indian coast and South East Asia; hence, he calls this seascape, ‘the corridor of al-Ṣīn (China)’.48 Yanbu c (or as it is called today [Yanbuat-the Sea]), described then in the tenth century as a large sea town and known earlier by the geographer Claudius Ptolemy (d. c. 170) as Iambia Vicus 49 is linked with the villages inland of (Yanbu-at-the-Palm Grove). Other coastal towns are listed in connection with inland settlements: AlḤawrā’, the port of Khaybar with ‘suburbs densely inhabited’; cAththar, port of San c ā’, capital of today’s Yemen; and Ghalāfiqa, port of Zabīd (Figure 5.1). What mattered in this inventory of coastal towns on the Ḥijāz and Tihāma Arabian coasts were wells and the availability of water. Thousands of wells exist in the Ḥijāz and many wells were found in the Yanbuc al-Nakhl region.50 Al-Muqaddasī did not fail to remark that among the other sea towns Yanbuc is ‘a splendid town’, blessed with ‘abundant water’.51 He not only documents in which port town wells and cisterns were found but notes

88  Dionisius A. Agius

FIGURE 5.1 Sea

towns identified in al-Muqaddasī’s tenth-century text.

Source: Drawn by Uma Bhattacharya.

the quality of water. Al-Qulzum has brackish but ‘drinking water’; it had to be fetched on boats or camels from ‘a place some six miles away called Suways’.52 At Ghalāfiqa, he says that there was ‘fresh water’ and Mukhā (Mocha) had ‘drinking water’. cAdan had many cisterns, although, as he reports, the water is ‘salty’.53 Watering stations for the pilgrim/merchant routes were found all along the coast as well as inland trails to Makka and Madīna, and al-Muqaddasī takes care to mention that some were well protected by forts. He records the economic benefits drawn from markets and the security they can offer to seafarers and ships’ crew; hence the mention of forts that kept the marauding desert Arabs at bay and the need to store safely the various cargoes. Some sea towns had granaries for grain brought by sea from Egypt set up in places such as Al-Jār and Yanbuc to provide food for the holy cities of Makka and Madīna as well as the pilgrims travelling by camel-caravan from the north or south of the Ḥijāz (that became in modern times the pilgrim railway trail). The importance of Al-Jār as a supplier of Egyptian grain appears to have occurred after famine had struck the Arabian coast in 634–635.54 Egypt continued to supply grain

Maritime landscapes of the Arabian Red Sea  89

for many centuries after this. Other sea towns such as Judda (Jeddah) and further south, Al-Sharja, al-Ḥirda and cAṭana, all with grain storages, suggest that the supply was coming from cAydhāb on the African coast or Yemeni ports.

Reflections on the tenth-century geographical knowledge systems The sections above cover salient points in placing al-Muqaddasī at the forefront of scientific inquiry in medieval Islam. His geography does not simply lie in the physical description of land and sea, mountains and rivers and towns and villages but, in general, it looks at how the people engage, directly or indirectly, in the environment they live in and occupy in the work they perform. He does not dwell on the different ḥadīths (sayings and deeds of the prophet Muḥammad) and legal practices (sharīca) of one community or the other but rather he describes the practicalities of the life of the ordinary people, how they dressed, what they ate, the language they spoke and their work, domestically or out in the street or fields. This he did with much detail, but when it came to the Arabian Red Sea landscape the information about human activity is almost not there. However, observations and comments on his involvement with maritime culture are worth noting: For a tenth-century merchant and pilgrim, he provides information as to where water can be accessed and whether the anchorage provides good markets and a resting place for travellers, camels and cargo. Further, it needs to be mentioned that as a merchant, he noted elsewhere imported and exported goods, commenting about the intangible culture through the varied customs and traditions, as well as the tangible culture of this tenth-century Islamic world. His description of buildings brings the places to life but in the case of the coastal landscape he fails to describe the forts or caravanserai of the sea towns he lists. In general, his approach to the physical and human inquiry resembles sections of the Periplus of the Erythraean Sea written in Koiné Greek by an anonymous Graeco-Roman merchant from Egypt in the first century Common Era.55 A periplus is ‘a voyage or journey round a coastline’;56 it includes, apart from imports and exports, a description of the coastal towns, navigation, geography, ethnography and trading opportunities, accompanied by maps.57 Of note, however, is that works of the Greek historians Herodotus (d. 426 BCE) and Thucydides (d. after 404 BCE) have sections that could be considered forerunners of the periplus genre and, indeed, the geographies of early medieval Islam. Herodotus’ Book II is about Egyptian and African history covering the geography of the regions with human activity, speaking often from first-hand experience and extensive sea and land travel.58 His contemporary, Thucydides, relied also on eye-witnesses accounts and his own experiences.59 Al-Muqaddasī and his contemporaries, it could be said, seem to have followed early Latin and Greek models, probably through translation, but the crucial difference is that the majority of Muslim geographers were concerned only with the world of Islam.

90  Dionisius A. Agius

Among these histories and periplus archetypes there are to be considered the early Islamic maritime manuals that are more or less contemporary to the Muslim geographies and, though embellished with anecdotes of the cajā’ib (marvels and wonders), genre of literature, they contain physical descriptions of maritime landscapes of different regions and information on human and economic activities on a number of Indian Ocean anchorages: These are the Silsilat al-tawārīkh (Chains of Narratives) with accounts collected by Sulaymān, the Sīrāfī merchant (early second/eighth century); while the second collection is the Akhbār bilād al-Hind wa al-Ṣīn (News of India and China), author unknown, though re-edited by a merchant Abū Zayd Ḥasan also of Sīrāf (fl. fourth/tenth century). Such works contain a wealth of data60 but need further investigation. They contain material that has been referenced by the medieval Muslim geographers Ibn Rusta (d. after 290/893–894), Ibn Khurradādhbih (d. c. 300/911) and Ibn Faqīh al-Hamadānī (fl. end of third/ ninth century).61 Returning to my last question: How significant is this information for the modern researcher? Al-Muqaddasī mentions shipwrecks at Fārān – ‘utter destruction of ships’ and at Jābir – ‘many a ship has been wrecked’.62 A number of land and underwater archaeological surveys have been conducted in the past 40 years, including those supported by the Saudi Commission for Tourism and National Heritage.63 A number of these surveys are inconclusive and excavations have not been complete. The location of landing places known to the geographer-traveller, the cisterns and wells and a number of forts, are now mostly hidden underground. For example, at El Wejh (north of Al-Ḥawrā’), one of the main anchorages from the Egyptian coast in the cAbbāsid (132–656/749–1258) and Fāṭimid (297– 567/909–1171) periods, no Islamic ruins have been found so far.64 Excavations at Al Jār revealed the ruins of a cistern and walls of a port with a wharf.65 Yanbuc is described by al-Muqaddasī as ‘well-fortified by a wall’. So far no foundation of a medieval wall has been excavated but efforts have been made to restore the early modern Sour district.66 Buried under this town, like other wealthy sea towns, are the medieval houses of merchants and ship owners, the wakālas or caravanserais where merchants stayed and perhaps left their cargoes in fear of the danger they might encounter in crossing the desert en route to the holy cities,67 just waiting to be discovered. Groundwater is an important feature; hence al-Muqaddasī’s mention of wells in several places and a study on their location and the causes for the water to dry out is necessary to understand the early medieval coastal and desert landscapes. Parallel to the Ḥijāz coastal route that al-Muqaddasī documents was the old desert pre-Islamic incense route,68 which became the pilgrim–trade route in the early years of Islam and then formed the route of the Ḥijāz railway at the beginning of the twentieth century. These routes were connected, and al-Muqaddasī’s information on wells and forts has relevance to the inland route. Wells and forts are landmarks of a coastal and desert landscape that tell a story of religious significance.69 Although al-Muqaddasī’s information

Maritime landscapes of the Arabian Red Sea  91

is terse and can be difficult to interpret, it may be possible to infer from his accounts the interchangeability of the two routes: water and protection being the essential needs for every traveller. Classical Roman and Greek literature record a number of ports on the Arabian Red Sea whose finds have been the subject of debate by archaeologists and maritime historians. The difficulty is to locate them against the list of ports provided by al-Muqaddasī, his contemporaries or later geographer and historian travellers. For example, the disputed Leukē Kōmē has been placed between Khurayba in the northern Ḥijᾱz, and Al-Wajh (El Wejh) and Yanbuc al-Baḥr in the south,70 or one other possibility is Al-Ḥawrā’, south of Al-Wajh, marked as the Roman port, according to Richard Burton (d. 1890).71 It is more likely that Al-Wajh was Leukē Kōmē, lying as it does opposite Myos Hormos,72 later Islamic Al-Quṣayr on the Egyptian coast. This southern location is mostly favoured by scholars, though the possibility of cAynūna in the north of the Ḥijᾱz has not been ruled out by some.73 International teams of terrestrial and underwater archaeologists would be required to excavate these sites fully and evaluate the finds. But archaeology alone cannot give answers to questions of historical facts and insights into the landscapes unless the sites and finds are verified against classical (Greek and Latin) and medieval Islamic texts such as those of the early geographers and those of the later period.74 This should be further strengthened by cartographic evidence such as the maps of al-Idrīsī75 and the anonymous Book of Curiosities (fifth/eleventh century),76 which offer unique insights to our understanding of coastal regions. Understanding the past by textual and archaeological evidence is not the complete answer to our scientific inquiry. It can only be achieved by supporting data with oral history, a technique I myself have used very successfully. In this more all-encompassing approach, underpinned by hard archaeology, we start to see patterns of continuity and change in the maritime landscapes and the material culture of the region.

Notes 1 From this genre I note here as background to the study, a few works by Muslim geographers of comparative importance in the pursuit of an inquiry into the physical nature of the maritime landscapes of peninsular Arabia. Travel experience formed the basis for al-Yacqūbῑ’s (d. 284/897) geography in his work, al-Tārīkh (The History); it is based on uninterrupted voyages in which he reports on the lands he saw and the people he interviewed, verifying the oral reports with the best and reliable authority on the subject; while Ibn Khurradādhbih’s (d. c. 300/911) Kitāb al-Masālik wa l-mamālik (The Book of Routes and Provinces) relied on auditory reports from the mercantile community and seafarers. Two other works by al-Iṣṭakhrī (d. c. 350/961), the Kitāb masālik al-mamālik (The Book of the Routes of the Provinces) and Ibn Ḥawqal (d. c. 380/990), the Kitāb ṣūrat al-arḍ (The Book of the Configuration of the Earth), are credited with advancing human knowledge about lands and people by observation and hearing; producing maps of the Mediterranean, the Red Sea, the Persian Gulf and what is known today as the Indian Ocean, showing the coastal towns and in some parts, highlighting the trade towns

92  Dionisius A. Agius

2 3 4 5 6

7

8 9

inland. Ibn Ḥawqal travelled extensively, and when revising al-Iṣṭakhrī’s geography, added much more material on economic matters to his work. See al-Yacqūbῑ, Kitāb al-Buldān, trans. Gaston Wiet, Cairo, 1937, p. 2; al-Iṣṭakhrī’s map of the Sea of Persia, in Kitāb masālik al-mamālik, ed. M. J. De Goeje, Leiden: Brill, 1927, pp. 28–29; and in Ibn Ḥawqal’s Kitāb ṣūrat al-arḍ, Beirut, 1992, p. 50; see also Houari Touati, Islam and Travel in the Middle Ages, trans. Lydia G. Cochrane, Chicago and London: University of Chicago Press, 2010, p. 130. Also known in the Islamic world as al-Maqdisī. Al-Muqaddasī, The Best Divisions for Knowledge of the Regions, trans. Basil Collins and rev. Mohammad Hamid Alta’i, Reading: Garnet, 2001, pp. xxiii–xxiv. Al-Muqaddasī, The Best Divisions for Knowledge of the Regions, p. 1. Al-Muqaddasī, The Best Divisions for Knowledge of the Regions, p. 8. See for example an inventory of terminology on a varied number of objects of domestic use: household furniture, kitchen utensils, textiles, weights, and measures, inns and markets and cultivated plants, al-Muqaddasī, Aḥsan al-taqāsīm fī macrifat al-aqālīm, M. J. De Goeje (ed.) [Bibliotheca Geographorum Arabicorum, III], Leiden: Brill, 1906; first edited 1877, pp. 30–32. Al-Muqaddasī also includes a list of sea and river craft, see Dionisius A. Agius, ‘Historical-Linguistic Reliability of Muqaddasī’s Information on Types of Ships’, in Dionisius A. Agius and Ian Richard Netton (eds.), Across the Mediterranean Frontiers: Trade, Politics and Religion, 650–1450, Turnhout: Brepols, 1997, pp. 303–329. Salah Zaimeche, ‘Al-Muqaddasi and Human Geography: An Early Contribution to Social Sciences’, http://muslimheritage.com/article/al-muqaddasi-and -human-geography-early-contribution-social-sciences (accessed 29 March 2019); see also Dionisius A. Agius, ‘River Craft and Ships of Medieval Iraq: al-Muqaddasī’s Account’, in Anne-Marie Eddé and Emma Gannagé (eds.), Regards Croisés sur le Moyen Age arabe. Mélanges à la memoire de Louis Pouzet s.j. (1928–2002), Mélanges de l-Université Saint-Joseph, 2005, 58: 392–393. Dionisius A. Agius, Classic Ships of Islam: From Mesopotamia to the Indian Ocean, Leiden: Brill, 2008, p. 15–16. Further details on his methodology and geographic inquiry, see the ‘Introduction’ to Al-Muqaddasī, The Best Divisions for Knowledge of the Regions, pp. xxiv–xxix. In the introductory statement of his book, al-Muqaddasī claims that ‘the scholars who formerly dealt with the sciences wrote the initial works: then their successors in turn commented on and summarized their works. Accordingly’, he continues, it occurred to me to direct my attention to a science which [my predecessors] had neglected, and to specialize in a branch of learning they had not dealt with, except defectively – I mean an account of the Islamic regions, with the deserts and the seas in them … (p. 1).

10 See Houari Touati, Islam and Travel in the Middle Ages, p. 114. 11 Al-Mascūdi, Murūj al-dhahab wa ma c ᾱdin al-jawhar, trans. Charles Barbier de Meynard and Pavet de Courteille as Les prairies d’or, volume 1, Paris: Imprimerie Impériale, 1861–1917, p. 9; see Houari Touati, Islam and Travel in the Middle Ages, p. 123. 12 The route to China had its dangers and many a story includes descriptions of strange creatures and folk belief in spirits haunting the coastal landscape and seascape. Such stories of the marvels and wonders were common among travellers and some are also found in works of geography. 13 Buzurg Ibn Shahriyar, The Book of the Wonders of India: Mainland, Sea and ­Islands; trans. and ed. G. S. P. Freeman-Grenville, London and Hague: EastWest Publications, 1981, p. 50. 14 See Suhanna Shafiq, Seafarers of the Seven Seas: The Maritime Culture in the Kitāb cajā’ib al-Hind by Buzurg Ibn Shahriyār (d. 399/1009), Berlin: Klaus Schwarz, 2013, pp. 64–65.

Maritime landscapes of the Arabian Red Sea  93

15 Al-Mascūdi, Murūj al-dhahab, volume I, p. 94; al-Muqaddasī, The Best Divisions for Knowledge of the Regions, p. 8; al-Muqaddasī, Aḥsan al-taqāsīm fī macrifat al-aqālīm, p. 43; see also Suhanna Shafiq, Seafarers of the Seven Seas, p. 20. 16 Houari Touati, Islam and Travel in the Middle Ages, p. 119, fn. 55. 17 Al-Muqaddasī, The Best Divisions for Knowledge of the Regions, p. 9. 18 It is an oblong-shape headdress: ‘a shawl-like garment worn over head and shoulders’, see Edward W. Lane, Arabic-English Lexicon, volume II, Cambridge: Islamic Texts Society, 1984, p. 1867; also Hans Wehr, A Dictionary of Modern Written Arabic, edited by J. Milton Cowan, Wiesbaden: Otto Harrassowitz, 1966, p. 580. 19 Al-Muqaddasī, The Best Divisions for Knowledge of the Regions, pp. 10–11. 20 Al-Muqaddasī, The Best Divisions for Knowledge of the Regions, p. 9. 21 Dionisius A. Agius, The Life of the Red Sea Dhow: A Cultural History of Seaborne Exploration in the Islamic World, London: I. B. Tauris, 2019, pp. 124–134. 22 Seán McGrail, Boats of the World. From the Stone Age to Medieval Times. Oxford: Oxford University Press, 2004, pp. 145–148. 23 See A. Raymond, Artisans et commerçants au Caire au XVIII siècle, Volume 1, Damascus: Institut Franϛais de Damas, 1974, p.115. 24 Agius, The Life of the Red Sea Dhow, p. 144. 25 Table contents cited from al-Muqaddasī, The Best Divisions for Knowledge of the Regions, p. 11. 26 Bad seamanship was noted by the Andalusian traveller Ibn Jubayr (d. 614/1217– 1218); he reports the sailing gear was faulty ‘which time and again became entangled and broke when sails were raised or lowered …’, Ibn Jubayr, The Travels of Ibn Jubayr, trans. R. J. C. Broadhurst, New Delhi: Goodword, 1952, p. 70. 27 Al-Muqaddasī, The Best Divisions for Knowledge of the Regions, p. 13. The Arabic name for ‘pirates’ is bawārij, see al-Muqaddasi, Aḥsan al-taqāsīm fī macrifat ­al-aqālīm, p. 14. The historian al-Balādhurī (d. 279/892) applies the name to mean ‘deep-sea ships’, recorded in an attack on Jeddah in 151/768, see al-Balādhurī, Liber expugnationis regionum (Kitāb futūḥ al-buldān), edited by M. De Goeje, Leiden: E. J. Brill, 1866, p. 440; al-Mas cūdī, Kitāb al-tanbīh wa l-ishrāf, edited by M. J. De Goeje [Bibliotheca Geographorum Arabicorum], VIII], Leiden: E. J. Brill, 1894, p. 55. It seems that the name for ships became associated with pirates, hence the term employed by al-Muqaddasī; see Agius, Classic Ships of Islam, p. 329. 28 Strabo notes Nabatean pirate rafts attacked vessels approaching Egyptian ports, see The Geography of Strabo, with Greek text, trans. H. L. Jones, Volumes I–VIII, London and New York: William Heinemann and G. P. Putnam, 1917, Bk. 16.4.18. 29 Robert G. Hoyland, Arabia and the Arabs from the Bronze Age to the Coming of Islam, London and New York: Routledge, 2003, pp. 40–42, 46–49. 30 The Periplus Maris Erythraei, trans. and commentary Lionel Casson, Princeton, NJ: Princeton University Press, 1989, 20:7. 6–11. 31 Ralph K. Pedersen, ‘A Preliminary Report on Coastal and Underwater Survey in the Area of Jeddah, Saudi Arabia’, American Journal of Archaeology, 119 (1), 2015: 128. 32 Al-Muqaddasī, The Best Divisions for Knowledge of the Regions, p. 71. 33 Agius, The Life of the Red Sea Dhow, p. 146. 34 The information in the table is extracted from Al-Muqaddasī, The Best Divisions for Knowledge of the Regions, pp. 63–77. 35 Al-Muqaddasī, The Best Divisions for Knowledge of the Regions, p. 11. 36 Al-Muqaddasī, The Best Divisions for Knowledge of the Regions, p. 9. 37 Gerald R. Tibbetts, ‘Arab Navigation in the Red Sea’, The Geographical Journal, 127 (3), 1961: 324, Fig. 1. 38 Al-Muqaddasī, The Best Divisions for Knowledge of the Regions, p. 11.

94  Dionisius A. Agius















­

Maritime landscapes of the Arabian Red Sea  95

61

62 63

64 65 66 67 68

69 70 71 72

73

74

à la Chine, trans. J. Toussaint Reinaud, Volumes I–II, Paris: Imprimerie Royale, 1845. The importance of comparative study of the works of early Muslim authors were pointed out by Gabriel Ferrand, see Voyage du marchand arabe Sulaymân en Inde et en Chine, trans. Gabriel Ferrand, Paris: Bossard, 1922, pp. 13–14; see also Gerald R. Tibbetts, A Study of the Arabic Texts Containing Material on SouthEast Asia, Leiden-London: Brill, 1979, pp. 5–6, 8–9. Al-Muqaddasī, The Best Divisions for Knowledge of the Regions, p. 11. See Cheryl Ward and Uzi Baram, ‘Global Markets, Local Practice: Ottoman-Period Clay Pipes and Smoking Paraphernalia from the Red Sea Shipwreck at Sadana Island, Egypt’, International Journal of Historical Archaeology, 10 (2), June 2006: 137; A. Al-Mughannam, S. Al-Helwa and J. Mursi, ‘Catalogue of Stations on the Egyptian (Coastal) and Syrian (Inland) Pilgrim Routes’, Al-Atlal, Journal of Saudi Arabian Archaeolog, 7, 1983: 42–49. Ralph Pedersen and his team have undertaken a survey in 2012, see Ralph K. Pedersen, ‘A Preliminary Report on Coastal and Underwater Survey in the Area of Jeddah, Saudi Arabia’, American Journal of Archaeology, 119 (1) January 2015: 125–136. Alî Ibrâhîm Al-Ghabbân, Les deux routes syrienne et égyptienne de pèlerinage au nord-ouest de l-Arabie Saoudite, Volumes I–II, Cairo: Institut Français d’Archéologie Orientale, 2011, Volume I, p. 193. Alî Ibrâhîm Al-Ghabbân, Les deux routes syrienne et égyptienne de pèlerinage au nord-ouest de l-Arabie Saoudite, Volume I, p. 355, Figs. 141, 142 a–b. Hisham Mortada, ‘A Traditional Settlement in a Modernized City: Yanbu, Saudi Arabia’, International Journal of Architectural and Environmental Engineering, 10 (10), 2016: 1258–1263. See comments on this subject in Richard Burton, A Secret Pilgrimage to Mecca and Medina, London: Folio Society, 2004, p. 151. See for example Isaiah 60: 6; Jeremiah 6: 20; Alessandro de Maigret, Arabia Felix: An Exploration of the Archaeological History of Yemen, London: Stacey International, 2009, p. 29. The Nabateans acted as middle men during the third century CE. The Periplus mentions ‘ships … loaded from Arabia’ came to the harbour of Leukē Kōmē on the Arabian Red Sea, ‘from which there is [a route inland] to Petra’ (The Periplus of the Erythraean Sea, trans. and ed. G. W. B. Huntingford, London: The Hakluyt Society, 1980, Ch. 19, p. 21), which, though not specifically stated, would have carried incense to be delivered to Syria and Mesopotamia. Aylin Orbaşli and Simon Woodward, ‘A Railway “Route” as a Linear Heritage Attraction: The Hijaz Railway in the Kingdom of Saudi Arabia’, Journal of ­Heritage Tourism, 3 (3), 2008: 159–175. R. K. Pedersen, ‘A Preliminary Report on a Coastal and Underwater Survey in the area of Jeddah, Saudi Arabia’, p. 126. Timothy Power, The Red Sea from Byzantium to the Caliphate AD 500–1000, Cairo and New York: American University in Cairo Press, 2012, p. 112. Caroline Durand, ‘Crossing the Red Sea: The Nabataeans in the Egyptian Eastern Desert’, in Dionisius A. Agius, John P. Cooper, Athena Trakadas, Chiara Zazzaro (eds.), Navigated Spaces, Connected Places, [Red Sea Project V], [BAR International Series 2346], Oxford: Archaeopress, 2012, p. 88. Michal Gawlikowski, ‘Looking for Leuke Kome’, in Andrea Manzo, Chiara Zazzaro and Diana Joyce de Falco, Stories of Globalisation: The Red Sea and the Persian Gulf from Late Prehistory to Early Modernity, [Red Sea Project VII], Leiden: Brill, 2019, p. 288. See for example, Ibn al-Mujāwir, A Traveller in Thirteenth-Century Arabia: Ibn al-Mujāwir’s Tārīkh al-Mustabṣir, trans. and annotated G. Rex Smith, London: The Hakluyt Society, 2007; Abū l-Fidā’, Géographie d’Aboulféda (Taqwīm ­al-buldān), edited by M. Reinaud and M. Le Baron Mac Guckin de Slane, Paris:

96  Dionisius A. Agius

Imprimerie Royale, 1840; al-Nuwayrī l-Iskandarānī, Kitāb al-ilmām bil-iclām fī mā jarat bihi l-aḥkām wa l-umūr al-muqḍiyya fī waqcat al-Iskandariyya, Volumes I–VII, Hyderabad: Dā’irat al-Macārif al-Uthmāniyya, 1968–76; Al-Maqrīzī, AlMawāciẓ wa l-ictibār fī dhikr al-khiṭaṭ wa l-āthār, edited by G. Wiet, Volumes I–II, Cairo: Institut Français d’Archéologie Orientale, 1911–1924. 75 Al-Idrīsī, Nuzhat al-mushtāq fī ikhtirāq al-āfāq (Géographie), trans. P. A. Jaubert, Volumes I–II, Paris: Imprimerie Royale, 1836–1880. 76 An Eleventh-Century Egyptian Guide to the Universe, trans. and ed. Yossef Rapoport and Emilie Savage-Smith, Leiden: Brill, 2014.

PART 2

The materiality of knowledge production Introduction Himanshu Prabha Ray

The chapters in the previous section examined the many conceptualizations of the ocean as evident from textual accounts. How are these representations in literature to be understood and what impact did this knowledge have on the ground? In recent writings on the eighteenth and nineteenth century, global interconnections have often been emphasised as a means of stimulating new knowledge and learning. Trade is credited with creating channels of communication, and along with commodities, providing opportunities for scholars, religious leaders and others to travel. At times traders themselves are described as agents who negotiated within several cultural domains and whose mediations gave rise to new intellectual material and technical practices. Kapil Raj argues that a new development in the second millennium CE was the emergence of cross- cultural mediation as a specialised activity that was autonomous from trading communities. One of the intermediary types suggested by Raj is that of the knowledge broker. Knowledge mediation, Raj argues, gained increasing status throughout the eighteenth century, especially as a result of sustained European presence in the Indian Ocean region and eventual colonization.1 Elsewhere I have demonstrated that this formulation does not take into account the development of new disciplines such as archaeology and oceanography in the eighteenth and nineteenth centuries in Asia. These disciplines were unconnected both with trading activity and with the development of intermediary knowledge brokers.2 Both these disciplines were instead related to understanding the past of Indian Ocean regions such as South and South East Asia and were connected with establishing and controlling new

98  Himanshu Prabha Ray

routes of maritime communication in the colonies and new knowledge systems for understanding the past. Early ‘archaeologists’ in India such as the first Director-General of the Archaeological Survey of India, Alexander Cunningham (1814–1893) was a military man and not a scholar who had travelled along trade routes to bring new knowledge to the colony. It is no coincidence that interest in the material heritage of South East Asia corresponded with British control, starting with the occupation of Penang in the Straits of Malacca by Captain Francis Light in 1786. The primary objective of James Rennell (1742–1830) considered the father of oceanography and best known for his Map of Hindoostan was not a search for knowledge, but of propagation of British colonial achievements to his countrymen back home. In 1760 Rennell secured a place as a midshipman on the frigate America and sailed from England for Madras. After reaching Madras he transferred to another ship and took part in the five-month naval blockade of Pondicherry by the British fleet comprising of 16 ships. While in the navy he lost no opportunity to survey the places at which the ships anchored and acquired the reputation of an enthusiastic and diligent surveyor. Clearly there is a need to delink trade from knowledge generation both in the eighteenth and nineteenth centuries as well as in the ancient period. The last six to seven decades have coincided with an expansion of archaeological techniques and methods and a vast corpus of archaeological data is now available for understanding island colonization and the use of watercraft to cross the seas.3 Recent writings date the beginnings of maritime activities in the Indian Ocean to the prehistoric period long before the existence of Empires and emphasise that these networks dealt with food stuffs and staples, such as cloth, medicinal plants and so on.4 Seafaring was handled by mobile small-scale seafaring communities travelling across the seas, rather than by elite merchant groups at the behest of states or empires, as briefly mentioned in the Introduction to the first Section. How are these small-scale mobile communities to be factored into a discussion of cross­cultural interactions? Archaeobotanical data supports the transfer of grain, processed fish, oil and wine and containers of spices across the ocean in third and second millennia BCE.5 Austronesian voyagers were known to have taken dogs, chickens, banana, yam and taro as they settled the islands of the Pacific. Indian craftspeople transplanted native crops across the Bay of Bengal in mainland South East Asia by the end of the first millennium BCE.6 Experts suggest that colonisation was one of the factors that led to introduction of new crops, as is evident from examples from more recent times when several plants were introduced by Europeans in the New World. Religious institutions also facilitated the movement of certain ritual items. For example, black pepper (Piper nigrum), native to southwestern India, was used in Egypt in the Bronze Age, as evident from its occurrence in the mummy of Ramses II, placing its date of import at around 1200 BCE.7 Another remarkable find of pepper was made at Berenike on the Red Sea coast where

The materiality of knowledge production  99

two large terracotta jars made in India were recovered from the courtyard floor of the first century CE temple of the Graeco-Egyptian god, Serapis. While one of them was empty, the other contained 7.55 kg of black peppercorns. The nature of religious ceremonies where pepper may have been used is not clear, nor are there any explanations in written accounts. The areca nut palm was probably introduced to South Asia around the mid-second millennium BCE, from South East Asia, together with several other South East Asian crops such as sandalwood, lime cultivars, ginger and galangal. Loan words for crops and nautical terms from Malayo–Polynesian languages have been detected, not only in the Andaman–Nicobar chain, but also in India, Sri Lanka, Maldives and east Africa, strongly suggestive of maritime crossings in antiquity.8 Despite the data to the contrary showing travel by diverse groups comprising sailors, adventurers, religious specialists and so on, the paradigm favouring trade activity remains dominant. This dependence of Southeast Asian societies and their economies on trade in raw commodities—in proto-historic as well as in modern times—no doubt reinforced their image of passive performers in earlier historical representations. This dissymmetry (manufactured goods against raw commodities) will remain one major characteristic of exchange patterns in the Indian Ocean (or the South China Sea).9

Much of the secondary literature continues to accept trade as a motivating factor based on finds of imported ceramics, beads and other objects in the archaeological record, though the context of the finds is seldom considered. K. Rajan, for example, argues for the use of Tamil Brahmi engraved on memorial stones, seals, rings and on ceramics as a ‘prerequisite to trade’. ‘This linguistic uniformity would not have been achieved without minimum continuous cultural interaction. The striking similarity observed in various Iron Age monuments clearly points to the emergence of cultural homogeneity’. Rajan thus conflates Tamil language with cultural homogeneity of the Iron Age monuments leading to ‘resource transactions through trade’.10 Singularly absent in these secondary writings is any discussion of the organization of trade and the social actors involved in resource mobilisation. Rigorous archaeological analysis of data from sites in the Tungabhadra River basin in north Karnataka, which was one of the more densely populated areas of the south starting from the prehistoric period onward, shows the creation of symbolically differentiated spaces in the third and second millennium BCE. This was further consolidated in the first millennium BCE when the practice of construction of megalithic monuments was in existence largely through new forms of large-scale labour mobilization and differential access to land and water resources. Analysis of surface remains over a long period from the third millennium BCE to 500 CE shows continuities in social differentiation but also changes in the institutionalization of earlier practices of privileged access.11

100  Himanshu Prabha Ray

Social differentiation as it emerged in small-scale Iron Age communities and the participation of these groups in cross- cultural contacts has larger implications for understanding and defining so- called ‘imports’ in the archaeological record both in South and South East Asia. For example, there is evidence for the earliest appearance of iron as well as onyx beads (of South Asian origin) at Taungthaman, Myanmar, in the fifth century BCE, and the appearance of the same two artefact types at Ban Chiang, Thailand, and in the Tabon caves in Palawan, the Philippines. As discussed elsewhere, the shared culture that extended across not only South Asia, but also the Indian Ocean from fourth century BCE to fourth century CE was part of a literate tradition, which was by no means controlled by the ruler or the brahmana, but included Buddhist and Jain monks, navigators and trading and crafts groups. Writing facilitated storing of information, and cumulative knowledge promoted a new genre of cultural and artistic expression and aided ordering of information under numeric and alphabetic heads and the use of maps. These networks may be identified in the archaeological record by specimens of writing on pottery, seals and sealings and by inscriptions on stone and copper plates, though there were further shifts and transformations in these.12 An underutilised primary source for the spread of cultural knowledge is data from writing on pottery or inscribed potshards. ‘Wherever Buddhism travelled, it fostered intimate associations with the written word’.13 The occurrence of inscribed shards at major centres along the east coast in the subcontinent is of relevance, as it helps define the nature of maritime networks. Thus, it is important to trace the social lives of the objects found in archaeological contexts for a comprehensive understanding of cross- cultural interactions. It is the three issues of (1) cultural homogeneity of the Iron Age megalithic monuments, (2) uniformity of transactions at so- called port sites in south India and (3) writing being only a tool for economic transactions that the chapters in this section challenge.

Notes 1 Kapil Raj, ‘Mapping Knowledge Go-betweens in Calcutta, 1770–1820’, in Simon Schaffer, Lissa Roberts, Kapil Raj and James Delbourgo (eds), The Brokered World: Go-Betweens and Global Intelligence, 1770–1820 (Sagamore Beach: Watson Publishing International, 2009), pp. 105–150. 2 Himanshu Prabha Ray, ‘Voyages of ‘Discovery’: Mapping the Bay of Bengal’, in Satish Chandra and Himanshu Prabha Ray (eds), The Sea, Identity and History: From the Bay of Bengal to the South China Sea (Delhi: Manohar Publishers, 2013), pp. 303–328. ­ 3 Himanshu Prabha Ray, Ray, Maritime Archaeology of the Indian Ocean, in David Ludden (ed.), Oxford Research Encyclopedia: Asian History (Oxford: Oxford University Press, 2016). Himanshu Prabha Ray, The Archaeology of Seafaring in Ancient South Asia (Cambridge: Cambridge University Press, 2003). 4 Himanshu Prabha Ray, ‘Writings on Maritime History of Ancient India’, in Sabyasachi Bhattacharya (ed.), Approaches to History: Essays in Indian Historiography (New Delhi: ICHR and Primus Books, 2011), pp. 27–54.

The materiality of knowledge production  101



­

­

​­

6 MAKING MEGALITHS AND CONSTITUTING COLLECTIVES Politics, places and historicity in prehistoric South India Andrew M. Bauer and Peter G. Johansen

Introduction The ‘past’ is a political resource. From the history of the Scottish kilt to contemporary American celebrations of Columbus, historical scholars have demonstrated that ostensibly deeply rooted cultural traditions and their associated historicities often selectively appropriate actors, actions and materials to facilitate the constitution of particular social collectives— producing narratives and practices that have considerable political salience as the bases for a variety of social identities, nationalist ideologies, counterclaims, or territorial disputes.1 As the late historian and anthropologist Michel–Rolph Trouillot has noted, the dual ability to affirm and silence is at the core of how the production of history is itself a critical dimension of the sociohistorical process; that which is ‘said to have happened’ recursively affects that which happens.2 It would be incorrect to assume that inhabitants of bygone social and cultural contexts understood temporal relations in a manner that is analogous to western historicism.3 Nevertheless, archaeologists have cogently called attention to how architectural forms and materials were deliberately fashioned in a range of cultural contexts as part of political strategies to influence perceptions of the temporal depth of particular empires or communities.4 To be sure, the materiality of landscape is a significant factor in shaping both how people understand temporal relations and the effectiveness of any claims about those relations: ‘What happened leaves traces, some of which are quite concrete—buildings, dead bodies…monuments… —that limit the range and significance of any historical narrative’.5 This chapter focuses on some of the ‘traces’ that comprise the archaeological landscape of the Deccan region of South India to investigate the intersections of prehistoric memorialization practices and politics during a

Making megaliths and constituting collectives  103

period of emergent social differentiation in an interior context that flanked the Indian Ocean. More specifically, we examine ­Neolithic Period and Iron Age mortuary practices and the development of a class of archaeological features that are typically glossed as ‘megalithic’ monuments in South Asian archaeology, and throughout Eurasia and northeastern Africa more generally (­Figure 6.1).6 Although these features have been the subject of a long history of scholarship on India, only recently has there been a concerted effort to understand the production of megalithic monuments—and the enduring places that they partly constituted—as elements of a broader archaeological and social context. By drawing on our recent research at the site of Maski (Raichur District, Karnataka), and contextualizing that evidence within other detailed studies of the variation and spatial distribution of megalithic features, we will suggest that the practices of creating these monuments established social collectives and ultimately articulated temporal relations to significant material and symbolic resources that were critical to the constitution

­FIGURE 6.1 

 eneralised distribution of megalithic monuments in South India, G and the location of the Maski Archaeological Research Project and Hire Benakal.

Source: District boundaries and underlying geographic data sourced from ESRI Online

104  Andrew M. Bauer and Peter G. Johansen

of social affiliations and differences—producing and signifying distinct kin, community or residential associations with places. Pushing beyond Trouillot and others we will argue that these monumental traces did not simply ‘set the stage’ for a politics of the past or the production of what many archaeologists and historians have often termed ‘social memory’; rather, they were part-and-parcel of the generation of collectivities and historicities, simultaneously calling both into being.7 To develop this argument we begin by briefly reviewing previous archaeological approaches to studying South Indian megalithic features.

Contexts of scholarship and megalithic contexts: a history of research In an early issue of Ancient India, Gordon Childe, then Director of the Institute of Archaeology at the University of London, published a generalised description of megalithic monuments constructed of ‘large slabs or blocks of stone’ that had been recorded in parts of northern Africa, western Europe and the Caucuses to highlight their ‘kinship’ with those that are ‘abundant in the Deccan’ of southern India.8 By the time of his writing, megalithic features of South India had become well known to both colonial and Indian scholars. In the Nizam’s Dominion of Hyderabad, for instance, the initial publications of the Journal of the Hyderabad Archaeological Society in 1916 would devote the highest number of pages to ‘Hyderabad Cairns’, and in the subsequent issue the Director of the Department of Archaeology of the Nizam’s territory, G. Yazdani, would himself pen an essay on the ‘Megalithic Remains of the Deccan’ to highlight an important avenue of research in the state’s burgeoning efforts to develop archaeological research and a particularised regional heritage.9 Fundamental to much of this early scholarship on South Indian megaliths was a concern for how the monuments of the Deccan were both similar and distinct from those that were found elsewhere in Eurasia and Africa. As a consequence, considerable emphasis was placed on developing and refining typologies of megalithic forms, which would subsequently become an important archaeological index of the ‘Megalithic Period’ in cultural-h istorical paradigms that became prevalent in South Asian archaeology.10 The term ‘megalith’ is now used in South Asian archaeology to refer to a general suite of monuments that include substantial stone elements. As a class of features they are far from uniform. They frequently involve the combination of a number of different kinds of architectural components (e.g. unmodified cobbles, percussion-shaped boulders, quarried stone slabs) that are arranged in a variety of forms that have been classified into numerous morphological types, including dolmens, cairns, rectilinear cists, menhirs, stone circles and even boulder-filled crack features.11 These forms are also often variously combined, such that identifying any monument as a singular ‘type’ often elides significant variation. For instance, Bauer has recognised dolmens that were enclosed by stone circles, which also included menhirs as

Making megaliths and constituting collectives  105

elements—meaning a single ‘monument’ simultaneously combined dolmen, stone circle and menhir forms.12 While most of these features are probably mortuary monuments, having been shown to contain subsurface burial cists with single or multiple interments, it is also the case that some of them do not have associated burials. Thus, to discuss ‘megaliths’ in general terms glosses over significant variability in both ‘form and function’ of a class of archaeological features that have been treated as similar. In fact, systematic analyses of monuments that even fit into a single ‘type’ category, such as the dolmens found at the site of Hire Benakal—one of the few South Indian megalithic complexes that has been fully mapped and recorded in detail— show remarkable variation in construction elements and size.13 The great variation in megalithic monuments of South India has been a central focus of prehistoric archaeology in the region for nearly 200 years. Early colonial scholars were keenly interested in the cairns, stone circle cists and dolmens of southern India largely because of their apparent similarity and ‘kinship’ to features that could be found in Europe, and especially the Caucasus, where they represented important lines of evidence in ethnolinguistic narratives of ancient history that foregrounded transcontinental migration and diffusion.14 Starting from that point, Bauer and Trivedi sketched the history of research on Indian megaliths in four principal periods.15 The first, associated with colonial scholarship, operated with Indo-European linguistic theories of ancient kinship and migration and ‘attributed these monuments to ancestral groups of people that informed contemporary British antiquarianism’.16 This work was exemplified by the nineteenth- century studies of Meadows–Taylor, but also underlies the later concerns of Childe for the ‘kinship’ of South India’s megaliths with those of northern Africa and Europe.17 ­Cultural-historical ​­ paradigms developed by the ­mid-twentieth ​­ century and archaeologists working in India would come to attribute megalithic monuments to a ‘Megalithic Period’ that was situated between that of the preceding Neolithic Period and the subsequent Early Historic Period.18 In more recent years, however, it has been demonstrated through chronometric dating techniques that megalithic mortuary features were produced during the Iron Age (ca. 1200–300 BCE) as well as during the Neolithic (ca. 3000–1200 BCE) and Early Historic Period (ca. 300 BCE–CE).19 Thus, ‘the Megalithic Period’ is a misnomer; and it is more accurate to consider the production of megalithic monuments as a variety of cultural practices that were evident in multiple archaeographic periods. Related to their concern for periodization, cultural-historical approaches still largely relied on migration models and diffusion to explain change in the archaeological record. The historical association of cultural ‘traits’ to specific groupings of people led to the identification and categorization of a range of megalithic ‘types’, which did point to regional differences in megalithic building practices. As a consequence, some scholars called attention to limitations of using the presence of megalithic monuments as a homogenous cultural trait that could be linked to a singular group or ‘wave’ of migratory people in

106  Andrew M. Bauer and Peter G. Johansen

India. For instance, Sundara suggested that regional differences in monument forms, such as the prevalence of dolmens and ‘port-hole’ cists in some areas of the southern Deccan, might reflect past cultural or ethnic differences in the archaeological record, perhaps resulting from underlying environmental or geographical determinants.20 According to Bauer and Trivedi, the third phase of scholarship departed from the cultural-historical emphasis of prior work by adopting an explicit concern of the ‘New Archaeology’ for a social evolutionary perspective on megalithic monuments and burial practices. In this context, researchers’ concerns became less about the diffusion and migration of ancient people and more about how the burial record could be used to make inferences about the social organization of the ancient societies that produced them. This development followed rather directly from a broader archaeological trend to conceptualise mortuary datasets as reflections of ancient social relationships more generally.21 In the context of South Asia, for instance, Moorti summarised a corpus of previous scholarship on megaliths and pointed to disparities in megalithic mortuary architecture and grave goods (e.g. ceramics, beads, iron tools, iron weapons and copper and bronze objects and adornments) to argue that socially stratified societies emerged during the period of their production.22 A comparable concern for social organization is also evident in the work of Brubaker, who synthesised evidence for megalithic burial inequalities while also aggregating the number of monuments at megalithic mortuary complexes as a proxy for settlement site size to ultimately infer that Iron Age societies most closely approximated regional ‘chiefdoms’ according to site rank-size models.23 The evolutionary approaches and methods espoused in these studies can be critiqued on a number of epistemological grounds,24 but they nevertheless arrived at some conclusions that have considerable merit. For instance, megalithic monuments do vary significantly in size and the mortuary record documented by previous scholarship points to vast disparities in the mobilization of labour in mortuary and commemorative rituals (Figure  6.2).25 It is evident that not all prehistoric inhabitants were commemorated with mortuary monuments involving the labour-intensive construction of stone and masonry architecture, a point that is attested by the presence of unelaborated pit burials alongside megalithic monuments at some Iron Age sites.26 As Bauer has summarised: while the evidence for such disparities in mortuary treatment cannot be taken as a simple reflection of Iron Age social organization and status differences, it nevertheless implies that new modes of labour mobilization emerged during the period and that not all inhabitants participated in the system equally.27 Yet, it is important to stress that the results of these early efforts to infer social organization from the South Indian megalithic record ‘relied heavily upon data from older generations of scholarship, which did not lay the same

Making megaliths and constituting collectives  107

­FIGURE 6.2 

 xamples of some ‘megalithic’ forms and elements documented E in the region of Hire Benakal and Maski: (A) c­ obble-​­and-boulder supported dolmen; (B) s­ lab-​­supported dolmen; (C) partitioned rectilinear cist; (D) boulder enclosure; (E) anthropomorph supporting element; (F) p ­ assage-​­chamber feature; (­G-​­H) boulder stone circles.

Source: Photographs by authors

emphasis on detailed enumeration, statistical investigation, or on recording variability systematically’.28 Indeed, much of this earlier data has now been shown to be heavily biased toward large, salient monuments at the expense of understanding other dimensions of the archaeological record that were likely produced contemporaneously with that of megalithic sites. The overemphasis on such monuments even led many scholars to wrongly suggest that other kinds of sites, such as Iron Age settlements, did not even exist in the archaeological record.29

108  Andrew M. Bauer and Peter G. Johansen

The most recent approaches to the study of megaliths in South India have sought to produce datasets that allow for rigorous statistical analyses and to contextualise megalithic sites as specific kinds of meaningful places within a broader cultural landscape of social activities related to a range of settlement, ritual and quotidian practices.30 For instance, Johansen has pointed to a number of contexts where it appears that megalithic complexes were deliberately constructed in association with ritualised piles of burned cattle dung, or ‘ash mound’ features, of the previous Neolithic Period as an effort to establish a social relationship to this valued substance.31 Yet, intensive study of megalithic sites, and the broader archaeological landscapes of which they are part, remains relatively nascent in South India. Only a handful of systematic and intensive archaeological surveys have sought to identify megalithic sites alongside other evidence for prehistoric occupation by recording prehistoric settlements and less salient elements of the archaeological record, such as ephemeral occupations, and even ‘off-site’ artefact distributions. Although there is evidence that systematic survey methods are beginning to be adopted more broadly,32 ‘village-to-village’ ­ ​­ survey remains the dominant mode through which archaeological sites in India have been documented. As a consequence, salient megalithic monuments are still frequently discussed and described with little concern for less obtrusive constituents of the archaeological record, much to the detriment of interpreting the social significance of even the large and salient megalithic sites.33 Moreover, there also remains a dearth of detailed spatial maps of monuments and forms even within single sites, some of which contain more than 1,000 features. Partly as a result of the historical research bias toward describing obtrusive megaliths, until recently there has also been very little attention to excavating Iron Age settlements and documenting a suite of quotidian consumption and production practices of the ancient inhabitants who constructed megaliths.34 Although more detailed assessments of prehistoric landscapes and settlement activities are relatively new in South India and comparatively few, a number of general conclusions can already be drawn about a range of cultural and social practices associated with megalithic production. Perhaps most importantly, systematic surveys have demonstrated that megalithic monuments are frequently found in association with other kinds of evidence for ancient settlement and land use and thus should not a priori be treated as secluded places of commemorative or mortuary rituals. For instance, intensive survey of the Benakal Reserve Forest (Koppal District, Karnataka) has shown that megalithic complexes occur alongside other prehistoric land-use features, inclusive of ephemerally occupied rock shelters, constructed hilltop terraces and modified rock pools or stock tanks.35 As Bauer and Trivedi summarise: The results of systematic pedestrian survey suggest that megaliths... were built as components of complex material, symbolic, and social landscapes. Megalithic sites in the study region do not appear to be

Making megaliths and constituting collectives  109

isolated ritual gravesites. Instead, the distribution of other prehistoric archaeological sites and features suggest that they were active places in the cultural geographies of their makers and appear to have marked and created important locales for a variety of other cultural activities.36 In fact, Bauer has noted that the production of megalithic monuments often involved the expansion of important water retention features through stone quarrying, such that the process of monument creation was simultaneously a process of expanding water resources and effectively merged spatial practices that many archaeologists have tended to separate as either ‘ritual’ or ‘economic’.37 Unsurprisingly in light of this convergence, excavations carried out at the nearby prehistoric settlement site of Kadabakele showed that megalithic features are associated with residential terraces and were periodically maintained and modified during the course of habitation at the settlement; furthermore, Johansen’s surface documentation of the settlements at Bukkasagara and Rampuram documented small megalithic features along the residential margins of each site.38 Taken together, recent excavations, surface documentation and survey results are beginning to shed important details on the cultural landscapes and activities within which the construction of megalithic monuments occurred in South India. We now know much about how certain dimensions of land use, such as settlement and herding activities, articulated with the construction of megalithic places in some regions. Yet, one should be careful to not generalise too broadly from these rather limited detailed analyses. Moreover, as Haricharan and others point out, understanding the social significance of monuments requires further refinement to their chronological variation.39 While architectural stratigraphy and sediment stratigraphy have made it possible at some sites to establish the relative contemporaneity or sequence of some monuments (see below), there remains a significant dearth of chronometric dating assessments associated with various monumental forms and elements. Nevertheless, at a limited number of sites sustained and systematic research has made it possible to evaluate both diachronic changes and contemporaneous differences in prehistoric mortuary practices on the Deccan. In an effort to continue to improve this situation, in the following we present an analysis of the age and development of some mortuary practices at, and around, the site of Maski, allowing us to build on more recent approaches that contextualise the production of megalithic monuments as elements of an emerging social landscape in prehistoric South India. Our recent research points to a subtle, but significant, relationship between megalithic burials and emergent social affiliations and distinctions. As a mid-twentieth century ‘type-site’ of the central Deccan,40 Maski has been shown to have archaeological deposits dating to all of the primary archaeographic periods of South India: the Neolithic, Iron Age, Early Historic, Medieval and Early Modern periods.41 Renewed research at the site by the Maski Archaeological Research Project has documented

110  Andrew M. Bauer and Peter G. Johansen

a multitude of prehistoric sites, including a cemetery (MARP-79) that was in use from the Neolithic Period until at least the early part of the Iron Age. Considered en mass, the Maski material sheds considerable light on the development of megalithic mortuary practices and their significance to the development of cultural and social landscapes in prehistoric South India.

The Maski archaeological research project In 1954 B.K. Thapar’s excavations with the Archaeological Survey of India exposed stratified remains at Maski that established a chronological framework for the central Deccan; Thapar related ceramic wares and forms from Maski to the ceramic typology developed by Wheeler at Brahmagiri, a multicomponent site approximately 150 km to the south.42 Thapar’s research questions were primarily cultural-historical in seeking to establish the sequence of occupation. In doing so, his study would become an important contribution to the development of artefact and feature chronologies in the region, effectively making Maski a southern Indian type site. Although much of the archaeographic periodization identified by Thapar has subsequently been updated by researchers working on the prehistoric Deccan more generally, and at Maski specifically, his work continues to bear on a variety of important research questions.43 Foremost among them for the purposes of this chapter, Thapar identified five classes of burials at Maski, potentially providing evidence for both developments in megalithic burial practices in the region and the relationships between such burial practices and the emergence of new modes of social distinction in prehistoric southern India.44 After the work of Thapar there was little archaeological research at Maski until the Maski Archaeological Research Project (MARP) was initiated in 2010 to build on Thapar’s important chronological foundations.45 The general goals of MARP were to collect archaeological and environmental data that would allow us to examine dynamic relationships among spatial practices, such as settlement and agropastoral land use, production (e.g. craft production of iron and gold) and consumption activities, and the political constitution of social affiliations, distinctions and inequalities from the Neolithic through the Medieval and Early Modern periods. To this end, MARP has conducted a systematic pedestrian survey of a 64 km2 centred on the multicomponent site of Maski and initiated test excavations at a number of locales to further establish evidence for past cultural practices and the occupational histories of these places. MARP’s documentation of prehistoric occupation in the region has largely corroborated diachronic patterns in settlement and land use identified elsewhere in the southern Deccan, such as Bauer’s survey of the Benakal Reserve Forest noted above. Prehistoric settlement in the MARP study region is marked by several small Neolithic Period occupations and a relatively high frequency of settlements that can be assigned to the Iron

Making megaliths and constituting collectives  111

Age. Most of the settlement sites are situated on the region’s residual, granitic hills, where there is also evidence that rock shelters were at least ephemerally occupied. Excavations elsewhere in the interior Deccan suggest that during these prehistoric periods inhabitants relied on a mixed agropastoral economy, primarily dry farming cereals and pulses and herding cattle, sheep and goat.46 Detailed recording of ‘off-site’ artefacts or ‘background’ scatters by MARP has also pointed to significant spatial patterns in prehistoric agricultural production and expansion. For instance, the earliest evidence for field manuring or fertilization practices on the peneplain date to the Iron Age or Early Historic Period, when inhabitants primarily made use of the moisture-retentive, clay-rich soils of the eastern portion of the study region. This pattern is contrasted with that of the preceding Neolithic Period, which lacks low- density artefact deposits on the peneplain. Agricultural expansion during the Iron Age was likely related to a diversification of cultigens, which came to include nonlocal cereals (e.g. wheat, barley, rice) by the end of the Neolithic Period, as well as attendant watermanagement strategies, such as new scheduled uses of moisture-retentive, ­clay-rich ​­ soils.47 MARP also documented the likely continuation of herding activities on residual hilltop terraces in the study region. These sites were identified by the presence of soil retention walls and modified rock pools, and occasionally also included associated megalithic monuments and rock art.48 Herd animals, such as sheep, goat and cattle, are also evident in excavated archaeological contexts dating to prehistoric periods.49 The documentation of MARP-79, a 2-hectare prehistoric cemetery located on the pediment slope of a low inselberg hill near the centre of the MARP study area, is of particular relevance to the development of megalithic burial practices in the region (Figure 6.3). In surveying this cemetery MARP recorded the profiles of more than 20 burials that had been exposed by contemporary sediment quarrying and stone mining.50 The mining activities allowed MARP to observe monument features, human skeletal remains, slipped, and slipped and polished ceramics, copper and iron objects and chipped and ground stone lithics within open quarry pits and in the exposed profiles of the graves. Consequently, we photographed the exposures, created section drawings and collected sediment and artefacts, including human remains, for future analysis. To date, we have also obtained 13 radiocarbon assessments from secure burial contexts. In addition, two burials that were largely still intact have been excavated. Taken together with Thapar’s excavation results of megalithic features nearly 1.5 km to the north of MARP-79, the data point to significant variation in burial practices during the latter part of the Neolithic period and the development of more prominent mortuary places during the Iron Age. These developments hold significance for the relationships among ritual practices and the coproduction of cultural places and social collectives—points that we will return to after presenting some of the Maski data in more detail.

112  Andrew M. Bauer and Peter G. Johansen

­FIGURE 6.3 

 ARP systematic survey results, showing the location of Thapar’s M site (Maski), ­MARP-​­79, and the distribution of Iron Age occupation and megalithic features near the study region’s central outcrop.

Source: Authors

Variation in prehistoric mortuary practices at Maski MARP’s survey results and excavations at the cemetery site of ­MARP-​­79 have greatly expanded Thapar’s earlier assessment of the c­ ultural-​­history of Maski and have provided important contextual details for reinterpreting the significance of the burial practices he documented. Thapar identified five burial classes at Maski that he differentiated primarily by the lithic assemblage, the orientations of the cist and the skeletal completeness of the graves. Types ‘i’ and ‘ii’ involved the excavation of an elliptical pit in which skeletal materials were encased in ash and situated in the pit along with a mortuary assemblage of pottery and iron objects and sometimes animal remains (e.g. sheep/goat), before being buried with excavated earthen fill. These burials show no ‘lithic demarcation’ on the surface; but one instance of a type ‘i’ burial does make use of tabular stones to cover the skeletal materials within the e­ arthen-​­filled pit. Thapar’s distinction between type ‘i’

Making megaliths and constituting collectives  113

and ‘ii’ burials was based on pit orientation (e.g. ­north-​­south vs. e­ ast-​­west) and skeletal completeness. However, it is important to note that Thapar indicates that ‘[o]nly one’ instance of a type ‘ii’ burial was ‘fully excavated’’;51 thus, one should question his generalizations about the class as a whole, especially with regard to skeletal content. Moreover, a careful reading of Thapar’s description of the skeletal materials from both types of burials suggests that most of the human remains he recorded among these burial types were first disincarnated and thus lacked completeness. This becomes even more probable in the context of MARP’s excavations of similar burials (see below). Thapar equated burial types ‘i’ and ‘ii’ with the ‘first occupation of the site by the ­megalith-​­people’ based on their stratigraphic relationships with prior Neolithic habitation deposits; however, burial types ‘iii’ through ‘v’ are those that are most frequently associated with the ‘megalithic’ period in South Indian cultural historical frameworks.52 Burial type ‘iii’ involved the excavation of a small cylindrical pit for the placement of an urn with accompanying skeleton materials and grave goods. Burial types ‘­iv-​­v’ are more conventionally referred to as stone circles or cairn megaliths,53 and were described by Thapar as excavated pits for the interment of skeletal materials and grave goods that also included the arrangement of large boulders in a circle above the pit to delineate the grave at the surface. Boulder circles were also filled with a cairn packing of cobbles and stone rubble to produce salient landscape features (­Figure 6.4). MARP has subsequently documented more than 20 burials at ­MARP-​­79, inclusive of burials that are very similar in morphology and content to those described by Thapar. The limitation of space does not permit us to detail the results of all documented burials and excavations at M ­ ARP-​­79 here; however, several significant observations can be made that speak to the range and development of burial practices at Maski, especially when taken together with Thapar’s results. To begin with, it appears that there is significant burial variation prior to the onset of the Iron Age. For instance, at least three different burial assemblages—(1) simple pit burials lacking a prepared cist and grave goods, (2) a burial in a prepared terracotta sarcophagus and (3) a disincarnated skeleton buried in a prepared ashy cist accompanied by a suite of storage jars and serving vessels that were placed to the south of the chamber—show significant overlap in their radiocarbon age assessments, all dating to the early second millennium BCE and together suggesting that all three mortuary practices were constitutive of the cemetery in the latter half of the Neolithic Period (­Figure 6.5).54 MARP excavations of the last of these burial assemblages documented artefacts that included a small number of both jars and bowls, which might relate to consumption practices such a feasting that were associated with mortuary rituals. The suggestion of mortuary feasting is much stronger in light of other evidence at Maski that appears to date to slightly later contexts. A second burial excavated by MARP (no. 19) showed the placement of a disincarnated skeleton in a similarly burnt ashy cist but with tabular granitic capstones

114  Andrew M. Bauer and Peter G. Johansen

­FIGURE 6.4 

 eneralised profiles of two of Thapar’s excavated megalithic types G at Maski. (Note that only one of four examples of a type ‘i’ burial described by Thapar included stone capstones above the skeleton.)

Source: Authors

across the top of the burial chamber and a much more considerable 21 ceramic vessels placed to the south of the skeletal casing (­Figure 6.6). The grave assemblage included several large storage jars, but the bulk of the vessels were a range of ­open-​­mouthed shapes, including carinated ­slipped-​­and-polished black bowls and shallow ‘­dish-​­on-stand’ forms. In addition to the high number of serving vessels, the artefact assemblage included fragments of an iron dagger and the partial remains of a m ­ edium-​ ­sized mammal that were found in association with several vessels. Taken as a whole, this assemblage is analogous to burials excavated earlier by Thapar, which likewise included fragmentary ashy cist burials accompanied by a limited number of iron objects, a high number of ­open-​­mouth vessels and the partial remains of animals identified as sheep.55 Radiocarbon determinations from our excavations of this burial yielded ­late-​­second millennium BCE results (e.g., 1­ 406-​­1216 BCE, 2σ), which are consistent with our other assays (e.g. 1279–1132 BCE, 2σ) from graves with nearly identical

Making megaliths and constituting collectives  115

­FIGURE 6.5 

 adiocarbon probability distribution assays from burial contexts R at ­MARP-​­79. 

Source: Authors

morphologies and indicate that these mortuary practices were present around the start of the Iron Age.56 Taken together, it appears that the practice of burying the dead in ashy cists occurred in the early second millennium BCE among a suite of other burial practices (e.g. sarcophagus burial, simple pit burial), and by the beginning of the Iron Age came to include a capstone lithic assemblage, a greater range and quantity of serving vessels and, in some instances, the disarticulated remains of a limited number of livestock, suggesting the possibility that the burial rituals also included feasting activities. The number of o ­ pen-​­mouth serving vessels and association with animal remains is highly suggestive that mortuary practices involved communal consumption events in the case of some burials—a point that we elaborate on below. In summary, the Maski data point to significant mortuary variation in the early second millennium BCE as well as chronological developments in mortuary practices. Based on stratigraphic relationships at Maski, Thapar

116  Andrew M. Bauer and Peter G. Johansen

­FIGURE 6.6 

 xcavation plan of burial no. 19 at ­MARP-​­79.  Note that capstones E sealing the cist are displayed transparently to show underlying skeletal remains.

Source: Authors

identified simple pit burials that largely lacked lithic elements (e.g. type ‘i’) as the first form in the sequence at Maski. Radiocarbon assays hitherto obtained by MARP from M ­ ARP-​­79 corroborate the conclusion that the earliest grave assemblages at Maski lack stone architectural features. Although MARP dated at least three different burial practices to the early second millennium BCE, none include significant lithic elements. In contrast, the latest radiocarbon assays from M ­ ARP-​­79 obtained come from contexts in which mortuary chambers were capped with granitic stone slabs. These contexts date to the latter part of the second millennium BCE and straddle the conventional date for the start of the Iron Age (ca. 1200 BCE). In addition, mortuary contexts that MARP has documented with granitic capping stones include iron tools as part of the artefact assemblage, corroborating that the graves date to the latter part of the second millennium BCE at the earliest.57 Taken together, Thapar’s excavations, MARP’s excavations and recent radiocarbon assays from other recorded burials at cemetery M ­ ARP-​­79 indicate that mortuary practices at Maski began to incorporate stone elements sometime during the second millennium BCE, as ­pit-​­cut graves and ashy cists would come to be capped with granitic stones. One such feature documented by MARP also incorporates a granitic boulder that demarcates the eastern edge of the burial pit, perhaps a precursor to the truly ‘megalithic’ ringed stone circles and cairns that are also evident at the site.

Making megaliths and constituting collectives  117

Making megaliths and constituting collectives in prehistoric South India The variation in grave assemblages at Maski implies that South Indian prehistoric mortuary practices came to involve more labour-intensive activities during the Iron Age, and in doing so became cultural contexts for generating new forms of social affiliations and distinctions as greater numbers of inhabitants differentially participated in producing specific monumental places. That mortuary rituals came to involve greater numbers of participants and sometimes incorporated feasting activities suggests that the making of monuments had significant social effects. In short, burial rituals and the production of mortuary monuments began to incorporate significantly more collective labour and also appear to have involved commensal politics, wherein multiple people shared in ritual consumption activities. Thus, rather than focus on how prehistoric mortuary architecture and grave assemblages may correlate with the status, rank or class of the deceased individuals—as has been suggested by others—we find it more appropriate to place emphasis on how mortuary practices and megalithic places contributed to the establishment and maintenance of social collectives. After all, it was living associates, not the deceased, who mobilised labour to produce megaliths. Considering the Maski material in the context of other well-studied regions of South India, it appears that the development of megalithic burial practices involved a high degree of labour mobilization by at least the end of the Iron Age. The sheer size of some of the megalithic features associated with the period suggests that their production involved a large number of participants, and probably draft animals. For instance, it is hard to imagine how some of the dolmens at Hire Benakal—which includes features comprised of granitic capstones weighing greater than 10 tons placed on massive boulder-sized uprights that extend more than 3 m above the ground surface—could have been erected without drawing on a pool of people and animals that extended to a broad range of kin or community relationships.58 The same could be said for many of the cists enclosed by boulder circles and capped with cairns, such as those that are common at Maski noted above. Indeed, experimental archaeological studies in South India have conservatively indicated that a minimum of 70 people would have been required to construct such features in less than three days and that, in all probability, the number of participants was much larger in antiquity.59 It is important to reiterate, however, that not all prehistoric megaliths are equally ‘mega’. As Bauer has demonstrated by quantifying the variation among dolmens he documented through pedestrian survey around Hire Benakal, the majority of these mortuary features are comprised of capstones less than 75 cm above the ground and are frequently supported by uprights composed of modified cobbles rather than massive stone slabs or orthostats.60 Moreover, a number of monuments at Hire Benakal are composed entirely of cobbles that could be easily assembled by a small group

118  Andrew M. Bauer and Peter G. Johansen

of people. While some of the variation in the size and morphology of these ‘megalithic’ features are possibly chronological, it is probable that significant contemporaneous variation in monument production also existed at these sites. For example, stratigraphic relationships among architectural elements of several dolmens at Hire Benakal indicate that approximately contemporaneous monuments show substantial disparities in size. In short, the spatial relationship among some features attest to considerable differences in forms and labour mobilization even among monuments that were produced at roughly the same time. This should be unsurprising in light of the burial evidence dated at Maski, where we now know that remarkable variation in mortuary practices occurred simultaneously in the early second millennium BCE. In short, it appears that significant inequalities existed in the abilities of inhabitants to mobilise labour to produce megalithic features. These differences are evident both within and between mortuary sites. In some instances, such inequalities were probably related to social access to pastoral resources, and consequently abilities to manage livestock, herd wealth and likely sponsor feasts.61 As noted above, in the Benakal forest megalithic monuments are frequently associated with archaeologically documented prehistoric herding locales, which appear to have been relatively restricted to places where vegetated terraces occurred alongside accessible surface water. Indeed, at several sites granitic rock pools were expanded to increase their holding capacities in the process of crafting megaliths.62 Given the common association of megalithic monuments and important modified rock pools and pasturage in this context, Bauer has suggested that the production of monuments established relationships among ancient inhabitants and these significant places of land use—linking commemorative practices with quotidian activities of herding.63 This argument also finds support among the residual hills in the Maski study area.64 Taken as a whole, it is thus possible to identify at least two modes of political practice associated with megalithic production in prehistoric South India: one related to the constitution of collectives of labour and shared consumption activities involved in the process of making monuments and a second related to the material legacy of monuments in constituting cultural and historical places of social affiliation. With respect to the first, the production of truly ‘megalithic’ monuments, no doubt, required the assembly of a large number of people. The making process would have (re)established relations of kin or community as inhabitants laboured toward a common cause. In other words, the practices of monument production were simultaneously practices of reconstituting political relations of affiliation and distinction as regional inhabitants differentially participated in mortuary rituals. That monument production appears to have frequently also involved feasting lends stronger support to this argument, given that feasts are well documented to highlight social distinctions (e.g. those sponsoring the feast) and simultaneously promote social cohesion among communal participants.65 In that sense, the cultural practices associated with megalithic

Making megaliths and constituting collectives  119

monument production were deeply related to the constitution and maintenance of social collectives. In well-studied regions, the monuments and the temporal durability of their material presence in constituting cultural places also appear to be integrally connected to the maintenance of these social relationships. The practice of making megaliths probably had political effects that transcended the occasion of their initial production, extending to how places were temporally perceived and experienced by inhabitants as part of a landscape of social difference and cultural resources. At a number of sites, for instance, it is evident that multiple cists or dolmens were enclosed in the same megalithic stone circle, materially and symbolically grouping distinct burial chambers.66 These data imply that collective associations with specific monumental places were also regularly renewed through additional rituals and ongoing monument maintenance. Environmental perceptions are shaped through a history of experiences and material practices that necessarily engender socially and symbolically distinctive places. ‘The recursive relationship between practice and environment becomes constitutive of both (meaningful) landscapes and differentiated subjectivities’ as people differentially constitute places and spaces through meaningful tasks and activities.67 Thus, prehistoric megalithic complexes of South India, monumental places that were constructed and maintained by particular associations of kin or community, would have held different meanings for those who participated in their production and regularly renewed them than they would for those who did not. Moreover, that megalithic places are often associated with valued substances, such as the water of rock pools or the dung of Neolithic ashmounds, as well as other evidence for land use on the region’s residual hills, suggests that in some contexts the production of these monumental places was an important dimension of establishing social relations to particular resources, and probably was a mode of generating privileged access to them.68 The significance of megalithic monuments to regional inhabitants was thus also rooted in the durability of their presence. As noted at the outset of this chapter, creating and manipulating landscape features can be an effective means of influencing perceptions of historicity that are significant to the sociohistorical process.69 While we are not concerned with speculating on the discursive traditions (e.g. origin myths, oral histories, etc.) associated with megalithic monuments during the Iron Age, it is nevertheless the case that megalithic monuments, in constituting enduring cultural places, would have been simultaneously symbolic and material evidence of social activities. As Morrison has insightfully noted, places have a way of persisting—not because their meanings are stable, but because their material constituents perpetuate them as distinct and perceptible places.70 Stated more strongly, one might recognise that megaliths are objects that necessarily ‘bear witness to a history’.71 Their physical instantiations constitute evidence that a place has a social history and simultaneously call it into being. In the case at hand, the persistence of megalithic monuments appears to have solidified temporal relationships to significant places, especially as

120  Andrew M. Bauer and Peter G. Johansen

they were regularly renewed through ongoing land use and additional monument construction and maintenance. In other words, the production of these monumental places—which frequently coupled social groupings with materially and symbolically valued resources—was critical to the maintenance and historicities of social collectives. Rather than merely reflecting social inequalities in status or access to resources in the past, the production of megalithic features and their material durability made the monuments instrumental elements to the process through which historical relationships and social affiliations were constituted, maintained and negotiated. This framing of the significance of megalithic monuments has some parallels to recent scholarship produced under the rubrics of ‘new materialisms’ and ‘posthumanist’ approaches to the social sciences that warrants a final point of clarification.72 In rearticulating the boundaries and associations of ‘the social’, for example, Latour has called attention to the role of materials as ‘mediators’ of human relationships—that is to say that they hold the potential to ‘transform’ and ‘modify’ the historical contexts that they partly comprise.73 Social collectives and relationships of inequalities become constituted through assemblages of human and nonhumans: ‘It's the power exerted through entities that don't sleep and associations that don't break down that allow power to last longer…—and, to achieve such a feat, many more materials than social compacts have to be devised’.74 To be clear, we are not suggesting that durable stone monuments were equal to humans as political agents or actants (sensu Latour) in prehistoric South India. The mediating effects of monuments existed because inhabitants recognised significance in their forms and practiced commensal politics in which herd animals, grazing places and settlement locales were culturally valued. While the presence of the monuments engendered perceptions of social places they surely did not fully determine their meaningful content. In other words, megalithic monuments were part-and-parcel of the production of historicities that were related to the constitution of social collectives and specific places and resources, but only as also mediated through cultural systems of value and signification.75 Indeed, with changes in cultural values and the modification of social collectives, novel historicities were instantiated through the continued production of these places. To underline a point we made at the outset of this chapter, in the ongoing production of monuments—as durable places of political landscapes—people may call upon their materiality to strategically affirm, silence or reconstitute temporal claims that influence the perception of a prospective yet uncertain futurity.

Conclusion: the politics of megaliths in prehistoric South India Based on an assessment of mortuary practices at Maski and a review of megalithic sites that have been investigated in detail in northern Karnataka, we have argued that megalithic production was associated with at least two modes of political practice in prehistoric South India: One related to the constitution of collectives of labour and shared consumption

Making megaliths and constituting collectives  121

activities involved in the process of making monuments and a second related to the material legacy of monuments in constituting cultural and historical places of social affiliation. Some megalithic monuments were thus part-and-parcel of the constitution of social associations. In short, large and salient features (e.g. boulder circles and dolmens) required the generation of collectives in the process of their production and simultaneously established enduring social and temporal relationships to places. Yet, the review of the production of megalith complexes presented above also indicates that not all ancient inhabitants exercised equal abilities to produce these relationships. Monumental places were unequally created, experienced and made meaningful. While this process potentially institutionalised some forms of collective difference and social inequalities— solidifying associations with places where some people commemorated their kin and grazed their herds and others did not—monumental placemaking was clearly not the only modality of political practice during the period. We have suggested elsewhere, for instance, that evident inequalities in the abilities to mobilise labour to build megaliths and other landscape features were probably related to herd wealth, differential capacities to promote commensal politics and a wide range of other consumption and production practices, such as selective control of aspects of craft production, evidenced at settlement sites.76 Our conclusions here remain subject to revision as systematic research on megalithic sites, and of equal importance, other less salient contexts associated with the quotidian practices of their producers, continues to develop.

Acknowledgements The Maski Archaeological Research Project discussed in this chapter is a collaboration between the Karnataka Department of Archaeology, Museums and Heritage (KDAMH) and the authors. We thank the Archaeological Survey of India for permission to conduct research at Maski (F.1/8/2009-EE). We also thank the National Science Foundation of the United States, the Fulbright–Nehru program of the United States–India Educational Foundation, the Social Sciences and Humanities Research Council of Canada, the University of Illinois, DePauw University, Stanford University and McGill University for financial support; the American Institute of Indian Studies for assistance and continuous support, and particularly Dr. Pradeep Mehendiratta, Purnima Mehta and Dr. Vandana Sinha; and friends, colleagues and project codirectors at KDAMH, including T. S. Gangadhar, Dr. R. Gopal, Dr. Gavisiddaiah and N. L. Gowda. Earlier drafts of this chapter were presented by Bauer at several venues, including the Stanford Archaeology Center, the annual meeting of the American Anthropological Association and a special conference on ‘Knowledge Traditions of the Indian Ocean World’ at Oxford University. Many thanks are owed to numerous friends and colleagues for feedback in those contexts, and especially Himanshu Prabha Ray.

122  Andrew M. Bauer and Peter G. Johansen

Notes 1 Eric Hobsbawm and Terence Ranger, The Invention of Tradition (Cambridge: Cambridge University Press, 2012); Michel-Rolph Trouillot, Silencing the Past: Power and the Production of History (Boston: Beacon Press, 1995). Note that by the early twentieth century the figure of Columbus had been selectively appropriated as a (Catholic) national hero of the US colonial state, while the annual celebration of ‘Columbus Day’ simultaneously silenced how Columbus established his personal fortune through the sustained practice of abducting, killing, maiming and enslaving inhabitants of Caribbean Indigenous communities during the late fifteenth and early sixteenth centuries. This latter legacy of Columbus is very much alive in contemporary historical scholarship and indigenous communities; e.g. Thomas King, The Truth about Stories: A Native Narrative (Toronto: House of Anansi, 2003); Linda Tuhiwai Smith, Decolonizing Methodologies: Research and Indigenous Peoples (London: Zed Books Ltd., 2013). 2 Trouillot, Silencing the Past. 3 Stephan Palmié and Charles Stewart, ‘Introduction: For an Anthropology of History’, HAU: Journal of Ethnographic Theory 6, no. 1 (2016): 207–236. 4 Adam T. Smith, The Political Landscape: Constellations of Authority in Early Complex Polities (Berkeley: University of California Press, 2003); Timothy R. Pauketat and Susan M. Alt, ‘Mounds, Memory, and Contested Mississippian History,’ in Ruth M. Van Dkyke and Susan E. Alcock (eds.), Archaeologies of Memory (Chichester: Blackwell, 2003), pp. 151–179; Carla M. Sinopoli, ‘Echoes of Empire: Vijayanagara and Historical Memory, Vijayanagara as Historical Memory,’ in Archaeologies of Memory, 2003, 15–33; Ruth M. Van Dyke and Susan E. Alcock, Archaeologies of Memory (Chichester: John Wiley & Sons, 2008). 5 Trouillot, Silencing the Past, ­p. 29. 6 V. Gordon Childe, ‘Megaliths’, Ancient India 4 (1948): 4–13. 7 Trouillot, Silencing the Past, p. 29; Richard M. Eaton and Phillip B. Wagoner, Power, Memory, Architecture: Contested Sites on India’s Deccan Plateau, 1300– 1600 (Oxford: Oxford University Press, 2014); Van Dyke and Alcock, Archaeologies of Memory. 8 Childe, ‘Megaliths’ p. 5. 9 G. Yazdani, ‘Megalithic Remains of the Deccan: A New Feature of Them’, Journal of the Hyderabad Archaeological Society 7 (1917): 56–79; G. Yazdani, ‘The Scope of Archaeology in the Hyderabad State,’ The Journal of the Hyderabad Archaeological Society 2 (1916): 1–13. H. E. Hunt, ‘Hyderabad Cairns, Their Problems,’ Journal of the Hyderabad Archaeological Society 1 (1916): 180–224. 10 D. H. Gordon and F. R. Allchin, ‘Rock Paintings and Engravings in Raichur, Hyderabad’, Man 55 (1955): 97–99; V. D. Krishnaswami, ‘Megalithic Types of South India,’ Ancient India 5 (1949): 35–45; Bal Krishen Thapar, ‘Maski 1954: A Chalcolithic Site of the Southern Deccan,’ Ancient India 13, no. 1957 (1957): 4–142; Adiga Sundara, The Early Chamber Tombs of South India: A Study of the Iron Age Megalithic Monuments of North Karnataka (Delhi: University Publishers (India), 1975); RE Mortimer Wheeler, ‘Brahmagiri and Chandravalli 1947: Megalithic and Other Cultures in the Chitaldrug District,’ Mysore State, Ancient India IV (1948): 81–308. 11 Andrew M. Bauer, Before Vijayanagara: Prehistoric Landscapes and Politics in the Tungabhadra Basin (New Delhi: Manohar and American Institute of Indian Studies, 2015); Robert Brubaker, ‘Aspects of Mortuary Variability in the South Indian Iron Age,’ Bulletin of the Deccan College Research Institute 60 (2000): 253–302; Krishnaswami, ‘Megalithic Types of South India’; Kathleen D. Morrison, Mark T. Lycett, and Mudit Trivedi, ‘Megaliths and Memory: Excavations at Kadebakele and the Megaliths of Northern Karnataka’, in Proceedings of the 20th Conference of the European Association for South Asian Archaeology and

Making megaliths and constituting collectives  123

12 13 14

15 16 17 18 19

20 21

22 23 24 25 26 27 28 29 30

Art, Contextualizing Material Culture in South and Central Asia in Pre-Modern Times, vol. 2, 2015. Bauer, Before Vijayanagara. Bauer, Before Vijayanagara. Childe, ‘Megaliths’; Meadows Taylor, A Student’s Manual of the History of India (Green London: Longmans, 1906); Meadows Taylor, Megalithic Tombs and Other Ancient Remains in the Deccan (New Delhi: Asian Educational Services, 1941); Trautmann Thomas, Aryans and British India (Berkeley: University of California Press, 1997). Andrew M. Bauer and Mudit Trivedi, ‘Contextualizing Megalithic Places: Survey, Mapping, and Surface Documentation in the Environs of Hire Benakal (Koppal District, Karnataka)’’ Man and Environment 38, no. 2 (2013): 46–61. Bauer and Trivedi, ‘Contextualizing Megalithic Places’, p. 47. Childe, ‘Megaliths’; Taylor, A Student’s Manual of the History of India; H. E. Hunt, ‘Hyderabad Cairns, Their Problems,’ Journal of the Hyderabad Archaeological Society 1 (1916): 180–224. H. D. Sankalia, ‘Megalithic Monuments in the Deccan and Karnataka,’ Bulletin of the Deccan College Research Institute 2, no. 1/2 (1940): 184–184; Thapar, ‘Maski’; Wheeler, ‘Brahmagiri and Chandravalli 1947’. Andrew M. Bauer and Peter G. Johansen, ‘Prehistoric Mortuary Practices and the Constitution of Social Relationships: Implications of the First Radiocarbon Dates from Maski on the Occupational History of a South India ‘Type Site’, Radiocarbon 57, no. 5 (2015): 795–806; Brubaker, ‘Aspects of Mortuary Variability in the South Indian Iron Age’; Smriti Haricharan, Hema Achyuthan, and N. Suresh, ‘Situating Megalithic Burials in the Iron Age-Early Historic Landscape of Southern India,’ Antiquity 87, no. 336 (2013): 488–502; Kathleen D. Morrison, ‘Brahmagiri Revisited: A Re-A nalysis of the South Indian Sequence’, South Asian Archaeology 2001 (2005): 257–261. Sundara, The Early Chamber Tombs of South India. Lewis R. Binford, ‘Mortuary Practices: Their Study and Their Potential,’ Memoirs of the Society for American Archaeology 25 (1971): 6–29; Lewis R. Binford, An Archaeological Perspective (New York: Seminar Press, 1972); Arthur A. Saxe, ‘Social Dimensions of Mortuary Practices in a Mesolithic Population from Wadi Haifa, Sudan,’ Memoirs of the Society for American Archaeology 25 (1971): 39–57. Udayaravi S. Moorti, Megalithic Culture of South India: Socioeconomic Perspectives (Varanasi: Ganga Kaveri Publishing House, 1994); Lawrence S. Leshnik, South Indian Megalithic Burials: The Pandukal Complex (Wiesbaden: Steiner, 1974). Brubaker, ‘Aspects of Mortuary Variability in the South Indian Iron Age’. pp. 287–292. ­ E.g. Bauer, Before Vijayanagara; Brubaker, ‘Aspects of Mortuary Variability in the South Indian Iron Age,’ pp. 287–292. Bauer and Trivedi, ‘Contextualizing Megalithic Places’. Thapar, ‘Maski’; Leshnik, South Indian Megalithic Burials. Bauer, Before Vijayanagara, ­p. 6. Bauer and Trivedi, ‘Contextualizing Megalithic Places’. p. 48. See discussions in Bauer, Before Vijayanagara; Moorti, Megalithic Culture of South India. Bauer and Trivedi, ‘Contextualizing Megalithic Places’; Peter G. Johansen, ‘The Politics of Spatial Renovation: Reconfiguring Ritual Places and Practice in Iron Age and Early Historic South India,’ Journal of Social Archaeology 14, no. 1 (2014): 59–86; R. K. Mohanty and V. Selvakumar, ‘The Archaeology of the Megaliths in India: 1947–1997,’ Indian Archaeology in Retrospect 1 (2002): 313–352; R. K. Mohanty, ‘A Preliminary Report of the Excavations at Mahurjhari, 2001: A Megalithic and Early Historic Site in Vidarbha, Maharashtra,’ Pratnatattva 9 (2003): 41–48; Morrison, Lycett, and Trivedi, ‘Megaliths and Memory’.

124  Andrew M. Bauer and Peter G. Johansen



­













­

Making megaliths and constituting collectives  125

48 49 50 51 52 53 54 55 56 57 58 59 60 61

62 63 64 65

66

67

68 69 70

­Sanganakallu-Kupgal,’ ​­ Archaeological and Anthropological Sciences 8, no. 3 (2016): 575–599. Peter G. Johansen and Andrew M. Bauer, ‘On the Matter of Resources and Techno-Politics: The Case of Water and Iron in the South Indian Iron Age,’ American Anthropologist 120, no. 3 (2018): 412–428. E.g. Thapar, ‘Maski’. Johansen and Bauer, ‘Beyond Culture History at Maski’; Bauer and Johansen, ‘The Maski Archaeological Research Project (2010–18)’; Bauer and Johansen, ‘Prehistoric Mortuary Practices and the Constitution of Social Relationships’. ­ Thapar, ‘Maski,’. p. 30. Thapar, ­p. 27. Cf. Brubaker, ‘Aspects of Mortuary Variability in the South Indian Iron Age’; Edmund Henderson Hunt, ‘Hyderabad Cairns, Their Problems,’ Journal of the Hyderabad Archaeological Society 1 (1916): 180–224; Thapar, ‘Maski’. Bauer and Johansen, ‘The Maski Archaeological Research Project (2010–18)’. Thapar, ‘Maski,’ pp. 127–9. ­ Johansen and Bauer, ‘Beyond Culture History at Maski’. Johansen, ‘Early Ironworking in Iron Age South India’. Bauer and Trivedi, ‘Contextualizing Megalithic Places’. R. K. Mohanty and S. R. Walimbe, ‘An Investigation into the Mortuary Practices of Vidarbha Megalithic Cultures,’ in C. Margabandhu and K.S. Ramachandran (eds.), Spectrum of Indian Culture (Delhi: Agam Kala Prakashan, 1996). Bauer, Before Vijayanagara. Andrew M. Bauer, Peter G. Johansen, and Radhika L. Bauer, ‘Toward a Political Ecology in Early South India: Preliminary Considerations of the Sociopolitics of Land and Animal Use in the Southern Deccan, Neolithic through Early Historic Periods,’ Asian Perspectives 46, no. 1 (2007): 3–35; Bauer, Before Vijayanagara. Bauer, ‘Producing the Political Landscape’. Bauer; Bauer, Before Vijayanagara. Johansen and Bauer, ‘On the Matter of Resources and Techno-Politics’. Michael Dietler and Brian Hayden, ‘Digesting the Feast—Good to Eat, Good to Drink, Good to Think: An Introduction,’ in Dietler and Hayden (eds.), Feasts: Archaeological and Ethnographic Perspectives on Food, Politics, and Power (Washington: Smithsonian Institute, 2001), pp. 1–20. Bauer, Before Vijayanagara; V. Sontakke, ‘A Review of Inner Layout of Megalithic Burials in Upper Wainganga Valley: A New Emergence in Vidarbha Megaliths,’ Heritage: Journal of Multidisciplinary Studies in Archaeology 2 (2014): 493–515. Andrew M. Bauer, ‘Urbanism and the Temporality of Materiality on the Medieval Deccan: Beyond the Cosmograms of Social and Political Space,’ in Susan M. Alt and Timothy R. Pauketat (eds.), New Materialisms Ancient Urbanisms (London: Routledge, 2019), pp. 184–217, p. 193. See also Tim Ingold, The Perception of the Environment: Essays on Livelihood, Dwelling and Skill (London: Routledge, 2000); David Harvey, The Condition of Postmodernity, vol. 14 (Cambridge, MA: Blackwell, 1989); Uzma Z. Rizvi, ‘Subjectivities and Spatiality in Indus Urban Forms: Mohenjo Daro, the Body and the Domestication of Waste,’ in The Archaeology of Politics: The Materiality of Political Practice and Action in the Past, 2011, pp. 221–244; Smith, The Political Landscape. Bauer, Before Vijayanagara; Johansen, ‘The Politics of Spatial Renovation’. Sinopoli, ‘Echoes of Empire’; Smith, The Political Landscape; Trouillot, Silencing the Past; Susan E. Alcock and Ruth M. Van Dyke, Archaeologies of Memory (London: Blackwell, 2003). Kathleen D. Morrison, Daroji Valley: Landscape History, Place, and the Making of a Dryland Reservoir System (New Delhi: AIIS and Manohar Publishers, 2009).

126  Andrew M. Bauer and Peter G. Johansen

71 Laurent Olivier, Dark Abyss of Time: Memory and Archaeology (Lanham, MD: Rowman & Littlefield, 2011), p. 6. 72 Diana Coole and Samantha Frost, ‘Introducing the New Materialisms,’ in Coole and Frost (eds.), New Materialisms: Ontology, Agency, and Politics (Durham: Duke University Press, 2010), pp. 1–43; Sarah Whatmore, Hybrid Geographies: Natures Cultures Spaces (London: Sage, 2002). 73 Bruno Latour, Reassembling the Social. An Introduction to Actor-Network-Theory (New York: Oxford University Press, 2005), p. 39. Latour, p. 70. ­ 74 75 See also Andrew M. Bauer, ‘Questioning a Posthumanist Political Ecology: Ontologies, Environmental Materialities, and the Political in Iron Age South India,’ Archeological Papers of the American Anthropological Association 29, no. 1 (2018): 157–174; Andrew M. Bauer and Steve Kosiba, ‘How Things Act: An Archaeology of Materials in Political Life,’ Journal of Social Archaeology 16, no. 2 (2016): 115–141. 76 Bauer, Johansen, and Bauer, ‘Toward a Political Ecology in Early South India’; Johansen and Bauer, ‘On the Matter of Resources and Techno-Politics’’.

7 BULLION, BAUBLES AND BOWLS Reconstructing networks of exchange in the Indian Ocean Uthara Suvrathan

The Indian Ocean world in antiquity was constituted of multiple spatial and temporal scales that overlapped in complex ways. Much has been written on the goods traded, the contexts of production and methods of manufacture, the nature and geography of trade routes, the modalities of trade and so on. Our preoccupation with trade often leads to simple narratives detailing the items traded out of one place and into another. We know that the sites participated in Indian Ocean trade but what does that really mean? When we start looking at how the ports or sites are embedded in smaller, regional networks and how these and other networks change over time, we can begin to think about the interplay between the global and the local as well as of temporal changes in the Indian Ocean system. In this chapter I want to move beyond the question of trade, although that is undeniably important, and examine meaning and value beyond the purely economic – that of the social lives of the objects traded. I study the ocean as a unified space and emphasise the need to analytically re- examine the not inconsiderable archaeological material found on the Indian Ocean rim. I compare archaeological data from published excavation reports from three sites – Arikamedu, India; Tissamaharama, Sri Lanka; and Berenike, Egypt (Figure 7.1). Arikamedu is an ancient port located in Tamil Nadu, southern India whose heyday dates from the third century BCE to the early centuries CE, although it continued to be intermittently occupied into the seventh century CE. Although initially considered important for its postulated links with the Roman Empire reanalysis of the ceramics at the site tie the foundations of the site as a trading station to developments within early historic south India, most notably the beginning of Mauryan Imperial contact with the region – and not necessarily only with Roman trading activity in the region. There has been a long history of archaeological research at the site1 and

128  Uthara Suvrathan

­FIGURE 7.1 The

three sites mentioned in the text.

Source: Author.

excellent material culture studies on the beads,2 bead technology,3 Roman glass4 and pottery.5 Tissamaharama was the main settlement of one of several regional kingdoms located in the southeast of the island in the third century BCE and continued to be occupied into the thirteenth century CE. While it is an inland city characterised by a citadel surrounded by numerous Buddhist monasteries and craft production areas, the site was close to the ancient ports of Kirinde and Godavaya. Since 1992, a series of excavations at this site were conducted under the joint direction of the Archaeological Department of Sri Lanka and the Commission for General and Comparative Archaeology (KAVA) of the German Institute of Archaeology. Several preliminary reports have been published.6 Berenike, an Egyptian port on the west coast of the Red Sea, was an early centre linking trade between the Mediterranean and the Indian Ocean. The site was established by the Greeks in the third century BCE, rising to special prominence under the Romans by the late centuries BCE and early century CE, before being abandoned in the sixth century CE. There have been several seasons of excavations and survey in and around the site by the joint University of Delaware–Leiden University expedition.7 The three sites are very different both in their cultural context and their roles in the Indian Ocean system. Berenike was a major transit point linking Egypt and the Mediterranean to the Indian Ocean. Arikamedu fulfilled a similar role in the Indian subcontinent but also had a longer regional history

Reconstructing networks of exchange  129

than previously believed. Tissamaharama was a very different site, not a port, and was located at one remove from the ocean. Nevertheless, as will be seen in the discussion below, this site also participated in what I call the Indian Ocean system, a transoceanic and multicontinent set of networks and routes of movement and exchange. The task of dealing with disparate data from very different sites is a daunting one. I examine three categories of artefacts – ceramics, coins and beads – and by tracing long term trends in this cross- cultural dataset, I reconstruct a section of the western Indian Ocean network, identifying temporal and spatial patterns that reflect changing emphases and frequencies of interaction in the movement of individuals or communities, in cultural exchange and trade. I seek to emphasise the multiplicity of these links and attempt to define and characterise the nature of some of them. I suggest that comparative and cross- cultural analysis begins to address the temporal and spatial divides that partition the study of the Indian Ocean world in antiquity and is one way of balancing the study of the ocean as a unified space as well as of the disparate terrestrial regimes that surround it.

Considering Value and Exchange Following Appadurai,8 I suggest that value is created by exchange and is embodied in the things that are exchanged. Not only can a study of the distribution of objects allow us to trace various cross- cultural linkages but a focus on the social life of artefacts can also explicate the commodities’ locations within diverse social and historical contexts  – objects have different values in different cultural contexts as they pass through different ‘regimes of value’.9 Therefore, in considering intercultural trade, it is important to explore distinct cultural patterns of consumption and meaning for a variety of objects in their diverse contexts. Is it possible for archaeologists to get at this level of meaning? In a discussion of exchange on the colonial peripheries of Oceania, Nicholas Thomas suggests that common sense and utilitarian interpretations of trade items based on a presumed economic value ignores ‘attributions of value arising from the associations of things rather than their material properties’.10 Instead, he suggests that western goods had a ‘cultural value’ beyond the merely utilitarian and took on a variety of meanings and significances in specific cultural contexts. Guns, for instance, were not just weapons but things to display in sacred places, special kinds of exchange items and ways of creating a history. In our case, beads or coins were not entirely novel things in the Indian Ocean sphere that would have needed to be adapted to an unfamiliar context. It may be that certain exotic beads or coins were viewed not just as foreign objects but as scarce commodities that were a foreign form of something that was all too familiar. Such objects thus may have been assigned distinctive value for several reasons, apart from the intrinsic value of the coin or bead itself (or the materials they were made of).

130  Uthara Suvrathan

That beads often have social and cultural value is evident from ethnographic cases, where they were an important way of displaying the wealth and social status of the wearer.11 But did this value derive solely from their physical properties? A similar question can be posed with respect to coins. Were they valued as sources of bullion? Or did the form and type play a role in enhancing their value and contributing to their widespread use across cultural boundaries? In an interesting study, Nicholas Saunders12 showed that the value of pearls lay in that they were considered by the indigenous people of the Americas to be the embodiment of the spiritual and creative power of light. Pearls, as items that had meaning because of this worldview, were later recontextualised when the advent of the Europeans led to their being judged by their physical characteristics alone and for their commercial value. This raises interesting questions about how the very materiality of an artefact acts not just as a bridge between the mental and the physical worlds, but also allows for its interpretation and reinterpretation in different contexts. In the case that Saunders discusses, native peoples initially exchanged pearls for shiny objects (like glass) that had little commercial value to European traders but were considered worth having because their brilliance embodied light as did the pearls. Therefore, apart from the intrinsic value of a bead or coin itself, a variety of other considerations can play a role in imbuing it with value. Recent studies on material culture tend to focus on how objects construct and express social identities, rather than on the object itself.13 For instance, its age and connection to the ancestors might be crucial in users’ assessment of an object’s value, as is the case for beads among the Kelabit of Sarawak, Borneo. People seek legitimacy by buying these beads which are considered to mark the ‘goodness’ of a person, a quality that is seen as inherited and is therefore proven through the ownership of ancestral beads and not necessarily the inheritance of the beads.14 On the other hand, the ethnohistorically known use of beads as currency among the Kelabit potentially delinks them from this specific set of meanings or social contexts and involves them in a web of impersonal economic exchange. Similarly, the use of Roman imperial issues in distant areas might not solely be linked to their value as bullion. In Sri Lanka, several forgeries of Roman issues, and even of North Indian Imperial series, have been identified. This could have something to do with the larger connotations of the use of these coins, perhaps providing legitimacy to local rulers. It is valuable then to differentiate between the ‘commodity phase’ of artefacts as things that can be exchanged and their phase as cultural objects within a particular social context, where their importance does not lie in its exchangeability but in the specific meanings they hold for a group or an individual. In this phase, more importance might be placed on an object’s value as an heirloom or as an object of display communicating a variety of messages and emphasis might be placed on its not ­re-entering ​­ a commodity phase. As Kopytoff15 points out, commodities might be considered to have life histories and to move in and out of a commodity state. Only a

Reconstructing networks of exchange  131

consideration of specific uses, forms and trajectories of objects can explicate their social contexts and meanings.16 It might be also fruitful, following Lesure, to explore how the evaluation of objects is closely tied to the ways their use contributed to the creation, display and maintenance of social relationships, cultural values and religious beliefs.17 To explore the social roles of valuable objects of jade and greenstone in early Mesoamerica, Lesure suggested first, the characterisation of the social uses of artefacts according to degree of alienability and second, an evaluation of the kinds and scales of interaction with which they were implicated.18 Assigning gradations of value is useful for organizing the data in categories that encoded meaningful variability among valuable objects. Thus, in Early Formative Mesoamerica (1600–1200/1100 BCE) gradations were not well developed, with greenstone not being made into inalienable objects but referencing individual and horizontal social differences. Greenstone was equally accessible to all households, and the high incidence of discard seemed to indicate that they were everyday items of personal adornment or utilitarian objects (celts) and not status markers. By the late Early Formative (1200/1100–850 BCE) an expansion in gradations of value was seen in that greenstone items were used as signifiers of vertical social differentiation (as seen in differential access to the material in graves) and in ritual. But it was not used extensively and was not the focus of iconographic elaboration.19 The Middle Formative (850–400 BCE) saw the development of hierarchical societies and greenstone was used to construct chiefly inalienable artefacts (e.g. celts) with elaborate life histories as well as in ritual. At the same time, however, its use in vertical and horizontal relationships at an intermediate scale continued, since the material was used (beads and pendants) in an extensive system of different kinds of gradations of value. While Lesure’s focus is on elite items, his framework is a useful starting point for the consideration of the different categories of artefacts studied in this essay. In the case of beads or coins, for instance, their degree of alienability is likely not a meaningful category to consider since they were part of an extensive exchange system where their very alienability was an important characteristic. However, the concept of gradations of value is useful in the consideration of their circulation and the kinds and scales of social interactions in which they participated. Gradations of value among artefacts might be indicated by differences in size; form (objects of one material but in different forms); composition (similar form but of different materials, as for instance, forgeries of Roman issues in locally available metals); and specificity (an assessment of uniqueness).20 The gradations of value thus identified could be used to search for variation in the social uses of the artefacts in question. Drawing from the discussion above, the next section will consider the specific context within which networks of exchange were embedded and the systems of value which were constructed. I will look both at the long- distance movement of the artefacts as well as the smaller scale social relationships

132  Uthara Suvrathan

of which objects were a part by assessing gradations of value within broad categories of goods. Differences in the distribution of goods of different values might distinguish social and horizontal relationships within producing and consuming communities. Similarly, scales of social relationships can be analysed by looking at the distribution and kinds of patterning within a site, the location of workshops, for instance. Widespread distribution of specific goods can indicate horizontal relations, while a differential distribution (within or between sites) may indicate some amount of vertical differentiation as Lesure suggests. In the latter case, additional data, especially on the ethnographically known use of beads and ceramics, and ethnoarchaeological studies of bead and ceramic production can further help us make meaningful analogies to assist in the reconstruction of past social systems.

Studying Ceramics On a broad scale, if the sources of specific ceramic types can be unambiguously identified then the proportions of these types in assemblages can tell us something of the modes and routes of movement, and also the changing nature of this contact. For this kind of analysis sites cannot be studied in isolation. But it is here that we run into the problem of the lack of consistent reporting across archaeological sites. Assemblage composition is not uniformly represented (by weight or by count) nor do we have detailed information on recovery techniques. Despite these limitations, certain interesting patterns and avenues for future research can be identified. First, comparative data clearly brings out the nonhomogenous nature of the Indian Ocean system and highlights the importance of examining variations in ceramic form and fabric that are useful in studying local or regional contexts of use. In fact, each site was first involved in regional networks before being drawn into the transoceanic system. Both Berenike and Tissamaharama were closely involved in local (Egyptian and south Asian) networks respectively before evidence for their participation in longer scale exchanges. At Berenike, for instance, in the earliest Ptolemaic period (midthird to second century BCE) settlement (located at a different area than the later settlement due to the silting of the harbour), locally produced Nile silt Amphorae and large spouted jars/kegs were the most common ceramic types recovered at the site.21 Tissamaharama, between the third and first century BCE, seems to participate in ceramic technologies widespread throughout south Asia, most notably Black and Red Ware (henceforth BRW). This ceramic category, produced through inverted firing techniques that generate a black interior and a red exterior, has a long history in south Asia (as in other regions of the world, where it is not an uncommon technological strategy), from pre-Iron Age/late Harappan contexts in north and central India22 to Chalcolithic and Early Historic contexts (i.e. from the mid second millennium BCE to the early centuries CE and later). Part of the problem is a definitional one, since the first identification of BRW as a unified group by Wheeler elides

Reconstructing networks of exchange  133

regional and temporal variations in manufacturing choices (form, fabric and surface treatment) that are becoming more apparent.23 Perhaps long-lived ceramic categories such as BRW can be seen as a ‘technology of meaning: a technology which allowed meanings to be produced’.24 That is, while certain aspects of the ceramics, such as surface treatment or firing technologies, may have remained stable over time and space, it might have had different meanings/significance at the local level, and may have been open to reinterpretation or re- encoding in different spatial or temporal contexts. As has been argued at Tissamaharama, BRW exhibited subcontinental shapes but was probably locally manufactured due to the large quantities of the ware that were excavated.25 A related point is that of the identification of the movement of techniques and technologies that were adapted to specific regional contexts both within and across land masses. In south Asia, many items were made in imitation of Hellenistic and Roman fashions (e.g. the terracotta oil lamps found at many sites in India). Amphorae have also been found in large numbers. The majority of those found at Arikamedu contained wine, oil and garum (fish sauce). This could imply either the presence of Roman consumers or the adoption by local elite of foreign tastes and fashions.26 Rouletted Ware (a fine, wheelturned, black-slipped ware) was initially regarded as an import from the Mediterranean region that dated to between the first and second centuries CE. But reanalysis by Begley27 suggests that this pottery can be dated much earlier to the second century BCE. Moreover, only the technique of rouletting shows clear evidence of classical ceramic techniques. In its fabric and shape the ware has close similarities to local pottery28 and mineralogical analysis suggests that the majority of Rouletted Ware was manufactured in the lower Ganges valley area of eastern India and Bangladesh.29 Thus, the distribution of Rouletted Ware reflects the existence of a network of movement throughout the eastern coast of India, extending up to northern Sri Lanka. Widening our discussion from ceramics alone, the distribution of nonceramic artefacts in Arikamedu’s hinterland reflects the existence of larger (inland) distributional patterns. Apart from ceramics, lapidary raw material recovered at Arikamedu seems to have originated from the interior of the subcontinent. While no Roman coins have been found at Arikamedu or at other port sites, several thousand coins (individually or in hoards) have been found inland in southern India with a special clustering in the Coimbatore region. This evidence hints at the existence of internal networks linked to the external trade system and the inland networks that sustained oceanic contacts bears further investigation. Second, apart from the spatial dimensions of the distribution of ceramics indicating connections between regions, a temporal consideration of ceramic distribution at each location also reflects changing participation in Indian Ocean networks. For instance, by looking at the nonlocal ceramics at Tissamaharama, it is possible to comment on how the site constructed and/or was situated in a changing different web of networks over time. During the last half of the first millennium BCE (third to fourth centuries

134  Uthara Suvrathan

­

Reconstructing networks of exchange  135

­FIGURE 7.2 Imported

ceramics at Tissamaharama, late centuries BCE.

Source: Author.

eastern section of the Indian Ocean and in Sri Lanka. The lack of historical sources (especially the presence of a useful maritime manual or a text of that nature) has often led to a neglect of this region. Nevertheless, as this cursory survey of ceramic traditions argues, it is possible to tease out some of the historical developments and changing linkages of sites. Third and finally, there is the possibility that the developmental trajectory of geographically distant sites was linked. At both Arikamedu and ­Berenike the late centuries BCE and early centuries CE represented the peak of their participation in the Indian Ocean system and were followed by a significant decline by the third century CE. At Arikamedu, by the end of the second century CE, the Mediterranean trade with southeast India came to an end, although the settlement itself continued and a few ­third-​­to-sixth century Mediterranean amphorae were recovered.36 At Tissamaharama too, connections with the east coast of India seemed to decline in the early centuries CE, the same period during which Arikamedu’s Mediterranean trade was also on the wane. We therefore seem to be dealing with ­pan-​­Indian Ocean changes in networks. The nature and source of these changes remains an important area for further research and probably has much to do with political and economic developments in all the regions under consideration. Ceramics (whether their technology of manufacture or the actual pots themselves) moved over vast distances, both overland and overseas, delineating the broad outlines of these ancient routes of movement. The evidence of the Periplus indicates that trade between India and Egypt primarily

136  Uthara Suvrathan

­FIGURE 7.3 Imported

ceramics at Tissamaharama, early centuries CE.

Source: Author.

centred on the western coast of India, and that merchant ships did not sail to the east coast but accessed this region through indirect overland trade. While there is a lack of ceramic sourcing analyses that could help directly address this issue, the distribution of some ceramic types and forms does provide interesting insights into trade networks in the Indian Ocean. The consideration of other categories of material culture, such as beads and coins, allow us to further investigate regional and subregional networks.

All that glitters: beads and coins The wide distribution of the I­ ndo-​­Pacific glass beads indicates the interconnectedness of the Indian Ocean trading system and have multiple meanings in varied cultural contexts. Recent work on beads argues for a move away from a mere descriptive classification to a consideration of their symbolic and social value.37 As Dubin points out, ‘[T]hey are kaleidoscopic, combined and recombined in an astonishingly wide range of materials; they express social circumstances, political history, and religious beliefs’,38 not to mention economic connections, trade routes and cultural preconceptions. In the ancient Indian Ocean too, beads had a variety of meanings and it is possible to examine some of the cultural and historical processes that led to variation in artefacts across space and time.

Reconstructing networks of exchange  137

By the second century BCE ‘Indo-Pacific’ monochrome glass beads were made at several sites throughout India using indigenous techniques, most notably the drawing (lada) process where a lump of molten glass with a trapped air bubble is pulled out to make a thin tube that is cut to make beads.39 These beads were traded far and wide, being found in east Africa, the Middle East, Southeast Asia and Egypt between 200 and 1600 CE. Recent research indicates a diversity in both manufacturing techniques and in the sources of the glass, with origins as diverse as the Mediterranean, China, the Near East and Central Asia, in addition to sources in south Asia.40 The Indo-Pacific beadmakers appear to have been in urban centres and ports that were often located close to the mouths of rivers. The bead-making industry therefore seems to have been located at prime locations, and at the juncture of inland and overseas trading networks. Berenike, for instance, was located at the juncture of several routes – from East Africa, from the Mediterranean, from the Indian subcontinent and from Sri Lanka. Arikamedu, on the other hand, was located between the eastern and western sectors of the Indian Ocean network, with connections both to Berenike and to Southeast Asia. Early glass bead making was also located in settlements with evidence for other specialised craft activity and therefore some level of social complexity. The fact that the earliest four sites with evidence of the Indo-Pacific bead manufacturing not only were major ports, but were also the first urban centres in their respective regions, has been used to suggest that there might have been a loose political organisation connecting these sites.41 These sites were Mantai in Sri Lanka; Arikamedu in India; Klong Thom, Thailand; and Oc-Eo, Vietnam. While the connections of the bead trade to early state formation in the region is open to justifiable questioning, the wide distribution of beads does highlight some interesting issues. Not only does the distribution of Indo-Pacific beads reflect larger scale trade routes and connections in the Indian Ocean, but a consideration of some individual sites brings out something of the gradations of value in the patterning of the beads at the three sites. In the first place, while IndoPacific beads could be recognised as a distinct category, the very fact that different regions around the Indian Ocean favour different colours in beads indicates that these beads were not a monolithic cultural category with a uniform meaning. Francis, in his detailed studies of Indo-Pacific beads, has argued that blue-g reen and orange beads were not made at Arikamedu but at Mantai and were also found at the Egyptian sites.42 At Arikamedu purple-blue and reddish brown were the dominant colours  – colours that were scarcer at Mantai and the Egyptian sites. Similarly, black and violet are also found in large numbers at Arikamedu but not at Mantai and are almost absent at the Egyptian sites. What is the importance of colour? Is it culturally determined? Or is it a functional result of the raw materials used at these sites? Could it be that they mark identity or connections to a particular place?

138  Uthara Suvrathan

Apart from distinctiveness in terms of colour, each site was different in terms of the stages of bead manufacture vs. consumption represented. While Arikamedu was clearly a glass bead-producing centre, the other two sites seem to have been mainly consumers, with there being limited (stone) bead production. The Periplus of the Erythraean Sea indicates that a variety of communities were involved in oceanic trade, including the Nabateans, Arabs and Indians. How is social or ethnic diversity to be analysed through material culture, especially in a context in which the groups in question might have been extremely mobile? In India, ethnographic work has indicated that glass-bead manufacture involved complex technological know-how mainly undertaken by men while women did the finishing, the polishing and the stringing,43 and that the same people who made glass beads made bangles as well.44 Abraham suggests that south Indian glass working should be situated within multicrafting contexts since they probably operated as marginal communities in a context where elites were limited.45 Interestingly, archaeological evidence from Arikamedu indicates some amount of spatial differentiation of the organisation of bead manufacture. The tube- drawing phase of bead production was located in the southern area of the site along with evidence of other craft activity in metal, semiprecious stones, ivory and shell, while the cutting of the tubes into beads was located in the northern sector.46 Could this imply that different social groups were involved? It can be noted that there were some differences between the ceramics of the two sectors, suggesting that the social composition might have been different, with a predominance of coarse ‘megalithic’ wares in the southern ‘workshop’ sector, while fine wares were more common in the northern sector. Kelly’s work on the beads from Arikamedu and other southern Indian locations suggests that there was a diverse and heterodox community of practice of bead makers employing a wide variety of techniques and chaînes opératoires.47 Use-value is not a stable phenomenon and depends on the properties of the object, its movement, as well as the contexts of origin and end, all of which need to be considered in any study of beads. While beads could be and often were made from simple objects, their value was linked also to the raw material used and possibly the amount of labour invested in its making. Since glass beads were often made to imitate natural stones; this attempt to recreate natural forms, to provide an alternative to a restricted and valuable raw material, might also indicate that glass was a technology that was more widely accessible. The small size of the Indo-Pacific beads and the possibility of their easy transportation over long distances would have added to their ease of dispersion. There is some evidence of copying higher- quality beads, translating the basic shape and colour into less expensive materials. There are some examples of glass beads made to imitate gems and semiprecious stones by judicious application of grinding into faceted forms and drilling from both sides (typical of stone beads): wound dark brown and black beads imitating onyx; pale bluish beads imitating beryl.48

Reconstructing networks of exchange  139

Coins, like beads, tend to move between different regimes of value. The frequent finds of Roman and other coins have bolstered the argument that currency was the principal means of exchange in long- distance trade networks. But is it necessary to look for a single medium of exchange for all transactions? It is important to consider the possibility that apart from currency, other items (cowries, beads) might have played a role and barter may have also been important. Instead of thinking of the Indian Ocean trade network as a monolithic system/structure, it is more useful to conceptualise it as constituted of several regional/local networks. An analysis of coins from Tissamaharama illustrates some of the ways in which we can take apart the networks that defined a site’s interaction with the Indian Ocean trade system. Most of the information on coins from Tissamaharama comes from a large private collection.49 There are also quite a few chemical analyses of Sri Lankan coins,50 as well as a catalogue of the coins found at the German excavations at the site.51 The evidence from the coins suggests the existence of at least three different geographical, spatial and social configurations of exchange: local (i.e. within Lanka); regional (the south Asian subcontinent); and pan-Indian Ocean. First, we have evidence of local issues of inscribed and uninscribed lead coins that date back to the second century BCE to second century CE.52 The use of lead, a metal that is not found on the island (the closest sources are western and central India and Myanmar), might indicate the import and control of minting. The inscriptions on the coins provide some insights into the diversity of socioeconomic groups that were involved in their issuance, including individuals identified by personal names, householders (gahapati) and royal or administrative groups. It is likely that only certain aspects of the trading network were controlled by centralised authority and the small sample of seven seals and sealings bear the names of kings for the most part.53 Apart from the local issues discussed above, the range of non-Sri Lankan coins found at the site reflects something of the wide variety of networks within which this site was located. In the first place are the coins from the Indian subcontinent, including third century BCE punch marked coins/Karshapana, some of which seem to have been locally minted as indicated by the presence of coin moulds for the Imperial Mauryan series.54 In the absence of a common local currency, does the Imperial Series represent an attempt to create a unified currency? Or was it adopted as a recognisable medium of exchange that was already well established in certain sections of the trading network? As Graeber notes, the establishment of the value of objects is often closely tied to issue of visibility and invisibility.55 As a currency that circulated over a vast area of India, probably under the supervision of a centralised authority,56 we could speculate that the Imperial series would carry some extra value that led to their use so distant from the region of their original use. And finally, not only were many Roman coins also found but many locally minted imitations were also recovered, often of extremely inferior workmanship. All the coins (originals and imitations) are of bronze. Moreover,

140  Uthara Suvrathan

the Roman coins found are essentially small currency as at Berenike. If they have no value as bullion due to the intrinsic value of the metal, then they might have served some purpose in cross- cultural trade. Roman imitations in Sri Lanka are interesting in this context. Can the form of the coins themselves have played a role in creating their value? Could a recognisably ‘Roman’ or ‘Indian’ coin have been valued as such and not necessarily for the intrinsic value of the metal of which it was constituted? Is there a way to tell? If ‘low-value’ but distinctive coins have a long use life (i.e. are found in archaeological contexts much later than the date of their manufacture), can this be taken as an indication of their ‘value’ ? This could suggest that these foreign coins were not merely used as bullion but had some other value/ meaning.57 This brief discussion brings out the different levels of exchange networks within which one site was implicated. It is unfortunate that we do not have many coins/sealings from Arikamedu that would have provided an interesting comparison. There is a dramatic difference between the four coins found in Arikamedu and the much larger number of coins found at the other two sites of Berenike and Tissamaharama. This could be a factor of the erosion and disturbance at the site of Arikamedu. However, given that there have been three excavations at this large site we must look at other reasons for the scarcity of coins at the site. Could it be that currency took a different form here? Were beads used as currency in this bead-manufacturing site? Another limitation of comparative data on coins lies in the inconsistencies and lack of uniformity in archaeological reporting. For instance, in the Berenike report, the coins that have a strong archaeological context are noted. Therefore, here we have a record of the coins (with a certain imperial date) and then also of the context of recovery (dated by the ceramics). This enables Sidebotham to estimate coin use-life. In the Ruhuna report, on the other hand, the coins are themselves dated by the context that they are from and in a few cases another coin (Roman) is used to date the context/layer.58 This poses a major problem if we accept that coins tended to be used over a longer period of time, not necessarily coeval with their date of manufacture/ minting. This also precludes the reader from asking similar questions of different datasets. This paper makes a plea for a cross- cultural consideration of several categories of material goods. Such an approach allows us to identify multiple levels of interaction in the premodern Indian Ocean. The Indian Ocean system can be conceptualised as comprised of a multiplicity of smaller networks and some of these were identified. Apart from tracing the development of long- distance exchange networks, it is also possible to look more closely at the different role of each site in these networks. Berenike, for instance, was clearly a transit point for commodities, and not a large importer of Mediterranean or South Asian luxury goods (beads or wine). On the other hand, the ceramic evidence from the site does point to long-term contact with South Asia and the possible movement of nonlocal merchants. Arikamedu was both a production centre (of beads)

Reconstructing networks of exchange  141

and an importer of Mediterranean goods. The level of Tissamaharama’s involvement in the western section of the Indian Ocean trade appears to be slightly different. While the ceramic evidence indicates close connections with the Indian subcontinent and little to no contact with the Mediterranean and Egypt (no amphorae), the presence of numerous Roman issues (and local copies) shows that the site did participate in this trade. Nevertheless, even though this site was at one remove, as it were, from the main system, its participation necessitated (or led) to its incorporation in a wider monetary system where Roman coins were valued in exchange. It is thus possible to consider and characterise the differential role of the three sites in the system, as a first step towards understanding how the Indian Ocean can be understood as a historical entity structured by a web of complex networks held in place by a variety of big and small nodes. Ultimately, this article represents a preliminary attempt to formulate frameworks for diachronic and synchronic comparative cross- cultural and interdisciplinary research that historicises linkages and disjunctures, porousness and rigidity and the nature of connectedness in the premodern Indian Ocean.

Notes 1 Vimala Begley et  al. (eds.), The Ancient Port of Arikamedu: New Excavations and Researches 1989–1992, Volume I, Paris: Ecole Francaise d'Extreme- Orient, 1996; Vimala Begley et al. (eds.), The Ancient Port of Arikamedu: New Excavations and Researches 1989–1992, Volume II, Paris: Ecole Francaise d'ExtremeOrient, 2004; R. E. M. Wheeler, A. Ghosh and Krishna Deva, ‘Arikamedu: An Indo-Roman Trading-Station on the East Coast of India’, Ancient India, 1947, 2, pp. 17–124. ­ 2 Peter Francis, ‘Some Observations on the Glass Beads of Arikamedu’, Revue Historique de Pondichery, 1984, 30, pp. 156–161; Peter Francis, ‘Beadmaking at Arikamedu and Beyond’, World Archaeology 1991, 23(1), pp. 28–43. 3 Leonard Gorelick and A. John Gwinnett, ‘Diamonds from India to Rome and Beyond’, American Journal of Archaeology, 1988, 92(4), pp. 547–552. 4 E. Marianne Stern, ‘Early Roman Export Glass in India’, in V. Begley and R. de Puma (eds.), Rome and India: The Ancient Sea Trade, Madison: The University of Wisconsin Press, 1991, pp. 113–124. 5 Vimala Begley, ‘Rouletted ware at Arikamedu: A New Approach’, American Journal of Archaeology, 1988, 92(3), pp.  427–440; Vimala Begley, ‘Ceramic Evidence for Pre-Periplus Trade on the Indian Coasts’, in V. Begley and R. de Puma (eds.), Rome and India: The Ancient Sea Trade, Madison: The University of Wisconsin Press, 1991, pp.  157–196; Vimala Begley, ‘Pottery from 1989–92 Excavations: Northern Sector’, in The Ancient Port of Arikamedu, Volume I, Paris: Ecole Francaise d’Extreme- Orient, 1991, pp.  115–286; Vimala Begley, ‘Pottery from the 1992 Excavations in the Southern Sector’, in The Ancient Port of Arikamedu, Volume II, Paris: Ecole Francaise d’Extreme- Orient, 2004, pp. 105–323; Vimala Begley and R. S. Tomber, ‘Indian Pottery Sherds’, in S. E. Sidebotham and W. Z. Wendrich (eds.), Berenike ‘97: Preliminary Report of the Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert, including Excavations at Shenshef, Leiden: Research School CNWS, 1999, pp. 161–182; Vimala Begley and R. S. Tomber, ‘Indian Pottery Sherds’, in S. E. Sidebotham and W. Z. Wendrich (eds.), Berenike ‘98: Preliminary Report of the Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the

142  Uthara Suvrathan

6

7

8 9 10 11

12 13 14 15 16 17 18 19 20 21

Eastern Desert, including Excavations in Wadi Kalat, Leiden: Research School CNWS, 2000, pp. 149–167. V. D. Gogte, ‘XRD Analysis of the Rouletted Ware and Other Fine Grey Ware from Tissamaharama’, in H.-J. Weisshaar, H. Roth, and W. Wijeyapala (eds.), Ancient Ruhuna: Sri Lankan- German Archaeological Project in the Southern Province, Mainz am Rhein: von Zabern, 2001, pp. 197–202; H. J. Weisshaar, H. Roth and W. Wijeyapala, eds., Ancient Ruhuna: Sri Lankan- G erman Archaeological Project in the Southern Province, Volume I, Mainz am Rhein: von Zabern, ­ pp. 197–202. Steven E. Sidebotham and Willemina Z. Wendrich (eds.), Berenike ‘94: Preliminary Report of the Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert, Leiden: Research School CNWS, 1995; Steven E. Sidebotham and Willemina Z. Wendrich (eds.), Berenike ‘95: Preliminary Report of the Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert, Leiden: Research School CNWS, 1996; Steven E. Sidebotham and Willemina Z. Wendrich (eds.), Berenike ‘96: Preliminary Report of the Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert, Leiden: Research School CNWS, 1998; Steven E. Sidebotham and Willemina Z. Wendrich (eds.), Berenike 1998: Report on the 1998 Excavations at Berenike and the Survey of the Egyptian Eastern Desert, Including Excavations in Wadi Kalalat, Leiden: Research School of Asian, African and Amerindian Studies (CNWS), 2000. Arjun Appadurai, ‘Introduction: Commodities and the Politics of Value’, in A. Appadurai (ed.), The Social Life of Things: Commodities in Cultural Perspective, Cambridge: Cambridge University Press, 1986, pp. 3–63. Arjun Appadurai, ‘Introduction: Commodities and the Politics of Value’, p. 4. Nicholas Thomas, ‘The Cultural Dynamics of Peripheral Exchange’, in C. Humphrey and S. Hugh-Jones (eds.), Barter, Exchange and Value: An Anthropological Approach, Cambridge: Cambridge University Press, 1992, p. 21. Margaret Carey, Beads and Beadwork of East and South Africa, Aylesbury: Shire Publications Ltd, 1986; Margaret Carey, ‘Gender in African Beadwork: An Overview’, in L. D. Sciama and J. B. Eicher (eds.), Beads and Bead Makers: Gender, Material Culture and Meaning, Cross- Cultural Perspectives on Women, New York: Berg, 1998, pp. 83–93. Nicholas J. Saunders, ‘Biographies of Brilliance: Pearls, Transformations of Matter and Being, c. AD 1492’, World Archaeology, 1999, 31(2), pp. 243–257. Margaret W Conkey, ‘Style, Design and Function’, in C. Tilley et  al. (eds.), Handbook of Material Culture, London: Sage Publications, 2005, pp. 357–359. Monica Janowski, ‘Beads, Prestige and Life among the Kelabit of Sarawak’, in L. D. Sciama and J. B. Eicher (eds.), Beads and Bead Makers: Gender, Material Culture and Meaning, New York: Berg, 1998, pp. 213–246. Igor Kopytoff, ‘The Cultural Biography of Things: Commoditisation as Process’, in A. Appadurai (ed.), The Social Life of Things: Commodities in Cultural Perspective, Cambridge: Cambridge University Press, 1986, pp. 64–69. Arjun Appadurai, ‘Introduction: Commodities and the Politics of Value’, p. 5. Richard G. Lesure, ‘On the Genesis of Value in Early Hierarchical Societies’, in J. E. Robb (ed.), Material Symbols: Culture and Economy in Prehistory, Carbondale: Southern Illinois University Press, p. 24. Lesure, ‘On the Genesis of Value in Early Hierarchical Societies’, pp. 24–28. Lesure, ‘On the Genesis of Value in Early Hierarchical Societies’, pp. 40–41. Lesure, ‘On the Genesis of Value in Early Hierarchical Societies’, pp. 30–31. R. S. Tomber, ‘The Pottery’, in S. E. Sidebotham and W. Z. Wendrich (eds.), Berenike ‘96: Preliminary Report of the Excavations at Berenike (Egyptian Red Sea Coast) and the Survey of the Eastern Desert, Leiden: Research School CNWS, 1998, 163–180.

Reconstructing networks of exchange  143







­

144  Uthara Suvrathan

40

41 42 43 44 45 46

47 48

49 50

51

52 53

pp. 447–604; Lois Sherr Dubin, The History of Beads: from 30,000 B.C., New York: Harry N. Abrams, 1987. C. Pion and B. Gratuze, ‘Indo-Pacific Glass Beads from the Indian Subcontinent in Early Merovingian Graves (5th–6th century AD)’, Archaeological Research in Asia, 2016, 6, pp. 51–64; Joanna Then-Obłuska and Laure Dussubieux, ‘Glass Bead Trade in the Early Roman and Mamluk Quseir Ports—A View from the Oriental Institute Museum Assemblage’, Archaeological Research in Asia, 2016, 6, pp.  81–103; Marilee Wood, ‘Glass Beads from Pre-European Contact SubSaharan Africa: Peter Francis’s Work Revisited and Updated’, Archaeological ­ Research in Asia, 2016, 6, pp. 65–80. Peter Francis, ‘Beads, the Bead Trade and State Development in Southeast Asia’, in A. Srisuchat (ed.), Ancient Trades and Cultural Contacts in Southeast Asia, Bangkok: The Office of the National Culture Commission, 1996, pp. 139–152. Peter Francis, ‘Human Ornaments’, in E. Sidebotham and W. Z. Wendrich (eds.), Berenike 1998, Leiden: CNWS, 2000, pp. 211–225. Peter Francis, ‘Glass Beads in Asia. Part I. Introduction’, Asian Perspectives, 1990, XXVIII(1), p. 16. Peter Francis, ‘Bead Report IX: Bangles and Beads, Part I’, Ornament, 1983, 6(4), pp. 36–37. ­ S. A. Abraham, ‘Glass Beads and Glass Production in Early South India: Con​­ textualizing ­Indo-Pacific Bead Manufacture’, Archaeological Research in Asia, 2016, 6, pp. 4–15. Peter Francis, ‘Beads and Selected Small Finds from the 1989–92 Excavations’, in Vimala Begley (ed.), The Ancient Port of Arikamedu, Volume II, Paris: Ecole Francaise d’Extreme- Orient, 2004, pp. 447–604; E. Marianne Stern, ‘Early Roman Export Glass in India’, in Vimala Begley and Richard Daniel De Puma (eds.), Rome and India: The Ancient Sea Trade, Madison: The University of Wisconsin Press, 1991, pp. 113–124. G. O. Kelly, ‘Heterodoxy, Orthodoxy and Communities of Practice: Stone Bead and Ornament Production in Early Historic South India (c. 400 BCE–400 CE)’, Archaeological Research in Asia, 2016, 6, pp. 30–50. Anne Sibylle Hannibal-Deraniyagala, ‘Beads from Tissamaharama: A Typology of Sri Lankan Glass and Semi-Precious Stone Beads’, in H.-J. Weisshaar, H. Roth, and W. Wijeyapala (eds.), Ancient Ruhuna, Mainz am Rhein: von Zabern, ­ 2001, pp. 203–226. Osmund Bopearachchi and Rajah M. Wickremesinhe, Ruhuna: An Ancient Civilization Re-visited. Numismatic and Archaeological Evidence on Inland and Maritime Trade, Nugegoda: R. M. Wickremesinhe, 1999. Reinhold Walburg, ‘Technical Aspects of the Lanka-Roman 5th Century Copper Coin Initations: Metallurgy and Manufacture’, Beitrage zur Allgemeinen und Vergleichenden Archaologie, 1994, 14, pp. 329–340; Reinhold Walburg, ‘The Preserved Part of the Matara Coin Hoard, Sri Lanka. Lanka- Roman Imitations of Late Roman Copper Coins in the British Museum, London’, Beitrage zur Allgemeinen und Vergleichenden Archaologie, 1996, 17, pp. 53–68; Reinhold Walburg, ‘The “Maneless Lion Coins” of Ancient Sri Lanka: A Revision’, Beitrage zur Allgemeinen und Vergleichenden Archaologie, 1997, 17, pp. 135–146. Reinhold Walburg, ‘The Coins from Tissamaharama, Godavaya and Ambalantota (1995–1998): An Annotated Catalogue’, in Hans-Joachim Weisshaar, Helmut Roth, and M. Wijeyapala (eds.), Ancient Ruhuna, Mainz am Rhein: von Zabern, 2001, pp. 261–274. Bopearachchi and Wickremesinhe, Ruhuna: An Ancient Civilization Re-visited, pp. 51–60. ­ Bopearachchi and Wickremesinhe, Ruhuna: An Ancient Civilization Re-visited.

Reconstructing networks of exchange  145

­

­

8 MATERIAL CULTURES OF WRITING IN THE INDIAN OCEAN WORLD A palm-leaf letter at the Mamluk court Elizabeth Lambourn

This contribution centres on writing technologies in the western Indian Ocean before 1500; more particularly it explores the communicative role of writing supports in long distance, transcultural interactions. The past decade has seen a welcome rise of interest in the role of writing within the institutions and informational networks of merchants and other elites in the Indian Ocean world.1 Much of this scholarship has focused on the period after 1500, a natural frustration to the medievalist that I am and for whom these later practices are part of an evolving continuum. Of greater concern, however, as a historian of material culture, is the altogether passing attention paid to the materiality of writing and its hermeneutic capabilities.2 It is remarkable to note how often, even among historians working with documents, the physical support takes second place to a primarily textual document. Even within the relatively developed area of scholarship on Mamluk chancery practice, with which this chapter engages, approaches remain distinctly linguistic. The editors and contributors to the volume Mamluk Cairo a Crossroads of Embassies follow essentially semiotic models based on the work of John Wansbrough, notably his 1996 Lingua Franca in the Mediterranean. The material qualities of a document are a ‘meta-language’,3 materials, layout or script are ‘external’ as opposed to ‘internal’ features of a document.4 As a historian of material culture, writing materials and the materiality  – the qualities of being material  – of documents are not only foremost on my mind, I also conceive of them as integral to the production and final interpretation of the document. No text exists without its support. Where better to explore these tensions and entanglements than through the study of a letter that was illegible? For my case study I turn to late thirteenth century Mamluk Cairo and accounts of the receipt at court of what was ultimately an illegible letter from a Sri Lankan ruler. Contemporary accounts of this event nevertheless

Material cultures of writing  147

make the point that writing materials – as well as the complex assemblage of things that sealed and wrapped finished letters, in addition to the envoys that carried them – are never simply passive carriers of a primary written meaning. Rather, the material qualities of writing are imbued with meaning and are themselves legible. This chapter explores the ways that the materiality of an ultimately illegible letter participated in the diplomatic mission, complementing and sometimes contradicting both the written text of the letter as well as the oral parsing of the text offered by the envoys who carried it. In the process, this chapter also contributes new material to our knowledge of the technologies of containment associated both with palm-leaf documents and with Mamluk decrees. The Indian Ocean is well-known for having been a dynamic area of cultural interaction and exchange. This chapter is a reminder that writing materials and writing cultures more broadly were an integral part of these exchanges and are deserving of our attention.

Writing cultures across the Indian Ocean – a macro view Before turning to the Mamluk sources and their accounts of this Sri Lankan embassy, it is useful to review briefly the very different writing cultures and materialities of writing present across the Indian Ocean area. The Indian Ocean is traversed by a series of fascinating interfaces between very different writing cultures: While dominantly paper-based writing cultures marked its very western and eastern ends – first in East Asia and from the mid-tenth century onwards in Egypt and the wider Islamic world – its core, South and South East Asia, retained strong paperless writing cultures well into the Colonial period, preferring a range of plant-based supports such as barks and leaves as well as cloth and metallic writing supports.5 Of course, these interfaces were complex and shifting. Before the adoption of paper across the Islamic world many areas  – from the Eastern Mediterranean, via the Arabian Peninsula and Iranian world to Central Asia – also relied on paperless supports, on papyrus, leather and tree barks. Nepal’s deep attachment to palm leaf, a support dependent on imports from elsewhere in the subcontinent,6 even as it developed indigenous paper making, is an important reminder that paper was not the only support imbued with innate prestige. Mobile individuals represent another layer of action at which these boundaries were confounded. Just as the silk roads facilitated the early circulation of paper and papermaking from China to Central Asia, the maritime silk roads were no less important in bringing paper into the bark and leaf cultures of South and South East Asia, and in some instances in carrying bark and plant-leaf writing cultures into the paper zones. Until systematic studies of interactions between writing cultures across the Indian Ocean area are undertaken, we remain at the level of tantalizing, but nevertheless isolated, anecdotes and rare survivals. Just two suffice to illustrate the rich material available. In 689, in Palembang, then the capital of the Buddhist kingdom of Srivijaya in southern Sumatra, the Chinese

148  Elizabeth Lambourn

monk Yijing – also known as I- ching or I-tsing – wrote a letter to Guanzhou requesting that Chinese paper and ink cakes be sent out to him in order for him to continue copying Buddhist texts.7 It is evident from this vignette that Yijing considered paper the only appropriate support for the copying of texts that were to be sent back to China, even if the originals would have been written on palm leaf. In the event, as Yijing boarded the ship to hand his letter to the merchant who was to courier it to China, the ship set sail, inadvertently taking the monk back to China – or at least so the story goes. Yijing eventually gathered scribes and materials and returned to Sumatra to continue his work, finally returning to China in 695 having copied over 400 Buddhist texts in the course of the 25 years he spent in South and South East Asia. We do not know whether Yijing was always able to obtain paper for copying texts during his travels, but if only a fraction of these were copied onto paper, the amounts of Chinese paper and ink circulating in luggage destined for Yijing in South and South East Asia would have been significant. It is easy to dismiss Yijing’s account as a symbolic vignette, aimed perhaps at underlining his retention of Chinese scholarly practice in spite of many years spent outside China. Yet 450 years later, documentary evidence from the Cairo Geniza – a unique corpus of documents relating to the activities of Jews from the Islamic Mediterranean in the Western Indian Ocean during the twelfth century – proves that Jewish merchants from the eastern Mediterranean sojourning along the western Indian seaboard had paper and ingredients for ink couriered in exactly the same fashion.8 It is unclear whether all travelling Chinese scholars planned their writing supports so meticulously but Yijing’s example points to at least one early route through which knowledge of paper and the material itself entered the Eastern Indian Ocean and perhaps even South Asia. If scholarly and mercantile activity present two important zones of contact between writing cultures, diplomatic correspondence represents a third and it is on this that I focus here.

The 1283 Sri Lankan embassy to Cairo On 14 Muharram 682 (14 April 1283) the Mamluk sultan al-Mansur Qalawun (r. 1280–1290) received Hajji Amin al-Din Abu cUthman and Shaykh cAli Lukanti envoys from a Sri Lankan ruler since identified as Bhuvanaika Bahu I (r. 1273–1284) of Dambadeniya, a polity in northern Sri Lanka.9 As research into this event by the French historian Éric Vallet has shown, this small embassy was charged with establishing direct trade between northern Sri Lanka and Mamluk Egypt, effectively bypassing the newly enlarged and politically active Rasulid state in the Yemen that was then seeking to regulate and dominate exchanges through the Red Sea.10 Although the initiative never had time to bear fruit as Bhuvanaika Bahu’s kingdom was overthrown the following year by the Indian Pandya king Maravarman Kulasekhara, the embassy itself captured the contemporary Egyptian imagination and subsequently that of Colonial and later scholars, with the result that it occupies

Material cultures of writing  149

a written space far larger than its size or political significance would normally warrant.11 I pick up where Éric Vallet left off, having established the geopolitical context of the delegation, to emphasise what might at first seem a more tangential matter, namely the materiality of the letter or letters presented by the Sri Lankan envoys and its reception in Cairo. While accounts of the Sri Lankan embassy to Qalawun have been known to western scholars since at least the early nineteenth century and frequently alluded to, it was only in 2015 that Éric Vallet published a full record of the Mamluk sources in which the incident is related together with a complete bibliography.12 Vallet’s work has shown that accounts of this embassy follow two, apparently independent, lineages: one grounded in official regnal histories, the other linked more explicitly to the Mamluk chancery. Shortened accounts of the embassy feature in the contemporary history of Qa13 lawun’s reign composed by Ibn cAbd ­al-Zahir ​­ and, almost verbatim, in the regnal chronicle of Baybars al-Mansuri (fl. 1260–1325).14 This textual lineage pays passing attention to the format of the diplomatic correspondence presented. Of far greater significance to the present discussion, however, is the ‘chancery lineage’ Vallet traced back to a now lost work, the Tadhkirat ­al-labīb ​­ ­wa-nuzhat ​­ ­al-​adīb of Muhammad b. Mukarram, a work described as a ‘chancery log-book’ of Qalawun’s reign.15 This provided the source of the earliest surviving extended narrative of the embassy by Ibn al-Furat (1334–1405),16 also repeated by the famous Egyptian historian al-Maqrizi.17 It should come as little surprise that it is the chancery account that gives the most detail about the materiality of the letter presented. Ibn al-Furat’s account describes the letter thus: the letters they delivered took the form of a thin sheet of gold (wa ṣūra ­al-kutub ​­ allatī waṣalat calá aydinihim ṣafīḥa dhahab raqīqa) three ­finger-breadths ​­ (āṣābic)18 across and half a cubit (dhirāgh)19 long or thereabouts, round, circular (mudawwara ḥalqa); and inside it [was] something resembling green palm leaves (shay’ shabīh bi-l-khūṣ ­ ​­ akhḍar) on which was writing resembling Greek (rūmī) or Coptic script (khaṭṭ).20 In both textual lineages recounting this embassy – that is, in both the regnal history and chancery lineages – no person could be found in all of Cairo to translate this missive into Arabic. The names of the two chief envoys – al-Hajj Amin al-Din Abu cUthman and Shaykh cAli ­Lukanti21 – indicate that both were Muslims of a certain standing, one an individual who had undertaken the pilgrimage to Mecca, the other using the honorific shaykh to denote a learned individual. Given their appointment as emissaries, we may guess that they were very possibly Sri Lankan residents or indeed natives. Both were clearly cognisant of the political and economic context in which the letter was composed and it was this that allowed them to relay the letter’s core content in effect to translate the letter, however loosely  – into spoken Arabic. Abu cUthman and cAli

150  Elizabeth Lambourn

Lukanti can be placed within a long line of cosmopolitan intermediaries – almost always merchants, often Muslim  – who had served as envoys between the Central Islamic Lands and the Indian Ocean world in preceding centuries. But it is not this lineage, nor their possible agency in the request itself, that I want to focus on here but on the materiality of the letter or letters presented. Even translated by the envoys, their otherness remains a core concern of Ibn Furat’s account, its packaging, materials and, in the chancery version, its dimensions minutely detailed.

­Palm-leaf ​­ correspondence Abu cUthman and cAli Lukanti presented what South and South East Asianists will recognise as a palm-leaf document, a letter written incised or inked onto the processed leaves of various species of palm, usually either the talipot (Corypha umbraculifera) or palmyra (also known as toddy) palms (Borassus flabellifer). This writing technology is very ancient and was particularly widespread in southern India, Sri Lanka and island South East Asia where it was employed for writing a range of texts.22 In northern India birch bark supports with inked writing were common. Distinct geographical patterns of use are also visible, at least in India and Sri Lanka, with palmyra leaf more commonly used in the south and talipot predominantly in Bengal and western India.23 However, it has long been noted that Talipot is far superior as writing material; longer, wider, lighter in colour with a smooth and supple surface, whereas the Palmyra or Toddy palm leaf is shorter, narrower, thicker, coarser and tends to become brittle and prone to physical damage.24 There seems little doubt that large transregional networks existed across the subcontinent to trade processed leaves. Palm leaves could be written on with ink, but in southern India and Sri Lanka they were incised with a hard stylus and the incised text was only ‘inked’ when it was required to be read. As Dominic Wujastyk succinctly expresses it, ‘in order to read the text, it would have to be wiped with a cloth soaked in oil and lampblack, that would fill the incised letters with dark colour and render the manuscript [or document] legible’.25 Once treated with vegetal oils and suitably wrapped, palm-leaf and bark writings were remarkably resistant to insect damage and the cyclical humidity of the monsoon. Palm-leaf writings also had the advantage that, without permanent ink, they were spared the corrosive effects of many ink ingredients. As one might expect of an account based upon chancery documentation, Ibn Furat’s description appears to show a greater familiarity with South Asian writing supports, correctly identifying the use of palm leaf (khūṣ) in contrast to historian Ibn cAbd ­al-Zahir’s ​­ rather more speculative qishr ­al-jawz, ​­ ‘nut bark’. While few very ancient palm-leaf documents or manuscript survive,26 by far the majority take the form of thin and elongated sheets (see Figure 8.1).

Material cultures of writing  151

­FIGURE 8.1 ­Palm-​­leaf

manuscripts and rolled legal documents, eighteenth and nineteenth centuries, from the family papers of Sreeramula Rajeswara Sarma, Guntur Distric, Andhra Pradesh. Photograph reproduced with kind permission of Dr. Sreeramula Rajeswara Sarma.

In the case of multileaf palm manuscripts, the flattened leaves are usually ‘bound’ by means of twin cords running through each leaf, the whole ‘sandwiched’ between rigid wooden covers and further wrapped in cloth for protection. One might easily assume that the Sri Lankan letter was written on flat p ­ alm-​­leaf sheets, but another common format for p ­ alm-​­leaf writings was to use the full length of a prepared palm leaf and simply roll it up for storage. This format was preferred for documentary and epistolary texts, few of which survive, which may explain why this format is substantially less ­well-​­known or studied compared to the fl ­ at-​­sheet format. Nevertheless, valuable work by Sreeramula Rajeswara Sarma on a variety of eighteenth- and ­nineteenth-​­century manuscripts and documents in his family’s possession indicates that it was the preferred format for legal documents as well as letters. ­Figure 8.1 shows the range of p ­ alm-​­leaf sizes and formats in Sarma’s family collection. Sarma describes six rolls, each one a single leaf that had been ‘inscribed and then rolled up into small rolls of about three to four cm diameter’.27 There is no doubt that this is an ancient format and Sarma was able to identify a small number of earlier references to the common use of this format for letters. In the 1870s the scholar A.C. Burnell noted that in S. India and Ceylon […] palm leaves have always been used for this purpose [letter writing] up to recent times. For this purpose a strip of ­palm-​­leaf is cut in the usual form, and smeared with turmeric or some similar colour for ornament. The ends are split a little way to secure the whole which is folded in a ring, and then fastened by a thread’.28

152  Elizabeth Lambourn

While many Portuguese and later European visitors remarked on the use of palm leaf for writing, Burnell also signaled João de Barros’s mention of the rolled format in his Décadas da Ásia: [regarding] the other things that serve [people] in the manner of our missive letters and common writing, it suffices for the leaf to be written on and rolled up; and to seal it, one binds it with any kind of line [or cord], or a piece of fibre from the same palm tree.29 Although the format is scarcely mentioned in literature on Sri Lankan palm-leaf writings, Sirancee Gunawardana records the continuing use of this format for horoscopes that were written on a long scroll of Talipot ‘and rolled inwards so that the diameter was approximately 1½ inches [38 mm]’.30 If it is clear from these references that the rolled format was widespread in southern India and Sri Lanka, Nepal is home to perhaps the largest collections of rolled palm-leaf documents in South Asia, with the Asa Archives in Kathmandu holding some 1,300 land-grant documents or tamsuk, the oldest dated to the equivalent of 1384 CE.31 These rolls are generally 18 mm to 60 mm in diameter and many are closed with a clay seal.32 The existence of this lesser known palm-leaf format begs the question of which format the palm-leaf letter delivered to Qalawun’s court followed. Although Ibn al-Furat’s chancery- derived account focuses on the form of the letter and its container it is the version in cAbd ­al-Zahir’s ​­ regnal account that offers a clearer clue. He describes the receipt at court of a letter in a golden case in which was something seemingly rolled up – said to be nut bark33 (kitāb fī ḥuqq min dhahab wa fī-hi ​­ shay’ yushabih 34 ­al-tawr ​­ qīla innahu qishr al-jawz). ­ ​­ c

Abd al-Zahir’s account appears to confirm what the cylindrical form of the container already points to, namely that this ‘something seemingly rolled up’ was none other than a typical rolled palm-leaf letter, effectively the appropriate format for epistolary exchanges. Letters written on such a support might extend over several palm leaves, and this perhaps explains Ibn alFurat’s use of the plural kutub, ‘letters’, at the start of his account and the switch to the singular kitāb, ‘letter’, by its end. He concludes with the phrase ‘This is the outline (ṣūra) of what the envoys presented as being in the letter (kitāb)’.35 The rolled palm-leaf format came with its own technologies of containment, quite distinct from the better-studied wooden covers of manuscripts proper. At their simplest these containers were made of palm leaf and Sarma describes how three of the rolled documents in his possession came ‘enclosed inside slightly larger rings of blank palm leaves which function[ed] like envelopes’.36 Figure 8.2 shows one such document with its cylindrical palm-leaf ‘envelope’. Gunawardana also notes how the horoscopes were held in a small ‘case’ made from talipot.37 Elite correspondence, by

Material cultures of writing  153

­FIGURE 8.2 Legal

document on rolled palm leaf (left) with its plain p ­ alm-​­leaf ‘envelope’ (right). From the family papers of Sreeramula Rajeswara Sarma, Guntur District, Andhra Pradesh. Photograph reproduced with kind permission of Dr. Sreeramula Rajeswara Sarma.

contrast, appears to have turned to ­purpose-​­made containers in other, often precious, materials. Song Chinese records offer additional evidence from the other end of the Indian Ocean for the use of p ­ alm-​­leaf rolls in diplomatic correspondence and their containers. The receipt in 977 of such a letter from the kingdom of Brunei was the occasion for equally close scrutiny of this unusual material and format, thus the ­Song-​­shi observes that the letter was enclosed in different small bags, which were sealed, and it was not written on Chinese paper, but on what looked like very thin bark of a tree; it was glossy, slightly green, several feet long and somewhat broader than one inch, and rolled up so tightly that it could be taken within the hand. The characters in which it was written were small and had to be read horizontally.38 A cylindrical gold c­ arry-​­case represented another solution for the transportation of documents in this format and it is this that is mentioned in all the Mamluk accounts of the Sri Lankan ­embassy – a cylindrical case made from a single rolled sheet of gold. Although no version gives the dimensions of the palm leaves themselves, Ibn ­al-​­Furat provides the precise measurements of the container at ‘three ­finger-​­breadths (āṣābic) across and half a cubit (dhirāgh) long or thereabouts’, in other words around 60 mm in diameter by 290 mm long.39 At 60 mm, its diameter represents the higher end seen among surviving rolls; as discussed previously, the Nepalese tamsuks were rolled between 18 mm and 60 mm in diameter, Sri Lankan horoscopes around 38 mm and Sarma’s family documents around 30–40 mm. It is more difficult to extrapolate meaning or comparisons from the length of the gold container. Half a cubit (dhirāgh) or around 290 mm represents several times

154  Elizabeth Lambourn

the height of the largest surviving palm leaf (recorded as around 101 mm), a confirmation perhaps of the earlier suggestion that this was a letter written over several palm leaves. We might speculate that the more elongated cylinder was designed to contain several rolled sheets stacked one on top of the other. A container half a cubit (dhirāgh) or around 290 mm long would be able to hold three rolled palm leaves of the larger size, but even as many as ten of the narrowest leaves.40 At least one early modern letter has survived with one of its original hard containers. This is a letter written on gold sheet, sent by the Burmese king Alaungpaya to King George II of Britain in 1756 and now held in the Gottfried Wilhelm Leibniz Bibliothek in Hanover, Germany.41 The gold sheet retains the oblong format of the palm leaves it referenced, and was rolled just like a ­palm-​­leaf letter. But whereas regular chancery correspondence in ­eighteenth-​­century Burma was carried in bamboo containers,42 this elite communication was wrapped in red paper and inserted in a hard cylindrical container carved from elephant ivory 32 mm high (foot to finial) by 50 mm in diameter. The container survives today in the Hanover collections (­Figure 8.3).43 This was, however, only the innermost layer of containment of a series of ‘envelopes’ not unlike the multiple layers that enclosed the missive from Brunei to the Song court and described earlier.

­FIGURE 8.3 Ivory

container carved for the gold letter written by King Alaungphaya of Burma to King George II of Great Britain (letter dated 7 May 1756). Dimensions: h: 132 mm; diameter: 50 mm. Image reproduced with kind permission of the Gottfried Wilhelm Leibniz ­Library – Lower Saxony National Library, Hannover. Shelf mark: The Golden Letter of the Burmese King Alaungphaya to King George II of Great Britain of 7 May 1756. (Letter accession number GWLB: Ms lV, 571a.)

Material cultures of writing  155

Historian Jacques Leider was able to reconstruct these further layers and describes how the ivory container was itself inserted into: a sealed shwe phyin-eit, ­ ​­ a ‘golden cloth pouch,’ a term which Than Tun interprets as ‘brocade bag’. This bag was kept in a lacquer casket described as being made of polished wood covered with red resin and bearing a golden design. The casket was sealed and a piece of paper with an English text was fixed to it. Finally it was put into what seems to have been a particularly solid, red- colored bag.44 Unfortunately only the gold letter and its ivory casket appear to have survived. The gold container presented at Qalawun’s court was, in other words, a precious equivalent of the simple palm-leaf ‘envelope’ described by Sarma and Gunawardana, the bamboo carriers known from eighteenth- c entury Burma, the fabric bags mentioned in the Song source and the ivory casket made for the Burmese gold letter. It also surely travelled in further boxes, bags and wrappings but of these we have no record. Another clue in the Mamluk sources may confirm this identification of the format and containment of Bhuvanaika Bahu I of Dambadeniya’s letter. Palm leaves not only had distinct qualities depending on whether they came from talipot (Corypha umbraculifera) or palmyra (Toddy) (Borassus flabellifer) palms, it is also clear that, exactly as with paper, different levels of processing were undertaken according to the leaves’ intended final use. The French scholar Gérard Colas has noted that in the past only leaves destined for the copying of manuscripts were subject to complex processing. Instead, ‘until at least the early twentieth century, in India palm leaves were used for correspondence as well as for private documents without any particular preparation’.45 Valuable ethnographic work by Panchanam Bhoi on the processing of palmyra leaves in contemporary Orissa observed how leaves: that are three-to-four months’ old – just when the leaves are spreading – are preferred. They are intact, greenish looking, longer and broader […] the longest and the broadest are considered the best. Next, the leaves are given what is termed balikasa. For days they are buried in sand over which some water is sprinkled. This seasons the leaves. Such leaves suffice, for ordinary purposes or for the zamindar’s account. But for pothis [books, manuscripts], which are intended to be more permanently kept, a further seasoning is necessary. This goes by the name of ­haldi-kasa. ​­ For this, the superior leaves are sorted out and put into a solution of turmeric and sour boiled rice water […]. After they are taken out, they acquire a darker colour and become more elastic.46 Sri Lankan processing appears to have been slightly simpler with leaves boiled, then dried and stored in a rolled format (known as puskola or prepared leaf) until they were sold and underwent the final stages of processing, notably cutting and smoothing.47 Within this context, Ibn al-Furat’s

156  Elizabeth Lambourn

description of the palm leaves as ‘green’ is perhaps not as surprising as it may first seem when we are used to seeing the heavily prepared flat sheets employed for copying texts. We can imagine that lightly seasoned leaves destined for ‘ordinary purposes’ such as correspondence would have retained something of their greenish tinge. The ­Song-shi ​­ also remarked on the ‘glossy, slightly green’ colour of the bark of the Brunei letter.

Envoys carrying letters, letters carrying envoys This vignette of Qalawun’s reign, however peripheral to the main axes of early Mamluk diplomacy, is a valuable reminder that letters carry meaning in a multiplicity of ways and can in fact succeed in their communicative task even when they are effectively unreadable to the intended recipient as was the case here. Before considering the ways that the materiality of the Sri Lankan letter ‘spoke’ to the Mamluk court I want to pause to consider the interrelationships between this letter and its couriers. As Éric Vallet already noted, these were complex and in the Mamluk sources the letter itself initially appears almost redundant. The communicative power of the Sri Lankan letter certainly did not reside primarily in its text and the envoys in fact play a far more central role in the Mamluk accounts than the letter itself. We do not know whether the letter Qalawun received required inking in order to become legible, although this seems likely, but Ibn al-Furat’s account in any case underlines that the text was illegible, with the alphabet redolent of Coptic or Greek but effectively illegible in Mamluk Egypt. The informational content of the letter travelled instead in the memories of the two envoys, whether they had read (or could read) the text of the letter itself, or ‘knew’ its general content because of their understanding of the current political situation. This content was delivered by them orally in Arabic. The complementarity of envoy and missive appears to be a feature of all diplomatic correspondence, and this was certainly the case in the Islamic world. Eva Grob has argued that couriers played an important part in high-level correspondence during the first Islamic centuries48 and while some historians of the Mamluk period such as Lucian Reinfandt see a gradual eclipse of the messenger by this period, in favour of reliance on the written communication,49 others, such as Marie Favereau, firmly maintain that letters were only ever one component of the diplomatic exchange. As she states: beyond the heavy responsibility of delivering letters and presents safely, these ‘ambassadors’ before their time (often the same individuals return), were charged with delivering an oral message that completed the written missive. Thus, the very text of letters was never complete and it can never be regarded as carrying all the information involved in a diplomatic exchange.50 In long- distance diplomatic interactions across writing cultures, as here, letters operated more as catalysts for the oral exchange of information than as primary vectors of information.

Material cultures of writing  157

It seems significant in this context too that, according to Ibn al-Furat’s ‘chancery’ version, the contents of the Sri Lankan ruler’s letter was delivered orally, not only because no one could be found in Cairo to read the letter, but because according to Sri Lankan custom, only the king himself was permitted to read the text of the letter aloud.51 As he states: ‘They indicated that no one was sufficiently worthy to read it [the letter] aloud except for the king (malik) himself’.52 Malik certainly refers here to the Sri Lankan ruler Bhuvanaika Bahu I, whose letter this was and who is termed malik earlier in al-Furat’s text, rather than to the receiving sovereign Qalawun, whose titles were of course al-Malik al-Sultan. This detail appears to point to particular conventions and protocols in South Asian, or at least Sri Lankan, diplomatic correspondence that require further study, but it suggests, at least at first reading, that a royal letter was understood to be the king’s voice. And yet, as the descriptions of the letter, and its case in the Mamluk sources indicate, the letter was far from passive in these exchanges. Before I continue to discuss the nontextual message of the Sri Lankan letter it is worth noting the presence of a second written document in these exchanges at the Mamluk court. Ibn al-Furat’s account indicates that the two envoys at one point presented a second written document. As he describes it: The envoys were asked what was in their letter (kitāb) and they presented a sheet [of paper] (waraqa) on which it was written …. [details of their mission from the ruler follow]. They clarified that the contents of the letter (kitāb) was as follows …. [a paraphrase of the letter’s contents follows]. This is the outline (ṣūra) of what the envoys presented as being in the letter (kitāb).53 The term waraq clearly designates paper  – waraqa ‘a sheet of paper’ – a surprising technological intrusion into what is otherwise a narrative of otherness. As a familiar writing technology this waraqa occasioned no commentary about its material or form; likewise it appears to have been legible to the court, presumably written in Arabic, since Ibn al-Furat simply summarises its contents. Indeed, Ibn al-Furat’s text is the only version to mention the production at court of this sheet of paper alongside the palm-leaf letter. The waraqa was certainly not a formal Arabic translation of the illegible royal letter since the summary of this, as given orally by the envoys, is recorded in the next part of Ibn al-Furat’s account. From what we can gather from this passing reference, the sheet of paper appears to have carried an account of the preparation of the mission and apparent exchanges the envoys had had with the Sri Lankan ruler about the choice of route. One clear possibility is that this was written by the envoys themselves in order to facilitate their safe passage at various stages of their long journey. Avoiding the Red Sea route and the Rasulid authorities their trade mission aimed to bypass, the envoys travelled instead via Hormuz and Baghdad (i.e. via the Il-Khanid territories and their Gulf dependencies). Ibn cAbd al-Zahir’s ­ ​­

158  Elizabeth Lambourn

account in fact lists all the stages of this part of the Sri Lankan envoys’ journey, 45 staging posts between Hormuz and Baghdad,54 an unusual level of detail as noted by Éric Vallet, which may corroborate the existence of this second, route-focused text. This sheet of paper seems not to have been an Il-K hanid ‘laissez-passer’ document as such, or the Mamluk sources would have signalled this, but a narrative explaining the choice of route and overall mission. If written in Sri Lanka before the envoy’s departure the existence of this waraq might provide some evidence for the use of paper among the Muslim community there in the late thirteenth century, however, it is perhaps more likely to have been written in Hormuz, where paper was more commonly available, at the start of what was a complex multistage journey by sea and river.

The material hermeneutics of an ‘illegible’ letter This last complicating element of the story supports to some extent Lucian Reinfandt’s analysis of Mamluk diplomatic practice in the fifteenth century in which he has suggested that ‘if, on the surface, human envoys, officeholders and dignitaries were the actors, their underlying paperwork was the real protagonist’.55 In the sections that follow I turn back to the more fully described and discussed palm-leaf letter in order to explore how even an illegible missive might be an active participant in the process of diplomatic exchange, a physical communicator every bit as voluble as its human couriers. The letter of Bhuvanaika Bahu I might be thought of as carrying the two envoys, as opposed to being carried by them, in the sense that it provided physical proof of their mission along the length of their journey, a tangible physical manifestation of the Sri Lankan ruler. If the writing material itself held little prestige in Mamluk Egypt, the gold of the case certainly communicated clearly across languages and scripts. As medievalist Charlotte Behr and scholars in other fields have demonstrated, golden objects were commonly ‘perceived as prominent instruments of power and rulership’56 and often imbued with additional political dimensions.57 As in Europe, both the Central Islamic Lands and South Asia used trimetallic currency systems with gold at their apex. While more culturally specific meanings of gold remain to be disentangled for both regions it is already possible to signal the fact that Sri Lanka and Southern India had been experiencing a severe shortage of gold since the early twelfth century. As H.W. Codrington noted in his classic study of Sri Lankan numismatics, from the reign of Parakrama Bahu I (r. 1153–1186) onwards the coinage ‘consists entirely of base metal, sometimes washed with silver’.58 Although one cannot help wondering whether the gold case was more simulacrum than the genuine article, gold must have enjoyed an enhanced status in Sri Lanka during this period, while it continued to be a high status material in the Mamluk world. Whatever the purity of the gold from which the case was made, we can be certain that it materialised not only the status of the letter’s author but Sri Lanka’s gemstone wealth and abundance in other rare luxury commodities.

Material cultures of writing  159

The Mamluk sources report that the Sri Lankan king’s letter set out the riches he had access to: Lots of jewels ( jawāhir), rubies (yāqūt) and pearls (lu’lu’) […]. I have brazilwood (baqam), cinnamon (qirfa) and everything the Kārim imports […]. I have 27 fortresses whose treasuries are full of jewels, rubies and pearls (mighāṣāt).59 Yet in dwelling on the physical form of the royal letter, Ibn Furat’s account with its chancery logbook lineage emphasises the degree to which this letter and its case also disrupted epistolary expectations and chancery norms. Most shocking of all, certainly, was the writing support. The reliance of both Ibn cAbd al-Zahir and Ibn al-Furat’s accounts on similes – the support was ‘seemingly rolled up’, it was written on something ‘resembling green palm leaves’, in writing ‘resembling Greek or Coptic script’ – underlines the otherness of this support as much as the script. In earlier centuries the use of palm leaf might not have presented such a cultural shock since we know that until the ‘paper revolution’ forever changed the hierarchy of writing supports in the Central Islamic Lands; in pre- and early Islamic Arabia palm-leaf stalks had commonly been used for writing while bark appears to have had a long history of use in Iran.60 As Jonathan Bloom and others have shown, following the first transfer of paper-making technology to Central Asia in the eighth century, paper quickly became the normative writing support across the Central Islamic Lands, supplanting papyrus, parchment and any other materials commonly used before, at least in principle.61 Once paper became established, however, such uses of plant leaves were clearly seen, if not as inappropriate, then at least as unusual curiosities. Certainly by the turn of the second millennium CE their adoption would have represented a radical departure from the customary writing habits of the medieval Central Islamic Lands. The very material otherness of the letter and its case may have acted in Bhuvanaika Bahu’s favour reinforcing, as Éric Vallet has noted, longstanding preconceptions of the Indian Ocean world as a place of vast riches and strange wonders or cajā’ib. But if gold spoke a common language, the use of palm leaves for writing must have resonated with equally prevalent ideas about the strange customs and wonders of the Indian Ocean islands. If the island of Waqwaq had trees upon which humans grew like fruit, writing on the leaves of trees was perhaps the less surprising custom. Form and content worked together to guarantee the emissaries the attention of Qalawun’s court and indeed the memorialization of this embassy and the letter itself in the Mamluk sources. More significantly perhaps, palm leaf disrupted the, by now widely established, directional and orientational norms linked to paper. Not only was Mamluk chancery practice entirely paper based, paper had also long since radically changed norms and expectations about the format and sizes of documents. Orientation mattered: With paper the writing surface was always longer than it was wide, what we would today call ‘portrait’ format.62

160  Elizabeth Lambourn

Palm leaves, by contrast, produced sheets that were many times wider than they were ‘long’, what might be described as an extreme ‘landscape’ format, quite unlike anything then current in the Islamic world. Frédéric Bauden and Malika Dekkiche have shown how codified the choice of paper size had become in Mamluk diplomatics in relation to the status of the addressee and the genre of document written, with width the paramount measure of status. Rolled documents (darj, pl. durūj) were the norm for official correspondence, made by pasting together multiple sheets of paper. Mamluk governmental decrees could be a full dhirāgh or 580 mm across and, by pasting sheets together, extend to several metres in length. For example, several chancery manuals specify that correspondence with the most prestigious foreign rulers (first category) was written on sheets of Baghdadi paper measuring approximately one cubit wide; less prestigious correspondents received narrower letters.63 Palm leaves were certainly wide, but they were of almost inconsequential length. Not only was the ‘landscape’ format of palm leaves incommensurate with the ‘portrait’ format for paper that had been preferred in the Central Islamic Lands since the tenth century, even the largest Sri Lankan palm leaf could never compete with the largest Mamluk sheet of paper, let alone a roll. In this environment, palm leaves could only ever underwhelm, and it may be this that explains the omission in all accounts of the dimensions of the palm-leaf letter and their emphasis instead on the container that enclosed it and that offered a far more universal measure of prestige: gold. The case itself may have been a familiar form. Although, as Bauden observed, chancery manuals generally pay little attention to ceremonials and practices surrounding the actual presentation of documents, four cylindrical document cases of the Mamluk period are known, all made of engraved brass.64 The oldest identified so far is dated to the beginning of the fourteenth century and now in the David Collection in Copenhagen (Figure 8.4). Though small at 212 mm long by 25 mm in diameter it carries three bands of inscriptions, the most important being a dedication to its patron, the name 65 ​­ The case is not tentatively read as al-Hajj Ahmad al-­Acsar ­al-Khazanchi. only the earliest surviving example of this type of object but also the most lavish of the four, being decorated with silver and gold inlay. While these cases have been described variously as cases for surgical instruments or pen cases, the consensus now seems to be that they were made as document cases. Precedents for the form in the Near East or elsewhere in the Islamic world have not been identified as yet but then the question of how documents were contained and carried has received comparatively little attention thus far. As with the cylindrical cases used to contain rolled palm-leaf documents, the cylindrical form of these brass cases announced that they carried a rolled document (darj, pl. durūj), the norm for official correspondence. The length of the example in the David Collection, at 212 mm, corresponds roughly to a third of an Egyptian cubit and it may eventually be possible to marry this to known widths recommended for correspondence with, or grants to, individuals of specific ranks. As an example of this, Bauden has

Material cultures of writing  161

­FIGURE 8.4 Document

holder, brass inlaid with gold and silver, Mamluk Egypt or Syria, early fourteenth century. Dimensions: l: 210 mm; diameter: 25 mm. David Collection, Copenhagen, 4/1976. Image reproduced with kind permission of the David Collection, Copenhagen. Photographer: Pernille Klemp.

been able to ascertain elsewhere that in the m ­ id-​­fourteenth century, a document awarding a land holding to an amir tablkhana was fixed at half a cubit’s width or 290 mm.66 Material, decoration but above all shape and dimensions, announced the status of the contents: The cylindrical form announced that it was designed to contain an official rolled document, a darj, and its length also spoke of the width of that document and thus the status of its recipient. Though the earliest surviving Mamluk document case is not exactly contemporary with the Sri Lankan embassy and the documentary practices of the 1280s, it comes shortly after and is unlikely to be an entirely novel form. I am not suggesting that Bhuvanaika Bahu consciously planned for his letter to provoke surprise and to prove impossible to read. While he might certainly have chosen to have an Arabic version written, this does not appear to have been done, and there is altogether little evidence for bilingual communications between polities at this period. In terms of writing surface, even had Bhuvanaika Bahu wished to write on a support more familiar to his intended correspondent, paper was certainly a rare commodity. Even in the fifteenth century the Genoese trader Niccolo De’ Conti reported that paper was universally absent in India with the exception of the port of Cambay in

162  Elizabeth Lambourn

Gujarat;67 as he observed, Cambay’s inhabitants were the only ones to use paper; Indians otherwise wrote on tree leaves (i.e. on palm leaf).68 Sri Lanka likely followed altogether similar patterns, the reason I suggest that the envoys’ waraqa describing their route between Hormuz and Baghdad likely originated in Iran.

Conclusions The Sri Lankan embassy to Mamluk Cairo left a remarkable literary trace for several generations afterwards, a sign if it were needed that an illegible letter can nevertheless communicate powerfully, perhaps almost more powerfully than a letter that can be read. The fact that even in the late thirteenth century, after centuries of diplomatic exchanges along the length of the Indian Ocean, a letter might arrive in a language that no one in Cairo could read is a healthy reminder first and foremost of the central importance of envoys and oral missives in diplomatic exchanges at this period. It is noticeable that among the many accounts of previous diplomatic exchanges between South Asia and the Islamic world gathered and analysed by Éric Vallet, none discusses in detail the format of the letter received and one must suspect that the main communication lay with the oral missive delivered by the envoys carrying the letter. It also underlines, however, the comparatively limited mobilities at this phase of Indian Ocean trade. While Sri Lankan products such as jewels, pearls or spices regularly reached Egypt and travelled north of the Mediterranean, into Europe, Sri Lankans on the whole did not. As Éric Vallet concluded in his thorough analysis of the historical context of this mission, the disproportionately large coverage of this modest embassy is due to the fact that it marked an altogether new development in the extent of western Indian Ocean diplomacy, a ‘first contact’ between the sovereigns of Egypt and Sri Lanka, at the threshold of a new era in long distance exchanges between East and West.69 As such this encounter likely belongs to a very particular subset of epistolary culture, one in which the usual hierarchies of legibility were substantially inversed and in which the envoys together with their verbal message took the preeminent role. Nevertheless, I hope to have shown that the form of this illegible letter did nevertheless communicate its own message. Another important point to emerge from this study is the relative neglect of the containment of epistolary texts, and more generally any form of writing containment that is not ‘book’ binding in the strictest sense, or related to the containment of written amulets. The above discussion underlines quite how many other forms of containment existed. Like the much neglected envelope, there is ample room for studies of other types of document container, and indeed text containers more broadly, from the regionally distinctive ­sub-Saharan ​­ tradition of baktar, the leather bags designed to contain loose leaf Qur’ans and other texts, to the elaborate silver cases made in India and elsewhere for Jewish marriage contracts, kettubot, or the textile bags and covers sewn for Jewish prayer books and the Torah. Too often dispersed

Material cultures of writing  163

across disciplines and museum collections, too often separated from the texts they once contained, these technologies of containment deserve their own dedicated study. The scholarly world’s overwhelming interest in text goes a long way to explaining this neglect, but it is impossible not to think of Zoë Sofia’s observation about the very gendered perception of containment in Western academia. Responding to the work of Lewis Mumford, the renowned historian and philosopher of technology during the 1930s, Sofia makes the point that the history of technology has always preferred ‘aggressive tools and dynamic machines’ to the passive, ‘quietly receptive and transformative ‘feminine’ elements of container technologies’.70

Notes 1 Spurred by Miles Ogborn’s path breaking study of the writing and print culture of early British India Indian Ink: Script and Print in the Making of the English East India Company (Chicago: Chicago University Press, 2007), a number of significant monographs and single articles have since developed this research for other mercantile groups. Chapters in Sebouh Aslanian’s 2011 From the Indian Ocean to the Mediterranean. The Global Trade Networks of Armenian Merchants from New Julfa (Berkeley CA: University of California Press) and most recently Gagan Sood’s India and the Islamic Heartlands. An Eighteenth- Century World of Circulation (Cambridge: Cambridge University Press, 2016) examine variously business correspondence, language and writing, and, of course, courier systems. Nancy Um’s recent study of the material culture of writing among European merchants in Yemen, ‘Chairs, Writing Tables, and Chests: Indian Ocean Furniture and the Postures of Commercial Documentation in Coastal Yemen, 1700–40‘, Art History, 2015, 38(4): 718–731 brings welcome light to the furniture and associated writing postures that formed an integral part of European company business while furniture is joined by writing utensils in the final chapter of her monograph Shipped but Not Sold: Material Culture and the Social Protocols of Trade During Yemen's Age of Coffee (Honolu’lu: University of Hawai’i Press, 2017). Another important body of work looks at the more obvious area of early diplomatic correspondence, for a particularly strong focus on the materiality of such letters (to be expected of a British Library curator) see: Annabel Teh Gallop’s ‘Gold, Silver and Lapis Lazuli: Royal Letters from Aceh in the Seventeenth Century’, in R.M. Feener, P.T. Daly and A. Reid (eds), Mapping the Acehnese Past, Leiden: Brill, 2011, pp.105–139. 2 The very important University of Heidelberg Collaborative Research Center for the study of the materiality and presence of writing in nontypographic societies. Backed by German Research Foundation (DFG) funding its affiliated scholars and doctoral candidates have produced since 2014 no less than 18 monographs and edited volumes on this area, all of them open access e-books. A full list and links available from the De Gruyter, the publisher’s, website; see https://www. degruyter.com/view/serial/428997. 3 Malika Dekkiche, ‘Diplomatics, or Another Way to See the World’, in Frédéric Bauden and Malika Dekkiche (eds), Mamluk Cairo a Crossroads of Embassies, Leiden: Brill, 2019, pp. 185–213. 4 Frédéric Bauden, ‘Mamluk Diplomatics: The Present State of Research’, in Frédéric Bauden and Malika Dekkiche (eds), Mamluk Cairo a Crossroads of Embassies, Leiden: Brill, 2019, pp.  47–50. See also Lucian Reinfandt, ‘Strong Letters at the Mamluk Court’, in Frédéric Bauden and Malika Dekkiche (eds), Mamluk Cairo a Crossroads of Embassies, Leiden: Brill, 2019, pp. 218–219.

164  Elizabeth Lambourn



​­

­



­

­



​­ ­

​­

​­ ­

­ ​­ ­

­

​­

­

­

​­

­

​­

­ ​­

­ ​­

​­

​­ ­

­

­

​­ ­

​­

​­

­

​­

­

­ ­

​­

­

​­

​­

Material cultures of writing  165





­







­

​­

­

​­

­

​­

­

­





­

​­

​­

166  Elizabeth Lambourn

34 Ibn cAbd ­al-Zahir, ​­ Tashrīf ­al-ayyām, ​­ pp. 51–52. I follow the French translation and the Arabic edition it follows which read ­al-tawr, ​­ thus ‘quelque chose d’enroulé’ (Vallet, ‘Messagers’, p. 120). An alternative reading is found in Codrington’s famous 1919 article about the embassy where he read al-­tūz, ‘birch bark’ (Codrington, ‘Sinhalese Embassy’, p.  83). Tūz is a non-A rabic word that describes a tree but also the bark products of that tree, including the sheets of bark processed for writing; the term is found in Persian, but its longer etymology is probably more complex. The most famous description of the use of this bark for writing comes from the Iranian polymath al-Biruni who described that way that in north and central India ‘people use the bark of the tūz tree, one kind of which is used as a cover for bows. It is called bhūrja [Betula utilis]’, Muhammad ­ ​­ ibn Ahmad al-Biruni, Taḥqīq mā li-l-hind ­ ​­ min maqūlah maqbūlah fī ­al- caql ​­ aw mardhūlah, E.C. Sachau (trans) as Alberuni’s India, repr., Delhi: Munshiram Manoharlal,1992, p. 171. Baybars al-Mansuri truncates this passage and simply records the receipt of a letter in a golden case. Ibn al-Furat, ­ ​­ 35 Ta’rīkh ­al-duwal, vol. 7, p. 262. ​­ 36 Sarma, ‘From my Grandfather’s Chest of Palmleaf Books‘, p. 224. 37 Gunawardana, Palm Leaf Manuscripts, ­p. 84. Book 489, cited in Carrie C. Brown, ‘An Early Account of Brunei by 38 ­Song-shi, ​­ Sung Lien’, Brunei Museum Journal 1972, 2(4), p. 227, n.3. I am grateful to Annabel Teh Gallop for sending me this reference. 39 See footnotes 18 and 19 above for this conversion. 40 Sircar, Indian Epigraphy, p. 61, n.2, ‘1¼ to 4 inches in height and one to three feet’ or ‘ca. 30 to 101 mm in height and ca. 304 to 914 mm’ in width. 41 On a near contemporary Balinese example, and the history of this writing support see the British Library Asian and African Studies blog entry for 14 April 2016 at https://blogs.bl.uk/asian-and-african/2016/04/a-gold-letter-from-bali. ­ ​­ ­ ​­ ­ ​­ html. As the blog’s author Annabel Teh Gallop notes ‘the use of gold as a writing material has a long tradition in South-East Asia. The National Museum in Jakarta has examples of Buddhist texts in Sanskrit from the 10th century inscribed on gold strips’. 42 ‘Individual rolled-up ­ ​­ sagyun could be carried without any risk of damage and transferred to the addressees in bamboo containers’, Leider: King Alaungmintaya’s Golden Letter, ­p. 97. 43 The golden letter is extensively published, see for example Jacques P. Leider: King Alaungmintaya’s Golden Letter to King George II (7 May 1756). The Story of an Exceptional Manuscript and the Failure of a Diplomatic Overture, Hanover, Gottfried Wilhelm Leibniz Bibliothek, 2009. The letter also has its own ­ ​­ website; see http://der-goldene-brief.gwlb.de/index.php?id=280. I am grateful to Ines Schuchardt and Anja Flek of the Gottfried Wilhelm Leibniz Bibliothek – Niedersächsische Landesbibliothek for sending me the dimensions and making this image available. 44 Leider, King Alaungmintaya’s Golden Letter, p. 103. ­ 45 ‘Jusqu’au début du XXe siècle au moins, les feuilles de palmier furent employées en Inde pour la correspondance ainsi que pour des documents privés et officiels sans avoir été préparées de façon particulière’, Gérard Colas, ‘La Feuille de Palmier, Support d’Écrit dans l’Inde Ancienne’, in Claude Laroque (ed.), Papiers et protopapiers: les supports de l’écrit ou de la peinture, Paris: HiCSA, 2017, p. 125. Available from: http://hicsa.univ-paris1.fr/documents/pdf/ ­ ​­ ­ PublicationsLigne/Actes%20Laroque%202017/07_Colas.pdf. The complete set of workshop papers are available online at http://hicsa.univ-paris1.fr/page.php? ­ ​­ r=133&id=873&lang=fr. 46 Bhoi, ‘Palm Leaf Manuscripts‘, p. 331. 47 Guide to the Collections in the Colombo Museum, Ceylon Part I Archaeology and Ethnology (Colombo: H.C. Cottle, 1912), p. 38. For an excellent video showing

Material cultures of writing  167

the two-stage processing of palm leaves, here into a fully prepared bound manuscript, filmed in 2009 by the Traditional Palm Leaf Manuscripts Preservation Project, Rangiri Technical Centre, Dambulla; see https://www.youtube.com/ watch?v=1G7Nd5Y6UCE with a link through also from the University of Iowa’s Cultural and Textual Exchanges site. 48 Eva M. Grob, Documentary Arabic Private and Business Letters on Papyrus: Form and Function, Content and Context, Berlin and New York: De Gruyter, 2010, ­p. 99. 49 Reinfandt, ‘Strong Letters’, pp. 218–219. 50 ‘Ces “ambassadeurs” avant l’heure (ce sont souvent les mêmes qui reviennent) avaient la charge, outre la lourde responsabilité de mener à bon port lettre et présents, de délivrer un message oral venant compléter la missive écrite. Ainsi, le texte même des lettres n’était jamais complet et il ne peut, en aucun cas, être considéré comme contenant toutes les informations de l’échange diplomatique’. (see Marie Favereau, ‘Comment le sultan mamlouk s’adressait au khan de la Horde d’Or. Formulaire des lettres et règles d’usage d’après trois manuels de chancellerie (1262-v.1430)’, ­ ​­ Annales Islamologiques, 2008, 41, p. 71.) 51 Ibn al-Furat, ­ ​­ Ta’rīkh ­al-duwal, ​­ vol. 7, p. 261. 52 Ibid., p. 261. In Vallet’s translation, ‘Ils indiquèrent que nul n’était digne de la lire à voix haute (yaqra’) sinon le roi lui-même’, ‘Messagers’, p. 122. 53 Ibn ­al-Furat, ​­ Ta’rīkh ­al-duwal, ​­ vol. 7, pp. 261–262. 54 French translation in Vallet, ‘Messagers’, p. 121. 55 Reinfandt, ‘Strong Letters‘, p. 214. 56 The subject has been studied far more rigorously for Late Antiquity and medieval Europe, see the broad ranging discussion and bibliography in Charlotte Behr, ‘The Working of Gold and Its Symbolic Significance’, in A. Pesch and R. Blankenfeldt (eds), Goldsmith Mysteries. Archaeological, Pictorial and Documentary Evidence from the 1st Millennium AD in Northern Europe, Neumünster: Wachholtz Verlag, 2012, pp. 51–58. 57 Behr, ‘Working of Gold’, p. 52. 58 H.W. Codrington, Ceylon Coins and Currency, Colombo: A.C. Richards, 1924, p. 73. With thanks to Shailendra Bhandare for his clarification of the situation (personal communication 12 September 2019). From Ibn cAbd al-Zahir’s account, rendered into English from Éric Vallet’s 59 French translation of the Arabic original, see ‘Messagers’, pp. 120–121, for Ibn Furat’s slightly different version see pp. 121–122. 60 For a useful overview see the section ‘Which materials were written upon?’ in Peter Stein, ‘Literacy in Pre-Islamic Arabia’, in A. Neuwirth, N. Sinai and M. Marx (eds), The Qur’ān in Context: Historical and Literary Investigations into the Qur’ānic milieu , Leiden: Brill, 2010, pp. 257–263. 61 Jonathan Bloom, Paper Before Print. The History and Impact of Paper in the Islamic World, New Haven, CT: Yale University Press, 2001. 62 I am grateful to Alison Ohta for pointing out the difference between chancery practice and Egyptian conventions in the copying of Qur’anic manuscripts where the ‘landscape’ format was retained until the late twelfth/early thirteenth century (personal communication 30 September 2019). 63 Dekkiche, ‘Diplomatics‘, pp. 200–201 and n.114. 64 See Carine Juvin, ‘Civilian Elite and Metalwork: A View from the Edge‘, ASK Working Paper 29, Bonn, 2018, p. 10 who identifies the example in the Musée des Arts Décoratifs, Paris, AD 5598 (fifteenth century, Egypt or Syria, L: 258 mm; Diameter: 46 mm) and another case in the Museum Civico Correr, Venice (inv. no. XII 7, Mamluk, L: 228 mm; Diameter: 47 mm) and the David Collection example

168  Elizabeth Lambourn

65 66 67

68 69 70

(4/1976, Mamluk, early fourteenth century, l: 210 mm; diameter: 25 mm). Sylvia Auld includes the Venetian example and notes another, making four, in the Victoria and Albert Museum, London (Inv. no. 1429–1855, l: 205 mm; diameter: 45 mm) (see Sylvia Auld, Renaissance Venice, Islam and Mahmud the Kurd. A Metalworking Enigma, London: Altajir World of Islam Trust, 2004, Catalogue 9.1 and 9.2, pp. 304–305, now available online at http://collections.vam.ac.uk/item/ O76489/box-unknown/). ­ ​­ The Paris example is now in the Louvre and is illustrated in Bauden, ‘Mamluk Diplomatics’, p.  63, Fig. 1.2; image also available online at http://cartelfr.louvre.fr/cartelfr/visite?srv=car_not&idNotice=34486. Reading by Manijeh Bayani, 1999, ‘mimmā cumila bi-rasm …’ ‘made by the order of ….’ catalogue entry kindly shared by Dr. Joachim Meyer at the David Collection, Copenhagen (personal communication 6 August 2019). Frédéric Bauden, ‘The Recovery of Mamluk Chancery Documents in an Unsuspected Place’, in Michael Winter and Aamalia Levanoni (eds), The Mamluks in Egyptian and Syrian Politics and Society, Leiden: Brill, 2004, p. 71. Stating that ‘nam papiri usus per uniuersam Indiam abest, excepta Combaita ciuitate’, see Poggio Bracciolini, De Varietate Fortunæ Livre IV, Michèle GuéretLaferté (ed. and trans.) as De l’Inde. Les Voyages en Asie de Niccolò De’Conti, Turnhout: Brepols, 2004, p. 106. ‘Combahitæ soli papyri usum habent: cæteri omnes Indi in arborum foliis scribunt’, Bracciolini, De Varietate, p. 158. ­ Vallet, ‘Messagers’, p. 129. ­ Zoë Sofia, ‘Container Technologies’, Hypatia, 2000, 15(2), p. 185.

PART 3

Anchoring the coasts Introduction Himanshu Prabha Ray

The chapters in the previous section moved the discussion beyond the paradigm of trade to the question of defining the world of sailing in the ancient period. In the context of practical navigation, the coastal features along the shore provided markers. It was the visibility of landmarks that created chains of perceptibility from one vantage point to the next and served both to express the relationship of individual localities to one another and to make sense of the wider world.1 These coastal features include shrines, forts, lighthouses and so on. In the 1980s, Michael Pearson argued that a history of the Indian Ocean needs to take into account littoral society and the outward-looking character of the communities who faced the sea. He reiterated this in a later publication: ‘Rather than look out at the oceans from the land… a history of an ocean has to reverse this angle and look from the sea to the land, and most obviously to the coast’.2 A term that Pearson used in the context of analysing connections of coastal centres was ‘umland’, defined as an area culturally, politically and economically related. In the context of the present work, where the focus is on knowledge traditions, the concept of umland helps link the sacred geography of religious architecture along the coast, with ancillary shrines, tanks, ritual networks and routes of pilgrimage located in the interior as well as on the littoral. It draws into dialogue riverine routes leading inland from the coast as well as to the ‘foreland’ across the seas. This would bring into the conversation not only the diverse usage of sites and shrines as centres of social integration but would also weave in data from ‘movable’

170  Himanshu Prabha Ray

or ‘portable’ ritual objects, such as bronze gongs, musical objects and religious imagery on coins.

Coastal shrines On the west coast of India, in Gujarat, coastal temples were dedicated to a variety of deities, ranging from the non-Sanskritic fertility goddess lajjagauri whose shrine dated to first century BCE was excavated from the site of Padri in the Talaja tahsil of Bhavnagar district of Gujarat hardly 2 km from the Gulf of Khambat to temples of Surya or Sun and other gods along the Saurasthra coast from the sixth century onwards. From the tenth to thirteenth century, the primary route was along the coast from Dwarka on the Gulf of Kachchh to Somnath on the Saurashtra coast and Bhavnagar at the head of the Gulf of Khambat. The coastal centres of Somnath and Dwarka were well known for their magnificent temples, though the beginnings of these sites date to the early centuries of the Common Era.3 Clearly this is an issue by no means restricted to western India, but needs to be examined in other parts of the coasts of the Indian Ocean world. It is evident that the coast itself underwent radical transformations in the colonial period as ‘port cities’ were built to meet the demands of steampowered cargo vessels in the nineteenth century. Given these dynamic transformations of the coast, the chapters in this book accept sea space as a plural assemblage of spatial representations, practices and imaginations related to the sea, seen as social space. Sea spaces may be categorised as coastal spaces; mobile space represented by the boat; island spaces; and the sea as an imagined space, as discussed in the first section. An important component of this landscape is coastal architecture and its interlinkage with travelling groups who moved both across the sea as well as on routes into the interior. This shift entails re-establishing the centrality of the sea and viewing it not merely as a space that permits movement, but as a site of intertwined cultural encounters and shared experiences, as expressed through representations on the walls of temples, but also through performance and music and coins. Worship in the Hindu temple from around the fifth to sixth centuries CE involved bathing of the image, application of sandal paste, flowers, lighting of oil lamps and performance of vocal and instrumental music and dances, cost of sacrifices and offerings and maintenance of servants of the gods. Recitation from sacred texts including the Epics, the Ramayana and the Mahabharata was a crucial part of ritual activities further reinforced by representations of themes from literature in narrative panels on temple walls. The Arthaśāstra of Kautilya mentions devadāsīs (Book II.23) connected with the temples who learnt the arts of music and dancing and makes a distinction between the dancer who performed for the gods (devadāsī) and the one who performed on a public platform (nartaki). The Nāṭyaśāstra treatise on dramaturgy ascribed to the Indian sage Bharata (ca. third to fifth century) discusses in detail related disciplines

Anchoring the coasts  171

such as dance, drama and music. A commentary on the Nāṭyaśāstra was added by Abhinavagupta in the tenth and eleventh century. In the first chapter of the Nāṭyaśāstra the author describes the performance of the samudramanthan or churning of the ocean thereby taking ocean imagery to a wider audience. It is suggested here that rather than ‘reading’ the visual narratives through the textual tradition, as has often been the case, these should instead be contextualised in terms of the social milieu of the monument itself. Visual representations are then to be read at multiple levels: as signifiers of the content of philosophical thought, as indicators of the vibrant traditions of the performing arts and as symbols of the social and cultural ethos.4 References in the Ramayana itself indicate that the story was performed ‘on the streets and royal highways’ of Ayodhya, a popular function that has continued to the present day and one that has a bearing on its visual representation (Uttarakanda 84.4).5 Valmiki taught the Ramayana to Kusa and Lava who recited it to their father Rama (I.4). Similarly, the eighth-century work Uttara-Rāmacaritam by Bhavabhuti commences with a scene in which Lakshmana invites Rama and Sita to view the newly painted murals depicting their own story. The amazing continuity of tradition and the tenacity with which these forms have survived at various levels of social strata also present a challenge for any serious researcher and it is here that visual representations provide invaluable clues. An example may be quoted from the ninth-century temple complex of Prambanan in central Java. In addition to its magnificent architecture, the temple is well known for the narrative sculptures on its walls. Sixty-two panels carved around the outer balustrade of the main temple portray dancers and musicians together with 70 reliefs showing a trio of figures in graceful standing poses, usually one male and two females, but sometimes three males together or three females together showing groups of dancers.6 These dance reliefs have drawn the attention of scholars since the 1950s. One of the first Indian scholars to write on them was Kapila Vatsyayan who identified it as one of the karaṇas mentioned in Bharata’s Nāṭyaśāstra.7 This initiative was further supported by the Indonesian stalwart Edi Sediyawati who discussed the relationship between Indian and Indonesian dance forms.8 The karaṇa of the Nāṭyaśāstra can be seen on several Indian temples and the best-known date from the tenth and eleventh century, such as Chidambaram, Brihadisvara at Thanjavur and Kumbakonam. As discussed by Alessandra Iyer, the 108 karaṇa until recent times have been mistaken for poses, without realising that what is depicted in sculpture is only a portion of a whole movement and may be equated to a still photograph of a moving dancer.9 These temple images were sculpted at a time when Abhinavgupta wrote his commentary of the Nāṭyaśāstra in the eleventh century.10 The Prambanan temple is significant also in other ways, as it denotes a crucial stage in the development of a Javanese idiom. The middle of the ninth century marked a transition in the use of Sanskrit for royal inscriptions in Java and the gradual increase in the use of Old Javanese, as

172  Himanshu Prabha Ray

is evident from the Siwagraha inscription of 856 CE. These developments have been attributed to political changes and the rise of the Sanjaya dynasty culminating in the 30-year long reign of Rakai Kayuwani Dyah Lokapala. The number of edicts (over 60) issued by this monarch ‘easily outnumbers the surviving inscriptions issued during the reigns of all his predecessors combined’.11 Another area where there is a shift in the ninth century is in musical terminology and the use of many types of musical instruments as represented on temple walls. Around the ninth century musical terms start to appear in inscriptions, especially in Java. In Java and Bali, the data suggests a shift from a bamboo music culture to a bronze music culture with the use of gongs and bells in the courts and temples, while the villages maintained a bamboo and wood tradition in music-making. ‘Gongs assumed a sacred and prestigious position as a technology of bronze casting developed and diversified into larger musical ensembles that are extant today’.12 Flat gongs of bell metal are well represented in temple reliefs in India, such as at Amaravati on the Andhra coast dated at least from the second century BCE as well as in Java, and were found in shipwrecks dating to the tenth, eleventh and thirteenth centuries. Two or three stringed boat lutes are represented on the temples at Borobudur and Prambanan in central Java and often follow Indic models. This pattern differs from that available from mainland Southeast Asia, especially from Khmer epigraphy, where Sanskrit musical terms associated with temples occur from seventh to fourteenth century CE. It is also significant that there does not seem to be an overlap between the pattern from Java and Bali and that with the temples at Angkor. Another theme favoured for representations on temple walls are watercraft and sailing ships.

Boats and ships on temple walls Representation of boats in art and architecture is an integral part of the importance attributed to seafaring activity by society and is a widespread practice at sites across the Indian Ocean. In the Indian subcontinent these representations are found mainly in two historical periods: the first from the second century BCE to the seventh century CE and the second from the eleventh to the fifteenth centuries CE.13 This trend is continued in the terracotta temples of Bengal dated to the eighteenth and nineteenth centuries CE. One of the largest preserved brick temple is that of Kantanagar on the banks of the Dhepa River in the Dinajpur District of Bangladesh. The temple was built in the eighteenth century (probably started in 1704, interrupted around 1722 and completed in 1752) by the Dinajpur Raja, a Hindu landlord family (zamindar) that possessed large territories in the Dinajpur area. The occurrence of a European ship on the north face and several figures with European dress sets the historical context of the scene, with the growth of European trade in eighteenth-century Bengal.14 Interesting data on representations of boats and ships comes from the eighth-century Buddhist monument of Borobudur in central Java with

Anchoring the coasts  173

exquisitely carved 1,460 stone panels. Of these carvings, 11 show representations of boats and ships associated with stories from the life of the Buddha, though the watercraft are of local origin, such as the masted ships with outriggers. Inglis has examined the cultural context of the representations in detail and indicates that it shows the unique Javanese perception of the seas and the dangers of seafaring activity.15 It is this local context that often fails to get attention in studies of the stupendous monuments of Southeast Asia. Since its discovery in 1862, all attention at Angkor has focussed on the beauty and aesthetics of the monumental architecture at the site with little understanding of its cultural landscape and its cultural connect with the local village communities. The local villagers believe in the protective spirits neak-ta that protect the land, ponds, forests and rice fields. The temples at Angkor are associated with these local spirits and ritual practices carried out even at present link the villagers with the ancient monuments. It is this local knowledge and cultural practices that sustain the link between the monuments and the people.16 It is the local and regional contexts of seafaring activity that are highlighted in this section. One final point has to be emphasised and that is the need to analyse religious imagery on coins, more so at the time of the introduction of new religions such as Islam in the Indian Ocean around the eighth century. The impact of Islam has been discussed from a range of topics,17 including prayer,18 law19 and literary works such as stories, poems, genealogies, histories and treatises on a broad range of topics. These ‘contributed to the rise of a common repository of images, memories and meaning that in turn fostered a consciousness of belonging to a trans-local community’.20 In this discussion, coins have seldom found space, which is a theme that this section hopes to address.

Notes 1 Peregrine Horden and Nicholas Purcell, The Corrupting Sea: A Study of Mediterranean History, Oxford: Wiley Blackwell, 2000: 125. 2 M.N. Pearson, The Indian Ocean, London and New York: Routledge, 2003, p. 5. 3 Susan Verma Mishra and Himanshu Prabha Ray, The Archaeology of Sacred Spaces: The Temple in Western India, London and New York: Routledge, 2017. 4 Himanshu Prabha Ray, ‘Narratives in Stone: The Ramayana in Early Sculptures’, in Anila Verghese and Anna L. Dallapiccola (eds), Art, Icon and Architecture in South Asia: Essays in Honour of Dr. Devangana Desai, New Delhi: Aryan Books International, 2015, pp. 201–222. 5 The Ramayana of Valmiki, Volume II, Ayodhyakanda, Introduction and Translation by Sheldon I. Pollock, Princeton: Princeton University Press, 1986. 6 Alessandra Iyer, ‘Prambanan Revisited: A Fresh Perspective on the Dance Sculptures of Candi Śiwa’, Bulletin de l’École française d’Extrême-Orient, 83 (1996): 157–184. Alessandra Iyer, Prambanan: Sculpture and Dance in Ancient Java: A Study in Dance Iconography, Bangkok: White Lotus Company, 1998. 7 Kapila Vatsyayan, ‘Dance Sculptures of Lara Djonggrang (Prambanan)’, Quarterly Journal of the National Centre for the Performing Arts, VI, 91 (1977): 1–14. 8 Edi Sedyawati, ‘The Dramatic Principles of Javanese Narrative Temple Reliefs’, in Bernard Arps (ed.), Performance in Java and Bali. Studies of Narrative

174  Himanshu Prabha Ray

9 10

11

12

13

14 15 16

17 18 19 20

Theatre Music and Dance, London: School of Oriental and African Studies, University, 1993, pp. 174–185. Iyer, ‘Prambanan Revisited’, p. 159. Earlier writings on Prambanan by other authors take for granted the relationship with the Natyaśastra but no analysis is ever attempted. See A.J. Bernet Kempers, Indonesian Art, Amsterdam: Van der Peet, 1959; Claire Holt, Art in Indonesia. Continuities and Change, Ithaca: Cornell University Press, 1967. J. Wisseman-Christie, ‘Revisiting Early Mataram’, in Marijke J. Klokke and Karel R. van Kooij (eds), Fruits of Inspiration; Studies in Honour of Prof. J.G. de Casparis, Groningen: Forsten. [Gonda Indological Studies 11] 2001, pp. 42 [25–55]. Arsenio Nicholas, ‘Early Musical Exchange between India and Southeast Asia’, in Pierre-Yves Manguin, A. Mani, and Geof Wade (eds), Early Interactions between South and Southeast Asia: Reflections on Cross-Cultural Exchange, Singapore: Institute of Southeast Asian Studies, 2011, pp. 343–365. Jean Deloche, ‘Iconographic Evidence on the Development of Boat and Ship Structures in India’, in Himanshu Prabha Ray and Jean-François Salles (eds), Tradition and Archaeology: Early Maritime Contacts in the Indian Ocean, New Delhi: Manohar, 1996, pp. 199–224. Sandrine Gill, ‘The Kāntānagar Temple: Hindu Temples in East Bengal under the Mughals’, in Himanshu Prabha Ray (ed.), Archaeology and Text: The Temple in South Asia, New Delhi: Oxford University Press, 2007, pp. 124–146. Douglas A. Inglis, ‘The Borobudur Vessels in Context’, Master’s thesis, Texas A & M University, 2014, http://hdl.handle.net/1969.1/153576, accessed on 12 April 2020. Senthilpavai Kasiannan, ‘Cultural Connections Amidst Heritage Conundrums: A Study of Local Khmer Values Overshadowed by Tangible Archaeological Remains in the Angkor World Heritage Site’, unpublished PhD thesis, The University of Sydney, 2011, pp. 154–160. Sebastian Prange, Monsoon Islam: Trade and Faith on the Medieval Malabar Coast, Cambridge: Cambridge University Press, 2018. Stephen Headley and David Parkin, Islamic Prayer across the Indian Ocean: Inside and Outside the Mosque, London and New York, 2000. Elizabeth Lambourn and Carol Symes (eds), Legal Encounters on the Medieval Globe, Kalamazoo: Arc Humanities Press, 2017. Ronit Ricci, ‘Islamic Literary Works in South and Southeast Asia’, Journal of Islamic Studies, 21, 1(2010): 1–28.

9 PRACTICES OF FAITH The coastal shrines of ancient South Arabia Salila Kulshreshtha

Frankincense! Frankincense! You are the one who is going to the heavens; keep away from us, the enemy, and protect us from the hatred of the friend and enemy.1

In Arabic the word for frankincense is luban, signifying whiteness or purity. The use of frankincense has been known from across civilizations for centuries. It is used in religious rituals, as also for purification, sanitation and for medicinal purposes as it continues to do even today. Varieties of frankincense and myrrh are extracted from trees that grow in southern Arabia, Somalia, Ethiopia, on the island of Socotra and in India. In ancient South Arabia frankincense was a sacred commodity and its harvesters worked under ritualistic restraints. It was stored in temples and burned as offerings to the gods, particularly in conjunction with sacrifices, at least since the third century BCE. Numerous incense burners of this time have been found from various archaeological sites of South Arabia; these were small cube-shaped altars with a cavity on the top and usually with four short legs.2 In 2000 UNESCO listed the Dhofar coast of Oman as the Frankincense Trail and a World Heritage Site, which was later renamed in 2005 as the Land of Frankincense. As per the UNESCO listing: The frankincense trees of Wadi Dawkah and the remains of the caravan oasis of Shisr/Wubar and the affiliated ports of Khor Rori and Al Baleed vividly illustrate the trade in frankincense that flourished in the region for many centuries as one of the most important trading activities of the ancient world. The site’s Outstanding Universal Value constitutes of four elements, which illustrate the history, significance and continuity of frankincense

176  Salila Kulshreshtha

FIGURE 9.1 Map

showing coastal shrines of ancient South Arabia (drawn by Uma Bhattacharya).

trade in the region. The group of archaeological sites in Oman represent the production and distribution of frankincense, one of the most important luxury items of trade of the Old World. The four components of the listing include three archaeological sites: (i) The Oasis of Shisr, lying 180 km north of Salalah, which was an agricultural oasis, and a caravan site and an important station for water supply on the routes from the hinterland to where the frankincense was brought on the port; (ii) Khor Rori and (iii) Al Baleed, which were successive ports and fortified settlements where frankincense was stored and traded. The fourth site in the listing is the region ofWadi Dawkah (the Dhofar region) where frankincense trees still grow and are harvested. The four sites in Oman (Figure 9.1) are nominated by virtue of embodying the cultural sites and landscape in which the trade of frankincense took place, from the Neolithic period to the Middle Islamic Period, northward through the desert routes and eastward and westward through the maritime routes going to India, and China and East Africa.3 In this unique listing UNESCO has demarcated a geographical area, which includes both cultural heritage as well as natural heritage. However, the UNESCO listing and delineation of what constitutes the heritage of

Coastal shrines of ancient South Arabia  177

these ancient sites limits the scope of understanding the complexity and networks of these very early settlements in light of the role of frankincense as a commodity of luxury trade without acknowledging its importance in ritual practices. The four sites, which are included in this listing, emphasise the procurement, storage, trade and export of this luxury commodity. A major lacuna with this listing is that there is no indication of the local consumption and circulation of frankincense. If one moves away from this narrow prism of overemphasising the centrality of trade as the defining feature of ancient cultural networks and understanding the history of settlements around the Indian Ocean as being port cities and ancient harbours, the sites can provide useful information about several cultural and ritual linkages. Coastal sites are significant not only as markers of connectivity, but as nodal points linking the different cultures of the Indian Ocean world, both over land and across the seas. The objects themselves – in this case frankincense – create social relationships and hence are a means of communication across a shared cultural ethos such as the Indian Ocean.4 The objective of this chapter is to examine the centrality of frankincense in the religious rituals performed in the shrines of ancient South Arabia where it was produced. The listing leaves out local ritual use of frankincense, and the shrines from the region provide crucial evidence of this. There are no inscriptions or religious records that outline the performance of rituals; evidence can only be gathered from archaeological records such as the remains of shrines and their cultural assemblage. The issue becomes more pertinent since at Khor Rori, one of the three archaeological sites listed by UNESCO, extensive remains of three shrines have been excavated that provide evidence for the use of frankincense in the rituals performed at these shrines. Through the course of this chapter I will examine the ancient, pre-Islamic shrines of South Arabia: the layout of the shrines, their chronology and geographical location; the deities to which these were dedicated and the distinctive rituals performed in these and then move on to discuss the three shrines found at Khor Rori. Finally, I will explore the ritual linkages between these shrines and argue how frankincense was an important ritual commodity, which circulated through the region along with pilgrims who travelled between these shrines and was not merely an item of luxury trade that was only exported outside ancient South Arabia.

The shrines of ancient South Arabia Several studies have focused on the almost three identical temple complexes found both within and outside of the city walls of Khor Rori, based on the epigraphic material and on surface exploration of the sacred structures.5 One of the earliest discussions was published by Frank P. Albright in 1953,6 who besides discussing in detail the layout of the temple found next to the building containing a freshwater well identified it as a temple belonging to the Hadramawt kingdom who ruled over Southern Arabia, in the region, which constitutes Yemen today. Subsequent works continued to draw

178  Salila Kulshreshtha

connections of the temples at Khor Rori with other Hadramawt temples. J.F. Breton of the French Archaeological Mission along with Yemeni Centre for Cultural and Archaeological Research in Aden made a general survey of Wadi Hadramawt of Southern Arabia in 1978–1979 when he explored seven temples, many of which were dedicated to Syn or Sayin, the moon god of the Hadramawt, as evident from the inscriptions found at the site.7 On the basis of these, Breton has argued that it was quite possible that in ancient Hadramawt an original temple plan in a ‘homogenous form’ was developed and these were attached to ancient villages and cities.8 He cites the examples of dedicatory slabs found in the temples of Ba Qutfah and Al-Hurayda, more than 100 miles apart, with inscriptions dated between fourth century BCE and first CE to show that these reflect a deep sense of cultural and religious unity to the whole area. Keeping in view this overall homogeneity, three kinds of religious monuments of the Hadramawt have been identified:9 (1) rock sanctuaries, (2) sanctuaries with rectangular ground plan and (3) sanctuaries with nonrectangular ground plan. Alexander V. Sedov, on the basis of epigraphy, architectural layout of the shrines and the cultural assemblage of the sites, has been able to reconstruct the religious architecture and give an overview of the religious rites performed at the shrines and the composition of the Hadramawt pantheon.10 One of the most important contributions of Sedov has been the way in which he has linked the different temples of the Hadramawt kingdom through a relationship of ‘interaction or cosubordination’, in which Sayin, the lunar deity, appears to be the supreme deity of the Hadramawt.11 Sedov established a network of temples and traced the cultural relation between kingdoms of north Arabia and south Arabia and also with others in the Near East.12 In a more recent study Christian Darles looked at the typologies of shrines found in ancient South Arabia based on the examination of more than 200 structures that were identified as sacred, out of which he says only 60 have been excavated and published.13 Darles has not only examined the Hadramawt shrines but also those belonging to the other kingdoms of South Arabia. He emphasises the fact that the shrines need to be studied not as isolated structures but in comparison to each other as well as in the context of their environment, inhabitants, chronology and architectural composition. By understanding the composition of the sanctuaries as well as their transformation over time, not only their regional, geographical and political role can be understood but also a change in their religious philosophies such as the shift from polytheism to monotheism. He has identified at least 34 temples associated with the Hadramawt of which he says the oldest can be dated to the ninth century BCE. The large number of shrines listed by Darles reaffirms the observations of Pliny the Elder in his Natural History where he recorded at least 60 temples in Sabota (later Shabwa), the capital of the Hadramawt kingdom and a heavily fortified town.14 Some of the shrines and temple complexes associated with the Hadramawt discovered so far include (Figure 9.1): a religious complex with several temples dedicated to different divinities

Coastal shrines of ancient South Arabia  179

excavated at the oasis of Raybun; three temples at Shabwa; one at Ba Qutfah in Wadi Hadramawt; several shrines at Makaynun; sanctuaries at Al Hurayda in Wadi Amd; and Juja in Wadi Hadrm; three temples in Sumhuram (Khor Rori); and a shrine in Bir Ali. The religious complex at most Hadramawt settlements included two kinds of shrines: the temple extra muros, located outside the city walls and the temple intra muros, located inside the city walls. The extra muros temples in most instances were located on the steep walls of wadis, generally not far from the settlements. The temples varied in size and comprised of a single temple or an elaborate religious structure with several buildings. At some sites, the sacred space could also extend to the unbuilt space comprising agricultural lands, sources of freshwater, etc.15 On the basis of excavations at different settlements it is possible to draw broad conclusions about the layout of the shrines dedicated to the different gods of the polytheistic Hadramawt pantheon. The core of each temple was a rectangular building with a hypostyle hall which served as the sanctum, where the cult connected to the god’s worship was officiated. This building was constructed on a high stone platform, the entrance to which was through a four-columned propylon reached through a stone staircase. The temples were on a west–east axis and the entrance mostly opened to the east. The cella or the sanctum was located in the central building, in the heart of the temple. Almost all excavated structures have revealed an altar located just opposite to the entrance and in all the shrines this position of the altar is clearly defined. The cella was divided into three parts with the use of columns – the altar being located in the middle of the central part. The main part of the altar was a cubic platform, and a three-stepped staircase led to this platform. In quite a few temples in front of the altar an offering table/sacrificial table in stone with a long gutter was found. The ritual arrangement around the altar also included stone slabs with relief images of animals such as goats and ibex, vessels for collecting freshwater or other liquids for libation and a large incense burner placed on a high column. In front of the altar, a multicolumned covered gallery was found with rows of stone benches, which in some temples were also found plastered. What is interesting to note about these shrines is that while the benches were covered with a roof placed on columns (the hypostyle temple), the altar was situated under open skies and this part of the temple had no roof. At several shrines, votive objects such as bronze spearhead, dagger, mortar, jewellery and stone incense burners have been found around the altar. Most temples have also shown the presence of a fireplace or some kind of oven. The temple of Syn at Raybun, called in inscriptions as the Khafas temple, for instance, has a large fireplace in the centre of the hypostyle hall with fragments of large vessels probably used for cooking of ritual meals. In other temples such as at the extra muros shrine at Khor Rori these ovens have been found in subsidiary rooms along with low, square and rectangular-shaped stone benches. These have been regarded as a part of a ritual kitchen where people came to take part in ritual banquets. The

180  Salila Kulshreshtha

indirect supposition that this could be a refectory could be found in the garbage layers, which includes heaps of leavings including bones of various animals and fish. The interior of temples seems to have been quite simple and monotonous with no imagery found on the walls. However, dedicatory inscriptions seem to have played an important role in almost all shrines and have been found inscribed on walls, pillars, benches, stairs and blocks of stone. Besides inscriptions, special dedicatory stelae, called musnad, were mounted inside and outside the structures. These are frequently found associated with shrines of Syn. In certain temples such as at Marib there were special rooms where these steles could be kept.16 In contrast to the plain walls, in some temples life-size human and animal statues could have been placed as evident from special pedestals found for these, which have small hooks to hold these statues. In addition, fragments of sculptures and animal figurines have also been found from various shrines. The podium and pedestals seen in temples may also have had bronze images of these animals mounted on them. For instance, at Sumhuram it has been suggested that the intra muros temple dedicated to Syn had a bull figurine. Archaeologists suggest that while these figurines could have been of sacrificial animals some of these might have been zoomorphic manifestations of the deities; for instance, Sayin had two animal manifestations: the eagle and the bull.17 Others have suggested that the life-like human figures such as those found at Shabwa were anthropomorphic representations of the main deity to whom the temples were dedicated18 or alternatively were statues of noblemen and aristocratic donors of the temples.19 A series of Hadrami coins found in several temples show an eagle on the reverse and a human figure with long ringlets on the obverse, and these have been identified as representations of Sayin.20 On the basis of inscriptions found from the shrines it is known that Sayin or Syn was the federal god of the Hadramawt. Syn is identified as a lunar deity and through inscriptions he can be traced as early as the sixth century BCE. Temples dedicated to Syn have been found at least in Raybun oasis, Wadi Amd and Wadi Idim, and Shabwa. It has also been suggested that he may have been worshipped in different regions and by different epithets. An examination of epigraphic evidence from different shrines has helped to reconstruct the polytheistic pantheon of the Hadramawt where each region had its own popular deity.21 Deities serving as patrons of a clan, family or tribal confederacy occupied different structural levels in the pantheon. In addition, there were three federal deities, Almaqah, Sayin and Amm, who were worshipped all over the kingdom.22 ‘Deities played a vital sociopolitical role and that their cults served as the focus and expression of tribal or federative cohesion and loyalty’.23 Appeals and dedication were made to the different deities for harvest, recovery from illness and wellbeing. The other important deities of the pantheon included Athtar, the father of Sayin who held a dominant position in the pantheon of entire South Arabia; Hawl, another lunar or solar deity; and a goddess, named

Coastal shrines of ancient South Arabia  181

Dhat Himyan, who was equated as a solar deity responsible for ripening of crops and hence also associated with fortune and fertility. It is possible that each of the deities of the pantheon had a temple dedicated to him or her somewhere in the kingdom. A temple dedicated to Syn has also been found inside the city walls of Khor Rori, along with at least two other shrines: an extra muros temple located at the cliff of the promontory and a second temple located inside the city walls near what has been called the Monumental Building, which contains the only freshwater well in the city. All these shrines have been excavated and their cultural assemblage has been subject to analysis. In the next section I will briefly discuss the architectural layout of the three shrines at Khor Rori and examine how much they conform to the patterns of other Hadramawt shrines and give an idea of the rituals performed in these sanctuaries, especially those involving frankincense. The temple extra muros at Khor Rori stands as an isolated building to the northwest of the city close to Wadi Darbat, only a few hundred metres above water level as is the case at many Hadramawt settlements. Dated between the third and the first centuries BCE, archaeologists have suggested that this small temple was in use only during the earliest period of Sumhuram’s history since being located so close to the water, it may have been prone to frequent flooding, which was the cause of its final damage and abandonment.24 In its layout and orientation, the temple conforms to the other Hadramawt shrines. Remains of a stone altar as well as a sandstone offering table have been found in the shrine.25 A hollow cavity on top of the altar suggests that it might have been meant for holding a container such as an incense burner. Material remains found from this shrine include a highly decorated incense burner, pottery shards, fragments of bronze, two spindle whorls in bone and shell, 10 chlamys townsendi shells used as oil lamps and a necklace made of 73 shells of different families. Five of a total of 12 stone objects found inside the temple were meant to crush products, which archaeologists have suggested were frankincense or other resins.26 It may be possible that because of its very early chronology, this shrine differs in many ways from the typical Hadramawt temples while it does have the characteristic of a ‘hypostyle temple’ with stone benches and podium though the grand staircase is missing.27 The temple intra muros also known as the temple of Syn has been found located inside the city walls against its northwest corner and has a similar layout as the extra muros temple. Excavations have suggested at least two phases of construction of the shrines: The first phase between the third century BCE to the third century CE and the second between the third to the fifth century CE when several modifications were made to the shrine. On the basis of examination of other South Arabian shrines archaeologists have suggested that these might have contained bronze sculptures – an animal manifestation of the deity (usually a bull), which is associated with Sin Syn.28 The structure of this temple is in the hypostyle form; the sanctum is divided by two rows of pillars with a podium on the axis of the entrance and

182  Salila Kulshreshtha

five additional chambers having different functions. In particular, one room with remains of food leftovers has been identified as the temple kitchen. A large number of ritual objects have been found from this shrine including animal figures, shell incense burners, ornaments, ceramic vase with South Arabian character and a part of an offering table. Several bronze objects including the fragment of a jug with a horse spout, non-South-Arabian coins, a camel pendant and a bronze bell have also been recovered. The two most significant finds include a bronze bowl and a bronze incense burner with inscriptions naming the temple of Syn in the city of Sumhuram. A bronze plaque carrying a human face and an inscription, which mentions the temple of Syn, has also been found. The architectural layout of this shrine was in conformity with other shrines of Syn elsewhere in the Hadramawt kingdom. The urban shrine - or the Palace-Temple is the third shrine at Khor Rori and is located just behind the huge structure called the Monumental Building, which protects a freshwater well, possibly the only source of freshwater within the city.29 This temple has been labelled as the ‘urban shrine of Sumhuram’30 with its structure possibly dating back to the most ancient levels of the city’s occupation, much earlier than the temple of Syn found inside the city walls. The ‘Monumental Building’ shrine stands adjacent to the large imposing building standing is located right in front of the city’s massive gateway, just like the temple of Syn at Shabwa, which stands in front of the city gates. This building was originally labelled as a temple31 but later called a Palace–Temple Complex and more recently as an ‘urban shrine’.32 The urban shrine, similar in the same style as the extra muros temple has a whitish plaster floor, which is lined up with benches also covered in plaster.33 The orientation of this shrine is similar to the temple of Syn, which is along an east–west axis, but their entrances open in opposite directions.34 The main focus of the shrine seems to be the altar, which is now badly preserved. Several architectural elements such as lintels, decorated blocks35 and pillars were discovered along with several incense burners. There is a striking serpent motif that appears in the decoration of small pillars set beside the altar.36 The snake as an ornamental motif is however not a typically Hadramawt element seen in other Hadramawt shrines. The presence of three elaborate shrines at Khor Rori indicates the continued importance of the shrine through different phases of the city’s history. The earliest shrine was located on the coast, accessed from the sea as well as from the city, but was later moved inside the city, protected within the heavy fortifications of the city walls. At a later phase within the protection of the city walls not one but two shrines would have been in existence – the shrine of Syn as well as the shrine housed adjacent to the Monumental Building.37 The shrines at Khor Rori as well as other Hadramawt shrines, lying protected in the centre of the town, were religious spaces, as well as places for enacting public ceremonies and ceremonial gatherings. but also delineate the geographical boundaries of the political territory of the Hadramawt. However, what makes Sumhuram vastly distinct from other Hadramawt

Coastal shrines of ancient South Arabia  183

settlements is that not only was it geographically detached from the heart of the Hadramawt kingdom but the three shrines here are the only Hadramawt shrines that lie on the coast. It is significant to note that all other Hadramawt shrines such as at Raybun, Shabwa and Makaynun, which were large ritual centres, all lay in the wadis in the inland and some even at the confluence of two wadis. Having said that, in what ways were the three distant shrines at Khor Rori connected with the other Hadramawt shrines? The key to this lies in tracing the threads of cultural and ritual connectivity, which link the shrines of Khor Rori with other shrines and settlements in South Arabia.

The linkages and interconnections of the coastal shrines I would like to apply two disciplinary models that have been used to study sacred spaces located on the coasts to trace the web of linkages and connectivity that the coastal shrines of Khor Rori were enmeshed in. The first model can be derived from Nicholas Purcell and Peregrine Horden’s work The Corrupting Sea: A Study of Mediterranean History38 where they explain how the sea provides an important channel of mobility where relationships are ‘anchored’39 in the landscape but through navigation is also paralleled in other cultic topographies. They also argue that the sacred spaces on the coasts as well as those on islands in the ancient world would have also served as crucial points of communication between cultures. The second significant study that I draw upon is by Barbara Kowalzig who has looked at the classical world and the cult of goddess Aphaia on the island of Aegina, (Greece) and suggested that religious sites should not be seen in isolation but as parts of a larger, interlocking system of cults tied together through myths and rituals on a local, regional or even on a macroregional level.40 She describes how littoral sanctuaries in particular should be thought of as ‘nodes’, while the myths relating them to one another function as ‘ties’, which link these shrines into clusters of religious practices and often materialised through the idea of pilgrimage. If we examine the shrines at Khor Rori by virtue of their geographical location highlighting their location on the coast and apply some of the issues raised in these studies at least three kinds of linkages can be deduced for the shrines: The first and the most obvious connectivity is seen in the similarity in the architectural layout of the three shrines at Sumhuram with other Hadramawt shrines. The main features of these shrines include the existence of a hypostyle hall divided into several naves by rows of pillars and benches made of small stones. A large cubed podium stood under the open sky on the wall opposite to the entrance of the sanctum, in front of which a large stone (or bronze as in the urban shrine) offering table was placed with a gutter for liquid to run and get collected in a basin underneath or in the city drains. A special large altar for incense burning was also placed in front of the podium along with several votive incense burners and other objects of donation.

184  Salila Kulshreshtha

It is known from the dedicatory inscriptions that the intra muros temple at Sumhuram was dedicated to the deity Syn. Several plaques made of bronze and stone plaques have been found from the temples, either hung on the walls or fallen on the floor, of this temple, inscribed with dedicatory verses as well as human figures. There is also sufficient evidence available from the three shrines that shows that Syn was depicted in his zoomorphic representation as a bull. A limestone plate with the image of a bull incised on it has also been found from the temple of Syn. It is thus possible to locate the temple of Syn at Khor Rori as a regional shrine within the larger network of shrines dedicated to the various deities of the polytheistic pantheon of South Arabia. From the available material evidence, it is not possible to know if the shrines of Syn were any different from the shrines dedicated to other deities. It is, however, known from inscriptions that Syn was a ‘federal’ deity and his main temple, known as ‘Sayyin dhu’Ilim’, was located at Shabwa, the capital of the kingdom, where a central street of the city began at the northeast gate, dividing the city into two parts and leading straight to the shrine. A similar arrangement is seen at Sumhuram where the shrine adjacent to the Monumental Building was located directly facing the city gates. In a series of inscriptions recorded at Raybun it is clearly illustrated that Sayyin dhu Mayfaan (Mayfaan was probably the name of the shrine of Syn at Raybun) completely submitted to Sayyin dhu’Ilim (Alim or shrine of Syn at Shabwa) thus showing the authority of the deity as well as the hierarchy of the deity and that of the priests of Shabwa.41 Among the sites excavated at Raybun, a the temple dedicated to Syn, called Mayfaan, contains the richest corpus of epigraphic evidence with over 1,100 inscriptions, which help recreate the cult of Syn.42 The inscriptions dedicated to Syn are remarkably different from those dedicated at the goddess temples located at the same site, in their outward appearance, vocabulary, drafting formulae and contents.43 While dedications to Syn are carved on stone stele, those at the goddess temples have been inscribed on blocks of stone. Frantsouzoff has drawn strong parallels between the stele at Mayfaan and bronze plaques found at the Temple of Alim in Shabwa. Based on these inscriptions Frantsouzoff has argued that there was a strong unity of culture in the field of religion, and the cult united the people and communities in one ethnic, cultural and religious community. While the inscriptions belong to different time periods, date from the first to the fourth century CE, worshippers of the national god composed their dedication to him in the same language in both Alim (Shabwa) and Mayfaan (Raybun). The cult of Syn thus may have been performed in a coherent way in clearly set out buildings with prescribed rituals and a system of dedicatory inscriptions. From the material remains found from the Hadramawt shrines it is also possible to trace the ritual interconnections between the various shrines dedicated to Syn, which may have involved fumigation and offerings of incense and libation, ritual ablution, sacrifices and feasting. Based on the evidence

Coastal shrines of ancient South Arabia  185

from the shrines of Sumhuram as well as from other shrines of South Arabia, ritual ablution before entering sacred places seems to be a common feature. In the temple of Makaynun one of the subsidiary rooms was a plastered bath intended perhaps for ablutions. At the temples in Raybun and Hadran stone basins for water have been found at the entrance along with wells in the vicinity. Similarly, at the three shrines at Khor Rori elaborate arrangements for storing water in the antechambers at the entrance of the shrines such as stone basins and large vessels have been excavated, which indicate ritual ablution and purification before entering the shrines.44 However, the most elaborate system of ablution comes from the shrine near the Monumental Building where a freshwater well is located at the centre of the temple.45 To the west of the well is a large stone tank with a hole near the bottom. Water was meant to enter this tank through a sloping platform between it and the well and ran out through a large hole at the bottom into a drain channel lower than the bottom of the tank. The building also contains a series of rooms that are plastered from wall to wall and have been identified as public baths.46 Freshwater was possibly also used in temple rituals as evident from some of the water vessels found in the sanctums of the shrines as well as for cleansing of the offering tables, which appear worn because of exposure to water.47 In addition, it is believed that other liquids were poured as offerings on tables and altars, which had gutters and runnels leading to basins sunk in the floor or draining liquids outside the temple.48 The importance of water to rituals can be reaffirmed by the fact that one of the shrines is situated in close proximity to the only freshwater well inside the city. What is intriguing about this shrine is the very narrow zigzag entrance built between its walls and those of the Monumental Building, which provided the only access to both these buildings to likely restrict access to the area where its freshwater source was located. It is evident from the layout of the shrines that the area near the podium, which included an altar and an offering table, was the most important part of the shrine.49 Blood sacrifices may have been performed here as evident from the findings of animal bones. Along with the bones of a pig, goat, chicken and fish the bones of a whale have been found from the urban shrine.50 Other offerings included fruit, dates and cereal. The floor of the shrine directly in front of the altar has been found covered with a thick deposit of ash, along with nuggets of frankincense, animal bones, small pieces of bronze and bronze objects. Similar evidence comes from the temple of Mayfaan where a small room was found with ash deposit and small pieces of nuggets. It has been suggested that the devotees walked around the altar and trampled upon this ash deposit. That the rituals involved congregations is also evident from the rows of benches found at all temples, where the devotees sat and watched the ritual enactment.51 In addition, several small-sized stone and bronze objects have been found from the temple of Syn and in the extra muros temple at Khor Rori such as mortars made of stone and bronze, other bronze objects

186  Salila Kulshreshtha

such as a bell, an inscribed bowl and pieces of jewellery and glass and shell beads and dices, which may have been a part of donations to the shrines. Remains of kitchens or refectories have been found in most Hadramawt shrines including from the three shrines at Sumhuram, with evidence of cooking including an oven in rooms outside of the sanctum, which indicates that ritual feasting formed a core part of temple rituals. As per inscriptions, Syn is believed to have been a lunar deity who was worshipped and celebrated at holy banquets held inside the temples.52 The epithet of the supreme god of Hadramawt who was worshipped in Shabwa as Sayyin dhu’Iilum has been interpreted as Sayyin of the banquets.53 Ryckmans substantiates his reading of the inscriptions with evidence from Pliny’s Natural History about the temple of Syn in Shabwa, that ‘the whole harvest of frankincense ought to be brought into the temple for the collection of the tithes’. Recent investigation of work has examined pottery assemblages from various sanctuaries of South Arabia and has divided them into three different groups of vessels – tableware, kitchen and cooking ware as well as storage and transport vessels – and indicate their usage in ritual banquets.54 The material evidence available from the shrines of Khor Rori thus indicate links with other temples in the region through routes of pilgrimage where devotees travelled between the different shrines to make offerings and for ritual feasting. The frankincense itself found in great quantities in the city of Sumhuram was not locally grown but brought through land routes from the regions in the interior possibly not only by traders but also by pilgrims. This third factor that also ties the shrines of Sumhuram with other Hadramawt shrines together is the circulation and use of frankincense in the shrines and fumigation being a central ritual. This is evident from the large numbers of incense burners of various types found at the shrines with residue of resin still stuck to their base. In addition, many shells and fragments of pottery have also been found that were used for incense burning. Large stone mortars were also found in the temples at Sumhuram, Raybun, Hadran and Kafas, which may have been used for crushing frankincense as evident from the resin still stuck inside. Large incense burners with pyramidal base have been found placed at the centre of both the extra muros and the intra muros shrines along with many transportable votive burners. Their façades are engraved with images of Syn represented as a crescent moon.55 The incense burners from Sumhuram are found not only from the ritual spaces but from a variety of archaeological contexts including not only religious domains, but also private, houses, streets, fortification walls as well as from rubbish dumps.56 It is also important to note that incense burners from other sites and geographical areas have also been found in Sumhuram, indicating that people from other regions came to worship at the shrine of Syn and offered him frankincense.57

Coastal shrines of ancient South Arabia  187

Conclusion This chapter has explored how unlike other temples dedicated to Syn in South Arabia, the three shrines at Khor Rori, located on the coast, would have served as nodal points that were not just connected with the hinterland but also accessed from the sea and visited by various kinds of people as attested by the material remains found there. The shrines represent a cultural entity that tied together various religious networks in the region, which included the harvest and procurement of frankincense, the ritual use of this frankincense and through notions of pilgrimage and ritual feasting the movement of various people and ritual objects within the region of South Arabia as well as from across the ocean. The objects found from the shrines of Khor Rori indicate widespread interactions with the outside world and a cosmopolitan character of the city. For instance, a substantial presence of Indian sailors at Sumhuram is evidenced by the presence of Indian ceramic types such as the Rouletted Ware and black and red ware found along with a pot-shard with Tamil Brahmi script. The fragments of Rouletted Ware, a variety of fine grey pottery from found at Sumhuram is the only evidence of this ceramic type found anywhere from the Arabian Peninsula and establishes direct maritime links with South Asia.58 One of the rarest objects to be found at Khor Rori is an Indian bronze figure of a female in a dancing pose believed to have come from western India.59 This further underscores the argument that the movement of people across the Indian Ocean did not merely involve the transport of trade commodities but also items of personal use or ritual objects such as the bronze female figurine. J.F. Salles has discussed how the geographical position of Sumhuram helped it to bridge two sailing circuits: one with India and the other with regions of the Gulf with sites such as Ed-Dur and Mleiha.60 This chapter has attempted to study the interrelated web of interactions of a shrine of ancient South Arabia as it was integrated in various regional sacred networks. Most secondary literature on the Indian Ocean has focused on maritime networks, which highlight the role of frankincense in international trade, the trading ports and harbours involved and so on, largely supported by evidence from texts such as the Periplus of the Erythraean Sea that have tended to completely overlook sacred networks. On the other hand, archaeological evidence shows extensive remains of the use of frankincense in the shrines as a ritual item thus connecting the shrines of South Arabia with those in the Persian Gulf, Red Sea and South Asia. There is hence a need to study the various shrines located on the coast of the Arabian Peninsula as facilitating linkages both inland as well as across the ocean through maritime routes.

Notes 1 A Dhofari prayer recited every morning by women before a locally made, terracotta cuboid incense burner, The Dhofar Ethnographic Survey Project, Field, Season January 2013, directed by W. Zimmerle.

188  Salila Kulshreshtha



­ ­







Coastal shrines of ancient South Arabia  189



­



190  Salila Kulshreshtha



10 ENTANGLED TRADITIONS The royal barges of Angkor Veronica Walker Vadillo

The aim of this chapter is to analyse the role of Indian iconography in the establishment of Angkor’s fluvial cultural landscape by leading a contextualised study of boats. This type of study proposed by Van de Noort1 and ­Adams2 revolves around the analysis of the interrelationship between people and boats, and how the resulting dialogue generates the traditions that give form to maritime, or in this case, fluvial cultures. With 34 contextualised nautical scenes depicting over 106 boats available for study, this analysis can be done in considerable detail. In what follows I will discuss riverine watercraft as the functional objects that allowed the people of Angkor to colonise the aquatic space for its resources. Particular attention will be given to the contexts of construction and usage, which seem to have a strong pre-Indian set of beliefs. The discussion will then move to analyse the transformation of functional watercraft into zoomorphic vessels used in culturally charged activities where boats act as liminal agents and are imbued with symbolic meaning. It is in this context of transformation that Indian iconography played a key role. In his study of canoes and sailing in Papua New Guinea, Malinowski3 wrote: ‘a craft, whether of bark or wood, iron or steel, lives in the life of its sailors, and it is more to a sailor than a mere bit of shaped matter [...]’. To Malinowski, the emotional attitudes of people toward their craft were the motor by which to know what boats meant in the eyes of their users.4 This anthropological approach to the study of watercraft results in a more meaningful understanding of boats within their social context, but to reconstruct this for past societies is a challenge, as users are no longer available for questioning. An alternative way of studying boats in their cultural context is by analysing the different elements that affect the life of boats. These can be defined as purpose, technology, tradition, materials, economics, environment and ideology.5 The purpose or function of a vessel is determined

192  Veronica Walker Vadillo

by the needs of a society, but their ability to deliver an efficient watercraft is constrained by technological advancements. Function and technology nonetheless exist within a tradition that embodies the idea of what a boat is and how it must be constructed, and thus boat construction is, above all, social practice.6 As a product of society, boats are carriers of symbolic and ideological constructs that have been put in place to ensure best practice. Failure of boat design comes at great personal and material cost, and therefore it has been noted that boat builders exerted great caution in the introduction of new designs.7 However, as Adams pointed out, barriers erected by ideology are not immutable and are subject to human desire to refine and innovate.8 Thus change occurs as a result of a dialectic process between the numerous factors involved in boat construction.9 Materials, economics and the environment are some of these factors; the availability of materials, the resources involved in the processing of boats and the constraints imposed by the environment limit the development of boat construction, and as a result, alterations to them can lead to changes in boat design.10 But perhaps the most distinctive factor that affects boat construction is the ideological concept that determines what a boat is.11 In this respect, it is the bow and the stern where researchers have to turn if they are to understand the philosophy that governs a society.12 These sections of a boat are probably one of the most prominent features in boat design, as they confer the visualization of what a society thinks a boat should look like. It also becomes an identifiable cultural trait that demarcates the boat-builder’s adherence to a specific group, a display that is as much for those belonging to the same group as it is to those outside the group.13 The production of boats is therefore entangled in multiple elements that affect the design of boats in different ways; however, it is ‘in the social aspects of life surrounding the production and use of ships that ideologies are played out and become most visible’.14 By studying the boats of Angkor as part of fluvial material culture embedded in the human–environment interaction debate it is possible to begin to understand the role of boats in Angkor’s fluvial cultural landscape. The approach taken here is to look at boats as culturally constructed entities ‘endowed with culturally specific meanings, and classified and reclassified into culturally constituted categories’.15 This approach focuses on the dialectic process that takes place between boats and people, and aims to discern how the former is redefined and used throughout its lifetime through the study of its cultural biography.16 As with any human being, the cultural biography of boats can be elaborated on by looking specifically at the construction, use, modification and demise of boats within their social context.17 The corpus of Angkorian nautical iconography is very wealthy both in the number of boats depicted and the detailed representation of nautical elements and boat usage. Of the 106 vessels depicted in the 34 scenes representing nautical iconography, 95 of them are dugouts or raised dugouts, two are planked vessels and the other two are bundle rafts, while the remaining seven are too deteriorated to determine their construction technique. The activities in which boats are centre stage are varied and they include warfare,

The royal barges of Angkor  193

FIGURE 10.1 Nautical

iconography in Angkor contains a great wealth of information, not only in terms of nautical technology, but also on boat usage and location. In this scene from the Bayon temple a noble lady picking a lotus flower appears. The action takes place in a pond, as marked by the stepped shoreline. Source: V. Walker.

pilgrimages to religious sites, state progresses by water, rituals, festivals and economic activities like fishing and trade.18 The environment where the boats are used differs from manmade pools to natural spaces like rivers or lakes. The vast majority of the boats are decorated with zoomorphic shapes (77) and are used by the elite, while only 16 correspond to undecorated boats (14 used by commoners, and two used by court ladies to pick up lotus flowers in manmade ponds; Figure 10.1). Except for one medium-sized vessel, the majority of these undecorated boats are small in size. Although this limits our ability to fully grasp the nautical technology of medium-to-large size vessels used by commoners, the iconographic data of small-size vessels are nonetheless instrumental in understanding the cultural processes that lead to the transformation of boats into mythical creatures from the Hindu pantheon. The general consensus is that the boats of Angkor were dugouts,19 or expanded dugouts,20 a theory that is supported by the iconography and the written account of Zhou Daguan, a Chinese emissary who visited Angkor at the end of the thirteenth century, wherein it is said that very large boats were made with hardwood broken into planks held together with iron nails, while smaller boats were made from a single very large piece of wood, hollowed out by chiselling and then stretched using heat from a fire and wooden timbers.21 According to this source, smaller boats were broad at the centre and pointed at either end. While there are no archaeological remains that can provide additional data on large vessels, the suggestion that smaller boats were made from a

194  Veronica Walker Vadillo

single log of hardwood was confirmed by the archaeological record when two dugouts were found in the area of Siem Reap: the Angkor Thom boat, dated to the thirteenth century,22 and the Kra Raleung boat, dated to the fifteenth century.23 The Angkor Thom boat is an unfinished vessel that shows the same process of wood carving as the one documented in contemporary Cambodia and in the delta region of Southern Vietnam, as will be discussed later on. On the other hand, the remains of the Kra Raleung boat show a series of holes drilled on the boat at equal intervals,24 which are marks that can be interpreted through the ethnographic data as chak bansak: A series of holes drilled on the hulls to even out the walls of the dugout. The use of embers and timbers to expand the walls of the dugout mentioned by Zhou Daguan also resembles the technique documented in contemporary dugouts, and the description of the hull design (large in the middle and pointed at either end) could easily be applied to contemporary Cambodian vessels. In Cambodia, boats are also made of koki trees (hopea odorata), like the Angkor Thom and Kra Raleung boats. Hence, the archaeological, historical and ethnographic record shows a certain degree of continuity. Conservatism among maritime communities is widely documented; since environmental conditions tend to remain persistent (i.e. currents, flow, winds), once an optimal solution has been achieved the introduction of new designs are met with skepticism.25 This is mostly due to what Hunter calls common sense: The optimised design has proven to be efficient in facing the dangers involved in navigating the waters that it was designed for, so change appears superfluous and risky.26 Given the accumulated archaeological, historical and ethnographic evidence, and taking into account the aforementioned conservatism, this study will use ethnographic data to provide a more comprehensive view of the archaeological material of common watercraft that forms the basis of decorated watercraft at the centre of this study.

Khmer boat-building in an ethnographic context In the cultural biographies of objects the first thing that needs to be addressed is who made the object ―in this case the boat― and why.27 In contemporary Cambodia it is usually a man, who will become a private owner, who orders the construction of the boat.28 The construction of a boat has an economic and apotropaic character, as it is expected to provide good business and bring good luck to the family.29 Hence, since its conception the boat is already expected to have its own agency that will affect the safety of the journeys and the prosperity of its owner. Before explaining the rituals surrounding boat-building in Cambodia, it is important to provide a context to these actions, as they form part of the Khmer traditional system of beliefs. According to this system, the world is divided into two different spheres: the world of the forests (prei) and the world of the village (srok). The former is governed by wild spirits while the latter corresponds to domesticated spaces.30 These are not immutable landscapes, as forests

The royal barges of Angkor  195

can be transformed into domesticated spaces, and the forest can take them over if the proper ritual is not followed.31 So even though there is a clear demarcation between the prei and the srok, there is also a fluid relationship between them. There are two ways in which this system of belief affects boat construction today: the processing of the wood and the location of boat-building activities. First, it is thought that the spirits inhabit trees,32 therefore there is a ritual procedure destined to get permission from the spirit before felling the tree.33 The way the wood is processed is also very revealing; the wood is given form with chiselling tools, and the fire used for opening the walls of the dugout is fuelled with the debris of the hollowing-out process. Helms has proposed that as living forms with ascribed supernatural potencies, trees that were transformed into boats likely retained the extraordinary powers of the parent tree,34 a proposition that seems to be in place in Khmer boat-building traditions. The ideological context surrounding work in the forest would have required the boat-builder to carefully manage the intangible forces contained in the material.35 In this sense, the consumption of the boat’s debris from the hollowing-out process in the fire on which the boat is later placed should be seen as part of this material processing wherein the discarded wood is burnt and made part of the boat again through the smoke and the heat. The way in which wood is processed in Cambodia for the construction of dugouts therefore ensures the continuation of the tree’s supernatural powers (and agency) into the boat. The need for other-than-human agency in watercraft that could contribute to safe and profitable journeys explains the survival of this type of wood processing for boat-building in Cambodia up until the early twenty-first century even though other technologies (i.e. plank-built ships) were available through contact with other cultures.36 The second way in which the Khmer traditional system of beliefs affects boat construction is the location of work. The tree is felled in the forest, which is a space controlled by wild spirits, so while the work takes place in this landscape there are a series of taboos related to behaviour (i.e. no loud conversations, no laughter) that need to be observed. Additionally, offerings and rituals have to be conducted to protect the boat-builders from the wild spirits.37 Once work in the forest is finished the boat is called touk kamrol, wild boat; then a ritual is performed to ask the wild spirits for permission to leave the premises and take the boat, and another ritual is performed in the village to ask the domestic spirits for protection.38 The polishing of the boat then takes place, and the village’s shaman takes over the rituals.39 The final transformation of the wild boat into a boat is marked by two actions and accompanying rituals. First, a pair of symmetrical eyes for the boat are carved from wood and placed on the hull. These will help the spirit of the boat see any dangers on the river and avoid collisions.40 The second action is the launching of the boat once it is finished; the owner and the family get on board and the shaman (acar) directly addresses the boat to let it know that in the forest it was a koki tree, but in the village it is now

196  Veronica Walker Vadillo

a boat and its name is luck and prosperity.41 The processes involved in the construction of dugouts in Cambodia are thus conducive to the creation of watercraft imbued with other-than-human agency destined to bring luck and contribute to the safe journey of the owners.42 Malinowski posited that the need for this type of apotropaic ritual stems from the anxiety that the sea provokes in humans.43 This shows that the Mekong River Basin can exert the same kind of apprehension in humans. The archaeological, historical and iconographic evidence clearly shows that dugouts were used during the Angkor era, and that these were made in a similar fashion to the ones documented in the ethnography. More importantly, wood processing was done in the same manner: chiselling the log and never using saws.44 Taking this into account, it seems likely that similar ritual practices and apotropaic attributions were therefore in place during the Angkor period (ninth to fifteenth century), but given the pre-Indian elements present in rituals it seems highly likely that the nautical tradition predates the arrival of Indian merchants. Similar practices have been documented in Borneo,45 which may point to regional pre-Indian traditions. The activities of these utilitarian vessels, such as fishing and transport,46 would have dangers associated with them47 that could result in the loss of cargo, loss of life or both. Creating apotropaic mechanisms that would contribute to safe and profitable journeys suggests that even though the people of Angkor used the Mekong River basin, they perceived it as a dangerous space for which other-than-human agency was needed. Furthermore, the identification of the boat as a living entity is furthered in the way boat parts are named in Khmer, as each part of the boat is named akin to the human body: the bow is called the khbal (i.e. head); the section between the bow and the first thwart is called the ka (i.e. throat); the first thwart is known as the sma (i.e. shoulder); the walls of the boat are called the sach (i.e. flesh); the central section of the boat is the troung (i.e. chest); the ribs are called the chhaeing (i.e. skeleton); and the section between the last thwart and the stern is known as the chrak ka kansay (i.e. heel).48 The ritual behaviour that surrounds contemporary boat construction in Cambodia indicates that the utilitarian boats of Angkor, which the evidence suggests were made in the same way from a technological perspective, possessed the same apotropaic and other-than-human agency attributions as their modern counterparts. The need to create a functional object with independent agency to bring good luck to the owner and help in the success of the journey was likely the result of the perceived dangers of navigating the Mekong River basin. However, the cooperation of the spirit came at a cost in terms of construction process, material processing and ritual behaviour that had to be observed throughout the life cycle of the boat. The perception of boats as living creatures that were owed respect not only marked the way ship-building traditions developed, but also set the tone for the modifications and uses that will be seen below in the context of social practice.

The royal barges of Angkor  197

FIGURE 10.2 Decorated

boats are used by the elite in a variety of settings; the scene located on the west wall of the pond lying just outside the northeastern wall of the Phimeanakas complex, in Angkor Thom, represents seven boats, five of which are carrying noble people toward the stairs in the middle of the pond’s wall. Containers are seen onboard the vessels, and in one instance a nobleman is reaching into a chest held by one of his servants. Source: V. Walker.

From plain forms to zoomorphic shapes Khmer nautical iconography, generally dated to the twelfth to thirteenth centuries,49 has 77 decorated boats out of 106 boat depictions: 45 are used by the elite in peaceful activities, while 32 take part in warfare activities.50 Focusing on barges involved in peaceful activities, these are often of small or medium size, and are found in areas of the temples where access could be controlled (i.e. inside enclosures or in pavilions with doors; Figure 10.2), which marks the private character of these depictions. The only representations of decorated boats in public spaces are found in Preah Khan, where the causeways that lead to the temple are decorated with nautical iconography, as well as the side doors of the west gopura. The most important features of Angkorian-decorated boats are on the one hand that the hulls have been modified and transformed into zoomorphic shapes tapping clearly into Indian iconography, and on the other hand that the boats are used exclusively by the elite in what can be interpreted as socially charged activities. It is therefore clear that the transformation of the boats is linked to social activities of the elite, likely of ritual character given the location of the representations in private spaces and the events that take place in the representations. The archaeological, historical and ethnographic data discussed above suggests that the utilitarian vessels of Angkor were imbued with otherthan-human agency, and that they acted as liminal apotropaic agents between nature and humans. The boats of Angkor’s elite are characterised

198  Veronica Walker Vadillo

by a zoomorphic transformation through decoration; boats are no longer utilitarian objects but fantastic mythical animals with prominent heads and tails, and in some instances, with bodies made of scales. Although there are no physical remains of the decorated boats of Angkor, the extensive iconographic corpus shows these decorations in minute detail. The modification of the decoration suggests a need to amplify the boats’ potency by transforming the common hull into a mythological animal, so decorations need to be considered within their context of usage. Hull decorations (i.e. the main body of the boat) are not very common in boats engaged in peaceful activities, and they only depict makara, a hybrid mix of crocodile, fish, tapir, bird and elephant, which is very common in Khmer architecture.51 Stempost decorations are more common, and they depict the garuda, the nāga and/or the reachisey. The garuda is a mythological bird that represents birth, ambrosia and heaven.52 This bird-like figure is the mount of Vishnu and is invoked in the Mahabharata as a symbol of martial prowess, speed and violent force.53 In Khmer art, its figure is tightly linked to the nāga, a serpent-like creature that is considered to be ‘the guardian of the treasures of the earth, the keeper of the energy stored in water, and the safeguarder [sic] of the prosperity of the region, traditionally related to water availability’.54 The most common combination of these two is with garuda mounting on the nāga, profusely depicted in the balustrades of Angkor Wat. This combination, however, is not present in boat decorations. The reachisey, on the other hand, is another mythical creature of the Khmer pantheon associated with water that has a head of a lion, a small trunk-like nose and is covered in scales.55 Although it seems to have common features with the gajasimha,56 it is thought to be a mutation of the makara.57 The nāga and the reachisey are hard to tell apart, as they both sport clenched teeth; however, the presence of a goatee and/ or scales have been interpreted in this analysis as part of the reachisey. It seems likely that the decorations were related to either the purpose of the event in which the boat took part, and/or to the rank of the user, but this discussion is out of the scope of this chapter as the main interest lies in the processes involved in the transformation of undecorated boats to fantastic animals.

Transforming the myth into reality The earliest depiction of zoomorphic boats in Angkor appears in the twelfth century as part of the mural decorations of Angkor Wat. The carving covers the majority of the wall and contains a single scene of two very large vessels set in a natural environment (Figure 10.3).58 The one in the upper section has a reachisey decoration with its distinctive goatee.59 A contiguous line of paddlers wearing livery and high chignons propel the boat toward the west. There are two elevated platforms filled with women, some with children. A grand central pavilion is located amidships where two men are playing chess;60 the traditional pieces of the game in the Khmer version are visible.61

The royal barges of Angkor  199

FIGURE 10.3 The

figureheads of the two barges depicted in Angkor Wat represent an idea of how the boats used during the Dvaravati Water Festival would have looked. The boat from the lower registry has a heavily ornate hamsa (goose). Source: V. Walker.

The vessel on the lower section has a hamsa, an avatar of Vishnu in the form of a swan or goose, decorating the bow.62 Paddlers are also wearing livery, but they are only spread at the fore and the aft. At the centre there is a couple depicted in a large size surrounded by children. On the left room there are two women embracing each other, while on the right-side room there is another large bearded male figure63 looking at a cock fight that is taking place on the aft platform. The scene seems to depict the mythical story of the Dvaravati Water Festival described in the Harivamsa II: 45–46.64 ­Dvaravati is the city of Vishnu from where he departs on a maritime pilgrimage to a city called Pindaraca.65 Vishnu encourages all those taking part in the nautical parade to indulge in the pleasures of life as they board magnificent vessels shaped like geese, peacocks and sea monsters, with golden pavilions decorated with gold and precious stones and crowned with garlands. Music, singing, dancing and love games are enjoyed by all on board.66 The identification of this scene with the Dvaravati Water Festival story suggests that the boats depicted here are the way in which the artisans of Angkor envisioned the boats described in the story. Political power and legitimacy in Angkor depended on the ability of the king to demonstrate his association with the divine world through ritual

200  Veronica Walker Vadillo

re-enactment and the configuration of centres of power that represented the divine macrocosm in the mundane world.67 These urban centres relied on the Hindu idea that there is a parallel between the macrocosm of the gods and the microcosm of humans. The power of the kings of Angkor thus depended on their ability to demonstrate to their subjects that they could tap into divine sources of power. This was done by creating symbolic associations that attempted to reconstruct the divine macrocosm in the mundane microcosm; in this way, the outer moat, the axial towers and the main tower of Angkor Wat are symbolic representations of Mount Meru surrounded by the ocean and the mountains. If the festival in the southwest corner pavilion of Angkor Wat is interpreted as the Angkorian interpretation of the boats used in the mythological story, then it seems logical that in attempting to reproduce the divine macrocosm into the mundane world the boat builders of Angkor would have copied these models for the boats of Angkor. While the agency of spirits in utilitarian vessels may have been good to exert influence on the mundane world, the transformation of boats into zoomorphic creatures from the realms of the gods suggests that these were more than apotropaic mechanisms to deal with the perils of navigating mundane waters. The transformation of boats into mythical creatures that are believed to inhabit the threshold of the divine world suggests that the elite of Angkor was seeking to colonise the metaphysical space attributed to the aquatic environment and reinforce that connection between the divine and the mundane.68 The correspondence between the Angkorian decorated boats and the divine creatures after which they were modelled would have transformed boats into important symbols of the elite since this symbolic correspondence would mean that by using these boats the owners or users would become the divine counterpart of the creature depicted in the bow. This phenomenon is well documented in Thailand: When the king, who is considered to be an avatar of Vishnu on earth, boards the royal barge Anantanāgarāj, boat and king are considered to be the living representation of the Vishnu Anantaśāyin myth.69 Vishnu Anantaśāyin 70 is a theme from Hindu mythology where Vishnu is represented sleeping on the serpent Ananta on the cosmic ocean from where all things will be created. Vishnu’s dream between creations and the birth of Brahma, who is the creative force that will generate a new universe, are the main elements of the Vishnu Anantaśāyin, while the role of Ananta is to be the floating bed of Vishnu, which to all effects could be seen as a zoomorphic boat (Figure 10.4). Although the first representations of this mythological theme in Cambodia depicted Ananta as a coiled snake, this form was eventually substituted first by the elongated form of the snake, and then eventually by a reachisey. The reasons for this change are unclear, but it may be related to the personification of the ocean in the figure of the dragon-makara-reachisey; Bénisti reckons that Ananta is one and the same with the ocean, and reminds that one of the names of the ocean is makarālaya, the residence of the makara.71

The royal barges of Angkor  201

FIGURE 10.4 One

of the carvings of Vishnu Anantasayin bathed by the waters of Kbal Spean in the Kulen Mountains. Seeing the water touch the body of the snake, it is easy to imagine the serpent Ananta as a zoomorphic boat. Source: V. Walker.

The round sculpture of Vishnu Anantaśāyin that was consecrated in the island temple of the West Mebon in the early twelfth century by Suryavarman  II72 could be construed as a precursor of this symbolic correlation. This colossal figure, which measures 6 m in length,73 was placed at the centre of the West Baray, the largest manmade body of water of Angkor.74 It has been suggested that by placing the round figure of Vishnu Anantaśāyin in the West Mebon Suryavaman II aimed to sanctify the waters, which were both a functional water mechanism and an intensely ritualised space.75 Although it is difficult to tell when the boats of Angkor were transformed into zoomorphic creatures, by the thirteenth century, the multiheaded nāga decoration was already used on the flagship in the festival scene of the Bayon, which suggests that this symbolic correspondence was being used at the time. The use of a multiheaded decoration to evoke the serpent Ananta and associate the user (i.e. the king) with Vishnu and transform the water where they navigate into the cosmic ocean (thus sanctifying them) is a clear example of how this correlation may have worked. The variety of stempost decorations suggests that these were linked to specific activities or specific ranks. In a highly hierarchical society like that of Angkor, where insignia were ascribed to people according to their rank,76 it would be surprising if the decorated barges were not regulated. Examples of this type of regulation are found in the royal barges of Thailand. During the Ayutthaya period the decoration of the barges were linked to the rank of the user: Those belonging to

202  Veronica Walker Vadillo

high-ranking officers had a small central pavilion, while painted and gilded barges were reserved for the two chief ministers.77 The king could gift such a boat to a high-ranking officer, but in that case the boat could only be used by the officer under exceptional circumstances.78 Having a fleet of decorated vessels for use within Angkor would have contributed to the illusion of reflecting the macrocosm of the divine world in the mundane microcosm, so it would have been to the king’s advantage to have their subjects travelling by these boats on official missions. The decorated boats would have therefore also been a reflection of Angkor’s hierarchically organised society. This is similar to the case of the royal barges of Thailand, where the king would send his royal fleet to bring foreign envoys up the Chao Phraya River in the seventeenth century.79 The Thai Royal Barges show that the zoomorphic features of the boats can be emphasised by the clever use of colours and movement, something that was noted by European travellers to Ayutthaya in the seventeenth century. Hence, crew and boat would have been part of a performance that would have brought the boats to life for anyone who saw them from the shorelines, adding further realism to the magnificence of the boats and their users.

Conclusions The message embedded into these zoomorphic barges was one of control and power. As has been shown, the otherworldly character of common boats was used by the elite in their search to reproduce the divine macrocosm in their mundane world, drawing on Indian iconography to transform utilitarian watercraft into zoomorphic creatures that were known to inhabit the threshold of the divine world. This modification of ship decorations allowed Angkorian rulers to colonise the metaphysical space of the aquatic environment and enabled them to tap into the divine as the source of royal power. The inspiration behind the modification of these boats had its origin in mythological stories from the Hindu and Buddhist traditions, like the Dvaravati Water Festival, but as Kopytoff points out, what defines an object’s biography ‘is not the fact that they are adopted, but the way they are culturally redefined and put to use’.80 The pre-existence of a well-established ship-building tradition with an ideology that ascribed agency to boats not only through ritual behaviour associated with the construction process but also through the nature and processing of the wood likely contributed to consolidate the incorporation of decorated boats into the ruler’s reconstruction of the divine macrocosms. The Angkorian utilitarian watercraft was the functional vessel needed to use the water environment for economic purposes, but the anxiety generated by the risks involved in these journeys resulted in a specific behaviour surrounding the creation and use of these boats, and the perception that a spirit imbued with energy living in the boat could help them overcome these dangers. The apotropaic character of this boat spirit appears to have been deeply entrenched in Angkorian society, as that would explain

The royal barges of Angkor  203

the modification of these vessels into zoomorphic beings and their use as part of the elite’s reconstruction of the divine macrocosm. As Stuart–Fox pointed out, the power of the Angkorian kings was dependent on their ability to reflect the heavenly realm of the gods on earth.81 The designs of the decorations may have stemmed from Hindu or Buddhist traditions, but the ability to infuse them with the necessary meaning was embedded in Angkorian culture. The Angkorian mastery of wood carving and the ability to add weight to the boats without affecting their stability was down to a long-standing ship-building tradition, while the savvy use of movement and colour in the rowers’ livery and paddles would have brought the zoomorphic shapes to life. This performative act hoisted the elite into the realms of the gods, a display that likely motivated and inspired those who partook in or witnessed the event. But while stage and performance were crucial, perhaps what really allowed the king to use boats as symbolic representations of magical creatures was the social behaviour surrounding boats. The attribution of agency to boats and the accompanying rituals would have been a fertile ground for the creation and use of a fleet of zoomorphic boats imbued with magical character. The careful choice of decoration shows creatures that acted as liminal agents between the human world and the world of the gods. Water was therefore seen as a threshold of the divine world within the context of boat use, which is consistent with Khmer cosmological traditions where the ocean features prominently.82 Starting from the conceptualization of the Angkorian utilitarian boat as an entity with other-than-human agency endowed with apotropaic qualities, this chapter has discussed how pre-existing nautical traditions intertwined with a selection of Indian motifs to create zoomorphic barges for an elite whose power depended on displaying their connection to the divine world. While the Khmer made a conscious choice to select aspects of Indian culture that best expressed their understanding of the maritime realm, travellers who shared the same Indian influences would have felt a sense of familiarity with – and understanding of – the choice of iconography used in the royal barges. This overlapping of local and foreign traditions to construct new nautical realities are but an example of how intensive interactions through trading routes resulted in the development of new practices at the same time that it supported the spread of an iconographic corpus that helped to anchor the coasts along the Indian Ocean.

Notes 1 Robert Van de Noort, North Sea Archaeologies: A Maritime Biography, 10,000 BC–AD 1500. New York: Oxford University Press, 2011, pp. 38–43. 2 J. Adams, ‘Ships and boats as archaeological source material’, World Archaeology, 2001, 32(3): 292–310. 3 B. Malinowsky, Argonauts of the Western Pacific: An Account of Native Enterprise and Adventure in the Archipelagoes of Melanesian New Guinea, 1978 edition. London: Routledge, 1922, p. 80.

204  Veronica Walker Vadillo











The royal barges of Angkor  205













­

206  Veronica Walker Vadillo





















The royal barges of Angkor  207

71 72

73 74

75

76 77 78 79

80 81 82

reclining over an elongated Ananta resting on the body of a reachisey (Bénisti, 1965: 99; Roveda, 2005: 58). The combination of these two creatures in the same image led Bénisti to suggest that there never was a transformation of the nâga into the reachisey, but rather a substitution that was likely connected to autochthonous traditions (Bénisti, 1965: 99). After Angkor Wat, the Vishnu–reachisey combination becomes the norm while the serpent disappears from the iconographic theme completely (Bénisti, 1965: 101). Bénisti, ‘Représentations khmères de Viu couché’, pp. 105–106. M. Feneley, D. Penny, and R. Fletcher, ‘Claiming the hydraulic network of Angkor with Viṣṇu: A multidisciplinary approach including the analysis of archaeological remains, digital modelling and radiocarbon dating: With evidence for a 12th century renovation of the West Mebon’, Journal of Archaeological ­Science: Reports, 2016, 9: 275–292, p. 290. Feneley, Penny and Fletcher, ‘Claiming the hydraulic network of Angkor with Viṣṇu’, p. 286. Incidentally, these waters were sanctified with representations of Visnu Anantaśāyin carved on the riverbed of the spring that feeds the reservoir as discussed in Feneley, Penny and Fletcher, ‘Claiming the hydraulic network of Angkor with Viṣṇu’, p. 290. Feneley, Penny and Fletcher, ‘Claiming the hydraulic network of Angkor with Viṣṇu’, pp. 290–291. Feneley, Penny and Fletcher have proposed a coiled Ananta as the likely platform of the reclining Vishnu, of which there are no remains. However, it is also possible that Ananta was depicted with its elongated body, making it look like one of the decorated boats of Angkor. See Zhou Daguan’s chapters 2, 3 and 4 where he discusses the different residences, clothing and insignia (palanquins and parasols) ascribed to different ranks in Harris, A Record of Cambodia, pp. 49–52. N. Gervaise, The Natural and Political History of the Kingdom of Siam. 1989 edition. Bangkok: White Lotus Press, 1688, p. 213. Gervaise, The Natural and Political History of the Kingdom of Siam, p. 96. see M. Ribadeneyra, Historia de las islas del archipiélago filipino y reinos de la gran China, Tartaria, Cichinchina, Malaca, Siam, Cambodge y Japón. 1947 edition. Madrid: Editorial Católica S.A., 1601; Gervaise, The Natural and Political History of the Kingdom of Siam; Tachard, G. A Relation of the Voyage to Siam Performed by Six Jesuits, trans. 1981 edition, A. Churchill, Bangkok: White Orchid Press, 1688. Kopytoff, ‘The cultural biography of things’, p. 67. M. Stuart-Fox, A Short History of China and Southeast Asia: Tribute, Trade and Influence, Crows Nest: Allen & Unwin Ltd, 2003, p. 51. See Ang, ‘In the beginning was the Bayon’, in Clark, J., Ang, C., Woodward, H., and Vickery, M. (eds.) Bayon: New Perspectives. Bangkok: River Books, 2007; Dolias, J. Le crocodile ou la nagî: l’océan dans l’imaginaire cambodgien. Paris: Les Indes Savantes, 2012.

11 MUSICAL NOMENCLATURE FOR THE SANSKRIT TERM KANGSA IN SOUTHEAST ASIA Arsenio Nicolas

The Sanskrit term kangsa Following a study of two Sanskrit musical terms kangsa (‘bell-metal’) and keccapi (‘tortoise’) as used in Southeast Asia in an earlier article,1 this chapter proceeds with a detailed discussion of terms for gongs and other bronze musical instruments in Khmer, Old Javanese, Cham and their contemporary meanings in Cambodia, Vietnam, Thailand, Indonesia and the Philippines. The term kangsa appears in dated inscriptions in Old Khmer in seventh century; Old Cham in the ninth, tenth and twelfth centuries; and as gangsa in Old Javanese from the ninth century, thereby establishing a firm reference for a historical sequence.

India: kamsa or kangsa2 ˙ The Sanskrit term kaṃsa or kangsa is a type of white-copper or bell-metal or brass, a drinking vessel of brass or a goblet. It is also a kind of musical instrument, a type of gong or plate of bell-metal struck with a stick or rod.3 In India, an early site attesting to the presence of gongs is a bas-relief in Amaravati, dated from at least the second century BCE to the seventh century CE. Murthy describes the panel as follows. The panel portraying ‘Monks adoring as the gong is sounded by dwarfs’ reveals two dwarfs carrying a gong. Two dwarfs are seen carrying the bar with the hanging gong on their shoulders, probably of brass or bronze. The gong is played with a kana. One of the dwarfs has raised his left hand high in the act of beating the gong vigorously.4 The shape or profile of the gong is distinct – its size is quite large, and given its proportion in relation to the body size of the musicians, it appears to have a diameter of about the length of the musician’s torso; the rim or sides of the gong is turned-in (i.e. the rim is slanted inwards towards

The Sanskrit term Kangsa in Southeast Asia  209

the inner part); and lastly, the relief is not clear on whether it is a flat gong or a bossed gong. Twelfth-century reliefs from Hoysala temples in India show a thick, evenshaped flat gong and two double-membraned drums beaten with two hands, hanging from the neck and shoulders with a strap.5 Similar illustrations of gongs in temples are found in Cambodia and Java, as well as numerous similar practices where bossed gongs are carried in processions in Java6 and Bali7 in Sulu8 and in Thailand. The following table illustrates the uses of the term kaṃsa and other terms for flat gongs or flat discs in India. Historical sources attest to the use of the term as bronze or alloy. In contemporary practices, however, the term is used for musical instruments, generally for flat gongs of various types or profiles, and others like struck plates and rice bowls. Such variation in the use of the term points to a wide distribution over time and the adaptation of the term other than for bronze or alloy. These historical sources clearly define the term kangsa as a type of bronze. Arthaśāstra Rasaratna Samuccaya (twelfth century) Cilappattikaram (third to fifth century)

kaṃsata-tala kaṃsya

Bronzes of different proportions9 Alloy10

kancam

Bronze (Tamil)11

In contrast, kangsa and its derivatives are mainly used for musical instruments in contemporary terminology in India in different regions and languages, along with other terms in local languages. Ratijana, Orissa

Orissa, West Bengal Santhal, Munda and Ho Northern India Malayalam Tamil

kansar kansa

Flat gongs Rice bowls

thali

Thin plates12

kaṃsar kaṃsa kansathal or bata

Gongs Rice bowls13

chenkala or chennala semmankalam

Flat gongs or ‘struck plates’

Chhotanagpur (Figure 11.2)14

Kannada

jagte or jagante

Hindi

thali

Rajastahani

ghadiyal

Benares

tala cennalam16

Small flat gong for Brahmins15

Kerala

cennalam

Played in kathakali with singers, drums, cymbals17

210  Arsenio Nicolas

In Rajasthan, the thali is a flat gong that has a straight rim. The flat gong is laid on the ground, the proximal end rests on the left knee of a female musician and the distal end on the ground,18 a technique similar to that employed in the playing of flat gongs called gangsa by the Kalinga, Itneg and Ifugao in northern Luzon. In the discussion below, it will be shown that similar transformation of the term kangsa in several areas of Southeast Asia begins from the adaption of the Sanskrit term ‘bell-metal’ and eventually is used as names of bronze musical instruments.

The term kangsa in Khmer Çoedes dated two Khmer inscriptions to the seventh century that mentioned kangsatala.19 He translates the term as ‘gongs en bronze’ (bronze gongs) and found that the term kaṃsatala does not appear in the Sanskrit dictionaries he consulted, but is more common in Pali texts.20 Saveros Pou classified kangsatal (cymbal) as Middle Khmer,21 and proposes that it may have been derived either from the Sanskrit kaṃsyatala or from the Prakrit kaṃsatala.22 Finot’s translation of the Sdok Kak Thom inscription dated 974 S/1052 CE renders kangsa-tala as cymbales de cuivre, mentioning lutes, flutes, 50 orchestras, cymbals and drums.23 These are therefore varying interpretations of the term kaṃsa over several decades of scholarship. A recent study of Khmer inscriptions noted that musical instruments, ritual items and jewellery are found predominantly in the context of temple-related transactions.24 The following musical terms are mentioned in Khmer inscriptions from the seventh to the fourteenth centuries CE.25 List of terms referring to gongs, brass, copper, cymbals and bell, in Sanskrit and Old Khmer inscriptions in Cambodia and Thailand (Siam) seventh to fourteenth centuries CE. 1 The Sanskrit terms kaηsa, kaṃsatAla and kaṃsatala are mentioned in three Sanskrit inscriptions and two Old Khmer inscriptions dated from the seventh to the ninth centuries. The term kaηsa refers primarily to bell metal (Monier Williams), and if tAla may refer to cymbals, the terms kaηsatAla and kaηsatala may therefore refer to bronze cymbals, instead of gongs as proposed by Çoedes, Finot and Pou above. The term tAla does not appear in Old Javanese and Old Balinese inscriptions. The term kala, however, appears in Old Javanese as a type of trumpet, drums, cymbals or a kind of gong (cf. kali).26 2 The Sanskrit term tAla, defined as cymbals or gong, is mentioned in three Old Khmer inscriptions dated from the tenth, eleventh and fourteenth centuries CE. 3 The term koη, defined as gong (?), is mentioned in two Old Khmer inscriptions dated to the tenth century CE. 4 The term katyAη, defined as small bell, is mentioned in one Old Khmer inscription dated to the tenth century CE.

The Sanskrit term Kangsa in Southeast Asia  211

Gongs in the bas reliefs of the North gallery – East wing of the third enclosure of Angkor Wat Of the eight galleries of the third enclosure of Angkor Wat, it is only on the North gallery – East wing that bossed gongs of two types are carved – big and small suspended bossed gongs and semicircle bossed gongs or ­circle gongs (Figures 11.1 and 11.2). The gallery portrays the narrative of the ­Victory of Krishna over Asura Bana.

FIGURE 11.1 A

row of musicians – three oboes, a two-stringed lute, cymbals, a big-barrel drum with stand, two bossed gongs carried in a pole by four persons, two oboes, an eight-gong circle, a conch-shell, a small cylindrical drum hanging from the neck, and an oboe.

Source: Arsenio Nicolas.

FIGURE 11.2 A

row of musicians – two long horns, a pair of cymbals, a small drum hanging from the neck, a conch-shell, a short horn, a big barrel drum – accompanying the procession of Suryavarman II.

Source: Arsenio Nicolas.

212  Arsenio Nicolas

This is of particular significance for the following reasons: First, the North Gallery – East Wing and the East Gallery – North Wing was carved in the sixteenth century during the reign of Ang Chan (1528– 1566). This is evidenced by the inscriptions on these two galleries that the bas reliefs were made starting Wednesday 8 September 1546 and was completed on Sunday 27 February 1564.27 Second, the recent discovery of ‘hidden paintings’ on the Bakan, the first enclosure of Angkor Wat, on the highest or third tier of the temple, confirms the carving of bossed gongs, semicircle gongs (known today as khong vong) and a type of oboe known today as sralai on the North Gallery – East Wing of the third enclosure. However, one musical instrument depicted in the painting is a xylophone, known today as a roneat, which is not depicted on the bas-reliefs.28 Third, the creation of the bas-reliefs of the eight galleries of the third enclosure of Angkor Wat did not happen in one single reign of the Angkor monarchs. Suryavarman II (r. ca. 770–ca. 830) founded Angkor and it was constructed by his successor, Jayavarman VII (r. 1182/1183–ca. 1220).29 The bas-reliefs along the northeast gallery enclosures were left unfinished, and it may be that the construction was abandoned after the death of Suryavarman II around 1150. This idea of abandonment is supported by the discovery of drawings at the northwest corner of the second enclosure.30 Fourth, the diverse artistic styles of the bas-relief carvings on the eight galleries also indicate that each of these was created by eight or more groups of carvers that were working on each gallery over an extended period of time, each one with another group of designers and artists. Only five of the galleries are decorated with musical instruments. The West Gallery – South Wing, with the story of the Battle of Kurukshetra, from the Mahabharata epic are adorned with drums, oboes, horns, cymbals; the South Gallery – West Wing with the story of the historic procession of Suryavarman II, the builder of the temple, whose processions are accompanied with drums, oboes, horns, trumpets, cymbals, conch shells, priest’s bells; and the East Gallery – North Wing portraying the Battle of Lanka from the Ramayana epic where Rama and his allies, including the monkey troops, defeat Ravana and rescues Sita is adorned with drums, horns and oboes. Bossed gongs are not etched on these galleries, where such grand narratives are usually associated with gong music. The South Gallery – East Wing (with the narrative Judgement of Yama, the God of Judgement and of the Underworld, and Heaven and Hells) has only conch shells illustrated. The remaining two – the East Gallery – South Wing (with the narrative Churning of the Ocean of Milk) and the East Gallery – North Wing (with the narrative the Victory of Vishnu over the Asuras) – do not have any musical illustrations. It is thus significant that the North Gallery – East Wing showing the Victory of Krishna over the asura Bana (carved much later between 1546 and 1564) is decorated with bossed gongs, half-circle gongs, drums, oboes, horns, buffalo horn, cymbals and conch shells, representing a wide array of musical instruments as compared to all the other galleries (Figures 11.3–11.5).

The Sanskrit term Kangsa in Southeast Asia  213

FIGURE 11.3 A

big-bossed gong carried on a pole, with the person at the back beating it.

Source: Arsenio Nicolas.

Fifth, the absence of bossed gongs in the five galleries mentioned above, which date to the twelfth century, seems to belie the presence of excavated bossed gongs in four shipwrecks on the Gulf of Thailand from the thirteenth to the sixteenth century. During this period, bossed gongs were already circulating in the Gulf of Thailand as evidenced by four shipwrecks with bossed gongs as part of their cargo. All these ships, with three originating from Thailand and one from Southern China, also carried Thai tradeware as part of the cargo. Some were also carrying Vietnamese and Chinese trade-ware, and without any Khmer ceramics.31 The first two shipwrecks, Java Sea and Rang Kwien, date from the Sukhothai period and the latter two, Phu Quoc and Sattahip, from the Ayutthaya Period in Thailand. The following lists the four shipwrecks with bossed gongs in their cargo, dated from the mid-thirteenth century to the sixteenth century. 1. Twelfth-thirteenth century Java Sea (Thai or Indonesian ship) with two bossed gongs; the ship originated from southern China carrying tradeware: southern China and Thai fine-paste ware. 2. Thirteenth-fourteenth century Rang Kwien, Thailand with one bossed gong and one bell originating from Thailand and heading for Indonesia carrying Thai and Vietnamese trade-ware with Chinese coins from the Hongwu period (1368–1402).

214  Arsenio Nicolas

FIGURE 11.4 A

big-bossed gong carried on a pole and a small barrel-shaped drum in front of it.

Source: Arsenio Nicolas.

FIGURE 11.5 Two bossed gongs, one small, one bigger. The player is at the back, his

left hand holding a beater with cloth wrapped at the tip to give off a loud, booming, nonmetallic sound when beaten, as practiced today in Cambodia, Thailand, Java and Bali, Borneo and the Philippines. Source: Arsenio Nicolas.

The Sanskrit term Kangsa in Southeast Asia  215

3. Fourteenth-fifteenth century Phu Quoc, Vietnam with 51 bossed gongs, originating from Thailand and heading towards Borneo and the Philippines with Sisachanalai, Sawankhalok and Vietnamese trade-ware. 4. Sixteenth-century Sattahip, Thailand with bossed gongs originating from Thailand carrying Sukhothai, Sawankhalok and Chinese trade-ware. Of particular relevance are the fourteenth-fifteenth century small, bossed gong fragments found in a ceremonial site in Sungai Lumut, Brunei with Ming, Sawankhalok, Sukhothai and Ming tradeware.32 The semicircle gongs on the bas-reliefs of the North Gallery – East Wing in Angkor Wat have the same dimensions. These gongs are small and are similar to the small bossed gongs of the present-day gulintangan, in Borneo, the kulintang in Mindanao and Sulu, the bonang in Java, the trompong in Bali (­Figures 11.6–11.9). The bas-relief carvings that adorn the walls of the outer temple enclosure add up to some 521 meters, with narratives taken from the Khmer dynastic history of Suryavarman VII (one gallery); the Hindu epics – the Mahabharata (two galleries) and the Ramayana (two galleries); as well as from the Bhagavata Purana, the great Hindu creation myth.33

FIGURE 11.6 A

gong circle with nine bossed gongs, the musician using two hand beaters.

Source: Arsenio Nicolas.

216  Arsenio Nicolas

FIGURE 11.7 A

gong circle with nine graduated bossed gongs, the musician uses two beaters with cloth wrapped around the tip, another person carries the instrument.

Source: Arsenio Nicolas.

FIGURE 11.8 A

gong circle with eight graduated bossed gongs, another person holding the frame

Source: Arsenio Nicolas.

The Sanskrit term Kangsa in Southeast Asia  217

FIGURE 11.9 A gong circle with eight bossed gongs, with two beaters without the

tip cloth wrapper. A conch-shell and a small cylindrical drum also are shown. Source: Arsenio Nicolas.

Similar to the twelfth-century bas-reliefs on Angkor Wat, the twelfthcentury Angkor Thom and late twelfth-century Bayon do not have bossed gongs etched on the walls of the temple. The musical instruments carved on their walls are mainly drums, cymbals, arched harps and bowed lutes.34 A recent study documents that early Ayutthayan artists resided in Angkor in the first half of the fifteenth century.35 During the reign of Ang Chan (1528–1566), Angkor Wat underwent a transformation from a monument to Vishnu into a Theravada Buddhist shrine. Ang Chan ordered the completion of the Vaishnavite bas-reliefs, and converted the Bakan, the first enclosure of Angkor Wat, into a sanctuary of Buddha images.36 Boisselier also noted that the rendition of the half-circle gongs on the bas-relief depicting the Victory of Krishna over the Asura Bana may have links to ­Siamese styles of art, locating the paintings to the post-Angkor period.37 As the Kingdom of Ayutthaya was founded in the mid-fourteenth century, it gradually supplanted Angkor as the dominant regional power, eventually becoming the most powerful polity in mainland Southeast Asia during the sixteenth century.38 Ringis (1990: 38) noted that the development of Siamese mural paintings may themselves have been influenced by the bas-reliefs of Angkor Wat (Table 11.1).39

218  Arsenio Nicolas TABLE 11.1 Distribution of bas-reliefs of musical instruments in the eight galleries

of Angkor Wat based on field research in 2009 and 201140

Gallery

Narrative story

Musical instruments

West Gallery, South Wing, 49 m long

Battle of Kurukshetra, from the Mahabharata epic Historic procession of Suryavarman II, the builder of the temple

Drums, oboes, horns or trumpets,41 cymbals

South Gallery, West Wing, 94 m long

South Gallery, East Wing, 66 m long

Drums, oboes, horns or trumpets, cymbals, conch shells, priest’s bells Conch-shells

Judgement of Yama, the God of Judgement and of the Underworld and Heaven and Hells None East Gallery, South Wing, Churning of the Sea 49 m long of Milk, from the Bhagavata Purana, the great Hindu creation myth None East Gallery, North Wing, Victory of Vishnu over 52 m long the Asuras (carved much later, between 1546 and 1564) Bossed gongs, circle North Gallery, East Wing, Victory of Krishna over gongs, drums, oboes, 66 m long the asura Bana horns or trumpets, (carved much later, buffalo horn, cymbals, between 1546 and 1564) conch-shells Drums, oboes, cymbals, North Gallery, West Wing, Battle of Dewas and conch-shells 94 m long Asuras, detailing the 21 important Gods of the Hindu Pantheon, with their vahana or mounts Drums, horn, oboes, North West Corner Rama returning to cymbals Pavilion West bay, north Ayodhya in triumph, side following the victory in the Battle of Lanka; to the right of the window, monkeys dance and play musical instruments in celebration Drums, horns, oboes East Gallery, North Wing, Battle of Lanka, Rama 51 m long and his allies, including the monkey troops, defeats Ravana and rescues Sita

The Sanskrit term Kangsa in Southeast Asia  219

In Khmer inscriptions, the following musical terms are mentioned. There are 13 Sanskrit derived musical terms, 6 Old Khmer.42 Sanskrit term

Description

Dates of inscriptions

kangsa, kangsatAla, kaMsatala tAla hUdUka gandharvva vINa vadya tUryya, turyya zikharA kinnara, kinara dundubhi vAditra

Brass; copper, alloy

628–968

Cymbals Tambourine, drum Musician, singer Lute Musical instrument Musical instrument Stringed instrument Stringed instrument Tambour, kettledrum Musical instrument

972–1308 960–1101 seventh c to eleventh c 879–972 910–928 seventh c to thirteenth c 960–1101 972, 994, 1024 1190, 1200 1002–1049

Old Khmer

Description

Dates of inscription

cheng katyAng chko, cko kluy, kloy sgar

Cymbals, small Bell, small Tambour Flute Drum

972 972 972, 994 994 1308

The list does not include any term that refers to gongs in the inscriptions dated from the seventh to the eleventh century, and as discussed above, the gongs on the North–East gallery of the third enclosure of Angkor Wat date to the sixteenth century. In Cambodia today, these musical instruments are known as follows: kong mong (bossed gong); kong vung (semicircular gong chime); peat (half-moon gong or flat disc); ching (small hand cymbals); kandoeng (bell); sampho (small, double-headed barrel drums); skor yol (suspended barrel drum); skor thom (large double-headed barrel drum); skor (cylindrical drum); memm (bowed monochord); ksaey mouy (musical bow or plucked monochord); tror (fiddle); chapey dong veng (long-necked lute); pinn (angular harp or arched harp); sampho sralai (quadruple-reed oboe) and saing (conch); and roneat (xylophone).43 Shorto provides the terms skɔ:, sgər, sgəər for drum along with cognates from other Mon-Khmer languages.44 The circle gongs with eight or nine gongs are known today in Cambodia as kong wong chamnet key or kong vong similar to the bas-relief illustrations in Angkor Wat. This ensemble is played for cremation ceremonies called wong dontree traming, together with a quadruple-reed oboe, srolay; barrel drum, skor thom muy; and suspended bossed gong, kong thom.45 In

220  Arsenio Nicolas

other places, the ensemble is called kong skor, and the instruments are called kong vong, sralai, kong and skor thom, respectively.46 In northeast Thailand, the circle gongs are simply called khong wong with eight gongs, which are also played during cremation ceremonies particularly in Surin, Buriram and Sisaket where the Khmer people settled a long time ago. Over a period of several centuries, from the seventh century to the present, the nomenclature of Khmer musical instruments has changed. Significantly, only the Old Mon-Khmer sgar exists today as skor, referring to drums of different profiles and types. Likewise, the Old Khmer kluy or kloy for flute is not known today in Cambodia but is known in Thailand as khlui.

The term kandgsa in Cham A ninth-century Cham inscription dated 18 May 875 mentions the term kangsam and is translated by Finot as ‘laiton’47 and by Golzio as ‘bell-metal’.48 In another inscription dated 918 CE, the term kamsa is mentioned and is translated by Huber as ‘bronze’49 and by Golzio as ‘bell metal’.50 In both these two inscriptions, kangsam appears together with other metals like gold (suvarṇa), silver (rajata), iron (loha), copper (tāmra), and kamsa with gold (suvarṇa), iron (loha) and copper (tāmra) as prized possessions of the ruling monarch. Another inscription dated 1156 CE mentions kangsa bhaja as copper pitchers.51 Po Dharma listed several terms in Cham manuscripts dated around the sixteenth and seventeenth centuries for musical instruments, which included gangsa and kangsa (Skt., also gamak: laiton [brass], as well as tembaga [also tamaga, tamagra, tamaka, kamada = cuivre; bronze]).52 Old Khmer Burma

kansatala gaza ganza

Khmer

kangsatala

Thai

gangsadan or kangsadan

Cymbals, Prakrit-loan word53 Sixteenth century54 Copper money55 Big flat and thick disc in a Buddhist temple56

In ancient Cham, the terms gangsa and kangsa are known (Malay, gamak; French, laiton; English, brass).57 In Vietnam today, two terms are generally used for flat gongs – cing or ceng and sar. Among the Mnong Gar, flat gongs are called cing.58 In Aymonier and Cabaton’s Cham-Francais dictionnaire, the term ceng refers to a small-bossed gong, while the term sar is used for a flat gong. While the current terms for flat gongs in highland Vietnam are not derived from the Sanskrit form kaṃsa, the similarities between the music of the flat-gong ensembles in highland Northern Luzon and highland Central Vietnam today has led Maceda to propose that flat gongs with narrow, straight rims, known as gangsa in northern Luzon, Philippines, may have come from Vietnam,59 although maritime records of flat gongs in shipwrecks

The Sanskrit term Kangsa in Southeast Asia  221

dating from the tenth to the twelfth centuries in Borneo, Java and Sumatra indicate that these flat gongs were transported by ships from south China.60 While there are many references to gongs in inscriptions and manuscripts, bas-reliefs in Cham temples dated from ninth to the fourteenth centuries do not include gongs. The illustrations are mainly drums, stringed-lutes, breast lutes with resonator, arched harps, horizontal flutes, cymbals, horns or trumpets.61

Thailand In the Isan region in northeast Thailand, flat gongs are called panghaq or panghat, which are derived from the Sanskrit term kangsa (Thai: bronze = samrit). The distribution of the term panghat covers the northeast region of Thailand, and in some regions on the borders with Cambodia and Laos. Panghat is played to accompany ceremonial dances in northeast Thailand called fon khong thum, ( fon=dance; khong thum = medium-sized cylindrical drum, hanging from a pole borne on the shoulders of two musicians with the one at the back beating the drum). It is possible that the flat gong, panghat, is a later addition, with the drum as the sole accompanying instrument in the past. A rare ensemble called wong khong seng also makes use of panghat, played together with large heavy wooden drums, hanging from a wooden bar borne in the shoulders of two men. In the town of Mukdahan, in northeast Thailand, by the border of Cambodia, this ensemble consists of a khong seng (long drum); thap yai (cymbals); khong thum (big drum); and panghat (flat gong). This ensemble is unique to this place.62 A unique type of big, flat circular disc or gong called kangsadan is housed in an open pavilion in the yard of the Wat Phrathat Haripunchai in Lamphun, central Thailand. This is a unique huge and heavy flat and circular disc, with a diameter of 2.28 m and a thickness of 3 cm.63 Unlike flat gongs of the panghat and gangsa type, the rim of the kangsadan is thick and is turned in towards the front, and the back side is simply flat. A long wooden log hanging from the ceiling is used to strike the disc, similar to the technique of striking big bronze bells in the temples in Japan, China and Korea. It is suspended from the beams of the pavilion by means of two thick metal clamps that are fastened to the upper rim of the gong. The gong has an inscription on the upper side where the clamps are fastened. In this inscription, the term kangsahtan64 is used (Pali: kamsatala).65

The term gangsa in Java and Bali Zoetmulder translates the Old Javanese term gangsa as derived from the Sanskrit form kaŋsa as bell-metal.66 In Java, the term gangsa first appears in Old Javanese inscriptions from the ninth to the tenth centuries, during the Central Javanese Period. The term cannot be found in Old Balinese inscriptions, and it does not appear anymore during the East Javanese period from the tenth to the fourteenth centuries.67

222  Arsenio Nicolas

In Old Javanese, gangsa particularly refers to bell-metal, preserving its derived meaning from Sanskrit. Casparis classifies the term as Middle Indian known in Java from the ninth century onwards. It corresponds to the Sanskrit kaŋsa (‘bell-metal, ‘braze’) found in other Indonesian languages. The changes of the initial k into g is, unlike that of the intervocalic k, are common in Prakrit. Similar forms with g- are known in Pegu, Burma in the forms ganza and gamça (s.v. ganza).68 In Old Javanese inscriptions, gangsa is mentioned in two contexts. First, it refers to bronze smiths, pandai gangsa, in at least six Old Javanese inscriptions (late ninth to early tenth centuries). Second, gangsa is either mentioned alone, or more commonly, in combination with three terms used for metals in Old Javanese: tambaga or tamwaga (copper); wsi or wesi (iron); and mas (gold). It is mentioned as a metal alloy in six inscriptions dated 862 and 91569 and in 862, 880, 904, 907, 909 and 915, occurring with the term pandai ‘smith’. Old Javanese inscriptions mention gangsa, dated late ninth and early tenth centuries, while the last two are from east Java, dated early tenth century.70 From the tenth- and eleventh-century inscriptions from the Brantas River area, two of the three inscriptions use gangsa as a single term to refer similarly to bronze, in association with three other metals most commonly mentioned – iron (wsi); copper (tambaga); and tin (timah). Two of these inscriptions are dated at 929 and 1021. The third is without a date, perhaps copied during the Majapahit Period from a tenth- or eleventh-century inscription.71 The term does not appear in later inscriptions but is mentioned in literary texts in Old Javanese72 and in Classical Malay.73 In this set of inscriptions, tambaga (‘copper, bronze’), appears together with gangsa (‘bell-metal’). In a study of Middle Indian terms in Old Javanese, Casparis provided a list of words that are not directly traceable to a Sanskrit, Hindi or Tamil prototype.74 The majority of the objects in the list were used by traders and artisans. Four of these are directly related to metalcraft: gangsa (Skt. kamsa, bell-metal, brass); gusali (blacksmith); pandai (smith); and tamwaga or tambaga (Skt. tamra, tamraka, copper).75 Panday is an old Austronesian word for a skilled worker, smith or gold smith, among others. The other terms refer to crafts, trade commodities and trader (banyaga).76 All these terms, numbering 34, are dated from 798 CE to 934 CE, and had all been incorporated into Old Javanese well before 1000 CE.77 Sedyawati notes that blacksmiths and coppersmiths are frequently mentioned in inscriptions as compared to goldsmiths. The former are ‘always mentioned in relation to the restriction of the number of producing smithies, which are free from taxation, within one village’.78 The following table summarizes the uses of the term gangsa im Old Javanese, Old Javanese literature

gangsa79

Ramayana kakawin Old Javanese

gangsa; Skt. kamsa (‘klokkemetaal’, bell-metal)80 gangsa gubar81

The Sanskrit term Kangsa in Southeast Asia

223

As in the Khmer inscriptions, gangsa or kangsa in Old Javanese do not refer to any musical instrument but to items related to metals. It is only later that the term gong first appears in Old Javanese literature in the classic panegyric poem, Negarakrtagama, dated fourteenth century.82

Bali There are no Old Balinese inscriptions that mention the term gangsa. However, contemporary usages of gangsa are found in a variety of ways in several Balinese ensembles. Name of ensemble gamelan luang gamelan gong gong ageng gamelan gambang gamelan pelegongan

gangsa (metallophones with seven keys) gangsa gede (big), gangsa chenik (small) saron or gangsa jongkok 83 gangsa jongkok (Batur, Kintamani) saron or gangsa, played with bamboo xylophones called gambang gangsa jongkok 84

McPhee defines gangsa as ‘bronze, a name interchangeable with saron indicating any one-octave metallophone with keys resting over a sound-box and struck with a single mallet’ with the following types: gangsa gantung gangsa jongkok gangsa siya

Metallophone, keys thinner and wider, suspended over Bamboo resonating tubes, one key on top of one tube. Thicker keys placed above a rectangular trough resonator gangsa gantung (nine keys, in gamelan gong kebyar)85

In contemporary Java and Bali, gangsa has different meanings. First, it refers to bronze musical instruments generally classified as metallophones (Bali: gangsa jongkok, Java: gambang gangsa). Second, it also means a musical ensemble made up of gongs, metallophones and drums (Java, kromo: gangsa, ngoko: gamelan).86 After about the sixteenth century, the term gangsa appears in literary texts and manuscripts. It is also mentioned quite frequently in classical Malay texts dating to at least the seventeenth century, which have been derived from Javanese.87 As compared to Bali and northern Luzon, the use of the term gangsa in Java is not widespread, as it is the high Javanese (kromo) term for gamelan. A unique type of metallophone is used in the palace of Surakarta, called a gambang gangsa,88 which is played only for the musical piece, the gending Colopito, when the Susuhunan (Javanese monarch) makes a ceremonial exit in the grand hall (pendopo ageng) of the palace during the Jumenengan, the anniversary of the installation to the

224  Arsenio Nicolas

throne of the Susuhunan, during which the sacred dance, Bedoyo Ketawang, is performed.89 In the Yogyakarta palace, the gambang gangsa is considered an archaic instrument together with the bendhe (a type of small gong with central boss) and the slentho (a knobbed metallophone).90

The terms gangsa and palayi in Northern Luzon In the Philippines, the term gangsa refers to flat gongs and flat-gong ensembles in highland northern Luzon. Flat gongs have not been found in any shipwreck along the western coasts of the Philippines, but in the three early shipwrecks, dated ninth to tenth century, in the coasts of north Borneo, Java Sea and Sumatra that transported flat gongs. Following this period was a chain of shipwrecks that transported bossed gongs from the twelfth to the seventeenth centuries.91 In the early twentieth century, Jenks observed that the Bontoc gong, or gang'sa, originates in China, though perhaps it is not now imported directly from there. It certainly does not enter the Island of Luzon at Manila, or (and) in Ilokos Sur, and, it is said, not at Vigan, also in Ilokos Sur.92 Later ethnographic studies by Conklin document Ifugao traders obtaining bronze gongs that go back to pre-World War II from Chinese merchants at the mouths of the Cagayan, Abra and Amburayan rivers on the northernmost and northwestern coasts of Luzon.93 Two terms for flat gongs were in use in Luzon in the seventeenth century. The term gangsa appears in the Cagayan province in a document dated 1640.94 Two Negrito groups in Luzon, among the Ayta Magbukun in Bataan and Ayta Magkunana in Bulacan, today use the term palay and paley respectively for flat gongs,95 and the term palayi also appears in an early seventeenth-century Tagalog dictionary in the province of Laguna. A mid-twentieth century dictionary of Ilongot language from the southeastern tip of the Northern Cordillera lists pala’e as cymbal,96 although the contemporary terms for flat gongs used by the Ilongot are paryuk and changsa, the latter being similar to that used by the Bontoc.97 In the twentieth century, the terms palay, paley and pala’e were then known in Luzon at least from the southern Cordillera in the north, to the eastern regions of Bataan, to the central province of Bulacan. The table below shows early references to flat gongs in northern Luzon, particularly those mentioning the terms ‘gangsa’ and ‘palayi’.

1609

Tagalog

1613 1640

Tagalog Cagayan

‘Metal bells’ shaped like ‘large pans brought from China’ 98 playi99 gaza100

The Sanskrit term Kangsa in Southeast Asia  225

1785 1865

Tagudin, Ilocos Pangasinan

1904 1918

Pangasinan Itawes (Cagayan)

‘Little cymbal of metal’ from Borneo101 gansa ‘cobre y significa tambien laton' (copper and may mean also as brass)102 pala-y ‘campana de china’103 ganza104

While the term gangsa is widely used as a common term for flat gongs and flat-gong ensembles in northern Luzon, what is more significant are the individual names assigned to each gong in one ensemble for the Kalinga and the Ifugao. These terms are local and not derived from any Sanskrit term.105 The following are the current terms for flat gongs and the individual gongs in ensembles in northern Luzon.106 Kalinga Ifugao Isneg Bontoc Ilongot Sagada and Bontoc Tingguian/Itneg

gangsa palook and gangsa topayya (six gongs: balbal, kadua, katlo, kapat, opop and anungos) gangha (three gongs: tobob, hibat, ahot) hansa (flat-gong ensemble); gangsa (flat gong); and ludag (drum). changsa, gangsa changsa kangsa gansa

The term bheri An Old Balinese inscription, Prasasti Blanjong, dated 835 Caka (913 CE), now kept in the village temple of Blanjong near Renon, south Bali, provides the only paleographical reference to bheri, written in a language that is predominantly Sanskrit.107 This stone inscription is written in two languages. One side is in nagari, which bears the date, and the other is in kawi, where the term bheri is mentioned. The Old Balinese term bheri is derived from the Sanskrit bheri, which means kettledrum.108 Bheri is not mentioned in Old Javanese inscriptions, but is referred to in at least six Old Javanese literary works dated from the end of the tenth century to the fifteenth century.109 The term is only found in Java and Bali and in Sanskrit. It occurs in a ninth-century Old Balinese inscription and very frequently in many literary Old Javanese texts from the tenth to the fourteenth centuries. It appears frequently with a modifier like mrdangga and murawa, which are also Sanskrit terms for drums; hence bheri may have meant a generic term for ‘drum’ during this period. According to Zoetmulder, however, who quotes Kunst, this term had already meant a small gong by the eleventh century, as it is used in modern Javanese.110 Since there are no extant inscriptions in Java that refer to bheri, it cannot be said, however, that the term was not known as early as 913, around the time when the Prasasti Blanjong was written in Bali.

226  Arsenio Nicolas

Gong bheri In Bali today, flat gongs are part of an ensemble called gong bheri, which accompanies the dance baris cina in the village temples of Renon and Semawang, both found along the eastern shores of the southern tip of the island. The instruments in this ensemble consist of two suspended, flat gongs (bar, ber); a suspended, medium-sized concave surfaced flat gong (bebende or teng); three medium-sized, suspended bossed gongs (tawa-tawa ageng or pung, tawa-tawa alit or pir, and kempli or pu); a pair of small-bossed gongs (kajar or klu, and klenang or nang); a large barrel-shaped drum (bedug or bebedug); a pair of cymbals (cengceng kopyak); and a conch-shell (sungu).111 In the performance that I witnessed in 1985 in Renon, the ensemble accompanied a trance dance.112 The ensemble gong bheri is a unique assemblage of different types of gongs. The two flat gongs, bar and ber, are similar to the flat gongs of northern Luzon, where they are commonly called gangsa. The bebende or teng is rarely found today in Bali. Kunst referred briefly to the present-day Javanese gong beri, which is ‘a small, often knob less gong’113 but did not give a full description. Kunst traced the term beri to the Sanskrit forms bheri, bhairi and bahiri, and noted that the term stood for drums during the first Hindu–Javanese centuries, which later on was used to refer to a small species of gongs. In Indonesia, this is one type of ensemble that makes use of flat gongs and bossed gongs in an ensemble. This particular mix of two types of gongs is also common to the highland peoples of central Vietnam,114 Cambodia,115 Sumatra and Borneo.116

Flat gongs in Southeast Asia In many scattered villages and towns in Southeast Asia, flat gongs are known by various names not derived from Sanskrit. Pakpak (Dairi) (Sumatra)

gerantung Four flat gongs played melodically by a single player Accompanied by a set of bossed gongs.117

Bidayuh

puum

(Borneo, Sarawak) Sasak (Lombok) oncer Kayan Uma’ Jalan mebang (Borneo, Sarawak) Kayan and Modang mehbiang (Borneo, Sarawak)

Two large flat gongs played in a gong ensemble with three types of bossed gongs tetawak, bendai, sanang118 Small flat hanging gong119 Flat gong120 Flat gong, Klinjau and Telen Rivers along the Mahakam121

In Borneo, the flat gongs without a boss are used as trays to place ritual paraphernalia and are not played as music instruments, except among the

The Sanskrit term Kangsa in Southeast Asia  227

Kayan, who play these instruments in an ensemble of three gongs.122 A flat gong and a two-headed drum called tawung are laid on the floor as part of paraphernalia in a ritual to honour the dead.123 Among the Ao-heng or Penihing, a Dayak group in east Kalimantan, a flat gong is part of the ritual offerings for nyari, a healing ritual that involves several households.124 Among the Bontoc in Sadanga, northern Luzon, flat gongs were used as plates in distributing the cooked buffalo meat to the owners of the flat gongs who participated in the grand communal ritual, chono.125 Kroeber wrote about a ritual sacrifice of a dog in Bontoc, showing the emptying of the blood from its almost beheaded body onto two flat gongs.126 These flat gongs were laid on the stone floor in the ceremonial stone circle of the dap-ay, or men’s house, where important village rituals were held, especially those that were related to headhunting. In my study of music and headhunting, rituals involving the sacrifice of dogs did not use flat gongs but were generally accompanied by bamboo or wooden musical ensembles.127

Flat gongs and bossed gongs Two types of gongs are found in Asia – flat gongs and bossed gongs. A great variety of names are used for gongs providing a general distribution outline and relations from these patterns of distribution.128 Four archaeological finds of flat gongs – in Tanjung Simpang, Borneo (tenth century); in Jambi, Sumatra dated 1231; on Lingga Island (early twelfth century); and on Java Sea (thirteenth century) – attest to the early spread of flat gongs in the Philippines and Indonesia, extending to the Thai–Malay peninsula. Flat gongs, however, are not found in bas-reliefs in Cambodia, Java and Champa, but are illustrated in several eleventh-thirteenth century Hoysala temples in Karnataka, India.129 Three shipwrecks dated thirteenth century with bossed gongs were found in Butuan (northern Mindanao), Rang Kwien (Thailand) and in Java Sea, all dated in association with trade ceramics. Bossed gongs are represented in temple reliefs in Cambodia and Java (dated to the sixteenth century), about the same period when bossed gongs and flat gongs begin to appear in 21 shipwrecks and inland archaeological sites from Thailand to Vietnam, Sumatra, Java, Borneo, Palawan and Mindanao from the tenth to the seventeenth centuries. It is during this period that various types of gongs with boss appear in shipwrecks on the Gulf of Siam, Java Sea, Mindanao and Palawan, in the bas-reliefs of Angkor Wat and several temples in east Java (Panataran, Kediri).130 In Brunei, a fourteenth-fifteenth century archaeological site yielded bossed gong fragments, similar to present-day gulintangan, which may have been used as ritual offerings.131

Music shifts, music loss The term kangsa provides an important link to the movement of Sanskrit musical terms into Southeast Asia. Its use for various types of bronze

228  Arsenio Nicolas

musical instruments in Southeast Asia brings into light the patterns of spread and distribution of music and language. The uneven distribution of flat gongs in Asia today, as compared to bossed gongs, invites speculation into the movement of this musical instrument into a wide area in this region. Three routes of movement of flat gongs can then be summarized as follows. One route originates from India, with flat gongs likely carried by Indian traders that sailed to China in the first millennium CE or earlier.132 The evidence so far for this route comes solely from similar flat gongs in India known today, primarily in the Hoysala temples dated eleventh to thirteenth centuries, and from the spread of the term kangsa (Skt. kamsa) as known in seventh-century Khmer inscriptions and ninth-century Cham inscriptions and as gangsa and kangsa in ninth- and tenth-century Old Javanese inscriptions. Another route starts from China, evidenced by several flat gong artefacts from the tenth-century shipwreck in Tanjung Simpang, north Borneo (inscribed with a Chinese name Guo on the flat surface of all the gongs); in the twelfth-century shipwreck in Pulau Bakau, Riau; in the late twelfth-century Lingga shipwreck; and a flat gong found in Jambi, Sumatra with a Chinese inscription bearing with the date 1231 mentioning the term for the gong as tongluo.133 While a number of Chinese written documents indicate very old provenance of flat gongs in China, it cannot be said that China is the origin of the flat gong – first, flat gongs come in different sizes and profiles; and second, China has one of the oldest writing traditions in the world, but other music cultures that are older do not have writing systems. A third route, which requires much further study, pertains to the diverse networks and linkages among the peoples of Southeast Asia that have engaged in maritime and inland trade for several thousand years. The terms for flat gongs in northern Luzon, Khmer, Cham and Isan (Thailand) are derived from the Sanskrit term kaṃsa. The current terms for flat gongs in highland Vietnam are not derived from the Sanskrit term kaṃsa, but point to the similarities between the music of the flat-gong ensembles in highland Northern Luzon and highland Central Vietnam today.134 The distribution of this term and its derivations invite further exploration into the spread of the term from Sanskrit and Indian sources to Thailand, Sumatra, Java, Bali, Borneo and the Philippines. In the case of flat gongs in the northern highland Philippines, the problem at hand is to explain how flat gongs, reportedly coming from Borneo, China or mainland Southeast Asia and perhaps India, had come to be been called gangsa, which is derived from a Sanskrit term. The use of Sanskrit terms for vocal forms, musical instruments and musical ensembles as found in Old Javanese and Old Balinese inscriptions, literature and in temple reliefs, as can also be found in Myanmar, Thailand, Laos, Kampuchea, Champa and Vietnam, are early records of music in the area. The subsequent use of kangsa for several types of musical ensembles and musical instruments illustrates the manner by which bronze musical instruments – flat gongs, bossed gongs, cymbals, flat discs

The Sanskrit term Kangsa in Southeast Asia  229

and metallophones – became pervasive in Southeast Asia, following the earlier spread of bronze drums from about the mid-first millennium BCE.

Gongs and bamboo musical instruments The appearance of bronze gongs, both flat gongs and bossed gongs, changed the musical geography and history in Southeast Asia. In previous studies, I noted that the music of bamboo musical ensembles in northern Luzon,135 central Java and Bali136 imitate the music of gong ensembles in northern Luzon, Java and Bali, respectively, and the older and more ancient music of bamboo musical ensembles have been integrated into the music of flat-gong and bossed-gong ensembles after the appearance of gongs in Southeast Asia. The idea of gongan and non-gongan structures in Javanese and Balinese gong ensembles, where a four-beat structure is punctuated at the fourth with a gong beat,137 can be further applied to the numerous types of bamboo, wooden and bronze musical ensembles in the region.138 These shifts in musical iconography, musical styles and structures further illuminate certain ideas of musical exchanges in Asia in ancient times, relating older village ensembles in India, Assam and south China to the village musical ensembles in Southeast Asia before the rise of court music cultures in the region in the first millennium CE. Music shift (i.e. a shift from one musical system to another upon the introduction of a new system) occurred in the context of the spread of Hinduism and Buddhism in Southeast Asia, and was centred in the temples and in the royal courts where rituals based on the new religions were instituted.139 Beginning in the late sixteenth-century Philippines, Spanish colonialism eradicated the many diverse musical traditions as the majority populace was converted to Christianity,140 a phenomenon still occurring in Borneo,141 Thailand142 and in many other parts of the world today.143 In West Malaysia, the emergence of local state (negeri) Islamic legislation has continually prohibited and then repermitted the performances of ancient pre-Islamic rituals, music and theatre.144 Music shift and music loss, like language loss, occur amidst increasing global mobilities and social movements, wherein slowly, indigenous musical traditions vanish and are replaced with music introduced from the outside by various means. The disappearance of the diversity of music cultures all over the world since the sixteenth century has not been thoroughly studied and analysed. The powerful forces of religion, economics, nationalism, globalization and the dominant system of music industries today have slowly replaced indigenous musical traditions with music that is generated from centres of religious, economic and political power.

The music of Southeast Asia and India – historical connections and contemporary practices The musical practices of the recent past in Southeast Asia are known only by inference from archaeological artefacts, temple bas-reliefs, inscriptions

230  Arsenio Nicolas

and literature in cultures and communities that developed writing and dynastic histories. The majority of the more ancient music cultures without writing in Southeast Asia are known today by recent documentation in ethnomusicology and studies in historical linguistics. Musical instruments and dances etched on the bas-reliefs of temple walls in Champa, Angkor, Prambanan, Borobudur or Bagan may have corresponding images in India, but localization and indigenization in artistic rendition betrays the idea of Indianization, as has long been argued. In contrast to the manner that musical practices in Chinese temples in Southeast Asia did not entirely diffuse into the local communities, the dances and rituals in Hindu and Buddhist temples with a square layout like Borobudur, Prambanan, Angkor or Bagan may have been performed in a pradaksina circumambulation, as practiced today in Hindu–Balinese temples, in Prambanan and Borobudur in Java, as well as in Buddhist temples like in Prathat Na Dun in northeast Thailand. The Indian epics Ramayana and Mahabharata were translated and written down in palm-leaf and paper manuscripts into Khmer, Thai, Javanese, Balinese and Malay,145 and were subsequently rendered into shadow puppet plays, accompanied with the music of gong ensembles. Among the Maranao in the Philippines, the Maharadia Lawana is a sung epic by a solo singer based on the Ramayana narrative.146 The wide distribution of gong ensembles in Southeast Asia, which are primarily played in Hindu and Buddhist temples and palaces, are not found in India, although music, dance and theater are integrated into sacred and religious ceremonies in Asia.

From Southeast Asia to India The movement of Sanskrit musical terms into Southeast Asia during the early period in the history of Southeast Asia is evident from the inscriptions. Contemporary ethnographic data in the courts, temples and villages further reinforce this. Little is known, however, on the movement of musical ideas from Southeast Asia to India. The long history of maritime links between China and India is certainly connected to the sea passages in Southeast Asia, as attested by the journeys of the monks Faxian in the fourth century and Yijing in the seventh century,147 and Indian or Arab shipwrecks, one of which carried musical instruments as part of the cargo – the ninth-century Belitung shipwreck with bells and cymbals.148 The eleventh-century Burmese ruler Kyanzittha sent musical instruments to Bodhgaya during the reconstruction of the temple – bronze drums, as well as musicians, singers and dancers.149 It has been also proposed that the bin and vina may have been derived from bamboo idiochords in Assam and Java.150 Flat gongs were circulating in maritime Asia since the tenth century following the routes from China, Vietnam, Philippines, Borneo, Java, Sumatra and perhaps to the west towards India and back. Overland routes on the mainland transported these gongs as well. Sanskrit terms were circulating following the earlier spread of Austronesian, Austroasiatic, MonKhmer and Tai-Kadai musical terms. However, musical terms do not move

The Sanskrit term Kangsa in Southeast Asia  231

synchronously with the movement of musical instruments, as communities have the choice to name them accordingly. Very little is known about the history of the music of the unwritten and oral traditions, and of bamboo and wooden musical instruments. Early contacts between South Asia and Southeast Asia were precursors to the spread of Indic religious ideas to Southeast Asia.151 Some ten small bells and bronze vessels were found in a fourth-century BCE site in Ban Don Ta Phet, Kanchanaburi, Thailand, using metal working technologies from India.152 The peoples of Southeast Asia were innovative farmers, metallurgists, musicians and mariners.153 The Austronesians sailed from their home islands during the Neolithic period reaching as far as Madagascar to the Easter Island of the Pacific.154 In these maritime crossings, the exchanges of musical ideas between India and China and the centres of Austronesian and Austroasiatic peoples certainly transpired before the historic period. Such exchanges are indeed themes for future studies on the history of music in Asia.

Notes 1 Arsenio Nicolas, ‘Early musical exchange between India and Southeast Asia’, in Pierre-Yves Manguin, A. Mani, and Geof Wade (eds.), Early Interactions between South and Southeast Asia: Reflections on Cross-Cultural Exchange. Singapore: Institute of Southeast Asian Studies, 2011, pp. 343–365. 2 kaṁsa, kaṃsa, kamsa or kāmsya are representations of a form with nasalized vowel preceding -s [kãsa], which today, in some languages, is reflected with a velar nasal pronunciation [kãŋsa]. The symbol m does not represent a bilabial nasal [m]. (Ritsuko Kikusawa and Lawrence Reid, personal communication). 3 Monier-Williams, A Sanskrit-English Dictionary. Oxford: Clarendon Press, 1905, p. 241. 4 Krishna K, Murthy, Archaeology of Indian Musical Instruments. Delhi: Sundeep Prakashan, 1985, p. 76, pl. 23. 5 Jean Deloche, ‘Musical instruments in Hoysala sculpture (12th and 13th centuries)’, BEFEO 1988, 77: 57–68, fig. 1 a, nos. 2, 3, 4. 6 Jaap Kunst, Music in Java: Its History, Its Theory and Its Technique. The Hague: Martinus Nijhoff, 1973. 7 Colin McPhee, Music in Bali. New Haven: Yale University Press, 1966, fig. 2–7. 8 Jose Maceda, Gongs and Bamboo. A Panorama of Philippine Musical Instruments, Tables, Maps, Glossary and Photo Layout by Marialita Yraola, Quezon City: University of the Philippines, 1998, p. 142, ill. 151, 153. 9 Sharada Srinivasan, ‘High-tin bronze bowl making in Kerala, South India, and its archaeological implications’, in Asko Parpola and Petteri Koskikallio (eds.), South Asian Archaeology, Proceedings, Twelfth International Conference of the European Association of South Asian Archeologists. Helsinki University, 5–9 July 1993. Vol. 2, pp. 695–706, 1994, p. 700, quoting Kangle (1972: 108). 10 Himanshu Prabha Ray, The Archaeology of Seafaring in Ancient South Asia, Cambridge: Cambridge University Press, 2003, pp. 156, 185. 11 Srinivasan, ‘High-tin bronze bowl making’, p. 701; Ray, The Archaeology of ­Seafaring, p. 109. 12 Meera Mukherjee, Metalcraftsmen of India, Anthropological Survey of India Memoir No. 44. Calcutta: Reliance Printing Works, 1978, pp. 349–350. 13 Srinivasan, ‘High-tin bronze bowl making’, p. 700.

232  Arsenio Nicolas









The Sanskrit term Kangsa in Southeast Asia  233

32

33

34 35 36

37 38 39 40

41 42 43 44 45 46 47 48

49 50 51

and Chinese tradeware (Nicolas 2009: 68); See also Roxanna Brown, ‘Guangdong: A missing link to Southeast Asia’, in M. Roxanna Brown (ed.), Guangdong Ceramics from Butuan and Other Philippine Sites. Manila: Oriental Ceramic Society of the Philippines, Oxford University Press, 1989, pp. 81–85. Arsenio Nicolas, ‘Gongs, bells and cymbals: The archaeological record in maritime Asia from the ninth to the seventeenth centuries’, Yearbook for Traditional Music, International Council for Traditional Music and Cambridge University Press, 2009, pp. 62–93. West Gallery, South Wing: 49 m long; South Gallery, West Wing: 94 m long; South Gallery, East Wing: 66 m long; East Gallery, South Wing: 49 m long; East Gallery, North Wing: 52 m long; North Gallery, East Wing: 66 m long North Gallery, West Wing: 94 m long; East Gallery, North Wing: 51 m long. Nicolas field data 2009, 2011. Polkinghorne, Pottier, and Fischer, ‘Evidence for the 15th century Ayutthayan’, p. 101. Ashley Thompson, ‘Pilgrims to Angkor: A Buddhist ‘cosmopolis’ in Southeast Asia?’, Bulletin of the Department of Archaeology, Royal University of Fine Arts, 2004, 3: 88–119. This recalls the Hindu-Buddhist period in Javanese history, exemplified by the Hindu Prambanan temple, and the several Buddhist temples in the area – Candi Sewu, Candi Plaosan, and Candi Kalasan. Jean Boisselier, ‘Note sur les bas-reliefs tardiffs d’Angkor Vat’, Journal Asiatique, 1962, 250: 244–246, esp. p. 244. Tan et al., ‘The hidden paintings of Angkor Wat’, p. 560. Rita Ringis, Thai Temples and Temple Murals. Singapore: Oxford University Press, 1990, p. 38. For a description with sketches of these musical instruments, see George Groslier, ‘Chapitre XII. Les instruments de musique’, in George Groslier (ed.), Recherches sur les Cambodgens. Paris, 1921, figs. 81 and 82; For some description of the musical and ritual life in Angkor, see Michael Coe, Angkor and the Khmer Civilization. London: Thames and Hudson, 2003, pp. 179–185. The terms horns and trumpets are used arbitrarily to describe the blown instruments illustrated on the bas-reliefs. Data provided by Ellen Rustig. Sam-Ang Sam, ‘Cambodia’, Grove Dictionary of Music and Musicians, 2011, 3: 861; 2002: 17–23. Harry L. Shorto, A Mon-Khmer Comparative Dictionary. Pacific Linguistics 579, Research School of Pacific and Asian Studies, The Australian National University, 2006, pp. 65, 421. ̊ Dassanīyabhāb Khmaer = Khmer Ganàkammakar ̣ Srāvjrāv Silpà Vappadham, performing arts (in Cambodian language). Phnom Penh: Ganà Kammakār ̊ 2003. Srāvjrāv Sīlpa Vappadharm, An Raksmey, Siem Reap, Personal Communication, December 2019. Louis Finot, Les inscriptions. vol. II. Le cirque de Mi-son (Quang-nam). Hanoi: F.H. Schneider, 1904, pp. 84–99; see also Ramesh Chandra Majumdar, Ancient Indian Colonies in the Far East. Lahore : Punjab Sanskrit Book Depot, 1927, pp. 74–88. Karl-Heinz Golzio, Inscriptions of Campa. (Based on the editions and translations of Abel Bergaigne, Etienne Aymonier, Louis Finot, Edouard Huber and other French scholars and of the work of R.C. Majumdar). Aachen: Shaker, 2004, pp. 68, 72. Edouard Huber, ‘Etudes Indochinoises’, BEFEO XI, 1911, 3–4: 259–311, esp. 15–22; see also R.C. Majumdar, Ancient Indian Colonies in the Far East. Lahore: Punjab Sanskrit Book Depot, 1972, pp. 139–143. Golzio, Inscriptions of Campa, pp. 118, 119. Ibid., pp. 178, 179. See also Finot, Les inscriptions, pp. 976–977; and Majumdar, Ancient Indian Colonies in the Far East, p. 207 (no. 87) and pp. 208–209 (no. 90).

234  Arsenio Nicolas

52 Po Dharma, Quatre lexiques malais-cam anciens, p. 360. Po Dharma listed several terms for musical instruments. The terms gangsa and kangsa (Skt., also gamak: laiton (brass), as well as tembaga (also tamaga, tamagra, tamaka, kamada = cuivre; bronze) are known (Po Dharma 1999: 360). The following musical terms are also found in these texts: canang or caneng (petit gong, small gong); gendang (also gunang, ginaong, gundang: tambour; drum); gong (also gaong, glaong: gong, cymbal); and bangsi (Skt.; also bengse: transverse flute) (ibid.: 358). 53 Saverous Pou (2003: 254). The Middle Khmer form is kansatal ‘cymbal’ and is based on the Skt. form kamsyatala. 54 Tome Pires, The Suma Oriental of Tomé Pires, trans. A. Cortesao. London: The Hakluyt Society, 1944, p. 96, n 5. 55 Pires, The Suma Oriental of Tomé Pires, pp. 99–100. Pires defines ganza as copper, or alloy of copper and tin.. Pires defines ganza as copper, or alloy of copper and tin. 56 Arsenio Nicolas field notes 1986, 2014. 57 Po Dharma, Quatre lexiques malais-cam anciens, p. 360. 58 George Condominas, Musique Mnong Gar du Vietnam, One 12″ LP record, OCORO OCR 80, Paris: Collection du Musee de l’Homme, ORTF, 1974. 59 Jose Maceda, ‘Introduction’, in J. Buenconsejo (ed.), A Search in Asia for a New Theory of Music. Quezon City: UP Center for Ethnomusicology, University of the Philippines, 2004, , pp. ix–xxv, xxii–xxiii. 60 Nicolas, ‘Gongs, Bells and Cymbals’, pp. 62–93. 61 Emmanuel Guillon, Cham Art. Treasures from the Da Nang Museum, Vietnam. Bangkok: River Books, 2001; George. Maspero, The Champa Kingdom: The History of an Extinct Vietnamese Culture. Bangkok: White Lotus Press, 2001; Henri Parmentier, Cham Sculpture of the Tourane Museum (Da Nang, Vietnam) Religious Ceremonies and Superstitions of Champa. Bangkok: White Lotus Press, 2001; Vibrancy in Stone: Masterpieces of the Da Nang Museum of Cham Sculpture. Bangkok, Thailand: River Books, 2018; Tranh Ky Phoung, Personal Communication, December 2, 2019. 62 Arthid Kamhongsa, Personal Communication. College of Music, Mahasarakham University, Thailand. 63 Hans Penth, ‘Kunst in Lan Na Thai: Der Grosse Gong im Kloster Phrathat Haripunchai’, Artibus Asiae, 1970, 32(4): 307–314, esp. pp. 308–309; Arsenio Nicolas field notes, 1986, 2014. 64 Penth, ‘Kunst in Lan Na Thai’, p. 314. 65 Ibid., p. 313, n. 12. 66 Zoetmulder, Old Javanese-English Dictionary. 1982. The Sanskrit term tamra is defined as copper or brass, and becomes tumbaga in Old Javanese (ibid.: 1924). 67 Jaap Kunst, Hindu-Javanese Musical Instruments. The Hague: M. Nijhoff, 1968. 68 Johannes G. de Casparis, ‘Some notes on words of Middle Indian origin in Indonesian languages (especially Old Javanese)’, in L. Santa Maria, F.S. Rivai, and A. Sorrentino (eds.), Papers from the III European Colloquium on Malay and Indonesian Studies, Naples, 2–4 June 1981. Naples: Dipartimento di Studi Asiatici, Instituto Universitario Orientale, 1988, p. 57. 69 Kunst, Hindu-Javanese Musical Instruments, pp. 91, 92. 70 Louis C. Damais, Répertoire onomastique de l'épigraphie javanaise, jusqu'a pu sindok ´Sri Isanawikrama Dharmmotungadewa; étude d'epigraphie indonésienne. Paris: École française d’Extrême-Orient, 1970, pp. 748, 749, 925. 71 Jan Wisseman-Christie, ‘Javanese markets and the Asian sea trade boom of the tenth to thirteenth centuries, AD’, Journal of the Economic and Social History of the Orient, 1998, 41(3): 344–381, esp. pp. 370–371. 72 Zoetmulder, Old Javanese-English Dictionary, p. 492.

The Sanskrit term Kangsa in Southeast Asia  235

73 Arsenio Nicolas, Alat Muzik Melayu dan Orang Asli (The Musical Instruments of the Malay and the Orang Asli), Occasional Papers No. 9. Bangi: Institut Alam dan Tamadun Melayu, Universiti Kebangsaan Malaysia, 1994; 2017. 74 Casparis, ‘Some notes on words of Middle Indian origin’, pp. 51–52. 75 An early Tagalog-Spanish dictionary (Pedro de San Buenaventura, Vocabulario de Lengua Tagala, El romance Castellano presta primera. Primera y segunda parte. Valencia: Librerías Paris-Valencia. (Originally published: 1613). Recorded the term panday with two meanings: in general it refers to an official, but in particular, a platero (silversmith or jeweller) and herrero (blacksmith or ironworker) and gives the following examples: panday ginto (goldsmith) and panday bacal (ironsmith), but in particular, a platero (silversmith or jeweller) and herrero (blacksmith or ironworker) and gives the following examples: panday ginto (goldsmith) and panday bacal (ironsmith). 76 Casparis, ‘Some notes on words of Middle Indian origin’, p. 68. 77 Ibid., p. 66. 78 Edi Sedyawati, ‘Preface’, in W.H. Kal (ed.), Precious Metals in early Southeast Asia. Proceedings of the Second Seminar on Gold Studies. Amsterdam: Tropenmuseum, 1999, pp. 7–8. 79 Zoetmulder, Old Javanese-English Dictionary, p. 492. 80 Hendrick Herman Juynboll, Oudjavaansch-Nederlandsche Woordenlijst. Leiden: E.J. Brill, 1923, p. 159. The reference to Ramayana is from XXVI, 13. The ninth century dating of this kekawin still remains problematic. Kunst has classified this kekawin under literary texts that remain undated or have yet to be dated. Kunst, Hindu-Javanese Musical Instruments, p. 111, note 60. 81 Zoetmulder, Old Javanese-English Dictionary, p. 492, mentioned in Kekawin Smaradahana, dated c. 1210 CE. 82 Theodoor G. Pigeaud, Java in the Fourteenth Century, Volumes I–V. The Hague, 1960, vol. I: 50, vol. II: 79; vol. III: 77. 83 McPhee, Music in Bali, p. 30. 84 McPhee, Music in Bali, fig. 40. In the gamelan gong in Sulahan, Bangli, central Bali, the metallophone is called saron, instead of gangsa. 85 McPhee, Music in Bali, p. 368. 86 Soepomo Poedjosoedarmo, ‘Wordlist of Javanese Non-Ngoko Vocabularies’, Indonesia 1969, 7: 165–190, esp. p. 175. Nicolas, Alat Muzik Melayu dan Orang Asli; Arsenio Nicolas, ‘Musical terms in 87 Malay classical literature: The early period (14th–17th century)’. Nalanda Sriwijaya Centre Working Paper No. 14. Singapore: Institute of Southeast Asian Studies – Yusof Ishak Institute, January 2017. 88 A similar instrument is found in the piphat ensemble in Thailand. This type of brass/bronze or iron bar xylophone is called ranat thong or ranat ek lek respectively, from the mid nineteenth century during the reign of King Rama IV (1851–1868) (David Morton The Traditional Music of Thailand. Berkeley: University of California Press, 1976: 68). 89 Nicolas field notes, 1979–1993; 2009. 90 Roger Vetter, ‘Music for the ‘lap of the world’: Gamelan performance, performers and repertoire in the Kraton Yogyakarta’. Unpublished Ph.D. Dissertation, University of Wisconsin-Madison, 1986, p. 120. 91 Nicolas, ‘Gongs, bells and cymbals’, pp. 62–93. 92 Albert E. Jenks, The Bontoc Igorot. Manila: Bureau of Printing, 1905, p. 190. 93 Harold C. Conklin, Ethnographic Atlas of Ifugao: A Study of Environment, Culture and Society in Northern Luzon (with the special assistance of Pugguwon Lupaih and Miklos Pinther). New Haven and London: Yale University Press, 1980, pp. 31–32, 98. Conklin proposed that all bronze gongs are of foreign origin, presumably southern Chinese or mainland Southeast Asia. Maceda, however, proposed that the music of flat gongs in Northern Luzon is related to that of the

236  Arsenio Nicolas



­

́

The Sanskrit term Kangsa in Southeast Asia  237



238  Arsenio Nicolas



The Sanskrit term Kangsa in Southeast Asia  239



12 CONNECTED WORDS Coins and maritime worlds of the Indian Ocean Shailendra Bhandare

The centuries following the fourth witnessed the emergence and establishment of two great Empires in the Middle East – the Byzantine Empire to the North and West and the Sasanian Empire to the East. This effectively shifted the route focus of the India trade to the North of the Arabian Peninsula. Now the Persian Gulf and the large river systems of Mesopotamia played a crucial part in the trade between what became the ‘New West’, with its core lying around Byzantium, Syria, Mesopotamia and Egypt and the old ‘East’. The Sasanians viewed this trade with great interest as their empire sat to the north of its main passage routes. Whitehouse & Williamson1 and Malekandathil2 have offered a detailed account of Sasanian involvement in the trade across the Persian Gulf with particular emphasis on emergence of port sites on the Iranian coast. Archaeological proof in the form of a reciprocity of Sasanian and India ceramic evidence has been well documented in finds in excavations of coastal sites such as Sanjan3 in India and Siraf and Rishahr on the Iranian coast of the Persian Gulf.4 There is ample evidence to indicate that the networks that were established during the ascendancy of the Byzantine and Sasanian polities were functional and thriving after the most significant political upheaval of the times – the advent of Islam and the fall of the territories that once belonged to these empires to a nascent Arab/Islamic state. After a turbulent century that saw conquest, assimilation and reformulation, the establishment of Baghdad as the seat of the Abbasid Caliphate was a marker for an epoch of stability and flourish that encouraged a reaffirming of older trace and commercial links. Baghdad was located on the River Tigris, which provided a direct point of access to the Persian Gulf through its delta ports like Basra. It meant that the network route north of the Arabian Peninsula retained its importance, with new nodal towns like Damascus and Aleppo located to the West of Baghdad, rising to prominence in its eventual linkage with

Connected words: coins and maritime worlds  241

the Eastern Mediterranean and the Byzantine worlds. As McPherson has indicated, the rise of Islam not only facilitated the trade between Islamic and non-Islamic lands, it also invigorated it. As he suggests, polities on both sides of the Islamic and non-Islamic divide ‘promoted commerce as a means of gaining wealth’, and its outcome was a ‘ripple effect’ in terms of an upsurge in the commercial activity in the Western Indian Ocean region, bordering the West coast of India.5

Coin circulation at the Islamic interface It is inconceivable to envisage trade and commerce without the mediation of money. It will be appropriate to note at the beginning, that the notion of India existing in an era where there was ‘paucity’ of circulating coinage during CE 500–1000 can be effectively put to rest. Indian economy was sufficiently monetised during this period. In a period where coins were worth the actual content of metal they were composed of, any foreign coins made of precious metals like gold could also enter into circulation and finds of Byzantine gold coins in India are a testimony to this. Both Byzantine and Sasanian empires were heavily monetised and each had a specific currency regime. The Byzantine monetary system was largely bimetallic, with the gold Solidus and the copper Follis (with Nummi, its subdivisions) forming its backbone.6 The Byzantine solidi were struck from a number of mints in the sixth to mid-eighth centuries, but as the empire declined under the advance of the Arabs, the number of mints gradually dropped after the eighth century. The Sasanian currency, on the other hand, was largely composed of silver drachmas with an occasional issue of gold.7 The Eastern and North-Eastern parts of Iran were served largely by a copper currency imitating Sasanian coin design. Much like the Byzantine, the Sasanian currency was issued from a number of mints and served as the main transactional avenue for collection of taxes and military payment. In their design, the most outstanding feature of these currencies was that they were focused on religion – the Byzantine currency design made ample use of Christian motifs, including a bust of Jesus Christ as the Pantokrator and features such as a globus cruciger (an orb surmounted by a crucifix) or a cross on a stepped pedestal. Some Byzantine coins were struck in the name of Christ, as ‘King of Kings’. The Sasanian coinage was also struck in the name of ‘King of Kings’, in this case the emperor. On the reverse of most Sasanian coins, a fire altar, served by armed attendants on either side, is shown as a representation of Zoroastrianism, which was the state religion of the empire. The fire in the altar was recognised as the ‘Fire of the King’ by an inscription, so there was a close connection between kingship and the fire. The image of the emperor in both cases was shown with royal attributes – in the case of the Byzantine coins, either a bust or a full/halflength portrait of the emperor appeared on coins. His depiction was that of a suitably authoritative figure, sporting attributes like a royal cloak, a flowing beard or a pair of starkly staring eyes. He was also depicted holding

242  Shailendra Bhandare

overtly Christian symbols like a crucifix sceptre or a globus cruciger. The Sasanian emperor was depicted sporting a massive headgear/crown that had celestial and ­divine attributes such as a large solar disc and wings. As the Islamic Empire grew in parts formerly controlled by these two empires, no doubt the overall design of the coins and the emphasis it laid on the divine right of kingship its issuers upheld came into a direct ideological conflict with the idea of the ‘king’ as defined in Islam. At an early time, Arab Muslims did not even have a ‘king’ as such; they had a ‘leader’, called the Caliph, who led the ‘Faithful’ against the ‘Infidels’ to spread the word and law as dispensed by Allah, the only God who is nonmanifest. Early Islamic democracy gave way to regard the office of the Caliph as hereditary and kingly. Two rights – the mention of the king in congregational sermons held on Fridays and the right to have his name mentioned on the coins of his realm – came to be regarded as the principal prerogatives that defined ­Islamic sovereignty.8 With such a clear emphasis and nexus between coins and kingship, it would seem surprising that the early Arab rulers of the Levant and Iran carried on with typology and design of the pre-Islamic coinages in those regions. However, coins are functional objects of great economic significance and a sudden change in their appearance is often greeted with suspicion, and ever worse, discounting. This harbours the potential for causing a threat to the circulatory and fiscal mechanisms of which coined money is an integral part. For this reason, as Islam spread through the erstwhile Byzantine and Sasanian lands, the tendency of the early Islamic rulers was to issue coins in the prevalent types. The coin types, however, were cleansed gradually of their pre-Islamic character by making salient, yet not too drastic, changes to the typology and by incorporating some distinctly Islamic features such as the words Allah or Bism-Allah or indeed a partial or full profession of the faith (Shahada). Heidemann9 has presented a survey of these changes, and their progress alongside the Islamic conquest. It all changed rather dramatically in the 77 year of Hegira (CE 696–697) when the Umayyad Caliph instituted a reform of coinage and issued the first gold coins of a fully Islamic character (Figure 12.1) – they were entirely aniconic and struck in the name of none other than Allah. Both these features were radical departures from the norm in terms of coin design. Until this time, coinage had always been issued either in the name or the image, or both, of a king. The inscriptions on these coins were in Kufic script and of a mixture of religious and secular nature. On the obverse, the inscription in the centre was the Kalimat at-tawḥīd, which went lā ilāha illā-llāhu waḥdahu lā sharīka lahu (‘there is no god except Allah, and one [is] He; (there is) no partner to him’) in three lines. In the margin, a statement appeared from the Quran – muḥammadur rasūlu-llāh arsalahu bi-'lhudā wa dīn al-haqq lī-yuzhirahu 'ala al-dīn kollihi walau kariha al-mushrikūn (‘Muhammad is the messenger of Allah; him He sent with guidance and true faith to make it prevail over all other faiths even though the polytheists may hate it’). This was the Sura 9:33, referred to as the Surat at-tauba. The reverse inscription

Connected words: coins and maritime worlds  243

FIGURE 12.1 Umayyad

gold Dinar, dated AH77 (Roma Numismatics Ltd., Auction XVIII, 29 September 2019, Lot 1362 (https://www.coin archives.com/w/openlink.php?l=4024961|4162|1362|c05fcb369ca8e 22cbfed9c68d765ecc2).

proclaimed the oneness and the indivisibility of Allah, as Allah ahd Allah al-samad lam yalīd wa lam yulad (‘Allah [is] One; Allah [is] the Eternal, the Absolute; not begetting and not begotten’) in three lines. This was also a Quranic quote, taken from Sura 112 (al-ikhlās). The only non-Quranic inscription was found in the margins of the reverse, and spelled out the authority and the date in which the coin was struck – b-ismi-llāh Darb hadhā al-dinar fī sanat sab'a wa sab’ain (in the name of Allah struck this dinar in the year 7 and 70). Two years later a silver coinage based on the same design precepts was also introduced into circulation. The reformed coinage, although radically different in appearance to its non-Islamic and partly Islamic precursors, was in the same league in weight standards. The gold coin, now called the Dinar, weighed an average of Byzantine solidi in circulation of that time, and the weight (4.25 gm) became the new standard as a Mithqal. The silver ‘Dirham’ was slightly lower in weight in comparison to the Sasanian drachms at about 2.97 gm. Gradually, other inscriptional features, such as the name of a mint, and occasional citations to the representative of the Caliph, were also added in the design programme.

The Arab/Islamic World and the Indian subcontinent in c. CE 700–1000 Towards the end of the first century of the Islamic Era, the conquering armies of the Arab governor al-Hajjaj ibn Yusuf under the command of Muhammad bin Qasim reached Sindh via Balochistan and Makran and established the first Islamic emirate in the Indian subcontinent. Bin Qasim marked his conquest of the port of Deybal on the mouths of the Indus by striking silver Dirhams of the Umayyad type, in the year AH95.10 The quick spread of Arab rule to the northwest and northeast from Sindh is attested by silver Dirhams bearing the mint-names of al-Qandal (identified as Gandava

244  Shailendra Bhandare

in East Balochistan)11 struck in the year AH96 and ‘al-Hind’, which was presumably struck in the city of Multan in the year AH97.12 The political situation further east from Sindh, into the Indian heartland, was characterised by a struggle for imperial supremacy. Three great polities vied against each other to dominate Northern and Peninsular India at this time. They were the Rashtrakutas of the Deccan, the Gurjara-Pratiharas of Northern and Western India, and the Palas of Northern and Eastern India (Bengal). The city of Kannauj, located in the upper Gangetic basin or Madhyadesha, often became a focus of this long-drawn power struggle. It was conquered and plundered many times by armies of warring dynasties. The Pratiharas were actively hostile to the Arab emirate established in Sindh;13 kings Nagabhata I (c.CE 730–760) launched a punitive expedition in response to an Arab raid into Central India and his descendant Nagabhata II (c.CE 800–833) is noted to have defeated the ‘Turks’.14 The hostility of the Gurjara-Pratiharas towards the Arabs meant that their rivals from South India, the Rashtrakutas, developed a friendly relationship with Arabs. The fact that much of the Northern half of the Western coastline of India rested firmly under the control of the Rashtrakutas or their feudatories meant that the centuries-long maritime trade networks fostered with the Indian Ocean region, particularly through the Persian Gulf route were actively supported by the Rashtrakutas. Indeed, as we will see later, the Rashtrakutas encouraged establishment of Arab settlements in their domains. Apart from the more traditional wares involved in the trade, such as timber and spices, there was a growing need for martial equipment for the Rashtrakutas to sustain their long and fierce campaigns against the Pratiharas and Palas to the North and the Pallavas and the Cholas to the South. Horses, imported from Arabia, were a mainstay of both Pratihara and Rashtrakuta armies.15 Horse trade was thus a particularly valuable component of the bulk of trade that passed across the Indian Ocean between Arabia, the Middle East and Indian Western coast. It is a matter of some coincidence that the establishment of Baghdad as the capital of the Abbasid Caliphate by the Caliph al-Mansur in CE 762 was almost contemporary to the establishment of the Rashtrakutas as an independent dynasty when Dantidurga (c.CE 735–756), the first Rashtrakuta king, overthrew his master the Chalukya king Kirtivarman II. The Rashtrakuas emerged as a major political power rather quickly after the reign of Dantidurga, under the illustrious careers of his successors, ­Krishna I (c. CE 756–774), Govinda II (c. CE 774–780), Dhruva (c.CE 780–793) and Govinda III (c. CE 793–814). The comparatively longer reigns of further two rulers, Amoghavarsha (c. CE 814–878) and Krishna II (c. CE 878–914), marked the apogee of the Rashtrakuta Empire. But during the tenth century, the dynasty saw increasing internal feuds, challenges posed by belligerent feudatories and defeats at the hand of resurgent polities like that of the Paramaras. Finally, at the end of the tenth century the Rashtrakutas were overthrown by Tailapa the Chalukya, a descendant of the same lineage that the Rashtrakutas had brought down nearly three and a half centuries before.16

Connected words: coins and maritime worlds  245

By far the most interesting source of information about the Rashtrakutas and their activities are Arabic texts written by Arab geographers. The most accessible source of these mentions is Elliot and Dowson.17 Noteworthy among these are the accounts of Sulaiman (the Silsilat ut-Tawārikh, chronologically the earliest, dated to AH237 or CE 851); Ibn Khurradādhbih (the Kitāb ul-Masālik wa al-Mamālik; he died in AH300/CE912); al-Mas‘udi (the Muruj ul-Zahab, c.AH332/CE 943); al-Istakhri (the Kitāb ul-Akālim, c.AH340/CE 951); and Ibn Hawkal (also known as the Kitāb ul-Masālik wa al-Mamālik, c.AH366/CE 976). The Rashtrakutas are referred to by the Arab geographers as ‘Balharā’, which is most likely a corruption of Vallabha, or Vallabha Rāja, a title or a part of titles the Rashtrakuta kings are known to have held (‘Balharā is the title borne by all kings of this dynasty’ – Sulaiman18). In most instances, the word is used both an individual/collective and in a territorial sense - firstly as with the definite article appended (‘the Balharā’) and secondly as a name to describe the country or region that ‘the Balharā’ rules. The Balharā is ‘king of kings’ and the ‘most powerful monarch’ of ­India. Sulaiman famously mentions that people of India and China regard the Balharā as one of the four most powerful kingdoms of their times, the other three being the Caliph of Baghdad, the king of China and the ‘king of the Greeks’ (Byzantium).19 The territory of the Balharā is often described as the stretch between ‘Kanbaya and Saimur’ (Khambayat and Chaul)20 and the country is named ‘Kamkam’ or ‘Makamkam’, which is conceivably a reference to ‘Konkan’. Al-Mas‘udi describes the capital of the Balharā to be the great city of ‘Mankir’, which is recognizably Malkhed, the Rashtrakuta capital.21 All the geographers make a special mention to the way the Balharā was very friendly towards Muslims – congregational mosques existed in towns like Kanbaya (Khambayat), Sindan (Sanjan) and Saimur (Chaul) where practices like calling the Azan were tolerated; Muslims were encouraged to settle and participate in statecraft, and the Balharā and his ministers always treated Muslim traders with respect and extended protection for their activities. The Balharā was at war with other kings in India, particularly the king of ‘Jurz’ (Gurjara, and also by extension, of Gujarat) whose territories ‘form a tongue of land’.22 Both kingdoms were very powerful and had ‘plenty of horses and camels’ in their armies and particularly impressive elephants. Sulaiman mentions that the Balharā gives regular pay to his armies, ‘as the practice is among the Arabs’.23 Most geographers mention that the coin current in the Balharā country was the ‘Tātariya Dirham’, which was equal to ‘a Dirham and a half of the coinage of the king’.24 Other authors, like al-Mas‘udi and Ibn Hawkal, also repeat this information, but in some manuscripts the name of the coin is read as ‘Tāhiriya’ or ‘Talatawiya’.25 On the basis of the former of these two variants it has been suggested that the coins were probably of the Tahirid dynasty,26 which ruled in Khurasan and Seistan almost contemporarily to Sulaiman, who is the earliest geographer to mention this term. However, the description about the equivalence of the ‘Tatariya/Tahiriya’ dirhams casts a shadow of doubt on this proposition. If at all, these dirhams were the same

246  Shailendra Bhandare

as those issued by the Tahirids, why would they be worth one and a half times of any other king’s dirhams remains inexplicable, because the Tahirid coins do not have any such distinction.

Rashtrakuta coinage: new discoveries When R.S. Sharma compiled a list of the coins of Indian dynasties from the medieval period, he did not find a single attributable coin of the Rashtrakutas in any of the museum collections he consulted. This led him to conclude that the Rashtrakutas never issued any coinage. The impact of his model of ‘paucity of metallic currency’ was such that this sentiment got expressed in several other commentaries on the history of early medieval India, even though some of these commentaries had been highly critical of Sharma’s propositions regarding deurbanisation and the concomitant d ­ ecline in trade.27 In my previous critique of Sharma’s methodology28 I underlined the significance of curatorial treatment of the coin collections when it comes to retrieval of and reliance on the data a repository like a museum can supply. There is always an element of curatorial bias and limitations in the way coin collections in a museum are acquired and arranged. This is not a criticism per se of the ability or expertise of the museum curator; it is more to do with practicalities like funding and the focus of academic interests of a curator as a specialist in particular areas or themes concerning the objects under his curatorial care. This can often mean that significant data, both in terms of quantity and quality, can remain hidden, particularly when the person dealing with the data is not a numismatist and has to rely exclusively on the curator’s expertise in retrieving the data. A good example of this is the often-overlooked category of coin ‘imitations’, which traditionally have drawn much less curatorial interest than coins that are well attributable and therefore can be assimilated in the main collection quite easily. The category of imitative coinages has come under scrutiny of numismatists and historians only very recently,29 but in past they have always been consigned to cabinets, which are subsequently overlooked (and even forgotten!) as repositories of material that is unattributable. Most imitative coinages are crude and not as ‘exciting’ as the prototypes they follow, so from an aesthetic viewpoint they are often considered of not great worth. Since aesthetic appreciation has been traditionally considered as a hallmark of the curator’s trade, objects that are not ‘up to the standard’ are often relegated to a curatorial oblivion. While looking through one such cabinet, labelled ‘Imitations’, in the British Museum I came across a gold coin30 (Figure 12.2), which is described as follows: OBV: a human figure seated cross-legged on a lotus. REV: inscription in proto-Nagari in a single line – Shri Shubhatunga with a

stylised representation of a conch shell below. There is a further inscription on either side of the conch shell – to the left it appears to read, Ane or Ani, and to the right, Kshato.

Connected words: coins and maritime worlds  247

FIGURE 12.2 Gold

coin with legend ‘Shubhatunga’, British Museum (photograph by author).

‘Shubhatunga’, the name mentioned in the coin inscription, is in fact a well-known Rashtrakuta title so there can be no doubt that this is a Rashtrakuta gold coin. The obverse figure can be identified as a Garuda and shows remarkable similarities with the Garuda emblem often seen on the seals accompanying Rashtrakuta copperplate grants. Ostensibly, not being correctly identified and having been consigned to the ‘Imitations’ cabinet, meant that it was never incorporated in the coin data that was supplied to Sharma by the British Museum. The Rashtrakutas often held titles that ended in -Tunga, -Varsha, -Vallabha and -Avaloka suffix. They show a peculiarity in the titles they adopt: the titles are often repeated for kings with the same name. To give an example, all kings named ‘Krishna’ in the dynasty – and they were three – were also called Shubhatunga. So while the attribution of this gold coin can be made undoubtedly to a Rashtrakuta ruler named ‘Krishna’, which of the three rulers named ‘Krishna’ does the coin belong to would be a worthwhile question and we will revisit it a little later. Two more coins with legend Shubhatunga have since come to light (­Figures 12.3 and 12.4). The first of these31 is pretty much of the same type as the British Museum coin, although struck with a different pair of dies. The other32 coin differs a little, inasmuch as it does not have the legend on either side of the conch shell on the reverse. There are only clusters of dots on either side of the conch. The other peculiarity is that the second letter in the legend Shri Shubhatunga is articulated as the retroflex ‘Sh’ letter, which is orthographically incorrect; however, the rapid Prakritization and ‘­Deshification’ of languages during the Rashtrakuta period would not necessarily preclude such an anomaly. A few more Rashtrakuta gold coins surfaced in 2018 on the coin market in India, which were said to have been found as a small hoard somewhere in the Uttara Kannada or adjoining districts of Karnataka. A total of eight to ten pieces, dispersed on the market, have been documented so far. Of these, the two most well-preserved are shown here (Figures 12.5 and 12.6). They belong to a private collector in Mumbai who wishes to remain anonymous, but I must record my gratitude to him for allowing me to publish them.

248  Shailendra Bhandare

FIGURE 12.3 Gold

coin with legend ‘Shubhatunga’ (Classical Numismatic Gallery, Auction 2, 18 December 2010, Lot 38) (https://www.coinarchives .com/w/openlink.php?l=2416121|2330|38|9712ab6f82bc0eadab0d 98bfc7ff0758).

FIGURE 12.4 Gold

coin with legend ‘Shubhatunga’ (Todywalla Auctions, Auction 118, 1 September 2018, Lot 70).

OBV: a figure of a horseman facing left, holding a hook-shaped weapon in

his hand. REV: proto-Nagari legend in two lines – Shri Jagatunga Apra/tihata Gata

Garudadhwaja; a row of symbols – a lotus bud, a Kalasha or vase of plenty and a conch shell – in a row are placed above the legend and a three-petal lotus bloom is below, as if the legend was resting on it. The second coin is of the same type, except the horseman here faces left. One coin of this variety with typical double piercing for use in jewellery has also been documented. Evidently, the inscription on the coins identifies its issuer with the title Shri Jagatunga. While this is also a well-known Rashtrakuta title, the legend here presents it as a pun: the suffix -Tunga means lofty, tall, erect, high, etc., and here it is appended to the word Jaga (the World). But the syntax of the legend suggests that it is the Garuda banner (dhwaja) to which this quality is ascribed. This Garuda banner, apart from standing tall and lofty in the world, is said also to have ‘become invincible’ (Apratihata Gata). It

Connected words: coins and maritime worlds  249

FIGURE 12.5 Gold coin with horse rider facing left (Classical Numismatic Gallery,

Auction 31, 11 August 2018, Lot 81 (https://www.coinarchives.com/w/ openlink.php?l=3549121|3614|81|58f3d28db955ef0c7fab5c36b6eda071).

FIGURE 12.6 Gold coin with horse rider facing right (private collection, M ­ umbai;

photograph by author).

is well-known that some Rashtrakuta rulers were devout Vaishnavas. The names ‘Krishna’ and ‘Govinda’, which were given to many rulers in the dynasty, or the use of the Garuda as the chief motif on their royal seal, are testimony to their devotion to Vishnu. However, since ‘Jagatunga’ was also a title, the legend could also be construed as referring to ‘Jagatunga’ the monarch who was become ‘invincible’ and a banner of Garuda himself. It is no doubt a dual statement of the monarch’s invincibility as well as devotion. By far the most interesting aspect about these coins is the pseudoinscriptions that surround the motifs on both sides. It can be quickly discerned that these are in fact crude copies of Kufic inscriptions seen as marginal legends on Islamic, particularly Abbasid, gold dinars. As we have seen, the marginal inscription on the obverse of Islamic dinars was the Surat al-Taubah or verse 9:33 of the Quran. If one compares the two, a number of forms show a close, yet illegible match. Conceivably, Abbasid dinars served as a model for these Rashtrakuta gold coins, and the engravers of the dies that were used to strike the Rashtrakuta gold coins copied the exergual inscription with an aim to imitate it.

250  Shailendra Bhandare

A close examination of the ‘Shubhatunga’ coins also reveals the same design trait. There are pseudo-Kufic inscriptions on both sides of the coins in Figures 12.3 and 12.4, much like those on the ‘horseman’-type coins of ‘Jagatunga’. However, here they appear to be more stylised and crude than those seen on the latter.

Significance of the coins The first gold coin attributed to the Rashtrakutas was published by Prabhu.33 This was gold gadyana, which he conclusively attributed to Govinda IV (c. CE 930–935). With the coins described here, the number of coins that can be attributed to the Rashtrakutas has gone significantly up and the view that the dynasty never issued any coins can be safely put to rest. Now the question arises to which Rashtrakuta rulers can these coins be ascribed to, because as we have seen the titles were held by more than one ruler. ‘­Shubhatunga’ was the title of Krishna I (c. CE 756–774), Krishna II (CE 878–914) and Krishna III (CE 939–967). When I published the British Museum coin for the first time I indicated that given the economic and political context, and the occurrence of the pseudo-Kufic inscriptions, K ­ rishna II fits better as the issuer of the British Museum coin and as such the other two coins newly published here should also be attributed to him. The same quandary is presented to us for ascribing the ‘horseman’ type coins – the title ‘Jagatunga’ was held by four members of the Rashtrakuta family: Govinda II (c. CE 774–780), Govinda III (c. CE 793–814), one of the sons of Krishna II (c. CE 878–914) who predeceased him and a younger brother of Krishna III (c. CE 939–967). The main feature of the coins that helps us to suggest the attribution is the pseudo-Kufic inscription on the reverse. It is clearly derived from the mint/date formula following the prototype of the reverse of Abbasid dinars. When its letter forms are compared it becomes evident that the prototype inscription is probably without any mint name, but the letters fit date forms of AH132–218, which correspond to CE 750–833.34 By far the most likely candidate for the issue of the ‘horseman’ type coins would therefore be Govinda III. He was the successor of Dhruva, the first Rashtrakuta ruler to launch the transformation of the Rashtrakuta kingdom into an empire. Like his father, Govinda III carried on the expansion drive and became the first Rashtrakuta ruler to raid the imperial capital city of Kannauj, making the Gurjara-Pratihara ruler Nagabhata II sue for peace. King Dharmapala of Bengal, an ally of the Pratiharas, was also humiliated. To the East, he subdued the Eastern Chalukya rulers of Vengi and installed his own representative on their throne, thus occupying a large tract of the Andhra country. Further to the south, he inflicted a defeat on the Pallava king Dantivarman and sacked the capital Kanchi in CE 803. To the west, he successfully wrested the region of South Gujarat from the feudatories of the Pratiharas and installed his brother Indra as his viceroy. Indra’s line of succession became a Rashtrakuta branch known as the Rashtrakutas of Lata.

Connected words: coins and maritime worlds  251

The attribution of the ‘horseman’-type coins to Govinda III would also fit in with the attribution of the seated Garuda type coins of Krishna II. On these coins the remnants of the exergual Kufic inscriptions have become more stylised than they are on the ‘horseman’-type coins. In the latter, the die engravers appear to make a closer attempt to match the legends on the prototypal Abbasid dinars, insomuch as some of the letter forms are recognizable and even be restored to their origins. But on the ‘Garuda’-type coins they are degraded enough to be a pseudo-inscription that cannot be backtracked to its prototype. This indicates that the ‘Garuda’-type coins must be later than the ‘horseman’-type coins. This chronological proposition suits the attribution of the former to Krishna II and the latter to Govinda III. It would be indeed interesting to discuss why elements of Abbasid ­dinars were copied in this manner on the Rashtrakuta coins. However, before we do that it is appropriate to note some other interesting data that sheds light not only on the processes involved in ‘trans-regional mimesis’ but also on the trade connections between India’s western coast and the Islamic world across the Arabian Sea. Alongside the ‘Garuda’-coin in the British ­Museum, in the same tray there were some very crude imitations of Islamic dinars (Figures 12.7 and 12.8). These were conceivably collected from South I­ ndia in the nineteenth century, as they belong to the collection of Sir W ­ alter ­Elliot, a colonial administrator in Madras Presidency and an early collector of South Indian coins, who presented his collection to the British ­Museum in 1875. These are very crude and bear slash marks, suggesting that they were tested at some point while they circulated to check their purity and authenticity. The fact that Islamic gold dinars circulated far and wide from their place of origin and came to be regarded as an elite trade currency that could be relied upon for its weight and purity has been well-established. This reliance and trust built around themselves led to them being imitated, officially or unofficially, to produce ‘coins’ that used stylistic and typological mimesis as a visual tool to generate trust as the coins circulated. Numismatists might

FIGURE 12.7 Indian

imitation of an Islamic Dinar, ex-Walter Elliot, British ­Museum (accession number BM 1876, 6.2.2; photograph by author).

252  Shailendra Bhandare

FIGURE 12.8 Indian

imitation of an Islamic Dinar, ex-Walter Elliot, British ­Museum (accession number BM 1876, 6.2.3; photograph by author).

turn their nose up at them, for being indescribable, crude and unattributable, but from the viewpoint of monetary history they are objects that circulated alongside the ‘real McCoy’ and undoubtedly produced their own histories of circulation, exchange and consumption. A brass die to produce such imitations, found at Prakashe in northern Maharashtra, was reported by the present author along with Heidemann.35 A number of such imitations have been reported in the numismatic trade.36 They vary in their style and execution; some are more truthful to their prototypes while others such as the pieces from the Elliot collection are very crude. It is interesting to note that imitating I­ slamic dinars was not a phenomenon that was confined in any way to the East of the Islamic World. To the West, in Europe, the Anglo-Saxon kings in Britain and the Carolingians in France also produced imitations of Islamic dinars. Imitative or ‘defective’ currency was very much a part of the circulatory picture of Europe in the same period that the Rashtrakutas ruled over great parts of Peninsular India. The most common name for these coins was ‘Solidus Mancus’ and as Grierson has indicated,37 the coinage that was designated by this term was essentially an imitative and sometimes debased coinage. The Anglo-Saxon king Offa is known to have promised to pay the Pope 365 ‘mancuses’ every year. Only one of these coins survive (in the British Museum) and it is a near-faithful copy of an Abbasid dinar, dated AH157 (which fall in Offa’s reign), with the Latin legend ‘Offa Rex’ inserted in its Kufic legends, albeit at an 180° orientation (Figure 12.9)! Alongside the imitative gold coins in the British Museum and bearing a very similar slash mark is another gold coin also from the Elliot collection (Figure 12.10). Its attribution is not yet certain, due to the fact that the slash has damaged the legends considerably and they have yet to be read satisfactorily. But given its type similarities with the ‘Garuda’ and ‘Horseman’-type coins discussed above, there can be little doubt that this is also yet another Rashtrakuta gold coin. On the reverse it has a very similar layout to the ‘Horseman’-type coins, with a legend in proto-Nagari arranged in two lines over a three-petal lotus motif. But above the legend, there is a prominent

Connected words: coins and maritime worlds  253

FIGURE 12.9 Gold ‘Dinar’ of

Offa, king of Mercia (AD 756–796), British Museum, accession number 1913, 1213.1 (https://research.britishmuseum. org/research/collection_online/collection_object_details.aspx? assetId=31108001&objectId=1093298&partId=1).

FIGURE 12.10 Gold

coin with seated goddess on obverse, ex-Walter Elliot, ­British Museum (accession number 1876, 6.2.1; photograph by author).

crescent mark. The crescent could well allude to the Rashtrakutas, who claimed their descent from the ‘Lunar Race’ of lore. On the obverse, we see the representation of a four-armed goddess, seated cross-legged on a lotus with one of the hands lowered in a gesture of benevolence. In all likelihood this goddess represents Lakshmi. What is even more interesting is the fact that this coin also shows traces of pseudo-Kufic inscriptions in the exergue, much like the ‘Horseman’ and ‘Garuda’-type coins we have just discussed. From whatever is visible, the legend appears to begin with Shri Ja ya tu Ga ru/da X X X X dha ja thus indicating that a reference to the ‘Garuda banner’ has been made here as well. The aspect of having crude and vestigial Kufic legends in the exergue is not limited to the coins issued in the name of Rashtrakutas alone. A small group of ‘boar’-type gold coins, which are also in the same collection that has the ‘horseman’-type Rashtrakuta coins published above, also exhibits this feature (Figures 12.11 and 12.12, published with the permission of the collector). Both these coins have a boar standing facing right on the obverse, with a short and not conclusively read inscription in Hale Kannada script

254  Shailendra Bhandare

FIGURE 12.11 ‘Boar’-type

gold coin with pseudo-Arabic inscriptions in the ­margins (private collection, Mumbai; photograph by author).

FIGURE 12.12 ‘Boar’-type

gold coin with stylised pseudo-Arabic inscriptions in the margins (private collection, Mumbai; photograph by author).

above. Around the boar motif, a pseudo-Kufic inscription can be made out. Although it is highly stylised, some words like Bism-Allah and Hadha can be made out from their rudiments. On another coin of the same type, the inscription is fully degenerated and looks like an ornamental border. Although these coins are not yet satisfactorily attributed, they could well be issues of Rashtrakuta feudatories ruling somewhere on Karnataka’s coastal fringe. The pseudo-Kufic inscriptions are not always limited to the marginal or exergual space in the design of the imitative coins. There exist unattributable and uninscribed coins that bear designs with a resemblance to the central inscriptional designs of Arab dinars as well. Two examples that the author has studied have illegible horizontal lines in the centre of the obverse. A comparison with Arab dinars makes it abundantly clear that these coins are trying to copy the Kufic legends in the centre of the dinars, but what the engraver has been able to achieve is only a semblance of the original inscription. The nature of these pieces as circulatory currency is beyond doubt as all of these imitative coins are struck to the same weight standard as that of the dinar. In fact, the prevalent standard of gold coinage in South

Connected words: coins and maritime worlds  255

and Peninsular India, which is many times referred to as the Gadyana or Suvarna standard, has a close bearing on the weight standard of the Islamic dinars and their precursors – the Byzantine solidi. The fact that these weight standards could influence each other and be copied across the Arabian Sea stands as a testimony to the impact that monetization of the transregional trade connections had on contemporary Indian coinage. Quite unlike the postulations of Sharma, who advocated a decline of trade during the period in which ‘paucity of metallic currency’ prevailed, these influences and imprints show that transregional trade must not only have survived but thrived during c.800–1000 CE.

The Arab settlement at Sanjan and its monetary evidence Sanjan, on the Western coast of India (location coordinates 20°12′4″N 72°48′4″E), was an important trading port during the early medieval period. It is presently located to the extreme south of Gujarat state and is mainly famous as the mythical landing site for India’s prosperous Parsi Zoroastrian community, who is said to have made landfall here, after being set adrift when they fled Iran carrying the Holy Fire, having suffered a defeat at the hands of the invading Islamic armies. For our period, Sanjan has a threefold importance: 1 it is described almost in every Arab geographic account mentioned above (referred to as Sindan);38 2 it finds mention in two sets of Rashtrakuta copperplate charters (found at Chinchani in Maharashtra, not very far south from Sanjan) as a site of an Arab settlement;39 and 3 the site was subjected to systematic archaeological excavations lasting a few seasons and a report of the findings has been published. It is therefore a unique site for which we have evidence from textual as well as material (inscriptional and archaeological) sources. The archaeological material unearthed at Sanjan gives a clear indication of its connections with the Persian Gulf, with Iranian (Sasanian and Islamic) ceramic and glass evidence40 proving beyond doubt that the port was in close trade connections with Iran in c. CE 600–1000. Numismatic evidence dates from the early medieval to the modern horizons in terms of chronology, but in particular three small copper coins are of great interest to us because they corroborate the inscriptional evidence well. They are illustrated in the BAR publication of the report (numbered as Sanjan SJN-B 2002TT1NE, weight 2.46 gm; Sanjan SJN-B2004TT4NE, weight 2.01 gm and Sanjan SJNK2003B4NE, weight 2.16 gm).41 All three probably belong to the same type: On the obverse, there is a seated figure of Garuda with his hands in Añjali Mudrā or gesture of salutation. The Garuda is flanked on either side by a Swastika. On the reverse there is a Kufic legend in three or four lines, which is not fully readable. On one of the coins (no. 11 in the report) it seems like

256  Shailendra Bhandare

the die has slipped while the coin was being struck, which has made it even more difficult to read fully in comparison to the other coins. The report describes the legend to be ‘al-Sultan or al-Malik…’; however, it is quite clear that the top word is Allah and the line below it is, in all probability, al-Mulk Lillah (‘the kingdom [belongs] to Allah’). Both these are invocations that are well-known on other series of Islamic coins. What follows in the line after this is most likely the name of the issuer; however, it is unfortunately not readable on any of the coins. The find of these three coins at Sanjan is very significant particularly in the context of what we know from other sources such as inscription. The copperplate charter of Rashtrakuta king Indra III (c. CE 914–929) found at Chinchani is dated 17 April CE 926. In verses 16–20 of these plates, a subordinate ruler is introduced. His name is Madhumati and he has a Tājika connection (Tājika-anvaya).42 Desanskritizing the two words, it is evident that the name was ‘Muhammad’ and he was described to be an Arab. This Muhammad the Arab had received the control of the entire Sanjan division (Samyana Mandala) from Krishnaraja – ostensibly, Krishna II, who was Indra III’s predecessor and grandfather. Verse 18 states that Muhammad had conquered the masters of all harbours in the vicinity and appointed his own officers in-charge. Verse 19 states that Muhammad was also known as Sugatipa and he was the son of Sahariyarahara. What precisely Sugatipa means is not clear, but the name of Muhammad’s father is most likely to have been ‘­Shahriyar’. Amongst the pious deeds of Muhammad were the establishment of a ‘soup kitchen’, where rice, curry and ghee was available free of cost and free ferry services across two streams in the Sanjan region. These were executed through Puvvaiya, who was an officer/minister of Muhammad. Further in the text, it is stated that Muhammad made the donation of a village and designated land in another village, towards the upkeep of a small monastery (Maṭhikā) that was founded by a Brahmin named Annammaiya, who was a friend of Puvvaiya, Muhammad’s minister. Details of what the proceeds from the donations should be utilised for are given in further verses. The contents of the Chinchani copperplate grants make it clear that the region of Sanjan was under the rule of an Arab governor during the reigns of at least two Rashtrakuta kings. Muhammad was appointed during the reign of Krishna II, which came to an end in CE 915. The charter itself is dated CE 926, so by the time it was issued, Muhammad must have been regnant as a governor for more than a decade. The feudatories of Rashtrakuta’s ruling over Konkan prior to Krishna II’s rule appear to have been members of the Shilahara family as evidenced by Shilahara records that acknowledge Amoghavarsha, the immediate predecessor of Krishna II, as their overlord. The establishment of an Arab governor at Sanjan most likely caused a temporary hiatus in the Shilahara rule over parts of Konkan.43 Conceivably, to have accorded such a high status to a person of foreign origin, preferring him over a dynasty that had served as feudatories for decades, must have meant that Muhammad was significantly important to the Rashtrakuta king Krishna II. What this significance might have been is not explicitly

Connected words: coins and maritime worlds  257

known, but a calculated guess would indicate the role Arabs played in bringing in valuable commodities like horses through the transoceanic trade. The mention in the copperplate that Muhammad was ‘like a jewel in battles’ that he took over the control of many harbours in the vicinity of Sanjan and appointed his deputies is important in this respect. No doubt, this alludes to the fact that the Arabs controlled vital entrepots for maritime trade in North Konkan. It is interesting to note that while M ­ uhammad is called by the generic name ‘Tajik’, which is most often translated as ‘Arab’, his f­ather’s name (Shahryar) is distinctly of an Iranian origin. So either he might have been an Arab of a sufficiently Persianised nature, or the word ‘Tajik’ ­referred to any person from the wider ‘Arab’ world, irrespective of a specific ethnic connotation. The coins excavated at Sanjan show that the authority of the Muslim governors extended to issuing their own money. The invocation on the coin proclaiming ‘the Kingdom (belongs) to Allah’ suggests that even though the governors participated in acts of piety like upkeep of a Brahmanic monastery, and quite clearly claimed allegiance to the Rashtrakuta king, the nature of their rule was Islamic. However, the final authority rested with the Rashtrakutas as the coins had the royal emblem of the Garuda on them. The coins therefore encapsulate neatly the hybrid nature of political organization practised under the Rashtrakutas. The accounts of Arab geographers mentioned above corroborate the reflections of this tolerance. Apart from the copper coins of the seated ‘Garuda’-type, the Sanjan excavations have also yielded some other important coins. One of them is a small silver coin, commonly referred to as a dramma44 of ‘elephant’ type. However, the coin is in a bad state of preservation, with bits chipped off, as reported in the BAR report (coin no. 15, Sanjan SJN-B2002TT1NE). Photographs of it in a much better state of preservation survive and are available on-line45 so it appears that it was broken due to poor storage and/ or handling. On the obverse of the coin is an elephant marching to right. On the reverse, a Brahmi legend in two lines Shri A mo / (gha) va ri sha is visible. Evidently, the coin was issued in the name of the Rashtrakuta ruler Amoghavarsha (c. CE 815– 877). The weight reported in the BAR report is only 0.14 gm, but large pieces of the coin are chipped away, and it is plausible that the coin in its unbroken state might have weighed closer to 0.35 gm, and that would make it half a dramma. The second important coin, or rather a clipped fraction of a coin, is an Abbasid silver dirham, cut into a small piece (coin no. 14; Sanjan SJN2002TT1NE, weight 0.75 gm). The BAR report does not identify the original coin type from which the bit is clipped, but an attribution to the Abbasid Caliph al-Rashid (commonly known as Harun al-Rashid, c. CE 786–809) was suggested by Dalal in an online article46 and with the comparative illustration he provides, this seems to be correct. The coin is of Muhammadiya mint, which was a very productive mint during the Abbasid period. It is now a part of the modern city of Tehran in Iran. The interesting aspect here is the fact that the coin appears to be clipped intentionally, not accidentally. While the full dirham might have weighed close to 3 gm, the clipped bit weighs

258  Shailendra Bhandare

0.75 gm, which is within the range of the circulatory weight standard of silver drammas. Al-Rashid was a contemporary of Govinda III, the predecessor of Amoghavarsha and the issuer of the ‘horseman’-type gold coins discussed above. Chronologically, therefore, these two silver coins from Sanjan excavations fit our frame of discussion perfectly well. The intentional clipping of an Abbasid dirham into a small fraction that fitted in with the local circulatory standards points to an interesting reappropriation of foreign coins that arrived with the trade into Sanjan.

Epilogue: connected worlds and Rashtrakuta coins The foregoing discussion clearly demonstrates two aspects: 1 First and foremost, Rashtrakuta coinage exists, so the oft-repeated statement that the Rashtrakutas never issued any coins needs to be consigned to history! The data published here, alongside that published in the small book by Jantakal et al.,47 conclusively proves that the Rashtrakutas issued a trimetallic currency. The main reason historians had been oblivious to these coins is that they were never collected and therefore were not represented in museum collections. Most were wrongly attributed or remained unattributed, having been consigned to categories such as ‘imitations’. The data published here sets the record straight. 2 Rashtrakuta precious metal coinage was intimately connected to the trading worlds of the wider Indian Ocean world. Some salient design elements of these coins suggest that at least a part of the stylistic inspiration they derived was from the regions under Islamic rule, located to the West. This influence arrived via the Persian Gulf route of the wider oceanic trading networks, which is corroborated by not only textual and inscriptional evidence, but also by archaeological evidence that is found at either end of this network branch. One may wonder why the Rashtrakutas, and some other issuing authorities, had a preference for the exergual inscriptions of the Islamic dinars when they copied them. This is a complicated question, but answers might be found in the direction provided by an intersection that the disciplines of anthropology and archaeology have followed in te recent past. Arjun Appadurai, in his path-breaking analysis of commodification, consumption and circulation of objects (‘things’ as he calls them),48 points us to the complex and multiple levels of understanding, appreciation and internalization they achieve and foster in the course of their ‘social life’. Coins are by nature objects that yield to exchangeability, and as such, they are exposed to processes of commodification. This creates different levels in which they are socially ‘understood’ – they can be seen and consumed as symbols for cultural appropriation, or markers of status, apart from serving their primary function as a medium of exchange. Darley49 discusses such trajectories in relation to Roman coins in Peninsular India. In a recent work, Smagur50

Connected words: coins and maritime worlds  259

discussed, with the help of a carefully assembled data set, how Roman coins prompted terracotta imitations (‘bullae’) in the Deccan and other parts of Peninsular India. Why aspects are copied and what precisely ‘stylistic ­m imesis’ achieves has been a focus of great anthropological interest, as evident from the trail of research significant contributions like ‘Mimesis and Alterity’ by Michael Taussig51 have generated. The answer to the question raised about Rashtrakuta gold issues may be found along the lines of these inquiries and analyses. From a purely functional viewpoint, it is clear that the Islamic dinars were a successful trade currency and as such commanded the trust and confidence of its consumers with regard to its value and acceptability. This meant that other currencies vied to acquire the same visual appeal as the Islamic dinars so as to create a ‘mimetic reciprocity’ partaking the elite circulatory character of these coins. In a numismatic sense, this phenomenon is found often repeated in India with regard to particular currencies – from the ancient Western Kshatrapas imitating the precursor Indo-Greek drachmae in appearance, to the early modern Marathas and rulers of Mysore designing their own coins in imitation of more successful currencies such as the Mughal rupees of Surat and Arcot mints.

Notes 1 D. Whitehouse, and A. Williamson, ‘Sasanian Maritime Trade’, Iran, 1979, 11: 29–49. 2 P. Malekandathil, ‘The Sassanids and the Maritime Trade of India during the Early Medieval Period’, Proceedings of the Indian History Congress, 2002, 63: 156–173. 3 R. Nanji, Mariners and Merchants: A Study of the Ceramics from Sanjan (Gujarat), Sanjan Excavation Report, vol. 1, Oxford: BAR International Series 2231, 2011. 4 Whitehouse and Williamson, ‘Sasanian Maritime Trade’, p. 39. 5 K. McPherson, ‘Processes of Cultural Interchange in the Indian Ocean Region: An Historical Perspective’, The Great Circle, 1984, 6(2): 84–85. 6 For a preliminary survey see P. Grierson, Byzantine Coinage, Washington, DC: Dumbarton Oaks Research Library and Collection, 1999. 7 The best introductory survey is Robert Göbl, Sasanian Numismatics, English edition, Braunschweig: Klinkhardt & Biermann, 1971. 8 See entry for Sikka in Encyclopedia of Islam, 2nd edition, eds. P. Bearman, Th. Bianquis, C.E. Bosworth, E. van Donzel and W.P. Heinrichs, Leiden: Brill (online edition: https://referenceworks.brillonline.com/entries/encyclopaedia-of-islam-2/ sikka-COM_1076?s.num=0&s.f.s2_parent=s.f.book.encyclopaedia-of-islam-2 &s.q=Sikka accessed 4.12.2019). 9 S. Heidemann, ‘The Evolving Representation of the Early Islamic Empire and Its Religion on Coin Imagery’, in A. Neuwirth et al. (eds), The Qur’an in Context: Historical and Literary Investigations into the Qur’anic Milieu (Text and Studies on the Qur’an 6), Leiden: Brill, 2010, pp. 149–196. 10 M. Klat, Catalogue of the Post-Reform Dirhams: The Umayyad Dynasty, London: Spink, 2002, no. 377. 11 Enrico Leuthold, Jr., Richerchi sui Dirham Umayyadi, Milan: Published by the author, 2008, p. 39; also A.M. Fishman, and I.J. Todd, The Silver Damma: On the mashas, daniqs, qanhari dirhams and other diminutive coins of India 600–1000CE, Mumbai: IIRNS Publications LLP, 2018, p. 92.

260  Shailendra Bhandare

12 S. Khan, and R. Babur, ‘Coins of al-Hind’, Journal of the Oriental Numismatic Society, no. 221, 2014, pp. 27–30. 13 Vibhuti Bhushan Mishra, The Gurjara-Pratīhāras and Their Times, Delhi: S. Chand, 1966, p. 17. 14 These details are glossed in the Sagar Tal inscription of Bhoja I. See H. Shastri, ‘Inscriptions from Gwalior’, Archaeological Survey of India Annual Report 1903–04, Calcutta, pp. 279–280. 15 R. Chakrabarti, ‘Horse Trade and Piracy at Tana (Thana, Maharashtra, India): Gleanings from Marco Polo’, Journal of the Economic and Social History of the Orient, 1991, 34(3): 168–169. 16 For an overview of Rashtrakuta history, see A.S. Altekar, The Rashtrakutas and their Times, Poona: Oriental Book Agency, 1934. 17 H.M. Elliott, and J. Dowson, The History of India as Told by Its Own Historians: The Muhammadan Period, vol. 1, London: Trübner & Co., 1867. 18 Ibid., p. 4. 19 Ibid., p. 3. 20 Ibid., p. 27. 21 Ibid., pp. 19–20. 22 Ibid., p. 4. 23 Ibid., p. 3. 24 Ibid., p. 3–4. 25 Ibid., p. 24. 26 Ibid., p. 3. 27 See, for example, the discussion by Andre Wink in A. Wink, Al-Hind: The ­Making of the Indo-Islamic World, vol. 1, Leiden: Brill, 1996, p. 308. 28 S. Bhandare, ‘Space for Change: Evaluating the “Paucity of Metallic Currency” in Medieval India’, in Himanshu P. Ray (ed.), Negotiating Cultural Identity: Landscapes in Early Medieval South Asian History, New Delhi: Routledge, 2015, pp. 159–202. 29 Paula Turner discusses imitations of Roman coins found in India in P. Turner, Roman Coins from India, New York: Routledge, 2016, pp. 37–40 ; Another worthwhile publication is P. Berghaus, ‘Roman Coins from India and Their Imitations’, in A.K. Jha (ed.), Coinage, Trade and Economy, Nashik: IIRNS publications, 1991, pp. 108–121. 30 Bhandare, ‘Space for Change’, p. 175, plate 7.14. 31 Classical Numismatic Gallery Auction 2, lot 38, 18 December 2010 (https:// www.coinarchives.com/w/openlink.php?l=2416121|2330|38|9712ab6f82bc0ead ab0d98bfc7ff0758). 32 Todywalla Auctions, auction no. 118, lot no. 70, 1 September 2018. 33 M. Mukunda Prabhu, ‘Karnataka Coins – Recent Finds and Research’, Quarterly Journal of the Mythic Society, April–June 1997, 88(2), p. 105 34 I am grateful to Stephen Album for this suggestion. 35 S. Bhandare, & S. Heidemann, ‘A Die for Imitation Umayyad Dinars Found in India’, Oriental Numismatic Society Newsletter, 2000, 162: 8. 36 One such example can be seen at https://www.coinarchives.com/w/openlink.php ?l=4247711|4449|224|66ac606d41d8bc955935c8b2483b3992. 37 P. Grierson, ‘Carolingian Europe and the Arabs: The Myth of the Mancus’, ­Revue belge de Philologie et d'Histoire Année 1954, 32–34: 1059–1074. 38 Elliot and Dowson, History of India, pp. 14, 27, 38. 39 Sircar, D.C., ‘Rashtrakuta Charters from Chinchani’, Epigraphia Indica, 1957–58, 32: 45–60. 40 R. Nanji, Mariners and Merchants, op. cit. 41 S. Rajaguru, et al., Early Medieval Sanjan: Aspects and Analysis, Sanjan Excavation Report, vol. 3, Oxford: BAR International Series 2509, 2013, p. 39 (coin 11), p. 46 (coin 54), p. 48 (coin 62).

Connected words: coins and maritime worlds  261

4 2 Sircar, ‘Rashtrakuta Charters’, p. 47. 43 R. Chakravarti, ‘Monarchs, Merchants and a Matha in Northern Konkan (c.  AD 900–1053)’, in Ranabir Chakravarti (ed.), Trade in Early India, New Delhi: OUP India, 2001, p. 267. 44 The extensive series of small silver coins called ‘Dammas’ was first discussed by Tye. See R. Tye, ‘Dammas, Daniqs and Abd al-Malik’, Oriental Numismatic Society Newsletter, 148: 7–10, 1996. 45 https://www.livehistoryindia.com/cover-story/2019/09/22/sanjan-diggingdeep-into-history (accessed on 2.2.2019). 46 Ibid. 47 G. Jantakal, B. Kapadia, and P. Jinjuvadiya, , History and Coinage of the Rashtrakutas, Nashik: IIRNS publications, 2019. 48 A. Appadurai, ‘Introduction: Commodities and Politics of Value’, in Arjun ­Appadurai (ed.), The Social Life of Things: Commodities in Cultural Perspective, Cambridge Studies in Social and Cultural Anthropology, Cambridge: CUP, 1986, pp. 3–63. 49 R. Darley, ‘Self, Other and the Use and Appropriation of Late Roman Coins in Peninsular India (4th–7th Century CE)’, in HimanshuPrabha Ray (ed.), Negotiating Cultural Identity: Landscapes in Early Medieval South Asian History, New York: Routledge, pp. 60–84, 2015. 50 E. Smagur, ‘From Coins to Bulla: A Cultural Response to the Influx of Roman Denarii into India’, Numismatic Digest, 42: 63–78, 2018. 51 M. Taussig, Mimesis and Alterity: A Particular History of the Senses, New York: Routledge, 1993.

Taylor & Francis Taylor & Francis Group http://taylora ndfra ncis.com

INDEX

Note: Page numbers followed by “n” denote endnotes. Abbasid (AH132-656/CE749-1258) 7, 74n42, 90, 240, 244, 249, 250–251, 257, 258 Abhara, Captain 81 abhaya-mudrā 4 Aden 15, 18n12, 57, 82, 84, 85, 87, 94n47, 94n53, 143n26, 178; Gulf of ~ 57, 84 Al Baleed 6, 175–176 Aleppo 240 Alexandria 15, 60, 61, 66, 72n21 Al-Ḥawrā’ 83–85, 87, 90, 91 Al-Hurayda (temple) 178, 179 al-Idrīsī 91, 96n75 al-Istakhrī 91n1, 245 Africa(n) 7, 14, 18n20, 36n48, 36n64, 68, 69, 73n36, 79, 86, 87, 94n49, 99, 101n5, 103–105, 137, 144n40, 178, 236n106; East ~ 6, 14, 56, 62, 63, 68, 69, 73n37, 78n89, 82, 84, 86, 99, 123n5, 137, 142n11, 176; North/ Northeast ~ 7, 69, 87, 103, 104, 105, 142n7, 144n40 Ajā’ib al-Hind (or Kitāb ᶜajā’ib al-Hind, Marvels of India) 16 Aksumite 5 al-Biruni 8, 166n34 al-Jār 83, 84, 85, 88 al-Masᶜūdī (d. AH345/CE946) 81, 82, 92n11, 93n15, 93n27, 245 al-Muqaddasī (al-Maqdisī) 79–94 al-Maqrizi 96n74, 149, 164n17

al-Qulzum (Ancient Clysma) 82, 83, 85, 87–88 Al-Quṣayr (Myos Hormos) 84, 91 al-Rashid, Harun 257–258 amphorae 132, 133, 135, 141 amṛta 23–25, 27, 28 Anantaśāyin 200, 201, 206n70; see also Vishnu Anantaśāyin Andaman 14, 99 Angkor 3, 10n9, 172, 173, 174n16, 191–199, 200, 204n18, 204n22, 204n23, 205n46, 205n49, 205n50, 206n58, 206n67, 206n70, 207n72, 211, 212, 217–221, 227, 230, 232n27, 232n29, 233n36, 233n37, 233n40, 237n130 Angkor Thom 194, 200, 217 Angkor Wat 3, 198, 199–200, 206n70, 211–212, 215, 217, 218, 219, 227, 232n27, 232n28 añjali mudrā 255 Anuradhapura 134, 143n30 Arabian/ peninsula 2, 9, 10n23, 17n11, 61, 62, 63, 74n41, 79, 80, 82–89, 91, 94n47, 95n63, 95n68, 101n7, 147, 181–182, 187, 188n5, 188n7, 191n25, 240 Arabian Sea 57, 62, 63, 251, 255 Arabic 2, 7–8, 15–16, 17, 63, 80, 87, 94n18, 94n27, 96n61, 151, 158, 159, 163, 164n10, 167n20, 167n34, 181, 169n48, 169n59, 177, 189, 248, 257

264 Index

Aramaic 6 archaeobotanical 98 areca nut 99 Arikamedu 2, 127, 133–135, 137–138, 140–143, 141n1, 141n2, 141n5, 143n26, 143n27, 143n30, 144n46 ash mound 108 Ashoka(n) 2, 14, 134 astronomy/studying stars/stellar navigation 7, 16, 94n45 Asura 23–24, 26, 28, 34n30, 211, 212, 217–218 Aurobindo, Sri 19, 20, 33n3 Australia 17, 237n136 Austronesian 73n37, 98, 222, 230, 231, 238n138, 239n154 Axumite 6, 69 Ayutthaya 201–202, 213, 217, 232n29, 233n35, 237n130 Bactria(n) 6, 7, 11n27 Baghdad(i) 7, 157–158, 160, 162, 240, 244, 245 Balkh/Balkhī 7, 8, 11n29 Barmakid 7, 11n29 Bay of Bengal 3, 8, 11n35, 11n40, 14, 15, 57, 62, 74n38, 74n40, 98, 100n2, 101n8 Ba Qutfah temple 178, 179 Bamiyan 7 Ban Chao, General 61 Ban Don Ta Phet 4, 231, 239n152 barrel drum 211, 219 Basra 82, 240 Bayon 193, 201, 205n50, 206n63, 206n64, 207n82, 217 bead(s) 8, 99, 100, 106, 128–132, 136– 140, 141n2, 142n11, 142n14, 143n38, 144n40, 144n41, 144n43, 144n45, 144n48, 145n55, 186 bead making/trade 101, 128, 137–140, 141n2, 142n11, 144n14 Belitung shipwreck 63, 74n42, 74n43, 230 Berenike 2, 6, 10n17, 56, 73n35, 84, 98, 127, 128, 132, 134, 136, 140, 141n5, 142n7, 142n21, 144n42 beryl 138 Bhāgavata Purāṇa 23, 34n22, 35n45, 215, 218 Bharata 3, 170, 171 bheri (kettledrum) 225, 226 bin Qasim, Muhammad 243 Black and Red Ware (BRW) 132, 133, 134, 143n22, 187

boat design/construction: building traditions 15, 194, 195, 205n36; bundle rafts 192; dugout 192–196; iconographic 174n13, 193, 196, 198, 203, 204n18, 207n70; plank stitched/ built 82, 192, 195, 205n36, 206n63; rituals (apotropaic) 194, 196, 197, 200, 202, 203; zoomorphic 191, 193, 197–198, 200, 201, 202, 203 boat remains (excavated): Angkor Thom boat 194; Kra Raleung boat 194 boat type (hawārī) 14, 17n6 Book of Curiosities 91 Book of Routes and Provinces 15–16, 91n1 Book of the Wonders of India 92n13, 92n14 bossed gongs 209, 211–217, 218–220, 224, 226–229, 232n31 Brahmagiri 14, 110, 122n10, 123n18, 123n19 brahman(a)/ic/brahmin 9, 39, 40, 44–46, 49, 54n35, 54n36, 100, 209, 256 Brahmi 2, 6, 7, 52, 99, 187, 257 Brantas River 222 Bṛhad-kathā 23 bronze figure 187 bronze plaques 184 Brunei 153, 154, 156, 166n38, 215, 227 Buddha: relics/cult 4, 5, 6; wooden image 4, 10n9 Buddhist Cosmopolis 2 Buddhist texts 5, 14, 148, 166n41 Buddhist mendicant 44, 49, 50 burial: assemblage 113, 114; chambers 114, 119; types/classes 112, 113–114; see also mortuary Burmese gold letter 155–156 Buzurg Ibn Shahriyār, Captain 81, 92n13, 92n14, 94n45 Cagayan 224, 225, 236n104 Cairo (Fustat) 15, 87, 94n41, 146, 148, 149, 157, 162, 163n3, 163n4 Cairo Geniza 3, 15, 148, 164n8 Cambay (Khambat) 161, 170 capstones/capping stones 113–114, 116, 117 caravanserais (wakālas) 85, 89, 90 ceramics 8, 15, 56, 74n42, 99, 106, 111, 124n46, 124n47, 127, 129, 132–136, 138, 140, 213, 227, 232n31, 259n3 Chalukya 244, 250 Cham 208, 220–221, 228, 232n31, 234n60, 234n61

Index  265

Cham temples 221 Chao Phraya River 202 China 8, 16, 57, 59–70, 73n27, 73n34, 75n57, 77n76, 77n80, 77n84, 77n87, 78n90, 81, 87, 90, 92n12, 94n60, 99, 137, 147–148, 176, 207n81, 213, 221, 224, 225, 228, 229, 230, 231, 237n132, 238n147, 238n148, 245 China and the Roman Orient 57, 71n5, 72n21 Chinese paper 61, 148, 153 Chinese silk 60, 62, 67–68, 77n79 Chinese texts 57, 58–59, 67, 70, 75n51 China Sea 57, 61, 62, 70, 74n40, 82, 99, 100n2 Chinchani copperplate grants 255, 256, 260n39 Cholas 244 Christian(ity) 5, 229, 238n142, 241–242 churning of the ocean (samudramanthan) 2, 3, 19, 22, 23–26, 28–29, 32, 33, 34n29, 35n41, 51, 171, 212, 218 cists 104–105, 107, 112–116 coastal shrines/temples 5, 10n6, 170, 175, 176, 183 coins: Abbasid 249, 250, 251, 257–258; Byzantine 241–242, 243, 255, 259n6; Chinese 213; dinars 16, 249, 250, 251, 252, 254, 255, 258, 259, 260n35; dirhams 243, 245–246, 259n10, 259n11; Garuda-type 247, 248, 249, 251, 252, 253, 255, 257; Hadrami 180; hoard 133, 144n50, 247; horsemantype 248, 250, 251–252, 253–254, 258; imitation/forgeries/local copies 130, 131, 139, 140, 144n50, 246–247, 251, 252, 258, 259, 260n29; Indian dynasties 246; Islamic 242, 256, 259n9; Kārshāpaṇa 139; Mauryan 139; pre-Islamic 242; punch marked 139; Rashtrakuta 246–250, 251–253, 258–259; religious imagery 170, 173; Roman 3, 62, 130, 133, 139–141, 145n57, 258–259, 260n29, 261n49; Sasanian 240–242, 243, 255, 259n7; South Indian (Elliot Collection) 245, 251, 252, 253; Sri Lankan 139; Tahiriya/Tahirid 245–246 coin moulds 139 compass card 16 conch (shell) 211, 212, 217, 218, 219, 226, 237n133, 246–248 copperplate charters/grants 4, 247, 255, 256, 257 Coromandel coast 134

Coptic 149, 156, 159 cosmology 14, 19, 20, 76n57 cosmos 14, 21, 22, 26, 27, 31 cowries 139 cymbals 210, 211–212, 217, 218, 219–221, 226, 228, 230, 233n32, 237n129 Damascus 240 Dambadeniya 148, 155 dangers in sea/river voyages 37, 39, 45, 46, 47, 51, 92n12, 173, 194, 195, 196, 202 Daoism 65, 68, 75n52 Da Qin 2, 56–61, 63, 64–70, 70n2, 72n18, 72n19, 73n27, 77n75, 78n90 Dhofar 6, 10n23, 175, 176, 187n1, 188n6 dinars 16, 249, 250, 251, 252, 254, 255, 258, 259, 260n35 dirhams 243, 245–246, 259n10, 259n11 dolmens 104–106, 117–119, 121 Dvaravati Water Festival 199, 202 dvīpa 14, 38, 40, 43 Ed-dur 9 Egyptian 6, 56, 84, 86, 88, 89, 90, 91, 93n28, 95n63, 95n72, 96n76, 99, 128, 132, 137, 141n5, 142n7, 142n21, 148, 149, 160, 165n19, 167n62, 168n66 epistemology 9n4, 20–22, 24, 29, 106 eunuch 16 Fan Ye 58 Fārān 83, 84, 90 Fāṭimid (AH297-567/CE909-1171) 90 feasting 113, 115, 117, 118, 184, 186, 187 figurine 180, 187 Fine Grey Ware 134, 142n6, 143n29, 143n30 Foucault, Michel 21, 22, 27, 33n5, 33n10, 34n14 frankincense 68, 79, 175–177, 181, 185–187 Funan 56, 57, 73n27 Fusṭāṭ see Cairo gamelan 223, 235n84, 235n90, 236n111 Gan Ying 57, 58, 59, 61 garuda 198, 205n52, 247, 249, 257 garuda-type coinage see coins garum 133 Geniza see Cairo Geniza glass: bangles 138, 144n44; beads 136–137, 138, 141n2, 141n4, 144n40, 144n45, 144n48, 186; palace 65; wares

266 Index

56, 62, 65, 67, 74n38, 77n75; working/ making 138, 140 Godavaya 128, 144n51 Go Thap (Vietnam) 4 Graeco-Roman 60, 63, 67, 68, 69, 89; see also Roman Greek: Indo-Greek 259; polity 7; script 6, 89, 149, 156, 159; sites 91, 130; speaking people (yavanas) 63, 69, 70n2, 72n17; texts 2, 7, 15, 74n37, 79, 84, 89, 91, 93n2 Grey Ware 134 groundwater/wells 80, 85, 87, 90, 185 Gulf of Oman 82 Gurjara-Pratiharas 244, 250, 260n13 Hadramawt 5, 6, 177, 179–184, 186, 188n7, 188n10, 189n33, 189n41, 190n51 Haixi 60, 66, 72n22 Hanoi 4, 232n23, 233n47 Harappa/ Harappan 15, 132 hawārī (boat type) 14, 17n6 Hellenistic 76n61, 133 Herodotus 72n15, 72n17, 89, 94n58 Ḥijāz 85, 87, 88, 90, 95n69 Hindu: cosmology 19, 200; deities 4, 10n14, 193, 218; epics 215; mythology 24, 33n1, 35n40, 200, 206n57, 215, 218; temple(s) 4, 170, 174n14, 230, 233n36; traditions 202, 203 Hinduism 4, 8, 34n29, 229 Hire Benakal 103, 105, 107, 117, 118, 123n15 Historia Augusta 61 historian/geographers: Greek 3–4, 87, 89; Muslim/Arab 8, 79, 81, 85–86, 89, 90, 91, 91n1, 245, 257 Hoq cave 2, 6, 10n20 Hormuz 157, 158, 162 horse trade 244, 260n15 Hou Hanshu 57–60, 65, 66, 67, 68, 69, 71n6, 71n11, 71n13, 72n14, 72n19, 76n58, 78n90 Hoysala temples/sculpture 209, 227, 228, 231n5 Ibn al-Furat 149, 152, 153, 155–156, 157, 159, 163n10, 163n16, 165n20, 166n35 Ibn Hawkal 245 Ibn ͨAbd al-Zahir 149, 150, 152, 157, 159, 164n13, 166n34, 167n59 Ibn Khurradādhbih 16, 90, 91n1, 245 Ibn Shahriyār (Buzurg) 81, 92n13, 92n14, 94n45, 256

incense burner 175, 179, 181, 182, 183, 186, 187n1, 188n2, 190n57 incense route 90 Indo-Pacific beads 136–137, 144n40, 144n45 inscribed potshards 100 inscriptions/texts: bilingual nagari-kawi 225; bilingual Tamil-Chinese 16; Cham/Old Cham 208, 220, 228; kawi 225; Kharoshti 6, 52; KharoshtiBrahmi 52; Khmer/Old Khmer 172, 206n70, 208, 210, 219, 220, 223, 228, 233n24; Kufic 249–251, 253, 254, 256; nagari/proto-nagari 225, 246, 248, 252; Old Balinese 210, 221, 223, 225, 228; Old Javanese 3, 49, 171, 210, 221, 222, 223, 225, 228; on plaques 184,; Pali 5, 210, 221; Sanskrit 2, 4, 7, 17, 171, 210, 225, 229; Sdok Kak Thom 210, 232n23; Siwagraha 172; Tamil 16; Tamil Brahmi 99, 187 Iron Age settlements 107, 108, 124n38 Islam 8, 11n27, 18n14, 36n57, 79, 81, 84, 86, 89, 90, 92n1, 92n8, 92n10, 168n64, 173, 174n17, 240, 241, 242 Island/ ‘island factor’ 2, 5, 6, 9, 10n19, 14–16, 17n5, 19, 38–47, 50, 51, 53n5, 53n18, 54n24, 54n28, 54n33, 69, 83, 84, 86, 95n63, 98, 128, 134, 139, 150, 159, 170, 175, 183, 201, 224, 226, 227, 231 ivory 61, 68, 138, 154–155 Jābir 83, 84, 90 Jade 65, 72n23, 131 Jain 4, 33n7, 100 Jambudvīpa 14, 19, 53n4 Jamnagar 14 Jew/Jewish 15, 16, 148, 162 Jiaozhi (Vietnam) 56, 57 Judaist 5 Kadabakele 109 Kafas 186 kaivartas (fishing communities) 48 Kalinga 55n59, 210, 225 Kannauj 244, 250 kaṃsa/kangsa/gangsa (bell-metal) 208–210, 219–221, 228, 231n2 kangsatala/kansatala 210, 219, 220, 232n19, 232n20 Kanheri 6 karṇadhāra (helmsman) 40, 45, 46, 50 Kataha (Kedah) 38, 43, 45, 46, 47, 53n5

Index  267

Kathāsaritsāgara (KSS) 2, 23, 37, 39, 44, 48–49, 53n1, 53n6 kawi 225 Kakawin/Kekawin 222, 235n80, 235n81 Kelabit (Sarawak) 130, 142n14 Kharava 13, 17n2 kharoshti see inscriptions Khin Ba (Myanmar) 5 Khom 5 Khmer 173, 174n16, 205n49; boatbuilding traditions 194–198, 199, 203, 204n23; ceramic culture 232n31; language/inscriptions 206n70, 210, 219, 223, 228, 232n21, 232n24, 233n44; music heritage 208, 210, 215, 219, 220, 228, 230, 233n40, 234n53 Khor Rori 3, 9, 175, 176, 177–179, 181–187, 188n5, 188n24, 189n50, 190n56 Kirinde 128 kitāb (letter) 152, 157 Kitāb ᶜajā’ib al-Hind (The Book of the Wonders of India) 81, 92n13, 92n14, 94n45 Kitab ul-Masālik w’al-mamālik (The Book of Routes and Provinces) 91n1, 245 Klong Thom 137 knowledge broker 97 koki tree (hopea odorata) 194, 195 Kra Raleung boat 194 Kuhn, Thomas 22, 32, 34n15, 34n16, 36n71 Lakshadweep islands 14 Lanka (Sri Lanka) 14, 15, 16, 44, 47, 62, 72n25, 74n39, 99, 127, 130, 133, 134, 135, 137, 139, 140, 144n50, 148, 150, 152, 158, 162, 165n30, 212, 219, 238n145 lapis lazuli 68, 163n1 Latin 74n37, 79, 89, 252 Laxmi (goddess of wealth) 51 Leukē Kōmē 91, 95n68, 95n73 Levi-Strauss, Claude 24, 26, 31, 34n19, 34n28, 35n43, 36n59 Liangshu (Book of Liang) 56, 57, 58, 61, 69, 70 Lijian 60, 66 Luzon 210, 220, 224, 225, 226, 227, 228, 229, 235n93, 236n106 Machilipatnam 3 magico-ritual/tantric practices 42, 48, 49

Mahābhārata 14, 17n9, 23, 34n22, 35n34, 170, 198, 206n53, 212, 215, 218, 230 Mahavira 21 Majapahit 222 makara 198, 200, 205n51, 206n56, 206n57, 206n70 Makaynun 179, 183, 185, 189n41 Makran (coast) 14, 243 Mamlūk 84, 144n40, 146–147, 148, 149, 153, 155, 156–160, 161–162, 163n3, 167n64 Mamlūk chancery/diplomacy 146, 149, 151, 156, 158, 159, 160, 163n4, 164n11, 168n64, 168n66 Mamlūk Islamic period (AH7–8/CE13– 14) 85, 156, 160 Mandvi 14, 17n2 Mangalore 16 Mantai 137 Maski 2, 103, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 120, 122n10, 123n19 Maski Archaeological Research Project (MARP) 109–113, 115, 116, 117, 124n41, 124n45, 124n47, 125n50, 125n54 Mauryan 2, 14, 127, 139 Mauryan rock edicts 2, 14 Mayfaan (temple) 184, 185 medicine 7 megalithic: cists 104, 105, 106, 115, 116, 117, 119; Iron Age sites 105–106, 107, 108, 111, 112, 115, 116, 122n10, 123n30, 124n32, 124n34, 124n35, 124n38, 125n57, 126n75; monuments 101, 102, 104–105, 106–107, 109, 110, 112, 119–120, 121, 122; mortuary/ burial practices 103, 106, 109–111, 112–117, 118, 120, 123n19, 123n21, 123n23, 125n59; “period” 105; stone circle 104–105, 107, 113, 116, 119, 227; types/forms 105, 106, 112–114, 116, 122n10, 122n11 Mekong: delta 4, 10n9, 56, 194; river 4, 196 menhirs 104 Meru 14, 19, 27, 200 Ming 16, 58, 215 Mleiha 9, 187 Mohinī 25, 35n46 monochord 219 monsoon 13–14, 17n1, 57, 62, 150 mortuary architecture 106, 117 mortuary/burial practices/rituals see under megalithic

268 Index

Muallim nī pothī 16 musical instruments 172, 208–210, 212, 217, 218, 219, 220, 223, 228–231, 231n4, 231n5, 231n8, 232n14, 232n15, 233n40, 234n52, 234n67, 235n73, 237n129 Muslim traders/merchants 61, 63, 82, 90, 150, 245 Myos Hormos 56, 84, 91 myrrh 68, 175, 190n51 myth(ology/ic): Chinese 58, 64–65, 68, 69; Indic 19, 20, 22–29, 31, 32–33, 33n1, 34n29, 35n40, 35n41, 119, 199, 206n57, 215, 218, 255; Khmer 193, 198, 200, 202, 205n49, 205n51, 205n52 Nabatean/Nabataean 84, 93n28, 95n68, 95n72, 138 nāga 198, 201, 206n57, 206n70 Nagarjunakonda 3, 4 Nakhon Pathom 4 Natyaśāstra 3, 170–171, 173n10 nautical iconography 3, 192–193, 197 nautical charts/guides/manuals 8, 13, 16, 18n17, 34n17, 82, 86, 87 nautical technology/design 191–192, 193 navigation techniques/traditions 16, 81, 86, 87, 191, 195, 196, 203, 205n36 Niccolo De’ Conti 161 nishada 40, 49 Northern Black Polished Ware (NBPW) 134 oasis 175, 176, 179, 180 oceanography 97–98 Oceanus Orientalis Indicus 13 Oc-Eo 137 Oman 15, 16, 74n41, 82, 175–176 ontology 20–22, 23–24, 27, 29, 34n14, 34n20, 126n72 onyx 100, 139 oral transmission 3, 5, 23 orpiment 67 orthostats 116 Oxus (Amu Darya) 7 Pagan 5, 16, 236n101, 238n149 Palas 244 Palembang 148 Pali 5, 17n7, 210, 221 Pallavas 247 palm leaf documents/manuscripts 149, 150, 152–155, 156, 161, 162, 163, 166n6, 167n23, 167n30, 168n37

palm leaves: Palmyra palm (Borassus flabellifer) 150; Talipot palm (Corypha umbraculifera) 150 Pandya king 148 paper 61, 142, 147, 148, 153, 157–160, 161–162, 163, 164n7, 167n61, 230 paper-making 147, 159 Parthia/n 57, 65, 77n80 pathsala 7 pearls 68, 74n38, 74n42, 130, 142n12, 159, 162 Pegu 222 pepper 53n16, 73n35, 98–99 Periplus Maris Erythraei 15, 60, 63, 74n38, 75n46, 75n47, 77n81, 84, 89–90, 93n30, 94n57, 95n68, 134, 135, 138, 187 Persian 2, 8, 63, 69, 82, 86, 94n44, 166n34, 257 Persian Gulf 2, 15, 61, 68, 74n41, 82, 91n1, 95n73, 187, 240, 244, 255, 258 phenomenology 21 pilgrim(age) 5, 9, 79, 80, 82, 84, 85, 88, 89, 90, 95n63, 95n67, 149, 169, 177, 183, 186, 187, 193, 199, 233n36 pirates/piracy 83, 84, 93n27, 93n28, 260n15 pit burials 106, 113, 116 plank built/stitched 48, 82, 84, 192, 193, 195, 205n36, 206n63 Pliny the Elder 178 Porbandar 14 portolan charts 86 pothīs 16, 155 pramāṇa 20, 33n7 Prambanan 53n16, 171, 172, 173n6, 173n7, 174n9, 174n10, 190n6, 190n7, 190n10, 230, 233n36 Pratiharas 244, 250, 260n13 Preah Khan 197 Ptolemaic period (mid 3rd to 2nd century BCE) 132 Ptolemaic 10n17, 66 Ptolemy (Claudius) 4, 60, 62, 63, 72n24, 72n25, 75n47, 87 Pundra (Paundra) 38, 45, 52 Purāṇas /Purānic 2, 14, 19, 23, 33n1, 34n22, 38, 53n4 Qalawun, al-Mansur 148–149, 152, 155, 156, 157, 159 Qin Lun 56, 61 Quanzhou 16 Qur’an(ic) 163, 167n60, 167n62, 259n9

Index  269

radiocarbon dating 4, 112, 113, 114–116, 123n19, 143n30, 207n72 rahnāma (sailing manual) 86 Raichur (District) 2, 103, 124n41 Ramanujan, A.K. 27, 31, 35n34, 35n37, 36n52 Rāmāyaṇa 23, 27, 28, 34n22, 35n23, 170, 171, 173n4, 173n5, 215, 215, 222, 230, 235n80, 238n145 Rashtrakuta 244–250, 251, 252–254, 255, 256–257, 258–259, 260n16, 260n39, 261n47 Rasulid 148, 157 Raybun 179, 180, 183, 184–186, 188n16, 189n42, 190n51 reachisey 198, 200, 206n57, 206n59, 206n70 realgar 67 Records of the Three Kingdoms (Sanguo zhi) 58 red coral 67, 68, 76n74 Red Polished Ware (RPW) 134 Red Sea 2, 5–6, 10n17, 13, 14, 15, 56–57, 61, 62, 71n3, 79, 80, 82–85, 87, 88, 89, 91, 91n1, 93n21, 93n37, 94n54, 95n63, 95n68, 95n71, 95n72, 95n73, 98, 128, 141n5, 142n7, 95n21, 148, 157, 187 relic/relic chamber/reliquary 4, 5 Rennell, James 98 rhinoceros horn 61, 68 riḥla (travel narrative) 80 Rishahr 240 Rinan (Annam) 56, 57 ritual banquets/meals 179, 186, 187, 190n53, 190n54 ritual objects 5, 170, 182, 187 rock art 111 Roman 2, 3, 6, 59, 60, 63, 64, 66, 69, 70, 71, 79, 84, 90, 91, 124n37, 128, 135, 136 Roman: coins/issues 3, 62, 130, 131, 133, 139–141, 144n50, 145n57, 258–259, 260n29, 261n49, 261n50; Empire 56–59, 61, 63, 64–65, 66–69, 70n3, 72n6, 72n12, 72n13, 73n18, 73n19, 73n25, 74n37, 76n58, 76n59, 76n71, 76n73, 76n74, 77n76, 77n79, 127; glass(ware) 63, 66, 68, 77n75, 128, 141n4, 143n4, 144n40, 144n46; port 10n17, 84, 91, 128; trade/merchants 2, 56, 57, 62, 64, 70, 70n2, 70n3, 72n25, 73n34, 73n35, 73n37, 74n39, 89, 127, 134, 141n1, 143n30 Romano-Chinese contacts 57, 58, 59–60, 64, 71n5, 71n6, 71n8, 71n10, 71n12, 71n13, 72n18, 72n19, 72n21, 72n22,

73n33, 75n50, 76n59, 77n85, 78n90, 78n91 roneat (xylophone) 212, 219 Rouletted Ware 133, 134, 141n5, 142n6, 143n28, 143n29, 187 Ruhuna report 140 Sabaeans 84 Sabota 178–179; see also Shabwa sacrificial animals 54n36, 55n41, 180, 185, 227 saffron 67 sāgara (ocean) 37, 38 Salalah 6, 176, 188n5 Salsette 6, 10n19, 14, 17n5 samudra(m) (ocean) 19, 26, 36n66, 38 samudra-manthan 19, 171 Sangam 69 Sanguo zhi (Records of the Three Kingdoms) 58 Sanjan (Sindan) 240, 255–258, 259n3, 260n41 Sanskrit Cosmopolis 2 Sanskrit knowledge transmission 7–8, 166n41, 172, 208–210, 219, 221–222, 227–230, 237n132 sarcophagus burial 113, 115 Saurashtra 14, 17n2, 170 Scientism 20, 33n6 sea hazards/dangers 8, 37, 39, 45–46, 47, 51, 52, 81, 83, 84, 86, 90, 92n12, 173, 194, 195–196, 202, 205n46 Sdok Kak Thom inscription 210 Seres (“Silk People”) Sērikē or Sinai/ Thinai 60–61, 72n25, 75n49 Shabwa 178–179, 180, 182, 183, 184, 186; see also Sabota Shahr-i Zohak 7 Shaktideva 39, 40–42, 49–50, 51 Shashankapura 38, 53n6 shell 61, 68, 138, 181, 182, 186, 212, 218, 237n133, 246, 247, 248 Shilaharas 256 Shinto 9 Shipwrecks: Belitung 63, 74n42, 74n43, 230; Butuan 227, 232n31; Fārān 83, 84, 90; Gulf of Thailand 213; Jābir (Jā’iz) 83, 84, 90; Java Sea 213, 224, 227; Lingga 227, 228, 238n129, 238n133; Phu Quoc 213, 215; Pulau Bakau 228; Rang Kwien 213, 227; Red Sea 84, 95n63; Sattahip 213, 215; Tanjung Simpang 227 Shisr oasis 175, 176 Shubhatunga 246–248, 250

270 Index

Sihaladīpa 14 Silsilat al-tawārīkh (Chains of Narratives) 90, 94n60, 245 Sikotar Mata (Goddess) 6, 14 “Silk people” 60; see also Seres Silk Route/Road 61, 72n23, 72n25, 73n34, 74n40, 76n58, 77n80, 147, 165n26; maritime ~ 147 Simhala 38, 43, 46, 53n5 Sinai 84, 167n60 Sindbad the Sailor 16 Sindh 243–244 Siraf 90, 240 Śiva 4, 16, 23, 26, 35n46 Siwagraha inscription 172 Socotra (Usqūtra) 2, 5–6, 10n19, 10n20, 14, 17n5, 17n6, 69, 83, 84, 175 Song court 154 Song-shi 153, 156, 166n38 spindle whorls 181 śramaṇa 44, 50 Srivijaya 43, 53n16, 101n13, 147 stele 180, 184, 189n41 storage jars/vessels 113–114, 186 storax 67, 77n75 Sufi 8, 9 Sukhothai 5, 213, 215 Sulaymān al-Mahrī 86, 90, 245 Sumhuram (Khor Rori) 179, 180, 181, 182, 183–187, 188n5, 188n24, 189n28, 189n30, 189n37, 190n56, 190n57, 190n58 Surakarta 223 sutra 5 Suvarnabhumi/Suvarṇadvīpa 5, 38, 43, 45, 46, 48, 52, 53n5, 53n18, 54n28 swastika 255 Syn temples (Sayin temples) 178, 179, 180–182, 184, 185, 186–187, 189n41 Ta Chin river 4 Tagalog 224, 235n75 Tahirid 245–246 Tamil 16, 69, 77n88, 99, 127, 189, 209, 222 Tamil Brahmi 99, 187 Tāmralipta/Tāmralipti 38, 44, 45, 47, 52 tamsuk (land grants) 152, 153, 164n6 terracotta 99, 113, 133, 172, 187n1, 259 thalassology 29–31, 36n49, 36n54 Thucydides 89, 94n59 Tigris River 240 Tissamaharama 2, 127–129, 132–135, 136, 139, 140, 141, 142n6, 145n25, 145n35, 144n48, 144n51

Tokharistan 7 Tungabhadra River 99, 101n11, 122n11, 124n34 U Thong 4 underwater archaeology 1, 15, 90, 91, 93n31 UNESCO 30, 36n51, 175, 176, 177, 188n3, 237n121 vāhana (ship) 46 vahana (mount) 218 Vanga 52, 55n59 Veda(s) 23, 33n3, 35n40, 35n59 Vedic 19, 24, 26, 35n20 Veraval 14, 17n2 vidyādhara/vidyādharī 39–42, 43, 48, 49, 51, 53n12 Vietnam(ese) 2, 4, 10n10, 15, 56, 57–58, 69, 70, 70n2, 137, 194, 208, 213, 215, 220, 226, 227, 228, 230, 234n58, 234n61, 236n93, 236n114, 237n134 Vikramaditya 38, 43, 44 Vishnu 35n40, 50, 198, 199, 200, 201, 206n70, 207n72, 212, 217, 218 Vishnu Anantaśāyin 200–201, 206n70 Viṣṇu Purāṇa 19, 23, 33n1, 34n22 Wadi Dawkah 175, 176; see also Dhofar Waqwaq 159 water resources management/retention 80, 89, 90, 99, 109, 185, 201 Weilüe 57–60, 64, 65, 66, 67–70, 71, 71n6, 71n10, 71n11, 71n13, 72n14, 72n19, 75n49, 78n90 writing 2, 3, 5, 7, 9, 15, 58, 100, 101n13, 146–148, 149–152, 156, 157, 158, 159, 161, 162, 163n1, 163n2, 164n5, 164n8, 165n22, 166n34, 166n41, 228, 230 Yahya al-Barmaki 7 yānapātra(m) (ship) 40, 46, 47 Yanbuᶜ 83, 84, 85, 87, 88, 90, 91 Yavana (yona) 69, 77n88, 143n26 Yemen (al-Yaman) 17n6, 82, 87, 89, 95n68, 148, 163n1, 177–178 Yijing (also I-ching or I-tsing) 148, 164n7, 230 Yogyakarta 224, 235n90 Yu Huan 58, 71n13 Zheng He 16, 18n16 Zhou Daguan 193, 194, 204n21 Zoroastrian(ism) 241, 255