Constructing Kanchi: City of Infinite Temples 9789048550913

Citation preview

Constructing Kanchi

Publications The International Institute for Asian Studies (IIAS) is a research and exchange platform based in Leiden, the Netherlands. Its objective is to encourage the interdisciplinary and comparative study of Asia and to promote (inter)national cooperation. IIAS focuses on the humanities and social sciences and on their interaction with other sciences. It stimulates scholarship on Asia and is instrumental in forging research networks among Asia Scholars. Its main research interests are reflected in the three book series published with Amsterdam University Press: Global Asia, Asian Heritages and Asian Cities. IIAS acts as an international mediator, bringing together various parties in Asia and other parts of the world. The Institute works as a clearinghouse of knowledge and information. This entails activities such as providing information services, the construction and support of international networks and cooperative projects, and the organization of seminars and conferences. In this way, IIAS functions as a window on Europe for non-European scholars and contributes to the cultural rapprochement between Europe and Asia. IIAS Publications Officer: Paul van der Velde IIAS Assistant Publications Officer: Mary Lynn van Dijk

Asian Cities The Asian Cities Series explores urban cultures, societies and developments from the ancient to the contemporary city, from West Asia and the Near East to East Asia and the Pacific. The series focuses on three avenues of inquiry: evolving and competing ideas of the city across time and space; urban residents and their interactions in the production, shaping and contestation of the city; and urban challenges of the future as they relate to human well-being, the environment, heritage and public life.

Series Editor Paul Rabé, Urban Knowledge Network Asia (UKNA) at International Institute for Asian Studies, the Netherlands

Editorial Board Henco Bekkering, Delft University of Technology, the Netherlands; Charles Goldblum, University of Paris 8, France; Xiaoxi Hui, Beijing University of Technology, China; Stephen Lau, University of Hong Kong, Hong Kong; Rita Padawangi, Singapore University of Social Sciences, Singapore; R. Parthasarathy, Gujarat Institute of Development Research, Gujarat, India; Neha Sami, Indian Institute of Human Settlements, Bangalore, India

Constructing Kanchi City of Infinite Temples

Emma Natalya Stein

Amsterdam University Press

Publications Asian Cities 16

Cover illustration: Festival at Ekāmbaranātha Temple, Kanchi (October 2013) Photograph by Emma Natalya Stein Cover design: Coördesign, Leiden Lay-out: Crius Group, Hulshout isbn 978 94 6372 912 3 e-isbn 978 90 4855 091 3 (pdf) doi 10.5117/9789463729123 nur 693 © Emma Natalya Stein / Amsterdam University Press B.V., Amsterdam 2021 All rights reserved. Without limiting the rights under copyright reserved above, no part of this book may be reproduced, stored in or introduced into a retrieval system, or transmitted, in any form or by any means (electronic, mechanical, photocopying, recording or otherwise) without the written permission of both the copyright owner and the author of the book. Every effort has been made to obtain permission to use all copyrighted illustrations reproduced in this book. Nonetheless, whosoever believes to have rights to this material is advised to contact the publisher.

For Bob and all the Babus



Table of Contents

Acknowledgments 9 List of Illustrations

13

Note on Transliteration, Translation, and Illustrations

23

Introduction 25 All Streets Lead to Temples

An Ancient City Layers of Time Kanchi Known and Unknown 1 Sandstone and the City

Building Pallava-Kanchi (ca. seventh through ninth century)

From Brick to Stone (the Seventh Century) Sandstone Temples in the City (the Eighth Century) The Temples of Pallava-Kanchi Everywhere but Kanchi (the Ninth Century) Conclusion: Foundations Laid 2 Realignment

Kanchi in the Chola Era (ca. tenth through thirteenth century)

Orienting the Gods Pilgrimage and Processions From Ancient Village to Temple Town Local Style Conclusion: Urban Logic 3 The City and its Ports Part 1: KṢETRA The River Networks Over the Hills The Coast Part II: KṢATRA Kanchi in a Buddhist World The City and its Mirrors Conclusion: From Kanchi to the Sea

30 34 38 43 46 63 68 92 101 103 105 126 136 148 151 153 156 156 165 172 182 182 193 196

4 Kanchi Under Colonialism What Happened in Kanchi while those Towering Gateways Arose? Embattled Territory William Daniell’s Most Considerable Temple James Wathen’s Soaring View Henrietta Clive’s ‘Hindoo Gods and Monsters’ Colonel Colin Mackenzie’s Search for the Jains Surgeon George Russell Dartnell James Fergusson’s Downward Spiral Prince Alexis Soltykoff’s ‘City of Infinite Temples’ Conclusion: Plastered Pasts

199 205 208 222 225 229 234 242 245 253 263

Epilogue 267 The Living Temple

Encounter Expansion Continuation

267 268 271

Bibliography 275 Index 295

Acknowledgments My affair with the city of Kanchi began with a whirlwind day in the summer of 2011. It was late into my first tour of South India. I had stayed in Māmallapuram far longer than planned, enchanted by its coastal landscape, and was left with only one day to visit Kanchi before flying back home. I was out of time. Knowing little about the city, I did what most visitors would do and followed my guidebook to Kanchi’s five most famous temples. I was instantly struck by the vitality of the city. People and traffic flowed in all directions simultaneously, even more so than in other Indian cities. Temples dotted every corner and the end of every street, and the city squeezed itself between them. The diversity of Kanchi’s temples in chronology and condition – let alone their sheer abundance – was staggering. Some were archaeologically preserved monuments, while others were gleaming with sacred substances, resounding with music, and teeming with devotees. Past and present coincided in rare harmony. I saw that Kanchi’s popular epithet, ‘City of One Thousand Temples’, was no mere hyperbole. I am a New York City girl, but Kanchi’s vibrant urban landscape stole my heart. Fieldwork during 2013-2016 would have been impossible without the generous assistance of Kanchi’s priests, residents, tea sellers, monument attendants, and other keepers of local knowledge. A lucky return visit in January 2020, just before the onset of the COVID-19 pandemic, enabled me to reconnect with this amazing community. The Annadurai family requires special mention. I thank A. Prabu, A. Valavan, and their father, Annadurai, for introducing me to many of the sites discussed in this book. My warm thanks also to Narayanasami, head monument attendant of the Archaeological Survey of India’s Kanchipuram division, and to Mallai Dilip, for a decade of friendship that included many excursions to temple-sites near Māmallapuram. Thanks also to drivers Balakuru of Kumbakonam and Kumar of Pondicherry, who patiently entertained my many requests to go temple-hunting in remote villages across the Tamil landscape. Special thanks go to Aditya and Buvana Vaidyanathan for their invaluable friendship and for giving me a home in Chennai. My deep gratitude also goes to the Reddy family of Hyderabad, for their sponsorship of large-scale rituals in Kanchi and Tiruvannamalai, and for allowing me to observe and photograph these rituals up close. I owe a debt of gratitude to Dr. S. Barathi and the teachers at the American Institute of Indian Studies language center in Madurai for giving me the ultimate tool for navigating Kanchi: the language of Tamil.

10 

Construc ting K anchi

Research was carried out with generous financial support from the South Asian Studies Council and the Department of the History of Art at Yale, the Yale University Art Gallery, the Paul Mellon Centre in London, and a yearlong Junior Research Fellowship from the American Institute of Indian Studies (AIIS), as well as affiliation from the Institut Français de Pondichéry (IFP) and the École Française d’Extrême-Orient (EFEO). Many thanks go to the tireless staff and scholars in Pondicherry, especially Ramaswamy Babu, for sharing his knowledge of how to access hundreds of temples in Tamil Nadu, and to Valérie Gillet for asking hard questions and for her friendship in a difficult time. I completed the book thanks to two fellowships at the Freer Gallery of Art and Arthur M. Sackler Gallery, the Smithsonian’s National Museum of Asian Art, where I now serve as a curator of South and Southeast Asian art. Here I received critical support and encouragement from my colleagues Debra Diamond, Louise Cort, and Massumeh Farhad. Most of my maps were made with generous assistance from Dan Cole, GIS Coordinator and Chief Cartographer at the Smithsonian. The international community of scholars and friends to whom I am indebted is too numerous to list in full. My dissertation advisor, Tamara Sears, deserves special thanks for teaching me to pay attention to patterns in the built environment – patterns that form some of the core theses of this book. Committee members Vidya Dehejia, Tim Barringer, and Ruth Barnes each contributed important comments and suggestions at crucial moments in the project. A symposium organized by Subhashini Kaligotla in Garden and Landscape Studies at Dumbarton Oaks moved my thinking profoundly beyond the temple walls. This book also greatly benefited from conversations, conferences, and workshops with Frederick Asher, Andrea Acri, Rebecca Bloom, Osmund Bopearachchi, Crispin Branfoot, Greg Bryda, Jonas Buchholz, Jean-Luc Chevillard, Amanda Culp, Eric Gurevich, Ute Hüsken, Padma Kaimal, Katherine Kasdorf, Dipti Khera, Risha Lee, David Ludden, Philip Lutgendorf, Padma Maitland, Giulia Nardini, Michael Meister, Charlotte Schmid, Holly Shaffer, Peter Sharrock, Heeryoon Shin, and Caleb Simmons. I am extremely grateful to Paul van der Velde, Publications Officer at the International Institute of Asian Studies, for inviting me to publish with Amsterdam University Press, and to the Asian Cities Series Editor, Paul Rabé, for his enthusiasm and collegiality. Special thanks to Mary Lynn van Dijk, Assistant Publications Officer at the International Institute for Asian Studies, as well as to Saskia Gieling, Commissioning Editor, and Jaap Wagenaar, Production Editor, at Amsterdam University Press. I would like to

Acknowledgments

11

thank the two anonymous reviewers whose careful reading and thoughtful suggestions greatly enhanced this book. Finally, I would like to thank my parents, Bob and Marilyn, for inspiring my interest in Art History and for infinite other teachings. And my husband, Ari – forever my editor and forever my love.



List of Illustrations

Illustration 1 Illustration 2 Illustration 3 Illustration 4 Illustration 5 Illustration 6 Illustration 7 Illustration 8 Illustration 9 Illustration 10

Illustration 11

Illustration 12

Ekāmbaranātha Temple and Sannathi Street, Kanchi, seventh century – the present Photograph by Emma Natalya Stein Map of Kanchi Temples Map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution Kailāsanātha Temple, Kanchi, ca. 700-725 CE Photograph by Emma Natalya Stein Buddha, Kanchi Police Station, twelfth century Photograph by Emma Natalya Stein Sīteśvara Temple, Kanchi, tenth century Photograph by Emma Natalya Stein Festival at Kāmākṣī Ammaṉ Temple, Kanchi (July 2014) Photograph by Emma Natalya Stein Panel 4, Tāṉtōṉṟīśvara Temple, Kanchi, seventh century Photograph by Emma Natalya Stein Lakṣita cave-temple, Maṇṭakappaṭṭu, ca. 580-630 CE Photograph by Emma Natalya Stein Brick shrine in quarry area, Ārpākkam Photograph by Emma Natalya Stein Ekāmbaranātha Temple, Kanchi, Left to right: Roof of Kacci Mayāṉam shrine (tenth century), Pillared Hall (twentieth century), Gateway (seventeenth century), Vṛṣabheśvara Shrine (ninth century), Gateway (sixteenth century) Photograph by Emma Natalya Stein (Vṛṣabheśvara Shrine at far right) Archaeological Survey of India, A view of the tank from the east, Ekambreswaraswami Temple, Conjeevaram, 1897, photographic print, 21.1 × 25.6 cm Source: © The British Library Board, Photo1008/3[325] Panel 1-4 (and see Illustration 7), Tāṉtōṉṟīśvara Temple, Kanchi, panels seventh century Photograph by Emma Natalya Stein

26 27 28 32 33 37 44 48 50

52

53 59

14 

Construc ting K anchi

Illustration 13 Pillar base with dancing figures, Vīraṭṭāṉeśvara Temple, Kanchi, seventh century Photograph by Emma Natalya Stein Illustration 14 Kailāsanātha Temple, Kanchi, ca. 700-725 CE Photograph by Emma Natalya Stein Illustration 15 Lion-based pillars in prākāra, Vaikuṇṭha Perumāḷ Temple, Kanchi, ca. 730-795 CE Photograph by Emma Natalya Stein Illustration 16 Kinnara and Kinnarī, Kailāsanātha Temple, Kanchi, ca. 700-725 CE Photograph by Emma Natalya Stein Illustration 17 Mackenzie Collector, Plan of the Kailasanath Temple, Conjeeveram, ca. 1800, watercolor with pen and ink, 65.6 × 37.5 cm Source: © The British Library Board, WD754 Illustration 18 Sluice, Māmaṇṭūr, seventh century Photograph by Emma Natalya Stein Illustration 19 Coronation scene, Vaikuṇṭha Perumāḷ Temple, Kanchi, ca. 730-795 CE Photograph by Emma Natalya Stein Illustration 20 Reclining Vishnu among nāga stones, Vaikuṇṭha Perumāḷ Temple, Kanchi Photograph by Emma Natalya Stein Illustration 21 Dharmarāja Ratha, Māmallapuram, seventh century Photograph by Emma Natalya Stein Illustration 22 Piṟavāttāṉeśvara Temple, Kanchi, ca. 700-725 CE Photograph by Emma Natalya Stein Illustration 23 Mukteśvara Temple, Kanchi, eighth/ninth century Photograph by Emma Natalya Stein Illustration 24 Map of Pallava sites, ninth century Map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution Illustration 25 Sundaravarada Perumāḷ Temple, Uttiramērūr, ninth century Photograph by Emma Natalya Stein Illustration 26 Sabhā, Uttiramērūr, base ninth century Photograph by Emma Natalya Stein Illustration 27 Kanchi District Geological Map, blue showing sandstone below ground

62 64 65 69

70 75 82 84 85 87 89 93

94 95 96

List of Illustr ations

Illustration 28 Illustration 29 Illustration 30 Illustration 31 Illustration 32 Illustration 33 Illustration 34 Illustration 35 Illustration 36 Illustration 37 Illustration 38 Illustration 39 Illustration 40 Illustration 41 Illustration 42

Source: Geological Survey of India, Chennai Quarry field, Ārpākkam Photograph by Emma Natalya Stein The Axis of Access, Kamarajar Salai, Kanchi, from North Photograph by Emma Natalya Stein Map of temple orientations, Kanchi Map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution Aruṇācaleśvara Temple, Tiruvannamalai Photograph by Emma Natalya Stein Procession for the Goddess, Kamarajar Salai, Kanchi Photograph by Emma Natalya Stein Map of maṇḍapa locations in central Kanchi Photograph by Emma Natalya Stein Maṇḍapas, juncture of Kamarajar Salai and South Raja Veedhi Photograph by Emma Natalya Stein Bṛhadēśvara Temple, Thanjavur, ca. 1010 CE Photograph by Emma Natalya Stein Durgā Shrine at Kacchapeśvara Temple, Kanchi, tenth century Photograph by Emma Natalya Stein Jvarahareśvara Temple, Kanchi, twelfth century Photograph by Emma Natalya Stein Map of old and new main roads Map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution Map of temple locations, road from Uttiramērūr to Tirupati Map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution Vyāghrapureśvara Temple, Tiruppulivaṉam, eleventh century Photograph by Emma Natalya Stein Goddess buried near Agnīśvara and Mātṛkā Temples, Kaḷakkāṭṭūr Photograph by Emma Natalya Stein Veṅkaṭeśvara Temple, Tirupati

15

99 104 105 107 110 112 113 115 117 118 121

122

124 125 127

16 

Illustration 43 Illustration 44 Illustration 45 Illustration 46 Illustration 47 Illustration 48 Illustration 49 Illustration 50 Illustration 51 Illustration 52 Illustration 53 Illustration 54 Illustration 55 Illustration 56

Construc ting K anchi

Photograph by Emma Natalya Stein Procession at Varadarāja Perumāḷ Temple, Kanchi Photograph by Emma Natalya Stein Pillar, Gandhi Road, Kanchi, sixteenth century Photograph by Emma Natalya Stein Map of a processional route, Kanchi Map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution Chola bronzes at Gōmuktēśvara Temple, Tiruvāvaṭutuṟai, tenth century Photograph by Emma Natalya Stein Map of sites in Ārpākkam Village Map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution Ayyaṉār, Tiruvālīśvara Temple, Ārpākkam, eleventh century Photograph by Emma Natalya Stein Festival at Ekāmbaranātha Temple, Kanchi, October 2013 Photograph by Emma Natalya Stein Inscriptions, Ulakaḷanta Perumāḷ Temple, Kanchi, eleventh century Photograph by Emma Natalya Stein Anēkataṅkāvatīśvara Temple, Kanchi, ca. 1070 CE Photograph by Emma Natalya Stein Interlocked chain motif, Anēkataṅkāvatīśvara Temple, Kanchi, ca. 1070 CE Photograph by Emma Natalya Stein Map of temple-sites from Kanchi to the sea Map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution Veṅkateśa Perumāḷ Temple, Tirumukkūṭal, ninth century Photograph by Emma Natalya Stein Lion-based pillar, Veṅkateśa Perumāḷ Temple, Tirumukkūṭal, ninth century Photograph by Emma Natalya Stein Sculptures near pond, Kumaravāṭi, ninth century and later Photograph by Emma Natalya Stein

130 132 134

135 137

137 142 144 149 150 154

157 158 161

List of Illustr ations

Illustration 57 Durgā, Kumaravāṭi, ninth century Photograph by Emma Natalya Stein Illustration 58 Ceṇpakeśvara Temple, NattamParamēsvaramaṅkaḷam, tenth century Photograph by Emma Natalya Stein Illustration 59 Vedagirīśvara and Bhaktavatsaleśvara Temples, Tirukkaḻukuṉṟam Photograph by Emma Natalya Stein Illustration 60 Caṅku Tīrtham from Vedagirīśvara Temple, Tirukkaḻukuṉṟam Photograph by Emma Natalya Stein Illustration 61 Vīṇādhara-Ardhanārīśvara, Vedagirīśvara Temple, Tirukkaḻukuṉṟam, seventh century Photograph by Emma Natalya Stein Illustration 62 Bhaktavatsaleśvara Temple, Tirukkaḻukuṉṟam, ninth century and later Photograph by Emma Natalya Stein Illustration 63 Dakṣiṇāmūrti, Bhaktavatsaleśvara Temple, Tirukkaḻukuṉṟam, ninth century Photograph by Emma Natalya Stein Illustration 64 Henry Salt, Pagodas at Trinchicunum, Plate 11 in ‘Twenty-Four Views in St. Helena, the Cape, India, Ceylon, the Red Sea, Abyssinia and Egypt’, 1804, wash, 46.8 × 62.2 cm Source: © The British Library Board, WD1305 Illustration 65 Coastal Māmallapuram at sunrise, Shore Temple, ca. 700-725 CE Photograph by Emma Natalya Stein Illustration 66 Great Relief during 2015 flooding, Māmallapuram Photograph by Mallai Dilip Illustration 67 Shore Temple, Māmallapuram, ca. 700-725 CE Photograph by Emma Natalya Stein Illustration 68 Quarried Boulders, Māmallapuram Photograph by Emma Natalya Stein Illustration 69 Ascetic and the Birth of Pallava Relief, Vaikuṇṭha Perumāḷ Temple, Kanchi, ca. 730-795 CE Photograph by Emma Natalya Stein Illustration 70 Three Buddhas, Paḷḷūr, eleventh/twelfth century Photograph by Emma Natalya Stein Illustration 71 Map of Buddhas and Buddhist remains in Kanchi

17

162 163 166 167 168 169 170

171 173 174 175 177 181 188 189

18 

Illustration 72 Illustration 73 Illustration 74 Illustration 75 Illustration 76 Illustration 77

Illustration 78

Illustration 79 Illustration 80

Construc ting K anchi

Map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution Buddha, Karukkiṉil Amarntavaḷ Ammaṉ Temple, Kanchi, eleventh/twelfth century Photograph by Emma Natalya Stein Buddha, formerly in Kāmākṣī Ammaṉ Temple, Kanchi, sixth century; present location: Government Museum, Madras Photograph by Emma Natalya Stein Temple-cities that flourished ca. eighth-thirteenth century Map by Emma Natalya Stein Bagan, Myanmar Photograph by Emma Natalya Stein Sir Joshua Reynolds, Lady Henrietta Antonia Herbert, Countess of Powis (1758-1830), 1777-1778, oil on canvas, 55¼ x 44¼ in Source: Powis Castle, © National Trust Images, 1181064 Henry Dixon, View Looking Towards an Unidentified Mandapa, Probably at Kanchi (ThousandPillared Hall, Ekāmbaranātha Temple, Kanchi), 1868, photographic print, 20.2 × 27.3 cm Source: © The British Library Board, Photo 1000/26[2559] Anonymous, Rough Sketch of the Fortified Pagoda of Great Cangivaram, 1872, pen and ink on paper, Archives nationales d’Outre-mer, Aix-enProvence, FR ANOM 25DFC 260B Source: Archives nationales d’Outre-mer, Aix-enProvence, FR ANOM 25DFC 260B Archaeological Survey of India, The Amman shrine, Ekambreswaraswami Temple, Conjeevaram, 1897, photographic print, 21.1 × 25.5 cm Source: © The British Library Board, Photo 1008/3[323] Battle of Puḷḷalūr 1780, in Four Plans of engagements between the British forces and those of Hyder Ally and Tippo Sahib [at] Congeveram,

190

191 194 195

200

204

209

211

List of Illustr ations

Illustration 81 Illustration 82 Illustration 83

Illustration 84 Illustration 85 Illustration 86 Illustration 87 Illustration 88 Illustration 89



Sholangur, Vellore [and] Veracundaloor [in the Presidency of Madras], 1780-1782 Source: © The British Library Board, MAPS 54570.[2.] Thomas Colman Dibdin after James Fergusson, Sketch Map of India, 1847, lithograph, 38.2 × 29.4 cm Source: © The British Library Board, X472[35] Henry Merke after James Hunter, A View from the Royal Artillery Encampment, Conjeveram, 1805, aquatint Source: © The British Library Board, X768/3[36] Thomas Daniell, The entrance of an excavated Hindoo Temple, at Mavalipuram, on the coast of Coromandel, Plate 2 from Oriental Scenery, Part 5, 1799, aquatint Source: © The British Library Board, X432/5[2] Thomas Fraser, Elevation of the Great Pagoda at Conjeveram, 1812, engraving, 68 × 50 cm Source: © The British Library Board, P250 J. Clark after James Wathen, Second View from the Great Pagoda near Conjeveram (Plate VI), 1814, hand-colored aquatint Source: Yale Center for British Art, Paul Mellon Collection, DS412.W38 After Anna Tonelli, Palanquin Bearers and a Messenger Camel, in the Journal of Lady Charlotte Florentia Clive, London, 1857 Source: © The British Library Board, WD4235 Mackenzie Collector, Conjeveram Before You Come to the Great Pagoda, 1804, wash, 32 × 50.2 cm Source: © The British Library Board, WD717 After Henry Salt, Pagoda in Conjeveram, Published: Valentia, 1811, opp. p. 437 Source: © The British Library Board, W5289 Mackenzie Collector, Hindoo Architecture: Granite Pillars in a Choultry at Conjeveram, 14th June 1807, 1807, folio 26 in Album of 156 drawings chiefly of architecture and sculpture in S. India (1803-1808), pen and ink Source: © The British Library Board, WD1064

19

214 217

221

224 225

226

230 233 235

237

20 

Construc ting K anchi

Illustration 90 Mackenzie Collector, A Jain at Conjeveram, 16th March 1800, folio 4 in Album of 82 drawings depicting the costume of various castes in Balaghat, Carnatic, 1800, watercolor Source: © The British Library Board, WD1069 Illustration 91 John Gould, Map of Conjeeveram, 23 April 1816, 1816, Pencil, pen-and-ink, and watercolor, 19 × 23 in Source: © The British Library Board, WD2701 Illustration 92 George Russel Dartnell, Stone Pillar in a Choultree at Conjeveram–near Wallajabad–Jany 1829, Pencil and sepia ink wash Source: © The British Library Board, 1997, 1109, 0.60, and .61 Illustration 93 Pillar in Vasanta Maṇḍapa, Kāmākṣī Ammaṉ Temple, Kanchi, sixteenth century Photograph by Emma Natalya Stein Illustration 94 Archaeological Survey of India, Southeast view, Kailasanatha Temple, Great Conjeeveram, Chingleput District, 1900, photographic print, 24.3 × 29.4 cm Source: © The British Library Board, Photo 1008/5[391] Illustration 95 Archaeological Survey of India, South-west view, Matangesvarasvami Temple, Great Conjeeveram, Chingleput District, (Mataṅgeśvara Temple, Kanchi), 1900, photographic print, 23.5 × 18.7 cm Source: © The British Library Board, Photo 1008/5[407] Illustration 96 Thomas Colman Dibdin after James Fergusson, Mahavellipore. The Five Raths, 1839, lithograph, 25.8 × 36.4 cm Source: © The British Library Board, X590 Illustration 97 L.H. Rudder after Prince Alexis Soltykoff, Voyage en poste dans les plaines du Punjab, entre Loodiana et Omritsar, Fevrier 1842. [‘The artist in a palanquin in the plains of the Punjab, between Ludiana and Amritsar, February 1842’], 1848, lithograph, 33.3 and 53.8 cm Source: © The British Library Board, P978

238 240

243

244

246

247

250

255

List of Illustr ations

Illustration 98 L.H. Rudder after Prince Alexis Soltykoff, Condgeveram, ville sainte dans le Carnatik aux environs de Madras, lieu de pélerinage des Hindoux [‘Conjeveram, sacred city in the Carnatic in the vicinity of Madras, a Hindu pilgrimage place’], 1848, lithograph Source: © The British Library Board, P969 Illustration 99 L.H. Rudder after Prince Alexis Soltykoff, Intérieur du couvent de Condgeveram, à 40 milles de Madras, Réunion journalière des Brames en l’honneur des deux divinités Conservatrice et Destructive, Juillet 1841 [‘Inside a temple at Conjeveram, 40 miles from Madras, Daily reunion of the Brahmins in honor of the two deities of Preservation and Destruction, July 1841’], 1848, lithograph, 46 × 59.8 cm Source: © The British Library Board, P970 Illustration 100 The goddess in her chariot, Kāmākṣī Ammaṉ Temple, Kanchi, 18 July 2014 Photograph by Emma Natalya Stein

21

256

257 262



Note on Transliteration, Translation, and Illustrations

Translations of Tamil and Sanskrit terms are given in parentheses the first time they appear in each chapter. Tamil words are transliterated according to the system of the Tamil Lexicon (University of Madras). For Sanskrit words I use the standard (Monier-Williams) system of transliteration. Where the Sanskrit equivalent of a Tamil word is more commonly known, I use the Sanskrit form (such as gopura rather than gopuram). I use conventional spellings for the names of relatively well-known places (Kanchipuram, Madurai, Chidambaram, etc.) and Tamil transliteration for smaller places (such as Tirukkaḻuṉṟam, Cevilimēṭu, Māmaṇṭūr). Where it cannot be avoided, I use hybrid Sanskrit/Tamil forms. I use conventional spellings for certain additional words (such as Chola rather than Cōḻa, and Shiva and Vishnu rather than Śiva and Viṣṇu.) For architectural descriptions, I follow the terminology employed in the Encyclopaedia of Indian Temple Architecture (Meister and Dhaky, 1983). Images and translations are my own unless otherwise noted. Since many of the temples discussed in this book were not previously documented, I have prioritized photographs and illustrations that are unique.

Introduction All Streets Lead to Temples On the northern border of the city of Kanchi stands the sprawling temple complex of Ekāmbaranātha. Shaded lakes and dense, jungly groves fringe the walled compound to its north and west, and a soaring gopura (‘gateway’) marks the main passage into the sacred space from a street on the southern side. Leading directly to the temple, this busy commercial street is lined with sweet-sellers, tea stalls, and merchants selling the city’s famed goldand-purple silk saris (Ill. 1). Some vendors have semi-permanent stalls set up along the street or built into maṇḍapas (‘pillared halls’), while others are itinerant peddlers who hawk their wares from movable carts. Between the shops are the houses of priests whose families have maintained the Ekāmbaranātha temple for generations. Closer towards the gopura, ladies sell garlands of jasmine and trays heaped with fresh lotus blossoms, coconuts, and bananas to be given as offerings to the gods inside the temple. From the early hours of the morning, the bustling street swells with auto-rickshaws, cars, buses, and all manner of vehicles carrying the thousands of visitors who arrive at Ekāmbaranātha each day. By nightfall, the temple traffic slows, the merchants cover their goods and close their shops, and the street returns to its local residents. When Xuanzang, the famous Chinese monk and traveler, visited Kanchi in the middle of the seventh century, he praised the city as a prosperous urban center surrounded by fertile paddy fields and filled with learned priests tending hundreds of sacred buildings.1 The city’s many temples greatly impressed this well-traveled Buddhist pilgrim. It is perhaps ironic, then, that Xuanzang visited Kanchi a century before its construction as a temple-city had even truly begun. During the eighth through thirteenth century, Kanchi served as the royal capital for two major South Indian 1 Xuanzang, Si-Yu-Ki. Buddhist Records of the Western World. Translated from the Chinese of Hiuen Tsiang (A.D. 629) by Samuel Beal, 2 vols., vol. 2 (London: Trubner & Co., (1884) 1969), 228-229. For Xuanzang, I follow spelling given in Robert E. Buswell and Donald S. Lopez, The Princeton Dictionary of Buddhism (Princeton: Princeton University Press, 2014).

Stein, Emma Natalya, Constructing Kanchi: City of Infinite Temples. Amsterdam, Amsterdam University Press 2021 doi: 10.5117/9789463729123_intro

26 

Construc ting K anchi

Illustration 1 Ekāmbaranātha Temple and Sannathi Street, Kanchi, seventh century – the present

dynasties – the Pallavas and then the Cholas – and was home to thousands of priests, literati, and landholding elites. The rulers and residents who dwelled in and around the city during these formative centuries sponsored the construction of more than 50 stone temples that still stand in varying states of preservation today (Ill. 2). Built from supple blocks of sandstone and smooth granite slabs, these temples were crowned with pyramidal towers, adorned with relief carvings of divine figures in forested and palatial landscapes, and elegantly inscribed with courtly Sanskrit and Tamil verse (Ill. 3). The dark inner sanctums sheltered stone liṅgas (‘signs of Shiva’) or icons of goddesses and gods. Other sanctums contained sensuous bronze deity figures with swaying hips, tapering limbs, and regal ornaments that flickered in the light of butter lamps. The construction of these temples was part of a series of larger acts of urban planning, which included the establishment of special avenues designed for processions of the festival bronze images. Kanchi’s temples attracted a wide network of merchants and devotional communities that flocked to the city. In the 500-year period of Pallava and Chola reign, temples throughout Tamil Nadu took on new roles within a widening range of cultural, economic,

27

Introduc tion

Illustration 2 Map of Kanchi Temples (map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution)

28  Illustration 3 Kailāsanātha Temple, Kanchi, ca. 700-725 CE

Construc ting K anchi

Introduc tion

29

legal, and political processes.2 Temples remained religious institutions, but they served more than religious functions. They acted as the legal sovereigns of lands and properties, and they were responsible for the management of community resources, such as water and rice. Temples were also responsible for various municipal services – they maintained schools, hospitals, and feeding houses.3 The rise of temples in Tamil Nadu was representative of a distinct change in South India’s socioeconomic makeup that included the consolidation of urban centers and the creation of robust agrarian estates.4 In previous centuries, temple worship focused on deities that were connected with landscape and sustenance. The seventh and eighth century saw instead the institutionalization of temples dedicated predominantly to Shiva and Vishnu.5 Animated by the communities of people who used and moved between them, temples became spaces where diverse groups fashioned, enacted, and negotiated their claims to prosperity, political authority, and cultural capital. Such negotiations can be seen through structural renovations and read in records of pious gifts that are documented in inscriptions on the temple walls. This book offers a first-ever understanding of Kanchi’s physical transformation from a relatively small settlement into a cosmopolitan urban center. The first half of the book geographically reconstructs the emergence and reconfiguration of the city around a major pilgrimage route that was also a great artery of commerce. It then broadens the scope of enquiry both geographically and temporally to consider networks of trade and devotion 2 A large body of literature on precolonial South India considers the role of temples. Burton Stein, Peasant State and Society in Medieval South India (Delhi: Oxford University Press, 1980); Kenneth R. Hall, Trade and Statecraft in the Age of the Cholas (New Delhi: Abhinav Publications, 1980); James Heitzman, Gifts of Power: Lordship in an Early Indian State (Delhi: Oxford University Press, 1997); R. Champakalakshmi, Trade, Ideology, and Urbanization: South India 300 BC to AD 1300 (Delhi: Oxford University Press, 1999); Kesavan Veluthat, The Early Medieval in South India (Delhi: Oxford University Press, 2009). 3 An inscription on Tirumukkūṭal’s Ādikeśava Perumāḷ temple records multiple functions. T.V. Mahalingam, ed. A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, 8 vols., vol. 3 (New Delhi: Indian Council of Historical Research, 1989), p. 336, Cg.-1313. 4 Veluthat makes a case for an early medieval period in South India (seventh through twelfth century) represented by the temple as an institution. Absent in previous centuries were the ‘sprawling Brāhmaṇa settlements with vast areas of agricultural land under their command and the temple as the pivot around which they functioned.’ Veluthat, The Early Medieval in South India, 5 and 62-63. 5 Ibid. Although goddesses (‘devī’) were also important in this period, it was not until the tenth and eleventh century that goddess worship became fully incorporated into mainstream practice, and separate temples for forms of Devī were constructed.

30 

Construc ting K anchi

that connected Kanchi with the rest of South India and a wider world. The final chapter considers the continued construction of Kanchi’s identity as a temple-city in the colonial era and onwards into the present. The research coordinates extensive first-hand field surveys with archival materials, including epigraphic and archaeological reports, early photography, colonialera travelers’ accounts, and contemporary practice. The aim of the book is twofold. First, it documents the city and maps its monuments spatially and chronologically in relation to each other, to the city, and to features of the natural environment. Second, it situates temples as functional establishments that continuously contributed to a growing urban landscape. At its core, the book demonstrates that Kanchi is structured with a unique urban logic that coordinates the placement of temples and roads. Overlaying this tightly woven urban plan, the building and renovation of temples in and around the city has enabled Kanchi to continuously thrive as a prosperous center from at least the eighth century up until the present.

An Ancient City Kanchipuram, or ‘Kanchi’, as it is more simply known, is an ancient city. When the Pallava dynasty moved its seat from Andhra Pradesh in the third century to establish a presence in Tamil Nadu, the royal family selected Kanchi for their new capital because it was already an important place.6 Kanchi was widely known throughout India during this era – nearly 2000 kilometers away in North India, the Allahabad pillar of Aśoka contains a ca. fourth-century inscription from the Gupta king, Samudragupta, in which he claims to have conquered one of Kanchi’s early Pallava kings.7

6 For discussion of the early Pallavas, see T.V. Mahalingam, Kāñcīpuram in Early South Indian History (Madras: Asia Publishing House, 1969), 25-53; Valérie Gillet, La Création d’une Iconographie Śivaïte Narrative: Incarnations du Dieu dans les Temples Pallava Construits (Paris: Institut Français de Pondichéry, Ecole Française d’Extrême-Orient Pondichéry, 2010), 23-28; ‘The Dark Period: Myth or Reality?’, The Indian Economic and Social History Review 51, no. 3 (2014). 7 Lines 19-20 read: ‘Whose magnanimity blended with valour was caused by (his) f irst capturing, and thereafter showing the favor of releasing, all the kings of Dakṣiṇāpatha, such as… Vishṇugōpa of Kāñcī…’ The Pallava king referenced is Viṣṇugopavarman. Ramesh Chandra Majumdar and Anant Sadashiv Altekar, The Vākātaka-Gupta Age: Circa 200-550 A.D. (Delhi: Motilal Banarsidass, [1967] 1986), 245. I thank Derek Mitchell for drawing my attention to this inscription.

Introduc tion

31

Kanchi appears frequently in South Indian epics and poetry from the Caṅkam era (ca. first century BCE-sixth century CE).8 Although the texts give us mere glimpses of its character rather than the kind of rich descriptions of the city of Madurai and port of Pūmpukār found in the epic Cilappatikāram [‘Tale of the Anklet’], Kanchi’s repeated inclusion in Caṅkam literature indicates its importance as a South Indian city. The Pattuppāṭṭu anthology of ten Tamil poems often mentions Kanchi.9 It does so through the literary technique of āṟṟuppaṭai in which a wandering bard, having received favor from a king in a distant land, returns to his homeland and advises another bard or artist to seek similar sponsorship. This gives occasion for the returned traveler to extol his royal patron and the lands in that king’s domain. In the Perumpāṇāṟṟuppaṭai, the bard praises Kanchi as a place populated by foreign merchants and alive with diverse festivals.10 Kanchi is the ultimate destination in the Maṇimēkalai, one of the five great Tamil epics.11 Also a Caṅkam-era work, this text describes a spiritual journey along the path of Buddhism that leads the heroine directly to Kanchi. The reader may long for a description of the city upon the heroine’s arrival, but instead the text concentrates on the many teachers of the Buddhist dharma she meets there. The great philosopher Śaṅkarācārya, who was probably active in the eighth century, is believed to have founded Kanchi’s main monastery and to have spent time there near the end of his life.12 Kanchi maintains a position of importance within literature of the late first and early second millennium. A great many of Kanchi’s temples are sanctified in the hymns of the Tēvāram and Divya Prabandham, poetic anthologies composed by the Tamil saints (Shaiva nāyaṉmārs and Vaishnava āḻvārs respectively), who probably lived between the sixth and ninth century. The life stories of the Shaiva saints were later compiled into a hagiographic anthology called the Periya Purāṇam, which was composed in the twelfth century at the Chola court.13 Kanchi’s prominent role in this corpus of 8 This is the range of dates for Caṅkam texts that is generally accepted by scholars of Tamil literature. 9 Pattuppāṭṭu, [‘The Ten Tamil Idylls’], trans. A. Dakshinamurthy (Kattankulattur: Tamil Perayam, SRM University, 2012). I have also consulted the original Tamil, with assistance from Jean-Luc Chevillard (CNRS, Paris). U.Vē. Cāminātaiyar, ed., Pattuppāṭṭu, V., ed. (Madras: Kabeer Printing Works, 1956). 10 Pattuppāṭṭu, 184. 11 Shattan, Maṇimēkalai [‘The Dancer with the Magic Bowl’], trans. Alain Danielou and T.V. Gopala Iyer (India: Penguin Books, 1993). 12 James Heitzman, The City in South Asia (Abingdom: Routledge, 2008), 57-58. 13 Versions consulted: Cēkkiḻār, Periya Purāṇam (Madras: C.K. Subiramaniya Mudaliyar, 1950); St. Sekkizhar’s Periya Purāṇam, trans. T.N. Ramachandran, 2 vols., vol. 1 (Thanjavur:

32 

Construc ting K anchi

Illustration 4 Buddha, Kanchi Police Station, twelfth century

bhakti (‘devotional’) literature attests to enduring cults of devotion that centered on the city. Archaeological evidence of Kanchi’s longevity supports the literary testimony. Megalithic burial sites and excavations in and around the city have revealed the area’s continuous inhabitation and its activity in Roman circuits of trade.14 In the majestic cave-temple site of Māmaṇṭūr, ten kilometers to the south, a first-century BCE inscription in Brahmi script reveals that by the early historic period, Kanchi and its hinterland served as an important center of religious and literary knowledge.15 Below the inscription is an ancient stone bed associated with early Jain ascetics, who would have used the caves for retreat. At that same site, a Grantha inscription credits the Pallava king Mahendravarman I (ca. 580-630 CE) with authorship of Tamil University, 1990); St. Sekkizhar’s Periya Purāṇam, trans. T.N. Ramachandran, 2 vols., vol. 2 (Thanjavur: Tamil University, 1995); The History of the Holy Servants of the Lord Śiva: A Translation of the Periya Purāṇam of Cēkkiḻār, trans. Alastair McGlashan (Victoria, Canada: Trafford Publishing, 2006). 14 Robert Bruce Foote, On the Occurrence of Stone Implements in Lateritic Formations in Various Parts of the Madras and North Arcot Districts (Madras: Graves, Cookson and Co., 1865); Prehistoric and Protohistoric Antiquities of India (Delhi: Leeladevi Publications, 1916). 15 ARE 1939-40, no. 171.

Introduc tion

Illustration 5 Sīteśvara Temple, Kanchi, tenth century

33

34 

Construc ting K anchi

the Mattavilāsa Prahasana, a Sanskrit drama that takes place in courtly Kanchi.16 Throughout the play, specific temples are associated with different sectors of society to give a picture of a multireligious place. Today, the various literary representations are borne out through Kanchi’s surviving temples and statues of Buddhas, Jinas, and plethora of Hindu gods (Ill. 4). Kanchi itself provides a rich and complex archive. At least eight full temples were established in the eighth century alone, under the auspices of the Pallavas. Dozens of small shrines and scattered fragments furnish evidence of additional Pallava-era temples. No less than 25 extant shrines date to the Chola period (ca. 850-1279 CE), and disengaged sculptures attest to the former presence of an even greater number (Ill. 5). Countless inscriptions and architectural fragments have gone unnoticed in official epigraphic and archaeological reports, and still more lie hidden beneath recent layers of paint and plaster.17

Layers of Time The archaeology in Tamil Nadu is largely horizontal – rather than being buried underground, the sculptures and architectural elements that belong to earlier iterations of sacred sites are typically hidden in plain sight. With some detective work, they can be found incorporated into later structures, scattered throughout open courtyards in temple complexes, or re-enshrined, sometimes as a different deity than first intended. Structural degradation and renovation often exhibit distinctive patterns. Over time, interiors become exteriors, as walls and superstructures disintegrate due to weather and ware. With renovation and rehabilitation, exteriors in turn become interiors. Pillared halls are sealed into fully walled structures, and additional maṇḍapas are appended to earlier sequences of entry halls. What previously had been an open-air ambulatory may later be covered with a roof, or a concentric cloister around a temple may subsume the main shrine, such that what originally was a freestanding building becomes a dark room. Since the founding of the Archaeological Survey of India (ASI) in the latter half of the nineteenth century, these transformations have sometimes been halted or 16 SII Vol. 4, p. 13, no. 138. The inscription is situated in the northernmost cave (Cave 1), on the south interior wall of the verandah. 17 Specialist readers can find more extensive listings and architectural descriptions of Kanchi’s sites in Emma Natalya Stein, ‘All Streets Lead to Temples: Mapping Monumental Histories in Kanchipuram, ca. 8th-12th centuries CE’ (PhD diss., Yale University, 2017).

Introduc tion

35

reversed in officially protected monuments.18 This book interrogates these processes in order to peel away the layers of time from temples built up over the course of what has often been more than a millennium of construction and renovation. Like temple walls, rituals too are diachronic and multilayered. Many of the sacred festivals that take place in the city today represent consolidations of what previously were expansive rituals or even daily practices. For example, during the Pongal festival that opens the Tamil month of Tai (January/February), a festival bronze of Shiva is carried in procession from the Ekāmbaranātha temple to the town of Dimasamudram, eight kilometers to Kanchi’s north. In Dimasamudram, local performers present song and dance, and the priests from Ekāmbaranātha conduct pūjā (‘worship’) for the gathered devotees. That same night, the festival bronze is returned to Kanchi for an elaborate abhiṣeka (‘ritual bathing’) inside the Ekāmbaranātha temple. The purpose of Shiva’s daylong journey is to extend darśan (“blessing through an exchange of gazes with the divine”) of Ekāmbaranātha as both temple and god to people in surrounding villages who are unable to travel to Kanchi. Although the practice of bronze icon processions dates back at least a millennium, this particular festival was instantiated only several decades ago, as a consolidated form of an earlier ritual.19 Formerly, the icon was taken to multiple villages in the surrounding area. Now it is brought only to one. Dimasamudram was selected as the singular destination for the sake of efficiency. It is comprised of a combination of previously independent settlements, and it is centrally located among a cluster of proximate villages. Devotees can go to Dimasamudram more easily than they can reach Kanchi, and the god can spread darśan in a more limited amount of time. The colonial period brought about extensive changes to official forms of ritual praxis. The differences can be seen through comparison between contemporary rituals and precolonial literature and artistic representations. Sectarian rivalry, both among competing Hindu groups and between Hindus and Jains, marks one such critical arena of change.20 While colonial of f icers often sought to emphasize tensions between devotional 18 After several attempts, the ASI was established in earnest in 1871 under Alexander Cunningham. 19 My thanks to Ekāmbaranātha pūjāri (‘priest’) Nagaswamy Aiyyer and his son for discussing this ritual with me (January 2014). 20 Sectarian rivalry continues to be a primary concern in scholarship. See for example Emmanuel Francis, Valérie Gillet, and Charlotte Schmid, ‘L’Eau et le Feu: Chronique des Études Pallava’, Bulletin de l’Ecole Française d’Extrême-Orient 92 (2005); Emmanuel Francis, ‘“Woe to Them!”: The Śaiva Curse Inscription at Mahābalipuram (7 th Century CE)’, in The Archaeology

36 

Construc ting K anchi

communities, scholars have recently argued against the notion of rivalry, instead emphasizing productive encounters between members of different religious communities, or drawing attention to the pluralistic textures of precolonial Hinduism.21 Leslie Orr has shown that temples with different dedications sometimes exchanged personnel, and that the corpus of Tamil inscriptions constitutes a literary genre of its own that transcends sectarian divides.22 Various forms of Tamil texts often situate Shiva and Vishnu as relatively equal recipients of devotion. Sandeśakāvyas (‘messenger poems’), for example, give greater importance to regional unity than to the author or sponsor’s sectarian affiliation.23 The Haṃsasandeśa, by the fourteenthcentury Vaishnava theologian Vedānta Deśika who was born and resided in Kanchi, extols temples dedicated to Shiva in the midst of what is supposedly an exclusively Vaishnava poem.24 Even in the Kāmākṣīvilāsa, a Sthalapurāṇa (‘legendary history’) of Kanchi that centers on the Goddess, Vishnu and Shiva are given a remarkable level of prominence.25 In practical terms, Kanchi once maintained a daily ritual in which Shiva and Vishnu visited with each other and were worshipped in tandem.26 A daily occurrence until the middle of the nineteenth century, this practice survives only in the form of a cursory ritual that takes place once a year during the ten-day festival of Garuda Sevai. Similarly, the present-day designations of Shiva-Kanchi (Big/ of Bhakti: Mathurā and Maturai, Back and Forth, ed. Charlotte Schmid and Emmanuel Francis (Pondicherry: École Française d’Extrême-Orient, 2013). 21 Richard H. Davis, ‘The Story of the Disappearing Jains: Retelling the Śaiva-Jain Encounter in Medieval South India’, in Open Boundaries: Jain Communities and Culture in Indian History, ed. John E. Cort (Albany, NY: State University of New York Press, 1998); Gary Michael Tartakov and Vidya Dehejia, ‘Sharing, Intrusion, and Influence: The Mahiṣāsuramardinī Imagery of the Calukyas and the Pallavas’, Artibus Asiae 45, no. 4 (1984); Elaine M. Fisher, Hindu Pluralism: Religion and the Public Sphere in Early Modern South India (California: University of California Press, 2017). 22 Leslie Orr, ‘Preface’, in Pondicherry Inscriptions, ed. Bahour S. Kuppusamy and G. Vijayavenugopal (Pondicherry: Institut Français de Pondichéry and École Française d’Extrême-Orient, 2006), XXVII; ‘Processions in the Medieval South Indian Temple: Sociology, Sovereignty and Soteriology’, in South-Indian Horizons: Felicitation Volume for Francois Gros on the Occasion of his 70th Birthday, ed. Jean-Luc Chevillard and Eva Wilden (Pondicherry: Institut Franc̨ a is de Pondichery and École Franc̨ a ise d’Extreme-Orient, 2004), 450-451. 23 Steven P. Hopkins, ‘Lovers, Messengers, and Beloved Landscapes: Sandeśakāvya in Comparative Perspective’, International Journal of Hindu Studies 8, no. 1-3 (2004): 40. 24 Ibid. 25 Śrīkāmākṣīvilāsaḥ (Bangalore: Bhāratalakṣmī Mudraṇālayam, 1968). I thank Ute Hüsken for pointing out this aspect of the text. 26 See Chapter Four. Prince Alexis Soltykov, Lettres Sur l’Inde (Paris: Amyot, 1848), 70-71.

Introduc tion

37

Illustration 6 Festival at Kāmākṣī Ammaṉ Temple, Kanchi (July 2014)

periya-Kanchi), Vishnu-Kanchi (Little/ciṉṉa-Kanchi), and Jina-Kanchi that divide the city into devotional zones were not set into place until the latter half of the nineteenth century.27 While early colonial sources distinguish ‘Little Kanchi’ from the rest of the city, they make no mention of affiliation with a particular god.28 Not all of Kanchi’s temples continued as active centers of devotion or economic exchange after the Chola period. By the time European travelers arrived in Kanchi, the Pallava temples in particular lay well outside of the urban focus. When the first colonial-era artists and travelers visited the city, they did not know about the great Kailāsanātha and Vaikuṇṭha Perumāḷ – Pallava temples that today are among the sole destinations for the rare foreign visitors to Kanchi. Photographs taken shortly after the Pallava temples came to light in European circles show that these temples remained, to varying extents, sites of devotion for local people who resided 27 Nilakanta Sastri also mentions an area called Buddha-Kanchi, but I have found no further reference to this. K.A. Nilakanta Sastri, A History of South India: From Prehistoric Times to the Fall of Vijayanagar (Madras: Oxford University Press, 1975), 437. 28 Colonial sources for the use of ‘Little Kanchi’ include the writings of Henrietta and Charlotte Clive. Nancy K. Shields, ed. Birds of Passage: Henrietta Clive’s Travels in South India, 1798-1801 (London: Eland, 2009), 109-110.

38 

Construc ting K anchi

in the temple’s very immediate vicinity. Far removed from their royal Pallava origins, however, the temples had become relatively quiet places that had long-since ceased to attract the attention of affluent residents and visitors to the city. Archaeological surveyors focused on these monuments to the exclusion of others in order to perpetuate the fiction of the empire’s own necessity. If the temples were in shambles, they argued, so too was India. Other temples in the city inadvertently benef ited from the colonial focus on the Pallava sites. The sprawling complexes of temples such as Ekāmbaranātha, Varadarāja Perumāḷ, and Kāmākṣī Ammaṉ, as well as dozens of other shrines throughout the city, remained active ritual centers that were uninterrupted by colonial efforts at preservation. As a result, they continue to function as Kanchi’s primary living temples and are still the recipients of local devotion (Ill. 6). Contemporary practice in these temples can shed important light on ritual processes that are depicted in the more ancient sculptural reliefs and described in early literature. I have observed daily rituals, as well as several large-scale pūjās, that were sponsored by collective donations or individual members of the South Indian urban elite. Many of the practices involved in these rituals show a remarkable level of continuity from the past.

Kanchi Known and Unknown Kanchi is well known to scholars through studies of Tamil literature, socioeconomic history, and religion. However, despite the city’s clear importance there has never been a single sustained study of Kanchi’s urban space or architecture, nor have the temples been systematically mapped. Portions of the city’s political history have been charted and certain monuments have been classified according to dynastic style.29 Some scholars have produced monographs on single temples.30 Others have conducted valuable analyses of 29 Mahalingam, Kāñcīpuram in Early South Indian History; C.R. Srinivasan, Kanchipuram through the Ages (New Delhi: Agam Kala Prakashan, 1979); K.R. Srinivasan, ‘Pallavas of Kāñcī: Phase I’, in Encyclopaedia of Indian Temple Architecture, ed. Michael W. Meister and M.A. Dhaky (New Delhi: American Institute of Indian Studies, 1983); ‘Pallavas of Kāñcī: Phase II’, in Encyclopaedia of Indian Temple Architecture, ed. Michael W. Meister and M.A. Dhaky (New Delhi: American Institute of Indian Studies, 1983). 30 C. Minakshi, ‘The Historical Sculptures of the Vaikuṇṭhaperumāḷ Temple, Kāñcī’, Memoirs of the Archaeological Survey of India 63 (1941); K.R. Venkataraman, Dēvī Kāmākshī in Kāñchī: A Short Historical Study (Tiruchirapalli: Sri Vani Vilas Press, 1973); K.V. Raman, Srī Varadarājaswami Temple, Kāñchi: A Study of Its History, Art and Architecture (New Delhi: Abhinav Publications, 1975); R. Nagaswamy, ‘Innovative Emperor and His Personal Chapel: Eighth Century Kanchipuram’,

Introduc tion

39

iconography and inscriptions using Kanchi as a case study.31 Among scholars of literature, Kanchi is known as a multilingual and multireligious place that fostered the development of regional cosmopolitan literatures.32 The city has also been mentioned in broader studies of premodern world systems.33 In focusing on the relationships among temples, local landscapes, and transregional networks of religion and power, this book contributes to a broader range of ongoing scholarship within area studies and cultural history.34 First, it enters into conversations on cosmopolitanism and the politics of cultural production.35 Second, it engages with urban studies that have described Kanchi as a driving force in South India’s socioeconomic integration.36 Third, it considers modes of encounter and reception by in Royal Patrons and Great Temple Art, ed. Vidya Dehejia (Bombay: Marg Publications, 1988); Padma Kaimal, ‘Learning to See the Goddess Once Again: Male and Female in Balance at the Kailāsanath Temple in Kāñcīpuram’, Journal of the American Academy of Religion 73, no. 1 (2005); D. Dennis Hudson, The Body of God: An Emperor’s Palace for Krishna in Eighth-Century Kanchipuram (Oxford: Oxford University Press, 2008); Padma Kaimal, Opening Kailasanatha: The Temple in Kanchipuram Revealed in Time and Space (Seattle: University of Washington Press, forthcoming 2021). 31 Michael Lockwood, Pallava Art (Madras: Tambaram Research Associates, 2001); James Heitzman, ‘Urbanization and Political Economy in Early South India: Kāñcīpuram During the Cōḻa Period’, in Structure and Society in Early South India: Essays in Honour of Noboru Karashima, ed. Kenneth R. Hall (New Delhi: Oxford University Press, 2001); Francis, Gillet, and Schmid, ‘L’Eau et le Feu: Chronique des Études Pallava’; Gillet, La Création d’une Iconographie Śivaïte Narrative: Incarnations du Dieu dans les Temples Pallava Construits; James Heitzman and S. Rajagopal, ‘Urban Geography and Land Measurement in the Twelfth Century: The Case of Kanchipuram’, Indian Economic and Social History Review 41, no. 3 (2004). 32 Hopkins, ‘Lovers, Messengers, and Beloved Landscapes: Sandeśakāvya in Comparative Perspective’, 37; David Shulman and Yigal Bronner, ‘“A Cloud Turned Goose”: Sanskrit in the Vernacular Millennium’, The Indian Economic and Social History Review 43, no. 1 (2006). 33 K.A. Nilakanta Sastri, Foreign Notices of South India: From Megasthenes to Ma Huan (Madras: University of Madras, 1972); Janet L. Abu-Lughod, Before European Hegemony: The World System A.D. 1250-1350 (New York: Oxford University Press, 1991); Kenneth R. Hall, ‘Local and International Trade and Traders in the Straits of Melaka Region: 600-1500’, Journal of the Economic and Social History of the Orient 47, no. 2 (2004). 34 Nachiket Chanchani, Mountain Temples and Temple Mountains: Architecture, Religion, and Nature in the Central Himalayas, Global South Asia (Washington: University of Washington Press, 2019). 35 Sheldon Pollock, The Language of Gods in the World of Men: Sanskrit, Culture, and Power in Premodern India (Berkeley: University of California Press, 2006); Shonaleeka Kaul, Imagining the Urban: Sanskrit and the City in Early India (London: Seagull Books, 2010). 36 R. Champakalakshmi, ‘The Urban Configurations of Toṇḍaimaṇḍalam: The Kāñcīpuram Region, C.A.D. 600-1300’, in Urban Form and Meaning in South Asia: The Shaping of Cities from Prehistoric to Precolonial Times, ed. Doris Meth Srinivasan and Hugh Urban (Washington: National Gallery of Art, 1993); Heitzman, ‘Urbanization and Political Economy in Early South India: Kāñcīpuram During the Cōḻa Period.’

40 

Construc ting K anchi

analyzing travel literature ranging from early first-millennium Tamil epics to nineteenth-century European accounts.37 Fourth, it continues the burgeoning interest in landscape studies both within and beyond South Asia.38 Finally, it makes a critical intervention in postcolonial studies by challenging narratives of deterioration and decay. Kanchi was never in a state of decline. Instead, the city continued to flourish through enduring processes of transformation and renewal. This book takes as its organizing structure the pairing of kṣetra and kṣatra, the dual principle of royal dominion known from South Asian literature.39 Kṣetra, which means ‘field’ in Sanskrit, is defined as a ruler’s terrestrial domain, and kṣatra represents his sphere of influence. The first two chapters map Kanchi’s kṣetra by tracing the city’s shifting contours and the emergence of a major pilgrimage route that led precisely through its urban core. The third chapter transitions outwards to Kanchi’s broader kṣatra by looking at patterns of movement that linked the city to its hinterland, and by considering connections with multireligious urban capitals across a wider South and Southeast Asian region. The fourth chapter focuses on afterlives and reception of the sites discussed in the first three chapters through colonial-era encounters with Kanchi and the city’s role in the production of colonial knowledge. Chapter One focuses on the last two centuries of the Pallava period, from ca. 700-900 CE, which represent a pivotal moment in the history of South Indian art. It begins with a landscape of brick and open-air shrines, which was soon supplanted by stone as the preferred building material for elite sacred architecture. Through a rigorous mapping of temples and architectural fragments from this period, I offer a new vision of the Pallava’s royal capital. I demonstrate that Kanchi’s urban core doubled in size during a single century, and that the early city was situated to the west of present-day Kanchi. I also recover a lost geological landscape, one that was rich in sandstone, the primary construction material in Pallava-Kanchi. 37 Iḷaṅkō Aṭikal, The Cilappadikāram [‘The Tale of the Anklet’], trans. V.R. Ramachandra Dikshitar (Madras: The South India Saiva Siddhanta Works Publishing Society, 1978); Shattan, Maṇimēkalai; Pattuppāṭṭu; Shields, ed. Birds of Passage: Henrietta Clive’s Travels in South India, 1798-1801; Lady Charlotte Florentia Clive, Journal of Lady Charlotte Florentia Clive (1787-1866), Copied by Professional Calligrapher, W.H. Ramsay (British Library, Visual Arts Collection, WD2435, 1857); Soltykov, Lettres Sur l’Inde. 38 Daud Ali and Emma Flatt, eds., Garden and Landscape Practices in Precolonial India: Histories from the Deccan (Delhi: Routledge, 2012). 39 Hermann Kulke, ‘Kṣetra and Kṣatra: The Cult of Jagannātha of Puri and the “Royal Letters” (Chāmu Ciṭaus) of the Rājās of Khurda’, in Kings and Cults: State Formation and Legitimation in India and Southeast Asia (New Delhi: Manohar, 1993).

Introduc tion

41

Throughout the book, the terms ‘Pallava-Kanchi’ and ‘Chola-Kanchi’ refer to the physical footprint of the city and the entire sociocultural milieu that lay within it during each dynasty’s reign. The chapter closes with the ninth century, when growing political instability within Kanchi prompted the Pallavas to concentrate their efforts on outlying areas, while new rulers vied for power inside the city. Chapter Two centers on a road that emerged as a major pilgrimage route in the Chola period. Passing directly through the urban core and extending far beyond the city’s boundaries, the road positioned Kanchi as one stop along an enduring pilgrimage network that connected northern Tamil Nadu with the holy hilltop temple at Tirupati. I identify a unique architectural pattern that exists within Kanchi alone – all temples face this main road. The chapter then follows the road to the villages south of Kanchi. Changes in structures of patronage and temple architecture in the rural hinterland supply important information concerning similar transformations that were taking place on a much larger scale and with greater complexity inside the borders of the city. This chapter represents the first time that Kanchi’s Chola-period temples – often hidden behind perpetually locked doors or embedded within much later temples – have been identified, mapped, given a chronological ordering, and analyzed in the context of their contribution to the growth and intensification of the urban space. Chapter Three widens the focus from the urban core to consider the creation of a rural-urban continuum that increasingly came to def ine Kanchi. Building on foundational studies by scholars such as Burton Stein, R. Champakalakshmi, and James Heitzman, it examines the more circumscribed settlements in the rural areas – being at a more manageable scale of two or three temples – to understand changes inside the city. Each of the settlements discussed shows evidence of a deep history, and each was then home to Pallava and/or Chola-era sacred architecture that received subsequent endowments and modifications over time. Together, these places became an interwoven network. The chapter traces interactions among the settlements through shared patterns in temple architecture, iconography, and inscriptions, as well as residential design. The chapter’s final section considers the possibility of a cultural landscape that expanded Kanchi’s kṣatra transregionally. Specifically, it looks at Kanchi’s role in extended Buddhist networks and then explores connections across urban centers in an integrated South and Southeast Asian region. 40 40 John Guy has demonstrated that South and Southeast Asia constituted an integrated region as early as the fifth century. John Guy, ed. Lost Kingdoms: Hindu-Buddhist Sculpture of Early

42 

Construc ting K anchi

Finally, Chapter Four uses colonial-era sources to examine how foreign visitors encountered Kanchi at the crucial moment when knowledge of Indian art and geography was first being systematized and transmitted to European audiences. This chapter investigates the ways in which a popular narrative was set in place that viewed Kanchi as a city in decline – most clearly articulated by architectural surveyor James Fergusson – and how it is somewhat ironically belied by other genres of representation. While early photographs of the Pallava temples show these once-prominent sites as crumbling structures in patches of wilderness, colonial-era prints, drawings, and travelers’ accounts tell a very different story – one that speaks not of decline or disrepair, but of a continuously flourishing urban center. Sources examined range from the collected manuscripts of Colonel Colin Mackenzie, to the private diaries and letters of Henrietta Clive (daughter-in-law of Robert Clive), to the drawings and letters of the Russian aristocrat Alexis Soltykoff, who described Kanchi as a city of ‘infinite temples.’41 Moving from kṣetra outward to kṣatra, the four chapters construct a layered vision of Kanchi, from its establishment as a royal capital, to its present-day preservation of a cultural heritage that includes not only archaeological monuments but also the abundant processions and festivals that keep the city pulsing with life. The book demonstrates the ways in which Kanchi has been shaped, reshaped, ordered, and reconfigured through the construction of temples and the allocation of space. Through Kanchi’s continual transformations, its temples have remained intrinsically connected with the vibrant urban landscape.

Southeast Asia (New Haven and London: Metropolitan Museum of Art and Yale University Press, 2014). 41 Soltykov, Lettres Sur l’Inde, 69.

1

Sandstone and the City Building Pallava-Kanchi (ca. seventh through ninth century) Abstract Chapter One focuses on the last two centuries of the Pallava period (ca. 700-900 CE), when Kanchi became a city marked by the towers of many sandstone temples. This period represents a pivotal moment in the history of South Indian art, when stone supplanted brick as the construction material for elite sacred architecture. Through a rigorous mapping of the city’s surviving temples and archaeological remains, this chapter reveals that Kanchi’s urban core doubled in size within a single century and demonstrates that the eighth-century city was actually situated west of present-day Kanchi. In the ninth century, growing political instability within Kanchi prompted the Pallavas to concentrate their efforts on outlying areas, while new rulers vied for power inside the city. Keywords: Kailāsanātha temple, Ekāmbaranātha temple, King Mahendravarman I, Tamil, Sanskrit, Mattavilāsa Prahasana

A set of seven sandstone relief carvings at the Tāṉtōṉṟīśvara temple in Kanchi depicts sensuous figures of courtesans, courtly elites, attendants, and ascetics, who stumble, dance, or wrestle playfully (Ill. 7). Male and female figures flirt and converse. A hand teasingly coils a lock of hair, a torso twists just enough to reveal a proffered flower, and a foot steps forward, its toes curling in a pleasing bend. A princely figure holding a hilted sword approaches a lady whose posture and gesture arrest his movement. The women’s torsos are slender, with sloping hips and low centers of gravity. Their flesh is supple and their ornaments thick and weighty. The figures are carved in high relief – they seem to emerge from the wall, twisting as they struggle or shy away from one another. Whimsical yet also highly refined, these sculptures draw on themes from performing arts and courtly literature, such as the Mattavilāsa Prahasana Stein, Emma Natalya, Constructing Kanchi: City of Infinite Temples. Amsterdam, Amsterdam University Press 2021 doi: 10.5117/9789463729123_ch01

44 

Construc ting K anchi

Illustration 7 Panel 4, Tāṉtōṉṟīśvara Temple, Kanchi, seventh century

Sandstone and the Cit y

45

[‘The Farce of Drunken Sport’], a comedic drama composed in Classical Sanskrit by the renowned king of the Pallava dynasty, Mahendravarman I (ca. 530-680 CE).1 The sculptures are situated on a main thoroughfare that leads to the bustling Ekāmbaranātha temple, a sacred monument that is mentioned in the Mattavilāsa Prahasana. Through their location in real space and their representation of real character types that were present in the streets of Pallava-era Kanchi, these sculptures straddle the multiple worlds at play in a city where the courtly and the urban intersect. This chapter examines the shaping of Kanchi as a cosmopolitan courtly capital in the seventh through ninth century – the period in which the city was reimagined as a sacred and administrative center. Although the Pallavas had resided in Kanchi for nearly half a millennium, little is known of the city’s physical layout. Sources for the early history of the Pallavas are primarily limited to royal birudas (‘titles of praise’) inscribed on slabs of stone or copper plates, rather than archaeological remains.2 These texts have enabled scholars to determine a plausible succession of kings, but the precise date of ascension to the throne for each ruler remains debatable. Throughought the Pallava period, inscriptions are dated not with a calendrical year but with the regnal year of the ruling king. Often, scholars have turned to comparison with more robust inscriptions from neighboring dynasties to establish dates for the Pallava kings.3 The seventh and eighth century represent a moment of self-presentation for the Pallavas in Tamil Nadu. After the sixth century, they concentrated their power entirely in northern Tamil Nadu (Toṇṭaimaṇṭalam), leaving their earlier residence of Andhra Pradesh (Āndhradeśa) – they ceased to mark their presence in Andhra with inscriptions and instead began incorporating Tamil language into their inscriptions in Toṇṭaimaṇṭalam.4 In and around Kanchi, kings and key members of the royal family began constructing prominent sacred and municipal institutions designed for specific communities and types of activities that contributed to their capital 1 A. Vishnu Bhatt and Michael Lockwood, Mattavilāsa Prahasana: ‘The Farce of Drunken Sport’ by King Mahendravarma Pallava (Madras: Christian Literature Society, 1981), 101. 2 K.R. Srinivasan, Cave-Temples of the Pallavas (Delhi: Archaeological Survey of India, 1964), 110-111 and 32-34; T.V. Mahalingam, Kāñcīpuram in Early South Indian History; Inscriptions of the Pallavas (Delhi: Agam Prakashan, 1988); Valérie Gillet, La Création d’une Iconographie Śivaïte Narrative: Incarnations du Dieu dans les Temples Pallava Construits, 23-28. 3 Valérie Gillet, La Création d’une Iconographie Śivaïte Narrative: Incarnations du Dieu dans les Temples Pallava Construits, 23. 4 Valérie Gillet, ‘Pallavas and Buddhism: Interactions and Influences’, in Sivasri: Perspectives in Indian Archaeology, Art and Culture, ed. D. Dayalan (2013), 107.

46 

Construc ting K anchi

city’s cosmopolitan life. They built temples that rivaled India’s greatest monuments in scale and architectural complexity, and they fashioned themselves as great patrons of the arts who authored or supported literary production in the elite courtly language of Sanskrit. In doing so, the Pallavas transformed Kanchi into an expansive urban center marked by the towers of many sandstone temples. This chapter’s three sections each examine Kanchi during a different historical moment in this critical period of change. The first section looks at Kanchi in the seventh century, the time that brought about the birth of stone, sacred architecture in South India. The second section maps out Kanchi’s eighth-century monuments in order to trace the city’s original geographical expanse. The third section moves into the ninth century, when temple construction burgeoned seemingly everywhere but in Kanchi. Although the city was an important place well before the seventh century and has endured until the present, Kanchi’s urban and sacred geography was primarily established in the eighth century. It was this Pallava city that established new patterns for royal town planning – patterns that would set the core framework for the city’s continued expansion.

From Brick to Stone (the Seventh Century) The sacred architecture of pre-eighth-century Tamil Nadu was primarily composed of open-air spaces or buildings made of perishable materials, such as timber and brick. We learn of their presence only through literary descriptions, a limited corpus of archaeological remains, and conspicuous absences in the built environment. By contrast, stone held mortuary associations, used in burial sites from as early as 4,000 BCE.5 Stone cairn circles and commemorative monuments known as hero stones are described in early Tamil literature and can still be found throughout the Tamil countryside.6 The Perumpāṇāṟṟuppaṭai, a second-century Tamil poem, attests to a practice 5 K.R. Srinivasan, ‘Āndhras, Ikṣvākus, and Literary Sources’, in Encyclopaedia of Indian Temple Architecture, ed. Michael W. Meister and M.A. Dhaky (New Delhi: American Institute of Indian Studies, 1983), 11. Stone circles and two dolmons dating to ca. 4000-1000 BCE were found at Janakipuram, 40 kilometers southeast of Kanchi between Uttiramērūr and Māmallapuram. Hema Achythan et al., ‘A Reconnaissance Study of Tafoni Development, Exfoliation, and Granular Disintegration of Natural and Artificual Rock Surfaces in the Coastal and Lowland Regions of Tamil Nadu, Southern India’, Zeitschrift für Geomorphologie 54, no. 4 (2010): 498. 6 Akananuru – the Akam Four Hundred, trans. A. Dakshinamurthy, 3 vols. (Thiruchirapalli: Bharathidasan University, 1999), Poem 53. An extant example is the burial site at Tiruppūr.

Sandstone and the Cit y

47

of large-scale construction in brick, which it says was used to build Kanchi’s city wall.7 However, Kanchi’s fluctuating boundaries and the absence of any remains suggest the wall was only a literary trope. James Heitzman has persuasively argued that Kanchi grew incrementally, and my research further demonstrates that villages were urbanized into Kanchi’s domain over the course of the eighth through the thirteenth century.8 Walls would have limited the city’s shifting footprint and fluctuating borders, and there is no local tradition of the city having ever had walls. Nevertheless, the inclusion of a detailed mention of brick construction in the Perumpāṇāṟṟuppaṭai is striking because it contrasts with a description of Madurai’s wall, which is said to be made from slabs of stone.9 An inscription on a set of copper plates found in the village of Kūram, near Kanchi, records the designation of land for a brick kiln.10 Inscriptions on temples throughout Tamil Nadu contain mentions of earlier brick shrines that were subsequently rebuilt in stone. For example, a Chola-period inscription in Mēlappaḻuvūr states that, having become neglected and abandoned, the brick temple was revitalized through conversion to stone.11 While northern and western India have far longer histories of rockcut and structural (constructed) temples, stone, sacred architecture was not introduced in South India until comparatively late. Cave-temples in Orissa and the western Deccan date as early as the second century BCE, and structural temples farther north to the fifth century CE. The earliest cave-temples in Tamil Nadu were not made until around the turn of the seventh century, and structural temples nearly a century later.12 In addition to stone’s use in burial contexts, Tamil Nadu’s natural landscape may have factored in the late introduction of sacred architecture made of stone. Unlike the western Deccan or Orissa, peninsular India is not home to natural caves. The local granite is especially hard, and its crystalline composition makes it exceedingly difficult to carve, even using modern tools.13 Rather 7 Tamil Poetry through the Ages: Pattuppāṭṭu, [‘Ten Idylls’], trans. A. Rajarathnam, 2 vols., vol. 2 (Chennai: Institute of Asian Studies, 2000), 184. 8 Heitzman, ‘Urbanization and Political Economy in Early South India: Kāñcīpuram During the Cōḻa Period’, 128. For a broader geographic scope, see Heitzman, The City in South Asia. 9 Tamil Poetry through the Ages: Pattuppāṭṭu, [‘Ten Idylls’], 169. 10 SII Vol. 1, p. 147-155, no. 51. The earliest surviving Pallava-era structural temple is situated in Kūram. Srinivasan, ‘Pallavas of Kāñcī: Phase I’, 44. 11 ARE 1924, no. 393. 12 A more definitive chronology cannot be ascertained, because Pallava inscriptions are dated to the regnal years of the ruling king, rather than a date within a calendrical era. 13 Degree of hardness is a characteristic of stone that affects workability and a sculpture’s outcome. Peter Rockwell, The Art of Stoneworking: A Reference Guide (Cambridge: Cambridge

48 

Construc ting K anchi

Illustration 8 Lakṣita cave-temple, Maṇṭakappaṭṭu, ca. 580-630 CE

than spalling (separating off in layers) like the Deccan basalt at the caves of Ellora and Ajanta, or being easily workable like the sandstone from which Orissa’s Udayagiri and Khandagiri caves are hewn or the Great Stūpa at Sanchi is constructed, Tamil Nadu’s granite, once broken, separates in uneven fragments that stubbornly resist attempts at control.14 The most notable reference to physical construction in South India is in an inscription on a pilaster on the façade of what is likely the region’s very first cave-temple in which the Pallava king Mahendravarman I (ca. 580-630 CE) famously proclaims something like a conquest of material (Ill. 8).15 No longer reserved for funerary purposes, stone would thenceforth become a University Press, 1993), 17-19. 14 I am grateful to many stone sculptors in Māmallapuram, whose deep knowledge of material is invaluable. Special thanks go to Mallai Dilip for continued conversations and Vijaya Jothi for a particularly informative discussion about quarries (January 2014). For the names of stones and their properties, see S. Gangopadhyay, ‘Monuments of India Built through Ages with Special Reference to the Construction Material’, in Proceedings of an International Symposium Organized by the Greek National Group of Iaeg: The Engineering Geology of Ancient Works, Monuments and Historical Sites, Preservation and Protection, ed. Paul G. Marinos and George C. Koukis (Athens: A.A. Balkema, 1998), 585-586. See also Richard Newman, The Stone Sculpture of India: A Study of the Materials Used by Indian Sculptors from ca. 2nd Century BC to the 16th Century (Cambridge: Harvard University Art Museums, 1984). 15 Dating of Pallava rulers is approximate because of the convention of dating inscriptions using the regnal year of the ruling kind. I generally use dates for rulers given by T.V. Mahalingam

Sandstone and the Cit y

49

suitable and even desirable material with which to build temples. For he, who is praised as vicitracitta (‘brilliant-minded‘), had made a home for Brahmā, Shiva, and Vishnu without the use of brick, timber, metal, or mortar.16 With stone established as a sacred material, Mahendravarman I and his immediate successors consolidated their power in places with plentiful resources of stone, and they marked their territory with sacred edifices. Cave-temples survive across a region spanning Māmaṇṭūr in northern Tamil Nadu to Tiruchirappalli in the South.17 Often situated on the banks of rivers or reservoirs, these sites were excavated from large boulders or hills of stone.18 In addition to cave-temples, the world-famous Pallava monuments at Māmallapuram also include monolithic shrines that were hewn from natural ridges of stone. Relatively flat as it is, Kanchi’s landscape was not conducive to rock-cut shrines. However, by the time the Pallava king Narasiṃhavarman II (ca. 690-728 CE), popularly known as Rājasiṃha, sponsored the construction of his renowned Kailāsanātha temple, Kanchi was already a place with many local shrines. The Tamil and Sanskrit literature and inscriptions that attest to Kanchi’s ancientness also describe the city as a place with a diverse religious landscape.19 The shrines to these various gods must have been built of brick, not stone. It was only in the late seventh century, under the reign of King Parameśvaravarman (ca. 672-700 CE) that architects developed techniques for building structural temples. The earliest structural temple to survive in Tamil Nadu is in the village of Kūram, eight kilometers to the north of Kanchi.20 Builders in Kanchi were quick to take up the practice. During the eighth through thirteenth century, open-air shrines and temples in Tamil Nadu made of brick were converted to stone – sometimes through a single renovation, but more often incrementally. From their inception, temples were frequently constructed with a stone foundation and brick walls. The walls and roof of the main sanctum were usually the f irst to be converted. Renovation of the maṇḍapa (‘pillared hall in in his Kāñcīpuram in Early South Indian History and A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District. 16 Rao Bahadur H. Krishna Sastri, ed., Epigraphia Indica, Vol. 17 (New Delhi: Archaeological Survey of India, 1923), p. 14, n. 5. 17 For a comprehensive survey, see K.R. Srinivasan, Cave-Temples of the Pallavas. 18 Minakshi points out the association of Mahendravarman’s cave-temple sites with rivers and tanks. C. Minakshi, Administration and Social Life Under the Pallavas (Madras: University of Madras, 1938), 96. 19 Pattuppāṭṭu, 184; Maṇimēkalai. See my Introduction for discussion, p. 6-7. 20 Srinivasan, ‘Pallavas of Kāñcī: Phase I’, 44 and Pl. 50.

50 

Construc ting K anchi

Illustration 9 Brick shrine in quarry area, Ārpākkam

front of the sanctum’) might wait for future sponsorship, which was often achieved through collective donations. Inscriptions such as those found in the Mucukundeśvara temple at Koṭumpāḷūr tell of donations of individual pillars for a maṇḍapa’s reconstruction. As Leslie Orr suggests, this points to a gradual conversion from brick to stone.21 Since brick deteriorates over time and stone does not, if a temple’s maṇḍapa is not rebuilt, often the stone basement is all that survives in front of the sanctum. Alternatively, when the maṇḍapa’s walls have been converted from brick to stone, inscriptions on the basement moldings often predate the walls and superstructure, which were built only later. The practice of building temples with brick and timber did not end when the fashion for stone, sacred architecture emerged. Throughout the period of transition from perishable to durable materials, temple builders continued to use brick both exclusively and in combination with stone – a material selection process that was determined by preference and practical necessity. This occurred even at the elite level. The ninth-century temples at Kālampākkam and Uttiramērūr have brick walls above granite foundations. Shrines made of brick or diverse materials with indeterminate dates still stand in the less urbanized areas of Kanchi and in many villages, even in places with plentiful stone quarries, such as Ārpākkam (Ill. 9). The prosperous port town 21 Leslie Orr, Archaeology of Bhakti III workshop presentation (August 2015).

Sandstone and the Cit y

51

of Kāvirippūmpatṭiṉam contained a Chola-era brick temple, dated between the tenth and twelfth century.22 Although stone had become highly coveted, brick remained an affordable material for construction and reparation that was available to a wide range of communities. Stone, which was difficult to quarry and expensive to employ, rapidly became the hallmark of the elite. The King’s Hut: Ekāmbaranātha Temple Today the Ekāmbaranātha temple is Kanchi’s vastest complex. Its main shrine – itself a tangle of vestibules and tiny chambers – is surrounded by no less than five prākāra-enclosures. However, it likely began as a simple shrine made of brick. As recounted in the Periya Purāṇam, the twelfthcentury hagiographic anthology of the Tamil Saints, Ekāmbaranātha is the place where Umā worshipped Shiva in the form of a liṅga made of sand.23 Responding to the warm embrace of his lover’s bosom, Shiva manifested himself in bodily form, and the divine couple married on the spot, beneath a sacred mango tree. A tree shrine within the complex marks the site of the legendary event.24 Because tradition holds that the temple’s main liṅga is the very one Umā made, the priests do not lustrate it, lest the sand should wash away. Although nothing remains of the original shrine, granite pillars in the style associated with the first king Mahendravarman and inscribed with his birudas (‘royal titles of praise’) were recovered from a ruined maṇḍapa within the complex in 1920.25 The pillars were taken off-site and placed on display at the Madras Government Museum, where they remain.26 Similar pillars were subsequently reported lying among the debris from a ruined brick maṇḍapa.27 This brick structure supported by granite pillars may have 22 IAR 1970-1971, p. 35. 23 Periya Purāṇam, line 1130-1146. 24 Schier suggests that the original name of Ekāmbaranātha points to an early association with a pillar deity (‘kampa’) rather than a mango tree (‘āmra’). Kerstin Schier, ‘The Goddess’s Embrace: Multifaceted Relations at the Ekāmranātha Temple Festival, Kanchipuram’ (PhD diss., University of Oslo, 2012), 37. I would like to thank Kerstin Schier for sharing her completed doctoral thesis with me at the final stages of my own. Her thesis has since been published as Kerstin Schier, ‘The Goddess’s Embrace: Multifaceted Relations at the Ekāmranātha Temple Festival, Kanchipuram’ (Wiesbaden: Harrassowitz Verlag, 2018). 25 ARE 1921, no. 82; SII Vol. 12, p. 8, no. 14. 26 The pillars are displayed near the entrance to the main sculpture gallery, with a thirteenthcentury Buddha from the Tiruchirappalli area in their midst. 27 K.R. Srinivasan, ‘The Pallava Architecture of South India’, Ancient India: Bulletin of the Archaeological Survey of India 14 (1958): 131.

52 

Construc ting K anchi

Illustration 10 Ekāmbaranātha Temple, Kanchi, Left to right: Roof of Kacci Mayāṉam shrine (tenth century), Pillared Hall (twentieth century),

Gateway (seventeenth century), Vṛṣabheśvara Shrine (ninth century), Gateway (sixteenth century)

been the earliest iteration of the Ekāmbaranātha temple, the shrine called Ēkampam in the poetic hymns of Saints Sambandar, Appar, and Sundarar.28 The Ekāmbaranātha temple complex as it appears today was not constructed at a unified moment. Rather, as can be seen from the range of architectural styles, materials, and recycled components, the site developed in a piecemeal fashion, with shrines converted, expanded, and rebuilt continuously over an incredibly long period of time (Ill. 10). Although it now contains later additions, the eastern portion of the complex appears to be an older subsection.29 This area includes a shrine called Kacci Mayāṉam, an ancient tree, the larger of the temple’s two ablutions tanks, and a temple called Vṛṣabheśvara that dates to the latter part of the Pallava period.30 28 Tēvāram hymns 1.122, 2.12, 3.41, 3.114, 4.7, 4.44, 4.99, 5.47, 5.48, 6.64, 6.65, 7.61. All hymn numbers refer to ‘Digital Tēvāram/Kaṇiṉit Tēvāram [Cd-Rom]’, ed. Jean-Luc Chevillard and S.A.S. Sharma, Collection Indologie n. 103 (Pondicherry: École Française d’Extrême-Orient/Institut Français de Pondichéry, 2007). www.ifpindia.org/digitaldb/site/digital_tevaram/INDEX.HTM 29 The hypothesis about Ekāmbaranātha’s original location emerged during a conversation with A. Prabu and Raja Sharma while we were at the Ekāmbaranātha temple (July 2014). 30 The Vṛṣabheśvara temple is called Vālīśvara in Srinivasan, ‘Pallavas of Kāñcī: Phase I’, 67 and Pl. 44. My thanks to Jonas Buchholz for the correction. I discuss Vṛṣabheśvara’s dating in

Sandstone and the Cit y

53

Illustration 11 (Vṛṣabheśvara Shrine at far right) Archaeological Survey of India, A view of the tank from the east, Ekambreswaraswami Temple, Conjeevaram, 1897, photographic print, 21.1 × 25.6 cm, British Library, Photo1008/3(325)

Although today the Vṛṣabheśvara is mostly encased in a thick layer of cement, an archival photograph reveals more of the Pallava-era sandstone construction (Ill. 11). The archaeological report of the Mahendravarman-era pillars only tells us that the ruined maṇḍapa was located near the thousand-pillar hall. This hall stands to the south of the main temple, and the maṇḍapa may have been east of it, near a small tank that is no longer in use. If so, it would have been situated within the older zone. The greatest concentration of reused architectural materials is also found within this area, primarily in the form of sandstone blocks that are embedded in the prākāra wall, and granite slabs that have been repurposed as paving stones. Some of the recycled elements carry inscriptions or relief sculptures. A sandstone shrine supported by previously unnoticed lion-based pillars characteristic of the Pallava period detail in subsequent pages.

54 

Construc ting K anchi

is also on the eastern side of the temple. This crumbling structure extends from the northern end of the eastern exterior prākāra into the street, and its sanctum is sunken deep below ground level. The sprawling expanse of Ekāmbaranātha, together with its great variety of structural components, suggests that the full area was originally a loose cluster of shrines only later grouped together into a single complex. In the process, some structures lost their original functions and meanings, while others gained new identities and assumed new purposes. Most of the large complexes in Kanchi seem to have followed a similar trajectory. However, Ekāmbaranātha’s presence in Pallava-era literature brings its early history more sharply into focus. Literary Temples and Esoteric Sects The very same king whose royal birudas are inscribed on a pillar that once stood at Ekāmbaranātha, and who is credited with conquering the medium of stone, was also a great contributor to the literary arts. The first Mahendravarman authored the one-act Sanskrit dance drama, the Mattavilāsa Prahasana, written in the early seventh century and set in courtly Kanchi. The play mentions specific temples and identifiable places that begin to map out the structure of the early city and to designate areas for particular communities and types of activity. The Mattavilāsa also gives us access to the range of participants who were active in the flourishing religious and cultural landscape of the city at that time. The particular ways in which the text parodies the characters reveal the author’s deep familiarity with each tradition, and satirical intertextual references reveal his sophisticated literary knowledge.31 The Mattavilāsa centers on characters belonging to a tantric sect of Shaiva ascetics, called Kāpālikas (‘skull-bearers’).32 In the play, a drunken Kāpālika and his Kāpālinī-consort are looking for the Kāpālika’s missing bowl, made of a human skull, which he lost somewhere in the streets of Kanchi. While on 31 Mahendravarman is generally given credit for inventing the genre of farce; Anne Monius, Imagining a Place for Buddhism: Literary Culture and Religious Community in Tamil-Speaking South India (New York: Oxford University press, 2001), 64 and 185 n. 32. For a survey of Prahasanas, see K.K. Malathi Devi, Prahasanas in Sanskrit Literature and Kerala Stage (Delhi: Nag Publishers, 1995). Some scholars believe that Mahendravarman was originally a Jain, in part because a verse in the Periya Purāṇam tells us Appar converted him to Shaivism. Periya Purāṇam, line 1411. The proposed conversion has raised some debate, summarized in Mahalingam, Kāñcīpuram in Early South Indian History, 75. 32 Andrea Acri, ‘Kāpālikas’, in Hinduism and Tribal Religions, ed. Pankaj Jain, Rita Sherma, and Madhu Khanna (Dordrecht: Springer Netherlands, 2020).

Sandstone and the Cit y

55

their search, they encounter a Buddhist, a Pāśupata-Shaiva, and a madman (‘unmattaka’), and they make reference to other communities in the city’s population, including the Jains.33 All of the characters are satired in illuminating ways. ‘Ah, Kanchipuram is incredibly beautiful!’ says the Kāpālika to his lover as they walk. ‘The sound of the mṛdaṅga drums is confused with that of the thundering clouds that rest on the tops of tall temples (‘vimāna śikhara’); the flower garland shops look like pegs in the construction of spring; the sounds of the girdles of wonderful young women proclaim the victory of the flower-arrowed god [Kāma, god of love].’ To which the Kāpālinī replies, ‘Lord! Just like Lady Liquor, Kanchi is indisputably sweet.’34 The Mattavilāsa specifically mentions two Hindu temples that can still be found in Kanchi and must have already existed by the time the play was composed. The first temple is Ekāmbaranātha, which is explicitly named in the play as Ekāmra. This temple is designated as a Kāpālika place when the Buddhist character recognizes the protagonist from his wild appearance and the attributes he carries as ‘that awful Kāpālika who dwells at Ekāmra.’35 I find another connection between the Ekāmbaranātha and a sect of Tantrikas that may be the Kāpālikas in a Tēvāram hymn by Appar.36 The Saint sanctifies Kacci Ēkampam (an older name for Ekāmbaranātha) as a place where devotees practicing extreme austerities at the feet of the Lord smear themselves in ashes and roll on the bare ground. Siva, who has the color of milk and has spreading dreadlocks coiled into a crown and a cobra tied at the waist – The Lord in Ēkampam in Kacci! Your devotees join together with folded hands and smear themselves with sacred ash, rolling on the ground for many days, feeling intense ardor that your feet are their refuge; please take pity on them (and grant your grace).37 33 The Mattavilāsa Prahasana contains one of the earliest accounts of certain aspects of Pāśupata doctrine and conduct, as well as the oldest mention of a Pāśupata individual. Ferstl further suggests that the character of the madman (‘unmattaka’) is also a Pāśupata. Chistian Ferstl, ‘Overlooked Material for the Study of Pāśupata Śaivism: The satirical play Mattavilāsaprahasana’, in Transmission and Tradition. On Quotations and Paraphrases from and Allusions to Ancient Texts on Indian Philosophy, ed. E. Prets and H. Marui (Vienna: VÖAW, 2015), 3 and 9-14. 34 Mattavilāsa Prahasana, line 43-44. 35 Mattavilāsa Prahasana, line 69. 36 Appar, hymn 4.99 37 pāmpu araic cērttip paṭarum caṭaimuṭippāl vaṇṇaṉē – Kacci ekampaṉē! kūmpālaic ceyta karatalattu aṉparkaḷ kūṭi cāmparaip pūci

56 

Construc ting K anchi

A third connection between the Ekāmbaranātha and the Kāpālikas is the Kacci Mayāṉam, a west-facing shrine concealed beneath a twentieth-century pillared maṇḍapa that precedes the main entrance to the Ekāmbaranātha temple.38 The earliest parts of the Kacci Mayāṉam are inscribed walls and floor slabs from the tenth and eleventh century, though the shrine has since been modified abundantly.39 The Periya Purāṇam tells us Appar visited Kacci Mayāṉam at Ēkampam. This means a shrine with this name may have stood on the Ekāmbaranātha’s premises even before the present structure would suggest.40 In Classical Tamil, the designation mayāṉam refers to a cremation ground, which could have been situated in the marshy area to the north of Ekāmbaranātha .41 The temple’s position on the outskirts of Kanchi’s settled area further points to a possible proximity to a cremation ground, and the somewhat remote location would have made the temple attractive to Kāpālikas. Known for wild, transgressive behaviors, such as consuming meat and liquor and dancing in cremation grounds, the Kāpālika’s objective was to overcome the fear of death and the attachment to purity through constant contact with the impure.42 Places conducive to such practices were situated on the peripheries of settlements, far from palaces and market centers. Kacci Mayāṉam and the neighboring Vṛṣabheśvara shrine may have had an original association with death, transgression, and the Kāpālikas. A final – if somewhat confusing – piece of evidence for the Ekāmbaranātha’s association with the Kāpālikas comes from a local belief that inside Kacci Mayāṉam’s sanctum is an icon of a Kāpālika, ‘named Umāsahita’, who carries a khatvaṅga (‘skull-topped staff”), sword, and skull.43 This tradition also holds that Kacci Mayāṉam was an independent temple that was later incorporated into the complex, as I have also suggested. While Kacci Mayāṉam in fact enshrines a liṅga, the Pallava-era Vṛṣabheśvara temple contains an Umāsahita pal nāḷ taraiyil puraṇṭu niṉ tāḷ caraṇ eṉṟu empalippārkaṭku iraṅku(kaṇṭāy) Appar, 4.99.8. 38 The pillared hall was built since the 1897 photograph (See Ill. 79). 39 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 121, Cg.-501; p. 122, Cg.-503; p. 129, Cg.-528; p. 134, Cg.-550. 40 Periya Purāṇam, line 1590. 41 I am grateful to Vijayavenugopal at the École Française d’Extrême-Orient, Pondicherry, for informing me of this terminology during the Classical Tamil Summer Seminar (August 2015). 42 David N. Lorenzen, ‘A Parody of the Kāpālikas in the Mattavilāsa’, in Tantra in Practice, ed. David Gordon White (Princeton: Princeton University Press, 2000), 81-82. For further discussion, see The Kāpālikas and Kālāmukhas: Two Lost Śaivite Sects (New Delhi: Thomson Press, 1972). 43 C. Chandramouli, Census of India 2001: Temples of Tamil Nadu, Kancheepuram District (Tamil Nadu: Directorate of Census Operations, 2001), 140.

Sandstone and the Cit y

57

relief panel on the sanctum’s rear wall. Tradition may have conflated the two shrines because they are only meters apart and they both face west in an otherwise east-facing complex. The tradition nevertheless indicates that there is a lingering association between Kacci Mayāṉam (or perhaps the Ekāmbaranātha temple more generally) and the Kāpālikas. The second temple to appear in the Mattavilāsa is not named and is therefore more difficult to identify. Near the end of the play, the Pāśupata ascetic takes his leave and declares he will go to a temple to await the evening service of a god he describes as bhagavataḥ pūrvasthalīnivāsinaḥ (‘the Lord who dwells in the east’). 44 The somewhat puzzling epithet has been taken to refer to Sūrya (the Sun God). 45 However, this seems unlikely because the scene takes place in the evening, and a Sūrya pūjā would typically take place in the morning. If a specific temple is being referenced here at all, it is probably the Kāyārohaṇeśvara temple. There are three main reasons for this. First, this temple has already been suggested as the main Pāśupata center in Kanchi because of its name. 46 Rajeshwari Ghose points out the association of Pāśupatas with temples named Kāyārohaṇa – a Sanskrit term used in Pāśupata literature to mean the ascent (‘ārohaṇa’) of the body (‘kāya’) of the Pāśupata founder, Lakulīśa, to join the body of Shiva. 47 Venkataranam suggests that Kāyārohaṇeśvara temples are associated with Lakulīśa because he was born in a place called Karvan, originally called Kāyāvarohaṇa or Kāyāvatāra. 48 In Tamil literature, there is a connection between the Pāśupatas and the word kārōṇam, which is used in the Tēvāram as an abbreviated version of kāyārohaṇam.49 The Kāyārohaṇeśvara temple’s local designation, ‘Guru Kōyil’, may be the distant legacy of a reference to the Pāśupatas’ principal guru and founder, Lakulīśa. Kanchi’s ācāryas (‘senior initiated priests’) know Kāyārohaṇeśvara to be an important and ancient

44 Mattavilāsa Prahasana, line 184. There is abundant literature on the Pāśupata sect throughout South and Southeast Asia. For a good introduction, see Andrea Acri, ‘Pāśupatas’, in Hinduism and Tribal Religions, ed. Pankaj Jain, Rita Sherma, and Madhu Khanna (Dordrecht: Springer Netherlands, 2020). 45 Bhatt and Lockwood, Mattavilāsa Prahasana: ‘The Farce of Drunken Sport’ by King Mahendravarma Pallava, 101. Ferstl is uncertain what the epithet refers to, in Ferstl, ‘Overlooked Material for the Study of Pāśupata Śaivism: The satirical play Mattavilāsaprahasana’, 14. 46 Rajeshwari Ghose, The Tyāgarāja Cult in Tamil Nadu: A Study in Conflict and Accomodation (Delhi: Motilal Banarsidas, 1996), 150. 47 Ibid. 48 Venkataraman, Dēvī Kāmākshī in Kāñchī: A Short Historical Study, p. 20 and p. 48, n. 3. 49 I thank Jean-Luc Chevillard for drawing my attention to this connection and providing references (Pondicherry, August 2014).

58 

Construc ting K anchi

place.50 The second reason is the temple’s location. The Kāyārohaṇeśvara temple is situated in an unsavory part of Kanchi, not far from the city’s main cremation grounds and the massive Tāyār Kuḷam, a public bathing area. This is a likely spot for ascetic activity. The third reason is that the temple has the right date for an early cult center in Kanchi. Although the temple structure dates to the Chola period, probably ca. eleventh century, a relief depicting Somāskanda (Shiva and Umā seated with their son Skanda) within the shrine predates the architecture by approximately two centuries. This icon closely compares with a lintel relief at the ninth-century Jalanātheśvara temple in Takkōlam. The name, architecture, and location together suggest that Kāyārohaṇeśvara is a sacred site of significant age, and that it once may have been the place for Kanchi’s Pāśupata community. A third sacred structure that is mentioned in the Mattavilāsa is a royally sponsored Buddhist monastery (‘rājavihāra’).51 Although no indication is given of its name or location, some scholars have identified it with a monastery and stupa mentioned by the Chinese pilgrim Xuanzang, said to be in the southwest part of the city.52 However, several factors suggest the Mattavilāsa’s royal monastery instead occupied the site that is now the Kāmākṣī Ammaṉ temple. Prior to 1915, three large-scale Buddha statues were discovered inside the innermost prākāras and the garden at the Kāmākṣī Ammaṉ temple.53 One of these Buddhas stands more than seven feet tall and dates to as early as the sixth century, while the other two are seated figures measuring over three and five feet each, their appearance consistent with other Cholaperiod examples.54 Excavations carried out inside the temple complex and its immediate vicinity in the 1970s yielded the remains of circular brick basement moldings, which are similar to brick stūpa structures unearthed at Nagapattinam and Kāvirippūmpatṭiṉam that date to between the first century BCE and fourth century CE.55 Unearthed from the same area was a ceramic fragment carrying five Brahmi characters that are paleographically datable to the first or second century CE. The archaeological report suggests 50 My sincere thanks to Nagaswamy Aiyyer at Ekāmbaranātha temple for first sending me to Kāyārohaṇeśvara (October 2013). 51 Mattavilāsa Prahasana, line 64. 52 R. Champakalakshmi, Religion, Tradition, and Ideology: Pre-colonial South India (New Delhi: Oxford University Press, 2011), 342. 53 T.A. Gopinatha Rao, ‘Bauddha Vestiges in Kanchipura’, Indian Antiquary 44 (June 1915). The largest Buddha is on display in the Madras Government Museum. 54 Rao mentions local knowledge that there are two more similar Buddha statues buried in the temple garden. Ibid., 129. 55 IAR 1969-1970, p. 35; IAR 1973-1974, p. 24.

Sandstone and the Cit y

59

Illustration 12 Panel 1-4 (and see Illustration 7), Tāṉtōṉṟīśvara Temple, Kanchi, panels seventh century

the inscription represents the name of a Buddhist monk, Putalatisa.56 The presence of the standing Buddha alone was enough to lead Vidya Dehejia and other scholars to conclude that the Kāmākṣī Ammaṉ temple was originally a Buddhist shrine, perhaps dedicated to Tārā or Vasudhārā, which was later taken over by Hindus.57 Remains of Buddhist statuary scattered throughout Kanchi indicate that the Mattavilāsa could refer to any one of several Buddhist places in the city (see Chapter Three). However, the scale and antiquity of the Buddhas found on the premises strongly suggests that the Kāmākṣī Ammaṉ temple’s previous incarnation was the location intended in the play. A Farce of Drunken Sport: Tāṉtōṉṟīśvara Temple The Mattavilāsa Prahasana’s inclusion of real places and real communities in Kanchi makes it tempting to interpret the sculptural programs on some of Kanchi’s Pallava temples as representations of the play. Indeed, the set of seven sandstone relief carvings at the Tāṉtōṉṟīśvara temple, mentioned 56 Ibid. 57 Vidya Dehejia, ‘The Persistence of Buddhism in Tamil Nadu’, in A Pot-Pourri of Indian Art, ed. Pratapaditya Pal (Bombay: Marg, 1988), 58.

60 

Construc ting K anchi

at the beginning of this chapter, are locally believed to represent scenes from the Mattavilāsa (Ill. 12).58 Although the characters in the reliefs do not correspond precisely to those in the play, there are intriguing moments of intersection between text and image. Transgressive behavior, such as drunkenness and cavorting, is central to both. Close inspection of the panels reveals that the oblong objects the characters hold while they frolic are liquor skeins (mostly empty) and the vessels are kamaṇḍalus – water pots that equally could be used for carrying alcohol. In panel 7, a drunken couple dances playfully. The man holds aloft a kamaṇḍalu in his right hand and curls his left arm around his lady’s shoulders. His hand rests on hers where she grasps one end of the skein draped behind her shoulders. Her ringlet curls frame her face as she cocks her head to the side, casting her lover a sidelong glance. Where their legs cross, the two lovers connect ecstatically. The reliefs are now affixed to the exterior walls of the mahāmaṇḍapa (‘front entrance hall’) at the Tāṉtōṉṟīśvara temple, a heavily renovated structure along the main street that leads to the Ekāmbaranātha temple. The panels are jammed low into a wall of red tiling in a graceless manner that obscures their original purpose as much as it does their ornamental borders. A sandstone upapīṭha (‘platform’) exposed at the base of the tiled wall indicates a Pallava date for the original structure, and photographs taken prior to the tiling confirm the wall itself is also made of sandstone and brick. Although the panels presently appear as unusually placed exterior niche figures, the scale and subject matter suggest that they originally formed a narrative. Similar panels can be found in situ along the prākāra walls at the Shore temple in Māmallapuram and the Vaikuṇṭha Perumāḷ temple in Kanchi. While the Tāṉtōṉṟīśvara panels could have been transported from a nearby site, it is more likely that they were originally situated in a former prākāra at this very building. A relief panel of similar size and proportion is housed inside the Bhadrakāḷī Ammaṉ temple, where the two figures it depicts are worshipped as a local god and goddess. The panel has been built into a wall inside the ardhamaṇḍapa (‘vestibule’). The ribbed ghaṭa (‘capital’) of a Pallava-style sandstone pillar is presently being used as the temple’s balipīṭha (‘offering pedestal’), and there are several other small sculptures within the temple that may well be Pallava in date. Besides the fragments, Bhadrakāḷī Ammaṉ is entirely a modern structure. However, additional factors suggest the site has a long history as a temple. First, whereas modern goddess temples in Kanchi are generally built facing east, Bhadrakāḷī Ammaṉ faces north, as 58 Local knowledge in Kanchi, and Pradeep Chakravarthy, ‘Era of the Opera’, The Hindu 13 February 2011.

Sandstone and the Cit y

61

is appropriate to the goddess in ritual texts and is true of Kanchi’s early goddess temples, such as the two Karukkiṉil Amarntavaḷ Ammaṉ temples. Second, the temple is situated on a street of the same name, ‘Bhadrakāḷī Ammaṉ Kōyil Street.’ Ancient sites in the city tend to employ this system of street names – Ekāmbaranātha is at the end of Ekāmbaranātha Sannathi (‘shrine’) Street, Vaikuṇṭha Perumāḷ is at the end of Vaikuṇṭha Perumāḷ Sannathi Street, and so forth. The connection in name suggests a coordination between the temple and the street that leads to it. In the reliefs at the Tāṉtōṉṟīśvara and Bhadrakāḷī Ammaṉ temples, the characters are theatrical. While their movements are the practiced gestures of the courtly world, the figures do not have the comportment and attire of the kings and gods that grace the walls of the royally sponsored Pallava monuments.59 Neither are they the anonymous attendants that make up the celestial supporting cast. Rather, they seem most closely to represent a range of individuals who would be classified among the marginal character types in the Kāma Sūtra and Artha Śāstra, India’s most important texts about courtly conduct and politics. Such antinomian figures include dancers, actors, musicians, itinerant merchants, mendicant monks, and various sorts of ascetics and attendants. These figures occupy liminal positions in society. They are able to transgress class and gender expectations, to cross-dress or undress, to disobey rules, to go where they should not go, and to undertake prohibited activities such as eating meat, drinking alcohol, and cavorting with members of the opposite sex. If employed with some degree of caution, they could therefore be of great advantage to persons in positions of power – particularly as spies, go-betweens, or even wielders of magic. In addition to their literary representations, marginal character types are peppered throughout South and Southeast Asian temple sculpture. They are among the erotic figures on the moldings of the jagatī (‘foundation platform’) at Khajuraho’s world-famous Lakṣmaṇa temple, and within the quotidian Karmavibhaṅga scenes that encircle the lowest level of Candi Borobudur, an important Buddhist monument in Central Java, Indonesia. Andrea Acri has identified figures of Brahmins and ascetics depicted as sūtradhāras (‘stage directors’) and gurus of dance in the upper tiers of Borobudur and in Candi Shiva at Prambanan.60 He has further proposed that a troupe of 59 Daud Ali explores courtly bodily comportment in Daud Ali, ‘Aristocratic Body Techniques in Early Medieval India,’ in Rethinking a Millennium: Perspectives on Indian History from the Eighth to the Eighteenth Century, Essays for Harbans Mukhia, ed. Rajat Datta (Delhi: Aakar Books, 2008). 60 Andrea Acri, ‘Birds, Bards, Buffoons and Brahmans: (Re-)Tracing the Indic Roots of Some Ancient and Modern Performing Characters from Java and Bali’, Archipel 88 (2014).

62 

Construc ting K anchi

Illustration 13 Pillar base with dancing figures, Vīraṭṭāṉeśvara Temple, Kanchi, seventh century

Balinese ascetic dancers called Caṇṭang Balungs, ‘the rattling bones’, has Indic roots in a group of itinerant Kāpālikas who became institutionalized overseas as performers.61 The Tāṉtōṉṟīśvara figures seem to belong to the antinomian character type. At the Tāṉtōṉṟīśvara and Bhadrakāḷī Ammaṉ temples, the absence of firm iconography or inscriptions leaves dating to stylistic comparison. Although the reliefs are carved from the same sandstone as all the Pallava temples in Kanchi, I find closer parallels with certain sculptures from Māmallapuram. Most similar are the side panels of the Mahiṣāsuramardinī cave-temple. In particular, at Tāṉtōṉṟīśvara, Bhadrakāḷī Ammaṉ, and the Mahiṣāsuramardinī cave-temple, faces are oblong and oval rather than square-shaped, carving is in deep relief, and bodies twist dramatically away from the background. The women have long, slender torsos with low centers of gravity and garments such as strapless breast-bands. They do not wear a sacred thread, and their ornaments include large plug-disk earrings that show the stretched earlobe wrapping around. The men wear a twisted sacred thread, no anklets, and similar earrings to the women. These features can also be found in pillar 61 Ibid., 37-38.

Sandstone and the Cit y

63

fragments carved with figures of dancers, which are now on the premises of the Kacchapeśvara and Vīraṭṭāṉeśvara temples in Kanchi (Ill. 13). These stylistic elements all conform to what Michael Lockwood has identified as typical of the seventh-century Pallava style.62 By contrast, at Kailāsanātha and other post-seventh-century Pallava temples, the faces are square-shaped and torsos shorter. The women’s breasts are placed higher on the body, their bands have shoulder straps, and they wear sacred threads and multiple anklets. Men also wear anklets and multiple sacred threads. Based on his careful stylistic analysis, Lockwood attributes the side panels at the Mahiṣāsuramardinī cave-temple to the period of Rājasiṃha’s immediate predecessor, Parameśvaravarman I (ca. 672-700 CE).63 A similar date would make the Tāṉtōṉṟīśvara and Bhadrakāḷī Ammaṉ reliefs, together with the Kacchapeśvara and Vīraṭṭāṉeśvara pillar bases, among the earliest Pallava remains in all of Kanchi. This group of fragments would have been crafted by expert artisans just after the Mattavilāsa was composed. The reliefs and the drama were animated by the same courtly world of Pallava-Kanchi at the birth of art in stone.

Sandstone Temples in the City (the Eighth Century) During the seventh century, the Pallava kings had sponsored Tamil Nadu’s very first cave-temples and rock-cut, monolithic temples. In doing so, they had moved stone architecture beyond the mortuary associations it had traditionally held in the region and established it as an elite material. Beginning at the turn of the eighth century, under the reign of King Rājasiṃha (ca. 700-725 CE), experimentations with architecture soon yielded unprecedently large temples that were built with blocks of local stone (Ill. 14). At Māmallapuram, the Shore temple’s basement moldings are carved into the living rock, while the rest of its structure is built from the plentiful charnockite in the area. Paṉamalai’s Tāḷagirīśvara temple is granite quarried from the hill on which it stands. Kanchi’s Kailāsanatha temple is made primarily of local sandstone. Temples built within Kanchi during Rājasiṃha’s reign share certain physical similarities that define a loosely grouped architectural style.64 Signif icantly, they are all made of sandstone with a light gray granite 62 Lockwood, Pallava Art, 15-16. 63 Lockwood, Pallava Art, 15-16. 64 A.H. Longhurst, ‘Pallava Architecture, Part II (Intermediate or Māmalla Period)’, Memoirs of the Archaeological Survey of India 33 (1928); ‘Pallava Architecture, Part III (the Later or Rājasiṃha Period)’, Memoirs of the Archaeological Survey of India 40 (1930); Srinivasan, ‘Pallavas of Kāñcī: Phase I.’; ‘Pallavas of Kāñcī: Phase II.’

64 

Construc ting K anchi

Illustration 14 Kailāsanātha Temple, Kanchi, ca. 700-725 CE

selectively incorporated. Fragments of Pallava temples that have become disengaged from their original contexts are now peppered throughout the city and its surrounds. They can be recognized through comparison with the Kailāsanātha and the other surviving Pallava monuments. By far the most frequently encountered Pallava-era elements are sandstone pillars with a shaft supported by a lion that is seated in stately elegance (Ill. 15). These lion-based (‘siṁhapāda’) pillars were used so frequently that they almost acquire the status of an emblem of Pallava construction. In addition to the pillars in known Pallava-era temples, I have identified fragments of such lion-based pillars in many other temples, either built into structures or lying loose in the compound. Temples with decontextualized lion-based pillars include Kacchapeśvara, Māṇḍukaṇṇīśvara, Rudrakoṭīśvara, Varadarāja Perumāḷ, Ekāmbaranātha, and Karukkiṉil Amarntavaḷ Ammaṉ in Kanchi, and the Narasiṃha Perumāḷ temple in nearby Cevilimēṭu. In many of these places, the lion-based pillar fragment is the sole surviving vestige of a history that extends back farther than the existing temple structure. The fragments are sometimes positioned prominently, as if to attest to the antiquity of the site.65 The lion-based pillars 65 It is not beyond the realm of possibility that some of the loose siṃhapāda pillars were brought from elsewhere and placed in their current locations. However, the use of local sandstone indicates that they originate in the immediate area. Pillars have been used as symbols of victory

Sandstone and the Cit y

65

Illustration 15 Lion-based pillars in prākāra, Vaikuṇṭha Perumāḷ Temple, Kanchi, ca. 730-795 CE

seem to have held a level of prestige that made them worthy of preservation even when the rest of the structure had deteriorated beyond repair. In addition to these pillars, I also attribute to the Pallava period several decontextualized full sculptures and panels bearing relief carvings. These include a sandstone figure of Ganesha at the Vahnīśvara temple near Okkappiṟantāṉ Kuḷam, a loose figure of a gaṇa (‘dwarf’) formerly at Tāṉtōṉṟīśvara, several sculptural fragments in the courtyard at one Karukkiṉil Amarntavaḷ Ammaṉ temple, and pillar capitals built into the floor at the other Karukkiṉil Amarntavaḷ Ammaṉ temple. These fragments are in addition to the granite pillar bases at the Kacchapeśvara and Vīraṭṭāṉeśvara temples, the single relief and other fragments at Bhadrakāḷī Ammaṉ, and the Tāṉtōṉṟīśvara’s beautiful set of reliefs. Like the lion-based pillars, these remains are fairly uniform in artistic style, as well as material. They are all made of sandstone, with only a few exceptions in the light gray granite that also appears in full Pallava-era temples. When figures are depicted, they display sensuous yet powerful figures with long, tapering limbs and a balance of exuberance and restraint.

and conquest throughout Indian history. Finbarr Barry Flood, ‘Pillars, Palimpsests and Princely Practices: Translating the Past in Sultanate Delhi’, RES 43 (2003).

66 

Construc ting K anchi

When placed on a map, it becomes clear that nearly all of the Pallava-era temples and all of the lion-based pillars and other fragments in Kanchi that date to the late seventh century or first half of the eighth century are situated in what is now the western part of the city. The only outlying Pallava-era fragments are the pillar bases at Vīraṭṭāṉeśvara, the lone lion-based pillar at Varadarāja Perumāḷ, and a Ganesha identified by Schier at the Āti Kāmākshi Ammaṉ temple.66 Given Varadarāja Perumāḷ’s importance as a Vaishnava center, it is unsurprising that a royal emblem should be found from an earlier iteration of the shrine, regardless of the building’s location. Although Vīraṭṭāṉeśvara presents no clear explanation, the temple is relatively near to the other Pallava monuments. If in fact the Āti Kāmākshi Ammaṉ was an early temple dedicated to a wild form of the goddess, as Schier and others suggest, it likely would have occupied a peripheral rather than a central position in the city.67 It is not uncommon to find local temples dedicated to the fierce goddess Kāḷiyammaṉ in less urbanized parts of Kanchi. The full Pallava temples built during Rājasiṃha’s reign include the Kailāsanātha, Iṟavāttāṉeśvara, Piṟavāttāṉeśvara, Airāvateśvara, and Amareśvara. From these temples alone, the contours of the Pallava city begin to emerge.68 Candraprabhā Jiṉālāya, a Jain temple in the relatively independent settlement of Tiruparuttikuṉṟam at the western outskirts of Kanchi, has also been assigned to late in Rājasiṃha’s reign, or soon after.69 It is perhaps more than coincidence that all the temples in Kanchi matching names in the Tēvāram are located on the western side of the city.70 These include Ekāmbaranātha, Oṇakānteśvara, Tirumēṟṟaḷinātha, and Aṉēkataṅkāvatīśvara. 66 Schier also reports that the figure of Vīralakṣmī in the temple’s courtyard is sandstone, and she suggests that fragmentary sculpture including the Mahiṣāsuramardinī and possibly the Ayyaṉār may be of significant age. I have also speculated about these particular sculptures during visits to the site. Schier, ‘The Goddess’s Embrace: Multifaceted relations at the Ekāmranātha temple festival, Kanchipuram’, 148-152. 67 Schier concludes that the Ādi Kāmākṣī Ammaṉ temple was rebuilt on the site of an earlier temple that was likely dedicated to the Goddess. Ibid. 68 Local sources call the area west of the Kacchapeśvara temple along the road to Kailāsanātha ‘Old Kanchi.’ I have been told that there were plans to begin excavations in this area that were never realized, but I have been unable to obtain confirmation (Pers. comm. December 2013). 69 Srinivasan, ‘Pallavas of Kāñcī: Phase I’, 74. Signif icant renovations to this temple make dating uncertain. 70 Schmid draws attention to the challenges of associating temples from the Tēvāram and Divya Prabandham with existing temples. Charlotte Schmid, ‘The Edifice of Bhakti: Towards an “Archaeological” Reading of Tēvāram and Periyapuranam’, in Mapping the Chronology of Bhakti: Milestones, Stepping Stones, and Stumbling Stones, ed. Valérie Gillet (Pondicherry: Institut Français de Pondichéry and École Française d’Extrême-Orient, 2014).

Sandstone and the Cit y

67

The divya deśam (site sanctified by the Vaishnava Saints in the Divya Prabandham) of Pāṇḍavadūta Perumāḷ is also in the western zone.71 The temples bearing names mentioned in early literature but not within this western zone are situated in areas that originally were not part of Kanchi. Tirukkālīśvara temple, extolled by the Shaiva Saint Sambandar, is in the village of Tirukkālimēṭu, which remains only loosely part of the city.72 The divya deśams of Viḷakkoḷi Perumāḷ, Aḻakiya Ciṅka Perumāḷ, Aṣṭabhuja Perumāḷ, Yathoktakārī, and Varadarāja Perumāḷ are all within the area formerly known as Attiyūr, which remained an independent settlement until the Chola period.73 As I will discuss in Chapter Two, temples in outlying areas do not necessarily conform to patterns evident in Kanchi. In the latter half of the eighth century, the establishment of another monumental Pallava temple began to shift the city eastwards. This is the three-storey shrine of Vaikuṇṭha Perumāḷ, the only Pallava temple in the city on the scale of the Kailāsanātha.74 At the time of its construction, the Vaikuṇṭha Perumāḷ temple would have been the tallest building in its area. Towering over its surrounds, the temple’s pinnacle would have oriented approaching travelers and visitors to the city. Even today, the building remains one of the tallest structures on that side of Kanchi. Not long after Vaikuṇṭha Perumāḷ, several additional temples were established, all of which faced west. Mukteśvara and Mataṅgeśvara were positioned to the Vaikuṇṭha Perumāḷ’s immediate north and south. 71 Whereas the divya deśams of Pāṇḍavadūta Perumāḷ and Vaikuṇṭha Perumāḷ occupy their original footprints, others in the city may have been relocated. Venkataraman proposed that each of the five divya deśams in Kanchi that is situated within another temple originally had its own temple. These include Nīlāttiṅgaḷtuṇḍam in the Ekāmbaranātha temple, Kaḷvar/Kaḷvaṉūr in the Kāmākṣī Ammaṉ temple, and Kārakam, Kārvāṉam, and Nīrakam in the Ulakaḷanta Perumāḷ temple. Venkataraman, Dēvī Kāmākshī in Kāñchī: A Short Historical Study, p. 50, n. 16. 72 Sambandar, 3.65 73 The first record to mention Kāñcipurattu-Tiruvattiyūr, meaning ‘Attiyūr, a part of Kanchi’, is as late as the thirteenth century. The inscription is reportedly on the north wall of the second prākāra at the Varadarāja Perumāḷ temple. SII Vol. 4, p. 292-293, no. 859-860. This inscription documents the conjoining of two parts of the city that would later find themselves uneasily divided again. 74 The dating for the Vaikuṇṭha Perumāḷ temple is secure as Pallava temples go. Contributing sources include inscriptions around the basement molding dated to the regnal years of Nandivarman II (second half of the eighth century) and of Dantivarman (ca. 813 CE), a poem by Tirumaṅkai Āḻvār, who is thought to have lived in the latter half of the eighth century, and the architectural style and iconography. See Emmanuel Francis, ‘Royal and Local Bhakti under the Pallavas’, in Mapping the Chronology of Bhakti: Milestones, Stepping Stones, and Stumbling Stones, ed. Valérie Gillet (Pondicherry: Institut Français de Pondichéry and École Française d’Extrême-Orient, 2014), 113 and Introduction to the volume.

68 

Construc ting K anchi

Within the Ekāmbaranātha complex, the Vṛṣabheśvara temple was likely constructed around this time or slightly later. Whereas Mukteśvara and Mataṅgeśvara preserve Rājasiṃha’s temple style, Vṛṣabheśvara compares more closely with ninth-century temples in Pallava areas farther afield in Tamil Nadu. For example, the absence of rearing lions on the wall pilasters, as well as the use of a curvilinear form of kumuda basement molding, are also seen on the ninth-century Varadarāja Perumāḷ temple at Ālampākkam. The inclusion of a panel of Umāsahita (Shiva and Umā without Skanda) rather than Rājasiṃha’s characteristic Somāskanda on the rear wall of the sanctum further supports a date for Vṛṣabheśvara outside the king’s era. The location of these monuments reveals that by the turn of the ninth century, the Pallava city extended all the way from the northern border of the modern city, where Ekāmbaranātha is found, south to at least the second Karukkiṉil Amarntavaḷ Ammaṉ temple. From the west, the city stretched at least from the Kailāsanātha temple, first to the Airāvateśvara and later to the Vaikuṇṭha Perumāḷ temple on the east. Whereas the Pallava temples and fragments cluster within these boundaries, all of Kanchi’s temples built after the ninth century are situated closer to the modern center of town and even farther to the east. The Kailāsanātha temple once had been Kanchi’s most prominent royal monument. The city’s shift towards the east – away from Kailāsanātha – would gradually lead this temple to fall out of use while enabling other temples to come to the fore.

The Temples of Pallava-Kanchi Art and Embodied Practice: The Kailāsanātha Temple The Kailāsanātha temple is a masterwork of sacred architecture. With a central tower fifteen meters high and a total footprint of more than one thousand square meters, it stood among the largest Hindu temples in India at the time of its construction.75 It is the most fully realized Pallava monument. In its elaborate sculptural and painted program, Shaiva and Vaishnava deities are placed in natural settings or given narrative context. They are sitting or wandering in placid forest hermitages, dancing wildly in cremation grounds, battling demons, or engaging in various episodes from familiar narrative and ritual texts. Between and around the deity figures, there are gandharvas and apsaras, kinnara-kinnarī couples and sensuous 75 Longhurst, ‘Pallava Architecture, Part III (The Later or Rājasiṃha Period)’, 13.

Sandstone and the Cit y

69

Illustration 16 Kinnara and Kinnarī, Kailāsanātha Temple, Kanchi, ca. 700-725 CE

cāmaradhāras, demons and gaṇas who are attendants and musicians, and animals both real and imaginary (Ill. 16). All remaining surfaces are covered in geometric and vegetal decorative ornament, encasing the temple in a skin of thick lotus petals and luxurious brocades. To this day, the natural setting and the people who maintain and use this temple give life to the images. In the evening, green parakeets flutter about, squawking among the figures of bulls and elephants that line the coping of the prākāra. When the birds perch above the lion-based pillars, their luminous green feathers stand out in sharp contrast to the sandstone’s natural tone and the stark white of recently applied plaster. However, in the eighth century, the birds would have blended with brilliantly colored paintings. Remains of the painted surfaces are still visible in recessed places that are sheltered from the ravages of South India’s tropical climate. Although only limited paintings survive, it is clear that they depict a cast of characters similar to the temple’s sculptural program, including figures of deities and the celestial court.76 The expansive temple-site has at least five distinct types of spaces. The layout directs and indexes patterns of movement into, out of, and through the sacred space and begins to articulate some of the ways in which the 76 A study of the paintings as a group, including dating by scientif ic analysis, would be a valuable scholarly contribution beyond the scope of this chapter. However, the paintings should be treated as an integrated part of the temple’s decorative program.

70 

Construc ting K anchi

Illustration 17 Mackenzie Collector, Plan of the Kailasanath Temple, Conjeeveram, ca. 1800, watercolor with pen and ink, 65.6 × 37.5 cm, British Library, WD754

temple may have originally been used (Ill. 17). While some parts of the temple seem to have been restricted, others could accommodate larger numbers of people. Composed of miniature, single-occupancy cells, the prākāra wall not only bounded the sacred space but facilitated private meditation or other solitary practices.77 The cells do not face the courtyard but are tucked behind the inner wall, accessible only from the east. Their open side faces the flat, back wall of the adjacent cell, which was originally painted with an image of a deity. These cell-like shrines echo the humble hermit’s hut 77 Sears has shown the ways in which monastic architecture indexes multiple ritual processes. The Kailāsanātha temple’s plan resembles a monastery, with individual cells surrounding an open courtyard, here centered by a shrine. Tamara I. Sears, Worldly Gurus and Spiritual Kings: Architecture and Asceticism in Medieval India (New Haven and London: Yale University Press, 2014).

Sandstone and the Cit y

71

that forms the basis of early sacred architecture in India.78 They could easily have been used for meditations that required a certain level of initiation.79 Even today, the restricted size and difficulty of access to the prākāra cells continues to suggest a function as meditative chambers. It is not uncommon to see devotees sitting cross-legged inside, facing east, and meditating, especially in the early morning while the sun gradually illuminates the monument. By contrast, a once-spacious courtyard between the prākāra and the complex’s other components provided a place for multiple worshippers to congregate or circulate. Situated between the main port of entry and the primary shrine, a pillared maṇḍapa served as a shelter in which roughly a dozen people could rest simultaneously, but it also facilitated darśan. Devotees could behold the main icon from the maṇḍapa’s elevated platform. When the maṇḍapa was joined to the sanctum in a transformative sixteenthcentury renovation, the visual connection between the two spaces was maintained. A screened window penetrates the maṇḍapa’s sixteenth-century west wall, through which the sanctum interior remains visible. The maṇḍapa is elaborately carved with figures of goddesses (Lakṣmī and Sarasvatī on the south and Jyeṣṭā and Durgā on the north), making it its own site of worship worthy of circumambulation. While circumambulating the maṇḍapa, the visitor would have also had the opportunity to worship the icons positioned in the niches of the prākāra. The maṇḍapa and the courtyard surrounding it may have been accessible to a broader sector of the population, at least at designated times. The important archaeologist of the 1930s and 1940s C. Minakshi has suggested that the maṇḍapa may have served as a congregation place for learned Brahmins of Kanchi’s ghaṭikā, usually translated as university or institution of higher learning.80 Many inscriptions from as early as the sixth century mention a ghaṭikā of great importance at Kanchi.81 Most often cited are the Kasakudi copper plates of Nandivarman II (ca. 731-796 CE), and a 78 Ananda K. Coomaraswamy, ‘Early Indian Architecture: IV. Huts and Related Temple Types’, RES: Anthropology and Aesthetics 15 (1988). 79 Scholars have suggested that the Kailāsanātha temple was a place used by ritual initiates. Nagaswamy, ‘Innovative Emperor and His Personal Chapel: Eighth Century Kanchipuram’; Kaimal, ‘Learning to See the Goddess Once Again: Male and Female in Balance at the Kailāsanath Temple in Kāñcīpuram.’ Gillet describes a sophisticated iconographic program that would have required a high level of knowledge to fully appreciate. Gillet, La Création d’une Iconographie Śivaïte Narrative: Incarnations du Dieu dans les Temples Pallava Construits. 80 Minakshi, Administration and Social Life Under the Pallavas, 192. 81 Ibid., 187.

72 

Construc ting K anchi

lengthy discussion of the ghaṭikā follows in the epigraphic report.82 While Minakshi acknowledges the absence of direct evidence for the ghaṭikā’s location, she cites several Pallava and Chalukya inscriptions that associate the royally sponsored Kanchi ghaṭikā with the Kailāsanātha temple.83 Since temples were associated with the preservation of knowledge from early on in South Indian history, it is possible that learned members of society used this and other temples, such as Ekāmbaranātha and Tirumēṟṟaḷinātha, as places of study.84 Perhaps we should understand the Kanchi ghaṭikā to be a gathering of individuals for the purpose of study rather than a location. At the center of the Kailāsanātha complex is the primary shrine, called Rājasiṁhēśvara, a monumental vimāna with an inner sanctum surrounded by a narrow circumambulatory passageway. Abutting the vimāna are seven subsidiary shrines called aṅga-ālāyas (‘limb shrines’). Although they are similar in design to the prākāra cells, the aṅga-ālāyas are large enough to accommodate more than one person. Each shrine is equipped with a drainspout in a practical location, and a stone bench against the back wall. These features suggest that the aṅga-ālāyas were used for abhiṣekas (‘ritual lustrations’) of liṅgas, portable icons, or perhaps for the initiation of practitioners.85 The Kailāsanātha’s sculptural program gives some indication of the learned nature of the audience it addressed. In La Création d’une Iconographie Śivaïte Narrative, Gillet persuasively argues that with the innovation of the Kailāsanātha temple, Shiva, who is represented in all the manifestations known at the time, is brought into being through a narrative sequence that moves him from the beneficent yogin to the vanquisher of enemies. In the process, Shiva moves from humanity to a divine form of royalty and usurps a role previously reserved for Vishnu.86 This transformation is activated when the viewer circumambulates the temple. 82 SII Vol. 2, Part 3, p. 346-353 (text), p. 42-61 (discussion and translation), no. 73. 83 Inscriptions that suggest an association between Kanchi’s ghaṭikā and the Kailāsanātha temple include the Kasakudi copper plates, the Vēlūrpāḷaiyam plates, and a Kannada inscription by the Chalukyan king Vikramāditya II, which is situtated on a pillar in the Kailāsanātha temple’s maṇḍapa. Minakshi, Administration and Social Life Under the Pallavas, 191-193. Vēlūrpāḷaiyam plates: SII Vol. 2, Part 5, p. 508, no. 98. Vikramāditya II’s inscription: Epigraphia Indica, Vol. 3: 1894-95 (Delhi: Archaeological Survey of India, 1979), p. 359-360, no. 48. 84 Minakshi, Administration and Social Life Under the Pallavas, 191. 85 Ronald Inden, ‘Ritual, Authority, and Cyclic Time in Hindu Kingship’, in Kingship and Authority in South Asia, ed. J.F. Richards (Delhi: Oxford University Press, 1998). I have observed the abhiṣeka of an initiate at the Iṟavāttāṉeśvara temple in Kanchi (July 2014). These ceremonies can be used to confer spiritual advancement and also political authority. 86 ‘Śiva assume donc la fonction jusqu’ici attribuée à Viṣṇu. C’est là un tournant tout à fait nouveau dans l’iconographie śivaïte. Les Pallavas créent en effet une iconographie narrative qui prend

Sandstone and the Cit y

73

Building and Using the Temple Constructing this monumental temple would have required consulting and mobilizing a wide array of people, including master architects and ritual specialists responsible for the temple’s design, as well as teams of stone masons, painters, scribes, bearers, and assistants.87 An entire community of artisans and laborers must have flowed into Kanchi in the early decades of the eighth century. As temples were designated for construction or renovation throughout the next hundred years, Kanchi’s population would have swelled in number and grown increasingly diverse. As each new temple went up, there would have been multiday festivals to accompany its consecration.88 Itinerant performers, merchants, attendants, and cooks would have been among those accompanying the various devotional communities that arrived to marvel at the newly erected buildings. Also present would have been the people tasked with cleaning up the refuse these hoards inevitably left in their wake. If during its construction and premiere the space was overrun with people from various sectors of society, at least parts of the Kailāsanātha temple were likely reserved for more exclusive audiences after its consecration. A back door (now sealed) at the center of the prākāra’s west wall may have served as the private entrance for priests and elite visitors, such as the royal family.89 At times, the temple may well have been used for esoteric forms of ritual available only to initiates. The foundation inscription suggests that Rājasiṃha was an early initiate into Śaiva Siddhānta. The word śaivasiddhāntamārge appears at the place where the text wraps from the northern to the western maintenant en charge tous les aspects de la divinité: du sauvage ou humain au purement divin et royal. Le passage délicat de l’un à l’autre est mis en scène dans le déroulement des représentations au cours de la pradakṣiṇā.’ Gillet, La Création d’une Iconographie Śivaïte Narrative: Incarnations du Dieu dans les Temples Pallava Construits, 283. Kaimal has further noted the ways in which the sculptural program directs movement in clockwise and counterclockwise directions. Padma Kaimal, ‘Circumambulation and Its Opposite: Visual and Verbal Cues to Movement Outside and Inside the Kailasanatha Temple Complex in Kanchipuram’, in Landscape and Sacred Architecture in Pre-modern South Asia (Dumbarton Oaks, Garden and Landscape Studies, 14 November 2014). 87 Vijaya Ramaswamy, ‘Vishwakarma Craftsmen in Early Medieval Peninsular India’, Journal of the Economic and Social History of the Orient, 47, no. 4 (2004). 88 The Cilappatikāram describes the city of Pukār’s celebration of a multiday festival of Indra, which gives some sense of the range of materials, people, and processes that may have been involved in Kanchi. Aṭikal, The Cilappadikāram, 110-121. 89 Secondary entrances are used in this way in contemporary practice. During the course of a privately sponsored ten-day pūjā that I attended at the Aruṇācaleśvara temple, the patrons and officiating priests entered and exited the east-facing complex through the western or northern gopuras, which are not accessible for the public (Tiruvannamalai, May 2015).

74 

Construc ting K anchi

side of the monument.90 The locative case refers to the spiritual pathway along which the king tread. The extent to which this Śaiva Siddhānta conformed to practices recorded in later texts remains questionable. However, the reference is nonetheless significant because it is the earliest known epigraphic record to explicitly mention this spiritual path.91 As it is described in ca. twelfth-century texts, the practice of Śaiva Siddhānta entailed ritual initiations that transmitted particular meditations, visualizations, breathing exercises, and physical yogas.92 These practices were designed to train the body and the mind to transcend physical reality and merge with the supreme yogin that is Shiva.93 Sitting inside the chambers of the prākāra and facing east like the god in the temple’s inner sanctum was a form of embodied practice that helped bring the practitioner closer to Shiva. While the shrines outside the temple complex house liṅgas, inside the private space of the prākāra cells it is the yogin rather than the liṅga that becomes Shiva. It is likely that when the temple was in its greatest ritual use, not everyone was permitted to meditate inside these chambers. Similarly restricted in space, circumambulating the main liṅga through the narrow inner passageway was also a limited privilege. Selecting the Site The Kailāsanātha’s present-day remoteness with respect to the city raises the question of what motivated the selection of this site for the monument. Today the temple is situated on the western outskirts of town, surrounded by agricultural lands, and accessed only by a single, quiet road. There is no miraculous account, no justifying dream or vision, nor any inscription that explains the selection of the site for the Kailāsanātha. However, the temple relates spatially to earlier Pallava monuments, such as the nearby Ekāmbaranātha temple and the more distant site of Māmaṇṭūr – both of which were explicitly royal Pallava monuments – and Pallava kings had a tendency to build or rebuild temples near to earlier Pallava sites.94 90 SII Vol. 1, p. 12-14, no. 24 (verse 5). 91 There are several possible references to Śaiva Siddhānta in copper plates and literature that predate the Kailāsanātha temple. Francis, ‘“Woe to Them!”: The Śaiva Curse Inscription at Mahābalipuram (7th Century CE)’, 184-185. For an early South Indian Śaiva Siddhānta text, see Dominic Goodall, The Parākhyatantra: A Scripture of the Śaiva Siddhānta (Pondicherry: École Française d’Extrême-Orient, 2004). 92 For an excellent study of Śaiva Siddhānta, see Richard H. Davis, Ritual in an Oscillating Universe: Worshiping Śiva in Medieval India (Princeton: Princeton University Press, 1991). 93 Ibid. 94 Orr, Gillet, and Kaimal have agreed that proximity to Māmaṇṭūr and Ekāmbaranātha were likely considerations in selecting the Kailāsanātha temple’s site (discussed August 2015).

Sandstone and the Cit y

75

Illustration 18 Sluice, Māmaṇṭūr, seventh century

At Māmaṇṭūr, a ridge of rocks houses four excavated cave-temples with inscriptions from the reign of Mahendravarman I, including the important biruda that furnishes evidence for the Mattavilāsa’s royal attribution.95 North of the rocky ridge is a massive ēri (‘reservoir’), called Cittiramegha Taṭākam, which is also attributed to Mahendravarman I through another inscribed biruda. The lake feeds extensive paddy fields for many neighboring villages through two granite sluices, original portions of which were part of Mahendravarman’s project (Ill. 18).96 A Brahmi inscription on a rock that shelters the type of pādukā (‘stone bed’) associated with early Jain ascetics shows that the site was already active near the turn of the first millennium.97 Inscriptions from the Chola era reveal Māmaṇṭūr’s continued importance, if not continuous patronage for architecture. Since the Kailāsanātha is situated ten kilometers directly north from Māmaṇṭūr, the location may have been intended to establish a connection with the earlier site. It is possible that a road connected Māmaṇṭūr with the Kailāsanātha’s part of Kanchi (see Chapter Two). Unlike royal Chola establishments, Rājasiṃha’s constructions tended to establish genealogies through the act of building a temple on or near the 95 The inscription praises Mahendravarman for his composition of the Mattavilāsa and treatises on music and painting. SII Vol. 4, p. 12, no. 136. 96 SII Vol. 4, p. 12, no. 138; ARE 1888, no. 40. 97 ARE 1939-1940, no. 171.

76 

Construc ting K anchi

site of an earlier Pallava monument. This is apparent at the Shore temple at Māmallapuram and the Tāḷagirīśvara temple at Paṉamalai. The Shore temple’s eighth-century vimānas, both built by Rājasiṃha, incorporate an earlier reclining Vishnu, which was carved into the living rock probably during the reign of Narasiṃhavarman (Mahāmalla) I (ca. 630-668 CE).98 Paṉamalai’s outlying location in the Villupuram district can be explained by its proximity to the important site of Maṇṭakappaṭṭu, where Mahendravarman’s first cave-temple conquered the medium of stone. At the Tāḷagirīśvara and Shore temples, Rājasiṃha seems to have been continuing a practice that his predecessors had begun. Prior to Rājasiṃha’s reign, Narasiṃhavarman I had excavated the Orukkal cave-temple at Tirukkaḻukuṉṟam (near Māmallapuram), where, according to an inscription, a temple already existed on the top of the hill. The original hilltop temple, which was probably a brick shrine founded by Mahendravarman I, was then rebuilt in stone under Parameśvaravarman I, Rājasiṃha’s immediate predecessor.99 Two systems of practice account for the selection of sites for temples throughout Tamil Nadu. The first is the rebuilding of local shrines, a process that often involved converting a structure from perishable to more durable materials, as in the movement from brick to stone. This is typical of temples sponsored by local, rather than royal, patrons.100 The second is the foundation of a royal establishment where there was not previously an existing local shrine. Examples of royal Chola foundations that were built as an assertion of state power include the large-scale temples at Thanjavur, Gangaikondacholapuram, Dārāsuram, and Tirupuvaṉam.101 There are also examples from other dynastic contexts, such as the Mūvar Kōyil, a royal foundation built by the Irukkuvēls at Koṭumpāḷūr. Therefore, Rājasiṃha’s construction of the Kailāsanātha temple either established the site as sacred or involved the reconstruction and expansion of a shrine that was already there. The relationship with earlier Pallava temples in the other two of Rājasiṃha’s three largest temples suggests that the Kailāsanātha was also designed as the continuation of an earlier sacred site. That several granite pillars from Mahendravarman I’s time have been incorporated into the Kailāsanātha’s construction further supports this hypothesis. The archaic pillars can be found in the maṇḍapa that originally stood separately from the 98 Lockwood, Pallava Art, 89, n. 29. 99 Srinivasan, ‘The Pallava Architecture of South India’, 132. 100 I return to this topic in Chapter Two. Emmanuel Francis, ‘Royal and Local Bhakti under the Pallavas’, in Mapping the Chronology of Bhakti: Milestones, Stepping Stones, and Stumbling Stones. 101 Veluthat, The Early Medieval in South India, 73.

Sandstone and the Cit y

77

main temple body. They have been interspersed with later replacements – some carry later inscriptions or recently have been given sandstone capitals, and the sole Pallava-Grantha inscription faces the sixteenth-century wall at nearly too narrow a distance to see, let alone photograph. This all may be why the pillars have gone virtually unnoticed in scholarship.102 Similar pillars, though smaller in scale, also support the maṇḍapa of the Shiva temple at Kūram. This apsidal temple, built by Parameśvaravarman in the late seventh century, is known to be the earliest surviving temple in Tamil Nadu to be built of blocks of stone.103 Given the pattern of establishing genealogy through sacred architecture that is evident in Pallava constructions not limited to Rājasiṃha’s reign, it is likely that the shrine on or near the site of the Kailāsanātha temple was originally established by a Pallava king, possibly Mahendravarman I. Revising the Temple Monumental and enduring as it may appear, the Kailāsanātha temple was far from immune to change. In the second decade of the eighth century, Rājasiṃha’s eldest son, Mahendravarman III (hereafter Mahendra), made the curious decision to sponsor the construction of an additional shrine in the Kailāsanātha’s eastern courtyard, and to enclose it with its own prākāra. Mahendra’s prākāra intercepts the eastern row of devakulikās (‘external shrines’) such that the sculptures on the abutting sides of the shrines are partially obscured. This detail makes it clear this prākāra was added after the monument’s original construction. Mahendra’s shrine fundamentally changed the pattern of movement into the main temple complex. Before his shrine was constructed, the Kailāsanātha was entered primarily through a single door in the east wall of the original prākāra. Mahendra’s shrine took the place of this door and forced the visitor instead to enter and exit the main space of the temple through narrow passages on the south and north sides of his shrine. Apertures in the east, south, and north walls of Mahendra’s prākāra further suggest two possible modes of accessing the temple. Whereas the eastern door leads the 102 The pillars but not the inscriptions are mentioned in Srinivasan, ‘The Pallava Architecture of South India’, 131; ‘Pallavas of Kāñcī: Phase I’, 63. In 1927, Jouveau-Dubreuil reported finding fragments of marble pillars west of the temple, which Minakshi confirmed at the site. Minakshi suggests they may have belonged to a Buddhist monastery on or near the Kailāsanātha temple. C. Minakshi, ‘Buddhism in South India’, in South Indian Studies, Vol. II, ed. R. Nagaswamy (Madras: Society for Archaeological, Historical, and Epigraphical Research, 1979), 98-99. This article was published posthumously – Minakshi passed away in 1940, tragically young. 103 Srinivasan, ‘Pallavas of Kāñcī: Phase I’, 44 and Pl. 50.

78 

Construc ting K anchi

visitor directly into and out of Mahendra’s shrine, the south and north doors promote circumambulation. Before the sixteenth-century renovation, there were three possible pathways for circumambulation from the east: around Mahendra’s shrine alone, around Rājasiṃha’s maṇḍapa, and all the way around the main vimāna. An opposing circuit led from the western door around Rājasiṃha’s vimāna or around all three structures. We cannot know precisely why Mahendra chose to build his shrine. It is possible that he desired to eclipse his father’s magnificent creation. By intercepting the main entrance, the shrine prevents a clear view of the main temple. However, there are other possible motivations. First, the addition of Mahendra’s shrine brings the plan of the Kailāsanātha closer to that of Māmallapuram’s Shore temple. The Shore temple is comprised of two Shiva shrines with a Vishnu sanctum wedged between them. Although the smaller Shiva shrine (Rājasiṃheśvara) both visually and physically obstructs the larger one (Kṣatriyasiṃheśvara), both structures were sponsored by Rājasiṃha. The larger shrine is circumambulated by entering through narrow apertures to the south and north of the smaller shrine, not unlike the entrances that flank Mahendra’s shrine at Kailāsanātha. Second, the position of Mahendra’s shrine increases possibilities for monitoring access to the temple. Father and son may have agreed that the temple had become too popular. Situated on a main road and doubtless a wonder to behold, the Kailāsanātha temple had the potential to become thronged with crowds. Physically restricting the point of access to the temple would have limited the number of people who could enter simultaneously. It also would have enabled a gatekeeper greater control by giving him more time to decide whether a person should be granted permission to enter the temple. It is hard to imagine that people would not have wanted to visit this temple. The city’s Shaiva population in particular would have wanted to come for devotional purposes, and the novelty of stone temple architecture may well have given the temple an even broader and more diverse appeal. The Kailāsanātha’s sculptures conform to ideals of beauty that were well codified throughout courtly India by the eighth century. God and king alike were supposed to cultivate a graceful appearance. They should be slender, adorned with ornaments, have a pleasant countenance, and favor sidelong rather than fully frontal glances.104 Such practiced appearance was cultivated 104 Codes of divine and courtly appearance are explored in Daud Ali, Courtly Culture and Political Life in Early Medieval India (Cambridge: Cambridge University Press, 2004); Vidya Dehejia, The Body Adorned: Dissolving Boundaries between Sacred and Profane in India’s Art (New York: Columbia University Press, 2009).

Sandstone and the Cit y

79

by courtiers and represented in art. The Pallava rulers were self-avowed connoisseurs. If their sponsorship of art at Māmallapuram is not testament enough to their care and concern with aesthetics, the accomplishments of king Mahendravarman I as an author, poet, and playwright must be. Mahendravarman prided himself on creating the first cave-temple in South India, as stated in an inscription that he himself composed. King Rājasiṃha seems to have employed the finest artisans available for the construction and adornment of his temples. There was a high level of attention to artistic detail, style, and materiality in every one of his projects. The reliefs on the Kailāsanātha temple’s walls represent the most refined artistic excellence in the corpus of Pallava art, according to their own conventions. Beautiful is not a subjective assessment here, nor are judgments of refinement and excellence. The tableau of Dakṣiṇāmūrti on the south wall is one of the most impressive compositions (see Illustration 3). Drawing on Buddhist imagery, Shiva is embodied as the divine teacher seated beneath a tree in a forest hermitage.105 In this manifestation he is benevolent and calm, yet his matted locks and fanged teeth reveal his proclivity for wandering in cremation grounds and his potential for ferociously slaughtering demons. At once, he bestows grace and destroys the universe, joined in a single beautiful form. The sculptures that adorn the temple use language drawn from elite Sanskritic culture represented in visual form. Icons of deities are positioned over the heads of elephants, symbols of royal authority. Supplicants bow at the feet of Shiva, who grants his blessing like a king absolving a crime. Strikingly sensuous women, musicians, and further attendants flesh out the celestial court, and between each panel, girding and expanding the corners, rearing vyālas (‘ithyphallic lions’) spring forth – symbols of the Pallava dynasty. Along the prākāra, royal couples sit enthroned with ritual water pots laid before them, attended by cāmaradhāras (‘flywhisk bearers’) and a host of entertainers. Rājasiṃha’s temple elevates the courtly world to the level of the divine. An Iconography of Kingship: The Vaikuṇṭha Perumāḷ Temple Kanchi’s second largest Pallava temple is the Vaikuṇṭha Perumāḷ. Like the Kailāsanātha temple, the Vaikuṇṭha Perumāḷ is one of the few temples in Kanchi that has received scholarly attention. Most notably, Dennis Hudson argued that the sculptural program directly represents books one through twelve of the Vaishnava text Bhāgavata Purāṇa in conjunction with a Divya Prabandham hymn by Tirumaṅkai Āḻvār. In his religious interpretation, 105 Francis, Gillet, and Schmid, ‘L’Eau et le Feu: Chronique des Études Pallava’, 594.

80 

Construc ting K anchi

Hudson proposed that the monument was designed as a three-dimensional maṇḍala that held an exoteric meaning (read horizontally as the devotee circumambulates) and an esoteric meaning (transmitted as the devotee ascends vertically).106 As Michael Meister has pointed out, Hudson valuably identified this temple as what George Kubler has called a ‘prime object’, a singular artistic production that stands apart from what has come before and that shapes what comes after in transformative ways.107 I would add that the countless hours Hudson spent at Vaikuṇṭha Perumāḷ during the course of more than 30 years gave him a deep understanding of the phenomenological experience of being inside that particular temple. This awareness comes through in fleeting observations, such as, ‘once inside, viewers stand inside a compact and bounded realm where the perspective is at all times partial, shifting, and intimate.’108 Architecturally as well as ritually, the Vaikuṇṭha Perumāḷ temple exists at a ‘node of change.’109 Prākāra Reliefs Encircling the Vaikuṇṭha Perumāḷ temple is a series of narrative reliefs that has drawn some scholarly attention. In 1941, Minakshi provided the first detailed descriptions, but no one has carried out a thorough or even provisional study of them since.110 Making sense of the narrative program becomes increasingly challenging. The reliefs have suffered wear from use, weathering, and poor restoration. Some of the panels were placed in the wrong order or even upside down during early twentieth-century renovations by the Archaeological Survey of India (ASI).111 Although Minakshi did not take notice of the mistaken ordering, her descriptions, together with drawings produced by Alexander Rea in 1909, remain an important source for the study of this monument. The series is significant to the corpus of Pallava artwork as a whole.112 Like texts such as the Nantikkalampakam (a history of Nandivarman III), the visual narrative supplies important information about Pallava history. It 106 Hudson, The Body of God: An Emperor’s Palace for Krishna in Eighth-Century Kanchipuram. 107 George Kubler, The Shape of Time: Remarks on the History of Things (New Haven and London: Yale University Press, 1962), 35; Michael W. Meister, ‘Navigating Hudson’, Journal of Vaishnava Studies 11, no. 1 (2002): 46. 108 Hudson, The Body of God: An Emperor’s Palace for Krishna in Eighth-Century Kanchipuram, 6. 109 Meister, ‘Navigating Hudson’, 46. 110 Minakshi, ‘The Historical Sculptures of the Vaikuṇṭhaperumāḷ Temple, Kāñcī.’ 111 Francis, Gillet, and Schmid, ‘L’Eau et le Feu: Chronique des Études Pallava’, 598. 112 Hudson, The Body of God: An Emperor’s Palace for Krishna in Eighth-Century Kanchipuram, 57-81.

Sandstone and the Cit y

81

also tells us about the fashioning of royal-religious ideology in the Pallava’s courtly world.113 The cycle presents a detailed depiction of the entire Pallava lineage, beginning with divine progenitors. Francis, Gillet, and Schmid have remarked on the likeness between the structuring of the Vaikuṇṭha Perumāḷ temple’s royal relief carvings and royal praśastis of the Pallava kings.114 Each king is illustrated with a selection of key accomplishments, just as he would be described with epithets in a praśasti. ‘Ici, l’eloge royal se fait image’ [‘Here, the royal elegy is the image’].115 Nicholas Dirks has identified Nandivarman’s geneaology as the establishment of a significant change in the structure of South Indian kingship, from sovereignty based on the right to perform sacrifice, to sovereignty based on divine origins.116 The reliefs cover the entire back wall of a cloistered prākāra. The sequence begins on the west wall, to the north of the entrance, and runs clockwise around the monument. The narrative proceeds along an upper and then lower register of panels on each side of the temple individually. To follow the story requires interrupting circumambulation and walking back to the beginning of a wall in order to view the continuation below. It seems that circumambulation was subordinated to historical memory in the more secularized space of the prākāra.117 Nevertheless, it is unlikely that a common visitor to the temple would have undertaken to view each panel individually and to follow the narrative, episode by episode.118 Even without following the sequence, the visual program of the Vaikuṇṭha Perumāḷ’s prākāra as a whole is explicitly imperial. Coronation scenes and images of enthroned kings at court, along with battles, processions, sacrifices, and royal ministers in council dominate the walls. Seen throughout the corpus of Pallava art, the coronation is repeated here to the extent that it 113 Emmanuel Francis, Le Discours Royal dans l’Inde du Sud Ancienne. Inscriptions et Monuments Pallava (IVème-IXème Siècles). Tome II: Mythes Dynastiques et Panégyriques (Leuven: Peeters Publishers, 2017). 114 Francis, Gillet, and Schmid, ‘L’eau et Le Feu: Chronique des Études Pallava’, 597. 115 Ibid. 116 Nicholas Dirks, ‘Political Authority and Structural Change in Early South Indian History’, The Indian Economic and Social History Review 13 (1976): 144. The claim to divine kingship was part of a broader shift in sovereignty throughout South India in the period often described as the early medieval. Veluthat, The Early Medieval in South India, 68-70. 117 Like the narrative reliefs, the inscription that wraps around the base of the monument does not accord with circumambulatory order. As Orr points out, reading the inscription would have conflicted with the expected mode of worshipping at the temple. Orr, ’Preface’, V. 118 Dehejia points out similar problems in following narrative sequences elsewhere in India. Vidya Dehejia, ‘On Modes of Visual Narration in Early Buddhist Art’, The Art Bulletin 72, no. 3 (1990).

82 

Construc ting K anchi

Illustration 19 Coronation scene, Vaikuṇṭha Perumāḷ Temple, Kanchi, ca. 730-795 CE

Sandstone and the Cit y

83

becomes an iconic image (Ill. 19).119 In each occurrence, the king is depicted seated on a throne facing frontally with two feet planted on a stool or other support and two hands firmly on his thighs. The elbows, bent out to the sides, broaden his chest and frame his body. The image evokes certain icons of deities, such as the prosperity goddesses Gajalakṣmī and Jyeṣṭā, both of which were associated with royalty. In the Vaikuṇṭha Perumāḷ reliefs, the central figure wears a crossed-chest ornament that loops over his upper arms to create a sense of stability and to draw attention to the musculature of his torso. Two attendants lustrate the figure with water. This abhiṣeka constitutes a performative act that confers his new status as king.120 The courtly world of the f irst Nandivarman covers the north wall. This is followed by the coronations of all the kings from Siṃhaviṣṇu to Parameśvaravarman II, which are compressed together on the east wall. Along the south and half of the west wall, the biography of Nandivarman II (ca. 731-796 CE) is displayed in the greatest level of detail. This is not surprising as he is the Vaikuṇṭha Perumāḷ temple’s patron. The final sequence shows the construction of his temple. Although the first thirteen panels on the south wall are labeled with inscriptions for added precision, the reliefs concerning Nandivarman II have been subject to much speculation, particularly with regard to the protagonist’s birthplace.121 The sequence presents the king as a military hero who had to travel a great distance in order to reach Kanchi. These reliefs constitute the ‘official’ version of Pallava history. However, the truth of their claims is of lesser importance than their claim to truth.122 Depicting the temple’s royal patron in sculpture in effect constructs the king through 119 Weschler suggests that the f igure of Rama becomes an icon at the Papanatha temple at Pattadakal through a similar repetition of a distinct image in an otherwise narrative sequence. Helen J. Weschler, ‘Royal Legitimation: Ramayana Reliefs on the Papanatha Temple at Pattadakal’, in The Legend of Rama: Artistic Visions, ed. Vidya Dehejia (Bombay: Marg, 1994). 120 Inden cites the Vaikuṇṭha Perumāḷ reliefs as visual evidence for the signif icance of the abhiṣeka coronation. Inden, ‘Ritual, Authority, and Cyclic Time in Hindu Kingship’, 58. 121 Mahalingam suggested Cambodia for the king’s origins. Mahalingam, Kāñcīpuram in Early South Indian History, 153; Hudson, The Body of God: An Emperor’s Palace for Krishna in Eighth-Century Kanchipuram, 72-73. 122 On the self-fashioning of South Indian kings and their imperial histories, see Francis, Royal dans l’Inde du Sud Ancienne. Inscriptions et Monuments Pallava (IVème-IXème Siècles). Tome Le Discours II: Mythes Dynastiques et Panégyriques. See also Daud Ali, ‘Royal Eulogy as World History: Rethinking Copper-Plate Inscriptions in Cōla India’, in Querying the Medieval: Texts and the History of Practices in South Asia, ed. Ronald Inden, Jonathan Walters, and Daud Ali (Oxford: Oxford University Press, 2000). On the claim to truth in the writing of medieval history more broadly, see Robert M. Stein, Reality Fictions: Romance, History, and Governmental Authority, 1025-1180 (Notre Dame: University of Notre Dame Press, 2006).

84 

Construc ting K anchi

Illustration 20 Reclining Vishnu among nāga stones, Vaikuṇṭha Perumāḷ Temple, Kanchi

visual means. In this version, the king is fashioned as an avatar of Vishnu who transcended the boundary between earthly and divine worlds.123 Architecture The architectural plan of the Vaikuṇṭha Perumāḷ temple is highly sophisticated and complex. The vimāna is a square, catustāla (‘four-storey’) tower with adjoining ardhamaṇḍapa.124 Both stand on a high, unornamented pādabandha-adhiṣṭhāna (a particular series of basal moldings) that supports walls that are fully sculpted. Visible through narrow windows in the exterior walls, an inner circumambulatory passage (‘pradākṣina’) can be entered from the east exterior through the central bhadra (‘projection’) or from inside the ardhamaṇḍapa on the northwest side of the main sanctum. The lower-level sanctum holds an image of a seated Vishnu. As the inner ambulatory encircles the first level, it passes large-scale relief panels of three forms of seated Vishnu before opening onto a staircase on the east side. The stairs make a quarter circumambulation as they rise, before directing movement into the second sanctum. This middle-level shrine is accessed from the west and can be circumambulated twice, once interiorly and then along an open-air balcony. 123 Francis, Gillet, and Schmid, ‘L’Eau et le Feu: Chronique des Études Pallava’, 605. 124 For the architectural plan and full description, see Srinivasan, ‘Pallavas of Kāñcī: Phase I’, 68-73.

Sandstone and the Cit y

85

Illustration 21 Dharmarāja Ratha, Māmallapuram, seventh century

Today, the second-storey sanctum contains a sculpture of Anantaśayana (reclining Vishnu) that post-dates the temple by many centuries. It replaces an earlier icon, which may be a sculpture of Anantaśayana that is presently installed beneath the temple’s sacred tree (Ill. 20). The coiled serpent couch upon which Vishnu lies and the serpentine shape of the cushions supporting his back have led local devotees to worship this icon together with nāga stones (auspicious serpent spirit images shallowly carved into stones). Another fragmentary Vishnu that once also may have been a sanctum icon is now on the floor near the doorway to the later-appended mahāmaṇḍapa.125 This figure may at some point have occupied the third sanctum, which originally should have contained a standing Vishnu but today is empty and inaccessible.126 A staircase leads down from the southeast side of the second storey back to the lower pradākṣina and the eastern entry. The main shrine of the Vaikuṇṭha Perumāḷ has its prototype in Māmallapuram’s Dharmarāja Ratha, a three-storey rock-cut temple (Ill. 21). A key difference in technique distinguishes the two. As a monolithic temple, the Ratha was carved from the top down, and its lower sections were left 125 The Vishnu torso is similar in style to a goddess beneath the tree at the Mataṅgeśvara temple. 126 The temple’s priests say the upper shrine originally was accessed by ladder, and that the ladder was damaged by lightning (Pers. comm. 2014). There are many such stories of lightning strikes in India.

86 

Construc ting K anchi

unfinished. This means that in order to build the foundations for the Vaikuṇṭha Perumāḷ temple, the architect had to deduce the monument’s intended footprint. This act was a tremendously creative enterprise that would not be without effect. Most immediately, the Vaikuṇṭha Perumāḷ temple greatly informed the architecture of the Sundaravarada Perumāḷ temple in Uttiramērūr, built less than a century later, but the Vaikuṇṭha Perumāḷ was destined to have a far longer legacy in triple-storey Vishnu temples across the Tamil lands.127 Smaller Temples: Piṟavāttāṉeśvara, Airāvateśvara, Iṟavāttāṉeśvara, and Amareśvara The Piṟavāttāṉeśvara and Airāvateśvara temples are situated on the eastern boundary of the original Pallava area of Kanchi. Both temples are small in scale, comprised of a vimāna with attached mukhamaṇḍapa, and both open west towards the Kailāsanātha. In a sense, they can be understood as distant shrines subsidiary to the larger monument. Piṟavāttāṉeśvara is formally quite similar to the Kailāsanātha’s row of devakulikās, as well as the miniature shrines in the temple’s prākāra (Ill. 22). It especially resembles the devakulikā furthest south, which is taller than the others and stands away from Mahendra’s prākāra. The brackets above the kapotas (‘cornices’), as well as the bulbous shape of the octagonal śikhara (‘superstructure’), are common to both shrines but otherwise are unusual features in Pallava temples.128 Since the inscriptions suggest that the eastern shrines at Kailāsanātha were sponsored by members of Rājasiṃha’s inner circle – specifically women – it seems possible enough that the Piṟavāttāṉeśvara temple results from similar patronage.129 Several decades later, the Iṟavāttāṉeśvara temple was constructed directly opposite Piṟavāttāṉeśvara. Iṟavāttāṉeśvara and Piṟavāttāṉeśvara are positioned to face each other on opposite sides of a narrow road that leads towards the early Pallava site of Kūram, eight kilometers away. The Airāvateśvara temple played an important role as a boundary marker on the southeastern corner of the early eighth-century Pallava city. The temple 127 Triple-storey Vishnu temples include the Kūṭal Aḻakar temple at Madurai, the Rajagopāla temple at Maṉṉārkuṭi (sixteenth century) and the Saumyanārāyaṇa temple at Tirukkōṣṭiyūr (seventeenth century). Crispin Branfoot, ‘Approaching the Temple in Nayaka-Period Madurai: The Kūṭal Aḻakar Temple’, Artibus Asiae 60, no. 2 (2000): 217. A three-storey Vishnu shrine is also found at the rear of Kanchi’s Kāmākṣī Ammaṉ temple. Potentially related examples include the Varadarājasvāmi temple at Ālampākkam, the Kṛṣṇa temple at Tiruveḷḷaṟai, the Tiruppērnakar Perumāḷ temple at Kōvilaṭi, and the Varadarāja Perumāḷ temple at Kanchi. The list continues to grow. 128 Srinivasan, ‘Pallavas of Kāñcī: Phase I’, 65. 129 Lockwood, Pallava Art, 149-153.

Sandstone and the Cit y

87

Illustration 22 Piṟavāttāṉeśvara Temple, Kanchi, ca. 700-725 CE

stands on a direct line of axis with the Kailāsanātha, at a distance of one kilometer to the east. It is a small temple, now missing its superstructure, but its walls are richly carved with sensuous celestial women, fierce dvārapālas, and other deities. The iconography and style of carving closely parallel the Kailāsanātha’s sculptural program, which suggests that Airāvateśvara was built not only within the reign of Rājasiṃha, but close to the time of Kailāsanātha. It is quite possible that the two temples were under construction at the same time. Airāvateśvara is also positioned in dialogue with the Kacchapeśvara temple, which is situated directly across the street. The present-day shrine of Kacchapeśvara is an eleventh-century structure. However, within the compound there are fragmentary sandstone sculptures and siṁhapāda pillars, white granite pillar bases with figures of dancing ascetics, a sandstone well, and a great many sandstone blocks with inscriptions that have yet to be read.130 These fragments show that the site was already sanctified in the Pallava period. The close proximity between Kacchapeśvara and Airāvateśvara facilitates a relationship between the two temples. Today, priests from Kacchapeśvara are responsible for maintaining the rituals at Airāvateśvara. 130 The inscribed blocks have been repurposed primarily for building a shrine to Sūrya that stands opposite the main body of the later temple.

88 

Construc ting K anchi

In his 1909 publication, Pallava Architecture, Alexander Rea made a series of drawings of Kanchi’s temples, including what he called alternately the Tripurantakesvara or Tirupurantakesvara, both of which refer to the Amareśvara temple.131 Today this temple’s walls are encased in multiple layers of thickly applied plaster and paint, leaving relief sculpture all but invisible beneath. Rea’s five drawings and a dozen archival photographs show that the temple was already severely damaged and poorly repaired with bricks at the time of plastering. Using these limited representations, Gillet was able to reconstruct a plausible sculptural program for the temple, which closely corresponds to other temples from Rājasiṃha’s reign.132 However, in the present condition, the architectural structure and the rearing-lion pilasters are the only indications that Amareśvara is a Rājasiṃha-period construction. Not all Pallava commissions were temples for Shiva or Vishnu. The Jain establishment of Candraprabhā Jiṉālāya is a quiet place in the hilltop village of Tiruparuttikuṉṟam, known today as Jina-Kanchi. The sprawling, thirteenth-century (and later) temple complex of Trilokya Jinasvāmi, dedicated to the Jina Vardhamāna (Mahāvīra), is just a few meters away. The site is a typical Jain pilgrimage place, with a cluster of monuments situated on top of a hill.133 Today Candraprabhā Jiṉālāya is an austere, archaeologically protected monument, surrounded by a well-maintained garden sequestered behind an often-locked gate.134 Corner pilasters in the shape of vyālas are the only sculptural elements to adorn the walls. It is possible that this temple was never finished, or that the Jain community for whom it was built preferred the absence of images on the exterior.135 The temple has been transformed through renovations, but an original upper ambulatory, along with the sizeable ardhamaṇḍapa and the slenderness 131 It is not clear if the name Rea published was ever associated with this building. Alexander Rea, Pallava Architecture, Vol. XXIV (Madras: Archaeological Survey of India, New Imperial Series, 1909), Pl.CIX-CXIII. The tri/tiru slippage is common in Tamil because the Sanskrit tri (‘three’) begins the epithets of many gods, and the Tamil tiru (‘sacred’) is a ubiquitous prefix to names of sages, temples, and places. 132 Gillet, La Création d’une Iconographie Śivaïte Narrative: Incarnations du Dieu dans les Temples Pallava Construits, Plan IV. 133 Julia Hegewald, Jaina Temple Architecture in India: The Development of a Distinct Language in Space and Ritual (Berlin: G&H Verlag, 2009). 134 A description of the temple, including the interior, can be found in T.N. Ramachandran, Bulletin of the Madras Government Museum: Tirupparuttikkuṉṟam and its Temples (Madras: Commissioner of Museums, 2002), 11-13. 135 North Indian Jain temples are often more ornamental on the interior than the exterior. The best-known example is the hilltop complex at Mt. Abu (ca. 1030 CE). Earlier examples can be found at Osian.

Sandstone and the Cit y

89

Illustration 23 Mukteśvara Temple, Kanchi, eighth/ninth century

of the wall pilasters, suggests that the temple may not date to Rājasiṃha’s reign but slightly later, perhaps to the period of Nandivarman II.136 Process and Technique: Mukteśvara and Mataṅgeśvara Temples The Mukteśvara and Mataṅgeśvara temples are positioned to the north and south of the Vaikuṇṭha Perumāḷ temple, respectively. They are smaller shrines, quite similar to each other in scale and architectural form, which suggests the two monuments were likely built at the same time or in rapid succession. However, whereas Mukteśvara carries an extensive decorative program and inscriptions that attest to the temple’s continued use in the centuries after its construction, Mataṅgeśvara is remarkably unfinished (Ill. 23).137 136 Srinivasan, ‘Pallavas of Kāñcī: Phase I’, 74; Ramachandran, Bulletin of the Madras Government Museum: Tirupparuttikkuṉṟam and its Temples, 16. 137 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 119, Cg.-491 and p. 127, Cg.-523. A Vijayanagara inscription is also mentioned in the ASI report. SII Vol. 4, p. 285-287, no. 827-829; Longhurst, ‘Pallava Architecture, Part III (the Later or Rājasiṃha Period)’, 19.

90 

Construc ting K anchi

A comparison between Mukteśvara’s relatively complete and Mataṅgeśvara’s incomplete decorative programs supplies important information about processes, techniques, and priorities for stone carvers.138 At the most basic level, the Mataṅgeśvara temple shows us that in Pallava architecture the structure was generally completed first and the decorative program was applied subsequently, rather than using sculpted blocks to construct the temple. This technique is similar to practices typically used in parts of Southeast Asia. Unfinished portions of monuments, such as Borobudur in Central Java and the Angkor complex in Cambodia, show that these temples were first constructed and then given their rich sculptural programs, which were carved in low relief.139 In the Southeast Asian monuments, important body parts – especially faces – are positioned so as to avoid crossing the seams between stones.140 In terms of carving techniques, Kanchi’s Pallava temples also have much in common with the seventh-century monolithic temples at Māmallapuram. Māmallapuram’s monuments were carved from the top down using a subtractive technology not dissimilar to excavating a cave. Extensive lack of finish at these monuments makes their processes of production visible.141 In Kanchi, the masons seem to have followed certain expectations for their work process, including the order in which to add sculptures.142 Through the Mataṅgeśvara’s unfinished condition, we can see that even before relief carvings were begun, all the exterior corners were girded with protective animals. Recumbent elephants support the exposed foundation platform, rearing lions leap from the walls, and seated lions lift the roof. The animals were put in place first, as if they were structural rather than decorative components to the fabric of the building. This may be because they were considered apotropaic symbols of royal and religious authority that were intended to protect the god inside the sanctum, and, by extension, the temple’s sponsor. In temples elsewhere in India, corners were seen as vulnerable points in need of fortification.143 138 Recent scholarship has examined the ways in which unfinished monuments reveal processes of making. Vidya Dehejia and Peter Rockwell, The Unfinished: Stone Carvers at Work on the Indian Subcontinent (New Delhi: Lustre Press, 2016). 139 Jinah Kim, ‘Unf inished Business: Buddhist Reuse of Angkor Wat and its Historical and Political Significance’, Artibus Asiae LXX, no. 1 (2010); Natasha Reichle, ‘Unfinished Business: Clues About Artistic Praxis from the Hidden Foot of Borobudur’, Artibus Asiae LXIX, no. 2 (2009). 140 ‘Unfinished Business: Clues About Artistic Praxis from the Hidden Foot of Borobudur’, 339. 141 Dehejia and Rockwell, The Unfinished: Stone Carvers at Work on the Indian Subcontinent. 142 For architectural work processes elsewhere in the world, see Rockwell, The Art of Stoneworking: A Reference Guide, 89-106. 143 Michael W. Meister, ‘Siva’s Forts in Central India: Temples in Daksina Kosala and Their “Daemonic” Plans’, in Discourses on Siva, ed. Michael W. Meister (Philadelphia: University of

Sandstone and the Cit y

91

With the corners complete with animal imagery, the reliefs inside the sanctum and interior walls of the mukhamaṇḍapa were the apparent next concern. The north wall and inner faces of the mukhamaṇḍapa are intricately carved. Turning to the exterior walls, reliefs depicting deities and scrolling vegetal ornament were carved between the pilasters. The imagery was lightly sketched and then brought toward completion. In figural carvings, the midsections and faces were reworked before the extremities. Reliefs were added in a clockwise circumambulatory direction, in this case beginning from the west (front) of the monument and continuing toward the north.144 Sculpture is entirely absent from half of the eastern and all of the southern exterior walls. It stands to reason that having a consistent work process allowed sculptors to slowly perfect their forms and proportions by refining and adjusting the details.145 The dating of these two monuments remains uncertain. An inscription at Mukteśvara suggests that the shrine was built by one of Nandivarman’s queens, but whether this Nandivarman was the second or third Pallava ruler with the name is not specified.146 At Mataṅgeśvara and Mukteśvara, the reliefs are deeply set into the niches, leaving them surrounded by a ‘scooped out’ recess.147 This technique of carving brings the sculptures closest in appearance to those at Kanchi’s Vaikuṇṭha Perumāḷ temple and the Kailāsanātha temple at Tiruppaṭṭūr (near Tiruchirappalli). Both of these monuments are generally assigned dates towards the end of the eighth century, during the reign of Nandivarman II or the early years of Dantivarman (ca. 796-846 CE).148 Most likely, Mukteśvara and Mataṅgeśvara were also constructed near the end of the eighth century.149 Even long after the Pallava period, the circumambulatory order of working that is evident at Mataṅgeśvara appears not to have been limited to Pennsylvania Press, 1983). 144 The expectation to add the decorative veneer in a clockwise direction was not always met. At the Piṟavāttāṉeśvara temple, the reliefs on the north and south walls are much more finished than on the east (rear) wall. 145 Rockwell, The Art of Stoneworking: A Reference Guide, 129. 146 The dates of the two Nandivarmans are ca. 731-796 and ca. 829-853 CE respectively. Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 119, Cg.-491; Francis, ‘“Woe to Them!”: The Śaiva Curse Inscription at Mahābalipuram (7th Century CE)’, 192. 147 Srinivasan, ‘Pallavas of Kāñcī: Phase I’, 68. 148 Ibid., 64; Gillet, La Création d’une Iconographie Śivaïte Narrative: Incarnations du Dieu dans les Temples Pallava Construits. 149 Gillet arrives at a similar conclusion along iconographic lines. The programs of Mataṅgeśvara and Mukteśvara are consistent with late eighth-century temples. La Création d’une Iconographie Śivaïte Narrative: Incarnations du Dieu dans les Temples Pallava Construits, 63-64.

92 

Construc ting K anchi

the carving of sculptural programs. It also applied to the constructing of enclosure walls. In large temple complexes, one finds that the outer prākāra is often made of mixed rubble. Various types of local granite and sandstone are sometimes included, and the walls are topped with brick and overlaid with cement. At the Kacchapeśvara temple, for example, entering from the main (north) gopura and examining the wall in a circumambulatory manner reveals that the first (east) wall consists of well-dressed blocks of more or less uniform size and of primarily a single material. The only serious interruptions to the east wall are a fragmentary inscription and two sandstone sculptures of unidentifiable Shaiva deities that are embedded sideways in the prākāra’s lower courses. By contrast, the far (western) half of the second (south) wall contains many variegated blocks of stone, and fragmentary architectural elements are freely incorporated. The final (west) wall is a disorienting pastiche. Reused architectural materials ranging from doorframes and basement moldings to pilasters and cornices are the dominant makeup of this wall. The outer prākāras of Ekāmbaranātha, Tirumēṟṟaḷinātha, Varadarāja Perumāḷ, Aṣṭabhuja Perumāḷ, and additional temples in Kanchi exhibit similar patterns. The consistency suggests it was standard practice for the prākāra to be built more or less in a clockwise direction, beginning from the main point of entry into the complex. An attempt was made to use stones of uniform size and shape, but the expanse of the project was often overambitious for the available materials. By the final wall, pre-decorated stones and other unideal fragments had to be incorporated as the supply of good stones became depleted.

Everywhere but Kanchi (the Ninth Century) Capping a century’s efflorescence of sandstone architecture, the Vaikuṇṭha Perumāḷ temple was the last major monument to be built in Pallava-Kanchi. During the ninth century, a serious reduction in temple building occurred inside the city. Although the Vṛṣabheśvara temple at Ekāmbaranātha may be compared with ninth-century temples, and the dates of Mataṅgeśvara and Mukteśvara remain uncertain, none of Kanchi’s temples can be firmly dated to the ninth century. Likewise, in contrast to the hundreds of eighthcentury inscriptions, less than ten ninth-century inscriptions are known from anywhere in the city.150 The only one of these found in situ encircles 150 Kanchi’s ninth-century inscriptions are on stone slabs built into paving stones at Tirumēṟṟaḷinātha, Ulakaḷanta Perumāḷ, and Varadraja Perumāḷ temples. SII Vol. 12, p. 18,

Sandstone and the Cit y

93

Illustration 24 Map of Pallava sites, ninth century (map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution)

the vimāna at the Vaikuṇṭha Perumāḷ temple. It details gifts of gold to the temple by the Pallava king Dantivarman, who was, meanwhile, sponsoring the construction of the great Sundaravarada Perumāḷ temple in Uttiramērūr, 40 kilometers outside the city.151 The great number of temples constructed in Pallava territories outside of Kanchi during the ninth century indicates that there was no shortage of support for temple architecture and iconic sculpture (Ill. 24). South of Kanchi, Uttiramērūr houses spectacular examples from this period, including the Sundaravarada Perumāḷ, Kailāsanātha, Kedāreśvara, and Mātiriyammaṉ temples, as well as a granite platform that served as the Brahmin council’s administrative meeting hall (‘sabhā’) (Ill. 25, Ill. 26).152 The tiny village of Kumaravāṭi boasts fragmentary ca. ninth-century sculpture, architecture, and inscriptions. The villages of Ārpākkam and Kaḷakkāṭṭūr are home to a no. 44; Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 120, Cg.-495; R. Nagaswamy, Viṣṇu Temples of Kāñcīpuram (New Delhi: D.K. Printworld, 2011), p. 150, Fig. 5.23. In a storeroom in the Kāmākṣī Ammaṉ temple, a loose slab carries an eighth-century record thought to have been re-engraved in the ninth century. Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 115, Cg.-480. 151 A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 118-119, Cg.-490. 152 Uttiramērūr’s tenth-century Murukaṉ temple houses a number of fragmentary sculptures that may date to the previous century.

94 

Construc ting K anchi

Illustration 25 Sundaravarada Perumāḷ Temple, Uttiramērūr, ninth century

Sandstone and the Cit y

95

Illustration 26 Sabhā, Uttiramērūr, base ninth century

range of fragmentary sculptures that could easily belong to the ninth century. Farther west is the handsome ninth-century Candramaulīśvara temple at Brahmadesam.153 Near Māmallapuram, apsidal ninth-century shrines and sculptures remain at Tirukkaḻukuṉṟam and Orakaṭam. North of Kanchi, the ninth-century Ādikeśava Perumāḷ temple stands at Kūram. At Puḷḷalūr’s brick-and-mortar Kailāsanātha temple, there are several granite pillars that closely resemble pillars at Uttiramērūr, as well as a possibly older disengaged sandstone figure of a male deity. Ninth-century Pallava temples also stand at Takkōlam, Tiruttaṇi, and Kālampākkam.154 Kāvērippākkam was yet another important center of temple construction in the ninth century.155 153 Srinivasan, ‘Pallavas of Kāñcī: Phase II’, 104. 154 Tiruttaṇi’s Vīraṭṭāṉeśvara temple is an important monument in the broader history of South Indian temple architecture. Ibid., 102-103; Valérie Gillet, ‘Gods and Devotees in Medieval Tiruttaṇi’, in The Archaeology of Bhakti II: Royal Bhakti, Local Bhakti, ed. Emmanuel Francis and Charlotte Schmid (Pondicherry: École Française d’Extrême-Orient/Institut Français de Pondichéry (Collection Indologie no. 125, 2016). 155 The ‘Kanchipuram Yoginis’, about which Kaimal has written extensively, are of the same deep gray stone as many sculptures in the Madras Government Museum that come from Kāvērippākkam. The stone is called metagabbro, according to Smithsonian conservation scientists. Padma Kaimal, Scattered Goddesses: Travels with the Yoginis (Ann Arbor, Michigan: Association for Asian Studies, 2012). Scholarship on these yoginis continues to develop. Katherine E. Kasdorf and Padma Kaimal, ‘Getting the Band Back Together: Reuniting the Kanchi Yoginis’,

96 

Construc ting K anchi

Illustration 27 Kanchi District Geological Map, blue showing sandstone below ground Courtesy GSI, Chennai

Sandstone and the Cit y

97

98 

Construc ting K anchi

While building activity by no means ceased in the ninth century, it seems to have taken place everywhere but Kanchi. The shifting political landscape from Pallava to Chola dominion accounts in part for the significant reduction in patronage for new temples inside the capital itself, but it may not be the only explanation.156 The availability of constructional materials may also have played a significant role. Several scholars have noted the change in primacy from sandstone to granite that occurred at the turn of the ninth century.157 They have not, however, questioned why the preferred material changed.158 High resolution maps produced at the Geological Survey of India (GSI) show that Kanchi is situated at the center of a basin of sedimentary rock, including sandstone and poor-quality shale (Ill. 27).159 However, sandstone that was once plentiful at surface level is now buried deep below ground. By the close of the eighth century, the Kanchi area was experiencing a diminishing supply of this natural resource. It might at first appear that the abandonment of sandstone was a matter of aesthetic preference or dynastic distinction, or that it stemmed from a desire to participate in a more widespread use of local granite in Tamil Nadu. However, the geological evidence shows that the shift in dominant construction materials was much more logistical in nature – the supply of sanstone was being exhausted. In scholarship, the question of identifying a stone source for Kanchi’s sandstone temples has been raised only in the context of conservation. in Annual Conference on South Asia (Madison, WI2019); ‘Reuniting the Tamil Yoginis: The Plans Take Shape’, Colgate University symposium, 16-17 October 2020. We now call them the Tamil Yoginis as opposed to the Kanchipuram Yoginis because they were not made in Kanchi. 156 For a summary of the turbulent transition primarily based on Pallava and Chalukya inscriptions, see Mahalingam, Kāñcīpuram in Early South Indian History, 186-227. 157 The EFEO scholars even cite the use of sandstone in support of an eighth-century date for the large-scale sandstone f igure of Vishnu that formerly stood overlooking the pond in the village of Ciṟutāvūr. Emmanuel Francis, Valérie Gillet, and Charlotte Schmid, ‘Trésors Inédit du Pays Tamoul: Chronique des Études Pallava II: Vestiges Pallava autour de Mahābalipuram et à Taccūr’, Bulletin de l’Ecole Française d’Extrême-Orient 93 (2006): 436-441. Sadly, the sculpture was absent at the time of my visit. The local priest reported it had been stolen, along with his temple’s electric generator and some lights (January 2014). 158 Srinivasan, ‘Pallavas of Kāñcī: Phase II’, 89. 159 I am grateful to Dr. S. Raju and Dr. Singanenjam at the GSI, Chennai, for their kind assistance (September 2015-February 2016), and to Mr. Muthusankar at the IFP and Mr. Kumar in Pondicherry for directing me there (May 2015). I am also grateful to Dr. Stefan Nicolescu, Collections Manager in the Division of Mineralogy at the Peabody Museum of Natural History, Yale University, and to Dr. Nicholas Christie-Blick, Professor in the Department of Earth and Environmental Studies, Columbia University, for first discussing this issue with me and suggesting the need for high-resolution geological maps (October-November 2014).

Sandstone and the Cit y

99

Illustration 28 Quarry field, Ārpākkam

The single study by D. Jayanthi left the precise location of the quarry sites inconclusive.160 ‘Authentic records on the source of sandstone procured for the construction of the monuments are not available’, writes the author.161 Jayanthi also makes note of the fact that no sandstone quarries are active anywhere in Tamil Nadu.162 Analysis confirms that Kanchi’s sandstone was local. Most of the stone’s grains are round and relatively large in size, which indicates a very short duration of transportation.163 While Jayanthi does not directly suggest a possible quarry site for Kanchi’s sandstone temples, he identifies five villages in the vicinity where sandstone can be found, buried below ground level.164 Whereas the vast quarry fields of granite used in the Chola period are situated in areas south of Kanchi, the sandstone supply

160 D. Jayanthi, ‘Conservation of Sandstone Monuments at Kanchipuram Tamilnadu India’ (PhD diss., Anna University, 2007). 161 Ibid., 110. 162 Ibid., 116. 163 Ibid., 64. 164 Tiruppakkuḻi, Kīḻāmpi, Mēlampi, Amarāvatipaṭṭaṇam, and Timmacamuttiram. The sandstone itself is ‘calcerous sandstone with fine-grained matrix of ferruginous material’ and belongs to the Upper Gondwana Age. Jayanthi’s focus being conservation, he suggests that the sandstone from these villages is suitable for use as replacement stones and for other repairs to the temples. Ibid., 113 and 123.

100 

Construc ting K anchi

lay to the northwest of the city – closest, in fact, to the sandstone Pallava temples and the location of the Pallava-era city. Since the flurry of construction throughout the eighth century seems to have used up the earlier supplies of local sandstone in Kanchi, in the ninth century the Pallavas consolidated their power by diverting funds to outlying areas that had plentiful resources of stone with which to build temples. Besides housing ninth-century monuments, what unites these sites is that they could each serve as their own quarry. The same was true of ninth-century sites much farther south with inscriptions dated to the regnal years of the Pallava kings, such as Tiruveḷḷaṟai and Ālampākkam (near Tiruchirappalli).165 Building in such resource-rich areas enabled the Pallavas to continue the cost-efficient practice of construction using local materials. This was not a new strategy for them. Built up over the course of the seventh and eighth century, Māmallapuram is a massive stone quarry.166 Experimentations with material then yielded new possibilities for architectural forms, giving rise to a wave of innovative designs. At Uttiramērūr, for example, architects combined the elite practice of building in stone with techniques familiar to them from domestic architecture – they topped granite basement moldings with brick-and-mortar walls. At Tiruveḷḷaṟai, temples were built on top of boulders, and the architects used the living rock as the floor. In doing so, they revived a practice introduced more than two centuries earlier at Māmallapuram’s Shore temple. The exhausted sandstone reserves did not dampen the verve for temple construction. Instead, patrons shifted their attention to places where elite sacred architecture could continue to flourish. When the Cholas gained control of Kanchi, they turned to the granite-producing areas that lay southeast of the city (Ill. 28).

165 Ālampākkam contains a Vishnu temple and a Saptamtrika temple both dating to the ninth century. The important site of Tiruveḷḷaṟi houses numerous temples and a spectacular well in the shape of a swastika with a long inscription dated to the regnal years of Dantivarman. Numerous additional sites in the Kaveri region contain Pallava-dated inscriptions and Pallavastyle sculptures from the ninth century. To what extent these places were directly influenced by the Pallavas is a matter of ongoing study, workshopped in Francis, Gillet, and Schmid, the Archaeology of Bhakti III: Minor Dynasties (Pondicherry, August 2015). 166 In addition to stone, quarry sites are typically rich in minerals and metals. They became settlements and sites of sacred activity because of these diverse economic resources. Deborah Stein, ‘Smelting Zinc and Housing the Divine at Jawar’, Artibus Asiae 72, no. 1 (2012). See also Mitch Hendrickson and Damian Evans, ‘Reimagining the City of Fire and Iron: A Landscape Archaeology of the Angkor-Period Industrial Complex of Preah Khan of Kompong Svay, Cambodia (ca. 9th to 13th Centuries AD)’, Journal of Field Archaeology (2015).

Sandstone and the Cit y

101

In the process, as I will discuss in the following chapter, they also shifted the city closer to their new sources of stone.

Conclusion: Foundations Laid The Pallavas emerged in the seventh and eighth century as one of the most powerful dynasties in South India. By building institutions in key locations, and by presenting themselves as elite patrons of architecture and art, they established Kanchi as a major cosmopolitan capital. Unlike the Cholas, who maintained multiple administrative headquarters, the Pallavas fixed their royal seat to a single place: Kanchi, the epicenter of their cosmos. From this primary urban node, the Pallavas asserted their claims to an expanding territory through military conquest and the construction of sacred architecture. In the process, they mapped out their own vision of a social and religious landscape. In the beginning of the eighth century, the construction or expansion of the Kailāsanātha, Ekāmbaranātha, Airāvateśvara, and Kacchapeśvara temples established the borders of the city. Less than a century later, the east-west extent of the city had nearly doubled. By the end of the eighth century, Kanchi stretched from the hilltop Candraprabhā Jiṉālāya to the Vaikuṇṭha Perumāḷ temple, which towered over the city’s new eastern frontier. The fragments in the two Karukkiṉil Amarntavaḷ Ammaṉ temples and the Rudrakoṭīśvara temple, as well as at Cevilimēṭu farther south, show that Kanchi also extended significantly in the direction of the Palar River. As more temples were established in the urban landscape, the city continued to grow. At the same time as the Pallavas gave shape to their capital, their sponsorship for architecture outside of Kanchi communicated the dynasty’s claim to a broader regional authority. Pallava-era temples are built on all sides of the city. Kūram, Takkōlam, and Tiruttaṇi lay to the north, Tirumukkūṭal and Māmallapuram to the east, Uttiramērūr and Paṉamalai to the south, and Brahmadesam and Kāvērippākkam to the west. The choice of locations for new temples reflects the Pallava’s political influence, as well as their desires to project their dynastic fame in the proverbial four directions.167 As 167 The conquest of the quarters (‘digvijaya’) is a pan-Indic trope in imperial praśasti. Ali has described Rajendra’s digvijaya, as it is elaborated in the Tiruvāḷaṅkāṭu copper plate charter, as ‘the Cōḻa family’s long courtship of the earth.’ Ali, ‘Royal Eulogy as World History: Rethinking Copper-Plate Inscriptions in Cōla India’, 173 and 211.

102 

Construc ting K anchi

Kanchi expanded into new territories, builders were tasked with selecting sites that had plentiful resources of water and stone. When Pallava power declined inside Kanchi over the course of the ninth century, places like Uttiramērūr played crucial roles. The Cholas would become Kanchi’s next rulers, but it was the Pallavas who laid the foundations for the city that still endures.

2 Realignment Kanchi in the Chola Era (ca. tenth through thirteenth century) Abstract Centering on the emergence of a royal road, Chapter Two reveals Kanchi’s underlying urban logic. All temples in the city faced this road, which established an enduring pilgrimage network that connected northern Tamil Nadu with the holy hilltop temple at Tirupati. Following the road to the villages north and south of Kanchi, the chapter demonstrates that changes in patronage structures in the rural hinterland supply valuable information concerning similar transformations taking place on a much larger scale and with greater complexity inside the borders of the city. Presented here for the first time, this chapter identifies, maps, and examines Kanchi’s Chola-era temples in the context of their contributions to the growth and intensification of the urban space. Keywords: Chola bronze, Procession, Varadarāja Perumāḷ/Varadaraja Perumal temple, Vaishnava, Shaiva, Goddess Kāmākṣī/Kamakshi

At the center of modern Kanchi, a large road called Kamarajar Salai runs north to south, bisecting the city along a broad avenue that is lined with shops, tea stalls, residences, and shrines. As Kanchi’s busiest thoroughfare, this road is plied with vehicles starting in the earliest hours of the morning (Ill. 29). However, Kamarajar Salai is more than a route through the urban core of Kanchi. It is also a main artery for religious processions. All streets, big and small, branch off from this avenue and lead to Kanchi’s many temples. In turn, the temples orient their principal gods towards the central avenue (Ill. 30). Temples to the west of Kamarajar Salai have sanctums that open towards the east. Temples to the east of Kamarajar Salai have sanctums that open towards the west. This centralized orientation is consistent in all temples built within the boundaries of the city during the last twelve hundred years, regardless of religious dedication, builder, date, or material. Stein, Emma Natalya, Constructing Kanchi: City of Infinite Temples. Amsterdam, Amsterdam University Press 2021 doi: 10.5117/9789463729123_ch02

104 

Construc ting K anchi

Illustration 29 The Axis of Access, Kamarajar Salai, Kanchi, from North

This chapter examines how Kanchi’s urban landscape was radically reoriented and reconfigured by this unique spatial logic during the Chola period (ca. 850-1275 CE).1 The city’s new rulers laid the main avenue and other new roads, and they allocated resources for the construction, expansion, and maintenance of dozens of temples within and around the city.2 In doing so, the Cholas realigned the old Pallava city to foster their own political connections and to establish new ritual networks. Unlike the Pallava monuments, which were built by members of the royal court, Kanchi’s Chola-era temples had more diverse sources of patronage. Acts of urban planning at the elite level connected the locally built temples with the imperial expansion of the city. Most notably, the northern extent of the main avenue connected Kanchi with Tirupati, an ancient pilgrimage site that remains the most holy destination for Vaishnava devotees in South India. By promoting pilgrimage and weaving an expanding web of political 1 The beginning of the Chola era is traditionally dated to ca. 850. The first recorded inscriptions in Kanchi to be dated in the regnal years of a Chola ruler correspond to ca. 922 CE, found on pillars in the maṇḍapa of the Kailāsanātha temple, followed by an inscription on the Yathoktakārī temple. Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 120/Cg.-497 and Cg.-98, and p. 153/Cg.-628. 2 Chola inscriptions designate roads of different scales, ranging from small paths in fields (‘vati’), to roads inside cities or between contiguous settlements (‘vaḻi’), to large highways that connected major places (‘peruvaḻi’). Heitzman, The City in South Asia, 48.

105

Realignment

Illustration 30 Map of temple orientations, Kanchi (map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution)

alliances, the Cholas cultivated the economic and social integration of their empire from a newfound seat in Kanchi.

Orienting the Gods In Italo Calvino’s fictitious city of Eudoxia, an underlying pattern is easily obscured by the busyness of daily life: In Eudoxia, which spreads both upward and down, with winding alleys, steps, dead ends, hovels, a carpet is preserved in which you can observe the city’s true form. At first sight nothing seems to resemble Eudoxia less than the design of that carpet, laid out in symmetrical motives whose patterns are repeated along straight and circular lines, interwoven with brilliantly colored spires, in a repetition that can be followed throughout the whole woof. But if you pause and examine it carefully, you become convinced that each place in the carpet corresponds to a place in the city and all the things contained in the city are included in the design, arranged

106 

Construc ting K anchi

according to their true relationship, which escapes your eye distracted by the bustle, the throngs, the shoving. All of Eudoxia’s confusion, the mules’ braying, the lampblack stains, the fish smell is what is evident in the incomplete perspective you grasp; but the carpet proves that there is a point from which the city shows its true proportions, the geometrical scheme implicit in its every, tiniest detail.3

Like Eudoxia, Kanchi’s urban structure adheres to a rigorous pattern that only becomes evident when the city’s layout is seen from above. Although some scholars have discussed the orientation of individual temples in Kanchi, none have noticed the consistent logic that governs temple orientation throughout the city. 4 Kanchi’s structure is unique within South India – a dense and wide distribution of temples defines the urban space, which is organized around the city’s central avenue. So-called temple-towns in Tamil Nadu tend to be organized around a single expansive temple complex (Ill. 31). A sequence of streets laid in concentric squares extends outwards from the main temple walls, like a ripple in water. This description applies to places such as Tiruvannamalai, Chidambaram, Madurai, and the island-city of Srirangam, but there are some notable exceptions.5 Often compared with Kanchi for its temples and its thriving sari industry, Kumbakonam is a multi-temple town between the Kaveri and Kollidam Rivers, with three large east-facing temple complexes built along an east-west road. However, Kumbakonam was never a royal capital, and it has far fewer temples than Kanchi. As I will discuss in Chapter Three, to find closer parallels to Kanchi’s urban form, one must turn farther afield within what was an integrated South and Southeast Asian region. Several places in Southeast Asia share important features with Kanchi. The great temple-cities of Polonnaruwa in Sri Lanka, Bagan in Myanmar, Si Thep in northeastern Thailand, Angkor, Koh Ker, and Sambor Prei Kuk in Cambodia, and Prambanan and Majapahit in Indonesia were contemporary with Kanchi’s urban efflorescence. Each of these places was a royal capital for the ruling dynasty, a center of commercial trade, and a multilingual and muiltireligious society. Each was 3 Italo Calvino, Invisible Cities (New York: Harcourt, Inc., 1972), 96. 4 The question of temple alignments across a broader South and Southeast Asian region has occupied scholars for generations. Recent work includes Veronique Degroot, Candi, Space and Landscape: A Study of Distribution, Orientation and Spatial Organization of Central Javanese Temple Remains (Leiden: Sidestone Press, 2009). 5 For an introduction to these sites, see George Michell, ed. Temple Towns of Tamil Nadu (Bombay: Marg Publications, 1993).

Realignment

107

Illustration 31 Aruṇācaleśvara Temple, Tiruvannamalai

also characterized by that same dense and wide distribution of temples, which were constructed primarily between the seventh and the end of the thirteenth century within an urban context. These cities were part of what Sheldon Pollock has described as a Sanskrit Cosmopolis, which was fostered by broader patterns of commercial trade, religious movements, and political enterprise.6 Kanchi’s parallels with this wider world may have to do with its status as a royal capital, its prominent Buddhist population, and its role in regional and transregional connections during a period of accelerated temple construction throughout South and Southeast Asia. Kanchi’s pluralistic structure has escaped scholarly notice. Some have accorded the Kāmākṣī Ammaṉ temple undue historical primacy because it is so popular today.7 However, the focus on Kāmākṣī Ammaṉ derives from conventional wisdom and certain expectations rather than from a more expansive and critical analysis of the city. Local histories centralize 6 Pollock, The Language of Gods in the World of Men: Sanskrit, Culture, and Power in Premodern India, esp. Chapter 3. 7 D. Dennis Hudson, ‘Kanchipuram’, in Temple Towns of Tamil Nadu, ed. George Michell (Bombay: Marg Publications, 1993), 23. In another essay, Hudson argues that the orientations of some of Kanchi’s Vishnu temples form a maṇḍala shape, ‘Ruling in the Gaze of God: Thoughts on Kanchipuram’s Maṇḍala’, in Tamil Geographies: Cultural Constructions of Space and Place in South India, ed. Martha Ann Selby and Indira Peterson (Albany: State University of New York Press, 2008).

108 

Construc ting K anchi

this temple and weave other sacred sites selectively into its story. Kanchi’s Māhātmya and Sthalapurāṇa texts (Sanskrit and Tamil legendary histories) give particular emphasis to the three largest complexes in the city and relegate the others to subsidiary positions, if mentioning them at all.8 Tamil mythology connects Umā (as Kāmākṣī Ammaṉ in Kanchi) with Shiva (Ekāmbaranātha) and Vishnu (Varadarāja Perumāḷ). Vishnu is understood to be Umā’s brother and therefore Shiva’s brother-in-law. In Kanchi, it is often said that both gods gaze towards the goddess. However, Ekāmbaranātha is situated to the northwest of Kāmākṣī Ammaṉ but opens directly east, and Varadarāja Perumāḷ is situated to the southeast of Kāmākṣī Ammaṉ but opens directly west. Therefore, neither god really faces the goddess.9 There are a mere handful of exceptions to Kanchi’s pattern of east-west temple orientations, each of which can be explained by one of a few factors. Exceptions can be due to the temple’s specific deity of dedication, its location in an area that was not yet part of Kanchi when the temple was constructed, or its more recent establishing or transformation.10 For example, three of the city’s goddess temples face north, which is the direction traditionally associated with the goddess throughout India. The only two significant exceptions to the pattern are the Vṛṣabheśvara and Airāvateśvara temples, both built in the Pallava period. As discussed in Chapter One, Vṛṣabheśvara’s western orientation may have to do with its original functions and specific sectarian affiliation. Also west-facing, Airāvateśvara once marked the eastern borders of the Pallava city, before the establishment of the Chola’s main road. 8 There are at least five Māhātmya and Sthalapurāṇa texts for Kanchi, in several editions. Dating such texts poses significant challenges, but scholars place them between the fourteenth and nineteenth century. Thanks to Ute Hüsken, Jonas Buchholz, and Malini Ambach for convening a workshop on Kanchi-related texts (Pondicherry 2020), and to Jonas especially for elucidating the sources. P.B.A. Charya, ed. Kāñcimāhātmya (Kanchipuram: Sri Sudarsana Press, 1906). A summary of the first book of Civañāṉa Muṉivar and Kacciyappa Muṉivar, Kāñcippurāṇam is in R. Dessigane, P.Z. Pattabiramin, and Jean Filliozat, Les Légendes Çivaïtes de Kañcipuram: Analyse de Textes et Iconographie (Pondicherry: Institut Français d’Indologie, 1964). Other texts include Kāñcīmāhātmya (Śrī Skāndapurāṇāntargataṃ Śrīkāñcīmāhātmyam Srīrudrakoṭimahimādarśaḥ), (Viijayavāḍa, Madras, Haidarabad: Vēṃkaṭrāma aṇḍ kō, 1967); Śrīkāmākṣīvilāsaḥ; Kaccālaiyar, ‘Kāñcippurāṇam’, in Āḷavantār Ātīṉak Kaccālaiyar Iyaṟṟiya Kāñcippurāṇam (Kampar Purāṇam), ed. S.K. Irāmarājaṉ (Chennai: Makāmakōpāttiyāya Ṭākṭar U. Vē. Cāminātaiyar Nūlnilaiyam, 1983). 9 For the mythical relationships among these temples and their deities, see Schier, ‘The Goddess’s Embrace: Multifaceted Relations at the Ekāmranātha Temple Festival, Kanchipuram’, as well as Ute Hüsken’s extensive research on the Varadarāja Perumāḷ temple and a forthcoming publication by Jonas Buchholz. 10 Stein, ‘All Streets Lead to Temples: Mapping Monumental Histories in Kanchipuram, ca. 8th-12th centuries CE’, 120-121.

Realignment

109

The Royal Center Kanchi’s Kamarajar Salai held the status and topographical position of the rājamārga – a royal road – of Sanskrit literature. The rājamārga axially defined the ideal city of Sanskrit drama and poetry (‘kāvya’) from as early as the first millennium CE.11 All sectors of society could be found on the rājamārga. It was where the poorest and the marginal wandered, and it was where royal and religious processions found greatest expression (Ill. 32). In the Mattavilāsa Prahasana, the seventh-century play by the Pallava king Mahendravarman I (ca. 580-630 CE), Kanchi’s royal avenue is described as a crowded place filled with good people.12 Watching the Buddhist monk cower nervously over his alms bowl, which is filled with roasted meat (Buddhists in South India were supposed to be vegetarian), the Kāpālinī remarks, ‘Lord! Look, look! This red-robe is cautiously walking quickly in the rājamārga, which is filled with trustworthy people, his limbs all hunched up and his glances shooting to both sides.’13 Like the rājamārga described in kāvya, Kanchi’s main road was ‘a site of urban rhythms.’14 In literary sources, such as Sanskrit kāvyas and Tamil epics, the southern end of the rājamārga meets an entrance to the city, which is often marked by a gate, and the northern end grants access to the royal palace. The axial nature of the rājamārga was therefore not only due to the road’s central position, but also its municipal function as a pathway to the palace.15 Little archaeological evidence of palaces survives anywhere in India from before the eighteenth century. This absence is in part because, as domestic structures, palaces were built of perishable materials, such as timber and brick. The fifteenth-century Man Mandir palace at Gwalior, Madhya Pradesh, made of stone, is an example made famous by its rarety. Another exceptional example of palace architecture survives at Gangaikondacholapuram, where the brick foundation structure has been exposed through excavation.16 In addition to having tenuous structures that did not endure, the role of palaces 11 Kaul, Imagining the Urban: Sanskrit and the City in Early India, 91. ‘King among roads’ is a more accurate translation for rājamārga or rājavīdhi than ‘royal road.’ Katherine E. Kasdorf, ‘Forming Dorasamudra: Temples of the Hoysala Capital in Context’ (PhD diss., Columbia University, 2013), 69-70. 12 For discussion of the Mattavilāsa Prahasana (MP), see Chapter One. 13 Mattavilāsa Prahasana, line 65. 14 Kaul, Imagining the Urban: Sanskrit and the City in Early India, 92. 15 Ibid. Kasdorf shows that Dōrasamudra’s royal road can be traced through a concentration of monuments built of permanent materials along a particular road in the eastern part of the Hoysaḷa capital. Kasdorf, ‘Forming Dorasamudra: Temples of the Hoysala Capital in Context’, 68-70. 16 The excavation is reported in IAR 1955-1956, p. 27.

110 

Construc ting K anchi

Illustration 32 Procession for the Goddess, Kamarajar Salai, Kanchi

as administrative headquarters for ruling dynasties made them attractive targets for attack in times of conflict. Representations in relief carvings and frequent mentions in a range of literary genres begin to flesh out the picture of palace architecture. Architectural treatises, epics, purāṇas, kāvyas, and other texts alike present the royal complexes as vast, sprawling compounds that are heavily fortified with guarded gates.17 The Sanskrit travel novel Daśakumāracarita (ca. eighth century), gives a wonderfully detailed description of a palace complex.18 This text holds special significance here because its celebrated author, Daṇḍin, was born in Kanchi during the period of the Pallava king Rājasiṃha (ca. 700-725 CE), builder of the great Kailāsanātha temple. In the Daśakumāracarita, desirous to meet with his beloved in the most secluded part of the palace – the harem – the hero first crosses a moat and skillfully 17 Aṭikal, The Cilappadikāram; Dandin, Daśakumāracarita [‘What Ten Young Men Did’], trans. Isabelle Onians, Clay Sanskrit Library (New York: New York University Press, 2005).; Bruno Dagens, ed. Mayamatam: Treatise of Housing, Architecture, and Iconography, 2 vols., vol. 1-2 (New Delhi and Delhi: Indira Gandhi National Centre for the Arts and Motilal Banarsidass Publishers Pvt. Ltd., 2007); and the Periya Purāṇam. 18 Dandin, Daśakumāracarita, esp. Chapter 8.

Realignment

111

scales a rampart.19 He then climbs up ‘a flight of steps built from piled baked bricks [that] led to the upper floor of the principal gate.’ Upon descending, he turns down avenue after avenue of tree-lined garden paths, while ‘feeling the bulging wall of the enormous palace.’20 In the twelfth-century hagiography of the Tamil saint Appar, petitioners assemble at Kanchi’s palace gate and must explain their purpose to the guard, who serves as an intermediary.21 The Caṅkam era (ca. first century BCE-sixth century CE) epic Cilappatikāram [‘Tale of the Anklet’] includes an episode in which the anklet’s thief encounters watchmen at the gates to Madurai’s palace.22 Across genre and period of composition, texts place the palace in a centrally located position in the northern part of the city.23 Could the same have been true in Kanchi? Several inscriptions affirm the existence of a Chola palace at Kanchi. An important copper plate charter records an order made by the king while seated in the citramaṇḍapa (‘painted hall’) of his golden palace at Kanchipuram, regarding the expenditures and income of the Ulakaḷanta Perumāḷ temple.24 A twelfth-century stone inscription on this temple records a visit by the king Kulottuṅga I, together with two of his wives, during which he made a gift of land to the temple for cultivation by a community of weavers.25 Having seen that the weavers had been negligent, the king regifted the land to the temple. An eleventh-century inscription that is prominently placed on the stately basement moldings of the Bṛhadēśvara temple at Gangaikondacholapuram records an order given by the Chola king Vīrarājendra while sitting in his palace in Kanchipuram.26 The content 19 Dandin, Daśakumāracarita, line 8.75-8.82, p. 301-303. Incidentally, the beloved has fallen in love with the hero after seeing his portrait (line 8.36-8.42). 20 Ibid. 21 Periya Purāṇam, line 1350. 22 Aṭikal, The Cilappadikāram, 224-225. The walled and gated palace formation seems even to have been emulated in temporary architecture associated with military travel. Elsewhere in the text, the Cēra king on a military campaign resides in a portable palace encircled by textiles supported by wooden stakes and guarded by gatekeepers. Ibid., 363. 23 The Cilappadikāram provides a long description of the port-city of Pukār (Canto V). Mapping out the different quarters reveals that the palace is situtated in the northern quarter of the city. Ibid., 122. 24 For discussion of this charter (Uttama Cōḻa, 986 CE), see Nagaswamy, Viṣṇu Temples of Kāñcīpuram, 19-24. 25 The inscription is dated in the fourth regnal year, day 40, of Kulottuṅga I (c.1110 CE). ARE 1921, no. 39, and part II, p. 98; Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 135, Cg.-553. 26 The record is on the northwest corner of the monument and is dated in the king’s fifth regnal year (ca. 1068 CE). ARE 1892, no. 82; SII Vol. 4, p. 156-157, no. 529; Mahalingam, A Topographical

112 

Construc ting K anchi

Illustration 33 Map of maṇḍapa locations in central Kanchi

relates to a number of villages assigned as devadānas (‘Gifts to the Gods’) for the expenditures of the temple at Gangaikondacholapuram. From these three inscriptions, we learn that Kanchi was counted among the places where the Chola ruler could temporarily reside, and from which he could conduct his courtly affairs. We also learn that the Chola-Kanchi palace was built and functional by at least the tenth century, though nothing of its structure survives. Heitzman and Champakalakshmi have extrapolated that Chola-era inscriptions identify a busy commercial area that more or less corresponds to Kanchi’s present-day urban core. It is unfortunate that neither scholar supplies details nor proposes a specific location for the palace, but the built environment may tell us more.27 At the place that has remained the immediate center of the northern part of Kanchi, the main avenue arrives at an area where the broad Raja Veedhis (‘kings among streets’) are laid in an approximate square around a noticeably elevated area (Ill. 33). Several factors suggest that the elevated

List of the Inscriptions in the Tamil Nadu and Kerala States: Tiruchchirappalli District, 8 vols., vol. 8 (New Delhi: Indian Council of Historical Research, 1991), p. 322-323, Tp.-1460. See also Daud Ali, ‘The Epigraphical Legacy at Gangaikondacholapuram: Problems and Possibilities’, in New Dimensions in Tamil Epigraphy, ed. Appasamy Murugaiyan (Chennai: Cre-A, 2012). 27 Champakalakshmi, ‘The Urban Conf igurations of Toṇḍaimaṇḍalam: The Kāñcīpuram Region, C.A.D. 600-1300’; Heitzman, ‘Urbanization and Political Economy in Early South India: Kāñcīpuram During the Cōḻa Period.’

Realignment

113

Illustration 34 Maṇḍapas, juncture of Kamarajar Salai and South Raja Veedhi

space enclosed by these broad streets was the site of the Chola palace.28 While palaces themselves were made of brick, the surrounding gates and walls were sometimes made of stone, as we know from Tamil literature and rare architectural remains.29 The southwest, northwest, and southeast corners of the Raja Veedhi square are marked with the remains of maṇḍapas. Two barely visible maṇḍapas also flank Kamarajar Salai as it crosses the southern Raja Veedhi (Ill.34). These maṇḍapas are now sealed-wall structures housing shops and street-side shrines, with walls nearly completely covered with paint and billboard advertisements. However, upon careful inspection, their composition using uneven blocks of stone in a range of stylistic developments indicates that they are built in part from recycled architectural materials, and that they have been built up and transformed over time. The position of the maṇḍapas at key junctures suggests that they may have first been constructed at approximately the same time that the Raja Veedhis were laid. The two maṇḍapas that flank Kamarajar Salai are perfectly positioned to be parts of an entryway to the palace precincts. The corner maṇḍapas also 28 Hudson also proposed this site as a palace area, but during the era of the Pallavas rather than Cholas. As I demonstrated in the first chapter, the Pallava city was not located this far to the east, which makes his suggestion unlikely. However, we cannot rule out the possibility that there was a Pallava palace on the same location. The Pallavas may have had several royal residences in the city. Hudson, ‘Kanchipuram’, 23. 29 Tamil Poetry through the Ages: Pattuppāṭṭu, 2, 169.

114 

Construc ting K anchi

originally may have been designed with reference to the royal establishment. Some scholars have suggested that the Kāmākṣī Ammaṉ temple is located at the highest point in the city. However, the greatest elevation (93 meters) lies between the Kāmākṣī Ammaṉ and Ulakaḷanta Perumāḷ temples, within an area that today is a dense labyrinth of shops, shrines, and residences.30 It seems unlikely that there would not have been something of significance there to justify the precise positioning of the Raja Veedhi square, yet no singular edifice survives to mark the center. Perhaps what endures is a royal presence by virtue of its conspicuous absence from Kanchi. A Royal Temple? In the Kaveri heartland, the Cholas established a royal temple in each successive place they considered their capital (Thanjavur, Gangaikondacholapuram, Dārāsuram, Tirupuvaṉam). In addition to the references to a Chola palace in Kanchi discussed above, inscriptions also referred to two distinct maṇḍalas (here meaning ‘zones’) of dominion – Cōḻamaṇḍalam in the Kaveri region, centering wherever their current capital was, and Jayaṅkoṇṭacōḻamaṇḍalam in the north, centering on Kanchi. The geographical nomenclature makes it clear that the Cholas considered Kanchi to be their northern capital for a period of several centuries, beginning at least by the year 922, when the first inscriptions in Kanchi dated in a Chola ruler’s regnal years are known.31 That the royal residence did not survive is not altogether surprising. What is surprising is that among the many temples to remain from the era of Chola reign in Kanchi, none are known to be royal foundations.32 Considering Kanchi’s legacy as a royal and ritual center, it would stand to reason that a Chola king would have wanted to establish a temple to mark his conquest of the past. One finds throughout Tamil Nadu an enduring contrast between the usage and upkeep of ‘local’ and ‘royal’ temples, as Francis and Schmid have shown.33 Kesavan Veluthat has also distinguished between temples built by royal patrons as explicit statements of power and others that developed as 30 ‘Kanchi Topographic Map’, https://en-ca.topographic-map.com/maps/87a/Kanchipuram/. 31 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 120/Cg.-497 and Cg.-98. 32 Many thanks to Valérie Gillet for discussing this topic with me. The question of a Chola royal temple in Kanchi emerged during a conversation with her in Pondicherry (September 2015). 33 Francis, ‘Royal and Local Bhakti under the Pallavas.’ For further explorations of this topic, see the essays in Charlotte Schmid and Emmanuel Francis, The Archaeology of Bhakti II: Royal Bhakti, Local Bhakti (Pondicherry: Institut Français de Pondichéry and École Française d’Extrême-Orient, 2016).

Realignment

Illustration 35 Bṛhadēśvara Temple, Thanjavur, ca. 1010 CE

115

116 

Construc ting K anchi

the centers of agrarian corporations and Brahmin settlements with claims to deep religious histories.34 Royal temples tend to become frozen in time as vestiges of an imperial past, and they lack the kind of local flavor so prominent at other temples. This can be seen in the royal Chola edifices, where the temple’s monumental scale far outweighs the local community’s limited ritual requirements (Ill. 35). Local temples vary widely in size. The scale of any given temple reflects the needs of the devotional community that sponsored its construction and continues to maintain its ritual life. Two ca. ninth-century temples in the village of Kīḻappaḻuvūr exemplify the distinction between royal and local temples. While the locally patronized Tēvāram temple of Ālantuṟai continues to flourish as the main village shrine, the royal Paśupatīśvara temple, founded by the Paḻuvēṭṭaraiyar dynasty, sits unused and forgotten. Foundation inscriptions are exceedingly rare in Tamil temples. But temples could be explicitly royal in other ways. In Kanchi, the Kacchapeśvara and Ekāmbaranātha temples, and the monumental but little-known Jvarahareśvara temple, present themselves as possibilities for a royal Chola temple. Fragmentary remains found on the premises of Kacchapeśvara and Ekāmbaranātha reveal that both of these temples have longer histories stemming back to the Pallava period. By contrast, Jvarahareśvara bears no evidence of a pre-Chola past. Any of these sites would have made a suitable place for the Cholas to stake their official presence in the city through the expansion of a Pallava foundation or the construction of a new royal temple. In the absence of further evidence, however, the question of which temple must remain a matter of speculation, as follows. The Kacchapeśvara’s central position, across from the Pallava Airāvateśvara temple and at the midpoint between the Kailāsanātha and Vaikuṇṭha Perumāḷ temples, would have made it a prime location for a major Chola monument. While Kacchapeśvara is less lofty in scale than the royal edifices at the other Chola capitals, it is among the largest and most complete temples in Kanchi. A history as a royal monument might help explain the site’s relative tranquility today. Despite Kacchapeśvara’s large footprint and its mention on street signs in the city, this temple consistently receives far fewer visitors than Kāmākṣī Ammaṉ, Ekāmbaranātha, and Varadarāja Perumāḷ.35 Spanning two centuries, Kacchapeśvara’s thirteen noticed inscriptions (ca. 990-1181 CE) reveal a level of continued interest in the temple, as well as 34 Veluthat, The Early Medieval in South India, 72-73. 35 One evening, a Kanchi resident told me that the Kacchapeśvara temple is where she goes when she needs some peace and quiet (January 2014).

Realignment

117

Illustration 36 Durgā Shrine at Kacchapeśvara Temple, Kanchi, tenth century

the establishment’s relative prosperity throughout the Chola period. The temple held tax-free land that was gifted by the king, and it also employed dancers who did well enough to make their own donations. An inscription on the south wall of the ardhamaṇḍapa (‘vestibule’) records gifts of lamps by one of the temple’s dancers together with her sister and her daughter.36 The primary recipient of donations at Kacchapeśvara, however, was not Shiva but the goddess. Durgā is enshrined in her own ca. tenth-century structure that stands a good distance from the main temple body, in the northeast corner of the complex (Ill. 36). Not only is this building covered in donative inscriptions, most of which date to the late tenth and eleventh century, but additional records of donations to this goddess appear on the walls of the main temple as well.37 Given the relative paucity of donations to Kacchapeśvara as opposed to the goddess, it is possible that the Durgā shrine was an earlier center of local devotion that was later converted to stone. The presence of a massive, ancient tree, many nāga stones, and a Pallava-era goddess enshrined in the northwest 36 The inscription is dated to the sixth regnal year of Rājādhirāja II (ca. 1169 CE). Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 139, Cg.-568. For a seminal study of the role of women as Chola patrons, see Leslie Orr, Donors, Devotees, and Daughters of God: Temple Women in Medieval Tamilnadu (New York: Oxford University Press, 2000). 37 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 125, Cg.-515 and p. 127-128, Cg.-524.

118 

Construc ting K anchi

Illustration 37 Jvarahareśvara Temple, Kanchi, twelfth century

corner of the complex also indicates a long history of the site, one particularly associated with the worship of goddesses and autochthonous deities. It is possible that the Durgā shrine was converted through local, collective donations into stone in the late tenth century. Shortly thereafter, the Shiva temple was established, but it would never receive the same level of devotion. The Ekāmbaranātha temple is the second possibility for a royal Chola edifice. It was already an important temple long before the beginning of the Chola period, and its close connections with the Pallava imperial past are seen through the presence of Mahendravarman-inscribed pillars and the Pallava-period Vṛṣabheśvara shrine on the temple premises (see Chapter One). Similarly, its proximity to the royal Kailāsanātha temple, but on a larger campus, as well as its inclusion in royal Pallava literature, would have made it an attractive contender for expansion or renewal in the post-Pallava period. Although the inner temple body was entirely reconstructed in the early twentieth century, a great many fragmentary inscribed slabs suggest that building activity, and possibly extensive renovations, occurred during the Chola period.38 However, the Ekāmbaranātha’s association with the 38 My thanks to Crispin Branfoot for drawing my attention to records of early twentieth-century reconstruction projects by the Nattukottai Chettiar community in Tamil temples. Crispin

Realignment

119

Tēvāram could have deterred the Cholas from using it as a royal site, as Tēvāram places tended to receive their patronage on the local level.39 Kanchi’s third possibly royal Chola temple contrasts strikingly with the other two monuments. With an imposing structure and ornate sculptural form, the Jvarahareśvara temple bears no pre-Chola fragments and remarkably few post-Chola additions (Ill. 37). The large-scale and elaborately carved structure of the all-stone temple reveals it was an elite commission from the highest echelons of society. Today the temple is under the protection of the Archaeological Survey of India (ASI). The virtuosic circular structure of its vimāna and the ornamental wall carvings throughout the temple earned it a place in the Encyclopaedia of Indian Temple Architecture. 40 ASI protection may have saved this temple from enduring a similar fate to royal commissions, such as the Paḻuvēṭṭaraiyar’s temple in Kīḻappaḻuvūr, which fell out of use with the dynasty’s end. Despite its location on the street that leads to Ekāmbaranātha, few devotees visit Jvarahareśvara today, and most of Kanchi’s residents do not even know of this temple. Those who do arrive have come specifically for pūjā of Kubera, a freestanding sculpture with indeterminate date inside the main sanctum. Images of Kubera – the god of prosperity who is often associated with kingship – are also carved in relief on the exterior wall of the ardhamaṇḍapa, prominently placed to flank the door. Similarly quiet in the Chola period, Jvarahareśvara was not the recipient of the abundant pious donations that flowed into other shrines in the city. Unlike most of Kanchi’s temples, the walls are all but devoid of inscriptions. Although it remains unfinished, the temple’s lack of finish was not the reason it is uninscribed. Nor would it have detracted from the temple’s status as a royal Chola monument. The Bṛhadēśvara temples at both Thanjavur and Gangaikondacholapuram are famously incomplete. Despite their lack of finish, their basal moldings are replete with inscriptions, which document the extensive staff and the holdings allotted to the temples. Only two inscriptions are present at Jvarahareśvara. The earlier of the two is situated on the jagatī molding of the north adhiṣṭhāna (‘basal moldings’),

Branfoot, ‘Remaking the Past: Tamil Sacred Landscape and Temple Renovations’, Bulletin of the School of Oriental and African Studies (January 2013). 39 Francis, ‘Royal and Local Bhakti under the Pallavas.’ 40 K.R. Srinivasan, ‘Cōḻas of Thanjavur: Phase III’, in Encyclopaedia of Indian Temple Architecture, ed. Michael W. Meister and M.A. Dhaky (New Delhi: American Institute of Indian Studies, 1983), 320-322.

120 

Construc ting K anchi

and it registers a gift of land to the temple.41 The twelfth-century date of this inscription matches well with the architectural style to give us an approximate date for the temple’s construction. The timing of Jvarahareśvara’s construction would have coincided with a prosperous period of Chola reign, when the dynasty had assumed power over vast territories, among them the city of Kanchi.42 Although none of Kanchi’s temples are explicitly royal Chola monuments, the Jvarahareśvara presents itself as most likely for a royal temple. The Royal Road As the Cholas expanded their influence in Kanchi, they built a main road that took the place of an older road, situated one kilometer to the west, that has since become more or less forgotten. The earlier route proceeded from the Pallava’s Kailāsanātha temple area southwards to the ancient settlement of Māmaṇṭūr, a site also important for the Pallavas (see Chapter One) (Ill. 38). The old road is lined with temples and archaeological remains that date to the Pallava period or even before. By contrast, the new road is lined with Chola-period temples all along the pathway from Kanchi south to Uttiramērūr. As the Cholas rose to prominence, the Pallava-era monuments and the associated road seems to have decreased in importance. The city of Kanchi then shifted eastwards and was reconfigured, such that the new road and new temples assumed the city’s new center. The road to Uttiramērūr did not entirely eclipse the Māmaṇṭūr road. Instead, both remained in use but with differing levels of activity. For example, the village of Kaḷakkāṭṭūr, which is situated closer to the Uttiramērūr road, contains both Pallava-period and Chola-period remains. Likewise, Perunakar, a megalithic site along the Māmaṇṭūr road, houses a relatively large-scale twelfth-century temple covered in Chola-era donative records. Nevertheless, there seems to have been a marked decrease in activity along one road and increase in activity along the other, as evidenced from the relative concentrations of architecture, sculpture, and inscriptions along the two roads. A large portion of the old road from the Pallava city to Māmaṇṭūr still exists today as the Kanchipuram-Vandavasi highway. The highway begins to 41 The inscription is dated to the fifth regnal year of Vikrāmacōḻa (ca. 1123 CE). Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 137, Cg.-559. The second inscription is on the south adhiṣṭhāna and registers a gift of tax on looms in Kanchi. Ibid., p. 148, Cg.-604. 42 For developments in the art of the twelfth century, see K.A. Nilakanta Sastri, The Cōḻas (Madras: University of Madras, (1935 [1955]); Vidya Dehejia, Art of the Imperial Cholas (New York: Columbia University Press, 1990), 93-126.

121

Realignment

Illustration 38 Map of old and new main roads (map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution)

the south of the Cheyyar River and proceeds north towards Kanchi, passing the megalithic burial site of Cēttuppaṭṭu. After traversing the river, it runs near to the cave-temples and lake at Māmaṇṭūr, and then to the village of Mēṉallūr, where a ca. seventh-century sandstone Durgā can be seen in front of the paddy-side Kāḷattīśvara temple. 43 The road continues towards the village of Kuraṅkaniḷmuttam, where a single cave-temple sponsored by Mahendravarman I (ca. 580-630 CE) sports the earliest representation of the horned dvārapāla, a type of Shaiva guardian specific to Pallava art. 44 Next follows a series of places established for different types of communities. 45 Pallavapuram is known as the place where the royal family lived, perhaps as a summer palace retreat rather than an administrative center. The village of Tuci is said to have been the military’s residence, the name referring to the dust (‘tūci’) stirred up in military marches. Cevilimēṭu 43 The temple itself is a modern concrete structure. 44 For discussion of dvārapālas as personifications of deities’ attributes (‘āyudhapurushas’), see Lockwood, Pallava Art, 7. 45 The local etymologies shed light on the function each settlement had during the Pallava period. I am grateful to A. Prabu, A. Valavan, and their father Annadurai for bringing me to Kuraṅkaṇilmuṭṭam and Māmaṇṭūr, and for discussing the village names (October 2013).

122 

Construc ting K anchi

Illustration 39 Map of temple locations, road from Uttiramērūr to Tirupati (Map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution)

– where the Narasiṃha Perumāḷ temple contains a Pallava-era lion-based pillar – is known as the servants’ village (from the Tamil cey, ‘to serve’). The presence of many Pallava-era archaeological remains and inscriptions within and between these places gives relevance to the local etymologies. The present-day Kanchipuram-Vandavasi highway (the Māmaṇṭūr road) forks off at Cevilimēṭu. The western segment continues to connect Pallava sites by bypassing Kanchi and proceeding northwest towards Kūram, which contains the earliest structural temple from the Pallava era. The eastern segment eventually joins Kamarajar Salai without passing any monuments along the way. However, it may not always have veered off in this manner. If the Kanchipuram-Vandavasi highway were to continue directly north from Cevilimēṭu, it would soon connect with present-day Krishnan Street, a north-south running avenue in the western part of Kanchi. Krishnan Street is lined with architectural remains of varying dates, including the Kāyārohaṇeśvara temple, the two Karukkiṉil Amarntavaḷ Ammaṉ temples, the Rudrakoṭīśvara and Cidambareśvara temples, Tirumēṟṟaḷinātha, and a pair of shrines dedicated to Ganesha and Ayyaṉār. Most of these sites contain

Realignment

123

Pallava-era fragments or inscriptions. After passing these monuments, Krishnan Street arrives just slightly east of the Kailāsanātha temple. In contrast to Krishnan Street and the Kanchipuram-Vandavasi highway, Kamarajar Salai splices the city and extends from the current center directly south to cross the Vegavathi and then the Palar and Cheyyar Rivers. All the way to Uttiramērūr, it passes through places that contain Chola-period (as opposed to Pallava) temples and inscriptions. Unlike Krishnan Street, it continues onwards as the Kanchipuram-Uttiramērūr highway without any interruption, hiatus, or diversion. This broad road leads to villages including present-day Kaḷakkāṭṭūr, Ārpākkam, Mākaṟal, Tiruppulivaṉam, and finally Uttiramērūr (Ill. 39). At Uttiramērūr, three magnificent temples and a vast inscriptional corpus preserved on the sabhā (‘assembly hall’) reveal the village’s gradual rise to prominence during the ninth century – precisely the time of Kanchi’s transition from Pallava to Chola dominion. 46 A remarkably vast ēri (‘reservoir’) called Vairamegha Taṭākam provides water for the village and its agrarian lands. 47 Whereas the other large lakes in the Kanchi area have bunds of no more than three kilometers in length, Uttiramērūr’s stretches seven kilometers along its eastern edge. A bund is a raised stretch of earth that separates the lake from the paddy fields – irrigation is controlled by sluices that enable water to pass through the bund in regulated quantities. Villages in Tamil Nadu were typically positioned to the east of lakes in order to take advantage of the land’s natural slope for irrigation.48 Satellite imagery reveals that Uttiramērūr’s lake nearly converges with the one at Tiruppulivaṉam and may well have done so in times of plentiful rain. That this tank is first mentioned in inscriptions from the ninth year of Dantivarman’s reign (ca. 805 CE) suggests that it was constructed in the early ninth century. 49 To create this ēri and build the temples would have required an enormous mobilization of resources. The nearby villages of Mākaṟal, Ārpākkam, and Tiruppulivaṉam also house important temples that received large endowments for construction and maintenance throughout the Chola period (Ill. 40). The primary 46 For an informative study of Uttiramērūr, see François Gros and R. Nagaswamy, Uttaramērūr: Légends, Histoire, Monuments (Pondicherry: Institut Français d’Indologie, 1970). This collaborative volume includes descriptions of temples, inscriptions, and ritual life, as well as useful maps. 47 The lake’s ancient name is given as a boundary marker in Uttiramērūr’s inscriptions. Ibid., 93. 48 Heitzman, ‘Urbanization and Political Economy in Early South India: Kāñcīpuram During the Cōḻa Period’, 122-124. 49 Minakshi, Administration and Social Life Under the Pallavas, 97.

124 

Construc ting K anchi

Illustration 40 Vyāghrapureśvara Temple, Tiruppulivaṉam, eleventh century

temples in these three villages are apsidal in shape, as is the temple at Perunakar. Through the use of a shared architectural language, the repeated structure creates a network in the mind of the traveling pilgrims who visited the sites. Not far west of the highway is the original site of Kaḷakkāṭṭūr village, where the ca. tenth-century Agnīśvara temple sits on the banks of another massive ēri.50 In addition to the Chola temple, Pallava- and Chola-era sculptures, architecture, and inscriptions can be found in this village in fragmentary form. Some are partially buried in fields and others haphazardly installed in quiet shrines (Ill. 41). One inscription mentions that a Subrahmāṇya temple (not surviving) was also in this village.51 The former presence of multiple temples, together with the impressive scale of the ēri, suggests 50 Balasubrahmanyam dates the Agnīśvara temple to the reign of Āditya I (ca. 871-907 CE) based entirely on the sculptural style of the Ardhanārīśvara. The architectural style and earliest inscriptions instead support a tenth-century date. S.R. Balasubrahmanyam, Middle Chola Temples: Rajaraja I to Kulottunga I (AD 985-1070) (Amsterdam: Oriental Press, 1975), 172-74; Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 113-14, Cg.-471-77. 51 A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 114, Cg.-471-477.

Realignment

125

Illustration 41 Goddess buried near Agnīśvara and Mātṛkā Temples, Kaḷakkāṭṭūr

Kaḷakkāṭṭūr was a prosperous place. The village is positioned between the old and new roads and both may have granted access to it at various times. In relatively recent history, however, the settlement relocated away from the ancient site to be positioned directly along the Uttiramērūr road. This move recapitulates earlier practices of relocalization that take place over time as roads come into and fall out of use.52 It may well have been that the road from Uttiramērūr to Kanchi grew to prominence as power was consolidated outside of Kanchi and the Pallava monuments within the city declined in importance. When the Cholas supplanted the Pallavas as Kanchi’s rulers, the change in the urban landscape 52 For discussion of a temple that was ‘relocalized’ in order to be closer to the road to Tirupati, see Gillet, ‘Gods and Devotees in Medieval Tiruttaṇi.’ Accessing the ancient site of Kaḷakkāṭṭūr now requires a two-kilometer walk along the high bund. According to the village priest, some three hundred years ago there were plans to pave this road, but the lake swelled in retaliation and broke the bund, flooding the village. Frightened, the residents fled and shifted the village to its present location, closer to the Uttiramērūr road. In the process, the village seems to have shrunk significantly, as the now-overgrown expanse of the earlier village is quite extensive. Like many of the sites discussed in this chapter, Kaḷakkāṭṭūr is seriously worthy of its own study. The Chola temple and various shrines are maintained by the single priest, who is both generous and highly knowledgeable about the temple history and architecture. I wish to express my gratitude to him for his careful maintenance of the site, and for his hospitality during my visit (February 2016).

126 

Construc ting K anchi

became permanent. Inscriptions in Kanchi began to be dated in the regnal years of the ruling Chola king in the first quarter of the tenth century.53 It was shortly thereafter that Kamarajar Salai must have begun to function as the centerpiece of the city in Chola-Kanchi. The road was so important that it is even specifically mentioned in an inscription at Kanchi’s Varadarāja Perumāḷ temple.54 From this road, all other streets branched, winding their ways through residential enclaves, and always leading to temples. For these reasons I call the road an ‘axis of access’ – a central artery that granted passage into and out of the heart of the city.55

Pilgrimage and Processions The axis of access continues beyond Kanchi in both southern and northern directions. On the northern side, it passes the villages of Kūram, Paḷḷūr, Puḷḷalūr, Takkōlam, and Tiruttaṇi. Like the areas along the road south from the city, each of these places houses important architecture, sculpture, and inscriptions that range in date from the ninth century well into the Chola period. Temples along the road are dedicated to a range of deities, including Shiva, Vishnu, Murukaṉ, and the goddess, as well as local gods. 60 kilometers beyond Tiruttaṇi, the road reaches the hilltop Veṅkaṭeśvara temple at Tirupati. This temple was established by the time of the Vaishnava Saints, and it was a devotional destination by the Chola period. Gillet has pointed out two Cholaperiod inscriptions in Tiruttaṇi village that mention the road to Tirupati.56 She has further demonstrated that an entire Vishnu temple was ‘relocalized’ in order to place it nearer to the road, which was quickly becoming a significant pilgrimage route.57 Tirupati then rose to major prominence as a Śrīvaiṣṇava center during the fourteenth century. Patrons from the Vijayanagara kingdom’s Saluva dynasty sponsored most of the structural fabric of the temple that 53 Kanchi’s first Chola-period inscriptions are on pillars in the Kailāsanātha temple’s maṇḍapa. Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 120, Cg.-497. 54 Heitzman, ‘Urbanization and Political Economy in Early South India: Kāñcīpuram During the Cōḻa Period’, 127, n. 10. 55 I borrow the expression ‘axis of access’ from Michael W. Meister, ‘The Hindu Temple: Axis of Access’, Concepts of Space, Ancient and Modern (1991). Meister describes the experience of the Hindu temple as a physical movement inwards along a horizontal axis and a conceptual movement upwards along a vertical axis that begins at the center of the inner sanctum and ascends to the pinnacle of the superstructure. 56 Gillet, ‘Gods and Devotees in Medieval Tiruttaṇi’, 470. 57 Ibid.

Realignment

127

Illustration 42 Veṅkaṭeśvara Temple, Tirupati

stands today.58 Popularly known as the Golden Temple, Tirupati has become the most holy destination for devotees of Vishnu in South India and one of the most prosperous religious establishments in the entire subcontinent (Ill. 42). The connection to Kanchi through patterns of pilgrimage remains strong. Large groups of devotees with shaven heads frequently arrive at Kanchi’s temples. They have been to Tirupati, where tonsure is a principal form of devotion. Other groups wear certain garments to announce that they are on their way. For Shaivas, pilgrimage assumed a new level of prominence in the twelfthcentury Tamil text, the Periya Purāṇam. Comprising over four thousand stanzas, this narrative anthology presents hagiographies of the 63 Shaiva saints (nāyaṉmārs, ‘leaders’). The nāyaṉmārs are traditionally believed to have wandered the Tamil countryside during approximately the sixth through ninth century, singing devotional hymns of praise to Shiva, the kings, and the Tamil landscape. However, Leslie Orr has convincingly argued that there is little to suggest that the Saints ever actually traveled.59 More likely, they remained in their local towns, where they received visitors and 58 Ranjeeta Dutta, ‘Pilgrimage as a Religious Process: Some Reflections on the Identities of the Srivaisnavas of South India’, Indian Historical Review 37, no. 1 (2010): 33. For an extensive study of Tirupati, see T.K.T. Viraraghavacharya, History of Tirupati: the Tiruvengadam Temples, 3 vols., vol.1-3 (Tirupati: Tirumala Tirupati Devasthanam Religious Publications, 1997). 59 Leslie Orr, ‘The Sacred Landscape of Tamil Śaivism: Plotting Place in the Realm of Devotion’, in Mapping the Chronology of Bhakti: Milestones, Stepping Stones, and Stumbling Stones, ed. Valérie

128 

Construc ting K anchi

wrote poetry based loosely on oral accounts of other peoples’ travels. The goal of their collective works (Tēvāram) was not to serve as a guidebook to real-world places but to express bhakti, devotion. Therefore, many of the descriptions of temples and landscapes in the poems read as somewhat generic or can be confusing when mapped onto the ground. The Periya Purāṇam sought to reinscribe the Tēvāram hymns onto Tamil geography by creating possible itineraries and life stories for the Saints. It also sought to reach an audience for whom the contemporary rather than the eighth-century landscape was familiar. The text was composed by the poet Cēkkiḻār, at the request of the Chola king Kulottuṅga II (ca. 1133-1150 CE). Although it is set within the period of the nāyaṉmārs (ca. sixth-ninth century), in many ways the Periya Purāṇam tells us more about the Chola period than about the time it purportedly records. For example, in Pūcalār’s story, the Saint constructs a temple in his mind.60 He furnishes the establishment with all the necessary elements, including a large gopura, a tank, and a well. Pūcalār’s temple conforms to the expected architectural norms of twelfthcentury Tamil Nadu. However, in the eighth century, the gopura and tank would not necessarily have been present in the manner he describes them. In the Periya Purāṇam, Kanchi features in the opening invocation chapter, ‘The Glory of the Divine Mountain’, and in the histories of at least nine Saints.61 The purāṇa of Tirukkuṟipput Toṇṭar, a washerman born in Kanchi, gives the city greatest attention. In the space of the text, Kanchi emerges not only as a city of multiple temples but also as a base from which to explore neighboring areas. When Appar visits, he first makes his way up the street that approaches the Ekāmbaranātha temple, singing praises of the many shrines that line its sides.62 After visiting temples in Kanchi, Saint Sundarar travels to proximate temples and then returns to the city before continuing on his journey.63 Pilgrimage also reached a new level of importance for Vaishnavas in the twelfth and thirteenth century.64 The concept of the divya deśam (sacred Gillet (Pondicherry: Institut Français de Pondichérry and École Française d’Extrême-Orient, 2014), 201. 60 Shulman has analyzed Pūcalār’s fascinating purāṇa for what it tells us about twelfth-century understandings of the imagination in South India. David Shulman, More Than Real: A History of the Imagination in South India (Cambridge, MA: Harvard University Press, 2012), 4-8. 61 Tirukkuṟipput Toṇṭar, Appar (Tirunāvukkaracar), Tiruñāṉa Sambandar, Ēyarkōṉ Kalikkāmar (Sundarar), Tirumūlar, Cākkiyar, Aiyaṭikaḷ Kāṭavar Kōṉ, Ceruttuṇai, and Pūcalār. 62 Periya Purāṇam, line 1586-1587. 63 Ibid., line 3344-3350. 64 The first text to elucidate the concept of pilgrimage in a Vaishnava context was the Divyasuricaritam, which Dutta dates to the twelfth century. Dutta, ‘Pilgrimage as a Religious Process:

Realignment

129

site sanctified by the Āḻvārs) emerged in the Śrīvaiṣṇava guruparamparā texts (hagiographies of the Āḻvārs). These places were systematized into a collection of 108 sites that constituted an ideal Vaishnava pilgrimage circuit.65 Like the Shaiva nāyaṉmārs, it is unclear to what extent the Vaishnava saints actually traveled. Dutta mentions that there are at least several sites that the Āḻvārs do not seem to have visited.66 Insofar as the divya deśams achieved scriptural and devotional importance, however, Kanchi, Tirupati, and Srirangam were the most sacred of places, and Kanchi and Tirupati were often linked. The biography of the Śrīvaiṣṇava theologian Rāmānuja (ca. 1077-1157 CE) describes a pilgrimage in which he proceeds from Kanchi to Tirupati, and then back to his birthplace of Srirangam.67 Along with pilgrimage, processions of bronze icons became a key practice during the Chola period. The road from Kanchi to Tirupati has long served as a processional pathway for the gods. Undertaken in shorter segments or for the full hundred-kilometer journey, processions that leave the city historically have been occasions for deities to survey their domain, visit other gods, and reside temporarily in other temples.68 In the process, these processions make worship available to a widely dispersed public. They enable people who were not granted access to the principal temples, or who resided in more remote locations, to see and be seen by the gods. Carrying a figure of a god through the city streets made divine vision (‘darśan’) available to the public, to the great many people who resided in or visited Kanchi in this important time. Through movement, the essence of the sacred space could be transmitted outwards, such that a greater number of people could see and be seen by the gods and simultaneously witness the splendor the patron had imparted. At the same time, devotees rushed inwards, towards the temple and towards the deities inside.69 Even today, a procession typically begins in the interior of a temple. The bronze icon Some Reflections on the Identities of the Srivaisnavas of South India’, 26. 65 Dutta, ‘Pilgrimage as a Religious Process: Some Reflections on the Identities of the Srivaisnavas of South India’, 26. 66 Ibid, 23. 67 Ibid, 28. 68 For good studies of South Indian processions, see Vidya Dehejia and Richard H. Davis, The Sensuous and the Sacred: Chola Bronzes from South India (New York: American Federation of Arts, 2002); Orr, ‘Processions in the Medieval South Indian Temple: Sociology, Sovereignty and Soteriology’; Crispin Branfoot, Gods on the Move: Architecture and Ritual in the South Indian Temple (London: Society for South Asian Studies, British Academy, 2007). 69 According to Orr, Chola-era inscriptions emphasize the inward rushing of devotees over the outward movement of the god. Orr, ‘Processions in the Medieval South Indian Temple: Sociology, Sovereignty and Soteriology’, 466.

130 

Construc ting K anchi

Illustration 43 Procession at Varadarāja Perumāḷ Temple, Kanchi

Realignment

131

or icons intended for transport are first bathed in sacred substances and dressed and adorned with pigments. Then they are mounted on silver or golden pedestals with long wooden poles as supports. Despite their heavy weight, the sculptures are actually designed for portability – the bases are fashioned with holes and protruding loops for lashing the sculpture to the transport with thick ropes. Once secured, the bronze is bedecked with silk and cotton textiles, flower garlands, and golden ornaments studded with rubies, diamonds, emeralds, and pearls (Ill. 43). The Chola-era bronzes that were made for processions are renowned for their finely cast, sensuous figures with extraordinarily beautiful bodies and regal ornaments. They are traditionally fabricated through a process of lostwax casting. The figures themselves are cast solid and include a disk-shaped base beneath the feet. The figure is then set into a separate, hollow-cast lotus base over a square or rectangular tiered pedestal. The final component is the prabhā (‘aureole of flames’), which may be either a separate element affixed by means of upright prongs or an integrated element that is attached to the lotus base, as in images of Naṭarāja. For each segment, the primary image is a wax figure composed of beeswax and resin.70 This is an essential component of the process, as it is the form the final bronze will take. The wax figure is then packed in clay with apertures called sprues giving access to the interior. The wax is melted out, leaving the clay mold intact, and then the molten metal mixture of copper alloys is poured in through the sprues. As it cools, the bronze solidifies inside. When the clay mold is cracked open to free the figure, the mold is destroyed and no identical sculpture can be produced. Each bronze icon made in this manner is therefore entirely unique. The Kanchi Bronzes Festival processions of bronze icons are central to Kanchi’s history, and they remain a daily occurrence to this day. However, the original bronzes are nowhere to be found within the city. At present I know of no Chola-period bronzes identified as having come from Kanchi.71 The dearth of detailed provenance information is not surprising for bronzes in museums outside 70 For discussion of lost-wax casting processes in the Chola period, see Dehejia and Davis, The Sensuous and the Sacred: Chola Bronzes from South India, 12-13. Elaborated in Vidya Dehejia, ‘The Thief Who Stole My Heart: The Material Life of Chola Bronzes from South India, c.855-1280’, lecture presented in A.W. Mellon Lectures in the Fine Arts (National Gallery of Art, Washington, DC, 2016), Lecture One, 3 April 2016. 71 A ca. sixteenth-century Somāskanda from the Ekāmbaranātha temple is in the collection of the San Francisco Asian Art Museum (Avery Brundage Collection, B60S157+).

132 

Construc ting K anchi

Illustration 44 Pillar, Gandhi Road, Kanchi, sixteenth century

Realignment

133

of India. Most Chola bronzes are labeled only as ‘Tamil Nadu’, or even more simply, ‘South India’, and scholars are just beginning the connoisseurly work of identifying workshops and matching styles to geography.72 However, museums within India – especially in Tamil Nadu – often label their sculptures with more specific findspots. It is remarkable that the single bronze from the Kanchi area on display at the Madras Government Museum is from a temple in Kūram, not Kanchi. The absence of Kanchi bronzes is made all the more striking by more than a dozen collected inscriptions from the Chola period that specifically mention sculptures gifted to temples in the city. For example, a twelfthcentury inscription at the Varadarāja Perumāḷ temple records the king’s gift of an image of an āḻvār (Vaishnava saint) along with land for its worship.73 An eleventh-century inscription at the Sīteśvara temple near Sarva Tīrtam Kuḷam mentions provisions for the daily worship of an image of Āṭavallāṉ, a dancing Shiva, yet no such sculpture remains on the premises today.74 Evidence of festival bronze processions that took place within Kanchi can be seen through architectural fragments scattered at passage points within the city. At the juncture of the parts of the city known as ‘Little’ and ‘Big’ Kanchi, fragmentary pillars stand on opposite sides of present-day Gandhi Road, just west of the Aṣṭabhuja Perumāḷ temple and east of the place where the colossal wooden chariot of Varadarāja Perumāḷ is kept. Today, they are hidden in plain sight – one is even built into a popular fruit stand (Ill. 44). These pillars likely once supported an arched gateway that marked the passage between the two city zones. Although the pillars date not before the sixteenth century, the processional pathway is firmly entrenched in the culture of the city and likely existed much earlier. Processions of bronze icons still follow this route, which leads from the Varadarāja Perumāḷ temple to Aṣṭabhuja Perumāḷ and Viḷakkoḷi Perumāḷ, and then turns north to pass the Ulakaḷanta Perumāḷ temple and onwards to Tirupati (Ill. 45). What happened to the Kanchi bronzes? It is likely that bronzes dating to the Chola era in Kanchi suffered the same fate as those elsewhere in Tamil Nadu. Although some bronzes from the Chola period remain in worship in Tamil temples, surviving examples 72 Dehejia has proposed a ‘coastal’ and a ‘capital’ workshop for Chola sculpture. Dehejia, ‘The Thief Who Stole My Heart: The Material Life of Chola Bronzes from South India, c.855-1280’, Lecture Two, 10 April 2016. 73 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 160-161, Cg.-654. 74 ARE 1955-1956, no. 265

134 

Construc ting K anchi

Illustration 45 Map of a processional route, Kanchi (map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution)

are exceedingly rare (Ill. 46). As Dehejia has demonstrated, it was common practice for ritual specialists to bury, or otherwise conceal, temple valuables in times of instability or threat.75 They carefully lined a pit with soft grass and laid the bronzes inside in a ceremonial manner. Several generations easily passed before the bronzes could be recovered. By that time, the few individuals who knew the secure location had died or otherwise departed, and the gods remained underground. Burial likely happened with increased frequency in the tumult surrounding the Chola dynasty’s gradual disintegration. Kanchi’s extensively urbanized condition means that stashes of bronzes likely lie beneath many layers of concrete. The practice of concealing precious bronzes continued into the colonial period. In 1879, district collector Charles Stewart Crole recorded that bronzes from the Kāmākṣī Ammaṉ, Ekāmbaranātha, and Varadarāja Perumāḷ temples had been taken to Uṭaiyārpāḷaiyam for protection when Kanchi was under threat from the Mysore ruler Haidar Ali.76 Reports of the incident 75 Dehejia, Art of the Imperial Cholas, 123-125; ‘The Thief Who Stole My Heart: The Material Life of Chola Bronzes from South India, c.855-1280’, Lecture Six, 8 May 2016. 76 Schier suggests an earlier date in the last decades of the seventeenth century for the bronzes’ removal, or perhaps that there were two similar incidents. Crole’s proposed date of 1780 was

135

Realignment

Illustration 46 Chola bronzes at Gōmuktēśvara Temple, Tiruvāvaṭutuṟai, tenth century

describe the bronzes being disguised as corpses for removal from the city. Continuing where Crole left off, Venkataraman further found that, after a lengthy period of itinerance, the Kāmākṣī Ammaṉ bronze was installed in a temple in Thanjavur, where it still remains.77 Still other Kanchi bronzes likely have disappeared into private hands or entered museums without the details of their findspots. Colonial-era travelers frequently collected antiquities. A ca. twelfth-century bronze Vishnu appeared on the New York art market in 2009, boasting its history of ownership by Louis Bonvin, French Governor-General of Pondicherry

based on a misreading of the date in an inscription from the Varadarāja Perumāḷ temple that records the bronze’s return to Kanchi. Schier, ‘The Goddess’s Embrace: Multifaceted Relations at the Ekāmranātha Temple Festival, Kanchipuram’, 173; Charles Stewart Crole, The Chingleput, Late Madras, District: A Manual Compiled under the Orders of the Madras Government, by Charles Stewart Crole, of the Madras Civil Service (Madras: Lawrence Asylum Press, 1879), 117. See also Ute Hüsken, ‘Gods and Goddesses in the Ritual Landscape of Seventeenth and Eighteenth-Century Kāñcipuram’, in Layered Landscapes: Early Modern Religious Space Across Faiths and Cultures, ed. Eric Nelson and Jonathan Wright (London: Routledge, 2017). 77 Venkataraman, Dēvī Kāmākshī in Kāñchī: A Short Historical Study, 46-47. The goddess is enshrined in the Paṅkāru Kāmākṣi temple (situated about a kilometer north of the Bṛhadīśvara temple in Thanjavur). For discussion, see Schier, ‘The Goddess’s Embrace: Multifaceted Relations at the Ekāmranātha Temple Festival, Kanchipuram’, 161-164. Hüsken has located shrines built for the other Kanchi bronzes in Thanjavur.

136 

Construc ting K anchi

from 1938 to 1945.78 Robert Clive’s daughter-in-law, Henrietta, was not so lucky. When she visited Kanchi in 1800, she and her daughter lamented the presence of many stone and wooden icons inside the temples, but not a single ancient bronze.79

From Ancient Village to Temple Town Kanchi’s urban landscape grew increasingly conducive to new forms of devotionalism as pilgrimage and processions became officially promoted practices. The central avenue and the square layout of the Raja Veedhis served as an ideal processional space, and the orientations of temples directed focus towards the new pathways. All streets in the city now led to temples. Enshrined within the inner sanctums of temples, the stationary gods (‘mūlavars’) made of stone or wood looked towards their bronze-cast mobile counterparts (‘utsavars’). Similar processes of redesign were at work in nearby villages outside of Kanchi, such as Ārpākkam, Mākaṟal, Tiruppulivaṉam, Uttiramērūr, Perunakar, Kaḷakkāṭṭūr, and Tirumukkūṭal to name a few. Each place had ample natural resources, including land suitable for rice cultivation (‘nel’), water that could be collected in artificial lakes (‘ēri’), and stone for temple construction (‘kal-curaṅkam’). These assets profitably sustained the expanded settlement as a tax-free endowment (‘brahmadēya’). At each of these places, an ancient village was transformed into a relatively prosperous, religiously plural town, conveniently situated along a growing pilgrimage route. Looking on the micro level at such individual villages can help us understand processes that were taking place on a much larger scale and with greater complexity in the nagarams, or urban areas. Near to Kanchi, Ārpākkam village serves as a key example of one such microstate. Nel, Ēri, Kal: How to Make a Sustainable Brahmadēya Twelve kilometers south of Kanchi along the Uttiramērūr road is a junction where a battered signboard points eastward towards the Jain temple at Ārpākkam. The well-paved road at the turnoff from the highway soon plunges into a sprawling wilderness area where paddy fields give way to 78 First offered at Christie’s, NY, sale 2271, 20 March 2009, lot 1292. 79 Clive, Journal of Lady Charlotte Florentia Clive (1787-1866), Copied by Professional Calligrapher, W.H. Ramsay, 48.

137

Realignment

Illustration 47 Map of sites in Ārpākkam Village (map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution)

Illustration 48 Ayyaṉār, Tiruvālīśvara Temple, Ārpākkam, eleventh century

138 

Construc ting K anchi

crater-like pockets of quarried-out rock spread in dark ringlets across the flat landscape. To the north is a large lake with a slender line of trees visible on the distant bank, and to the south is a tangle of quarry followed by paddy, with the village of Mākaṟal two kilometers beyond. As the signboard promises, the paved road terminates one kilometer east of the highway in front of the Jain temple in the tiny village of Ārpākkam. Situated in the fertile tract between the Palar and Cheyyar Rivers, the now-remote Ārpākkam houses four extant temples, each dedicated to a different god (Ill. 47). All of these temples are constructed fully of stone with architectural form, sculpture, and inscriptions dating to the eleventh through thirteenth century. Inscriptions given in the regnal years of the greatest eleventh- and twelfth-century Chola kings – Rājarāja I, Rājēndra I, Kulottuṅga I, and Rājādhirāja II – form veneers of words across the temple walls. The Shiva temple contains a superb ca. eleventh-century figure of Ayyaṉār seated on a plinth (Ill. 48). A sculpture of the Buddha, stylistically datable also to the eleventh/twelfth century, and an earlier stone slab (half buried) depicting the goddess Jyeṣṭā are also in the village.80 Along the road to Ārpākkam from the Kanchi-Uttiramērūr highway is a ruined brick shrine embellished with stone and containing fragmentary granite sculptures of goddesses (see Ill. 9). Numerous additional archaeological remains from as early as the megalithic period can also be found nearby.81 Although Ārpākkam is now a modest village with a population of less than three thousand, the archaeological and epigraphic records indicate that it was an important place during the Chola era.82 Inscriptions refer to Ārpākkam as lying within Mākaṟal-nāṭu. While this territorial designation tells us Mākaṟal was the more prominent place, Ārpākkam was an ancient settlement that had its own prestige during the eleventh and twelfth century. An extensive inscription dated to the fifth regnal year of Rājādhirāja II (ca. 1168 CE) on the southwest wall of the Shiva temple reveals that Ārpākkam was even considered so important that it was selected as a gift of gratitude to a Chola general for his heroic acts in a major battle 80 Minakshi mentions additional Buddhist statuary and local knowledge of a Buddhist presence. Minakshi, ‘Buddhism in South India’, 111-112. 81 Cairn circles are well documented in Ārpākkam. IAR 1978-1979, p. 21. An Indian newspaper reported the discovery of a possibly pre-megalithic (ca. 1800-1500 BCE) urn burial site in the nearby village of Maṇḍapam, which is situated amid the quarry pockets near the junction of the Kanchipuram-Uttiramērūr road and the Ārpākkam turnoff. T.S. Subramanian, ‘Urn Burial Site Discovered near Kancheepuram’, The Hindu 9 April 2012. 82 Population data courtesy ‘Census of India’, http://www.census2011.co.in/data/village/629732Ārpākkam-tamil-nadu.html.

Realignment

139

between the Cholas and the army of Ceylon (present-day Sri Lanka). 83 Considered together, Ārpākkam’s remains indicate that the village was home to a diverse cult history with strong enough patronage to sponsor fine carvings in stone. In addition to being a microstate, Ārpākkam is also a paradigmatic example of a Chola-period brahmadēya – an ancient settlement in an arable area that was legally established as a tax-free endowment, with a Brahmin lineage given responsibility for maintaining temple ritual and overseeing the cultivation of the land.84 The act of creating a brahmadēya could be either the foundation of a settlement or a transfer of its legal ownership. At its core, the brahmadēya was a technique for expanding rice cultivation through official donations of land ownership and taxes on land that was underproductive.85 It instantiated new incentives for productivity. Through systems of irrigation and cultivation, areas that otherwise would have comprised discrete administrations became knit together. By extension, these newly connected areas became integrated incrementally into more extended networks. Such transactions across the countryside carried with them the potential for significant physical transformations to the landscape, its architecture, and its sociopolitical composition. They also inevitably led to certain amounts of resettling or redistribution of population and changes in personnel.86 The most significant physical transformations were the conversion of wilderness or fallow fields into paddy, and the rebuilding of village shrines in the durable material of stone. Ārpākkam’s initial (re)building seems to have taken place in the early eleventh century. The first major expansions to the temples were made around two centuries later. When a king established a brahmadēya, the terms of the agreement were usually recorded on a set of copper plates bound on a ring bearing the royal seal. The agreement stipulated that the king would grant certain tax exemptions and, in exchange, the villagers and local council would meet 83 ARE 1899, no.20; SII Vol. 6, p. 188-190, no. 456. 84 Stein, Peasant State and Society in Medieval South India. 85 James Heitzman, ‘Temple Urbanism in Medieval South India’, The Journal of Asian Studies 46, no. 4 (1987): 815. 86 The eleventh-century Karāṇṭai copper plates record that over a thousand Brahmins were brought from Andhra Pradesh and elsewhere in Tamil Nadu to a new settlement near Thanjavur. Burton Stein, ‘Circulation and the Historical Geography of Tamil Nadu’, The Journal of Asian Studies 37, no. 1 (Nov. 1977): 17. Some Pallava copper plate inscriptions use the expressions kuṭi nīkki and mun-peṟṟārai māṟṟi, which mean removing earlier occupants, either physically or in terms of their rights over the land. Veluthat, The Early Medieval in South India, 84-85.

140 

Construc ting K anchi

their responsibility of cultivating the land and maintaining the temple.87 Support for the temple was siphoned through the settlement assembly (‘ūrār’), which consisted of leaders of local agricultural groups.88 Implicit in this arrangement was that the residents also agreed to publicly recognize the king’s authority by such methods as calculating the date according to his regnal year and, if available, employing an architectural format and any emblems or insignia associated with his dynasty for the reconstruction of their temple and the installation of the gods. The appearance, for example, of the emblematic Chola Shiva-Naṭarāja, and an increasingly standardized Shaiva iconographic program in village temples across the Tamil countryside suggest that the overseeing architect in each place was familiar with the various forms and designs in currency inside the king’s domain.89 Even in the very remote village of Putūr-Brahmadesam, the image of Shiva as Lord of the Dance appears repeatedly on the Rudrakoṭīśvara temple, where an inscription dated to the reign of Kulottuṅga I (ca. 1091 CE) records the endowment of the village as devadāna, a gift of tax-free land to the god.90 It seems to have been the local workshop’s responsibility to accept the choices that the head architect ordered, while they worked within their own local crafts tradition.91 The establishment of brahmadēyas was not an exclusively Chola royal activity, nor were brahmadēyas the only type of village. As Karashima and Veluthat remind us, although there are more records of Brahmin settlements, 87 Daud Ali has demonstrated the ways in which copper plate charters constitute their own discursive field. Ali, ‘Royal Eulogy as World History: Rethinking Copper-Plate Inscriptions in Cōla India.’ 88 Heitzman, The City in South Asia, 47; Veluthat, The Early Medieval in South India, 31. 89 Kaimal has elucidated the function of the Naṭarāja as a royal Chola emblem. Padma Kaimal, ‘Shiva Nataraja: Shifting Meanings of an Icon’, The Art Bulletin 81, no. 3 (1999). One iconographic program seen throughout Chola territory became standardized during the period of Sembiyan Mahādēvī. Dehejia, ‘The Thief Who Stole My Heart: The Material Life of Chola Bronzes from South India, c.855-1280’, Lecture Three, 17 April 2016. The most frequent program places Ganesha and Dakṣiṇāmūrti on the south wall, Vishnu in the rear, and Brahmā and Durgā on the north wall. 90 ARE 1915, no. 269. My thanks to Anbarasu for bringing me from Kanchi to Brahmadesam, and to the Candramaulesvara’s monument attendant for informing us of the little-known Putūr temple (February 2015). 91 Michael Meister differentiates ‘style’ from ‘idiom.’ Whereas style was a temporal designation that could hold dynastic connotations, ‘idiom’ refers to the local workshop’s particular craft tradition, which transcends time and political context. Michael W. Meister, ‘Style and Idiom in the Art of Uparāmala’, Muqarnas 10 (1993). I would add that idiom often extends to material, as different materials yield different artistic forms and conventions.

Realignment

141

there were far more non-Brahmin settlements, called veḷḷāṉvakai.92 The infrequent references to veḷḷāṉvakai in inscriptions reveal that these nonBrahmin villages included residential areas, cremation grounds, sources of drinking water, irrigation channels, and cultivated fields and pastures.93 The difference was that they were not driven by a temple. With the innovation of temple endowments for agrarian settlements in the Pallava period, the establishment of a brahmadēya became the ‘gift par excellence.’94 From a Pallava context, an important set of copper plates that records the endowment of a brahmadēya village was recovered from the coastal town of Kasakudi, four kilometers north of Karaikal, in 1891.95 The Kasakudi plates provide an example of a typical record of this type. The eleven plates are dated to the 22nd year of the reign of Nandivarman II (sponsor of the Vaikuṇṭha Perumāḷ temple in Kanchi). The first 104 lines are in Sanskrit (Grantha script), and a shorter segment of 30 lines in Tamil follows. The final lines are again in Sanskrit with a short parenthetical note in Tamil. The Sanskrit portion constitutes the royal decree. It opens with an elaborate praśasti that gives the Pallava genealogy, followed by praise for the Brahmin who is to be endowed with custodianship of the village. The subsequent Tamil portion is the village community’s formal acceptance. In order to acknowledge understanding, despite linguistic differences, the Tamil portion restates the terms and details the tax exemptions, perhaps in order to prevent subsequent dispute. The final lines in Sanskrit and Tamil confirm the contract. In the Kasakudi plates, the village in question, called Koḍukoḷḷi, is to be gifted to the Brahmin Jyeṣṭhapāda-Somayājin, who resides at Pūni.96 After extensive praise of this Brahmin – who is said not only to be expert in all forms of knowledge, sacred and worldly, but also to be courteous, unique, and radiant like the sun – the boundaries of the village are specified. The grant then states that the brahmadēya is to include all the lands within the given boundaries, excluding previous grants to temples and Brahmins and 92 Noboru Karashima, South Indian History and Society: Studies from Inscriptions, A.D. 850-1800 (Delhi: Oxford University Press, 1984), xx; Veluthat, The Early Medieval in South India, 30. 93 Veluthat, The Early Medieval in South India, 31. 94 Nicholas Dirks, ‘Political Authority and Structural Change in Early South Indian History’, 153. 95 SII Vol. 2, Part 3, p. 346-353 (text), p. 42-61 (discussion and translation), no. 73. 96 Hudson has proposed that this Brahmin was the ācārya (‘priest’) to Nandivarman II and mastermind behind the Vaikuṇṭha Perumāḷ temple’s design. Hudson, The Body of God: An Emperor’s Palace for Krishna in Eighth-Century Kanchipuram, 77.

142 

Construc ting K anchi

Illustration 49 Festival at Ekāmbaranātha Temple, Kanchi (October 2013)

the houses of people who had acquired a right to hold land for the purpose of cultivating it, but including all other houses, fields, gardens, and groves, as well as the use of water by means of digging channels from nearby water bodies. In his discussion of the inscription in the epigraphic report, Hultzsch describes the village as near (likely between) the Vegavathi and Cheyyar Rivers and not far from the Tiraiyaṉ (present-day Tennēri) Tank, which is situated roughly fifteen kilometers east of Kanchi and would have enabled plentiful irrigation for the fields.97 To acknowledge the transfer of ownership, Koḍukoḷḷi is given a new name, Ekadhīramaṅgalam. The change in name and the mention of exclusions indicates that the village itself was already an established settlement with a longer history, possibly including certain areas that already had been made tax-free. As at Ārpākkam, brahmadēyas were very often places that had been inhabited from as early as the megalithic period. Archaeological evidence of megalithic settlements in Tamil Nadu generally comes in the form of urn-burial sites surrounded by cairn circles, together with black and red rouletted pottery, sarcophagi, and sometimes terracotta f igurines and coinage. Remains distributed across various stratified levels suggest to archaeologists that the sites have been continuously inhabited for thousands of years. 97 SII Vol. 2, Part 3, p. 345, no. 73. Minakshi identifies the tank as Teṉṉēri, Minakshi, Administration and Social Life Under the Pallavas, 95.

Realignment

143

While villages with long histories lying outside of Kanchi, such as Ārpākkam and Koḍukoḷḷi, were being transformed into religiously plural temple-towns, a similar process was at work inside of Kanchi during the Chola era. Localities within the urban area that held loose clusters of shrines were being built up, rebuilt, and consolidated into temple complexes that would continue to expand.98 Some of these sites would endure another millennium of renovation and renewal, to become the city’s most popular devotional centers (Ill. 49). Structures of Patronage in the Era of the Cholas: Inscribing Devotion A shift in dominant patronage away from the royal and toward the local and communal gave rise to an impressive number of temples in Kanchi during the tenth through thirteenth century.99 In Pallava areas during the seventh and eighth century, temples in and around Kanchi had been predominantly royal structures – that is, they were sponsored by kings or queens, as is explicitly expressed in royal inscriptions displayed prominently on the temple exterior. By contrast, temples built in the very same areas while under Chola control bear little to no evidence of direct involvement from kings. Instead, there was a notable widening of involvement in temple 98 Heitzman describes the inner-city process as a multiplication of the independent brahmadeya. ‘The more concentrated patterns of this “temple urbanism”, based on the developing religious institutions, allowed the integration of wider agrarian zones within the larger complexes… The “urban” character of these larger administrative units rested, however, on the integration of a number of individual settlements, grouped around ritual centers, that preserved in themselves the characteristics of the village… But the administrative recognition of this extended area as one place, the extensive and integrated commercial or manufacturing networks, and especially the ritual interactions of the many temples formed the complex infrastructure of a major central place. In the absence of defensive walls, the settlement patterns and intense regional interactions of the capital shaded into the local networks preserved as peripheral centers.’ Heitzman, ‘Temple Urbanism in Medieval South India’, 817. 99 Chola patronage is currently under reconsideration by scholars. I am grateful to Leslie Orr, Padma Kaimal, and others for continued conversations, especially during the Archaeology of Bhakti III workshop (August 2015). Publications on this topic include: Padma Kaimal, ‘Early Cōla Kings and “Early Cōla Temples”: Art and the Evolution of Kingship’, Artibus Asiae 56, no. 1/2 (1996); ‘The Problem of Portraiture in South India, Circa 870-970 AD’, Artibus Asiae 59, no. 1/2 (1999); ‘The Problem of Portraiture in South India, Circa 970-1000 AD’, Artibus Asiae 60, no. 1 (2000); Orr, Donors, Devotees, and Daughters of God: Temple Women in Medieval Tamilnadu; Padma Kaimal, ‘A Man’s World? Gender, Family, and Architectural Patronage, in Medieval India’, Archives of Asian Art 53 (2002/2003); Orr, ‘Preface’; Risha Lee, ‘Constructing Community: Tamil Merchant Temples in India and China, 850-1281’ (PhD diss., Columbia University, 2012); Dehejia, ‘The Thief Who Stole My Heart: The Material Life of Chola Bronzes from South India, c.855-1280.’ Work continues.

144 

Construc ting K anchi

Illustration 50 Inscriptions, Ulakaḷanta Perumāḷ Temple, Kanchi, eleventh century

construction, maintenance, and the performance of ritual throughout Tamil Nadu. The king’s removal from the active role of building temples may be symptomatic of a larger shift in governance that had its seeds in the Chola period. As Arjun Appadurai describes, the king had become an administrator rather than a legislator by the fourteenth century.100 Decisions were typically reached communally, and the king’s responsibility was little more than to formalize the resolution. Appadurai’s argument closely corresponds to a proposal Burton Stein made one year earlier.101 Stein argued that during the course of the Chola period, a network emerged in which highly localized powers surrounded an overlord, whose authority was primarily ritual in nature. The royal figurehead held legislative duties only within a relatively small domain (‘kṣetra’). The same held true in the context of temple patronage. When a king is mentioned in temple inscriptions (beyond the use of his name for establishing the date), it is often in his capacity as the overarching landowner who could legally gift land to the temple-estate. At 100 Arjun Appadurai, Worship and Conflict under Colonial Rule: A South Indian Case (Cambridge: Cambridge University Press, 1981), 68. 101 Stein, Peasant State and Society in Medieval South India.

Realignment

145

Kanchi’s Aṉēkataṅkāvatīśvara temple, for example, the two inscriptions from the reign of Kulottuṅga I (late eleventh and early twelfth century) record donations of land by the king. Gifts by the king and the resident community progressively allocated land and local resources to the local shrine, which led to the formation of an ever-expanding temple-estate that was owned by a deity.102 By and large, the many inscriptions that encase the walls of Chola-period temples record pious donations by individuals, families, and communities outside the royal court. According to the work of scholars including Heitzman and Orr, the majority of donations recorded from the Chola period were in the form of perpetual lamps that served to make the temple visible from afar at night, and to light and enliven the otherwise dark interior (Ill. 50). The donor could be an individual or a group, but gifts often required collaborative efforts. Although not always specified in the often-brief inscriptions, donating a lamp involved not only making a gift of sheep to provide the daily supply of ghee, but also designating at least one person to look after the sheep and possibly another to maintain the lamp’s flame. In an inscription at the Murukaṉ temple in Uttiramērūr, for example, a group of devotees gifts sheep for a lamp in the temple, and the sheep are entrusted to a local resident.103 A comprehensive epigraphical survey of all of Kanchi’s Chola-period inscriptions would provide valuable insights into Kanchi’s political economy and the relative involvement in temple activity of the more affluent members of the city’s diverse communities.104 Of the myriad visible inscriptions dating to ca. 900-1300 CE, both in situ on temples and reused as architectural materials, less than three hundred have been noticed in the epigraphic reports, and fewer still have been published. Working from epigraphic reports and archival estampages rather than on-site research, Heitzman began this important work with an article in which he gathered 277 records that had been collected earlier from 21 temples in Kanchi, including 167 102 For the expanding role of the temple, see Veluthat, The Early Medieval in South India, especially ‘The Temple in Medieval South India’ and ‘Land Rights and Social Stratif ication.’ Veluthat has elaborated upon the role of the temple and productively challenged Burton Stein’s model of the segmentary state for its failure to recognize multiple social strata within the peasant class. 103 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 283, Cg.-1124. 104 The desired product would be a comprehensive publication akin to Bahour S. Kuppaswamy and G. Vijayavenugopal, eds., Pondicherry Inscriptions, Part 1: Introduction and Texts with Notes (Pondicherry: École Française d’Extrême-Orient and Institut Français de Pondichéry, 2006).

146 

Construc ting K anchi

records from the Varadarāja Perumāḷ temple alone.105 Heitzman’s article and Mahalingam’s Topographical List of the Inscriptions in the Tamil Nadu and Kerala States have been invaluable resources for my work. However, countless unreported inscriptions lie hidden beneath modern painting and other renovations. The records of donations from any sector of the population were quite literally ‘written in stone’, both to remember the act and to bring the donor permanently into proximity with the god of the temple.106 The range of communities represented in Kanchi’s noticed inscriptions alone include Brahmin councils, queens, merchants, weavers, village headmen, women, daughters, shepherds, maidservants, wives, gardeners, dancing girls, temple servants, and collective residents of villages. This breadth of patronage was by no means limited to Kanchi’s vicinity. Rather, the shift from ‘royal’ to ‘local’ that took place in northern Tamil Nadu was part of a more widespread change in patronage structures spanning from Kanchi in the North to Koṭumpāḷūr in the South. What is meant by patronage is not necessarily the sponsoring of temple construction, but provision for the temple’s maintenance, both in the sense of the temple’s physical structure and the continuity of its ritual life. In fact, unlike the epigraphic corpus elsewhere in India, the known Chola-era inscriptions rarely give any details regarding the temple’s foundational moments. This striking lacuna has led Orr to suggest that when, as 21stcentury scholars, we ask the inscriptions to tell us who built the temple, we are simply asking the wrong question.107 Rather than being founded by an individual, the inscriptions suggest that temples in the Chola period across Tamil Nadu were often the results of collective endeavors that gave rise to the existing monument over a span of time. Gifts of bronze images also could be the product of collective donations.108 Even segments of temples were frequently funded by more than one individual, or an individual paid for a specific building block. Inscriptions on pillars in the thirteenth-century maṇḍapa of the Mucukundeśvara temple at Koṭumpāḷūr bear the names of their individual donors.109 The same 105 Heitzman, ‘Urbanization and Political Economy in Early South India: Kāñcīpuram During the Cōḻa Period.’ 106 Orr, ‘Preface’, IV and XVI. 107 Orr, ‘Preface’, IV and XVI. Developed further in Orr’s concluding remarks to The Archaeology of Bhakti III workshop (August 2015). 108 Dehejia, ‘The Thief Who Stole My Heart: The Material Life of Chola Bronzes from South India, c.855-1280’, Lecture Four, 24 April 2016. 109 Noticed by Leslie Orr, The Archaeology of Bhakti III workshop (August 2015).

Realignment

147

is true on fifteen pillars also dating to the thirteenth century within the thousand-pillared hall at Kanchi’s Ekāmbaranātha temple.110 There is also an inscription that appears to be a name on a pillar in the thirteenth-century maṇḍapa at the Viḷakkoḷi Perumāḷ temple in Kanchi. These sites suggest that the gifting of individual pillars was a growing and widespread practice, especially in the thirteenth century, when there was the greatest flexibility in structures of patronage. Expanding Edifices in Kanchi Kanchi’s tenth-century temples were among the first to be built or converted to stone, not under royal patronage like the Pallava temples had been, but as expressions of local bhakti. The numerous inscribed blocks built into paving slabs and later walls suggest that there was a much higher degree of patronage for religious monuments in tenth-century Kanchi than the surviving examples alone would indicate. Only four temples retain known tenthcentury inscriptions in situ.111 These are the Paṇāmuṭīśvara, Cokkīśvara, and Tirukkālīśvara temples, and the Durgā shrine in the Kacchapeśvara temple complex. The main body of the Karukkiṉil Amarntavaḷ Ammaṉ temple (east of Kāyārohaṇeśvara) and the Sīteśvara temple near Sarva Tīrtam Kuḷam closely correspond to dated tenth-century examples from the Kaveri region, but they bear no inscriptions dated in the tenth century. Although these six temples are spread over an area that spans from central Kanchi all the way east to the village of Tirukkālimēṭu, there may have been specific reasons why each one was selected for early reconstruction. Often, the rationale seems to have been a connection with an older history for the site. The sponsorship of temples would continue to increase exponentially in the eleventh century, when Kanchi truly became the ‘city of one thousand temples’, as it is popularly known today. The number of extant temples with eleventh-century architectural features amounts to at least one dozen, and the number of eleventh-century inscriptions to no less than 50.112 As is true throughout the Tamil country, the efflorescence of temple construction was primarily due to the large number of donors who increasingly filled the world of the temple. To accommodate this widening devotional base, 110 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 152, Cg.-623. 111 I distinguish inscriptions in situ from those on slabs built into paving stones or apparently later walls. 112 C.R. Srinivasan provides a useful appendix of inscriptions grouped by king in Srinivasan, Kanchipuram through the Ages, 285-295.

148 

Construc ting K anchi

temple layouts became more expansive and the structures more complex and multifaceted. Porches and cloistered enclosures sheltered the growing numbers of visitors, and multiple sanctums distributed the worshippers and kept them moving throughout the space. As at Kacchapeśvara, the appending of a mahāmaṇḍapa (‘entry hall’) in front of the ardhamaṇḍapa (‘vestibule’) was typically among the first additions. Besides expanding the interior, it also provided a shrine for the goddess, positioned on the northern side of the mahāmaṇḍapa. Unlike the constricted ardhamaṇḍapa and main sanctum, the mahāmaṇḍapa was made large enough to accommodate multiple people. This way, while one group of devotees was engaged in worshipping the temple’s primary god, another could have darśan of his female counterpart. While earlier temple complexes included shrines for subsidiary parivāra devatās (‘associated deities’) that stood separately from the main structure, in the eleventh century, builders carved out a space for the goddess inside the temple body.113 As goddesses grew in importance in subsequent centuries, full shrines were added on the northern side of the temple, still within the same compound but again separated from the main shrine. Rather than replacing the interior goddess sanctum, these extra structures were simply added to the complex. Changes in religious practice and changes in temple architecture worked in tandem throughout the tenth and eleventh century.

Local Style Chola-period builders in Kanchi looked not only to satisfy the demands of their widening patronage base within the parameters of the dominant architectural trends, but also to incorporate imagery found locally in temples in the city. At times, translating forms that were established in the Chola heartland of the Kaveri region across the great distance to Kanchi resulted in a somewhat awkward appearance for certain deities and structural elements. However, for the ornamental veneers, the Pallava temples provided a local roster of imagery that is not encountered outside the city. Positioned nearly adjacent to the Kailāsanātha temple, the Aṉēkataṅkāvatīśvara temple serves as an informative example (Ill. 51).

113 Ninth-century aṣṭaparivāra devatā shrines can be seen in the Pudokkottai district, such as at the Cuntarēśvara temple at Tirukkaṭṭaḷai. The eight deities include Sūrya, the Saptamātṛkās, Ganesha, Murukaṉ, Jyeṣṭhā, Candra, Caṇḍeśa, and Nandi, but not Umā.

Realignment

Illustration 51 Anēkataṅkāvatīśvara Temple, Kanchi, ca. 1070 CE

149

150 

Construc ting K anchi

Illustration 52 Interlocked chain motif, Anēkataṅkāvatīśvara Temple, Kanchi, ca. 1070 CE

The architectural and sculptural style situate the Aṉēkataṅkāvatīśvara temple in the late eleventh or early twelfth century, a date that agrees with its earliest inscriptions.114 The inscriptions are from right around the year 1100, which makes Aṉēkataṅkāvatīśvara among the earliest temples anywhere in Tamil Nadu to exhibit its specific form of potikā (‘corbel, bracket’). The particular way in which this element is rendered can also be seen at proximate ca. twelfth-century temples, including Puṇyakoṭīśvara and Brahmapureśvara. The potikā further bears close resemblance to vernacular architecture in Kanchi. The inner courtyards of traditional houses are often lined with well-carved wooden pillars that have very similarly shaped brackets. Although these houses date back at most two hundred years, the woodcarvings likely follow more ancient traditions. Aṉēkataṅkāvatīśvara’s f inely executed ornamental reliefs display a wide range of figural and geometric elements. Featuring scrolling foliage supporting miniature lions and apsaras (‘flying goddesses’), pendent jewels embossed with flowers, waves of sea foam, and interlocking chains, the 114 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 156-157, Cg.-640-641.

Realignment

151

imagery has more in common with temples within the local area than it does with temples in the Kaveri region. Close comparisons can be found at Pallava sites including Māmallapuram’s Shore temple, Kanchi’s Kailāsanātha temple, and especially the Sundaravarada Perumāḷ temple at Uttiramērūr.115 These patterns in stone almost certainly held longer histories in local traditions of woodcarving. They may also have been applied to bronze sculptures that were fabricated in Kanchi. Noticeably unique to Kanchi is a motif of interlocked chains that first appears prominently on the basement moldings of the Kailāsanātha temple, just meters away from Aṉēkataṅkāvatīśvara (Ill. 52). Although to my knowledge it does not feature on other Chola-era temples, this motif graces the upper and lower bands along the vṛtta-sphuṭita (‘overflowing vase’) on the western side of Aṉēkataṅkāvatīśvara. Perhaps this crucial detail pays homage to the temple’s Pallava-era architectural predecessor. Similarly, at Aṉēkataṅkāvatīśvara, a pādabandha-adhiṣṭhāna has been selected for the basement moldings from a range of possible options that were very much in circulation by the earliest date this temple could have been built. The pādabandha-adhiṣṭhāna was the type most frequently met with in Pallava temples. Overall, the Aṉēkataṅkāvatīśvara temple is a typical example of a middle- to late-Chola style temple constructed in this particular locality. It displays a Kanchi-specific aesthetic that balanced newness with continuity.

Conclusion: Urban Logic By using Kanchi as an administrative outpost, the Cholas expanded their territory into northern Tamil Nadu and claimed the royal legacy of the Pallava city for themselves. After the reduction in temple construction that took place during the transition to Chola dominion in the ninth century, the city’s new rulers reconfigured the more ancient royal capital. Along with a broader, more diverse base of patronage, they transformed the city into a processional space that was replete with sacred monuments. They laid avenues that connected key monuments and marked out the dynasty’s own northward pathway of conquest. Arriving from the south and extending 115 Parallels also can be drawn with decorative motifs in Chalukyan temples that are contemporaries of the Pallava monuments. The rendering of motifs such as birds, apsaras, and makaras with riders seen both on the Sundaravarada Perumāḷ temple at Uttiramērūr and the Virupaksha temple at Pattadakal are noteworthy comparisons. These similarities show the circulation of imagery across a broader region. Some even say the Chalukyan monument was inspired by Kanchi’s Kailāsanātha.

152 

Construc ting K anchi

to the north, the royal avenue through the midline of the city reinforces the Cholas’ history and forecasts their political aspirations.116 Traversed by merchants and the military, as well as pilgrims and processions, the road functioned as a theater for public and political spectacle and for the practice of everyday life.117 The road connected the city with the palace, which was centrally located at the heart of the urban core. A set of square streets created a processional space around the palace, and temples framed the square from all sides. Throughout the city, new temples were constructed and old sites were renovated and expanded. Sponsorship was not necessarily royal. Rather, it was tied to the new political and economic structures that facilitated local patronage. These temples established a pattern that would remain entrenched for centuries to follow. The shrines were positioned to face the central avenue, the focal point for the gods. Kanchi’s connection with villages in surrounding areas – most notably to the south of the city in places rich with stone and minerals – reveals the complex hierarchies of political and economic integration that characterized the Chola period. While Kanchi was rebuilt, so too were villages outside the city. In the hinterland, sponsorship for temples entailed new incentives for economic productivity and new destinations for devotees. Urbanization processes gave shape to outlying areas, and the rural settlements became incrementally recognized as belonging to the Chola’s city.

116 Rajēndra’s eleventh-century copper plate charter boasts of the king’s conquest of the Ganges, more than a thousand kilometers to the north. Ali, ‘Royal Eulogy as World History: Rethinking Copper-Plate Inscriptions in Cōla India’, 207. 117 I borrow this expression from Michel de Certeau, The Practice of Everyday Life, trans. Steven Rendall (Berkeley: University of California Press, 1984).

3

The City and its Ports Abstract Chapter Three explores the creation of a broader geographical ruralurban continuum that increasingly came to define Kanchi. It examines settlements in the rural areas surrounding the city that became home to Pallava and Chola-era sacred architecture and received subsequent endowments and modifications over time. Together, these settlements formed a devotional and economic network. The chapter traces interactions among temple-sites through shared patterns in architecture, iconography, and inscriptions, as well as residential design. The chapter’s final section considers the possibility of a cultural landscape that expanded Kanchi’s kṣatra (‘sphere of influence’) transregionally. Specif ically, it looks at Kanchi’s role in extended Buddhist networks and examines connections across urban centers in an integrated South and Southeast Asian region. Keywords: Buddhism, Tamil merchants, Indian Ocean, Rurban, Epigraphy, Cultural landscape

The twelfth-century hagiography of Saint Tirukkuṟipput Toṇṭar opens with generous praise for the lands that surround Kanchi.1 The poem takes us on a journey across the hinterland, from Māmallapuram on the seacoast (‘neytal’), through Tirukkaḻukuṉṟam in the hills (‘kuṟiñci’), to the farms and granaries (‘marutam’) around Vallam, and finally to the urban center (‘nagaram’) of Kanchi itself. Along the way, the Palar River is described as a source of joy.2 A Chola-era reader would have been familiar with these landscapes from classical, Caṅkam-era Tamil literature (ca. first century BCE-sixth century CE), which was still in circulation when the poem was

1 2

In the Periya Purāṇam, 110 of the poem’s 128 stanzas pass before we even meet the Saint. Periya Purāṇam, line 1098-1100.

Stein, Emma Natalya, Constructing Kanchi: City of Infinite Temples. Amsterdam, Amsterdam University Press 2021 doi: 10.5117/9789463729123_ch03

154 

Construc ting K anchi

Illustration 53 Map of temple-sites from Kanchi to the sea (map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution)

composed.3 More than just poetic concepts, these landscapes are grounded in the actual experience of travel through Tamil Nadu. This chapter widens the focus from Kanchi’s urban core to consider the ways in which the city’s hinterland was transformed over the course of the eighth through the thirteenth century (Ill. 53). As Kanchi’s rulers consolidated and expanded the city, the relationship between landscapes and the built environment grew increasingly dynamic. The jungle was converted incrementally into settled terrain, and the new villages became integrated together as people traveled from place to place with ever greater degrees of frequency. Natural resources were manipulated to serve the settlements – rivers and streams were diverted to craft irrigation networks that watered the fertile tracts, and stone quarries were exploited to build temples. The positions of monuments mark out patterns of movement and reveal clusters of settlements that were gradually brought into Kanchi’s kṣetra 3 For studies of landscape and emotion in Caṅkam-era Tamil literature, see George L. Hart III, Poets of the Tamil Anthologies: Ancient Poems of Love and War (Princeton: Princeton University Press, 1979); Martha Ann Selby and Indira Peterson, eds., Tamil Geographies: Cultural Constructions of Space and Place in South India (Albany: State University of New York Press, 2008); Martha Ann Selby, Tamil Love Poetry: The Five Hundred Short Poems of the Ainkurunuru (New York: Columbia University Press, 2011).

The Cit y and its Ports

155

(‘territorial domain’). As its kṣetra grew to connect the inland capital with the broader maritime world, Kanchi’s kṣatra (‘sphere of influence’) also expanded. A cultural landscape extended the purview of Kanchi far beyond the boundaries of the city. Inscriptions functioned in part as a means to establish a growing geographical expanse and sphere of influence for the dynasty that controlled the natural resources and ruled over the land from a seat in the nagaram (‘urban area’). The networks of places that lay beyond the city limits show how Kanchi connected with a wider world system. 4 Kanchi played an important role in long-distance Buddhist networks and connected with other urban centers across South and Southeast Asia. On the micro level, the sites examined in this chapter share certain fundamental qualities. Each has archaeological evidence of an ancient settlement, Pallava and/or Chola-era sculpture or architecture on-site, and at least one temple that received endowments and modifications over time. Each site also shows evidence of significant regional interactions that can be traced through patterns in architecture, iconography, and inscriptions. Tamil inscriptions from these areas designate constellations of places that came to define a rural-urban continuum. Inscriptions also establish boundaries, which distinguished the settled areas from the surrounding wilderness (‘araṇiyam’).5 Contiguous lands were arranged into a hierarchy of subdivisions, which were designated with specific terminology.6 Taṉi-ūrs (‘independent villages’) were clustered together into nāṭus, which were typically named for the primary village in the group. The nāṭus were grouped into kōṭṭams, which in turn lay within the larger category of maṇḍala. For example, eleventh- and twelfth-century inscriptions on Ārpākkam’s Shiva temple refer to Ārpākkam as lying within Mākaṟal-nāṭu, a subdivision of Eyiṟ-kōṭṭam, a district of Jayaṅkoṇḍacōḻa-maṇḍalam.7 This means that Ārpākkam was a taṉi-ūr situated within the cluster of villages in which Mākaṟal was considered the main place. Mākaṟal belonged to the consortium called Eyiṟ. And all were under the dominion of the Cholas in northern Tamil Nadu, called 4 I refer to the formative phases of Abu-Lughod’s world system. Abu-Lughod, Before European Hegemony: The World System A.D. 1250-1350. 5 Heitzman, ‘Temple Urbanism in Medieval South India’, 799. See also Heitzman and Rajagopal, ‘Urban Geography and Land Measurement in the Twelfth Century: The Case of Kanchipuram.’ 6 For longer explanations of the hierarchies of lands and the terminology used in inscriptions, see the works of Heitzman, Hall, Champakalakshmi, and Burton Stein cited throughout this book. 7 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 110, Cg.-461-462.

156 

Construc ting K anchi

Jayaṅkoṇḍacōḻa-maṇḍalam. Whereas Cōḻa-maṇḍalam designated the Kaveri River area, Jayaṅkoṇḍacōḻa-maṇḍalam centered on Kanchi. In Indian literature about kingship and inheritance, the dual concept of kṣetra and kṣatra define a ruler’s domain.8 Both kṣetra and kṣatra derive from the Sanskrit verbal root kṣi. While kṣetra is from the root’s second meaning, ‘to abide, stay, dwell, reside’, kṣatra is from the verb’s first meaning, ‘to possess, have power over, rule, govern, be master over.’9 Kṣetra most directly translates to ‘field’, as in the field a farmer (‘kṣetrika’) ploughs, his soil or landed property, but it can also refer to the body, as in the field for the in-dwelling soul. In the context of kingship, kṣetra refers to the physical terrain within a ruler’s jurisdiction. By contrast, kṣatra holds the less tangible meaning of a ruler’s sphere of influence. Demarcating the kṣetra was thus a critical aspect of governance, but kṣatra was equally important. India’s geographical and political landscapes were both defined by circles of dominion that were forged through techniques of governance, such as allegiances, taxes, and gifts.10 As with the ordering of lands described in inscriptions, hierarchies of kingship were conceptually ordered into maṇḍalas.11 While landed territories (‘kṣetra’) near the urban center may have shifted in and out of the ruler’s direct control, there was also a less proximate zone (‘kṣatra’) in which his power fluctuated.

Part 1: KṢETRA The River Networks Like the north-south roads of Māmaṇṭūr and Uttiramērūr discussed in the previous chapter, the east-west routes from Kanchi are lined with ancient monuments. These east-west routes track along the now dry bed where the Palar River flowed at least through the last quarter of the nineteenth

8 Kulke, ‘Kṣetra and Kṣatra: The Cult of Jagannātha of Puri and the “Royal Letters” (Chāmu Ciṭaus) of the Rājās of Khurda.’ 9 Sir M. Monier-Williams, A Sanskrit-English Dictionary (Springfield, VA: Nataraj Books, [1899] 2010), 327. 10 Stein, Peasant State and Society in Medieval South India; Ronald Inden, ‘Hierarchies of Kings in Early Medieval India’, Contributions to Indian Sociology 15, no. 99 (1981); Ali, Courtly Culture and Political Life in Early Medieval India. 11 For an explanation of the maṇḍala system of kingship, see Ali, Courtly Culture and Political Life in Early Medieval India.

The Cit y and its Ports

157

Illustration 54 Veṅkateśa Perumāḷ Temple, Tirumukkūṭal, ninth century

century.12 To the east of Kanchi, temples are plentiful. The Pallava- and Chola-era temples that stand at Tirumukkūṭal, Kumaravāṭi, Vallam, Orakaṭam, Tirukkaḻukuṉṟam, Tiruviṭantai, and Māmallapuram mark out the paths taken by ancient travelers. In the process, these temples engage local topographies, such as rivers, hilltops, and the sea. Tirumukkūṭal On the southern bank of the confluence of the Palar, Cheyyar, and Vegavathi Rivers stands the ninth-century Veṅkateśa Perumāḷ temple at Tirumukkūṭal (Ill. 54). Now an unimposing shrine set within a modest compound, this temple occupies a critical position halfway between Kanchi and the eastern seaboard. Unlike most other Vishnu temples, which face either east or west, this Vishnu temple faces north towards the confluence, such that the waters of the three rivers metaphorically swirl at the Lord’s feet. When encountered today, the Veṅkateśa Perumāḷ temple is a highly layered building with the later history much more visible than the original 12 In 1874, district collector Charles Stewart Crole reported that the Palar ‘continues in fresh, more or less, high, for six months on end.’ Crole, The Chingleput, Late Madras, District: A Manual Compiled under the Orders of the Madras Government, by Charles Stewart Crole, of the Madras Civil Service, 2.

158 

Construc ting K anchi

Illustration 55 Lion-based pillar, Veṅkateśa Perumāḷ Temple, Tirumukkūṭal, ninth century

structure. Like many Vishnu temples in Tamil Nadu, the exterior is now a sealed compound wall, which functions as a closely set cloistered prākāra that subsumes the main body of the temple. The enclosure’s smooth walls, interrupted only by shallow pilasters, provided ample surface area for inscriptions. Together with the structural fabric of the temple, the multitude of inscriptions reveals the temple’s enduring lifespan and diverse functions.13 A slab bearing a Pallava inscription dated to the reign of Nṛpatuṅgavarman (ca. 893 CE) records a gift for the establishment of a perpetual lamp.14 This tells us a shrine was already present at the rivers’ confluence by the last decade of the ninth century. Portions of the northern wall around the sanctum’s door likely belonged to the Pallava-period temple, as did the elegantly styled lion-based (‘siṁhapāda’) pillars supporting the eastern face of the mukhamaṇḍapa (‘entry hall’), which was added later (Ill. 55). The Chola-period inscriptions provide a wealth of information regarding the establishment’s function as a municipal institution. A 55-line inscription 13 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 332-338, Cg.-1299-1316. 14 Ibid., p. 332, Cg.-1299.

The Cit y and its Ports

159

records the temple’s connection to a sizeable hospital and school in the eleventh century.15 This gives us the sense that Tirumukkūṭal was an important and prominent place, situated along a well-traveled route. The site’s importance is further reinforced through other records. A connection specifically with Kanchi is mentioned in another extensive eleventh-century inscription, which concludes with the stipulation that bundles of hay were to be collected from the village residents and used only for the purposes of cultivating a temple flower garden, rather than being sold privately or sent off to Kanchi.16 The mention of the city in this context suggests that Tirumukkūṭal and Kanchi were linked through frequent commerce. At Tirumukkūṭal, the icon of Vishnu as Veṅkateśa Perumāḷ speaks to larger networks. Facing north, the deity establishes a connection with Tirupati, where the main icon is the same form of Vishnu. The site’s connection with Tirupati is in evidence from at least the Chola period. The long inscription of Vīrarājendra (ca. 1069 CE) mentions provisions for feeding pilgrims journeying to Tirupati.17 The association between the two sites remains strong in local knowledge.18 Up until the most recent renovation, Vishnu’s view north towards Tirupati was preserved. Before the outermost prākāra was added, a screened window through the northern wall allowed for Vishnu’s gaze to pass through uninterrupted. Kumaravāṭi and PV Kaḷattūr Leaving Tirumukkūṭal towards the east, one encounters significant sites at nearly regular intervals. While many of these places today are tiny villages, their Pallava- and Chola-era remains display the same features seen in temples in the most prominent urban centers. These close parallels reveal the former presence of elite patrons, and the transmission of knowledge throughout the Tamil countryside. Most proximate to Tirumukkūṭal, Ciṅkaperumāḷkōyil and the rocky outcrop at Vallam both contain royally sponsored cave-temples that date to the Pallava period.19 Continuing southeast along the Palar, the villages of Kumaravāṭi and PV Kaḷattūr both 15 Ibid., p. 336-337, Cg.-1313. 16 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 334, Cg.-1305. 17 Balasubrahmanyam, Middle Chola Temples: Rajaraja I to Kulottunga I (AD 985-1070), 343. 18 Tirumukkūṭal’s ASI monument attendant replied after some thought that the temple is oriented north because it faces Tirupati, where there is a Tirumukkūṭal gopura (February 2015). 19 Srinivasan assigns the single cave-temple at Ciṅkaperumāḷkōyil to the reign of Narasiṃhavarman I or Parameśvaravarman I, and the upper cave at Vallam to Mahendravarman

160 

Construc ting K anchi

house temples, fragments, and inscriptions that date to the ninth through twelfth century. Situated on opposite sides of the river, these two villages give us important insights into the history of the Palar River. Several scholars have proposed relatively recent changes to the river’s course.20 Geological experts instead contend that although the Palar once flowed north of its present location, it had already shifted its course long before the Pallava era.21 Together with the geological data, the temples at Kumaravāṭi and PV Kaḷattūr show that the river has been steady for upwards of one thousand years. If it had shifted beyond the eight-kilometer tract of land between the villages, it would have flooded either site. Both Kumaravāṭi and PV Kaḷattūr were able to maintain fertile agrarian lands, reservoirs (‘eri’), and rectangular tanks (‘kuḷam’) thanks to the presence of the river. The twelfth-century Muṉkuṭumīśvara temple at PV Kaḷattūr stands on the bank of a large eri. The temple’s god looks out across lush paddy fields that are supplied with water through sluices in the high bund. PV Kaḷattūr has continued to employ its lands for agriculture. The quality of rice, said to be like gold (‘poṉ’), is what gives the village its name, poṉviḷainta kaḷattūr. A ninth-century inscription on a loose slab shows the site has a history that predates the temple, and indeed PV Kaḷattūr is mentioned as a main place (both nāṭu and kōṭṭam) in an eleventh-century inscription at Orakaṭam.22 Today, the brilliant green fields and shimmering waterworks contrast sharply with the many surrounding villages, where lands perpetually change hands for the purported purpose of tax-free development. To the west of the Palar, Kumaravāṭi is surrounded by rich reserves of water. The large kuḷam (here ‘stepped pond’) at the southern edge of the settlement must have once sustained far more than the humble present-day village would suggest. Sculptures and architectural fragments unearthed during the laying of roads in the area have been installed on the reservoir’s eastern bank (Ill. 56).23 The Aḻakeśvara temple, dedicated to Shiva, remains I. The two lower caves at Vallam belong to the later Pallava period, possibly during the reigns of Rājasiṃha and Nandivarman II. Srinivasan, Cave-Temples of the Pallavas, 110-111 and 32-34. 20 Heitzman, ‘Urbanization and Political Economy in Early South India: Kāñcīpuram During the Cōḻa Period’, 127, n. 8. 21 Dr. Singanenjam, ‘Pātai Māṟiya Pālāṟu’, Tiṉamaṇi 72 (2005). Warm thanks to Dr. Singanenjam for providing this valuable reference. 22 The inscriptions are dated 860 CE and 1016 CE respectively. Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 9, Cg.-40 and p. 30, Cg.-140. 23 Most of Kumaravāṭi’s archaeological remains were reported (in list form only) in IAR 1990-1991, p. 64.

The Cit y and its Ports

161

Illustration 56 Sculptures near pond, Kumaravāṭi, ninth century and later

intact and is situated at the far northeastern edge of the village. This shrine faces east towards lush paddy fields populated by wandering cattle and their herdsmen. The temple has been rebuilt over time, but the granite basal moldings appear to be original and made of local stone. The upper face of the tripaṭṭa-kumuda molding on the south side is lined with inscriptions that have been attributed to the ninth century.24 A figure of Durgā inside a separate, west-facing shrine is likely contemporary with the main temple (Ill. 57). Kumaravāṭi can be accessed by means of a small road from Uttiramērūr, which is fourteen kilometers due west. I suspect this road is not an ancient route, or at least not one that was well traveled, because it crosses several sizeable expanses that appear in satellite imagery to have been bodies of water not long ago. These include the ghost of a distributary that flowed southwards off the Cheyyar, beginning at a point near the village of Putali. Running parallel to the Kanchi-Uttiramērūr Road, this arm of the river extended southwards to Marakkāṇam, where it emptied into the sea. Marakkāṇam’s Bhūmīśvara temple, built in the late tenth or very early eleventh century, stands at the juncture of the former river and the coast.25 24 IAR 1990-1991, p. 64. 25 The Bhūmīśvara temple’s first inscription is 1002 CE. Srinivasan, ‘Cōḻas of Thanjavur: Phase II’, 249.

162 

Construc ting K anchi

Illustration 57 Durgā, Kumaravāṭi, ninth century

Whereas Tirumukkūṭal was situated at a confluence of rivers, Kumaravāṭi was conveniently and perhaps strategically placed between the Palar and this southern distributary of the Cheyyar. Irrigation channels from these rivers would have kept its pond full and fields plentifully watered, while enabling the village and its temple to achieve relative prosperity. Nattam-Paramēsvaramaṅkaḷam Southeast from Kumaravāṭi is the tiny settlement of NattamParamēsvaramaṅkaḷam.26 The site is home to the beautiful Ceṇpakeśvara temple (Ill. 58). Unlike the temples at Kumaravāṭi and PV Kaḷattūr, much of Ceṇpakeśvara’s original appearance is preserved. The earliest inscription is from the third regnal year of Rājēndra Chola (ca. 1015 CE), but the architectural features and sculptural style are consistent with tenthcentury temples from the period of the great temple patron, Queen Sembiyan

26 The village is not to be confused with nearby Paramēsvaramaṅkaḷam. My warm thanks to A. Valavan for informing me of this temple.

The Cit y and its Ports

163

Illustration 58 Ceṇpakeśvara Temple, Nattam-Paramēsvaramaṅkaḷam, tenth century

(Cempiyaṉ) Mahadevi.27 It may be more than coincidence that the temple’s name, Ceṇpakeśvara, resembles the name of the queen, Cempiyaṉ – the temple’s architectural integrity, as well as the high quality of carving in the sculptures, suggests a prosperous source of patronage. The niche icons are original and follow a pattern of placement imported from the Kaveri area that spread throughout Chola territories in Sembiyan Mahadevi’s time.28 Bhikṣāṭaṇa, Ganesha, and Dakṣiṇāmūrti occupy the south wall, Liṅgōdbhavamūrti the west, and Brahmā and Durgā the north. Details were not omitted. Dakṣiṇāmūrti has a medicine bag slung over a branch in the tree above him, and Durgā has a parakeet delicately balanced on her wrist. The techniques of production are also the same as those used in the Kaveri area. Carved on single blocks of stone, the figures fill the niches completely. On the north wall, a scene of liṅga worship also recapitulates similar images from tenth-century temples in the Kaveri region, such as the Siddhanātha temple in Tirunaraiyūr.

27 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 347, Cg.-1354. 28 For images of the niche icons, see Stein, ‘All Streets Lead to Temples: Mapping Monumental Histories in Kanchipuram, ca. 8th-12th centuries CE’, Figs. 3.16-3.19.

164 

Construc ting K anchi

Despite its elite appearance, Ceṇpakeśvara does not seem to have been the recipient of continuous sponsorship for construction. Besides the addition of an eastern maṇḍapa and several very recent, separately standing shrines made of concrete, few significant modifications to the temple have been made. The main temple body stands exposed in an open field – no prākāra wall or external structure has been added. The location close enough to the river for the fields to be fertile would have made this area attractive for elite developers at the time of the temple’s construction. I find it curious that its upkeep did not continue. Nattam simply means residential area or village hamlet, but today the epithet is locally said to designate the place where the outcastes from the main place dwell. If this tradition has a longer history, perhaps it is the reason Nattam-Paramēsvaramaṅkaḷam became overshadowed by the proximate village called Paramēsvaramaṅkaḷam. Vāyalūr The Tiruppulīśvara temple complex in the village of Vāyalūr marks the place where the Palar River splays out into an estuarine delta. The multireligious complex consists of a ca. eleventh-century apsidal Shiva temple and a square shrine dedicated to Vishnu. Vāyalūr has been the subject of scholarly attention primarily for its contested Pallava inscription.29 Dating to the reign of Rājasiṃha (ca. 700-725 CE), the inscription is located on a pillar in the style of Mahendravarman I that has been built into the entry gate on the eastern side of the complex. Historians such as Mahalingam have used this inscription as evidence for an antiquated theory of overseas imperialistic contacts between the Pallavas and maritime Southeast Asia.30 The inscription was originally translated: ‘May he exercise the royal prerogative and take up the vow of administering (his) subjects up to the extremities of his kingdom, as even to include the thousand islands.’31 The basis of the argument for imperialism hinged on a single word, read as dvīpalakṣam, meaning ‘thousand islands.’ The interpretation was that Rājasiṃha’s influence extended far into the seas, and this mythology was perpetuated in subsequent scholarly literature.32 29 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 105-106, Cg.-445. 30 Mahalingam, Kāñcīpuram in Early South Indian History, 116. 31 Ibid. 32 Citing the Vāyalūr inscription as evidence, Nanditha Krishna writes of the ‘seafaring Pallavas.’ Nanditha Krishna, ed. Kānchi: A Heritage of Art and Religion (Madras: C.P. Ramaswami Aiyar Institute of Indological Research, 1992), 4.

The Cit y and its Ports

165

The notion that India engaged in maritime imperialism in the premodern era is a particular paradigm of historiographic analysis. A new reading of the inscription effectively dismisses the possibility that the contents of the text refer to maritime activity. Upon their firsthand examination of the Vāyalūr pillar, Francis, Gillet, and Schmid found that the crucial word reads varṣalakṣam, meaning ‘thousand years’, rather than dvīpalakṣam/ lakṣadvīpam, ‘thousand isles.’33 Through their revised translation, the only definitive mention of foreign territories in the corpus of royal Pallava inscriptions disappears.34 Vāyalūr’s proximity to the delta suggests it could have functioned as a harbor or port, and local knowledge tells us that the village of Vasavasamudram, two-and-a-half kilometers upstream, was a port in the Pallava era.35 However, archaeologists have determined that the elevation with respect to sea level at both of these locations is too great for them to have held such a function.36 Excavations at Vasavasamudram further revealed a relatively short, if ancient, period of inhabitation, dating to around the second century.37 One cluster of ruins in a peanut field is all that remains of its old temple. Yet even without the specific function of a port, both Vāyalūr and Vasavasamudram were deemed important enough for elite and even royal patronage of sacred architecture in the Pallava and Chola eras.

Over the Hills A second sequence of monuments extends east and slightly south from Kanchi to the sea. This overland route passes the rocky cave-temple site of Vallam before winding through hillocks that rise suddenly from an otherwise flat landscape. In places where residences and markets cluster around the foot of a hill, the summit is often marked with a temple. In many cases, that temple’s role as the main local shrine has been supplanted by a more 33 Emmanuel Francis, Valérie Gillet, and Charlotte Schmid, ‘De Loin, de Près: Chronique des Études Pallava III’, Bulletin de l’Ecole Française d’Extrême-Orient 94 (2007): 302. 34 Ibid. 35 Champakalakshmi, ‘The Urban Conf igurations of Toṇḍaimaṇḍalam: The Kāñcīpuram Region, C.A.D. 600-1300’, 197. 36 All sites used as ports in the Pallava period are situated at present-day elevations lower than six meters above sea level. Vāyalūr and Vasavasamudram are elevated significantly higher. D. Dayalan, ‘Punceri, an Ancient Seaport in Tamil Nadu: New Light on the Location of the Ancient Seaport of Mamallapuram’, (unpublished). 37 IAR 1970-1971, p. 33.

166 

Construc ting K anchi

Illustration 59 Vedagirīśvara and Bhaktavatsaleśvara Temples, Tirukkaḻukuṉṟam

accessible temple at the base of the hill (Ill. 59). Although the hilltop site may not receive much foot traffic, it still serves an important function – visible in the distance, it signals the approach to a settled area and aligns with other hilltop destinations. For example, Orakaṭam’s ninth-century temple is situated four kilometers directly north of an earlier hilltop monument at Tirukkaḻukuṉṟam. Tirukkaḻukuṉṟam The Vedagirīśvara temple stands on the southernmost summit of a dramatic series of hills looming over the small town of Tirukkaḻukuṉṟam. At the base of the mountain, the village’s densely packed market street stretches from the more conveniently situated temple complex of Bhaktavatsaleśvara to the large ablutions tank of Caṅku Tīrtham (Ill. 60). At first, the massive scale of the tank appears to grossly outweigh the Vedagirīśvara temple’s relatively small expanse. However, the tank’s immensity is in balance if one considers the entire mountain as a sacred edifice.38

38 Thanks to Amanda Culp for this perceptive observation (28 February 2016).

The Cit y and its Ports

167

Illustration 60 Caṅku Tīrtham from Vedagirīśvara Temple, Tirukkaḻukuṉṟam

Multiple techniques of intervention have shaped the mountain over time. Carved into the living rock, steep flights of stone stairs lead up to the hilltop temple along two pathways. Halfway up the circuitous route, the Orukkal cave penetrates the slope’s eastern side. Just beside the Vedagirīśvara temple, a trench in the rock serves as a water encatchment that once provided daily libations for the legendary pair of sacred eagles (‘tirukkaḻu’) that gives the town its name. Tirukkaḻukuṉṟam’s chronology ranges from at least the seventh century through to the present.39 The site saw a period of increased temple construction during the Pallava era, particularly in the seventh century and then again in the ninth. The Orukkal cave-temple’s walls are replete with finely carved relief sculptures of dvārapālas and deities.40 An inscription dated to Mahāmalla’s reign (ca. 630-668 CE) records a grant of land for performing worship to the god on top of the hill, which suggests there was already a shrine there (likely made of brick) at that time. 41 Inscriptions on both the 39 For a study of the town, see K.V. Gopalakrishnan, Tirukkaḻukkuṉṟam (Pakṣitīrtham) and its Temples: A Study of their History, Art, and Architecture (New Delhi: Abhinav Publications, 2005). 40 Srinivasan argues the Orukkal cave was probably excavated under Narasiṃhavarman (Mahāmalla) I (ca. 630-668 CE) in the style of Mahendravarman I. Srinivasan, Cave-Temples of the Pallavas, 40-43. 41 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 67, Cg.-294. See also Srinivasan, ‘The Pallava Architecture of South India’, 132.

168 

Construc ting K anchi

Illustration 61 Vīṇādhara-Ardhanārīśvara, Vedagirīśvara Temple, Tirukkaḻukuṉṟam, seventh century

Vedagirīśvara and Bhaktavatsaleśvara temples record donations of sheep for a great many lamps, giving us the sense that Tirukkaḻukuṉṟam was bustling with life and flooded with light throughout the Pallava and Chola era. The town is specifically connected with Kanchi through a twelfth-century inscription from the Bhaktavatsaleśvara temple’s shrine for the Tamil Saint Māṇikkavācakar, where a guild of oil merchants from Kanchi is said to have come together and agreed to supply the daily provision of oil for a lamp before the image. 42 From its reconstruction to its present-day use, the Vedagirīśvara temple is quite literally a multilayered building. It is a constricted, intimate space, with a Pallava-era core sheathed in a Chola-era exterior, and further surrounded by a closely set covered prākāra. The main shrine is thought to be one of the earliest structural temples in the region, built by Parameśvaravarman I, Rājasiṃha’s predecessor, in the last quarter of the seventh century.43 Its walls are simple slabs, carved minimally with divine images and perhaps originally 42 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 75, Cg.-324. 43 Srinivasan, ‘The Pallava Architecture of South India’, 132.

The Cit y and its Ports

169

Illustration 62 Bhaktavatsaleśvara Temple, Tirukkaḻukuṉṟam, ninth century and later

supplemented with brick and stucco. 44 In the Chola era, walls were built around this sanctum as a kind of shell. Niche-shaped punctures on three sides reveal the Pallava-era relief carvings behind them (Ill. 61). VīṇādharaArdhanārīśvara seated against the bull is on the south wall, Somāskanda on the west, and Shiva seated in royal ease (‘rājamūrti’) on the north. The same figures grace the corresponding walls on the inside of the sanctum.45 Enlivened by the flickering butter lamps of an evening pūjā, these are some of the most beautifully executed relief carvings to survive from the Pallava era. At the base of the hill, the pristine Bhaktavatsaleśvara temple is a very different type of structure (Ill. 62). It is a standard, full complex, complete with concentric enclosure walls penetrated by towering gopuras, a main temple comprised of multiple shrines, and several gardens and water features including a square tank in the northeast quadrant of the complex. In contrast 44 ‘Pallavas of Kāñcī: Phase I’, 45. 45 Lockwood’s identification of this early Somāskanda has demonstrated that the practice of placing Somāskanda on the rear wall of the inner sanctum, behind the liṅga, was initiated by Parameśvaravarman rather than (as is more popularly believed) Rājasiṃha. Lockwood, Pallava Art, 21-45.

170 

Construc ting K anchi

Illustration 63 Dakṣiṇāmūrti, Bhaktavatsaleśvara Temple, Tirukkaḻukuṉṟam, ninth century

to the Vedagirīśvara’s quiet intimacy, Bhaktavatsaleśvara is a constant flurry of activity. The eastern gopura announces the temple’s presence at the end of the bustling market street, and devotees pass continually in and out of the temple. Despite their differences, however, a strong connection is maintained between the two temples, as each is immediately visible from the other. Bhaktavatsaleśvara also houses Pallava-era architecture (Ill. 63). The oldest surviving portion of the complex is a south-facing, apsidal shrine with its niche figures intact. This shrine was completed by the end of the ninth century, according to the date of an inscription from the reign of Āditya I (ca. 898 CE).46 A heavily worn figure of Caṇḍeśa on the premises also appears to be of significant age. It is possible that this Caṇḍeśa originally occupied the south-facing shrine, which is in the proper position for the deity.47 Since the 46 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 67, Cg.-295. The inscription is reportedly on the wall of the strong room of the Vedagirīśvara temple, which does not have such a room. I assume this is a confusion with the Bhaktavatsaleśvara. 47 For a seminal study of Caṇḍeśa, see Dominic Goodall, ‘Who Is Caṇḍeśa?’, in Genesis and Development of Tantrism, ed. Shingo Einoo (Tokyo: Institute of Oriental Culture, University of

The Cit y and its Ports

171

Illustration 64 Henry Salt, Pagodas at Trinchicunum, Plate 11 in ‘Twenty-Four Views in St. Helena, the Cape, India, Ceylon, the Red Sea, Abyssinia and Egypt’, 1804, wash, 46.8 × 62.2 cm, British Library (BL_WD1305)

shrine now functions as the temple’s locked strong room, the icon could have been brought out and installed in the main complex, where it is presently mounted on a concrete block to the north of the ardhamaṇḍapa’s door. Tirukkaḻukuṉṟam remained prominent enough throughout the subsequent centuries to have made its way into a colonial-era guide by the Indian Army Major Herbert Andrew Newell.48 As a member of the military, Newell was particularly impressed by the Vedagirīśvara’s external appearance, which he described as a fortified, prison-like structure.49 Newell’s choice to visit Tirukkaḻukuṉṟam was unusual. Colonial travelers whose journeys were motivated by antiquarian interests more frequently stayed in Māmallapuram (see Chapter Four), near the Dutch fort at Sadras.50 Another unusual visitor Tokyo, 2009). 48 H.A. Newell, Seven Pagodas, Chingleput and Conjeeveram: An Illustrated Guide with History (Madras and Bangalore: Higginbothams, 1914), 7. 49 Ibid. 50 First to describe and illustrate Māmallapuram was the Dutch traveler Jacob Haafner, in 1786. Paul van der Velde, Life Under the Palms: The Sublime World of the Anti-colonialist Jacob Haafner (Singapore: Ridge Books, National University of Singapore, 2020), 115. I thank Paul van

172 

Construc ting K anchi

was the artist Henry Salt, who traveled alone between Pondicherry and Māmallapuram in 1804 while his employer, the Viscount Valentia, kept to the coastal road. Salt stopped to see several sites along the way, among them Tirukkaḻukuṉṟam. In his illustrated memoir of the journey, Salt describes the Pagodas at Trinchicunum (an Anglicization of Tirukkaḻukuṉṟam) as follows: The Chief Brahmin resides on the hill, where the Pagoda has the appearance of a fortification and would in fact be extremely difficult of approach, were it not for a handsome flight of steps cut in the rock. The Pagodas below consist, as usual, of a lofty wall of stone, and having gateways of a pyramidal form, ending in something like a sarcophagus. These are surrounded by groves of coconut, mango and tamarind trees.51

Salt’s representation shows the temple in a picturesque landscape from the northeast corner of Caṅku Tīrtham (Ill. 64). One of the Bhaktavatsaleśvara’s gopuras rises behind it, with rolling hills in the distance. The Vedagirīśvara temple’s prominent position on the top of the main hill is emphasized by the rocky outcroppings that lead up to it, ringed at the base with lush trees bursting with foliage. In the foreground, clusters of people are neatly dispersed across the picture plane. They prepare to enter the water, or they sit at ease on low-lying boulders. The hovering clouds and pale sky suggest an early evening scene, when the light begins to turn from gold to pink.

The Coast Like thickened darkness, cool moist shadows lie on sand as white as gathered moonlight. With its black-branched puṉṉai trees the flowering grove is empty. Still he does not come, but the boats draw near of my brothers who search for fish. – Kuṟuntokai 12352

der Velde for showing me the illustration, J.E. de Moor and Paul van der Velde, eds., De Werken van Jacob Haafner, Vol 3 (Zutphen: Walburg Pers., 1997), 402. 51 Henry Salt, Twenty-Four Views in St. Helena, the Cape, India, Ceylon, the Red Sea, Abyssinia and Egypt (London: William Miller, 1809), Pl. 11. 52 Hart III, Poets of the Tamil Anthologies: Ancient Poems of Love and War, 64.

The Cit y and its Ports

173

Illustration 65 Coastal Māmallapuram at sunrise, Shore Temple, ca. 700-725 CE

Māmallapuram Ten kilometers to the east of Tirukkaḻukuṉṟam, paddy fields and forest wilderness give way to the scrubby screw pines and bleached sands of a coastal landscape. At Māmallapuram, colorful vessels line the shores and fishermen string their nets and set out at sunrise (Ill. 65). Māmallapuram’s landscape has been shaped by centuries of intervention. Its sacred architecture has fundamentally manipulated and engaged the natural elements. Most of the principal monuments, including the widely known Great Relief, are carved into or built from a single north-south running ridge of charnockite granite that is situated just inland from the sandy beach. The famous Five Rathas are in a separate area, carved from a cluster of monumental boulders to the south of the main ridge, and another three rathas are situated to the west. The lily-covered Kōṉēri Lake – a quarried-out tank – stretches between the ridge and the western rathas. Grooves and punctures in the living rock throughout the site divert and direct the flow of water. It is highly likely that water was used for ritual purposes here, as many scholars have also suggested.53 Specifically, it may 53 Padma Kaimal, “Playful Ambiguity and Political Authority in the Large Relief at Mamallapuram,” Ars Orientalis 24 (1994). Kaimal has also suggested that an image of Gajalakṣmī was carved on the rear wall of the so-called Tiger Cave in Cāḷuvaṉkuppam, where another pit fills with water during times of flood. “Lakṣmī and the Tigers: A Goddess in the Shadows,” in The

174 

Construc ting K anchi

Illustration 66 Great Relief during 2015 flooding, Māmallapuram (photo by Mallai Dilip)

have been poured through the central cleft in the Great Relief to bathe the auspicious nāga-deities and fill the cistern below.54 In times of heavy rain, the cistern still becomes full, and the elephants seem to splash in the water, their feet submerged just below the surface (Ill. 66).55 A great deal of scholarly attention has been lavished upon Māmallapuram’s many monuments, yet many questions remain unanswered. Debates have revolved around the chronology of the temples, the identities of their patrons, the iconography, and the interpretation of particular sculptural elements Archaeology of Bhakti: Mathurā and Maturai, Back and Forth, ed. Charlotte Schmid and Emmanuel Francis (Pondicherry: École Française d’Extrême-Orient, 2013). Lutzker placed great emphasis on the importance of water at this monument. Mary-Ann Lutzker, “A Reinterpretation of the Relief Panel at Māmallapuram,” in Chhavi-2: Rai Krishnadasa Felicitation Volume (Bhanaras: Bharat Kala Bhavan, 1981). Kumar-Dumas has suggested that the cascade issued through the spout that now lies partially buried in the base of the cistern. Divya Kumar-Dumas, “Understanding Difference in Landscape Architectural History of Sri Lanka and India,” (Paper presented in 45th Annual Conference on South Asia, Madison, WI, 2016). 54 Kaimal, “Playful Ambiguity and Political Authority in the Large Relief at Mamallapuram.” 55 Thanks go to Mallai Dilip for photographing during flooding, when I was unable to reach the site (November-December 2015).

The Cit y and its Ports

175

Illustration 67 Shore Temple, Māmallapuram, ca. 700-725 CE

– most notably the Great Relief at the center of the ridge.56 Recent scholarship has focused on what the monuments’ lack of finish can tell us about processes of production. For decades, scholars have acknowledged the likely presence of multiple groups of masons and artists.57 Most recently, Dehejia and Rockwell identified at least six teams to account not only for the excavation and carving of rock-cut monuments but also for the clearing of removed stones.58 Connecting technique with symbolic meaning, Samuel Parker suggested that the unfinished nature of Māmallapuram’s monuments was intended to convey the continuously manifesting nature of the universe.59 But Māmallapuram was also deeply connected with Kanchi, and it played an important role as a final stop along the pathway of temples from the urban capital to the sea. 56 Srinivasan, Cave-Temples of the Pallavas; Michael Lockwood, G. Siromoney, and P. Dayanandan, Mahabalipuram Studies (Madras: Diocesan Press, 1974); Michael D. Rabe, ‘The Monolithic Temples of the Pallava Dynasty: A Chronology’ (University of Minnesota, 1987); Kaimal, ‘Playful Ambiguity and Political Authority in the Large Relief at Mamallapuram’; Michael D. Rabe, ‘The Māmallapuram Praśasti: A Panegyric in Figures’, Artibus Asiae 57, no. 3/4 (1997); The Great Penance at Māmallapuram: Deciphering a Visual Text (Chennai: Institute of Asian Studies, 2001); Francis, Gillet, and Schmid, ‘L’Eau et le Feu: Chronique des Études Pallava.’ 57 Srinivasan, Cave-Temples of the Pallavas, 26-27. 58 Dehejia and Rockwell, The Unfinished: Stone Carvers at Work on the Indian Subcontinent. 59 Samuel K. Parker, ‘Unfinished Work at Mamallapuram or, What is an Indian Art Object?’, Artibus Asiae 61, no. 1 (2001).

176 

Construc ting K anchi

The extensive monumentification of Māmallapuram under the Archaeological Survey of India (ASI) hinders our vision of the site as a whole. Most notably, the ASI has continually removed what are referred to throughout their reports as ‘accretionary structures’, with the intention to more fully expose the stone monument (Ill. 67). At the Shore temple in 1990, for example, ‘several accretionary structures like walls, steps, etc. were removed.’60 In the process, the ASI has demolished and failed to document structures that would have supplied important insights into the ways in which the site was used at various moments in history. Bricks of varying age can be found encrusted throughout the entire area. Some are the remains of the residential and municipal buildings that once populated the spaces between stone monuments. Large-scale bricks that date to the early centuries of the first millennium lie to the west of the Shore temple.61 Their discovery in the 1980s pushed the date of the site’s earliest activity back by more than five hundred years. Ironically, the large bricks were found only two years before the removal of more ‘accretionary structures.’ A great many additional bricks are concentrated especially at the ridge’s mesa-like western summit, near a flat stone bed carved with a regal lion. Near to the lion is an excavated pool and the remains of a square, stone foundation built on top of the living rock. This area was long thought to have contained a palace, but it could have held any kind of prosperous establishment – royal, religious, or some combination of the two. We know that Māmallapuram attracted elite patronage. The multitude of stone monuments, royal inscriptions, and elegant sculptures could hardly have resulted from anything less. Factors that attracted such patronage include the site’s geology and its natural resources – the period of time in which Māmallapuram’s builders crafted the landscape into a rich web of sacred edifices was precisely when stone was becoming a powerful commodity. In Māmallapuram, laterite and charnockite of excellent quality were so plentiful that the site could function as its own stone quarry. Stake holes, sheer faces, and incisions can still be seen across nearly every visible surface (Ill. 68). While some grooves in the rock may have been used for drainage, for ritual purposes, or for the insertion of brick and timber structures, many reveal processes of stone extraction.62 Near the western group of rathas, 60 IAR, 1990-1991, p. 140. 61 Archaeologists believe the ancient brick ruin was originally a wall that jutted far out into the sea. IAR 1988-1989, p. 155. 62 Dayalan, ‘Punceri, an Ancient Seaport in Tamil Nadu: New Light on the Location of the Ancient Seaport of Mamallapuram’; Dehejia and Rockwell, The Unfinished: Stone Carvers at Work on the Indian Subcontinent.

The Cit y and its Ports

177

Illustration 68 Quarried Boulders, Māmallapuram

an imposing block of stone sits in a circular pocket in the earth. This pit closely resembles the punctured landscape at Ārpākkam, among other stone-rich sites throughout Tamil Nadu. The technique for extracting rock also appears to be the same as in other parts of southern Asia, such as the so-called stone pools at Dimbulagala in Sri Lanka. The blocks of stone hewn from Māmallapuram’s main ridge were used locally for building temples, but no temples made of Māmallapuram charnockite can be found outside the immediate area. However, more stone was removed than can be accounted for by the three structural temples that survive at the site alone.63 The remains of an additional temple dedicated to Murukaṉ were recently revealed near the group of monuments in Cāḷuvaṉkuppam, and the basement of another large monument was excavated just south of the Shore temple, situated directly on the beach. The Mukunda Nāyaṉār temple, which survives primarily through excavation and reconstruction by the ASI, is sunken deep below ground level. Unlike the Ōlakkāṉeśvara and Shore temples, which were built on sturdy floors of living rock, these temples did not have solid foundations and were unable to maintain their structural integrity. There are likely many more sacred monuments that 63 Māmallapuram’s three structural temples include the Olakkaṇṇeśvara, Mukuṇṭa Nāyaṉar, and Shore temples.

178 

Construc ting K anchi

have been buried in sand or swallowed by the sea.64 The abundance of locally available stone, in addition to water, made Māmallapuram a fitting place in which to bring about the birth of stone temple architecture in South India. Māmallapuram is commonly believed by scholars of Tamil literature and history to be the coastal port-town of Nīrppeyaṟṟuttuṟai described in the Perumpāṇāṟṟuppaṭai. This Caṅkam-period text traverses the landscape from the seaside town to Kanchi.65 A similar pathway is also described in the Periya Purāṇam, in the story of Tirukkuṟipput Toṇṭar. In this text, Māmallapuram, called Māmallai, is said to exemplify the connection between coastal and mountainous landscapes. From the sea whence roll waves enfolding chanks, The coral-creeper spreads and twines with its shoots The branches of sandal-tree; thus are there Many lands linking neytal with kuṟiñci Where thrive great towns like Māmallai Rich in mansions of pennant-flaunting balconies66

Since the colonial era, Māmallapuram has been believed to be the so-called royal port of the Pallava dynasty.67 If it held this role, the coastal town would have functioned in tandem with the inland urban capital of Kanchi, in much the same way that the early Chola’s inland capital at Uṟaiyūr worked 64 There is a long mythology to a city beneath the sea off the coast of Māmallapuram. It stems in part from the presence of a rocky ridge that runs parallel to the shore and protrudes in certain places during low tide. Māmallapuram’s early epithet, ‘Seven Pagodas’, has also been a subject of much speculation. The earliest source is the writings of the sixteenth-century Venetian merchant Gasparo Balbi. Balbi’s remarks have gone untranslated from the Italian: ‘All that night we sailed downwind, holding the bow to the north. The next morning at 3:00 we came to a place known as the Seven Pagodas. On top of the pagodas are eight small hills that are lovely and not especially tall. The place is at a distance of seven leagues from San Thome, where we arrived at midday on 30 May, greeting our destination with three shots of artillery.’ I am grateful to Giulia Nardini for translation assistance. Gasparo Balbi, ‘Viaggio dell’Indie Orientali di Gasparo Balbi, Gioielliero Venetiano (1579-1588)’, in Viaggi di C. Federici e G. Balbi alle Indie Orientali, ed. Olga Pinto (Rome: Instituto Poligrafico dello Stato, 1962), 160-161. See also an account by the eighteenth-century traveler William Chambers, in M.W. Carr, ed. Descriptive and Historical Papers Relating to the Seven Pagodas on the Coromandel Coast (New Delhi: Asian Edicational Services, (1869) 1984), 1-29. In the early nineteenth century, a similar mythology of an underwater city seems to have been attached to the coastal town of Cuddalore, referenced in the papers of Henrietta Clive. Shields, Birds of Passage: Henrietta Clive’s Travels in South India, 1798-1801, 251. 65 My thanks to Valérie Gillet for directing me to this reference (April 2014). 66 Periya Purāṇam, line 1118. 67 Krishna, Kānchi: A Heritage of Art and Religion.

The Cit y and its Ports

179

together with the port of Pukār. Merchant ships were harbored and the goods were unloaded at Pukār for sale in the local bazaars, and for transport to the capital and elsewhere in the Chola’s domain. The Cilappatikāram [‘Tale of the Anklet’] provides rich descriptions of the vast Pukār bazaars. The pairing of inland urban capital and coastal port was in fact a standard formula throughout premodern South Asia, including Sri Lanka.68 However, Māmallapuram was probably not the most practical site for a port-town designed to distribute large shipments of merchandise. Colonial-era records tell us that Māmallapuram was disconnected from the mainland by at least the nineteenth century, if not long before.69 Visitors who approached by land describe reaching a place that had to be crossed by means of a small boat.70 The contours of the island are clear from satellite imagery – the backwaters at Kōvaḷam and Kuṉṉattūr marked its northern and southern extremes, respectively.71 In 1914, Newell provided an informative description of the approach to Māmallapuram in his illustrated guide. As the seacoast draws near, the road degenerates into a mere sandy track winding across a desert sparsely grown with thorn bushes and the ubiquitous cactus. At the Buckingham Canal, the shallow, sluggish waters of which stretch between Lakes Pulicat and Ennore, the jhadka [‘bullock cart’] is exchanged for a large, flat-bottomed barge. Towards this primitive vessel the traveler wades knee deep in slime to be finally ferried across, partly by means of a pole and partly by touring.72

If Māmallapuram were an island during the Pallava period, it would have made little sense for it to function as a port. Moreover, the waters between Kōvaḷam and Māmallapuram were dangerous. They contain a reef close to shore that posed a signif icant challenge for approaching ships.73 D. Dayalan has suggested that Pūñcēri, a village west of Māmallapuram on 68 Osmund Bopearachchi, ‘Sites Portuaires et Emporia de l’Ancien Sri Lanka’, Arts Asiatiques 54 (1999). 69 James Fergusson, Picturesque Illustrations of Ancient Architecture in Hindostan, by James Fergusson (London: J. Hogarth, 1847), 57; Newell, Seven Pagodas, Chingleput and Conjeeveram: An Illustrated Guide with History, 9. 70 Seven Pagodas, Chingleput and Conjeeveram: An Illustrated Guide with History, 9. 71 Kumar-Dumas, ‘Understanding Difference in Landscape Architectural History of Sri Lanka and India.’ 72 Newell, Seven Pagodas, Chingleput and Conjeeveram: An Illustrated Guide with History, 9. 73 The East India Company’s ship, Rockingham, wrecked against the reef in 1776. Crole, The Chingleput, Late Madras, District: A Manual Compiled under the Orders of the Madras Government, by Charles Stewart Crole, of the Madras Civil Service, 10.

180 

Construc ting K anchi

the opposite side of the backwater (and therefore on the mainland) was a likely alternative location for the port.74 Boats could have docked in the backwater and unloaded at Pūñcēri before being distributed to the markets and the city. Māmallapuram may not have been a central port, but it was not isolated either. Although accessing it may have required travel by boat, the plethora of Pallava- and Chola-period monuments suggests that the journey did not deter patrons from sponsoring the construction of temples nor their continued maintenance, expansion, and ritual life. In addition to Māmallapuram, the temple sites of Tiruviṭantai and Cāḷuvaṉkuppam are also situated on the island. Māmallapuram, Cāḷuvaṉkuppam, and Tiruviṭantai all house architecture that post-dates the Pallava period by many centuries. Among the most important post-Pallava remains at Māmallapuram are extensive expansions and additions to the Sthalaśayana Perumāḷ temple in the center of town, a ca. eleventh-century set of Saptamātṛkās with Shiva, and two prominently situated inscriptions in the Shore temple.75 Dating to the eleventh century, the Shore temple inscriptions record divisions of land and taxes to be levied on various communities, which tells us that Māmallapuram was a lively and productive settlement.76 At Cāḷuvaṉkuppam, architectural components of the Murukaṉ temple can also be dated to the eleventh century, and Tiruviṭantai is home to the tenth-century Nitya Kalyāṇa Perumāḷ temple.77 Newell made a perceptive observation about the connection between Māmallapuram and Kanchi’s temple architecture in the introduction to his guide. It seems that a visit to the ancient Pallava seaport would be incomplete unless the old Pallava capital were seen as well, and the relationship traced between the celebrated Kailāsanātha Temple at the latter place and the world-famed rock shrines on the seashore.78

74 Dayalan, ‘Punceri, an Ancient Seaport in Tamil Nadu: New Light on the Location of the Ancient Seaport of Mamallapuram.’ 75 Mahalingam, A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District, p. 22-23, Cg.-105-106. 76 Ibid. 77 Of the numerous inscriptions at the Nitya Kalyāṇa Perumāḷ temple at Tiruviṭantai, at least nine are dated to the tenth century. The architectural style of the vimāna and ardhamaṇḍapa agree with this date. Ibid., p. 92-95, Cg.-392-400. 78 Newell, Seven Pagodas, Chingleput and Conjeeveram: An Illustrated Guide with History, II.

The Cit y and its Ports

181

Illustration 69 Ascetic and the Birth of Pallava Relief, Vaikuṇṭha Perumāḷ Temple, Kanchi, ca. 730-795 CE

Indeed, close links between individual sculptural and architectural forms in Kanchi and Māmallapuram connect the monuments in both places. These formal parallels lay the parameters for the ‘Pallava style’ of art within and beyond the region. Most explicitly, the Vaikuṇṭha Perumāḷ temple in Kanchi looks to Māmallapuram for its architectural design and sculptural motifs. Māmallapuram’s Dharmarāja Ratha is the prototype for the Vaikuṇṭha Perumāḷ temple’s vertically stacked, triple-shrine plan. A figure of an ascetic practicing yoga in Māmallapuram’s Great Relief and in a panel at the Shore temple again reappears in the Vaikuṇṭha Perumāḷ temple’s prākāra (Ill. 69).79 The ascetic at the Vaikuṇṭha Perumāḷ temple appears within the important scene depicting the birth of Pallava, progenitor of the Pallava lineage, which is prominently positioned to be one of the first images to come into view when the visitor turns towards the standard direction of circumambulation. This representation was likely intended to bring the Great Relief to mind for the viewer, in order to connect eighth-century Kanchi with the Pallava’s coastal domain.

79 The Great Relief f igure has been subject to numerous interpretations. See for example Kaimal, ‘Playful Ambiguity and Political Authority in the Large Relief at Mamallapuram’; Rabe, ‘The Māmallapuram Praśasti: A Panegyric in Figures.’

182 

Construc ting K anchi

The Vaikuṇṭha Perumāḷ temple’s arrangement of narrative reliefs further connects this particular monument with Māmallapuram. The Shore temple also contains a set of reliefs that are similarly positioned along the interior face of a prākāra wall that closely surrounds the central shrine. Whereas the Shore temple’s reliefs function primarily as distinct courtly scenes now too eroded to read in full, the Vaikuṇṭha Perumāḷ’s reliefs appear to have been designed as a narrative sequence.80 Deities, kings, and courtiers both terrestrial and divine become characters ordered by mythology, hagiography, and praśasti.81

Part II: KṢATRA Kanchi’s kṣetra can be mapped by plotting temples, rivers, and roads within a relatively circumscribed locality. However, more ineffible and difficult to trace is its kṣatra, which extended across a broader region and far beyond the borders of the Tamil land. People and goods moved systematically within the kṣetra, but what circulated across the kṣatra were less tangible assets – Hindu-Buddhist religious philosophies, ideologies of kingship, artistic iconographies, architectural motifs. Their agents of transmission were merchants, ritual specialists, artisans, entrepreneurs, and pilgrims who traveled widely, circulating goods, objects, texts, and knowledge. Tamil merchants were among the earliest traders to be active in Southeast Asia, and other Shaiva and Muslim communities also played important roles in the spread of Indic culture overseas.82 Mentioned in inscriptions and other forms of early literature, Southeast Asia was very much a part of the integrated cultural zone that came to define the Tamil kṣatra.

Kanchi in a Buddhist World Existing literature on maritime networks of the first millennium is vast, and it is well known that Buddhists were among the many active participants in

80 Gillet has argued for the development of a narrative sequencing to Shiva’s iconography that took place in Pallava temples during the course of the eighth century. Gillet, La Création d’une Iconographie Śivaïte Narrative: Incarnations du Dieu dans les Temples Pallava Construits. 81 Francis, Gillet, and Schmid, ‘L’Eau et le Feu: Chronique des Études Pallava’, 597. 82 Guy, Lost Kingdoms: Hindu-Buddhist Sculpture of Early Southeast Asia, 10.

The Cit y and its Ports

183

long-distance trade.83 Buddhists warrant particular attention here because of their explicit ties to Kanchi.84 Monks associated with Kanchi are accredited with translating and transmitting the religious and philosophical systems that would form the core teachings of Buddhist tantra.85 Buddhist travelers who were not monks often carried texts and small bronze sculptures along with their merchandise. It was largely due to these portable goods that artistic styles and iconographies were transmitted and entrenched in areas across a wider South and Southeast Asian world.86 The idea of a widespread Buddhist network was central to the forging of a collective identity for Buddhist practitioners in an increasingly Shaiva-centric religious landscape.87 This was true for Buddhists in Tamil Nadu as much as anywhere else, especially while the Shaiva and early Vaishnava saints articulated the supremacy of their own faith through sensuous devotional

83 Important relevant works include, George F. Hourani, Arab Seafaring in the Indian Ocean in Ancient and Early Medieval Times (Princeton, NJ: Princeton University Press, 1951); K.N. Chaudhuri, Trade and Civilisation in the Indian Ocean: An Economic History from the Rise of Islam to 1750 (Cambridge: Cambridge University Press, 1985); Abu-Lughod, Before European Hegemony: The World System A.D. 1250-1350; Jan Wisseman Christie, ‘The Medieval Tamil-Language Inscriptions in Southeast Asia and China’, Journal of Southeast Asian Studies 29, no. 2 (1998); Himanshu Prabha Ray and Jean-Francois Salles, eds., Tradition and Archaeology: Early Maritime Contacts in the Indian Ocean (New Delhi: Manohar, 1996); Pierre-Yves Manguin, A. Mani, and Geoff Wade, eds., Early Interactions between South and Southeast Asia: Reflections on Cross-Cultural Exchange (Singapore: Institute of Southeast Asian Studies, 2011); Rila Mukherjee, Vanguards of Globalization: Port-Cities from the Classical to the Modern (Delhi: Primus Books, 2014). An exhibition at the Metropolitan Museum of Art highlighted the role of art in these networks. Guy, Lost Kingdoms: Hindu-Buddhist Sculpture of Early Southeast Asia. 84 Khmer inscriptions refer to Sanskrit texts that guided the development of Shaivism in Cambodia. Alexis Sanderson, ‘The Śaiva Religion among the Khmers, Part I’, Bulletin de l’Ecole Française d’Extrême-Orient 90-91 (2003-2004): 353. 85 Kanchi is referred to repeatedly in Andrea Acri, ed. Esoteric Buddhism in Mediaeval Maritime Asia (Singapore: ISEAS Publishing, 2016), see for example p. 10 and 237-252. See also Jeffrey Sundberg and Rolf W. Giebel, ‘The Life of the Tang Court Monk Vajrabodhi as Chronicled by Lü Xiang (呂向): South Indian and Śrī Laṅkān Antecedents to the Arrival of the Buddhist Vajrayāna in Eighth-Century Java and China’, Pacific World: Journal of the Institute of Buddhist Studies 13 (2011). 86 A growing body of literature exists on this topic. For Indonesia, see the seminal work by Pauline Scheurleer and Marijke J. Klokke, Divine Bronze: Ancient Indonesian Bronzes from A.D. 600 to 1600 (Leiden: E.J. Brill, 1988). The research continues with Mathilde Mechling, ‘The Buddhist and Hindu Bronze Statues from Indonesia in their Religious, Cultural and Artistic Contexts (7th-11th centuries)’ (University Sorbonne Nouvelle-Paris 3 and Leiden University, Dissertation in Progress). 87 Monius, Imagining a Place for Buddhism: Literary Culture and Religious Community in Tamil-Speaking South India, 84.

184 

Construc ting K anchi

poetry – as is often cited, the tenth verse of each of Saint Sambandar’s hymns typically contains some denunciation of Buddhists. Anne Monius has drawn attention to the calculated ways in which the sixth-century Tamil epic Maṇimēkalai worked to balance an idea of transregional Buddhism with a more immediate, localized Buddhist vision. The linguistic choice of Tamil as opposed to Sanskrit for the text was a careful move, Monius argues, intended ‘as a means of claiming a place for Buddhism in a religiously diverse landscape and literary culture and making that immense community envisioned also local, familiar, close at hand.’88 However, it should not be overlooked that the re-centralized, ‘local’ Buddhist landscape, as Monius defines it, was itself transregional in the text. While the near-exclusion of North India is significant, the Maṇimēkalai’s world encompassed both South India and maritime Southeast Asia.89 The Maṇimēkalai aimed to define this South and Southeast Asian cultural landscape in Buddhist terms. However, the same geographical reconception was also applied within Shaiva and Vaishnava communities.90 As we know, especially from surviving temple-cities in Indonesia and Cambodia, Buddhist and Hindu cults existed simultaneously and were the recipients of both royal and local patronage. Himanshu Prabha Ray cautions us not to stumble into the pitfall of imagining an earlier Buddhist landscape that was entirely supplanted by a later Hindu dominance.91 To do so would be to follow the narrative of the disappearing Buddhists and Jains that has loitered in scholarly literature since its formulation during the early colonial era.92 While texts like the Maṇimēkalai endeavored to inscribe one religion’s prominence over another, the evidence is often not borne out by the material world. 88 Monius, Imagining a Place for Buddhism: Literary Culture and Religious Community in Tamil-Speaking South India, 83. 89 Ibid., 88 and 102. 90 Alexis Sanderson, ‘The Śaiva Age: The Rise and Dominance of Śaivism During the Early Medieval Period’, in Genesis and Development of Tantrism, ed. Shingo Einoo (Tokyo: Institute of Oriental Culture, University of Tokyo, 2009); ‘The Śaiva Religion among the Khmers, Part I’; Ronald Inden, ‘Imperial Purāṇas: Kashmir as Vaiṣṇava Center of the Words’, in Querying the Medieval: Texts and the History of Practices in South Asia, ed. Ronald Inden, Jonathan Walters, and Daud Ali (Oxford University Press, 2000). 91 Himanshu Prabha Ray, ‘A “Chinese” Pagoda at Nagapattinam on the Tamil Coast: Revisiting India’s Early Maritime Networks’, IIC Occasional Publications 66 (2015). 92 Colin Mackenzie, ‘The Jains’, Asiatic Researches 9 (1809); Fergusson, Picturesque Illustrations of Ancient Architecture in Hindostan, by James Fergusson; Davis, ‘The Story of the Disappearing Jains: Retelling the Śaiva-Jain Encounter in Medieval South India.’ See also Chapter Four in this book.

The Cit y and its Ports

185

Kanchi emerges as an important connector between regions and a place of intersection for different religions, not only in the realm of literature but also in real-world terms. At the story’s culmination, Maṇimēkalai’s miraculous bowl is enshrined in Kanchi on a lotus pedestal that replicates an original on the distant island of Maṇipallavam. The bowl’s previous travels serve as a medium for reimagining the extent of the Buddhist world, as it is envisioned in the text.93 Monius argues that the acts of building a shrine and installing the object in Kanchi ‘mark this city as an important site in the Buddhist world’, and that – purely for the purposes of the text – Kanchi assumes the place of central importance.94 The Maṇimēkalai may well have operated in particular ways and for particular purposes within the realm of literary imagination. However, there is also much to indicate that within a thriving, well-connected Buddhist community in Southeast Asia and South India, Kanchi’s status as an important center was very real indeed. The Chinese pilgrim Xuanzang’s mid-seventh century visit shows us that Kanchi was widely known in broader Buddhist circuits.95 His disciple, HwuiLih, adds that Kanchi was a place of refuge for Sri Lankan Buddhist monks.96 Xuanzang reports that the sixth-century Buddhist teacher Dharmapāla was born in Kanchi, and several Chinese sources on the life of Vajrabodhi (ca. 671-741 CE) claim the same origin for the great monk.97 Others say that the Pallava king Rājasiṃha invited Vajrabodhi from Pandya country to Kanchi so that he could use his powers to bring rains during a time of drought.98 The sixth-century Buddhist teacher Bodhidharma is commonly believed to have been a Kanchi prince, and there is a Bodhidharma meditation center to honor him in Kanchi today.99 Abhayadatta’s foundational text for Vajrayāna Buddhism, the hagiographies of the 84 Mahāsiddhas (the 93 Monius, Imagining a Place for Buddhism: Literary Culture and Religious Community in Tamil-Speaking South India, 101. 94 Monius, Imagining a Place for Buddhism: Literary Culture and Religious Community in Tamil-Speaking South India, 100-102. 95 Xuanzang, Si-Yu-Ki. Buddhist Records of the Western World. Translated from the Chinese of Hiuen Tsiang (A.D. 629) by Samuel Beal, 2, 228-229. 96 The Life of Hiuen-Tsiang, by the Shaman Hwui Li. (London: Kegan Paul, Trench, Trubner, 1911), 138-139. 97 Xuanzang, Si-Yu-Ki. Buddhist Records of the Western World. Translated from the Chinese of Hiuen Tsiang (A.D. 629) by Samuel Beal, 2, 229. C.D. Orzech, ‘Vajrabodhi’, in Esoteric Buddhism and the Tantras in East Asia, ed. C.D. Orzech, H.H. Sorensen, and R.K. Payne (Leiden and Boston: Brill, 2011), 346. 98 R. Champakalakshmi, Religion, Tradition, and Ideology: Pre-colonial South India, 348. 99 For the tradition of Bodhidharma, Minakshi, Administration and Social Life Under the Pallavas, 222 and 224; Champakalakshmi, Religion, Tradition, and Ideology: Pre-colonial South India, 348.

186 

Construc ting K anchi

Caturaśīti-siddha-pravṛtti, ca. late eleventh/early twelfth century), claims that Sakara and the prolific master Nāgārjuna were both born in Kanchi, and that Tilopa practiced austerities in a cremation ground in Kanchi.100 On a more local level, the Periya Purāṇam tells us the Shaiva saint Cākkiyar was originally a Buddhist and went to Kanchi to study Buddhism.101 In the ninth through twelfth century, Kanchi was the point of origin for a great many monks who donated stone and bronze sculptures to the important Buddhist monastery at Kurkihar, in Bihar, Northeast India, situated more than 2,000 kilometers away.102 These sculptures were commissioned on-site from the consumate craftsmen in the area, and they were often inscribed with the donor’s name and birthplace. Of the 56 inscribed bronzes published by Picron, at least sixteen mention donors who hail from Kanchi.103 These industrious Buddhist pilgrims may have come from several different monastic institutions in Kanchi. Xuanzang describes a monastery and a large stupa, and the Mattavilāsa Prahasana mentions a royally sponsored Buddhist monastery in the city.104 Chola-period inscriptions in Kanchi’s temples tell us that there were multiple Jain and Buddhist monasteries (‘paḷḷiccantam’ and ‘bauddha-paḷḷikku-paḷḷiccantam’).105 The archaeologist C. Minakshi mentions an inscription that tells of a buddhapaḷḷi in the vicinity of the Varadarāja Perumāḷ temple, and, by correlating literary references with archaeological remains, she concludes that there were no less than 100 Abhayadatta, Buddha’s Lions: The Lives of the Eighty-Four Siddhas (Caturaśīti-Siddha-Pravṛtti), trans. James B. Robinson (Berkeley: Dharma Publishing, 1979), 75, 99, and 227. This important Sanskrit text (the Caturaśīti-siddha-pravṛtti) was translated into Tibetan (the Grub thob brgyad cu rtsa bzhi’i lo rgyus) by sMon-grub Shes-rab. Thanks go to Rebecca Bloom for drawing my attention to Kanchi’s inclusion in this reference. Jeff Watt has also mentioned a Vajrayana tradition that the eighth-century monk Shantaraksita was born in Kanchi (Pers. comm. January 2017). 101 The extent to which the Periya Purāṇam derides Buddhism shows that in the twelfth century Buddhism was still perceived as a threat to the text’s ideal readership. Periya Purāṇam, line 3636. 102 Pal has noted a Kurkihar bronze Avalokiteshvara gifted by the Kanchi monk Viryavarman in the collection of the Los Angeles County Museum of Art (M.81.8.3). Pratapaditya Pal, Indian Sculpture, 2 vols., vol. 2 (Los Angeles: Los Angeles County Museum of Art, 1988). For a longer study of Kurkihar sculptures, see Claudine Bautze-Picron, The Forgotten Place: Stone Images from Kurkihar, Bihar (New Delhi: Archaeological Survey of India, 2015). 103 My thanks to Sayantani Pal for discussion. 104 Xuanzang, Si-Yu-Ki. Buddhist Records of the Western World. Translated from the Chinese of Hiuen Tsiang (A.D. 629) by Samuel Beal, 2, 230; Mattavilāsa Prahasana, line 64. Bhatt and Lockwood, Mattavilāsa Prahasana: ‘The Farce of Drunken Sport’ by King Mahendravarma Pallava. See Chapter One in this book. 105 Heitzman, ‘Urbanization and Political Economy in Early South India: Kāñcīpuram During the Cōḻa Period’, 128, n. 11.

The Cit y and its Ports

187

six Buddhist monasteries in the city.106 It is possible that elite sponsorship for such establishments became unstable during the change of regime from the Pallavas to the Cholas, thus prompting significant portions of the monastic community to leave the city in the tenth century and make the journey to Kurkihar.107 The Kanchi Buddhas Recent scholarship has shifted the date of Buddhism’s disappearance in South India substantially forward, from the eighth to the thirteenth century or even later.108 Along with the Chola-period inscriptions and literary references to Buddhism, a multitude of Buddhist statuary that survives in Kanchi supports the revised chronology. Spread throughout the city, large-scale stone sculptures dating from as early as the sixth century up to the late Chola period attest to the enduring presence of an affluent Buddhist community. The post-thirteenth century images of Buddhas that can be found in Kanchi are relief carvings on pillars in which the Buddha is depicted among other avatars of Vishnu. Although the Buddha was first included in lists of Vishnu’s avatars as early as the seventh or eighth century, the fact that the pre-thirteenth-century figures of the Buddha in Kanchi are found in isolation, and that nothing comparable from another religious affiliation exists, suggests they represent the Buddha himself rather than an avatar of Vishnu.109 Buddhist bronzes and archaeological evidence from Nagapattinam have been reasonably well published, yet with the exception of Minakshi’s research, sources on Buddhist remains in Kanchi have tended only to discuss 106 Minakshi, ‘Buddhism in South India’, 86 and 98. 107 The Kurkihar bronzes remain an important subject for anyone undertaking a longer study of the Buddhist world. Questions that may be pursued include what route these monks took, where they were going, and where they may have stopped along the way. The distance from Kanchi to Kurkihar is vast, roughly the same as the distance from Kanchi to Palembang across the Indian Ocean, but the journey would have been undertaken by foot rather than by ship. Gerd Mevissen suggested that the monks were traveling to Nalanda. Deborah Klimburg-Salter proposed they were on pilgrimage to the Buddhist sites in North India, as Kurkihar is only 25 kilometers from Bodh Gaya. I thank these scholars for discussing the topic with me during the European Association of South Asian Archaeology and Art conference in Cardiff (July 2016). 108 Dehejia, ‘The Persistence of Buddhism in Tamil Nadu’; Davis, ‘The Story of the Disappearing Jains: Retelling the Śaiva-Jain Encounter in Medieval South India.’ 109 The Buddha is listed among avatars of Vishnu in an inscription at the Ādivarāha cave-temple at Māmallapuram, dating to the seventh or eighth century on paleographic grounds. Gillet, ‘Pallavas and Buddhism: Interactions and Influences’, 108.

188 

Construc ting K anchi

Illustration 70 Three Buddhas, Paḷḷūr, eleventh/twelfth century

two of the city’s numerous surviving examples.110 The relative prevalence of Nagapattinam in scholarly literature is perhaps due to the site’s economic function during the Chola period, as well as its geographical position within the better-studied Kaveri area. Buddhist statues exist both within Kanchi and in proximate villages – not only beneath ground level, but also hidden in plain sight. Here we move from kṣatra back onto the firmer ground of kṣetra. In the 1980s, a farmer in the village of Paḷḷūr (fifteen kilometers northwest of Kanchi) was digging in his field when his blade struck something hard. With the help of his wife and other villagers, he proceeded to unearth three monumental seated Buddhas made of a gray-green granite stone. Each was subtly different in appearance – all seated with the right leg crossed over the left and spine upright, two with the hands at rest in their laps and one in the classic earth-touching gesture, the largest Buddha’s face framed by a 110 John Guy, ‘The Lost Temples of Nagapattinam and Quanzhou: A Study in Sino-Indian Relations’, Silk Road Art and Archaeology 3 (1993-1994); Ray, ‘A “Chinese” Pagoda at Nagapattinam on the Tamil Coast: Revisiting India’s Early Maritime Networks.’ The Kanchi Buddhas that have been discussed by several scholars are those found at the Kāmākṣī Ammaṉ and Karukkiṉil Amarntavaḷ Ammaṉ temples. Dehejia, ‘The Persistence of Buddhism in Tamil Nadu’; Rao, ‘Bauddha Vestiges in Kanchipura’; Venkataraman, Dēvī Kāmākshī in Kāñchī: A Short Historical Study. A more comprehensive listing of Buddhas in and around Kanchi can be found in Minakshi, ‘Buddhism in South India’, 109-113. Many of the Buddhas Minakshi mentions can no longer be found on-site.

The Cit y and its Ports

189

Illustration 71 Map of Buddhas and Buddhist remains in Kanchi (map by Emma Natalya Stein and Daniel Cole, Smithsonian Institution)

flaming arch. From that moment onwards, the farmer and his wife devoted themselves to the Buddhas, tending for them as if they were not statues but people. They relocated the Buddhas from the field, placing them at first in a clearing on the northern edge of the village. But when it rained, mudslides would flood the clearing and the statues would fall backwards. ‘He’s fallen!’ the farmer would cry, and the villagers would come rushing to help him put the Buddhas upright. In 1998, using their own money, the farmer and his wife completed the construction of a maṇḍapa, where they installed the three massive statues on a cement pedestal (Ill. 70). Similar events have transpired in the proximate villages of Kaṇikiluppai, Ārpākkam, and Kaḷakkāṭṭūr, where large-scale statues of the Buddha have been found in what is usually described to me as a ‘fallen place’, from which they were recovered by local residents and often installed in a nearby Hindu shrine. Of the great many Buddhas within Kanchi, some have been placed inside temples, while others reside in rather unexpected locations throughout the city (Ill. 71). Like the Buddhas at Ārpākkam and Kaṇikiluppai, two seated Buddhas were discovered in the southern part of Kanchi in the first decade of the twentieth century by a local farmer, who placed them among an eclectic group of deities in the courtyard of the Karukkiṉil Amarntavaḷ

190 

Construc ting K anchi

Illustration 72 Buddha, Karukkiṉil Amarntavaḷ Ammaṉ Temple, Kanchi, eleventh/ twelfth century

Ammaṉ temple (Ill. 72).111 A group of seven seated Buddhas carved in relief is built into the exterior prākāra wall of the Ekāmbaranātha temple, near the parking lot.112 An exceptionally beautiful Buddha lives beneath a tree in the yard of the Subaraya Mudaliar Boys School, where the headmaster ensures that the students treat it with respect. At the Shiva-Kanchi Police Station, near the Ekāmbaranātha temple, a Buddha is installed in a recently constructed pavilion opposite a shrine for Ganesha (see Ill. 4). Although the shrine is new, the Buddha has been at the Police Station for nearly a century. When C. Minakshi inquired about its origins in the 1930s, she was told it had been brought from a Kanchi suburb known as Pusivakkam.113 I have heard that it was either found nearby and brought into the police station for safekeeping, or that it was unearthed more or less where it sits today. In any case, the statue comes from the local area. 111 Rao, ‘Bauddha Vestiges in Kanchipura.’ 112 Minakshi has published a photograph of a two-foot-long reclining Buddha, also from Ekāmbaranātha, which unfortunately has been either removed or obscured since the time of publication. C. Minakshi, Kanchi: An Introduction to its Architecture (Delhi: Publications Division, Ministry of Information and Broadcasting, 1954). 113 Minakshi, ‘Buddhism in South India’, 110.

The Cit y and its Ports

191

Illustration 73 Buddha, formerly in Kāmākṣī Ammaṉ Temple, Kanchi, sixth century (present location: Government Museum, Madras)

Most dramatic among the findings in Kanchi is a monumental standing Buddha recovered from the inner prākāra of the Kāmākṣī Ammaṉ temple, reported by Gopinatha Rao in 1915 in the journal Indian Antiquary.114 After many years behind locked doors, the figure is now on display in the Amarāvatī wing of the Madras Government Museum, where it was quite literally stuck in a corner (Ill. 73). Unlike the corpus of Chola-period seated Buddhas, the Kāmākṣī Ammaṉ Buddha stands nearly seven feet tall and is represented in the style associated with later Amarāvatī. Scholars have therefore dated the sculpture to as early as the sixth century, which makes it contemporary with Monius’s dating for the Maṇimēkalai text.115 In 1970, excavations inside the Kāmākṣī Ammaṉ temple complex and near the Shankaracharya Mutt revealed remains of circular brick structures, dating to the second or first century BCE, that closely resemble known Buddhist stūpas found elsewhere in Tamil Nadu.116

114 Rao, ‘Bauddha Vestiges in Kanchipura.’ 115 Dehejia, ‘The Persistence of Buddhism in Tamil Nadu’, 54. 116 IAR, 1969-1970, p. 34-35.

192 

Construc ting K anchi

The presence of Buddhas and Buddhist architectural remains in and immediately around the Kāmākṣī Ammaṉ temple suggests that this location was previously a Buddhist site that was reappropriated at a later time by Hindus, as scholars including Rao and Dehejia have suggested. In his 1915 report, Rao considered the possibility that the Kāmākṣī Ammaṉ Buddha was placed inside the temple for safekeeping. However, from the Buddha’s abandoned state in the innermost prākāra and the figure’s monumental scale, he concluded instead that the temple had been built around the statue.117 Through a detailed study of this temple, the historian K.R. Venkataraman persuasively argued that the Kāmākṣī Ammaṉ temple was originally a cluster of Buddhist and Jain shrines that were later consolidated into the vast complex we see today.118 As he demonstrates, the conversion process was not abrupt but gradual, with the various deities removed from the site and relocated to other temples or museums over the course of the last four hundred years.119 The Kanchi Buddha statues may not be the only Buddhist artistic remains in the city. The series of sandstone relief carvings at the Tāṉtōṉṟīśvara temple, and the f igures of ascetics on fragmentary pillar bases at the Kacchapeśvara and Vīraṭṭāṉeśvara temples bear no attributes except water pots and liquor skeins.120 Situated in sites that have had multiple renovations and reconfigurations over time, and fragmentary as they are, these sculptures resist sectarian identification – they could have belonged to a Buddhist or Jain context as easily as to a Hindu one. Gillet has shown that Buddhism was so influential in the Pallava period that new forms of Shiva were developed in response to Buddhist imagery.121 Whether or not we include the Tāṉtōṉṟīśvara, Kacchapeśvara, and Vīraṭṭāṉeśvara sculptures, the multitude of Buddhist remains gives a sense of the large and thriving Buddhist community that existed in Kanchi at least through the late twelfth century.122 117 Rao, ‘Bauddha Vestiges in Kanchipura.’ 118 Venkataraman, Dēvī Kāmākshī in Kāñchī: A Short Historical Study. I am indebted to this publication, which I accessed at the AIIS library in Gurgaon. 119 Tradition holds that the Jain goddess called Dharmadevī installed in the Trilokya Jinasvāmi temple in Tirupparuttikkuṉṟam was brought from the Kāmākṣī Ammaṉ temple. Ramachandran, Bulletin of the Madras Government Museum: Tirupparuttikkuṉṟam and its Temples, 20. 120 See Chapter One. 121 Shaiva forms that absorb elements of Buddhist iconography include Bhikṣāṭaṇamūrti, Liṅgodbhavamūri, and Dakṣiṇāmūrti. Gillet, ‘Pallavas and Buddhism: Interactions and Influences.’ 122 The abrupt end to the artistic evidence after the twelfth century bears further interrogation. When Buddhas appear in reliefs from the post-Chola era in Kanchi, it is in their capacity as avatars of Vishnu. They are typically placed on pillars among the other daśavatara.

The Cit y and its Ports

193

With the exception of the Buddha from the Kāmākṣī Ammaṉ temple, all the Kanchi Buddhas are iconographically and stylistically consistent. They are large-scale granite f igures of Shakyamuni, seated simply in meditation or in bhūmisparśa mudra, calling the earth to witness at the moment of his enlightenment. They wear a close-fitting robe with a narrow fold over the shoulder that terminates in a rippling wave. The border of the ankle-length garment is gently upraised, leaving the feet exposed. Significantly, the legs are crossed in vīrāsana, with one foot laid on top of the other, rather than in the full lotus pose more usually seen in figures of Buddhas from elsewhere in India and from the Himalayas. The flaming uṣṇīṣa that surmounts the head is also a noteworthy departure from Indian prototypes. The style of these Buddhas bears close resemblance to sculptures in bronze and stone from nearby Nagapattinam, but also from Sri Lanka and both maritime and mainland Southeast Asia. The similarity is especially evident in the position of the legs and the rendering of the flaming uṣṇīṣa. The region in which similar Buddhas are found maps neatly onto the Buddhist world that is imagined in the Maṇimēkalai. The parallel landscapes further indicate that the literary and artistic worlds were part of a shared cultural zone that spanned both sides of the Indian Ocean.123

The City and its Mirrors Kanchi was a key participant in the integrated South and Southeast Asian world, not only for its connections to religious and material culture, but also for the structure of its urban landscape. Kanchi is not the only place deserving of its popular epithet, ‘city of one thousand temples’ (Ill. 74). Although not in Tamil Nadu, there are places farther af ield that share important characteristics with Kanchi, such as Bhubaneshwar in Orissa, Polonnaruwa in Sri Lanka, Bagan in Myanmar (Burma), Prambanan in Indonesia, and Angkor, Koh Ker, and Sambor Prei Kuk in Cambodia. Each of these cities is peppered with a dense but wide distribution of temples, and the majority of the temples was established or converted to stone during the course of the eighth through thirteenth century – a period that seems to constitute a widespread ‘temple boom’ that spanned South and Southeast

123 Ray, ‘A “Chinese” Pagoda at Nagapattinam on the Tamil Coast: Revisiting India’s Early Maritime Networks.’

194 

Construc ting K anchi

Illustration 74 Temple-cities that flourished ca. eighth through thirteenth century

Asia.124 In addition to having many temples, each city was simultaneously a political, commercial, and religious center with an elite courtly culture and a multireligious population. With that said, Kanchi is unique among these places in certain ways. Gorgeous and pleasant to visit as they can be, sites like Bagan in Myanmar and Angkor in Cambodia today have been designated as protected archaeological parks, with ticketed entries and only a handful of active temples (Ill. 75). At places where urban life continues in close proximity to the temples, such as Bhubaneshwar in Orissa and Prambanan in Java, the old temples have become archaeologically preserved monuments, frozen as a particular vision of their time and no longer in use as centers of devotion. By contrast, Kanchi continues to function as an active city and commercial center, as well as an important pilgrimage destination. Consequently, Southeast Asian temple-cities can supply important insights into aspects of Kanchi’s appearance in earlier centuries. At the most granular level, specific parts of temples in Southeast Asia supply missing information about original structures and materials. Some of the perishable materials even survive, preserved by jungle and forest overgrowth. In a living city like Kanchi, what were originally corresponding architectural 124 I refer also to the period of increased trade in the Indian Ocean that took place precisely at the same time. See Jan Wisseman Christie, ‘Javanese Markets and the Asian Sea Trade Boom of the Tenth to Thirteenth Centuries AD’, Journal of Economic and Social History of the Orient 41, no. 3 (1998).

The Cit y and its Ports

195

Illustration 75 Bagan, Myanmar

elements have long since been either demolished or rebuilt in stone or concrete. For example, whereas in rural Tamil Nadu one finds the remains of stone-based maṇḍapas that once supported brick walls, in Polonnaruwa’s ruins, portions of the actual brick walls survive.125 More broadly, the Southeast Asian cities preserve clear evidence of their religious plurality. Whereas Kanchi’s plethora of Buddhist buildings survive only in the form of fragmentary statuary, fully complete Hindu and Buddhist monuments stand side-by-side at Bagan and Angkor. A shrine containing a liṅga is situated inside the grounds of a Buddhist monastery in Polonnaruwa. The Shaiva and Buddhist establishments once coexisted as active religious sites within close proximity or even a single shared space. Kanchi remains unique also for its pattern of orienting temples towards the city’s central road (discussed in Chapter Two). I have traveled to Bhubaneshwar, Polonnaruwa, Bagan, Prambanan, Angkor, Koh Ker, and Sambor Prei Kuk to map and note the locations and orientations of temples. Nowhere but Kanchi does this convention hold. It existed in the fabric of the city. Resisting transmission, this knowledge remained deeply local.

125 An example is the ninth-century Shiva temple at Kāḷiyāpaṭṭi.

196 

Construc ting K anchi

Conclusion: From Kanchi to the Sea Described as timeless and eternal throughout the colonial period, villages in India were increasingly treated as evidence of India’s stagnation in a distant, utopian past. Scholars of no less prominence than Henry Maine and Karl Marx maintained the fiction of the self-sustaining village community, which was stagnant both physically and developmentally.126 For Marx, the village was a unitary place where agriculture and manufacturing converged.127 Throughout such colonial-era discourse runs the belief that the village constituted an idealized connection with a truly authentic India.128 The emphasis on the village as a sedentary space was part of a broader trend during the colonial period that sought to fix people in place and minimize the extent to which they wandered.129 This drive to locate, segment, and classify was applied not only to physical space, but also to categories of race, class, and gender. In the Gandhian era, when nationalist movements within India subverted the European vision and instead celebrated the village, they still fetishized it as the bastion of ancient tradition – a keeper of the past. This chapter has argued instead that the rural and the urban in Tamil Nadu did not constitute distinct entities. Rather, the city and its hinterland stretched along a continuum of incrementally settled landscape. Temples constructed in Kanchi and its surrounds during the eighth through thirteenth century reveal patterns of frequent movement that created constellations of settled and productive lands. Groundbreaking epigraphical studies by Noboru Karashima were largely responsible for reformulating scholarship on South India towards a more mobile vision of society. Uprooting the fiction of stagnation, Karashima used inscriptions to demonstrate that village life in South India was not confined to the village.130 The movement of people and the circulation of knowledge kept Kanchi’s boundaries permeable. Kanchi’s kṣetra extended into proximate sites in the agrarian hinterland through commerce, pilgrimage, and the use of shared 126 The Marxian view of the Indian village is productively discussed in Karashima, South Indian History and Society: Studies from Inscriptions, A.D. 850-1800, xxii. 127 Ibid. 128 A large body of literature exists on the village in colonial discourse. For a good discussion of the village as nationalist icon, see Mary Elizabeth Hancock, The Politics of Heritage from Madras to Chennai (Bloomington: Indiana University Press, 2008), 179-188. 129 Irschick has argued that the colonial government sought to spatialize, reconstruct, and resacralize Tamil Nadu’s rural society during the late eighteenth and early nineteenth century. He calls this process ‘sedentarization.’ Eugene F. Irschick, Dialogue and History: Constructing South India, 1795-1895 (Berkeley: University of California Press, 1994). 130 Karashima, South Indian History and Society: Studies from Inscriptions, A.D. 850-1800, xxll.

The Cit y and its Ports

197

artistic and architectural practices. Donors who contributed to multiple temples moved along the ancient routes and had their names inscribed in stone at each temple they visited. Legal agreements recognized these sites as participants in a single political ecology. Set in distinct natural landscapes, the temples also engaged their surrounding environments, as well as connecting with other monuments across a broader geography. Vishnu surveyed his city, gazed towards Tirupati, or watched as the river’s waters bathed his feet. Hilltop temples served as beacons for approaching travelers. Both inside and outside of Kanchi, architectural plans were selected and adjusted in order to meet the shifting needs of their users. Places farther afield reveal less palpable connections that can be understood as the city’s kṣatra. This landscape was connected through religious networks, portable assets, and a broadly shared urban structure. Aspects of Kanchi in turn are revealed through these distant places. In particular, Kanchi’s Buddhist history can be seen in its scattered sculptures and archaeological evidence. But the city’s legacy in intellectual networks further points to a flourishing base of Buddhist patronage in Kanchi throughout the Chola period. Architectural practices were among the systems of knowledge that circulated across South and Southeast Asia. Agents of this spread traveled along the deeply etched trade routes that spanned the broader region. Although I have yet to find another city in which all temples orient towards a central road, Kanchi’s urban structure was only unique in terms of temple orientation. As a whole, the city shares its general pattern of temple distribution with ancient urban sites in Sri Lanka and Southeast Asia that saw the bulk of their construction at the very same time as Kanchi. I would not propose that one place was directly responsible for another. Rather, these cities belonged to a shared cultural zone that extended across the Indian Ocean and even beyond the littoral areas. Kanchi’s efflorescence constituted an astonishing process of urbanization that transcended its local context.

4

Kanchi Under Colonialism Abstract Using colonial-era sources, Chapter Four examines Kanchi at the crucial moment when Indian art and geography was first being systematized, interpreted, and transmitted by Europeans. A popular narrative was established that characterized Kanchi as a city in decline – yet this narrative is belied by Colonial-era prints, drawings, and travelers’ accounts. While photographs show Kanchi’s Pallava temples in various states of ruination, these alternative sources tell a story of Kanchi as a flourishing urban center. The original sources include the collected manuscripts of Colin Mackenzie, the architectural writings of James Fergusson, the private diaries of Henrietta Clive (Robert Clive’s daughter-in-law), and the drawings and letters of the Russian aristocrat Alexis Soltykoff, who described Kanchi as a city of ‘infinite temples.’ Keywords: Nineteenth century, Photography, Clive of India, Panorama, Mysore Wars, Pallava temple

The multitude of people was quite amusing, but the noise and the dust were very great. -- Lady Henrietta Clive, 18001

On 11 March 1800, Lady Henrietta Antonia Clive and her entourage arrived in Kanchi.2 A portrait by Sir Joshua Reynolds, painted two decades earlier, shows her in stately elegance, donning a wide-brimmed, peacock-blue hat and swathed in a heavy dress the color of cream, with a slightly dissatisfied look across her face (Ill. 76). Armed with Robert Orme’s recently published 1 Shields, Birds of Passage: Henrietta Clive’s Travels in South India, 1798-1801, 109. Warm thanks to Siddhartha Shah for suggesting this invaluable source to me (London 2016). 2 Ibid.

Stein, Emma Natalya, Constructing Kanchi: City of Infinite Temples. Amsterdam, Amsterdam University Press 2021 doi: 10.5117/9789463729123_ch04

200 

Construc ting K anchi

Illustration 76 Sir Joshua Reynolds, Lady Henrietta Antonia Herbert, Countess of Powis (1758-1830), 1777-78, oil on canvas, 55¼ x 44¼ in, Powis Castle, National Trust, 1181064

A History of the Military Transactions of the British Nation in Indostan as her guidebook, restless Henrietta struck out to see places associated with her father-in-law, Robert Clive – the famous ‘Clive of India’ – while her husband Edward remained stationed in Madras. According to the diary of Charlotte Clive, Henrietta’s then twelve-year-old daughter, the party totaled 750 people, plus oxen, camels, and no less than fourteen elephants. Charlotte’s description is worth quoting in full for the sense it gives of how they traveled and the impressions of India they held.

K anchi Under Colonialism

201

Fourteen elephants were employed to carry our tents, which consisted of two large round tents, six field officers, three captains, and several smaller tents for the cavalry, infantry, etc. by whom we were escorted. Four elephants were employed in carrying a part of our baggage; two were not loaded that had been trained for carrying howdahs, which we sometimes rode when the weather was not too oppressive. We had two camels, which were mostly used for carrying messages, and one hundred bullocks to draw the bandies in which all the rest of our baggage was to be conveyed. Our party consisted of Mamma, Signora Tonelli, my sister and myself, Captain Brown, and Dr. Horsman… We found it necessary to take all our servants, for traveling in India is not like traveling in Europe, as we were obliged to take every article for cooking etc. etc. that could possibly be wanted. This of course occasioned a great number of followers, as all our servants took their wives, and those of higher caste their slaves to prepare their meals, which do not give great trouble, as they only eat boiled rice and curry, the latter of which is made up of meat and vegetables, which they never vary, and only drink water. When every soul was assembled they amounted to 750 persons!! […] which is not in India a very great number, and it is not to be wondered at when all is considered.3

Henrietta and her daughters were among the very first Europeans to record a journey to Kanchi in the colonial period. Their private letters and diaries provide an essential resource for the era. Writings and illustrations by a small number of artists, the first British surveyors of Indian architecture, and a Russian aristocrat further flesh out the picture of Kanchi’s reception across an expanding globe. Illustrations reveal major transformations to Kanchi’s built environment and its social landscape that have taken place over the course of the last two hundred years, and reports by government officials and colonial-era military historians lend weight to, or complicate, the images. Gazetteers furnish important details, such as population size, local industries, and principal communities – when Kanchi was constituted a municipality in 1866, it had a population nearing 50,000 and was very much a Hindu temple-town. 4 In this chapter, I focus on the colonial era because the European visitors to Kanchi were interested in the structures and histories that the first three 3 Clive, Journal of Lady Charlotte Florentia Clive (1787-1866), Copied by Professional Calligrapher, W.H. Ramsay, 37-38. 4 Sir William Stevenson Meyer, ‘Conjeeveram Tāluk and Conjeeveram Town’, in The Imperial Gazetteer of India: Vol. X, Central Provinces to Coompta (Oxford: Clarendon Press, 1908); Madras District Gazetteers: Statistical Appendix for Chingleput District (Madras: Superintendent, Government Press, 1915).

202 

Construc ting K anchi

chapters of this book addressed. Their records lay the groundwork for the construction of Kanchi’s historic past. Colonial-era sources on Kanchi are made all the more valuable by their relative scarcity. Unlike northern India, which was well traveled and illustrated by European artists, Tamil Nadu remained obscure. Of the few artists who explored the region, most went directly to the British strongholds of Tiruchirappalli and Thanjavur (known to Europeans as Trichy and Tanjore). The more ambitious among them ventured to Māmallapuram, which was then a sparsely populated island that was separated from the mainland by a sizeable salt marsh.5 Māmallapuram’s rich archaeological heritage made it a particularly attractive setting for European artists in search of picturesque views. There, the monuments in states of partial ruination, half-buried in sand, set in quiet landscapes backed by rolling seas, enabled European artists to aestheticize the all-too-often unfamiliar Indian landscape through familiar rubrics of natural beauty and wonder amid decay.6 This chapter examines these often-competing narratives and situates them within the political and cultural context of early-colonial South India. Within the larger theme of the city as a changing religious landscape, the chapter first addresses physical changes to the city, such as the transformations to temples brought about by military use. It then turns to the aesthetic theories that determined what it is from the past that would be preserved and the concomitant shift from a picturesque orientation in artists’ depictions of temples to a more scientific one during the course of the nineteenth century. It also examines some of the negotiations that ensued between worshippers and archeologists or company officials, as well as the growing tensions between the worlds of living temples and archeological monuments. Despite the rich firsthand accounts by artists and travelers including James Wathen, William Daniell, Anna Tonelli, and Charlotte and Henrietta Clive, an insidious line of narrative that expressed decline from a golden past to a dilapidated present became prevalent in colonial discourse about India and about culture, art, politics, and heritage more broadly. No narrative is without its prejudices – even the most generous accounts use words like ‘grotesque’ to describe the beautiful divine images encountered along the 5 Fergusson, Picturesque Illustrations of Ancient Architecture in Hindostan, by James Fergusson, 57; and see my Chapter Three. 6 The picturesque – literally the state of being like a picture – resists a singular definition. Nevertheless, across the art of the British empire there were certain common tropes and elements, such as ‘roughness, decay, wildness, and unexpected contrasts.’ Romita Ray, Under the Banyan Tree: Relocating the Picturesque in British India (New Haven: Yale University Press, 2013), 4.

K anchi Under Colonialism

203

way, yet the insistence on decline was particularly troublesome. For India, this line of thought held at its core a belief in a distinct hierarchy of religions. In a vision of a utopian pre-Hindu past, Buddhism and then Jainism had been dominant. The perceived emphasis on a single divine preceptor, as well as a certain level of constancy in these two religions, made them less threatening to European audiences than the vast and capricious pantheon of Hindu gods. In a colonial manual of Madras, published in 1879, the author reports that Buddhism, Jainism, and Vaishnavism were ‘nothing more than three descending stages of one superstition.’7 Of these, ‘Buddhism was the most ancient and the purest form, to which succeeded Jainism, similar in origin, but cloaked with local extravagance.’8 Vaishnavism was believed to have emerged from Jainism being ‘altered in form to suit modern ideas in India, so as to compete with the established Shivite [sic.] faith.’9 Also in the nineteenth century, Henry Maine and Karl Marx would theorize that Indian society was timeless and stagnant, unresponsive to broad regional changes.10 In Kanchi, the Pallava temples in particular conformed to and helped reify the narrative of decline because they had fallen out of use long before Europeans arrived in the city. As royal edifices, by the year one thousand, the Pallava temples stood testament to a regime no longer in power. Throughout the Chola period and the centuries thereafter, ritual concentrated in temples that grew through the support of the local populations. Temples such as Kāmākṣī Ammaṉ, Ekāmbaranātha, Kacchapeśvara, Varadarāja Perumāḷ, and Aṣṭabhuja Perumāḷ continuously received endowments for maintenance and expansion, leading their earlier portions to become obscured by more recent structures and layers of plaster and paint. Photographs taken as early as 1868 and during more extensive surveys around 1900 depict Kanchi’s Pallava temples nearly exclusively. Other temples to be included are the less active parts of Ekāmbaranātha and Varadarāja Perumāḷ, such as the thousand-pillared halls where foreigners were permitted access (Ill. 77). The photographic medium lent an aura of truth that resonated beyond the words of artist-travelers. Colonial photography endeavored to strip away all ‘accretions’ to the ‘original’ or 7 Crole, The Chingleput, Late Madras, District: A Manual Compiled under the Orders of the Madras Government, by Charles Stewart Crole, of the Madras Civil Service, 26. 8 Ibid. 9 Ibid. 10 Nicholas Dirks, ‘Political Authority and Structural Change in Early South Indian History.’

204 

Construc ting K anchi

Illustration 77 Henry Dixon, View Looking Towards an Unidentified Mandapa, Probably at Kanchi (Thousand-Pillared Hall, Ekāmbaranātha Temple, Kanchi), 1868, photographic print, 20.2 × 27.3 cm, British Library, Photo 1000/26(2559)

‘authentic’ building, including people, ritual paraphernalia, and landscape features.11 By omitting information that would disclose the subject’s vitality, photographs of the Pallava temples from the early surveys of Kanchi would lead us to believe Kanchi was in a serious state of disrepair and desolation. However, travelers who took interest in the city as a whole record a very different Kanchi – a vibrant place that teemed with music, dance, and the bustle of urban life. 11 I refer the reader to a growing body of literature on photography and the colonial archive, see for example Christopher Pinney, Camera Indica (Chicago: University of Chicago Press, 1988); Vidya Dehejia and Charles Allen, India through the Lens: Photography 1840-1911 (Washington, DC: Freer Gallery of Art and Arthur M. Sackler Gallery, Smithsonian Institution in association with Mapin Publishing, Ahmedabad, 2000); Maria Antonella Pelizzari, ed. Traces of India: Photography, Architecture, and the Politics of Representation, 1850-1900 (Montreal and New Haven: Canadian Centre for Architecture and Yale Center for British Art, 2003); Guha-Thakurta, Monuments, Objects, Histories; Zahid R. Chaudhary, Afterimage of Empire: Photography in Nineteenth-Century India (Minneapolis: University of Minnesota Press, 2012); Deepali Dewan and Deborah Hutton, Raja Deen Dayal: Artist-Photographer in 19th-Century India (Ahmedabad: Mapin Publishing, 2013).

K anchi Under Colonialism

205

What Happened in Kanchi while those Towering Gateways Arose? Tamil Nadu’s history between the fourteenth and eighteenth century is only beginning to be given its due scholarly attention, and no one has yet focused specifically on Kanchi.12 The following brief discussion – and this book as a whole – should serve as a point of departure for a greater concentration on Kanchi in future studies of this period, and it is my hope that a scholar already versed in the shifting tides of the Vijayanagara Empire (ca. 1336-1646 CE) will one day make this vital contribution.13 Vijayanagara rulers from the western Deccan sponsored the construction of the outer layers of many of the vast, iconic temple complexes that attract visitors to Kanchi today. The famed Tuluva king of Vijayanagara, Kāyāvarohaṇa (ca. 1509-1529 CE), is credited with constructing the southern gopura (‘gateway’) in the Ekāmbaranātha’s fourth enclosure and the Kacchapeśvara’s north gopura, among many other portions of buildings in the city.14 Structural changes to temples during the Vijayanagara period follow certain widespread patterns that can readily be seen in Kanchi. At the exterior, builders often reoriented the temple complex to facilitate direct access from the city, and they constructed an outer prākāra (‘enclosure wall’) with a towering gopura to announce the entrance. They also added an array of shrines and maṇḍapas (‘pavilions’) into the outer prākāras, including shrines for the āḻvārs (Vaishnava saints) in Vishnu temples and a festive vasanta maṇḍapa in Shiva temples. In larger complexes, the renovations might also include the construction of a thousand-pillared hall. 12 Scholars working on this period include Crispin Branfoot, Anna Seastrand, Archana Venkatesan, Elaine Fisher, V. Narayana Rao, David Shulman, and Sanjay Subramanyam. See works including Velcheru Narayana Rao, David Shulman, and Sanjay Subrahmanyam, Symbols of substance, court and state in Nāyaka period Tamilnadu (Delhi: Oxford University Press, 1998); Velcheru Narayana Rao, David Shulman, and Sanjay Subrahmanyam, Textures of time: writing history in South India, 1600-1800 (Delhi: Permanent Black, 2006); Crispin Branfoot, ‘Processions and Presence: Bronze Sculptures from the Temples of Southern India’, Arts of Asia 36, no. 6 (2006); Branfoot, Gods on the move: architecture and ritual in the South Indian temple; Branfoot, ‘Remaking the past: Tamil sacred landscape and temple renovations’; Anna Lise Seastrand, ‘Praise, Politics, and Language: South Indian Murals, 1500-1800’ (PhD diss., Columbia, 2013); Elaine M. Fisher, Hindu Pluralism: Religion and the Public Sphere in Early Modern South India (California: University of California Press, 2017). 13 Studies on Vijayanagara are not lacking. John Fritz and George Michell have done the most extensive, in-depth work on Vijayanagara itself. See their numerous publications and the website of the Vijayanagara Research Project: http://www.vijayanagara.org/ 14 For a biography of this famous king, see Srinivas Reddy, Raya: Krishnadevaraya of Vijayanagara (New Delhi: Juggernaut Books, 2020).

206 

Construc ting K anchi

Renovations often left the vimāna (‘main shrine’) and ardhamaṇḍapa (‘vestibule’) relatively untouched, but added an inner prākāra with a roof and an elevated cloister supported by pillars to enclose the original building. Sometimes they attached an additional maṇḍapa to the older body of the temple as well. Stylistic and material differences between Chola- and Vijayanagara-era constructions visibly distinguish parts of buildings. For example, a smiling, squatting lion was a characteristic motif repeated again and again in the Vijayanagara period. In Kanchi, a lighter colored granite was favored over the darker stone of the Chola period. Numerous temples small and large received these kinds of renovations. As in earlier centuries, inscriptions document only a limited number of structural developments, but they reveal political processes and methods of governance that were in place in the Vijayanagara period. C.R. Srinivasan has usefully compiled Kanchi’s epigraphic reports from these centuries and organized them chronologically by king.15 In her study of religious dynamics, Elaine Fisher references a fourteenth-century inscription on the Varadarāja Perumāḷ temple in which Kāyāvarohaṇa’s successor, Acyutadevarāja (ca. 1529-1542 CE), ruling from afar through his vassal Sāḷuva Nāyaka, makes an edict intended to equalize the endowments to the Varadarāja Perumāḷ and Ekāmbaranātha temples.16 Fisher discusses the rise of a complex form of sectarianism during this period, and the growing importance of the Śrīvaiṣṇavas and their monastic institutions, one of which was centered in Kanchi.17 These negotiations would continue to play out through the colonial era and into the present day.18 At the local level, the fourteenth through eighteenth century was an era of great religious literary production. Across Tamil Nadu, Sanskrit Māhātmyas and Tamil Sthalapurāṇas (also called Talapurāṇam) were written, often with patronage from prosperous local sponsors. These texts comprised legendary temple histories that extol the virtues of specific temples, as well as their locations. They also create larger temple networks through shared deity forms, interpolated narratives that reach back to earlier Purāṇas, and the use of recognizable literary tropes.19 Scholars are currently exploring no less than five Kanchi Māhātmyas and Sthalapurāṇas that were written during this period.20 15 Srinivasan, Kanchipuram through the Ages. 16 Fisher, Hindu Pluralism: Religion and the Public Sphere in Early Modern South India, 8-9. 17 Ibid. 18 Hüsken, ‘Gods and Goddesses in the Ritual Landscape of Seventeenth and Eighteenth-Century Kāñcipuram.’ 19 Ute Hüsken, Jonas Buchholz, and Malini Ambach convened a workshop on Kanchi-related texts (Pondicherry 2020) and launched the project, ‘Networks of Temples and Networks of Texts in South India.’ 20 For references to the five texts, see Chapter Two, fn. 8.

K anchi Under Colonialism

207

Diverse sources from the seventeenth and eighteenth century give us brief insights into Kanchi’s social landscape and status as a South Indian city. The Jesuit records of 1683 tell of the ‘celebre cidade de Canjaburão’ [‘the famous city of Kanchipuram’].21 The papers of Anandarangappoullé, a mutaliyār (‘Courtier’, Persian ‘diwan/divan’) to the French Company from 1747 to 1756, give a glimpse of Kanchi during that pivotal decade leading up to the Battle of Plassey (1757) in which Robert Clive played an instrumental role in the formative phases of the British Empire.22 Anandarangappoullé encountered a great many people from a wide range of social strata in his role as intermediary between the French and the local communities. On 2 November 1746, a Kanchi resident reported that the army of Mafouz Khan had entered the city the previous week and ordered the urban landholders to flee. This morning Vengadaçalaaya arrived. His brother-in-law is from Kanchipuram. He reported that on Wednesday afternoon, the army of Mafouz Khan, composed of four hundred horsemen, a million infantry, four or five hundred cannons, and seven, eight, or nine elephants, arrived outside of Kanchipuram. Then, he descended into the heat of the chettys and sent an order to the paléagars and gilédars to gather their people and move to Mylapore. He said that tomorrow, leaving this place, he would descend upon Madras.23

A footnote explains that paléagars (Tamil ‘pāḷaiyakkārar’) are local chiefs or feudatories (French ‘hommes du camp’, or ‘hommes du district’) and that gilédars (Persian ‘zilahdar’) have a higher status as chiefs of districts appointed directly by the sovereign. Chettys (‘Chettiars’) are a Tamil caste of moneylenders and merchants. Although the record is of an invasion, it reveals the presence of significant communities of prosperous, urban, landholding elites residing in the city.

21 The Madurai Mission Annual Letter of 1683, drafted by João de Brito on 9.5.1684, in Archivum Romanum Societatis Iesu, Goa 54, f. 436v. Warm thanks to Giulia Nardini for providing this interesting reference. 22 Les Français dans l’Inde: Dupleix et Labourdonnais, Extraits du Journal d’Anandarangappoulé, 1736-48, trans. Julien Vinson (Paris: Ernest Leroux, 1894), LXXIII. 23 ‘Ce matin est arrivé Vengadaçalaaya. Son beau-frère est arrivé de Kāndjipuram. Il rapporte que mercredi, dans l’après-midi, l’armée de Mafouzkhan, composée de quatre cents cavaliers, un millier de fantassins, quatre ou cinq canons, et sept, huit ou dix éléphants, est arrivée au-delà de Kāndjipuram. En outré, il est descendu à la chauderie des chettys et a envoyé l’ordre aux paléagars et aux gilédars de réunir leurs hommes et de se porter sur Méliapour. Il dit que demain, quittant cet endroit, il descendra sur Madras…’ ibid.

208 

Construc ting K anchi

Embattled Territory By the middle of the eighteenth century, Kanchi had found itself in a precarious position at the crossroads of embattled French and British forces as they competed for territory and influence in southern India. During the course of the 1750s, Kanchi changed hands repeatedly as Lord Clive carried out his notorious rivalry with the celebrated French Governor-General Dupleix.24 The events are detailed in Orme’s A History of the Military Transactions of the British Nation in Indostan (published in 1799) – the publication Henrietta Clive selected as a guidebook for her tour through South India.25 The action centered on the Ekāmbaranātha temple, where the British had established a fort. Until this time, Clive had been a secondary figure. But in the 1750s, as Martineaux so aptly puts it, ‘his star was rising.’26 Robert Clive in his Fortified Temple Both British and French troops in the colonial period made use of temples for a variety of purposes, none of which were religious. While they used some structures as military storehouses, such as a shrine at Vellore recorded by Henrietta, elsewhere they more violently appropriated entire complexes by converting them into fortresses.27 From the colonizers’ perspective, the presence of multiple protective enclosures, towers for surveillance, and rooftops that could support cannons made temple complexes ideal for use in military defense. Originally, the enclosure walls had multiple functions. On a spiritual level, they were intended both to protect the gods inside the sanctums and to prevent their awesome power from overwhelming the cosmos.28 On a practical level, walls were a kind of fortification – they 24 Robert Orme, A History of the Military Transactions of the British Nation in Indostan: From the Year MDCCXLV, 2 vols., vol. I-II (London: Printed for John Nourse, Bookseller in Ordinary to his Majesty, 1799); Alfred Martineaux, Dupleix et l’Inde Française: Vol. 3, 1749-1754, 3 vols. (Paris: Société d’Éditions Géographiques, Maritimes et Coloniales, 1927). See especially Martineaux, Chapter 11, ‘Opérations autour d’Arcate, Conjivaram, Chinglepet, et Tirnamallé’, p. 261-292. 25 Shields, Birds of Passage: Henrietta Clive’s Travels in South India, 1798-1801, 227. 26 ‘C’est alors que parut un de ces homes dont la mission parait être de détourner le cours normal des événements et d’ouvrir aux peoples des destinées insoupçonnées. Cet homme était Robert Clive.’ Dupleix et l’Inde Française: Vol. 3, 1749-1754, 262-263. 27 Shields, Birds of Passage: Henrietta Clive’s Travels in South India, 1798-1801, 115; Jean Deloche, Ancient Fortifications of the Tamil Country as Recorded in Eighteenth-Century French Plans (Pondicherry: École Française d’Extrême-Orient, 2013). 28 David Shulman, Tamil Temple Myths: Sacrifice and Divine Marriage in the South Indian Saiva Tradition (Princeton: Princeton University Press, 1980).

K anchi Under Colonialism

209

Illustration 78 Anonymous, Rough Sketch of the Fortified Pagoda of Great Cangivaram, 1872, pen and ink on paper, Archives nationales d’Outre-mer, Aix-en- Provence, FR ANOM 25DFC 260B, Image courtesy: EFEO/Jean Deloche

guarded the temple’s wealth and the treasuries of bronze sculptures, gold, and jewelry it held in store.29 The British practice of using temples as fortresses was widespread. In the Cuddalore district, the Pallava-era Vīraṭṭāṉeśvara temple at Tiruvatikai was fortified, or at least there were plans for its fortification.30 At Kāvērippākkam and Arcot near Kanchi, fortresses were erected, or existing forts were reinforced. The same was true of Chola-era sites in areas farther south, including the massive temple complexes at Chidambaram, Srirangam, and, most famously, Tiruchirappalli.31 The British also converted the East Javanese city of Majapahit into a fortress in the nineteenth century.32 29 Veluthat historicizes the role of temples as treasuries and the concomitant architectural developments. Veluthat, The Early Medieval in South India, 6. 30 Deloche, Ancient Fortifications of the Tamil Country as Recorded in Eighteenth-Century French Plans, 19. 31 Ibid. 32 My sincere thanks to Hadi Sidomulyo for sharing his encyclopedic knowledge of Majapahit (August 2016).

210 

Construc ting K anchi

Plans for the Ekāmbaranātha temple’s transformation into a military fortress are visualized in a map drawn by a French officer in 1752 (Ill. 78). They are also described in a letter from Véry de Saint Romain to Dupleix. It would be necessary to totally knock down the first wall of the temple and entrench ourselves within the second wall; the debris from the first [wall] will serve to make four good, small bastions, at little cost; all around the second wall is a nice gallery, whose terrace would serve to place the muskets; the third wall towers above the second and also the same gallery. With a little bit of gunpowder we can easily knock down the whole of the first wall.33

Indian treatises on dwelling, such as the Mayamata, refer to enclosure walls as numbered from inner to outer – from the vantage point of the god inside the sanctum.34 By contrast, Europeans typically counted from their own perspective, from the outside in towards center. Véry de Saint Romain’s ‘first wall’ is therefore likely the Ekāmbaranātha’s outermost (fifth) enclosure. The fortified area would be the wall that extends from the now-closed southern gopura and encloses the ruined courtyard around the vasanta maṇḍapa and the Naṭarāja shrine. The ‘terrace’ upon which to place the guns might be this prākāra’s clerestory, which is no longer extant but is visible in archival photographs (Ill. 79). Now closed to the public, the portion of the Ekāmbaranātha seen in the photographs reveals extensive damage and renovations that remain incomplete to this day. Although his troops were few in number, Lord Clive was able to find protection within the three outer enclosures of the Ekāmbaranātha temple when he first arrived in Kanchi.35 Orme’s account of the siege is evocative, The army waited some days for two 18 pounders, which were coming from Madras; and as soon as they arrived began to batter in breach at the distance of 200 yards. The enemy had no cannon, but f ired very 33 ‘Il faudrait totalement abattre le premier mur de la pagode et se fortifier dans la seconde enceinte; les debris de la première serviroient à faire quatre bons petits bastions, à peu de frais; la seconde enceinte a tout autour une belle gallerie dont la terrasse serviroit à mettre de la mousquetterie; la troisième enceinte domine la seconde et la même gallerie. Avec un peu de poudre l’on pourra aisément faire sauter tout le premier mur.’ Letter from Véry de Saint Romain to Dupleix, 1752, BN, NAF 9152, 9153; reproduced in Deloche, Ancient Fortifications of the Tamil Country as Recorded in Eighteenth-Century French Plans, 13. 34 Dagens, Mayamatam: Treatise of Housing, Architecture, and Iconography. 35 ‘La pagode avec ses trios enceintes successives, lui offrait un abri assure.’ Martineaux, Dupleix et l’Inde Française: Vol. 3, 1749-1754, 269.

K anchi Under Colonialism

211

Illustration 79 Archaeological Survey of India, The Amman shrine, Ekambreswaraswami Temple, Conjeevaram, 1897, photographic print, 21.1 × 25.5 cm, British Library, Photo 1008/3(323)

smartly with their musketry, which killed several men at the battery, and lieutenant Bulkley, reconnoitering the pagoda over a garden-wall in company with captain Clive, was shot through the head close by his side. The wall resisted three days before it began to give way, when the garrison, conscious of their demerits, and dreading the just resentment of the English, abandoned the pagoda in the night, but left behind the two prisoners. After ruining the defenses of Conjeveram, Captain Clive sent 200 Europeans and 500 Sepoys to Arcot.36

In January of 1752, immediately after Dupleix’s first encounter with Clive in Kanchi, the Governor-General decided to move his base of operation. 36 Orme, A History of the Military Transactions of the British Nation in Indostan: From the Year MCDDXLV Vol. I-II, 199-200. The portion of the temple that was damaged in the initial encounter between the French and English at the end of 1751 was very likely the east wall of the outermost prākāra enclosure (see below).

212 

Construc ting K anchi

He was determined to divert the English away from the coast before their upcoming expedition to Tiruchirappalli, and he decided that Kanchi was too far away from the seacoast. Indeed, it was its inland location that had protected the city for millennia before. Certain that the French were in the vicinity of Arcot, in March 1752 Clive marched on Kanchi, where he returned the fort to British control. Like so many visitors to the city, the Captain did not rest, but continued onwards to Arcot.37 Clive and Dupleix’s presence in Kanchi have been all but erased from local memory. It is possible, however, that the legacy of their interventions lives on in the built environment. Orme and the French sources both describe damage to a prākāra.38 The area affected by the siege was probably the eastern wall of the outermost enclosure – a wall that would be rebuilt at the end of the eighteenth century.39 Kanchi in the Mysore Wars In the decades that followed, conflicts between the British and the French would continue to escalate. At the conclusion of the Seven Years War in 1763, the Treaty of Paris established colonial territories in India for Britain and France. 40 Then came the Mysore Wars, which were fought in four distinct phases between 1767 and 1799. The Mysore Wars were led by the British army against the sultanate of Haidar Ali and his infamous son, Tipu. The Mysore sultanate turned against the East India Company in 1779, when the British wrested the French port of Mahe from Haidar Ali’s dominion. 41 Mysore was allied in battle with the French, who supplied them with weapons, in part 37 ‘Sans s’attarder à Vendalore, Clive se mit donc en marche sur Conjivaram, dont il importait de s’assurer à nouveau la possession. La garnison de la pagode se rendit à la première sommation et, sans prendre de repos, Clive marcha sur Arcot.’ Martineaux, Dupleix et l’Inde Française: Vol. 3, 1749-1754, 274. According to Malreaux’s fn.2, Vendalore is 22 miles south of Madras and 11 miles from St. Thomas Mount. 38 ‘Mais soit que La Volonté ait été, dit-on, trahi par un notable de la place, soit qu’il ait pris peur de quelques coups de canon qui firent une breche dans les murs de la pagode, il évacua la ville pendant la niut au bout de trois ou quatre jours, et la laissa sans résistance aux mains de Clive (29 ou 30 décembre [1751]). Brenier, arrivé quarante-heures plus tard, en reprit possession le 3 janvier; les Anglais l’avaient spontanément abandonée après avoir démantalé une partie de la première enceinte.’ Ibid., 269-270. 39 See below and fn. 73 in this chapter. 40 Simmons points out the importance of the Treaty of Paris for the subsequent fate of South India. Caleb Simmons, Devotional Sovereignty: Kingship and Religion in India (New York: Oxford University Press, 2020), 9. 41 Kaushik Roy, War, Culture and Society in Early Modern South Asia (1740-1849) (New York: Routledge, 2011), 83.

K anchi Under Colonialism

213

through Mahe. 42 The Mysore rulers were known to the British as fierce and tyrannical, and Tipu was famed across the seas for his obsession with tigers. The conflicts played a critical role in consolidating British power in the subcontinent, and in constructing new notions of sovereignty in South India. 43 Two crucial battles were fought just ten kilometers north of Kanchi, in the village of Puḷḷalūr (called variously ‘Polliloore’ or ‘Pollilur’ in colonial sources). This was not the first time Puḷḷalūr’s open fields had served as a major battleground. The Kasakudi copper plates tell of a battle fought at Puḷḷalūr in the seventh century, during which the Pallava king Mahendravarman I defeated a consortium of opposing powers. 44 A set of maps preserved at the British Library records the plans for the two colonial-era battles (Ill. 80). The first took place on 10 September 1780. The Mysore army was led by Tipu Sultan, and the British by Colonel William Baillie. Diagrams of the organization of troops in the days leading up to the clash show the British camps progressing along the road from Kanchi to Puḷḷalūr. The Ekāmbaranātha temple is marked on Kanchi’s northern border. At the center is Baillie’s position, surrounded by his army in a protective square formation. Tipu nevertheless captured Baillie and achieved a major victory over the British. This defeat had financial and moral consequences for the British that were disproportionate to the actual losses they suffered in battle. 45 Tipu celebrated his triumphs in vast mural paintings that are installed at his Dariya Daulat palace at Śrīraṅgapaṭṭaṇa (known to the British as Seringapatam). The murals depicting the Puḷḷalūr victory stretch across the summer palace’s western exterior wall, sheltered by a pillared verandah made of teakwood. 46 In the southern half of the mural, Baillie is seen carried in his palanquin at the center of the battle square. The Sultan’s army 42 Ibid. On dynamics of trade during this period, see Liza Oliver, Art, Trade, and Imperialism in Early Modern French India (Amsterdam: Amsterdam University Press, 2020). 43 Caleb Simmons, Devotional Sovereignty: Kingship and Religion in India. 44 SII Vol. 2, Part 3, p. 346-353 (text), p. 42-61 (discussion and translation), no. 73. Although the ‘chief enemies’ that Mahendravarman defeated were originally taken to refer to the Chalukya ruler Pulakeśin II, Mahalingam suggests it was instead Nallaṭi, who sought to re-establish the power of the early Cholas, but to no effect. According to Mahalingam, Pulakeśin II did not enter the Kanchi area until the reign of Mahendravarman’s successor, Narasiṃhavarman I. Mahalingam, Inscriptions of the Pallavas, p. xlvii. The area north of Kanchi was an embattled territory throughout history. The Rashtrakutas and Cholas fought another legendary conflict at Takkōlam in the mid-tenth century. 45 Irschick, Dialogue and History: Constructing South India, 1795-1895, 16-17. 46 For detailed discussion of the murals, see Simmons, Devotional Sovereignty: Kingship and Religion in India, 79-87.

214 

Construc ting K anchi

Illustration 80 Battle of Puḷḷalūr 1780, in Four Plans of engagements between the British forces and those of Hyder Ally and Tippo Sahib [at] Congeveram, Sholangur, Vellore [and] Veracundaloor [in the Presidency of Madras], 1780-1782, British Library, MAPS 54570.(2.)

approaches from the northern half and attacks the square from all sides, with horses galloping and their riders brandishing swords. The orderly British lines diminish quickly, leaving soldiers struggling to load cannons as the Mysore army rushes by. The protective square around Baillie appears meager in defense, as if to satirize the British strategies. Massive in scale and brilliantly colored, the murals would have overwhelmed any visitor who entered the otherwise tranquil garden. So powerful were they, that Tipu ordered the paintings whitewashed as Lord Cornwallis approached in 1792, for fear of further inciting British rage. 47

47 Colonial-era sources including Valentia (1803) and Walter (Cambell) Campbell (1833) mention other paintings at the palace that caricatured the British as drunkards and rapists. The Puḷḷalūr murals were partially restored under Col. Wellesley, and subsequently repainted in 1854 under Lord Dalhousie. Anne Buddle, Tigers Round the Throne: The Court of Tipu Sultan (1750-1799) (London: Zamana Gallery, 1990), 48.

K anchi Under Colonialism

215

The second battle at Puḷḷalūr took place just over a year later, on 27 August 1781. The Mysore forces were led by Haidar Ali, and the British by Sir Eyre Coote. This time the British were more successful. While Haidar Ali’s losses amounted to more than two thousand men, Coote lost less than five hundred. 48 Two of the fallen British soldiers are memorialized on a monument that still stands amid the placid paddy fields that today comprise the village of Puḷḷalūr. 49 After the battle, Haidar Ali’s troops pressed onwards towards Kanchi, while Coote’s retreated to Tiruppattūr to resupply.50 The final battle of the Mysore Wars culminated in a decisive victory for the British in 1799, most famously depicted in a panorama by Robert Ker Porter (1800) and years later in a portrait-cum-history painting by Sir David Wilkie (1839). Despite a report to the contrary by Kanchi’s collector, Lionel Place, it does not seem that Haidar Ali or Tipu Sultan themselves ever successfully entered Kanchi, though they may well have known of Kanchi’s important temples. Caleb Simmons discusses mural paintings in the citramaṇḍapa  (‘painted hall’) of the Veṅkaṭaramaṇasvāmi temple in Mysore, built during the reign of Tipu’s successor, Kṛṣṇarāja III (r. 1799-1868 CE), that include Kanchi among other important South Indian sacred sites.51 A section of the eastern wall maps a version of Kanchi – the Varadarāja Perumāḷ and a procession of the deity appear at left, with Sarva Tīrtham and the Ekāmbaranātha temple above, and Kāmākṣī Ammaṉ in the middle.52 The Palar River flows through the composition, and houses with tiled roofs are interspersed among the temples. The murals show that these three temples and the large ablutions tank had become Kanchi’s primary religious destinations by the early nineteenth century. In his 1799 Report on the Jagir, Lionel Place writes that the Varadarāja Perumāḷ temple was significantly damaged when the Mysore army searched for buried treasure inside.53 After fires were set and floors torn apart, ‘the whole interior was nothing more than what had escaped the dilapidations

48 Roy, War, Culture and Society in Early Modern South Asia (1740-1849), 85. 49 Buddle mentions that Pollilur is ‘often confused with nearby Pullalur.’ Anne Buddle, The Tiger and the Thistle (Edinburgh: National Gallery of Scotland, 1999). The field with the monument that memorializes the battle is, however, actually in Puḷḷalūr, and ‘Pollilur’ is simply the Anglicized spelling. 50 Roy, War, Culture and Society in Early Modern South Asia (1740-1849), 85. 51 Simmons, Devotional Sovereignty: Kingship and Religion in India, 214. 52 Ibid., p. 219, Fig. 7.4b. My thanks to Caleb Simmons for sharing photographs. 53 Irschick, Dialogue and History: Constructing South India, 1795-1895, 80.

216 

Construc ting K anchi

of time and neglect or the ravages of a devastating Enemy.’54 In contrast to the many records of sieges within the Ekāmbaranātha temple, Place’s Report stands as the only reference I have found to the role of the Varadarāja Perumāḷ temple in the Mysore Wars. While other Kanchi temples in fact may have been damaged during the course of conflict, it is possible that Place singled out the Varadarāja Perumāḷ temple and either exaggerated or invented its destruction in order to justify his own aims in Kanchi. As collector, Place supported Hindu religious life because he believed this would make the population ‘happier’, and therefore more productive, useful, and cooperative.55 Responding to reports that much of the local population had fled the region during the Mysore Wars, he sought to draw people back to their original residences. As Eugene Irschick has shown, Place’s efforts were part of a larger colonial project that aimed at sedentarizing the Tamil population.56 Between 1795 and 1798, Place funneled money into the Varadarāja Perumāḷ temple in order to ‘revitalize’ its rituals. Although he supported both of Kanchi’s largest temples, he gave greater attention to Varadarāja Perumāḷ, even gifting the towering gilt lampstand to the temple himself.57 At one point he requested that Colonel Baird supply 30 sepoys to accompany the Garuda Sevai procession.58 The collector’s extensive involvement in this particular festival led the missionary Alexander Duff to accuse him of having been the greatest Christian champion of ‘idolatry.’59 Besides Place’s own testimony, there is nothing to suggest that the temple was in need of fiscal support in the late eighteenth century. In fact, the revitalization he implemented led to accelerated costs and crowds that inevitably outweighed the temple’s capacity. In 1801, under Place’s successor, overcrowding during Garuda Sevai caused an unfortunate incident in which devotees were trampled. Perhaps as a result, British support of temple ritual was discontinued in 1818.60 Kanchi’s legacy as a place associated with the Mysore Wars invited attention from European audiences. Yet it did not receive the same level of visitorship as other sites of antiquarian interest. Most European travelers who went to South India followed the pathway taken by the British military campaigns, beginning with arrival by sea in Madras. The travelers then 54 Ibid. 55 Irschick, Dialogue and History: Constructing South India, 1795-1895, 79. 56 Ibid. 57 Lady Clive remarked on the lampstand gift. Clive, Journal of Lady Charlotte Florentia Clive (1787-1866), Copied by Professional Calligrapher, W.H. Ramsay, 51. 58 Irschick, Dialogue and History: Constructing South India, 1795-1895, 79 and 83. 59 Duff visited Kanchi in 1849 and found the city to be a thriving Hindu center. Ibid., 85. 60 Ibid., 84-85.

K anchi Under Colonialism

217

Illustration 81 Thomas Colman Dibdin after James Fergusson, Sketch Map of India, 1847, lithograph, 38.2 × 29.4 cm, British Library, X472(35)

moved west overland, crossing northern Tamil Nadu to Bangalore, and then turning southwest. After proceeding to Mysore, they continued south to Coimbatore and the surrounding sites. Tiruchirappalli lay directly east of Coimbatore and was then followed by Thanjavur and finally seaside Tranquebar. From there it was directly northward along the coast to Madras, with Portuguese Porto Novo and French Pondicherry providing convenient

218 

Construc ting K anchi

stops along the way. This itinerary left many opportunities for interesting diversions, as it passed through concentrated areas of historical and colonial-era significance. Kanchi was situated directly along the road that led from Madras to important sites of battle, such as Kāvērippākkam, Arcot, and Vellore. It would have been a logical place for an extended stay. However, colonial visitors who even bothered to stop in Kanchi did not linger long. It may have been the vibrancy of Hindu ritual that they encountered in Kanchi, as well as the lack of accomodations suited to European tastes, that deterred them from more extensive stays.61 Or it may have been something more general about the difficulty of travel in Tamil Nadu. Passages from the writings of the Viscount Valentia, James Wathen, and James Fergusson provide useful insights into why Kanchi was usually omitted from colonial travel itineraries. Not only did the city’s vibrant and soaring temples pale in comparison to Māmallapuram’s deserted ruins for some Europeans, Kanchi and its vicinity simply may have been too challenging to visit. As he passes from Kanchi and heads northwest towards Vellore, the Viscount Valentia makes note of the thinning of villages and the destruction wrought by Tipu Sultan’s forces. He found some relief in the sheltered pavilions (‘maṇḍapas’, ‘choultries’) along the road, which – although in a ruinous condition by his assessment – were often maintained by locals who gave subsistence to passersby. Always interested in the military histories of places, Valentia mentions damage that was inflicted upon the tank at Kāvērippākkam, known to him as the largest tank in southern India, and the fact that many villages in the area had erected compound walls to protect themselves from the Mysore army.62 It seems that the journey immediately west from Kanchi presented an especially significant challenge. Valentia, who was not prone to complain, writes, The heat of the sun was extremely oppressive, the thermometer being at 96 degrees; we were therefore able to go but slowly. Villages are thinly scattered, the jungle is more frequent, and the soil a dry gravelly sand, which being raised by the wind, nearly suffocated me.63 61 There were only two travelers’ bungalows in Conjeeveram Taluk, each with modest accommodations, and none in the city itself. Madras District Gazetteers: Statistical Appendix for Chingleput District, 4. 62 George Viscount Valentia, Voyages and Travels to India, Ceylon, the Red Sea, Abyssinia, and Egypt in the Years of 1802, 1803, 1804, 1805, and 1806, 1 (Text) and 4 (Plates) (London: Rivington, 1811), 396. 63 Ibid.

K anchi Under Colonialism

219

In this episode we find some indication that the difficulties travelers encountered in the South outweighed those in North India. James Wathen also lodged certain dissatisfactions. He began his narration of the journey from Madras to Kanchi with laments of repeated delays. Although he had intended to embark at 6 pm, the party did not set out until three hours later. Once underway, they were delayed a further two hours by a festival procession in which Wathen first encountered the ‘strangely-grotesque figures’ of the portable Hindu gods.64 The following day, Wathen hastened to depart from Kanchi, but found that his bearers were unwilling to leave until the midday heat had subsided. Artists, scholars, and adventurers paid little attention to Kanchi. Henrietta’s visit to Tiruchirappalli spanned more than a week, yet she rested in Kanchi for only one night. The famous uncle-and-nephew artist duo, Thomas and William Daniell, spent just one morning sketching in the city – of the hundreds of illustrations they produced during their South India tour, none of the Kanchi drawings have come to light.65 The Russian Prince Alexis Soltykoff similarly allotted only two days to Kanchi. The city remains absent even from the vast corpus of drawings and writings by the architectural surveyor, James Fergusson. Although he included Kanchi on the map of places he saw, no drawings are known from his visit, and the surviving portion of his diary concludes long before he entered Tamil Nadu (Ill. 81).66 Fergusson provides the most extensive explanation as to why he spent so little time in Tamil Nadu. Lamenting the brevity of his engagement with South Indian temples, he explains that his travels in Tamil Nadu, specifically within the territory of the Madras Presidency, were arduous.67 He received little assistance from the East India Company, he was often hungry, and he had trouble finding bearers to carry his palanquin. At Tiruchirappalli, Fergusson finally hired his own bearers, who carried him all the way back to Madras – a distance of more than 300 kilometers. Despite traveling at night, he found the progress ‘cruelly slow.’ Frustration may account for the absence of sketches to survive from the end of his tour.68 On the map 64 James Wathen, Journal of a Voyage, in 1811 and 1812, to Madras and China; Returning by the Cape of Good Hope and St. Helena (London: J. Nichols, Son, and Bentley, 1814), 54. 65 It is known only from William’s diary that they went to Kanchi and made sketches there. William Daniell, ‘William Daniell’s Diary’, in Walker’s Quarterly: Thomas Daniell, R.A.; William Daniell, R.A., ed. Martin Hardie and Muriel Clayton (1932). 66 James Fergusson, Diary of James Fergusson During Tours in India, When Collecting the Materials for his Picturesque Illustrations of Ancient Architecture in Hindostan: 18 June-5 July 1837 and 26 August 1838-22 March 1839 (British Library, Western Manuscripts Collection, Add MS 35282: 1848). 67 Fergusson, Picturesque Illustrations of Ancient Architecture in Hindostan, by James Fergusson, iv. 68 Ibid.

220 

Construc ting K anchi

that concludes Fergusson’s volume of Ancient Architecture, ‘Conjeveram’ (a colonial spelling of Kanchipuram) marks the farthest point west in northern Tamil Nadu along the red line that tracks the author’s journey. Having begun his trip in Calcutta and concluding it with a departure from Madras, Kanchi must have been among the very last places he visited. It is possible that other artists suffered similar hindrances to Valentia, Wathen, and Fergusson. Stories of such sufferings may have dissuaded additional potential travelers from visiting Tamil Nadu. Henrietta Clive and Prince Alexis Soltykoff mention the hospitality of Mr. Hodson, Lord Elphinstone, and unnamed individuals who assisted them. But they also bemoan the heat, the dust, and the slowness of their progress. It is possible that the overall atmosphere in Tamil Nadu was even less accommodating than they realized, and that if they had traveled extensively in North India before experiencing the South, they would have been more inclined to complain. Lieutenant James Hunter In light of Kanchi’s involvement in the Mysore Wars, it is not surprising that military officers were the first artists to make illustrations of Kanchi under colonialism. With the demands of their service, military artists throughout India had the opportunity to venture to places far removed from typical traveler circuits.69 Lieutenant James Hunter, an amateur draftsman and artilleryman in the British army, spent the final year of his life in Kanchi (1791-1792) while serving under Lord Cornwallis in the Mysore Wars. Lieutenant Hunter’s illustrations are the earliest colonial-era representations of Kanchi and its surrounds, filtered through the picturesque aesthetic that was in growing currency throughout the eighteenth century. During his time in India, Hunter executed 40 sketches. The originals were sent to London and posthumously worked into etched aquatints by artists including Henry Merke, who never visited the subcontinent. An album of Hunter’s illustrations was published in 1804 by Edward Orme, under the title, Picturesque Scenery in the Kingdom of Mysore.70 In A View from the Royal Artillery Encampment, Conjeveram, oblique perspectives framed by trees recall pastoral ruins, tents are pitched in the shadowed verdant landscape, British officers and local residents populate the interstitial spaces at pleasing 69 Ray, Under the Banyan Tree: Relocating the Picturesque in British India, 9. 70 Also called A Brief History of Ancient and Modern India in the records of the British Museum. The complete album is preserved at the British Library; 23 of the unbound pages are at the British Museum; and a selection of the original watercolors have appeared on the London art market and were sold at Sotheby’s and Christie’s between 1969 and 2003.

K anchi Under Colonialism

221

Illustration 82 Henry Merke after James Hunter, A View from the Royal Artillery Encampment, Conjeveram, 1805, aquatint, British Library, X768/3(36)

distributions, and the gopuras of the Ekāmbaranātha temple tower sublimely in the distance (Ill. 82).71 Officers in blue and white British uniforms stand and converse, while Indians wearing lungis sit on the ground, some tending fires. The lone woman in the picture holds a clay pot on her head in a classical pose that is a leitmotif of the colonial picturesque, seen with great frequency in the work of William Hodges.72 Dusky pink light and long shadows suggest an evening scene, the officers back at camp relaxing at the end of the day. The calm, quietness of Hunter’s scene reveals a harmony between the British and the local populace in a time of war with another Indian ruler, yet it also masks the unequal power relations inherent in the image. The Indian people the artist includes do not have the same status as the European officers. Rather, they serve the British as cooks and attendants in the camp. 71 For theorization of the sublime, see Edmund Burke, A Philosophical Enquiry into the Origin of our Ideas of the Sublime and Beautiful (Oxford: Oxford University Press, [1759)]2014). See also Partha Mitter, Much Maligned Monsters: History of European Reactions to Indian Art (Oxford: Clarendon Press, 1977). 72 See for example The Marmalong Bridge, ca.1783, Yale Center for British Art (B1974.3.8) published in Geoff Quilley, ed., William Hodges, 1744-1797: The Art of Exploration (New Haven and London: Yale University Press, 2004), cat. no. 42, p. 140-141.

222 

Construc ting K anchi

The position of the setting sun to the left, and the location of the temple at the drawing’s center, situates the camp distinctly to the southwest of the Ekāmbaranātha temple. The large, south gopura appears at right and the smaller south gopura left. Between them, the main gopura of the eastern entry was visible to Hunter but is now obscured by recent construction. Easily mistaken for a tall tree, the top of yet another gopura peeks out from the tree line at what would be the east side of the outermost prākāra, seen at the mid-right of the page. Although today there is no gopura in this place, inspection of the wall reveals neatly aligned granite blocks peppered with reused sculpture and architectural materials – sure signs that this wall has been transformed. This is the part of the prākāra that was likely damaged in the 1751 conflict between the British and the French, mentioned above. One source tells us that in 1799 a stretch of approximately 30 yards was rebuilt by Chingleput District Collector Mr. Hodson – who would host Henrietta Clive and her entourage the following year – and that the reconstruction was recorded with an inscription.73 The reconstruction constituted a significant change to the temple. It even swallowed an ancient shrine that is supported by now barely detectable Pallava-era, sandstone, lion-based pillars near the northeast corner of the complex. After all the transformations, it remains entirely possible that a low gopura marked an aperture that led through the middle of this eastern wall at the time of Hunter’s visit.

William Daniell’s Most Considerable Temple The Ekāmbaranātha temple would continue to draw attention from foreign visitors to Kanchi, perhaps because of its role in the wars in addition to its imposing scale. Over time, it seems to have become common practice for visitors to seek out or be brought to this temple, to the exclusion of others in the city. Still today, foreigners generally include Ekāmbaranātha on their daylong tour of Kanchi, if they visit any temple besides Kailāsanātha. On 12 April 1792, a month before Lieutenant Hunter’s death, Thomas and William Daniell arrived in Kanchi.74 They spent most of the day making drawings of the Ekāmbaranātha temple, which William Daniell described with his frequently used expression, ‘the most considerable Hindoo temple we

73 M. Gopalakrishnan, History of Sri Ekambaranathar Temple, Kancheepuram, South India (Kanchipuram: Madhan Printers, 1978), 4. For Mr. Hodson, see Shields, Birds of Passage: Henrietta Clive’s Travels in South India, 1798-1801, 109. 74 Daniell, ‘William Daniell’s Diary’, 77.

K anchi Under Colonialism

223

have seen.’75 By far the most celebrated of the artists who visited South India, the Daniells had embarked on their journey the month before, having already completed an extended tour of North India. Their self-proclaimed interests lay less in the aesthetic merit of their artworks than in the veracity of the view, and they utilized a camera obscura, an optical device that aided them in their depictions of architecture in landscape.76 The six-volume publication Oriental Scenery, which resulted from their India tours, was a magnum opus. As Douglas Fordham has written, Oriental Scenery departed from earlier colonial representations of India and instead merged ‘antiquarianism, landscape, and architectural history into a highly detailed, wide-ranging vision.’77 Upon returning to England in 1793, the Daniells began transforming their drawings into full paintings and aquatints, which, unlike Lieutenant Hunter and other artists, they produced themselves. Many of the artists who translated firsthand drawings into prints had never visited India, and thus they had to rely on other printed media in circulation to supply missing information. By contrast, the Daniells were able to complete their pictures on the basis of their own memories, with presumably fewer and less dramatic distortions. Although the camera obscura imparted a level of precision in scale and organization of space, it was also conducive to the creation of outlines rather than fleshed-out forms. This can be seen in the original drawings made on-site. The use of outlines nevertheless meant that a significant amount of detail had to be added long after the artists had departed India. The Daniells’ trip to Kanchi is well recorded in William’s diary in which the artist provides details about the places they visited and the drawings they produced. However, it is most unfortunate that none of the Kanchi drawings can be found. The diary states that the Daniells drew the Ekāmbaranātha temple from multiple vantage points.78 While Thomas sketched the towering southern gopura from the street side, William positioned himself to the north. The Daniells visited Kanchi near the very beginning of their South India tour, when they were more concerned with picturesque scenery than with monuments.79 At that point, they also would have found themselves relatively unfamiliar with Tamil temple architecture. If the sketches even survived the journey back to England, it is possible they were considered 75 Ibid. 76 Mildred Archer, Early Views of India: The Picturesque Journeys of Thomas and William Daniell, 1786-1794 (London: Thames and Hudson, 1980), 226. 77 Douglas Fordham, ‘Oriental Scenery’, in Aquatint Worlds: Travel, Print, and Empire, 1770-1820 (London: Paul Mellon Centre for Studies in British Art, 2019), 117. 78 Daniell, ‘William Daniell’s Diary’, 77. 79 Archer, Early Views of India: The Picturesque Journeys of Thomas and William Daniell, 1786-1794, 141-142.

224 

Construc ting K anchi

Illustration 83 Thomas Daniell, The entrance of an excavated Hindoo Temple, at Mavalipuram, on the coast of Coromandel, Plate 2 from Oriental Scenery, Part 5, 1799, aquatint, British Library, X432/5(2)

unsuccessful. As the Daniells moved deeper into the South, however, they turned to architecture with greater attention. 80 In their visits to Māmallapuram at the conclusion of their time in South India, they would have found the ideal synthesis of architectural interest and picturesque landscape (Ill. 83). This may explain why, whereas their representations of Māmallapuram are prolific, the Kanchi drawings remain unknown. Thomas Daniell’s drawings of the southern gopura would have provided a view of Ekāmbaranātha that may have been unprecedented but would not remain unique. In 1812, the little-known artist Thomas Fraser produced an elevation of the same gopura (Ill. 84). The architectural precision of Fraser’s drawing reveals the artist’s technical training as a military engineer, and he may have used a camera obscura or other optical device as well.81 Like the Daniells, Fraser prepared his own prints from the sketches he made on-site. 80 Ibid., 143. 81 Fraser served with the Madras Engineers from 1796 through 1823 and became Superintending Engineer of Public Works for the Madras Presidency Division in 1803. British Library online catalogue of India Office Select Materials, WD512, www.bl.uk/catalogues/indiaofficeselectpd/ PrintandDraw.aspx

K anchi Under Colonialism

225

Illustration 84 Thomas Fraser, Elevation of the Great Pagoda at Conjeveram, 1812, engraving, 68 × 50 cm, British Library, P250

Comparison between Fraser’s and the Daniells’ works shows that the picturesque aesthetic was of secondary importance to capturing the proportions. However, the focused attention on architectural design led Fraser to significantly misrepresent the sculptures that adorn the gopura’s exterior. Various figures replace the pairs of dvārapālas that consistently flank the central windows up to the summit. Just beneath the barrel-vaulted roof, the iconography is also entirely transformed. On the roof itself, the central Garuda flanked by Vishnu and Shiva is replaced with a somewhat abstracted Shiva and Umā that more closely resembles Ravana’s abduction of Sita. It is possible that Fraser simply noted the presence of sculptures while he stood on the busy Kanchi street, planning to insert them when he had time to work at leisure. In order to determine which deities to depict, he could have consulted other gopuras he encountered in less populated places, or other representations in circulation at the time.

James Wathen’s Soaring View The artist and writer James Wathen also included an illustration of the Ekāmbaranātha’s outermost, southern gopura in his Journal of a Voyage, in 1811 and 1812, to Madras and China, published in 1814 upon his return

226 

Construc ting K anchi

Illustration 85 J. Clark after James Wathen, Second View from the Great Pagoda near Conjeveram (Plate VI), 1814, hand-colored aquatint, Yale Center for British Art, Paul Mellon Collection, DS412.W38

to England. The image shows the tower in isolation, its awe-inspiring structure dwarfing the people who gaze up in wonder. Nothing of the busy bazaar street is seen in the picture. However, Wathen did not work at the Ekāmbaranātha temple from ground level only. He also ascended another gopura, where he was rewarded with a soaring view.82 Together with his descriptive text, Wathen’s views from the top of the Ekāmbaranātha tower achieve an ideal panoramic experience that was highly sought by European travelers (Ill. 85).83 The panorama was at its peak in popularity in England precisely at the time of Wathen’s travels in India. As a result, Wathen’s images are somewhat formulaic – the vantage point is high, the mountains convey a sense of sprawling vastness, and the lakes shimmer in the foreground; the shrines, houses, and people that populate 82 If counted from the outside, Wathen would have ascended the second southern gopura, which is accessed from the thousand-pillared hall. The Clives climbed the same tower (see below). The stairs remain present but inaccessible. 83 On the panorama, see Stephen Oettermann, The Panorama: History of a Mass Medium (New York: Zone Books, 1997).

K anchi Under Colonialism

227

the landscape give a sense of infinitely receding depth. Although only two illustrations were published, the artist tells us that he made drawings of all four directions, which would have immersed the viewer in full surround. The text is similarly sublime. ‘Almost entranced in its contemplation’, Wathen writes, I forgot all the world beside, and felt as if I could have continued on this elevated spot forever. To whichsoever point of the compass I turned, the view was equally wonderful, new, and enchanting. The eye of man, I am persuaded, never could, from any other spot in the universe, survey a scene more grand, beautiful, and interesting… To the West and South the view was bounded by majestic gauts or mountains. To the Northeast was the open country, Madras, and the sea. The gauts beyond Arcot and Vellore were lofty, and plainly perceptible.84

Wathen himself was a teetotaling vegetarian, known for his remarkable stamina.85 His unconventional character may have prompted him to visit lesser-known parts of India, including Kanchi. Wathen spent three days in Kanchi at the recommendation of an unnamed friend in England who told him he would see ‘many Hindoo temples, or pagodas; particularly, two very large ones, which were much reverenced by the natives, and visited often at their festivals by persons residing at a great distance.’86 Wathen traveled by the favored method for Indian elites and European visitors alike – carried in a palanquin in the cool of night. He was accompanied by a friend named Mr. Parkin, an English gentleman who had chosen to remain in India after he retired from company office. Parkin had taken up residence in ‘an elegant house’ in a village he calls Persewachum, just outside Madras.87 Wathen and his party set out from Madras to Kanchi at 9 pm on 12 August 1811.88 An Indian associate had employed 24 bearers for the travelers’ palanquins and procured the provisions for the journey. These included ‘tea-equipage, cold meats, bread, wine, sugar, etc.’, all of

84 Wathen, Journal of a Voyage, in 1811 and 1812, to Madras and China; Returning by the Cape of Good Hope and St. Helena, 67-68. 85 David Whitehead, ‘Wathen, James (Bap.1752, D.1828)’, in Oxford Dictionary of National Biography (Oxford: Oxford University Press, 2004). 86 Wathen, Journal of a Voyage, in 1811 and 1812, to Madras and China; Returning by the Cape of Good Hope and St. Helena, 52. 87 Ibid. 88 Ibid.

228 

Construc ting K anchi

which originally would have been imported to India through the sprouting networks of empire.89 As night fell, Wathen made note of the many small villages with mud houses and the soothing melody of his bearers’ song. At daybreak he found himself in a wilderness area, with little evidence of settlements and ‘scarcely any appearance of a road.’90 The paddy fields would have been brown in the dry of August, the Tamil landscape withholding her emerald greens until the monsoon’s arrival a month later. As the party drew nearer to Śrīperumputūr, however, they began to see evidence of the settled area: a large train of bullocks loaded with sacks of cotton on the way to Madras, a spacious stone building surrounded by stables, various animals grazing in a field. Wathen was pleased to fulfil his friend’s prediction by encountering a group of people from Madras traveling to the Ekāmbaranātha temple. Amid these informative descriptions, Wathen offers explanation of the ‘choultries’ that are mentioned with great frequency in colonial narratives. Choultry was itself a word with a questionable etymology. In Hobson-Jobson, the authoritative dictionary for Anglo-Indian terminology, ‘choultry’ is described as ‘a place where four roads meet.’91 The term is further defined as, ‘a hall, a shed, or a simple loggia, used by travelers as a resting-place, and also intended for the transaction of public business.’92 Some structures called choultries were ancient maṇḍapas. Others were modern constructions, large enough in some cases to accommodate a regiment.93 These were built in a European style and supplied with stables and other amenities.94 Certain choultries were outfitted with facilities, such as tables, chairs, and stands for placing mattresses should a traveler wish to spend the night. According to Wathen, the more robust choultries were established by the British government, ‘for the convenience and comfort of travelers, as there are no public inns upon the roads in India.’95 In the early twentieth century, district gazetteers would list the available rest houses in each town and detail the nature of the accommodations, including cost.96 However, in 89 Ibid., 53. 90 Ibid., 54. 91 Sir Henry Yule and A.C. Burnell, Hobson-Jobson: A Glossary of Colloquial Anglo-Indian Words and Phrases, and of Kindred Terms, Etymological, Historical, Geographical and Discursive (London: J. Murray, 1903), 211. 92 Ibid. 93 Wathen, Journal of a Voyage, in 1811 and 1812, to Madras and China; Returning by the Cape of Good Hope and St. Helena, 55. 94 Ibid. 95 Ibid. 96 Madras District Gazetteers: Statistical Appendix for Chingleput District, 3-7.

K anchi Under Colonialism

229

earlier times when foreign travelers were fewer, choultries seem to have been free of charge. The Dutch traveler Jacob Haafner describes staying in choultries along with his palanquin bearers at no cost in 1786, and he illustrates his description with a simple wooden house with a tiled roof.97 Access to choultries generally was monitored by a local resident responsible for keeping the keys, much like watchmen guard keys to temples today.98 Journeys were timed in conjunction with the distances between choultries, and the newer ones were presumably positioned according to frequent travel patterns. The Indian associates who accompanied travelers were adept at readying breakfast when the party arrived at the destined choultry in the morning. Like Wathen, Henrietta Clive makes frequent note of her ‘breakfasting’ locations, as the visitors would have arrived hungry from traveling overnight.

Henrietta Clive’s ‘Hindoo Gods and Monsters’99 A decade before Wathen’s three-day visit to Kanchi, Henrietta Clive and her entourage also had attempted to ascend one of Ekāmbaranātha’s soaring gopuras. Strong as she was to travel in South India, Henrietta found herself unable to manage the tower’s tremendous height. She writes, The steps are steep and the staircase totally dark. I went up the first two stories with great surprise. The ladder was more than I could handle. I went up a few steps, but my head grew so giddy I was obliged to return. Signora Anna, the girls and everybody else went to the top. The view is very extensive. We saw a good country with a great deal in cultivation. The hills in the distance are wild and fine and I was much pleased with the whole scene. From the top there is a more commanding view, but I saw as much as satisfied me.100

For the Clive ladies, travel to Kanchi was fueled by a thirst for adventure. The family had left from England, traveled around the Cape of Good Hope, and landed in Madras at the end of the eighteenth century. Bored with the 97 van der Velde, Life Under the Palms: The Sublime World of the Anti-colonialist Jacob Haafner, 102 and 103, Fig.18. 98 Wathen, Journal of a Voyage, in 1811 and 1812, to Madras and China; Returning by the Cape of Good Hope and St. Helena, 55. 99 Shields, Birds of Passage: Henrietta Clive’s Travels in South India, 1798-1801, 110. 100 Ibid., 109-110.

230 

Construc ting K anchi

Illustration 86 After Anna Tonelli, Palanquin Bearers and a Messenger Camel, in the Journal of Lady Charlotte Florentia Clive, London, 1857, British Library, WD4235

minimal and formulaic social life of a colonial wife, Henrietta soon desired to see more of India. Taking her two daughters and their governess Anna Tonelli with her, she first moved westwards, which meant that – like the Daniells and other visitors to the Arcot area – the temples of Kanchi were among the very first South Indian architecture she encountered. The party’s eyes were not yet acclimated to what they saw, and although they enjoyed their visit to the temples, they did not remain in the city more than one night. The artist Anna Nistri Tonelli was an important member of Henrietta Clive’s extensive entourage. Tonelli was trained as a portraitist in Florence, where she had met Edward Clive, the Earl of Powis.101 She achieved significant success as an artist, and her works were shown at the Royal Academy in 1794 and 1797. Despite exhibiting among professionals, Tonelli likely never achieved the full recognition her artworks deserved, in part because as a tutor and drawing teacher she would have been regarded first

101 Biographical details on Anna Tonelli are from Neil Jeffares, ‘Tonelli, Sig.ra Luigi, Née Anna Nistri’, Dictionary of Pastellists before 1800 (2006), www.pastellists.com/Articles/TONELLI.pdf.

K anchi Under Colonialism

231

and foremost as a servant in eighteenth-century polite society.102 From 1794 to 1798, she was also the Clive daughters’ drawing instructor in London, after which she remained with the family as the girls’ governess during their travels in India. Remaining a practicing artist, she illustrated their journeys along the way. Tonelli’s works survive in limited form, primarily within a reproduction of Charlotte’s journal at the British Library, as well as ten portraits in the collection of Powis Castle.103 Coupled with the substantial written accounts by Henrietta and Charlotte, Tonelli’s illustrations tell us a great deal about how colonial-era Europeans traveled in South India, though they do not show us architecture. A painting in Charlotte’s journal depicts a group of palanquin bearers at rest in a rural landscape (Ill. 86). They sit or stand at ease, with the palanquin poles beside them or held loosely in their hands. Three bearers are in a shady area in the foreground. A larger group is clustered around the palanquin, which is laid on the banks of a pond. A British officer stands and converses casually with them, and a camel with a rider stands behind him. Swift on their feet and resilient in hot climates, camels were often employed as message bearers in the colonial period. In addition to Charlotte’s journals, camels are also mentioned as messengers throughout accounts from William Daniell and Prince Soltykoff, and a camel appears in James Fergusson’s representation of the Five Rathas at Māmallapuram (see Ill. 96). Describing her family’s entourage, Charlotte writes, We had two camels, which were mostly used for carrying messages, and one hundred bullocks to draw the bandies in which all the rest of our baggage was to be conveyed.104

Although Tonelli leaves us no illustrations from Kanchi itself, Henrietta and Charlotte’s writings provide rich accounts of the city. Most importantly, they describe Kanchi as a prosperous and industrious place. Although they found the multi-armed deities strange, the ladies were greatly impressed by the temples, both as architectural edifices and for their capacity as sites of communal activity. We learn from the journals that Henrietta and her 102 Kim Sloan, A Noble Art: Amateur Artists and Drawing Masters, C.1600-1800 (London: British Museum Press, 2000), 176. 103 Clive, Journal of Lady Charlotte Florentia Clive (1787-1866), Copied by Professional Calligrapher, W.H. Ramsay. 104 Ibid., 103.

232 

Construc ting K anchi

entourage passed through a rural area and entered the well-kept city streets through Little Kanchi, where their first architectural sight was a white gopura that very likely belonged to the Varadarāja Perumāḷ temple. Upon arrival, they were greeted by a welcome party in a scene that uncannily recalls the episode in the twelfth-century Periya Purāṇam, when Kanchi’s residents rush forth to greet the Tamil saint Sambandar (who had also arrived by palanquin) a thousand years earlier.105 Leaving their own palanquins behind, the Clives entered Kanchi with the pomp and circumstance of a royal party. They rested for a while in the garden house of the district collector, Mr. Hodson, and then set off to see ‘the Great Pagoda’, as colonial sources often refer to the Ekāmbaranātha temple.106 Although neither author specifically mentions that the Great Pagoda is in Big Kanchi, they are careful to distinguish Little Kanchi from the rest of the city. This shows that like Shiva-Kanchi and Vishnu-Kanchi, Big (‘periya’) Kanchi is a later appellation. Ekāmbaranātha was of particular interest to Henrietta because of the victory her father-in-law had achieved there in 1752 and the role the temple had played throughout the Mysore Wars. According to Charlotte’s adept description, the Clive party’s route through the Ekāmbaranātha complex took them f irst up the southern gopura of the fourth enclosure (over the thousand-pillared hall), then into the temple itself, and then back out through the thousand-pillared hall.107 The ladies both write that in the same morning they visited a mosque, which was likely either the Sunnath Jamath Jumma Masjid or the Hajrath Hameed Hauliya Dharga. Both of these establishments are situated near the Ekāmbaranātha temple along the North Raja Veedhi and remain principal sites of Muslim activity in the city. Although the Jumma Masjid is more prominently located, Henrietta’s mention that the structure was built over the grave of a holy man suggests it was the dargha rather than the mosque that they visited. That at least the mosque was known to early European travelers by the first decade of the nineteenth century is evident from a drawing by one of Mackenzie’s collectors that presents a view of the mosque from the northeast (Ill. 87).108 However, the drawing’s title refers not to the 105 Periya Purāṇam, line 2884-2888. 106 The collectorate grounds remain a pleasant area shaded by banana groves, not far from the Varadarāja Perumāḷ temple. 107 Clive, Journal of Lady Charlotte Florentia Clive (1787-1866), Copied by Professional Calligrapher, W.H. Ramsay, 47-48. 108 The drawing is dated 1804 in the British Library record, though the folio itself is undated. British Library online catalogue of India Office Select Materials, WD717, www.bl.uk/catalogues/ indiaofficeselectpd/PrintandDraw.aspx.

K anchi Under Colonialism

233

Illustration 87 Mackenzie Collector, Conjeveram Before You Come to the Great Pagoda, 1804, wash, 32 × 50.2 cm, British Library, WD717

mosque but to the building’s position in relation to the Ekāmbaranātha temple. Although this is one of the busiest intersections in the city, the artist has left the scene devoid of people in order to isolate the architecture. The Clives, by contrast, were interested in recording all aspects of their experience. Throughout their writings, the Clive ladies express a commonly held colonial preference for Islamic architecture over ‘Hindoo gods and monsters.’109 Nevertheless, they continued to visit temples and concluded their Kanchi tour with the Varadarāja Perumāḷ temple, about which they write little. At least one of Colonel Colin Mackenzie’s collectors was in Kanchi only days after the Clives’ visit. He left us a drawing of a Jain ascetic, dated 16 March 1800. A second drawing from Kanchi in the same Mackenzie album was likely made at the same time as the surrounding folios, all of which are dated in the year 1800. Inscribed ‘A Buljawar Merchant, drawn by a Brahman at Conjee’, it shows a prosperous merchant who had come to the city, presumably for trade. The two Kanchi folios in this album suggest that either Mackenzie employed a Kanchi resident, or that someone associated with Mackenzie was in the city when the Clive ladies arrived.

109 Shields, Birds of Passage: Henrietta Clive’s Travels in South India, 1798-1801, 110.

234 

Construc ting K anchi

Colonel Colin Mackenzie’s Search for the Jains Colonel Colin Mackenzie was a Scottish officer in the British Army who served in Śrīraṅgapaṭṭaṇa during the Mysore Wars. After the defeat of Tipu Sultan, Mackenzie assumed the prestigious role as the first Surveyor General of India in 1815.110 With the Mysore Sultanate no longer an obstacle, Mackenzie embarked on a grand survey of South India. Unlike other colonial collectors, he worked closely with a network of Indian associates whom he employed to gather materials on architecture, epigraphy, culture, language, and methods of economic and social exchange. The importance of these assistants cannot be understated, not only in the accumulation of information, but also for their role in the production and systemization of colonial knowledge.111 Now housed at the British Library, the drawings, manuscripts, sketches, maps, and inscriptions that Mackenzie’s collectors gathered comprise one of the largest and most valuable resources for the study of colonial-period South Asia. It was largely through the collected papers of Colin Mackenzie that Kanchi’s Pallava temples began to come into favorable light for European audiences, yet this effect was unintentional. Mackenzie’s project seems to have aimed at collecting anything and everything from antiquity, without prioritizing any one period over another.112 He did not hold the same prejudices as some of his contemporaries, and he carried out his project largely before discourse that denied India a sense of history became prevalent.113 The process worked through local chains of contact – when someone directly employed by Mackenzie arrived in a new place, he would typically ask to be taken to the assembly of Brahmins, who could then engage their own local networks to assist the collector.114 Scholars have refered to Mackenzie’s 110 A good summary of Mackenzie’s career can be found in, David Blake, ‘Colin Mackenzie: Collector Extraordinary’, British Library Journal, no. 21 (1991). 111 See Jennifer Howes, Illustrating India: The Early Colonial Investigations of Colin Mackenzie (1784-1821) (New Delhi: Oxford University Press, 2010); Phillip B. Wagoner, ‘Precolonial Intellectuals and the Production of Colonial Knowledge’, Comparative Studies in Society and History 45, no. 4 (October 2003); and the important works on colonial knowledge and collecting by Nicholas B. Dirks, Castes of Mind: Colonialism and the Making of Modern India (Princeton: Princeton University Press, 2001) and ‘Colonial Histories and Native Informants: Biography of an Archive’, in Orientalism and the Postcolonial Predicament: Perspectives on South Asia, ed. Carol A. Breckenridge and Peter van der Veer (Philadelphia: University of Pennsylvania Press, 1993). 112 Orr points out that Mackenzie’s methodology was quickly abandoned in the formative phases of Indology. Orr, ‘Preface’, XXIII. 113 Dirks, ‘Colonial Histories and Native Informants: Biography of an Archive’, 292 and 303. 114 Ibid., 297.

K anchi Under Colonialism

235

Illustration 88 After Henry Salt, Pagoda in Conjeveram, Published: Valentia, 1811, opp. p. 437, British Library, W5289

collaborators as assistants, members of a team, or an establishment. However, we might more productively think of them as a network of direct assistants to Mackenzie who collaborated with local subnetworks of people in each area. Over the course of the f irst two decades of the nineteenth century, Mackenzie’s network of collectors made representations of Kanchi’s art, people, and diverse religious monuments. One collector managed to procure a detailed architectural plan of the Kailāsanātha temple (see Ill. 17). Unfortunately, we do not know if the draughtsman was a Kanchi resident or the traveling collector. This document may well have been the reason why the Viscount Valentia and artist Henry Salt later made a point of seeking out the remote Kailāsanātha temple when they visited the city. The drawing of the Kailāsanātha by Henry Salt was published in 1809, under the title Pagoda in Conjeveram, as part of Valentia’s multivolume opus, Voyages and Travels to India, Ceylon, the Red Sea, Abyssinia, and Egypt (Ill. 88).115 The image depicts the temple obliquely, from outside the prākāra on the southwest side. The prākāra wall appears damaged and overgrown, and the compound sits in uncultivated f ields. Two Indian men stand 115 My thanks to Siddhartha Shah for drawing my attention to this image (London 2016).

236 

Construc ting K anchi

beside the temple and gaze up at it in wonder. The contrast between their diminutive size and the temple’s monumental scale is awe-inspiring. Salt’s drawing of the relatively abandoned Kailāsanātha contrasts wonderfully with Valentia’s description of the active Ekāmbaranātha temple, [From] the great pagoda, dedicated to Iswara, the priests, and numerous dancing-girls were drawn out to pay their compliments. The latter were very numerous, and some of them pretty.116

It is unlikely that Kanchi locals would have directed European visitors to the Kailāsanātha, which was hardly a point of civic pride at that time. The temple was not in regular use, it lay outside the center of the city, and the visitors would not have passed by it on their way into or out of Kanchi. It is possible that Valentia and Salt knew of the Kailāsanātha simply from the plan in the Mackenzie papers, which was made around the turn of the nineteenth century. Although the folio bears no date, the drawing must have been made between 1796, when Mackenzie began his collecting activities, and 1809, when he concluded them in the area.117 The Mackenzie manuscripts also include the first facsimiles of the Kūram copper plates, an important set of Pallava records. Bound in a volume of inscriptions titled The Ancient Written Characters of India, the Kūram plates were recorded not for their historical import but for their value as specimens of an Indian script. Similarly, a Mackenzie collector documented the fragmentary Pallava-era pillar bases that remain at Kanchi’s Kacchapeśvara and Vīraṭṭāṉeśvara temples (Ill. 89). The single drawing, dated 1807, is among the only archival representations of these important objects.118 Titled in the plural, Hindoo Architecture: Granite Pillars in a Choultry at Conjeveram, the drawing in fact shows three rotated views of the same pillar. While the figures are similar in type to those on the pillars at Kacchapeśvara and Vīraṭṭāṉeśvara, the combination does not conform to any one pillar individually. The album does not indicate specifically where the depicted pillar was found, and it is possible that the fragment was removed, destroyed, or covered by construction since the time it was documented. Although the Mackenzie team may not have had reason to believe the Pallava vestiges 116 Valentia, Voyages and Travels to India, Ceylon, the Red Sea, Abyssinia, and Egypt in the Years of 1802, 1803, 1804, 1805, and 1806, 4 vols., Vol. 1 (Text) and 4 (Plates), 396. 117 Blake, ‘Colin Mackenzie: Collector Extraordinary’, 130 and 32. The British Library catalogue dates the plan to 1800. 118 I know of only two mid-twentieth-century photographs in the collection of the French Institute of Pondicherry, IFP_6000-6003 and IFP_2689-2612.

K anchi Under Colonialism

237

Illustration 89 Mackenzie Collector, Hindoo Architecture: Granite Pillars in a Choultry at Conjeveram, 14th June 1807, 1807, folio 26 in Album of 156 drawings chiefly of architecture and sculpture in S. India (18031808), pen and ink, British Library, WD1064

were of special significance, their collecting project constitutes the earliest wave of consolidating and transmitting knowledge of Kanchi’s Pallava history to European audiences. Mackenzie held a special interest in India’s Jain history, in part as an alternative to Hinduism, and in part because one of his primary associates was Jain. His network may have been especially receptive to Jain vestiges in Tamil Nadu because they had begun their work in the Deccan, where many Jain monuments can be found.119 As a result, Jainism was among the first traditions to be addressed in the surveys. Mackenzie’s dedication to Jain studies can be seen through a portrait by Thomas Hickey (1816) in which Mackenzie is depicted along with his three most important Indian associates – Dharmaiah, a Jain and expert in ancient languages; Kavali Venkata Lakshmayya, Mackenzie’s interpreter; and Kistnaji, a harkara (‘expert in geography’) – with the monumental Jina statue at Karkala behind him.120 119 Thanks to Crispin Branfoot for this suggestion, and for an interesting conversation about Mackenzie’s project. 120 Tobias Wolffhardt, Unearthing the Past to Forge the Future: Colin Mackenzie, the Early Colonial State, and the Comprehensive Survey of India, trans. Jane Rafferty, Studies in British and Imperial

238 

Construc ting K anchi

Illustration 90 Mackenzie Collector, A Jain at Conjeveram, 16th March 1800, folio 4 in Album of 82 drawings depicting the costume of various castes in Balaghat, Carnatic, 1800, watercolor, British Library, WD1069

K anchi Under Colonialism

239

A great many images in the Mackenzie collections depict artworks and individuals believed to be from the Jain tradition. In ‘The Jains’, a short travel essay published in the journal Asiatic Researches in 1809, Mackenzie records so-called obscure narratives and scattered mentions of a community that was ‘formerly a powerful people, who contested the sway with the Brahmans.’121 The perceived connection to the past would later lend these images an even greater significance in the colonial archive. The very first drawing from Kanchi in the Mackenzie collections, dated 1799, is labeled Statue dug up at Conjeveram, supposed to belong to the Jain worship. The deity’s tight snail shell curls and lack of cranial protuberance (‘uṣṇīṣa’) or added adornments suggests the sculpture belonged to the Jain tradition, but it very well may have been a Buddha. Mackenzie often identified Buddhist remains as Jain, even including the now-famous stupa at Amarāvatī. A second drawing, inscribed ‘A Jain at Conjeveram, 16th March 1800’, sensitively renders a barefoot ascetic with the attributes of a South Indian Jain, including the sacred marks on his forehead (Ill. 90). These early Mackenzie sources seem to have established what would become an enduring reputation among Europeans that Kanchi was a place where a substantial Jain legacy could be discovered. European artists and travelers studied existing documentation of the regions they planned to visit in order to determine their itineraries and market their experiences – not unlike we do today with digital resources and popular guidebooks.122 Individual documents or publications likely provided the impetus for travelers to seek out particular places. Reading the words of Robert Orme, Henrietta marvels at the accuracy of his description as she explores Samiaveram (present-day Camayapuram, ten kilometers north of Tiruchirappalli), Upon examining this spot with Orme’s book it exactly answered in every respect to the distance from the lesser and greater pagoda. In the evening I went to the place and have not the least doubt it was right.123

History (New York: Berghahn Books, 2018), 4-5. For identification of the site, Crispin Branfoot, ‘Review of Illustrating India: The Early Colonial Investigations of Colin Mackenzie (1784-1821)’, Bulletin of the School of Oriental and African Studies 75, no. 1 (2012): 181. 121 Mackenzie, ‘The Jains’, 227. Here he uses ‘Brahmans’ in the sense of Shaiva and Vaishnava Hindus. Tamil Jains also could be members of the Brahmin caste. 122 Romita Ray cites examples from travelers elsewhere in India. Ray, Under the Banyan Tree: Relocating the Picturesque in British India, 8. 123 Shields, Birds of Passage: Henrietta Clive’s Travels in South India, 1798-1801, 227.

240 

Construc ting K anchi

Illustration 91 John Gould, Map of Conjeeveram, 23 April 1816, 1816, Pencil, penand-ink, and watercolor, 19 × 23 in, British Library, WD2701

The first and one of the only attempts to map Kanchi was made in 1816 by the artist John Gould (Ill. 91).124 Drawn to scale, Gould’s early map presents a wide-sweeping view of Kanchi surrounded by a verdant landscape. While most of the map is not labeled, the prominence accorded to Tiruparuttikuṉṟam as ‘A Jain Temple and Establishment’ may have been due to Mackenzie and other colonial agents’ special interest in recovering a fictive past in which Jainism had been India’s dominant religion. It also may have been the case that Gould made use of local Jain informants.125 In whatever circulation it may have had, this map would have perpetuated the city’s Jain reputation. 124 The map’s author is not the renowned ornithologist and natural history artist of the same name. Besides the maps I produced for this book, I am aware of only two other maps of the city: One produced by the EFEO Pondicherry in 1972 (PY523) and the other labeled in Tamil and found on display in select temples in Kanchi. I was fortunate enough to procure the Tamil map through the off ice of the Kacchapeśvara temple and to scrutinize the EFEO map with Ramaswamy Babu, head archivist in Pondicherry. 125 I thank Tamara I. Sears for this suggestion.

K anchi Under Colonialism

241

Kanchi may well have an important Jain history. Xuanzang’s description of the city includes the presence of many Jains, whom he calls Nirgranthas.126 The Pallava-era Candraprabhā Jiṉālāya temple and Chola- through Vijayanagara-era Trilokya Jinasvāmi temple show that there was significant support for Jain establishments, at least through the Vijayanagara period. There are other newer, active Jain temples throughout the city as well. Kanchi’s inland position further supports the possibility of a substantial Jain population in the early historic period. While Buddhist populations tended to concentrate in coastal areas to facilitate their involvement in maritime trade, Jain ascetics gravitated towards inland areas where hills provided shelter for meditation.127 The Brahmi inscription at Māmaṇṭūr shows that Jains were established in the Kanchi area by the first century CE.128 However, the continued prominence of Jainism seems to have been somewhat exaggerated in the Mackenzie manuscripts. Early twentiethcentury records that provide more detailed information on population and communities tell us Kanchi was very much dominated by Hinduism. The Imperial Gazetteer of India, published in 1908, records that within a population totaling 46,164, there were 44,684 Hindus, 1,313 Muslims, 49 Christians, and only 118 Jains.129 Nevertheless, the gazetteer goes on to discuss the Jain temple of Trilokya Jinasvāmi before any other temples in the city, thus perpetuating the hyperbole. It mentions that donative inscriptions by successive dynasties reinforce the temple’s importance. However, it is more likely that rulers gave land grants to this temple as a way to ensure that the outskirts of Kanchi and the non-Hindu populations fell under their control. Mackenzie’s extensive documentation of South India circulated with decreasing frequency throughout the nineteenth century. The most robust references are given in Mark Wilks’ 1817 publication, Historical Sketches of the South of India; later Mackenzie would appear in the context of gazetteers and colonial manuals, including The Imperial Gazetteer of India.130 District collector Charles Stewart Crole refers frequently to the Mackenzie

126 Xuanzang, Si-Yu-Ki. Buddhist Records of the Western World. Translated from the Chinese of Hiuen Tsiang (A.D. 629) by Samuel Beal, 2, 229. 127 Champakalakshmi, Religion, Tradition, and Ideology: Pre-colonial South India, 338. 128 ARE 1939-1940, no. 171. 129 ‘Conjeeveram Tāluk and Conjeeveram Town’, in The Imperial Gazetteer of India: Vol. X, Central Provinces to Coompta, 377. 130 Dirks, ‘Colonial Histories and Native Informants: Biography of an Archive’, 304.

242 

Construc ting K anchi

manuscripts throughout his 1879 report.131 Today, the archive gives us access to a wide range of individuals whose voices cannot otherwise be heard.132

Surgeon George Russell Dartnell Over the course of the nineteenth century, artists and travelers would continue to seek out fragments of the past, and to present India’s landscape through picturesque views. The British Army Surgeon, George Russell Dartnell, worked in the foreign service in India, Ceylon (Sri Lanka), and Burma (Myanmar) during the 1820s and early 1830s and made several representations of Kanchi in 1829. He completed several paintings based on his India drawings while posted in Canada over the course of the next two decades, yet further records of his time in India are scarce.133 Two original drawings from the unfortunately titled album Worthless Scraps from Many Lands depict an elaborately carved pillar (Ill. 92). Both images are inscribed and dated, Stone Pillar in a Choultree at Conjeveram–near Wallajabad–Jany 1829, but no indication is given as to which temple the pillar belongs. The pillar type is not uncommon in post-Chola architecture in Kanchi, seen for example at the Kāmākṣī Ammaṉ temple’s vasanta maṇḍapa (Ill. 93). However, given that European visitors tended only to visit Ekāmbaranātha, most likely it is located in that temple’s thousand-pillared hall, described with such wonder by Charlotte Clive.134 In Dartnell’s 1829 drawing, the artist has isolated the pillar and also distorted it in ways that are not altogether unexpected of a colonial artist. Most notably, he has amplified the size of the figure at the pillar’s base. Whereas the real Kanchi pillar of this type has two registers of deities above the figure’s head, Dartnell’s has room for only one. This transformation was in fact a typical colonial artistic practice, which, while adding to a sense of the grotesque, also indicates the ways in which foreign visitors grappled with unfamiliar forms. Large figures on pillars are characteristic of the architecture of Greco-Roman Antiquity that would have been familiar to European travelers and are still typical features of modern Hellenistic-style 131 Crole, The Chingleput, Late Madras, District: A Manual Compiled under the Orders of the Madras Government, by Charles Stewart Crole, of the Madras Civil Service. 132 Ibid., 308. 133 Honor de Pencier, Posted to Canada: The Watercolours of George Russell Dartnell (Toronto and Oxford: Dundurn Press, 1987), 26. 134 Clive, Journal of Lady Charlotte Florentia Clive (1787-1866), Copied by Professional Calligrapher, W.H. Ramsay, 48.

K anchi Under Colonialism

243

Illustration 92 George Russel Dartnell, Stone Pillar in a Choultree at Conjeveram– near Wallajabad–Jany 1829, Pencil and sepia ink wash, British Museum, 1997,1109,0.60 and .61

architecture throughout the Western world. The impulse to Hellenize Indian sculpture was in keeping with lines of discourse that viewed India as an archaic version of a pre-classical ancient past.135 Related narratives of stagnation and timelessness were crystallizing throughout the nineteenth century. Dartnell’s stay in Kanchi was bracketed by crucial transitions to official systems of temple administration. In 1817, the East India Company passed Regulation VII, which transferred custodianship of all Tamil temples in EIC-controlled areas to the Board of Revenue, managed through the district collectors.136 The collectors’ responsibility was soon taken over by 135 Mitter, Much Maligned Monsters: History of European Reactions to Indian Art. 136 The following discussion is drawn from Venkataraman, Dēvī Kāmākshī in Kāñchī: A Short Historical Study, 55-56.

244 

Construc ting K anchi

Illustration 93 Pillar in Vasanta Maṇḍapa, Kāmākṣī Ammaṉ Temple, Kanchi, sixteenth century

K anchi Under Colonialism

245

committees, which received fixed allowances from the central government. When this system failed as well, the Madras Government created the Hindu Religious and Charitable Endowments Administration Department, which still controls all temples in Tamil Nadu that are not under the auspices of the Archaeological Survey of India or private individuals. In 1842, committees in Kanchi were appointed to manage the Ekāmbaranātha and Varadarāja Perumāḷ temples. However, the Kāmākṣī Ammaṉ temple instead was entrusted to the Śaṅkarācārya of Kumbakonam, at his request. Amid protests by the Vaṭama Kāmakōṭṭiyār families, who cite their origins to the valley of the Narmada River and who had previously controlled the temple, the Śaṅkarācārya from Kumbakonam assumed responsibility in 1843. More than one hundred years later in 1960, the Śaṅkarācārya’s establishment resigned its position, and the Hindu Religious and Charitable Endowments Administration again appointed the Kumbakonam Śaṅkarācārya (now under the name Sri Kanchi Kamakoti Peethadhipati Jagadguru Sri Shankaracharya Swami Mutt) to administer the temple. The temple remains under the auspices of the Mutt today. With the establishment of the Hindu Religious and Charitable Endowments Administration came a distinct differentiation between archaeological monuments that were protected by the ASI and active centers of worship. These two temple administrations coexist uneasily in Kanchi and throughout Tamil Nadu today. In most of the ASI-protected temples, such as the Kailāsanātha, the Hindu Religious and Charitable Endowments Administration controls the inner sanctum. This means that the exterior of the temple cannot be touched, but the priests have full permission to use the interior for worship.

James Fergusson’s Downward Spiral With the exception of the Mackenzie drawing of the Kailāsanātha temple, the great Pallava monument would not make its lasting entrance into European circuits until the final decade of the nineteenth century. By this point in time, rather than recording all elements of India’s archaeological heritage, as the Mackenzie project had done, surveyors were endeavoring to distinguish between what was old and what was even older. In a letter to the Journal of the Royal Asiatic Society of Great Britain and Ireland in 1891, James Burgess, who was then Superintendent of the recently founded ASI,

246 

Construc ting K anchi

Illustration 94 Archaeological Survey of India, Southeast view, Kailasanatha Temple, Great Conjeeveram, Chingleput District, 1900, photographic print, 24.3 × 29.4 cm, British Library, Photo 1008/5(391)

corrected an error.137 In the Journal’s previous issue, the discovery of Kanchi’s Kailāsanātha temple had been wrongly attributed to Robert Sewell, who had in fact missed the temple entirely when he surveyed the city. Sewell had failed to notice this large monument because of its remote location on the far western edge of Kanchi, about a kilometer’s walk down the quiet street that leads to it. A photograph of the Kailāsanātha taken a decade later during the first extensive photographic survey of Kanchi shows the temple amid an overgrown field pocked with uneven surfaces (Ill. 94).138 A small cluster of villagers stands in front of the temple, posing for the photograph, and a sculpture of Nandi gazes forlornly towards an inaccessible god. The Kailāsanātha looms in a state of ruination. The monument’s scale 137 J. Burgess, ‘The Temple of Kailāsanātha’, Journal of the Royal Asiatic Society of Great Britain and Ireland (1891): 337. 138 While there are individual photographs of Kanchi from as early as the 1860s, the f irst extensive photographic survey did not take place until 1900.

K anchi Under Colonialism

247

Illustration 95 Archaeological Survey of India, South-west view, Matangesvarasvami Temple, Great Conjeeveram, Chingleput District, (Mataṅgeśvara Temple, Kanchi), 1900, photographic print, 23.5 × 18.7 cm, British Library, Photo 1008/5(407)

far outweighs the limited ritual requirements of the local residents who are gathered in the foreground. Like so many other royal temples in Tamil Nadu, what was once a courtly mansion had become a village hut.139 139 For discussion of the fate of royal temples in Tamil Nadu, see my Chapter Two.

248 

Construc ting K anchi

Other archival photographs show that the manicured lawns and gardens that today surround many of Kanchi’s Pallava temples were patches of relative wilderness at the turn of the twentieth century. Although he primarily emphasized Kanchi’s urban activity, in 1841, Prince Soltykoff described Kanchi as situated ‘in a sacred forest.’140 He may have been referring not only to the wilderness he encountered as he approached, but also to the unkempt jungle-ridden pockets that remained well within the city. The temples in these spaces were desolate, crumbling buildings, patched with poorly executed brick repairs and plastered surfaces that already belonged to the distant past. Photographs show the Mataṅgeśvara temple half buried, Mukteśvara shrouded in overgrowth, and Amareśvara badly damaged overall (Ill. 95). Slightly later photographs of a soot-encrusted Iṟavāttāṉeśvara and a partially buried Piṟavāttāṉeśvara show that these temples too were not in use.141 By the late nineteenth century, a distinct preference for these relatively abandoned sites began to be articulated in official sources. Ignorant of its antiquity or royal foundation, in 1879 district collector Crole cited a sandstone temple that is most likely the Kailāsanātha as an example of ‘the best and most brilliant period’ of Hindu architecture.142 ‘Standing in a field south-west of the new dispensary’, he reports, ‘the carvings and sculptures on this waif of the past, though florid, are full of verve and spirit.’143 It may have been because the Pallava temples were quiet places, away from the frenzied bustle of the active ritual centers, that colonial surveyors began to gravitate towards these sites. A study of Kanchi’s Pallava temples exclusively, using representations produced in the colonial era, would easily lead one to conclude that the city had followed a steadily declining pathway. The eighth century would have marked the pinnacle of the city’s efflorescence, and the ninth century would have begun its continuous slide into dilapidation and decay. Such was precisely the type of trajectory that James Fergusson articulated for the history of India as a whole. By shifting their gaze from Kanchi as a city to the Pallava temples alone, archaeological surveyors created a distorted picture of Kanchi. In the opening lines to his Picturesque Illustrations of Ancient Architecture in Hindostan, published in 1847, Fergusson emphasizes the notions of both decline and decay. He writes, 140 Soltykov, Lettres Sur l’Inde, 71. 141 Piṟavāttāṉeśvara partially buried is seen in a photograph in the collection of the French Institute of Pondicherry, IFP_5754-5756. 142 Crole, The Chingleput, Late Madras, District: A Manual Compiled under the Orders of the Madras Government, by Charles Stewart Crole, of the Madras Civil Service, 111. 143 Ibid.

K anchi Under Colonialism

249

I know of no one characteristic that can be predicated with perfect certainty of all the styles of architecture in Hindostan, except the melancholy one that their history is written in decay. For whenever we meet with two buildings or two specimens of art of any sort, in the whole country between Cape Comorin and the Himalayas, if one is more perfect or of a higher class than the other, we may at once feel certain that it is also the more ancient of the two; and it only requires sufficient familiarity with the rate of downward progress to be enabled to use it as a graduated scale, by which to measure the time that must have elapsed before the more perfect could have sunk into the more debased specimen.144

An independent scholar who sought to establish the first chronology of Indian architecture, Fergusson traveled extensively on long, self-motivated tours during the 1830s and 1840s. Although his work remains an essential resource for South Asian art historical scholarship, it is also the source of lasting misconceptions. At the root of Fergusson’s understanding of India was the deep belief in the existence of two distinct races, defined linguistically as the Sanskritic and the Dravidian, the former being invading conquerors and the latter indigenous.145 He divided architecture along these lines, with the Sanskritic giving rise to the Nāgara, and the Dravidian to the Drāviḍa classes of temples. Fergusson’s architectural theories gained traction within the Royal Asiatic Society and played a crucial role in the formative years of the ASI, before its official establishment in 1871. Faulty as it was, his architectural chronology and the qualitative judgments he attached to it would continue to reverberate as scholars established an art historical canon for South Asia.146 Fergusson maintained that there was a distinct hierarchy among Indian religions, with Buddhism followed by Jainsim, then Vaishnavism, and finally Shaivism in descending order. He argued that architecture could be classified, and thus identified, along religious lines – an assumption that would persist well into the postcolonial era. Fergusson’s preference for Buddhism and what he understood to be ‘a Buddhist architecture’ had great impact on the ASI’s decisions about which monuments to document and restore. Key among these was his belief that the shape of the Buddhist caitya hall was 144 Fergusson, Picturesque Illustrations of Ancient Architecture in Hindostan, by James Fergusson, 11. 145 Ibid., 9. 146 Tapati Guha-Thakurta, Monuments, Objects, Histories (New York: Columbia University Press, 2004), Chapter 1.

250 

Construc ting K anchi

Illustration 96 Thomas Colman Dibdin after James Fergusson, Mahavellipore. The Five Raths, 1839, lithograph, 25.8 × 36.4 cm, British Library, X590

the basis for all Hindu apsidal temples. In 1930 for example, it was colonial surveyors’ lasting preference for presumably Buddhist-based, apsidal forms that likely inspired ASI Superintendent A.H. Longhurst to take particular interest in the modest but apsidal Shiva temple at Orakaṭam.147 With Shaivism and Vaishnavism actively practiced all around him, Fergusson privileged Buddhism and Jainism so heavily in part because he viewed them as timeless and unchanging religions. They were ‘tolerably uniform in their tenets’, he surmised, while Shaivism and Vaishnavism, by contrast, existed in a constant state of flux, divided into ‘innumerable sects, some of which are continually going out of fashion, while new ones are springing up.’148 Nevertheless, Fergusson viewed the divisions of race and religion as sufficient for identifying architectural typologies. Ironically, he argued that these typologies were consistent with the very subdivisions whose stability he denied – although he considered Shaivism and Vaishnavism as ephemeral classifications, he thought the architecture they produced was definable and constant. 147 Longhurst, ‘Pallava Architecture, Part III (the Later or Rājasiṃha Period)’, 20-21. 148 Fergusson, Picturesque Illustrations of Ancient Architecture in Hindostan, by James Fergusson, 8.

K anchi Under Colonialism

251

Fergusson did not consider himself to be an artist. Instead, he used a camera lucida and kept his focus monument-centered with the intention of producing what he deemed an authentic representation of the site as he encountered it. Despite this stated intention, he could not thoroughly escape the artistic predilections of his predecessors. He even began the title of a seminal album of illustrations with the word ‘Picturesque.’ Fergusson’s images show the tendency to amplify the scale of relief sculpture and to select oblique views conducive to picturesque framings. In Illustrations of the Rock Cut Temples of India, Fergusson depicts the Five Rathas at Māmallapuram from an oblique angle on the northwest side (Ill. 96). Still half-buried in sand as they were at the time, the monolithic structures rise from the coastal landscape as evocative ruins. Fergusson likely selected this particular view because the roof structures of all five rathas were visible. Although elements of his view were documentary, we should not give the image complete credibility. As we know from the author’s own writings, the foreground, sky, and any additional scenery were left to the fantasy of Thomas Colman Dibdin, an artist in London who had never visited India.149 It is unfortunate that decades upon decades of interventions by the ASI have rendered this vantage point impossible today. It may have been the city’s Jain reputation that drew Fergusson to visit Kanchi. He mentions that the Jains left ‘abundant proofs of their existence at Conjeveram.’150 It is possible that he had heard tell of the temples in Tiruparuttikuṉṟam or knew of the city’s Jain population through Xuanzang’s account or elsewhere. He could have been familiar with Mackenzie’s 1809 publication or the original drawing of the Statue dug up at Conjeveram, or he may have known of the map on which John Gould so prominently labeled the Jain temples. Fergusson would have taken interest in a place with significant Jain history because of Jainism’s perceived proximity to Buddhism, both in terms of chronology and philosophical orientation. Despite his relatively short time in South India, Fergusson’s descriptions of Tamil temple architecture are remarkably accurate. Rather than curvilinear, superstructures are ‘rectilinear.’151 He points out that the tower above the main shrine is the highest point in the complex only at Thanjavur and a select number of other examples. While temples in urban areas or ancient capitals have accrued additional structures over time, as he tells us, many 149 Ibid., iv. 150 Ibid. 151 Fergusson, Picturesque Illustrations of Ancient Architecture in Hindostan, by James Fergusson, 20.

252 

Construc ting K anchi

village temples remain more or less as they were in their initial construction. He also demonstrates a remarkable sensitivity to variations among temples of similar design. For example, he notes that while the gopuras are generally placed on axis with the main sanctum, there are many exceptions, among them the Ekāmbaranātha temple at Kanchi. He also notes the basic layout of South Indian temples, including the presence of a thousand-pillared hall, which he had assumed mere hyperbole until he counted at Chidambaram.152 It is also in his writings on Chidambaram that Fergusson famously asserts that, were he to have a week to devote to it, he could ascertain the age of every portion of the temple.153 The author’s descriptions of temple landscapes are similarly informative. From his account we learn that Māmallapuram was separated from the mainland by a salt marsh at the time of his visit, and that there were several rocky outcroppings within the sea itself, at least one of which carried a pillar.154 However, Fergusson was not concerned with depicting landscape in his illustrations. His interests resided almost exclusively in the buildings. As stated without compunction, his insistence on the veracity of the image extended only as far as the monument’s exterior walls. At the conclusion of Ancient Architecture, he once again reiterates his theory of decline, The practiced eye can also always detect a certain elegance of outline and design in the older structures, which at once reveals their antiquity, repeats the sad lesson that the history of India and her arts is written in decay, and presents a picture of steady progressive degradation from the earliest period at which we knew her to the present hour.155

The deep irony here is that, throughout his study, Fergusson gets his architectural chronology nearly entirely backwards. He attributes much earlier dates to the late- and post-Chola-period temples that he so admires than, for example, the Shore temple at Māmallapuram, which in fact was built half a millennium earlier. Despite his insistence on a downward progression, what he was witnessing was instead – according to his own aesthetic proclivities – a steady ascent.

152 Ibid. 153 Fergusson, Picturesque Illustrations of Ancient Architecture in Hindostan, by James Fergusson, 60. 154 Ibid., 57-58. 155 Ibid., 67-68.

K anchi Under Colonialism

253

Prince Alexis Soltykoff’s ‘City of Infinite Temples’ It must have been chilly in Paris on 14 February 1841, when the Russian Prince Alexis Soltykoff boarded the Great Liverpool at Southampton to begin his first voyage to India.156 Traveling through high seas and distant lands, the Prince arrived in Bombay a month later and carried onwards to Sri Lanka and then to the Tamil country. The writings and illustrations he produced along his journeys would capture not just a visual image but the multisensory experience of the places he visited.157 In the Kanchi records, he depicts the interiors of even single temples as the disorienting complexes of long promenades, vast courtyards, shrines, and pillared halls that they actually are. He remarks on the city’s dense population, its noise, its dust, and its relentless activity. Whereas the Clives and other travelers mention only the Ekāmbaranātha and Varadarāja Perumāḷ temples, the Prince tells us he was bewildered by Kanchi’s seemingly infinite number of temples. Unabashedly, he allowed himself to be taken from one temple to another – and then yet another. He found them all to be ‘superb, magnificent, incomparable.’158 A prolific writer, Prince Soltykoff recorded the details of his journeys in letters to his dear and only surviving brother Pierre (Petr).159 Entitled Lettres sur l’Inde, the Prince’s correspondence was published in its original elegant French upon his return to Paris. It was translated soon after into Russian, but never into English, and this valuable resource remains virtually unknown to scholars.160 I include my own translation of the Kanchi episode interspersed with my analysis.161 156 Ibid., 1. 157 I choose to use ‘Soltykoff’ from the range of other possible spellings (Saltykov/Saltykow/ Soltikoff/Saltuikov) because this is how the Prince signed his artworks. The suffix –off indicates a Russian family who emigrated to or spent significant time living in France. Richard Walding, Helen Stone, and S. Nair Achuthasankar, ‘The Russian Prince and the Maharajah of Travancore’, Journal of Kerala Studies 36 (2009): 51 and 81 n. 11. 158 Soltykov, Lettres Sur l’Inde, 70. 159 Incidentally, Petr was an important collector of Renaissance ceramics (Maiolica), and his name appears in catalogues from a range of institutions. Warm thanks to Jamie Gabbarelli for drawing my attention to this aspect of the family’s eclectic history. 160 Soltykov, Lettres Sur l’Inde; Saltykoff, Pis’ma Ob Indii. The original French text, which fortunately has been digitized, can be found in a limited number of rare book collections, including the crumbling volume I accessed at the British Library. Besides brief and scattered references to the Russian translation of the Lettres, the Prince’s India travels have come to scholarly attention only once before. Walding, Stone, and Achuthasankar, ‘The Russian Prince and the Maharajah of Travancore.’ 161 Many thanks to Martine Tramus for reviewing my French translations.

254 

Construc ting K anchi

In addition to his skills as a writer, the Prince was an adept artist with an insatiable appetite for drawing. The sketches from his trip were worked into lithographs by L.H. Rudder and published in Paris at the same time as the text. As an artist, the Prince’s primary interests lay in processions, festivals, entertainment, and the paraphernalia of cultural life, but he also depicted people quite faithfully. He sketched gatherings, military and political transactions, transportation, and entertainment. He also drew portraits, for which he sometimes subjected the sitter to multiple views. He often appears in his own images, depicted as a man at leisure graced with the talent and the time to absorb the world around him. The Prince had first studied art in Russia with his tutor, Alexandr Orlowski, a Polish-born Jew who fled from Warsaw to St Petersburg in 1802, where he became a high-level court painter. Orlowski became known for his portraits of nobles in uniform, and for his Romantic history paintings. The tutor’s direct impact on the Prince can be seen both in portraiture and in military scenes. The Prince’s eye was further filtered during his travels with the Russian Foreign Service, which took him to Istanbul, Athens, London, Florence, Rome, and Tehran. In many of these places he encountered the kinds of settings that would fuel the most celebrated Orientalist painters, such as Gérôme, only decades later. A dramatic print depicting the Prince in his palanquin, journeying across the plains of the Punjab, is a scene that would have appealed to any Romantic artist (Ill. 97). The torch-lit Prince awakens in the night while fiercely armed horsemen pass at close proximity to his party. The bearded, muscular palanquin bearers are swathed in white turbans, loincloths, and cloaks. The moon is barely visible through the dust and smoke. Startled, the Prince sits up from his slumber, white light flickering all around him, and pushes open his palanquin’s door to witness the excitement. Upon reaching Tamil Nadu, the Prince was stripped of his habitual vocabulary. He did not retreat, however, but imbibed this new and very different land. Conjeveram, 27 June, Evening. I spent last night in my palanquin, and, this morning, I arrived in Conjeveram. I will make just a tour, and return to Madras the day after tomorrow in the morning. Conjeveram is a sparse village in a sacred forest, with fifty thousand residents (of which ten thousand are Brahmins, they say here), and containing an infinite number of pagodas. I stopped in an isolated house, all neat and tidy, established for foreigners. The Governor of Madras, Lord Elphinstone, knowing approximately when I would arrive, had sent three servants loaded with provisions, who awaited me.

K anchi Under Colonialism

255

Illustration 97 L.H. Rudder after Prince Alexis Soltykoff, Voyage en poste dans les plaines du Punjab, entre Loodiana et Omritsar, Fevrier 1842. [‘The artist in a palanquin in the plains of the Punjab, between Ludiana and Amritsar, February 1842’], 1848, lithograph, 33.3 and 53.8 cm, British Library, P978

Before lunch, I had a bath which they also prepared for me, just as I do twice each day; and then the Brahmins, who were aware of my arrival, sent to my door three elephants and a troupe of dancers and musicians: the dancers were not very pretty, albeit young. There was, however, one who was very good. A miserable little girl playing an instrument made of coconuts, the sound of which recalls that of the oboe and resembled the voice of a crying child; and, while playing, she writhed in a horrible manner, whilst her mother, a large black shrew, kept treating her harshly. I expressed my distaste to the mother, upon banishing her from my room, and I gave the little girl a few rupees, all the while telling myself that a word to Lord Elphinstone would suffice to moderate this old woman’s behavior towards the young girl.162

In the illustration titled Condjévéram, ville sainte dans le Carnatik aux environs de Madras, lieu de pélerinage des Hindoux [‘Conjeveram, sacred city in the Carnatic in the vicinity of Madras, a Hindu pilgrimage place’], a loose cluster of people stands amid the wooded area on the banks of the temple 162 Soltykov, Lettres Sur l’Inde, 69-70.

256 

Construc ting K anchi

Illustration 98 L.H. Rudder after Prince Alexis Soltykoff, Condgeveram, ville sainte dans le Carnatik aux environs de Madras, lieu de pélerinage des Hindoux [‘Conjeveram, sacred city in the Carnatic in the vicinity of Madras, a Hindu pilgrimage place’], 1848, lithograph, British Library, P969

tank, the pillared halls and towers of the sacred complex unfolding behind them (Ill. 98). Large-scale wooden effigies of deities and relief sculptures emerge from the distance, and devotees descend the stepped ghats for ritual ablutions. The dancer and the little girl described in the Prince’s writings appear on the left side in the foreground. The woman holds the girl by the wrist in a familiar gesture, and the bare-chested youth clutches the flute at her waist. Neither malicious nor older than the other performers, the gentle mother in the picture belies the Prince’s interpretation. Whereas the author’s cultural prejudices at times come through in his writings, his drawings seem to give a less inhibited, and perhaps more honest, presentation of what he saw. He continues, The elephants, there were four attached to these pagodas, created a superb effect against the background of palm trees and grass; and these men and women who filled my apartment were not without character. All the people, as one might expect, had come to get a few coins. I was forced to give handfuls of rupees to many hands. I went to one temple, then to another.

K anchi Under Colonialism

257

Illustration 99 L.H. Rudder after Prince Alexis Soltykoff, Intérieur du couvent de Condgeveram, à 40 milles de Madras, Réunion journalière des Brames en l’honneur des deux divinités Conservatrice et Destructive, Juillet 1841 [‘Inside a temple at Conjeveram, 40 miles from Madras, Daily reunion of the Brahmins in honor of the two deities of Preservation and Destruction, July 1841’], 1848, lithograph, 46 × 59.8 cm, British Library, P970

All were superb, magnificent, incomparable. The architecture was the most grandiose, the details the most extravagant, the throng of fantastic animals, all entangled with palm trees and enormous banyans, and so many galleries, colonnades, courtyards, forests, esplanades, filled with a crowd of Brahmins, elderly people, youths and children, nearly naked, besmeared in yellow and white on their foreheads and chests, some horizontal, serving Shiva, the destroyer, the others perpendicular, serving Vishnu, the preserver. All adore Brahma, the supreme being, the creator. The dancers were also there and they danced to the sound of their raucous music; the elephants followed me everywhere like shadows. The full ensemble was marvelous, and each detail was an attraction, a profound interest.163 163 Soltykov, Lettres Sur l’Inde, 70-71.

258 

Construc ting K anchi

The elephants are seen among a throng of people in a second lithograph, entitled Intérieur du couvent de Condjévéram, à 40 milles de Madras, Réunion journalière des Brames en l’honneur des deux divinités Conservatrice et Destructive, Juillet 1841 [‘Inside a temple at Conjeveram, 40 miles from Madras, Daily reunion of the Brahmins in honor of the two deities of Preservation and Destruction, July 1841’] (Ill. 99). The image depicts an assembly of Indian men and women of various caste and occupational status within the courtyard of a temple in Kanchi. Although the setting does not precisely conform to any one specific temple, it is most likely either the Ekāmbaranātha or the Varadarāja Perumāḷ. Not only were these the two temples most frequently visited by European travelers, but both contain the relatively rare form of pillared hall with a clerestory that is seen at the image’s left.164 Two tall elephants bearing the auspicious markings of Shiva and Vishnu on their foreheads stand amid shaven-headed priests in white vēṣṭis. To the left are dancers and musicians. The two camels reclining in the shadows at the lower right likely belong to the Prince’s entourage as messengers. Like other foreign artists, the Prince has exaggerated the height of the pillar figures and the sculptures inside the shrines, but he has also included details that reflect careful and sensitive observation. Among the crowd behind the shadowed camel, a woman casually balances a pot on her hip. Far from an exotic trope like Hodges’ and Hunter’s pot-carrying women, the naturalistic swerve of her body, the tilt of her head, and the way in which she slings her arm across the pot, create a specific posture that women in Tamil Nadu often take when carrying a vessel. The ritual illustrated, or something very close to it, is mentioned in collector Crole’s district report. Based largely on local histories and an indeterminate amount of eyewitness testimony, Crole’s report weaves together historical events with the Sthalapurāṇas (mythical temple histories) that his informants would have recounted to him from the Tamil texts and oral traditions. Even so, he cannot withhold his own judgments, and his opinions often interrupt the narrative flows. In his discussion of Kanchi, Crole records that the city’s largest festival is that of Garuda Sevai in which the god from the Varadarāja Perumāḷ temple is transported on various vāhanas (‘vehicle’) through the broad city streets.165 On the sixth day of the ten-day festival, which takes place in the Tamil month of Vaikāci (May/ June), Varadarāja Perumāḷ’s image is taken to visit Shiva at Ekāmbaranātha. 164 The Ekāmbaranātha’s clerestory only survives through archival images (see above). 165 Crole, The Chingleput, Late Madras, District: A Manual Compiled under the Orders of the Madras Government, by Charles Stewart Crole, of the Madras Civil Service, 115-116.

K anchi Under Colonialism

259

Since the Prince visited in July rather than May or June, he was not fortunate enough to witness the full festival. However, it seems that a similar procession that brought the two gods together was enacted on a daily basis and on a smaller scale at the time of his visit. The consolidation of this ritual into a single annual affair – Garuda Sevai – is typical of changes to ritual praxis that took place under colonial rule. So too is an insistence on the separation of Shiva and Vishnu. One of the ways in which the colonial government inscribed the fiction of its own necessity was to publicize sectarian rivalry. In 1786, priests from the Ekāmbaranātha and Varadarāja Perumāḷ temples were in conflict over rights to use certain streets for a festival procession.166 The local British officers became involved and chose to announce their decision by beating drums in what they perceived to be the traditional manner for Tamil rulers to make proclamations. In doing so, the colonial officers drew public attention to what otherwise probably would have remained little more than a petty dispute. The Prince continues with informative details not only about the material culture he encountered in Kanchi, but also the local food. They also showed me the temple’s jewels (footnote: Some Englishmen had recommended that I see them, as they were the only thing still ‘worth seeing’, worthy of making the effort to see in Conjeveram, all the rest was nothing but ‘native dirt and beastliness’; one of the many evidences of the little faith one should accord what one hears, because these objects were very little compared with the infinite interest of all the rest.) which are part of the costume of the idols that they carry from time to time on the chariots that I saw and that are extraordinarily large, made of astonishingly sculptured wood, formed of unfamiliar ornaments swarming with small, grotesque deities. There again it was necessary to hand out dozens of rupees. Besides this, I had no other expenses, because I am lodged and fed for nothing by the government or rather by Lord Elphinstone. One becomes a freeloader in this country, and it would be difficult for it to be otherwise, because if Lord Elphinstone had not had the generous attention to send me some products from his splendid French kitchen and a detachment of bottles from his cave, admirable and unique perhaps in India, to what would I find myself reduced? Coconuts, perhaps, because, moreover, the culinary preparations that one can find in the Conjeveram bazaar, 166 The incident is reported in a letter from the Jaghire Committee to the Resident in the Jaghire, preserved in the Tamil Nadu State Archives and discussed in Irschick, Dialogue and History: Constructing South India, 1795-1895, 21-22 and 211 n.26.

260 

Construc ting K anchi

or the fabulous dishes that one might pull from the Brahmin kitchens, are of a taste so opposite to all that we regard as edible, that anyone who is not Indian cannot think of it as food. Their food comprises fruits and vegetables, but prepared with such smells, such oils and sweetness to make one’s heart sink. Sometimes one believes one is eating the musk of lamp oil, sometimes the rancid pomade of carnation or of soap from Naples. Here then is a country where it is necessary to renounce doing as the natives do. I have scrupulously followed that rule in Turkey, in Italy, in Persia, and in Georgia, alas! And even in Sicily; but here, in India, there is no way.167

The Prince’s lack of enthusiasm for temple jewels in this passage stands in sharp contrast to more official genres of colonial-era reportage. For example, Crole paid great attention to the relative levels of wealth that each temple possessed. Unlike the Varadarāja Perumāḷ temple, which remains among the most prosperous of Kanchi’s temples, Ekāmbaranātha was by no means financially well-off by the latter half of the nineteenth century.168 Whereas Crole priced a still often-cited gift of a necklace from Lord Clive to the Varadarāja Perumāḷ temple at 3,682 rupees, he also recorded that 10,000 rupees worth of ornaments were stolen by dacoits from the Ekāmbaranātha temple in April 1845. The theft took place just four years after the Prince saw the temple bustling with devotees, dancers, musicians, and temple elephants.169 The detailed nature of Crole’s account of the theft and subsequent discovery of what may have been the stolen property in the house of a man in a nearby district suggests the events were true.170 Although it remains a popular devotional site, Ekāmbaranātha receives little financial support today compared with the other two of Kanchi’s largest temples. The Ekāmbaranātha’s vast size makes maintenance a perpetual challenge. Under restoration since the colonial era, a long-faded appeal for funds to repair the now-closed thousand-pillared hall and the courtyard that once contained a clerestory languishes on the temple’s official website.171 Comparison between photographs taken around 1900 and the present condition of the courtyard show significant further deterioration rather

167 Soltykov, Lettres Sur l’Inde, 71-72. 168 Crole, The Chingleput, Late Madras, District: A Manual Compiled under the Orders of the Madras Government, by Charles Stewart Crole, of the Madras Civil Service, 113. 169 Ibid., 115. Today the only temple in Kanchi to possess elephants is the now-prosperous Kāmākṣī Ammaṉ temple. 170 Ibid., 113. 171 www.ekambaranathartemple.org/needs.php.

K anchi Under Colonialism

261

than improvement.172 Perhaps had it not been mutilated into a fortress in the first place, the Ekāmbaranātha’s fate would have been different. It is testament to the strength of its mythological history and the integrity of its architecture that the temple has carried on as a major pilgrimage place and site of daily activity to as great an extent as it has done. Drawing his account of Kanchi to a close, the Prince describes the city further, giving particular attention to the local fauna. They maintain in Conjeveram quite the quantity of monkeys. One sees them in large numbers in the houses, on the roofs and inside the temples, and they are more or less venerated. The first thing that I saw upon arriving here, it was first the monkeys, next an immense pond surrounded by an elegant and vast staircase of granite. Beside this pond walked a respectable elephant. The elephants also have on their foreheads the horizontal and vertical marks, according to whether they are consecrated to Vishnu or Shiva. This one carried on his back a shaven man, a devil-worshipper, who held before him a copper vase with green leaves inside. Before them walked a few musicians, and between some others a young boy mounted on a cow, holding kettledrums, and banging them together as if possessed, evidently enchanted by his task; then there were many bells and two cymbals suspended on the elephant, and some kind of horn or trumpet. I stopped and exited my palanquin in order to contemplate this serious farce and wait for it to pass in front of me. It is more than time to finish. Adieu!173

The author, however, cannot stop, and he continues with a postscript, P.S. – I remained in Conjeveram until the evening. There was precisely a festival with a procession. Imagine this, in the middle of this strange architecture, illuminated by a hundred torches, the colossal, gold idol (it is silver, I think), appearing all of a sudden, decorated with flowers, on an immense scaffolding pulled by a crowd of men, and advancing by itself, in the midst of Brahmin people. Add to this, maddened musicians mounted on cows. First the idol made a long tour within the interior of the 172 Schier refers to a mid-nineteenth-century German travel account in which the author reports that Ekāmbaranātha was considerably less prosperous than the Kāmākṣī Ammaṉ and Varadarāja Perumāḷ temples. Karl Graul, Reise Nach Ostindien Über Palästina und Egypten von Juli 1849 bis April 1853: der Süden Ostindiens und Ceylon, 2. Abt., 5 vols., vol. 5 (Leipzig: Dörffling & Franke, 1856), 187; Schier, ‘The Goddess’s Embrace: Multifaceted Relations at the Ekāmranātha Temple Festival, Kanchipuram’, 56. 173 Soltykov, Lettres Sur l’Inde, 72-73.

262 

Construc ting K anchi

Illustration 100 The goddess in her chariot, Kāmākṣī Ammaṉ Temple, Kanchi (18 July 2014)

K anchi Under Colonialism

263

monastery and through its spacious courtyards; next it went out through the immense doorway surmounted by a tower that is higher and larger than the grand tower of Moscow, but older by four thousand years, they told me, and made of minutely sculptured granite. Thus it passed next into the streets of Conjeveram and into the forests, amid cries, chants, prostrations, and fireworks.174

The Prince does not specify which temple held the procession, nor does he mention which deity was celebrated. However, the event he witnessed is strikingly similar to a festival I experienced at the Kāmākṣī Ammaṉ temple. On just such an evening in the month of July, the Goddess was installed on a golden chariot and pulled by a crowd of temple Brahmins and devotees in a procession around the complex.175 The only difference was a reversal in the sequence of events. Whereas the Prince saw the deity circumambulate the temple and then exit into the streets, I saw her first make her procession through the city and then around the temple. However, just as the Prince described, Goddess Kāmākṣī emerged as if by magic, and, mounted on her chariot surrounded by a throng of people, she made a spectacular sight (Ill. 100).

Conclusion: Plastered Pasts The colonial travelers’ accounts and illustrations reveal the various ways in which Kanchi as a city has been continually transformed and renewed through entangled practices of renovation and preservation. Ongoing processes of renovation and reconstruction have kept temples vital, yet officially protected monuments have maintained important aspects of India’s rich aesthetic history. For some local devotees, preserving a temple entailed a good deal of reconstruction. At nearly the same time that ASI Superintendent Longhurst was documenting Orakaṭam’s apsidal Shiva temple, Tamil Shaiva Chettiar patrons were tearing down the Bhaktavatsaleśvara temple’s apsidal vimāna in the nearby town of Tirukkaḻukuṉṟam and rebuilding it completely, preserving the original apsidal form.176 At that same temple, the patrons did 174 Ibid., 73. 175 The festival took place on 18 July 2014. 176 Crispin Branfoot is actively researching reconstruction projects by the Nattukottai Chettiar community in Tamil temples. Branfoot, ‘Remaking the Past: Tamil Sacred Landscape and Temple Renovations’; ‘Renovations and Ruins: Collecting Sculpture in Early 20th-century Colonial

264 

Construc ting K anchi

not touch the also apsidal shrine for Caṇḍeśa, built under Pallava reign in the ninth century. The turn of the twentieth century brought about a wave of similarly sponsored reconstructions to the inner portions of temples. In Kanchi, the Ekāmbaranātha temple’s vimāna and first prākāra were rebuilt under Chettiar patronage in 1900, and a sign near the eastern gopura records the donation, just as inscriptions had in the past. Efforts at preserving Kanchi’s cultural heritage are inextricably linked with local devotional practices. During the colonial era, plaster assumed an unprecedented level of prominence in both arenas. People have plastered temples for the same reasons they plaster their homes – to seal and protect the surface from water and other effects of a monsoon climate.177 Today, one encounters a great quantity of plaster in various stages of deterioration on many of Kanchi’s temples, but it is especially prominent on the walls of the sandstone Pallava temples. Thick and intractable, this is not the coating that originally primed the surfaces for frescoes. The paintings that survive in the recesses of Kanchi’s Kailāsanātha and Paṉamalai’s Tāḷagirīśvara temples are instead laid over plaster that is eggshell-thin. The thicker plaster was applied at some time long after the temples had been completed, possibly in the service of repair. In the process, the plaster obscured and effaced the earlier details. At Kanchi’s Amareśvara temple, the surface is so thickly covered in plaster that the sculptural program is only known from limited archival sources.178 In 1903, Alexander Rea made drawings of Amareśvara and described Kanchi’s Pallava temples as including ample plaster, yet even more plaster has been added since then. The owners have filled up the hollows with brick and covered the figures with plaster. So complete, indeed, is their decay, that their successful preservation offers a problem of the gravest difficulty, and it is only by painstaking care that a satisfactory solution can be found to it.179

As part of its emerging importance, plaster became a key ingredient in colonial arts education during the nineteenth century. Supplies for the plaster of Paris used in the Madras Art School were gathered from the Madras Presidency’ (Paper presented at Reuniting the Tamil Yoginis: The Plans Take Shape, Colgate, 17 October 2020). 177 Srinivasan, ‘The Pallava Architecture of South India’, 118. 178 Gillet, La Création D’une Iconographie Śivaïte Narrative: Incarnations Du Dieu Dans Les Temples Pallava Construits, Plan IV. 179 Archaeological Survey of India Annual Report, Vol. 3 (1903-1904) (Calcutta: Superintendent of Government Printing, 1996), 64-65.

K anchi Under Colonialism

265

‘clayey estuarine beds’ north of Madras, where masses of gypsum and crystals of selenite are found.180 Casts of important archaeological sites were created and mobilized by British officials in an effort to collect and command India’s heritage. For example, in 1830, Babington praised the large Mahiṣāsuramardinī relief panel in Māmallapuram as, ‘the most animated piece of Hindu sculpture’ he had ever seen, and he recommended that a plaster cast be taken for the Royal Asiatic Society.181 The practice was widespread across colonial India. A photograph taken in 1870 depicts H.H. Cole observing a team of Indian artisans taking plaster casts of pillars at the Quwwat al-Islam Mosque in Delhi.182 That same year, a full-size plaster cast of the eastern toraṇa of the Great Stūpa at Sanchi was transported to London for display in the South Kensington Museum (now the V&A).183 Intended to serve as an emblem of grand imperial custodianship, the cast later languished in the bowels of museum storage. Throughout the ASI’s efforts at conservation, there has been a continuous oscillation between the application and removal of plaster. In 1973, the dead plaster on the eastern wall of the Tāḷagirīśvara temple was removed and reapplied, and the terrace was given two courses of flat tiles and further coats of plaster.184 At Uttiramērūr in 1986, damaged plaster on the sabhā (‘assembly hall’) was removed and the wall was replastered.185 At Kanchi’s Kailāsanātha temple in 1996, after nearly a decade of consistent plaster removal, there was an application of ‘a very thin coat of lime’ on the weathered outer surface of the prākāra, ‘as a protective measure, as was done in the past.’186 In 2000, the Kailāsanātha’s superstructure was entirely replastered, and then in 2008 that plaster was cleaned and more was applied. The notion of perpetually remaking and repairing represents a thread of continuity that cuts through a seemingly discontinuous colonial past. As patrons have done since at least the Chola era, families or foundations periodically donate sums of money in order to sponsor restorations to temples that have not come under official protection. For example, Kanchi’s 180 ‘Chingleput District’, in The Imperial Gazetteer of India, Provincial Series: Madras Presidency, Vol. I (Madras: Usha, [1908] 1985), 532. 181 Carr, Descriptive and Historical Papers Relating to the Seven Pagodas on the Coromandel Coast, 49. 182 Tapati Guha-Thakurta, ‘The Production and Reproduction of a Monument: The Many Lives of the Sanchi Stupa’, South Asian Studies 29, no. 1 (2013): 87. As Curator of Ancient Monuments in India, H.H. Cole was responsible for some of the earliest reports for the ASI. 183 Ibid., 85. 184 IAR, 1973-1974, p. 75. 185 IAR, 1986-1987, p. 161. 186 IAR, 1996-1997, p. 273.

266 

Construc ting K anchi

Airāvateśvara temple was completely revitalized through local sponsorship during 2014-2015.187 An important aspect of these renovations is the replastering and repainting of the gopuras and temple towers. The process typically concludes with a large festival, culminating in a mahā-kumbhabhiṣeka in which the pinnacle of the sanctum’s roof is adorned and lustrated with sacred substances. This ceremony effectively reconsecrates the temple, as if it were first being put into worship. As I prepared to depart from Kanchi in March 2016, the Kāmākṣī Ammaṉ temple was undergoing an extensive, multiyear renovation. The paving stones were to be relaid, the surfaces cleaned, and the gopuras replastered and painted. Although ritual activity would continue inside the temple throughout the duration of the work, the towers had to remain covered and the bronze festival icon was not to come out of her sanctum because the goddess ‘should not see the temple in such a condition.’188 As Sankar, a young Kāmākṣī Ammaṉ priest, told me, when the renovation would be complete, the whole temple would be ‘new.’189

187 The renovation was supported by a generous private donor. Work was administered by a local team organized by Anbarasu, a tea seller whose stall is next to the Airāvateśvara temple. 188 Pers. comm., Sankar, Kāmākṣī Ammaṉ Temple, 21 July 2015. 189 Ibid.

Epilogue The Living Temple Encounter It is a pleasant evening in October 2013, day four in my first week of fieldwork. I wander through the rain-soaked streets of Kanchi, where sprawling puddles make mud that sticks to my feet and shoes. I’m on my way to check on the Kauśikeśvara temple, a single-storey granite structure behind a gate that has been locked each time I visited. I wonder if it is ever open. The Archaeological Survey of India’s official green fence and blue signboard had already become my touchstones for preserved monuments, but none of the ASI monument attendants in Kanchi seemed to know anything about a temple with this name. The gate is locked as usual. ‘Kauśikeśvara’, as it is written on the banner over the entry to the temple grounds, is a local name. Cokkīśvara is the name written on the blue sign, which is tucked inconspicuously behind the locked gate. Cokkīśvara is indeed under the auspices of the ASI, but the temple boasts no monument attendant. It is situated in the front yard of a large house, and a privately hired watchman guards the property. While the banner presents the local title, the blue signboard posts the temple’s historic name. The date is given as the ninth century in both labels. However, the blue signboard attributes the temple to the reign of Uttama Cōḻa, who in fact ruled in the tenth century (ca. 969-985 CE).1 The confusion in names and dates that I encountered almost immediately upon arrival in Kanchi reveals some of the dynamics at play in a living city with temples that range in condition from preserved monuments to active ritual centers. Whereas some temples enjoy continuous endowments for their maintenance, the lives of others progress cyclically, with periods of relative disuse interleaved with periods of prominence. Some temples have 1 There is frequent confusion between the ninth century and the nine-hundreds, even in official documents in India.

Stein, Emma Natalya, Constructing Kanchi: City of Infinite Temples. Amsterdam, Amsterdam University Press 2021 doi: 10.5117/9789463729123_epil

268 

Construc ting K anchi

become all but stripped of their sacrality through conservation efforts by the ASI. Others have had their ritual functions diminished due to neglect rather than monumentification. Still others have endured as relatively private spaces for devotion. The largest and most popular temples in the city also have been subjected to change. Temple renovation remains a near-constant component of life in the city of one thousand temples. Restoration and preservation are often at odds. Individual temples that were once important can fall out of use for a variety of reasons. The priest who maintained the temple dies and no one takes up the task. Ownership of the property comes under dispute and gets locked in court. The family that usually gifts money for repairs has hardships or decides to allocate their funds differently. Under any of these circumstances, the temple is left at risk. It can effectively disappear both from sight and from common knowledge. Conversely, a large endowment can carry the potential to revitalize a temple that has fallen out of use.

Expansion Like its many temples, Kanchi’s role as a node within an extended pilgrimage and commercial network has fluctuated over the course of time. By contrast, the city’s wide-reaching reputation as a sacred center and a courtly-urban capital has remained remarkably consistent. After the close of the Chola period, Kanchi maintained its status as an important locus of multilingual literary production. This was especially true in post-Rāmānuja Śrīvaiṣṇava circuits.2 Foremost among authors was the prolif ic poet-philosopher, Vedānta Deśika (ca. 1268-1369 CE), who was born in Kanchi and is honored with a shrine next to the Viḷakkoḷi Perumāḷ temple.3 The city also functioned continuously as a model for courtly elites in far-reaching places that were connected to the subcontinent through maritime networks. Textual sources attest to connections between Kanchi and 2 Hopkins, ‘Lovers, Messengers, and Beloved Landscapes: Sandeśakāvya in Comparative Perspective’; Yigal Bronner, ‘Singing to God, Educating the People: Appayya Dīkṣita and the Function of Stotras’, Journal of the American Oriental Society 127, no. 2 (2007); Marcus Schmücker, ed. Viṣṇu-Nārāyaṇa: Changing Forms and the Becoming of a Deity in Indian Religious Traditions, Studies in Hinduism V (Vienna: 2016). 3 Frieldhelm Hardy, ‘The Philosopher as Poet: A Study of Venāntadeśika’s “Dehalīśastuti”, Journal of Indian Philosophy 7, no. 3 (1979); Shulman and Bronner, ‘“A Cloud Turned Goose’: Sanskrit in the Vernacular Millennium.’

Epilogue

269

Southeast Asia over the course of many centuries. A good example is the late fourteenth-century Deśavarṇana (Nagarakṛtāgama), which was composed in the court of king Rājasanagara (ca. 1350-1389 CE) in Majapahit, East Java.4 This text was unique among Old Javanese poetic texts (‘kakawin’) in that it was a eulogy of the king rather than the customary recreation of an episode from an Indian epic. Literally translatable as the ‘Description of the Lands’, the text is given a real-world setting that provides an insightful picture of fourteenth-century Majapahit as a consciously crafted, highly literary and artistic courtly culture that was fashioned on the basis of Indian models. One way for Mahajapahit’s upstart kings to establish their empire’s prestige was to claim connections with the specific courts they sought to emulate. Whereas numerous places are mentioned in the Deśavarṇana in order to claim Java’s importance on a global scale, the author deems only India and Java to be truly excellent.5 The use of eulogy, familiar from Indian literature but previously unknown in kakawin, marks a direct connection with the subcontinent in both content and literary form. In Canto 93, a Brahmin called Śrī Mutali Sahṛdaya – a Tamil name – is said to have brought his own eulogy for the king.6 So too has a Buddhist monk called Śrī Buddhāditya, who came to Java from none other than Kanchi. Although the Deśavarṇana includes Kanchi among the great lands with which the Majapahit kingdom is closely associated, the Kanchi that the text conjures is not the politically volatile fourteenth-century city, but rather the glorious courtly center of the Pallava past. The Majapahit king Rājasanagara fashioned himself after the renowned Pallava rulers who had used patronage of the literary, dramatic, and visual arts in order to establish the cosmopolitanism of their capital. In doing so, the East Javanese king may have preserved or reinstated practices and conventions that had gradually fallen out of 4 Mpu Prapañca, Deśawarṇana (Nāgarakṛtāgama) by Mpu Prapañca, trans. Stuart Robson, Verhandelingen van het Koninklijk Instituut voor Taal-, Land- en Volkenkunde 169 (Leiden: KITLV Press 1995). I am grateful to Hadi Sidomulyo for informing me of this important East Javanese reference to Kanchi, and for discussing it with me in long email exchanges (beginning August 2016). Monius has also noted this text in support of a Buddhist literary connection between South India and Java. Monius, Imagining a Place for Buddhism: Literary Culture and Religious Community in Tamil-Speaking South India, 110-111 and p. 205 n. 164. 5 Prapañca, Deśawarṇana (Nāgarakṛtāgama) by Mpu Prapañca, 85, canto 83.2. Places mentioned include India, China, and Cambodia. See especially Canto 83 in which foreigners arrive at Rājasanagara’s court. 6 Kern first identified this character’s name with the Veḷḷāḷa (landholding) caste in Tamil Nadu. H. Kern and N.J. Krom, Het Oud-Javaansche Lofdicht Nāgarakṛtāgama van Prapañca (1365 AD) (Weltevreden, Java: Drukkerij Volkslectuur, 1922). See Robson p. 148-149, note 1d.

270 

Construc ting K anchi

currency in the fragmented courts of fourteenth-century South India. Texts like the Deśavarṇana indicate some of the ways in which South India functioned in transregional networks of intellectual, cultural, and material exchange. In concluding this book, I would suggest that future lines of inquiry continue to examine the local and transregional contexts that enabled Kanchi to hold such an enduring legacy, and on so vast a geographical and temporal scale. More research remains to be done inside Kanchi itself. The large corpus of inscriptions has yet to be studied in full. More extensive ethnographies of Kanchi’s present-day communities would also provide fruitful lines of inquiry. The temples are intimately connected with the keepers of local crafts traditions who are responsible for making the ritual objects, weaving flower garlands, and stocking the local bazaars. The movement of objects into and out of the sacred complexes is yet another way in which the temples engage their city and the wider world. Kanchi has long been known for its role in the global textile trade. Although the city has been associated with weaving for at least a millennium, a detailed study of the industry remains to be undertaken.7 Kanchi is mentioned as a place famed for fine textiles in early Sanskrit and Tamil literature, such as the ca. sixth-century Sanskrit poem, Jānakīharaṇa by Kumāradāsa.8 However, the present-day tradition is dominated by craftsmen who trace their origins to Zoroastrian silk weaving families (‘paṭṭunūlkārar’) that emigrated to Kanchi from western India more recently.9 Some of Kanchi’s priests also claim distant regional origins. A pūjāri at the Ekāmbaranātha temple says it was a Maharaja of Mysore who brought his family to Kanchi and installed them at the temple.10

7 Nesa Arumugam, Silk Sarees of Tamil Nadu (New Delhi: Abhinav Publications, 2011). See especially Chapter 8, ‘The Noble Kanchipuram’, p. 123-135. 8 This best of men enjoyed the southern quarter where the snakes sought after its honored shores and merchant (-caravans) were attracted by the ( fine) threads of Kāñcī (textiles) and profits were made by efforts of the sword/difficult efforts, like a skillful courtesan from whom johns request appointments of pleasure, to whom the traders are attracted by her beltstrings, and who can be enjoyed by the efforts of the sword. (Jānakīharaṇa 1.18). My thanks to Victor d’Avella for providing the reference and translation. 9 The renowned weaver Krishnamoorthy’s family came to Kanchi during the Vijayanagara period. I am grateful to him for showing me the looms and discussing his history with me (February 2016). 10 Nagaswamy Aiyyer (January 2014). My deep gratitude to Nagaswamy Aiyyer for many interesting conversations on his verandah, and for his patience as my Tamil slowly improved.

Epilogue

271

Continuation It is getting dark on that same October evening, and the ASI monuments will now certainly be closed. I walk down a small street behind the Kāmākṣī Ammaṉ temple and encounter a fruit seller who offers items that are colorful, round, ridged, and prickly. I buy some bananas and green citrus fruits – ‘“sweet lemons’, a customer informs me, raising an eyebrow as if she were letting me in on an exquisite secret. I carry the fruits in a plastic bag towards Babu’s house, which is situated just beyond the eastern corner of the Kāmākṣī Ammaṉ temple complex.11 The house is painted pistachio green and is fronted by a small verandah with a gate that is nearly always open. Babu and his family are important priests at the Kāmākṣī Ammaṉ temple. They comprise one of the two lineages responsible for maintaining temple rituals. Babu is at home this evening, and his younger brother Ramesh welcomes me at the door. Babu tells me to sit after accepting the fruits in the plastic bag that I offer him and passing them along to his wife, who takes them into the kitchen. ‘I have come to thank you’, I say. A couple of nights earlier, Babu had allowed me inside the Kāmākṣī Ammaṉ temple, where foreigners were generally not permitted. It had been the final day of Śarada Navarātri, the nine-day goddess festival that culminates with a special pūjā for Sarasvatī, goddess of music and knowledge, in Tamil Nadu. The sanctum for the festival bronze of Kāmākṣī Ammaṉ was flung open, and the goddess, mounted on a swing inside, was outfitted with the attributes of Sarasvatī and lovingly adorned with flower garlands in pink, white, and saffron orange. A group of musicians seated opposite the sanctum adored the goddess and thrilled the crowd with fervent devotional music. ‘So, what is your project?’ Babu asked me with great interest. We discussed Kanchi’s Pallava and Chola temples in as much Tamil as I could manage so early in my fieldwork. In moments of silence he just stared at me with a little smile. His eyes were disarmingly kind – so much so that I couldn’t think of anything to say in any language. At some point I turned the conversation to him. ‘How many generations has your family been pūjāris at this temple?’ I asked. His eyes sparkled and he sat back in his chair and folded his arms across his bare chest and Brahmanical thread. ‘Since before the time of the Rāmāyaṇa’, was his reply.

11 T.R. Dhakshinamurthy Sastrigal (‘Babu’).

272 

Construc ting K anchi

Babu passed away in the Tamil month of Kārttikai, December 2015, and his son has since entered the priesthood. I take my leave and decide to circumambulate the Kāmākṣī Ammaṉ temple since I am right nearby. The gopura is still brightly lit from the Navarātri festival. Above the main sanctum that shelters the stone mūlavar of Kāmākṣī Ammaṉ, the golden superstructure glows in the flickering lights. While I am sitting on the high steps of the vasanta maṇḍapa, writing in my notebook, a Kanchi resident named Narayanan Subramanian, whom I had met through Babu, calls me on my cell phone. He tells me to go to the Ekāmbaranātha temple immediately, where a pūjāri named Nagaswamy Aiyyer is expecting my arrival. I quickly return to the fruit seller and purchase some more bananas before hailing an auto-rickshaw. The driver drops me in front of the Ekāmbaranātha’s large southern gopura, where Thomas Daniell and James Wathen had stood sketching a century earlier. Entering the vast complex, I soon find Nagaswamy Aiyyer standing inside the cloistered second prākāra amid a group of younger acolytes. He accepts my offerings and we converse more easily in his style of Tamil, which makes frequent use of Sanskrit terminology in discussions of temples and ritual. Several days later, the spectacular Ati Rudra mahāyajña would commence. The patron was a prominent Hyderabad family that regularly sponsors similarly impressive multiday festivals in various temples throughout the year. The Ekāmbaranātha would be bedecked with flowers, palm fronds, bronze lamp stands, and platters of sacred items. Priests from temples not only in Kanchi but also its surrounding areas would gather inside the prākāra near the shrine of the sacred mango tree to recite the Veda and offer a wide array of sacred substances into fire altars. For the conclusion of the ten-day pūjā, the altars were to be dismantled and the coconuts, flowers, fruits, and medicinal herbs carried in a grand procession around the temple. At the culmination, the priests, patrons, and assembled visitors would accompany the sacred items into the inner sanctum, to present the offerings to Kanchi’s gods. At that time, I did not know that the Kāpālikas once had covered themselves in ashes and rolled on the Ekāmbaranātha temple’s floors, nor that in the middle of the eighteenth century, Clive of India had sought refuge within its walls. Ekāmbaranātha existed for me in the present, as a living site of devotion. For devotees everyday, the Ekāmbaranātha functions as the place for the perpetual recapitulation of Shiva and Umā’s marriage beneath the sacred mango tree. This mythological history enables the temple to connect with an era as distant in the past as when Babu’s family

Epilogue

273

began their service at the Kāmākṣī Ammaṉ temple, and to bring that era continually into the present. Simultaneously, a tangible connection with the historical past undergirds the structural fabric of the temple, in the form of fragmentary sculptures, inscriptions, pillared halls, and other evidence of changes to the complex. The same is true in so many of Kanchi’s temples. As active centers of devotion, protected monuments, or sites of local worship, Kanchi’s temples participate in the continuous reimagining of their city.

Bibliography Abbreviations ARE IAR PP SII

Annual Reports on Indian Epigraphy: 1887-1905 ‘Indian Archaeology: A Review’ Periya Purāṇam (in Tamil) South Indian Inscriptions

Primary Sources Abhayadatta. Buddha’s Lions: The Lives of the Eighty-Four Siddhas (Caturaśīti-SiddhaPravṛtti). Translated by James B. Robinson. Berkeley: Dharma Publishing, 1979. Akananuru – the Akam Four Hundred. Translated by A. Dakshinamurthy. 3 vols. Thiruchirapalli: Bharathidasan University, 1999. Aṭikal, Iḷaṅkō. The Cilappadikāram. [‘The Tale of the Anklet’]. Translated by V.R. Ramachandra Dikshitar. Madras: The South India Saiva Siddhanta Works Publishing Society, 1978. Balbi, Gasparo. ‘Viaggio dell’Indie Orientali di Gasparo Balbi, Gioielliero Venetiano (1579-1588)’, in Viaggi di C. Federici e G. Balbi alle Indie Orientali, edited by Olga Pinto, 69-234, 95-366. Rome: Instituto Poligrafico dello Stato, 1962. Bhatt, A. Vishnu, and Michael Lockwood. Mattavilāsa Prahasana: ‘The Farce of Drunken Sport’ by King Mahendravarma Pallava. Madras: Christian Literature Society, 1981. Burgess, J. ‘The Temple of Kailāsanātha.’ Journal of the Royal Asiatic Society of Great Britain and Ireland (1891): 337. Burke, Edmund. A Philosophical Enquiry into the Origin of our Ideas of the Sublime and Beautiful. Oxford: Oxford University Press, (1759) 2014. Cāminātaiyar, U.Vē., ed. Pattuppāṭṭu. Madras: Kabeer Printing Works, 1956. Carr, M.W., ed. Descriptive and Historical Papers Relating to the Seven Pagodas on the Coromandel Coast. New Delhi: Asian Educational Services, 1869. Reprint, 1984. Cēkkiḻār. The History of the Holy Servants of the Lord Śiva: A Translation of the Periya Purāṇam of Cēkkiḻār. Translated by Alastair McGlashan. Victoria, Canada: Trafford Publishing, 2006. —. Periya Purāṇam. [in Tamil]. Madras: C.K. Subiramaniya Mudaliyar, 1950. —. St. Sekkizhar’s Periya Purāṇam. Translated by T.N. Ramachandran. 2 vols. Vol. 1, Thanjavur: Tamil University, 1990.

276 

Construc ting K anchi

—. St. Sekkizhar’s Periya Purāṇam. Translated by T.N. Ramachandran. 2 vols. Vol. 2, Thanjavur: Tamil University, 1995. Clive, Lady Charlotte Florentia. Journal of Lady Charlotte Florentia Clive (1787-1866), Copied by Professional Calligrapher, W.H. Ramsay. British Library, Visual Arts Collection, WD4235, 1857. Crole, Charles Stewart. The Chingleput, Late Madras, District: A Manual Compiled under the Orders of the Madras Government, by Charles Stewart Crole, of the Madras Civil Service. Madras: Lawrence Asylum Press, 1879. Dagens, Bruno, ed. Mayamatam: Treatise of Housing, Architecture, and Iconography. 2 vols. Vol. 1-2. New Delhi and Delhi: Indira Gandhi National Centre for the Arts and Motilal Banarsidass Publishers Pvt. Ltd., 2007. Dandin. Daśakumāracarita. [‘What Ten Young Men Did’]. Translated by Isabelle Onians. Clay Sanskrit Library. New York: New York University Press, 2005. Daniell, William. ‘William Daniell’s Diary.’ In Walker’s Quarterly: Thomas Daniell, R.A.; William Daniell, R.A., edited by Martin Hardie and Muriel Clayton, 3-106, 1932. ‘Digital Tēvāram/Kaṇiṉit Tēvāram [Cd-Rom]’, edited by Jean-Luc Chevillard and S.A.S. Sharma. Pondicherry: École Française d’Extrême-Orient/Institut Français de Pondichéry, 2007. Fergusson, James. Illustrations of the Rock Cut Temples of India. London: John Weale, 1845. —. Picturesque Illustrations of Ancient Architecture in Hindostan, by James Fergusson. London: J. Hogarth, 1847. —. Diary of James Fergusson During Tours in India, When Collecting the Materials for his Picturesque Illustrations of Ancient Architecture in Hindostan: 18 June5 July 1837 and 26 August 1838-22 March 1839. British Library, Western Manuscripts Collection, ADD MS 35282, 1848. Fergusson, James, and James Burgess. The Cave Temples of India. London: W.H. Allen & Co., 1880. Les Français dans l’Inde: Dupleix et Labourdonnais, Extraits du Journal d’Anandarangappoulé, 1736-48. Translated by Julien Vinson. Paris: Ernest Leroux, 1894. Goodall, Dominic. The Parākhyatantra: A Scripture of the Śaiva Siddhānta. Pondicherry: École Française d’Extrême-Orient, 2004. Graul, Karl. Reise Nach Ostindien Über Palästina und Egypten von Juli 1849 bis April 1853: der Süden Ostindiens und Ceylon, 2. Abt. 5 vols. Vol. 5, Leipzig: Dörffling & Franke, 1856. Kaccālaiyar. ‘Kāñcippurāṇam.’ In Āḷavantār Ātīṉak Kaccālaiyar Iyaṟṟiya Kāñcippurāṇam (Kampar Purāṇam), edited by S.K. Irāmarājaṉ. Chennai: Makāmakōpāttiyāya Ṭākṭar U. Vē. Cāminātaiyar Nūlnilaiyam, 1983.

Bibliogr aphy

277

Kāñcīmāhāt mya (Śrī Skāndapurāṇāntargataṃ Śrīkāñcīmāhāt myam Srīrudrakoṭimahimādarśaḥ). Viijayavāḍa, Madras, Haidarabad: Vēṃkaṭrāma aṇḍ kō, 1967. Kern, H., and N.J. Krom. Het Oud-Javaansche Lofdicht Nāgarakṛtāgama van Prapañca (1365 AD). Weltevreden, Java: Drukkerij Volkslectuur, 1922. The Life of Hiuen-Tsiang, by the Shaman Hwui Li. London: Kegan Paul, Trench, Trubner, 1911. Mackenzie, Colin. ‘The Jains.’ Asiatic Researches 9 (1809): 272-278. Madras District Gazetteers: Statistical Appendix for Chingleput District. Madras: Superintendent, Government Press, 1915. Martineaux, Alfred. Dupleix et l’Inde Française: Vol. 3, 1749-1754. 3 vols. Paris: Société d’Éditions Géographiques, Maritimes et Coloniales, 1927. Meyer, William Stevenson, Sir. ‘Chingleput District.’ In The Imperial Gazetteer of India, Provincial Series: Madras Presidency, Vol. I, 522-562. Madras: Usha, 1908. Reprint, 1985. —. ‘Conjeeveram Tāluk and Conjeeveram Town.’ In The Imperial Gazetteer of India: Vol. X, Central Provinces to Coompta, 376-379. Oxford: Clarendon Press, 1908. Orme, Robert. A History of the Military Transactions of the British Nation in Indostan: From the Year MCDDXLV. 2 vols. Vol. I-II. London: Printed for John Nourse, Bookseller in Ordinary to his Majesty, 1799. Pattuppāṭṭu. [‘The Ten Tamil Idylls’]. Translated by A. Dakshinamurthy. Kattankulattur: Tamil Perayam, SRM University, 2012. Prapañca, Mpu. Deśawarṇana (Nāgarakṛtāgama) by Mpu Prapañca. Translated by Stuart Robson. Verhandelingen van het Koninklijk Instituut voor Taal-, Land- en Volkenkunde 169. Leiden: KITLV Press, 1995. Salt, Henry. Twenty-Four Views in St. Helena, the Cape, India, Ceylon, the Red Sea, Abyssinia and Egypt. London: William Miller, 1809. Saltykoff, Prince Alexis. Pis’ma Ob Indii. Moskva: Nauka, (1851) 1985. Shattan. Maṇimēkalai. [‘The Dancer with the Magic Bowl’]. Translated by Alain Danielou and T.V. Gopala Iyer. India: Penguin Books, 1993. Shields, Nancy K., ed. Birds of Passage: Henrietta Clive’s Travels in South India, 1798-1801. London: Eland, 2009. Soltykov, Prince Alexis. Lettres Sur l’Inde. Paris: Amyot, 1848. Śrīkāmākṣīvilāsaḥ. Bangalore: Bhāratalakṣmī Mudraṇālayam, 1968. Sundberg, Jeffrey, and Rolf W. Giebel. ‘The Life of the Tang Court Monk Vajrabodhi as Chronicled by Lü Xiang (呂向): South Indian and Śrī Laṅkān Antecedents to the Arrival of the Buddhist Vajrayāna in Eighth-Century Java and China.’ Pacific World: Journal of the Institute of Buddhist Studies 13 (2011): 129-222. Tamil Poetry through the Ages: Pattuppāṭṭu. [‘Ten Idylls’]. Translated by A. Rajarathnam. 2 vols. Vol. 2. Chennai: Institute of Asian Studies, 2000.

278 

Construc ting K anchi

Valentia, George Viscount. Voyages and Travels to India, Ceylon, the Red Sea, Abyssinia, and Egypt in the Years of 1802, 1803, 1804, 1805, and 1806. 4 vols. Vol. 1 (Text) and 4 (Plates). London: Rivington, 1811. Wathen, James. Journal of a Voyage, in 1811 and 1812, to Madras and China; Returning by the Cape of Good Hope and St. Helena. London: J. Nichols, Son, and Bentley, 1814. Xuanzang. Si-Yu-Ki. Buddhist Records of the Western World. Translated from the Chinese of Hiuen Tsiang (A.D. 629) by Samuel Beal. 2 vols. Vol. 2. London: Trubner & Co., (1884) 1969. Yule, Sir Henry, and A.C. Burnell. Hobson-Jobson: A Glossary of Colloquial AngloIndian Words and Phrases, and of Kindred Terms, Etymological, Historical, Geographical and Discursive. London: J. Murray, 1903.

Epigraphic and Archaeological Sources Achythan, Hema, K. Ashok Kumar, Sunil K. Tiwari, and Stephen Norwick. ‘A Reconnaissance Study of Tafoni Development, Exfoliation, and Granular Disintegration of Natural and Artificual Rock Surfaces in the Coastal and Lowland Regions of Tamil Nadu, Southern India.’ Zeitschrift für Geomorphologie 54, no. 4 (2010): 491-509. Annual Reports on Indian Epigraphy: 1887-1905. Delhi: Archaeological Survey of India. Archaeological Survey of India Annual Report, Vol. 3 (1903-1904). Calcutta: Superintendent of Government Printing, 1996. Balasubrahmanyam, S.R. Middle Chola Temples: Rajaraja I to Kulottunga I (AD 985-1070). Amsterdam: Oriental Press, 1975. Dayalan, D. ‘Punceri, an Ancient Seaport in Tamil Nadu: New Light on the Location of the Ancient Seaport of Mamallapuram.’ (unpublished). Evans, Damian, Christophe Pottier, Fletcher Roland, Scott Hensley, Ian Tapley, Anthony Milne, and Michael Barbetti. ‘A Comprehensive Archaeological Map of the World’s Largest Preindustrial Settlement Complex at Angkor, Cambodia.’ Proceedings of the National Academy of Sciences of the United States of America 104, no. 36 (2007): 14277-14282. Foote, Robert Bruce. On the Occurrence of Stone Implements in Lateritic Formations in Various Parts of the Madras and North Arcot Districts. Madras: Graves, Cookson and Co., 1865. —. Prehistoric and Protohistoric Antiquities of India. Delhi: Leeladevi Publications, 1916. Gangopadhyay, S. ‘Monuments of India Built through Ages with Special Reference to the Construction Material.’ In Proceedings of an International Symposium Organized by the Greek National Group of Iaeg: The Engineering Geology of Ancient

Bibliogr aphy

279

Works, Monuments and Historical Sites, Preservation and Protection, edited by Paul G. Marinos and George C. Koukis, 583-591. Athens: A.A. Balkema, 1998. Hultzsch, E., ed. South Indian Inscriptions, Vol. 1: Tamil and Sanskrit, from Stone and Copper Plate Edicts at Mamallapuram, Kanchipuram, in the North Arcot District, and Other Parts of the Madras Presidency, Chiefly Collected in 1886-87. Madras: Archaeological Survey of India, 1890. —, ed. Epigraphia Indica, Vol. 3: 1894-95. Delhi: Archaeological Survey of India, 1979. —, ed. South Indian Inscriptions, Vol. 2, Part 3: Tamil Inscriptions in the Rajarajesvara Temple at Tanjavur and Other Miscellaneous Records of the Chola, Pallava, Pandya and Vijayanagara Dynasties, New Imperial Series. Madras: Archaeological Survey of India, 1992. ‘Indian Archaeology: A Review.’ New Delhi: Director General, Archaeological Survey of India, 1953-2005. Jayanthi, D. ‘Conservation of Sandstone Monuments at Kanchipuram Tamilnadu India.’ PhD diss., Anna University, 2007. Krishna Sastri, Rao Bahadur H., ed. Epigraphia Indica, Vol. 17. New Delhi: Archaeological Survey of India, 1923. —, ed. South Indian Inscriptions, Vol. 2, Part 5: Tamil Inscriptions in the Rajarajesvara Temple at Tanjavur and Other Miscellaneous Records of the Chola, Pallava, Pandya and Vijayanagara Dynasties, Archaeological Survey of Southern India Series. Madras: Archaeological Survey of India, 1983. —, ed. South Indian Inscriptions, Vol. 4: Miscellaneous Inscriptions from the Tamil, Telugu, and Kannada Countries and Ceylon. New Delhi: Archaeological Survey of India, 1986. Krishnamacharlu, C.R., ed. South Indian Inscriptions, Vol. 12: The Pallavas. Madras: Government Press, 1943. Kuppaswamy, Bahour S., and G. Vijayavenugopal, eds. Pondicherry Inscriptions, Part 1: Introduction and Texts with Notes. Pondicherry: École Française d’ExtrêmeOrient and Institut Français de Pondichéry, 2006. Longhurst, A.H. ‘Pallava Architecture, Part II (Intermediate or Māmalla Period).’ Memoirs of the Archaeological Survey of India 33 (1928). —. ‘Pallava Architecture, Part III (the Later or Rājasiṃha Period).’ Memoirs of the Archaeological Survey of India 40 (1930): 1-27 and Plates I-XIII. Mahalingam, T.V., ed. Inscriptions of the Pallavas. Delhi: Agam Prakashan, 1988. —. A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Chingleput District. 8 vols. Vol. 3. New Delhi: Indian Council of Historical Research, 1989. —. A Topographical List of the Inscriptions in the Tamil Nadu and Kerala States: Tiruchchirappalli District. 8 vols. Vol. 8. New Delhi: Indian Council of Historical Research, 1991.

280 

Construc ting K anchi

Majumdar, Ramesh Chandra, and Anant Sadashiv Altekar. The Vākātaka-Gupta Age: Circa 200-550 A.D. Delhi: Motilal Banarsidass, (1967) 1986. Minakshi, C. Administration and Social Life under the Pallavas. Madras: University of Madras, 1938. —. ‘The Historical Sculptures of the Vaikuṇṭhaperumāḷ Temple, Kāñcī.’ Memoirs of the Archaeological Survey of India 63 (1941): 1-61 and Plates I-XXIV. —. ‘Buddhism in South India.’ In South Indian Studies, Vol. II, edited by R. Nagaswamy, 83-131. Madras: Society for Archaeological, Historical, and Epigraphical Research, 1979. Newman, Richard. The Stone Sculpture of India: A Study of the Materials Used by Indian Sculptors from ca. 2nd century BC to the 16th century. Cambridge: Harvard University Art Museums, 1984. Nilakanta Sastri, K.A. The Cōḻas. Madras: University of Madras, (1935) 1955. —. Foreign Notices of South India: From Megasthenes to Ma Huan. Madras: University of Madras, 1972. —. A History of South India: From Prehistoric Times to the Fall of Vijayanagar. Madras: Oxford University Press, 1975. Rao, T.A. Gopinatha. ‘Bauddha Vestiges in Kanchipura.’ Indian Antiquary 44 (June 1915): 127-129. Rea, Alexander. Pallava Architecture. Vol. XXIV. Madras: Archaeological Survey of India, New Imperial Series, 1909. Rice, Lewis B., ed. Epigraphia Carnatica, Vol. 12: Inscriptions of the Tumkur District. Mysore: Government Central Press, 1904. Singanenjam, Dr. ‘Pātai Māṟiya Pālāṟu.’ Tiṉamaṇi 72 (23 July 2005): 1. Srinivasan, K.R. ‘The Pallava Architecture of South India.’ Ancient India: Bulletin of the Archaeological Survey of India 14 (1958): 114-138. —. Cave-Temples of the Pallavas. Delhi: Archaeological Survey of India, 1964. —. ‘Āndhras, Ikṣvākus, and Literary Sources.’ In Encyclopaedia of Indian Temple Architecture, edited by Michael W. Meister and M.A. Dhaky, 3-22. New Delhi: American Institute of Indian Studies, 1983. —. ‘Cōḻas of Thanjavur: Phase II.’ In Encyclopaedia of Indian Temple Architecture, edited by Michael W. Meister and M.A. Dhaky, 223-260. New Delhi: American Institute of Indian Studies, 1983. —. ‘Cōḻas of Thanjavur: Phase III.’ In Encyclopaedia of Indian Temple Architecture, edited by Michael W. Meister and M.A. Dhaky, 289-330. New Delhi: American Institute of Indian Studies, 1983. —. ‘Pallavas of Kāñcī: Phase I.’ In Encyclopaedia of Indian Temple Architecture, edited by Michael W. Meister and M.A. Dhaky, 23-80. New Delhi: American Institute of Indian Studies, 1983.

Bibliogr aphy

281

—. ‘Pallavas of Kāñcī: Phase II.’ In Encyclopaedia of Indian Temple Architecture, edited by Michael W. Meister and M.A. Dhaky, 87-106. New Delhi: American Institute of Indian Studies, 1983. Subrahmanya Aiyer, K.V., ed. South Indian Inscriptions, Vol. 6: Miscellaneous Inscriptions from the Tamil, Telugu and Kannada Countries, Archaeological Survey of India, New Imperial Series. Madras: Archaeological Survey of India, 1928.

Secondary Sources Abu-Lughod, Janet L. Before European Hegemony: The World System A.D. 1250-1350. New York: Oxford University Press, 1991. Acri, Andrea. ‘Birds, Bards, Buffoons and Brahmans: (Re-)Tracing the Indic Roots of Some Ancient and Modern Performing Characters from Java and Bali.’ Archipel 88 (2014): 13-70. —, ed. Esoteric Buddhism in Mediaeval Maritime Asia. Singapore: ISEAS Publishing, 2016. —. ‘Kāpālikas.’ and ‘Pāśupatas.’ In Hinduism and Tribal Religions, edited by Pankaj Jain, Rita Sherma and Madhu Khanna, 1-6. Dordrecht: Springer Netherlands, 2020. Ali, Daud. ‘Royal Eulogy as World History: Rethinking Copper-Plate Inscriptions in Cōla India.’ In Querying the Medieval: Texts and the History of Practices in South Asia, edited by Ronald Inden, Jonathan Walters, and Daud Ali, 165-229. Oxford: Oxford University Press, 2000. —. Courtly Culture and Political Life in Early Medieval India. Cambridge: Cambridge University Press, 2004. —. ‘Aristocratic Body Techniques in Early Medieval India.’ In Rethinking a Millennium: Perspectives on Indian History from the Eighth to the Eighteenth Century, Essays for Harbans Mukhia, edited by Rajat Datta, 68-93. Delhi: Aakar Books, 2008. —. ‘The Epigraphical Legacy at Gangaikondacholapuram: Problems and Possibilities.’ In New Dimensions in Tamil Epigraphy, edited by Appasamy Murugaiyan, 3-34. Chennai: Cre-A, 2012. Ali, Daud, and Emma Flatt, eds. Garden and Landscape Practices in Precolonial India: Histories from the Deccan. Delhi: Routledge, 2012. Appadurai, Arjun. Worship and Conflict under Colonial Rule: A South Indian Case. Cambridge: Cambridge University Press, 1981. Archer, Mildred. Early Views of India: The Picturesque Journeys of Thomas and William Daniell, 1786-1794. London: Thames and Hudson, 1980. Arumugam, Nesa. Silk Sarees of Tamil Nadu. New Delhi: Abhinav Publications, 2011.

282 

Construc ting K anchi

Bautze-Picron, Claudine. The Forgotten Place: Stone Images from Kurkihar, Bihar. New Delhi: Archaeological Survey of India, 2015. Blake, David. ‘Colin Mackenzie: Collector Extraordinary.’ British Library Journal, no. 21 (1991): 128-150. Bopearachchi, Osmund. ‘Sites Portuaires et Emporia de l’Ancien Sri Lanka.’ Arts Asiatiques 54 (1999): 5-23. Branfoot, Crispin. ‘Approaching the Temple in Nayaka-Period Madurai: The Kūṭal Aḻakar Temple.’ Artibus Asiae 60, no. 2 (2000): 197-221. —. ‘Processions and Presence: Bronze Sculptures from the Temples of Southern India.’ Arts of Asia 36, no. 6 (2006). —. Gods on the Move: Architecture and Ritual in the South Indian Temple. London: Society for South Asian Studies, British Academy, 2007. —. ‘Review of Illustrating India: The Early Colonial Investigations of Colin Mackenzie (1784-1821).’ Bulletin of the School of Oriental and African Studies 75, no. 1 (2012): 179-181. —. ‘Remaking the Past: Tamil Sacred Landscape and Temple Renovations.’ Bulletin of the School of Oriental and African Studies (January 2013): 1-27. —. ‘Renovations and Ruins: Collecting Sculpture in Early 20th-Century Colonial Madras Presidency.’ Paper presented at Reuniting the Tamil Yoginis: The Plans Take Shape, Colgate, 17 October 2020. Bronner, Yigal. ‘Singing to God, Educating the People: Appayya Dīkṣita and the Function of Stotras.’ Journal of the American Oriental Society 127, no. 2 (2007): 1-18. Buddle, Anne. Tigers Round the Throne: The Court of Tipu Sultan (1750-1799). London: Zamana Gallery, 1990. —. The Tiger and the Thistle. Edinburgh: National Gallery of Scotland, 1999. Buswell, Robert E., and Donald S. Lopez. The Princeton Dictionary of Buddhism. Princeton: Princeton University Press, 2014. Calvino, Italo. Invisible Cities. New York: Harcourt, Inc., 1972. Chakravarthy, Pradeep. ‘Era of the Opera.’ The Hindu, 13 February 2011. Champakalakshmi, R. ‘The Urban Conf igurations of Toṇḍaimaṇḍalam: The Kāñcīpuram Region, C.A.D. 600-1300.’ In Urban Form and Meaning in South Asia: The Shaping of Cities from Prehistoric to Precolonial Times, edited by Doris Meth Srinivasan and Hugh Urban, 185-207. Washington: National Gallery of Art, 1993. —. Trade, Ideology, and Urbanization: South India 300 BC to AD 1300. Delhi: Oxford University Press, 1999. —. Religion, Tradition, and Ideology: Pre-Colonial South India. New Delhi: Oxford University Press, 2011.

Bibliogr aphy

283

Chanchani, Nachiket. Mountain Temples and Temple Mountains: Architecture, Religion, and Nature in the Central Himalayas. Global South Asia. Washington, WA: University of Washington Press, 2019. Chandramouli, C. Census of India 2001: Temples of Tamil Nadu, Kancheepuram District. Tamil Nadu: Directorate of Census Operations, 2001. Charya, P.B.A., ed. Kāñcimāhātmya. Kanchipuram: Sri Sudarsana Press, 1906. Chaudhary, Zahid R. Afterimage of Empire: Photography in Nineteenth-Century India. Minneapolis, MN: University of Minnesota Press, 2012. Chaudhuri, K.N. Trade and Civilisation in the Indian Ocean: An Economic History from the Rise of Islam to 1750. Cambridge: Cambridge University Press, 1985. Christie, Jan Wisseman. ‘Javanese Markets and the Asian Sea Trade Boom of the Tenth to Thirteenth Centuries AD.’ Journal of Economic and Social History of the Orient 41, no. 3 (1998): 344-381. —. ‘The Medieval Tamil-Language Inscriptions in Southeast Asia and China.’ Journal of Southeast Asian Studies 29, no. 2 (1998): 239-268. Coedes, George. The Indianized States of Southeast Asia. Hawaii: East-West Center Press, 1968. Coomaraswamy, Ananda K. ‘Early Indian Architecture: IV. Huts and Related Temple Types.’ RES: Anthropology and Aesthetics 15 (Spring 1988): 5-26. Davis, Richard H. Ritual in an Oscillating Universe: Worshiping Śiva in Medieval India. Princeton, NJ: Princeton University Press, 1991. —. ‘The Story of the Disappearing Jains: Retelling the Śaiva-Jain Encounter in Medieval South India.’ In Open Boundaries: Jain Communities and Culture in Indian History, edited by John E. Cort, 213-224. Albany, NY: State University of New York Press, 1998. de Certeau, Michel. The Practice of Everyday Life. Translated by Steven Rendall. Berkeley: University of California Press, 1984. de Moor, J.E., and Paul van der Velde, eds. De Werken Van Jacob Haafner, Vol 3. Zutphen: Walburg Pers., 1997. de Pencier, Honor. Posted to Canada: The Watercolours of George Russell Dartnell. Toronto and Oxford: Dundurn Press, 1987. Degroot, Veronique. Candi, Space and Landscape: A Study of Distribution, Orientation and Spatial Organization of Central Javanese Temple Remains. Leiden: Sidestone Press, 2009. Dehejia, Vidya. ‘The Persistence of Buddhism in Tamil Nadu.’ In A Pot-Pourri of Indian Art, edited by Pratapaditya Pal, 53-74. Bombay: Marg, 1988. —. Art of the Imperial Cholas. New York: Columbia University Press, 1990.

284 

Construc ting K anchi

—. ‘On Modes of Visual Narration in Early Buddhist Art.’ The Art Bulletin 72, no. 3 (September 1990): 374-392. —. The Body Adorned: Dissolving Boundaries between Sacred and Profane in India’s Art. New York: Columbia University Press, 2009. —. ‘The Thief Who Stole My Heart: The Material Life of Chola Bronzes from South India, c.855-1280.’ Lecture series presented in A.W. Mellon Lectures in the Fine Arts. National Gallery of Art, Washington, DC, 2016. Dehejia, Vidya, and Charles Allen. India through the Lens: Photography 1840-1911. Washington, DC: Freer Gallery of Art and Arthur M. Sackler Gallery, Smithsonian Institution in association with Mapin Publishing, Ahmedabad, 2000. Dehejia, Vidya, and Richard H. Davis. The Sensuous and the Sacred: Chola Bronzes from South India. New York: American Federation of Arts, 2002. Dehejia, Vidya, and Peter Rockwell. The Unfinished: Stone Carvers at Work on the Indian Subcontinent. New Delhi: Lustre Press, 2016. Deloche, Jean. Ancient Fortifications of the Tamil Country as Recorded in Eighteenth-Century French Plans. Pondichérry: École Française d’ExtrêmeOrient, 2013. Dessigane, R., P.Z. Pattabiramin, and Jean Filliozat. Les Légendes Çivaïtes de Kañcipuram: Analyse de Textes et Iconographie. Pondicherry: Institut Français d’Indologie, 1964. Dewan, Deepali, and Deborah Hutton. Raja Deen Dayal: Artist-Photographer in 19th-Century India. Ahmedabad: Mapin Publishing, 2013. Dirks, Nicholas. ‘Political Authority and Structural Change in Early South Indian History.’ The Indian Economic and Social History Review 13 (1976): 125-157. —. ‘Colonial Histories and Native Informants: Biography of an Archive.’ In Orientalism and the Postcolonial Predicament: Perspectives on South Asia, edited by Carol A. Breckenridge and Peter van der Veer, 279-313. Philadelphia: University of Pennsylvania Press, 1993. —. Castes of Mind: Colonialism and the Making of Modern India. Princeton, NJ: Princeton University Press, 2001. Dutta, Ranjeeta. ‘Pilgrimage as a Religious Process: Some Reflections on the Identities of the Srivaisnavas of South India.’ Indian Historical Review 37, no. 1 (2010): 17-38. Ferstl, Chistian. ‘Overlooked Material for the Study of Pāśupata Śaivism: The Satirical Play Mattavilāsaprahasana.’ In Transmission and Tradition. On Quotations and Paraphrases from and Allusions to Ancient Texts on Indian Philosophy, edited by E. Prets and H. Marui, 1-19. Vienna: VÖAW, 2015. Fisher, Elaine M. Hindu Pluralism: Religion and the Public Sphere in Early Modern South India. Oakland University of California Press, 2017.

Bibliogr aphy

285

Flood, Finbarr Barry. ‘Pillars, Palimpsests and Princely Practices: Translating the Past in Sultanate Delhi.’ RES 43 (2003): 95-116. Fordham, Douglas. ‘Oriental Scenery.’ In Aquatint Worlds: Travel, Print, and Empire, 1770-1820, 59-117. London: Paul Mellon Centre for Studies in British Art, 2019. Francis, Emmanuel. ‘“Woe to Them!”: The Śaiva Curse Inscription at Mahābalipuram (7th Century CE).’ In The Archaeology of Bhakti: Mathurā and Maturai, Back and Forth, edited by Charlotte Schmid and Emmanuel Francis, 176-224. Pondicherry: École Française d’Extrême-Orient, 2013. —. ‘Royal and Local Bhakti under the Pallavas.’ In Mapping the Chronology of Bhakti: Milestones, Stepping Stones, and Stumbling Stones, edited by Valérie Gillet, 97-134. Pondicherry: Institut Français de Pondichéry and École Française d’Extrême-Orient, 2014. —. Le Discours Royal dans l’Inde du Sud Ancienne. Inscriptions et Monuments Pallava (IVème-IXème Siècles). Tome II: Mythes Dynastiques et Panégyriques. Leuven: Peeters Publishers, 2017. Francis, Emmanuel, Valérie Gillet, and Charlotte Schmid. ‘L’Eau et le Feu: Chronique des Études Pallava.’ Bulletin de l’Ecole Française d’Extrême-Orient 92 (2005): 581-611. —. ‘Trésors Inédit du Pays Tamoul: Chronique des Études Pallava II: Vestiges Pallava autour de Mahābalipuram et à Taccūr.’ Bulletin de l’Ecole Française d’Extrême-Orient 93 (2006): 431-481. —. ‘De Loin, de Près: Chronique des Études Pallava III.’ Bulletin de l’Ecole Française d’Extrême-Orient 94 (2007): 253-317. Ghose, Rajeshwari. The Tyāgarāja Cult in Tamil Nadu: A Study in Conflict and Accomodation. Delhi: Motilal Banarsidas, 1996. Gillet, Valérie. La Création d’une Iconographie Śivaïte Narrative: Incarnations du Dieu dans les Temples Pallava Construits. Paris: Institut Français de Pondichéry, Ecole Française d’Extrême-Orient Pondichéry, 2010. —. ‘Pallavas and Buddhism: Interactions and Influences.’ In Sivasri: Perspectives in Indian Archaeology, Art and Culture, edited by D. Dayalan, 105-36, 2013. —. ‘The Dark Period: Myth or Reality?’ The Indian Economic and Social History Review 51, no. 3 (2014): 283-302. —. ‘Gods and Devotees in Medieval Tiruttaṇi.’ In The Archaeology of Bhakti II: Royal Bhakti, Local Bhakti, edited by Emmanuel Francis and Charlotte Schmid, 443-494. Pondicherry: École Française d’Extrême-Orient/Institut Français de Pondichéry (Collection Indologie no. 125), 2016. Goodall, Dominic. ‘Who Is Caṇḍeśa?’ In Genesis and Development of Tantrism, edited by Shingo Einoo. Tokyo: Institute of Oriental Culture, University of Tokyo, 351-425, 2009.

286 

Construc ting K anchi

Gopalakrishnan, K.V. Tirukkaḻukkuṉṟam (Pakṣitīrtham) and its Temples: A Study of their History, Art, and Architecture. New Delhi: Abhinav Publications, 2005. Gopalakrishnan, M. History of Sri Ekambaranathar Temple, Kancheepuram, South India. Kanchipuram: Madhan Printers, 1978. Goudriaan, T. ‘Vaikhānasa Daily Worship: According to the Handbooks of Atri, Bhṛgu, Kāśyapa, and Marīci.’ Indo-Iranian Journal 12, no. 3 (1969-1970): 161-215. Gros, François, and R. Nagaswamy. Uttaramērūr: Légends, Histoire, Monuments. Pondicherry: Institut Français d’Indologie, 1970. Guha-Thakurta, Tapati. Monuments, Objects, Histories. New York: Columbia University Press, 2004. —. ‘The Production and Reproduction of a Monument: The Many Lives of the Sanchi Stupa.’ South Asian Studies 29, no. 1 (2013): 77-110. Guy, John, ‘The Lost Temples of Nagapattinam and Quanzhou: A Study in SinoIndian Relations.’ Silk Road Art and Archaeology 3 (1993-94): 291-310. —, ed. Lost Kingdoms: Hindu-Buddhist Sculpture of Early Southeast Asia. New Haven and London: Metropolitan Museum of Art and Yale University Press, 2014. Hall, Kenneth R. Trade and Statecraft in the Age of the Cholas. New Delhi: Abhinav Publications, 1980. —. ‘Local and International Trade and Traders in the Straits of Melaka Region: 600-1500.’ Journal of the Economic and Social History of the Orient 47, no. 2 (2004): 213-260. Hancock, Mary Elizabeth. The Politics of Heritage from Madras to Chennai. Bloomington: Indiana University Press, 2008. Hardy, Frieldhelm. ‘The Philosopher as Poet: A Study of Venāntadeśika’s “Dehalīśastuti”.’ Journal of Indian Philosophy 7, no. 3 (1979): 277-325. Hart III, George L. Poets of the Tamil Anthologies: Ancient Poems of Love and War. Princeton, NJ: Princeton University Press, 1979. Hegewald, Julia. Jaina Temple Architecture in India: The Development of a Distinct Language in Space and Ritual. Berlin: G&H Verlag, 2009. Heitzman, James. ‘Temple Urbanism in Medieval South India.’ The Journal of Asian Studies 46, no. 4 (1987): 791-826. —. Gifts of Power: Lordship in an Early Indian State. Delhi: Oxford University Press, 1997. —. ‘Urbanization and Political Economy in Early South India: Kāñcīpuram During the Cōḻa Period.’ In Structure and Society in Early South India: Essays in Honour of Noboru Karashima, edited by Kenneth R. Hall, 117-156. New Delhi: Oxford University Press, 2001. —. The City in South Asia. Abingdom: Routledge, 2008.

Bibliogr aphy

287

Heitzman, James, and S. Rajagopal. ‘Urban Geography and Land Measurement in the Twelfth Century: The Case of Kanchipuram.’ Indian Economic and Social History Review 41, no. 3 (2004): 237-268. Hendrickson, Mitch, and Damian Evans. ‘Reimagining the City of Fire and Iron: A Landscape Archaeology of the Angkor-Period Industrial Complex of Preah Khan of Kompong Svay, Cambodia (ca. 9th to 13th Centuries AD).’ Journal of Field Archaeology (2015): 644-664. Hopkins, Steven P. ‘Lovers, Messengers, and Beloved Landscapes: Sandeśakāvya in Comparative Perspective.’ International Journal of Hindu Studies 8, no. 1-3 (2004): 29-55. Hourani, George F. Arab Seafaring in the Indian Ocean in Ancient and Early Medieval Times. Princeton, NJ: Princeton University Press, 1951. Howes, Jennifer. Illustrating India: The Early Colonial Investigations of Colin Mackenzie (1784-1821). New Delhi: Oxford University Press, 2010. Hudson, D. Dennis. ‘Kanchipuram.’ In Temple Towns of Tamil Nadu, edited by George Michell, 18-39. Bombay: Marg Publications, 1993. —. The Body of God: An Emperor’s Palace for Krishna in Eighth-Century Kanchipuram. Oxford: Oxford University Press, 2008. —. ‘Ruling in the Gaze of God: Thoughts on Kanchipuram’s Maṇḍala.’ In Tamil Geographies: Cultural Constructions of Space and Place in South India, edited by Martha Ann Selby and Indira Peterson, 87-116. Albany, NY: State University of New York Press, 2008. Hüsken, Ute. ‘Gods and Goddesses in the Ritual Landscape of Seventeenth and Eighteenth-Century Kāñcipuram.’ In Layered Landscapes: Early Modern Religious Space across Faiths and Cultures, edited by Eric Nelson and Jonathan Wright, 63-81. London: Routledge, 2017. Inden, Ronald. ‘Hierarchies of Kings in Early Medieval India.’ Contributions to Indian Sociology 15, no. 99 (1981): 99-125. —. ‘Ritual, Authority, and Cyclic Time in Hindu Kingship.’ In Kingship and Authority in South Asia, edited by J. F. Richards, 41-91. Delhi: Oxford University Press, 1998. —. ‘Imperial Purāṇas: Kashmir as Vaiṣṇava Center of the Words.’ In Querying the Medieval: Texts and the History of Practices in South Asia, edited by Ronald Inden, Jonathan Walters, and Daud Ali, 29-98. Oxford: Oxford University Press, 2000. Irschick, Eugene F. Dialogue and History: Constructing South India, 1795-1895. Berkeley, CA: University of California Press, 1994. Kaimal, Padma. ‘Playful Ambiguity and Political Authority in the Large Relief at Mamallapuram.’ Ars Orientalis 24 (1994): 1-27.

288 

Construc ting K anchi

—. ‘Early Cōla Kings and “Early Cōla Temples”: Art and the Evolution of Kingship.’ Artibus Asiae 56, no. 1/2 (1996): 33-66. —. ‘Shiva Nataraja: Shifting Meanings of an Icon.’ The Art Bulletin 81, no. 3 (1999): 390-419. —. ‘The Problem of Portraiture in South India, Circa 870-970 AD.’ Artibus Asiae 59, no. 1/2 (1999): 59-133. —. ‘The Problem of Portraiture in South India, Circa 970-1000 AD.’ Artibus Asiae 60, no. 1 (2000): 139-179. —. ‘A Man’s World? Gender, Family, and Architectural Patronage, in Medieval India.’ Archives of Asian Art 53 (2002/2003): 26-53. —. ‘Learning to See the Goddess Once Again: Male and Female in Balance at the Kailāsanath Temple in Kāñcīpuram.’ Journal of the American Academy of Religion 73, no. 1 (March 2005): 45-87. —. Scattered Goddesses: Travels with the Yoginis. Ann Arbor, MI: Association for Asian Studies, 2012. —. ‘Lakṣmī and the Tigers: A Goddess in the Shadows.’ In The Archaeology of Bhakti: Mathurā and Maturai, Back and Forth, edited by Charlotte Schmid and Emmanuel Francis, 142-175. Pondicherry: École Française d’Extrême-Orient, 2013. —. ‘Circumambulation and Its Opposite: Visual and Verbal Cues to Movement Outside and inside the Kailasanatha Temple Complex in Kanchipuram.’ Paper presented in Landscape and Sacred Architecture in Pre-modern South Asia. Dumbarton Oaks, Garden and Landscape Studies, 14 November 2014. —. Opening Kailasanatha: The Temple in Kanchipuram Revealed in Time and Space. Seattle, WA: University of Washington Press, forthcoming 2021. Karashima, Noboru. South Indian History and Society: Studies from Inscriptions, A.D. 850-1800. Delhi: Oxford University Press, 1984. Kasdorf, Katherine E. ‘Forming Dorasamudra: Temples of the Hoysala Capital in Context.’ PhD diss., Columbia University, 2013. Kasdorf, Katherine E., and Padma Kaimal. ‘Getting the Band Back Together: Reuniting the Kanchi Yoginis.’ Paper presented in Annual Conference on South Asia. Madison, WI, 2019. Kaul, Shonaleeka. Imagining the Urban: Sanskrit and the City in Early India. London: Seagull Books, 2010. Kim, Jinah. ‘Unfinished Business: Buddhist Reuse of Angkor Wat and its Historical and Political Significance.’ Artibus Asiae LXX, no. 1 (2010): 77-122. Krishna, Nanditha, ed. Kānchi: A Heritage of Art and Religion. Madras: C.P. Ramaswami Aiyar Institute of Indological Research, 1992. Kubler, George. The Shape of Time: Remarks on the History of Things. New Haven and London: Yale University Press, 1962.

Bibliogr aphy

289

Kulke, Hermann. ‘Kṣetra and Kṣatra: The Cult of Jagannātha of Puri and the “Royal Letters” (Chāmu Ciṭaus) of the Rājās of Khurda.’ In Kings and Cults: State Formation and Legitimation in India and Southeast Asia, 51-65. New Delhi: Manohar, 1993. Kumar-Dumas, Divya. ‘Understanding Difference in Landscape Architectural History of Sri Lanka and India.’ Paper presented in 45th Annual Conference on South Asia. Madison, WI, 2016. Lee, Risha. ‘Constructing Community: Tamil Merchant Temples in India and China, 850-1281.’ PhD diss., Columbia University, 2012. Lockwood, Michael. Pallava Art. Madras: Tambaram Research Associates, 2001. Lockwood, Michael, G. Siromoney, and P. Dayanandan. Mahabalipuram Studies. Madras: Diocesan Press, 1974. Lorenzen, David N. The Kāpālikas and Kālāmukhas: Two Lost Śaivite Sects. New Delhi: Thomson Press, 1972. —. ‘A Parody of the Kāpālikas in the Mattavilāsa.’ In Tantra in Practice, edited by David Gordon White, 81-96. Princeton, NJ: Princeton University Press, 2000. Lutzker, Mary-Ann. ‘A Reinterpretation of the Relief Panel at Māmallapuram.’ In Chhavi-2: Rai Krishnadasa Felicitation Volume, 113-118. Bhanaras: Bharat Kala Bhavan, 1981. Mahalingam, T.V. Kāñcīpuram in Early South Indian History. Madras: Asia Publishing House, 1969. Malathi Devi, K.K. Prahasanas in Sanskrit Literature and Kerala Stage. Delhi: Nag Publishers, 1995. Manguin, Pierre-Yves, A. Mani, and Geoff Wade, eds. Early Interactions between South and Southeast Asia: Reflections on Cross-Cultural Exchange. Singapore: Institute of Southeast Asian Studies, 2011. Mechling, Mathilde. ‘The Buddhist and Hindu Bronze Statues from Indonesia in Their Religious, Cultural and Artistic Contexts (7th-11th centuries).’ University Sorbonne Nouvelle-Paris 3 and Leiden University, Dissertation in Progress. Meister, Michael W. ‘Siva’s Forts in Central India: Temples in Daksina Kosala and Their “Daemonic” Plans.’ In Discourses on Siva, edited by Michael W. Meister, 119-143. Philadelphia, PA: University of Pennsylvania Press, 1983. —. ‘Style and Idiom in the Art of Uparāmala.’ Muqarnas 10 (1993): 344-354. —. ‘The Hindu Temple: Axis of Access.’ Concepts of Space, Ancient and Modern (1991): 269-280. —. ‘Navigating Hudson.’ Journal of Vaishnava Studies 11, no. 1 (2002): 45-50. Michell, George, ed. Temple Towns of Tamil Nadu. Bombay: Marg Publications, 1993.

290 

Construc ting K anchi

Minakshi, C. Kanchi: An Introduction to Its Architecture. Delhi: Publications Division, Ministry of Information and Broadcasting, 1954. —. ‘Buddhism in South India.’ In South Indian Studies, Vol. II, edited by R. Nagaswamy, 83-131. Madras: Society for Archaeological, Historical, and Epigraphical Research, 1979. Mitter, Partha. Much Maligned Monsters: History of European Reactions to Indian Art. Oxford: Clarendon Press, 1977. Monier-Williams, Sir M. A Sanskrit-English Dictionary. Springfield, VA: Nataraj Books, (First Edition 1899] Nataraj Edition 2010. Monius, Anne. Imagining a Place for Buddhism: Literary Culture and Religious Community in Tamil-Speaking South India. New York: Oxford University press, 2001. Mukherjee, Rila. Vanguards of Globalization: Port-Cities from the Classical to the Modern. Delhi: Primus Books, 2014. Nagaswamy, R. ‘Innovative Emperor and His Personal Chapel: Eighth Century Kanchipuram.’ In Royal Patrons and Great Temple Art, edited by Vidya Dehejia, 37-60. Bombay: Marg Publications, 1988.

—. Viṣṇu Temples of Kāñcīpuram. New Delhi: D.K. Printworld, 2011.

Narayana Rao, Velcheru, David Shulman, and Sanjay Subrahmanyam. Symbols of Substance, Court and State in Nāyaka Period Tamilnadu. Delhi: Oxford University Press, 1998. —. Textures of Time: Writing History in South India, 1600-1800. Delhi: Permanent Black, 2006. Newell, H.A. Seven Pagodas, Chingleput and Conjeeveram: An Illustrated Guide with History. Madras and Bangalore: Higginbothams, 1914. Oettermann, Stephen. The Panorama: History of a Mass Medium. New York: Zone Books, 1997. Oliver, Liza. Art, Trade, and Imperialism in Early Modern French India. Amsterdam: Amsterdam University Press, 2020. Orr, Leslie. Donors, Devotees, and Daughters of God: Temple Women in Medieval Tamilnadu. New York: Oxford University Press, 2000. —. ‘Processions in the Medieval South Indian Temple: Sociology, Sovereignty and Soteriology.’ In South-Indian Horizons: Felicitation Volume for Francois Gros on the Occasion of his 70th Birthday, edited by Jean-Luc Chevillard and Eva Wilden, 437-470. Pondicherry: Institut Franc̨ a is de Pondichery and École Franc̨ a ise D’Extreme-Orient, 2004. —. ‘Preface.’ In Pondicherry Inscriptions, edited by Bahour S. Kuppusamy and G. Vijayavenugopal, I-XXVII. Pondicherry: Institut Français de Pondichéry and École Française d’Extrême-Orient, 2006.

Bibliogr aphy

291

—. ‘The Sacred Landscape of Tamil Śaivism: Plotting Place in the Realm of Devotion.’ In Mapping the Chronology of Bhakti: Milestones, Stepping Stones, and Stumbling Stones, edited by Valérie Gillet, 189-220. Pondicherry: Institut Français de Pondichérry and École Française d’Extrême-Orient, 2014. Orzech, C.D. ‘Vajrabodhi.’ In Esoteric Buddhism and the Tantras in East Asia, edited by C.D. Orzech, H.H. Sorensen, and R.K. Payne, 345-50. Leiden and Boston: Brill, 2011. Pal, Pratapaditya. Indian Sculpture. 2 vols. Vol. 2. Los Angeles: Los Angeles County Museum of Art, 1988. Parker, Samuel K. ‘Unfinished Work at Mamallapuram or, What is an Indian Art Object?’ Artibus Asiae 61, no. 1 (2001): 53-75. Pelizzari, Maria Antonella, ed. Traces of India: Photography, Architecture, and the Politics of Representation, 1850-1900. Montreal and New Haven: Canadian Centre for Architecture and Yale Center for British Art, 2003. Pinney, Christopher. Camera Indica. Chicago: University of Chicago Press, 1988. Pollock, Sheldon. The Language of Gods in the World of Men: Sanskrit, Culture, and Power in Premodern India. Berkeley: University of California Press, 2006. Quilley, Geoff, ed. William Hodges, 1744-1797: The Art of Exploration. New Haven and London: Yale University Press, 2004. Rabe, Michael D. ‘The Māmallapuram Praśasti: A Panegyric in Figures.’ Artibus Asiae 57, no. 3/4 (1997): 189-241. —. ‘The Monolithic Temples of the Pallava Dynasty: A Chronology.’ PhD diss., University of Minnesota, 1987. —. The Great Penance at Māmallapuram: Deciphering a Visual Text. Chennai: Institute of Asian Studies, 2001. Ramachandran, T.N. Bulletin of the Madras Government Museum: Tiruparuttikuṉṟam and its Temples. Madras: Commissioner of Museums, 2002. Raman, K.V. Srī Varadarājaswami Temple, Kāñchi: A Study of Its History, Art and Architecture. New Delhi: Abhinav Publications, 1975. Ramaswamy, Vijaya. ‘Vishwakarma Craftsmen in Early Medieval Peninsular India.’ Journal of the Economic and Social History of the Orient, 47, no. 4 (2004): 548-582. Ray, Himanshu Prabha. ‘A “Chinese” Pagoda at Nagapattinam on the Tamil Coast: Revisiting India’s Early Maritime Networks.’ IIC Occasional Publications 66 (2015): 1-18. Ray, Himanshu Prabha, and Jean-Francois Salles, eds. Tradition and Archaeology: Early Maritime Contacts in the Indian Ocean. New Delhi: Manohar, 1996.

292 

Construc ting K anchi

Ray, Romita. Under the Banyan Tree: Relocating the Picturesque in British India. New Haven, CT: Yale University Press, 2013. Reddy, Srinivas. Raya: Krishnadevaraya of Vijayanagara. New Delhi: Juggernaut Books, 2020. Reichle, Natasha. ‘Unfinished Business: Clues About Artistic Praxis from the Hidden Foot of Borobudur.’ Artibus Asiae LXIX, no. 2 (2009): 335-355. Rockwell, Peter. The Art of Stoneworking: A Reference Guide. Cambridge: Cambridge University Press, 1993. Roy, Kaushik. War, Culture and Society in Early Modern South Asia (1740-1849). New York: Routledge, 2011. Sanderson, Alexis. ‘The Śaiva Religion among the Khmers, Part I.’ Bulletin de l’Ecole Française d’Extrême-Orient 90-91 (2003-2004): 349-463. —. ‘The Śaiva Age: The Rise and Dominance of Śaivism During the Early Medieval Period.’ In Genesis and Development of Tantrism, edited by Shingo Einoo, 41-350. Tokyo: Institute of Oriental Culture, University of Tokyo, 2009. Scheurleer, Pauline, and Marijke J. Klokke. Divine Bronze: Ancient Indonesian Bronzes from A.D. 600 to 1600. Leiden: E.J. Brill, 1988. Schier, Kerstin. ‘The Goddess’s Embrace: Multifaceted Relations at the Ekāmranātha Temple Festival, Kanchipuram.’ PhD diss., University of Oslo, 2012. —. ‘The Goddess’s Embrace: Multifaceted Relations at the Ekāmranātha Temple Festival, Kanchipuram.’ Wiesbaden: Harrassowitz Verlag, 2018. Schmid, Charlotte. ‘The Edif ice of Bhakti: Towards an “Archaeological” Reading of Tēvāram and Periyapuranam.’ In Mapping the Chronology of Bhakti: Milestones, Stepping Stones, and Stumbling Stones, edited by Valérie Gillet, 241-286. Pondicherry: Institut Français de Pondichéry and École Française d’Extrême-Orient, 2014. Schmid, Charlotte, and Emmanuel Francis. The Archaeology of Bhakti II: Royal Bhakti, Local Bhakti. Pondicherry: Institut Français de Pondichéry and École Française d’Extrême-Orient, 2016. Schmücker, Marcus, ed. Viṣṇu-Nārāyaṇa: Changing Forms and the Becoming of a Deity in Indian Religious Traditions, Studies in Hinduism V. Vienna, 2016. Sears, Tamara I. Worldly Gurus and Spiritual Kings: Architecture and Asceticism in Medieval India. New Haven and London: Yale University Press, 2014. Seastrand, Anna Lise. ‘Praise, Politics, and Language: South Indian Murals, 15001800.’ PhD diss., Columbia, 2013. Selby, Martha Ann. Tamil Love Poetry: The Five Hundred Short Poems of the Ainkurunuru. New York: Columbia University Press, 2011.

Bibliogr aphy

293

Selby, Martha Ann, and Indira Peterson, eds. Tamil Geographies: Cultural Constructions of Space and Place in South India. Albany, NY: State University of New York Press, 2008. Seshadri, A.K. Kancheepuram and its Temples. Tirupati: Tirumala Tirupati Devasthanams, 2003. Shulman, David. Tamil Temple Myths: Sacrifice and Divine Marriage in the South Indian Saiva Tradition. Princeton, NJ: Princeton University Press, 1980. —. More Than Real: A History of the Imagination in South India. Cambridge, MA: Harvard University Press, 2012. Shulman, David, and Yigal Bronner. ‘“A Cloud Turned Goose”: Sanskrit in the Vernacular Millennium.’ The Indian Economic and Social History Review 43, no. 1 (2006): 1-30. Simmons, Caleb. Devotional Sovereignty: Kingship and Religion in India. New York: Oxford University Press, 2020. Sloan, Kim. A Noble Art: Amateur Artists and Drawing Masters, C.1600-1800. London: British Museum Press, 2000. Srinivasan, C.R. Kanchipuram through the Ages. New Delhi: Agam Kala Prakashan, 1979. Stein, Burton. ‘Circulation and the Historical Geography of Tamil Nadu.’ The Journal of Asian Studies 37, no. 1 (Nov. 1977): 7-26. —. Peasant State and Society in Medieval South India. Delhi: Oxford University Press, 1980. Stein, Deborah. ‘Smelting Zinc and Housing the Divine at Jawar.’ Artibus Asiae 72, no. 1 (2012): 141-158. Stein, Emma Natalya. ‘All Streets Lead to Temples: Mapping Monumental Histories in Kanchipuram, ca. 8th-12th centuries CE.’ PhD diss., Yale University, 2017. Stein, Robert M. Reality Fictions: Romance, History, and Governmental Authority, 1025-1180. Notre Dame: University of Notre Dame Press, 2006. Subramanian, T.S. ‘Urn Burial Site Discovered near Kancheepuram.’ The Hindu, 9 April 2012. Tartakov, Gary Michael, and Vidya Dehejia. ‘Sharing, Intrusion, and Influence: The Mahiṣāsuramardinī Imagery of the Calukyas and the Pallavas.’ Artibus Asiae 45, no. 4 (1984): 287-345. van der Velde, Paul. Life under the Palms: The Sublime World of the Anti-Colonialist Jacob Haafner. Singapore: Ridge Books, National University of Singapore, 2020. Veluthat, Kesavan. The Early Medieval in South India. Delhi: Oxford University Press, 2009.

294 

Construc ting K anchi

Venkataraman, K.R. Dēvī Kāmākshī in Kāñchī: A Short Historical Study. Tiruchirapalli: Sri Vani Vilas Press, 1973. Viraraghavacharya, T.K.T. History of Tirupati: The Tiruvengadam Temples. 3 vols. Vol. 1-3. Tirupati: Tirumala Tirupati Devasthanam Religious Publications, 1997. Wagoner, Phillip B. ‘Precolonial Intellectuals and the Production of Colonial Knowledge.’ Comparative Studies in Society and History 45, no. 4 (October 2003): 783-814. Walding, Richard, Helen Stone, and S. Nair Achuthasankar. ‘The Russian Prince and the Maharajah of Travancore.’ Journal of Kerala Studies 36 (2009): 10-87. Weschler, Helen J. ‘Royal Legitimation: Ramayana Reliefs on the Papanatha Temple at Pattadakal.’ In The Legend of Rama: Artistic Visions, edited by Vidya Dehejia, 27-42. Bombay: Marg, 1994. Whitehead, David. ‘Wathen, James (Bap.1752, D.1828).’ In Oxford Dictionary of National Biography. Oxford: Oxford University Press, 2004. Wolffhardt, Tobias. Unearthing the Past to Forge the Future: Colin Mackenzie, the Early Colonial State, and the Comprehensive Survey of India. Translated by Jane Rafferty. Studies in British and Imperial History. New York: Berghahn Books, 2018.

Index abhiṣeka (‘ritual bathing’) 35, 72, 83, 266 ācārya (‘senior initiated priest’) 57 Acyutadevarāja (king) 206 Āditya I (king) 170 Airāvateśvara temple (Kanchi) 66, 68, 86–87, 101, 108, 116, 266 Ajanta 48 Aḻakiya Ciṅka Perumāḷ temple (Kanchi) 67 Ali, Haidar (sultan) 134, 212, 215 Amareśvara temple (Kanchi) 66, 86, 88, 248, 264 Ancient Written Characters of India, The 236 Andhra Pradesh 30, 45, 139 n. 86 Aṉēkataṅkāvatīśvara temple (Kanchi) 66, 145, 148–151 Angkor (Cambodia) 90, 106, 193–195 Appar, Saint 52, 54 n. 31, 55–56, 111, 128 Archaeological Survey of India (ASI) 34, 80, 119, 176–177, 245, 249–251, 263, 267–268, 271 archaeology 30, 32, 34–35, 38, 42, 45–46, 53, 58–59, 88, 109, 120, 122, 138, 142, 155, 165, 186–187, 194, 197, 202, 245, 248, 265 architecture adhiṣṭhāna (‘basal molding’) 119–120, 161 aṅga-ālāya (‘limb shrine’) 72 ardhamaṇḍapa (‘vestibule’) 51, 60, 84, 88, 117, 119, 148, 171, 180 n. 77, 206 balipīṭha (‘offering pedestal’) 60 basement moldings 50, 58, 63, 67 n. 74, 68, 92, 100, 111, 151 caitya halls 249–250 catustāla (‘four-storey’) tower 84 choultry 218, 228–229, 236–237, 242–243 citramaṇḍapa (‘painted hall’) 111, 215 devakulikā (‘external shrine’) 77, 86 ghaṭa (‘capital’) 60, 65, 77 gopura (‘gateway’) 25, 73 n. 89, 92, 128, 159 n. 18, 169–170, 172, 205, 210, 221–226, 229, 232, 252, 264, 266, 272 jagatī (‘foundation platform’) 61, 119 kapota (‘cornice’) 86, 92 mahāmaṇḍapa (‘entry hall’) 60, 85, 148 maṇḍapa (‘pavilion’) 25, 34, 49–53, 56, 71, 72 n. 83, 76–78, 104 n. 1, 112–113, 126 n. 53, 146–147, 164, 189–190, 195, 204–206, 218, 228, 253, 256, 258, 273 mukhamaṇḍapa (‘entry hall’) 86, 91, 158 pādabandha-adhiṣṭhāna (‘basal molding’) 84, 151 paving stone 53, 92 n. 150, 147 n. 111, 266 pillars 50–51, 53–54, 60, 62–66, 72 n. 83, 76–77, 95, 97, 104 n. 1, 118, 126 n. 53, 132–133, 146–147, 150, 164–165, 187, 192, 206, 213, 236–237, 242–244, 252, 258, 265

pillars, lion-based (‘siṁhapāda’) 53, 64–66, 69, 122, 158, 222 potikā (‘corbel, bracket’) 86, 150 prabhā (‘aureole of flames’) 131 pradākṣina (‘inner circumambulatory passage’) 84–85 prākāra (‘enclosure wall’) 51, 53–54, 58, 60, 65, 67 n. 73, 69–74, 77, 79, 80–84, 86, 92, 158–159, 164, 168, 181–182, 190–192, 205–206, 210, 211 n. 36, 212, 222, 235, 264–265, 272 reliefs 26, 38, 43, 53, 57–63, 65, 79–84, 88, 90–91, 110, 119, 150, 167, 169, 173–175, 181–182, 187, 190, 192, 218, 251, 256, 265 sabhā (‘assembly hall’) 93, 95, 123, 265 śikhara (‘superstructure’) 55, 86 thousand-pillar hall 203–205, 226 n. 82, 232, 242, 252, 260 tripaṭṭa-kumuda molding 161 vasanta maṇḍapa 205, 210, 242, 244, 272 vimāna (‘main shrine’) 34, 51, 72, 76, 78, 84–86, 93, 119, 148, 168, 180 n. 77, 206, 251, 263–264 vṛtta-sphuṭita (‘overflowing vase’) 151 see also under Buddhism and Tamil Nadu temples Arcot 209, 211–212, 218, 227, 230 Ārpākkam 50, 93, 99, 123, 136–139, 142–143, 155, 177, 189 Aṣṭabhuja Perumāḷ temple (Kanchi) 67, 92, 133, 203 Āti Kāmākshi Ammaṉ temple (Kanchi) 66 Ayyaṉār (god) 66 n. 66, 122, 137–138 Bagan (Myanmar) 106, 193–195 Battle of Plassey (1757) 207 Bhadrakāḷī Ammaṉ temple (Kanchi) 60–63, 65 Bhaktavatsaleśvara temple (Tirukkaḻukuṉṟam) 166, 168–170, 172, 263 bhakti (‘devotional’) literature 32, 128, 147; see also individual titles Bhubaneshwar (Orissa) 193–195 biruda (‘title of praise’) 45, 51, 54, 75 Bodhidharma 185 Borobudur (Java) 61, 90 Brahmā (god) 49, 140 n. 89, 163, 257 brahmadēya 136, 139, 141, 143 n. 98 Brahmi 32, 58, 75, 241 Brahmins 29 n. 4, 61, 71, 139, 141, 172, 239, 254, 257, 260–261, 263, 269 councils 93, 146, 234 settlements 116, 140–141 Bṛhadēśvara temple (Thanjavur) 115, 119

296 

Construc ting K anchi

bronzes 26, 131, 136, 146, 183, 186–187, 193, 209, 266, 271–272 Kanchi bronzes 131–136, 151 procession of 26, 35, 129, 131, 133 Buddha, depictions of 32, 34, 51 n. 26, 58–59, 138 Kanchi Buddhas 187–193 Paḷḷūr seated Buddhas 188 Buddhism 31, 55, 61, 109, 182, 187, 192–193, 197, 203, 241, 249–251, 269 art and architecture 191–192, 195, 249–250 buddhapaḷḷi 186 and Kanchi 41, 107, 155, 182–187, 192–193, 197 monasteries (‘rājavihāra’) 58, 77 n. 102, 186–187, 195

Dantivarman (king) 67 n. 74, 91, 93, 100 n. 165 darśan 35, 71, 129, 148 Dartnell, George Russell 242–245 Daśakumāracarita (Daṇḍin) 110–111 Deccan 47–48, 205, 237 Deśavarṇana (Nagarakṛtāgama) (Mpu Prapañca) 269–270 devadāna (‘Gift of the Gods’) 140 Dharmapāla 185 Divya Prabandham 31, 67, 79 Dupleix, French Governor-General 208, 210, 212 Durgā (goddess) 71, 117–118, 121, 140 n. 89, 147, 161–163 dvārapāla 87, 121, 167, 225

Cāḷuvaṉkuppam 173 n. 53, 177, 180 cāmaradhāra (‘flywhisk bearer’) 69, 79 Cambodia 90, 106, 184, 193–194 Caṇḍeśa (god) 170, 264 Candraprabhā Jiṉālāya temple (Kanchi) 66, 88, 101, 241 Caṅkam-era literature 31, 111, 153, 178 Caṅku Tīrtham (Kanchi) 166–167, 172 cave-temples 32, 47–49, 62–63, 75–76, 79, 121, 159, 165, 167, 187 n. 109 Cevilimēṭu 64, 101, 121–122 Chidambaram 106, 209, 252 Chola dynasty 31, 76, 98, 104, 114, 116, 120, 123, 126, 134, 138, 140, 143, 151 palace at Kanchi 111–113 royal temple 114, 116, 118, 119–20 see also individual names of kings Cidambareśvara temple (Kanchi) 122 Cilappatikāram (‘Tale of the Anklet’) (Aṭikal) 31, 73 n. 88, 111, 179 Clive family 230, 232 Charlotte 200, 202, 229–230, 232–233, 243 Henrietta Antonia 42, 199, 202, 220, 222, 229, 233 Robert, Lord Clive 200, 207, 208–212, 260, 272 Cokkīśvara temple (Kanchi) 147, 267 Condgévéram, ville sainte dans le Carnatik aux environs de Madras, lieu de pélerinage des Hindoux (Prince Alexis Soltykoff) 256 Conjeveram Before You Come to the Great Pagoda (Mackenzie Collector) 233 copper plates 45, 74 n. 91, 101 n. 167, 111, 131, 139, 140 n. 87, 141, 152 n. 116 Kasakudi copper plates 71, 213 Kūram copper plates 47, 236 Crole, Charles Stewart 134–135, 157 n. 12, 241, 248, 258, 260

East India Company 212, 219, 243 Ekāmbaranātha temple (Kanchi) 25–26, 35, 38, 45, 51–57, 60–61, 64, 66, 68, 72, 74, 92, 101, 108, 116, 118–119, 128, 131 n. 71, 134, 142, 147, 190, 203–206, 208, 210, 213, 215–216, 221–226, 228–229, 232–233, 236, 242, 245, 252–253, 258–261, 264, 270, 272 Ēkampam 52, 55–56 Elevation of the Great Pagoda at Conjeveram (Thomas Fraser) 225 Encyclopaedia of Indian Temple Architecture 119 ēri (‘reservoir’) 75, 123–124, 136, 160

Daṇḍin 110 Daniell, Thomas and William 202, 219, 222, 224, 231, 272

Fergusson, James 42, 217, 218–220, 248–252 Fraser, Thomas 224–225 Gajalakṣmī (goddess) 83, 173 n. 53 Ganesha (god) 65–66, 122, 140 n. 89, 148 n. 113, 163, 190 Gangaikondacholapuram 76, 109, 111–112, 114, 119 Garuda Sevai festival 36, 216, 258–259 Geological Survey of India (GSI) 98 ghaṭikā 71–72 ‘Great Pagoda’ see Ekāmbaranātha temple Great Relief (Māmallapuram) 173–175, 181 Hajrath Hameed Hauliya Dharga (Kanchi) 232 Haṃsasandeśa (Vedānta Deśika) 36 Hindoo Architecture: Granite Pillars in a Choultry at Conjeveram (Mackenzie Collector) 236–237 Hinduism 34–35, 55, 68, 182, 184, 189, 192, 195, 201, 203, 216, 218–219, 241, 248, 250, 265; see also Shaivism; Shiva; Vaishnavism; and Vishnu Hindu Religious and Charitable Endowments Administration 245 Historical Sketches of the South of India (Mark Wilks) 241

Index

History of the Military Transactions of the British Nation in Indostan, A (Robert Orme) 200, 208 Hodges, William 221, 258 Illustrations of the Rock Cut Temples of India (James Fergusson) 251 Imperial Gazetteer of India, The 241 Indian Antiquary 191 Indonesia 61, 106, 184, 193 Intérieur du couvent de Condjévéram, à 40 milles de Madras, Réunion journalière des Brames en l’honneur des deux divinités Conservatrice et Destructive, Juillet 1841 (Prince Alexis Soltykoff) 257, 258 Iṟavāttāṉeśvara temple (Kanchi) 66, 86, 248 Jain at Conjeveram, 16th March 1800, A (Mackenzie Collector) 238–239 Jainism ascetics 32, 75, 233, 239, 241 colonial study of 237–241, 249, 250–251 population 35, 55, 184, 241 temples 66, 88, 136, 138, 192, 240–241 Vardhamāna (Mahāvīra) (Jina) 88 worship 90, 186, 192, 203 ‘Jains, The’ (Colin Mackenzie) 239 Journal of a Voyage, in 1811 and 1812, to Madras and China (James Wathen) 225 Journal of the Royal Asiatic Society of Great Britain and Ireland 245 Jvarahareśvara temple (Kanchi) 116, 118–120 Jyeṣṭā (goddess) 71, 83, 138 Kacchapeśvara temple (Kanchi) 63–65, 87, 92, 101, 116–117, 147–148, 192, 203, 205, 236 Kacci Mayāṉam shrine 52, 56–57 Kailāsanātha temple (Kanchi) 28, 37, 49, 63–64, 66–79, 86–87, 101, 104 n. 1, 110, 116, 118, 120, 123, 126 n. 53, 148, 151, 180, 222, 235–236, 245–246, 248, 264–265 Kailāsanātha temple (Puḷḷalūr) 95 Kailāsanātha temple (Tiruppaṭṭūr) 91 Kailāsanātha temple (Uttiramērūr) 93 Kaḷakkāṭṭūr 93, 120, 123–125, 136, 189 Kāḷiyammaṉ (goddess) 66 Kāma (god) 55 Kāmākṣī Ammaṉ temple (Kanchi) 37–38, 58–59, 67 n. 71, 86 n. 127, 93 n. 150, 107–108, 114, 116, 134–135, 188 n. 110, 191–193, 203, 215, 242, 244–245, 262, 266, 271–273 Kamarajar Salai 103–104, 109–110, 113, 122–123, 126 Kāma Sūtra 61 Karukkiṉil Amarntavaḷ Ammaṉ temples (Kanchi) 61, 64–65, 68, 101, 122, 147, 188 n. 110, 190 Kauśikeśvara temple (Kanchi) 267 Kāvērippākkam 95, 101, 209, 218

297 Kaveri region 100 n. 165, 114, 147–148, 151, 156, 163, 188 kāvya 109–110 Kāyārohaṇeśvara temple (Kanchi) 57–58, 122 Khajuraho 61 Koṭumpāḷūr 50, 76, 146 Kṛṣṇarāja III (king) 215 Kubler, George 80 kuḷam (‘ablutions tank’) 58, 65, 133, 147, 160 Kulottuṅga I (king) 111, 138, 140, 145 Kulottuṅga II (king) 128 Kumaravāṭi 93, 157, 159–162 Kumbakonam 106, 245 Kūram 47, 49, 77, 86, 95, 101, 122, 126, 133, 236 La Création d’une Iconographie Śivaïte Narrative (Valérie Gillet) 72 Lady Henrietta Antonia Herbert, Countess of Powis (1758–1830) (Joshua Reynolds) 200 Lakṣmī (goddess) 71 Lettres sur l’Inde (Prince Alexis Soltykoff) 253 liṅga see under Shiva Mackenzie, Colin (Colonel) 42, 233–242, 245, 251 Madras 200, 203, 207, 210, 216–220, 227–228, 245, 254–258 Madras Government Museum 51, 133, 191 Madurai 31, 47, 106, 111 Mahāmalla (king) 76, 167 Māhātmya 108, 206 Mahavellipore. The Five Rathas (James Fergusson) 250 Mahendravarman I (king) 32, 45, 48–49, 51, 53–54, 75–77, 79, 109, 118, 121, 164, 213 Mahendravarman III (Mahendra) (king) 77 Mahiṣāsuramardinī cave-temple (Māmallapuram) 62–63, 66 n. 66, 265 Mākaṟal 123, 136, 138, 155 Māmallapuram 49, 62, 76, 79, 85, 90–91, 100–101, 153, 157, 171–182, 202, 218, 224, 231, 251, 252–265 Māmaṇṭūr, cave-temples at 32, 49, 74–75, 121 Māṇḍukaṇṇīśvara temple (Kanchi) 64 Māṇikkavācakar, Saint 168 Maṇimēkalai (Shattan) 31, 184–185, 191, 193 Maṇṭakappaṭṭu, cave-temple at 48, 76 Map of Conjeeveram (John Gould) 240 Mataṅgeśvara temple (Kanchi) 67–68, 89–92, 247–248 Mattavilāsa Prahasana (‘The Farce of Drunken Sport’) (Mahendravarman I) 34, 43, 45, 54, 55 n. 33, 59, 109, 186 Mayamata 210 Minakshi, C. 71–72, 77 n. 102, 80, 186–187, 190 Mukteśvara temple (Kanchi) 67–68, 89–92, 248 Murukaṉ (god) 126, 145, 148 n. 113, 177, 180 Mysore sultanate 134, 212, 234 Mysore Wars 212–218, 220, 232, 234

298  Nagapattinam 58, 187–188, 193 Nāgārjuna 186 nāga stones 84–85, 117 Nandivarman I (king) 83 Nandivarman II (king) 67 n. 74, 71, 83, 89, 91, 141 Nandivarman III (king) 80–81, 91 Narasiṃhavarman I (king) see Mahāmalla Narasiṃhavarman II (king) see Rājasiṃha Newell, Herbert Andrew 171, 179–180 Nṛpatuṅgavarman (king) 158 Oṇakānteśvara temple (Kanchi) 66 Orakaṭam, temple at 95, 157, 160, 166, 250, 263 Oriental Scenery (Thomas and William Daniell) 223–224 Orissa 47–48, 193–194 Pagoda in Conjeveram 235 Pagodas at Trinchicunum 171–172 Palanquin Bearers and a Messenger Camel (Anna Tonelli) 230–231 Palar River 101, 153, 156, 160, 164, 215 Pallava Architecture (Alexander Rea) 88 Pallava dynasty 26, 30, 34, 38, 40–41, 45–46, 63, 74, 79, 81, 98, 100–102, 120, 141, 264, 271; see also individual names of kings Paṉamalai 63, 76, 101, 264 Paṇāmuṭīśvara temple (Kanchi) 147 Pāṇḍavadūta Perumāḷ temple (Kanchi) 67 Parameśvaravarman I (king) 49, 63, 76–77, 159 n. 19, 169 n. 45 Parameśvaravarman II (king) 83 Pāśupata sect 55, 57–58 Pattuppāṭṭu 31 Periya Purāṇam (Cēkkiḻār) 31, 51, 56, 127–128, 178, 186, 232 Perumpāṇāṟṟuppaṭai 31, 46–47, 178 Picturesque Illustrations of Ancient Architecture in Hindostan (James Fergusson) 248 Picturesque Scenery in the Kingdom of Mysore (James Hunter) 220 pillars see under architecture Piṟavāttāṉeśvara temple (Kanchi) 66, 86–87, 91 n. 144, 248 Place, Lionel 215 Polonnaruwa (Sri Lanka) 106, 193, 195 Prambanan (Java, Indonesia) 61, 106, 193–195 pūjā (‘worship’) 35, 38, 57, 119, 169, 270–272 Pukār 73 n. 88, 111 n. 23, 179 Puḷḷalūr 95, 126, 213–215 Puṇyakoṭīśvara temple (Kanchi) 150 purāṇa 110, 128, 206 PV Kaḷattūr 159–160, 162 Rājādhirāja II (king) 138 Rājarāja I (king) 138 Rājasanagara (king) 269 Rājasiṃha (king) 49, 63, 66, 68, 73, 75–79, 86–89, 110, 164, 168, 185

Construc ting K anchi

Raja Veedhis 112–113, 136 Rājēndra I (king) 138, 152 n. 116, 162 Rāmānuja 129, 268 Rea, Alexander 80, 88, 264 Report on the Jagir (Lionel Place) 215 reservoirs see ēri Reynolds, Sir Joshua 199–200 Rough Sketch of the Fortified Pagoda of Great Cangivaram, 1872 209 Royal Asiatic Society 245, 249, 265 Rudrakoṭīśvara temple (Kanchi) 64, 101, 122, 140 Śaiva Siddhānta 73–74 Salt, Henry 171–172, 235 Saluva dynasty 126 Sambandar, Saint 52, 67, 128, 184, 232 Sandeśakāvya (‘messenger poem’) 36 Śaṅkarācārya 31, 245 Sanskrit 26, 34, 40, 45–46, 49, 54, 57, 79, 107–110, 141, 156, 183 n. 84, 184, 206, 249, 270, 272 Sarasvatī (goddess) 71, 271 Sarva Tīrtham (Kanchi) 133, 147, 215 Second View from the Great Pagoda near Conjeveram (James Wathen) 226 Sembiyan Mahadevi (queen) 162–163 Shaivism 35, 203, 249–250 ascetics (Kāpālikas) 54–57, 62, 272 dieties in art 68, 92, 118 iconography 140, 182 n. 80, 192 n. 121 nāyaṉmār (saint) 31, 67, 127–129, 183–184, 186 population 78, 127, 182 Shiva (god) 36, 51, 57, 72, 74, 108 Dakṣiṇāmūrti 79, 140 n. 89, 163, 170, 192 n. 121 images of 35, 58, 68, 79, 133, 140, 169, 180, 182 n. 80, 225 liṅga (‘sign of Shiva’) 26, 51, 56, 74, 163, 195 Naṭarāja (god) 131, 140, 210 worship 29, 35–36, 49, 117, 127, 192, 257–259, 261, 272 Shore temple (Māmallapuram) 60, 63, 76, 78, 100, 151, 173, 175–177, 180–182, 252 Siṃhaviṣṇu (king) 83 Sīteśvara temple (Kanchi) 33, 133, 147 Sketch Map of India, 1847 (James Fergusson) 217 Soltykoff, Prince Alexis 42, 219–220, 231, 248, 253–263 Somāskanda 58, 68, 131 n. 71, 169 Sri Lanka 106, 139, 177, 179, 193, 197, 242, 253 Srirangam 106, 129, 209 Śrīraṅgapaṭṭaṇa 213, 234 Śrīvaiṣṇava 126, 129, 206, 268 Sthalapurāṇa (‘legendary history’) 36, 108, 206, 258 Sthalaśayana Perumāḷ temple (Māmallapuram) 180

299

Index

Stone Pillar in a Choultree at Conjeveram–near Wallajabad–Jany 1829 (George Russel Dartnell) 242–243 Sundarar, Saint 52, 128 Sundaravarada Perumāḷ temple (Uttiramērūr) 86, 93–94, 151 Sunnath Jamath Jumma Masjid (Kanchi) 232 Sūrya (god) 57, 87 n. 130, 148 n. 113 Tāḷagirīśvara temple (Paṉamalai) 63, 76, 264, 265 Tamil 35, 127, 182, 216, 258–259, 263 inscriptions 36, 47, 146, 155 language 26, 45, 56, 88 n. 131, 122, 141, 184 literature 31, 36, 38, 39–40, 46, 49, 57, 108–109, 113, 127–128, 141, 153, 178, 184, 206, 270 saints 31, 51, 111, 127, 168, 232; see also under Shaivism and Vaishnavism Tamil Nadu 30, 34, 41, 45–48, 106, 123, 133, 140, 142, 144, 151, 153, 155, 183, 196, 202, 205, 218, 219–220, 237 Tamil Nadu temples architecture 68, 128, 158, 223, 251 construction 30–31, 47, 49, 63, 76–77, 98–99, 146–147, 150, 177, 191, 195 control of 243, 245 inscription 46, 116, 146, 150 usage and upkeep 114, 247 see also names of individual temples Tāṉtōṉṟīśvara temple (Kanchi) 43, 44, 59–63, 65, 192 Tēvāram 31, 55, 57, 66, 116, 119, 128 Thanjavur 76, 114–115, 119, 135, 202, 217, 251 Tipu Sultan 212–215, 218, 234 Tiruchirappalli 49, 202, 209, 212, 217, 219 Tirukkālīśvara temple (Tirukkālimēṭu) 67, 147 Tirukkaḻukuṉṟam 76, 95, 153, 157, 166–172, 263 Tirukkuṟipput Toṇṭar, Saint 128, 153, 178 Tirumēṟṟaḷinātha temple (Kanchi) 66, 72, 92, 122 Tirumukkūṭal 101, 136, 157–159 Tiruparuttikuṉṟam (Kanchi) 66, 88, 240, 251 Tirupati 41, 104, 122, 126–127, 129, 133, 159, 197 Tiruppulivaṉam 123–124, 136 Tirupurantakesvara see Amareśvara temple Tirupuvaṉam 76, 114 Tiruttaṇi 95, 101, 126 Tonelli, Anna Nistri 201–202, 230, 231 Topographical List of the Inscriptions in the Tamil Nadu and Kerala States, A (Mahalingam) 146 Trilokya Jinasvāmi temple (Kanchi) 88, 241 Ulakaḷanta Perumāḷ temple (Kanchi) 92 n. 150, 111, 114, 133, 144 Umā (goddess) 51, 58, 68, 108, 225

Uttama Cōḻa (king) 267 Uttiramērūr 50, 86, 93–95, 100–102, 120, 122–123, 125, 136, 145, 151, 156, 265 Vaikuṇṭha Perumāḷ temple (Kanchi) 37, 60–61, 65, 67–68, 79–86, 89, 91–93, 101, 116, 141, 181–182 Vaishnavism 66, 68, 104, 128–129, 184, 203, 249–250 āḻvār (saint) 31, 67, 79, 126, 129, 133, 183, 205 divya deśam 67, 128–129 texts 36, 79 Vajrabodhi 185 Valentia, Viscount 172, 218, 220, 235–236 Varadarāja Perumāḷ temple (Kanchi) 38, 64, 66–68, 92, 108, 116, 126, 130, 133–134, 146, 186, 203, 206, 215–216, 232–233, 245, 253, 258–260 Vāyalūr 164–165 Vedagirīśvara temple (Tirukkaḻukuṉṟam) 166–168, 170–172 Vedānta Deśika 36, 268 Vellore 208, 218, 227 Veṅkateśa Perumāḷ temple (Tirumukkūṭal) 157–159 Veṅkaṭeśvara temple (Tirupati) 126–127 View from the Royal Artillery Encampment, Conjeveram, A (James Hunter) 220–221 View Looking Towards an Unidentified Mandapa (Henry Dixon) 204 Vijayanagara period 126, 205–206, 241 Viḷakkoḷi Perumāḷ temple (Kanchi) 67, 133, 147, 268 Vīrarājendra (king) 111, 159 Vīraṭṭāṉeśvara temple (Kanchi) 62–63, 65–66, 95 n. 154, 192, 205, 236 Vishnu (god) 36, 72, 108, 187, 192 n. 122, 259 images of 76, 84–85, 98 n. 157, 140 n. 89, 159, 197, 226 worship 49, 80, 127, 135, 257–258, 261 Voyage en poste dans les plaines du Punjab, entre Loodiana et Omritsar, Fevrier 1842 (Prince Alexis Soltykoff) 255 Voyages and Travels to India, Ceylon, the Red Sea, Abyssinia, and Egypt (Viscount Valentia) 235 Vṛṣabheśvara temple (Kanchi) 52–53, 56, 68, 92, 108, 118 Wathen, James 202, 218–220, 225–229, 272 Worthless Scraps from Many Lands (George Russel Dartnell) 242 Xuanzang 25, 58, 185–186, 241, 251 Yathoktakārī temple (Kanchi) 67, 104 n. 1 yoga 74, 181