Temple to Love : Architecture and Devotion in Seventeenth-Century Bengal 0253344875, 2004016533

Table of contents :
Cover
Contents
List of Illustrations
Acknowledgments
Note on Transliteration
Introduction
1 Desire, Devotion, and the Double-Storied Temple
2 A Paradigm Shift
3 Acts of Accommodation
4 Axes and the Mediation of Worship
Epilogue: A New Sacred Center
Glossary of Architectural Terms
A
B
C
E
G
J
K
L
M
N
P
R
S
T
U
Notes
Bibliography
Index
A
B
C
D
E
F
G
H
I
J
K
L
M
N
O
P
Q
R
S
T
U
V
W
Y
Z

Citation preview

Temple to Love

Illustrations

•

i

Contemporary Indian Studies Published in association with the American Institute of Indian Studies Susan S. Wadley, Chair, Publications Committee/general editor

AIIS Publications Committee/series advisory board John Echeverri-Gent Brian Hatcher David Lelyveld Martha Selby Books in this series are recipients of the Edward Cameron Dimock, Jr. Prize in the Indian Humanities and the Joseph W. Elder Prize in the Indian Social Sciences awarded by the American Institute of Indian Studies and are published with the Institute’s generous support. A list of titles in this series appears at the back of the book.

ii

•

Illustrations

Temple to Love Architecture and Devotion in Seventeenth-Century Bengal PIKA GHOSH

INDIANA UNIVERSITY PRESS Bloomington and Indianapolis

Illustrations

•

iii

This book is a publication of Indiana University Press 601 North Morton Street Bloomington, IN 47404-3797 USA http://iupress.indiana.edu Telephone orders

800-842-6796

Fax orders

812-855-7931

Orders by e-mail

[email protected]

© 2005 by Pika Ghosh All rights reserved No part of this book may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying and recording, or by any information storage and retrieval system, without permission in writing from the publisher. The Association of American University Presses’ Resolution on Permissions constitutes the only exception to this prohibition. The paper used in this publication meets the minimum requirements of American National Standard for Information Sciences— Permanence of Paper for Printed Library Materials, ANSI Z39.48-1984. Manufactured in the United States of America Library of Congress Cataloging-inPublication Data Ghosh, Pika, date Temple to love : architecture and devotion in seventeenth-century Bengal / Pika Ghosh. p. cm. — (Contemporary Indian studies) Includes bibliographical references and index. ISBN 0-253-34487-5 (cloth : alk. paper) 1. Temples—India—Bengal. 2. Architectural terra-cotta—India—Bengal. 3. Terracotta sculpture, Indic—India—Bengal. 4. Architecture—India—Bengal—17th century. 5. Architecture and religion. I. Title. II. Series. NA6007.B4G55 2005 726'1'09541409032—dc22

iv

•

1 2 Illustrations

2004016533

3 4 5 10 09 08 07 06 05

For Didu (Reba Basu)

Illustrations

•

v

CONTENTS

vi

•

Illustrations

List of Illustrations Acknowledgments Note on Transliteration

ix xiii xv

Introduction

1

1 • Desire, Devotion, and the Double-Storied Temple

39

2 • A Paradigm Shift

65

3 • Acts of Accommodation

108

4 • Axes and the Mediation of Worship

137

Epilogue: A New Sacred Center

185

Glossary of Architectural Terms Notes Bibliography Index

201 203 229 247

Illustrations

•

vii

viii

•

Illustrations

Illustrations All photographs are by the author unless noted otherwise; drawings are by G. Murugan, The Landscape Company, Bangalore, India. Illustrations are grouped at the end of each chapter.

Introduction 0.1. Map of South Asia Showing Major Sites Discussed 25 0.2. Map of Bengali Cultural Region 26 0.3. South Façade, Shyam Ray Temple, Vishnupur 27 0.4. Temple No. 4, Barakar 28 0.5. Radha Ballabh Temple, Krishnanagar 29 0.6. South Façade, Keshta Ray Temple, Vishnupur 29 0.7. Kala Chand Temple, Vishnupur 30 0.8. Murali Mohan Temple, Vishnupur 31 0.9. Radha Vinod Temple, Vishnupur 32 0.10. South Façade, Madan Mohan Temple, Vishnupur 33 0.11. Radha Shyam Temple Compound, Vishnupur 33 0.12. Temple-Types of Bengal 34 0.13. East Façade, Keshta Ray Temple, Vishnupur 35 0.14. Domestic Hut, Vishnupur 35 0.15. Temples No. 1 and 2, Barakar 36 0.16. Celebration of Ratha at Madan Gopal Temple, Vishnupur 37 0.17. European Ships, West Façade, Keshta Ray Temple, Vishnupur 38 0.18. Dedicatory Inscription, South Façade, Shyam Ray Temple, Vishnupur 38

Chapter 1 1.1. Central Upper Pavilion, Shyam Ray Temple, Vishnupur 57 1.2. Terra Cotta Panel Depicting Kirtan, Central Upper Pavilion, Shyam Ray Temple, Vishnupur 57 1.3. Gokul Chand Temple, Gokulnagar 58 1.4. Central Upper Pavilion, Gokul Chand Temple, Gokulnagar 59 1.5. Plan of Shyam Ray Temple, Vishnupur 60 1.6. Kalanjay Shiva Temple, Patrasayer 61 1.7. Ratha Procession, Madan Gopal Temple, Vishnupur 62 1.8. Priest Sushanta Mukhopadhyay Attending upon Madan Mohan and Radha during the Annual Celebration Commemorating the Arrival of the Saint, Srinivas, Who Initiated the Vaishnava Transformation of the Region, Madan Mohan Temple, Vishnupur 63 Illustrations

•

ix

1.9. Priest Subrata Pujari Carrying Madan Mohan from the Altar to His Dining Room, Madan Mohan Temple, Calcutta 63 1.10. Panel Depicting Double-Storied Temple, South Façade, Shyam Ray Temple 64

Chapter 2 2.1. Tantipara Masjid, Gaur 95 2.2. Qadam Rasul, Gaur 95 2.3. Goaldi Masjid, Sonargaon 96 2.4. Jami Masjid, Bagha 96 2.5. Masjid at Kushumba 97 2.6. Jami Masjid, Atiya 98 2.7. Egaroshindur, Sadi’s Mosque 98 2.8. Adina Masjid, Hazrat Pandua 99 2.9. Motichura Masjid, Rajnagar 100 2.10. Ruin, Kulut 100 2.11. Domestic Hut, Birbhum 101 2.12. Domestic Hut, Birbhum 101 2.13. Plan of Keshta Ray Temple, Vishnupur 102 2.14. Porch Ceiling, Shyam Ray Temple, Vishnupur 103 2.15. Eklakhi Mausoleum, Hazrat Pandua 104 2.16. Jami Masjid, Salban 104 2.17. Terra Cotta Wall Panels, Jami Masjid, Bagha 105 2.18. Jami Masjid Interior, Bagha 105 2.19. Court Scene, South Façade, Keshta Ray Temple 106 2.20. Tiger Taming, South Façade, Keshta Ray Temple 107

Chapter 3 3.1. Radha Damodar Temple, Ghutgeriya 125 3.2. Radha Damodar Temple Doorway, Ghutgeriya 126 3.3. Mathurapur Deul, Madhukeli 127 3.4. Terra Cotta Ornamentation, Mathurapur Deul, Madhukeli 128 3.5. Ratneshvar Temple, Jagannathpur 129 3.6. Malleshvar Shiva Temple, Vishnupur 130 3.7. Jadab Ray Temple, Jadabnagar 131 3.8. Terra Cotta Ornamentation of South Façade, Madan Mohan Temple, Vishnupur 132 3.9. Nandakishor Temple, Dvadasbari 133 3.10. Malla Estate Garden Pavilion, Vishnupur 134 3.11. Family of Shiva, North Porch, Shyam Ray Temple 135 3.12. Goddesses, West Porch, Shyam Ray Temple 136

x

•

Illustrations

Chapter 4 4.1. Worshippers Gathered in the Courtyard, North-South Axis, Madan Gopal Temple, Vishnupur 158 4.2. Plan of Madan Mohan Temple Compound, Vishnupur 159 4.3. Plan of Gokul Chand Temple Compound, Gokulnagar 160 4.4. Natmandir, Madan Mohan Temple, Vishnupur 161 4.5. Natmandir, Gokul Chand Temple, Gokulnagar 161 4.6. Priestly Activities in the Kitchen, East-West Axis, Gokul Chand Temple, Gokulnagar 162 4.7. Plan of Madan Mohan Temple, Vishnupur 163 4.8. Plan of Gokul Chand Temple, Gokulnagar 164 4.9. Plan of Radha Madhav Temple, Vishnupur 165 4.10. North Sanctum Wall, Shyam Ray Temple, Vishnupur 166 4.11. West Sanctum Wall, Shyam Ray Temple, Vishnupur 167 4.12. Images of Chaitanya and Nityananda, Altar on North-South Axis, Radha Shyam Temple, Vishnupur 168 4.13. Image of Krishna as Radha Shyam, Altar on East-West Axis, Radha Shyam Temple, Vishnupur 169 4.14. Kirtan Performance, South Façade, Madan Mohan Temple 170 4.15. Drummers, South Façade, Madan Mohan Temple 171 4.16. Blind Doorway, North Porch, Shyam Ray Temple, Vishnupur 172 4.17. South Façade, Lower Story, Shyam Ray Temple, Vishnupur 173 4.18. Wall Frieze, East Façade, Shyam Ray Temple, Vishnupur 174 4.19. Culmination of the Mahabharata, South Entrance, Madan Mohan Temple 175 4.20. Bamboo Frame of Hut, Outskirts of Vishnupur 176 4.21. Mihrab, Tantipara Masjid, Gaur 177 4.22. Mihrab, Jami Masjid, Bagha 178 4.23. Mihrab, Jami Masjid, Kushumba 179 4.24. Central Mihrab, Sadi’s Mosque, Egaroshindur 180 4.25. South Doorway into Sanctum, Radha Vinod Temple, Vishnupur 181 4.26. South Doorway into Sanctum, Madan Mohan Temple, Vishnupur 181 4.27. Animal-Headed Motif, Madan Mohan Temple, Vishnupur 182 4.28. Rasamandala, South Façade, Shyam Ray Temple, Vishnupur 183 4.29. Krishnalila Panels, South Façade, Keshta Ray Temple, Vishnupur 184

Epilogue 5.1. Govindadeva Temple, Vrindavan 200 5.2. Laterite Ratha, Vishnupur 200

Illustrations

•

xi

This page intentionally left blank

xii

•

Illustrations

Acknowledgments

The writing of this book was as humbling an experience as it was exhilarating. This project would not have been possible without the assistance of several institutions. The research and writing of the dissertation was funded by a Social Science Research Council Dissertation Research Grant, an American Institute of Bangladesh Studies Doctoral Dissertation Fellowship, and a School of Arts and Sciences Fellowship and Schapiro Weitzenhofer Fellowship from the University of Pennsylvania. Subsequent revisions were facilitated greatly by an American Institute of Indian Studies Senior Fellowship, a J. Paul Getty Postdoctoral Fellowship in the History of Art and Humanities, and at the University of North Carolina from a Junior Faculty Development Grant, Research and Study Leave, and grants from the University Research Council and the College of Arts and Sciences. Various people and organizations in Bangladesh made the research possible. Institutional support was provided by the Government of Bangladesh Department of Archaeology and the field museums, particularly at Mahasthangarh. I owe an immense debt of gratitude to Perween Hasan, who has always been a role model, a mentor, and a friend. Susan Lee, Khaled Ashraf, Mrs. Amina Chowdhury, and Dr. Rokiya Kabir assisted in various ways. In India I am grateful to the Archaeological Survey of India for the generous access to monuments provided by Mrs. Kasturi Gupta Menon, Mr. Bimal Bandopadhyay of the Bengal Circle, and Shekhar Datta, Bholanath Chatterjee, and the many members of the Vishnupur Subdivision. I thank Chittaranjan Dasgupta, Secretary of the Vishnupur Sahitya Parishad, for his insights into terra cotta iconography, and his family, who recounted the tales of Vishnupur’s gods over many cups of tea. Achintya Banerjee shared his knowledge of the town and its monuments and copied many local publications for me. Without the assistance of Dilip Datta I would not have been able to conduct many of the trips to remote temples in Bankura. Shyambhu Mitra, Dr. Ramakanta Chakravarti, Debarshi Nandi, Dr. Doel Mukerji, and Dr. Nimai Choudhury Illustrations

•

xiii

facilitated the field trips and my thinking about the monuments I examined. The curators of the Asutosh Museum of Calcutta University and the Gurusaday Dutta Museum provided generous access to their collections. Prashant Bhat and G. Murugan of The Landscape Company, Bangalore, spent many hours helping to make the drawings that sometimes express the ideas of the book more cogently than the text. The monuments could not have been measured, nor their layers of reconstruction unveiled, without the enterprise and meticulousness of Gangadhar Das of the Archaeological Survey of India. His goodwill in the town got us access to the interiors of several living temples. He also helped me discover the town and made our stay during that project extremely enjoyable. This book is based on my doctoral dissertation, which was supervised by Michael Meister at the University of Pennsylvania and generously advised by Tony K. Stewart of North Carolina State University. I owe them an enormous debt of gratitude for their encouragement, thoughtful suggestions, and advice. My intellectual debt to Michael Meister in my understanding of the monuments is clear throughout this work. Tony Stewart guided my reading of Gaudiya Vaishnava literature and deeply shaped my understanding of that material. The revision of that dissertation has been enriched by wonderful suggestions and insightful readings provided by various scholars: Catherine Asher, Janice Leoshko, Padma Kaimal, Pallabi Chakrabarty, Leela Prasad, Joanne Waghorne, Dorothy Verkerk, Mary Sheriff, Ajay Sinha, Margaret Ewalt, Sarah Weiss, David Gilmartin, David Curley, Richard Eaton, Rebecca Manring, Tracy Pinchtman, Donna Wulff, Cynthia Atherton, and Debby Hutton. Rebecca Brown deserves special thanks for painstakingly going through the entire manuscript. Many scholars generously shared their knowledge and advice about publication, including Frederick Asher, John Cort, and Romila Thapar. I thank the anonymous readers and the committee of the American Institute of Indian Studies, particularly Susan Wadley and Brian Hatcher, who nominated the manuscript for the Edward C. Dimock Prize. My family has supported this project with enthusiasm since its inception. My grandmother accompanied me on most of the field trips to deserted temples and spent many months in Vishnupur, asking the most basic questions, which made me stop and reconsider my assumptions. My sister and brother-in-law took on several adventurous field trips and ran numerous errands toward the completion of this project. To my mother and my mother-in-law I owe a huge debt for doggedly pushing me to fulfill dreams they did not have the opportunity to pursue. Rai arrived at the tail end of the process and gave me a new perspective on the whole enterprise. Branavan read, heard, and responded to innumerable versions, and shared his faith from our graduate school years to the end. xiv

•

Acknowledgments

Note on Transliteration

For the sake of accessibility, diacritics have been avoided, and the names of sites and monuments are standardized in accordance with current popular use and the Archaeological Survey of India. The names of wellknown historical figures, deities, communities, and texts (for example, Chaitanya, Krishna, Vaishnava, and Ramayana) appear in their generally accepted anglicized form. The diacritized versions of Sanskrit technical architectural terms are provided in parentheses when they are used for the first time.

Illustrations

•

xv

This page intentionally left blank

xii

•

Illustrations

Temple to Love

Illustrations

•

xvii

This page intentionally left blank

xii

•

Illustrations

Introduction For the pleasure of Sri Radhika and Krishna a new bejeweled temple was given by Maharaja Sri Raghunatha Singh, son of Sri Vir Hambir, the king in the Malla Saka year 949 [1643].1

This proclamation greets us as we approach the main entrance to the Shyam Ray Temple, one of the earliest monuments in the town of Vishnupur in Bankura district, West Bengal (figures 0.1, 0.2, 0.3, 0.18). The inscription boldly asserts that the temple is new by using the phrase “new bejeweled temple” (navaratna ratnam), and unveils the new Ratna (jewel) typology.2 As architectural innovations reformulated the region’s traditional temples, this new multi-faceted form emerged, reflecting—like a jewel—the remarkable creativity and cultural fluidity of the period. These seventeenth-century monuments have remained unnoticed for the most part, occupying the margins between architectural formations that have been given pride of place in the mapping of South Asia’s artistic heritage. As one art historian recently observed, they are no more than a footnote in South Asian architecture, and indeed few introductory textbooks deem them worthy of an illustration. (In chapter 2 I address the historical conditions that produced such a canon.) However, cultural interstices are often the site of some of the most exciting and creative interactions, as much recent scholarship, new historicism, and subaltern studies, to name a couple of strands, have shown us. In the case of seventeenth-century Bengal, the energetic architectural experimentation left its mark on two major empires: that of the Mughals, who dominated north India during the sixteenth and seventeenth centuries, and later that of the British. To the Mughals it provided the bangla3 that Emperor Shah Jahan employed to frame his image as a cosmopolitan world ruler, and for

Introduction

•

1

the British it inspired the ubiquitous bungalows that gradually dotted, and even plotted, their domination of South Asia through the eighteenth and nineteenth centuries. This book is my attempt to bring these temples to the fore, to promote an appreciation of the originality that their inscriptions marveled at. From the late sixteenth to the seventeenth century, architects experimented with the region’s traditional curvilinear-towered temple form, which belongs within mainstream north Indian Nagara (Na¯gara) temple construction (figure 0.4). These experiments culminated in the development of a completely new temple-type that stands apart from that preexisting tradition. Features of mosques and local thatched huts were combined to form a double-storied temple-type that served newly developing religious traditions. It also served the needs of local Hindu rulers for visual and symbolic self-definition in a world no longer dominated by Hindu monarchs. As Mughal presence was gradually established after the conquest of Bengal in 1575, local Hindu landholders took advantage of their distance from the imperial capital at Delhi and the delays and difficulties of establishing a provincial administration. During this time they enjoyed a remarkable degree of independence. They employed sacred architecture to assert their role in the dynamic Indo-Islamic political culture of the region, and to redefine it for themselves. The inscription cited above also points to the reconceptualization of temples as the pleasure grounds of the gods, where Krishna and his beloved Radha could rekindle their passion. Krishna is the focus of Gaudiya Vaishnavism, the bhakti (devotional) movement led by the Bengali saint Chaitanya (1486–1533) that swept up Bengal and Vrindavan in north India in a frenzy of ecstatic devotion and passionate song and dance over the sixteenth and seventeenth centuries.4 Chaitanya advocated a deeply emotional and intensely personal engagement with Krishna, modeled on Radha’s single-minded dedication. The inscriptions’ deployment of terms such as mudita (pleasure, happiness, delight) and rasa (taste, deliciousness), given specific connotation in a century of literary texts, leaves no doubt that the monuments were dedicated for the expression of this intensely passionate love shared by Radha and Krishna. Two terra cotta depictions of the divine lovers flank the text. On the left Krishna plays his flute for Radha, while a devotee kneels at their feet. On the right, however, is a less conventional image of Krishna seated with Radha on his lap. He lifts her chin with one hand to draw her closer. The intimacy of the amorous scene makes the text’s claim explicit. The inscription thus heralds the new architectural form as the immanent site of Krishna and Radha’s divine play (lila), the object of the devotee’s aspiration.5

2

•

Temple to Love

This new epigraphic convention marks a shift in what a temple meant to Bengal’s Gaudiya Vaishnava community at the time. It diverges from earlier inscriptions that typically emphasized the role of the king as patron and his personal relationship to the primary deity, praised the donor’s generosity in funding the monument, and indicated the specific end sought through the act of patronage.6 The formula embraced at Vishnupur, one focused on Radha and Krishna’s mutual pleasure, was repeated for more than a hundred years with minor variation, emphasizing through the epigraphs the purpose of the new “bejeweled” architectural form. Over a hundred red brick and terra cotta Ratna temples studded the lush green delta, from Midnapore and Bankura districts in southwestern West Bengal to Jessore and Khulna in southern Bangladesh, during the seventeenth and eighteenth centuries.7 Monuments surviving in relatively good condition at sites such as Vishnupur in Bankura district, Ghurisha in Birbhum district, and Krishnanagar, Bansberia, and Guptipara in Hooghly district in West Bengal share significant commonalities and can be treated as a regional architectural corpus (figures 0.2, 0.5). The end of Mughal rule can be conveniently used as an approximate date to separate these temples from those built subsequently, when increased European contact and then British political hegemony introduced further complexities in temple form and style. This study revolves around a smaller group revealing the early burst of architectural experimentation. I focus on the structures standing at Vishnupur, where over thirty temples were built in the seventeenth and eighteenth centuries. The town is thus an experimental laboratory, driven by the religious enthusiasm of Gaudiya Vaishnavism and sponsored by the Mallas, a dynasty of local kings that came to power at the end of the sixteenth century and held sway over most of Bankura district during the next hundred and fifty years.8 Most of these monuments are dated and therefore facilitate discussions about the beginnings of this new architectural style. In addition, the circumstances surrounding their patronage and use are better known than at other sites. Therefore, while I will discuss the broad range of Ratna temples in the Malla territories and beyond for comparative reasons, the primary monuments addressed in this book are the Shyam Ray Temple of 1643 (figure 0.3), the Keshta Ray Temple of 1655 (figures 0.6, 0.13), the Kala Chand Temple of 1656 (figure 0.7), the Lalji Temple of 1658, the Madan Gopal and Murali Mohan Temples of 1665 (figure 0.8), the Radha Vinod Temple of 1659 (figure 0.9), the Madan Mohan Temple of 1694 (figure 0.10), the Radha Madhav Temple of 1737, and the Radha Shyam Temple of 1758 (figure 0.11). Spanning the years from the mid-seventeenth century through the mid-eighteenth

Introduction

•

3

century, these temples enable investigation into the formation of the Ratna temple form while they illustrate the new space for Krishna and Radha’s pleasure gardens within the religion, politics, and social organization of Bengal during these years. This book examines the monumental transformations in the purpose and formulation of Hindu temples claimed in the inscription’s assertion of originality. The architecture reveals radical changes from the way in which temples were organized earlier. As a series, the temples disclose various experiments that the architects conducted to incorporate a festival pavilion, and therefore point to the transition from a single- to a twostoried form. The addition of a second altar in the sanctum below resulted in a reorientation of temples from the conventional east-west axis to a north-south one that is unusual. Facing a central courtyard, the new south-facing altar in the sanctum provided the focus of collective ritual activity, while the east-west axis was converted into one of priestly services. Further, porches were gradually closed off to provide space for more private worship. These changes culminated in the organization of a temple compound with multiple structures. A close scrutiny of the buildings hand in hand with observation of current practice thus also reveals the ways in which the new temples shaped ritual worship. The choices in the content and organization of terra cotta imagery on the monuments’ façades draw attention to these novel elements, as do the confident declarations of their inscriptions. These distinctive architectural features have never been analyzed before, not least because entry into many parts of the temple complexes is difficult. To protect the sanctity of the deity’s space, priests demarcate the threshold beyond which devotees may not enter, and the attendants of the Archaeological Survey of India stand guard to protect the monuments from vandalism. Unprotected monuments have often fallen into such ruin that their ceilings and stairs have given way. Further, some sites are remote and difficult to reach, off the arterial roads and railway lines. They require an art historian to wade across streams with a camera and tripod, or to crawl through the dense undergrowth of teak forests. Consequently, the inaccessibility of sites and of the entire temple compound has contributed to the neglect of their architectural properties in both popular discussion and scholarly literature. In attempting to correct this, this study fills a gap in the scholarship on eastern Indian art and architecture, between the literature on the earlier Pala sculpture and Sultanate mosques and that on later Bengali art that expresses a colonial world and then resists British power. There are, however, many other problems in attempting to produce a narrative surrounding these temples, and these too may well have de-

4

•

Temple to Love

terred scholarly commitment to them. For a start, the examples that remain standing are chance survivors from a much larger corpus of monuments that have perished for a variety of reasons, including the merciless monsoon rains and the dense vegetation they bring with them. When I walked around Vishnupur for the first time in the summer of 1994, I recall feeling distinctly dismayed at the vigorous banyan trees erupting through brick and mud ruins. I seemed to encounter them in larger numbers than the beautiful, well-preserved temples. The ruins are living testimony that there were probably many more innovations and variations that we can no longer discern or appreciate. Likewise, three hundred years have left their mark on the survivors. Most have been altered, modified by generations of users who made the space their own. A summer of measuring the structures with an architect-surveyor revealed that doors have been closed off, porches blocked, and halls transformed into shelters for the local homeless and “madmen.” Scratching through the lush green gardens and floral beds, I found rubble that marked courtyard floors and foundations of accessory structures that have fallen away. Such changes hide the architectural patterns that only a neardecade of searching had taught me to look for. Likewise, the temples disclosed the dramatic shifts in appearance caused by the stripping of the heavy white lime plaster that protected the terra cotta surfaces. However, a few temples that continue to be owned privately are repainted regularly, in vivid colors that highlight the terra cotta sculptural ornamentation (figure 0.5).9 They remind us of the aesthetic shifts that occur over three hundred years. They also point to the divergent interests of devotional communities, who wish to maintain a ritually potent building, and archaeological preservation, which is driven by a modernist preference for terra cotta rather than brightly colored paint. They reiterate the lessons learned from the Parthenon—that although monuments standing before us offer a tangible experience of the past, they yet remain tantalizing, for there is no unmediated view into the time and place from which they came. Inevitably, my narrative is a conscious selection, arrangement, and interpretation from this fragmentary and often altered archaeological record. Precarious though the process admittedly is, in conjunction with material available from a variety of other sources, and drawing upon my art history training and lived experience in Bengal, I began to discern broad patterns. Though some monuments inevitably seem to contradict any attempt to identify patterns in cultural phenomena, as a group they allowed me to imagine how architecture was defined by and also shaped the devotional activities around which a community cohered at Vishnupur in the seventeenth century. After my first encounter with these monuments and their inscriptions,

Introduction

•

5

I looked for texts that could help me understand the historical conditions of their making. Those who work on premodern Bengal are only too aware of the lack of documentary evidence regarding who produced buildings and images, and under what circumstances. No contracts or memoirs survive, and we have to correlate tangential references in the literary genres and inscriptions with the visual evidence of architecture and terra cotta imagery. For this period these textual sources are several. Among them, the sacred literary works produced by the Vaishnava religious community were devoted to asserting its distinctive theological propositions, producing hagiographies that celebrated Chaitanya and other spiritual leaders, and composing songs that in their lyricism match the emotions of Krishna and Radha that they describe. In addition, what is now called the mangalkavya genre was a performance tradition glorifying deities such as Chandi and Manasa. Its oral component was gradually transformed into written texts that provide material for uncovering the processes of social change that the authors comment upon. The sacred biographies provide glimpses of the larger religious processes and networks within which Vishnupur’s architectural formation occurs, and the mangalkavya narratives suggest the transition from a forested to an urban world, and from a tribal to a “Hindu” one. Although none of the major works were composed at Vishnupur, they provide a broad background against which I situate the architectural experimentation. Because I am an art historian, my perusal of these texts is preliminary, particularly when compared to the work of scholars who devote lifetimes to these texts, and indeed, I turned to these scholarly interpretations as guideposts in my forays. When I returned to Vishnupur with some of the larger historical processes in mind, I followed the inscriptions’ direction and sought out the gods for whose pleasure these monuments were created. The presence of Vishnupur’s many forms of Krishna is felt deeply to this day, and the town’s geographical and architectural features are tied closely to narratives of their miracles. I learned these stories when priests or attendants narrated the episodes that occurred at their temple to gatherings of pilgrims, and from local devotees who regularly came to the temple, usually when the crowd dispersed. Various other residents of the town also related them to me. Mrs. Salil Singh, one of the four daughters-in-law of Vishnupur’s last king, for example, stood at the doors of the palace, pointed to the dusty red road winding from the stone gateway to the palace, and told me how one of these Krishna images, Madan Mohan, had come galloping on his horse down this path. I also became familiar with the town’s oral lore from local tour guides, librarians, schoolteachers,

6

•

Temple to Love

Archaeological Survey attendants, and some of the women who befriended me. Gangadhar Das of the Archaeological Survey office cautioned me to tread lightly in the seemingly abandoned temples, for their deities had still not forsaken them. Through him I learned of their nocturnal activities and people’s unsuspecting encounters with them. Later, when I began to collect the cheap paperbacks and pamphlets circulating in Vishnupur’s bazaars, I encountered other versions. These local Bengali texts range from pilgrimage guidebooks that focus on describing the sites of the gods’ accomplishments to mahatmyas, verses of praise glorifying Vishnupur’s major deities. Together, they allowed me glimpses into how the temples provided suitable sites for the gods’ enjoyment. I also observed daily worship and festivals performed in living temples to understand the role of the rituals and the monuments in the life of the local community. I spent many lunch hours eating bhog, the leftovers of the gods’ meal, and numerous evenings attending singing sessions and trailing festival crowds and processions with my camera. At these occasions I also had the opportunity to ask the participants about their experience and how they understood it. I have relied on these traces from current practice to illuminate the architectural forms raised over three hundred years ago because this kind of material is unavailable from any texts surviving from the period. Many of my conclusions about the ways in which architecture defined, and was in turn dictated by, ritual practice are drawn from piecing together this ethnographic evidence and other less direct clues gleaned from the region’s folklore and textual descriptions. Some of the temples’ communities claim a history of unbroken affiliation, as succeeding generations of priests, caretakers, musicians, garland makers, and potters claim to have served the gods from the time the Malla rulers established the institutions and appointed their families. Such links to the past suggest that despite inevitable changes in practice and meaning, the shared elements begin to allow for a cautious reconstruction of the conditions that motivated the dramatic spatial reorganization undertaken in temple architecture of the seventeenth century. The devotional songs performed in the temples also yield glimpses into the meanings of the activities and the monuments that are unavailable from other sources. A wealth of songs written in Bengali even before the temples were built give ideal descriptions of Radha and Krishna’s divine love affair as it was conducted in the idyllic Vrindavan. The vividness of these descriptions provides verbal counterparts to the images that sheath the temple walls and allowed me to imagine how the visual imagery may have been deployed for the re-creation of Krishna’s pleasure. The songs are themselves products of, and continue to be used for, the

Introduction

•

7

mental re-creation of these events by devotees. Together with the depictions of devotional experience in the Vaishnava literature of the sixteenth and seventeenth centuries, the words of the songs provide an invaluable resource for gaining access to the cultural imagination that also produced the monuments where the events were ritually re-enacted. Before approaching that cultural process, however, it is first necessary to identify the architectural forms that constitute this study.

Temple Forms Three major categorizations of Bengali temples are in popular use at present and have been absorbed into the scholarly literature. These are called Chala, Nagara (locally called rekha), and Ratna (figure 0.12). Perhaps the region’s most distinctive contribution to temple architecture is the chala, a form of temple based very closely on the Bengali village hut with its thatch roof (compare figures 0.13, 0.14). The simplest such hut (chala) has a square or rectangular base and a bamboo skeleton in its walls, which is filled with reeds or matting and reinforced with mud. The thatched roof is also supported on a bamboo frame. This bamboo frame provides the distinctive curved upper ridge and bottom rim to the roof that is mimicked in Bengali mosques and temples. While this thatch roof is an ancient form, in existence at least from Vedic times, its use as a source for monumental architecture does not seem to have preceded the period of Sultanate rule, beginning in the thirteenth century, in Bengal. Thatched roof-types are imitated in the construction of Sultanate monuments from the fourteenth century, and by the seventeenth century the Vaishnava temples also begin to reproduce the basic features of the huts. Chala, one name for these huts, is applied to these temples in popular usage and local literature. Specifically, chala is thatch, and a thatched roof or covering.10 With reference to these temples, it is used to describe a sloping roof with curved ends made in stone or brick that imitates the thatch original. The two basic roof forms used in chala temples are the two-sided (that is, a gable roof consisting of two elongated eaves that converge at a curved ridge) and the four-sided (where four triangular eaves converge to a point at the top). The first is used to cover a rectangular structure, the second a square. Common parlance calls these forms do-chala (two eaves) and char-chala (four eaves). More complex combinations are used for seventeenth-century temples. An at-chala (eight eaves) is constructed in two levels, with a four-eaved roof on the lower structure surmounted by a smaller four-eaved upper structure (figure 0.5). The Radha Vinod temple at Vishnupur imitates this type (figure 0.9). When two gable-roofed huts

8

•

Temple to Love

are juxtaposed, sharing a common long wall, the name given is jor bangla, meaning twin Bengali huts (figure 0.14). The Keshta Ray Temple at Vishnupur is an example of this type (figure 0.13). A second category is a curvilinear-towered temple called Nagara in the scholarly literature on South Asian temples. This tower is ornamented with vertical bands of vines or lata (lata¯). This latina Nagara temple was dominant in Bengal and throughout north India from the seventh century onward. Michael Meister has defined the form in the following way: Curvilinear in outline, made up of laminated planes, this type of tower is marked into “storeys” by ribbed a¯malaka stones on the corners and divided into vertical “creepers” (lata¯s) by offsets, established in the plan and extending into the superstructure. This unified “Latina” (formula with lata¯s) temple spread rapidly across north India and into the Deccan by the seventh century as a viable and vital symbol of emergent Hinduism.11

In Bengal, only a handful of these single-towered temples survive from the reign of the Pala and Sena dynasties (eighth–twelfth centuries) and from the following period of Sultanate rule, the thirteenth to sixteenth centuries (figures 0.4, 0.15). The few that remain are located mostly in the southwestern part of Bengal, between the Ajay and Damodar rivers in the north and the Rupnarayan in the south (the districts of Puruliya, Bankura, and Burdwan, and parts of Midnapore, Hooghly, and Howrah). In this region, these temples are called Rekha Deul (curved-spired palaces of the gods), a terminology shared with the neighboring region of Orissa. In this case, the curvilinear shape of the tower is the defining characteristic used to name these monuments, rather than the vertical ornamental bands of creepers. Ratna, the third categorization commonly used for seventeenth-century temples, is less easy to define. The term is often used to identify temples with spires, but is also used to label the shikhara (´sikhara, spire) as well. In this book I turn to the dedicatory inscriptions of these monuments, where the designation ratna is applied fairly consistently, preferred over other general terms such as devakula, prasada, mandapam, and mandiram (house of god, palace, pillared hall, and temple, respectively) that are used in inscriptions on temples of other forms. Conversely, this word ratna does not appear to have been used for monuments that do not take the two-storied form. Such usage suggests that the term was likely used to differentiate these two-storied temples from the chala and curvilinear-spired Nagara at the time of their construction. The inscriptions seem to specifically differentiate Ratna as a new temple-type. The dedicatory inscription on the five-spired Shyam Ray

Introduction

•

9

Temple proclaims the structure “navaratna ratnam.” The profound significance of that phrase, defining Ratna as a new temple-type, has been obscured by its current loose interpretation as simply “exquisite gem of a temple,” without any rigorous analysis of the architectural nature of the structure. Such a translation evades the dilemma created by the words nava (which means both new and nine), and ratna (which means both jewel and ornament). In earlier scholarship, the number of spires has been a major concern in the discussion of these monuments. Because nava also means nine, prior translations called these temples nine-spired, when in fact they have one, five, or any other number of spires. The double meaning of the term nava thus diverted scholarly attention from the truly innovative upper shrine to a counting of spires. A. K. Bhattacharyya was the first to translate nava as “new” and read the inscribed phrase navaratna ratnam as “‘jewel’ among ‘new jewels’ i.e., a best one among the new temples.”12 Coming to the material from an epigrapher’s perspective, he did not attend to the crucial architectural distinction between a spire and a whole temple. He therefore did not recognize the full significance of this phrase in proclaiming a new architectural form. In keeping with Bhattacharyya’s translation of nava as “new,” I interpret the adjective navaratna as “new-spired,” which leads to a rereading of the phrase navaratna ratnam as “new-spired temple.” Here the noun ratnam refers to the temple. The adjective new (nava) thus applies to both the spire (ratna) and the temple (ratnam), that is, a new-spired Ratna temple-type. Consequently, I read the inscription to suggest a new temple-type resulting from the use of a new upper-storied structure that is a full temple in miniature rather than simply a spire. The inscription thus clarifies that this additional full story is what makes these temples new and distinctive as double-storied formations, and worthy of a new name. I therefore understand the Ratna temple as having a single lower story that acts as a base, with one or more miniaturized shrines, each with its own roof, placed on the upper terrace above the sanctum (figures 0.3, 0.5, 0.6, 0.7). The upper shrine acts as a cap for the sanctum. More significantly, the upper structure of the double-storied temple is a complete miniature temple, with its own spire. The Ratna temple’s upper structure has an interior space that is approachable by stairs, and arches open this shrine to the outside. Accessibility through the stairway and visibility through the arches allow the upper structure to function in an unprecedented way as a shrine for ritual performances. And perhaps because of this unique feature, the temple that incorporates it has received the designation of “new.” This second level takes a wide variety of different forms and is thus distinct from the lower story, which developed a consistent formula based

10

•

Temple to Love

on indigenous hut and Sultanate architectural traditions in the seventeenth century. Similarly, in ritual use, the lower story retains the conventional functions of a Hindu temple as a cave-like womb (garbha) for the deity residing in its sanctum (garbhagr. ha), while the upper structure takes on particular ritual festival roles for the Gaudiya Vaishnava tradition. In chapter 1 I suggest that the ritual use of this additional structure contributed as much to the creation of the new Ratna temple-type as its doublestoried form. This function at least partly dictated the development of the form, and this newly available extra space was then put to creative use. I use the term Ratna for this distinctive temple-type, with its corresponding specific functions. In redefining Ratna in this way, I have rejected the importance given to shape, number, and origin of the roofing elements, which were previously used as the primary criteria for defining seventeenth-century templetypes. David McCutchion, for example, used the term ratna to identify and count the number of spires. In distinguishing the ratna temple from the chala, he wrote, The pinnacled or ratna design has the same lower structure as the chala series—a rectangular box with curved cornice—but the roof is more or less flat (following the curvature of the cornice), and is surmounted by one or more towers or pinnacles called ratna (jewel).13

Here he significantly recognizes that the Ratna temple must have an upper structure built above a complete lower temple-structure with its own roof. However, he does not investigate its functional consequence for the region’s architectural and religious traditions, choosing instead to focus on a formal typologization: The principle of decorating the tower with miniature temple duplicates is very ancient in Hindu art; in the southern temple style they are boldly outlined, in the northern temple style they merged into the curvilinear shikhara.14

Rather than pursuing the significance of the ratna temple as a whole, he first explains its spires as derivations from latina and chala models and then classifies them by the numbers of sub-spires: ekaratna (one ratna), pancharatna (five ratnas), navaratna (nine ratnas), etc. In contrast, my study draws attention to the whole structure, the new two-storied temple formula, with separated upper shrines that accommodate specific ritual practices. The upper shrines can be of different forms, and there can be many of them. In redefining Ratna in these terms, I incorporate the heretofore distinct at-chala and jor bangla types built at seventeenthcentury Vishnupur, focusing on their double-storied organization rather than roof-type (figures 0.5, 0.13).

Introduction

•

11

Gaudiya Vaishnavism and Temple Ritual Almost all the surviving seventeenth-century monuments are dedicated to Krishna, indicating the critical role that Gaudiya Vaishnavism played in the emergence of this new temple-type.15 The forms of Krishna enshrined in these temples express the Vaishnava emphasis on his relationship with Radha. The names of Vishnupur’s temple deities, such as Radha Vinod, emphasize Krishna’s attractiveness (vinod) to Radha. Likewise, Radha Raman is the captor of Radha’s heart. Others, like Madan Mohan and Madan Gopal, assert his inherently alluring nature, which attracts not only Radha but also the devotional community. As Madan Mohan, Krishna seduces even Madan, the god of love; as Madan Gopal he is the cowherd (Gopal) identified as the god of love himself. In giving names to the temples, these deity images, like the epigraphic statements, point to the new understanding of temples as the sites for exploring Radha’s relationship with this charming cowherd god. Devotees seek to comprehend this divine love affair through worship practices that enable them to engage in an intimate relationship with Krishna. Gaudiya Vaishnava worship conceptualizes five emotional or relational states for experiencing Krishna, and these five states are graded in a hierarchy of increasing intimacy. They are santa, the peaceful condition in which the worshipper regards herself or himself as low and insignificant in relation to the supreme, omnipotent deity; dasya, the condition of devotion toward one’s master (as Arjuna served Krishna in the Bhagavad Gita); sakhya, the state of friendship; vatsalya, a relationship of caring affection between mother and child; and madhurya, the sweetness of love, such as that Radha had offered him. These five states are modeled on Krishna’s relationships with the people in his life, as described in the stories making up the Bhagavata Purana, the tenth-century text regarded as divine revelation by the Gaudiya community. They include the parental affection of his foster parents Yashoda and Nanda, the brotherly love of Balarama, the comradeship of the gopas (cowherding men), and the passionate love of the gopis (cowherding women). Of these, Radha’s passion was upheld as the most complex and intense, and thus the most satisfying form of love to both the devotee and Krishna. It is the polar opposite of aishwarya, the awe experienced before the magnificent but remote Vishnu-Narayana. Such an omnipotent and omniscient deity is not loved intimately but honored through ritual activity. Radha’s capacity to long intensely for Krishna, to serve him, and to risk everything for him was instead regarded as the ideal emotional experience for devotees to emulate.

12

•

Temple to Love

Married to another man, Radha had to undergo great difficulties, take huge risks, and compromise her position in society for the sake of her love for Krishna. As Edward Dimock has pointed out, love has an element of self-sacrifice: “only when the heart is given unconditionally . . . is such love pure and thus efficacious.”16 A second aspect of Radha’s love emphasized in this tradition is the pining inherent in the separation (viraha) from her lover. Because an adulterous relationship lacks predictability, much of the experience involves anticipation and expectation.17 This longing and sacrifice was identified as the path to salvation. Parakiya, the doctrine celebrating these especially delectable qualities of an illicit love, was celebrated in Vishnupur’s temple as the greatest sweetness to be relished, and temple architecture and ornamentation was oriented toward articulating these values. Chaitanya initiated a distinctive style of worship that provided devotees with the opportunity to participate in Radha and Krishna’s playful activities, and this devotional practice was so potent that it gradually culminated in the reorganization of temples as the arena for its performance. He led his followers in kirtan, songs praising Krishna and describing the episodes and emotions of his relationship with Radha. These sessions lasted through the night, and the songs were so powerful that they evoked visceral responses of not only dance, but even tears, trembling, and fainting. Gradually the processions overflowed through the streets of his hometown, Nabadwip, and soon the singing (nagar samkirtan) made sixteenth-century Bengali towns resound with the name of Krishna: When Prabhu gave the order to the people of the city, they began to make kı¯rtana from house to house. “I bow to Hari, I bow to the Ya¯dava Kr·s.n.a, to Gopa¯la, Govinda, Ra¯ma, and S´rı¯ Madhusudana.” Sam . kı¯rtana, with cymbals and mr·danga drums rang out, and nothing could be heard except for the sound “Hari Hari!”18

In the seventeenth century, Vishnupur’s temple courtyards provided an alternative venue for these ecstatic expressions of devotion and played a pivotal role in bringing the Vaishnava community together. These practices were brought to Vishnupur by Srinivas Acharya, one of the foremost leaders of the movement in the seventeenth century. Here he converted Vir Hambir,19 the earliest historical ruler of the local Malla dynasty, his family, the court, and a substantial segment of the population of Mallabhum, the Land of the Mallas.20 Srinivas was of the third generation of Vaishnava leaders, who were given the responsibility of consolidating the religious community in Bengal at the end of the sixteenth century. As Tony Stewart has recently

Introduction

•

13

elaborated, the movement had lost momentum as Chaitanya’s various companions proposed their own interpretations and organized their own lineages. Yet the tradition flourished in Vrindavan, where the goswamis, disciples handpicked by Chaitanya, compiled the theological texts of the tradition. In anticipation of the imminent dissolution of the religious tradition in Bengal, these leaders sent three energetic and charismatic students to rectify the situation. The trio, Srinivas, Shyamananda, and Narottama, accomplished this mission in part by aligning themselves with Hindu chiefs who were also looking for opportunities to establish their authority in the flux created by the collapse of the Bengal Sultanate.21 At Vishnupur, Srinivas’s collaboration with Vir Hambir culminated in the transformation of forests into the bowers of Krishna’s play. Here he taught about Radha’s selfless and all-consuming love and led the local community in kirtan that brought her experiences alive. Such was the intensity of his devotion that he would be transported to the green groves of Vrindavan, where Radha and Krishna danced or bathed in the waters of the River Yamuna. In one mystical vision he observed Radha’s nose ring slip off while they were splashing around, and her maid Manimanjari search assiduously through the grains of sand underfoot. Srinivas seems to have been so inert, engrossed in this visualization of the divine realm, that the king and his physicians believed he was dying. However, three days later he awoke, exclaiming with tears in his eyes that Radha’s ornament, trapped in the roots of a water weed, had been retrieved.22 His reveries also brought the eternal play of the gods into Vishnupur’s landscape, and his experiences provided a model for the emergent community of followers. Inspired by his devotion, the kings constructed a succession of temples that created the space and aesthetic for celebrating the activities of the divine lovers. The new temples, with their capacity to house large gatherings, were sites of the collective devotional activities that assisted in the formation of a Gaudiya community. Alongside temple construction, Srinivas had also mobilized his patrons’ resources toward organizing festivals, feasts, and fairs (mahotsav) commemorating events in the life of Krishna, Chaitanya, and other key members of the Gaudiya Vaishnava leadership. Following the success of the early gatherings in the western part of Bengal, Srinivas and his colleagues Narottama and Shyamananda prepared a more elaborate affair at Kheturi, Rajshahi district, where goddess worship held sway.23 Here they demonstrated how to serve the images, perform arati (the evening service), and sing kirtan honoring Chaitanya as well as Krishna. Doljatra, the festival of colors, was celebrated by smearing the images and then participants with red powder.24 The Ratna temples, in

14

•

Temple to Love

their capacity as local congregational centers, facilitated the replication of these activities at the local level. The success of the strategy can be discerned from Srinivas’s declaration that three new Vaishnava centers had emerged at Kheturi, Jajigram, and Vishnupur.25 By this time a three-layered mapping of the relationship between Krishna, Chaitanya, and Srinivas Acharya was developed. Chaitanya was identified as an avatar or incarnation of Krishna in his own lifetime, and Gaudiya Vaishnavas were thereby given immediate access to Krishna. At the end of the sixteenth century, after Chaitanya passed away, Srinivas Acharya was recognized as an incarnation of Krishna-Chaitanya.26 The immediate access afforded to the devotees present when Chaitanya was physically alive was now made available to devotees of the next generation, while the layering of identity simultaneously legitimated the authority of Srinivas. This Krishna, embodied by Chaitanya and later by Srinivas Acharya, resides in the Ratna temple’s sanctum and revels in the adoration of his devotees. The temples operate as a site for the reinforcement of these conflated identities, which are expressed through visual imagery and ritual practice. The populist style of worship with exuberant song and dance that Chaitanya launched and Srinivas brought to Vishnupur is modeled on Krishna’s own playful activities (lila) in Vrindavan. These paradigms are invoked in temple rituals, and they create sensory linkages with the original divine performance. Such overlaying of beliefs and practices made the divine visible to devotees with knowledge and divine grace. In the deity images, terra cotta vignettes, songs, and consecrated food, they are able to see Krishna and participate in his eternal play. These visual and verbal elements of temple activity are therefore fundamentally commemorative in nature, re-enacting the original activities of the gods.27 The architectural form, terra cotta ornamentation, and ritual use of the new temples stimulate and refine the imagination and emotion of the devotional community. Uninitiated viewers, however, do not have such access to the divine lila. As institutional sites, temples therefore define the nature of participation and set up boundaries that demarcate devotional communities by exclusion as well as inclusion. In examining the layering of meanings, this study thus also explores how the architecture, iconography, and worship practices of Ratna temples mark the threshold between two alternate perceptions and mediate between the manifest (prakat) and the hidden (aprakat). The triumph of these practices and of the new temples where they were undertaken considerably undercut the region’s older conventions and the architecture associated with them. The menace these new spaces

Introduction

•

15

and their personal style of devotion posed to conservative Brahmanical Hinduism, with its focus on priestly ritual, suggests the success of the new Vaishnava approach. As Vaishnava worship curtailed practices such as blood sacrifice, it threatened the traditional authority of Brahman priests. An orthodox Brahman community, for example, complained to its ruler, Whence have these upstarts brought the Vais.n.ava creed? The worship of our gods and goddesses is abandoned. The sacrifice of animals at the altars of our temples is stopped. The sacred rites are abolished, and we are undone. They do not touch meat or fish and live on vegetable food. They form Kı¯rtana parties; they dance and cry like mad men. The rites enjoined by the Vedas and the Tantras are abandoned. They have charmed the people by their songs and music.28

Their list of protests offers a dramatic comparison between the new rituals performed in Ratna temples and the older ones. The Brahman orthodoxy felt the Vaishnava practices of song and dance lacked the scriptural authority of the Vedas, Puranas, and Tantras. The rhetoric of deeply personal devotion and of equal access to the divine for everyone through song, dance, and mystical visions flouted the caste hierarchy that affirmed Brahman authority.29 Further, the songs were composed in the Bengali vernacular rather than the more esoteric Sanskrit that was unfamiliar to the singers. The Vaishnava shift to vegetarianism implies abandoning of animal sacrifices, an arena over which Brahmans presided. The diminished role of sacrifices also suggests that the new devotional community no longer gave to Brahmans the offerings that had constituted their livelihood.30 As local goddess cults had enjoyed wide popular support, the Vaishnavas cut into the Brahmans’ traditional clientele. Such observations reflect the resentment and anxiety amongst the more conservative Brahmans of Bengali society, who discerned a threat to their privileged status.31 In this expression of dislike, however, lies a grudging acknowledgment of Gaudiya Vaishnavism’s infectiousness and sensual appeal that bewitched throngs of people. By making space for the rousing kirtan sessions and mystical experiences to which the Brahmans objected, the new temple form played a significant role in establishing Gaudiya Vaishnavism as a populist movement. The redesigned temples with their increased capacity provided the major venue for public activities that brought the community together in the seventeenth century. The construction of temples and the performance of kirtan and festival celebrations therein are therefore processes that transformed the Bengali landscape as irrevocably as did the concurrent consolidation of the Mughal empire.

16

•

Temple to Love

Such expression of personal devotion did not eliminate the role of Brahman priests, however. Through the following chapters I analyze how the redesigned temple mediated between the two primary constituencies by providing separate arenas where the lay community and priests played their roles in the lives of the gods. As Ramakanta Chakravarti has noted, implicit in the act of temple patronage is the affirmation of the priest’s position as caregiver, because the temples house deity images that require daily acts for their maintenance.32 Daily service was extended in the seventeenth century to musical re-enactment of the divine lila by the community alongside priestly offerings of food and flowers. The spatial reorganization of Ratna temples, by the addition of an upper pavilion, an enclosed courtyard, and a second altar in the lower sanctum to incorporate these group activities, finely balances the roles of the two constituencies in the daily and festival cycles of worship.

Kings and Construction The architectural experimentation that gave rise to the Ratna temple form was made possible by the enthusiasm and generosity of local Hindu rulers. Constructing temples affirmed their authority in various ways. Contemporary literature such as Mukundaram’s Kavikankan Chandi suggests that temple building must be understood as part of the process of staking a claim to the urban cosmopolitanism and Brahmanical ritual order, a process often termed Sanskritization. Kalketu, the poem’s hero, is a tribal chief who scraped a precarious living from hunting. With the collaboration of the goddess Chandi, he constructed the mythical kingdom of Gujarat, over which he presided as ruler. This tale of upward social mobility resembles the narrative of Malla consolidation of a frontier principality with Vaishnava collaboration in the forests of southwestern Bengal. As the author, Mukundaram, lived in this area, he is likely to have witnessed the beginnings of that historical process.33 Their patronage of monumental architecture and affiliation with Gaudiya Vaishnavism would have enhanced the image of the local Hindu dynasties in their own eyes as well as in those of their peers and rivals. The Mallas controlled the dense forests of southwestern Bengal from their center at Vishnupur. British District Gazetteers characterized the inhabitants of these jungles, located between the tribal groups of the Chotanagpur Plateau to the west and the mainstream Hindu society of the Gangetic plain further east, as “semi-Hinduized aboriginals.”34 Contemporary literary works also locate these Bagdis, Doms, Hadis, Bauris, and Chandals, which formed the bulk of this population, at the margins of Hindu society, scraping an existence as palanquin bearers, broom and

Introduction

•

17

basket weavers, armed guards, and soldiers. They are described as low caste or even untouchable, uncivilized, consumers of flesh and wine, probably because their customs and social organization offended orthodox Brahmanical sensibility.35 The Vishnupur rajas themselves likely arose from these communities. The Kusmetia Bagdis, for example, claim that the Mallas were originally Bagdis.36 The Mallas, though, seem to have felt the need to dissociate themselves from such lowly origins, as the town’s lore maintains that they belonged to a royal house of north India. These narratives take pains to mystify specific historical roots, relying instead on familiar tropes of Hindu kings’ divinely ordained status. One version traces the family to a north Indian prince, who at a rest stop in the forests surrounding Vishnupur, on the way to the Jagannatha Temple at Puri, abandoned his heavily pregnant wife.37 This woman gave birth to a boy, Adi Malla, to whom the dynasty attributes its origins. He was rescued and raised by a Bagdi family, who were out collecting kindling. A local Brahman recognized his innate royal status when the boy was offered shade by a cobra’s hood and when he raised golden bricks in his fishing net. The Brahman took him to the funeral ceremony of the local king. Here the king’s elephant spotted the boy, lifted him in his trunk, and seated him upon the now empty throne. This act was recognized as a sign and the boy was crowned ruler of the region. The employment of tropes like the cobra’s and elephant’s detection of royalty implies divinely sanctioned authority, while the tenuous ties to an unidentifiable north Indian princely lineage grants the Mallas the requisite caste background to be respected as rajas of Vishnupur.38 Their enthusiasm for architectural patronage thus suggests an attempt to elevate their social standing, probably to set themselves apart from their subjects and their own past.39 These local rulers also recognized the potential of temples as mechanisms for fostering loyalties and binding communities through patronage of ritual activity. The Mallas had inaugurated political rituals such as Indparab or Indradhaj, the flag-hoisting ceremony that commemorated their coronation and renewed the trustworthiness of their tribal allies.40 Likewise, the spring hunting festival of Ekhan brought the king and his vassals together, and the distribution of yellow turbans reaffirmed their ties. Through carefully calculated diplomatic and symbolic acts, the Mallas thus worked to win the affections of the smaller tribal leaders who mediated their authority. It is hardly surprising that these local kings would appreciate the powerful populist appeal of Vaishnavism and embrace the celebration of Ratha and Rasa with equal enthusiasm. Having entered into an alliance with Srinivas, they constructed temples to pro-

18

•

Temple to Love

mote not only the pleasure of Radha and Krishna, but also their own upward mobility and control over their territory. Malla authority was also renewed in the provisions they made for the daily upkeep of the temples. The amounts of food offered to the gods each day indicate that large numbers of people were fed in these royal temples, and the redistribution of prasad (blessed food) provided the primary meals for many. The Madan Mohan Temple’s daily offerings of cooked rice, greens, lentils, and rice pudding continue to be distributed among visiting pilgrims, local devotees, and homeless people. As Dimock has noted, the sharing of food as a communal and participatory activity is an important dimension of the bhakti experience.41 In providing the food distributed both daily and during celebratory feasts, temples offered their royal patrons an opportunity to express their authority. Malla influence was also channeled through their appointment of subjects to specific roles in the lives of the temples. The families of Brahmans, potters, garland makers, conch-shell workers, silk weavers, drum makers, and musicians were assigned to each temple to supply their wares and services to these institutions. Many of these families carry on doing so even today, creating continuities over many generations despite shifts in the structure of patronage in independent India. The rulers also donated rent-free land for the upkeep of temples and as payment for services.42 The various caste groups appointed to perform particular tasks typically settled around each temple, giving rise to neighborhoods. To this day, these families flock to the temples for the evening musical sessions. Hand in hand with the emergence of these residential neighborhoods, markets developed near temple walls to channel the economic activity the temples generated. The Madan Mohan Temple, with its twenty-two neighborhoods and market, formed Baispara. Conch-shell workers, for example, established workshops and homes behind this temple and the chiseling of shell resounds there still, giving the neighborhood the name of Shankharibazar. Likewise, silk threads are prepared in the temple courtyards not only for clothes for the gods, but also for human wearing. These settlements extended the frontiers of the town and of the kings’ control from the fortified palace-center (darbar). The temple of Lalji, with its market and eight residential neighborhoods, forms the community of Krishnaganj, which competes with the Madhabganj community, formed by the nearby Madan Gopal Temple’s eleven neighborhoods and market, to organize the town’s most spectacular Ratha celebration (figure 0.16). The institutions sponsored by the Mallas thus continue to provide space for Gaudiya Vaishnavism to be practiced as a popular, broadbased religious movement. Locality is thus mapped spatially as neighborhoods, organized

Introduction

•

19

around royal patronage of temples. As Arjun Appadurai has observed, because such mapping is inherently tenuous, it is bounded and renewed ritually, even in opposition to individual elements, as in the case of the Ratha celebrations.43 Architecture thus helped shape the community and the social organization of the town as much as it responded to those needs. Like Mukundaram’s mythical principality of Gujarat, Vishnupur’s temples point to larger processes of forest clearing, urban planning, trade and economic growth, and organized ritual. In this case, the process is initiated by the collaboration of the Vaishnavas with Raja Vir Hambir and might be called Vaishnavization, as a more specific form of the general umbrella term “Sanskritization.” It is also useful in contrasting the Gaudiya populist emphasis with the Brahmanical domination associated with Sanskritization.

Temples in an Indo-Islamic Environment Unlike the Nagara form of temples, which emerged during the formation of regional Hindu principalities in north India, the configuration of the Ratna form took place in an Indo-Islamic world, in which Hindu rulers were no longer the highest authority of the land. The temples respond to the political ascendancy of Islam in Bengal, as does Gaudiya Vaishnava bhakti to the powerful presence of Sufi mysticism and Sultanate patronage. Architecture and terra cotta ornament express the political negotiations undertaken by Hindu local rulers to thrive in these new circumstances. This book suggests that the Ratna temple must also be seen as a product of the wealthy, highly cosmopolitan, diplomatic order created during the reign of the independent sultans. These dynasties of primarily Turkish, Persian, Afghan, and even Ethiopian migrants had moved eastward in search of political and economic opportunities, and controlled the region from the thirteenth century until the Mughal conquest. Particularly during the fifteenth and sixteenth centuries, the foreignness of Sultanate rule was localized in a variety of ways, and offered a cultural environment that fostered creativity in both mosque and temple architecture. The temples are therefore a vital record of the simultaneous processes of Sanskritization and Islamicization in the rise of small Hindu kingdoms as Sultanate political authority itself came to an end. The architectural experimentation undertaken over the seventeenth century was made possible by a century of prosperous trade in agricultural commodities, silk, and cotton. Bengal was the envy of contemporary sultans, Hindu kingdoms such as Vijayanagar, and foreign travelers in the sixteenth century. João de Barros (1496–1570) of the Portuguese mission

20

•

Temple to Love

suggests that their contemporaries compared themselves to the wealthy Bengali sultans, implying the respect that they commanded among their peers: Sultan Badur [Bahadur, 1526–37], being himself one of the richest Kings in that Orient, and very arrogant, used to say that he was one, and the King of Narsinga [Vijayanagar] two and the King of Bengal was three, meaning to say that the King of Bengal had alone, as much as he, and the King of Bisnaga [Vijayanagar] had together.44

This traveler’s account, and the opinion of contemporary courts he reflects, indicates that a very powerful state had been established in the first half of the sixteenth century. After the collapse of the Bengal Sultanate, Hindu local kings participated as successful middlemen in this maritime trade to Europe and the Middle East. The extraordinarily long-standing success of that trade relationship is reflected in, for example, the cotton fabrics sent from the Malla processing center at Sonamukhi to the former Taliban government in Afghanistan until its collapse in 2002. Ships with animal-headed prows evoke that international maritime trading activity on the temples’ walls (figure 0.17). Armed guards aiming long-barreled guns alternate with the seated crew members who row the boat. Their short upper garments, puffy pants, and pointed shoes mark them as European traders, like the Portuguese envoys who document these activities. Their presence on Ratna temples reminds us of the cosmopolitanism of the ports and markets where these boats docked. They also refer to the thriving mercantile system that bolstered the economy and provided the surplus that could be lavished on architectural experimentation. Temple construction thus expresses Malla connections to these trading networks and to the Indo-Islamic political culture of Bengal and most of north India. Ratna temples are closely related to the architectural style established by the Bengal Sultanate through repeated construction of mosques, shrines, and tombs. I interpret the architectural appropriation as an attempt to establish continuity with the recently deposed Sultanate. In building temples that look like modified mosques these rajas took on the sultans’ role as patrons of cultural forms alongside their mercantile and administrative roles. They attempted to create what Velcheru Narayana Rao, David Shulman, and Sanjay Subrahmanyam have called “vertical linkages” in discussing the formation of the Nayaka state in south India: The entire process of creating a Na¯yaka state seems to depend upon establishing a linkage, articulated in terms of personal loyalty, with a higher centre of authority—here embodied in the Vijayanagara overlord. No Na¯yaka king can do without this empowering source from

Introduction

•

21

above. On the other hand, this vertical linkage is optimally activated under conditions that effectively undermine its controlling power.45

The Malla kingdom made a place for itself alongside a host of other local kingdoms, both Hindu and Muslim, as successors of the Bengal Sultanate at a time when Mughal authority was tenuous. Although the Mughal armies deployed by the Emperor Akbar had destroyed the power base of Bengal’s last independent sultans in 1575, local Hindu and Muslim leaders organized rebellions independently and also in collaboration. Under these circumstances, the creation of a Sultanate style of temple can be interpreted as an invocation of a vertical linkage to the Bengal Sultanate, whose impotence may have made the association particularly attractive to the younger local dynasties. The Mallas thus seem to have been pragmatic and creative in maximizing the opportunities opening up in a changing world through their cultural activities. The form of the Ratna temple suggests that their patronage of monuments conferred upon them the authority of the Sultanate, which no longer posed the imminent threat that the Mughals did. The terra cotta images of elite life on the temple walls suggest a court now imagined on an Islamic rather than the ancient Hindu model and a desire to participate in a global Islamic order. Architectural patronage can therefore be interpreted as a political maneuver for control. Earlier scholars often conceptualized Vishnupur as a rise of Hindu power against the overwhelming might of Islam.46 In demonstrating how the temples point to the political pragmatism of Hindu authorities and their negotiations with their Muslim counterparts, this study challenges these histories that have overemphasized hostilities and polarized Hindu and Islamic communities by projecting current tensions onto the premodern period. Instead, it reconstructs architectural strategies for achieving cultural continuities while preserving plurality.

Organization of the Book Each chapter of the book probes an aspect of the temples’ architectural originality that is celebrated in their inscriptions, and examines the ways in which the innovations articulate the political and religious complexities of this period in Bengal’s history. Chapter 1 brings together previously ignored evidence—textual descriptions, epigraphical information, terra cotta depictions, contemporary rituals, and oral reminiscences—to argue that the Ratna temple was exceptional as a two-storied form, segregating complementary spaces for daily rituals in the sanctum below and a festival pavilion above. The temples’ newly incorporated

22

•

Temple to Love

pleasure pavilion shapes the ritual re-enactment of the passionate love modeled by the divine couple Radha and Krishna. Through song, dance, and worship, as well as the terra cotta imagery, the monument is transformed into the site of the gods’ amorous encounters. The vertical axis created between the upper pavilion, where priests take the deity images, and the courtyard, where throngs of devotees sing and dance, brings the community into that divine love play. Chapter 2 addresses the temples’ bold appropriation of the form of shrines, tombs, and mosques patronized by the earlier Bengal sultans. The mosque’s organization offers a model for a congregational space for the collective expression of divine love. The box-like mosque also provides a base for the temple’s new upper pavilion. The politico-religious environment that enabled, even stimulated, such intense architectural interaction across religious communities has been overlooked in most of the scholarship on these monuments. The temples assert the political agenda of these Hindu patrons as heirs to the authority of the sultans and their desire to participate in a global Islamic order. Chapter 3 traces the role of the region’s older temple typology in the newly organized temples and the new Indo-Islamic political order. Although the older Nagara temples were rejected when a new paradigm was sought, the development of the upper pavilion is partly rooted in that earlier form. Sporadic experiments demonstrate the transformation from a single-storied to a two-storied temple. The upper pavilion often takes the form of a miniature older temple as well. I argue that the new architectural pastiche achieved in the seventeenth century makes a statement about cultural accommodation in placing a miniature temple upon a mosque-base, an ideology that is corroborated in the other arts of the period. However, such inclusiveness implies hierarchy, and a pre-existing framework in which the incorporation takes place. Indeed, the size of the temple-based pavilion atop the Sultanate base is suggestive of the political reorientation of the structures’ patrons toward the Indo-Islamic world. Chapter 4 turns to the spatial organization of the compound within which the double-storied temple is placed. The arrangement of structures around the central shrine departs dramatically from that of older temple complexes. A second axis of worship is incorporated into the traditional east-west ground plan of Nagara temples. The new north-south axis is reserved for community performance, which complements the conventional axis of priest-conducted services. The reorientation of the temple compound to integrate the second axis and community participation is another example of the negotiations that characterize the plurality of the period.

Introduction

•

23

The epilogue situates the architectural formation in the new sacred center that emerged at Vishnupur. The new pilgrimage center is called Gupta Vrindavan, that is, a hidden or invisible Vrindavan. Vrindavan, the mythical site of Krishna’s amorous encounters, is the cosmological center of the Vaishnava religious community. Architecture and terra cotta ornament, along with landscape architecture, oral lore, and devotional poetry, shaped the new sacred center and manifested the play of the gods to the Bengali devotional community in a new Indo-Islamic global order. While focusing on seventeenth-century monuments from one region of South Asia, the book thus offers a study of the broader cultural issues central to determining the roles of art and religion in the self-definition of communities.

24

•

Temple to Love

Figure 0.1. Map of South Asia Showing Major Sites Discussed

Introduction

•

25

Figure 0.2. Map of Bengali Cultural Region

26

•

Temple to Love

Figure 0.3. South Façade, Shyam Ray Temple, Vishnupur

Introduction

•

27

Figure 0.4. Temple No. 4, Barakar

FACING PAGE: (TOP) Figure 0.5. Radha Ballabh Temple, Krishnanagar

(BOTTOM) Figure 0.6. South Façade, Keshta Ray Temple, Vishnupur

28

•

Temple to Love

Introduction

•

29

Figure 0.7. Kala Chand Temple, Vishnupur

30

•

Temple to Love

Figure 0.8. Murali Mohan Temple, Vishnupur

Introduction

•

31

Figure 0.9. Radha Vinod Temple, Vishnupur

FACING PAGE: (TOP) Figure 0.10. South Façade, Madan Mohan Temple, Vishnupur

(BOTTOM) Figure 0.11. Radha Shyam Temple Compound, Vishnupur

32

•

Temple to Love

Introduction

•

33

Figure 0.12. Temple-Types of Bengal

FACING PAGE: (TOP) Figure 0.13. East Façade, Keshta Ray Temple, Vishnupur

34

•

Temple to Love

(BOTTOM) Figure 0.14. Domestic Hut, Vishnupur

Introduction

•

35

Figure 0.15. Temples No. 1 and 2, Barakar

36

•

Temple to Love

Figure 0.16. Celebration of Ratha at Madan Gopal Temple, Vishnupur

Introduction

•

37

Figure 0.17. European Ships, West Façade, Keshta Ray Temple, Vishnupur Figure 0.18. Dedicatory Inscription, South Façade, Shyam Ray Temple, Vishnupur

38

•

Temple to Love

1. Desire, Devotion, and the Double-Storied Temple

The radical new design embraced for Krishna’s temples provided the architectural environment for stimulating the intensely emotional relationship with the divine that Chaitanya upheld. This chapter explores the reorganization of temple space by the addition of a pavilion and a courtyard to make the divine more accessible to the community. The new temples provided an exclusive conjugal space for Krishna and his beloved in the upper chamber. My research suggests that during festivals the attending priests took the images of Radha and Krishna from the sanctum where they resided to the upper shrine, where they could enjoy each other’s company through the performance of special rituals. The temple thus functioned as a house with multiple rooms so that the gods could move about and engage in various activities. Concurrently, the openness of this pavilion, with its multiple doors and its careful alignment with the courtyard below, would have made the ritual enactment of their passion available as a model for gathered devotees gazing up from the courtyard and allowed them to participate through song in that divine play. This spatial reorganization thus accommodated the style of collective devotional activity embraced by Gaudiya Vaishnavism and gave the community a more active role in the life of the temple. It therefore suggests a renegotiation of the balance of authority between Brahman priests and lay devotees. The new architecture and festivities broke away from earlier practices in the region. In providing an upper pavilion and a courtyard for gatherings, these newer temples diverged from Bengal’s earlier Nagara structures. The dark, constricted sanctum of the older temple-type had made the deities less visible to large numbers of people standing at the doorway, and the absence of a large courtyard had denied the devotional community a gathering area. The innovative Ratna form thereby enabled the shift Desire, Devotion, and the Double-Storied Temple

•

39

from priest-controlled worship in Nagara temples to a community experience of the gods at play. I analyze this transformation in how a temple functioned for the Vaishnava community of seventeenth-century Vishnupur through the changes embodied in the architecture, the new form of deity images that could move from one room to another, the nature of the festivals, the songs performed at these occasions, and depictions on the temples’ walls.

The Second Shrine The upper pavilion provides a key to understanding how Ratna temple space is organized and how the structure facilitated the new rituals celebrating the love of Krishna and Radha. Although the upper shrines of Vishnupur’s temples are no longer in use, traces of their function are preserved in the architecture and in vignettes of ritual activity on the terra cotta panels that draw attention to the double-storied form. Since ritual use of the upper shrine is scarcely discussed in religious texts from the period, I turned to ethnographic research to fill out my understanding of the processes that shaped and were shaped by the new temple formation. I draw upon the oral accounts I collected from residents of Vishnupur and other towns that were once part of the Malla domain, as some of them had witnessed the use of the second level or inherited stories of it.1 Significant continuities also emerged when this evidence was correlated with my own observation of current use of the temples’ many spaces. I use the living tradition to probe the archaeological record because several temple communities claim a history of undisrupted use into the present. Their deity images are served by succeeding generations of the original families appointed by the Malla rajas. These connections suggest the likelihood of some continuities, despite inevitable adaptations and shifts in practice and meaning, that can illuminate the spatial reorganization undertaken in the seventeenth century. Ratna upper pavilions are typically octagonal or square, surmounted by a domical spire (figure 1.1). The enclosed space is limited, about the same size as the sanctum below, suggesting that only a few people could have fit into the pavilion at a time. Iron rungs and hooks still attached to the domical ceilings indicate that a swing-like throne was probably suspended for the images of Krishna and Radha. Some temples still have ropes hanging from these hooks. A second set of rungs at the corners likely served to secure a canopy over their heads as they do in the sanctum below. These arrangements suggest that even though they are today the homes of bats and snakes, the upper chambers provided an alternate venue for ritual organized around the deity images.

40

•

Temple to Love

The Shyam Ray, Vishnupur’s earliest double-storied temple, has an octagonal upper shrine with a covered pathway wrapping around it. The inner surfaces of the pilasters supporting the arches of the outer pathway and the walls of the inner chamber are lined with bands of terra cotta figures of dancers and drummers that rise up the length of the walls (figure 1.2). In its ritual setting, the ornamented interior with the deity images swinging at the center would have re-created a three-dimensional rasamandala, the circular formation of Krishna’s erotic dance with the gopis. On the banks of the Yamuna, under a full autumnal moon, Krishna had replicated himself to embrace each of his female companions and satisfy her desire to be with him alone. In the Gaudiya Vaishnava tradition the rasalila is revered as the essence of Krishna’s numerous lilas. Devotees recreate Krishna’s dance through meditation and aspire to the utter abandon of the gopis in their passionate relationship with the deity. As an architectural expression of Krishna’s play, the upper pavilion with its rows of dancing women around the central space for the deity images would have served to concretize the object of the aspirant’s contemplation. It would have functioned as an architectural analog to the rasamandala depicted in two-dimensional form on the temple’s south façade and discussed further in chapter 4 (figure 4.28). The temple built earlier at Gokulnagar in 1638 offered a slightly different arrangement for worship (figure 1.3). An elevated altar platform at the center of the octagonal upper shrine would have supported a metal or wooden throne for the divine images (figure 1.4). This platform is identical in profile to the one in the sanctum below and bears the same moldings and decorative elements, making the upper altar conceptually parallel to the lower one. As in the sanctum, a drainage spout makes its way from under the wall of this shrine to the exterior for unobtrusive disposal of ritual fluids. Indented niches in the walls would have served as shelves for oil lamps and other ritual implements in the same way that they do in the lower-level sanctum today. The niches and drains point to the role of the upper level as an equivalent venue for ritual activity. In feel, however, the two spaces are dramatically different. Four doorways punctured in the cardinal walls of the upper pavilion allow sunlight to flood in and create an airy space, in contrast to the dark sanctum that is bounded by walls on three sides. The openness of the upper shrine also makes it visible from below so that Krishna, seated high above, can observe the throngs gathering in the courtyard and, reciprocally, devotees below can participate in the activities of the gods above. This early temple thus reveals the basic elements of what would become Vishnupur’s convention for the second shrine above the sanctum. Only the immovable altar platform is replaced by swings in the later

Desire, Devotion, and the Double-Storied Temple

•

41

temples, likely because the swing allowed movement back and forth, increasing visibility from below, and also enhancing the pleasure of the gods. I also read this shift to swings as heightening the contrast between structures and practices above and below. The undulation of the swing would complement the stasis associated with worship in the sanctum, as the light and airiness of the pavilion would differentiate it from the darkness of the sanctum, conceived as a womb-chamber (garbhagr·ha) from which divinity emerges. The upper shrines are approached through a staircase that is usually located in the southwestern corner of the building. The stairs are accessible from the passage leading into the sanctum from the south porch (figure 1.5), suggesting that only priests who had access to this passageway would also have access to the stairs and upper level. The layout therefore also indicates that members of the devotional community, who were not allowed access to the inner areas beyond the porch, would not be able to go upstairs. Rather, they would enjoy the proceedings from the courtyard outside, creating a vertical axis of devotional activity. Perforations in the upper part of the porch wall adjacent to the staircase serve as a window, providing light for these dark stairwells. The window is often treated as a large terra cotta panel, not unlike the others adorning these inner porch walls. The grid of the window is thus embellished with floral forms. Some have indented niches for placing an oil lamp for additional light. Aside from being dark, the stairs are typically steep and narrow, clearly not intended for large numbers of people to traverse. They lead to the northwestern corner of the upper story, that is, its back wall, from darkness into the light. This fundamental organization is fairly consistent in Ratna temples from the earliest Malla-sponsored temple at Gokulnagar to mid-eighteenth-century monuments such as the Radha Shyam Temple (figure 0.11).2 Conformity to the established pattern for over a century indicates that it was a successful response to the buildings’ functional needs, one that persisted despite variations in the ornamentation of these structures. Thus the reorientation of the temple to incorporate an additional shrine provided an appropriate space for specific Gaudiya Vaishnava rituals. Only after a century of proliferation was such practice gradually abandoned.3 Corresponding changes in building conventions demonstrate that most temples built from the second half of the eighteenth century onward no longer incorporate stairs to the upper level. Consequently, the upper space could not be used for ritual activity. The Ratna thus was transformed from a functional building-type to an icon. Later temples, such as the Bose family’s Sridhar Mandir in Vishnupur, preserve memory of the multi-leveled structures of the earlier period. This modified Ratna form

42

•

Temple to Love

was extended to temples dedicated to other deities, such as the Kalanjay Shiva Temple at Patrasayer to the north of Vishnupur (figure 1.6). This adapted form is even used for the Dakshineshwar Kali Temple, built in the nineteenth century outside Calcutta, where the popular saint Ramakrishna presided. Divested of its distinctive upper pavilion, and hence of its function as an alternative venue for the gods’ pleasure, the temple form therefore became available for other religious communities with different priorities, and with deity images that are not as conveniently portable as the Vaishnava ones discussed below. The co-option of the Ratna form also signaled that Gaudiya Vaishnavism itself had become mainstream enough by this time to provide a resource for other emergent traditions.

Portable Gods The rise of distinctive Gaudiya practices is consonant with the emergence of the additional structure, and the new theological emphasis on love and intimacy and the new worship style may well have stimulated the need for an additional shrine. The construction of an alternate shrine that Radha and Krishna can periodically inhabit is based on the premise that ritual practice is organized around their movement from one place to another. Mobility enables the gods to be present at multiple locations both to perform miracles and to receive the services that constitute worship. In this way, the movement of Krishna’s forms from temple altars makes the deity more accessible to his following. The gods’ journeys are a recurrent theme in the biographies of deity images, and mobility also features in modern festivals. Portability thus is a key to the restructuring of temples in the seventeenth century. The deity did not permanently inhabit the miniature temple on the terrace above the sanctum of double-storied temples. Instead, Krishna lived in the altar in the sanctum on the lower level, and the upper structure served as a pavilion for his occasional pleasure. Evening arati, the final ritual of the day, was conducted on a grand scale in the upper shrine during royally sponsored festivals celebrating the lives of Krishna, Radha, and the major Gaudiya Vaishnava saints and leaders. At Gokulnagar, for example, celebrations involving ritual worship of Krishna in the upper pavilion were Rathajatra (the chariot procession), Doljatra (the spring celebration of colors), Jhulan (the swing festival), and Rasalila (the celebration of the famous circular dance in which Krishna multiplied himself to satisfy the passion of each of his female companions).4 On these festive occasions, temple priests removed the deity images from the throne-like metal or wooden shrine where they were permanently installed in the sanctum below and took them up the curving

Desire, Devotion, and the Double-Storied Temple

•

43

stairway to the upper pavilion. Today it is the anthropomorphic forms of Krishna and Radha—either carved from monolithic dark gray phyllite or cast in astadhatu, an alloy of eight sacred metals—that are typically moved, while the shalagram, an ammonite fossil that represents Vishnu in a more abstract, non-anthropomorphic form, is not transported for such occasions. Such worship practices differ from the south Indian convention of using a separate set of festival images (utsavamurti) that are reserved solely for such occasions. These temporary images are kept in a storage shed in the temple compound and brought out only at festival times. Frequent transportation of the enthroned deity in Bengali temples is possible only because the typical images worshiped in this region are smaller than those in many other parts of South Asia (figures 1.7, 1.8). The permanent deity images are between fourteen and twenty inches high on average, small enough to allow for frequent and convenient transportation. Metal images are also less fragile than lithic ones. These images are never affixed permanently to the ground as are stone forms such as the lingam, Shiva’s phallic pillar. Consequently they can simply be lifted from the altar in the sanctum downstairs and carried by the priest up the stairs, or outside to the edge of the plinth where they are today taken to enjoy the festivities taking place in the courtyard. That ease of portability underlies the events that constitute the biographies of Vishnupur’s deities. Mobility is at the heart of the popular narrative of Vishnupur’s acquisition of the image of Madan Mohan, probably in the early part of the seventeenth century, by Raja Vir Hambir.5 Madan Mohan is said to have come to Vir Hambir in a dream and requested the raja to serve him in a suitable manner, as the impoverished Brahman who kept him was unable to do so. The king then traced the image to the house of the Brahman in Birbhum district, draped Madan Mohan in his shawl, and smuggled him back to Vishnupur. While this narrative speaks to the history of the region in a number of ways, here I want to emphasize the fundamental assumption that the divine image is portable. It is such portability that allows for the ambulatory nature of images such as Radha Raman, the family deity of Srinivas Acharya. They have no permanent home, but rather move from shrine to shrine, from the hospitality of one branch of the devotee family to that of another. Portability makes relocating ritual activity to multiple shrines possible and, conversely, the new architecture and ritual practices require deity images that can travel without too much difficulty. Even the heavier stone deity images of the Gaudiya tradition began to travel concurrently, although unlike those of Vishnupur’s images, these long and arduous journeys were motivated by political uncertainty. Govin-

44

•

Temple to Love

dadeva, the image discovered by the Gaudiya theologian Rupa Goswami in 1534, has a very well documented travel itinerary. He departed from Vrindavan in 1669 and, after six temporary residences throughout the Delhi-Ajmer region, he was permanently installed in Jaipur in western India as the lineage deity of the Kacchwaha Rajputs, the most powerful Hindu dynasty in north India.6 Similarly, contemporary literature describes the migration of other Vaishnava deities, such as Sri Nathji, the central focus of the Pushtimargi community’s worship, to Nathdwara amidst negotiations in pursuit of power.7 Closer to Vishnupur in the east, Lord Jagannatha of Puri had moved to the southern Orissan hinterlands, seeking refuge during the struggles between the local Hindu rajas, the Afghan governors, and the Mughal emperors.8 While portability of deities is not extraordinary, it is given specific ritual significance in Bengal. Many daily rituals undertaken today still require that the image be able to move easily. When Madan Mohan was established in the merchant Gokul Mitra’s household in Calcutta in the eighteenth century, the god was given his own quarters.9 By this time, the temple had become more overtly like a house, with different rooms for different activities. In the marble front reception area, he sits on a silk-lined silver shrine, where he receives darshan from beyond the cusped, pointed triple arches on Shah Jahani baluster columns that separate visitors from the sanctum. However, he also has an attached dining room, and a bedroom with bathing and dressing areas where priests tend to his daily needs (figure 1.9). The sanctum is thus reserved exclusively as a reception area, its Mughal features comparable to those of the emperors’ halls for public audience in the Delhi Red Fort, for instance. His priest moves him about this suite every day, and to additional festival pavilions within the temple complex for the celebrations of rasalila and annakot, the autumnal offering of mounds of rice. Through their miraculous and mundane daily activities, gods such as Madan Mohan illustrate a tradition of movement initiated in the seventeenth century that is still in force. That ease with which he is moved today would have allowed him to take a seat in the upper shrine of his two-storied temple at Vishnupur for special celebrations. As a vital prerequisite for use of multiple shrines, the peripatetic nature of Gaudiya Vaishnava deity images and their regular and more intermittent journeys significantly inform temple reorganization in the seventeenth century.

A Festival Pavilion At Vishnupur today, the upper shrines of the two-storied temples are no longer in use. In fact, most Ratna structures in the town are no longer living temples. They are protected monuments, presently under the super-

Desire, Devotion, and the Double-Storied Temple

•

45

vision of the Vishnupur subdivision of the Archaeological Survey of India. As part of the Survey’s caretaking function, it has closed off general access to the upper level. Moreover, the Vishnupur Raj, disempowered and impoverished first by the British government and then by Indian independence, cannot afford to maintain its innumerable temples. The stairs of many temples, as for example at Jadabnagar and Dwadasbari, have simply fallen away. The deities from most of these temples are now installed in the family’s home shrine and worshipped collectively by the family’s priest in order to control costs. The Madan Mohan and Madan Gopal Temples, which still retain their deities, are supported by their neighborhoods. Here too the upper shrine is not in use, in part because the stairs and the upper story are in a dangerous state of disrepair. Instead, speakers are attached here to broadcast announcements. Elaborate arati and kirtan, the rituals that were performed upstairs during special occasions, are now conducted below. At Gokulnagar, the memory of ritual use of the upper story is alive, as the practice was abandoned only after the deity images were stolen and replaced with framed calendar paintings, about thirty years ago according to the temple caretaker. The Kantaji Temple at Kantanagar (1704–22), further east in modern Bangladesh, is one of the few Krishna temples where the upper shrine is used, but only under particular contingencies such as when the monument is undergoing restoration work below. Such intermittent use of the upper story probably persists only in a few rare double-storied temples across the plains of Bengal. Although ritual worship in the upper story appears to be a dying practice, there are indications of its flourishing in earlier times. The fact that the two-storied temples were constructed repeatedly over a significant length of time and were predominantly patronized by the Gaudiya Vaishnava community suggests that the Ratna design successfully incorporated the ritual requirements of the Vaishnava community. The sanctum on the lower level accommodated rituals of daily service that animated the image and transformed the lifeless metal or stone into embodiments of Krishna and his beloved Radha, while the shrine on the upper level was reserved for occasional use, for Vaishnava festivals. At Vishnupur today the deity images are still transformed for festive occasions (figure 1.8). They are dressed particularly lavishly, usually in silk- and gold-embroidered brocades, often ordered from Varanasi by pious and generous devotees.10 Yellow is preferred for draping the darkhued Krishna,11 and red for the golden-complexioned Radha.12 Blue is also popular for Radha’s garments, but it is the only color forbidden for Krishna’s garments, possibly because he is believed to have been blueskinned like the peacock’s throat.13 Their jewelry is brought out from

46

•

Temple to Love

bank storage vaults. Expensive gemstones and other pieces of jewelry are also given by devotees on such occasions to express gratitude for vows taken and prayers that have been fulfilled, or to mark particular events in their personal lives.14 Special garlands of seasonal flowers are also offered.15 Priests adorn the deity images in all this finery in preparation for the festivities. Together with the change of place, the fancy attire transforms the images of Krishna and Radha from what they are in their more ordinary clothing and daily lives. For celebrations like Jhulan, the swing festival, priests continue to take the bedecked divine images to the shrine on the second level at Kantanagar when the lower level is undergoing conservation work. The images of Kantaji (Krishna) and Radhika are placed on a metal swing suspended by coir ropes wreathed with flowers. The upper shrine is the space of entertainment, and the images are rocked gently so that Radha and Krishna may enjoy the evening breeze at the end of the hot and humid monsoon day. Devotees partake of the deities’ pleasure from the courtyard below and offer flowers or holy water. The evening ritual of arati, the final service before the night sets in that concludes the daily cycle of worship, is also performed in the upper shrine on this day. Ratha, the chariot festival, is the one occasion on which current residents of Vishnupur remember the upper pavilion being used. On this day, at the designated time, the town’s primary deity, Madan Mohan, and Radhika were taken up to the second story, from where they presided over the entire town. This act signaled the commencement of the ceremony. The town’s other temples took their cue from him and their gods rode out from their temples in flower-bedecked chariots to the outskirts of the town and back (figure 0.16). Today, however, Madan Mohan is only brought out from the sanctum interior to the edge of the plinth in front of the covered porch (figure 1.8). The new location accomplishes the same purpose of reaching out to his devotees, gathered in the courtyard, to enjoy their attention. Deities such as Lalji and Madan Gopal, who ride out of their temples each year, occupy the upper pavilions of chariots that are modeled very closely on temples, reflecting that older practice of using the temples’ upper shrine. In his study of Raja Rajballabh (1714–63), David Curley has noted similar use of the temples at Rajballabh’s estate of Rajnagar in present Dhaka district. A biography from the early twentieth century paints a picture of festival celebration in such multi-tiered temples: On the full moon of Basanti the Disk [cakra] and/of Laksmi-Narayan used to be seated on a golden throne and colored with kunkum, and then kept swinging in the fourth story [i.e., in the open portico of the central dol-manca]. At this time Goddess Prakrti used to wear the abundant

Desire, Devotion, and the Double-Storied Temple

•

47

green leaves and multi-colored flower ornaments of Spring, and She would stand before the eyes of the world in the guise of Mohini [the enchantress]. Sweet nights poured forth pure moonlight arranged like a dreamworld of rare grace. Dressed in the beautiful garments of springtime, the inhabitants of Rajnagar—boys and old men and wives alike— used to play with Holi powder, [sprinkling it] with a whirling motion. The whole edifice of this temple and the space on all four sides used to be reddened from the continual sprinkling of Holi powder and the whole mansion used to tremble and echo with the ecstatic dancing and singing of “Horis.”16

This account suggests that these deities were still taken to a special pavilion at the top of the temple on festival nights at the turn of the twentieth century. Swinging high above, they observed the performance of kirtan, accompanied by dance, as had Madan Mohan in Vishnupur. By the eighteenth century, these temples were often more than twotiered, the stacking of pavilions probably competing for height and grandeur in the same way that Gothic cathedrals in Europe had vied to outdo each other. The Kantaji Temple has nine pavilions, the Rajnagar dolmancha (festival pavilion for the celebration of Dol/Holi) has seventeen, and the Krishnachandra Temple at Ambika-Kalna has twenty-five. Despite variation in the number of pavilions and stories, the custom of elevating the deity to the highest level for festivals remained consistent. In their day such temples were the highest structures of their towns and thus afforded the deity, the symbolic ruler of the land, the best vantage point for viewing the domain as well as the community crowding below. As Rajballabh’s biographer goes on to observe, If one stood on the fourth story, groves of large trees would look like very small sprouts, and Kirttinasa River would seem to be a small piece of beautiful blue cloth. The height of this temple was not less than 125 cubits (187 feet). By well arranged stairways one could ascend from each story to the next higher one, and descend to the next lower one.17

As at Rajnagar, the Vishnupur temples offer a view of the entire town from the upper level. The upper shrine of the Keshav Lalji Temple at Patpur, for instance, offers a panoramic view of the entire Darbar (Vishnupur Raj estate) area, including the palace towers, the double stone gates, the distinctive profiles of the Shyam Ray, Keshta Ray, Radha Shyam, and Lalji Temples, the mud walls of the town, and the lotus-filled waters of the Shyam and Krishna bandhs. The pleasures of the pavilion thus included not only the pleasant breezes on a hot and humid afternoon, but also a spectacular view for the gods, and of course the attending priests. The biographer’s sense of wonder and delight at the view, however, contrasts with the awe and distance created by the elevation of

48

•

Temple to Love

Gothic cathedrals. Rather, Krishna desired to participate in the lives of his devotees, to partake of their devotion as well as their everyday experiences of enjoying the view and the cool breeze on a hot summer evening. The upper pavilion brings him closer to his devotees.

Festivals and Celebration of Divine Passion In the Gaudiya tradition the festivals of Holi, Ratha, and Jhulan are first and foremost celebrations of Radha and Krishna’s passionate union. Not only is the brightness of the springtime blossoms matched by the smearing of colored powders at Holi, but the trees’ fertility is also celebrated with revelry. Through the carousing, devotees also partake in the divine lovers’ union. Emphasizing its carnivalesque elements such as the inversion of conventional societal norms, McKim Marriott observed, “The idiom of Holı¯ thus differed from that of ordinary life both in giving explicit dramatization to specific sexual relationships that otherwise would not be expressed at all and in reversing the differences of power conventionally prevailing between husbands and wives.” His experiences in the Vrindavan area led him to interpret Holi as a “feast of love” in his well-known essay of the same title.18 At the core of the Ratha celebration is another amorous encounter of the gods. During the chariot festival, Jagannatha, who is identified with Krishna, rides out from his residence in the temple at Puri to the Gundica Temple, about two miles away, where he resides for a week. Significantly, Jagannath goes to Gundica without his wife, Lakshmi. Here he is served exclusively by the adoring gopis, enacted by the devadasis who serve at the temple. As Frédérique Apffel Marglin has noted, The Vaishnavites consider Gundica¯ temple, or rather that temple when the deities reside in it, to be Brundaba¯n [Vrindavan], the place where Krishna performed his divine play. Having crossed the river of time, the deities reside in another place which is not involved in normal time since it remains empty and devoid of daily rituals the rest of the year.19

The sight of this divine play had caused the saint Chaitanya to express overflowing emotion through rapturous dance, and even fits of fainting.20 The intensity of his response to the gods’ love play provided a model for devotees to engage in the local Ratha celebrations in Vishnupur’s temple courtyards. The origins of this celebration, as described by local lore, also suggest similar emotional response to the deity. The Gundica Temple takes its name from a legendary queen who, unable to curb her curiosity, peeked into the craft workshop where the divine image was being made. Enamored of his form, she then prevailed upon her husband, Indradyumna, to

Desire, Devotion, and the Double-Storied Temple

•

49

initiate the building of the temple and the ceremonial procession.21 Jagannath’s return from Gundica is initiated through an equally impassioned encounter, this time with his irate wife, Lakshmi. When she finds him, Lakshmi sprinkles a magic dust that turns Jagannath’s thoughts to conjugal bliss. His body lurches forth, pushed by priests, expressing through this ritualized activity his longing to reunite with her.22 Like the empty Gundica temple, activated only at the time of the festival, the upper shrine of the Madan Mohan Temple at Vishnupur, empty during the rest of the year, was transformed into Vrindavan for the celebration. During Ratha it became the site of Krishna’s play with Radha, his favorite gopi. Swinging at Jhulan is another way in which the Gaudiya Vaishnava community rejoices in their love. In the month of Shravan, when Radha returns to her parents’ home, Krishna disguises himself as a girl so that he can join her there. The festival is a passionate celebration of their reunion. As John Stratton Hawley has observed, “The swinging season suggests wild freedom and delight, a swing ride in the sky while luscious clouds gather to relieve the drought and heat of May and June, but it also suggests rhythm and balance as the swing passes back and forth over the ground.”23 Swinging the deities back and forth in the upper pavilion is thus a metaphorical expression of their erotic engagement as well.24 The deeper layers of religious reality invoked through these celebrations of eroticism are understood only by practitioners, whose belief, knowledge, and imagination are not available to an uninitiated viewer such as myself. The Gaudiya community has a long history of protecting such knowledge from those outside the religious community, because it may be misunderstood. Indeed, at Kantanagar the priest of the Kantaji Temple, whom I interviewed, alluded to the “invisible” and “secret” aspect of worship practices, but of course did not describe them. As its significance did not dawn upon me then, I did not pursue the subject of his cryptic remarks; in retrospect, I suspect that it would have been inappropriate for the elderly man to discuss it further with me, a younger Bengali woman. Secrecy and esotericism in the transmission of powerful knowledge is found as early as the sixteenth century in this religious tradition. The Prem Vilas, for example, records that Jiva Goswami, one of the tradition’s six primary theologians, advocated careful selectivity when instructing his three young disciples how to disseminate the theological precepts that he and the other goswamis of Vrindavan had distilled.25 In keeping with his injunction, texts from the sixteenth and seventeenth centuries do not provide elaborate interpretations of temple practice. Moreover, ritual use of the upper shrine of Ratna temples expresses the powerful concept of the simultaneously prakat (visible) and aprakat (invisible). According to this notion, only skilled devotees have the depth

50

•

Temple to Love

of vision required to see what is invisible to uninitiated people. It is expressed, for example, in the fundamental belief that the saint Chaitanya was the deity Krishna in a human body. The recognition of his incarnation led to the aggregation of a religious community around him, and distinguished this collectivity from other religious traditions. Consequently, the priest’s discretion is not surprising. Sukhamay Bandyopadhyay came to a similar conclusion about the use of the upper shrine from his research in neighboring Birbhum district; he decided that these chambers are used as “the sleeping rooms of the deities.”26 Although he does not elaborate on the rituals surrounding “sleep,” he is most likely referring to similar celebrations of the passionate love of Radha and Krishna, because these pavilions are certainly not quarters for sleeping. Ratna upper shrines do not permanently house bedroom furniture or other accoutrements. On the contrary, the lowerlevel sanctum is provided with a bed, a mosquito net, pillows, and sheets, and the gods are put into bed every night and awakened in the morning. The sheer lack of any permanent ritual implements points instead to the transient nature of the rituals performed in the festival pavilions. Traditions that shroud their powerful truths in silence to protect them from misunderstanding and manipulation by outsiders inevitably create both methodological and ethical constraints for scholarship.27 Other visual and verbal sources, however, help us retrieve traces of practices that may have been transformed over time and are now fragmentary. Kirtan, the Vaishnava songs of praise, for example, revel in the delight experienced by the gods, and when sung by participating devotees in the temple courtyard, they enact the night trysts between Radha and Krishna: With the last of my garments shame dropped from me, fluttered to earth and lay discarded at my feet. My lover’s body became the only covering I needed. With bent head he gazed at the lamp like a bee who desires the honey of a closed lotus. The Mind-stealing One, like the chataka bird, is wanton, he misses no chance to gratify his thirst: I was to him a pool of raindrops . . .28

To view such unions of Radha with Krishna through their re-enactment during festival celebrations is one of the highest aspirations of devotees. Krishna is the beloved object sought by his human lover, the worshipper. The ritual expression of Radha and Krishna’s love in the upper chamber articulates the Gaudiya value of madhurya. Verses of

Desire, Devotion, and the Double-Storied Temple

•

51

kirtan evoke and uphold the qualities of their relationship as a model for the singers: Beloved, what more shall I say to you? In life and in death, in birth after birth you are the lord of my life. A noose of love binds my heart to your feet. My mind fixed on you alone, I have offered everything.29

Jiva Goswami taught his pupil Srinivas to prize this selfless and allconsuming love, and it was to inspire this love that Srinivas Acharya brought Gaudiya Vaishnavism to Bengal.30 That ardor is made tangible in the style of temple rituals he initiated. The rajas of Vishnupur, in response, provided funds for ritually rekindling the divine passion through festival performance. They erected monuments with a pavilion for the celebration of divine love. That theology of love and longing propagated by Srinivas is also heralded in the terra cotta ornamentation of Vishnupur’s temples.

Depictions of Divine Desire The terra cotta plaques adorning the exterior walls of these temples reveal that their upper stories were indeed built to host Radha and Krishna’s love play. A sequence of panels on the south façade of the Shyam Ray temple represents the distinctive profile of a two-storied temple with one larger shrine creating the base for a smaller second one above (figure 1.10). The two are almost identical in form, with a pair of segmented columns supporting a roof with a pronounced curve. I argue that these terra cotta tiles also articulate the separation of functions for the two levels of double-storied temples and they thereby give visual expression to the goals of Gaudiya worship. In one such relief panel, deity images are depicted installed in both shrines of a double-storied temple. However, it is in the representation of these images that the nature and function of the two spaces are distinguished. In the sanctum on the ground floor, the images of Radha and Krishna are shown standing upright, their feet planted firmly on the ground. Only their hands and heads are turned toward each other, much as in the typical three-dimensional images used for worship. This relatively more conventional and formal depiction of the deities seems to represent Krishna as Vishnu’s incarnation. In some of the panels he wears the vanamala, a long garland associated with the iconography of Vishnu. He is the omnipresent lord who demands obeisance from the devotee (aishwarya). One such worshipper is shown standing with folded hands at the entrance to the shrine. In dramatic contrast, in the upper pavilion

52

•

Temple to Love

on the second floor, Radha and Krishna are seated, in profile. Radha sits on Krishna’s lap with her leg folded back. Their bodies are intricately intertwined, fitting snugly within the frame of the shrine. Their faces are turned toward each other and they embrace, their hands reaching out to each other. They are immersed in each other rather than interacting with the viewer. If the couple in the lower shrine represents the more awesome aspect of god, the one in the upper shrine depicts the intimate and erotic dimensions of the divine. In her discussion of the iconography, mythology, and rituals of divine sexual union in Orissan popular paintings and icons in the Jagannatha Temple at Puri, Marglin has interpreted the conjugal union of Lakshmi and Vishnu as a different typology from the extramarital relationship of Radha and Krishna.31 The first is hierarchical, with the male dominating through size and centrality, while the second is nonhierarchical, with the figures given more or less equal size and depicted interlocking, much like those in the upper shrine of the terra cotta panel. Indeed the visual traditions of the two regions may well have interacted sufficiently to share stylistic and iconographic features; Vishnupur is very close to the modern border of West Bengal with Orissa. Marglin reads the former as a procreative union, emphasizing Lakshmi’s role as the married woman whose husband is alive, who is sexually active, bears children, and is the primary feeder of the family. Hence her sexuality is associated with auspiciousness, bountifulness, and fertility. Kama, the procreative union of Vishnu and Lakshmi, is associated with birth and death, and with the structures that maintain cosmic order. Krishna, on the other hand, is Acyuta, “the unfallen,” who enjoys sexual relations with the gopis without begetting progeny. His pleasure is unfettered by worldly cares.32 Radha and Krishna’s union is reciprocal and non-hierarchical. This distinction between Vishnu and Lakshmi’s relationship and that of Radha and Krishna seems to be expressed in visual form in the two terra cotta shrines on Vishnupur’s temples. The formal and frontal representation on the lower level suggests the conventional relationship of marriage and worldly attachment. The intertwined and intimate depiction on the upper level likely expresses the paradigm of passionate love and sexuality unburdened by worldly concerns. Devotees in Vishnupur’s temples aspire to attain this complete union and equivalence through singing and dancing like the gopis in Vrindavan. This reciprocity in Radha and Krishna’s love affair also finds expression in the Gitagovinda, a twelfth-century composition by the poet Jayadeva.33 This is the first literary work to suggest that Radha was perceived and worshipped as Krishna’s divine consort.34 Marglin reads the first part of the poem as describing a sexual union in which the male

Desire, Devotion, and the Double-Storied Temple

•

53

is the active partner. In the middle section she finds the symmetry of both partners expressed in their pangs of separation. In the last portion Radha is the active partner, performing “inverse sexual union,” in which she is the dominant partner, physically on top of Krishna. At Puri these verses are performed in the day’s final ritual of putting Lord Jagannatha to bed.35 At Vishnupur they are sung in the courtyard of the town’s temples following evening arati. This verbal exploration of the nuances of their relationship thus finds a visual counterpart in the terra cotta imagery on the walls. The terra cotta representation of the divine couple in the upper shrine also differs from the ways in which Krishna and Radha are typically depicted in three-dimensional images used for worship (figure 1.8). Such intimate interaction is never represented in cult images. In contrast with their embrace in the upper shrine of the relief panel, metal images for worship are always frontal. They engage with the viewer, only attending to each other with tilted heads and expressive gestures. Devotional activity connects the two visually distinct forms of the gods. When the small, portable metal images were brought from the sanctum to the upper pavilion for the festival celebrations of divine love, devotees standing below in the courtyard would have these more formal images before their eyes. Their active and intense imagination would re-create the union of Radha and Krishna during their secret, nightly trysts. The terra cotta depiction of that union can be interpreted as a concretization of mental images activated by ritual celebrations. In animating the metal icons, the terra cotta imagery expresses the Gaudiya notion of the simultaneously visible and invisible. The depiction of the shrine also typically includes a devotee who leans into the pavilion with outstretched hands as if to enter this space. This figure is clearly trying to access this most intimate union of the gods, which is the ultimate goal of Gaudiya Vaishnava worship. He or she stands at the threshold with hands folded in adoration (anjalimudra). The hands are depicted in front of the pillars that separate the space of the devotee from that of the deity, suggesting an attempt to break the barriers, to attain union with the divine. The figure is thus both a visual expression of entry into the lila, the aspiration of Gaudiya practice, and a model for the viewer. The figure probably represents a gopi or a human devotee transformed through acts of ritual worship and meditation into a gopi. Alternatively, the figure may be a manjari, one of the young and beautiful female servant-companions of Radha who hold a special position in the lila because, unlike the other companions who must leave when Radha and Krishna commence their lovemaking, the manjari stay and serve every need of the divine couple. Emulating and ultimately identifying

54

•

Temple to Love

completely with these privileged characters in the divine union not only allows the practitioner to serve them but, as David Haberman notes, also enables him or her to have direct access to the gods, to witness the divine union first-hand.36 Services performed by these young maidens include serving betel nuts, fetching water, fanning them, massaging their bodies, adorning them by combing and braiding their hair, and entertaining them with song and dance. Sometimes Radha may even send a manjari to Krishna for his pleasure. Manjari sadhana involves the participant’s transforming himself or herself by taking this role in order to participate in the eternal lila.37 Padma Kaimal, in her discussion of portraits of devotees in ninth- and tenth-century temples in the Kaveri region of south India, has pointed to the necessity of uncovering the specific cultural constructions of the body in interpreting its representations. In particular, she addresses the bhakti conception of the devotee’s body as not inferior to that of the deity. They are bound by a shared beauty, which attracts them to each other and calls for their union in Gaudiya poetry and temple terra cottas.38 Here, as on the walls of the south Indian temples, no priests mediate between worshipper and deity. The location of these panels at eye level suggests an entry point for the viewer into the cosmic lila. The viewer would strike a posture similar to that of the depicted figure to approach the object of worship in the upper shrine. Such figures thus serve to communicate with the deity and with the audience, who share the aspiration for access to the divine lila. These terra cotta panels on the Shyam Ray Temple’s south façade are enframed in elaborate bands of vegetal ornament with inset depictions of shrines in the four corners. These indented shrines depict Radha and Krishna in different poses denoting their different activities: Radha sitting on Krishna’s lap; Radha pressing his feet; the divine couple standing with Krishna’s hand on Radha’s breast; and Krishna departing, among others. Such representations of the divine couple may be read as a sequential development of their relationship, or of a single trysting cycle, similar to the sequences of acts described in devotional songs (kirtan).39 The desire depicted in terra cotta is that of both Radha and the viewer who aspires to attain her state of selflessness. These panels are the visual counterparts of the verbal expressions of such experience in ritual practice and song and they may well have helped shape devotion. Their location, adjacent to the main public entrance to this royal temple, is significant. They proclaim the ritual significance of the new two-storied temple. They also suggest the theological emphasis on parakiya relationships and manjari sadhana for the newly converted Malla rajas and local community. In the predominantly oral culture of seventeenth-century

Desire, Devotion, and the Double-Storied Temple

•

55

Bengal, such images likely provided a way of not only revealing but also storing new knowledge, a new theology, and a mode of expressing devotion in visual form.40 The creation of the new temple-type followed one of the most significant historical moments in the consolidation of the Gaudiya community in Bengal. Srinivas Acharya brought the central doctrinal texts to Vishnupur, persuaded the Malla raja and local population to a passion for Krishna, and introduced festivals with expressive rituals such as the singing of kirtan. The chronological sequence of these two phenomena would indicate that a new ritual requirement had arisen, to which the addition of an upper shrine provided a response. The small, portable physical forms in which Krishna manifested himself could be moved easily to the swing in this second shrine. From here he could enjoy the ritual celebrations undertaken below. This space is reserved for special rituals of entertainment and pleasure, distinct from the daily acts of service performed by priests in the sanctum downstairs. It is the site for the expression of divine desire, which the devotee can glimpse from the courtyard below. Festivals thus re-enact the play of the divine lovers, and it is the Ratna form that makes that play physically accessible to the community. Such temporary and special rituals differentiate what is distinctively Gaudiya from the daily services conducted in the lower-level sanctum that are common to the various Hindu communities of the region. The new rituals and their architectural space thus also shaped the Vaishnava community of the region.

FACING PAGE: (TOP) Figure 1.1. Central Upper Pavilion, Shyam Ray Temple, Vishnupur

(BOTTOM) Figure 1.2. Terra Cotta Panel Depicting Kirtan, Central Upper Pavilion, Shyam Ray Temple, Vishnupur

56

•

Temple to Love

Desire, Devotion, and the Double-Storied Temple

•

57

Figure 1.3. Gokul Chand Temple, Gokulnagar

58

•

Temple to Love

Figure 1.4. Central Upper Pavilion, Gokul Chand Temple, Gokulnagar

Desire, Devotion, and the Double-Storied Temple

•

59

60

•

Temple to Love

Figure 1.6. Kalanjay Shiva Temple, Patrasayer

FACING PAGE:

Figure 1.5. Plan of Shyam Ray Temple, Vishnupur

Desire, Devotion, and the Double-Storied Temple

•

61

Figure 1.7. Ratha Procession, Madan Gopal Temple, Vishnupur

62

•

Temple to Love

(TOP) Figure 1.8. Priest Sushanta Mukhopadhyay Attending upon Madan Mohan and Radha during the Annual Celebration Commemorating the Arrival of the Saint, Srinivas, Who Initiated the Vaishnava Transformation of the Region, Madan Mohan Temple, Vishnupur (BOTTOM) Figure 1.9. Priest Subrata Pujari Carrying Madan Mohan from the Altar to His Dining Room, Madan Mohan Temple, Calcutta

Desire, Devotion, and the Double-Storied Temple

•

63

Figure 1.10. Panel Depicting Double-Storied Temple, South Façade, Shyam Ray Temple

64

•

Temple to Love

2. A Paradigm Shift

The intense creativity and resources invested in developing a new type of temple indicate that the pre-existing Nagara form, used for temple construction for over a thousand years, was no longer adequate in the seventeenth century. Major changes had taken place in Bengal during the late sixteenth century. As Gaudiya Vaishnavism spread swiftly, fueled by the dynamism and charisma of motivated young leaders like Srinivas, a new devotional community emerged. The Ratna form responded to the need for a suitable space for the expressive style of collective worship he demonstrated. When the Malla rajas championed Vaishnavism, they employed prolific temple construction in an effort to transform the jungles of Vishnupur into the earthly site of Krishna’s lila. However, these patrons were no longer omnipotent, as the Pala and Sena dynasties had been up to the twelfth century. Rather, they were landholders with restricted authority under the Mughal empire, making their place within an Islamic courtly culture that had evolved over several centuries. Architectural patronage gave them an opportunity to establish and express their role within this reconstituted political culture. The new requirements arising from these changed circumstances motivated architectural experimentation with alternative paradigms. In abandoning the Nagara convention, the architects, masons, and terra cotta artisans had multiple options available for putting the parts of a temple together. They chose to draw from vernacular huts, and from the mosques and mausoleums of the late Sultanate period that gave such a model monumental form. Although we know little about the architects and builders aside from their handiwork, we can assume that they selected these particular models at least in part because they possessed the requisite technology and skills, inherited from previous generations who

A Paradigm Shift

•

65

had worked on the antecedent structures. The rapid spread of Ratna temples during the seventeenth century and into the middle of the eighteenth century testifies that the available expertise and materials provided a viable solution to the political and ritual needs of the period. This chapter traces the creative process of adaptation that produced a congregational temple by comparing the forms shared by Ratna temples and the preceding brick mosques, shrines, and thatch-roofed hut compounds. The striking resemblance between the façades of the Ratna temples’ cubical sanctums and the last of the Sultanate shrines and mosques indicates that in embracing these forms to build a space for community worship, the artisans appropriated aesthetic and symbolic features as well. Yet the scholarship on these temples overlooks this aspect of the monuments’ originality. By imposing modern criteria that divide Hindus from Muslims and separate temples from mosques, the literature neglects to consider how Sultanate forms, even mosques, became suitable models for Hindu temples. A study of the innovative Ratna form suggests that alternative criteria may explain the architectural innovation more satisfactorily than does the role of religion as it has been understood in colonial and postcolonial scholarship. By examining how the buildings function we can begin to appreciate why Ratna temples would find a more useful paradigm in mosques than in Nagara temples. They share with Sultanate mosques the capacity to house large gatherings for ritual worship. The Nagara paradigm, on the other hand, designed primarily to hold only the Brahman priest who performed the daily services, did not suit the Vaishnava style of interactive, collective worship. Locating the architectural changes within the particular political context of their patronage equally indicates that the Hindu religious identity of the sponsors does not explain their motivations. Instead, Vishnupur’s fashioning of its self-image in relation to the deposed sultans and the Mughals who had just conquered Bengal suggests that political pragmatism sometimes played a bigger part in cultural constructions than we have heretofore conceded.

Problems of Classification The close formal relationship between Sultanate mosques and Ratna temples’ lower stories has been ignored in most histories of Bengali art. Early British studies of Bengali architecture defined the extant monuments as either Hindu or Islamic, treating them as disparate traditions, the former indigenous and the latter imposed. The new Ratna formation slipped between these sectarian categories that were constructed in the nineteenth century and perpetuated until very recently. By overemphasizing the Hindu identity of the monuments, these scholars neglected the

66

•

Temple to Love

architectural inventiveness of Ratna temples as modifications of Sultanate structures such as mosques. A political climate that could allow for interaction among the region’s different religious and artistic communities could not be imagined within the available colonial and postcolonial conceptual frameworks. In his 1891 History of Indian and Eastern Architecture, James Fergusson, for instance, introduced a brief discussion of these monuments, focusing specifically on the Kantaji Temple at Kantanagar, as follows: In addition to the great Indo-Aryan style of temple-building, . . . there are a number of small aberrant types which it might be expedient to describe in a more extensive work; but, except one, none of them seem of sufficient importance to require illustration in a work like the present. The exceptional style is that which grew up in Bengal proper on the relaxation of the Mahomedan severity of religious intolerance, and is practised generally in the province at the present day.1

Fergusson’s characterization of the style of these monuments as “aberrant” and “exceptional” epitomizes the inadequacy of colonial epistemological and administrative frameworks to situate these structures. Although his choice of words reveals an uncomfortable awareness that these structures did not fit “the great Indo-Aryan style of temple-building,” the form associated with a Hindu temple, neither Fergusson nor the scholars who followed him could acknowledge just how closely these temples did in fact look like Bengali mosques.2 What held them back? The problem lies in the prevailing characterization of Bengal’s cultural history, which gives primacy to race and religion, and usually conflates them, in defining group identities. With the production of colonial knowledge during the nineteenth century such identification of people and monuments had come to dominate. British administrators invoked these categories in part to justify and maintain colonial institutions. Their historians, ethnographers, site surveyors, and cartographers propagated them in the documentation and description they compiled toward consolidating British territorial acquisitions.3 For example, the role of the late-nineteenth-century censuses in this process of classification has received much scholarly attention, beginning with Benedict Anderson’s work in Southeast Asia and Bernard Cohn’s writing in the South Asian context.4 These colonial perceptions of religion did not acknowledge the fluidity of identity that scholars like Richard Eaton have begun to uncover.5 On the basis of Judeo-Christian assumptions about the nature of religion, colonial scholars endowed a body of texts, beginning with the Vedas, with the authority of scriptures analogous to the books of the Semitic traditions, and classified a vast range of agrarian communities under the newly coined term “Hinduism,” itself created by the addition of

A Paradigm Shift

•

67

the Greek suffix -ism.6 This monolithic religious tradition was positioned in opposition to Islam, whose particular South Asian dimensions were ignored.7 Such an approach to religion as a formalized, organized, and bounded community cannot imagine that religious identity itself may have been constructed in different terms or that it may have played different roles in social organization at particular historical junctures. Sources from the sixteenth and seventeenth centuries hint at alternative priorities. Joseph O’Connell notes that Vaishnava biographies do not seem to have identified communities exclusively or even primarily by their religious affiliation, and when they did so identify them, the terms used did not necessarily carry the same connotation as they do today. Instead this literature suggests that the labels “Yavana” and “Mleccha” were more popular than “Muslim,” emphasizing the foreign nature rather than religious orientation of the community. Both of the former are Sanskritic terms that had gained currency far earlier than the coming of Islam. “Yavana,” the Sanskrit pronunciation of “Ionic,” was subsequently applied to other incoming groups of foreigners, including, by the sixteenth century, Turks, Arabs, Persians, and Afghans. “Mleccha” indicates lack of purity, a status relative to Brahmanical society, to which these writers typically belonged. It is typically used pejoratively, and applied to tribal and low-caste groups such as the indigenous inhabitants of the Vishnupur region as well as the elite Muslims. Gaudiya Vaishnava texts also use more specific ethniccultural designators such as “Turk” and “Pathan,” honorific titles such as “Sayid” and “Khan,” and occupational ones such as qazi or wazir, but rarely “Muslim” or “Mussalman.” O’Connell concludes, Nor is there any suggestion that the Hindu-Yavana/Mleccha difference is primarily a religious one; it is bound up with political and economic interests. Religious difference is perceived as a by-product of such socioeconomic division.8

Correspondingly, “Hindu” and “Hindustan,” referring to the peoples and land beyond the Sindhu or Indus River, were introduced in the Achaemenid inscriptions of Darius, and in circulation after that in Greek and Aramaic sources.9 The Persians, Afghans, Arabs, and Turks who later crossed that river and entered the Gangetic plain seem to have employed these terms to distinguish the natives from themselves, suggesting that the designation was primarily used by outsiders.10 In Gaudiya texts the term “Hindu” only appears in rare situations of conflict with the Muslims, but, O’Connell notes, these are treated as differences of jati, a social rather than religious distinction. Such perceptions differ from the formal and bounded religious identities presumed by colonial administrators.

68

•

Temple to Love

Unlike the unified religion imagined in later colonial documents, the Vaishnavas, Shaivas, Shaktas, and numerous more marginal communities, in Bengal as in most of South Asia, did not share a single historical founder, unifying beliefs and rituals, or a common ecclesiastical authority.11 Their rivalry with the Shaktas seems to have preoccupied Vaishnava authors more than their sporadic clashes with the Muslim authorities who quashed their more exuberant processions. The Vaishnavas challenged the caste-based authority of the Brahmans, and criticized their blood sacrifices. In fact, it was to the Yavana rulers that the Vaishnavas turned for protection from the invective of these other older devotional communities that we now collectively call Hindu. O’Connell concludes, “the Vais.n.avas in Bengal did not place their religious commitment in the solidarity of the Hindu people, nor in the sacred ideals, if there is such, common to Hindus. Their religious faith was in Krishna, a mode of faith that in principle a non-Hindu could share.”12 It follows then that these sixteenth-century texts do not sustain the perception of a Hindu alliance against the Muslims’ intolerance. An episode noted by Jayananda in his sixteenth-century biographical work Chaitanya Mangal suggests that far from challenging Muslim hegemony, Chaitanya affirmed it. This exercise of unofficial diplomacy occurred when the Orissan ruler Prataparudra contemplated an invasion of Bengal and came to the saint seeking his blessings. However, Chaitanya warned, “I see no way for you to conquer Gaur·. The sovereign of Gaur· will come to Nila¯cala (Purı¯) certainly; You will flee and Utkala (Orissa) will be ruined.” Hearing the master refuse (his blessing) Prata¯pa Rudra went to fight in Vijaya¯nagara.13

Such incidents, their veracity aside, suggest that the Vaishnavas did not deny the reality of Islamic political ascendancy. Rather, they embraced the stability of the Sultanate, manipulating situations to their advantage to make a place for themselves on the political map of Bengal. It is noteworthy that Chaitanya did not support the campaign, although it was designed by a Hindu ruler, a powerful devotee who offered him the support that the Muslim king of Bengal did not. Such a venture would probably have inconvenienced the cause of Vaishnavism in creating obstructions for pilgrims on the path to Puri. In the end, Prataparudra’s ambitions were redirected toward another Hindu state, Vijayanagar, to the south of Orissa. Such episodes indicate the need to re-examine unqualified assumptions of Hindu cohesiveness or political organization against Islamic groups. If the Hindus do not seem to have displayed the fixed religious identity

A Paradigm Shift

•

69

that nineteenth-century scholarship assumes, Bengal’s Muslims pose an equal challenge to tidy and orderly categorization. Not all of Bengal’s Muslims were easily ethnically distinguishable from the Hindus. The very British census reports that produced these categories seem to also have had trouble untangling them. In the 1901 census E. A. Gait, for example, had to qualify his claims: . . . it may be said generally that almost the whole of the functional groups, such as Jolaha and Dhunia, throughout the province, the great majority, probably nine-tenths of the Sheikhs in Bengal proper . . . are of Indian origin. The foreign element must be looked for chiefly in the ranks of the Saids, Pathans, and Mughals. Even here, there are many who are descended from Hindus, and . . . high caste converts are often allowed to assume these titles, and in some cases, to intermarry with those who are really of foreign descent, their number however, possibly only a small proportion of the total, and may be neglected. If the above estimates be taken as a basis, it would appear that the strength of the foreign element amongst the Muhammadans of Bengal cannot, at the most, exceed four millions, or say, one-sixth of the total number of persons who profess the faith of Islam.14

Gait’s observation was probably even more true of the population of the sixteenth and seventeenth centuries, prior to the Wahabi efforts at purification of the Muslim community. Yet Fergusson, like his compatriots, gave precedence to religion for segregating local populations and identifying their architectural formations and understood their religious identities in mutually exclusive terms. He did not recognize that the two communities could interact on any terms other than hostile. By portraying what he calls the Mahomedan Period, probably the period of Sultanate rule from the thirteenth to sixteenth centuries, as one of religious segregation and discord followed by a gradual truce, achieved presumably under Mughal imperialism, Fergusson rendered himself unable to appreciate the appeal of local architectural forms for the construction of buildings for both Hindu and Islamic ritual practices. The Ratna temples’ sophisticated adaptations of these available elements therefore remained unrecognized. Fergusson’s inability to address these monuments results, in part, from constricting these seventeenth-century structures to a narrow definition of Hindu temples based on the earlier Nagara type, “the great IndoAryan style.” He is not alone in struggling to locate these monuments in the canon that was under construction. The assistant director of the Archaeological Survey of India, J. D. Beglar, declared the Vishnupur temples unworthy of scholarly attention for their lack of antiquity: “Bishanpur is famed as an old place, and certainly contains very many

70

•

Temple to Love

temples and other old remains, but their age is not such as to merit detailed notice of them.”15 Not surprisingly, he was preoccupied with translating their inscriptions to determine the dates of their construction. Such treatment of the structures is strikingly different from his high praise for the suitably ancient Nagara monuments in nearby Bahulara and Sonatopal. Vishnupur partly failed to meet the colonial expectation of a Hindu temple because these scholars associated temples almost exclusively with the Nagara form and with earlier periods of Hindu sovereignty. Those assumptions did not allow an exploration of the complex relationships between structures, patrons, and artisans that evolved during the seventeenth century.16 The fragmentary record of Bengali architecture from the period of Sultanate rule contributed to such perceptions. The extant record of Sultanate architecture is limited to a vast number of small local mosques and a few tombs and shrines. Little by way of urban civic structures survives other than the crumbling boundary walls with imposing gates at Gaur. Yet the description of an imposing palace and luxuriant court within the ramparts of the citadel in European visitors’ accounts attests to the presence of monumental Sultanate residences: We arrived at the second gate and we were searched as at the first, and we passed by this and by others, as many as nine and we were searched each one, and arriving the last gate, we saw a great courtyard the length of a great track, and half hollow, and so wide, or wider than [it was] long on which twelve men were playing choqua-polo on horseback, and at the end of the said courtyard a great dais was set upon thick props of sandalwood, and those above, upon which the roof rested, were not as thick, all carved with maçanarya-joinery and [with] many gilded branches and small birds, and the ceiling above in the same manner, and [with] a moon and a sun, with [a] very great number of stars, and all gilded. We arrived to where the King was seated on a very great cateredivan, likewise gilded, with [a] very large store of great and small pillows, all embroidered, and with many precious stones and seed-pearls on them.17

Despite such tantalizing glimpses, no palaces or any other, more modest, residences remain.18 The overwhelming presence of explicitly religiously oriented monuments has helped make the terms “Islamic” and “Sultanate” synonymous despite the religious plurality of the population. What little does survive at Gaur also indicates stylistic continuity across secular and sacred structures. The heavy buttresses of Gaur’s massive gateways are also used for its mosques. Terra cotta ornamental details such as the flowers encased in pointed arches extend in horizontal bands across the

A Paradigm Shift

•

71

bastions of both gateways and mosques. The evidence of shared architectural elements and vegetal motifs therefore suggests that the missing secular monuments would have been significant for defining a Sultanate style rather than an Islamic one. Nineteenth-century art historical methodology also skewed the early understanding of Ratna temples. The assurance that visual analysis alone was enough to understand these buildings led nineteenth-century scholars away from grasping the originality of their architectural form. Their looking, of course, was predetermined by the prevailing categories discussed above. Moreover, colonial architectural historians, who devoted themselves to compiling material for vast encyclopedic histories, typically did not read the local languages. The significance of the temples’ inscriptions, which herald their originality in phrases such as navaratna ratnam, therefore eluded them. Decontextualized formal analysis discouraged any examination of that originality of form within the particular political and cultural circumstances of their production. The Radha Binodji Mandir at Perua, Birbhum district, for example, was built by the local Muslim landholder, Alilaki Khan of Rajnagar.19 Likewise, later epigraphs offer names of architects, masons, and terra cotta artisans that are both Hindu and Muslim. A closer look at the inscriptions than simply a search for dates and patrons’ names, hand in hand with visual examination, would have alerted them to the complexities of artistic production during this period. Writing in the middle of the twentieth century, Percy Brown remained reluctant to discard the preconceived categories that he had inherited, despite the failure of the visual evidence to fit them: “in no part of India are the two great cultural movements, the Hindu and the Muhammedan and the manner in which the one superseded the other more vividly illustrated than in some of the ancient remains of Bengal.”20 He continued to group monuments by ruling dynasties, which he conflated with religious affiliations, treating early Bengal as Hindu, ignoring the rich and complex interaction between the patronage and production of Hindu and Buddhist temples. He then characterized the subsequent history of Bengal as Islamic; although this is true of the rulers, it cannot of course be said of the population, which was substantially Hindu as well as Muslim. Nor do the labels “Hindu” and “Muslim” satisfactorily distinguish the newly converted Muslim populace from the Hindu Bengalis, as discussed above. Moreover, the scattered temples built by smaller minority religious communities such as Jains, Naths, and Dharmapanthis have no place in such linear narratives that reconstruct a progression from Hindu to Muslim to British rule. His privileging of the royal perspective over those of the other social strata does not give him the flexibility to address the Ratna temple

72

•

Temple to Love

typology satisfactorily, as this form developed during a period of Muslim overlordship.21 It also averts any discussion of temples that looked to mosques as sources of their form. His brief discussion of these structures is located in a chapter titled “The Brahmanical Buildings of Bengal (8th to 17th centuries)” in his Buddhist and Hindu Period, volume 1 of his Indian Architecture, while the second volume, titled The Islamic Period and allegedly devoted to later material, contains the discussion of the mosques that preceded these temples and profoundly shaped them. Brown thus conflates religious practices, the religious orientation of monuments, building techniques, and art historical styles under the rubric of religion and ignores chronology. By giving primacy to religion, he remained unable to examine the complex relationships between structures, patrons, and artisans and the particular historical circumstances that informed the formal choices for monumental construction. Such watertight categories cannot incorporate the complex architectural interactions revealed by a survey of standing monuments. Continuities in artistic traditions despite changing political regimes and religious movements, and therefore divergent sources of patronage, have begun to be explored only in the past few decades. Recent scholars of Bengal Sultanate monuments have noted that they drew upon the sculpture of the Pala period for decorative details,22 and for their architectural framework on Pala-period Buddhist architecture, known primarily from representations in sculpture and painted manuscripts, and from Burmese Buddhist temples at sites such as Pagan.23 Continuities from Sultanate architecture to the Vaishnava temples built during Mughal rule have remained unexplored, with the notable exception of a preliminary but critical article by David McCutchion. After compiling the first nearly comprehensive survey of these temples, he began to examine common architectural and decorative features among Pala-Sena monuments, Sultanate buildings, and seventeenth-century temples, and to consider the impact of Mughal and later European architecture on Bengali temples. However, McCutchion did not probe further to challenge the pre-existing framework of labels that separated Hindu from Muslim. As a result, he identifies architectural and ornamental elements by religious association, according to their use in temples and mosques. He uses terms such as “Hindu pillars,” “Hindu niches,” “Hindu trabeate system,” “typical Muslim embellishments,” and “Islamic features such as the dome and pointed arch.”24 Consequently, while his work is groundbreaking in its acknowledgment of the complex interactions among different architectural modes, its conceptualization now needs to be reconsidered. Contrary to McCutchion’s characterization of architectural features

A Paradigm Shift

•

73

and decorative motifs as Hindu and Muslim, I interpret shared use of artistic forms as an indication of a substantial common vocabulary without restricted sectarian meanings. Artisans working for either patron community probably drew from the available resources to satisfy the needs of their sponsors in the same way that they do today. The makers of lace-like banana pith taziya for the Shiite Muslim celebration of Moharram in Malda, northern West Bengal, for example, also use the same nonfigural motifs on rathas for the local Hindu community. Similarly, painters of long narrative scrolls called patas adapt from their repertoire of images to suit the ends of their patrons. Gurupada Chitrakar, a painter from Noya, Midnapore district, for example, uses the same animal-type for the goddess Durga’s lion as he does for the saint Satya Pir’s tiger. The animal’s form, he explains, is neither Hindu nor Muslim. It illustrates various narratives and satisfies the expectations of his different audiences. Shared forms for mosque and temple decoration leave little reason to believe that such practices were disjunct in the past. An ethnographic approach helps us begin to uncover the nature of older artistic practices that do not seem to have been documented, presumably because they were found unexceptional at the time. The above examples also indicate that these modern artisans do not necessarily belong to the same religious group as their patrons. Probably the seventeenth-century terra cotta workers, architects, and masons similarly held a variety of religious beliefs. In Mukundaram’s Kavikankan Chandi, Muslims were included among the workers constructing the mythical city of Gujarat inspired by the goddess Chandi.25 Likewise, Muslim composers of Vaishnava devotional poetry of the sixteenth and seventeenth centuries signed their verses with their names. Clearly they did not need to adopt Vaishnava pseudonyms; nor did Vaishnava compilers find their contributions inappropriate.26 Further, ethnographic studies of current practitioners of Bengali art forms indicate that the religious practices and identities of these artist communities are more flexible than nineteenth-century scholarship was prepared to concede. For example, Lee Horne’s work among the Malhar, metal-workers who produce a distinctive type of lost-wax threadware in southern Birbhum district, reveals that this community celebrates daily rituals and festivals that we now classify as exclusively Hindu or Muslim.27 They, however, do not perceive the conflicts that scholars experience in attempting to classify these communities. If such current practices have persisted since the seventeenth century, we can begin to discern some of the complexities that a linear narrative creating segregated Hindu and Muslim artistic traditions cannot accommodate. The early censuses, however, labeled such communities Hindu with-

74

•

Temple to Love

out qualification, confining them to fit colonial administrative agendas. The epistemological frameworks developed hand in hand with such approaches to living communities thus constricted Ratna temples to a monolithic Hindu temple model derived from the Nagara paradigm. The early scholars of these monuments were consequently unable to appreciate the ingenuity of adapting the basic mosque form for the ritual purposes of Hindu devotion. A study of the discourses that gathered around Ratna temples thus uncovers the ideologies circulated through literature and institutional practices until such notions about history, culture, religion, and community were accepted unquestioningly. It reminds us that the production of discourses cannot be divorced from the history of power and that such sectarian ideology, when nurtured, is capable of mobilizing communities with the violence that we have witnessed recently in Ayodhya, Kashmir, and Gujarat.

Making a Congregational Temple The importance that Fergusson and the following generations of scholars placed on the Hindu and Muslim identity of the monuments does not allow us to understand how public buildings constructed by the Sultanate, including its mosques, provided a model for Ratna temples. Their common function as spaces for collective activity, however, begins to explain the significant correspondence in their patterns of distribution, scale, and spatial organization. To grasp how a congregational space was created for Vaishnava temples we have to turn to the monuments erected during the reigns of the restored Ilyas Shahi (1433–86) and Husain Shahi (1493–1538) sultans, from which Ratna temples adopted architectural elements. This late Sultanate style developed during the late fifteenth century and continued into the middle of the seventeenth century, contemporaneous with the Ratna temples.28 It is embodied in structures such as the Tantipara and Chota Sona Mosques (1493–1519) and the Qadam Rasul (1519) at Gaur, the Goaldi Masjid (1519) at Sonargaon, Dhaka district, the Jami Masjid at Bagha (1523) and the 1558 mosque at Kushumba in Rajshahi district, the Jami Masjid at Atiya (1609) and Sadi’s Mosque at Egaroshindur in Mymensingh district (1652), and the midseventeenth-century mosque at Shura in Dinajpur district (figures 2.1, 2.2, 2.3, 2.4, 2.5, 2.6, 2.7).29 As Perween Hasan has noted, their shallow domes, low façades, curved cornices, and brick and terra cotta decoration are better integrated with local landscape than the earliest Sultanate constructions from the fourteenth century, such as the minar pronouncing victorious conquest at Chota Pandua, or the enormous Adina Masjid asserting an alien authority (figure 2.8). Later Sultanate architecture

A Paradigm Shift

•

75

discarded the minar and the imposing interior courtyard, shedding with these forms their association with state power and the grandeur of Islam. Rather, they turned to the curved cornices of indigenous thatched huts. As monuments in this style spread through the region, they provided the Vaishnava community with a model for creating a space for public worship.30 Today most late Sultanate structures survive along what is now the Bangladesh-India border, with the majority situated on the Bangladesh side. The majority of Ratna temples, on the other hand, remain standing in West Bengal. While the greater distribution of Hindus in the western areas and Muslims in the east, as revealed in the early censuses, partially explains the extant distribution of temples and mosques, modern politics has also affected the fate of these monuments.31 Consequently we have to look further east, to the deltaic part of Bengal, for the Sultanate monuments that best explain the Ratna form. In West Bengal, few Sultanate buildings survive aside from those in the cities of Gaur and Hazrat Pandua and the early Tribeni Mosque, which are protected by the Archaeological Survey of India. None remain standing in modern Bankura district, the area where the Malla dynasty held sway. The scattering of less permanent tree-side shrines to Satya Pir and dargahs of local holy men tucked away on side streets does, however, suggest the presence of a sizeable Muslim population in Vishnupur.32 The seventeenth-century Motichura Masjid at Rajnagar, in neighboring Birbhum district, offers glimpses of a vibrant tradition of terra cotta decoration (figure 2.9). Its size suggests that a substantial Muslim population used the building for prayer in the late Sultanate period. A giant banyan tree grows through another seventeenth-century structure at Kulut, in neighboring Burdwan district, whose shattered porch and interior chamber may have belonged to a resthouse (figure 2.10). An inscription found in the village of Siwan records the construction of a Sufi khanaqah as early as the thirteenth century.33 The presence of a stable Muslim zamindari in Birbhum was probably responsible for the construction of these monuments. Ambika-Kalna, further south in Burdwan district, flaunts the ruins of five substantial late Sultanate structures.34 The city of Burdwan, to the east of Vishnupur, was famous for the dargah of Pir Bahram Saqqa, a typical late Sultanate single-domed building with a curved cornice and terra cotta ornamentation (1562). The ruins of a fort and the dargah of Shah Ismail Ghazi at Garh-Mandaran, south of Vishnupur, attest to early Sultanate architecture in southwestern Bengal. The sheer proximity of these sites and Vishnupur, where Ratna temples were raised in the seventeenth century, suggests that the patrons and architects of the latter would have been aware of the former. Hints of

76

•

Temple to Love

their contact with these outlying areas can be gleaned from various sources. Mughal writers identify the Mallas as rulers of Birbhum and Pachet, suggesting that their territory extended beyond Bankura.35 Beglar, during his tour through Bengal in 1872–73, noted that the ancient seat of the Mallas was Labhpur in south Birbhum district, and the Malleswar Parganas were named after them.36 The narrative of the acquisition of Madan Mohan discussed in the previous chapter also suggests that Raja Vir Hambir used to go hunting in these parts, for it was on one such expedition that he encountered the deity. The ease of travel between south Birbhum and Vishnupur also suggests that architects and masons possessing the requisite skills were available to build similar structures for the Mallas. Likewise, the Mallas had connections with the north Midnapur area, as their daughters were married to the rulers of this region.37 The fragmentary condition of these buildings, however, does not allow any substantial analysis of their form. Instead, we have to turn to the betterconserved late Sultanate mosques and shrines of Gaur and Rajshahi to trace the late Sultanate sources of Ratna temples’ size, distribution, and organization of space. Seventeenth-century temples are relatively modest in scale compared to the earlier monumental constructions of the Pala dynasty in Bihar and the Gangas in Orissa. Shifts in patronage partly explain why smaller temple complexes were constructed in large numbers. Vaishnava temples were built by local landholders such as the Mallas of Vishnupur, who had risen to prominence under Mughal rule, rather than by the highest authorities in the land, which the Pala dynasty had represented. The new scale of construction was probably attractive within this new patronage system for a number of reasons. Monuments on this scale were less expensive to finance while they enabled their patrons to map out territory and establish control over their populations. At a time of significant political and administrative renegotiations following the Mughal conquest, rising Hindu zamindars probably found temple construction a useful strategy for asserting their authority. Scattered through the forested terrain of Bankura, these institutions must have helped consolidate heterogeneous tribal groups under the flexible umbrella of Hindu culture, as represented by Gaudiya Vaishnavism, as well as that of Malla reign.38 The new congregational style of worship is equally critical for understanding the preference for smaller-sized structures in large numbers. At the village level, a local temple probably served rural communities better because it was immediately accessible, rather than a remote though monumental one at a faraway urban center. This concern becomes a priority now because the new practices of worship introduced by the Vaishnavas are collective, like those of the Muslim communities. Unlike

A Paradigm Shift

•

77

the puja conducted in Shaiva and Shakta temples, the performance of kirtan meant that groups of people gathered nightly in Vaishnava temples to sing and dance. Consequently, easy access was vital for the maintenance of the emergent Vaishnava community. Sultanate architecture shares this preoccupation with easy access in its inclination toward numerous smaller mosques rather than a single colossal structure. Hasan attributes this predilection to the geographical and climactic peculiarities of the delta with its annual floods, pointing to the rejection of the monumental hypostyle courtyard-mosque following the failure of the Adina Masjid of 1375, the biggest mosque of its time in South Asia (figure 2.8). Instead, many smaller structures, without interior courtyards, were built across the Bengali countryside from the fifteenth century. The mosque at Goaldi, built in Dhaka district in 1519, is one of innumerable examples of this mosque-type that came to dominate the later Sultanate period (figure 2.3).39 The internal organization of Ratna temple complexes also indicates that accommodating group activity was now a vital concern. Individual structures for the deity’s permanent residence, housing of the sacred basil plant, daily food preparation, entertainment, and festivities allow different groups to discharge their various responsibilities. The spaces are clustered around an open courtyard within a walled enclosure (figures 4.2, 4.3). Today, many of the additional structures and boundary walls have begun to disappear as smaller buildings fall to ruin, leaving the central temple standing rather baldly, sometimes in the midst of paddy fields. In the case of protected monuments, the temples have recently been wrapped in colorful gardens planted by the Archaeological Survey of India that erase all traces of those accessory structures. Fortunately, however, a few nearly complete enclosures survive to indicate the prevailing pattern. In the Vishnupur region, the Madan Mohan Temple in town and the Gokul Chand Temple at Gokulnagar have almost intact compounds that remain functional. The sanctum, where Radha and Krishna preside, is of course the ritual focus, and the monuments’ various activities are oriented toward it. The two doors of the sanctum open to the east and to the south, onto the courtyard. From their throne at the north end of the sanctum, the gods have a direct view of the courtyard and the natmandir (entertainment hall) on the southern end of the complex. Gazing through the south entrance, they can enjoy the music and dance performed there. The tulsimancha (basil planter), a requirement in Vaishnava temples, is located in the courtyard, as the basil plant is believed to be an ardent devotee of Vishnu who cannot be parted from him. The east entrance into the sanctum provides convenient access for the priest, particularly when carrying food offerings from the kitchen. (The ritual

78

•

Temple to Love

significance of this reorganization of the temple complex will be explored further in chapter 4.) Such spatial organization of the Ratna temple complex differs radically from the Nagara convention, with its axial alignment of subsidiary halls in front of the sanctum.40 Access to the sanctum was controlled and hierarchically organized, which suited the priest-mediated practices conducted therein. The Ratna division of spaces by activities around a courtyard and within a walled enclosure is instead characteristic of the way domestic compounds are laid out to this day (figures 2.11, 2.12). The average Bengali village homestead is composed of a group of huts around an interior courtyard, typically bounded by a low mud-brick wall. Each hut is a single-roomed rectangular or square structure with mud walls supported by a sheet of woven bamboo, palm leaves, reed matting, or, rarely, wood posts and beams. Thatch roofs are piled upon a bamboo frame overhead. The hut closest to the entrance, often no more than a covered porch, is usually where male visitors are received, while inner ones are for sleeping and cooking. Additional huts may function as animal pens, storehouses, and domestic shrines in more elaborate homes. Such organization allows large joint families to reside and perform multiple functions within a walled compound. This division of spaces for numerous daily activities within the domestic complex may have provided a more suitable model for the new courtyard temple enclosures that were now sites of collective devotional activity. An analogous compound with designated spaces for various activities also occurs in Sultanate mosques, with an ablution area, a burial ground, a shrine for a local holy man, and perhaps a hall for ceremonial meals aside from the main prayer chamber, all within the compound gates. Most of these additional structures have fallen away, and the Jami Masjids at Bagha and Shura are rare examples of Sultanate walled complexes with multiple structures to accommodate the varied needs of the community (figure 2.4).41 Shallow rectangular covered porches flank the temple sanctum (figure 2.13). Doorways on the south and east walls of the sanctum open to these porches, where worshippers gather. The variety in the numbers of porches and their functions indicates experiments toward making the gods accessible to larger numbers of people. Smaller neighborhood temples, such as the Radha Vinod in Vishnupur or the Sharabhuja42 in the suburb of Tejpal, have only one porch, in front of the public south entrance, giving refuge from sun and rain to gathered devotees. More elaborate ones have porches along the other walls as well. The Shyam Ray Temple sanctum is completely enclosed by covered porches, all accessible directly from the plinth through triple-arched entryways (figure 1.5).43 Regardless of variations in the number of porches, the fundamental square or rectangular

A Paradigm Shift

•

79

unit with adjacent porches corresponds to the plan of Sultanate mosques and huts, indicating that domestic architecture likely provided a source for such plans in both temples and mosques in Bengal. The porches provide an intermediary space before the prayer chamber and facilitate the movement of large numbers of people. Smaller mosques throughout the delta have a porch attached to the east side of the prayer chambers, while the Qadam Rasul at Gaur has porches on three sides.44 Such plans likely provided an antecedent for the reorganization of temples from exclusively priestly space to centers of community worship.45 In huts, these porches are open, consisting merely of a sloping thatched roof with curved eave and edge, held up by bamboo stalks. The Murali Mohan Temple at Vishnupur refers explicitly to such huts in its use of open porches on all sides, supported merely by piers in the same way that the huts have porch roofs elevated upon bamboos (figure 0.8). It also draws attention to the careful imitation of the sloping cross-section of thatch roofing. The closed porches of the Keshta Ray Temple reproduce the hut roof on the exterior (figures 0.13, 0.14). The lower story of this temple takes the jor bangla form, a juxtaposition of two rectangular huts aligned on their longer side (figure 2.13). Each hut has two long sloping eaves with curved spine and ends. On the inside, the curved eaves and spine of the porch vaults give them the appearance of hut interiors with do-chala or double-eaved thatched roofs. Arcuate roofing technology, made popular through monumental Sultanate construction, is thus deployed toward meticulously simulating the hut’s porch. The Shyam Ray Temple’s porch ceilings even reproduce the crossing bamboo or timber frame that supports the thatch as a decorative terra cotta veneer (figure 2.14). The adaptations based on vernacular huts and Sultanate mosques point to the importance of collective worship in the Vaishnava tradition. Builders and patrons experimented to find a different arrangement than the Nagara temple’s more constricted space. Domestic compounds and Sultanate structures, which had already turned to domestic models, now provided a model for the arrangement of structures for the temple’s various activities around a courtyard.

The Sultanate-Style Façade The architects of the Ratna compound also adopted the materials, construction techniques, and design elements that gave Sultanate mosques, shrines, and tombs their distinctive appearance. They adapted heavily, even playfully, from the late Sultanate style to create the form and aesthetic of a bold new Hindu temple. The architectural elements appropri-

80

•

Temple to Love

ated for the lower story included the characteristic curvature of the chala, the dome, the triple-arched entrance, and reinforced corners. Some of these choices, such as use of the curved cornice, seem to be more symbolic than, for example, the more utilitarian preference for brick and terra cotta. The thatch roof piled upon mud huts is reproduced painstakingly in brick for both the cornice of the Ratna lower story and near-contemporary mosques, such as those at Atiya (figure 2.6). This curve of the hut’s thatch roof is due to its bamboo frame. Because bamboo is exceptionally strong and flexible, it provides adequate support to the pile of thatch overhead. While there are many local variations in hut construction, the central bamboo or wood post that bears the chala roof is often made a little longer than the others. The resulting upward curve of the bamboo frame compensates for any sagging at the ends. While the slope of the thatch roof allows rainwater to drain off efficiently, the curved cornice at its end does not have any such utilitarian function. Instead, it derives from the adjustments made to accommodate the pliable nature of the bamboo. Hasan interprets the replication of this distinctive feature in brick mosques as an attempt to make a foreign structure familiar to a newly converted community.46 Such appropriation of local vocabularies, masking and domesticating an alien Muslim presence, is a very familiar approach from the earliest Islamic monuments developed in the Middle East. The curved cornice seems to have acquired value through its perpetuation in mosques and then temples, distinguishing Bengali architecture from that of other South Asian regions. Its earliest tentative use in brick occurs in the early-fifteenth-century Eklakhi Mausoleum at Hazrat Pandua (figure 2.15), a bold attempt at an indigenous form by the first native dynasty of sultans. Subsequent rulers developed this local form, rather than importing mosques from further west, as part of their efforts to popularize and legitimize their regime and create a cultural identity distinct from both Delhi and the neighboring Sultanate of Jaunpur, which posed periodic threats to their sovereignty.47 Sufi pirs who had traveled east also perceived a local idiom as a successful way of presenting Islam, a faith alien to the Bengali fishing and farming population.48 In fact the thatch huts themselves continue to provide congregational mosques for small local populations, as at Salban, Comilla district, and it is virtually impossible to recognize them as mosques when they are not marked as such by signage or occupied at prayer times (figure 2.16). Over time, that curve became more pronounced. In seventeenthcentury temples and mosques the curve bows down, the movement of the line contrasting with the vertical, upward movement of the upper shrine

A Paradigm Shift

•

81

and spire. A terra cotta lintel above the triple-arched doorway echoes the cornice and adds to the low horizontal profile of the building’s lower story, as does the stringcourse in Sultanate mosques (figure 0.3). From this time that stylish curve became integral to the façades of Ratna temples, with all other elements of the façade modified to accommodate this rhetorical flourish. It is hardly surprising that the term chala, employed for huts, is also extended to temples in the Bengali language.49 Like the curved cornice, the corners of the Ratna temple’s lower story are differentiated from the rest of the façade. Although not heavy or protruding in the Vishnupur temples as in late Sultanate mosques, they are invariably differentiated in their terra cotta ornamentation (figure 0.3). Engaged pilasters and vertical bands of terra cotta typically separate the ends from the wall frieze. These ends are decorated in narrow bands of floral pattern or horizontal figural scenes, quite distinct from the vertical format adopted for the panels along the length of the wall frieze. Sultanate mosques have heavier, turret-like projections with octagonal or rounded profiles (figures 2.4, 2.5). Such strengthening of the corners of temples and mosques, like the curved cornice, was primarily decorative. It does, however, suggest a close relationship with the urban architecture of gateways and city walls, such as the extant structures at Gaur, where buttressing would have served to protect the capital.50 Flanked by these reinforced corners, the characteristic Ratna entrance consists of three arches supported by stout pillars. This distinctive entrance had already developed during the Sultanate period, as at the Qadam Rasul (figure 2.2). Temples appropriate the cusped, pointed arches from late Sultanate structures.51 These entry arches are supported on squat, segmented, and multi-faceted pillars (figures 0.3, 2.2). The columns stand on wide square bases, and develop a rounded profile as they rise to the central section. The shaft is short, and segmented by projecting bands of terra cotta. The upper section again becomes wider and square at the top, where it takes the weight of the arches. Sixteenth-century Sultanate structures such as the Qadam Rasul at Gaur have similar short, complex pillars and provide an immediate source for Ratna temples.52 Earlier, the Adina Masjid’s royal platform displays stunted pillars with smooth round surfaces, from which the later, more complex pillar-type likely developed. These pillars, in turn, relate to the engaged pilasters represented in sculptures of deity figures created during the Pala-Sena period. Such a comparison suggests that the Sultanate pillars may have a source in Pala architectural forms that no longer survive.53 While certain features of the Sultanate façade are adopted with hardly any changes, the architects are more circumspect and inventive with others. They are discerning as well as resourceful in their use of domes, for

82

•

Temple to Love

example, perhaps acknowledging the long association of domes with mosque construction. While the temples exploit domical technology for roof construction, they carefully adapt the appearance of the dome for a Hindu temple. The small dome capping the square Ratna sanctum is never visible from the outside, as it is concealed by the elevated platform upon which the upper pavilion stands. In contrast, the hemispherical domes atop Sultanate mosques are often many and flamboyant. While an interior dome is used to cover the temple’s sanctum, the exterior domes are rendered as inconspicuous as possible. In the Madan Mohan Temple at Vishnupur and the Nandakishor Temple at Dvadasbari, the domical profile of the upper shrine is downplayed by the kalasha (kalas´a, waterpot) and amalaka (a¯malaka, myrobalan fruit), conventional Nagara finial elements (figures 0.10, 3.9). In the Kala Chand Temple it is disguised by the horizontal laminations of a Nagara tower and in the Radha Shyam by vertical striations that recall the Nagara spire’s bands of creepers (figure 0.7).54 Ratna temples also share with Sultanate monuments the use of wellfired and finely laid brick.55 As in the earlier mosques and tombs, bricks are laid in horizontal courses as stretchers, using a mortar consisting of powdered brick, fine sand, and lime from snails’ shells. Ruined structures such as the Mahaprabhu Temple at Vishnupur reveal that mud was used to lay the inner core. A plaster of sand and lime is used to protect brick surfaces of walls, roofs, and vaults from the fungus and mold that spread during the damp monsoons.56 Their treatment of the brick is also similar. The entrance arches, for example, result from the dexterous manipulation of smaller and longer bricks to create alternating projections and recessions that impart an impression of lobes or cusps. In later temples, however, cusping is achieved by cutting the brick.57 Brick is probably such a popular medium for construction in this deltaic region because clay is readily available and therefore probably inexpensive compared to the cost of quarrying and transporting stone.58 A preference for monumental construction in brick was already manifested in Sultanate construction, and earlier in the Pala-period monastic complexes at Paharpur, Mahasthangarh, and Lalmai-Mainamati in Comilla. This continuity suggests that generations of artisans in the region probably passed down a facility for handling this medium and its peculiar characteristics.

Patronage and Politics Why do Vaishnava temples flaunt a Sultanate-style façade? Why did the Hindu rajas rising to power at the end of the sixteenth century establish such conceptual links with the Sultanate after it was overthrown

A Paradigm Shift

•

83

by the Mughals? Although the Vishnupur rajas have left us no text that reveals the precise motivations for their architectural patronage, implicit in their act of looking to Sultanate architecture is the power that the Sultanate continued to hold in the imagination of Bengal’s elite even after its deposition by imperial Mughal armies. The cultural and political circumstances of the late sixteenth and early seventeenth centuries, when the architectural experiments toward the Ratna form were being undertaken, yield the conditions under which such acts of appropriation become meaningful. The patterns of patronage across media suggest that the interests of these rising little kingdoms were compatible with those of the preceding sultans despite their differences of religious orientation. If the sultans had calculated to expand the base of their authority through patronage acts, the Hindu rajas sought to participate in that transformed cultural context as part of constructing their pedigree. Moreover, the resources commanded by the sultans indicate a very wealthy and powerful state, one which smaller chiefs aspiring to recognition in a predominantly Islamic political arena might well wish to associate themselves with. The choices made by Vishnupur, alongside numerous other Hindu kingdoms rising to power or maintaining their positions after the Mughal conquest, thus suggest that pragmatism often outweighed other concerns in political decisions. Contrary to Fergusson’s insistence upon Mahomedan intolerance, most historians today seem to agree that the later Ilyas Shahi and especially Husain Shahi sultans, while retaining some of the Persian etiquette of their forebears, interacted with their subjects and enthusiastically embraced the Bengali community and culture.59 Through overt displays of generosity toward their heterogeneous population they broadened the base of their authority. They appointed Hindus to high-ranking government posts, such as chief minister and supervisor of the mint, as well as to trusted personal positions like bodyguard and private physician.60 In fact, Rupa and Sanatana Goswami, who later became theologians of Chaitanya’s devotional movement, participated in these administrations as departmental head and private secretary, respectively. This is a noticeable shift from the earlier Sultanate, which had remained more aloof and exclusive. The orthodox Muslim nobility and Sufi saints who supported the earlier Sultanate court had denounced Hindu ascension to power.61 Asim Roy’s use of the term “cultural mediators” to describe the poets who successfully negotiated between the native and the newly arrived northern Indian, Turkish, Arabic, Persian, and Afghan cultural traditions can perhaps be applied equally well to the later sultans.62 As flamboyant display of Islamic religiosity and political symbolism diminished, the smaller Hindu successor states emerging from the reshuffle of political

84

•

Temple to Love

authority in the late sixteenth century do not seem to have been alienated by the rulers or averse to emulating their cultural forms in creating their own principalities. Having been part of the landscape and community for over a hundred years, later Sultanate architecture provided a suitable model for temples in the seventeenth century. Their shallow domes, low façades, curved cornices, brick, and terra cotta made these structures accessible to the multi-ethnic and multi-religious population. Bengal’s Muslims were predominantly recently converted rural farmers and fishermen, presumably unfamiliar with rising towers and hypostyle courtyards. They retained indigenous ideas and practices alongside newer Islamic ones. Even the Muslims who had entered Bengal from further west had adapted to local conditions as they became estranged from their Turkic, Persian, Afghan, and even north Indian origins.63 Familiar forms would also have alleviated the foreignness of the rulers and their faith in the eyes of the substantial Shaiva, Shakta, and Vaishnava communities, as well as other minority communities such as Sahajiyas, Dharma and Nath Panthis, and Bauls, who survived on the fringes and drew from the dominant traditions. The initiative to encourage an architectural formation drawing upon local sources thus suggests a shrewd political decision on the part of the sultans in the service of consolidating their realm and maintaining authority. Recognizing the success of this move, local landholders rising to power in the seventeenth century patronized new temples that built upon this style of monument. Like their architectural style, their literary patronage indicates an effort on the part of the sultans to nurture the indigenous cultural forms that were coming to the forefront. They actively encouraged the use of Bengali rather than Persian, both as a literary language and as the language of communication at court. In fact, the history of literature written in the Bengali language begins at this time. This literature explores themes that were local rather than Persian or Arabic. The Husain Shahi rulers, particularly Ala al-din Husain Shah (1493–1519) and Nasir al-din Husain Shah (1519–31), sponsored significant works about the snake-goddess Manasa, who was particularly popular in this swampy region. Vijaya Gupta’s Manasa-Mangala and Vipradas’s Manasa-Vijaya offer lavish eulogies to their enlightened patrons. Gaudiya Vaishnava literary production also began at this time. Chandidas composed Srikrishna Kirtana, the earliest extant corpus of devotional poetry describing Krishna’s lila in the Bengali language, and Krittibas translated the Ramayana into Bengali. The earliest Vaishnava padas (short devotional poems) emerged under the direct patronage of these sultans. Yashoraj Khan, a Husain Shahi official, is one famous composer of such poems. A signed pada survives from

A Paradigm Shift

•

85

another poet, Sheikh Kabir, also intimately connected with Nusrat Shah.64 Significant along with the patronage of Vaishnava literature by an Islamic court are the Muslim names of the poets themselves, suggesting that local expressive forms were encouraged regardless of the religious affiliations of patrons, artists, and subject matter. Correspondingly, Islamic literary works sponsored under late Sultanate rule were also relocated to Bengal, in part through skillful analogies. Historical figures of Islamic origin were endowed with the features of Bengali or South Asian characters, and mythological events were transformed into local narratives. The Prophet Muhammad’s daughter Fatima was described as the mother of the world, following the mangalkavya convention of praising local goddesses; Ali was portrayed as a celebrated archer, no less skilled than Rama; the battles between the early Muslims and “infidels” were compared with the wars of the epics Ramayana and Mahabharata; and Yusuf and Zulaikha were transplanted to Vrindavan on the banks of the Yamuna, a site already hallowed by the dalliances of Krishna and Radha. Fundamental theological concepts from the multiple traditions thriving in Bengal were interwoven with equal facility through this body of literature. The Islamic notion of a living prophet as a messenger of god, for example, was equated with the pre-existing concept of an avatar (the incarnation of a deity). In his Nabi-bangsa, a history of creation and the prophets from Adam to Mohammad, the Chittagong poet Saiyid Sultan treated major Brahmanical deities such as Brahma, Vishnu, Rama, and Krishna as consecutive prophets like Jesus and Mohammad, each succeeding another with an appropriate scripture. This new chain of progression thus embedded the new religious system within the old and familiar one.65 Since the later sultans were committed to indigenous elements over imported ones, the environment they fostered became more inclusive of their diverse constituent communities, who in turn could participate in this culture. The Vishnupur rajas were thus able to relate to this deeply transformed culture and identify with its patron sultans. In choosing to do so, the Mallas were also following a longstanding Vaishnava trend of seeking Sultanate support for their activities. Their early poets had mostly been sponsored by Husain Shahi patronage, as their works indicate. Some, such as Maladhar Basu, author of Srikrishna Vijaya, had also participated in the Sultanate’s administrative structure as revenue collectors. Important leaders of the movement, such as Rupa and Sanatana Goswami, also had longstanding relations with the sultans, who facilitated their activities in various ways.66 Their allegiance to the Sultanate had already been established by the time the Mallas embraced Vaishnavism. Hence Vishnupur’s transformation into a Sultanate-style capital would hardly have been surprising from the Vaishnava point of view. 86

•

Temple to Love

The success of these later sultans is reflected in descriptions of the Husain Shahi years as a golden age.67 The historian Nizamuddin Ahmad seems to have been the first to apply this characterization, applauding Sultan Alauddin Husain Shah in 1594 for having “summoned learned, great and pious men from different parts of the kingdom, and showed kindness to them. He made very great efforts and exertions for enriching and improving the conditions of the country.”68 By the time the Mallas were rising to power in Vishnupur, these sultans were memorialized in the region’s literature for stimulating an upsurge of creativity, supporting pluralism, and maintaining political stability. Bengalis, Hindus and Muslims alike, could look back with pride at this recent past during a time of political uncertainty when threats from outside forces made their own future uncertain. Hindu rajas thus fashioned their self-image by reaching out to this past to appropriate particular aspects of the cultural efflorescence. Nizamuddin praised the Husain Shahis not only for their ability to recognize and cultivate their more erudite subjects, but also for their creation of a highly prosperous state, which provided the material conditions for such cultural production.69 That affluence attracted foreign trade missions, which further glutted the brimming coffers of the sultans. Portuguese embassies, for example, were dispatched from their base in south India to establish favorable terms for maritime trade to partake of this sumptuousness. These foreign visitors’ accounts equally attest to the flourishing condition of Bengal under the sultans. They repeatedly extol the variety, abundance, and cheapness of agricultural produce and processed goods such as cotton cloth, which their boats loaded and took away. They give us vivid details of the architectural opulence of the capital at Gaur with its royal palaces and hierarchy of elite residences, and of a cosmopolitan mercantile capital with bustling thoroughfares where a variety of foreign traders, including Arabs, Turks, Europeans, and Chinese as well as other South Asians, congregated to conduct business. An anonymous Portuguese manuscript preserved in Lisbon provides one example: I saw only that the land is very fruitful and has many groves of trees and great fields of sugar. It has great numbers of rabaos-plants of rape and cabbage, and many other fruits of the land, and [the] location of the city is upon a very great flat plain because all the land is such. All the cross roads are paved with bricks as the new road in Lisbon, and along all these roads and byways are great crowds of people, at which all things are sold in abundance, thus of provisions, as of all [other] things, and all very inexpensively.70

These travelers were also dazzled by the conspicuous displays of wealth in which the sultans periodically indulged. Their gestures of generosity both A Paradigm Shift

•

87

satisfied their subject population in material terms and at the same time enhanced their reputation beyond the boundaries of their land, as word of their actions spread. The same anonymous Portuguese manuscript goes on to describe one such occasion: . . . they told me that the King would give things to eat for the soul of his mother, and I went to see it, or how he would do it, and I saw in the same field one hundred and fifty cars loaded with cooked rice, and many loaves of bread, and rabaos-turnips and onions and figs and many other fruits of the land, and another fifty carts loaded with cooked and roasted beef and mutton, and thus great supplies of fish, also cooked, all to give to the poor.71

The description is rendered more powerful by the skepticism about hearsay, followed by the unknown traveler’s trip to see for himself. The sultans seem to have carefully created and projected this image of themselves both internally and externally through a variety of means, including public displays of wealth, urban planning and architectural innovation, and literary patronage, as well as skillful administrative and trade policies. Rather than feeling estranged from it, the Hindu rajas were drawn to the power of the Sultanate, as their adoption of elements of Sultanate courtly culture indicates. Through their acts of architectural patronage they chose to associate their works with the form of royal tombs, mosques, and shrines that were visible symbols of the sultans’ authority. Bengali Hindus were by no means exceptional in constructing their cultural lineage from the Islamic rulers who dominated most of South Asia from the twelfth century onward. In his analysis of the rise of the state of Vijayanagar amidst the Deccani Sultanates of south India, Phillip Wagoner has pointed to the Vijayanagar rulers’ selective adoption of the long tunics and pointed caps that were popular elite attire in the Islamicate world to express their political commitment to it. Drawing upon Marshall Hodgson’s formulation, he has used the term “Islamicization” to emphasize the social and cultural complex associated with Islam but also extending across religious boundaries to non-Muslims, as distinct from Islamization, which pertains to Islam in the religious sense. He has observed that the “local Indic elite publicly adopted these forms . . . as a means of effecting their symbolic participation in the more universal culture of Islam, thereby enhancing their political status and credibility in the eyes of other participants in the Islamicate cultural system.”72 Like the Hindu rulers of Vijayanagar, the Mallas of Vishnupur seem to have oriented themselves to the culture of the Sultanate court in order to establish their authority in the context of Islamic political preponderance in Bengal as well as most of north India. Terra cotta representations

88

•

Temple to Love

of elite figures wearing similar apparel on the base moldings of Ratna temples suggest aspects of an Islamicate court imagined at Vishnupur (figure 2.19). On Malla temples such as the Keshta Ray are several depictions of male figures seated with their legs drawn up on a divan, supported by cushions, perhaps like the sultan on the gilt sofa receiving Portuguese emissaries described earlier. The figures are engaged in various activities, including embracing female companions, receiving attendants, enjoying a dancer’s performance, and smoking a hookah. They wear the long outer garment identified by Wagoner as the kaba¯yi described in Telugu and Kannada texts. It fits the body snugly at the top, held together at the waist by a sash, then flares out below it to end variously at or below the knee. The sleeves are long and fitted all the way to the wrists. This attire is distinct from that typically worn by the divine or devotee figures occupying the length of the temple’s walls. Krishna and his dancing devotees, for example, are always bare-chested, wearing only a loose lower garment that covers the legs and is gathered at the waist with the end flowing down, probably a single length of fabric, unstitched like the modern dhoti. A shorter version of this loose lower garment, ending above the knee, is worn by figures of lower social standing, such as palanquin bearers who attend upon the tunic-clad elite. As Wagoner suggested in his analysis of Vijayanagar dress codes, the selective depiction of the tunic suggests that it was appropriate for specific social ranks in courtly contexts, whereas conventional Indic attire was preferred in the religious context of the Krishna lila. The significance of the tunic as an integral part of Sultanate political culture is attested by several Portuguese visitors to the Husain Shahi court at Gaur during the sixteenth century. João de Barros documents the sultans’ practice of giving symbolic gifts of tunics, which both legitimated them as donors and honored the recipient vassals: “they dress him [in] a cabaia-Turkish tunic, which the King bestows, with which he departs more honored . . . for it is a sign that he is now reconciled with the King, and with that honor of the cabaia-Turkish tunic he dispatches him to serve in his office.”73 Another anonymous author, who was part of the 1521 embassy of Gonsalo Tavares, notes that once the sultan decided to receive them, his officers dressed the envoys in richly brocaded clothing to prepare them for a formal appearance at court.74 Here then is another example of one group accommodating the rules of propriety of another to obtain political favor. Its depiction on Vaishnava temples indicates the prevalence of this cultural form beyond the Islamic courts by the seventeenth century. If we can assume that the Pala-period artistic convention of including small figures of donors at the base of large deity images

A Paradigm Shift

•

89

continues in the organization of figural ornamentation in these later temples, then the depicted Islamicate attire likely represents a Sultanate courtly practice imagined at Vishnupur. These tunic-clad elite figures may suggest a desire on the part of emergent Bengali Hindu rajas to engage with the Sultanate. Glimpses of elite life given visual expression in the terra cotta panels at the base of seventeenth-century Ratna temples suggest courtly pleasures akin to those enjoyed by the sultans. The base molding of the Keshta Ray Temple, for instance, describes a vivid scene of tiger hunting and taming (figure 2.20). If the molding is read as successive stages of a single activity, then the animal is being captured by a hunter poised in the trees above to the right. The figures on the extreme right seem to be chaining the legs of the struggling animal and suspending it from a pole to carry it off. To the left are stages in a fight, with the man spearing the tiger from below and then offering it a large item, perhaps food, which the tiger twists back to grab in its mouth. Such tiger taming and fighting must have been a popular pastime among the Sultanate elite, as Chinese travelers describe it in some detail: [Thus there are people who] go about the market places and to the homes with a tiger held up by an iron chain. They undo the chain and the tiger lies down in the courtyard. The naked man then strikes the tiger who becomes enraged and jumps at him and he falls with the tiger. This he does several times, after which he thrusts his fist in the tiger’s throat without wounding him. After this performance he chains him up again and the people of the house do not fail to feed the tiger with meat and reward the man with money. So the tiger tamer has a promising business.75

Such terra cotta depictions of courtly entertainment previously popularized by the Sultanate suggest another upper-class leisure activity now envisioned at the Malla capital emerging at Vishnupur. The depictions can be read as attempts on the part of the Mallas to elevate their social standing, to set themselves apart from their subjects, and to participate, instead, in the courtly paradigm set in place by the sultans.

The Mallas and the Mughals Following the Mughal invasion of Bengal in 1575, as successive Mughal governors struggled to extend the imperial administrative system amidst sustained resistance into the early decades of the seventeenth century, the Mallas consolidated their authority over the forest tracts of southwestern Bengal. Their political activities suggest wariness of the Mughal empire while their cultural patronage indicates allegiance to the

90

•

Temple to Love

deposed Sultanate. The adherence to the Sultanate expressed in their monuments must also be understood in this context of contemporary politics in the recently annexed Mughal province. The Ratna temple emerged in the wake of protracted insurrection and turmoil that followed the conquest. When Akbar’s army crushed the Karranis, the last independent dynasty to rule the region, the absorption of this region into the Mughal empire was far from accomplished. Instead, Akbar’s victory initiated a period of uprisings spearheaded by dissident Afghan chiefs. As the historian Jadunath Sarkar writes, When Daud Karrani laid down his sword at the feet of Munim Khan in the darbar tent at Katak, the ceremony merely proclaimed the de jure annexation of Bengal to Akbar’s empire; but the actual imposition of imperial peace and orderly Mughal administration on Bengal was still far off.76

From the last quarter of the sixteenth century into the first quarter of the seventeenth, the local rulers who had served under the Sultanate, both Hindu and Muslim, rallied their troops and came together to cooperate as allies against the invading Mughal generals. The best-known of these vigorous resistance movements against imperial Mughal power was a coalition organized by twelve local kings called Bara Bhuiyans, all former governors of the Bengal Sultanate, gathered under the leadership of the Afghan chief Isa Khan, who controlled a sizeable part of modern Bangladesh.77 The Goanese Jesuit mission, in its annual letter of December 1600, described the situation in Bengal as follows: Twelve princes, however, called Boyones [bhuyans] who governed twelve provinces in the late King’s name, escaped from this massacre. These united against the Mongols, and hitherto, thanks to their alliance, each maintains himself in his dominions. Very rich and disposing of strong forces, they bear themselves as Kings, chiefly he of Siripur [Sripur], also called Cadaray [Kedar Rai], and he of Chandecan [Raja Pratapaditya of Jessore], but most of all the Mansondolin [“Masnad-i Ali” title of Isa Khan].78

Like the last sultan, Daud Karrani, they nominally submitted to the Mughal governors when cornered by Mughal military might. However, when the Mughal army moved to quell another area, and the threat was no longer imminent, they shook off the allegiance and reasserted independence by organizing insurrections or refusing to pay tribute. With expediency now a priority, the Bengali landholders, patrons of this new architecture, seem to have cast aside their religious differences, if this was ever critical to their political agendas, to rally together against external threat.

A Paradigm Shift

•

91

Vir Hambir, the Malla ruler of Vishnupur, responded to the Mughals in similar fashion. Akbar’s official biographer, Abul Fazl, approvingly noted his cooperation with Jagat Singh, son of the great Mughal general Raja Man Singh, in the pursuit of Qutlu Khan Lohani, the Afghan ruler of northern Orissa. Not only did he offer intelligence, he went further to send troops to the rescue of the impulsive young man when his life was in danger during the campaign of 1590: . . . the wicked Qutlu suddenly fell upon him with a large force and prevailed over him. Though the land-holder, Hamir warned Jagat of Bahadur’s craft and of the dispatch of an army to his assistance he did not accept the news. After thousands of efforts he sent some scouts. At the end of the day the enemy arrived. Owing to the breaking of the thread of deliberation and arrangement most of the men fled fighting. Hamir brought away that infatuated young man and took him to Vishnupur.79

At the beginning of Jahangir’s rule (1605–27), however, the tide turned. Bengal’s new Mughal governor, Islam Khan, sent troops against Vir Hambir, which suggests insubordination on the local ruler’s part. Precisely what form this insubordination took, however, was not documented. As Mirza Nathan, a Jahangiri loyalist who followed the governor to Bengal, writes, . . . then he despatched a force of two thousand cavalry, and four thousand infantry, consisting of the imperial and his personal contingents, under the command of Shaykh Kamal to fight against Bir Hamir, Shams Khan, and Salim Khan whose territories lay adjoining to one another. Instructions were given that if they submitted, they should be given protection and brought with comfort; but if unluckily they took recourse to impertinence and violence, their country should be conquered, and if they fell into their hands they should be given proper punishment and brought to him as prisoners. If they were killed in battle, their heads should be brought to his presence.80

At this time, however, Vir Hambir surrendered without too much resistance, and the imperialist Mirza Nathan commended him for it: Bir Hambir, due to his foresight and sense of honour, did not take recourse to deception; he came out and met Shaykh Kamal and led him to the territory of Shams Khan.81

Such gestures of submission alternating with acts of resistance continued through the seventeenth and eighteenth centuries in this southwestern periphery of the province.82 Mughal governors remained cautious and kept a watchful eye on the Mallas. Nawab Murshid Quli Khan (1697– 1712) made separate revenue arrangements for Vishnupur through a

92

•

Temple to Love

representative of the Malla Raj posted at his capital in Murshidabad.83 Subsequently, Nawab Shuja Khan (1727–39) sent out an army under his commander Zafar Khan against Malla Raja Gopal Singh (1713–52). However, the general was repelled. Shuja Khan himself then marched against the Mallas. At this time Gopal Singh, like Vir Hambir before him, acknowledged the nawab’s suzerainty. He was ritually incorporated with the presentation of a turban called a shiropa, much as Akbar’s allies had been assumed into the body of the emperor with a robe.84 It is hardly surprising that imperial Mughal accounts characterized the Malla rulers as rebellious. The attempts to resist and to organize forces to overthrow the Mughals in the early seventeenth century reflect the fact that Bengal’s zamindars were not convinced of the legitimacy and stability of Mughal rule. Rather, they regarded themselves as the political heirs to Bengal in a situation they must have perceived as a power vacuum after the defeat and death of Sultan Daud Karrani. By challenging Mughal governors, refusing to pay tribute, and harboring enemies of the Mughals, these local landholders sought to shake the Mughals off and to fill the perceived void by assuming the position of the sultans as independent rulers of Bengal. As the Mallas entered the political arena, through acts of war and diplomacy, amidst the turbulence following the demise of the Sultanate, they made an equally bold move in sponsoring experimental temples in a style that clearly signaled the power of the sultans. It now provided visual expression of Malla authority. In looking to the Sultanate to legitimate their position, these emerging local dynasties rejected Mughal political authority. Implicit in the choice of Sultanate mosques for their architectural patronage is a rejection of the Mughal style of buildings, introduced into Bengal by Jahangir’s governors at the regional capital established at Dhaka by Raja Man Singh in 1602. By the mid-seventeenth century, a distinct provincial Mughal style of monuments could be recognized in that city’s mosques and the fort at Lalbagh, for example. The Ratna temples therefore developed concurrently as a parallel architectural formation.85 In this political context, the startling similarities between mosques of the sixteenth century and temples of the mid-seventeenth century become comprehensible as a statement of autonomy. The originality proclaimed in the Ratna temples’ inscriptions, addressed in the introduction, lies partly in the dual strategy of distinguishing themselves from their Nagara precursors and audaciously asserting themselves in the Indo-Islamic political culture in which they emerged. They articulate this new orientation in the vignettes of courtly life in their base-moldings, and even seek out a new model for their spatial organiza-

A Paradigm Shift

•

93

tion and aesthetic in the monuments built by the sultans and the local hut forms that the sultans had integrated into their buildings. This paradigm shift both served the needs of the devotional community for collective worship and expressed the political ambitions of their patron rulers. That Sultanate frame is tailored for Vaishnava purposes with a sheath of terra cotta figural imagery that maps the Gaudiya cosmology and mirrors the community’s ritual activities. The adapted form of the Sultanate mosque now served as a foundation for the pleasure pavilion of the gods. Gaudiya Vaishnava identity is expressed through this pastiche of architectural elements and stylistic features and the manner in which the architecture is then displayed for ritual purposes to the devotional community.

FACING PAGE: (TOP) Figure 2.1. Tantipara Masjid, Gaur

(BOTTOM) Figure 2.2. Qadam Rasul, Gaur

94

•

Temple to Love

A Paradigm Shift

•

95

96

•

Temple to Love

Figure 2.5. Masjid at Kushumba

FACING PAGE: (TOP) Figure 2.3. Goaldi Masjid, Sonargaon

(BOTTOM) Figure 2.4. Jami Masjid, Bagha

A Paradigm Shift

•

97

98

•

Temple to Love

Figure 2.8. Adina Masjid, Hazrat Pandua

FACING PAGE: (TOP) Figure 2.6. Jami Masjid, Atiya

(BOTTOM) Figure 2.7. Egaroshindur, Sadi’s Mosque

A Paradigm Shift

•

99

(TOP) Figure 2.9. Motichura Masjid, Rajnagar (BOTTOM) Figure 2.10. Ruin, Kulut

100

•

Temple to Love

(TOP) Figure 2.11. Domestic Hut, Birbhum (BOTTOM) Figure 2.12. Domestic Hut, Birbhum

A Paradigm Shift

•

101

102

•

Temple to Love

Figure 2.14. Porch Ceiling, Shyam Ray Temple, Vishnupur

FACING PAGE: Figure 2.13. Plan of Keshta Ray Temple, Vishnupur

A Paradigm Shift

•

103

104

•

Temple to Love

(TOP) Figure 2.17. Terra Cotta Wall Panels, Jami Masjid, Bagha (BOTTOM) Figure 2.18. Jami Masjid Interior, Bagha

FACING PAGE: (TOP) Figure 2.15. Eklakhi Mausoleum, Hazrat Pandua

(BOTTOM) Figure 2.16. Jami Masjid, Salban

A Paradigm Shift

•

105

Figure 2.19. Court Scene, South Façade, Keshta Ray Temple

106

•

Temple to Love

Figure 2.20. Tiger Taming, South Façade, Keshta Ray Temple

A Paradigm Shift

•

107

3. Acts of Accommodation

When turning to Sultanate forms for the lower story, the architects of Ratna temples did not entirely ignore the region’s pre-existing Nagara (Rekha) temples. One of the ways in which Sultanate mosques and shrines were modified for the ritual requirements of the emerging Gaudiya Vaishnava community was through the addition of an upper pavilion. A range of experiments were conducted on the single-storied Nagara temples to incorporate a second story for hosting Vaishnava festivals and simultaneously creating an opportunity to expand the developing ritual process. The Ratna and the Nagara are, however, distinct typologies, and they were built concurrently in Bengal during the seventeenth century. The innovations visible in a group of Nagara temples from the end of the sixteenth century mark phases in the transition from a single- to a double-storied structure and show varying degrees of antecedence to the Ratna form. One variation for the Ratna form involved setting the Nagara structure upon a Sultanate base (figure 0.7). I interpret this choice as accommodating the older model within the new Ratna frame developed for Vaishnava practice. Such employment of the Nagara included the architectural and cultural past of the patrons and allowed them to embrace a larger audience. The architectural experiments thus point to the porousness of the emergent Vaishnava community and its ties to the region’s older Brahmanical practices associated with Shaiva and Shakta worship conducted in Nagara temples. However, the diminutive Nagara shrine is located upon the visually dominant Sultanate base. These choices and the manipulation of material from an architectural heritage shared with the patrons’ Brahmanical Hindu past make the monuments suitable for Vaishnava ritual and express the political aspirations of the Hindu patron kings.

108

•

Temple to Love

It is also important to remember that not all Ratna temples are equally rooted in the Nagara tradition, and this study does not mean to overlook the other architectural strains that came together in the formation of the Ratna temple. In fact, Ratna temples from the middle of the seventeenth century demonstrate an astounding variety in their upper shrines. They range from Sultanate mosques and chala huts in the seventeenth century to flat upper terraces and even steeply tapering Gothic spires in the nineteenth century.1 The diversity displayed in the upper shrine points to the willingness to experiment with the rich variety of architectural models available in the seventeenth century. No two temples have identical upper levels despite the remarkable uniformity of the Sultanate base of the temple’s lower story. Instead, the lower story, developed from Sultanate sources with its hut accents and veneer of Vaishnava imagery, seems to have provided a satisfactory common denominator that distinguished the typology. The multiplicity of upper shrines indicates that the various options were probably attempts to embrace and subsume a wider audience within the fold of Vaishnavism. The pastiche-like juxtaposition of modes in the upper and lower levels also accommodated the reality of Hindu landholders in an Islamic political system.

Theories of Origin and Development Beginning in the earliest documents of the colonial period and continuing for the next two hundred years, scholars encountering the Ratna temple have commented that it is noticeably different from the Nagara typology. Fergusson, as noted in the previous chapter, characterized these monuments as aberrations in relation to his tripartite classification of Indian architecture. Aside from distinguishing the Ratna temple as a separate type, however, these scholars did not analyze the nature of the Ratna’s relationship to other architectural forms. They responded, instead, to the simulation of hut elements on the Ratna façade: Its leading characteristic is the bent cornice, copied from the bambu huts of the natives. To understand this, it may be as well to explain that the huts in Bengal are formed of two rectangular frames of bambus, perfectly flat and rectangular when formed, but when lifted from the ground and fitted to the substructure they are bent so that the elasticity of the bambu, resisting the flexure, keeps all the fastenings in a state of tension, which makes a singularly firm roof out of very frail materials. It is the only instance I know of elasticity being employed in building, but is so singularly successful in attaining the desired end, and is so common, that we can hardly wonder when the Bengalis turned their attention to more permanent modes of building they should have copied this one.2

Acts of Accommodation

•

109

Others were so taken by this copying that they speculated that Ratna temples might have originated from these huts. Francis Buchanan-Hamilton, a surgeon and botanist employed by the Bengal Service to undertake general surveys from 1794 to 1816, seems to have been the first to propose such a theory of the origin of Ratna temples. On the basis of his travels around the Bengali countryside and observation of architectural forms, he reads their development from temporary to permanent and from more modest to elaborate structures. At the bottom of his evolutionary chain are sthans, transient outdoor shrines consisting of heaps of earth or terraces under trees. The more lasting structures were “gradual improvements” made to these impermanent roadside shrines. The enhanced versions were thatch-roofed mud huts, decorated with paintings and banana pith carving, used for housing clay images during festivals. His next stage involved the use of brick, which was more enduring than mud and thatch. These brick pavilions were flat-roofed buildings, which housed deity images that were tended by priests. Of the double-storied temples of the seventeenth century he writes, The next step is to add a kind of pyramid [pavilion] to the roof of the temple, which then becomes a Mondir. The Mondirs are often cased with carved tiles, and at any rate are plastered on the outside, and the ornaments on the plaster in general possess some taste. Many Mondirs built of late, instead of the pyramid, have adopted the dome of the Mosque.3

From this single-towered Ratna form he sees the pancharatna developing by the enlargement of the central “pyramid” and addition of others, and the navaratna by the division of the roof in two stages with a “pyramid” at the corner of each. He does not directly compare these to the Nagara tradition, however. The generations of scholars who succeeded him only offered variations on his trajectory, but like him they did not offer any substantial evidence to support their assumptions.4 The only scholar to probe the relationship of the Nagara and Ratna typologies is David McCutchion. He proposed three modes of sacred structures that might have contributed to the development of the Ratna temple: panchayatana, a five-shrined version of the Nagara temples of north India; Sultanate mosques and mausoleums; and the single-towered Nagara temples of Bengal.5 McCutchion perceived the Ratna form as essentially a Nagara temple with porches wrapped around the sanctum, in spite of noting connections also to Sultanate mosques: “The porch or corridor on all four sides suggests that the ekaratna may have originated from a rekha deul [Nagara temple] provided with a covered circumambulatory veranda all round.”6 He does not take notice of the transition

110

•

Temple to Love

from a single- to a two-storied structure. The upper structure of the double-storied temple is in fact a complete miniature temple, with its own spire. When that upper structure is a miniature Nagara temple, as it often is, the spire takes the form of a Nagara shikhara. Further, the lower level of the double-storied Ratna temple rejects the offset form of the Nagara temple tradition completely. It draws from a different set of sources— Sultanate mosques and vernacular huts—and is not simply a Nagara temple with porches as McCutchion suggests. For the complexities of the upper structure he found panchayatana origins to be the “more plausible model.” Panchayatana temples have groupings of four smaller shrines around the central temple, all on a single plinth. He observed that when the corner shrines of panchayatana temples are connected by verandas and surmounted by chatris (pavilions), their appearance is quite similar to pancharatna temples (Ratna temples with a central structure and four corner structures on the upper story), and he proposed that the panchayatana may have been a source for the replication of multiple shrines found on the upper level of Ratna temples. He argued that this origin explained why Ratna temples with multiple upper structures developed earlier, in his view, than those with single upper structures, and postulated, but without elaboration, that ritual requirements must have dictated the development of the multi-shrined Ratna temples. Temples with a single upper structure, he posited, were obtained by elimination of the corner elements of the five-shrined form. Because he did not recognize the Ratna temple as a two-storied form, he did not distinguish between the Nagara tower and the variety of structures that were used for the double-storied temple’s upper shrine from the mid-seventeenth to mid-eighteenth centuries. He also did not address the difference between the sanctum on the lower level and the multiple shrines on the upper story of the Ratna form. His comparison depends more on analogy than analysis of unique formal features. While McCutchion and the colonial scholars before him noted the odd amalgamation of the Ratna morphology, neither provided satisfactory explanations of its development. Examining the broader picture of Sultanate architecture, Nagara temples, and the chronology of Ratna temples instead reveals a more complex and experimental development of the Ratna form hand in hand with the ritual needs and political maneuverings of the patron community.

Addition of a Second Story The two dominant models of sacred architecture available for Ratna temples to draw upon, Nagara temples and Sultanate shrines and mosques,

Acts of Accommodation

•

111

were both single-storied forms. The transition from a single- to a twostoried temple to fulfill the ritual needs of the Gaudiya Vaishnava community involved experiments upon both. While the Sultanate type provided a paradigm for the lower story that was reproduced steadily through the seventeenth century, experiments conducted on the Nagara form attempted to incorporate a second story above the sanctum. A sequential development in architectural forms that culminated in the Ratna structure cannot, however, be traced from the fragmentary record of standing monuments. Rather, the experimentation visible in these structures suggests that the process was intermittent and not unidirectional; many of the experiments were in fact never followed through in later temples. Early grafting of elements from one mode onto the other is easy to spot in late-sixteenth-century temples built across West Bengal and Bangladesh. These experimental temples tentatively graft an architectural element or a decorative motif from the Sultanate repertoire or hut architecture onto the Nagara temple. These additions are typically not very well integrated into the Nagara structure and in some cases the monument draws attention to the substitutions. Such eclectic choices show that Bengali architects at this time not only were aware of the typical Nagara and Sultanate forms of architecture, they also created local variants by merging typological features of these dominant regional architectural modes. The sixteenth-century Radha Damodara Temple at Ghutgeriya, Bankura district, reveals such modifications on the basic Nagara frame developed in the region (figure 3.1). In keeping with the Nagara convention, the square sanctum is surmounted by a tall curvilinear spire.7 The temple stands on a broad, low plinth, with unadorned base moldings. The plain wall and tower are divided into five vertical offsets. Deep double moldings, with a shallow recess between them, separate the sanctum from the spire. The entrance to this temple, in contrast to the typical rectangular doorways of earlier temples, takes the form of a cusped, pointed arch supported on engaged pilasters (figure 3.2). The doorframe thus looks to Sultanate sources. High-relief carvings of lotus blossoms dominate these arch spandrels as they do the Sultanate-period gates and mosques of Gaur. In addition, figural forms are introduced into the ornamentation of this Vaishnava temple. Radha and Krishna, devotees, drummers, and curling vines complete with lotus blossoms inhabit the arch spandrel. The arch is enclosed within a rectangular frame consisting of small square panels bounded by a band of alternating circles and diamonds with flowers. These panels contain figures familiar from the Puranas, including Brahma on his goose, Shiva riding his bull, Vishnu sleeping on the coils of the snake Shesha, and Indra on his elephant above. Episodes from the life of Krishna are arranged vertically on either side of the doorway, juxtaposing

112

•

Temple to Love

the emerging visual imagery of the Gaudiya Vaishnava tradition with the pre-existing repertoire. This doorframe, its ornamental motifs and figural forms, and their style resemble those of later Ratna temples. The new type of door thus draws from different sources—the region’s popular Vaishnava and non-Vaishnava figures grafted onto a familiar Sultanate frame. The juxtaposition of the older and newer traditions is reiterated in a contrast of materials—the structure is built of buff-colored sandstone, while the doorway is carved from pink laterite. Such combination of features from the region’s Nagara tradition with elements from other architectural sources came to characterize the Ratna temples of the midseventeenth century. Like the Radha Damodara Temple at Ghutgeriya, the late-sixteenthcentury Mathurapur Deul at Madhukeli, further east in Faridpur district, inserts Sultanate decorative features into the Nagara body of the temple (figure 3.3). The Nagara surface is sheathed with delicate terra cotta decoration like that found in the Sultanate tradition. The horizontal bands of organic and geometric patterns are similar in treatment, for instance, to those of the sixteenth-century Jami Masjid at Bagha. These rows of ornament wrap in continuous bands around the twelve-sided walls and tower. A few narrative sequences, including scenes from the Ramayana, are interjected high above eye level (figure 3.4). These figures represent early experiments in the transition from the non-figural terra cotta motifs suitable for a mosque to the display of iconography required for Hindu temples. Like Ghutgeriya, this temple inserts a Sultanate-type cusped and arched doorway into the Nagara frame. Other Nagara temples adopt elements of hut architecture that were also used in contemporary Sultanate structures. The 1598 Krishna Temple at Baidyapur, Burdwan district, replaces the very top of the Nagara tower with a four-part hut roof.8 In doing so, it follows the roofing experiments of earlier Sultanate structures such as the Chota Sona Masjid at Gaur. The char-chala, with its characteristic curved eave, substitutes for the amalaka and kalasha, the finial elements typical of the Nagara tradition. Expanding upon such experiments with the tower, some Ratna temples ultimately replace it with a diminutive version of the entire hut. The Shiva Temple at Nandi also employs the distinctive roof of the thatched hut.9 The hut’s curved cornice replaces the horizontal laminations of the Nagara temple’s tower. In later, more fully developed Ratna temples, that curved eave is used boldly to create a transition between the upper pavilion and the base in Ratna temples, as discussed in the previous chapter. These bold experiments make sense within the context of a prosperous group of Hindu local kings taking over from their Sultanate forebears. The combination of thriving international trade, initiated under

Acts of Accommodation

•

113

the sultans and stimulated by these successors, and the political agendas of these local rulers, as discussed in the previous chapter, created the ground for architectural play with forms. The substitutions across modes visible in the standing monuments suggest that the Bengali architects took advantage of the opportunities afforded by these conditions. Aside from the borrowings from Nagara, Sultanate, and hut forms in the details of individual monuments, some structural changes in the Nagara form foreshadow the development of the two-storied Ratna temples. The variety of experiments concentrated in the vicinity of Vishnupur suggests a focused search for a new organizing principle fueled by royal patronage and religious enthusiasm. None of these temples can be defined as purely Nagara, as they no longer have the curvilinear tower characteristic of that typology. These “transitional” temples instead reveal various architectural approaches to the problem of setting a freestanding upper structure on top of the cubical sanctum. The Ratneshvar Shiva Temple at Jagannathpur, Bankura district, built probably at the beginning of the seventeenth century, reveals an early moment in the separation of the tower from its Nagara base (figure 3.5). The walls of its sanctum are six feet deep, this extraordinary thickness supporting the weight of the tower. The spire is offset from the corners of the sanctum, which acts as the upper structure’s socle. A prominent set of base moldings on the superstructure makes the spire conceptually independent of the substructure, as if a separate upper structure had been superimposed on the sanctum roof. The laterite Malleshvar Shiva Temple, built by the Malla ruler Vir Singh (who is possibly the same as Vir Hambir) in 1622 in Vishnupur, is not a Nagara temple, although it has been classified as such in the literature.10 Only its base moldings faintly recall Nagara conventions (figure 3.6). The temple stands on a raised plinth with five unadorned moldings. The exterior wall of the sanctum has very slight offsets, but remains basically square. This division of the wall into five parts is only a slight step away from Ratna temple walls whose five sections are flat, separated only by ornamental devices such as bands of floral pattern and thin pilasters. On top of this cubical base, an octagonal upper structure is stepped back and centered on the roof of the sanctum. There is no staircase leading to this upper pavilion, however. The lack of access indicates that this upper space was not yet functional, although the suggestion of an open upper chamber must have had symbolic importance for the Gaudiya Vaishnava community. This second story was heavily repaired in the early twentieth century, making further analysis of its present form precarious. The Jadab Ray Temple at Jadabnagar, built by the Malla Raja Raghu-

114

•

Temple to Love

nath Singh in 1650, is a true double-storied temple, with both a stairway, located in the southwestern corner of the building, and a functional upper story (figure 3.7). The upper structure, however, is barely set back from the edge of the lower story, giving this monument a heavier appearance than most contemporary Ratna temples. This slight recession is another experiment in the separation of the tower. However, the two shrines were conceived as parallel structures, as the walls of the upper shrine closely imitate those of the lower level. Both stories have central arched openings set within rectangular frames. Both shrines have central spaces capped by shallow domes supported on pendentives. The upper dome is broken by a series of vertical projections related to offsets in the wall below, as in Nagara towers. Although this domical superstructure has been given some Nagara elements, and the recession of its upper tower is almost imperceptible, this temple can no longer be called Nagara. The square lower story reproduces the characteristic features of the Ratna type. The sanctum lacks the typical Nagara set of base moldings; instead, a single, broad molding is used. The walls do not have projecting offsets, but are divided by pilasters into a broad central section enclosing the arched entrance, a wall frieze, and a corner, like the walls of the Malleshvar Temple. The doorway’s pointed arch is supported on squat faceted columns and is set within a rectangular frame with rosettes in the spandrels. The upper cornice sports the characteristic curve of the chala tradition. Shallow porches with vaulted ceilings border the southern and eastern sides. This temple therefore manifests the distinguishing traits of the Ratna typology. These early experiments in the thickening of the sanctum walls, the separation of the tower from the base, the creation of the upper level as a parallel to the lower story, and the furnishing of a functional space with the introduction of stairs developed a form incorporating all the basic elements of a Ratna temple. The typical Ratna temples of the midseventeenth century, such as the Shyam Ray and Kala Chand Temples at Vishnupur, the Gokul Chand Temple at Gokulnagar, and the Radha Ballabh Temple at Krishnanagar, elaborate on the innovative features represented in the Jadab Ray Temple (figures 0.3, 0.5, 0.7, 1.3). While the lower story is consistently a cubical sanctum flanked by vaulted porches following the Sultanate model, the form of the shrine placed on the upper terrace for festivals varies. Separating the two levels is typically a chaladerived cornice with a pronounced curve. Stairs located at the southwest corner of the building, between the sanctum and the porch, lead to the upper story, which is often an octagonal space. However, it may be enclosed by an outer porch whose exterior walls determine whether it appears from outside as a square or an octagon. In some cases four corner

Acts of Accommodation

•

115

spires are also supported on the upper terrace, as in the Shyam Ray and Gokul Chand Temples. Terra cotta narrative panels on Vishnupur’s Ratna temples often illustrate the various options that had become available for temple construction in the seventeenth century.11 A series of panels on the Madan Mohan Temple’s south façade display both single- and double-storied structures and a variety of roofing techniques (figure 3.8). These include a single-storied hut roof with a curved cornice and typical Nagara finials; a single-storied form with a tall, layered Nagara tower; a single-storied structure on the Sultanate model with an arched entrance and lotuses in the spandrels, but Nagara finial elements on top; and a double-storied form with a curved lower cornice with a small shrine following the hut paradigm stacked above it. The architectural representations serve as background for the events of Krishna’s life, suggesting that the differing architectural possibilities were given coherence by their ritual use as Krishna’s abode.

Variation in the Upper Shrine A survey of the standing monuments at Vishnupur reveals that while the base of the two-storied temples is consistently a Sultanate cubical form embellished with a curved upper eave and terra cotta panels, the diminutive shrines placed above display no such regularity of features. Rather, the upper shrines test out the dominant architectural models available in the region. In the seventeenth century these were Nagara temples, huts, Sultanate tombs, shrines, and mosques, and, by the eighteenth century, Mughal chatris (pavilions) as well. Ratna upper shrines either imitate these models in their entirety, or selectively appropriate their distinctive features and flaunt them in new and different combinations. The most popular choice for an upper structure at Vishnupur seems to have been a miniature version of the curvilinear-towered Nagara temple, first used in the Kala Chand Temple (figure 0.7). Like older Nagara temples such as the fifteenth-century monuments at Barakar, this upper shrine stands on its own plinth (figure 0.15). Each wall has a wide central projection, flanked by three others that continue into the spire. A stringcourse divides the wall at the level of the arch springing on the central offset. The juncture between the spire and the wall consists of a wide recess flanked by a pair of moldings. The spire’s squat curvature, horizontal laminations, and traditional finial elements all belong to the region’s Nagara conventions. In fact, this upper shrine diverges only in its doorway, which is a Sultanate arched entrance rather than the Nagara post-

116

•

Temple to Love

and-lintel frame. This basic formula is repeated for many temples, including the Radha Madhav, Radha Govinda, and Nandalal, and the Jor Mandir trio at the southern end of the town. Other upper shrines adhering to the Nagara convention display the diversity present in the Nagara temples of the region. Elements of Sultanate and hut construction are also inserted in the basic Nagara frame. Some upper structures, like that of the Ghutgeria Temple, have shikharas with a rounded tower profile rather than a flat upper altar (uttaravedı¯ ). In some cases, the upper story has a smooth shikhara, without the horizontal laminations that distinguish the Kala Chand Temple. The number and width of the vertical projections running along the length of the shrine and tower also vary. On the Murali Mohan Temple (figure 0.8) the vertical offsets of the walls are slender pilasters in the tradition of the Nagara temples at Telkupi and Banda, in Purulia district, rather than graded wall sections as in Barakar No. 4. The springing of the spire also varies. The Kala Chand Temple displays deep moldings with a recess, which is eliminated on the Radha Shyam Temple (figures 0.7, 0.11). Others mark the separation of the tower from the wall with curved hut cornices in a manner that makes the miniature temple above parallel to the Sultanate-based sanctum below. Some upper stories have corbeled interior ceilings like older Nagara temples; others use a dome on pendentives. In some cases the upper shrine has more than four sides, mimicking many-sided structural temples such as the Mathurapur Deul. The upper shrines of the Malleshvar Shiva Temple at Vishnupur and the Nandakishor Temple at Dvadasbari are octagonal, as are the central shrines of the Shyam Ray Temple at Vishnupur and the Gokul Chand Temple at Gokulnagar (figures 3.6, 3.9, 0.3, 1.3). The additional arched openings in these cases create a light and airy effect like a chatri (pavilion) rather than the dark, enclosed space of the lower level’s sanctum. This change is particularly noticeable at Dvadasbari, where the walls are replaced by arches supported on columns. The resulting upper shrine, like the chatris atop Delhi Sultanate tombs and Mughal palaces, is a pavilion for the enjoyment of the deity (compare with figure 3.10). Some Ratna temples rejected the Nagara model in favor of other superstructural types for the upper pavilion. The upper shrine of the Madan Mohan Temple is virtually a single-domed Sultanate structure like the Goaldi Masjid. Only its walls reveal Nagara projections and the smooth, rounded dome displays Nagara finial elements (figure 0.10). On Jadabnagar’s Jadab Ray Temple and Vishnupur’s Radha Shyam Temple, a series of vertical projections on the domical surface of the upper structure simulate the vertical bands of a latina Nagara temple (figure 3.7). It is as if the architects were testing the analogousness of the available

Acts of Accommodation

•

117

architectural models. Other Ratna upper shrines draw from hut architecture. On the Keshta Ray Temple at Vishnupur and the Radha Ballabh Temple at Krishnanagar, the upper shrine is a square structure with four pointed and tapered roof sections with curved cornices (figures 0.5, 0.6). The number of upper shrines is variable as well. In the Vishnupur region the oldest surviving monuments have additional shrines at the four corners of the upper terrace, supported by the wide base created by the addition of four porches around the sanctum.12 The resulting five-shrined upper story gave these temples the name pancharatna. The corner shrines of these temples are four-sided pavilions, while the central ones are octagonal (figures 0.3, 1.3). During the eighteenth century, more ambitious elaborations were based on this five-shrined form. The rectilinear base of the lower story was sometimes doubled by a smaller square or rectangular one centered above it. This additional structure constituted a base for a second set of five shrines, one central and four corner ones. These nine-pavilion temples are called navaratna. Up to twenty-five such miniature temple forms have been fitted above the temple’s base by adding sets of four corner shrines on additional stories. Although popular in other parts of Bengal, these proliferations do not seem to have been favored by the Malla rulers, who moved to the single-shrined version.13 As single and multiple upper shrines were built simultaneously, it is likely that the choice was based not on ritual use but rather on the degree of ostentation desired by the patron. Such playful combination of the distinctive elements of older architectural traditions, set above the unifying Sultanate base, created a composite structure parallel to the complex structures of power and political affiliation in seventeenth-century Bengal. By physically reproducing formal and aesthetic elements from a variety of available architectural modes, Ratna temples absorb and interweave the potency of each. These multiple signals suggest accommodation of a variety of regional authorities—Sultanate, Mughal, Malla, Islamic, Brahmanical—on the part of the Gaudiya Vaishnava community that predominantly patronized the Ratna form. The monuments give us visual clues that supplement the limited textual accounts of the processes of social and political negotiation between competing authorities in this period. A dome for the upper story, as discussed in the previous chapter, can be read in conjunction with the evolving authority of the Mallas in southwestern Bengal and their selfperception as heirs to the political authority of the independent sultans. The visual correspondences also parallel bhakti and sufi theological and liturgical equivalences in Bengal.14 The miniature Nagara upper structure points to the complex relationship of the Gaudiya Vaishnavas with their Brahmanical forebears. While the Sultanate base of the two-storied Ratna

118

•

Temple to Love

temple asserts the separation of Gaudiya Vaishnava architecture from that of older Hindu communities, the Vaishnava leaders clearly did not establish such a clean break. The form of the thatched hut as an upper shrine instead transcends religious boundaries. Huts were used as chambers for Hindu rituals, for religious monuments in the Islamic tradition, and also for domestic purposes. Associated with the Bengali countryside and rural communities, the hut proclaims a regional identity regardless of religious beliefs, as discussed in chapter 2.15 The Gaudiya Vaishnava community, like that of the Sultanate rulers before, sought a potent transsectarian architectural form. Their quest must be understood in conjunction with the need of a young religious community to create a wide popular base in Bengal. In the eighteenth century, the employment of chatris—the pavilion form associated with the palaces of Mughal emperors and their governors in Bengal—similarly expresses both architectural modernity and the local Hindu kings’ later attempts to reconcile with Mughal power. After Mughal domination of the region had become an incontrovertible reality, the later Malla rajas used Mughal features liberally within the established Sultanate framework of the Ratna temple. Pavilion forms for their gardens and also the experimental upper shrines of temples, the gatehouses of these temples, and their palace complex point to sharing across sacred and secular architecture as also in imperial Mughal architecture. Ironically, by the time the architectural integration of Mughal elements was achieved in the Ratna form, imperial political authority was collapsing in Bengal. While we cannot easily discover the specific goals of the patrons, their use of such a variety of upper structures suggests the political pastiche of the times. Tested out during a period of flux following the initial Mughal conquest of the region, these shrines likely represent the negotiations of the Vaishnava kings with other powerful groups in the region in efforts to establish their identity. At a moment of change and flexibility, multiple architectural and sculptural forms seem to have been used to signal a pastiche whose parts index the range of competing authorities amidst which the Gaudiya Vaishnava community coalesced. The variety does not seem to have posed a problem; rather, it seems to assert that the communities represented by these architectural models could be subsumed under the umbrella of Krishna worship.

Inclusiveness in Vaishnava Expressions How are we to understand the relationship of the Ratna to the older Nagara form? Both were built simultaneously in southwestern Bengal during the seventeenth century. The record of extant monuments and the

Acts of Accommodation

•

119

inscriptions of ruined temples do suggest, though, that the Ratna form was used only for the construction of Krishna temples at this time. The Nagara typology continued to be patronized, most often for Shiva and Devi temples, although a few Krishna temples were also built in this form.16 At Baital, for example, the Mallas chose the Nagara type for the Jhagrai Chandi goddess temple of 1659 and the Ratna for the temple to Krishna as Shyam Chand in the following year. Likewise at the later site of Ambika-Kalna, the Burdwan rajas built a Nagara temple to Shiva as Pratapeshwara alongside several multi-spired Ratna temples dedicated to Krishna. As late as 1811, Reverend W. Ward of the Serampore Mission continued to identify the Ratna form with Krishna worship in his ethnographic work on Hindu temples, images, and rituals. He distinguished five types of temples on the basis of the different deities worshipped in them: There are five kinds of temples among the Hindoos, one of which is dedicated exclusively to the lingu [lingam], another to Jugunnathu [Jagannatha], and another is appropriated to the images of any of the gods or goddesses. The first of these is called by the general name of Mundiru [mandir]; the second Daool [deul, i.e., Nagara], and the third Yorubangala [jor bangla]. The names of the other two are Punchu-rutnu [pancharatna], and Nuvu-rutnu [navaratna].17

He associates these last Ratna temples with forms of Krishna, as the pancharatna and navaratna “contain forms of Vishnoo, as Radhabullubhu, Gopalu, Mudunu-mohunu, Govindhu.” The Ratna therefore never displaces the Nagara; instead, the two co-exist, interacting intensely during this period. After the Malla rulers championed Vaishnavism, the majority of temples built at Vishnupur, their seat, were in the Ratna form. It seems to be associated with the formation of a cosmopolitan, urban center drawing on Sultanate models, while the older Nagara form was patronized in the rural hinterlands for the most part. With so many choices available, the selection of architectural forms for temple construction likely became politically motivated in the seventeenth century. Despite the assertion of independence from the region’s older architecture in the Ratna inscriptions’ insistence on originality, and the choice of Sultanate architectural features for the lower story, the Nagara remains one option for the upper shrine of the Ratna temple. The Ratna therefore does not entirely reject the Nagara convention but sporadically draws attention to its roots in the older architectural convention. However, when the Nagara is selected for the upper structure, its diminutive scale suggests that it is only one of several constituent parts of the overall Ratna framework. The use of the Nagara in the Ratna, in turn, validated the former and spurred on fresh experiments within the Nagara convention.

120

•

Temple to Love

The employment of the Nagara temple form as an alternative for the upper shrine of the Ratna temple can perhaps be understood in association with the complex relationships among the patrons of these monuments. To some extent, the emergent Gaudiya Vaishnava community defined itself in relation to Brahmanical practices, now associated primarily with Shakta and Shaiva worship and with the Nagara form, as the Ratna form was devoted exclusively to the bhakti style of Krishna worship. While the Vaishnava rhetoric is about difference, a closer look at Vaishnava beliefs and practices inevitably uncovers continuities alongside the newer ways to which the rhetoric draws attention. From his study of the sixteenth-century outpouring of hagiographical literature, Joseph T. O’Connell has observed that the bhakti challenge posed to the caste system and the authority of Brahmans by the encouragement of personal devotion without intercessors, for example, is contingent upon social context. While all devotees were welcome to attend festivals and kirtan sessions and could share in the consumption of prasada regardless of their caste status, this unrestricted sharing was confined to devotional ceremonial situations. In other circumstances devotees of different caste backgrounds did not dine together; nor did they tend to intermarry. He concludes that The devotees of Chaitanya’s Vaishnava movement thus maintained a two-faced or two-tiered stance toward social relationships: egalitarian affectivity with fellow devotees in devotional (sacred) situations; inegalitarian functionality (even with fellow devotees) in mundane (profane) situations. This stance may not have been heroic from the viewpoint of social ethics, but it was neither confused, nor inconsistent, nor for that matter, peculiar.18

His interpretation suggests that Vaishnava continuities with the prevailing social order were provisional, based on the adjustments and shifts in emphases necessitated by the devotees’ new bhakti perspective. The conditional accommodation of the caste-based social structure within the new Vaishnava worldview is paralleled in the qualified architectural expression of the Nagara within the Ratna framework. Like these architectural and social practices, acts of accommodation are discernible in the choice of terra cotta images that adorn Vaishnava temples. Although priority is given to depicting Krishna on the terra cotta panels, we also find a range of Puranic deities, identifiable by their particular weapons, their mounts, and other markers of their exploits. Sets of goddesses and the family of Shiva, for example, are very popular on these early Krishna temples. However, these deities are now repositioned within a new Vaishnava order in which Krishna reigns supreme. Their relative location on the monuments’ walls reflects their subordina-

Acts of Accommodation

•

121

tion in this new hierarchy. In the rare temples that have all four walls adorned with terra cotta panels, these gods are typically relegated to the side and back walls rather than the primary south façade; in those with terra cotta decoration only on the main entrance wall, they are placed in secondary locations such as base-moldings rather than at eye level, where Krishna dominates. The Vishnupur Shyam Ray Temple’s north porch displays the family of Shiva as a set of six panels that adorn the inner wall (figure 3.11). The uppermost left depicts Shiva with matted locks. He sits on a large bull, holding a drum and horn in his hands. In the adjacent panel is Durga in her martial role as slayer of Mahisha rather than as the more tranquil Parvati. She is ten-armed, bearing down upon the demon with her weapons and her feet. Her lion rears at him as well. Ganesha on a gigantic rat and Kartikeya on his peacock occupy the panels below them. At the bottom are the two goddesses Lakshmi (displaying lotus buds in each hand) and Saraswati (bearing a veena). In Bengal, these two goddesses are the daughters of Durga and Shiva. While the figures are exquisitely detailed, their location on the wall behind the sanctum suggests they were not central to the temple’s Vaishnava conceptualization. Likewise, the goddesses also feature on the Shyam Ray Temple, but again, on the inner wall of the west porch (figure 3.12). Here a set of female figures, who seem to correspond most closely to the Mahavidyas or Tantric wisdom goddesses, occupy nine panels.19 The relatively benevolent forms of the Goddess fill the three panels on the uppermost row. To the left, Saraswati stands with her slender veena in her hands; in the middle, seated upon a full lotus blossom, is a four-armed goddess who may be Tripurasundari; on the right Kamala, also known as Lakshmi and Sri, grasps the stalks of lotus buds in both hands. Kali dominates the middle tier, brandishing a sword and skull and trampling Shiva’s corpse. She is flanked by Kartikeya on the left and Ganesha on the right. Relegated to the bottom are the violent goddesses, suggesting their position in the new Vaishnava cosmology. On the left, Tara/Kali wears a garland of decapitated heads and a skirt of severed arms, and like Kali she stands on Shiva’s body. Next, Durga stabs the demon Mahisha with her trident. Finally, the most terrifying-looking, Chinnamasta, the headless goddess, spurts blood from her severed neck. The goddess panels are complemented by the ten incarnations of Vishnu on the other side of the porch, suggesting that they were conceived as parallel, co-existing, cosmologies. Even before the visual imagery for Ratna temples developed, Shiva and the goddesses had made their appearance in the growing body of Bengali Vaishnava literature. The Ramayana, for example, as recounted by the fifteenth-century poet Krittibas, reflects the role of the goddess in

122

•

Temple to Love

the Vaishnava fold as the arbiter of Rama’s destiny,20 and Shiva as the protector of Rama’s enemy Ravana. The prominence of the goddesses in Bengal is also indicated in the gradual conceptualization of Radha as one of the numerous forms of Devi. The Gitagovinda is among the earliest texts to hint at Radha’s status as a goddess, and subsequently the Vrindavan goswamis described Radha as Krishna’s shakti, the source of his power, as did the authors of the later Puranas.21 These works therefore treated her as analogous to the older goddesses Parvati and Lakshmi, the energies of the male gods Shiva and Vishnu. She is then associated with the Mahavidyas in the later Radha Tantra.22 Likewise, Subhadra, who accompanies Lord Jagannatha of Puri, is often worshipped as a form of Durga.23 The Vaishnava texts thus assert their worldview in relation to the region’s older cosmologies and practices, as do the images on the Krishna temples. The presence of Shiva and the goddesses in Vaishnava visual and literary forms parallels the process of self-definition on the part of the emergent Vaishnava community in relation to these older deities and Brahmanical worship practices, from which they remained inseparable. The earliest generation of poets and spiritual leaders identified with these older traditions prior to their persuasion to Vaishnavism. Chandidas, one of the early poets and author of the Srikrishna Kirtana, claimed to be a worshipper of Bashuli, to whom he offered his work.24 Bashuli is a goddess who continues to be popular in the western region of Bengal, and temples are dedicated to her at Chatna in Bankura district and Nanur in Birbhum. Despite their ostensible resistance, the early Vaishnava leadership also displayed interest in goddess worship, and maintained numerous Brahmanical practices alongside the new Vaishnava orientation.25 At Vishnupur, the worship of the goddesses, Shiva, and Krishna remains intertwined to this day. Like Chandidas, the Malla rulers had been devout goddess worshippers before Srinivas Acharya inspired Vir Hambir to embrace Vaishnavism. Their lineage deity was Mrinmoyi, a form of Durga, and they lavishly celebrated the autumnal festival of Durga Puja.26 They had also built Nagara temples to the goddesses and to Shiva before they patronized Ratna temples for Krishna worship, and continued to do so alongside the newer form. Their patronage of both deities and the different styles of worship associated with them may begin to explain the interweaving of features from the worship and temple styles associated with the one into those of the other. The goddesses shifted to a vegetarian diet as Vaishnavism condemned blood sacrifice, and conversely Krishna sometimes even ate fish, one of the five “m” offerings given to goddesses. Likewise, Mrinmoyi is believed to visit Shyam Ray in his temple on the night of the Jhulan festival.27 The depiction of these

Acts of Accommodation

•

123

older deities on the back walls of the Krishna temples suggests that the growing Vaishnava community acknowledged the prominence that Shiva and the goddesses enjoyed at Vishnupur. However, they were now relegated to a subsidiary role in the new Gaudiya Vaishnava cosmology mapped on the temple walls. Similarly, the Nagara temple form was acknowledged but reconfigured in the architectural expression of the emergent Vaishnava worldview. It was assimilated within the doublestoried frame of the Ratna temple as one of several options for its upper shrine. The new architectural pastiche, like the religious devotional movement with which it is associated, thus defined itself in relation to the matrix from which it emerged. Upon a consistent lower story derived from Sultanate and hut sources, the architects of the Ratna temple tested out various options for the upper pavilion. The most popular choice from the mid-seventeenth century seems to have been a miniature version of the region’s Nagara temple, which was itself the object of experiments by this time. I have suggested that the parts of this architectural pastiche point to the plural ancestry of the Gaudiya Vaishnava community and kings, and express the political aspirations of the latter. By the eighteenth century, the Ratna temple-type was firmly established as a formal alternative to the Nagara. Architects were not only aware of the two as distinctly different temple-types, but were sufficiently well versed in both modes to attempt playful experiments combining them. After the Ratna form became mainstream, temples like the Kalanjay Shiva Temple at Patrasayer, Bankura district, placed a full curvilinear-spired Nagara temple above a large square base with curved cornices (figure 1.6). However, there are no stairs to provide access to the upper space and the second level is a blind story, only imitating a Ratna temple’s upper shrine. That lack of functionality points to ritual shifts, and to the co-option of the Ratna form for housing other deities from the late eighteenth century onward. Unlike the small metal images of Radha and Krishna residing in the true double-storied temples, Shiva’s lingam is fixed permanently in the floor of the octagonal sanctum of the Kalanjay Shiva Temple. The immovable object of worship in this temple had rendered a functional pavilion on the upper level unnecessary. Vaishnavism came to dominate in the seventeenth century in part by accommodating the older religious practices, which it simultaneously challenged. During this time, the kings patronized an architectural form that incorporated the visual markers of those rival groups. After its successful establishment, Vaishnava visual imagery and architectural forms were absorbed by those rival communities for their own revitalization. The direction of appropriation was thus reversed from the middle of the eighteenth century. 124

•

Temple to Love

Figure 3.1. Radha Damodara Temple, Ghutgeriya

Acts of Accommodation

•

125

Figure 3.2. Radha Damodara Temple Doorway, Ghutgeriya

126

•

Temple to Love

Figure 3.3. Mathurapur Deul, Madhukeli

Acts of Accommodation

•

127

Figure 3.4. Terra Cotta Ornamentation, Mathurapur Deul, Madhukeli

128

•

Temple to Love

Figure 3.5. Ratneshvar Temple, Jagannathpur

Acts of Accommodation

•

129

Figure 3.6. Malleshvar Shiva Temple, Vishnupur

130

•

Temple to Love

Figure 3.7. Jadab Ray Temple, Jadabnagar

Acts of Accommodation

•

131

Figure 3.8. Terra Cotta Ornamentation of South Façade, Madan Mohan Temple, Vishnupur

132

•

Temple to Love

Figure 3.9. Nandakishor Temple, Dvadasbari

Acts of Accommodation

•

133

Figure 3.10. Malla Estate Garden Pavilion, Vishnupur

134

•

Temple to Love

Figure 3.11. Family of Shiva, North Porch, Shyam Ray Temple

Acts of Accommodation

•

135

Figure 3.12. Goddesses, West Porch, Shyam Ray Temple

136

•

Temple to Love

4. Axes and the Mediation of Worship

The experimentation in architectural form explored in the previous chapter is not limited to the superstructure of Ratna temples. Unlike their Nagara contemporaries, Ratna temples are bi-axial.1 They incorporate a second axis, which is devoted to the lay community’s activities (figure 4.1). This spatial reorganization of the temple is articulated not only in the ground plans but also in terra cotta ornament and ritual practice. The traditional east-west axis serves priestly practices just as it did in the older Nagara type of temples. The new north-south axis celebrates the community’s role in the daily life of the temple. Together they represent the major constituent elements of the Gaudiya Vaishnava community. In addition, a third space emerges within the Ratna temple for the renouncers who seek shelter in the rainy season. I therefore interpret the Ratna temple form as an attempt to negotiate between priests and laity on the one hand and the laity and mendicants on the other. The architectural frame and patronage structure house and hence mediate the activities of these groups. Thus, alongside the variations in upper shrine and mosque-based lower sanctum, Ratna temples create a new worship space for their major constituencies.

Incorporation of a Bhakti Axis While the east-west axis still functions in the traditional Nagara manner, now the primary axis is north-south, with the main entrance of the temple reoriented to the south (figures 4.2, 4.3). The sanctum opens onto a large courtyard suitable for devotees to gather for arati on a daily basis, oriented to the sanctum on the lower level and also to the upper pavilion for festivals. Beyond the courtyard is the hall for entertainment

Axes and the Mediation of Worship

•

137

(natmandir), where musicians perform (figure 4.4). They sit in two rows, facing each other, aligned with the deity images in the sanctum across the courtyard, so that the gods may observe their performance.2 This hall was typically quite large. At Gokulnagar, the covered hall, along with the open deck in front, is larger than the shrine itself (figure 4.5). In the Vishnupur temples, the central area of this hall is flanked by smaller, secondary rooms on either side, where the drums, cymbals, and other instruments for kirtan are stored. The courtyard often includes a basil planter (tulsimancha), sometimes in the shape of a ratha, between the shrine and the hall. With the north-south axis emphasizing the participation of the community, the east-west pathway is now transformed into a service conduit. The stairs from the east door of the sanctum lead down and across the courtyard to the kitchen (bhogamandapa) where the daily meals of the gods are prepared. In Vishnupur’s living temples, such as the Madan Mohan and Madan Gopal, priests use this passage to conduct the daily acts of service for the deity images, leaving the north-south axis for community worship. At Gokulnagar, the region’s earliest extant Ratna temple highlights this axis of priestly service by providing an elevated pathway connecting the shrine’s east platform to the kitchen (figure 4.6). Sometimes this east-west pathway is extended to the compound boundary. Temples retaining their original compound walls, such as the Gokul Chand, Madan Mohan, Lalji, and Radha Shyam, have a secondary doorway located near the kitchen, leading to a pond where the gods’ cooking utensils and clothes are washed. At the culmination of the temple’s two axes in the sanctum are two sets of indented niches that house altars (figures 4.7, 4.8, 4.9). It is almost as if separate sanctums have been created to coordinate the two ritual spheres. The larger size of the recess in the north wall usually indicates that this was the primary altar, while a smaller one in the west wall houses a second altar, as in the Radha Madhav Temple. Several variations developed in the creation of the secondary altar on the west wall. In some temples, such as the Madan Gopal, the west wall has two small niches instead of a single altar niche. Others, such as Gokul Chand and Radha Lalji, have identical deep, elevated platforms for two altars, but the indented niches are unequal in size. The Shyam Ray Temple, the only one at Vishnupur with an ornamented sanctum interior, expresses the relative significance of the two altars in the terra cotta elaboration of these niches (figures 4.10, 4.11). A large circular rasamandala fills the north wall, while a smaller one is depicted in the center of the west wall. On the north, a deep indentation in the wall with an elevated platform provides the architectural frame for a permanent altar. It faces the south entrance,

138

•

Temple to Love

providing the backdrop for the images of Radha and Krishna during kirtan. The absence of a similar niche on the west wall, despite the same visual imagery, suggests the potential for a secondary altar, but one that may have been temporary in this early temple. In order to understand how these dual axis spaces are used, we must turn to later temples to observe current practice. What we find is that these two altars are not entirely separate but complement each other. The Radha Shyam Temple’s sanctum images articulate the complementarity of the dual axes of Gaudiya Vaishnava devotional worship (figures 4.12, 4.13). Devotees approaching the temple from the south confront the golden-yellow images of Chaitanya and Nityananda in the niche on the north wall. Here they dance with their hands raised as if immersed in the performance of kirtan. When approaching from the east side, they see instead the dark-skinned, flute-playing Krishna as Radha Shyam. While both sets of images are central to Gaudiya Vaishnava devotion, the understanding of Chaitanya as an incarnation of Krishna had just emerged in the sixteenth century as his closest followers began to assert his identity with the god. This knowledge distinguishes the Gaudiyas from the other communities of Krishna worshippers emerging in north India at this time. Chaitanya, at the culmination of the north-south axis, was both the leader and, as a divine incarnation, the recipient of the passionate expressions of song and dance in the courtyard when this temple was the center of the local community’s devotion. The north-south orientation of this temple thus articulates the fundamental theological premises and ritual practices of the Gaudiya Vaishnava community while Krishna, placed on the temple’s east-west axis, continues to provide the focus for the daily course of Brahmanical ritual.3 Others, like the Madan Mohan Temple, have a deep arched recess in the west wall, and today the wooden beds of the gods are placed before it (figure 4.7). The deities spend the day in the altar on the north wall of the sanctum and all activity, both priestly and popular, is directed here. At night, they are put to sleep in their bed on the western end of the sanctum, moving from one axis to the other. The deity images are small and particularly easy to transport, as for example during festival celebrations, when the images were formerly taken up to the pavilion above and today are brought to the edge of the plinth, as described in chapter 1. The portability of images and the presence of two altars in a single sanctum would have satisfied the needs of the lay community for kirtan and still allowed the priests’ daily service. The changes in ground plan point to these shifts in ritual practice. The reorganization of the temple to create two complementary axes underscores the importance given to devotional performance and the growing

Axes and the Mediation of Worship

•

139

prominence of the community in the daily life of the Gaudiya temple. Circumambulation, a conventional activity performed by pilgrims visiting Nagara temples, is now replaced by the performance of kirtan in the courtyard in the evening (figures 4.14, 4.15). The shift away from circumambulation has a direct effect on the terra cotta decoration of the Ratna temple’s exterior. Unlike the walls of the Nagara temple, which project sculptural emanations of the central deity in the cardinal directions, the back walls of the Ratna temple do not do so. Instead, the north and west walls of the sanctum display terra cotta simulations of closed wooden doors (figure 4.16). These blind doors are shaped like the arched doors retained by numerous smaller temples throughout the town. The terra cotta imitates the grid of wooden strips and the circular heads of the nails that hold them together. On more elaborate temples, such as the Shyam Ray, floral forms fill these grids as they do on the wooden doors of the Vishnupur Raj palace. These carefully wrought imitation door panels bar entry to the sanctum and hence signal the lesser status of these two directions compared to the east and south, where open doors invite, respectively, priestly ritual and community darshan. This unequal symbolic significance of the four directions points to the shift away from the older practice of circumambulation. In the traditional Nagara form, all four walls are visible to pilgrims, and thus the cardinal directions are weighted equally. Here the terra cotta decoration emphasizes the shift to expressions of bhakti in the courtyard, and viewing of the sanctum images during arati and kirtan rather than circumambulation. The placement and elaboration of porches flanking the sanctum draws further attention to the new north-south axis for community worship. Small temples like the Radha Vinod invariably have a porch on the southern, community side of the sanctum to shelter visitors during the rainy season. In larger temples that have porches flanking the sanctum on all four sides, the south porch is privileged in its treatment. The Shyam Ray Temple’s south porch, for example, distinguishes three bays, unlike the single continuous space of the east porch. Individual vaulted ceilings are highlighted in rings of terra cotta, expressing the significance of the community’s devotion along the south axis. Such treatment gives precedence to the new bhakti axis leading from the hall for entertainment, through the courtyard, to the sanctum.

Performance in the Courtyard Although the courtyards of many of Vishnupur’s temples have recently been transformed into lush gardens by the Archaeological Survey of India, living temples such as the Madan Mohan still use the space for

140

•

Temple to Love

congregational gatherings. The devotional community gathers nightly to observe arati, the priests’ offerings to the gods in the sanctum. This is followed by the collective singing of kirtans in the courtyard, accompanied by musicians seated in the natmandir. The complementary practices of arati and kirtan are dramatic performances that re-enact Krishna’s play and transform the space into the site of the gods’ lila with the devotees as participants in it. The activities of the north-south axis thus express the emphasis on personal, interactive devotion that characterizes bhakti belief and practice throughout South Asia. In fact, the term bhakti is derived from the Sanskrit root bhaj, meaning “to take part in” or “to share,” pointing to the participatory quality of such worship. Singing kirtan is perhaps the most expressive of the performative activities that transform devotees from audience into participants in the play of the gods. The words of these songs evoke the various activities, moods, and feelings of Radha and Krishna, making their passion so palpable through sensuous description that devotees feel they partake of the divine experience. Most songs take Radha’s perspective and describe her desire. The songs are classified into two broad categories: vipralambha, the poignancy of separation, and sambhoga, the divine passion in consummation. In turn, these are further categorized so that separation has four subgroupings: the arousal of desire in each of the lovers by hearing descriptions of and seeing the other; the feeling of offense when Radha’s pride has been injured by Krishna’s neglect or pursuit of other women; the condition of simultaneous satiation and awareness of imminent separation; and the pain of separation created by Krishna’s departure. In the more elaborate kirtan sequences, songs are selected carefully from these groupings to re-create the divine love affair, from the earliest stage of Krishna’s desire for Radha upon first seeing her beautiful, delicate, golden body, through their first shy encounters, to the progression of their passion, Radha’s jealousy at his involvement with other gopis, her anger, his repentance, and the renewal of their longing.4 The performance of kirtan enables devotees to participate in the divine lila in various ways. Chanting the names of Krishna is believed to be so powerful that it brings the god into the temple courtyard and gives the singer the opportunity to enjoy direct contact with him. Vividly describing the divine realm and its activities also helps bring that remote reality into existence in the temple courtyard. Further, the singer can relate intimately to the emotion expressed because this divine experience is modeled on human love affairs. Moreover, the songs are typically in the first person, so the singer in effect takes on Radha’s voice as she or he cries out with intense emotion. Her longing for Krishna is therefore re-enacted within the temple compounds through the words of the songs.5

Axes and the Mediation of Worship

•

141

At Vishnupur’s Madan Mohan and Madan Gopal Temples, the singing is usually led by designated, experienced singers (kirtaniyas) and accompanied by musicians on instruments including the harmonium, drums (khol), and cymbals. For festivals, professional troupes complement the local community’s singers, and they may even alternate with the local singers in leading the performance. As the kirtan progresses, devotees often rise and dance to the music, swaying and throwing up their hands with increasing abandon, much like the terra cotta figures on the temple walls (figures 4.14, 4.15). The performance can induce ecstatic emotional states among worshippers and enable their transition to the world of Krishna’s play in the forests of Vrindavan. At the Madan Mohan Temple I have observed tears roll down the cheeks of passionate worshippers. They weep in Radha’s anguish and pain at being parted from Krishna. In so doing, they also emulate the leaders of the Gaudiya devotional movement, such as Chaitanya, who often moaned, trembled, and fainted from uncontrollable emotion during the performances of kirtan he led in his hometown of Nabadwip. The north-south axis from the sanctum to the hallway across the courtyard is the site of this performance. Devotees must be standing in the courtyard to see the gods of whom and to whom they sing. Their performance resituates the architecture in the space of the eternal lila, transforming brick and terra cotta into the forests of Vrindavan. The profuse vegetal ornamentation that frames the terra cotta panels depicting Krishna in his various activities (figure 4.18) also frames the performers who dance and sing like Krishna and the gopis. Performance thus incorporates the participants in Krishna’s lila and resignifies the architecture as the site of the eternal lila. Along with kirtan, the arati helps re-create the divine lila, reinforcing the bhakti orientation of Ratna temples to their courtyard and northsouth axis. Arati involves a series of offerings to the deities by temple priests. As these offerings were a part of the accoutrements of the original experiences of Krishna and Radha in the groves of Vrindavan, arati renews that primordial performance. Gaudiya Vaishnava temple practice thus co-opts conventional ritualized offerings to express its distinctive worldview. With elegant, sweeping hand movements priests offer lighted oil lamps, water in conch shells, incense, camphor, flowers, and flywhisks to the images of Radha and Krishna at the northern end of the sanctum. The sinuous motions highlight the heads, waists, and feet of the images in the darkness of the temple. These actions are performed exclusively by priests. Priests, however, act with the authority invested in them by the community.6 They also engage the gathering of devotees through their activity.

142

•

Temple to Love

Devotees observe these proceedings from the courtyard, following the flame of the lamp, the visual encounter affording them intimate contact with the divine in the darkness. Devotees then share in these offered items. They touch the flame with their palm to receive its blessing by raising the hand to the head. Priests sometimes sprinkle the water from the conch shell onto the crowd. Alternatively, the offered water is received in the hand, along with the flower petals, and devotees consume the prasada. At some temples, devotees also bring their own oil lamps from their homes, and light them from the temple’s offered flame. The oil lamps are then kept lit under the tulasi plant to illuminate the temple courtyard so that Krishna can observe the kirtan that follows. This aspect of the lamp’s function also draws attention to the interactive quality of such worship. The lights, fragrances, and flowers offered are articles especially pleasing to the divine lovers enthroned in the sanctum. Radha and Krishna had originally offered these things to each other as part of their erotic experiences in the darkness of the woods, as described in verses of kirtan sung in the temple courtyards. Krishna, for example, entices Radha with such offerings: My moon-faced one, I am waiting to make our bed ready, to gather lotus petals— your body will press them . . . Come, the sweet breeze from the sandalwoods censes our trysting place . . .7

Similarly, Radha cries out in the throes of passion in the final verses of the Gitagovinda, calling for such offerings as physical expressions of Krishna’s devotion to her: Fix flowers in shining hair loosened by foreplay, Krishna! Make a flywhisk outshining peacock plumage to be the banner of Love.8

As these offerings are sensory enhancements that make their nightly trysts more pleasurable, they are incorporated into Gaudiya Vaishnava precepts for worship. Arati in this tradition can therefore be understood as an attempt to re-create that experience, both for the gods and for their devotees. It is no coincidence that these activities are conducted at the end of the day, in the darkness of the evening, as Krishna had seduced Radha under the night sky, protected by it. As a re-enactment, the performance of arati and kirtan gives the

Axes and the Mediation of Worship

•

143

devotional community the opportunity to participate in the original activities of the gods. Stressing the commemorative nature of ceremonies, Paul Connerton coined the term “rhetoric of re-enactment” to address their calendrical, verbal, and gestural aspects.9 Through the words and actions of arati and kirtan, archetypal acts and actors, actual or imagined, are re-presented, causing what had disappeared to reappear before the eyes of the devotees. Through ritual performance, the “community is reminded of its identity as represented by and told in a master narrative.”10 The temple, reoriented by the addition of a north-south axis and public courtyard, provides the space and aesthetic for these expressive acts of devotion, effecting religious transformations, changing devotees into gopis and the scene into an enactment of Krishna’s lila.

Terra Cotta Elaboration of the South Façade The decorative program of the Ratna temple is a major departure from that of the older Nagara form in the articulation of the surface and axial emphasis. The organization of terra cotta ornamentation on the exterior walls expresses the changes in architecture and ritual. Simultaneously, it reinforces the relationship between the decoration of Sultanate mosques and shrines and that of Ratna temples. That continuity, seen earlier in their architectural form, reflects their shared purpose of creating congregational spaces. The terra cotta is not distributed equally over the façades of these temples but instead is concentrated on the entry walls, reiterating that all four façades are not of equal visual importance. This marks a rupture from the equivalent treatment of the four walls in Nagara temples, all of whose walls are integral to the experience of the circumambulating devotees. Ornamentation is typically lavished more generously on the Ratna south façade, which acts as a backdrop for the community activities in the courtyard. If we recall that this is also the direction the deity images face to participate in the public celebrations, we can begin to understand the role of the terra cotta decoration as a frame for the images. The Shyam Ray Temple indicates shifts in ornamentation to express the change in ritual practice from circumambulation to kirtan. All four façades are encased in terra cotta imagery, but it is the south façade, where the community gathered, and the east entrance, used by priests for daily ritual, that are most elaborately treated and given dedicatory inscriptions (figures 4.17, 4.18). The dual axes, highlighted in terra cotta, would have visually oriented the new Gaudiya community to the two complementary functions of the new Ratna temple—daily Brahmanical service of the deity and the new distinctively Gaudiya devotional performances. Fifty

144

•

Temple to Love

years later, the north, east, and west walls of the Madan Mohan Temple are unarticulated sheets of brick wall, indicating the completion of the transition toward the framing of deity images for the courtyard (figure 0.10). All later temples consistently prefer to devote ornate terra cotta work to the public south façade. Changes in ritual practice and the new dual orientation of the Ratna temple thus help explain the shift from a temple with all four walls sheathed in terra cotta ornament to a temple with only one adorned façade. The Madan Mohan Temple uses its south doorway as the central visual focus (figure 4.19). Above the triple-arched entrance, a broad band of geometric pattern delineates a tympanum area. The frame also echoes the pronounced curve of the roof. The enclosed field is crowded with martial figures confronting each other with raised bows and arrows. Amidst this crowd of warriors is the dying Bhishma, and this horizontal figure on his bed of arrows allows us to identify the battle as the culminating war of the Mahabharata. The pillars supporting the south entrance are divided into three segments, separated by projecting moldings with lotus bud fringes. Panels with dancers and drummers are packed to fit the column’s width.11 The surfaces on either side and above the arched entrance are organized into a sequence of figurative panels, embedded in vegetal frames and demarcated by pilasters. “Filler” pieces interlock with them in such a way that the entire terra cotta surface is woven tightly. The panels fit each architectural unit snugly, adapting their shapes and sizes to the monument’s curves. While covering and concealing the temple’s hut-like frame, the terra cotta simultaneously highlights the architectural skeleton, like a skin. Horizontal and vertical bands divide the wall frieze into a grid of framed figurative scenes. A decorative veneer is created by the consistent and diminutive size of the figures, and their confinement within elaborate frames. Repetition of a few popular scenes adds to the uniformity of the terra cotta surface.12 This rhythm of the gridded surface is central to the new aesthetic developed in the seventeenth century and is preserved in all later Ratna temples in the region.13 This flat ornamental grid is radically different from the offset wall of the Nagara temple (figures 0.4, 0.15).14 The rhythm of projections and recessions created by the framing grid and figurative panels is distinct from the sequential procession of the Nagara temple’s projections. The organization of the Ratna south façade’s ornament instead draws upon the formula developed for the main entrance of Sultanate shrines and mosques. The elements used to privilege that entrance are transferred onto the south façade of Ratna temples, making the south, and not the east, façade of Ratna temples equivalent to the primary community

Axes and the Mediation of Worship

•

145

entrance of the mosques and tombs. Rectangular panels with motifs such as suspended lamps and purna ghata (auspicious water pots with overflowing plants) in foliate frames are repeated on the Sultanate surfaces, creating a continuous decorative surface (figure 2.17).15 Similarly, the façade of the Radha Ballabh Temple at Krishnanagar, Hooghly district, is decorated with large rectangular patterned frames with pendant nonfigural motifs (figure 0.5). Most Ratna façades retain the elaborate vegetal frames, replacing the inset motifs with figural scenes. In the Vishnupur temples, this is accompanied by a reduction in the size of individual panels to create a more densely packed façade (figures 4.18, 4.19). Their rhythmic composition and the crowding of panels on the brick surface create an intensely busy and energetic effect.16 This parallels developments in the last Sultanate-style mosques built in the seventeenth century, where a similar dense appearance was also preferred over the sparser look of sixteenth-century monuments. The Atiya Jami Masjid and the Motichura Masjid in Rajnagar reduce the size of framed panels and increase their numbers to fill the wall surface in a grid-like design like that of the Vishnupur temples (figures 2.6, 2.9). Vernacular hut construction offers another instance of careful framing on which these façades may have been modeled (figure 4.20).17 The parallel exploration of terra cotta in Ratna temples and contemporary mosques indicates that artisans adopted some fashions with enthusiasm, applying them to the most potent façade of both building-types. It also points to the shared concern with creating successful congregational spaces in mosques and Ratna temples.

Doorways to the Divine Reiterating the connection between mosque and temple decoration, the primary doorway through the Ratna north-south axis into the sanctum is ornamented in the manner of the Sultanate mosque’s mihrab frame. Like the mosque mihrab, this entry is the visual focus of the devotional community using the Ratna temple. It is therefore appropriate that mihrab ornamentation is modified to frame the temple sanctum and the deity images therein. The organization of vegetal ornament remains almost unchanged and equally appropriate in marking the buildings’ most potent parts. The late Sultanate style of mihrab ornamentation shared by Ratna temples and mosques of the sixteenth and seventeenth centuries emerged from the combination of features tested out in the terra cotta decoration of the Tantipara Mosque at Gaur (figure 4.21). The pointed arched openings of that mosque’s multiple entrances are bounded by a rectangular frame. The space between the arch and its outer frame is adorned with

146

•

Temple to Love

vegetal motifs. A water pot (purna kalasha) at the apex of the arch gives rise to a profusion of curling vines with delicate leaves and flowers that fill the arch spandrels. A lotus blossom with many layers of petals appears on either side of the arch. The basic formula developed in the Tantipara Mosque recurs, with minor variations, in the mihrabs of sixteenth-century mosques such as Bagha, Egaroshindur, and Shialghuni in terra cotta, and at Kushumba in stone (figures 4.22, 4.23, 4.24), and it clearly provided a template for the terra cotta doorways of the Vishnupur temples (figures 4.25, 4.26). The stylized lotus, for example, differs in detail on each monument. At Kushumba, the central seedpod projects, and pointed petals spread outward from it. On the Radha Vinod Temple the lotuses are embedded in raised rings that suggest the outer layers of petals, and that resemble the concentric rings of petals used in the mosques at Egaroshindur and Bagha. The Madan Mohan Temple, however, uses rasamandalas to replace the lotus medallions. Two frames contain this profusion of ornament within the arch spandrels and separate the doorway from the rest of the wall. The outer of the two bands usually repeats the curling vine that lines the arch. The inner band consists of alternating square and rectangular panels with full lotus blossoms, purna ghatas, and animal heads (figure 4.27). Short engaged pilasters support these ornamented arches. Each consists of two vertical sections, separated by raised patterned moldings. This organization remains unchanged into the seventeenth century, although the motifs selected for their decoration vary. At Bagha, for example, the upper section is ornamented with hanging garlands of pearls with suspended bells (ghantamala). At Kushumba, these bells are replaced by tassels. At the Radha Vinod Temple, the pearls are eliminated in favor of a larger bell. In the Keshta Ray Temple, a vine fills this space. In the Shyam Ray Temple, dancing figures of Krishna occupy the facets. Thus despite small differences, the stylized lotus and indeed the entire doorway pattern is repeated across the region, regardless of whether the monument is dedicated to Islam or Gaudiya Vaishnavism. The individual motifs that make up the ornamentation of both mihrab frames and temple doorways—the water pot, hanging bell, curling vine, lotus,18 and leonine heads19—can be traced back to Bengali architecture and sculpture of the pre-Sultanate period. The niche-frames of Pala sculptures in effect were transformed creatively into the frames of temple doors and mihrab niches.20 The fertility, auspiciousness, and prosperity signified by these forms remained appropriate when they were reformulated for Sultanate mosques or Gaudiya temples. The transfer of ornamentation reserved for Pala deity images and mosque mihrabs to the Ratna south doorway signals the primacy of this portal as the architec-

Axes and the Mediation of Worship

•

147

tural frame for the altar. In comparison, the eastern doorway is usually unadorned. Like Pala sanctum images and Sultanate mihrabs, the south doorway frames the gaze of the devotee standing in the courtyard or porch seeking a glimpse (darshan) of the enthroned deity. The doorway from the south porch into the sanctum thus focuses the gaze of the devotee, serving as the most important locus for decoration.

Imagery and Axis The subject matter of the terra cotta panels embellishing the walls is coordinated with the new bi-axial organization and the particular roles of each axis. Thus the visual imagery on the south wall, usually framing the image of Chaitanya, typically articulates the central tenets of the Gaudiya community. The terra cotta thus offers the community a point of coherence, and simultaneously distinguishes the Gaudiyas from their rival Vaishnava groups in north India. The panels on the east wall, framing the secondary altar, tend to be more generic and instead invoke the identity of the Gaudiyas as both Vaishnavas and Hindus. Like the common priestly activities performed along this axis, the imagery reiterates the base they share with the other Brahmanical communities of the time. The earliest of the brick and terra cotta monuments at Vishnupur, the Shyam Ray Temple, reveals this organizational logic in the choice of imagery for the decoration of its walls. An oversized panel with a rasamandala dominates each side of the south entrance (figures 4.17, 4.28). Not only are these circular forms visually arresting and strategically situated to draw the viewer’s eye, but they are an embodiment of the distinctive theology and ritual practices that differentiate the Gaudiyas from contemporary Vaishnava groups. At the center of the circular formation, Krishna sways to the sound of his flute with a gopi at each side. Two concentric rings of dancers encircle them. Outstretched arms, bent knees, soaring sashes, and flying hair create angular patterns, and the reduplication of each perfectly poised body completes the bands surrounding the central trio. The dynamic rhythm of their dance is rendered infinite by the circular composition. Drummers and flute players provide the music for their performance from the corners of the square panels. Creepers, deer, and peacocks, drawn by the music, also hover at the periphery. They remind us of the pastoral location of this performance on the wooded banks of the Yamuna River. They also echo the way in which the devotee is drawn in through the terra cotta decoration to worship Krishna-Chaitanya. The rhythm and lyricism of the circular dance is optimized by the rasamandalas’ bold placement beside the main public entrance. Set sym-

148

•

Temple to Love

metrically on either side of the triple arches, these medallions are linked by the projecting arc of the cornice. Here the mandalas serve to visually anchor the temple’s primary dedicatory inscription, which is embedded directly above the central arched entrance. The form of the roundels also echoes the elegant curve of the upper cornice and completes the hemispherical arches of the entrance. These compelling forms are given such a visually significant location because they also epitomize some of the Gaudiya community’s most deeply held convictions. In this tradition the rasalila is revered as the essence of Krishna’s numerous lilas.21 The Bhagavata Purana, a tenthcentury text venerated as divine revelation by the group, offers the first full description of this episode. The tenth book narrates how Krishna lured the women of Vrindavan into leaving their families and homes to join him on the banks of the Yamuna. Here, protected by the woods, they danced the circular rasa dance on a full-moon autumn night. As their fingers interlocked to form a circle, Krishna appeared between each pair of gopis so that he danced simultaneously with each woman, fulfilling her passion: Thoroughly exhausted by (participation in) the ra¯sa dance and with her bangles slipping (from her wrists) and jasmine flowers dropping (from her braid), another Gopı¯ caught hold by her arm, the shoulder of Kr.s.n.a who was standing by . . . another Gopı¯ smelt how Kr.s.n.a’s arm placed on her shoulder, was fragrant like a lily and was anointed with sandal paste, (and losing herself) she (actually) kissed it . . . To another Gopı¯ who rested on Kr.s.n.a’s cheek . . . rocking in the course of dancing Kr.s.n.a gave his half-chewed betel . . . Another Gopı¯, standing by his side was fatigued with singing and dancing, making all the while a jingling sound of her anklets and girdle . . . pressed to her bosom his blissful lotus-hand. Obtaining Acyuta ( . . . Kr.s.n.a), . . . as their beloved, the cowherd women with his arms round their neck extolled him in song and played (danced) with him.22

This description offers a textual precedent for the basic iconographic elements of the roundels depicted on the temple walls. It also provides a model for the Gaudiya community’s most distinctive ritual practice: singing devotional kirtan and dancing to that music. Krishna’s divine play is therefore re-enacted through the visual imagery on the south façade, and in song and performance along the bhakti axis. Like the gopis, Vishnupur’s devotees gather to invoke Krishna and their love for him through song and dance performed in the courtyard in front of the terra cotta rasamandalas. According to the Bhagavata Purana, the gopis sang of their love for Krishna in chorus while each experienced his exclusive attention. This emphasis on collectivity, while affirming the deeply per-

Axes and the Mediation of Worship

•

149

sonal nature of bhakti worship, is invoked in the performances. It is also signaled in the terra cotta representation of the lila, in which each gopi is depicted as distinct yet also part of a pattern. The stress on collectivity in the panels points to the relationship between the ideal community and the living one performing it. The depictions can be understood as sanctioning and providing a paradigm for the emergent community and its attempts to replicate the original event as described in the Bhagavata Purana. In overlooking the temple’s new north-south axis, these two large and prominent medallions reflect and provide a foil for the community’s activities in the courtyard. The Gaudiya Vaishnava reinterpretation of the rasalila in the sixteenth century primarily involves the interjection of the figure of Radha, so vital to their ideal of passionate, selfless love (madhurya), into this circle.23 As Graham Schweig has noted, the invocation of the movement’s paradigmatic devotee figure makes the Bhagavata image of the dance shared by the various Vaishnava sampradayas explicitly Gaudiya.24 The Bhagavata Purana had described Krishna disappearing in the middle of the dance, and the dancers discovered that he had run off with one special gopi: Indeed, he was worshipped perfectly by this maiden (anaya¯ra¯dhitah) He who is Bhagavan, Hari, the supreme Lord. Abandoning us, being pleased with her, Govinda led her to a secret place.25

The Gaudiya community identifies this gopi, chosen by Krishna because she had worshipped him perfectly, as Radha.26 The Gaudiya Vaishnavas further interpolate the figure of Radha at the core of the dance and hence the center of the depicted circle. If we look closely at the two figures flanking Krishna in the middle of the terra cotta mandalas, they are in fact differentiated by their poses. To the right is an adoring devotee, while the figure on the left echoes the pose of the metal icons of Radha worshipped in the sanctum. She turns to gaze up at Krishna, raising one hand, perhaps offering a lotus bud as the iconic versions do. Such individuation of Radha at the hub of the dance formation is not prefigured in the Bhagavata Purana. Rather, the Gaudiya Vaishnavas give authority for this interpretation to Chaitanya himself. In his biographical work Caitanya Carita¯mr.ta, Krishnadas Kaviraj suggests that the revelation came to Chaitanya in a dream: One day Maha¯prabhu [Chaitanya] was lying down, and he saw in a dream Kr.s.n.a at the ra¯sa-lı¯la. He was beautiful in the tribhan.ga pose, his flute to his lips, wearing yellow garments and garlands of forest flowers, the charmer of Madana. The gopı¯s were dancing around him in a circle,

150

•

Temple to Love

and in the center Vrajendranandana [Krishna] danced with Ra¯dha¯. Seeing this, Prabhu [Chaitanya] became absorbed in this rasa, and he realized “I have gained Kr.s.n.a in Vr.nda¯vana.”27

This account further suggests that Chaitanya was a participant in the dance as he awoke to the realization that he had attained union with the divine. Like Radha, Chaitanya had worshipped Krishna with perfect devotion, and in fact he was identified with Radha, the favored gopi, by this time. Like the golden-skinned Radha, he was called Gauranga (goldenlimbed). The monumental medallions thus also point to Chaitanya, who is believed by this community to be the incarnation of both Krishna and Radha, as well as Krishna’s ideal devotee, and who is housed in iconic form in the south altar inside. The iconography of the rasamandala was probably employed to visually herald this concept of the dual incarnation, a radical proposition that distinguished the Gaudiyas from all their rival Vaishnava groups. The interjection of the familiar figure of their leader and the object of their worship would have made the Bhagavata description of the rasalila more immediate and accessible to the Bengali community. The Caitanya Carita¯mr.ta elaborates on the vision of Radha and Krishna at the core of the dance through a conversation between Chaitanya and Ramananda Raya, the minister of the Orissan king who had turned devotee. In discussing the nature of Radha’s love, Raya observes, “The bodies of the sakhı¯s—Lalita¯ and the rest—are simply extensions of her body.”28 Here the gopis are interpreted as emanations from Radha’s body. Raya describes her body further as draped with various emotional states including delight, humility, anger, madness, anxiety, inconsistency, dejection, restraint: “her body is covered with the ornamentation of all these bha¯vas.”29 The eight gopis forming the inner ring of the rasamandala can therefore be understood as personifications of Radha’s emotional states and body parts in the form of her eight closest female companions. Each dancing figure in the terra cotta mandala is a fraction of Radha. In dancing with every one of them, according to the Gaudiya understanding of the rasalila, Krishna embraces and savors each of her emotions and their bodily expressions that combine to embody her allconsuming and selfless love for him. Fundamental to the Gaudiya tradition is the belief that Krishna came down to earth in the form of Chaitanya to satisfy his desire to experience this love. As radial expressions of Radha’s qualities the medallions are an elucidation of the ideal devotee, the model for the gathered community to emulate in their activities, and they provide an explanation for the incarnation of Krishna as Chaitanya. The Caitanya Carita¯mr.ta therefore expresses the particular Gaudiya beliefs infused into the rasalila described in the Bhagavata Purana.

Axes and the Mediation of Worship

•

151

The theologians in Vrindavan had sanctioned this work as the authoritative version of Chaitanya’s life. Its understanding of Chaitanya as Radha, as part of the dual incarnation, was disseminated by the next generation of leaders, such as Srinivas, who brought this message to Vishnupur. Many versions of the text were copied and circulated at Vishnupur over the seventeenth and eighteenth centuries. The Malla ruler Gopal Singh, for example, personally created a copy that survives in the collection of the Vishnupur Sahitya Parishad. We can therefore assume that its interpretation of the rasalila was available to devotees who gathered in their neighborhood temples, through exegesis if not through direct encounter with the work. The bold and dynamic rasamandalas depicted on the Shyam Ray Temple are the earliest surviving representations of this subject in an architectural setting in Bengal. The basic composition likely entered Bengal as part of the pictures accompanying texts such as the Bhagavata Purana and the Caitanya Carita¯mr.ta. One of the earliest painted depictions of the rasalila occurs in a Bhagavata Purana manuscript from the first half of the sixteenth century, from Palam, not far from the Vaishnava center of Vrindavan.30 Such texts were likely copied and illustrated under the patronage of the growing Vaishnava movement in north India and disseminated through Gaudiya territories. Pivotal to Vishnupur’s foundation narrative is Srinivas Acharya’s arrival with a cartload of books, and the visual imagery may have arrived along this route with the texts. Located on the public south façade of Vishnupur’s earliest terra cotta temple, the rasamandalas provide a daytime counterpart to the nighttime singing and dancing performed as part of daily kirtan, when the darkness obscures the panels. For the devotee standing on the temple’s plinth or in the courtyard during the day, the rasamandala would likely have served as a mnemonic device to trigger recollection of the original divine dance. Radha’s reincarnation through Chaitanya’s golden-hued body, and his subsequent embodiment in the form of Srinivas, would have provided an immediate model for the devotee’s personal transformation into that ideal worshipper. As visual mechanisms the images would surely have prompted the memory to associate these layers of past action with present performance. We might recall that remembrance was a central component of worship in this devotional tradition. The tradition’s major texts reiterate the role of memory in attaining union with the divine. In the Bhagavata Purana Krishna explains his desertion of the dancing gopis as a lesson in overcoming separation through the power of recollection: My disappearance is only a veil that I draw in order to nourish yearning and love for me by the fire of separation . . . I wish that my devotee should not for a moment forget me.31

152

•

Temple to Love

Later, in the Gitagovinda, Jayadeva invited his audience to share in his reminiscence of Radha and Krishna’s secret tryst: If remembering Hari [Krishna] enriches your heart, If his arts of seduction arouse you, Listen to Jayadeva’s speech In these sweet soft lyrical songs.32

Like his songs, which are performed in these temples as part of kirtan, the terra cotta rasamandalas provide links to that ideal past. If Walter Ong’s premise that repeated, rhythmic, and balanced patterns were widely used for aiding recollection in predominantly oral cultures holds, then the rasalila with its repeated figures in concentric circles would have stimulated remembrance, as did the additive quality of repeated phrases in the songs.33 In Bengal, diagrams emphasizing symmetrical and repetitive geometric forms such as concentric circles have a long history in Shakta and Buddhist mandalas. For tantric practitioners the forms typically represent the self or the body as well as the cosmos. Visually tracing the lines of the mandala with the goal of creating a mental image of the cosmos is believed to aid concentration and stimulate meditation upon the relation of the self and the macrocosm. The presence of mandala-like Pala-style deity sculptures in the Vishnupur Sahitya Parishad would indicate that many in the Vishnupur area were familiar with such visual implements for yogic meditation. The rasamandala would have offered them an opportunity to transfer these visual skills to the contemplation of the new cosmology of Krishna’s eternal realm and their role in it. Below these roundels are the rows of panels depicting the distinctive two-storied temple form discussed in chapter 1 (figure 1.10). The south wall is therefore devoted to the distinctive beliefs, rituals, and architectural form of this community. If we turn to the east wall of the Shyam Ray Temple, smaller versions of the rasamandala alternate with the doublestoried temple form to create a stringcourse that divides the wall into upper and lower segments (figure 4.19). The difference in the size of the same motifs on the two façades points to their varying importance for the two different constituencies of the two axes. On the south wall, the monumental versions likely encouraged devotees to reflect upon the Gaudiya cosmology and the part that they played in it; on the east, this same Gaudiya belief system is situated within the larger Hindu pantheon that is the subject matter of these walls. Below the rasamandala stringcourse on the east wall, at eye level, are the avatars of Vishnu interspersed with other Puranic deities. Left of the entrance is the tortoise incarnation (Kurma), four-armed Vishnu riding Garuda, and Kalki on a horse. In the middle row is Indra riding an elephant, Shiva on his bull Nandi, and a

Axes and the Mediation of Worship

•

153

figure riding a deer that is most probably Chandra, one of the nine planetary deities. Below is the four-headed Brahma on his goose, Vishnu riding Garuda, and Shiva on his bull. Such a grid reflects the interweaving of different narratives so characteristic of the Puranas. To the right of the doorway are various Vaishnava deities. Uppermost is the Jagannatha triad: Balarama holding a plow, Subhadra with a pair of lotus buds, and Jagannatha. In the middle tier is Vishnu bearing his mace, disc, conch, and lotus, with a female consort on the panels on either side. Lakshmi bears a lotus in each hand, while Saraswati has an attenuated veena. The bottom three panels each display Krishna playing his flute. Here he is the incarnation of Vishnu, one of several embodied forms Vishnu takes to protect his followers and preserve the welfare of the world. Krishna, who dominates the south wall, is therefore situated in the larger Brahmanical world on the east façade. Here he is appropriately located at the bottom, as one of the younger deities, and his role is much smaller than it is on the south façade.

Stories on the South Façade The only other temple at Vishnupur with all four façades adorned in terra cotta attempted a different strategy to privilege its south façade and bhakti axis over that of orthodox priestly worship. Rather than employing the compelling circular form of the rasamandala, as the Shyam Ray had done, the Keshta Ray Temple devoted its south wall to a narration of Krishna’s primary exploits (figure 4.29). Rising up the length of the wall is a progression of paired episodes in Krishna’s life, from his early childhood to the years of his adolescence, that are the central concern of Gaudiya Vaishnava belief and practice.34 At the bottom are his confrontations with Putana and Trinivarta. Next the kicking of the cart (Sakatasura) is contrasted with the peaceful scene of the nursing of the baby boys. The third tier pairs the beloved scene of the stealing of butter with the powerful figure of the boy pulling the beak of Bakasura open. Above this, the crawling baby Krishna uproots the trees to release Nalakuvera and Manigriva (Jamalarjuna), and on the right he takes on the snake Aghasura. He then encounters the demon bull Arista and the ass Dhenukasura. The drinking of the forest fire is represented by one of several flames entering his mouth. Pralambha and Shankhachura are featured next, followed by the gopis, who are pursued by Shankhachura. Krishna’s departure for Mathura, the killing of Keshi, and the death of Kamsa are depicted in turn. Uppermost of the surviving panels is the vastraharana. Under the spreading branches of a leafy tree, Krishna and Balarama proceed toward the gopis whose clothes he has stolen. Panels depicting Krishna’s dalliance with the gopis dominate the eastern edge of the south 154

•

Temple to Love

wall. These include Krishna playing his flute atop a tree while the gopis in the stream attempt to cover themselves (vastraharana), taking the cows to pasture (gosthalila), and accompanying the gopis on a boat ride on the rain-swollen waters of the Yamuna (naukavilas). This gridded decorative surface, a bamboo frame hung with narrative scenes, must have acted as a launching pad for storytelling. Each scene represents a climactic moment in a particular episode of a long narrative. When a sequence of panels is organized vertically, as on the south wall of the Keshta Ray Temple, the formal similarity to the long painted scrolls of Bengal (jorano pata), with their enframed segments depicting Puranic episodes, is striking. These molded narratives were probably painted in vivid colors like the scrolls. The terra cotta sequences are still used by community elders and local tour guides as visual aids to storytelling. Such use of the images is not unlike the unrolling of scrolls by itinerant minstrels for a performance under a tree or in a domestic compound. The visual narratives also take on particular significance along the axis of bhakti devotion.35 The sheets of terra cotta panels must have functioned as visual counterparts to the songs sung in the courtyard. A typical sequence of kirtan sung in the Madan Mohan Temple begins with episodes of Krishna’s childhood, for example, describing the boy frolicking through his foster-mother’s house, while the village women flock to her courtyard, mesmerized by his good looks and grace. Typically, Radha is then introduced as golden-skinned, delicate, languorous of movement, and desired by Krishna. Their first shy meetings are described, followed by the messages passed back and forth by gopis, the disapproval of Radha’s older kinswomen, leading to her jealousy and rage at Krishna’s diversions with other women, his repentance, her longing. The evening culminates in songs of Krishna’s departure and Radha’s anguish at their separation, which serve also to express the singing devotee’s experience. Such performances are initiated at night when the imagery is invisible in the darkness. The rhythm of the terra cotta may well have contributed to the emotional experience of engaging with the deity during the daytime and stimulating the remembrances that are so vital to this tradition. It is thus no coincidence that these visual sequences are laid out on the south façade, before the courtyard where devotees gathered.

A Third Axis Ratna temples indicate experiments in their spatial organization and visual imagery to host mendicants and recluses (bairagi). These spaces are typically small and relegated to the back of the monuments, reflecting their smaller numbers in the Gaudiya Vaishnava community and their lesser role compared to the lineages of goswamis, who are householders Axes and the Mediation of Worship

•

155

and heirs to the original founders of the tradition.36 A closer look at the way the porches around the Ratna sanctum are laid out suggests that accommodating these wandering ascetic members of the Vaishnava community became a concern over time. The earliest temples, the Gokul Chand at Gokulnagar and the Shyam Ray and Kala Chand at Vishnupur, do not provide a collective space for bairagi (figures 1.5, 4.8). Around the square sanctum are open porches. At the corners where porches meet are small rooms, and the presence of door frames suggest that wooden doors would have provided secure storage space. Later temples, however, remove the doors that separate the corner rooms to convert the entire length of the porches into covered rooms, provided with light and drainage. The screening of porches and manipulation of the corner rooms created a viable space for the changing needs of the monuments without disturbing the fundamental bi-axial organization of the sanctum and courtyard (figure 4.9). The Keshta Ray Temple is the first to block off its porches at the center of the east and west walls. This creates two covered porches (figure 2.13). The one along the south wall, extending east and west, probably housed pilgrims. The porch on the north end, also extending west and to the east entrance, is accessible only from the narrow single-arched east door. The relatively unceremonious treatment of this doorway, along with its size, suggests that the east entrance to the porch provided a more restricted entry for smaller numbers of people than the prominent public entrance on the south side. Entering through this east door one encounters an image of Sharabhuja, the six-armed form of Krishna. This terra cotta relief is molded in a niche indented in the north porch, perhaps forming the focus of more private worship that revealed this deity’s more esoteric meanings to an exclusive group. Along with representing the flute-playing Krishna and the armed Rama, Sharabhuja displays the waterpot and rudraksha beads of a mendicant. In this tradition the paradigmatic ascetic was Chaitanya, who retired from his home and family to embrace the divine at the Jagannatha Temple at Puri. This mendicant aspect of Sharabhuja would have provided a role model, helping subsequent generations of ascetics to contemplate the nature of Chaitanya’s quest in the privacy of the quiet area at the back of the temple. Curtains suspended from the iron rungs above would have screened this space from the sanctum. Terra cotta jali (screens) provided light, and spouts to the exterior drained ritual fluids. Rows of seated figures with long matted locks framing the east entrance demarcate the interior as a space for ascetics. The north porch of the abandoned Keshav Lalji Temple in the village of Patpur, adjoining Vishnupur, is similarly disconnected from the others. Instead, a doorway in the center of the north wall provides a separate

156

•

Temple to Love

third entrance through the back of the building and also a source of light for the long covered porch.37 Likewise, the Madan Mohan and Madan Gopal, Vishnupur’s two living temples, have a dark, secluded space at the back of the temple, protected from the public’s gaze (figure 4.7). The north porches are blocked off in the northwestern corner and entry is provided only from the east priestly doorway. The Madan Mohan Temple’s priest, Sushanta Mukhopadhyay, indicated that this private porch was used for yogic meditation during the monsoons.38 By the eighteenth century, the Radha Madhav Temple extends this space for ascetics to include the west porch, suggesting that their numbers had increased over the century (figure 4.9).39 Such adjustments in the Ratna temple’s ground plan may represent attempts to incorporate and perhaps even contain yogic and ascetic practices by mainstream institutions, a concern that surely increased over the seventeenth century as the tantric dimensions of Gaudiya Vaishnavism developed. In its distribution, refinement, and iconographic coordination, the terra cotta decoration of these temples participates in the new spatial organization of the temple compound. The terra cotta ornament marks the north-south axis as primary by employing the motifs and organization of Sultanate mosque and shrine decoration. This assimilation of an additional north-south axis of worship must be understood as a response to the shared preoccupation in Islam and Gaudiya Vaishnavism with the community’s participation in the ritual activities of the temple. Like the rituals of the upper pavilion discussed in the first chapter, this spatial arrangement accomplished its purpose while maintaining the traditional role of priests as caretakers of the body of god. The availability of a space for public gatherings, in turn, must have made it easier to develop an elaborate ritual cycle involving the community and the compound. The Ratna temple’s new ground plan and ornamentation also express the third axis, whereby the Gaudiya Vaishnava community attempted to accommodate the practices of yogis within these temples. Temple architecture, decoration, and ritual space therefore provide a way to negotiate the relationships among the various elements composing the Vaishnava community in the seventeenth century.

Axes and the Mediation of Worship

•

157

Figure 4.1. Worshippers Gathered in the Courtyard, North-South Axis, Madan Gopal Temple, Vishnupur

FACING PAGE: Figure 4.2. Plan of Madan Mohan Temple Compound, Vishnupur

158

•

Temple to Love

Axes and the Mediation of Worship

•

159

Figure 4.3. Plan of Gokul Chand Temple Compound, Gokulnagar

160

•

Temple to Love

(TOP) Figure 4.4. Natmandir, Madan Mohan Temple, Vishnupur (BOTTOM) Figure 4.5. Natmandir, Gokul Chand Temple, Gokulnagar

Axes and the Mediation of Worship

•

161

Figure 4.6. Priestly Activities in the Kitchen, East-West Axis, Gokul Chand Temple, Gokulnagar

FACING PAGE: Figure 4.7. Plan of Madan Mohan Temple, Vishnupur

162

•

Temple to Love

Axes and the Mediation of Worship

•

163

Figure 4.8. Plan of Gokul Chand Temple, Gokulnagar

164

•

Temple to Love

Figure 4.9. Plan of Radha Madhav Temple, Vishnupur

Axes and the Mediation of Worship

•

165

Figure 4.10. North Sanctum Wall, Shyam Ray Temple, Vishnupur

166

•

Temple to Love

Figure 4.11. West Sanctum Wall, Shyam Ray Temple, Vishnupur

Axes and the Mediation of Worship

•

167

Figure 4.12. Images of Chaitanya and Nityananda, Altar on North-South Axis, Radha Shyam Temple, Vishnupur

FACING PAGE: Figure 4.13. Image of Krishna as Radha Shyam, Altar on East-West Axis, Radha Shyam Temple, Vishnupur

168

•

Temple to Love

Axes and the Mediation of Worship

•

169

Figure 4.14. Kirtan Performance, South Façade, Madan Mohan Temple

170

•

Temple to Love

Figure 4.15. Drummers, South Façade, Madan Mohan Temple

Axes and the Mediation of Worship

•

171

Figure 4.16. Blind Doorway, North Porch, Shyam Ray Temple, Vishnupur

172

•

Temple to Love

Figure 4.17. South Façade, Lower Story, Shyam Ray Temple, Vishnupur

Axes and the Mediation of Worship

•

173

Figure 4.18. Wall Frieze, East Façade, Shyam Ray Temple, Vishnupur

174

•

Temple to Love

Figure 4.19. Culmination of the Mahabharata, South Entrance, Madan Mohan Temple

Axes and the Mediation of Worship

•

175

Figure 4.20. Bamboo Frame of Hut, Outskirts of Vishnupur

176

•

Temple to Love

Figure 4.21. Mihrab, Tantipara Masjid, Gaur

Axes and the Mediation of Worship

•

177

Figure 4.22. Mihrab, Jami Masjid, Bagha

178

•

Temple to Love

Figure 4.23. Mihrab, Jami Masjid, Kushumba

Axes and the Mediation of Worship

•

179

Figure 4.24. Central Mihrab, Sadi’s Mosque, Egaroshindur

FACING PAGE:

(TOP) Figure 4.25. South Doorway into Sanctum, Radha Vinod Temple, Vishnupur (BOTTOM) Figure 4.26. South Doorway into Sanctum, Madan Mohan Temple, Vishnupur

180

•

Temple to Love

Axes and the Mediation of Worship

•

181

Figure 4.27. Animal-Headed Motif, Madan Mohan Temple, Vishnupur

182

•

Temple to Love

Figure 4.28. Rasamandala, South Façade, Shyam Ray Temple, Vishnupur

Axes and the Mediation of Worship

•

183

Figure 4.29. Krishnalila Panels, South Façade, Keshta Ray Temple, Vishnupur

184

•

Temple to Love

Epilogue: A New Sacred Center

Vishnupur’s architectural formation, including the Ratna temples and their placement in the new urban organization, was legitimized by the mapping of Vrindavan onto the younger town. Distinctive physical qualities of this older holy site were selectively reproduced, transforming the Malla capital into “Gupta Vrindavan,” that is, a hidden Vrindavan. Krishna’s play in this new Vrindavan was visible to the devotional community this program shaped. The extant material culture and forms of devotional expression also suggest that the consolidation of this first Gaudiya Vaishnava foothold involved careful negotiation in the heterogeneous political culture of seventeenth-century Bengal. Uncovering the layering of Vrindavan’s distinctive cultural forms and institutions upon Vishnupur reveals that this process followed immediately from their initial establishment at Vrindavan. This strategy not only linked but also legitimated both sites, reinforcing the mythic status mapped historically upon Vrindavan and giving Vishnupur the authority the Gaudiya Vaishnavas wished it to have. A variety of mechanisms were employed in overlaying a hidden Vrindavan upon Vishnupur. The process was enabled partly by the mapping of the local deity Madan Mohan to Krishna, the universal and ultimate object of Gaudiya Vaishnava devotion. Royal temple construction supporting worship of newly installed deities brought a new community together. The forested terrain of Vishnupur was transformed into a cultural center by launching expressive forms and institutions comparable to those established over the previous century at Vrindavan. A dominant mode for linking the two sites was the creation of narratives about the town, its deities, and its distinctive features, and the composition of devotional poetry by the rulers and courtiers. Landscaping of the

Epilogue

•

185

immediate surroundings of Vishnupur revealed the sacred topography of this new Vrindavan. The dominant telling of this remapping of Vrindavan illuminates the strategizing necessary to establish the religious community. When Srinivas Acharya was sent to Bengal to refocus the movement there, he was provided with a cartload of newly written manuscripts containing the vital precepts of the tradition to help him accomplish this task. At Vishnupur Srinivas lost these precious manuscripts in his charge. Tracing their theft to the local landholder, Raja Vir Hambir of the Malla dynasty, he visited the Malla court at Vishnupur and swept the court with his insightful recitation and interpretation of episodes from Krishna’s life. The raja was so entranced by Srinivas’s passion for Krishna that he fell at his feet and confessed to organizing a theft of the books he had mistaken for worldly riches. Through the raja’s desire to make amends, Srinivas gained his first and possibly most powerful devotee.1 The significance of this alliance reverberated rapidly through Mallabhum in the southwestern corner of Bengal. Seventeenth-century Gaudiya Vaishnava texts provide details that illuminate our understanding of the replication of Vrindavan at this site. According to the Bhaktiratnakar, at the moment of his initiation Raja Vir Hambir entreated his guru to stay longer at his court: “Without Prabhu Vishnupur would be nothing more than the forest.”2 The emerging sacred center, that is, would lapse back to its former state of undifferentiated jungle. As the term prabhu (lord), used to address Srinivas in the original, was also applied to Chaitanya by his followers, and Krishna was the ultimate lord, it is the absence of not just Srinivas but also Chaitanya and Krishna that he feared would threaten his territory. Gaudiya literature thus claims that its adherents urbanized and civilized Vishnupur from the wilderness and transmuted the rude ways of its book-thieving raja into spiritual energy. This account, percolated through text and oral narrative, indicates the mechanisms employed to map Vrindavan’s sacred territory upon Vishnupur. While unabashedly propagandistic, the narrative underscores the reality of collaboration with the local political authority for the success of the venture. This partnership assured that the kings would support the process of transforming Vishnupur into a sacred center. The patronage of oral poetry is foreshadowed in the critical role the tale gives to Srinivas’s recitation, which softens the hardened ruler. The importance of landscaping and temple building is anticipated in the jungle-clearing implied by the imminent danger that the newly constituted Vrindavan will relapse and revert to wilderness. The narrative highlights the significance accorded to both royal support and Vishnupur’s natural surroundings.

186

•

Temple to Love

The Recovery of Vrindavan The original sacred site of Vrindavan had itself been recovered over the previous century from undifferentiated forests. Gaudiya Vaishnava literature from the sixteenth and seventeenth centuries gives Chaitanya responsibility for rescuing Vrindavan from the state of wilderness into which it had degenerated. The Bhagavata Purana describes how Krishna had enjoyed the company of local women in Vrindavan’s forests: Like unto a young elephant exuding temporal fluid sauntering in the company of female elephants, the Lord, surrounded by a swarm of black-bees and a bevy of young ladies, roved in a grove on the bank of the Yamuna, where blew on all sides a gentle breeze bearing the fragrance of aquatic flowers as well as of the flowers on lands.3

As this divine play is believed to be re-enacted perpetually in these groves, it offers a taste of the pleasures that can ultimately be enjoyed by the devotee in the afterlife. This text describes how the sites of his divine lilas had become celebrated during his own lifetime, by pilgrimages through the forests of Vrindavan undertaken by his companions and famous sages while he resided there. Local tradition attributes the loss or disappearance of these sacred spots to Mahmud of Ghazni’s onslaughts of the eleventh century.4 Gaudiya Vaishnava texts describe how Chaitanya unveiled this now invisible site of Krishna’s pleasures to the devotional community once again.5 Specifically, they ascribe agency to the saint’s visit to Vrindavan in 1514, when he mobilized the reclamation of the land of Krishna’s biography, a restoration process that continued to the end of that century. The Caitanya Carita¯mr.ta claims that when the saint reached Vrindavan he entered an ecstatic state while wandering through its forests and bathing in its ponds.6 Through meditation, dreams, and delirious fits, he had visions of the divine couple and identified the sites of their erotic encounters in what was then dense, sparsely inhabited forest. Since, according to the text, Chaitanya was a dual incarnation of Krishna and his beloved Radha, the divine lovers were able to enjoy their lilas once more when the saint meandered through the forests. Chaitanya’s wanderings through the woods thus re-sited these events on the topography of the region.7 The imagined locale of an original narrative documented in the Bhagavata Purana, subsequently transformed into an idealized sacred cartography in the form of a meditative mandala, was now located geographically. As Krishna’s life was mapped on the ground, the site was built up for pilgrimages such as the annual Bana Yatra.8 The cultural construction of Vrindavan, which David Haberman called “externaliza-

Epilogue

•

187

tion” to draw attention to the process of revealing the hidden, ensued when the six goswamis recovered deity images, constructed temples, and institutionalized ritual worship.9 They were also responsible for memorializing the natural surroundings of Krishna’s play; for instance, they paved the embankments lining the sacred ponds where Radha and Krishna had bathed and refaced the steps on the riverside. This project was initiated in 1517, with the arrival of the first of the goswamis, Rupa, to these jungles. He is credited with recovering the image of Krishna as Govindadeva from under the ground at a spot that was revealed to him in a vision. The location was believed to be the very spot where Radha met her beloved for their nightly trysts.10 Here Rupa constructed the first temple in 1533. The goswamis’ project was made possible by imperial and subimperial financial sponsorship and personal interest. The present temple of Govindadeva was completed in 1590, under the patronage of Raja Man Singh, one of Akbar’s highest officials (figure 5.1). This Rajput dynasty, among the highest Hindu rulers of north India at this time, played a vital role in the construction of Vrindavan’s most important temple. As Catherine Asher has argued, the temple’s alliance with Mughal style expresses the political allegiance of its Hindu patrons to the emperor.11 The site also received significant imperial patronage.12 As early as 1565 the temple’s priest obtained a land grant from Akbar to collect revenue at this site. A second imperial farman of 1568 recognized Jiva Goswami as keeper of the temples and permitted him to claim all offerings and donations. The emperor, according to local lore, himself visited the site in 1573 and, moved by a vision in one of the sacred groves, issued land grants for the maintenance of the major temples built at this time.13 His biographer Abul Fazl lists thirty-five major temples receiving such land grants. This imperial development of Vrindavan in the sixteenth century gave physical expression to the eternal realm of Krishna, the ultimate object of bhakti.14 It is not surprising, therefore, that the buildings that defined the environment of this sacred Hindu center bear the imprint of Mughal architecture.15 The series of five early temples (Madan Mohan, Gopinath, Radha Ballabh, Jugal Kishor, and Govindadeva) from the closing years of the sixteenth century flaunt the red sandstone that unmistakably echoes its immediately precedent use at Akbar’s capital at Fatehpur Sikri. Not only are construction techniques used conspicuously in Sultanate and Mughal architecture, such as vaults and domes, employed here, but features of early Indic architecture made popular at Fatehpur Sikri, including the lotus bud fringes of the four-pointed arches, elaborate brackets for columns, and deep chajja (projecting eaves), also proliferate.16 The architectural construction of Vrindavan, contingent upon the

188

•

Temple to Love

political realities of sixteenth-century north India, therefore looks to Mughal fashion. Krishna’s residence was modeled in part on that of the emperor. And it must be recalled that the political rituals at Fatehpur Sikri drew liberally upon recognized forms and customs honoring Hindu deities, such as the daily darshan given by Akbar to establish his authority in a majority Hindu realm.17 Drawing upon imperial Mughal resources would consequently not have been an inappropriate way of marking the bowers of Krishna’s eternal realm and making them visible. Imperial and subimperial patronage also highlighted the distinctive geographical elements of Vrindavan as pilgrimage destinations. While they gave permanent form to Chaitanya’s visions, as with the Vrindavan temples, acts such as paving embankments of ponds and rivers and defining bowers can be understood in conjunction with other imperial Mughal gardening projects. The emperors were famous for their passion for verdant gardens with flowing water as frames for their palaces and tombs.18 Babur, the founder of the dynasty, laid the first of these TimuridPersian gardens, subdivided into quarters by walkways and canals, on the banks of the Yamuna at Agra, downstream from Vrindavan.19 The foundation for the development of Agra as a “riverbank” city, with a succession of such gardens on both banks of the river, was thus concurrent with the project at Vrindavan. While Akbar’s palace complex at Fatehpur Sikri was laid with tanks and waterways, Jahangir is perhaps best known for the magnificent gardens of Shalimar, Nishat Bagh, Achabal, and Vernag with their pools, pavilions, shade trees, flowers, and shrubs. Although the landscaping of Vrindavan does not conform explicitly to Mughal style, its ponds, pavilions, temples, and groves of flowering trees retracing Krishna’s love play were developed contemporaneously with these secular pleasure gardens for the enjoyment of earthly rulers. Both paradises, eternal and temporal, Islamic and Vaishnava, share the same imperial sources of patronage. By the time this project was accomplished at Vrindavan, the Goswami leadership had dispatched Srinivas on his mission back to Bengal to consolidate Gaudiya Vaishnava territory there. The process of reconstructing Vishnupur as a hidden Vrindavan at the margins of the empire thus began immediately after the earlier generation of leadership made the true Vrindavan itself visible in the Mughal heartland and with direct Mughal patronage.

Temples for Vishnupur’s Deities Temple construction and establishment of deity images shaped Vishnupur’s landscape to the contours of Krishna’s divine realm. These temples

Epilogue

•

189

are an original architectural form, distinct from the imperial Mughal style of Vrindavan. Like Vrindavan’s, however, their style is contingent upon the political dimensions of their patronage. The temples are a carefully coordinated pastiche of local architectural styles that express the multiple political agendas of their patrons. Built upon a Sultanate base, they declare their patrons’ participation in the Indo-Islamic order that provided the economic basis for urbanization at both Vrindavan and Vishnupur. That Sultanate lower story is sheathed with narratives of Krishna’s exploits in Vrindavan. The terra cotta figural panels give visual expression to the core beliefs and myths undergirding the new Gaudiya Vaishnava tradition while their floral frames link the Gaudiya Vaishnava sacred imagery to that of mosque mihrabs. The terra cotta program highlights the temples’ dual axes, which celebrate the highly expressive devotional style of the Gaudiya community in the courtyard while accommodating the priest-mediated practices of a traditional Hindu temple. Upon that consistent and recognizable Sultanate lower level is placed a pavilion of variable form for festivals that re-enact Krishna’s erotic encounters with his beloved. The popularity of the traditional Nagara temple for this pavilion articulates the shared base of the Gaudiya Vaishnavas with the older Hindu communities that patronized the older temple form. It also expresses the origins of the Malla dynasty’s authority in goddess worship prior to the Vaishnava collaboration. The new architectural form proclaimed in the inscriptions and described in the terra cotta thus articulates the foundations of Malla power in southwestern Bengal while it houses the new ritual practices of the Gaudiya Vaishnava community that re-create the original performances in Vrindavan’s groves. The proliferation of temples probably gave Vishnupur the reputation as a religious center that is attested in British site surveys and administrative reports from the mid-eighteenth century. Over a century and a half, Vishnupur had become so abundant in temples that in 1765, the local administrator, J. Z. Holwell, observed, “There are, in this precinct, no less than three hundred and sixty considerable Pagodas or places of worship, erected by the Rajah and his ancestors.” Similarly, Hesilrige, in 1789, commented, “In the whole district there are four hundred and fifty three Hindoo temples. The principal one in the town of Bissenpore formerly contained the idol of Mudun Mohun.”20 Today this profusion of monuments is offered by its residents as an explanation for its renown as Gupta Vrindavan. The deities installed in the Ratna temples were named in accordance with those of Vrindavan. The most notable example is Madan Mohan, one of the earliest enshrined deities in Vrindavan, who became the primary deity of Vishnupur. Temples were also dedicated to Shyam Ray,

190

•

Temple to Love

Radha Raman, Madan Gopal, and other deities established previously at Vrindavan. The stories of the discovery and installation of some of these images are analogous to those of Vrindavan. They share narratives of miraculous revelation comparable to Rupa Goswami’s vision of the cow spontaneously pouring milk and revealing the buried image of Govindadeva. Others are related to Vrindavan by contiguity. The daughter of Srinivas, Hemlata Devi, who successfully led the religious community in Vishnupur after his demise, went to Vrindavan and chose as her personal deity Radha Raman, from the selection offered by Jiva Goswami, after she was informed in a dream that she would be mother to Krishna exclusively and no mortal son.21 The connections between the generations of leadership at Vishnupur and Vrindavan were thus renewed through the bestowal of images.

Naming, Mythmaking, and Landscaping toward Consecrating a Vaishnava Realm The transformation of the Land of the Mallas into Vishnupur (Vishnu/ Krishna’s center) was achieved in part through the systematic renaming and consecrating of the territory held by this local dynasty. Villages surrounding the capital city of Vishnupur were appropriated to this newly discovered “Brajamandala” (sacred diagram delimiting the territorial extent of Braj) and given the names Gokulnagar, Jadabnagar, and Vasudevpur, just as parts of Benares, for example, had earlier symbolically incorporated other major tirthas (pilgrimage sites) toward the establishment of a pilgrimage center.22 And Vrindavan itself had appropriated within its forested terrain the four major dhams (sacred abodes of Krishna/Vishnu) that mark the corners of India and thus define the great circumambulation of the Hindu world.23 The superimposition of a Gaudiya Vaishnava mandala on Vishnupur was achieved partly by naming and giving new significance to the region’s geographical features. The rivers Damodar and Kangsabati are perceived as male and female lovers in eternal union like the divine couple, Krishna and Radha. Damodar is in fact another name of Krishna. According to the region’s oral lore, the female river, the smaller tributary, moves toward union with him. In one of the origin narratives of Madan Mohan, the Dwarakeshwar River is identified as the source of the god’s image.24 The king retrieved the metal image from the water in his net while fishing, along with implements for his worship and gold bricks to sustain the public performance of rituals. Again the process of Gaudiya Vaishnava appropriation of territory is comparable with that carried out at Vrindavan during the previous century. There, distinctive geographical fea-

Epilogue

•

191

tures such as the river and mountain had made the site suitable as a natural tirtha, and the place itself had been a primary locus of devotion.25 Through acts of renaming topographical elements and investing them with significance in the biography of the deity, a realm suited for Krishna’s sports had been mapped onto it. A physical transformation of the land and mapping of the sacred physical markers of Vrindavan onto the Malla capital accompanied this process. The land of Vrindavan is itself believed to be the embodiment of the divine and is far more important for Gaudiya Vaishnava pilgrims than the worship of temple images. Numerous local legends point to Mt. Govardhana’s status as a living being, as Krishna himself in the form of a mountain.26 While the mountain would have been difficult to replicate in the low-lying deltaic plains of Bengal, the kundas (ponds) scattered throughout the forests of Vrindavan, such as the famous paired Radha and Shyam kundas at the foot of Mt. Govardhana, were simulated at Vishnupur. These water bodies became stops for worship on Gaudiya Vaishnava pilgrimage wanderings through the forests of Vrindavan from the sixteenth century. The artificial water sources excavated by the rajas at Vishnupur, called bandhs, are situated near the perimeters of the town. Located near clusters of temples, they provided water for ritual purposes and an arena for re-enacting the water sports of Krishna during festivals.27 Their names, such as Jamunabandh, Lalbandh, Kalindibandh, Shyambandh, and Krishnabandh, suggest the newly discovered secret Vrindavan. Accounts of miracles accrued around them. While Hemlata Devi bathed in the Krishnabandh, the drops of her bath water were turned to rose petals by Radha herself.28 Another narrative recalls how Srinivas, when calling his daughter to return from her bath, witnessed her play with a boy he recognized as Radha Raman, her chosen deity.29 Such narratives complement the events witnessed by the previous generations at Vrindavan’s famous ponds, the most important of which is the Radhakunda. Originally dug by Krishna himself to clean up after killing the demon bull Arishta before he touched Radha, this pond was recognized by Chaitanya in the midst of the fields, and excavated by the goswami Raghunath Das in 1546.30 Such superimposition of an older sacred cartography onto a newly developing site to invest it and its agents with the authority of the former was by no means an unfamiliar strategy. The inland pools of Benares had been named for the sacred rivers of the four ends of the land, which had thereby been invoked within the confines of the city;31 the medieval Pallava and Chola kings of south India had claimed to bring the River Ganga down to their domain. Like the Radha and Shyamkunda on the edges of Vrindavan, these water bodies in Vishnupur were used for bathing. Dipping in the waters

192

•

Temple to Love

of the Radhakunda and Shyamkunda, where the deities had themselves sported, is believed to afford direct access to them. Like other parts of the forests of Vrindavan, these waters are considered to be the divine manifested in natural form. The Caitanya Carita¯ mr. ta claims, In that kun.d.a Kr.s.n.a eternally with Ra¯dhika¯ plays games in the water, and dallies with her on its banks. Whoever bathes in that kun.d.a, Kr.s.n.a grants to him a prema like that of Ra¯dha¯, and the sweetness of the kun.d.a is like the sweetness of Ra¯dha¯ and the greatness of the kun.d.a is like the greatness of Ra¯dha¯.32

In their replication at Vishnupur, that sweetness of Radha and Krishna is transmitted to the emergent religious center and made physically accessible to the newly converted local populace. The rajas also landscaped gardens, planted flowers and fruit trees from Vrindavan, and constructed pavilions around them. The trees of Vrindavan are as significant as its ponds in the mapping of Krishna’s eternal realm. They were believed to be the members of Krishna’s entourage in arboreal form. While restoring the Radhakunda, the goswami Raghunath Das, for example, was informed in a dream that the five Pandava brothers, the heroes of the Mahabharata, now resided on the banks of the Radhakunda in the form of trees.33 Their transplanting in the soil of Vishnupur physically connects this later site with not only the earthly Vrindavan but also the mythical abode of the deity and his entourage. The strategy of transplanting flora to replicate older sacred sites was also familiar; it has a long history in South Asia. The distribution of branches from the Bodhi tree under which the Buddha had attained enlightenment, for example, initiated and provided the center for the development of a Buddhist cosmic mandala at such sites as Anuradhapura in Sri Lanka.34 Such planning at Vishnupur seems part of a larger attempt to recreate, in a fairly general way, the features marking the sites of Krishna’s activities discovered by Chaitanya, most likely to create a pilgrimage path through forests, ponds, ghats, and temples similar to the one established at Vrindavan. One popular episode that Vishnupur’s tour guides tell about Madan Mohan points to the emergence of the site as a pilgrimage center. The deity is believed to have miraculously mobilized the stone chariot, a permanent and immobile architectural feature of the town (figure 5.3). This ratha, built by Vir Hambir’s successor, Raghunath Singh (1627–56), is located at the foot of the double stone gateways of the old fort. In his compassion, Madan Mohan stopped the wheels of his chariot from moving as he waited for his last devotee, an old woman who was traveling on foot from a faraway village, to arrive for the celebration of

Epilogue

•

193

Ratha Yatra, the chariot festival. The episode expresses the conviction that Madan Mohan is jagrata (awake and alert) to attend to the needs of his followers, the conviction that forms the basis of the pilgrimage to Vishnupur. His priest is convinced that the deity’s far-reaching influence alone maintains this community-patronized temple, drawing generous donations not just from devotees of neighboring districts but from Benares and Vrindavan in north India.35 Pilgrimage thus contributes to the local economy when kingly support is no longer available.

Madan Mohan of Vishnupur The Madan Mohan Bandana, an eighteenth-century poem extolling the town’s primary deity, Madan Mohan, seems to be the earliest surviving text to use the term “Gupta Vrindavan.”36 The description of each well-known miracle performed by Madan Mohan culminates in the couplet “This Vishnupur became Gupta Vrindavan / Where resides Lord Madan Mohan.” These lines provide an irregular refrain that strings the numerous episodes into a narrative. The poem reflects Madan Mohan’s ties to the local landscape and its distinctive topographical and architectural features, and thereby reaffirms the conflation of Vrindavan with Vishnupur. Vishnupur, as depicted in the poem, is a town imbued with the presence of Madan Mohan, much as Vrindavan had been inhabited by Krishna and retains the imprint of his miracles. This presence is demonstrated in two ways: by Madan Mohan’s direct involvement with the townspeople and by his interaction with the town’s natural topographic features and architectural forms. Perhaps the best known of Madan Mohan’s feats is his defense of the town from the historic bargi (Maratha) raids of the 1740s. While the townspeople gathered in his temple to chant the names of Krishna, Madan Mohan is believed to have galloped on a red horse to the edge of the fort and fired the royal cannon Dalmadal that dispersed the light artillery of the Maratha troops. Again the deity is integrally tied to the extant canon and the urban fabric through narrative, as Krishna had been to Mt. Govardhana, which he raised, and to Radha Kunda, the site of his nightly amorous exploits with Radha. Their association with the deity invests material forms and historical events with an authority that elevates them from their location in history to mythic status. By extension, Vishnupur is afforded the same standing. The invocation of these narratives by local residents maintains Madan Mohan’s identity with Krishna and the status of the town as Krishna’s abode. His accomplishments at Vishnupur are compared by the women of the Vishnupur Raj family to Krishna’s feats in Mathura, such as the killing of Kamsa, the slaying of the serpent Kaliya, and the raising of Mt. 194

•

Temple to Love

Govardhana.37 Other devotees perceive them as identical and interchangeable. During the annual celebration of the coming of Srinivas and ensuing Vaishnavization of the town at the Madan Mohan Temple, a local resident observed, “As Krishna had lived and played in the groves of Vrindavan, Madan Mohan resides in Vishnupur.”38 She explained that Krishna’s realm is here because all the temples’ deities ride their chariots to Madan Mohan’s courtyard each night, just as the gods had all flocked to Vrindavan. Hence this was the “Gupta Vrindavan.” Narrative is thus employed to retain the town’s distinctive status. The Madan Mohan Temple’s south façade, sheathed with terra cotta depictions of Krishna’s adventures, reinforces the identity of the local lord Madan Mohan as the universal deity Krishna. The metal image of Madan Mohan in the sanctum is virtually indistinguishable from the iconography of Krishna, or from that of any of the other specific images, such as Radha Raman, Radha Mohan, Murali Mohan, and Gopinath. Like them, Madan Mohan lifts his hands to play the silver flute that enchanted the gopis. Like them, he sways with his weight on one hip, his body curving rhythmically. The right leg is poised momentarily in front of the left, as if pausing in his dance. His head is bent, mesmerized by the sound of his flute, as is his beloved Radha. The identity of Madan Mohan as Krishna thus reverberates iconographically in sculpted metal and terra cotta. By introducing the designation “Gupta Vrindavan,” Madan Mohan’s devotees at Vishnupur explicitly claimed for the younger town the sacred authority of Vrindavan. The “hidden” aspect of this identification probably made the site itself a tool for Gaudiya Vaishnava initiates, who alone could comprehend its true significance. Such re-creation of the ultimate sacred realm in Bengal may indeed have fulfilled the ritual prerequisite for devout Gaudiya Vaishnavas to reside within the mystical mandala of Mathura, the site of Krishna’s eternal play.39 Conversely, the selective features of Vrindavan that were transported to Vishnupur determined the most important criteria for a sacred center for Vaishnavas. As this new Vrindavan remapped the original one, the replication process elevated Vrindavan’s status to myth, eliding the historical reality of its creation only a century earlier.

Composition of Poetry and Song Parallel to Vrindavan’s established reputation as a center for composition of the central texts of the tradition, Vishnupur gradually emerged as a locus for composition and performance of Vaishnava poetry and song in Bengal. In fact, part of the town’s present claim to fame rests on the foundation of a music school, the emergence of local musical geniuses, and the production of musical instruments to accompany kirtan perEpilogue

•

195

formances. Srinivas himself is credited with the introduction of the Manoharshahi Gharana, a school of music with a distinctive compositional style. Local singers and musicians perpetuated this style by leading public recitals in the temple courtyards during festival celebrations. This process was initiated by the enthusiastic personal sponsorship of Gaudiya Vaishnava song and poetry by Vir Hambir, and succeeding generations of Malla rulers continued the trend. The Karnananda describes Vir Hambir’s keen interest in these compositions and his active participation in their dissemination and preservation: The Raja recites the pada (lyric) dreaming, and on hearing it Pattadevi [his queen] was moved and burst into tears of ecstasy. At daybreak the raja looked delighted and at that moment the queen implored him to recite the padas to her again. She would die if the raja denied her the rare pleasure.40

Several of these rulers, including Vir Hambir, Raghunath Singh, and Chaitanya Singh (1752/3–1802), are credited with actively composing kirtan. Gopal Singh (1713–52) wrote a more ambitious work called Krishnamangal Kavya. By personally copying the rich local collection of poetry as well as translating core Vaishnava texts into Bengali, the Malla rulers set an example of making these texts accessible to the emergent Vaishnava community they supported. Gopal Singh, for example, copied the Caitanya Carita¯mr. ta. Large numbers of these manuscripts, with their finely painted wooden covers, survive in the Vishnupur Sahitya Parishad. In this way, Vishnupur developed a reputation as a local center for composing poetry, thereby preserving and disseminating the movement’s theological premises. However, this was not simply a matter of copying existing texts of the Gaudiya Vaishnava tradition. The rich corpus of devotional poetry that emerged in Bengal at this time was grounded locally to make it familiar and recognizable. Such a strategy served the dual purposes of strengthening the emergent community and facilitating proselytization, thereby assuring the continuity sought by the tradition’s leaders. Written versions of the songs reveal shared technical devices and imagery, with the poetry burgeoning in conjunction with the region’s flourishing sufi culture. To cite one instance, the following description of Yusuf, from the sufi story of Yusuf and Zulaikha, for example, bears startling similarities to those of Radha and Krishna: . . . your eyes are black as if bees are buzzing around them. Your eyebrows are like the bow of Kama and your ears like lotuses which grow on shore. Your waist is as slim as that of a prowling tigress. Your step is as light as a bird’s and when they see it even sages forget all else.

196

•

Temple to Love

Your body is as perfect as a well-made string of pearls. A maid, therefore, cannot control herself and longs for your embrace.41

If the sufi context were unknown, such lines could be identified with vivid descriptions of the beloved in Vaishnava verses such as the following one by Sheikh Kabir: What a wonderful beauty the maiden possesses! Her movement is as slow as that of an elephant. Her collyrium-coloured eyes on the white forehead Look like bees in a beautiful lotus garden. She is devoid of pride and her waist is thin The beautiful maiden possessing a moon-like face speaks smiling Like the full autumnal moon, it pours nectar.42

Resources shared by Gaudiya Vaishnavas and Muslims in oral and written compositions, as in architecture, thus suggest that both communities emphasized established regional patterns in order to smoothly introduce new traditions. It also suggests a second Vaishnava strategy for endurance in Bengal’s political situation. Rather than ignoring or challenging Islamic hegemony, Vaishnavas adapted to that reality and subversively appropriated relevant aspects of Islam’s message and strategies, thereby ensuring the success of Vaishnavism in the seventeenth century. Thus it is by affinity with contemporary developments as much as by differentiation from them that a local Vaishnava community and sacred center was secured.

Pilgrimage Routes The selection of Vishnupur as a new site for Vrindavan must have been due partly to its geographical location in southwestern Bengal, strategically situated on several important pre-existing routes through north India, probably mapped for trade, which were also traversed by Gaudiya Vaishnavas on pilgrimage. The ancient road from Tamralipta (Tamluk) in southwest Bengal to Vrindavan passed through Vishnupur.43 Similarly, the town lay on the road from Orissa to Monghyr in Bihar. The pilgrimage route between Puri and Vrindavan, called “Nilachaler Path” or “Jharikhander Path” and used by Chaitanya on his travels to the temple of Lord Jagannath at Puri, also went via Vishnupur.44 The first of these routes was critical to the creation of a Gaudiya Vaishnava center at Vishnupur because Srinivas journeyed from Vrindavan with the cartload of manuscripts compiled by the goswamis. This initial journey was renewed by Raja Vir Hambir himself, who traveled back and forth. Ingredients that went toward the making of a new Vrindavan were transported along this route. The raja, for example, brought kirtan composed at Vrindavan to be sung in his temples’ courtEpilogue

•

197

yards. He also brought back fruit and flowering trees on such trips and planted them himself to transform his capital into the green groves where Krishna could sport with the gopis once again.45

Layering The analogy between Vishnupur’s founding and Vrindavan’s rediscovery becomes even more powerful when it is recalled that Srinivas was himself regarded as an incarnation of Chaitanya during the seventeenth and eighteenth centuries. The association of Vishnupur with Vrindavan was thus strengthened to the point of identification, both being mapped as part of a single Gaudiya Vaishnava world. Such strategies deployed across multiple media suggest a deliberate attempt to yoke the two Gaudiya Vaishnava centers together and present Srinivas’s Vishnupur as a reincarnation of Chaitanya’s Vrindavan. By the mid-eighteenth century, when the last Ratna temples were built at Vishnupur, this replication of the newly reconstituted sacred center at Vrindavan may have developed other implications as well. Vrindavan was on the brink of being lost a second time. Now promoting this new Vrindavan may have been an attempt to replace it and thereby preserve it and maintain continuity in a different sense. Many of Vrindavan’s temples shut their doors and dispatched their deities to safer sites in anticipation of difficult times during Aurangzeb’s reign. The image of Govindadeva was removed from the temple (whose sanctum may have been razed at this time), transported to various temporary homes, and finally installed in a permanent temple at Amber. This transfer of the most important image initiated the creation of yet another Vrindavan, this one called Kanak (golden) Vrindavan, that is, one superior to even the original one.46 Vishnupur was therefore not unique in replicating an earthly site and thereby replenishing the eternal realm. The Vaishnavas’ shifting of the center of their sacred cartography to a more protected area at the empire’s periphery in Bengal, where the turbulences in the heart of the Gangetic Valley would reverberate only slowly, may have been a response to the political contingencies of the period. Invoking the past as an instrument for change in the present is a mechanism that is used widely in the construction of sacred centers. The consecutive creation of multiple sacred centers in rapid succession, replicating an ideal realm by repeated mappings that legitimate each other, is less common. In this case, no absolute mapping of Vrindavan can be recovered from the materials available at Vishnupur. Rather, through temple construction, oral lore, ritual worship, landscaping, and the production of texts Vishnupur was transformed into a parallel sacred center.

198

•

Temple to Love

The careful mediation of the forms and institutions of the original and the presentation of the message of this religious formation in local forms for its new hub suggests a shrewd awareness of and attempt to survive within the political realities of seventeenth-century Bengal. That it was acceptable for this Hindu holy site to be re-created in an Indo-Islamic manner is testimony to the fluidity of the cultural environment of this period. It also reminds us of the negotiated and mutable nature of markers erected to articulate and unify a new religious formation, which simultaneously allow it to define itself by specifying its boundaries. In the current political climate of South Asia, when fervent religious nationalisms are rendering such interplay between the region’s dominant religious cultures inconceivable, the historical moments when Hindu cultural centers could draw on mosques to give shape to earthly manifestations of the divine realm offer hope for the future.

FOLLOWING PAGE: (TOP) Figure 5.1. Govindadeva Temple, Vrindavan

(BOTTOM) Figure 5.2. Laterite Ratha, Vishnupur

Epilogue

•

199

200

•

Temple to Love

Glossary of Architectural Terms

amalaka bangla

myrobalan fruit; a finial on a Nagara temple a Bengali hut with a curved thatch roof; the stylized curve of a vault simulating the hut roof; chala bhogamandapa a temple hall reserved for the preparation of food for the deities chajja a sloping or horizontal projection from the top of a wall, supported by brackets, to provide protection from rain or sun chala a Bengali hut with a curved thatch roof; the stylized curve of a vault simulating the hut roof; bangla chatri a small open pillared pavilion ekaratna a Ratna temple with a single upper shrine garbhagriha “womb-house”; sanctum ghantamala an ornamental motif of bells suspended from a garland of pearls jor bangla twin chala huts kalasha a finial in the shape of a water pot (purna ghat) or pitcher lata a creeper; a vertical band rising up the shikhara in a mesh latina a north Indian curvilinear superstructure mihrab an arched niche in the qibla wall of the mosque Nagara a north Indian single-storied temple-type with a curvilinear tower natmandir a hall for entertainment, located across the courtyard from the sanctum in Ratna temples navaratna a Ratna temple with nine miniature shrine forms above the sanctum: one in each corner, and one in the center that supports four smaller ones above

Axes and the Mediation of Worship

•

201

pancharatna panchayatana

prasada rasamandala ratha Ratna

Rekha, Rekha Deul sardula shikhara tulsimancha uttaravedi

202

•

a Ratna temple with five miniature shrine forms above the sanctum: one in each corner, and one in the center “five-shrined”; a temple complex with a shrine in each corner and one in the center, typically all on a single plinth a palace, a mansion; a temple a circular form depicting the dance of Krishna with the gopis a chariot a double-storied Bengali temple-type, with additional structures creating a bi-axial organization on the lower level a curved shikhara; a Nagara temple-type a mythical composite leonine beast a spire, a temple tower a basil planter, located in a Ratna temple courtyard, often in the shape of a chariot an upper altar, serving as a base for the finials on a Nagara shikhara

Glossary of Architectural Terms

Notes

Introduction 1. Most temples at Vishnupur, and many constructed at other sites during the seventeenth century, bear dedicatory inscriptions with precise dates and names of patrons. This epigraphic information has been collected in A. K. Bhattacharyya, A Corpus of Dedicatory Inscriptions from Temples of West Bengal, c. 1500 A.D. to c. 1800 A.D. (Calcutta: Nabhana, 1982). The translation of the Shyam Ray Temple’s inscription is my own, based upon Bhatta. . charyya’s transcription: “S´rı¯ Ra¯dhika¯-Kr.s.n.a-mude S´ake=’nka-Ved=a¯nka-yukte · ´ ´ navaratna-ratnam Srı¯-Vı¯ra-Hamvı¯ra-nares´a-su¯nur=dadau nr.pa-Srı¯-Raghuna¯tha · hah Malla Sa(S´a)ke 949 S´rı¯ ra¯ja¯ Vı¯ra Sim · ha[h],” 70–71. Sim . . 2. The phrase is used repeatedly on Vishnupur’s temples, such as the Kala Chand (1656) and Lalji (1658). 3. The term bangla reflects how closely the hut-based form was associated with the region. Mughal writers preferred this term to others, such as chala, in discussing the adaptation of the temples’ distinctive roof form for imperial palatial architecture in their capitals at Delhi and Agra; see my “Space and the New Temple Vernacular,” in Traditional and Vernacular Architecture, ed. Deborah Thiagarajan and Subashree Krishnaswamy (Chennai: Madras Craft Foundation, 2003), 16–30. 4. The movement gets its name from Gaur (Gaud.), the sixteenth-century capital of the independent Sultanate of Bengal. This community dedicated to the worship of Krishna/Vishnu is thereby distinguished by its Bengali roots from other contemporary developments in north India such as the Pushti Margis, Radha Vallabhis, and Nimbarkis. For overviews of the bhakti tradition that coalesced under Chaitanya, see Sushil Kumar De, Early History of the Vaishnava Faith and Movement in Bengal, from Sanskrit and Bengali Sources (Calcutta: Firma KLM, 1986); Melville T. Kennedy, The Chaitanya Movement (New York: Garland, 1981); Ramakanta Chakravarti, Vais. n.avism in Bengal, 1486–1900 (Calcutta: Sanskrit Pustak Bhandar, 1985). 5. For a discussion of the term lila in the Gaudiya Vaishnava context, see Edward C. Dimock, “Lı¯la¯,” History of Religions 29.2 (November 1989): 159–73. 6. The epigraph of the temple sponsored by King Vijaya Sena (1095–1158) at Deopara, Rajshahi district, for instance, reads, “That sacrificer [the king]

Notes to pages 000–000

•

203

calling down the immortals from the slopes of Meru [the cosmic mountain] . . . brought about an interchange of heaven and earth . . . [The king] built a lofty edifice of Pradyumneshvara [the god Shiva], the wings, and plinth and the main structure of which occupied the several quarters, and the middle and uppermost parts stretched over the great oceanlike space—the midday mountain of the rising and setting Sun . . . the king regularly replaces the ascetic god Shiva’s elephant skin garment with multi-colored silk clothes, puts round the god’s breast a pearl necklace instead of the huge serpent, applies sandal powder to his skin instead of ashes . . . what could the moon-crested god bestow on him? When the end of the king’s days come, may Shiva grant him final union with himself!” Nani Gopal Majumdar, Inscriptions of Bengal (Rajshahi: Varendra Research Society, 1929), 3:52–53. The monument is no longer standing; its epigraph is now housed in the Indian Museum, Calcutta. 7. David McCutchion, an English professor of comparative literature at Jadavpur University, Calcutta, during the 1960s, conducted the most comprehensive survey of these monuments undertaken to date. He included temples in Bengal from the Pala-Sena period (that is, eighth to twelfth centuries) to the twentieth century in his study, Late Medieval Temples of Bengal: Origins and Classification (Calcutta: Asiatic Society, 1972). The arduousness of the process took its toll on his health and after his death, his research was collected, further interpreted, and published by George Michell as Brick Temples of Bengal: From the Archives of David McCutchion (Princeton, N.J.: Princeton University Press, 1983). 8. From the mid-eighteenth century, increasing British control over Bengal deprived them of their status in favor of the Burdwan Raj. 9. The Radha Ballabh Temple at Krishnanagar is painted in vivid colors, highlighting the terra cotta lines. At Guptipara, the porch of the Vrindavan Chandra Temple is painted rather than ornamented in terra cotta. The painted panels are organized in a grid, like the terra cotta surfaces. A wide variety of colors are used and the style is similar to that of nineteenth-century Kalighat paintings. Traces of paint also survive on the Vishnupur Radha Lalji Temple’s porch and the gates of the Radha Shyam Temple. (Whether this paint dates from the time of construction cannot be ascertained, however, without chemical analysis.) Mosque construction from the late fifteenth to the sixteenth centuries at Gaur also suggests that color was a significant component of the period aesthetic. Colored and glazed ceramic tiles survive on the late-fifteenth-century Lattan Masjid at Gaur. By the nineteenth century, stucco had replaced terra cotta for the most part, responding to the preference in Bengal’s Mughal mosques. 10. Sailendra Biswas, Samsad Bengali-English Dictionary (1968; reprint, Calcutta: Sahitya Samsad, 1996). 11. M. W. Meister, “Pra¯sa¯da as Palace: Kut.ina¯ Origins of the Na¯gara Temple,” Artibus Asiae 49.3 (1988–89): 254–80. 12. See Bhattacharyya, Corpus of Dedicatory Inscriptions, 22. 13. David McCutchion, “Origins and Developments,” in Brick Temples of Bengal, ed. Michell, 24. 14. Ibid., 24–25. 15. Only at a later date is this new form adapted for constructing temples to the gods of other Hindu groups, such as Shiva and Durga. 16. Edward C. Dimock, “Bhakti,” in Cooking for the Gods: The Art of

204

•

Notes to pages 000–000 3–13

Home Ritual in Bengal, ed. Michael W. Meister, cat. Pika Ghosh (Newark, N.J.: The Newark Museum; Philadelphia: University of Pennsylvania Press, 1995), 29. 17. Rupa Goswami defined the terms in his Ujjvala-nilamani: “A parakiya woman is she who, having no dependence on ordinary dharma, belonging to another, is attracted to a man and causes him to be attracted to her, but who does not enter into marriage with him. A swakiya woman is she who has been taken in marriage according to the accepted rites, who is obedient to the wishes of her husband, and she does not depart from the dharma of her wifely vows.” Edward C. Dimock, The Place of the Hidden Moon: Erotic Mysticism in the Vais.n.avasahajiya¯ Cult of Bengal (Chicago: University of Chicago Press, 1966; reprint, Chicago: University of Chicago Press, 1989), 17. 18. Caitanya Carita¯mr.ta of Kr.s.n.ada¯sa Kavira¯ja, trans. and comm. Edward C. Dimock, ed. Tony K. Stewart (Cambridge, Mass.: Dept. of Sanskrit and Indian Studies, Harvard University; distributed by Harvard University Press, 1999), 1.17.115–22, 321. For a description of nagar samkirtan in a Bengali village in Midnapore district, see Ralph W. Nicholas, “Vaisnavism and Islam in Rural Bengal,” in Bengal Regional Identity, ed. David Kopf (East Lansing: Asian Studies Center, Michigan State University, 1969), 34. 19. Recently reconstructed Malla family trees and oral tradition locate Vir Hambir at the end of a long line of rulers, but there is little corroborating evidence to confirm this lineage. The precise dates of Vir Hambir’s reign and conversion remain uncertain and contested among historians. However, most agree that he ruled from the last decade of the sixteenth to the second decade of the seventeenth century. The only contemporary source to mention his name is the Akbarnama. A detailed discussion of this issue is available in Chakravarti, Vais.n.avism in Bengal, 222–28. See also Prabhat Kumar Saha, Some Aspects of Malla Rule in Bishnupur, 1590–1806 (Calcutta: Ratnabali, 1995), 24–28. 20. Srinivas Acharya stopped in Vishnupur on his return to Bengal from Vrindavan, bearing the texts formulated there by the six goswamis. According to local legend, the king employed bandits to steal the treasures carried by Srinivas, who subsequently recovered them. Srinivas then proceeded to impress the king and convert him to Vaishnavism, giving him a new name, Haricharan Das (servant of, or at the feet of, Krishna). This legend is recorded in Bhaktiratnakar, chapter 7, and Prem Vilas, chapter 13. See Dinesh Chandra Sen, The Vais.n.ava Literature of Mediaeval Bengal (Calcutta: University of Calcutta, 1917), 83– 183; Chakravarti, Vais.n.avism in Bengal, 213; Saha, Some Aspects of Malla Rule, 172–78. 21. I am grateful to Tony Stewart, who deeply shaped my understanding of this material during discussions at the University of Pennsylvania in 1997, guided my reading of Gaudiya Vaishnava literature, and shared parts of his manuscript in progress, The Final Word. See also Steven Rosen, The Lives of the Vaishnava Saints (New York: Folk Books, 1991). 22. Nityananda Das, Prem Vilas, ed. R. Vidyaratna (Murshidabad: Radharaman Press, 1298 BS.), 19.174–75. 23. For a discussion of this festival, see Chakravarti, Vais.n.avism in Bengal, 229–43. 24. Das, Prem Vilas, 19.184–88. 25. Chakravarti, Vais.n.avism in Bengal, 236. 26. Sen, Vais.n.ava Literature, 88–89, 121.

Notes to pages 000–000 Notes to pages 13–15

•

205

27. Paul Connerton, How Societies Remember (Cambridge: Cambridge University Press, 1989), 61. 28. Prem Vilas, quoted in Sen, Vais.n.ava Literature, 152–53. 29. The poet Mukundaram, in his 1589 narrative poem Kavikankan Chandi, described a highly stratified society in the late sixteenth century, dominated by the Brahmans and Kayasthas who controlled the land and ritual. They were followed by the artisans and agriculturalists, then by the lower castes of fishermen, and lastly by untouchable groups, including sweepers and cremation ground workers. Mukundaram Chakrabarty, Kavikankan Chandi, ed. Srikumar Bandyopadhyay and Visvapati Chaudhuri (Calcutta: Calcutta University, 1974), 343–61. 30. Gifts to Brahmans for officiating over rituals included tax-free land. Lakshmana Sena, for example, recorded the “fee for the ceremony of the Great Gift in which a golden horse and chariot were given away, on this auspicious day, after duly touching water and in the name of the illustrious god Narayana, for the merit and fame of my parents as well as myself, for as long as the moon, sun and the earth endure, according to the principle of Bhumichchhidra, to Is´varadevas´armman, who officiated as the Acharya in the ‘Great Gift of gold horse and chariot.’” Majumdar, Inscriptions of Bengal, 3:104. 31. For discussions of the consolidation of Brahman authority in Bengal, see Kunal Chakrabarti, Religious Process: The Puranas and the Making of a Regional Tradition (New Delhi: Oxford University Press, 2001); and Ronald B. Inden, Marriage and Rank in Bengali Culture: A History of Caste and Clan in Middle Period Bengal (Berkeley and Los Angeles: University of California Press, 1976). 32. Ramakanta Chakravarti has also noted the tension between the orthodox preference to preserve the status and authority of priests and Brahman castes, and the anti-caste and anti-intellectual position that remembrance of the sports of Krishna and Chaitanya could be performed without priestly intervention. For example, in Midnapur, the seventeenth-century leader Shyamananda introduced the sacred thread associated with Brahmans into initiation among Vaishnavas, regardless of their caste. This act would have constituted a cooption of a powerful symbol of Brahman authority for lower-caste devotees. Chakravarti, Vais.n.avism in Bengal, 314–16, 320–27. 33. Hitesh Ranjan Sanyal, “Literary Sources of Medieval Bengali History: A Study of a Few Mangalkavya Texts,” in Indian Studies: Essays Presented in Memory of Prof. Niharranjan Ray, ed. Amita Ray, H. Sanyal, and S. C. Ray (Delhi: Caxton, 1984), 205–15. 34. L. S. S. O’Malley, Bengal District Gazetteers: Bankura (Calcutta: Government of West Bengal, 1908; reprint, Calcutta: Government of West Bengal, 1995), 56–57. 35. For a description of the livelihoods and lifestyles of these social groups, see Chakravarti, Vais.n.avism in Bengal, 60–61. 36. E. T. Dalton, Descriptive Ethnology of Bengal (Calcutta: Thacker, Spink, 1872; reprint, Calcutta: K. L. Muhkopadhyaya, 1973), 172, 181–82. 37. William Wilson Hunter, Annals of Rural Bengal (London: Smith, Elder, 1897), appendix E, 439–41; for another version, see O’Malley, Bengal District Gazetteers, 26–27. 38. Some historians of the Vishnupur Raj, such as Prabhat Kumar Saha,

206

•

15–18 Notes to pages 000–000

however, argue that the Mallas were in fact descended from Rajput princes, and belonged to the appropriate Kshatriya caste. Saha, Some Aspects of Malla Rule, 21–23. For an opposing view, see Hitesh Ranjan Sanyal, “Mallabhum,” in State Formation in Eastern India, ed. Surajit Sinha (Calcutta: Proceedings of the Seminar on State Formation in Eastern and Northeastern India at the Centre for Studies in Social Sciences, 1981), 73–142. 39. On parallel processes, see Surajit Sinha, “State Formation and Rajput Myth in Tribal Central India,” Man in India 42.1 (1962): 35–80; and Hermann Kulke, “Legitimation and Townplanning,” in Kings and Cults: State Formation and Legitimation in India and Southeast Asia (New Delhi: Manohar, 1993), 93– 113. 40. On the origins and development of this celebration, see Chakravarti, Vais.n.avism in Bengal, 246–47; and Joseph T. O’Connell, “Chaitanya Vaishnava Movement: Symbolic Means of Institutionalization,” in Organizational and Institutional Aspects of Indian Religious Movements, ed. Joseph T. O’Connell (Shimla: Indian Institute of Advanced Study, 1999), 215–39. 41. Dimock, Place of the Hidden Moon, 27. 42. See table of villages given for religious purposes and revenue acquired from land given in the name of the temple deities, and to the priests, in Saha, Some Aspects of Malla Rule, 76. 43. Arjun Appadurai, Modernity at Large: Cultural Dimensions of Globalization (Minneapolis: University of Minnesota Press, 1996), 198. 44. João de Barros, Da Asia, decade 4, book 9, chapter 1, quoted in Ronald Bishop Smith, The First Age of the Portuguese Embassies and Peregrinations to the Ancient Kingdoms of Cambay and Bengal (1500–1521) (Bethesda, Md.: Decatur Press, 1969), 133. 45. Velcheru Narayana Rao, David Shulman, and Sanjay Subrahmanyam, Symbols of Substance: Court and State in Na¯yaka Period Tamil Nadu (Delhi: Oxford University Press, 1992), 55. 46. R. C. Majumdar, History of Mediaeval Bengal (Calcutta: G. Bhardwaj, 1973), 335; J. C. French, “The Land of Wrestlers (A Chapter in the Art of Bengal),” Indian Art and Letters 1 (1927): 15–29; Gurusaday Dutt, “Bengali Terracottas,” Journal of the Indian Society of Oriental Art 6 (1938): 169–80; and Tarapada Santra, “Cataloguing and Documentation of Mediaeval Temples of Bengal: An Analysis of Problems and Methods,” Calcutta Review, new series 3.3 (January–March 1972): 239–45. 1. Desire, Devotion, and the Double-Storied Temple 1. Over the course of four fieldwork trips from 1996 to 2002, I have discussed this issue with Sushanta Mukhopadhyay, priest of the Madan Mohan Temple; Gangadhar Das and Mathur Majhi, of the Archaeological Survey of India; Mrs. Salil Singh, of the Vishnupur Raj family; and Achintya Banerjee, a local scholar and tour guide, as well as with residents who participate in the lives of these temples, including Banshi Kar Mukhya, Chhaya Das, Tara Banerjee, and Chandi Chakraborty. Beyond Vishnupur, I am grateful to the caretaker of the Gokul Chand Temple for our discussions of the situation at that site in December 1996. 2. Exceptions include the Radha Vinod Temple in Vishnupur and the

Notes to to pages 000–000 Notes pages 18–42

•

207

Sharabhuja Temple in Tejpal, a village adjoining Vishnupur. These structures conform to the appearance of a double-storied monument by simulating an upper pavilion on the exterior, but the structure is placed above the roof and there is no access to it. Their epigraphs reflect this distinction, as they do not use the term navaratna ratnam, which is employed in true double-storied monuments’ inscriptions. 3. An exception is the Hamsheshvari Temple, built at Bansberia, Hooghly district, in 1799. It is six-storied, the stacking of shrines now used for mapping tantric conceptualizations of the cosmos, and the verticality as a metaphor for mapping the practitioner’s aspirations. See N. N. Bhattacharyya, History of the Tantric Religion (1982; reprint, Delhi: Manohar, 1992), 375–78. 4. An enormous brick pyramidal rasamancha was also constructed at Vishnupur in the seventeenth century to host the town’s deities for the celebration of rasa. 5. Oral versions of this narrative were recounted to me by several people, including Achintya Banerjee (December 1996) and the mother and wife of Chittaranjan Dasgupta (May–June 2000). 6. See Alan W. Entwistle, Braj: Centre of Krishna Pilgrimage (Groningen: E. Forsten, 1987), 185; and R. Nath, “S´rı¯ Govindadeva’s Itinerary from Vr.nda¯vana to Jayapura, c. 1534–1727,” in Govindadeva: A Dialogue in Stone, ed. Margaret H. Case (New Delhi: Indira Gandhi National Centre for Arts, 1996), 161–84. 7. See Entwistle, Braj, 184. On worship of Pushtimargi images and their representation in painting, see Amit Ambalal, Krishna as Shrinathji: Rajasthani Paintings from Nathdvara (Ahmedabad: Mapin, 1987); and Stuart Cary Welch, ed., Gods, Kings, and Tigers: The Art of Kotah (Munich: Prestel, 1997). 8. See Hermann Kulke, “Early Patronage of the Jagannatha Cult” and “The Struggle between the Rajas of Khurda and the Muslim Subadars of Cuttack for Dominance of the Jagannatha Cult,” in The Cult of Jagannath and the Regional Tradition of Orissa, ed. Anncharlott Eschmann, Hermann Kulke, and Gaya Charan Tripathi (New Delhi: Manohar, 1978), 139–55, 321–42. One of the fascinating aspects of the complex power struggles surrounding Jagannatha in the late sixteenth century is that “the initial fight of the Afghans against the Jagannatha cult turned in the events during 1735 when a new Muslim Subahdar, against the embittered resistance of the Khurda raja, but with the obvious support of the priests of Puri, forcibly brought back the image of Jagannatha from its hiding place in South Orissa and established the cult of Puri.” Kulke, “Struggle,” 323. 9. I discuss the duplication of Vishnupur’s Madan Mohan image in my “Sojourns of a Peripatetic Deity,” RES Anthropology and Aesthetics 41 (spring 2002): 104–26. 10. Oral communication from Sushanta Mukhopadhyay, Vishnupur, December 1996. 11. Yellow is associated with Krishna as early as the twelfth-century songs of the Gitagovinda. Images elsewhere, such as Jagannath of Puri, are also typically dressed in yellow for festivals; see Katherine F. Hacker, “Dressing the Lord Jagannatha in Silk: Cloth, Clothes, and Status,” RES Anthropology and Aesthetics 32 (autumn 1997): 109. 12. Bengali songs also describe her as wearing a brick-red sari; see Donna M. Wulff, “Radha: Consort and Conqueror of Krishna,” in Devi: Goddesses of

208

•

Notes to pages 42–46 000–000

India, ed. John S. Hawley and Donna M. Wulff (Berkeley and Los Angeles: University of California Press, 1996), 127. Red is the color of a wedding sari, indicating the paradoxes in interpretations of their relationship. On the theological discussions about the relationship, see De, Early History of the Vaishnava Faith, 204–206, 348–51; and Dimock, Place of the Hidden Moon, 200–14. 13. Oral communication from Sushanta Mukhopadhyay, Vishnupur, December 1996. His behavior is also like a peacock’s mating dance. Like the male bird, which spreads out his tail and dances joyously in the rain, Krishna beguiles the women of Mathura to the woods with his haunting flute and dances with them. Like the peacock attempting to catch the eye of each hen in sight, Krishna multiplies himself to satisfy each of the gopis’ desires. 14. The patron family of the Gopinath Temple at Dasghara, Midnapore district, for example, gives jewelry to Gopinath, a form of Krishna, when their children pass difficult exams or get jobs, or at the birth of a child in gratitude for the well-being of the family. Oral communication from Paramita Biswas, Calcutta, May 1995, as well as Sushanta Mukhopadhyay, Vishnupur, December 1996. 15. These are usually seasonal flowers. Madan Mohan, for example, is believed to prefer frangipani blossoms at the beginning of the rainy season. Oral communication from Chhaya Goswami, Vishnupur, May 11, 2000. 16. David Curley’s translation from Rasik’lal Gupta’s Raja Raj’ballabh, in his “Temples, Tanks, and Maths: Royal Authority in the Zemindari Seat of Raj’nagar” (presented at the University of North Carolina, March 2000). 17. Ibid. 18. McKim Marriott, “The Feast of Love,” in Krishna: Myths, Rites, and Attitudes, ed. Milton Singer (Chicago: University of Chicago Press, 1966), 206. David Kinsley has also emphasized Krishna’s abandoning of social and moral convention: “It comes from another world where this-worldly morality and conduct have no place.” Kinsley, The Divine Player: A Study of Krishna Lila (Delhi: Motilal Banarsidass, 1979), 99–100. 19. Frédérique Apffel Marglin, Wives of the God-King: The Rituals of the Devadasis of Puri (Delhi: Oxford University Press, 1985; reprint, Delhi: Oxford University Press, 1989), 276. 20. According to the Caitanya Carita¯mr. ta, from then on he took it upon himself to prepare the ground for Krishna’s play with the gopis by sweeping the temple and washing it out with pots of water: “That Gaura [Chaitanya], having washed and cleansed with his followers the temple of S´rı¯ Gun.d.ica¯, and having made it cool as his own mind, made it fit for receiving Kr.s.n.a.” Caitanya Carita¯mr. ta 2.12. 21. Deula Toka, quoted in Anncharlott Eschmann, “The Vaishnava Typology of Hinduization and the Origin of Jagannatha,” in Cult of Jagannath, ed. Eschmann, Kulke, and Tripathi, 99. 22. Marglin, Wives of the God-King, 276–77. 23. John Stratton Hawley, At Play with Krishna: Pilgrimage Dramas from Brindavan (Princeton, N.J.: Princeton University Press, 1981), 27. 24. On swings and sexual symbolism, see Tryna Lyons, “The Simla Devı¯ Ma¯ha¯tmya Illustrations: A Reappraisal of Contents,” Archives of Asian Art 45 (1992): 36; and Marglin, Wives of the God-King, 235. The goddess Sitala, who is associated with heightened sexual energy and a capacity for angry violence, is

Notes pages 46–50 Notes to to pages 000–000

•

209

also cooled by swinging. See Katherine F. Hacker, “Traveling Objects: Brass Images, Artisans, and Audiences,” RES Anthropology and Aesthetics 37 (spring 2000): 150. 25. Sen, Vais.n.ava Literature, 103. 26. Sukhamay Bandyopadhyay, Temples of Birbhum (Delhi: B. R. Publishing Corporation, 1984), 114. 27. Scholars of more esoteric Vaishnava groups have come to recognize such limitations to the information available to non-practitioners. See Dimock, Place of the Hidden Moon, 234; Dominique-Sila Khan, Conversions and Shifting Identities: Ramdev Pir and the Ismailis in Rajasthan (Delhi: Manohar and Centre de Sciences Humaines, 1997), 20, 268; and Hugh B. Urban, The Economics of Ecstasy: Tantra, Secrecy, and Power in Colonial Bengal (Oxford: Oxford University Press, 2001), 16–22. 28. Edward C. Dimock and Denise Levertov, In Praise of Krishna: Songs from the Bengali (Chicago: University of Chicago Press, 1965), 27. 29. Ibid., 57. 30. Karnananda, verse 4, quoted in Ramakanta Chakravarti, Vais.n.avism in Bengal, 220. See also Chittaranjan Dasgupta, Bishnupurer Mandir Terakota (Bishnupur: Bangiya Sahitya Parishad, 1980), 69; and Dimock, Place of the Hidden Moon, 201–205. 31. Frédérique Apffel Marglin, “Types of Sexual Union and the Implicit Meanings,” in The Divine Consort: Ra¯dha¯ and the Goddesses of India, ed. John Stratton Hawley and Donna Marie Wulff (Boston: Beacon Press, 1982), 298–315. 32. Ibid., 306. Prema is defined by the devadasis (temple courtesans) of Puri as intercourse characterized by the withholding of semen. It is without consequence since it involves no birth. 33. Jayadeva introduces himself as a poet from the village of Kenduli. On the debate about the location of this village, since there are villages of the same name in both Bengal and Orissa, see The Gitagovinda of Jayadeva: Love Song of the Dark Lord, ed. and trans. Barbara Stoler Miller (Delhi: Motilal Banarsidass, 1984), 4. 34. Wulff, “Radha,” 109. 35. These highly erotic verses are sung at the end of the day, after the evening meal, when Jagannatha is dressed in sringara vesa (erotic clothing). Participants interpret this sequence of ritual activities to allude to sexual union. Hacker, “Dressing the Lord Jagannatha,” 112; Frédérique Apffel Marglin, “Refining the Body: Transformative Emotion in Ritual Dance,” in Divine Passions: The Social Construction of Emotion in India, ed. Owen M. Lynch (Delhi: Oxford University Press, 1990), 225–30. As Barbara Stoler Miller elaborates, “The lyrical techniques of Jayadeva’s songs combine with the conventional language of Sanskrit erotic poetry to express the intimate power of divine love. As Jayadeva elaborates the passion of Ra¯dha¯ and Krishna, he creates an aesthetic atmosphere of erotic mood (´s.r n.ga¯rarasa) that is bliss for devotees of Krishna.” Miller, Love Songs, 14. 36. Devotees may develop individual roles in the divine drama and gradually transcend their material reality and external bodies to inhabit the true or perfected body and identity (siddha-deha). This identity may be inherent, in which case it is recognized by the initiate’s guru, who imparts the techniques to induce

210

•

Notes to pages 000–000 50–55

it to the forefront, or a suitable role may be assigned by the guru in accordance with the initiate’s disposition, to which the initiate then aspires through the training of the guru. The shifting of identity proceeds from the original state, in which the awareness of the perfected form is slight while that of the ordinary self is dominant, to its reversal, when the pure inner self is realized. See David L. Haberman, Acting as a Way of Salvation: A Study of Ra¯ganuga¯ Bhakti Sa¯dhana (New York: Oxford University Press, 1988), 65–81; June McDaniel, The Madness of the Saints: Ecstatic Religion in Bengal (Chicago: University of Chicago Press, 1989). 37. The goal of worship in these Gaudiya temples is an internal transformation that is stimulated and aided by visual as well as mental imagery. See Gary B. Palmer and William R. Jankowiak, “Performance and Imagination: Toward an Anthropology of the Spectacular and the Mundane,” Cultural Anthropology 11.2 (May 1996): 225–58. 38. Padma Kaimal, “Passionate Bodies: Constructions of the Self in South Indian Portraits,” Archives of Asian Art 48 (1995): 9. 39. These four additional shrines, together with the inset scene showing the central upper shrine of the double-storied temple, might constitute an explicit reference to the Shyam Ray Temple’s particularly elaborate upper story with five pavilions. However, it is uncertain if the corner shrines were ever used for ritual. At Vishnupur and at Gokulnagar, where the Gokul Chand Temple is also fiveshrined, there seems to be general agreement that the corner shrines did not have distinctive functions. This would suggest that the corner shrines are primarily an attempt to impart a sense of grandeur to the monument. 40. Walter J. Ong, Orality and Literacy: The Technologizing of the Word (London: Routledge, 1988), 41; Connerton, How Societies Remember, 61. 2. A Paradigm Shift 1. James Fergusson, History of Indian and Eastern Architecture, vol. 2 (1891; reprint, New York: Dodd, Mead, 1899), 81. 2. The terra cotta figural surface ornamentation, Fergusson felt, displayed “an immeasurable inferiority to the carvings on the old temples in Orissa.” In discussing the Kantaji Temple’s upper structure, he did observe that “The centre pavilion is square, and, but for its pointed form, shows clearly enough its descent from the Orissan prototypes; the other eight are octagonal, and must, I fancy, be derived from Mahomedan originals. The pointed arches that prevail throughout are certainly borrowed from that style.” Ibid. 3. Such conceptual frameworks persisted in the early nationalist histories written in independent India, and are being resurrected in the present polarized political climate. For historiographic discussion of this material, see Tapati Guha-Thakurta, The Making of a New Indian Art (Cambridge: Cambridge University Press, 1992); Partha Mitter, Much Maligned Monsters: A History of European Reactions to Indian Art (Oxford: Clarendon Press, 1977; reprint, Chicago: University of Chicago Press, 1992); Nicholas B. Dirks, “Guiltless Spoliations: Picturesque Beauty, Colonial Knowledge, and Colin Mackenzie’s Survey of India” and Gary Michael Tartakov, “Changing Views of India’s Art History,” both in Perceptions of South Asia’s Visual Past, ed. Catherine B. Asher

Notes to to pages 000–000 Notes pages 55–67

•

211

and Thomas R. Metcalf (New Delhi: American Institute of Indian Studies and Swadharma Swarajya Sangha, 1994), 211–32, 15–36; and Arjun Appadurai, “Making the National Geographic: Changing Images of Territory in Colonial India” (paper presented at the Ethnohistory Workshop, University of Pennsylvania, 1995). 4. Benedict Anderson, Imagined Communities: Reflections on the Origin and Spread of Nationalism (London: Verso, 1991); and Bernard S. Cohn, “The Census, Social Structure, and Objectification in South Asia,” Folk 26 (1984): 25–49. 5. Richard Eaton has sensitively explored the complex nature of and motivations for conversion to Islam during this period in his The Rise of Islam and the Bengal Frontier, 1204–1760 (Berkeley and Los Angeles: University of California Press, 1993). Beyond Bengal, recent studies include David Gilmartin and Bruce B. Lawrence, eds., Beyond Turk and Hindu: Shaping Islamicate Identity in South Asia, 1400–1800 (Gainesville: University Press of Florida, 2000); Sila-Khan, Conversions and Shifting Identities; and Harjot Oberoi, The Construction of Religious Boundaries: Culture, Identity, and Diversity in the Sikh Tradition (Delhi: Oxford University Press, 1994). 6. Wilfred Cantwell Smith, The Meaning and End of Religion: A New Approach to the Religious Traditions of Mankind (New York: Macmillan, 1963), 64; and David Lorenzen, “Who Invented Hinduism,” Comparative Studies in Society and History 41.4 (October 1999): 630–59. 7. On the collaboration of South Asians who served as intermediaries and translators, see Rosane Rocher, “British Orientalism in the Eighteenth Century: The Dialectics of Knowledge and Government,” in Orientalism and the Postcolonial Predicament, ed. Carol A. Breckenridge and Peter van der Veer (Philadelphia: University of Pennsylvania Press, 1993), 215–49. 8. Joseph T. O’Connell, “Vais.n.ava Perceptions of Muslims in Sixteenth Century Bengal,” in Islamic Society and Culture: Essays in Honour of Professor Aziz Ahmad, ed. Milton Israel and N. K. Wagle (New Delhi: Manohar, 1983), 294. See also Joseph T. O’Connell, “The Word ‘Hindu’ in Gaud.ı¯ya Vais.n.ava Texts,” Journal of the American Oriental Society 93.3 (1973): 340–44. 9. A. L. Basham, The Wonder That Was India: A Survey of the Indian Subcontinent before the Coming of the Muslims (reprint, New York: Grove Press, 1959), 1. 10. Al Biruni in the eleventh century, for example, employed the terms in this way. Edward C. Sachau, trans. and ed., Alberuni’s India (London: Routledge and Paul Kegan, 1888). For a discussion of the changing meanings of the term “Hindu” in Arabic and Persian literature, see Carl W. Ernst, Mysticism, History, and Sufism in a South Asian Center (Albany: State University of New York Press, 1992), 22–37. 11. Romila Thapar, “Syndicated Moksha?” Seminar 313 (September 1985): 14–22; and Heinrich von Stietencron, “Hinduism: On the Proper Use of a Deceptive Term,” and Robert E. Frykenberg, “The Emergence of Modern ‘Hinduism’ as a Concept and as an Institution: A Reappraisal with Special Reference to South India,” both in Hinduism Reconsidered, ed. Gunther D. Sontheimer and Hermann Kulke (Delhi: Manohar, 1989), 32–53, 82–101. 12. O’Connell, “The Word ‘Hindu,’” 342. 13. Quoted in O’Connell, “Vais.n.ava Perceptions of Muslims,” 303.

212

•

Notes to pages 000–000 67–69

14. E. A. Gait, Report: Census of India, 1901, vol. 6, part 1 (Calcutta: Bengal Secretariat Press, 1902), 171. 15. J. D. Beglar, Report of a Tour through the Bengal Provinces of Patna, Gaya, Mongir, and Bhagalpur; the Santhal Parganas, Manbhum, Singhbhum, and Birbhum; Bankura, Raniganj, Bardwan, and Hughli in 1872–73, Archaeological Survey of India Reports, old series, vol. 8 (Calcutta, 1878; reprint, Varanasi: Indological Book House, 1966), 203. 16. For recent scholarship that has begun to examine these complexities, see Catherine B. Asher, “The Architecture of Raja Man Singh: A Study of Subimperial Patronage,” in The Powers of Art: Patronage in Indian Culture, ed. Barbara Stoler Miller (New Delhi: Oxford University Press, 1992), 183–202; Asher, “Authority, Victory, and Commemoration: The Temples of Raja Man Singh,” Journal of Vaisnava Studies 3.3 (summer 1995): 25–36; and Michael W. Meister, “The Two-and-a-Half-Day Mosque,” Oriental Art, new series 18.1 (spring 1972): 57–63. 17. Lembrança dalgumas coussas que se passaram quando Antonio de Bryto e Dyogo Pereyra foram a Bemgalla asy em Bengala como em Tanaçaiym e em Pegu onde tambem fomos, 11:47–88 (São Vicente), quoted in Smith, First Age, 91. For a Ming ambassador’s description of the earlier Sultanate palace at Pandua, see W. W. Rockhill, “Notes on the Relations and Trade of China with the Eastern Archipelago and the Coast of the Indian Ocean during the Fourteenth Century,” T’oung Pao 16.2 (1915): 441–42. 18. The absence of these grand palaces at the capitals of Lakhnauti, Pandua, and Gaur suggests that buildings associated explicitly with political authority were the primary targets of competing groups who came to power during the Sultanate period, rather than the mosques, shrines, and temples, contrary to what modern scholarship has assumed. 19. Bandyopadhyay, Temples of Birbhum, 126. 20. Percy Brown, Indian Architecture, vol. 1, Buddhist and Hindu Period (1942; reprint, Bombay: D. B. Taraporevala Sons, 1965), 149. 21. These temples only entered discussions of Mughal architecture in the late twentieth century. Catherine B. Asher, Architecture of Mughal India (Cambridge: Cambridge University Press, 1992). 22. Ralph Pinder-Wilson, “Stone Sculptures of Gaur,” in Indian Art Connoisseurship: Essays in Honor of Douglas Barrett, ed. John Guy (Delhi: Indira Gandhi National Centre for Arts and Mapin, 1996), 251–61; and Naseem Ahmed Banerji, “The Architecture and Architectural Decoration of the Adina Mosque, Pandua, West Bengal, India: The Problem of the Conjoined Buddhist, Hindu, and Islamic Motifs in the Mihrab Niches” (Ph.D. diss., University of Iowa, 1993). 23. Perween Hasan, “Sultanate Mosques and Continuity in Bengal Architecture,” Muqarnas 6 (1989): 68–73; P. Hasan, “The Footprint of the Prophet,” Muqarnas 10 (1993): 335–47; and P. Hasan, “Temple Niches and Mihrabs in Bengal,” in Islam and Indian Regions, ed. Anna Libera Dallapiccola and Stephanie Zingel-Avé Lallemant (Stuttgart: Franz Steiner Verlag, 1993), 86–94. 24. David McCutchion, “Hindu Muslim Artistic Continuities,” in The Islamic Heritage of Bengal, ed. George Michell (Paris: UNESCO, 1984), 213–30. 25. Chakrabarty, Kavikankan Chandi, 343–61. 26. Edward C. Dimock, “Muslim Vaisnava Poets of Bengal,” in The Sound

Notes to to pages 000–000 Notes pages 70–74

•

213

of Silent Guns and Other Essays (London: Oxford University Press, 1989), 23– 32. 27. Lee Horne, “Making in Metal,” in Cooking for the Gods: The Art of Home Ritual in Bengal, ed. Michael W. Meister, cat. Pika Ghosh (Newark, N.J.: The Newark Museum; Philadelphia: University of Pennsylvania Press, 1995), 32–38. 28. For the dating of these monuments I have relied upon Catherine B. Asher, “Inventory of Key Monuments,” in Islamic Heritage of Bengal, ed. Michell, 37– 140. 29. Perween Hasan has undertaken the most significant recent work on Sultanate mosques and tombs, “Sultanate Mosque-Types in Bangladesh: Origins and Development” (Ph.D. diss., Harvard University, 1984). She has defined the elements of Bengal Sultanate style and established its relationship to hut and older Pala-period architecture. An earlier study was conducted by Ahmad Hasan Dani: Muslim Architecture in Bengal (Dacca: Asiatic Society of Pakistan, 1961). 30. P. Hasan, “Sultanate Mosques,” 72. 31. No systematic history of premodern temple or mosque destruction on religious grounds can be gleaned from either literary accounts or the standing monuments of Bengal. While the 1643 Kabilaspur temple curses lowly or ignorant (avarna) folk with consuming beef, and even Yavanas, a king, with consuming pork as retribution for temple destruction, these statements must be balanced against a history of actual practice. Alauddin Husain Shah’s demolition of temples in Puri, mentioned in the Madala Panji, and in Navadvip, noted by Jayananda in his Chaitanya Mangal, for example, were political acts directed at Prataparudra and the state cult at Puri and the Brahmans of Navadvip, who were rumored to have planned to unseat the sultan. On the complexities of this issue, see Richard H. Davis, Lives of Indian Images (Princeton, N.J.: Princeton University Press, 1997); and Richard M. Eaton, “Temple Desecration and Indo-Muslim States,” in Beyond Turk and Hindu, ed. Gilmartin and Lawrence, 246–81. 32. The shrine of Pir Shah Kauban Ali is perhaps the best-known dargah in Vishnupur. 33. This stone tablet, dated July 29, 1221, is probably the earliest Muslim inscription in Bengal. Z. A. Desai, “An Early Thirteenth-Century Inscription from West Bengal,” Epigraphia Indica, Arabic and Persian Supplement (1975): 6–12. 34. Kalna was a flourishing river-port in the fifteenth and sixteenth centuries. Chaitanya broke his journey here, and his enshrined oar is the focus of a local Vaishnava cult. In the eighteenth century the town became a center of the rajas of Burdwan, who built a series of temples. 35. Mirza Nathan, Baharistan-i-Ghaybi: A History of the Mughal Wars in Assam, Cooch Behar, Bengal, Bihar, and Orissa during the Reigns of Jahangir and Shahjahan, vol. 1, trans. M. I. Borah (Gauhati, Assam: Dept. of Historical and Antiquarian Studies, Narayani Handiqui Historical Institute, 1936), 238. 36. Beglar, Report of a Tour, 146. 37. The inscription on the Lalji Temple at Raghunathgarh Thakurbati, Chandrakona, for example, identifies its patron Sri Lakshmanavati as the sister of the Malla king. See Bhattacharyya, Corpus of Dedicatory Inscriptions, 80. 38. This preference for large numbers of smaller structures continues into the later eighteenth century, after the collapse of Mughal authority, when patron-

214

•

Notes to pages 000–000 74–77

age shifts to merchants enriched by the European trade. As Hitesh Ranjan Sanyal points out, even one successful merchant or a wealthy subcaste such as the oil pressers could get together to set up their own temple as an attempt at upward social mobility. “Regional Religious Architecture in Bengal: A Study in the Sources of Origin and Character,” Marg 27.2 (March 1974): 31–43. 39. P. Hasan, “Sultanate Mosques,” 68–73. 40. On Nagara temple organization, see Michael W. Meister, “The Hindu Temple: Axis and Access,” in Concepts of Space, Ancient and Modern, ed. Kapila Vatsyayan (New Delhi: Indira Gandhi National Centre for Arts and Abhinav, 1991), 269–80. 41. Note that Perween Hasan has pointed to the individual thatched hut within the walled compound as the source for the single-domed mosque unit within the complex. “Sultanate Mosques,” 68–73. 42. This six-armed form of Vishnu bears Rama’s bow and arrow in the upper pair of hands, raises Krishna’s flute to his lips with the second pair, and displays the ascetic’s water pot and rudraksha bead necklace in the hands below. The paradigmatic ascetic in the Gaudiya tradition is of course Chaitanya, who renounced his family and worldly ties when he left Nabadwip. This new iconography can be understood as a visual reiteration of the goswamis’ proposition that Chaitanya was Krishna’s reincarnation. 43. One can walk all the way around the deserted sanctum in the porches. At the corners where these long corridors meet are four small square chambers. The two on either side of the main entrance have indented niches, some bearing deity images molded permanently in terra cotta, suggesting they may once have functioned as subsidiary shrines. 44. P. Hasan, “Sultanate Mosques,” 68–73. 45. The entrance arches of the east porch of Sultanate mosques have corresponding mihrab niches on the qibla wall on the west. While this feature could not be duplicated in Vaishnava temples, it appears to have been adapted so that the triple-arched entrance is aligned with a single-arched entrance from the porch into the sanctum, which in turn directs the viewer’s gaze to the altar located at the end of the sanctum. However, such axial arrangement of doorways to the altar is hardly new in temple construction at this time. Both Sultanate mosques and Ratna temples find a precedent in Nagara temples sponsored by the Pala and Sena dynasties. See P. Hasan, “Temple Niches.” 46. P. Hasan, “Sultanate Mosques,” 68–73. 47. On Sultanate architecture of Jaunpur, see Alois Anton Fuhrer, The Sharqi Architecture of Jaunpur (1889; reprint, Calcutta: Superintendent of Government Printing, 1994). 48. Eaton, Rise of Islam, 71–94. 49. The successful transformation of the thatch eave into brick seems to have motivated its dissemination to other building-types and regions, layering further meanings upon the form. That elegant curve of Bengal’s huts, monumentalized in brick, was imported from this peripheral province to the heart of the Mughal empire, where it was reproduced in white marble for palatial structures for key political rituals such as Shah Jahan’s Bangala-i Darshan at Agra, and the throne in the Diwan-i-Am of the Delhi Red Fort. The integration of this exotic element into imperial architecture unmistakably signified the final successful absorption of the rebellious province. The accruing significance of the chala as signifier of

Notes to to pages 000–000 Notes pages 78–82

•

215

region was thus reinforced by such flamboyant espousal beyond Bengal. This bangala, as it was called in Mughal literature, reflecting its regional distinctiveness, was then transported to Rajasthan for Rajput palaces and ultimately brought back to Bengal in this particular reincarnation by Marwari merchants in the overhanging balconies of their eighteenth-century mansions in north Calcutta. This highly stylized version also ornamented the form of emergent nationalistic Bengali cultural institutions as well as the British bungalow, giving the colonial structure local reference. 50. Heavy fortress-like mosques and mausoleums have a long history in South Asia. Rusticated and battered structures like the tomb of Ghiyasuddin Tughluq in Delhi indicate the preference for a fortified façade. In neighboring Jaunpur, monumental mosques such as the Atala Masjid have a heavily fortified appearance. This inclination probably reflects one strand in the development of mosque architecture from ribats, places of prayer in military encampments constructed at frontier zones during the early spreading of Islam. 51. Broad three- and five-lobed arches framing Pala figural sculpture suggest sources in the buildings of the eighth through twelfth centuries. See McCutchion, “Artistic Continuities,” 222. Naseem Ahmed Banerji corroborates his comparisons in her examination of the Adina Masjid’s mihrabs, “Architectural Decoration.” 52. McCutchion compared the east façade of the Qadam Rasul with the Madan Mohan Temple at Vishnupur in “Artistic Continuities,” 217. 53. Ibid. 54. The pendentives that typically effect the transition from the square walls of the sanctum to the circular base of the dome are also downplayed through terra cotta ornamentation. The complex patterns of these forms and of the terra cotta ornamentation highlighting their surface indicate that this transition zone was another area for experimentation. One of the most elaborate is the Shyam Ray Temple’s combination of squinches, which creates the appearance of an open ring of lotus petals around the central dome, filled with rows of terra cotta dancers and drummers. 55. In the southwestern part of West Bengal, particularly in the district of Bankura, a local laterite, which is dark red and coarse-grained, is also used for temple construction, as well as brick. However, this stone is treated as if it were brick, laid in horizontal courses and plastered with sand and lime. Laterite panels imitate the terra cotta figures and floral ornamentation on brick. In color, too, this stone matches the local brick, and laterite is used to build the platform on which brick temples stand. 56. Oral communications from Shankar Das, attendant at the Shyam Ray Temple, Vishnupur, from conservation workers employed by the Archaeological Survey of India at this temple, and from terra cotta artisans at Panchmura, Bankura district (December 1996, January 1997). See also George Michell, “Materials and Techniques,” in Brick Temples of Bengal, ed. Michell, 63–67. 57. The sizes of bricks vary both between seventeenth- and nineteenthcentury monuments and within each monument, in the construction of its different parts. The careful fitting of bricks in transitions and curved surfaces such as arch voussoirs suggests that while some standard shapes were used in bulk, awkward junctures were handled individually. 58. Stone is rare in the area, and probably had to be transported from

216

•

82–83 Notes to pages 000–000

southern Bihar. The source of this dark phyllite, used prolifically for Pala sculpture, has been discussed by Frederick M. Asher in “Stone and the Production of Images,” East and West 48.3–4 (December 1998): 313–28. 59. M. R. Tarafdar, Husain Shahi Bengal, 1494–1538 A.D.: A Socio-political Study (Dhaka: Asiatic Society of Pakistan, 1965); and Eaton, Rise of Islam, 63–69. 60. Jadunath Sarkar, ed., The History of Bengal, vol. 2, Muslim Period, 1200–1757 (Dacca: University of Dacca, 1947; reprint, Patna: Janaki Prakashan, 1977), 151–52. 61. Raja Ganesh, for example, seized the throne of Gaur in 1400. However, upon vociferous protest from the conservative ruling elite, he abdicated, and only by converting his son Jadu to Islam and placing him in the care of the local Chisti saints was he able to preserve control. As Jalaluddin, his son was able to retain authority, carefully balancing the orthodoxy of the court with the interests of the subject Bengali population. 62. Asim Roy, The Islamic Syncretistic Tradition in Bengal (Princeton, N.J.: Princeton University Press, 1983), 58. 63. Roy, Islamic Syncretistic Tradition, 6; Eaton, Rise of Islam; and Tarafdar, Husain Shahi Bengal. 64. In fact, aside from the Srikrishna Kirtana, most of the major Vaishnava works of the period refer to either the kings or other Muslim officials as patrons. Following the sultans, their governors of Chittagong, Paragal Khan and his son Chhuti Khan, sponsored the Bengali translation of the Mahabharata by Kavindra Parameshvar and Shrikar Nandi. Tarafdar, Husain Shahi Bengal, 239–51. 65. Roy, Islamic Syncretistic Tradition, 95. Sila-Khan has pointed to a similar trend in the ginanic literature of the Ismaili tradition in western India: Conversions and Shifting Identities, 46–61. 66. Joseph T. O’Connell, “The Social Implications of the Gaudiya Vaishnava Movement” (Ph.D. diss., Harvard University, 1970), 352–63. 67. Bimanbehari Majumdar, Govindadaser Padavali o tahar Yug (Calcutta: Calcutta University, 1961), 403. 68. Khwajah Nizamuddin Ahmad, Tabaqat-iAkbari, ed. Brajendranath De and M. Hidayat Hosein, trans. Brajendranath De, 3 vols. (Calcutta: Bibliotheca Indica, 1927–39), pt. 1, 443, quoted in Eaton, Rise of Islam, 63. 69. Earlier, Ibn Battuta had noted the same prosperity: “Bengal is a vast country and abounds in rice. In the whole world I did not see a country where commodities were cheaper than in Bengal.” Mahdi Husain, trans. and comm., The Reh.la of Ibn Bat. .t u¯.t a (Baroda: Oriental Institute, 1953), 234. 70. “Notices of the City of Gaur from an Anonymous Manuscript Preserved in the Torre do Tombo, Lisbon, in the Collection Entitled Sam Vicente,” quoted in Smith, First Age, 134. 71. Ibid., 136. 72. Phillip B. Wagoner, “Harihara, Bukka, and the Sultan: The Delhi Sultanate in the Political Imagination of Vijayanagara,” in Beyond Turk and Hindu, ed. Gilmartin and Lawrence, 316; and Wagoner, “‘Sultan among Hindu Kings’: Dress, Titles, and the Islamicization of Hindu Culture at Vijayanagara,” Journal of Asian Studies 55.4 (November 1996): 851–80. 73. João de Barros, Da Asia, decade 4, book 9, chapter 1, quoted in Smith, First Age, 130.

Notes to pages 000–000 Notes to pages 84–89

•

217

74. Lembranòa dalgumas, 47–88, quoted in Smith, First Age, 91. 75. P. C. Bagchi, “Political Relations between Bengal and China in the Pathan Period,” Visva-Bharati Annals 1 (1945): 96–134. 76. J. Sarkar, Muslim Period, 193. 77. Nalinikanta Bhattasali attempted to identify the twelve chieftains; “Bengal Chiefs’ Struggle for Independence in the Reigns of Akbar and Jahangir,” Bengal Past and Present 35 (January–June 1928): 30–36. 78. Quoted in Eaton, Rise of Islam, 147. 79. Abul Fazl, Akbarnamah, vol. 2, trans. H. Beveridge (Calcutta: Royal Asiatic Society of Bengal, 1934; reprint, Delhi: Rare Books, 1972–73), 878. 80. Nathan, Baharistan-i-Ghaybi, 18. 81. Ibid., 19. 82. Local rulers also seem to have provided asylum to rebels fleeing from the Mughals. When Kar Talab Khan, diwan of Orissa, made his escape from the pursuing forces of Aurangzeb, for example, he chose to make his way through Vishnupur. Ghulam Hussain Salim, Riazu-s-Salatin, trans. M. A. Salam (1903; reprint, Delhi: Idarah-I Adabiyat-I Delli, 1975), 251. 83. Abdul Karim, Murshid Quli Khan and His Times (Dacca: Asiatic Society of Pakistan, 1963), 83; see also O’Malley, Bengal District Gazetteers, 30. 84. Described in Ratan Kaviraj, Madan Mohan Bandana, MS. 324 (Vishnupur: Vishnupur Shahitya Parishad). 85. Architectural interaction between Mughal monuments and Ratna temples in the eighteenth century points to the consolidation of Mughal political authority and the local landholders’ reconciliation to their new overlords. The Madan Mohan Temple displays the merlons of the province’s Mughal mosques; the Radha Shyam flaunts Shahjahani baluster columns, arches, and floral decoration like that of the Taj Mahal. A visual vocabulary developed for the imperial ritual of darshan was now transferred back to temples for seeing the divine. 3. Acts of Accommodation 1. The paired Kali and Shiva Temples (1838–43) at Sonarang Tangibari, Dhaka district, use Gothic-style spires for their towers. On these temples, see George Michell, “Neo-classicism in Bengali Temple Architecture,” Art and Archaeology Research Papers 11 (June 1977): 28–31. The Hamsheshvari Temple (1799) at Bansberia, Hooghly district, has pointed spires with scalloped stucco decoration, also inspired by European contact. 2. Fergusson, History of Indian and Eastern Architecture, 2:81. 3. Robert Montgomery Martin, ed., The History, Antiquities, Topography, and Statistics of Eastern India (London: W. H. Allen, 1838; reprint, Delhi: Cosmo, 1976), 925. 4. E. B. Havell outlined a linear development for Indian architecture from more fragile bamboo origins to wood and finally more permanent materials such as stone and brick. In this developmental sequence he saw Bengali brick temples as exceptional in South Asia for preserving a direct connection to the curves produced by the bamboo originals and their replication in the vaulted roofs of an “early Magadhan style” of architecture patronized under Ashoka Maurya, represented in the sculptures of Bharhut and Sanchi. It is as historical artifacts, he argued, that these temples were important: “The modern Bengali style of temple, so far from belonging to what Fergusson calls an ‘aberrant type,’ is the 218

•

89–110 Notes to pages 000–000

lineal descendant of the early Magadhan style.” Havell, Indian Architecture: Its Psychology, Structure, and History from the First Muhammadan Invasion to the Present Day (London: John Murray, 1913), 92. On the other hand, Monmohan Chakravarti, among the earliest Bengali scholars of these monuments, posited a linear development from single-storied structures with hut roofs, upon which other forms were gradually placed: “Further development is marked by the substitution either of a spired tower or of a spired duplicate on the roof.” He then sketched the progression of these types: “Single towers gradually developed into multi-towers,” ranging from five to twenty-five. He confesses, however, “the intervening steps, if any, cannot be traced, in the existing remains as the oldest are the multi-towered.” Chakravarti, “Bengal Temples and their General Characteristics,” Journal of the Asiatic Society of Bengal 5 (May 1909): 141–62. 5. McCutchion, Late Mediaeval Temples, 11. 6. Ibid. 7. I date this monument by comparing the superstructure to that of the 1461 Nagara temples at Barakar, Burdwan district. The Radha Damodara Temple shares with them a tall profile, with a sharp inward curve at the very top. Bhumi amalakas mark the corners, although there are no repeated horizontal moldings along the length of the tower. Their somewhat arbitrary location suggests that while this Nagara feature is remembered, its symbolic function may no longer have been understood. Rampant lions spring from the central offsets on each side as on the older temples, but they are now situated far higher than before and no longer serve the same symbolic purpose. By the sixteenth century, many of its Nagara features therefore no longer fit within the Nagara configuration. 8. See Michell, Brick Temples of Bengal, pls. 322–26. 9. Ibid., pl. 17. 10. T. S. Bloch, Superintendent of the Archaeological Survey of India, offered this structure as an example of the Nagara temple in “The Temples of Vishnupur,” Annual Report of the Archaeological Survey of India, 1903–1904 (1906): 49–53. He classified the monuments solely on the basis of their variety of superstructures: a single square tower (Nagara); the single tower resting on a larger square building with a curved Bengali roof (that is, the Ratna temple with a single upper structure); five towers on the same building-type (pancharatna); and the jor bangla (two buildings shaped like rectangular huts sharing a common long wall). Likewise, David McCutchion classified this temple as Nagara in The Temples of Bankura District (Calcutta: Writers Workshop, 1972), 17. 11. Etchings by the Belgian artist Balthazar Solvyns at the end of the eighteenth century also record various forms of temples co-existing at that time. Solvyns, A Collection of Two Hundred and Fifty Colored Etchings: Descriptive of the Manners, Customs, Character, Dress, and Religious Ceremonies of the Hindoos (Calcutta: Mirror Press, 1799). 12. This is not to suggest, though, that the multiple-shrined version developed before the single-shrined ones, as David McCutchion proposed. The many piles of rubble and ruined temples in the region are a poignant reminder of the incompleteness of the existing archaeological record, making discussion about origins and developments inconclusive. 13. Only one temple with nine upper shrines was built in the Vishnupur area in the nineteenth century. The Sridhar Temple, however, was not a royal Malla temple, but built by a local merchant family. 14. Numerous scholars have noted parallels between Islamic mysticism and Notes Notes to to pages pages 000–000 110–118

•

219

the emotional philosophy and practice of Gaudiya Vaishnavism. M. R. Tarafdar posited a causal relation, tracing the influence of Sufism on Gaudiya Vaishnavism in Husain Shahi Bengal, 225–26. See also Nicholas, “Vaisnavism and Islam.” 15. P. Hasan, “Footprint of the Prophet,” 338–39. 16. At Vishnupur two pairs of Vaishnava Nagara temples were built to Chaitanya and Nityananda and to Krishna and Balarama in the durbar area. Nagara temples were also dedicated to Krishna, such as Shyam Chand at Dharapat (1693/1704) and Gopal at Bikrampur (1664). 17. Reverend W. Ward, Account of the Writings, Religion, and Manners, of the Hindoos: Including Translations from Their Principal Works (Serampore: Mission Press, 1811), 3:345–60. He describes the Ratna typology in these terms: “The temple called Punchu-rutnu has a single arched roof, with a large pinnacle or turret on the dome, and a smaller one on each corner . . . [The navaratna form has] a double roof . . . , with a small turret on each corner of the lower roof, and the same on each corner of the upper one, and a larger turret to crown the dome.” He thus differentiates Ratna temples from Nagara on the basis of the arched roof of the square base and their different superstructures. 18. Joseph T. O’Connell, Religious Movements and Social Structure: The Case of Chaitanya’s Vais.n.avas in Bengal (Shimla: Indian Institute of Advanced Study, 1993), 23. 19. On the Mahavidyas, see David Kinsley, The Ten Mahavidyas: Tantric Visions of the Divine Feminine (Berkeley and Los Angeles: University of California Press, 1997; reprint, Delhi: Motilal Banarsidass, 1998). These goddesses may also relate to the mother goddesses (matrikas), usually a set of seven, who are often depicted with Kartikeya or Ganesha. 20. I have discussed this narrative further in “Unrolling a Narrative Scroll: Artistic Practice and Identity in Late-Nineteenth-Century Bengal,” Journal of Asian Studies 62.3 (August 2003): 835–71. 21. C. Mackenzie Brown, “The Theology of Ra¯dha¯ in the Pura¯n.as,” in Divine Consort, ed. Hawley and Wulff, 57–71; and Yarina Liston, “The Ra¯dha¯ Cycle: From Lila¯ to Bhakti,” Journal of Vais.n.ava Studies 8.2 (spring 2000): 7– 26. Correspondingly, the goddesses are shaped by Vaishnava conceptions; see Rachel Fell McDermott, “The Vaisnavized Uma of Bengali Devotionalism,” Journal of Vais.n.ava Studies 8.2 (spring 2000): 105–24. 22. Tripurasundari created Radha, and Kali informed Krishna that she was his shakti and commanded him to perform ritual sex with her. June McDaniel, “The Tantric Radha: Some Controversies about the Nature of Radha in Bengali Vaishnavism and the Radha Tantra,” Journal of Vais.n.ava Studies 8.2 (spring 2000): 131–46; Chakravarti, Vais.n.avism in Bengal, 312. 23. Charlotte Vaudeville, “Krishna Gopa¯la, Ra¯dha¯, and the Great Goddess,” in Divine Consort, ed. Hawley and Wulff, 1–12. Jagannatha is himself worshipped as a form of Kali by Shakta devotees; see Marglin, “Types of Sexual Union,” 302. 24. Bar.u Cand.ida¯sa, Singing the Glory of Lord Krishna: The S´rı¯kr.s.n. akı¯rtana, trans. and annot. Miriam H. Klaiman (Chico, Calif.: Scholars Press, 1984), 11– 17. Chandidas describes how the goddess instructed him to choose Rami, a lowcaste sweeper at the temple, as his companion in sahaja worship practices. 25. Chaitanya is said to have brought texts from his pilgrimage to south India that discuss yantras and the metaphysical dimensions of sexual union of

220

•

Notes to pages 000–000 119–123

the lingam and yoni. Dimock, Place of the Hidden Moon, 56–57. The Vrindavan goswamis’ texts, which laid out the theological and ritual underpinnings of the movement, prescribe initiation rites around a pit in the shape of a yoni, the form associated with the goddesses. Tapan Kumar Raychaudhuri, Bengal under Akbar and Jahangir: An Introductory Study in Social History (Calcutta: A. Mukherjee, 1953; reprint, Delhi: Munshiram Manoharlal, 1966), 120–21, 129. 26. On Durga Puja at Vishnupur, see Akos Ostor, The Play of the Gods: Locality, Ideology, Structure, and the Festivals of a Bengali Town (Chicago: University of Chicago Press, 1980). 27. Guruprasad Sarkar, Mallasamskritira Patabhumikaya Bishnupurera Rasotsaba (Bishnupur: The New Minerva Press, 1980), 3, 65. 4. Axes and the Mediation of Worship 1. I thank Michael Meister for the discussions in Vishnupur and Philadelphia in December 1998 that helped me recognize the ways in which the architecture privileges both axes. 2. While most natmandir are single-storied, that of the 1660 Shyam Chand Temple at Vishnupur is double-storied, suggesting that the upper and lower levels provided separate spaces for daily and festival performance of kirtan. 3. However, this particular arrangement of images is not consistent throughout the living temples of Vishnupur. The Madan Gopal Temple offers an alternate placement of its small, portable, metal images. Here the images of Radha and Krishna and the shalagram occupy the altar at the end of the north-south axis, while smaller images of Chaitanya and his companion Nityananda preside in the two niches of the west wall. These variations in the location of images for ritual practice are ultimately insignificant, as Chaitanya is understood as an incarnation of Krishna. 4. Dimock and Levertov, In Praise of Krishna, xix. 5. See Stanley Jeyaraja Tambiah, “A Performative Approach to Ritual,” in Culture, Thought, and Social Action: An Anthropological Perspective (Cambridge, Mass.: Harvard University Press, 1985), 128; and John L. Austin, How to Do Things with Words (Oxford: Clarendon, 1962). Austin uses the term “speech acts” for words that by their very utterance execute deeds; in this case the words of the songs re-create the lila. 6. Pierre Bourdieu qualifies: “The performative magic of ritual functions fully only as long as the religious official who is responsible for carrying it out in the name of the group acts as a kind of medium between the group and itself: it is the group which, through its intermediary, exercises on itself the magical efficacy.” Bourdieu, Language and Symbolic Power, trans. Gino Raymond and Matthew Adamson (1991; reprint, Cambridge, Mass.: Harvard University Press, 1994), 116. 7. Dimock and Levertov, In Praise of Krishna, 55. 8. Gitagovinda 12.16–17, quoted in Miller, The Gitagovinda of Jayadeva, 124. 9. Connerton, How Societies Remember, chapter 3. 10. Ibid., 70. 11. Central panels along the lower segment of each pillar and the walls on either side of the entrance depict popular Krishnalila scenes. The two pillars

123–145 Notes to pages 000–000

•

221

depict the navanarigunjara (a fantastic elephant composed of gopis) and the vastraharana (Krishna sitting atop a tall tree, withholding the clothes of the bathing gopis). 12. These figural terra cotta panels bear some general similarities to the decorative panels found on Pala-period Buddhist viharas. Even though the panels at Paharpur are far larger, the initial impression of a sequential arrangement of panels is not dissimilar. Each panel is a self-contained unit set within bands of geometric or foliate pattern, as in the Gaudiya Vaishnava temples. There seems, however, to be no direct linkage between the seventeenth-century terra cotta panels and this earlier tradition. Details of both the enframed images and framing patterns differ according to the aesthetic preferences of the historical period and theological frameworks of patron religious communities. Their differences in style are as significant as any shared formulae. Whereas the panels of the early monasteries contain single figures, possibly without iconographic sequence, the later temple panels often incorporate narratives and are grouped in sets reflecting iconographic relationships. Further, the sculpted images from the early Buddhist sites are far less refined. The figures are rudimentary and bold, without the elaborate delineation of physical features, clothing, jewelry, and setting found on seventeenth-century temples. In their more delicate treatment, particularly in their vegetal ornament, the temple panels draw more directly upon sixteenth-century craft practices discussed below. 13. Contemporary Baluchari saris woven in Vishnupur demonstrate a similar love for a grid-like surface, although they restrict the grid to the borders and the end section. Their resemblance to the temple decoration and the persistence of the grid are particularly striking when the weavers select motifs such as the flute-playing Krishna and the drummers and dancers, which are so popular in the town’s terra cotta. 14. A gradual flattening of the wall can be observed in comparing the Ratneshvar and the Malleshwar with the later two-storied Jadab Ray and Shyam Ray Temples. 15. Terra cotta decoration is used sparingly on mosques from the late fifteenth and sixteenth centuries, relieving the uniform brick surface. It is concentrated in symbolically potent areas, such as the mihrab niche, the heavy corner turrets, and panels between the entrance arches. 16. This transformation in aesthetic coincides with variation in technique. On the walls of late Sultanate mosques such as Bagha, a complete decorative motif is a fairly large unit, spread over several bricks, while in the seventeenthcentury temples the individual terra cotta panel is often a self-contained unit. 17. The skeletal system of most chala huts consists of sheets of woven bamboo, matted hay, and reeds, strengthened by a grid-like frame of split bamboos. These frames and the inset woven panels reveal a wide variety of patterns, depending on local construction techniques. Today this type of construction can be seen in southeastern Bangladesh, where it is left exposed rather than covered with mud. This practice is probably popular here because the terrain is more watery and the land level is lower than in the rest of Bengal. A mud surface would simply wash away during the heavy monsoon rains. In Comilla district, for example, I observed the use of a second inner band to seal each woven panel or sheet within the grid. Such double and triple framing of square panels is reminiscent of the careful organization of the surface of the

222

•

Notes to pages 000–000 145–146

Shyam Ray Temple, where several frames enclose the figures. Like the rhythmic projection and recession of the temple’s gridded surface, the hut’s doubled frames vary in their projection, so that the structural bamboo grid projects farthest, the inner frame less, and the inset woven sheet the least. Like these bamboo framing strips, those of temples and mosques are woven into distinct patterns, coordinated throughout the façade. 18. The lotus blossoms in the spandrels of mihrab arches, for example, are abstractions of the more naturalistic, large lotuses of the Pala period. A comparison with Pala representations of Surya is particularly appropriate because lotuses rise from his hands to the upper corners of the stele, as they do on the spandrels of mihrab and Gaudiya temple doorway arches. The Philadelphia Museum of Art’s Sun God, for instance, holds ripe flowers, displaying their rings of petals and their seedpods. Incised lines articulate the fine veins of petals and the seeds. The stalk bears leaves and buds. The Goaldi Mosque’s central stone mihrab, one of the few original Sultanate features of this monument, is ornamented with a pair of many-petaled lotuses that closely resemble Pala depictions. The flowers retain their stalk and leaves, as in Surya images, in contrast with later, more abstract mihrab lotuses. The position of the lotuses also suggests that the mihrab niche itself may have been conceived by elimination of a standing figure like that of Surya. 19. The paired lion (sardula) heads in the rectangular frame around the mihrab niche and temple door also have sources in Pala steles. Hindu and Buddhist deity sculptures often have small rampant lions on either side of the central figure. In Ratna doorways, these animals are paired so that they spring in opposite directions, one upward and the other downward, and the pairs are repeated to create a decorative pattern. These abstracted representations of the sardula are then further overlaid with vines emerging from a central waterpot. The Sultanate mosque at Ashtagram seems to be the only one to employ such animal motifs. 20. On axial correspondences between doorways, mihrabs, and Pala-period altars, see Hasan, “Temple Niches,” 87–94. 21. Caitanya Carita¯mr. ta 2.21.44. 22. Bhagavata Purana, trans. and annot. Ganesh Vasudeo Tagare, vol. 10 of Ancient Indian Tradition and Mythology, ed. J. L. Shastri (1978; reprint, Delhi: Motilal Banarsidass, 1994), 10.33.11–16. 23. On the emergence of Radha, see Shashibhushan Dasgupta, Sri Radhar Kramabikas (Calcutta: Firma K. L. Mukhopadhyay, 1952); and De, Early History, 7. 24. Graham Schweig, “Ra¯salı¯la as a Portrait of Ra¯dha¯ and Krishna,” Journal of Vais.n.ava Studies 8.2 (spring 2000): 43–69. 25. Ibid. (Schweig’s translation), 53. 26. Jiva Goswami, for example, points to the source of Radha’s name in the phrase anaya¯ra¯dhitah. Cited in ibid., 56. 27. Caitanya Carita¯mr.ta, 3.14.15–19. 28. Ibid., 2.8.126. 29. Ibid., 2.8.135. 30. John Seyller, “Rasalila,” in Intimate Worlds: Indian Paintings from the Alvin O. Bellak Collection, ed. Darielle Mason (Philadelphia: Philadelphia Museum of Art, 2001), 46. 31. Bhagavata Purana 10.32.20, quoted in Barbara Stoler Miller, “The

Notes to to pages pages 000–000 147–152 Notes

•

223

Divine Duality of Ra¯dha¯ and Krishna,” in Divine Consort, ed. Hawley and Wulff, 13–26. 32. Gitagovinda 1.4, quoted in Miller, The Gitagovinda of Jayadeva, 69. 33. On the role of mnemonics in primarily oral cultures, see Ong, Orality and Literacy, 34. 34. David McCutchion has identified this vertical sequence. McCutchion, “Krishnalila on the Temples of Bengal,” Journal of the Indian Society of Oriental Art, new series 7 (1975–76): 33–51. 35. Later temples have only their south façade adorned in terra cotta. They combine the Keshta Ray’s narrative with the Shyam Ray’s more iconic mode of decoration to highlight the bhakti axis. 36. See Monika Thiel-Horstmann, “Bhakti and Monasticism,” in Hinduism Reconsidered, ed. Sontheimer and Kulke, 127–40. 37. The seventeenth-century Radha Ballabh Temple at Krishnanagar has a similar isolated porch at the back. Although it is today used to store fodder for the caretaker’s cattle, its seclusion and darkness again suggest a space for private activity. 38. Personal communication from Sushanta Mukhopadhyay, Vishnupur, December 1995. 39. Later temples in other locations, such as the Kantaji Temple at Kantanagar, Dinajpur district, have a series of porches. The Radha Govinda Mandir at Atpur (1786) has three successive fully enclosed porches that are wide enough to accommodate large numbers of people. Today this area takes the place of the courtyard as the site of kirtan. Such experiments suggest that other regions expanded porches to create public spaces analogous to the spacious covered porches of thakur dalans of contemporary urban mansions in north Calcutta. Epilogue 1. I synthesize and summarize this account from several versions that I collected from Sushanta Mukhopadhyay (priest of the Madan Mohan temple), Mathur Majhi (caretaker of the Radha Shyam Temple), Banshi Kar Mukhya, Chhaya Das, Tara Banerjee, Chandi Chakraborty, and Sukamal Goswami of Nakaijuri during May and June 2002. 2. Narahari Chakravarti, Bhaktiratnakar, ed. R. Vidyaratna, 3rd ed. (Murshidabad, 1332 BS.), 7.476. 3. Bhagavata Purana, trans. Tagare, 10.33.35. 4. K. D. Vajpeyi, Braj ka Itihas (Mathura: Akhil Bharatiya Braj Sahitya Mandal, 1955), 141. 5. Some scholars have argued that if the absence of earlier monumental construction is taken into account, this project was actually the initial establishment of a Vaishnava center at the site. Charlotte Vaudeville has pointed to the prevalence of cults of nagas, hills and natural phenomena, as well as evidence for Shiva and goddess worship, but not a significant Krishna cult, with the exception of the Keshava Temple at Mathura; see Vaudeville, “Braj, Lost and Found,” Indo-Iranian Journal 18 (1976): 195–213. 6. Caitanya Carita¯mr. ta 2.17–18. 7. As his journey gave material form and cohesion to what had been lost or

224

•

153–187 Notes to pages 000–000

had previously remained inaccessible, it is often called the Braj-prakash lila (the play that revealed Braj/Vrindavan). 8. See David L. Haberman, Journey through the Twelve Forests: An Encounter with Krishna (New York: Oxford University Press, 1994). 9. Ibid., 55. 10. For the narrative, see N. Thakur, “The Building of Govindadeva,” in Govindadeva, ed. Case, 68. 11. Asher, “The Architecture of Raja Man Singh.” 12. T. Mukherji and I. Habib, “Akbar and the Temples of Mathura and Its Environs,” Proceedings of the Indian History Congress 48 (1987): 234–50. 13. Haberman, Twelve Forests, 35. 14. Its maintenance was ensured by Akbar’s successor Jahangir through grants in 1598 and the construction of two more temples. He too paid a personal visit to Govindadeva, in 1620. Later, Shah Jahan issued supportive farmans and decreed that temple bells could resound freely throughout Vrindavan. T. Mukherji and I. Habib, “The Mughal Administration and the Temples of Vrindavan during the Reigns of Jahangir and Shah Jahan,” Proceedings of the Indian History Congress 49 (1988): 289–99. 15. The most recent discussion of this architecture is in Case, ed., Govindadeva. 16. Ernest Binfield Havell concluded, “The craftsmanship is that of FatehpurSikri, but the Hindu builders working on their own ground could deal with structural problems more freely and confidently than they were able to do under the restrictions of Musulman ritual and custom, with the result that they achieved a structural harmony and decorative unity.” Havell, Indian Architecture, 194. 17. See Asher, Architecture of Mughal India, 62–63; and Ebba Koch, “The Architectural Forms,” in Fatehpur-Sikri, ed. Michael Brand and Glenn D. Lowry (Bombay: Marg, 1987), 121–48. 18. See Sylvia Crowe et al., The Gardens of Mughul India: A History and a Guide (London: Thames and Hudson, 1972); James L. Wescoat, Jr., and Joachim Wolschke-Bulmahn, eds., Mughal Gardens: Sources, Places, Representations, and Prospects (Washington, D.C.: Dumbarton Oaks Research Library and Collection, 1996); and Elizabeth B. Moynihan, Paradise as a Garden in Persia and Mughal India (New York: G. Braziller, 1979). 19. A plan of Agra, with this fourfold garden, survives in the Jaipur Palace Museum; see Ebba Koch, Mughal Architecture: An Outline of Its History and Development, 1526–1858 (Munich: Prestel, 1991), 33. 20. Hesilrige, Burdwan, Letters Received, 1788–1802, West Bengal District Records, new series, vol. 2 (Calcutta: Government of West Bengal Printing), 43; and J. Z. Holwell, Interesting Historical Events, part 1 (London: T. Becket and P. A. De Hondt, 1766–71), 197–98. 21. Oral communication from Sukamal Goswami, Nakaijuri, June 2000. 22. See Diana L. Eck, Banares: City of Light (New York: Knopf, 1982). 23. A hill west of Dig is called Adibadri and marks the original Badrinath; Baldeo is considered the deity corresponding to Jagannath of Puri; Setubandh Kund at Kaman is described as the site where Krishna re-enacted Rama’s deeds at Rameshwaram; and points at Kosi are named after Dwaraka. See Entwistle, Braj, 307.

Notes to pages 000–000 187–191

•

225

24. Bishnupurer Madanmohaner Adimahatya (Calcutta: Rajendra Library, n.d.), 8. 25. Diana L. Eck, “India’s Tirthas: ‘Crossings’ in Sacred Geography,” History of Religions 20.4 (May 1981): 323–44. 26. See Charlotte Vaudeville, “The Govardhan Myth in Northern India,” Indo-Iranian Journal 22 (1980): 1–45. 27. Located along the length of escarpments on the east and west peripheries of the town, they probably also provided defensive measures and supplied drinking water. 28. Oral communication from Mrs. Chittaranjan Dasgupta, Vishnupur, June 2000. (The Dasgupta family is descended from Srinivas’s most important disciple, Ramchandra Kaviraj.) 29. Oral communication from Sukamal Goswami of Nakaijuri, Vishnupur, June 2000. (Sukamal Goswami is a descendant of Srinivas through his son.) 30. “In this way Maha¯prabhu went along dancing, and when he came to A¯rit.agra¯ma he recovered his senses. At A¯ rit.a he asked people the whereabouts of the Ra¯dha¯kun.d.a, but no one could say; not even his bra¯hman.a companion knew. Prabhu was the all-knowing Bhagavan, and he knew where the lost tı¯rtha was. He went to bathe in a small pool of water between two fields of paddy. When they saw this, all the people of the village were astonished; in prema Prabhu praised the Ra¯dha¯kun.d.a, ‘Among all the gopı¯s Radha is the most beloved of Kr.s.n.a; so Ra¯dha¯kun.d.a is beloved, the tank of his beloved.’” Caitanya Carita¯mr. ta 18.2–6. 31. Diana L. Eck, Dar´san: Seeing the Divine Image, 3rd ed. (New York: Columbia University Press, 1996), 48–49. 32. Caitanya Carita¯mr. ta 18.7–9. 33. Haberman, Twelve Forests, 102. 34. Ananda Wickremeratne, “Shifting Metaphors of Sacrality: The Mythic Dimensions of Anuradhapura,” in The City as a Sacred Center: Essays on Six Asian Contexts, ed. Bardwell Smith and Holly Baker Reynolds (Leiden: E. J. Brill, 1987), 44. 35. Personal communication from Sushanta Mukhopadhyay, Vishnupur, December 1995. 36. R. Kaviraj, Madan Mohan Bandana, MS. 324, Vishnupur: Vishnupur Shahitya Parishad. 37. Personal communication from Mrs. Salil Singh, Vishnupur, December 1995. 38. Oral communication from Gauri (Chhaya Goswami), Vishnupur, June 2000. 39. The performance of sixty-four acts of worship was prescribed to inspire and stimulate devotion to Krishna and anchor Vaishnava practice. Among these sixty-four, the five principal vaidhi instructions are to chant the name of Krishna, remember the stories of Krishna’s life, serve the divine image in the temple, live in the presence of holy men, and live within the realm of Mathura. Caitanya Carita¯mr. ta 2.22.74–75. 40. Quoted in Saha, Some Aspects of Malla Rule, 185. 41. Quoted in Qazi Abdul Mannan, The Emergence and Development of Dobhasi Literature in Bengal up to 1855 (Dacca: Dacca University, 1966), 86– 102. 42. Tarafdar, Husain Shahi Bengal, 251.

226

•

Notes to pages 000–000 191–197

43. The road from Tamluk wound its way through Patpur, Uramya, Sapur, Birsinghapur, Jote-behar, Rajhat, and Bahulara, to Telkupi in Purulia, and thence to Rajgir and Patna. See Beglar, Report of a Tour, 202. 44. Prem Vilas, cited in Saha, Some Aspects of Malla Rule, 20. 45. A. P. Mallik, History of Bishnupur Raj (Bishnupur: S. P. Mallik, 1921), 123. 46. Personal communication from Catherine Asher, Chapel Hill, N.C., March 2000.

Notes to pages 000–000 197–198 Notes

•

227

This page intentionally left blank

xii

•

Illustrations

Bibliography

Aall, Ingrid. “Bengali Terra-Cotta Temples: New Lines of Inquiry.” In Bengal: Change and Continuity, ed. Robert Paul Beech and Mary Jane Beech, 69–83. East Lansing: Asian Studies Center, Michigan State University, 1969. Abid Ali Khan, M. Memoirs of Gaur and Pandua. Calcutta: Bengal Secretariat Book Depot, 1931. Ahmad, Khwajah Nizamuddin. Tabaquat-iAkbari. 3 vols. Ed. Brajendranath De and M. Hidayat Hosein. Trans. Brajendranath De. Calcutta: Bibliotheca Indica, 1927–39. Quoted in Richard M. Eaton, The Rise of Islam and the Bengal Frontier (Berkeley and Los Angeles: University of California Press, 1993). Ahmed, Shamsud-din, ed. and trans. Inscriptions of Bengal. 4 vols. Rajshahi: Varendra Research Institute, 1960. Akhtar, Shirin. The Role of the Zamindars in Bengal, 1707–1772. Dhaka: Asiatic Society of Bengal, 1982. Ambalal, Amit. Krishna as Shrinathji: Rajasthani Paintings from Nathdvara. Ahmedabad: Mapin, 1987. Anderson, Benedict. Imagined Communities: Reflections on the Origin and Spread of Nationalism. London: Verso, 1991. Appadurai, Arjun. “Making the National Geographic: Changing Images of Territory in Colonial India.” Paper presented at the Ethnohistory Workshop, University of Pennsylvania, 1995. ———. Modernity at Large: Cultural Dimensions of Globalization. Minneapolis: University of Minnesota Press, 1996. Archer, Mildred, and W. G. Archer. Indian Painting for the British, 1770–1880. London: Oxford University Press, 1955. Archer, W. G. The Loves of Krishna in Indian Painting and Poetry. London: Allen & Unwin, 1957. Asher, Catherine B. Architecture of Mughal India. Cambridge: Cambridge University Press, 1992. ———. “The Architecture of Murshidabad: Regional Revival and Islamic Continuity.” In Islam and Indian Regions, ed. Anna Libera Dallapiccola and Stephanie Zingel-Avé Lallemant, 2 vols., 1:61–74. Stuttgart: Franz Steiner Verlag, 1993.

Bibliography

•

229

———. “The Architecture of Raja Man Singh: A Study of Sub-imperial Patronage.” In The Powers of Art: Patronage in Indian Culture, ed. Barbara Stoler Miller, 183–202. New Delhi: Oxford University Press, 1992. ———. “Authority, Victory, and Commemoration: The Temples of Raja Man Singh.” Journal of Vaisnava Studies 3.3 (summer 1995): 25–36. ———. “From Anomaly to Homogeneity: The Mosque in Fourteenth- through Sixteenth-Century Bihar.” In Studies in Art and Archaeology of Bihar and Bengal, ed. Debala Mitra and Gouriswar Bhattacharya, 67–84. Delhi: Sri Satguru, 1989. ———. “Inventory of Key Monuments.” In The Islamic Heritage of Bengal, ed. George Michell, 37–140. Paris: UNESCO, 1984. ———. “The Mughal and Post-Mughal Periods.” In The Islamic Heritage of Bengal, ed. George Michell, 193–212. Paris: UNESCO, 1984. ———. “Sub-Imperial Palaces: Power and Authority in Mughal India.” Ars Orientalis 23 (1993), 281–302. Asher, Catherine B., and Thomas R. Metcalf, eds. Perceptions of South Asia’s Visual Past. New Delhi: American Institute of Indian Studies and Swadharma Swarajya Sangha, 1994. Asher, Frederick M. The Art of Eastern India, 300–800. Minneapolis: University of Minnesota Press, 1980. ———. “Stone and the Production of Images.” East and West 48.3–4 (December 1998): 313–28. Austin, John L. How to Do Things with Words. Oxford: Clarendon Press, 1962. Bagchi, P. C. “Political Relations between Bengal and China in the Pathan Period.” Visva-Bharati Annals 1 (1945): 96–134. Bandopadhyaya, Amiyakumara. Bankura Jelara Purakirti. Calcutta: Purva Bibhag, Pascim Banga Sarkar, 1971. Bandyopadhyay, Bimal. Metal Sculptures of Eastern India. Delhi: Sundeep Prakashan, 1981. Bandyopadhyay, Sukhamay. Temples of Birbhum. Delhi: B. R. Publishing Corporation, 1984. Banerjea, J. N. The Development of Hindu Iconography. Calcutta: University of Calcutta, 1956. Banerjee, Sumanta. Appropriation of a Folk-Heroine: Radha in Mediaeval Bengali Vaishnavite Culture. Shimla: Indian Institute of Advanced Study, 1993. Banerji, Naseem Ahmed. “The Architecture and Architectural Decoration of the Adina Mosque, Pandua, West Bengal, India: The Problem of the Conjoined Buddhist, Hindu, and Islamic Motifs in the Mihrab Niches.” Ph.D. diss., University of Iowa, 1993. Banerji, R. D. Eastern Indian School of Medieval Sculpture. Archaeological Survey of India Reports, new imperial series, vol. 47. Delhi: Manager of Publications, 1933. Basham, A. L. The Wonder That Was India: A Survey of the Indian Subcontinent before the Coming of the Muslims. Reprint, New York: Grove Press, 1959. Bar. ru Cand.ida¯sa. Singing the Glory of Lord Krishna: The S´rı¯kr.s. n.akı¯rtana. Translated and annotated by Miriam H. Klaiman. Chico, Calif.: Scholars Press, 1984.

230

•

Bibliography

Baxandall, Michael. Patterns of Intention: On the Historical Explanation of Pictures. New Haven, Conn.: Yale University Press, 1985. Beeman, William O. “The Anthropology of Theater and Spectacle.” Annual Review of Anthropology 22 (1993): 369–93. Beglar, J. D. Report of a Tour through the Bengal Provinces of Patna, Gaya, Mongir, and Bhagalpur; the Santal Parganas, Manbhum, Singhbhum, and Birbhum; Bankura, Raniganj, Bardwan, and Hughli, in 1872–73. Archaeological Survey of India Reports, old series, vol. 8. Calcutta, 1878. Reprint, Varanasi: Indological Book House, 1966. Bell, Catherine. Ritual Theory, Ritual Practice. New York: Oxford University Press, 1992. Bernier, François. Travels in the Mogul Empire, A.D. 1656–1658. Translated by Archibald Constable. Westminster: Constable, 1891. 2nd ed., New Delhi: S. Chand, 1968. Bertocci, Peter J. “Elusive Villages: Social Structure and Community Organization in Rural East Pakistan.” Ph.D. diss., Michigan State University, 1970. Bhagavata Purana. Trans. and annotated by Ganesh Vasudeo Tagare. Vol. 10 of Ancient Indian Tradition and Mythology, ed. J. L. Shastri. Delhi: Motilal Banarsidass, 1978. Reprint, Delhi: Motilal Banarsidass, 1994. Bhardwaj, Surinder Mohan. Hindu Places of Pilgrimage in India: A Cultural Geography. Berkeley and Los Angeles: University of California Press, 1973. Bhattacharyya, A. K. A Corpus of Dedicatory Inscriptions from Temples of West Bengal, c. 1500 A.D. to c. 1800 A.D. Calcutta: Nabhana, 1982. Bhattacharyya, N. N. History of the Tantric Religion. 1982. Reprint, Delhi: Manohar, 1992. Bhattasali, Nalinikanta. “Bengal Chiefs’ Struggle for Independence in the Reigns of Akbar and Jahangir.” Bengal Past and Present 35 (January–June 1928): 30–36. ———. Iconography of Buddhist and Brahmanical Sculptures in the Dacca Museum. Dacca: Rai S. N. Bhadra Bahadur, 1929. Birdwood, G. C. M. The Industrial Arts of India. 2 vols. London: Chapman and Hall, 1880. Bishnupurer Madanmohaner Adimahatya. Calcutta: Rajendra Library, n.d. Biswas, Sailendra. Samsad Bengali-English Dictionary. 1968. Reprint, Calcutta: Sahitya Samsad, 1996. Bloch, T. S. “The Temples of Vishnupur.” Annual Report of the Archaeological Survey of India, 1903–1904 (1906): 49–53. Bourdieu, Pierre. Language and Symbolic Power. Translated by Gino Raymond and Matthew Adamson. 1991. Reprint, Cambridge, Mass.: Harvard University Press, 1994. Brand, Michael, and Glenn D. Lowry, eds. Fatehpur-Sikri. Bombay: Marg, 1987. Breckenridge, Carol A., and Peter van der Veer, eds. Orientalism and the Postcolonial Predicament: Perspectives on South Asia. Philadelphia: University of Pennsylvania Press, 1993. Brown, C. Mackenzie. “The Theology of Ra¯dha¯ in the Pura¯n.as.” In The Divine Consort: Ra¯dha¯ and the Goddesses of India, ed. John Stratton Hawley and Donna Marie Wulff, 57–71. Boston: Beacon Press, 1982. Brown, Percy. Indian Architecture. 2 vols. 1942. Reprint. Bombay: D. B. Taraporevala Sons, 1965.

Bibliography

•

231

Caitanya Carita¯mr. ta of Kr.s.n.ada¯sa Kavira¯ja. Trans. and with commentary by Edward C. Dimock. Ed. Tony K. Stewart. Cambridge, Mass.: Dept. of Sanskrit and Indian Studies, Harvard University; distributed by Harvard University Press, 1999. Case, Margaret H., ed. Govindadeva: A Dialogue in Stone. New Delhi: Indira Gandhi National Centre for Arts, 1996. Casey, Jane Anne, ed. Medieval Sculpture from Eastern India: Selections from the Nalin Collection. Livingston, N.J.: Nalini International, 1985. Chakrabarti, Kunal. Religious Process: The Puranas and the Making of a Regional Tradition. New Delhi: Oxford University Press, 2001. Chakrabarty, Mukundaram. Kavikankan Chandi. Ed. Srikumar Bandyopadhyay and Visvapati Chaudhuri. Calcutta: Calcutta University, 1974. Chakravarti, Monmohan. “Bengal Temples and Their General Characteristics.” Journal of the Asiatic Society of Bengal 5 (May 1909): 141–62. Chakravarti, Narahari. Bhaktiratnakar. Ed. R. Vidyaratna. 3rd ed. Murshidabad, 1332 BS [1925]. Chakravarti, Ramakanta. Vais.n.avism in Bengal, 1486–1900. Calcutta: Sanskrit Pustak Bhandar, 1985. Chatterjee, Anjali. Bengal in the Reign of Aurangzeb. Calcutta: Progressive, 1967. Cohn, Bernard S. “The Census, Social Structure, and Objectification in South Asia.” Folk 26 (1984): 25–49. Connerton, Paul. How Societies Remember. Cambridge: Cambridge University Press, 1989. Coomaraswamy, Ananda K. The Indian Craftsman. London: Probsthain, 1909. ———. “The Nature of ‘Folklore’ and ‘Popular Art.’” Indian Art and Letters 10.2 (1937): 76–84. Crowe, Sylvia, et al. The Gardens of Mughul India: A History and a Guide. London: Thames and Hudson, 1972. Crowe, Yolanda. “Reflections on the Adina Mosque at Pandua.” In The Islamic Heritage of Bengal, ed. George Michell, 155–64. Paris: UNESCO, 1984. ———. “Some Aspects of the Tree Motif in Medieval Indian Architecture.” In Dimensions of Indian Art: Pupul Jayakar Seventy, ed. Lokesh Chandra, Jyotindra Jain, and Agam Prasad, 79–83. Delhi: Agam Kala Prakashan, 1986. Cunningham, Alexander. Report of a Tour in Bihar and Bengal in 1879–80, from Patna to Sunargaon. Archaeological Survey of India Reports, old series, vol. 15. Calcutta: Office of the Superintendent of Government Printing, 1882. Curley, David. “Temples, Tanks, and Maths: Royal Authority in the Zemindari Seat of Raj’nagar.” Paper presented at “Orienting Architectural Form, Differentiating Meanings: Bengal Architecture from the Fourteenth to the Seventeenth Centuries,” workshop held at the University of North Carolina at Chapel Hill, March 2000. Dacca Museum. First International Congress on Bengal Art. Bangladesh: Dacca Museum, 1976. Dalton, E. T. Descriptive Ethnology of Bengal. Calcutta: Thacker, Spink, 1872. Reprint, Calcutta: K. L. Muhkopadhyaya, 1973. Dani, Ahmad Hasan. Muslim Architecture in Bengal. Dacca: Asiatic Society of Pakistan, 1961.

232

•

Bibliography

Das, Jadunandan. Karnananda. Ed. R. Vidyaratna. 2nd ed. Murshidabad, 1355 BS [1948]. Das, Nityananda. Prem Vilas. Ed. R. Vidyaratna. Murshidabad: Radharaman Press, 1298 BS [1891]. Das, Veena. “On the Categorization of Space in Hindu Ritual.” In Text and Context: The Social Anthropology of Tradition, ed. Ravindra Jain, 9–27. Philadelphia: Institute for Study of Human Issues, 1977. Dasgupta, Chittaranjan. Bishnupurer Mandir Terakota. Bishnupur: Bangiya Sahitya Parishad, 1980. Dasgupta, Prodosh. Temple Terracotta of Bengal. New Delhi: Crafts Museum, 1971. Dasgupta, Shashibhushan. Obscure Religious Cults. 3rd ed. 1969. Reprint, Calcutta: Firma KLM, 1995. ———. Sri Radhar Kramabikas. Calcutta: Firma K. L. Mukhopadhyay, 1952. Datta, Bimal Kumar. Bengal Temples. New Delhi: Munshiram Manoharlal, 1975. Datta, K. K. Alivardi and His Times. Calcutta: World Press, 1963. ———. Siraj ud-daulah. Bombay: Orient Longman, 1971. ———. Survey of India’s Social Life and Economic Condition in the Eighteenth Century, 1707–1803. Calcutta: Firma K. L. Mukhopadhyay, 1961. Datta, Sarojit. Folk Paintings of Bengal. New Delhi: Khama, 1993. Davis, Richard H. “Introduction: A Brief History of Religions in India.” In Religions of India in Practice, ed. Donald S. Lopez, Jr., 637–48. Princeton, N.J.: Princeton University Press, 1994. ———. Lives of Indian Images. Princeton, N.J.: Princeton University Press, 1997. De, Sushil Kumar. Early History of the Vaishnava Faith and Movement in Bengal, from Sanskrit and Bengali Sources. 1961. Reprint, Calcutta: Firma KLM, 1986. Desai, Vishakha N., and Darielle Mason, eds. Gods, Guardians, and Lovers. New York: The Asia Society Galleries, in association with Mapin, Ahmedabad, 1993. Desai, Z. A. “An Early Thirteenth-Century Inscription from West Bengal.” Epigraphia Indica, Arabic and Persian Supplement (1975): 6–12. Dey, Mukul. Birbhum Terracottas. New Delhi: Lalit Kala Akademi, 1959. Dikshit, K. N. Excavations at Paharpur, Bengal. Memoirs of the Archaeological Survey of India, no. 55. Reprint, 1955. Dimock, Edward C. “Bhakti.” In Cooking for the Gods: The Art of Home Ritual in Bengal, ed. Michael W. Meister, catalogue by Pika Ghosh, 27–32. Newark, N.J.: The Newark Museum; Philadelphia: University of Pennsylvania Press, 1995. ———. “Lı¯la¯.” History of Religions 29.2 (November 1989): 159–73. ———. “Muslim Vaisnava Poets of Bengal.” In The Sound of Silent Guns and Other Essays, 23–32. London: Oxford University Press, 1989. ———. The Place of the Hidden Moon: Erotic Mysticism in the Vais.n.avasahajiya¯ Cult of Bengal. Chicago: University of Chicago Press, 1966. Reprint, Chicago: University of Chicago Press, 1989. ———. The Sound of Silent Guns and Other Essays. London: Oxford University Press, 1989.

Bibliography

•

233

Dimock, Edward C., and Denise Levertov. In Praise of Krishna: Songs from the Bengali. Chicago: University of Chicago Press, 1965. Dirks, Nicholas B. “Guiltless Spoilations: Picturesque Beauty, Colonial Knowledge, and Colin Mackenzie’s Survey of India.” In Perceptions of South Asia’s Visual Past, ed. Catherine B. Asher and Thomas R. Metcalf, 211–32. New Delhi: American Institute of Indian Studies and Swadharma Swarajya Sangha, 1994. Donaldson, Thomas E. Hindu Temple Art of Orissa. Studies in South Asian Culture, vol. 12. Leiden: E. J. Brill, 1985–87. Drewal, Margaret Thompson. Yoruba Ritual: Performers, Play, Agency. Bloomington: Indiana University Press, 1992. Dube, S. C., ed. Tribal Heritage of India. New Delhi: Vikas, 1977. Dutt, Gurusaday. “Bengali Terracottas.” Journal of the Indian Society of Oriental Art 6 (1938): 169–80. ———. Catalogue of Folk Arts. Calcutta: Indian Society of Oriental Art, 1932. ———. Folk Arts and Crafts of Bengal: The Collected Papers. Calcutta: Seagull, 1990. ———. “The Indigenous Painters of Bengal.” Journal of the Indian Society of Oriental Art 1 (1933): 17–25. ———. “A Wood Carving from a Bengal Village.” Journal of the Indian Society of Oriental Art 5 (1937): 29–31. Eaton, Richard M. “Islam in Bengal.” In The Islamic Heritage of Bengal, ed. George Michell, 23–36. Paris: UNESCO, 1984. ———. “Mughal Religious Culture and Popular Islam in Bengal.” In Islam and Indian Regions, ed. Anna Libera Dallapiccola and Stephanie Zingel-Avé Lallemant, 75–86. Stuttgart: Franz Steiner Verlag, 1993. ———. The Rise of Islam and the Bengal Frontier. Berkeley and Los Angeles: University of California Press, 1993. ———. “Temple Desecration and Indo-Muslim States.” In Beyond Turk and Hindu: Shaping Islamicate Identity in South Asia, 1400–1800, ed. David Gilmartin and Bruce B. Lawrence, 246–81. Gainesville: University Press of Florida, 2000. Eck, Diana L. Banares: City of Light. New York: Knopf, 1982. ———. Dar´s an: Seeing the Divine Image. 3rd ed. New York: Columbia University Press, 1996. ———. “India’s Tirthas: ‘Crossings’ in Sacred Geography.” History of Religions 20.4 (May 1981): 323–44. Edwards, Holly Frances Cameron. “The Genesis of Islamic Architecture in the Indus Valley.” Ph.D. diss., New York University, 1990. Eliade, Mircea. Cosmos and History: The Myth of Eternal Return. New York: Harper and Row, 1959. ———. “Mythologies of Memory and Forgetting.” In The History of Religions: Essays in Methodology, ed. Mircea Eliade and Joseph Kitagawa, 329–44. Chicago: University of Chicago Press, 1959. Reprint, Chicago: University of Chicago Press, 1969. Entwistle, Alan W. Braj: Centre of Krishna Pilgrimage. Groningen: E. Forsten, 1987. Ernst, Carl W. Mysticism, History, and Sufism in a South Asian Center. Albany: State University of New York Press, 1992.

234

•

Bibliography

Eschmann, Anncharlott, Hermann Kulke, and Gaya Charan Tripathi, eds. The Cult of Jagannath and the Regional Tradition of Orissa. New Delhi: Manohar, 1978. Fazl, Abul. Akbarnamah. Vol. 2. Trans. H. Beveridge. Calcutta: Royal Asiatic Society of Bengal, 1934. Reprint, Delhi: Rare Books, 1972–73. Fergusson, James. History of Indian and Eastern Architecture. Vol. 2. 1891. Reprint, New York: Dodd, Mead, 1899. Firminger, W. K. “A Note on Bansberia: Its Raj, Its Temples, Its Missions and Schools.” Bengal Past and Present 2 (1908): 73–83. Fischer, Klaus. “Bengal Brick Temples.” In Studies in Indian Temple Architecture, ed. Pramod Chandra, 179–96. New Delhi: American Institute of Indian Studies, 1975. Fish, Stanley. “Interpreting the Variorum.” In Reader Response Criticism: From Formalism to Post-structuralism, ed. Jane P. Tompkins, 164–84. Baltimore: Johns Hopkins University Press, 1980. French, J. C. “The Land of Wrestlers (A Chapter in the Art of Bengal).” Indian Art and Letters 1 (1927): 15–29. Frykenberg, Robert E. “The Emergence of Modern ‘Hinduism’ as a Concept and as an Institution: A Reappraisal with Special Reference to South India.” In Hinduism Reconsidered, ed. Gunther D. Sontheimer and Hermann Kulke, 82–107. Delhi: Manohar, 1989. Fuhrer, Alois Anton. The Sharqi Architecture of Jaunpur. 1889. Reprint, Calcutta: Superintendent of Government Printing, 1994. Gait, E. A. Report: Census of India, 1901. Vol. 6, part 1. Calcutta: Bengal Secretariat Press, 1902. Gangoly, O. C. “Art of the Pala Empire of Bengal.” Rupam 37 (1929): 59. Geertz, Clifford. Local Knowledge: Further Essays in Interpretive Anthropology. New York: Basic Books, 1983. Ghose, A. “Figured Fabrics of Old Bengal.” Marg 3.1 (1949–50): 38–43. ———. “Old Bengali Paintings.” Rupam 27–28 (1926): 98–104. Ghosh, Benoy. Pascimbanger Sanskriti. Vol. 1. Calcutta: Prakash Bhaban, 1957. ———. Traditional Arts and Crafts of West Bengal: A Sociological Survey. Calcutta: Papyrus, 1981. Ghosh, D. P. “An Illustrated Ramayana Manuscript of Tulsidas and Pats from Bengal.” Journal of the Indian Society of Oriental Art 13 (1945): 129–39. Ghosh, Pika. “Sojourns of a Peripatetic Deity.” RES Anthropology and Aesthetics 41 (spring 2002): 104–26. ———. “Space and the New Temple Vernacular.” In Traditional and Vernacular Architecture, ed. Deborah Thiagarajan and Subashree Krishnaswamy, 16– 30. Chennai: Madras Craft Foundation, 2003. ———. “Unrolling a Narrative Scroll: Artistic Practice and Identity in LateNineteenth-Century Bengal.” Journal of Asian Studies 62.3 (August 2003): 835–71. Gilmartin, David, and Bruce B. Lawrence, eds. Beyond Turk and Hindu: Shaping Islamicate Identity in South Asia, 1400–1800. Gainesville: University Press of Florida, 2000. Ginzburg, Carlo. Clues, Myths, and the Historical Method. Translated by John Tedeschi and Anne C. Tedeschi. Baltimore: Johns Hopkins University Press, 1989.

Bibliography

•

235

Gopinatha Rao, T. A. Elements of Hindu Iconography. Madras: Law Printing House, 1914–16. Reprint, New York: Paragon Book Reprint, 1968. Goswami, K. G. Excavation at Bangarh. Calcutta: Calcutta University Press, 1948. Growse, F. S. Mathura: A District Memoir. 1883. Reprint, Ahmedabad: New Order Book Co., 1978. Guha-Thakurta, Tapati. The Making of a New Indian Art. Cambridge: Cambridge University Press, 1992. Haberman, David L. Acting as a Way of Salvation: A Study of Ra¯ganuga¯ Bhakti Sa¯dhana. New York: Oxford University Press, 1988. ———. Journey through the Twelve Forests: An Encounter with Krishna. New York: Oxford University Press, 1994. ———. “Vraja: A Place in the Heart.” Re-discovering Braj: International Association of the Vrindaban Research Institute Bulletin 14 (1988): 19–25. Hacker, Katherine F. “Dressing the Lord Jagannatha in Silk: Cloth, Clothes, and Status.” RES Anthropology and Aesthetics 32 (autumn 1997): 107–24. ———. “Traveling Objects: Brass Images, Artisans, and Audiences.” RES Anthropology and Aesthetics 37 (spring 2000): 147–65. Haq, M. E. A History of Sufism in Bengal. Dhaka: Asiatic Society of Bangladesh, 1975. Haque, Zulekha. Terracotta Decorations of Late Medieval Bengal: Portrayal of a Society. Dacca: Asiatic Society of Bangladesh, 1980. Hasan, Perween. “Eight Sultanate Mosques in Dhaka District.” In The Islamic Heritage of Bengal, ed. George Michell, 179–92. Paris: UNESCO, 1984. ———. “The Footprint of the Prophet.” Muqarnas 10 (1993): 335–47. ———. “Sultanate Mosques and Continuity in Bengal Architecture.” Muqarnas 6 (1989): 58–74. ———. “Sultanate Mosque-Types in Bangladesh: Origins and Development.” Ph.D. diss., Harvard University, 1984. ———. “Temple Niches and Mihrabs in Bengal.” In Islam and Indian Regions, ed. Anna Libera Dallapiccola and Stephanie Zingel-Avé Lallemant, 86–94. Stuttgart: Franz Steiner Verlag, 1993. Hasan, Syed Mahmudul. The Adina Masjid at Hazrat Pandua. Dhaka: Amina Art Press, 1970. ———. Gaud and Hazrat Pandua: Romance in Brick and Stone. Dhaka: Islamic Foundation Bangladesh, 1987. ———. Mosque Architecture of Pre-Mughal Bengal. Bangladesh: University Press Limited, 1979. ———. Muslim Monuments of Bangladesh. 1971, as Ancient Monuments of Bangladesh. Reprint, Dacca: University Press, 1980. Havell, E. B. Indian Architecture: Its Psychology, Structure, and History from the First Muhammadan Invasion to the Present Day. London: John Murray, 1913. Hawley, John Stratton. At Play with Krishna: Pilgrimage Dramas from Brindavan. Princeton, N.J.: Princeton University Press, 1981. Hawley, John S., and Donna M. Wulff, eds. Devi: Goddesses of India. Berkeley and Los Angeles: University of California Press, 1996. ———. The Divine Consort: Ra¯dha¯ and the Goddesses of India. Berkeley: Graduate Theological Union, 1982. Reprint, Boston: Beacon Press, 1986.

236

•

Bibliography

Hayes, Glen. “The Necklace of Immortality.” In Tantra in Practice, ed. David Gordon White, 308–25. Princeton, N.J.: Princeton University Press, 2000. ———. “Shapes for the Soul: A Study of Body Symbolism in the Vais.n.avaSahajiya¯ Tradition.” Ph.D. diss., University of Chicago, 1985. Hein, Norvin. The Miracle Plays of Mathura. New Haven, Conn.: Yale University Press, 1972. Hesilrige, ed. Burdwan, Letters Received, 1788–1802. West Bengal District Records, new series, vol. 2. Calcutta: Government of West Bengal Printing, 1955. Holwell, J. Z. Interesting Historical Events. Part 1. London: T. Becket and P. A. De Hondt, 1766–71. Horne, Lee. “Brasscasters of Dariapur, West Bengal: Artisans in a Changing World.” Expeditions 29 (1987): 39–46. ———. “Making Metal in West Bengal.” In Cooking for the Gods: The Art of Home Ritual in Bengal, ed. Michael W. Meister, catalogue by Pika Ghosh, 33–38. Newark, N.J.: The Newark Museum; Philadelphia: University of Pennsylvania Press, 1995. Hunter, William Wilson. Annals of Rural Bengal. London: Smith, Elder, 1897. Huntington, Susan L. The “Pala-Sena” Schools of Sculpture. Leiden: E. J. Brill, 1984. Husain, Mahdi, trans. and comm. The Reh.la of Ibn Bat. .t u¯.t a. Baroda: Oriental Institute, 1953. Husain, Shahanara. The Social Life of Women in Early Medieval Bengal. Dhaka: Asiatic Society of Bengal, 1985. Ikram, S. M. Muslim Civilization in India. New York: Columbia University Press, 1964. Inden, Ronald B. Marriage and Rank in Bengali Culture: A History of Caste and Clan in Middle Period Bengal. Berkeley and Los Angeles: University of California Press, 1976. Inden, Ronald B., and Ralph W. Nicholas. Kinship in Bengali Culture. Chicago: University of Chicago Press, 1977. Irwin, John. “Indo-Portuguese Embroideries of Bengal.” Art and Letters, new series 16.2 (1952): 65–73. Israel, Milton, and N. K. Wagle, eds. Islamic Society and Culture. New Delhi: Manohar, 1983. Kaimal, Padma. “Passionate Bodies: Constructions of the Self in South Indian Portraits.” Archives of Asian Art 48 (1995): 6–16. Karim, Abdul. Murshid Quli Khan and His Times. Dacca: Asiatic Society of Pakistan, 1963. Kaviraj, R. Madan Mohan Bandana. MS. 324. Vishnupur: Vishnupur Shahitya Parishad. Kennedy, Melville T. The Chaitanya Movement. New York: Garland, 1981. Khan, F. A. Mainamati: A Preliminary Report on the Recent Archaeological Excavations in East Pakistan. Karachi: Dept. of Archaeology, Ministry of Education and Information, Government of Pakistan, 1963. Khan, Muhammad Abid Ali. Memoirs of Gaur and Pandua. Ed. and rev. H. E. Stapleton. Calcutta: Bengal Secretariat Book Depot, 1931. Khan, Muhammad Hafizullah. Terracotta Ornamentation in Muslim Architecture of Bengal. Dhaka: Asiatic Society of Bangladesh, 1988.

Bibliography

•

237

King, Anthony. “The Bengal Peasant Hut: Some Nineteenth Century Accounts.” Art and Archaeology Research Papers 12 (1977): 70. Kinsley, David. The Divine Player: A Study of Krishna Lila. Delhi: Motilal Banarsidass, 1979. ———. Tantric Visions of the Divine Feminine: The Ten Mahavidyas. Berkeley and Los Angeles: University of California Press, 1997. Reprint, Delhi: Motilal Banarsidass, 1998. Koch, Ebba. “The Architectural Forms.” In Fatehpur-Sikri, ed. Michael Brand and Glenn D. Lowry, 121–48. Bombay: Marg, 1987. ———. Mughal Architecture: An Outline of Its History and Development, 1526–1858. Munich: Prestel, 1991. Kopf, David, ed. Bengal Regional Identity. East Lansing: Asian Studies Center, Michigan State University, 1969. Kramrisch, Stella. The Hindu Temple. Calcutta: University of Calcutta, 1946. Reprint, Delhi: Motilal Banarsidass, 1976. ———. “Pala and Sena Sculpture.” Rupam 40 (1929): 107–26. ———. Unknown India: Ritual Art in Tribe and Village. Philadelphia: Philadelphia Museum of Art, 1968. Kulke, Hermann. “Early Patronage of the Jagannatha Cult.” In The Cult of Jagannath and the Regional Tradition of Orissa, ed. Anncharlott Eschmann, Hermann Kulke, and Gaya Charan Tripathi, 139–55. New Delhi: Manohar, 1978. ———. Kings and Cults: State Formation and Legitimation in India and Southeast Asia. New Delhi: Manohar, 1993. ———. “The Struggle between the Rajas of Khurda and the Muslim Subadars of Cuttack for Dominance of the Jagannatha Cult.” In The Cult of Jagannath and the Regional Tradition of Orissa, ed. Anncharlott Eschmann, Hermann Kulke, and Gaya Charan Tripathi, 321–42. New Delhi: Manohar, 1978. Liston, Yarina. “The Ra¯dha¯ Cycle: From Lila¯ to Bhakti,” Journal of Vais.n.ava Studies 8.2 (spring 2000): 7–26. Lopez, Donald, Jr., ed. Religions of India in Practice. Princeton, N.J.: Princeton University Press, 1994. Lorenzen, David. “Who Invented Hinduism?” Comparative Studies in Society and History 41.4 (October 1999): 630–59. Lynch, Owen M., ed. Divine Passions: The Social Construction of Emotion in India. Delhi: Oxford University Press, 1990. Lyons, Tryna. “The Simla Devı¯ Ma¯ha¯tmya Illustrations: A Reappraisal of Contents.” Archives of Asian Art 45 (1992): 29–41. Majumdar, A. K. Bhakti Renaissance. Bombay: Bharatiya Vidya Bhavan, 1979. Majumdar, Bimanbehari. Govindadaser Padavali o tahar Yug. Calcutta: Calcutta University, 1961. Majumdar, Nani Gopal. Inscriptions of Bengal. Vol. 3. Rajshahi: Varendra Research Society, 1929. Majumdar, R. C. The History of Bengal. Vol. 1, Hindu Period. Dacca: University of Dacca, 1943. ———. History of Mediaeval Bengal. Calcutta: G. Bhardwaj, 1973. Mallik, A. P. History of Bishnupur Raj. Bishnupur: S. P. Mallik, 1921. Mandala, Satyananda. Dakshina Cabbisa Paraganara Lokasilpa. Calcutta: Firma KLM, 1984.

238

•

Bibliography

Mannan, Qazi Abdul. The Emergence and Development of Dobhasi Literature in Bengal up to 1855. Dacca: Dacca University, 1966. Marglin, Frédérique Apffel. “Refining the Body: Transformative Emotion in Ritual Dance.” In Divine Passions: The Social Construction of Emotion in India, ed. Owen M. Lynch, 225–30. Delhi: Oxford University Press, 1990. ———. “Types of Sexual Union and the Implicit Meanings.” In The Divine Consort: Ra¯dha¯ and the Goddesses of India, ed. John Stratton Hawley and Donna Marie Wulff, 298–315. Boston: Beacon Press, 1982. ———. Wives of the God-King: The Rituals of the Devadasis of Puri. Delhi: Oxford University Press, 1985. Reprint, Delhi: Oxford University Press, 1989. Marriott, McKim. “The Feast of Love.” In Krishna: Myths, Rites, and Attitudes, ed. Milton Singer, 200–12. Chicago: University of Chicago Press, 1966. Martin, Robert Montgomery, ed. The History, Antiquities, Topography, and Statistics of Eastern India. London: W. H. Allen, 1838. Reprint, in 5 vols., Delhi: Cosmo, 1976. McCutchion, David. “Hindu-Muslim Artistic Continuities.” In The Islamic Heritage of Bengal, ed. George Michell, 213–30. Paris: UNESCO, 1984. ———. “Krishnalila on the Temples of Bengal.” Journal of the Indian Society of Oriental Art, new series 7 (1975–76): 33–51. ———. Late Mediaeval Temples of Bengal: Origins and Classification. Calcutta: Asiatic Society, 1972. ———. “Origins and Developments.” In Brick Temples of Bengal: From the Archives of David McCutchion, ed. George Michell, 15–30. Princeton, N.J.: Princeton University Press, 1983. ———. “The Ramayana on the Temples of Bengal.” Eastern Railway Magazine 15.8 (August 1967): 9–15. ———. “Styles of Bengal Temple Terracotta: A Preliminary Analysis.” In South Asian Archaeology: Papers from the First International Conference of South Asian Archaeologists Held in the University of Cambridge, ed. Norman Hammond, 265–78. London: Duckworth, 1973. ———. The Temples of Bankura District. Calcutta: Writers Workshop, 1972. McDaniel, June. The Madness of the Saints: Ecstatic Religion in Bengal. Chicago: University of Chicago Press, 1989. ———. “The Tantric Radha: Some Controversies about the Nature of Radha in Bengali Vaishnavism and the Radha Tantra.” Journal of Vais.n.ava Studies 8.2 (spring 2000), 131–46. McDermott, Rachel Fell. “The Vaisnavized Uma of Bengali Devotionalism.” Journal of Vais.n.ava Studies 4.4 (winter 1995–96): 105–24. McLane, John R. Land and Local Kingship in Eighteenth-Century Bengal. Cambridge: Cambridge University Press, 1993. Meister, Michael W. “Fragments from a Divine Cosmology: Unfolding Forms on India’s Temple Walls.” In Gods, Guardians, and Lovers: Temple Sculptures from North India, A.D. 700–1200, ed. Vishakha N. Desai and Darielle Mason, 94–115. New York: The Asia Society Galleries; Ahmedabad: Mapin, 1993. ———. “The Hindu Temple: Axis and Access.” In Concepts of Space, Ancient and Modern, ed. Kapila Vatsyayan, 269–80. New Delhi: Indira Gandhi National Centre for Arts and Abhinav, 1991.

Bibliography

•

239

———. “Indian Islam’s Lotus Throne: Kaman and Khatu Kalan.” In Islam and Indian Regions, ed. Anna Libera Dallapiccola and Stephanie Zingel-Avé Lallemant, 2 vols., 445–52. Stuttgart: Franz Steiner Verlag, 1993. ———. “Pra¯sa¯da as Palace: Kut. ina Origins of the Nagara Temple.” Artibus Asiae 49.3 (1988–89): 254–80. ———. “The Two-and-a-Half-Day Mosque.” Oriental Art, new series 18.1 (spring 1972): 57–63. Meister, Michael W., ed., and Pika Ghosh, cat. Cooking for the Gods: The Art of Home Ritual in Bengal. Newark, N.J.: The Newark Museum; Philadelphia: University of Pennsylvania Press, 1995. Michell, George, ed. Brick Temples of Bengal: From the Archives of David McCutchion. Princeton, N.J.: Princeton University Press, 1983. ———. The Islamic Heritage of Bengal. Paris: UNESCO, 1984. ———. “Neo-classicism in Bengali Temple Architecture.” Art and Archaeology Research Papers 11 (June 1977): 28–31. Miller, Barbara Stoler. “The Divine Duality of Ra¯dha¯ and Krishna.” In The Divine Consort: Ra¯dha¯ and the Goddesses of India, ed. John Stratton Hawley and Donna Marie Wulff, 13–26. Boston: Beacon Press, 1982. ———, ed. The Powers of Art: Patronage in Indian Culture. New Delhi: Oxford University Press, 1992. ———, ed. and trans. The Gitagovinda of Jayadeva: Love Song of the Dark Lord. Delhi: Motilal Banarsidass, 1984. Mills, Margaret. Rhetorics and Politics in Afghan Storytelling. Philadelphia: University of Pennsylvania Press, 1991. Mitra, Debala. “Inscription from Wari, Saka 1467.” Epigraphia Indica 35 (1963–64): 179–84. Mitra, S. K., ed. East Indian Bronzes. Calcutta: Calcutta University Press, 1979. Mitter, Partha. Much Maligned Monsters: A History of European Reactions to Indian Art. Oxford: Clarendon Press, 1977. Reprint, Chicago: University of Chicago Press, 1992. Monier-Williams, Monier. A Sanskrit-English Dictionary. Oxford: Clarendon Press, 1872. Reprint, Delhi: Motilal Banarsidass, 1990. Mookerjee, Ajitcoomar. Folk Art of Bengal: A Study of an Art for and of the People. Calcutta: Calcutta University Press, 1946. Morinis, E. Alan. Pilgrimage in the Hindu Tradition: A Case Study of West Bengal. Delhi: Oxford University Press, 1984. Moynihan, Elizabeth B. Paradise as a Garden in Persia and Mughal India. New York: G. Braziller, 1979. Mukherji, T., and I. Habib. “Akbar and the Temples of Mathura and Its Environs.” Proceedings of the Indian History Congress 48 (1987): 234–50. ———. “The Mughal Administration and the Temples of Vrindavan during the Reigns of Jahangir and Shah Jahan.” Proceedings of the Indian History Congress 49 (1988): 289–99. Nath, R. “S´rı¯ Govindadeva’s Itinerary from Vr.nda¯vana to Jayapura, c. 1534– 1727.” In Govindadeva: A Dialogue in Stone, ed. Margaret H. Case, 161– 84. New Delhi: Indira Gandhi National Centre for Arts, 1996. Nathan, Mirza. Baharistan-i-Ghaybi: A History of the Mughal Wars in Assam, Cooch Behar, Bengal, Bihar, and Orissa during the Reigns of Jahangir and Shahjahan. Vol. 1. Translated by M. I. Borah. Gauhati, Assam: Dept. of

240

•

Bibliography

Historical and Antiquarian Studies, Narayani Handiqui Historical Institute, 1936. Nazimuddin, Ahmed. Mahasthan: A Preliminary Report on Archaeological Excavations. Karachi: Dept. of Archaeology and Museums, Ministry of Education and Scientific Research, Government of Pakistan, 1971. Newman, Richard. The Stone Sculpture of India: A Study of the Materials Used by Indian Sculptors from ca. Second Century B.C. to the Sixteenth Century. Cambridge, Mass.: Center for Conservation and Technical Studies, Harvard University Art Museums, 1984. Nicholas, Ralph W. “Vaisnavism and Islam in Rural Bengal.” In Bengal Regional Identity, ed. David Kopf, 33–47. East Lansing: Asian Studies Center, Michigan State University, 1969. Oberoi, Harjot. The Construction of Religious Boundaries: Culture, Identity, and Diversity in the Sikh Tradition. Delhi: Oxford University Press, 1994. O’Connell, Joseph T. “The Impact of Devotion upon the Societal Organization of Bengal.” In Studies in Bengali Literature, History, and Society, ed. Edward C. Dimock Jr., 33–123. East Lansing: Asian Studies Center, Michigan State University, 1975. ———. “Ra¯ma¯nanda Ra¯ya: A Sahajiya¯ or a Ra¯ga¯nuga¯ Bhakta?” Journal of Vais.n.ava Studies 1.3 (spring 1993): 36–58. ———. Religious Movements and Social Structure: The Case of Chaitanya’s Vais.n.avas in Bengal. Shimla: Indian Institute of Advanced Study, 1993. ———. “The Social Implications of the Gaudiya Vaishnava Movement.” Ph.D. diss., Harvard University, 1970. ———. “Vais.n.ava Perceptions of Muslims in Sixteenth-Century Bengal.” In Islamic Society and Culture: Essays in Honour of Professor Aziz Ahmad, ed. Milton Israel and N. K. Wagle, 289–313. New Delhi: Manohar, 1983. ———. “The Word ‘Hindu’ in Gaud.ı¯ya Vais.n.ava Texts.” Journal of the American Oriental Society 93.3 (July–September 1973): 340–44. ———, ed. Bengal Vaisnavism, Orientalism, Society, and the Arts. East Lansing: Asian Studies Center, Michigan State University, 1985. ———. Organizational and Institutional Aspects of Indian Religious Movements. Shimla: Indian Institute of Advanced Study, 1999. O’Flaherty, Wendy Doniger. Women, Androgynes, and Other Mythical Beasts. Chicago: University of Chicago, 1980. O’Malley, L. S. S. Bengal District Gazetteers: Bankura. Calcutta: Government of West Bengal, 1908. Reprint, Calcutta: Government of West Bengal, 1995. Ong, Walter J. Orality and Literacy: The Technologizing of the Word. London: Routledge, 1988. Ostor, Akos. The Play of the Gods: Locality, Ideology, Structure, and the Festivals of a Bengali Town. Chicago: University of Chicago Press, 1980. Pal, M. K. Asutosh Museum of Indian Art: An Introduction. Calcutta: Calcutta University Press, 1962. Palmer, Gary B., and William R. Jankowiak. “Performance and Imagination: Toward an Anthropology of the Spectacular and the Mundane.” Cultural Anthropology 11.2 (May 1996): 225–58. Paterson, J. C. K. Bengal District Gazetteers: Burdwan. New Delhi: Logos Press, 1910. Patua Art: Development of the Scroll Paintings of Bengal Commemorating the

Bibliography

•

241

Bicentenary of the French Revolution. Calcutta: Alliance Française of Calcutta and Crafts Council of West Bengal, 1989. Paul, Robert, and Mary Jane Beech, eds. Bengal: Change and Continuity. East Lansing: Asian Studies Center, Michigan State University, 1969. Pinder-Wilson, Ralph. “Stone Sculptures of Gaur.” In Indian Art Connoisseurship: Essays in Honor of Douglas Barrett, ed. John Guy, 251–61. Delhi: Indira Gandhi National Centre for Arts and Mapin, 1996. Ramanujan, A. K. Hymns for the Drowning: Poems for Vis.n.u by Namma¯lvar. Princeton, N.J.: Princeton University Press, 1981. Rao, Velcheru Narayana, David Shulman, and Sanjay Subrahmanyam. Symbols of Substance: Court and State in Na¯yaka Period Tamil Nadu. Delhi: Oxford University Press, 1992. Rashid, M. Harunur. “Bangladesh: A Profile in the Light of Recent Archaeological Discoveries.” Bangladesh Historical Studies 3 (1978): 11. Rawson, Philip. The Art of Tantra. London: Thames and Hudson, 1973. Reprint, London: Thames and Hudson, 1995. Ray, Amita, H. Sanyal, and S. C. Ray, eds. Indian Studies: Essays Presented in Memory of Prof. Niharranjan Ray. Delhi: Caxton, 1984. Ray, S. K. “The Artisan Castes of West Bengal and Their Craft.” In The Tribes and Castes of West Bengal: Census of India, 1951: West Bengal, ed. A. Mitra, 293–350. Calcutta: West Bengal Government Press, 1953. ———. “Primitive Statuettes of West Bengal.” Journal of Arts and Crafts 1.1 (January 1939): 1–8. Raychaudhuri, Tapan Kumar. Bengal under Akbar and Jahangir: An Introductory Study in Social History. Calcutta: A. Mukherjee, 1953. Reprint, Delhi: Munshiram Manoharlal, 1966. Rocher, Rosane. “British Orientalism in the Eighteenth Century: The Dialectics of Knowledge and Government.” In Orientalism and the Postcolonial Predicament: Perspectives on South Asia, ed. Carol A. Breckenridge and Peter van der Veer, 215–49. Philadelphia: University of Pennsylvania Press, 1993. Rockhill, W. W. “Notes on the Relations and Trade of China with the Eastern Archipelago and the Coast of the Indian Ocean during the Fourteenth Century.” T’oung Pao 16.2 (1915): 435–55. Rosaldo, Michelle. “Toward an Anthropology of Self and Feeling.” In Culture Theory: Essays on Mind, Self, and Emotion, ed. Richard A. Shweder and Robert A. LeVine, 137–57. Cambridge: Cambridge University Press, 1984. Rosen, Steven. The Lives of the Vaishnava Saints. New York: Folk Books, 1991. Roy, Asim. “The Interface of Islamization, Regionalization, and Syncretization: The Bengal Paradigm.” In Islam and Indian Regions, ed. Anna Libera Dallapiccola and Stephanie Zingel-Avé Lallemant, 95–128. Stuttgart: Franz Steiner Verlag, 1993. ———. The Islamic Syncretistic Tradition in Bengal. Princeton, N.J.: Princeton University Press, 1983. Sachau, Edward C., trans. and ed. Alberuni’s India. London: Routledge and Kegan Paul, 1888. Saha, Prabhat Kumar. Some Aspects of Malla Rule in Bishnupur, 1590–1806. Calcutta: Ratnabali, 1995. Salim, Ghulam Hussain. Riazu-s-Salatin. Trans. M. A. Salam. 1903. Reprint, Delhi: Idarah-I Adabiyat-I Delli, 1975.

242

•

Bibliography

Salomon, Carol. “The Cosmogonic Riddles of Lalan Fakir.” In Gender, Genre, and Power in South Asian Expressive Traditions, ed. Arjun Appadurai, Frank J. Korom, and Margaret A. Mills, 267–304. Philadelphia: University of Pennsylvania Press, 1991. Santra, Tarapada. “Cataloguing and Documentation of Mediaeval Temples of Bengal: An Analysis of Problems and Methods.” Calcutta Review, new series 3.3 (January–March 1972): 239–45. Sanyal, Hitesh Ranjan. “Literary Sources of Medieval Bengali History: A Study of a Few Mangalkavya Texts.” In Indian Studies: Essays Presented in Memory of Prof. Niharranjan Ray, ed. Amita Ray, H. Sanyal, and S. C. Ray, 205–15. Delhi: Caxton, 1984. ———. “Mallabhum.” In State Formation in Eastern India, ed. Surajit Sinha, 73–142. Calcutta: Proceedings of the Seminar on State Formation in Eastern and Northeastern India at the Centre for Studies in Social Sciences, 1981. ——— “Regional Religious Architecture in Bengal: A Study in the Sources of Origin and Character.” Marg 27.2 (March 1974): 31–43. ———. “Religious Architecture in Bengal (Fifteenth–Seventeenth Centuries): A Study of the Major Trends.” Indian History Congress, Proceedings, ThirtySecond Session, Jabalpur 1 (1970): 413–22. ———. “Social Aspects of Temple Building in Bengal: 1600–1900 A.D.” Man in India 48.3 (July–September 1968): 201–24. ———. “Transformation of the Regional Bhakti Movement (Sixteenth and Seventeenth Centuries).” In Bengal Vaishnavism, Orientalism, Society, and the Arts, ed. Joseph T. O’Connell, 59–70. East Lansing: Asian Studies Center, Michigan State University, 1985. Saraswati, S. K. Architecture of Bengal. Book 1, Ancient Phase. Calcutta: G. Bhardwaj, 1976. ———. Early Sculpture of Bengal. Calcutta: Sambodhi, 1962. ———. “Indo-Muslim Architecture in Bengal.” Journal of the Indian Society of Oriental Art 9 (1941): 12–36. Sarkar, Guruprasad. Mallasamskritira Patabhumikaya Bishnupurera Rasotsaba. Bishnupur: The New Minerva Press, 1980. Sarkar, Jadunath. The History of Bengal. Vol. 2, Muslim Period: 1200–1757. Dacca: University of Dacca, 1947. Reprint, Patna: Janaki Prakashan, 1977. Schechner, Richard. Between Theater and Performance. Philadelphia: University of Pennsylvania Press, 1985. Schweig, Graham M. “Ra¯dha¯ and the Ra¯salı¯la¯: The Esoteric Vision of Caitanyaite Vaishnavism.” Journal of Vais.n.ava Studies 8.2 (spring 2000): 43–71. Sen, Dinesh Chandra. History of Bengali Language and Literature. Calcutta: University of Calcutta, 1911. ———. The Vais.n.ava Literature of Mediaeval Bengal. Calcutta: University of Calcutta, 1917. Sen, Sukumar. Bangla Sahityer Itihas. 3rd ed. Burdwan: Burdwan Sahitya Sabha, 1940–56. Sen Gupta, Sankar. The Patas and the Patuas of Bengal. Calcutta: Indian, 1973. Seyller, John. “Rasalila.” In Intimate Worlds: Indian Paintings from the Alvin O. Bellak Collection, ed. Darielle Mason, 46–47. Philadelphia: Philadelphia Museum of Art, 2001. Sharma, R. The Religious Policy of the Mughal Emperors. New York: Asia, 1972.

Bibliography

•

243

Shokoohy, Mehrdad, and Natalie H. Shokoohy. “The Architecture of Baha aldin Tughrul in the Region of Bayana, Rajasthan.” Muqarnas 4 (1987): 114– 32. Shweder, Richard A., and Robert A. LeVine, eds. Culture Theory: Essays on Mind, Self, and Emotion. Cambridge: Cambridge University Press, 1984. Siddiq, Mohammad Yusuf. “An Epigraphical Journey to an Eastern Islamic Land.” Muqarnas 10 (1993): 83–108. Sila-Khan, Dominique. Conversions and Shifting Identities: Ramdev Pir and the Ismailis in Rajasthan. Delhi: Manohar and Centre de Sciences Humaines, 1997. Singer, Milton, ed. Krishna: Myths, Rites, and Attitudes. Chicago: University of Chicago Press, 1966. Sinha, Surajit. “State Formation and Rajput Myth in Tribal Central India.” Man in India 42.1 (January–March 1962): 35–80. Sirajuddin, Muhammad. Lyrics in Terracotta at Kantajeer Mandir. Dhaka: Markup International, 1992. Sircar, D. C. Some Epigraphical Records of the Medieval Period from Eastern India. New Delhi: Abhinav, 1979. Skelton, Robert, and Mark Francis. Arts of Bengal: The Heritage of Bangladesh and Eastern India. London: The Whitechapel Art Gallery, 1979. Smith, Bardwell, and Holly Baker Reynolds. The City as a Sacred Center: Essays on Six Asian Contexts. Leiden: E. J. Brill, 1987. Smith, Ronald Bishop. The First Age of the Portuguese Embassies and Peregrinations to the Ancient Kingdoms of Cambay and Bengal (1500–1521). Bethesda, Md.: Decatur Press, 1969. Smith, Wilfred Cantwell. The Meaning and End of Religion: A New Approach to the Religious Traditions of Mankind. New York: Macmillan, 1963. Solvyns, Balthazar. A Collection of Two Hundred and Fifty Colored Etchings: Descriptive of the Manners, Customs, Character, Dress, and Religious Ceremonies of the Hindoos. Calcutta: Mirror Press, 1799. Sontheimer, Gunther D., and Hermann Kulke, eds. Hinduism Reconsidered. Delhi: Manohar, 1989. Stewart, Tony K. “The Architectonics of Community and Space: The Mandala as Maker of Worldly Utopias in Early Gaudiya Vaisnavism.” Unpublished manuscript. ———. “The Final Word.” Unpublished manuscript. Tambiah, Stanley Jeyaraja. Culture, Thought, and Social Action: An Anthropological Perspective. Cambridge, Mass.: Harvard University Press, 1985. Tarafdar, M. R. Husain Shahi Bengal, 1494–1538 A.D.: A Socio-political Study. Dhaka: Asiatic Society of Pakistan, 1965. ———. “An Indigenous Source for Bengali Sufism.” In Islam and Indian Regions, ed. Anna Libera Dallapiccola and Stephanie Zingel-Avé Lallemant, 129–36. Stuttgart: Franz Steiner Verlag, 1993. Tartakov, Gary Michael. “Changing Views of India’s Art History.” In Perceptions of South Asia’s Visual Past, ed. Catherine B. Asher and Thomas R. Metcalf, 15–36. New Delhi: American Institute of Indian Studies and Swadharma Swarajya Sangha, 1994. Thakur, N. “The Building of Govindadeva.” In Govindadeva: A Dialogue in Stone, ed. Margaret H. Case, 11–68. New Delhi: Indira Gandhi National Centre for Arts, 1996. 244

•

Bibliography

Thapar, Romila. “Syndicated Moksha?” Seminar 313 (September 1985): 14–22. Thiel-Horstmann, Monika. “Bhakti and Monasticism.” In Hinduism Reconsidered, ed. Gunther D. Sontheimer and Hermann Kulke, 127–40. Delhi: Manohar, 1989. Urban, Hugh B. The Economics of Ecstasy: Tantra, Secrecy, and Power in Colonial Bengal. Oxford: Oxford University Press, 2001. Vajpeyi, K. D. Braj ka Itihas. Mathura: Akhil Bharatiya Braj Sahitya Mandal, 1955. Van Lohuizen de Leeuw, Johana E. “The Early Muslim Monuments at Bagerhat.” In The Islamic Heritage of Bengal, ed. George Michell, 165–78. Paris: UNESCO, 1984. Vaudeville, Charlotte. “Braj, Lost and Found.” Indo-Iranian Journal 18 (1976): 195–213. ———. “The Govardhan Myth in Northern India.” Indo-Iranian Journal 22 (1980): 1–45. ———. “Krishna Gopa¯la, Ra¯dha¯, and the Great Goddess.” In The Divine Consort: Ra¯dha¯ and the Goddesses of India, ed. John Stratton Hawley and Donna Marie Wulff, 1–12. Boston: Beacon Press, 1982. von Stietencron, Heinrich. “Hinduism: On the Proper Use of a Deceptive Term.” In Hinduism Reconsidered, ed. Gunther D. Sontheimer and Hermann Kulke, 32–53. Delhi: Manohar, 1989. Wagoner, Philip B. “Harihara, Bukka, and the Sultan: The Delhi Sultanate in the Political Imagination of Vijayanagara.” In Beyond Turk and Hindu: Shaping Islamicate Identity in South Asia, 1400–1800, ed. David Gilmartin and Bruce B. Lawrence, 300–326. Gainesville: University Press of Florida, 2000. ———. “‘Sultan among Hindu Kings’: Dress, Titles, and the Islamicization of Hindu Culture at Vijayanagara.” Journal of Asian Studies 55.4 (November 1996): 851–80. Ward, W. Account of the Writings, Religion, and Manners, of the Hindoos: Including Translations from Their Principal Works. Serampore: Mission Press, 1811. Welch, Stuart Cary, ed. Gods, Kings, and Tigers: The Art of Kotah. Munich: Prestel, 1997. Wescoat, James L., Jr., and Joachim Wolschke-Bulmahn, eds. Mughal Gardens: Sources, Places, Representations, and Prospects. Washington, D.C.: Dumbarton Oaks Research Library and Collection, 1996. White, Hayden. The Content of the Form: Narrative Discourse and Historical Representation. Baltimore: Johns Hopkins University Press, 1987. Wickremeratne, Ananda. “Shifting Metaphors of Sacrality: The Mythic Dimensions of Anuradhapura.” In The City as a Sacred Center: Essays on Six Asian Contexts, ed. Bardwell Smith and Holly Baker Reynolds, 45–59. Leiden: E. J. Brill, 1987. Williams, Joanna G. “Criticizing and Evaluating the Visual Arts in India: A Preliminary Example.” Journal of Asian Studies 47.1 (February 1988): 3–28. ———. “From the Fifth to the Twentieth Century and Back.” Art Journal 49.4 (winter 1990): 363–69. ———. “S´iva and the Cult of Jagannath: Iconography and Ambiguity.” In Discourses on S´iva: Proceedings of a Symposium on the Nature of Religious Imagery, ed. Michael W. Meister, 298–311. Philadelphia: University of Pennsylvania Press, 1984. Bibliography

•

245

Wulff, Donna M. Drama as a Mode of Religious Realization. Chico, Calif.: Scholars Press, 1984. ———. “Radha: Consort and Conqueror of Krishna.” In Devi: Goddesses of India, ed. John S. Hawley and Donna M. Wulff, 109–27. Berkeley and Los Angeles: University of California Press, 1996. Zbavitel, D. Bengali Literature. Wiesbaden: Otto Harassowitz, 1976.

246

•

Bibliography

Index

Italicized page numbers indicate illustrations. Abul Fazl, 92, 188 Achabal gardens, 189 Adi Malla, 18 Adina Masjid (1375), 75, 78, 99, 216n51 Agra, 189 Ahmad, Nizamuddin, 87 aishwarya, 12, 52 Akbar (1556–1605), 22, 91, 93, 188, 189, 225n14 Akbarnama, 205n19 Ala al-din Husain Shah (1493–1519), 85, 87, 214n31 Alilaki Khan of Rajnagar, 72 amalaka, 83, 113, 201, 219n7 Amber, 198 Ambika-Kalna: Sultanate structures in, 76 Anderson, Benedict, 67 anjalimudra, 54 annakot, 45 Appadurai, Arjun, 20 aprakat, 15, 50 Arab traders, 87 arati, 43, 46–47, 54, 137, 140–43 Archaeological Survey of India, 46, 70, 76, 78, 140, 207n1, 216n56 arches: pointed, 82, 112, 115, 146 architects: religious affiliation of, 72, 74 artisans, 83; religious affiliation of, 72, 74 ascetics, 155, 157, 215n42 Asher, Catherine, 188 Ashtagram: mosque at, 223n19 astadhatu, 44 Atala Masjid at Jaunpur, 216n50

Aurangzeb (1618–1707), 198, 218n82 axis, 215n45; bhakti, 137, 140, 149, 154, 224n35; dual, 23, 139, 148, 153, 190; east-west, 137–38, 162, 169; north-south, 137–38, 140, 142, 144, 146, 157, 158, 168 Babur (1483–1530), 189 bairagi, 155 Baispara, 19 Balarama, 12, 154 Baluchari saris, 222n13 Bana Yatra, 187 Banda Temple, 117 bandh, 192 Bandyopadhyay, Sukhamay, 51 Banerjee, Achintya, 207n1, 208n5 Banerjee, Tara, 207n1, 224n1 Bangala-i Darshan at Agra, 215n49 bangla, 201, 203n3; as bangala in Mughal context, 215–16n49 Bankura, 77 Bara Bhuiyan Kings, 91 Barakar: Temple 1, 36; Temple 2, 36; Temple 4, 28, 117 bargi, 194 Basanti, 47 Basu, Maladhar, 86 Beglar, J. D., 70, 77 Benaras, 191–92, 194 Bengali, 16, 85 Bhagavata Purana, 12, 149–51, 187 bhakti, 2, 19–20, 118, 121, 137, 140–42, 150, 155, 188, 203n4. See also axis: bhakti

Index

•

247

Bhaktiratnakar, 186 Bhattacharyya, A. K., 10 bhava, 151 Bhishma, 145 bhogamandapa, 138, 201 Biswas, Paramita, 209n14 Brahma, 86, 112, 154 Brahmanical Hinduism: relation to Gaudiya Vaishnavism, 69, 108, 118, 121, 139, 148, 154, 190 Brahmans, 17, 39, 69 Brajamandala, 191 Braj-prakash lila, 224–25n7 brick, 83, 85, 110, 215n49, 216nn55,57, 218n4, 222nn15,16 Brown, Percy, 72 Buchanan-Hamilton, Francis, 110 Buddhism, 153, 193, 218, 222n12 Caitanya Carita¯mr.ta, 150–52, 187, 193, 196 Chaitanya (1486–1533), 2, 6, 13, 39, 49, 69, 84, 121, 142, 151–52, 156, 186– 87, 189, 192–93, 197–98, 203n4, 209n20, 214n34, 215n42, 220n16, 220–21n25; as avatar of Krishna, 15, 51, 139, 148, 151, 187; images of, 139, 148, 168, 221n3; as incarnation of Radha, 152, 187 Chaitanya Mangal, 69, 214n31 Chaitanya Singh (1752/3–1802), 196 chajja, 188, 201 Chakraborty, Chandi, 207n1, 224n1 Chakravarti, Monmohan, 218–19n4 Chakravarti, Ramakanta, 206n32 Chala, 8, 34 chala, 8, 35, 76, 81–82, 109, 115, 201, 203n3, 215–16n49, 222–23n17 Chandi, 6, 17 Chandidas, 85, 123, 220n24 Chandra, 154 char-chala, 113 chatri, 111, 116–17, 119, 201 Chinese traders, 87, 90 Chinnamasta, 122 Chola Dynasty, 192 Chota Pandua, 75 Chota Sona Masjid at Gaur, 75, 113 circumambulation, 140 Cohn, Bernard, 67 colonial period, 109 colonialism: effect on scholarship, 67, 71; perception of religion during, 67 congregational space, 23, 75, 77, 141, 144, 146, 158

248

•

Index

Connerton, Paul, 144 courtly life, 22, 65, 89–90, 106–107, 196 courtyard, 142, 144, 155–56, 158, 190 Curley, David, 47 Dakshineshwar Kali Temple, 43 Damodar River, 191 dargah, 76; of Pir Bahram Saqqa, 76; of Shah Ismail Ghazi, 76 darshan, 140, 148, 189 Das, Chhaya, 207n1, 224n1 Das, Gangadhar, 207n1 Das, Raghunath, 192, 193 Das, Shankar, 216n56 Dasgupta, Chittaranjan, 208n5 Dasgupta, Mrs. Chittaranjan, 226n28 dasya, 12 Daud Karrani, 91, 93 de Barros, João (1496–1570), 20, 89 Delhi, 81, 117, 216n50 deul, 120 devadasis, 49, 210n32 Devi, 120, 123 dham, 191 dhoti, 89 Dimock, Edward, 13 do-chala, 80 Doljatra, 14, 43 dolmancha, 48 dome, 40, 81, 83, 85, 115, 117–18, 188 domestic architecture, 79–80, 101, 109 Doms, 17 Durga, 122 Durga Puja, 123 Dwarakeshwar River, 191 ekaratna, 110, 201 Eklakhi Mausoleum at Hazrat Pandua, 81, 104 entrance: triple arched, 81–82, 116, 145, 215n45, 222 ethnographic method, 7, 40, 74 European traders, 21, 87 experimentation: architectural, 20, 112, 120, 137, 155 Fatehpur Sikri, 188–89 Fergusson, James, 67, 70, 75, 84, 109 festivals, 11, 43–44, 49, 108, 137, 142. See also Doljatra; Durga Puja; Holi; Jhulan; Rasa; Rasalila; Ratha folklore, 7 food offerings (prasada), 19 foreignness, 68, 85 furniture: in Ratna temples, 51

Gait, E. A., 70 garbhagriha, 201 Garh-Mandaran, 76 garments: of the gods for ritual, 46, 89, 210n35; Sultanate, 89 Gaudiya Vaishnavism, 2, 14, 19, 43, 56, 68, 109, 118, 124, 139, 154, 189, 192; architectural spaces for, 11; ascetics in, 155; and congregational worship, 39; cosmology of, 24, 94, 153, 191, 198; esoteric aspects of, 50–51, 156–57, 219–20n14; goals of worship, 52, 54; imagery of, 113, 148, 190; Indo-Islamic context of, 119, 185; and krishnalila, 41, 149– 50; literary works, 6, 55, 68, 85–86, 123, 143, 152, 185–87, 194–96, 206; relation to Bengal Sultanate, 69, 86; relation to Brahmanical Hinduism, 16, 20, 69, 108, 121, 123, 148, 151; relation to Islam, 20, 197; and temple ritual, 12, 40, 42, 108, 112, 114, 137, 142, 144 Gaur, 69, 71, 76–77, 82, 87, 89, 95, 112, 203n4, 204n9, 213n18, 217n61 Gauranga, 151 ghantamala, 147, 201 Gitagovinda, 53, 123, 143, 153, 208n11 Goaldi Masjid at Sonargaon (1519), 75, 96, 117, 223n18 Gokul Chand Temple at Gokulnagar, 58– 59, 78, 115–17, 138, 156, 161–62, 207n1, 211n39; plan of, 160, 164 Gokulnagar, 41, 43, 191 Gopal Singh (1713–52), 93, 152, 196 gopas, 12 Gopinath (deity), 195 Gopinath Temple, 188, 209n14 gopis, 12, 41, 49, 54, 141–42, 144, 148, 151, 155, 195, 209nn13,20 Goswami, Chhaya, 209n15 Goswami, Jiva, 50, 52, 188 Goswami, Rupa, 45, 84, 86, 188, 191, 205n17 Goswami, Sanatana, 84, 86 Goswami, Sukamal, 224n1, 225n21, 226n29 goswamis, 155, 188 Govindadeva: image of, 44–45, 188, 191, 198 Govindadeva Temple (1590), 188, 200, 225n14 Gujarat (mythical kingdom), 17, 20, 74 Gundica, 50, 209n20 Gundica Temple, 49

“Gupta Vrindavan,” 24, 185, 190, 194, 195 Haberman, David, 187 Hambir, Vir. See Vir Hambir Hamsheshvari Temple at Bansberia (1799), 208n3 Haricharan Das, 205n20 Hasan, Perween, 75, 78, 81 Havell, E. B., 218–19n4 Hawley, John Stratton, 50 Hazrat Pandua, 76, 81, 99 Hemlata Devi, 191–92 Hesilrige, 190 Hindu: as label for architecture, 66 Hindu temples: access to, 4; condition of, 5, 78; economic impact of, 19; Muslim patronage of, 72; orientation, 4; and political authority, 77 Hinduism: Brahmanical, and relation to Gaudiya Vaishnavism (see Gaudiya Vaishnavism); as a term, 67 Hodgson, Marshall, 88 Holi, 48–49 Holwell, J. Z., 190 Horne, Lee, 74 Husain Shahi sultans (1493–1538), 75, 84–87, 89 Ilyas Shahi sultans (1433–86), 75, 84 Indo-Islamic context, 20, 23, 65–66, 84, 93, 109, 114, 118, 185, 190, 197, 199 Indparab, 18 Indra, 112, 153 Indradhaj, 18 inscriptions, 72, 120, 144, 149, 203nn1,2, 203–204n6 Isa Khan, 91 Islam Khan, 92 Islamic: as label for architecture, 66 Islamicate, 89 Islamicization, 88 Jadab Ray Temple at Jadabnagar (1650), 114–15, 117, 131, 222n14 Jadabnagar, 191 Jagannatha, 45, 49, 120, 123, 154, 208n8, 210n35 Jagannatha Temple at Puri, 18, 53, 156, 197 jagrata, 194 Jahangir (1605–27), 92, 93, 189, 225n14 Jami Masjid at Atiya (1609), 75, 81, 98, 146

Index

•

249

Jami Masjid at Bagha (1523), 75, 79, 96, 105, 113, 147, 178, 222 Jami Masjid at Salban, 104 Jami Masjid at Shura, 79 jati, 68 Jaunpur, 81 Jayadeva, 53, 153, 210n33 Jayananda, 69, 214n31 Jesuits: Goanese mission of, 91 Jhagrai Chandi Temple (1659), 120 Jhulan (swing festival), 43, 47, 50, 123 jor bangla, 80, 120, 201, 219n10 Jor Mandir, 117 jorano pata, 155 Jugal Kishor Temple, 188 kabayi, 89 Kabilaspur Temple (1643), 214n31 Kala Chand Temple (1656), 3, 30, 83, 115–17, 156, 203n2 Kalanjay Shiva Temple at Patrasayer, 43, 61, 124 kalasha, 83, 113, 201 Kali, 122 Kali and Shiva Temples at Sonarang Tangibari (1838–43), 218n1 Kalki, 153 Kalna, 214n34 kama, 53 Kamala, 122 Kangsabati River, 191 Kantaji (deity), 47 Kantaji Temple at Kantanagar, 47–48, 50, 67, 211n2, 224n39 Kantanagar, 50 Karnananda, 196 Kartikeya, 122, 220n19 Kavikankan Chandi, 17, 74, 206n29 Kaviraj, Krishnadas, 150 Kaviraj, Ramchandra, 226n28 Kedar Rai, 91 Keshav Lalji Temple at Patpur, 48, 156 Keshava Temple at Mathura, 224n5 Keshta Ray Temple (1655), 3, 9, 29, 35, 38, 48, 80, 89–90, 106–107, 118, 147, 154–55, 184, 224n35; plan of, 102 Khan, Yashoraj, 85 kirtan, 13, 16, 46, 48, 51–52, 55–56, 78, 121, 138–44, 149, 152, 155, 195– 97, 221n2, 224n39; images of, 57, 170 kirtaniyas, 142 Krishna, 41, 149, 215n42; abode of, 116,

250

•

Index

189, 194; as Achyuta, 53; as Chaitanya, 15, 151; in disguise as girl, 50; as focus of Gaudiya Vaishnavism, 2, 119, 186; iconography of, 52; images of, 6, 43, 112, 121–22, 124, 139, 147–48, 154; images of story of, 154, 184, 195; as Madan Mohan, 6; medium for images of, 44; as Radha Shyam, 169; relationship with Radha, 12; ritual garments of, 46; as Sharabhuja, 156; worship of, 141–44. See also Radha and Krishna Krishna Temple at Baidyapur (1598), 113 Krishnachandra Temple at AmbikaKalna, 48 Krishnaganj, 19 krishnalila, 41, 184, 221n11; and architectural form, 15 Krittibas, 85, 122 Kulut, 76, 100 kunda, 192 Labhpur, 77 Lakhnauti, 213n18 Lakshmi, 49–50, 53, 122–23, 154 Lalbagh, 93 Lalji (deity), 47 Lalji Temple (1658), 3, 19, 48, 203n2 Lalji Temple at Raghunathgarh Thakurbati, 214n37 lata, 9, 201 latina, 9, 117, 201 Lattan Masjid at Gaur, 204n9 lila, 17, 54–55, 65, 85, 89, 141–42, 144, 149, 187, 203n5; Braj-prakash, 22; relation to architectural form, 15 lingam, 120, 124 lotus, image of, 150, 188, 223n18 Madala Panji, 214n31 Madan Gopal (deity), 47, 191 Madan Gopal Temple (1665), 3, 19, 37, 62, 138, 142, 157, 158, 221n3 Madan Mohan (deity), 45, 47–48, 190– 91, 193–94, 209n15; as avatar of Krishna, 185, 194–95; images of, 44, 77, 195, 208n9; ritual use of images of, 63 Madan Mohan Bandana, 194 Madan Mohan Temple (1694), 3, 19, 33, 50, 78, 83, 116–17, 132, 138–40, 142, 145, 147, 155, 157, 161, 170– 71, 175, 181–82, 188, 195, 207n1,

216n52, 218n85; Krishna image at, 6; plan of, 159, 163; worship at, 63 madhurya, 12, 51, 150 Mahabharata, 86, 145, 175, 193, 217n64 Mahaprabhu Temple, 83 Mahavidyas, 122–23 Mahisha, 122 Mahmud of Ghazni, 187 Majhi, Mathur, 207n1, 224n1 Malla Dynasty, 3, 17, 22, 40, 55, 86, 88, 90, 118, 120, 123, 186, 190, 196, 205n19, 206–207n38; Gopal Singh of (1713–52), 93; historical roots of, 18; patronage of architecture by, 93; patronage of Gaudiya Vaishnavism, 65; relation to Husain Shahi rule, 87 Malla Estate Garden Pavilion, 134 Malla Kingdom, extent of, 77 Mallabhum, 13, 186 Malleshvar Shiva Temple (1622), 114– 15, 117, 130, 222n14 Man Singh, Raja, 92–93 Manasa, 6, 85 Manasa-Mangala, 85 Manasa-Vijaya, 85 mandala, 150, 153, 187, 195 mangalkavya, 6, 86 manjari, 54–55 manjari sadhana, 55 Manoharshahi Gharana, 196 Maratha, 194 Marglin, Frédérique Apffel, 49, 53 maritime trade, 21, 38, 87 Marriott, McKim, 49 Masnad-i Ali. See Isa Khan masons: religious affiliation of, 72, 74 Mathura, 194, 209n13, 226n39 Mathurapur Deul at Madhukeli, 113, 117, 128 matrikas, 220n19 McCutchion, David, 11, 73, 110–11, 204n7 memory, 152, 155 mendicants, 155–56 Midnapur, 77, 206n32 mihrab, 201, 215n45, 222n15, 223nn18,20 minar, 75 Mitra, Gokul: and Madan Mohan, 45 Mleccha, 68 Moharram, 74 Mohini, 48 Mosque at Goaldi (1519), 78

Mosque at Kushumba (1558), 75, 97, 147, 179 Mosque at Shialghuni, 147 Mosque at Shura, 75 Motichura Masjid at Rajnagar, 76, 100, 146 Mount Govardhana, 192, 194 Mrinmoyi, 123 mudita, 2 Mughal architecture, 116–17, 188–89, 204n9, 218n85 Mughal Empire, 22, 65, 70, 84, 91–93, 214–15n38; architectural influence, 45; 1575 invasion of Bengal, 90; Gaudiya Vaishnava relationship with, 118–19 Mukhopadhyay, Sushanta, 63, 157, 207n1, 208n10, 209n13, 224nn38,1 Mukhya, Banshi Kar, 207n1, 224n1 Mukundaram, 17, 74, 206n29 Murali Mohan (deity), 195 Murali Mohan Temple (1665), 3, 31, 80, 117 Murshid Quli Khan (Nawab, 1697– 1712), 92 Murshidabad, 93 Nabadwip, 142, 215n42 Nabi-bangsa, 86 naga, 224n5 nagar samkirtan, 205n18 Nagara, 2, 8, 34, 75, 83, 93, 111–12, 114–15, 117, 120, 123–24, 145, 215n45, 219nn7,10; definition of, 9, 201; as discussed by James Fergusson, 70–71; incorporation of Sultanate decoration on, 113; innovations in, 108; source for upper story of Ratna, 39, 108, 116, 118, 121, 140, 190; spatial organization of, 23, 79–80 Nandakishor Temple at Dvadasbari, 83, 117, 133 Nandalal Temple, 117 Narottama, 14 Nasir al-din Husain Shah (1519–31), 85 Nathan, Mirza, 92 natmandir, 78, 138, 141, 161, 201, 221n2 Navadvip, 214n31 navanarigunjara, 221–22n11 navaratna, 110, 118, 120, 201 navaratna ratnam, 1, 10, 72, 207–208n2 Nayaka, 21

Index

•

251

Nishat Bagh, 189 Nityananda: images of, 139, 168, 221n3

purna kalasha, 147 Putana, 154

O’Connell, Joseph T., 68, 121 Ong, Walter, 153 oral narratives, 6, 40, 143, 155, 186 Orissa, 69, 92, 151, 218n82

Qadam Rasul at Gaur (1519), 75, 80, 82, 95, 216n52 Qutlu Khan Lohani, 92

padas, 85 Paharpur, 83 Pala Dynasty: sculpture of, 147, 153, 216n51, 222n12, 223nn18,19 Pala-Sena period (8th–12th centuries), 204n7, 215n45 Pallava Dynasty, 192 pancharatna, 110, 118, 120, 202, 219n10 panchayatana, 110–11, 202 Pandava brothers, 193 Pandua, 213n18 parakiya, 13, 55 Parvati, 122–23 pata paintings, 74, 155 Pathan: as label, 68 patronage, 19, 56, 83, 193; of architecture, 18, 65, 71, 76–77, 84, 88, 93– 94, 119, 121, 186, 189–90; of festivals, 52; of literature, 85–86; motivation for, 17; Mughal, 90; Muslim, of Hindu temples, 72; of performance, 52; of ritual, 18, 123 patrons, religious affiliation of, 74, 86 performance, 140–41 phyllite: as medium for images of deities, 44 pilgrimage, 187, 189, 191–94, 197 Pir Bahram Saqqa: dargah of, 76 Pir Shah Kauban Ali, 214n32 plaster, 83; painting of, 5 political authority: and Hindu temples, 77 portability: of deities, 43, 56, 63, 221n3 Portuguese, 20–21, 87–89 prakat, 15, 50 Prakrti, 47 prasada, 19, 121, 143, 202 Pratapaditya of Jessore, 91 Prataparudra, 69, 214n31 Prem Vilas, 50 prema, 210n32 Pujari, Subrata, 63 Puranas, 112, 123, 154 Puranic deities, 121, 153, 155 Puri, 69, 214n31 purna ghata, 146–47

252

•

Index

Radha, 2, 6, 23, 43, 49–50, 53, 55, 86, 112, 123, 141, 143, 150–51, 155, 192; desires of, 55, 141; images of, 124, 139; medium for images of, 44; relationship with Krishna, 12; ritual garments of, 46; ritual use of images of, 63. See also Radha and Krishna Radha and Krishna, 23, 49, 53, 55, 191; ritual use of images of, 39; songs about, 7; trysts, 51, 53–55, 86, 153, 188, 190, 210n35 Radha Ballabh Temple at Krishnanagar, 29, 115, 118, 146, 188, 204n9, 224n37 Radha Binodji Mandir at Perua, 72 Radha Damodara Temple at Ghutgeriya, 112–13, 117, 125–26, 219n7 Radha Govinda Mandir at Atpur (1786), 224n39 Radha Govinda Temple at Vishnupur, 117 Radha Kunda, 192, 194 Radha Lalji Temple, 138, 204n9 Radha Madhav Temple (1737), 3, 117, 138, 157; plan of, 165 Radha Mohan (deity), 195 Radha Raman (deity), 44, 191–92, 195 Radha Shyam (deity), 83 Radha Shyam Temple (1758), 3, 33, 48, 117, 138–39, 168–69, 204n9, 218n85 Radha Tantra, 123 Radha Vinod Temple (1659), 3, 8, 32, 79, 140, 147, 181, 207–208n2 Radhakunda, 193 Radhika, 47 Raghunath Singh (1627–56), 115, 193, 196 Raja Ganesh, 217n61 Raja Man Singh, 92, 93 Raja Rajballabh (1714–63), 47–48 Rajnagar, 47–48, 72 Rajshahi, 77 Rama, 86, 122–23, 156, 215n42 Ramakrishna, 43 Ramayana, 85–86, 113, 122 Rao, Velcheru Narayana, 21 rasa, 2

Rasa (festival), 18 rasalila, 41, 45, 149–53 Rasalila (festival), 43 rasamancha, 208n4 rasamandala, 41, 138, 147–48, 151–54, 183, 202 ratha, 138, 193, 200, 202; relation to Shiite taziya, 74 Ratha (festival), 18, 37, 47, 49, 194 Rathajatra, 43, 62 Ratna, 4, 8, 22, 34, 48, 111, 115; access to upper story, 42, 114; and brick, 83; and colonial scholarship, 72, 75; definition of, 9–10, 202; emergence of, 91, 110; furniture in, 51; and Gaudiya Vaishnavism, 39, 46, 185; and Indo-Islamic context, 20–21, 70; kitchen in, 138, 162; locations of, 3; and Mughal architecture, 93; naming of, 190; relation to mosque architecture, 23, 66, 75, 108, 113; relation to Nagara, 108, 119–20, 215n45; relief sculpture of, 153; and ritual space, 56, 65; sanctum of, 156; south façade of, 144–45, 173; spatial organization, 78, 137; spread of, 66; terra cotta sculpture on, 64, 116, 122, 144; upper story of, 39– 40, 42, 46, 48, 53, 108–109, 111, 114–18, 124, 137, 157, 208; use for Krishna temples, 120; use in other traditions, 43 Ratneshvar Shiva Temple at Jagannathpur, 114, 129, 222n14 Raya, Ramananda, 151 Red Fort, Delhi, 215n49 rekha (Nagara), 34, 202 Rekha Deul, 9, 110, 202 religious identity: and architecture, 73; colonial misconceptions of, 68; and ethnicity, 72; Hindu 70; Islamic, 70 roof: char-chala, 8; do-chala, 8; thatch, 8, 66, 79–81, 119, 215n49 Roy, Asim, 84 rudraksha, 156, 215n42 Sadi’s Mosque at Egaroshindur (1652), 75, 98, 147, 180 sahaja, 220n24 Sahajiyas, 85 Saiyid Sultan, 86 sakhya, 12 sambhoga, 141 sampradaya, 150

sanctum, 11, 41, 47, 52, 79, 111–12, 114–15, 137–43, 146, 156, 215nn43,45; orientation of, 78; walls of, 115, 166–67 Sanskritization, 17 santa, 12 Saraswati, 122, 154 sardula, 202, 223n19 Sarkar, Jadunath, 91 Satya Pir, 76 scholarship: colonial period, 109 Schweig, Graham, 150 Sena, Lakshmana, 206n30 Serampore Mission, 120 Shah Ismail Ghazi: dargah of, 76 Shah Jahan, 215–16n49, 225n14 Shah, Nusrat, 86 Shaiva, 85, 108, 121 Shakta, 69, 85, 108, 121, 153 shakti, 123 shalagram, 44, 221n3 Shalimar, 189 Sharabhuja (deity), 156 Sharabhuja Temple at Tejpal, 79, 207– 208n2 Sheikh Kabir, 86, 197 Shiism, 74 shikhara, 111, 117, 202 shiropa, 93 Shiva, 44, 112, 120–21, 123–24, 153–54, 203–204n6, 204n15, 224n5; representations of, 122, 135 Shiva Pratapeshwara Temple at AmbikaKalna, 120 Shiva Temple at Nandi, 113 Shuja Khan (Nawab, 1727–39), 93 Shulman, David, 21 Shyam Chand Temple, 120, 221n2 Shyam Ray (deity), 190 Shyam Ray Temple (1643), 1, 3, 27, 41, 48, 57, 115–17, 123, 154, 203n1, 222n14, 224n1; and ascetics, 156; dedicatory inscription on, 38; goddess imagery on, 122, 136; kirtan imagery on, 144; as navratna ratnam, 9; plan of, 60; sanctum of, 79; terra cotta panels on, 52, 55, 64, 135, 140, 144–48, 152–53, 166–67, 172–74, 183, 216n54; upper story of, 211n39; and vernacular architecture, 80, 103, 222–23n17 Shyamananda, 14, 206n32 Shyamkunda, 192 Singh, Jagat, 92

Index

•

253

Singh, Mrs. Salil, 207n1 Solvyns, Balthazar, 219n11 spatial organization, 23, 39, 42, 75, 78, 157 Sri Nathji, 45 Sridhar Temple, 42, 219n13 Srikrishna Kirtana, 85, 123, 217n64 Srikrishna Vijaya, 86 sringara vesa, 210n35 Srinivas Acharya, 13–14, 18, 44, 52, 56, 65, 123, 152, 186, 189, 191–92, 195–96, 198, 205n20, 226nn28,29; as avatar of Krishna-Chaitanya, 15; celebration of, 63 Stewart, Tony, 13 Subhadra, 123, 154 Subrahmanyam, Sanjay, 21 Sufism, 20, 76, 81, 84, 118, 196–97 Sultan Bahadur (1526–37), 21 Sultanate (Bengal), 20, 213nn17,18; courtly life of, 89–90; extant monuments of , 71, 76; Gaudiya Vaishnava relation to, 22–23, 69, 88, 94, 118; Pala Buddhist art and, 73; relationship with Malla Dynasty, 90 Sultanate, Deccani, 88 Sultanate, Delhi, 117 Sultanate architecture: as source for Nagara, 113; as source for Ratna, 21–23, 65–66, 75, 82–83, 85, 108, 111–14, 117, 120, 124, 145, 188, 190; terra cotta on, 71; and vernacular structures, 8 Sultanate mausoleums, 110, 116 Sultanate mosques, 108–11, 116, 146, 157; mihrab decoration in, 146–47, 177–80, 190, 215n45; spatial organization of, 79–90; terra cotta decoration on, 144 swing, 41, 47, 50, 56; festival of ( Jhulan), 43; introduction of, 42; ritual reasons for, 42; as throne, 40 Tantipara Mosque at Gaur, 75, 95, 146– 47, 177 Tantra, 122, 153 Tara, 122 Tavares, Gonsalo: 1521 embassy of, 89 taziya, 74 Telkupi Temple, 117 terra cotta, 52, 94, 128, 138, 150, 190, 211n2, 215n43, 216nn54,55; and circumambulation, 140; images of

254

•

Index

courtly life in, 89–90; images of kirtan in, 142; jalis in, 156; narrative panels in, 116, 132, 154; painted, 155, 204n9; relation of architecture to, 145; representation of architecture in, 116; representation of deities in, 121; representation of Krishna and Radha in, 54–55; on south façade of Ratna temples, 144–48, 195, 224n35; on Sultanate monuments, 71, 75–76, 85, 105, 113, 147, 222nn12,13,15,16 thakur dalan, 224n39 tirtha, 191–92 Tribeni Mosque, 76 Trinivarta, 154 Tripurasundari, 122 tulasi, 143 tulsimancha, 78, 138, 202 Turk: as label, 68 Turkish traders, 87 Ujjvala-nilamani, 205n17 utsavamurti, 44 uttaravedi, 117, 202 Vaishnava, 85 Vaishnavization, 20 vanamala, 52 vastraharana, 221–22n11 vatsalya, 12 vernacular architecture, 35, 101, 176, 203n3, 222n; as source for Mughal architecture, 215n49; as source for Nagara, 113; as source for Ratna, 65–66, 76, 79–80, 94, 109, 111–12, 114, 116–19, 124, 145, 219n10; use as mosque, 81 Vijaya Gupta, 85 Vijaya Sena (1095–1158), 203–204n6 Vijayanagara, 20–21, 69, 88 vipralambha, 141 Vir Hambir, 13–14, 20, 77, 92–93, 123, 186, 193, 196–97, 205n19; and Madan Mohan image, 44; as Vir Singh, 114 Vir Singh (possibly same as Vir Hambir), 114 viraha, 13 Vishnu, 53, 86, 112, 123, 153–54, 215n42; iconography of, 52 Vishnupur, 3; courtly life at, 90; current status of temples in, 46; indigenous

inhabitants of, 68; location of kirtan singing at, 54; Muslim population in, 76; as seat of Mallas, 120; as Vrindavan, 24, 185–86, 189, 191– 95, 198; worship of goddesses at, 123 Vishnupur Raj, 46, 194, 206–207n38; palace of, 140 Vishnupur Sahitya Parishad, 152–53, 196 Vrindavan, 24, 49–50, 86, 123, 142, 149, 152, 185, 189, 194, 224–25n7; Holi celebrations at, 49; mapped onto Vishnupur, 14, 186–87, 191

Vrindavan Chandra Temple at Guptipara, 204n9 Wagoner, Phillip, 88–89 Ward, Reverend W., 120 Yamuna River, 41, 86, 148–49, 189 Yashoda, 12 Yavana, 68 yogic meditation, 157 Zafar Khan, 93

Index

•

255

This page intentionally left blank

xii

•

Illustrations

Contemporary Indian Studies Published in association with the American Institute of Indian Studies The Edward Cameron Dimock, Jr. Prize in the Indian Humanities Temple to Love: Architecture and Devotion in Seventeenth-Century Bengal Pika Ghosh The Joseph W. Elder Prize in the Indian Social Sciences The Regional Roots of Developmental Politics in India: A Divided Leviathan Aseema Sinha

Bibliography

•

257

This page intentionally left blank

xii

•

Illustrations

PIKA GHOSH is Associate Professor of Art at the University of North Carolina at Chapel Hill. She is co-editor (with Michael W. Meister) of Cooking for the Gods: The Art of Home Ritual in Bengal.

Bibliography

•

259