The Archaeology of Tibetan Books Brill’S Tibetan Studies Library

The Archaeology of Tibetan Books Brill’s Tibetan Studies Library

Edited by

Henk Blezer Alex McKay Charles Ramble

VOLUME 36

The titles published in this series are listed at brill.com/btsl The Archaeology of Tibetan Books

Agnieszka Helman-Ważny

LEIDEN | BOSTON Cover illustration: Monks reciting books at Bya mang po in Western Tibet. Photo by author, June 2013.

Library of Congress Cataloging-in-Publication Data

Helman-Wazny, Agnieszka. The Archaeology of Tibetan books / by Agnieszka Helman-Wazny. pages cm. — (Brill’s Tibetan studies library, ISSN 1568-6183 ; volume 36) Includes bibliographical references and index. ISBN 978-90-04-27504-1 (hardback : acid-free paper) — ISBN 978-90-04-27505-8 (e-book) 1. Books— Tibet Region—History. 2. Books—Conservation and restoration—Tibet Region—History. 3. Bookbinding— Tibet Region—History. 4. Book design—Tibet Region—History. 5. Printing—Tibet Region—History. 6. Papermaking—Tibet Region—History. 7. Manuscripts, Tibetan—History and criticism. 8. Tibet Region— Antiquities. 9. Archaeology and history—Tibet Region. 10. Arts, Tibetan—History. I. Title.

Z8.T53H46 2014 002.0951’5—dc23 2014011156

This publication has been typeset in the multilingual ‘Brill’ typeface. With over 5,100 characters covering Latin, ipa, Greek, and Cyrillic, this typeface is especially suitable for use in the humanities. For more information, please see brill.com/brill-typeface. issn 1568-6183 isbn ��-��-��-�7504-1 (hardback) isbn ��-��-��-�7505-8 (e-book)

Copyright 2014 by Koninklijke Brill nv, Leiden, The Netherlands. Koninklijke Brill nv incorporates the imprints Brill, Brill Nijhoff, Global Oriental and Hotei Publishing. All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by Koninklijke Brill nv provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910, Danvers, ma 01923, usa. Fees are subject to change.

This book is printed on acid-free paper. For Tomasz and Olga

∵

Contents

Acknowledgements ix List of Illustrations xi

1 Introduction 1 The Starting Point 1 Borrowing the Term “Archaeology” 2 Books Written in Tibetan 3 Statistics in Book History 4 My Fieldwork, Sources, and Experiments 5

2 Methods: An Uneasy Alliance of Science and History 12 Many Typologies of Book Features 13 Tools to Read What is Not Written 13 Survey of Disciplines Used for Dating 17 Study of Paper in Books 21

3 Tibetan Books: A Craft and Artistry 47

4 Indigo, Gold, and Human Blood: Tibetan Illuminated Manuscripts 76 Manuscripts versus Prints 76 Patronage and Ownership 77 Gold Manuscripts 79 Format and Book Binding Style 81 Illuminations and Decorations 85 Calligraphy 95 Page Layout 96 Ink 99 Writing Tools 101 Paper 102

5 Tibetan Woodblock Printing Culture 116 Origin of New Technology and the Early Tibetan Woodblock Prints 116 Production of the Tibetan Buddhist Canon 125 Edition, Re-edition, and Reprint: Technical Identity of Hundred-Volume Sets 132 viii contents

Early Editions of Tibetan Kanjur Produced in Beijing 136 Mongolian Kanjurs 158 Tibetan Kanjurs Produced in Tibet and Tibetan Borderland 163 Patterns of Physical Features in Selected Editions of Tibetan Kanjur in a Regional and Temporal Perspective 176

6 A Survey of Tibetan Paper 179 History of Paper in Central Asia and Tibet 179 Himalayan Papermaking Plants and the Localization of Tibetan Paper and Book Origins 183 Papermaking Methods in Tibet Traceable in Books 191 Papermaking Tools and Technology 192

7 Conservation 201 Typical Damage to Tibetan Books 206 Selected Procedures of Conservation Treatment 216

Appendix 1 Selected Items from the Tibetan Collections of the British Library 225 Appendix 2 Features of Paper in Selected Sets of Tibetan Kanjur 252 Appendix 3 Transcription of the Yongle Covers Inscriptions 262 Appendix 4 Selected Items from the Tibetan Collection of the Berlin State Library 265 Appendix 5 Selected Items from the Tibetan Collection of the Library of Congress, Washington DC, the Asian Reading Room 276 Bibliography 288 Index 299 Acknowledgements

Research on the codicology and materiality of Tibetan books is very much in its infancy. Creating this book required considerable groundbreaking preparation before many research tasks could be started—much less completed. Without the generous support, guidance, and encouragement I have received from many people and institutions, I could not have successfully confronted the enormity and complexity of such a challenge. My research had to be conducted on original objects; I could not study features of paper from photographs, even if these were high quality digitized resolutions. Selecting which books to choose was yet another problem. Library catalogues rarely contained useful information on the material aspects of books; some collections were extremely difficult to approach—especially those that were not yet properly catalogued. Thanks to very professional curators at many institutions, my work was made easier. I am enormously grateful to all who gave me essential advice on the selection of books for my study. Despite all the difficulties, I was fortunate to always meet the right people and profit from their professional help and knowledge. This book represents the culmination of the generosity of many individuals, and it is with pleasure that I am able finally to share it. With this in mind, my gratitude is directed first to the librarians and curators who supported my research on material aspects of Tibetan books with professional advice and organizational help. Thus, I would like to thank Dr. Susan Whitfield (British Library), Dr. Sam van Schaik (British Library), Mr. Burkhard Quessel (British Library), Dr. Michael Balk (Berlin State Library), Dr. Susan Meinheit (Library of Congress), Prof. Zdzisław Pietrzyk (Jagiellonian University Library), Dr. Monika Jaglarz (Jagiellonian University Library), Dr. Ellen Avril (Johnson Museum of Art, Cornell University), Dr. Lauran Hartley (Columbia University Library, NYC), Gene Smith, Jeff Wallman, Lobsang Shastri and Kelsang Lhamo (The Tibetan Buddhist Resource Center, Cambridge, Massachusetts, USA), and Dr. Peggy Daub (Special Collections Library, University of Michigan). I am grateful to Professor Richard Ernst for allowing me to examine many exceptional examples of Tibetan books in his private collection, and also for sharing his knowledge and advice on the material aspects of books and paintings. I am very much in debt to Mr. Arthur Leeper for enabling me to undertake technical and material studies on Yongle covers from his collection. Further, I would like to thank my husband Prof. Tomasz Ważny, of the University of Arizona, for identifying the wood of these covers. I would also like to thank Yudru Tsomu, Associate Professor of the Center for Tibetan Studies of Sichuan Province and Bugang Chashingtsang for their help with the Yongle covers inscriptions and the translation and identification of Tibetan terms. x acknowledgements

I would also like to direct my warmest thanks to Tibetologist colleagues with whom I could always discuss ideas related to Tibetan books. I thank Dr. Hildegard Diemberger (University of Cambridge), Dr. Amy Heller (CNRS), Dr. Kirill Alexeev (University of St Petersburg), Dr. Alexander Zorin (Russian Academy of Sciences), Dr. Elena Pakhoutova (Rubin Museum), Dr. Orna Almogi (University of Hamburg), and Prof. Dorji Wangchuk (University of Hamburg) for their kind interest and constant support. For financial assistance, I am deeply grateful for the support I received from several sources over the course of the last six years. My gratitude is directed to the Ministry of Science and Higher Education of Poland for supporting my three-year project from 2007 through 2009, “The lost fragment of Wanli Kanjur in the Jagiellonian Library? The value of authenticity of Tibetan books from Pander Collection in Poland.” This project was realized together with Prof. Marek Mejor and Dr. Thupten Kunga Chashab, colleagues from Warsaw University. I am also very grateful to the Library of Congress for granting me the 2010 Florence Tan Moeson Award to conduct research at the Tibetan Collection of the Library of Congress, Washington DC and the Frederick Williamson Memorial Fund, Museum of Archaeology and Anthropology, University of Cambridge for funding the travel grant: Mapping Tibetan Paper. Finally, my gratitude is also directed to the Provost’s Author Support Fund at the University of Arizona for granting me award toward the publication of this book. I would like to thank Reeder (Wick) Dossett for reading the entire book. His many valuable comments made my text more fluent and consistent. I also thank Elizabeth Green and Kamila Janiszewska for reading fragments of the book and offering useful advice on English-language matters and my sister Dorota Dominiak for technical help with copy editing. Particular thanks are directed to Dr. Cynthia Col who helped prepare the final manuscript. Her help with final copyediting, indexing, and checking and correcting translations of Chinese and Tibetan terms is much appreciated. Last but not least, I would like to thank Patricia Radder of Brill for her support and understanding. Without the expert assistance and generous support of my Family and everyone involved, this book would not have been possible. In the end, I have to realize that there is no end to the corrections and additions. In order to share this work, I accept responsibility for any flaws that remain. Thank you to all! List of Illustrations

FIGURE Caption

1 Wood-working tool traces observed in UV light on the book cover 16 2 Dating of a wooden cover. The edge of the plank with a visible tree-ring structure 20 3a, b, c, d Fibers of Daphne or Edgeworthia sp. observed under microscope in polarized light 25 4a, b Stellera chamaejasme fibers observed under microscope in polarized light 27 5a, b Fibers of Ramie (Boehmeria sp.) stained with Herzberg from Tibetan Dunhuang manuscripts (IOL Tib J 1254 and IOL Tib J 1357) observed under a microscope in polarized light at 600× magnification 28 6a, b Fibers of paper mulberry (Broussonetia sp.) found in a Tibetan scroll (IOL Tib J 1560, vol. 85 bi-f. 2) from China’s Dunhuang caves under a microscope in polarized light 29 7 Paper mulberry fiber and its separated primary wall observed by SEM in 600× magnification 30 8 Thick-walled hemp fiber from a raw plant colored with Herzberg stain and observed under a microscope in polarized light at 200× magnification 31 9a, b Woven paper made with floating type of papermaking mold 34 10 Laid paper made with movable type of papermaking mold 37 11 Embossed stamp with inscription: Фабрика Наследников Сумкина (factory of Sumkin’s heirs) found on the machine-made paper of the Tibetan manuscript 38 12 Embossed stamp with inscription: Фабрики Платунова (factories of Platunov) found on the machine-made paper of the Tibetan manuscript 39 13a, b Cross-section of Daphne bast tissue. a. Irregular fiber walls and lumens visible on cross-section under a microscope observed at 1200× magnification. b. Starch in Daphne identified in fluorescence observed at 1800× magnification 41 14 Highly sized and polished surface of a paper made of Thymeleaceae family plants fibers in Tibetan manuscript 45 15 Tibetan Prajñāparāmitā manuscripts used as an element of a ritual ensuring a good harvest in Kyirong Valley, Western Tibet 48 16 Pothi book on palm leaves 50 xii list of illustrations

17 Typical Tibetan pothi format books in Braga Monastery, Nepal 54 18 Tibetan pothi format book wrapped in cloth 54 19 Loose wooden book cover originally from Tibetan book 57 20 Palm-leaf pothi book with covers which have holes through which a single binding cord was drawn and knotted on their outer sides 57 21 A seventeenth century handwritten London Selkar (Sel dkar) Kanjur (OR 6724) 58 22 Large pothi format the Tibetan Book of the Dead 59 23 Concertina-format of Sino-Tibetan books from the Pander Collection 61 24 Format of unfolded Tibetan historical document 62 25a, b Tibetan stitched book on paper from Dunhuang (IOL Tib J 401) 64 26a, b Sino-Tibetan stitched book on paper from Dunhuang (IOL Tib J 530) 65 27a, b, c, d The Tibetan manuscript written in dbu med script, made of pothi leaves folded in half, and stitched with red thread (OR 15193) 66 28 A small prayer printed in Tibetan language and found in Khara Khoto (Kh-Tib.67) 68 29a, b, c Bound book written in gold with Tibetan religious text from Bhutan (MS 13162) 70 30a, b Bound manuscript composed of four gatherings and stitched with linen cord (CC 114-115) 72 31a, b, c, d Tibetan bound manuscript composed of four gatherings (BL 14727 1-2) 73 32a, b, c, d Bound manuscript of four gatherings written on one-layered paper (CC 101) 74 33 Text printed in gold on blue paper imitating an aesthetic of writing in gold on indigo dyed paper 78 34 Verso side of a wooden cover wrapped with paper dyed indigo with two silk curtains attached to it at the top (OR 15190) 83 35 The internal frontispiece of a book with the silk curtains attached at the top (OR 15190) 83 36 Verso of the inner cover of the Tibetan manuscript of the Prajñāpāramitā, Sherabkyi Paröltu Chinpa written in gold on black-inked paper from Western Tibet, Fifteenth century (2006.028a,b) 84 37 Illuminated inner cover written in gold on black-inked paper and made of layered paper joined together with leather strips/clasps. Prajñāpāramitā Cover and Frontispiece (2006.028a,b) 84 38 Illuminated inner cover written in gold on layered black-inked paper mounted with leather or paper strips Prajñāpāramitā Cover and Frontispiece (2006.028a,b) 85 list of illustrations xiii

39 The front page of Tibetan Selkar (sel dkar) Kanjur (OR 6724) 87 40 Text written in ‘raised gold’ technique on the front page of volume 65 of Tibetan Selkar (sel dkar) Kanjur (OR 6724) 87 41 Front page decorated with ornamented frames in volume 65 of the Tibetan Selkar (sel dkar) Kanjur (OR 6724) 88 42a, b Miniatures mounted on the front page of volume 65 of Tibetan Selkar (sel dkar) Kanjur (OR 6724) 88 43 The curtains stitched together on the bottom corners and attached on the top with three leather clips on the verso of the front page of Tibetan Selkar (sel dkar) Kanjur (OR 6724) 89 44 Frontispiece of the Tibetan manuscript of the Prajñāpāramitā, Sherabkyi Paröltu Chinpa from Western Tibet, Fifteenth century (2006.028a,b) 90 45 Verso of the inner cover of the Tibetan manuscript of the Prajñāpāramitā, Sherabkyi Paröltu Chinpa from Western Tibet, Fifteenth century (ET 77) 90 46 Frontispiece of the Tibetan manuscript of the Prajñāpāramitā, Sherabkyi Paröltu Chinpa from Western Tibet, Fifteenth century (ET 77) 91 47 Illuminated fourteenth-century manuscript in gold ink on a black background from Western Tibet 92 48 Illuminated manuscript ’phags pa shes rab kyi pha rol tu phyin pa rdo rje gcod pa written in gold ink on a black background (MAP 4323) 92 49 Front page from manuscript describing the life of Gshen rab mi bo in gold and silver ink and illuminations representing four stupas on both sides 93 50 Illuminated manuscript with text from the Diamond Sutra written in gold ink on a black background (DSB 1) 93 51 Linear sketch of a dragon in Tibetan bound manuscript (SKAZ 1813) 95 52 The manuscript (the Lotus Sutra) features passages of text marked with gold ink among silver written text 97 53 The manuscript with passages of text marked with gold ink among silver written text that features a golden triangle composed of gold-written words 98 54 Yellow scribal guidelines differentiating a manuscript from a print in the Tibetan Selkar (sel dkar) Kanjur 99 55 Edgeworthia/Daphne sp. fibers at 50× magnification observed in polarized light in Tibetan Prajñāpāramitā volume from the Bicher temple in Dolpo, Nepal 105 xiv list of illustrations

56 Edgeworthia/Daphne sp. fiber at 200× magnification observed in polarized light in the Tibetan Prajñāpāramitā from the Bicher temple, Dolpo, Nepal 106 57 Daphne/Edgeworthia sp. fibers at 150× magnification observed in polarized light. Tibetan Selkar Kanjur 106 58 Daphne/Edgeworthia sp. fibers stained olive-grey with Herzberg at 600× magnification observed in polarized light. Tibetan Selkar Kanjur 107 59 Broussonetia sp. fibers at 50×magnification in the Tibetan vertical scroll from Dunhuang (Dx 178) 109 60 Broussonetia sp. fibers stained reddish with Herzberg in 200× magnification in the Tibetan vertical scroll from Dunhuang (Dx 178) 109 61 Stellera chamaejasme (re lcag pa) fibers with wide-irregular lumen and narrow cell-walls at 600× magnification observed in polarized light in the Tibetan manuscript of Prajñāpāramitā, Sherabkyi Paröltu Chinpa 111 62 Edgeworthia/Daphne sp. and Stellera chameajasme fibers at 600× magnification observed in polarized light in manuscript from Western Tibet 112 63 Edgeworthia/Daphne sp. and Stellera chameajasme fibers at 600× magnification observed in polarized light in manuscript from Western Tibet 112 64 Daphne/Edgeworthia sp. and Stellera sp. fibers at 200× magnification from one of the prayer pages of the Prajñāpāramitā volumes 113 65 Daphne/Edgeworthia sp. fibers at 200× magnification observed in polarized light in one of the prayer pages of the Prajñāpāramitā volumes from the Bicher sanctuary, roughly dated to the fifteenth and sixteenth centuries 114 66 Tibetan book printed in the 1960s in India using lithography 118 67 An example of a block print from Turfan in Uighur language executed in six vertical text lines per page on a single-layered, woven type of paper (U 4753) 120 68 Volume of The Kalachakra Tantra printed from woodblocks carved for the funeral ceremonies of Kublai Khan (1215–1294) sponsored by Kublai Khan’s wife and son (TBRC W4CZ75) 124 69 A fragment of folio from The Kalachakra Tantra printed from old woodblocks that were carved for the funeral ceremonies of the Kublai Khan (1215–1294) under the sponsorship of Kublai Khan’s wife and son (TBRC W4CZ75) 124 list of illustrations xv

70 Fragment of a woodblock for Narthang Monastery prepared for carving by craftsman from Nyemo 128 71 Fragment of a woodblock for Narthang Monastery being carved by craftsman from Nyemo 128 72a, b Recto and verso side of folio 12 of the Yongle Kanjur volume 138 73a, b The miniatures of the Wanli Supplement volume’s cover depict the Emperor, and the Tibetan master, in the lower outer corners (TTib 1803.7 13 vol. 2) 141 74 The verso of the Wanli Supplement volume’s cover (TTib 1803.7 13 vol. 2) 142 75 The fragment of the Yongle cover of Tibetan Kanjur (originally Sera Monastery set) containing a curved seal on its verso side (ET 107) 143 76 The fragment of the Yongle cover of Tibetan Kanjur (originally the Sakya Monastery set preserved later in Potala, Lhasa) containing a curved seal on its verso side (ET 87) 144 77a, b Microscopic view of sandal wood in the Sakya type cover of the Yongle Tibetan Kanjur from the collection of Arthur Leeper 145 78 The cover of the Yongle Tibetan Kanjur from the Potala set (originally Sakya Monastery set) of the Yongle Kanjur, which has a flaming jewel in the middle flanked by four of the eight auspicious symbols, such as the wheel of law, victory banner, two fish, and treasure vase (ET 87) 146 79 The Sera copy of Yongle Kanjur with a ‘precious pearl in flames’ carved in the middle of the top cover (ET 107) 146 80 Types of painted long fore-edges in volume 54 of the Pander Collection containing Wanli Kanjur 148 81 Types of painted short fore-edges in volume 54 of the Pander Collection containing Wanli Kanjur 148 82 Types of painted fore-edges in volume 47 of the Pander Collection containing Wanli Kanjur 149 83 The long fore-edges of the Wanli Supplement volumes from the Pander Collection painted with a flower ornament against a red background 150 84a, b Painted fore-edges in volume 1 (ka, ’dul ba, bzhi) of the Berlin Manuscript Kanjur 151 85 Folio 3 layout of volume 26 from the Wanli set from Pander Collection 152 86 A close-up of text in the Yongle folio 12 printed in red ink 152 87 The Berlin manuscript Kanjur page layout (folio 4, ka, ’dul ba, bzhi) 153 xvi list of illustrations

88 Page image from the Wanli Kanjur Supplement (TTib 1803.7 13 vol. 1) 154 89 A close-up of text in folio 3 of volume 26 from the Wanli set from Pander Collection 155 90 The parts of headings visible in folio 4 of vol. 1 of the manuscript Kanjur of Berlin feature a straight line of text and characteristic ‘spurs’ in the right side of the upper part of letters such as ‘a’ ‘ka.’ 155 91 The paper mulberry fibers enveloped by a transparent membrane in magnification 600× magnification in the paper of Yongle folio 12 157 92 Paper mulberry and bamboo fibers composing the paper of the Mongolian Kanjur 160 93 A large bamboo vessel at 600× magnification in the paper of the Mongolian Kanjur 160 94 Bamboo fibers composing the paper of the printed Mongolian Kanjur in polarized light visible on an image at 100× magnification 161 95 Bamboo fibers composing the paper of the printed Mongolian Kanjur at 200× magnification 161 96 The Cone Kanjur acquired by Joseph Rock in 1928 and preserved in the Library of Congress 164 97 Page outline of the Cone Kanjur folio 1 of volume 92 yum (Prajnaparamita Sutras) 165 98 Fragment of text from volume 1 mdo (sutra) of the Cone Kanjur 166 99 Close-up of text from volume 35 ’dul ba (vinaya) of the Cone Kanjur 167 100 Paper of the Cone Kanjur volume 108 dkar chag (index) 168 101 Paper of the Cone Kanjur in volume 1 mdo (sutra) shows the presence of paper mulberry and straw fibers in 60× magnification 169 102 Paper of the Cone Kanjur in volume 1 mdo (sutra) composed of the paper mulberry fiber enveloped by a transparent membrane in 600× magnification 169 103 Page outline features of the Derge Kanjur 171 104 Woven type of paper made in a floating mold viewed against light in copy of the Derge Kanjur 171 105 Copy of the Derge Kanjur 172 106 Daphne or Edgeworthia sp. and Stellera sp. fibers observed in polarized light at 100× magnification in the paper of volume 77 of the Derge Kanjur 173 107 Daphne or Edgeworthia sp. fibers observed in polarized light at 600× magnification in the paper of the Narthang Kanjur 176 list of illustrations xvii

108 Daphne or Edgeworthia sp. fibers observed in polarized light at 600× magnification in the paper of the Lhasa Kanjur 177 109 Edgeworthia gardineri growing in the Khumbu region, Nepal 185 110 Stellera chameajasme found in Kyirong, Western Tibet 186 111 The root of Stellera chamaejasme shown by a former papermaker in Dobe Shang, Western Tibet 187 112 Drying process during papermaking in the Kullu Valley, Himachal Pradesh, 1940s 192 113 The process of making paper in the Kullu Valley, Himachal Pradesh, 1940s 197 114 Process of shaping a sheet of paper in the Kullu Valley, Himachal Pradesh, 1940s 197 115 Drying process during papermaking in the Kullu Valley, Himachal Pradesh, 1940s 198 116 A large size ‘floating’ mold, constructed with a wooden frame and attached woven textile, placed in water (a stream) in Gyantse, c. 1910–1920 199 117 Monks reciting books at Bya mang po in Western Tibet 202 118 Discoloration caused by animal glue used for joining several layers of paper together in Tibetan manuscript (MAP 14136) 208 119 Discoloration caused by interaction of light and other components of paper 210 120 Tibetan books stored in a general room at the Nyingma monastery in Nelung, Western Tibet 214 121 The results of water damage on a book from the collection of the Museum of Asia and Pacific in Warsaw (MAP 12385) 215 122 Collection of texts kept together as mini-library 216 123 Traditional shelving storage of books at Tengboche monastery, Nepal 217 124 Traditional shelving storage of books at Nyingma monastery in Nelung, Western Tibet 218 125 Traditional storage of books in Namche Basar Monastery, Nepal 219 126 Tibetan scriptures abandoned in caves near Chingkar, Western Tibet 219 127 Fragment of a scroll taken out from praying mills during conservation treatment 222 128 Leaves of a Tibetan book (MAP 14144) restored with a leaf-casting method 223 129 Deterioration of Tibetan gold manuscript ‘Life of Gshen rab mi bo’ caused by components of the paints 224 xviii list of illustrations

Photographs Credits Photographs used as illustrative materials are made by an author, otherwise indicated in caption. Author thanks to following individuals and Institutions for permission to publish photographs taken from book collections:

The British Library, London, UK The Library of Congress, Washington DC, USA The R.R.E. Collection (Professor Richard Ernst Collection), Winterthur, Switzerland The Herbert F. Johnson Museum of Art, Cornell University, USA The Asia and Pacific Museum in Warsaw, Poland The Pomeranian Library in Szczecin, Poland The National Museum in Warsaw, Poland The Columbia University Library, New York, USA The Library of Tibetan Works and Archives in Dharamsala, India The Berlin Sate Library, Germany The Tibetan Buddhist Resource Center, Cambridge, Massachusetts, USA The Harvard Yenching Library, Cambridge, Massachusetts, USA The Special Collections Library, University of Michigan in Ann Arbor, USA The Jagiellonian University Library, Cracow, Poland The Institute of Oriental Manuscripts of the Russian Academy of Sciences, St. Petersburg, Russia CHAPTER 1 Introduction

The Starting Point

In the beginning, my interest in Tibetan books was initiated by the wish to preserve materials of Tibetan culture. While studying paper conservation at the Fine Arts Academy in Warsaw, I was struck with the thought of choosing a Tibetan book for my diploma project. I already possessed basic practical skills to be a conservator of European books and papers and wanted to explore new areas. In 1997 a wood scientist and dendrochronologist at the Fine Arts Academy in Warsaw, Dr. Tomasz Ważny, who later became my husband, proposed that I assist his work in Nepal. This invitation is how I joined the German dendrochronological expedition to Mustang and Manang organized within the Nepal- German High Mountain Archaeology Project. During the trip, I documented in writing and photography the taking of samples from trees, fortresses, temples and other traditional architecture, and created many photos, drawings, and paintings of Nepalese people, landscapes, works of art and practically every- thing that caught my eye along the way. Fascinated, I seized another chance to go to the Khumbu region a couple of months later. Tomasz and I took a bus from Kathmandu to Jiri and spent a large part of this trip on the roof of the bus, looking from its height at the landscape, villages, and people who left and entered the bus over the ten-hour drive across the 120 km route. Later, as walked by foot from Jiri to Namche Basar, we came across Nuntala, a small village located about three thousand meters above sea level. There I directly observed methods of making paper in the Himalayas; as I watched, I was reminded of methods in articles and books about the very beginning of papermaking history. At that time I began to collect information more purposefully and made a commitment about my future career. Back to Polish reality for another two years I worked on the conservation of the Tibetan Diamond Sutra from the collection of the Asia and Pacific Museum in Warsaw, brought from Nepal by the director of the museum, Mr. Andrzej Wawrzyniak. That work afforded me many pleasures and opportunities to learn about the art and material culture of Tibet and Nepal. It also initiated my long-term cooperation with the museum and its staff. Mr. Wawrzyniak, in the past a Polish diplomat and sailor, was a magical personality; sitting in his office with a smoking pipe, he attracted guests with many Asian curiosities that he

© koninklijke brill nv, leiden, ��4 | doi 10.1163/9789004275058_002 2 chapter 1 had collected over the years while in Nepal, Indonesia, and other Asian countries. The pleasant aroma of tobacco, incense, and various fragments of wooden artefacts from the museum continues to waft through my mind bringing back pleasant my memories. The chirpings of birds and cicadas—very unusual for Poland—that he had brought with him from Asia enlivened his office provided yet another source of fascination for me. Wawrzyniak even had Vietnamese craftsmen build a small wooden temple among the trees in front of the museum; amidst the surrounding high, grey, communist-style block buildings, this provided a welcome oasis and respite from the hustle and bustle of Warsaw’s busy Solec Street. Over the course of more than ten years, I enjoyed visiting the museum regularly. After I realizing that the museum possessed a collection of more than a hundred Tibetan and Nepalese books that very few scholars had studied, the chance to develop this subject beckoned. This is how the next stage of my interest in Tibetan books began. I planned to base my doctoral dissertation on the Tibetan books preserved in Polish collections. Aside from the above-mentioned collection of the Asia and Pacific Museum, I was not sure whether I could find anything else; at that time, Tibetan art in general was not catalogued in my home country. This is why my work began with a search for other Tibetan books in Poland, usually by asking whomever I could if he/she was aware of the existence of any books with loose leaves. In fact, this method turned out to be the most effective way to proceed.

Borrowing the Term “Archaeology”

The term ‘archaeology of the book’ is used here in the context of studying the physical make-up and production of a given volume. It comprises analyses of the structures of books and an interpretation of technological aspects. It is sometimes said that real history is based on written documents, which implies that oral tradition is not as reliable. Even though written documents play an important role in the study of history, important evidence comes from a variety of sources, such as documents, artefacts, oral tradition, etc. Written documents such as manuscripts, printed books, or inscribed pillars carry statements but are also artefacts, and their physical properties can be studied using methods similar to those used to study sculpture, painting, or common material objects. Indeed, ignoring the non-linguistic aspects of books may deprive us of important historical evidence. While the history of Tibetan literature has received some attention within Tibetan Studies, a systematic study of the physical features of books, such as formats, binding styles, and materials, is almost completely lacking. Introduction 3

Contributions by Cristina Scherrer-Schaub, Amy Heller, and Kurtis R. Schaeffer are notable exceptions.1 This essay is about the physical features of Tibetan books, which exist in several formats and incorporate a wide variety of materials. This essay will not use the term ‘codicology’ in the context of old Tibetan manuscripts. Codicology is related to the study of the ‘codex’, which is a bound book. Between the second and fourth centuries, the preservation of writing in the Latin world moved from papyrus stored in rolls to parchment stored in a “codex,” or pages bound as books. We might say that codicology is the archaeology of the book defined as the codex, with ‘archaeology’ understood as the search for and interpretation of artefacts. Another reason for using the term ‘archaeology’ is that the majority of Tibetan books feature the pothi format, a construction of loose leaves.

Books Written in Tibetan

For the purpose of this book, the Tibetan language was the most general criterion for the selection of items, and served as a starting point for a preliminary typology of Tibetan manuscripts and printed books. The physical appearance of Tibetan books is a result of a chain of interactions between culture, econ- omy, art, and craftsmanship developed by communities affiliated with a variety of nations and cultures. Thus, when constructing a typology of particular features, such as page layout, format, binding style, and materials, it is possible to classify the majority of books into groups and relate them to particular local ‘book cultures.’ We can look at a book from many angles; with respect to physical structure alone, we can discuss ‘national scripts’, ‘national fonts’, ‘national characteristics of book arrangement (layout), binding, material used’—each of these features might represent different meanings. And indeed, this study shows that Tibetan books are composed of a variety of elements (issues) that have a multicultural character. Also, Tibetan language has never been unified and cannot be characterized by a common territory. The main reasons are poor economic development as well as inconvenient transportation and communication systems. Tibet has been scattered into four provincial prefectures and an autonomous region. The establishment of Songcen Gampo’s Tibet did not cause language unification; this linguistic diversity continued during the period of decentralization after

1 For example see: Scherrer-Schaub 1999: 3–36; Scherrer-Schaub and Bonani 2002: 184–215; Schaeffer 2009; Heller 2009: 77–194. 4 chapter 1 the fall of the kingdom in the ninth century—not only when Buddhism flourished in Tibet but also when printing appeared. Throughout the periods during which ancient Tibetan texts were being developed, the geographical and cli- matic diversity of highly mountainous areas, the character of nomadic cultures, and the availability of materials all conditioned the production of books written in Tibetan. Additionally, Tibetan and other great religious languages throughout history have served different nations at the same time. Tibetan has been used in literature across a large area including the Himalayan regions of South Asia. The full geographical range in which Tibetan has served as a language of learn- ing, however, is much greater even than this. With the promulgation of Tibetan Buddhism among the Mongols and Manchus, literary Tibetan became a common medium of communication among Central Asian Buddhists by the end of the seventeenth century, and was used at the beginning of the twentieth century as far west as Astrakhan, where the Volga River flows into the Caspian Sea, and as far east as Beijing. There has been a revival of the study and use of literary Tibetan to varying degrees in Buryatia and Tuva in the present Russian Federation, among ethnic Mongols and Yi in China, and in Mongolia itself. This is why books written in Tibetan differ widely in their form and in materials used—the various communities that used the Tibetan language did not equally adapt all Tibetan bookmaking techniques and did not use the same materials. The methods of each local ‘book culture’ were preferable to any outside standard.

Statistics in Book History

Viewed as part of world heritage, ancient books are not only vehicles for content, but also very important sources of information about technology, materials used for production, and also associated art techniques. There were several reasons for this study to be undertaken. At a micro-level, the study of individual books provides us with valuable information about a particular book and possibly its provenance. However, from a broader perspective, the examination of many books and book fragments aids in the reconstruction of the history of crafts connected with bookmaking, such as papermaking, ink production, and the art of scribing. And even more broadly, the examination of each book facilitates the collecting of evidence on book production from particular periods of time and regions in Tibet; this to enable us to understand better the role books play in Tibetan culture. In other words, this kind of material research, when performed on enough representative groups of books, will set the stage for writing the history of the book in Tibet. This book presents the Introduction 5 results of the current stage of my research with an emphasis on Tibetan manuscripts from Dunhuang, early printed editions of Tibetan Kanjurs, and richly illuminated manuscripts from Western and Central Tibet. In any kind of study of book collections, a statistical approach is important. It is not possible to base one’s conclusions only on a small group of books taken from an accidentally selected single collection, place, owner, etc. Statistically, results should be presented only for particular groups of books that are drawn from the same regional culture of book production during a unified period of time. When attempting a reconstruction of the history of the book in Tibet, it does not make much sense to compare books that are unrelated to each other. Although every observation is important, individual observations can only contribute to the database as a whole, and should not be understood as representative of all Tibetan book history. Statistics are nearly as important for historians and experts as information about individual items. With hundreds of Tibetan books viewed, it becomes necessary to provide means for visualizing the statistical properties of examined collections. Clearly, more research is needed before this field of study can fully contribute to the reconstruction of the history of the Tibetan book. However, I hope this book will inspire scholars of other disciplines to contribute to such a true history based on preserved objects.

My Fieldwork, Sources, and Experiments

While carrying out my research on Tibetan books over the course of more than ten years, I undertook a couple of trips to Nepal, Tibet (China), and Northern India. However, the majority of my fieldwork research was conducted in the reading rooms and storage areas of libraries, museums, and private collections. Through a questionnaire and detailed library research I found many interesting and valuable objects; since 2000, I have gradually studied and described a number of Tibetan books that I have tracked down (Appendices 1–5). For single works as well as for books from collections of various sizes, I viewed the chosen collections on the spot and singled out particular books for detailed research. Later, I also explored written sources, mostly secondary literature; additionally, I supplemented this research by interviewing Tibetan artisans and monks concerning scribing the texts (calligraphy), ink production, the creation of blue paper, carpentry, and metalworking on book covers. Interviewing Tibetan craftsmen was an important component of the reconstruction of practical work, which, of course, might differ regionally. Independent of regional differences, the contributions of Tibetan artisans and monks sheds considerable light on the procedures of book production and organization of the work 6 chapter 1 of craftsmen and artists. Finally, I did some ‘experimental manuscriptology’ by personally reconstructing art techniques and copying fragments of selected books. I established the following guidelines in my selection of books for research: search for collections brought from Central Asia and study their history and provenance; select Tibetan books with particular textual content (canonical texts); and, finally, search for books with technological or material significance. The last criterion led to the examination of diverse manuscripts and printed books with documented or textual clues regarding dating and place of origin. I furthermore examined Kanjur volumes from different editions to compare books that were executed both in manuscript and print formats, and products of different places and from different time periods. Another criterion for selecting a book fragment was the range of different types of paper and variety of book formats. For the purpose of this essay, the most essential surveys I undertook were in the British Library in London, UK (2005–until now), the Jagiellonian University Library in Cracow, Poland (2002–until now), the University of Michigan Special Collections Library, Ann Arbor, MI, USA, the Harvard-Yenching Library, Cambridge, MA, USA, the State Library in Berlin, Germany, the Library of Congress in Washington, DC, USA (May 2010), the Library of Tibetan Works and Archives in Dharamsala, India (March 2005), the Herbert F. Johnson Museum of Art at Cornell University, Ithaca, NY, USA (January–April 2008), the Rare Book and Manuscript Collection of Columbia University Library, NY, NY, USA (July 2007 and September 2010), and the National Library in Beijing, China (October 2007). The studied collections originated in a variety of time periods and regions of Tibet and Central Asia. This seemingly wide range of books helped me to achieve greater objectivity. Since the character of a particular collection usually depended on the purpose of and availability to the collectors, I tried to avoid being trapped by the view of a particular collector. I wanted to examine books that would later allow me to write their cross-sectional history through regions of Tibet, and also, in time, allow for some general chronology. This is why I examined Tibetan manuscripts from Dunhuang, which are the oldest preserved Tibetan books known, for their particular type of paper and book format (Appendix 1). The cave where the books were found was created in the early ninth century by a Chinese monk called Hongbian, the leader of the Buddhist community at Dunhuang.2 After his death in 862, the cave became a shrine to Hongbian; a statue of the monk was installed, and

2 For Tibetan manuscripts from Dunhuang preserved in Russian collections see: Savitzkiy 1991. Introduction 7 manuscripts belonging to him were deposited there. Over the following 150 years, increasing numbers of manuscripts were deposited in the cave. At some point, the statue of Hongbian was relocated to another cave, and manuscript storage became its primary function.3 The almost fifty examined manuscripts written in Tibetan found in Library Cave 17 in Dunhuang are dated to around the eighth to tenth centuries and representative of the oldest period of Tibetan book history and development of crafts such as papermaking and calligraphy. I conducted this research in cooperation with Sam van Schaik and the International Dunhuang Project at the British Library, and the IDP is adding all my results to its database. As my research on Tibetan book collections continued, new subjects for inquiry arose; in particular, at the British Library I realized that not all Tibetan books are in the pothi format. Since that experience, therefore, I also documented bookbinding diversity. I selected representative books from the British Library to illustrate different types of Tibetan binding (Appendix 1).4 My next experimental group of objects comprises illuminated manuscripts executed with a variety of techniques, especially writing in gold. These later evolved into a chapter on ‘Tibetan illuminated manuscripts.’ For this purpose I viewed the British Library Tibetan collection together with its curator Burkhard Quessel, and I chose two manuscripts written in gold against a blue or black background to represent this special form of Tibetan craftsmanship.5 For research on manuscripts written in gold I also examined the illuminated cover and frontispiece from the collection of the Herbert F. Johnson Museum of Art at Cornell University, and the frontispiece and cover from the private R.R.E. collection in Winterthur.6 Additionally, I examined five gold manuscripts from Columbia University Library, Rare Book and Manuscript Collection, and one in the Library of Congress in Washington, DC. Finally, I approached the most studied aspect of Tibetan book history— printing culture. I focused on the early history of Tibetan printing and its development. I started from the early history and technology of printing in Central Asia, and later I devoted a chapter to repetitive material features

3 See, Rong 2000; Imaeda 2008; van Schaik and Galambos 2012; and Helman-Ważny and van Schaik 2012. 4 Books singled out to represent different styles of Tibetan binding are: 13162, MS 13092, OR 11376, OR 14727 (1–2), OR 14728, OR 15190, OR 15193, OR MS 12163, TIB CC 074, TIB CC 101, TIB CC 114–115. 5 These books were MS 13162 and OR 15190. 6 Agnieszka Helman-Ważny. Examination of the Tibetan illuminated book cover and frontispiece of Eight Thousand Lines Perfection of Wisdom Sūtra (Acc. Number: 2006.028 a,b). Report for the H. Johnson Museum of Art, Cornell University prepared in 2008. 8 chapter 1

(patterns) in selected editions of Tibetan Kanjurs (Appendices 2–5). The idea for this choice developed when I was surveying the Jagiellonian University Library in Cracow, which has fortunately preserved the Pander Collection, originally from the Berlin State Library.7 Eugen Pander’s 865 books constitute a collection of immense value. It contains the Berlin Wanli Kanjur, deemed to be lost during World War II.8 Eugen Pander was a German professor born in 1854 in Livonia as a Russian citizen, and was a nineteenth-century pioneer of Tibetan Buddhist studies, especially known for his work in the field of Tantric Iconography.9 His collection of Tibetan books purchased during his stay in Beijing is one of a few important early Western collections of Tibetan literature. This discovery created new possibilities for reviewing the history of the early stages of Tibetan canonical printing. After preliminary research on the authentication of the Wanli Kanjur in Cracow, and study of the history of the Pander Collection, I expanded research towards the examination of all the available editions of the Tibetan Kanjur produced in China. These comprise one folio of the Yongle Kanjur preserved in the Special Collections Library, University of Michigan; fragments of the Wanli Kanjur edition and its supplement and a volume from one of the later editions of the Kangxi Kanjur both preserved in the Jagiellonian University Library in Cracow, Poland; two volumes of the supplement to the Wanli Kanjur located at the Harvard-Yenching Library; and the Berlin Kanjur (manuscript Beck), which was copied from the Wanli Kanjur set and preserved in the Berlin State Library. Next, I conducted a supplemental study of Kanjur sets produced in Tibet. These included the Narthang and Lhasa Kanjurs preserved in the Berlin State Library. Recently, I have also examined the Cone and Derge Kanjur editions preserved in the Library of Congress, Washington, DC. The Library of Congress holds the Derge Kanjur acquired by William Rockhill in 1908 and the Cone Kanjur redaction acquired by Joseph Rock in 1928. Both editions are important for comparing different editions of Tibetan Kanjurs. While surveying the Tibetan book collections preserved in a variety of places I heard many amazing stories about the provenance of these books. Although these stories usually concerned book fragments only, sometimes an adventure-filled tale of provenance was shared by an entire collection—as in the case of the Pander books in Cracow, which I discuss below. These stories opened my eyes to a new area of study: the history of Tibetan books collec-

7 For the history and preliminary examination of this collection, see: Helman-Ważny 2009; or Mejor et al. 2010. 8 Eimer 2000: 27–51. 9 Pander 1890; Lohia 1994: 33–270. Introduction 9 tions. In the further perspective, this also greatly helped in the identification of particular book fragments. Unidentified objects exist in a surprisingly large numbers even in known and catalogued collections. One of the most interesting stories concerns the Pander Collection, which passed from Chinese and Tibetan hands, to German and Polish hands. Eugen Pander brought his books to Berlin in 1889.10 Pander never clearly states where he obtained his collection; however, he relates that among his acquisitions, he was very lucky to obtain fifty-nine volumes of the Yongle, Wanli, and Jiajing editions of the Kanjur from an imperial monastery in Beijing.11 Pander had a connection with Yonghegong, or the Temple of Eternal Peace, in Beijing, and this was probably where he obtained his books.12 While in Beijing during October 2007, I made an attempt to find more information about Eugen Pander and his acquisitions; unfortunately, all I discovered was a former library associated with Yonghegong that has not functioned as a library since the Cultural Revolution. Located in the quarter behind Yonghegong, at present, this building is abandoned—only traces of its former greatness remain. The Pander Collection was first donated to the Museum für Völkerkunde in Berlin13 in 1889, and then to Königliche Bibliothek.14 During the closing stages of World War II, collections from the Prussian State Library (Preußische Staatsbibliothek) in Berlin were evacuated to the Silesian Castle Fürstenstein (Książ) and later to the Cistercian Brothers Monastery Grüssau (Krzeszów) to protect them.15 When Lower Silesia became Polish territory after the war, the Polish state claimed the collection as abandoned property. Thus, the collection found its way to Poland in 1946–47. A group of researchers led by Dr. Stanisław Sierotwiński, delegate of the Ministry of Education from the Jagiellonian University Library, transported the Pander Collection, along with other items, to the main library seat in Cracow.16 This collection is still situated in this library in accordance with the rights of deposit of the Polish Government.17

10 For details see: Helman-Ważny 2009 and Mejor et al. 2010: 15–16. 11 Eimer 2000: 29. 12 Pander 1890: 7; Lohia 1994: 62. 13 Museum für Völkerkunde, Berlin was renamed for Ethnologisches Museum in Berlin from 1993. 14 Nowadays Staatsbibliothek zu Berlin (former Preußische Staatsbibliothek). 15 The circumstances of the Berlinka collection acquisition, including the Pander Collection, by the Jagiellonian University Library are described in: Pietrzyk 2005: 81–87. 16 Jaglarz 2008: 49. 17 The status quo of Pander’s books: the part of the former Prussian State Library collection called Berlinka in Poland is referred to as property of the Polish nation now. Poland claims that it should retain ownership of the Berlinka as compensation for Polish historical 10 chapter 1

Referred to as the Berlinka collection, the Polish Government kept its existence a secret until 1977; then, Polish First Secretary of The Polish United Workers’ Party Edward Gierek gave East German leader Erich Honecker seven pieces of music manuscripts from the Berlinka collection, including Mozart’s original manuscript of Die Zauberflöte and Beethoven’s notes for his Ninth Symphony, as a gift.18 It took more than sixty years to make this collection accessible to scholars and a wider audience. I was the first to view it in 2003 after the lengthy period when Pander’s books were forgotten in the storage area of the Jagiellonian University Library. Just four years earlier, Helmut Eimer had described this collection as lost during World War II. Another interesting story concerns a mysterious folio (307 from Brgyad ston ’pa) in the Library of Congress. As curator of Tibetan collection Susan Meinheit told me, in 1941 it was reported to Poleman, head of the Oriental Department in the Library at that time, that this folio was found by cleaners on the floor. The folio was in an envelope sent to the Library of Congress in 1923 with a post- age stamp from Hasting, UK (with a note: “registered Hasting”). At first glance this folio looked very familiar to me, and I found that it closely resembled the Sel dkar edition set in the British Library. However, after checking all measurements and comparing these with my previous research on the Sel dkar (London) Kanjur set, I realized that it could not be the British Library set. This raises the question of whether any other set of Sel dkar Kanjur exists in the British Library, and opens a new path of inquiry; that is, the examination of preserved books should lead to understanding the characteristic of particular book production centers in Tibet. In March 2005, I surveyed the Library of Tibetan Works and Archives founded in 1970 by the XIV Dalai Lama of Tibet. This is one of the most important institutions in the world dedicated to the preservation and dissemination of Tibetan culture. There are more than seventy thousand Tibetan manuscripts and documents in this library, and the collection is still growing. The purpose of my visit was carry-out a conservation survey for this collection. Additionally, three manuscripts, a couple of freelance wooden covers, and a selection of administrative documents were particularly examined.19 I also had the oppor-

collections destroyed or looted by Germans during World War II. The German media refers to the Berlinka as the ‘last German prisoner of war,’ and claims that Poland is in violation of the Hague Convention of 1907. To this day, each side claims this collection. 18 Pietrzyk 2008: 16. 19 The manuscripts under study have acc. No. 14459, 14532, 23585. At the same time, about 300 historical and administrative documents were catalogued and examined by Prof. Schwiegers’s team of researchers from Bonn University. Introduction 11 tunity to examine writing tools and two gold manuscripts from the collection of the Tibetan Museum situated in the same building. I studied the paper of most of the examined books mentioned above. This research was supported by a field study concerning papermaking technology and the associated plants used as raw material in the production of Tibetan paper. CHAPTER 2 Methods: An Uneasy Alliance of Science and History

The everyday essence of cultures from the past and also the heritage and achievements of peoples and nation are coded in extant books. This deposit of knowledge can be textual, but messages from the past can also be sealed within the structure and form of books; to decode and read this material we need interdisciplinary tools. Research on the materiality of Tibetan books has not been undertaken in a systematic manner until now. Moreover, in the case of Tibetan books the interdisciplinary character of manuscript studies gives rise to many methodological problems. The lack of obvious or widely agreed-upon methods makes methodology itself still a field of study and experiment; as yet, there are no ready-to- apply guidelines. Additionally, the terminology is not unified, and is often dependent on the particular background of individual scholars. The obvious truth is that results obtained from scientific study and historical records should complete and inform each other. To achieve such a level of communication between disparate disciplines, however, is challenging. Historians rarely undertake the scientific examination of books, and tend to rely only on their textual context; at the same time, material scientists are usually not equipped with adequate knowledge of the source language and culture to interpret the results of their analyses without the help of linguists and historians. In carrying out the research necessary to establish a preliminary typology of the features of Tibetan books and to contribute to their material history, I drew on methods borrowed from different disciplines of both science and history, such as library science, bibliology, bibliography, codicology, paleography, material science, statistics, art history, the history of crafts, archaeometry, and archaeology. My main purpose was to group all examined books according to their physical features. It often proved helpful to link particular items to already-identified groups characterized by particular technological or material features. It proved to be easier to classify particular books into categories on a visual basis than by textual content only.

Many Typologies of Book Features

The initial stages of my investigations have always entailed library research, selection of books, and preliminary descriptions of the physical features of the books based on visual inspection, such as page outline features, format, and, in particular, examination of types of paper. Secondly, the genre of the book was preliminarily recognized based on the library catalogue entry or, in the case of un-catalogued items, with the help of a regional specialist. The documentation and research carried out on particular books shows the variety of techniques and materials used—such pieces of information are the best references for research on Tibetan books and are very useful in dating and finding their places of origin. While much more research needs to be done to achieve higher precision of regional attribution, my codicological results on paper organized into typology also helps in answering some questions about the trade and import of paper and manuscripts in the Himalayas and Central Asia. Next, I considered other sources of information such as textual content, layout of page, and manuscript form. This showed whether all mentioned features originate from the same area, whether that area is best understood as a cultural context, country, or region. Many aspects of the history and crafts of Tibetan books were indicated.

Tools to Read What is Not Written

The information taken from texts is verified (critically evaluated) by study of techniques of book production—consecutive descriptions of objects which contain elements of paleography, codicology, technical art history, and material analyses. A paleographical approach allowed for comparative analyses of page outlines in different types of books with special attention to the text composition, script types, and the style of rubrics and rulings.1 Elements of codicology were used when identifying book formats and bookbinding styles, and also when I examined techniques of book making. Further, I used technical art history

1 In Medieval Manuscript Studies ‘rubric’ refers to lines of text written in a different ink. Most often, but not always, red rubrics served as instructional guides to the reader, providing descriptive headings and marking divisions in the text. The term comes from Latin rubrica, meaning red. ‘Ruling’ refers to the horizontal lines applied to the paper to guide the scribe’s hand. 14 chapter 2 methods to study decorations, iconography, and paints within the scope of my abilities. The majority of my work was concerned with studying book forms and material analyses since results of this research offered a wealth of information about book origins and sometimes helped with dating. In addition, material analyses often provided clues on economic conditions in a particular region and about the methods of artists and craftsmen, such as whether their choice of materials was conscious, demanded by a patron, or less deliberate. Examination mainly entailed the study of the paper, ink, paints, textile, and wood that were the main components of Tibetan books. Some data was extracted by means of analytic methods, such as Optical Microscopy (OM), Scanning Electron Microscopy (SEM), etc. This data included paper components, ink/paints composition, and wood in covers among other materials. Because some research requires microsamples, material analyses are conducted only with the agreement of collection owners. Analyses of the paper included fiber composition, techniques for creating particular types of paper, paper structure and transparency—such as patterns of chain and laid lines, and the study of paper textures (how the paper surface absorbs ink). The identification of fiber composition allows for distinguishing the Tibetan type of paper from others. Identifying raw material used for paper production can also help in discovering paper origin. When comparing the results of fiber analyses of ancient books with the location of occurrence of the same plant, we can obtain information about the possible region of book origin.2 The area suggested by plant occurrence can be critically evaluated by other sources of information, such as textual content, layout of page, and manuscript form. This way, we can find out whether all mentioned features originate from the same area, whether that area is best understood as a cultural context, country, or region. The typology established regarding the differences in technology can also point to the region of a given book’s origin, since we generally know the geographical range in which a particular type of papermaking mold was used, and differences in sheet formation. This criteria alone is not reliable enough, since there are many areas where both types of papermaking molds (sieves) were used, and we definitely cannot take anything for granted based on this information. Nevertheless, technological typology does provide an important hint. Even though the information will still point to quite a wide region, by comparison with the two other criteria we can achieve much higher precision. In this

2 Helman-Ważny 2009: 173–186. Methods 15 way, handmade woven paper, handmade laid paper, and machine-made paper can be distinguished.3 A further level of criteria regarding the preparation of book leaves before writing or printing includes construction of the leaf and visual properties of the surface of the leaf, such as the presence of dyes in the paper (or of insect- repellant substances, which also change the color of raw paper), the size of the paper, how many layers of paper were glued together, and polishing of the surface in case of manuscripts. This level of criteria is particularly important in showing the difference between paper prepared for manuscripts and that used for prints. The inks and paints in written and printed books were manufactured according to diverse recipes at different times. Thus, chemical tests are an important part of the identification of inks. Because of a lack of portable equipment for testing, precise compositions of inks and paints, however, were determined only for a few samples. Adam Cupa of Nicolaus Copernicus University in Torun, Poland conducted pigment analyses for this project. Black ink is often described as Chinese or Indian, but rarely is there an indication whether the particular ink was produced in China or India. Additionally, both Chinese and Indian inks are based on carbon and they are almost undis- tinguishable with the usual scientific apparatus used for testing manuscripts, and certainly not to the naked eye. Most different is their form as a product. Chinese ink is essentially an intimate mixture of finely divided carbon and animal glue, often with a variety of additions for specific purposes. As John Winter from the Freer and Sackler Galleries of Art of the Smithsonian Institution, Washington DC has said, the apparent simplicity is deceptive.4 In fact we are dealing with a countless number of products called ‘Chinese ink’ that vary not only in type and quality of ingredients but also in the thorough- ness of their mixing. After the carbon and glue solution, with or without secondary components, has been combined and thoroughly mixed, portions are molded into sticks or cakes. Most of pigments used in Tibetan books have a natural mineral origin.5 Many are found in Tibet, others are imported. Typical traditional mineral pigments are: cinnabar (HgS) for red; realgar (As2S2) for orange; orpiment (As2S3) and yellow ochre (Fe2O3∙nH2O) for yellow; malachite (CuCO3∙Cu(OH)2) and ver- digris (mCu(CH3COO)2∙nCu(OH)2∙pH2O) for green; azurite (2CuCO3∙Cu(OH)2)

3 Methods of paper examination and record of paper features are discussed in: Meisezahl 1958: 17–28; Helman-Ważny 2006: 3–9; van Staalduinen, et al. 2006: 346–353. 4 Winter 2008: 45–58. 5 Jackson and Jackson 1994; Ernst 2001: 902. 16 chapter 2 for blue; chalk (CaCO3), gypsum (CaSO4∙2H2O) and kaolin (Al2(Si2O5)(OH)4) for white; and carbon black for black. Pigments produced synthetically were also occasionally used. These were for example orange red lead (Pb3O4), yellow massicot (PbO), smalt, a blue cobalt glass, and cinnabar. Lapis lazuli and artifi- cial ultramarine were rarely used. The most important traditional pigment of organic origin was indigo (C16H10O2N2), which was imported from India and Nepal. Others were dyes produced locally, such as red lac, red sandalwood, yellow saffron, and yellow dye produced from petals of a wild rose. Whenever possible I also examined elements in books made of textiles, such as book cloths or curtains attached to the covers or frontispieces. Optical microscopy identified the raw material of the textiles. Tomasz Ważny from the University of Arizona, Tucson, USA identified wood species in selected book covers. The tool traces and methods of splitting and finishing surfaces was also observed and documented (Figure 1).

Figure 1 Wood-working tool traces observed in uv light on the book cover from the collection of the British Library. Methods 17

Survey of Disciplines Used for Dating

Estimating the date and origin of Tibetan books is a difficult challenge, but it is one that may be fundamental for all fields of Tibetan research. One possibility, however rare, is if a date appears in text or we can identify donor, artist, owner or other persons whom are certain about the time when they lived. Studies of the text should always be the first step of the procedure leading to dating a book. Unfortunately, these days, this often tends to be both the first and the last step used for identification and all dating efforts tend to be limited to the content of the text. Thus, sometimes the identification of individuals, tracing lineages, and determining names of teachers may lead to an approximate date. However, this method will not distinguish an original text from a copy or forg- ery. This is especially true in the case of Tibetan books because the textual content and book itself can be from two different periods, depending on the habits used to copy Buddhist texts. Xylographical texts were often reprinted many times from the same wooden blocks over the course of dozens of years. This is why other possible sources of information should be taken into consideration as well. Additionally, there are many books that do not have any chronological information in the text or colophon. In such cases the most reliable results could be provided by technological, and material analyses. In Western museums and libraries, dating Tibetan book collections is usually difficult. There is a lack of ‘unambiguous determinants’ for dating books and their contexts. At present, possibilities for finding ‘fingerprints’ of time and origin in materials are promising, but much more research and data are still needed. We are still in the process of collecting data; hopefully, we will eventually have enough information to use as a reference for identification of other books from a regional and temporal perspective. It is very important to collect as much information and as many observations as possible about each book. The type and function of a particular book should be recognized based on linguistic studies—text identification and colophon translation. The identification of the Tibetan script or, ideally, the hand of the scribe may provide helpful clues in placing the book in its correct time period. However, with the exception of a limited number of Dunhuang manuscripts, our knowledge of Tibetan scripts, calligraphies, and scribal practices is not sufficient. Routine art-historical methods seem to be less effective in the case of dating Tibetan books. Miniatures and ornaments occurring in Tibetan manuscripts have usually been painted according to the same formal rules they have always followed, 18 chapter 2 and they are not signed by an artist.6 With the notable exception of work by David Jackson and Amy Heller, there is also no clear chronology of painting styles for these miniatures and ornaments that could be used as a reference.7 In this case, it is also more efficient to combine this method with other possible. However, in practice, stylistic analysis is often the only method applied by ‘experts’; relying on what they have seen—including some Tibetan books that may not have been correctly dated before—they use their instinct and intu- ition rather than knowledge and evidence. On the other hand, iconography and painting styles can suggest the Buddhist school or lineage, and further provide hints to the region of origin.8 The study of materials comes later; this includes the analyses of paper, pigments, the identification of wood, and dendrochronology of book covers. Finally, radiocarbon method can be useful, but the results may be ambiguous and might not provide additional information; thus, it is essential to check all other possibilities first to avoid high costs. To find new sources of information, technical aspects of bookmaking have been tracked historically. Archaeological findings confirm that Tibetan books exist as both paper and silk scrolls. Chinese-style binding exists, as well as concertina forms and many other methods of Tibetan binding. The differences in binding styles seem to be more regional than temporal.9 They are also connected to the function and usage of particular types of books. This may explain why we are yet to have a chronology of binding for Tibetan books. Still, since the dating of Tibetan books (or lack thereof) is the key problem in developing a chronology, the possibility presenting a chronology of Tibetan book formats depends on having more books dated precisely. In my study of paper, as mentioned above, I focused on the optical characteristics of the material including any features, such as chain and laid lines that are visible beneath the surface, as well as fiber composition, the presence of glue in the paper, and the finish. Fiber compositions in the paper compared with the regional distribution of plants used for papermaking may be helpful in determining the origin of particular books.10 The consistency of pigments and other media analyses can also help, if one traces this historically and compares it with the dates found by new methods discovered in the history of craftsmanship, such as papermaking and printing. Generally these written sources are as important as research on historic objects.11 Concerning the other

6 Gega 1983. 7 Jackson 1996; Heller 2009. 8 Pal and Meech-Pekarik 1988; Jackson 1996; Heller 1999. 9 This subject is approached in chapter 3 of this essay. See also: Helman-Ważny 2007: 4. 10 Helman-Ważny 2006; Helman-Ważny 2007. 11 Hunter 1978; Ilvessalo-Pfäffli 1995. Methods 19 materials, such as pigments, we do have clues as to when they were made. A wide variety of analytical techniques can be employed—preferably more than one—to obtain unambiguous results. The conducted tests showed that traditional mineral and organic pigments were used for miniature painting in the past. In the case of religious texts, the character of the material used was an important element connected with the principles of Buddhist aesthetics. We can also assume that the same kind of material was used, but research on pigments in more contemporary art shows that these rules did not restrict Tibetans to using only traditional pigments. I interviewed a few thangka painters at the Norbulinka Institute, and followed their instructions to try to find mineral pigments; all the thangka painters, however, mentioned places that did not sell mineral pigments. The limitation in dating here is that the number of traditional pigments and dyes identifiable on all kinds of paper-based materials from antiquity until the eighteenth century is relatively small, in total about 25–30 compounds. Many are of mineral origin or synthetic reproductions of minerals, such as cinnabar, vermilion, or azurite. Even if we precisely identify those materials, we cannot identify their origins until we link them to particular deposits in the case of natural minerals, or time of production in the case of synthetic reproductions. Unfortunately, the search for unusual materials that may serve as a fingerprint very rarely yields successful results. Book covers are potential sources of information, but loose-leaf construction of books can pose limitations. To draw clues from the covers we have to be sure, first of all, that they are truly part of the book (Figure 2). Dendrochronology is successfully used for dating works of art and historic wooden objects in general in Europe and America. Tree-ring measurements taken from the wooden object are compared to master chronologies which may provide a date with annual precision. Dendrochronology can be applied only to wood with a readable and continuous series of at least fifty rings. Dendrochronology’s requirements are compensated for by its high precision of dating and the independence of the method. The book covers and xylographic blocks, and eventually wooden slips, can be considered for dendrochronology. The small number of chronologies, however, available for Central Asia remains a limitation.12 Estimation of wood origin is another possibility of dendrochronology. Radiocarbon dating is applicable to manuscripts composed of organic materials using milligram-size samples in the case of the Accelerator Mass

12 Locations of the longest tree-ring chronologies from Tibet and Nepal: Dolpo (1556–1998); Mustang (1465–1992); Lukuchi Khola (856–1996); Dulan (515 BCE–2000 CE); Reting (1081– 1998); Chamdo (449–1994); Densa Thel (1214–1998). For further information see: Schmidt, Ważny, Malla, Höfs, and Khalessi 1999; Sheppard, Tarasov, Graumlich, Heussner, Wagner, Österle, and Thompson 2004: 869–881. 20 chapter 2

Figure 2 Dating of a wooden cover from the collection of the British Library (or 15190). a. The edge of the plank with a visible tree-ring structure. b. Measurement of tree-rings’ widths with Image Analysis. c. Graph visualizing the measurement. © Tomasz Ważny

Spectrometry (AMS) method. It can certainly be used to determine the age of Tibetan books. It is an independent, objective, and the only truly quantitative method for organic materials. However, as many scholars have pointed out, this technique alone cannot usually resolve issues of authenticity and precise dating. However, accuracy, which is in a range of 50–300 years depending on the time period, is a potential problem. Typically there may be an error of +/– 30 years on an uncalibrated date. This is usually calibrated to obtain a date in calendar years with a similar error. However, this error can be wider or narrower depending on the shape of the calibration curve resulting from fluctua- tions of 14C isotope in the atmosphere. From 1650–1955, however, there are many intercepts with the calibration curve and, as a result, several possible calendar dates exist for a single sample. The reason is that the so-called plateau of the calibration curve caused by increasing consumption of fossil fuel that did not contain 14C. From 1955 to the present, the resolution can be to the near- est year owing to elevated 14C levels from the atmospheric testing of atomic weapons until the Partial Test Ban Treaty in 1963. If the book was made in 1640, the date would not be quite so ambiguous. Methods 21

Additionally, this will tell the date that the cellulose molecules in the paper were formed rather than when the paper was made. The papermaking process will sometimes have blended fibers from a wide variety of original sources. Further, other substances such as dyes or fillers may have been added when processing paper before writing, and can contaminate samples by stable 12C isotope (‘dead carbon’). This makes results of 14C dating disputable in the case of singular samples.13 From the other side large scale of sampling from old manuscripts is also questionable and not recommended. This is why a combination of historical and scientific dating methods is the only way to obtain reliable evidence, and the radiocarbon dating method is best reserved as a control. Furthermore, the greatest limitation of radiocarbon dating is its cost and size of the sample required. The radiocarbon method is relatively expensive and it is important to have preliminary results indicating the range of time anticipated. Sample sizes depend on the type of material to be pretreated, the amount of material available, and, where apparent, the degree of preservation of the material. The required sample size is also a limiting factor. In contrast, AMS (Accelerator Mass Spectrometry) technique requires only 1–2 mg of paper, but the price is higher still.14

Study of Paper in Books

Knowing how paper is made and which factors determine its properties opens new prospects for the study of Tibetan books far beyond their textual content. The identification and classification of Tibetan ancient papers becomes much easier thanks to knowledge about papermaking tools and terminology. The common view that the study of paper is not useful as a source of information for identifying a book is encapsulated by Jonathan A. Silk’s assertion: “The quality and size of the paper, and even the color of the ink used, tell us nothing; it is quite easy to change these while printing from the same set of blocks, thus producing exactly the same contents.”15 This claim has the appeal of a simple truth but is contradicted by systematic research on paper. One of the reasons why paper has been underestimated as a source of information is lack of interdisciplinary work; Tibetan studies is a difficult field for paper historians and other researchers without a background in Tibetan culture and language.

13 On radiocarbon dating of manuscripts, see Pollard 2011. 14 Brock et al. 2010. 15 Silk 1996: 153–200. 22 chapter 2

Another problem for the study of paper is posed by the lack of clear terminology to describe the paper in works of art. There are further obstacles. One of them is the common, misleading description of Tibetan paper as ‘rice paper.’ Tibetans never used actual rice paper. First, it should be noted that ‘rice paper’ has nothing to do with rice. This type of paper is made of the inner part (pith) of the tree Tetrapanax papyriferus, which grows in Taiwan and parts of China; the paper is also known under the name ‘pith paper.’16 Hunter mentions the Chinese name of this paper—kung shu—as well as the earlier name under which the plant was known—Fatsia papyrifera.17 Rice paper has a distinctly white color and spongy wafer-like structure. There is a hypothesis that rice paper could be the pre-paper material.18 There is, however, no evidence for this idea. When rice paper was first produced is not known; significantly, rice paper is first mentioned in 1634 as a material produced in the south of China.19 Because its structure and small format is limited by the pith size of the tree, rice paper was primarily used for water painting and letters. Today, its production has almost entirely ceased; it is preserved in Taiwan on a small scale. Rice paper is different from all the papers produced in the Himalayan region in every respect. Also, Tibetans never made a paper from rice straw. If they had, it would excuse the use of such a term. Rice-straw paper was only made and used in China and was one of the lesser-quality papers (rice straw is not a durable material for paper). Without precise terminology, the relevance of the study of paper lacks con- viction. Let us, therefore, turn to the issue of terminology. I use a three-level set of criteria for the typology of paper in Tibetan manuscripts: (1) raw material used, (2) technology of papermaking (type of papermaking mold), and (3) preparation of leaves before writing or printing. For the first criterion, I adapted and simplified the already existing differen- tiation for Tibetan, Chinese, and Russian types of paper. In Tibet, the Thymelaeaceae family of plants were the primary materials for papermaking.20 These are very distinctive from the materials in Chinese types of paper, which are composed of plants, such as ramie, paper mulberry, hemp, bamboo, straw,

16 Bell 1983: 105–119. 17 Hunter 1943: 20–21. 18 Hunter 1943: 20. 19 Bell 1983: 105. 20 Hunter 1943; Meisezahl 1958: 17–28; Trier 1972; Koretsky 1986; Helman-Ważny 2005: 27–37; Helman-Ważny 2006: 3–9. Methods 23 and many others;21 they are also distinctive from materials in Russian types of paper that are composed of flax, hemp, and wood pulp. The second criterion allows us to create a typology by differences in technology. The categories are: handmade woven paper, handmade laid paper,22 and machine-made paper. The area range of particular types of papermaking molds and differences in papermaking technology (sheet formation) are generally known; this information can be very helpful in establishing the region of book origin. The area delineated on the basis of this criterion alone would be rather broad, yet put together with the other two criteria, this information allows us to make a much more precise estimate. The third criterion, the method used to prepare the paper leaves, includes construction of the leaf and visual properties of its surface, such as details of paper dyeing or introducing insect-repellent substances, which also change the color of raw paper, paper sizing, gluing of paper layers (and number of layers), and surface polish. This criterion illuminates the difference between paper prepared for manuscripts and that prepared for prints.

Raw Material Raw material is like a fingerprint preserved in ancient material. However, today, this information is only partly readable. The plant species that are the most difficult to distinguish between in paper analysis are those of the Thymelaeaceae family; this family of plant species are most typically found in Tibetan paper. The methods of fiber analysis allow for clear-cut identification only in a limited number of cases. Therefore, I refer to these fibers collectively as Thymelaeaceae-family plant fibers. Although this cautious phrasing may seem vague, it refers to a small number of plant species: only a few species of this family are used for making paper, and only one, Stellera chamaejasme, has been established as a paper component used exclusively in Tibetan papermaking. Since the 1970s, there have been efforts to draw a clear distinction between the Daphne and Edgeworthia species in ancient papers but without much success. The research results were usually left at a preliminary level. After trying to find a method that would enable me to make the distinction between these fibers, I turned to questioning the distinction itself. It is important to note that Tibetans, the Nepalese, and the Bhutanese do not often differentiate between papers made of either of these two fibers; also, they use the same common name for the few species available in the Himalayan region that produce similar quality paper. That is not to say, however, that the distinction is of no use,

21 Hunter 1932; Hunter 1943; Tsien 1973: 510–519; Tsien 1985; McClure 1986. 22 van Staalduinen, et al. 2006: 346–353. 24 chapter 2 only that more refined methods, such as DNA analysis, are necessary for its delineation. At this stage of research on paper fibers, I would rather emphasize the differences between various types of Tibetan and Chinese papers developed in particular regions as well as distinctions from issues of paper imported from further regions (e.g., Russia). With this in mind, I turned away from the question of differences between Daphne, Wikstroemia, and Edgeworthia fibers to regional characteristics of papers. As was increasingly clear in the examinations of Tibetan books, many of the raw materials in the Tibetan books that I analyzed were not native to Tibet and had never grown on the Tibetan plateau. Thus, components of papers greatly help in recognition of their areas of origin, especially when cross-referenced with plant occurrence. Another significant feature of Tibetan papers is that they are often composed of just one or a maximum two types of fiber (a mix of Daphne and Edgeworthia fibers, which I classify as one type). This feature sets Tibetan papers apart from Chinese rag, later mixed-fiber papers, and Russian papers composed of hemp/flax and, later, wood pulp. Most of the papers from the selection of papers I studied were composed of fibers from the Thymelaeaceae family (Daphne sp. and Edgeworthia sp.). I found these fibers also in documents and books from the Dunhuang caves, of which the oldest were dated to the ninth century. This finding allows us to establish that Tibetans had already developed their own paper production at that time. Both species from the Thymelaeaceae family were the basic material in papers from the foot of the Himalayas.23 These species were very difficult to tell apart in samples of historic papers. However, after the fibers of both species were stained with Herzberg olive-grey, they were distinguishable from other bast fibers by the broad central portions of the fibers characteristic of the Thymelaeaceae family (Figures 3a, 3b, 3c, 3d).24 The Daphne fibers seemed more rigid in their general shape and more irregular in thickness than the Edgeworthia fibers. The broad portions were not associated with cross-markings, which occurred densely and were often visible as a ‘v’ or zigzag shape in polarized light (Figures 3a and 3b). Dislocations and irregularities in the fiber walls and lumens were clearly visible (Figures 3b and 3d). Fiber dimensions of both species are also very similar; however, Edgeworthia sp. fibers are narrower (apart from the wider central portions), and the shape of the fibers is more gently curved than the Daphne sp., whose fibers are more gnarled in shape.

23 Characteristics of the plants used for papermaking in Nepal and other references: Trier 1972: 50–59. 24 Ilvessalo-Pfäffli 1995: 351. Methods 25

Figure 3a, b, c, d Fibers of Daphne or Edgeworthia sp. observed under microscope. a. Fibers stained with Herzberg in magnification 150× from paper of Tibetan manuscript from The Asia and Pacific Museum in Warsaw, Poland (MAP 14144). Broad portions visible. b. Fibers broad fragment in magnification 600× from paper of a Tibetan manuscript from The Asia and Pacific Museum in Warsaw, Poland (MAP 14144). c. Fibers stained with Herzberg in magnification 300× from paper of a Tibetan manuscript from the Stein Collection, The British Library (M58, vol. 1, f.58). d. Fibers broad fragment in magnification 600× observed in polarized light from paper of a Tibetan manuscript from the Stein Collection, The British Library (M58, vol. 1, f.58).

These features mark the difference in fiber placing. The bast of Edgeworthia is usually whiter and softer than that of Daphne, but paper made of Daphne was generally considered the better paper in Tibet. Paper made of Daphne is known by Tibetans as dug shog ‘poisonous paper’ because the bark of this tree contains a substance that deters insects,25 making it naturally resistant to insect infestation and deterioration; it is even thought to have medicinal properties.26 The addition of yellow sulphate of arsenic is not the only the reason why the paper is called dug shog—the plant itself has poisonous properties.

25 Imaeda 1989: 410. 26 Upreti 2004: 236–243. 26 chapter 2

Stellera chamaejasme is another representative of the Thymelaeaceae family—one that is more readily distinguishable from the Daphne and Edgeworthia plants of the same family used for making paper. It took me a long time to detect this type of fiber in Tibetan books. I was only able to recognize Stellera fibers after preparing reference samples from plant tissues. Elaine Koretsky kindly supported me in this effort and supplied me with the root bark of Stellera that she had collected in Tibet. I performed an analysis of plant tissue to determine where the fibers were located in the plant and what other types of cells could be found in the pulp. The cross- and longitudinal sections of plant tissues helped to recognize the shape of particular cells. Stellera fibers are distinguished by their ribbon-like shape and an impression of transparency, similar to flabby cotton fibers. There are broad sections within the length of Stellera fiber cells that are very similar to those in Daphne and Edgeworthia sp. The staining color with Herzberg stain is also olive-grey to blue, the same as others from this family. However, the fiber walls and lumen are much more irregular when compared to Daphne and Edgeworthia. The lumen is often delineated by wavy fiber walls, which creates very irregular space inside the fiber cell (Figure 4a, b). Ramie Boehmeria sp. is a typical component of Chinese paper. I found it in early Dunhuang books and documents, as ramie textiles were often used for making paper since at least the third century. It is also present in many later Chinese papers—though quite at random and together with other components. Herzberg stained the fibers reddish brown. Their most characteristic feature is a ribbon-like shape placement associated with having a wide lumen and the flatness of the fibers, which had a tendency to twist or tangle together. A cross-section of fibers shows a closer resemblance to a flat ribbon cross- section than to a circular one. The lumen and fiber walls have many variations in size. The distinguishing feature of ramie is the great thickness of fibers and the clearly visible longitudinal stratifications of the wall, as well as its many cross-markings (Figure 5a, b). The fiber ends are mostly rounded, but they are rarely found. Many fibers are damaged in the pulp and broken into longitudinal stratifications; this situation can make it difficult to distinguish between separate fibers and bundles of fibers (Figure 5a). Ramie is indigenous to China and adjacent countries, such as Korea and Japan, and has been used since ancient times as a textile. This is one of the oldest textile plants because its use would have started more than six thousand years ago (it was used in the manufacture of strips for the preparation of mummies in Egypt). The Boehmerias form one of the genera of the Urticaceae family; the genus includes one hundred species. There are about sixty-five tropical or subtropical Methods 27

Figure 4a, b Stellera chamaejasme fibers observed under microscope in polarized light. a. Fibers observed in magnification 600× from the root bark of Stellera collected in Tibet by Elaine Koretsky. Irregularities in both fibers walls/lumen and ends are visible. b. Fibers observed in magnification 600× from the above mentioned sample. Exceptionally wide lumen in between irregular fiber walls characterizes Stellera fibers. species of which six are endemic in China. These are flowering plants, but unlike other members of the same family, this genus bears no stinging hairs. Boehmeria nivea, or white ramie, is a large perennial herb that spreads by frag- mentation of rhizomes. With stems that range in height from 2–3 m, this plant likes sandy and well drained soils. It is cultivated for its resistant bast fibers. It is possible to obtain two to four harvests per year depending on the climate of the region. Because the stems of ramie plants contain a high proportion of gums and pectins, special methods are needed for the separation of the fibers. The fibers are difficult to dye due to their high molecular crystallinity. Their shiny appearance qualifies them as a silk plant. Ramie fibers reached European markets in the form of fiber ribbons stripped from the stem, known as China grass.27 Ramie fibers consist of pure cellulose, and are remarkable because of their length and width. They are the strongest and most durable of vegetable fibers. Ramie fibers are shorter than those of hemp, but they are very strong making them suitable for the manufacture of ropes and nets. The longest fibers are reserved for the textile industry while the shortest are for paper manufacturing. Paper mulberry (Broussonetia sp.) is another typical component of Chinese paper. It is used alone or with other components. I found paper mulberry in many samples from Dunhuang documents (e.g., IOL Tib J 310.1208; vol. 69 f. 9 and 16), in Yongle and Wanli Kanjur volumes (fifteenth and seventeenth centuries), and also in an early gold manuscript from Western Tibet (Guge, fifteenth

27 Ilvessalo-Pfäffli 1995: 346–347. 28 chapter 2

Figure 5a, b Ramie fibers stained with Herzberg from Tibetan manuscripts from Dunhuang (iol Tib J 1254 and IOL Tib J 1357) observed under a microscope in polarized light at 600× magnification. a. Both cross-markings and longitudinal stratification typical for ramie are clearly visible. b. Damaged fibers broken into longitudinal stratifications; this situation can make it difficult to distinguish between separate fibers and bundles of fibers. century). Together with ramie, it was consider one of the most useful fibers due to its strength and length, which create the best types of paper. With Herzberg the fibers of paper mulberry stained reddish brown and a pale purple color. Blunted fiber ends were found; the lumen was marked but interrupted at intervals. Cross-markings were clear but irregular. The transparent membrane identified as a primary wall, stained by Herzberg violet to bluish purple, is clearly visible on all pictures (Figure 6a, b). This transparent membrane usually envelops the fibers, which is visible on the SEM photo (Figure 7). In historical papers, this membrane was torn in some cases into small pieces or separated from fibers as a ribbon. In some slides the membrane was found in only a small quantity, and in such cases the addition of hemp fibers should be considered during identification. Hemp fibers can be very close to paper mulberry, although the primary wall is not visible in this case. The condition of the membrane depends on processing of pulp during papermaking and the quality of the raw material. The membrane, or, primary wall, associated with paper mulberry fibers is usually considered to be their most characteristic feature. While this element exists in other species, it is not as clearly visible as in paper mulberry; the diameter of the membrane of paper mulberry is visibly larger than the diameter of its secondary-wall fiber. Fiber endings were possible to find during observation. Broussonetias are a genus of the Moraceae family. The main species in this family are papyrifera, kazinoki, kaempferi, and kurzii. Broussonetia papyrifera (and Broussonetia kazinoki) is a deciduous shrub which grows in the wild but can be cultivated; it can reach 15m in height. It grows as well on flat ground, allowing an easy harvest, as on hillside, where land is less desirable for other Methods 29

Figure 6a, b Fibers of paper mulberry (Broussonetia sp.) stained with Herzberg stain found in a Tibetan scroll (IOL Tib J 1560, vol. 85 bi-f. 2) from China’s Dunhuang caves under a microscope in polarized light. a. Fibers observed at 300× magnification. b. Fibers observed at 600× magnification. Transparent membrane envelops the fibers. cultures. It accommodates both dry lands as wetlands. Its cultivation is quite easy but because Broussonetia kazinoki is simpler, the two species are often planted in mixed fields. Cultivated shrubs provide better fiber than the wild ones; they are thick, long and strong. Its growth is slow, however, harvesting can take place each year. Shrubs provide bark used forever in East Asia, China, Korea, Japan and the Pacific area. They are mainly used in paper manufacturing, but also in traditional medicine as are the leaves and fruits of the shrub. In the past and until now, paper mulberry is native to or has been cultivated throughout China. Paper mulberry was harvested, grown, or sold for the manufacture of cloth very early, under the Shang (ca. seventeenth–eleventh century BCE). The earliest mention of the use of bark in the manufacture of paper can be found in the biography of Cai Lun, in the Hou Han Shu (History of the Later Han) dating from the fifth century CE; the use seems to have started with Cai Lun in the second century. The use of Broussonetia barks in the manufacture of paper, originated probably in the ancient production of bark cloth (tapa), especially in the South and Southwest. Papers made with paper mulberry are already popular under the Jin (265–420); under the Tang (618–907), these papers are quite common for writing. Papers for calligraphy and painting in the Ming Dynasty (1368–1644) are often made from a mixture of bark and bamboo or sometimes straw. At all times, they are also used for making money paper, clothes, for bookbinding, in interiors of buildings, etc. At the end of the Tang Dynasty lightweight armor was made of paper that was strong enough to stop arrows; this type of armor continued to be used even during the Qing (1644–1911) by populations of the south in what today is Yunnan, Guizhou, and Guangxi. 30 chapter 2

Figure 7 Paper mulberry fiber and its separated primary wall observed by SEM in 600× magnification.

I found hemp (Cannabis sativa) only in the Tibetan manuscripts from Dunhuang; in most cases, hemp was one of the components of rag paper. The genus Cannabis consists of only one species, C. sativa, grown in temperate climates. The common name ‘hemp’ is often used for several other plants, such as Crotalaria sp. (sunn hemp) and Musa sp. Textile production has employed hemp since prehistoric time in China, Japan, and Persia. Thus, the hemp I found in papers of Tibetan manuscripts from Dunhuang was made from old rags; rags were used in China in the early period of papermaking. Hemp fibers are long, thick-walled, and wider than those of flax (Figure 8). The fiber ends of hemp are usually blunted and its lumen is wide and flat. Fibers show dislocations, cross-markings, longitudinal striations, and swellings. The presence of associated cells depends on the raw material. Pulp produced from rags does not usually contain cells from other tissues. Pulp of a lesser quality (unclean) may contain short pitted fibers, pitted vessel elements, spiral and netlike vessels, spiral thickening, parenchyma, and even epidermis with hairs. Hemp cultivation is probably one of the oldest and most durable in China. Hemp was grown all over China but particularly in the north. It probably Methods 31

Figure 8 Thick-walled hemp fiber from a raw plant colored with Herzberg stain and observed under a microscope in polarized light at 200× magnification. evolved in northern China and was the first fiber plant to be cultivated. The oldest Chinese pharmacopeia book, Bencao jing ca. 100 BCE, mentions that Taishan (Mount Tai in Shandong province) was one of the oldest locations where hemp was grown in historical times. Hemp is the oldest plant to be used in making clothes before the use of cotton fibers were introduced in textiles during the Ming Dynasty (1368–1644). Hemp, flax, ramie, jute are called ‘ma’ in old Chinese documents. The use of hemp as fiber in papermaking is the oldest and dates back to the Western Han (206 BCE–8 CE). The first papers are made from either fiber cloths or raw bast fibers. The oldest paper fragments from the Western Han period, discovered in Minfeng (Xinjiang province) date from the beginning of the third century BC, the fragments found in Baqiao (Shaanxi province) date 140–87 BC and the fragments found in 1934 in Lobnor, near Loulan (Xinjiang province) date from the first century BC, were made with hemp fibers. Other common papermaking materials were used such as paper mulberry and ramie in the Han and later as rattan introduced in the third century, but hemp remained the primary material for paper manufacture. During the Tang period (618–907), hemp paper is considered as the best for calligraphy, bookbinding and official documents (including at Imperial Court’s) 32 chapter 2 because it is light, strong and waterproof. The use of hemp in papermaking declined after the Tang and was replaced progressively by ramie and bamboo. I found bamboo fibers typically in Chinese papers used for Tibetan books, for example in the supplement for Wanli Kanjur from the seventeenth century or in The Book of the Dead from the British Library (OR 15190). Typically, the fibers are long, fairly narrow, thick-walled, and have many pointed ends. There are also wide, thin-walled pitted fibers, which are fairly long, and have tapering ends. The pitted vessel elements are not very long but may be very wide (maximum dimensions being about 1 mm × 250 μm). The piths are in longitudinal rows, similar to those of cane. Parenchyma cells are small, barrel-shaped, or rectangular, and mostly thick-walled. Among the bamboo genera used for papermaking are Bambusa, Dendrocalamus, Melocanna, and Phyllostachys, but precise identification of each species was not possible. Cotton was found only as addition in contemporary papers, mostly machine- made. It consists of pure cellulose. Cotton is distinguished from the common bast fibers by the ribbon-like, twisted fibers, which are smooth-looking. Fiber walls show longitudinal and spiral striations. Each fiber has a somewhat tapering tip and a broken base where the fiber was parted from the seed. Cotton fibers are unicellular seed hairs of several species of the genus Gossypium. The cotton plant is a perennial plant that grows to a height of 0.5–1.5 m. It is, however, often cultivated as an annual for cotton-producing industries, or in regions where annual freezes appear or where the plant is intentionally killed and plowed under. Cotton plants become tall shrubs if allowed to grow freely. It is grown in subtropical and tropical zones throughout the world. Cotton is believed to originate in India, where it was probably grown as far back as 3000 BCE. The oldest existing piece of cotton is supposed to have originated in the Indus valley, in today’s Pakistan. It is a fragment of cloth stuck onto a silver vase, which was dated between 2350 and 1850 BCE. Cotton fibers fall into two groups: longer fibers known as lint and shorter called linters. Cotton lint is used in textiles. Paper was made from old rags, new textile cuttings, and later linters. The fiber properties of paper have strongly shaped the aesthetics of a given culture based on the raw material used. At the same time, artists have been choosing materials to match their creative impetus. It is thus challenging to draw conclusions about the relationship between material and aesthetics because so many of these observations are subjective. Beyond the raw materials, there are other aspects of the papermaking process which influence the nature of the resultant paper, such as the degree of fiber blending or type of papermaking mold applied, and the preparation of the paper surface before a drawing, calligraphy, or a print is set on it—questions that I will address in coming chapters. Methods 33

Scientific examination of Asian paper specimens28 revealed old methods, materials and techniques, as well as differences in aesthetics directly attributable to the papers themselves. In this way papers from the past are a trove of information, and their fiber components are one of the most important criteria for paper classification. Moreover, analysis of the fiber in the papers compared with the geographic distribution of the source plants, could suggest the origin of the paper.

Papermaking Sieve Printing In general it was possible to single out three kinds of papers: woven paper hand-made on the textile type of sieve, laid paper hand-made on bamboo, reed, grass or wired type of sieve, and machine-made paper.

Woven Paper Handmade paper produced with floating molds is the most traditional and the most ancient form of production as far as major Central Asian techniques of papermaking are concerned (Figure 9a, b). This type of paper made of Thymelaeaceae-family plants was mainly used for books, which consisted of loose leaves usually glued together in a few layers.29 The color of the papers examined usually ranged from yellowish to dark brown, though no organic substance that could be responsible for the color was detected. Initially, newly made paper of this type has a creamy color and contains small pieces of solid, dark brown, pulp impurities. This type of paper was identified as having been used in the most decorated handwritten books and xylographs. The leaves of the examined books usually comprised three to five layers of thin-woven paper with a rough surface adhered together to make thicker sheets.30 Microscopic study clearly indicated that the paper under study contained rather long fibers with many irregularities in the shapes of fiber walls and lumens.31 Leaves of this type of paper had a highly sized structure, as they were prepared for writing; this contrasts with the paper manufactured in the Himalayas, which was not sized. The paper was usually in good condition due to its physical and chemical stability and strength. The pH of most samples was in the range of

28 Collings and Milner 1978: 51–79; Ilvessalo-Pfäffli 1995: 348–349; Meisezahl 1958: 17–28. 29 Bhutanese papers are produced from the same plant as well. 30 The first extant manuscripts in Central Asia on the glued base of several layers are dated from the 5th century AD. See: Vorobieva-Desiatovskaia 1988. 31 The fiber identification was carried out by the author in keeping with the following standards: PN-72/P-04604; PN-76/P-50125. Comparative material: Collings & Milner 1978: 51–79; Ilvessalo-Pfäffli 1995: 348–349; Meisezahl 1958: 17–28. 34 chapter 2

Figure 9a, b Woven paper made with floating type of papermaking mold. a. Papermaking sieve print observed against light in a historical document from the Library of Tibetan Works and Archives in Dharamsala, India. Traces of pulp stir direction during paper production process are visible. b. Papermaking sieve print of loosely woven textile. Methods 35

6–7, which is close to the optimum of 7.32 The binding media of the analyzed objects of this kind of paper showed that starch paste or animal glue was used to stick the paper sheets together.

Laid Paper Paper with chain and laid lines was found in a large number of books. These features suggest that it was made using a dipping mold with a movable, bamboo sieve. In Tibet, the woven type of floating mold was used since the beginning of papermaking. However, as Elaine Koretsky observed during her extensive papermaking fieldwork in Asia, in addition to the dipping mold, the floating mold with a movable bamboo sieve was also used in some regions of Tibet.33 The sieve print of the floating mold on the paper (the water marking on the paper’s surface) is impossible to distinguish from the print left by the sieve of the dipping mold. Thus, all we can do is characterize the sieve types. This could be very close to Bhutanese technology with a floating or dipping mold equipped also with a movable bamboo sieve characterized usually by 12 laid lines in 3 cm (Figure 10).34 It is likely that papermaking with the floating mold and the pulp pouring method had been introduced into Bhutan from Tibet and that the dipping mold, on the other hand, might have come from China. Papers of this type are characterized by distances between chain and laid lines. Additionally, these papers are usually composed of Thymelaeaceae- family fibers, in contrast to Chinese papers and others with the print of laid and chain lines from the bamboo, reed, grass or wire papermaking sieve. Yet this technique could also suggest that the paper comes from China, where both molds were invented and used.35 As a rule, the leaves of large size made of this type of paper marked with tiny laid lines have a minimum of four layers, since the paper is very thin. This laid type of paper was traditionally composed of various kinds of fibers, which also allows for a further classification into papers of Bhutanese or Chinese origin. The recently dated papers with chain and laid lines characteristic were glued of one or at least two layers of paper and were made of many different components, such as straw, bamboo, softwood, hardwood and wood pulp. Such a fiber composition definitely dated those papers for not earlier than end of the nineteenth century.

32 The pH surface measurements were taken with a pH-METR type 517 according to standard number PN-84/P-50109. 33 Koretsky 1986: 2–6. 34 Imaeda 1989: 409–414. 35 Dąbrowski and Siniarska-Czaplicka 1991: 33. 36 chapter 2

Examples of papers of Chinese origin can be found in the Tibetan Buddhist canon (tsal pa line of transition), such as paper used for the Yongle folio from Michigan, Wanli volumes in the Jagiellonian Library in Cracow and Wanli Supplement volumes in the Harvard-Yenching Library. The Yongle folio is written on very thin Chinese-type paper with small visible laid lines (27–30 in 3 cm) composed of paper mulberry fibers, and consist of six or more layers glued together, each layer is made with a dipping technique by using a papermaking mold with a movable type of sieve. Chain lines were hardly visible due to the gluing together of many layers of relatively soft and absorbent paper. The paper used for Wanli volumes from Pander Pantheon from Berlin/ Cracow is a very thin Chinese-type paper with visible, tiny laid lines, made of six or more layers glued together and composed of paper mulberry fibers. The sheets were prepared in the same way as in the previous cases, with a dipping technique and by using a mold with a movable sieve. Chain lines were hardly visible due to gluing of many layers of paper together, but still possible to discern. However, there are some differences between paper types in particular volumes. Most of the volumes contain laid paper characterized by twenty-four laid lines in 3cm. Only the volumes from the Prajnaparamita section (vol. 29–37) are written on paper characterized by 15–18 laid lines in 3cm. This study indicates clearly that the mold used for the paper in Wanli Kanjur incorporated a movable bamboo sieve that left the laid- and chain-lines fingerprint pattern. That type of mold was not used in Tibet, which confirms the Chinese origin of the paper. The Chinese origin of this paper is further confirmed by its fiber composition: the paper mulberry fibers that compose most of the volumes together with bamboo and straw are typical Chinese paper components. I observed three grades of paper: the best (i.e., longest) fibers were used in volume 60 (dkar chag), and the worst were found in volumes 29–37, which contain Prajnaparamita texts. Tantra volumes were on medium-quality paper. The exception is volume 58, where the structure is hardly visible, but I was able to identify the paper components as three types of cells in its structure: paper mulberry fiber cells (short cut), pitted wood/bamboo tracheids or vessels stained blue-grey by Herzberg, and narrow straw fibers with pointed ends stained olive-grey-yellow. Harvard-Yenching Library volumes of Wanli Sup plement are written on paper made of paper mulberry, straw, and bamboo.

Machine-made Paper Machine-made paper makes up a large number of the manuscripts and xylograph books. Despite using a contemporary material, the form of the books is traditional—they are in pothi format. Blind embossed factory stamps from the Sumkin factory in Afamovo, Russia: Фабрика Наследников Сумкина (the Methods 37

Figure 10 Laid paper made with movable type of papermaking mold observed in a historical document from the Library of Tibetan Works and Archives in Dharamsala.

factory of Sumkin’s heirs) were found on some of the paper (Figure 11). The same type of stamp with the name ‘Platunov’ was discovered on a few other samples (Figure 12). Both paper mills were located in the Vjatka province, in the north of the European part of Russia.36 The Sumkin factory was founded in 1829 in the village of Afamovo in the Lalsk city district by the merchant Stepan Sumkin. Machine production was begun by Alexiey Sumkin in 1854. The use of the embossing stamps ceased in 1917. The only known successors, the grandson and granddaughter of Alexiey Sumkin, Innokentiy Shestakov and Ekaterina Shestakova, inherited the paper mill in 1867 and owned it until the Revolution in 1917, when all private property was nationalized. It follows that the papers embossed by the Sumkin heirs stamp can be dated from 1867 to 1917. Many of such examples can be found in R.R.E. private collection in Winterthur, Switzerland, Asia and Pacific Museum Collection in Warsaw, Poland, and Russian Collections. Identifying and dating the Platunov paper mill stamp was not as straightforward (Figure 12). The only known information is that the mill had been founded

36 This papermaking mill was identified with the generous assistance of Dr. Elena Mikolaychuk, a paper history specialist from the Hermitage, St. Petersburg, Russia. 38 chapter 2

Figure 11 Embossed stamp with inscription: Фабрика Наследников Сумкина ( factory of Sumkin’s heirs) found on the machine-made paper of the Tibetan manuscript. The Asia and Pacific Museum in Warsaw, Poland (MAP 10022).

in 1812 on the Kordjaga River in the same province.37 No date has been found when papermaking machines were introduced into the mill. It, however, is assumed that it was some time in the second half of the nineteenth century, and production continued until the 1917 nationalization. Historic papers of this type that were analyzed consist of flax/hemp, straw, and softwood fibers with an addition of wood pulp. The condition of these objects was different from those bearing the Sumkin heirs stamp: in many cases the paper structure was relatively well preserved. The very fact that machine paper had been used points to a more recent date of production than indicated in the description. The largest group of Tibetan books dates thus from the second half of the nineteenth century.

Leaf Preparation for Writing and Printing The procedure for preparing the sheet of paper is different for manuscripts and for prints. However, this difference is not entirely discernible in Tibetan papers in ancient books, since those papers were processed in many different ways. The color, texture, thickness, and condition of paper are the features notice-

37 Uchastkina 1972: 111. Methods 39

Figure 12 Embossed stamp with inscription: Фабрики Платунова ( factories of Platunov) found on the machine-made paper of the Tibetan manuscript. The Asia and Pacific Museum in Warsaw, Poland (MAP 10022).

able on first glance when we examine papers in ancient books. The following sections consider the relationship between these features and methods of leaf preparation.

Sizing At the last stage of production, the paper was still not sized. In Europe, paper that had not been sized was considered unsuitable for writing due to its high absorbency. Gelatine was used as a sizing agent to allow for writing with a sharply pointed tool. In East Asian countries, a plant extract was usually added to fibers mixed with water (called paper pulp); this additive protected the paper against fiber tangling. The fibers would become more water-resistant and the paper would thus become less absorptive. However, in traditional papermaking in most of the workshops in Central Asia, no such substance had to be added thanks to the properties of Daphne fibers (Figure 13a, b). Unsized paper devoid of other forms of processing could even be used for printing, though it was much too absorbent for writing. Thus, Tibetan paper meant for book production underwent further processing. One of the Tibetan formulas for sizing glue requires sieving wheat flour and stirring it with water until the mixture attains the consistency of buttermilk; 40 chapter 2 the foam on the surface would be removed. The slightly clarified liquid was divided between two vessels. One part of the liquid was boiled on an open fire. Both mixtures were mixed and applied onto paper in the hot sun but in a slightly damp place protected against air and dust. The paper sheers were subsequently left to dry. Old wheat flour that was kept for more than twenty days lost its firmness and was no longer suitable for this purpose. It was also important not to leave the mixture standing too long before applying it; if left too long, the liquid quickly lost its thickness—a sign that biodegradation had begun. There was also another method, which involved preparing a clear liquid out of wheat flour, a quart of buttermilk (without oil) and a bit of liquid glue. The concoction had to be lukewarm. If it was too hot, the mixture could damage the paper; if too cold, it would not stick to the paper and the resulting product would lose its consistency. Maintaining a balanced temperature was crucial. Each ingredient plays a role in this process: the buttermilk makes the paper white and clean, the liquid flour makes it hard and flexible. The liquid glue not only prepares the paper for receiving ink so that the ink does not diffuse or blot, it also makes the paper hard, shiny, and bright. It was also important not to exaggerate the additives—if too much dairy-product liquid was added, the paper would become heavy; too much liquid flour would render it impossible to draw lines on the paper; liquid glue that was too thick would make the paper rough and resistant to ink absorption. In such cases the proportions of the ingredients had to be changed. The best way to adjust the proportions was to make a preliminary test of each combination of ingredients on a few pages of paper. At the preliminary stage, it was also very important to remove any stains from the paper to avoid contaminating the surface of the paper.

Flattening Dried paper might not be evenly flat, which makes writing or printing on it difficult. Dusted, cleaned paper should be put in a pile, in which every ten dried pages are separated by one wet sheet. The pile should be pressed for not longer than one day and not shorter than forty minutes. The forty minute minimum is determined by the time needed to make and drink a cup of Tibetan tea. The moisture must have time to spread equally to all paper sheets in the pile.

Polishing Finally, the process of polishing is applied to make the paper surface even and slightly glossy, which is the surface most conducive for writing. The methods of polishing are the same for paper as for fabric prepared for painting thangkas. To make paper smooth, it should be placed on a hard, even stone without Methods 41

Figure 13a, b Cross-section of Daphne bast tissue. a. Irregular fiber walls and lumens visible on cross-section under a microscope observed at 1200× magnification. b. Starch in Daphne identified in fluorescence observed at 1800× magnification.

cracks. In traditional Tibetan terms, the stone should be smooth like liver. The paper should then be beaten with a piece of wood devoid of any oily substances or with a smooth, wooden yardstick. This must be done carefully to avoid damage to the paper. For the best result, the paper should not be entirely dry during the process. The best tool for smoothening paper is a conch shell, which makes the paper soft and smooth but leaves a bluish tinge. A smoothening stone makes the material more even than a shell, but with a stone there is a greater risk of dam- aging the paper; this especially the case with short-fibered paper. If agate (mchong) is used to smooth the surface of the paper, it leaves a different mark on its surface than the two tools mentioned above. But, if a bronze or iron tool is used, it may easily stain or tarnish the paper. The best smoothening stone is yellow dzi (gzhi). Highly prized in Tibet for its spiritual qualities, it is featured in amulets and jewelry and sometimes ground into medicines; it features by patterns of small circles with dots, so-called ‘eyes’—the more ‘eyes’ a stone has, the more valuable it is. The maximum number of ‘eyes’ on one stone is twelve. When a yellow dzi is used for the smoothing process, the paper gains a desired color, a smooth surface, and softness. However, finding such a stone is rather difficult. Cow or yak horns and porcelain make good replacements for yellow dzi, as do smooth and hard precious stones.

Paper Color The color of raw paper is not very significant for the identification of papers in ancient books, as the change of color during aging is not easy to control and depends on a variety of environmental factors. Frequently, neighbouring book leaves may differ in color to a great extent. This comes sometimes from different glues or other components and later from the activity of microorganisms. 42 chapter 2

There are too many unpredictable reasons for color difference to mention it as a qualification standard. Intentional change of color, however, into blue and black or yellow by means of paint or ink is undoubtedly significant for paper identification.

Methods of Processing Blue/Black Paper (Mthing Shog) My recipe for blue and black papers is based on an English translation of Dungkar Lozang Trinlé’s “How to Make Blue Paper” and interviews with Tibetan craftsmen. Written down by craftsmen from Dharamsala, this text is described as showing “how the blue paper is made according to ancient Tibetan science.” To make the paper layers adhere together, wheat flour should be boiled with water and then spread on paper. Once the paper has dried, it should be cleaned with a piece of cloth. Then, one piece of wet paper should be placed on top of a pile of dry sheets. The entire pile should be pressed with the wet sheet placed on the top of the stack. Next, black ink made of ye-leaf fir (sgron shing)38 is blended by hand on an ink slab. The boiled brains of yak and sheep are mixed into the ink and this ink mixture is boiled. Next, melted bean- starch fibers are added to the concoction. The warm ink preparation is ready to be applied to the pressed pieces of paper. After the ink is applied, the paper is placed on a smooth birch-wood surface and a dzi is used to smooth it out. After the first smoothening with the dzi stone is complete, the ink is re-applied and the paper is smoothened once more. Another formula describes how blue dye is made: “To obtain blue paper on which scriptures are written in gold or silver, purify black vitriol (nag mtsur) liquid and let the sediment settle until the liquid is clean; the sediment should not contaminate the fluid. Brew the thar nu39 but do not let the steam get into your eyes. Once it is ready, mix it with borax (thsa la) and filter it through a cloth that is similar to what is used for filtering tea. After that, add the mixture to sour beer made of wheat and boil it. Remember to add less black vitriol and more wheat beer and thar nu; take great care to maintain the balance of the ingredients. The obtained liquid should be tested on white paper: if the paper turns darkish, this is a certain sign that the substance was prepared well. Let the mixture cool for some time, and when it becomes lukewarm apply it on good traditional Tibetan paper, not the modern short-fibered one, and let it get dry. The next stage requires boiling myrobalan (a ru ra) pulp; filter the juice that is rendered in this process through a cloth similar to the one for filtering tea and apply it on paper to give it the desired blue color. When making ordi-

38 Sgron shing—Pinus longifolia in Sikkim. 39 Thar nu—a type of medical root used as a purgative. Methods 43 nary paper, the methods for preparing mixtures and applying them are similar.” According to the Chinese custom, blue paper is made by applying a solution to the paper; the solution consists of an extract of Tibetan wheat beer, ground cardamom (dz’a it), and burned and ground white cowry shell (’gron bu). Paper prepared with this method should be left to dry and subsequently folded and sprinkled with myrobalan juice. Blue paper can be also prepared by dyeing with indigo or by painting with a brush, or by a combination of both methods—blue margins and center-inked. I also observed stamping on the paper and papers with a clearly discernible finishing polish.

Yellow Paper Yellow arsenic, known in Europe as orpiment (Arsen sulphuric, As2S3), has been known and used for many purposes in most of Asia since early times. It is generally believed that Tibetan books were successfully protected against insects by an addition of some orpiment or other insect-deterring substances derived from plants to the paper structure. I, however, have encountered very few Tibetan books treated this way. Most such volumes, in fact, were from Nepal and India rather than Tibet. It is worth mentioning here again that Stellera, Euphorbia, and Daphne fibers act as insect repellents due to their poisonous properties. Besides yellow sulphate of arsenic for protection from pests, Chinese paper was soaked in plant extracts that repel insects. This procedure also gave the paper a yellowish color. Gibbs and Seddon identified this yellow color as a natural huangbo dye derived from Phellodendron amurense tree.40

Texture Because a very simple method of production is used initially, the texture of the paper sheets is not smooth at first. To enable convenient writing, the paper requires further processing. To smooth the texture, the paper would be soaked in starch glue and sometimes covered with a layer of plaster or chalk and subsequently polished. In Tibet, barley powder was also used for this purpose. Paper treated with such preparations become smooth and impervious to oily substances and ink penetration; only paper with surfaces prepared in this way would be allowed for hand-lettering and printing of text using xylographic blocks. Analyses of the binding media in the paper in the Tibetan books show a presence of animal glue and starch. Most probably, this means that the paper sheets (of three to five layers) were adhered with animal glue and finished with

40 Gibbs and Seddon 1998: 20. 44 chapter 2 a flour and starch paste, or that just starch paste was used to glue and finish off many thin sheets of paper of uneven thickness. The adhesives identified must have been introduced to the paper during the process of book preparation, not during papermaking as was done in Europe and in East Asian countries, such as Korea and Japan. The main property of paper texture is its absorbency (how ink spreads on the surface). Absorbency is a significant feature to observe when identifying and classifying ancient papers. In contrast to both the Chinese and Japanese traditions, Tibetan calligraphy in ancient books was created with a stylus (bamboo pen), not a brush. The characters resemble those in medieval European manuscripts on parchment rather than Chinese or Japanese characters. The features of traditional Tibetan paper undoubtedly contributed to this lettering style. Tibetan paper made with Daphne and Edgeworthia fibers was traditionally rather thick, with an uneven surface, when compared to thin, silky Japanese gampi and mitsumata paper made with similar fibers from the Thymelaeaceae family. Unlike the Japanese papermaking process, Tibetan sheet formation does not involve the use of a formation aid, and thus results in paper with uneven fiber dispersion. To obtain a smooth surface, Tibetans glued a few layers of paper together and finished the surface with additional substances, such as wheat powder (tsampa), barley powder, black ink, or even chalk and plaster.41 The paper was then polished, making the surface smooth and impervious to oil and ink penetration, in fact virtually non-absorbent (Figure 14). The differences in ink drawings or calligraphy from China stem from basic paper properties such as absorbency. Chinese calligraphy, which was printed on short-fibered bamboo/rice-straw paper, has an exacting quality in its geometric, square-shaped figures, and implies a greater sense of control over the medium. Thus, the distinctive character of Chinese and Japanese calligraphy is directly related to how the ink absorbs into the paper during writing or painting. Traditional East Asian paper was not sized, although Chinese paper was generally more absorbent than Japanese and Korean paper due to an extra stage in the papermaking process. In all three countries, the raw material was beaten to fibrillate the fiber. The Chinese then added another step to the process—the cutting of the fibers with a knife, which resulted in a paper with short, blunt-ended fibers—thereby affecting the paper structure, the absorbency, and the overall writing style.

41 Vorobieva-Desiatovskaia 1988: 334. Methods 45

Figure 14 Highly sized and polished surface of a Tibetan type of paper made of Thymeleaceae family plants fibers in Tibetan manuscript from The Asia and Pacific Museum in Warsaw, Poland (MAP 14144).

Thickness The leaves of most of the examined books comprised three to five layers of thin-woven paper glued together into thicker sheets. Often the sheets of such paper made in the Himalayas do not have a uniform thickness. This is the 46 chapter 2 result of the conditions and the methods of papermaking used, which do not always allow for adequate separation of fibers and for precise pouring of sheets of the same thickness. In most cases, it was possible to use only one side of such a sheet of paper, due to the high absorbency of the reverse. There were also books that contained paper sheets made by gluing two to eight or more layers of paper together. chapter 3 Tibetan Books: A Craft and Artistry

It is generally accepted that a significant number of books and documents representing the highest level of artistic achievement were produced in Tibet. Created by people of a specific culture dedicated to preserving sacred Buddhist teachings, the existence of these books as physical objects may be overlooked if one focuses on their use as offerings, tools of persuasion and conversion, objects of religious devotion, and symbols of status (Figure 15). As Schaeffer states, Tibetan books are also the embodiment of the Buddha’s words, a medium of Buddhist written culture, a symbol of the religion itself, a principal means of education, and a source of tradition and authority.1 This study, however, will project Tibetan books first of all as crafted aesthetic works—objects which can be commissioned, collected, traded, read, admired, or desired by people. Importance is given to the work of artists and craftsmen—their conscious or more deliberate choice of materials and artistic expression as reflected in differences in style, arrangement, formats, and other techniques used in book creation regionally and during particular periods of time. Tibetan books provide evidence of the great artistic skills of Tibetan craftsmen—the skills of scribes, painters, papermakers, carvers, printers, and editors, together with the support of patrons and owners. In this regard, both the function of the books and how they are perceived in social practice often reflect the expense of their production. This production entails not only precious materials but also—and perhaps most importantly—requires diverse human creativity and effort in order for particular books to garner spiritual merit. Over a period of more than one thousand years, Tibetans developed and refined the arts of reproducing manuscripts and printing books from wooden blocks. They continued the tradition of manuscript production even after the invention of xylography. The manuscripts were still being produced in the second half of the twentieth century, when Tibet had become part of China and the editing process was automated. The monks reproduced many simple copies in black ink on paper, as well as lavishly illuminated editions written in gold or other precious materials on black or blue paper. Both the materials composing a book and its format reflect its region of origin and cultural affiliation, but

1 Schaeffer 2009: 128–140.

Figure 15 Tibetan Prajñāparāmitā manuscripts used as an element of a ritual ensuring a good harvest in Kyirong Valley, Western Tibet in June 2013.

can also be related to the function and use of the book, the donor’s expectations of attaining spiritual merit from creating it, and the availability of materials. Tibetans adopted block-printing technology as early as the twelfth century and continued to carve woodblocks for printing into the twentieth century. From the seventh century when Tibetans were able to write their most significant achievements in their own language, the process of making books has been initiated particularly by the production of numerous copies of the Buddhist canon, but eventually has been spurred by other works as well. By that time, a variety of writing materials were already invented and available to Tibetan craftsmen. In China brushes and bamboo pens were used to write on silk and paper, respectively. The scroll is one of the earliest formats of books produced on silk and paper. Dating as early as the third century, the oldest preserved books on paper in scrolls in Central Asia are from the Mogao caves in Dunhuang. Scrolls not only contain texts but also are used to mount art objects such as paintings and calligraphy. In China and East Asia, the production of thin, high-quality paper was developed specifically for such purposes. Silk textiles were often attached to the ends and reinforced with wooden dowels for rolling. Books on bamboo sticks, wooden slips, or birch bark were known in specific regions of East and Central Asia. Birch bark, in particular, was available across the Tibetan Books: A Craft and Artistry 49

Himalayas and was favored by the Kashmiris, and was known and used by the Tibetans from at least the eighth century.2 In India and the countries of Southeast Asia, palm leaves were the most popular medium for writing Buddhist sutras since at least the first century BCE. Exactly when manuscripts began to be written in India is unknown. The oral tradition of recitation and memorization was extraordinarily strong in early Indian culture, and has continued into modern times. In spite of the strong privileging of memorization through all periods of Indian history, it is clear that manuscripts were written and copied in increasingly large numbers, probably from the last few centuries BCE. Written text and its main carrier, the manuscript, had become indispensable. For example, when the great Buddhist universities were founded by the late Gupta and early Pāla kings after the fifth century CE, libraries and scriptoria were an essential part of these establish- ments. In his biography of the famous Chinese pilgrim Xuanzang, Huili reports that when Xuanzang returned from India to China in 644, he brought 520 manuscripts with 657 different texts from Nālandā.3 The consistent use of this set of books created a very specific and unified format in book production. Called grantha—or later, pustaka (glegs bam)— they are comprised of long, narrow, rectangular leaves held together by a string that ran through holes in the middle and was bound around two wooden covers (Figure 16). The texts of Indian books were first inscribed with a wooden pen or a sharp stick on the surface of the leaf; next, ink rubbed on the leaf that would remain in the scored depressions. Palm leaf is very fragile material that is easily destroyed by moisture and insects. This is probably why the oldest examples of palm-leaf books have been found outside of India; palm-leaf books preserved in India are dated after the sixteenth century. Most of the earlier surviving Indian and Nepalese palm-leaf manuscripts with illustrations that date from the eleventh to twelfth centuries, have been found in Nepal; in India itself, few books survived the Islamic invasion. Both Indian and Nepalese palm-leaf books follow a horizontal format; up to the eighteenth century, palm-leaf books continued to be used by Nepalese scribes simultaneously with paper.4 The survival of ancient texts was secured only by continuous copying, which was viewed as a meritorious act for both scribe and sponsor. Paper began to supplant palm leaf as the most abundant writing support from about the twelfth century; some of the earliest paper manuscripts have been found in

2 Reynolds 1991: 21; Chiodo and Sagaster 1995. 3 Steinkellner 2004. 4 Reynolds 1991: 21. 50 chapter 3

Figure 16 Pothi book on palm leaves. The National Museum in Warsaw, Poland (SKAZ 47669).

Jaina libraries in Gujarat and Rajasthan.5 Before the twelfth century, the principal writing supports in India were birch bark and palm leaf; wood, copper, or textiles were also used occasionally. After the Samye [bsam yas]6 debate between 792 and 794, when Tibet embraced the Indian form of Buddhism and created its own script, Tibetans also inherited Indian formats for their books together with the holy texts.7 It is possible that a variety of formats appeared later, influenced by China, the Arab world, or other cultures present along the Silk Roads. However, when Tibetan interest in once again returning to the source of Buddhism brought Indian pandit Atīśa (982–1054) to Western Tibet (Guge) in 1042, and to Central Tibet

5 Macdonell 1968: 19; Wujastyk 2012. 6 The Samye dabate took place at Samye monastery at the request of Tibetan king Khri srong lde btsan. The point was to decide whether Tibetans should learn Buddhist teachings from Indian or Chinese scholars. See: Snellgrove and Richardson 1978: 70–71. 7 Grönbold 1991: 16–18. Tibetan Books: A Craft and Artistry 51 in 1045, the loose-leaf format became the most widely adopted.8 Looking at the origin of Tibetan book formats and materials used from the practical point of view, neither palm leaves nor bamboo grew in Tibet and clearly would not be very convenient for executing the vast Tibetan canonical literature. Silk was not only expensive but also it was not possible to write on it with a wooden or reed pen in accordance with Tibetan habits. The new medium of paper allowed for larger and more varied formats that could hold more text and decorations on each page. According to archaeological records, paper was introduced to the Xinjiang area very early via the Silk Road. Manuscripts written on paper as early as the third century have been found at Kaochang, Loulan, Kusha, Kotan, Dunhuang, and Turfan sites. The art may eventually have reached Tibet around 650 and possibly then spread to India; because the range of disagreement on the date for the beginning of paper production in India is more than a thousand years, India is one of the most difficult areas for the reconstruction of paper history. Most scholars assume two possibilities for how papermaking reached India: from China through Central Asia, Tibet, and Nepal; or through Islamic traders in the Indian Ocean and, later, via the Muslim invaders who entered India from the thirteenth century onwards.9 It is notable that several of the later centers of papermaking are Islamic-founded cities such as Ahmedabad, Hyderabad, Faizabad, and Aurangabad.10 And, knowledge of papermaking may well have arrived from more than one source. Recent findings on manuscripts support the view that Tibetans were able to make paper at least by the ninth century and that they apparently adapted plants of the Thymelaeaceae family—which occur widely in the Himalayas— as raw material for the technology invented and practiced by Chinese communities.11 Chinese paper was primarily created for writing with a brush; without additional preparation, it was not very suitable for text executed with a sharp- pointed stylus. Tibetans, similarly to Arabs, used pens with the nib cut out of a reed or bamboo shaft. Writing with such pen requires an even surface and virtually non-absorbent paper. This is why Tibetan paper was glued in layers and polished to resemble a thick leaf, sometimes scholars have even called this layering paper-mache—which it is not. This aesthetic could also reflect a desire to make Tibetan paper resemble Indian palm leaves, which are naturally thick with a glossy surface. Thus, the use of thick and glossy paper in Tibetan

8 Snellgrove and Richardson 1978: 108–111. 9 Losty 1982 makes important remarks about the earliest history of paper manufacture in Nepal, which predates the importation of paper technology from China. 10 Karabacek 2001. 11 Helman-Ważny and van Schaik 2013. 52 chapter 3 books could also be the fulfilment of a wish to make a perfect imitation of a palm leaf. Keeping the idea of the faithful copying of Indian books in mind, it should be emphasized that Tibetans wanted to copy the material and formal aspects of these books as well as their textual content. When creating a Tibetan book, we might assume that the primary concern was how to maintain all possible elements of pustaka. However, the Dunhuang manuscripts written in Tibetan, which are in fact the oldest preserved Tibetan books, include a wide variety of formats. The most common types are pothi, scroll, concertina, codex (thread booklet), and single sheets. These various types of bindings add to our understanding of the history and culture of the book, and offer insights into the study of readership, the book trade, cultural connections, and the provenance of books. This record of bindings thus offers an integral understanding of the social history of the use of books, their production, and the history of craft in Tibet. Tibetan names for books related to their format include dpe cha, glegs bam, or deb ther. As van Schaik recorded we have numerous examples from the Dunhuang manuscripts of the early use of dpe or dpe’ for ‘book.’12 Even more common is glegs bam, which is found in Buddhist scriptures translated in the Tibetan imperial period, including sūtras copied at Dunhuang in the early to mid-ninth century. In scribal notes among the Dunhuang manuscripts, van Schaik found variations including glegs bu (small book) and glegs tshas (book covers, used by scribes as spare notepaper). For examples of the terms dpe and glegs in scribal colophons, see the catalogues of Tibetan Dunhuang manuscripts by Marcelle Lalou and Louis de La Vallée Poussin.13 Along one side of Pelliot tibétain 405, a scroll, the label, dar ma gser can gyi dpe, provides evidence that dpe could refer to a scroll. The colophon of IOL Tib J 220 refers to the Indian and Chinese books (rgya gar dang rgya’i dpe) that were used by the translator Chos grub, who lived and worked in Dunhuang.14 In these early sources we also find another term for Buddhist books that did not survive into the later era: dar ma, presumably from the Sanskrit dharma. R.A. Stein has discussed the use of the term dar ma to refer to books of Buddhist scripture.15 The loan-word pothi (and its derivative pod) has not been found in Dunhuang but appears as early as the eleventh century in transmitted literature.16 Also Stoddard states that, according to oral tradition, Atīśa was a figure who specifi-

12 Helman-Ważny and van Schaik 2013. 13 Lalou 1939–1961; de La Vallée Poussin 1962. 14 For the study of terms done by Sam van Schaik. See: Helman-Ważny and van Schaik 2013. 15 Stein 2010: 47–49. 16 van der Kuijp 2006: 5–6. Tibetan Books: A Craft and Artistry 53 cally recommended the pothi format as the most convenient book form for Buddhist teaching and practice exactly at that time.17 However, the format itself was adopted by Central Asian Buddhist centers like Kucha and Khotan already in the fourth century CE. Dunhuang collections indicate that the pothi was the most common book form for Tibetan manuscripts by the tenth century, but other forms were in use as well. This loose-leaf format found in palm- leaf manuscripts, and usually known as pothi (Sanskrit: pustaka), was adapted to paper instead of palm leaves. In the Tibetan pothi format, the elements are leaves—one or a few layered papers, a cloth cover of cotton or silk, labels of brocade or other cloth, and covers of wood with a wood and metal fitting (Figure 17). A traditional Tibetan book comprises two outer covers, usually made of wood, that hold numerous unbound folios each made of several layers of paper glued together; texts are written in calligraphic script. The title page sometimes serves as an inner cover and is always richly ornamented with beautiful calligraphy and often decorated; the paper layers of Tibetan books are joined together with paper or leather clips. The leaves are wrapped in cloth called dpe ras and secured with a flat cord or sometimes a buckle (Figures 18). Some volumes have carved and gilded wooden book covers. The volume labels—made of silk, cotton, or brocade—identify the placement of particular books within the collection. The idea of Tibetan binding is that loose leaves are kept together by wrapping them in cloth and by placing the entire book in between wooden boards (covers). Indian pothi, which are also on loose palm leaves, have a string that passes through all the leaves to keep the leaves of the book together (bound). A scroll can be bound by gluing all of the panels together and mounting them with wooden rolls; however, not all scrolls have these binding elements. The definition of binding in professional literature is not really unified and even within the same discipline of conservation, scholars cannot agree on this term. The role of a book is to keep a text/content together ‘bound’ into one piece. For most speakers of English, a binding is only what holds the other parts together and it is difficult to imagine how loose-leaf construction can have a binding. In spite of this linguistic challenge, in this essay, I use the term ‘binding’ to refer to different levels and elements of the construction of the book. By different levels, I mean the total assembly vs. parts within it. In conservation terminology, for example, all of the elements of a medieval book (codex)—such as wooden boards, leather, the sewing, bands strips, etc.—are understood as elements of binding. Despite the variants in the case of loose-leaf books, however, the term ‘binding’ can still be applied and understood in very much the same way, as a

17 Stoddard 2010: 375. 54 chapter 3

Figure 17 Typical Tibetan pothi format books seen in 1997 by an author in Braga Monastery, Nepal.

Figure 18 Tibetan pothi format book wrapped in cloth, Library of Tibetan Works and Archives, Dharamsala. Tibetan Books: A Craft and Artistry 55 general qualification for the variations, whether of color, fabric, lettering, or decoration, between different copies of the same edition of a book cased in publisher’s cloth.18 When most of the parts of this type of binding are separated and easy to replace, the loose-leaf construction of books can pose limitations on the definition of binding style. Even within the same type of book construction (pothi format), there is still variety—various elements, techniques, and materials, such as single- or multi-layered paper or different elements of binding style, are found. The size of pothi book leaves varies. The Tibetan pothi books from Dunhuang fall into two groups: a larger (20.5 × 72–73 cm) and a smaller size (7.5–10 × 26.5–46.5 cm). The larger size is used only for the Śatasāhasrikā prajñāpāramitā manuscripts. The smaller size includes Buddhist sūtra and tantra texts, as well as a ritual text.19 Vorobieva-Desiatovskaia documented leaves that were 7–14 cm in height and 20–65 cm in width;20 according to Grönbold, the largest known book measures 130 cm high × 170 cm wide (preserved in Sakya [Sa-skya] monastery).21 The large size of Tibetan canonical volumes exemplifies the change from long and narrow palm leaves to the new medium of paper that allowed for larger and more varied formats that could hold more text and decorations per page. The Kanjur volumes from the Pander Pantheon collections exemplify two sizes: most of the leaves measure 23.8–24.5 cm in height and 68.5 cm in width and nine volumes are slightly larger with leaves 25–25.5 × 71 cm (vol. 29–37 and 58). The Yongle folio preserved in Michigan (pothi format) measures 24.2–24.5 × 68.7–69 cm, which is very close to the size of most of the volumes from the Pander Collection (vol. 23–28, 38–57 and 59–60). The volumes from the Harvard-Yenching Library measure 24 × 68.5 cm. All volumes are very close in size and reflect the characteristic size of Tibetan Kanjur volumes produced in Beijing. As mentioned above, the majority of the Tibetan books in all examined collections are in a loose-leaf format, but in a variety of sizes. Tibetan scribes did alter the size of the page, largely because they were using materials different from palm leaves. The sizes range from very small pothi books that are similar in proportion to palm leaves, to quite large volumes of the Kanjur measuring 26.8 cm in height and 73.2 cm in width. This large size is exemplified by the ‘Handwritten Kanjur of Berlin’ acquired in 1889 by V. Brandt (Peking Kanjur);

18 Carter and Barker 2004: 41. 19 Helman-Ważny and van Schaik 2013. 20 Vorobieva-Desiatovskaia 1988: 336. 21 Grönbold 1991: 10. 56 chapter 3 handwritten with black and red ink and preserved in the Berlin State Library, it measures 26.8 cm in height and 73.2 cm in width. The layout of the text on the page depends very much on book format and page size, and can provide clues as to the function, date, and geographical origin of a manuscript. Thus far, there have been no general studies of page layouts in Tibetan books. Features, however, such as the number of lines per page and the presence or absence of margins and guidelines are helpful for making typologies and identifying particular editions of text. Among pothis, the presence of string-holes and whether or not a circle surrounds or represents these holes are characteristic elements that refer to Indian palm-leaf books— pustaka, despite their lack of practical use in Tibetan books on paper. They probably provide some hints to the appearance of the master copy. The wooden covers of pothis were often skilfully carved, painted, richly gilded, and sometimes decorated with metal fittings. The covers of Tibetan Buddhist canons produced in China were often lacquered and painted. Tibetan and Nepalese covers were often deeply carved with floral or figural reliefs, and gilded (Figure 19). Painted covers frequently display geometric patterns and have some shallow carving. They are usually varnished and appear glossy, although created with the same mineral pigments as Tibetan scroll paintings. For example, the longer and narrower covers usually belonged to palm-leaf Hindu manuscripts on palm leaves, representative also of Nepalese book formats. These covers have holes through which a single binding cord was drawn and knotted on their outer sides. The cord then ran through the palm leaves of the manuscript, allowing it to be opened, read, and then closed and tied together (Figure 20). The covers being a part of a book written on paper are usually larger and have different proportions because of the shape of the paper leaves. Painting the inner sides of the manuscript covers with scenes related to the content of the books is common to the bookmaking traditions of India and Nepal. The wood used for the covers varies depending on the region, depending on which trees grow in particular areas. In tropical regions shorea, sal, and sandal trees were favored. In Central Asia various fruit trees and jojoba were preferred. In the northern parts of the Himalayas cedar, juniper, pine, fir and other coniferous trees were used. The construction of the covers also varies from a single piece of wood to a composite plate with five wooden parts wedged together without nails. Determining the origin of the wood helps in tracing where the manuscript was produced. Specific manuscript attributions are difficult, however, because covers often become separated from manuscripts, occasionally are produced Tibetan Books: A Craft and Artistry 57

Figure 19 Loose wooden book cover from the Library of Tibetan Works and Archives, Dharamsala, India.

Figure 20 Palm-leaf pothi book with covers which have holes through which a single binding cord was drawn and knotted on their outer sides. The cord then ran through the palm leaves of the manuscript, allowing it to be opened, read, and then closed and tied together. The National Museum in Warsaw, Poland (SKAZ nr 237). 58 chapter 3

Figure 21 A seventeenth century handwritten London Sel dkar Kanjur. Notice that the covers are made of cut-to-size tropical wood with the slightly careless application of lacquer and gold decorations compared to those used for book leaves produced in Tibet. The British Library Collection (OR 6724 vol. 56).

at a later date than the text, are made from wood purchased elsewhere, or are damaged and then renewed. To draw clues from the covers, we first have to be sure that they are truly part of the book. For example, a seventeenth century handwritten London Sel dkar Kanjur in the British Library, brought to England by the Younghusband expedition, contains covers that at first sight seem properly suited in size. But the application of the lacquer and gold decorations was slightly careless compared to that used for the book leaves produced in Tibet (Figure 21). Identification of the wood showed that the covers were made of tropical wood indigenous to India, not Tibet. Such wood was usually not transported over long distances. From this we can deduce that the wood was brought to Tibet from quite a distance and thus that the covers were probably made later than the book leaves, possibly added by one of the previous owners. For example, in the Tibetan Book of the Dead from the British Library (OR 15190), the upper cover is made of a wooden plank wrapped with paper dyed indigo and glued to the cover under a top layer of silk (Figure 22). The same cover is carved and painted on the inside with two silk respect curtains attached to the top. The plank used for it is made of softwood cut radially, whose surface is finished. A good example of gleaning hints on dating and origin is the Tibetan Book of the Dead from the British Library. This is a Tibetan manuscript produced on Tibetan Books: A Craft and Artistry 59

Figure 22 Large pothi format the Tibetan Book of the Dead. The upper cover is made of a wooden plank wrapped with paper dyed indigo and glued to the cover under a top layer of silk. The British Library Collection (OR 15190).

Chinese-type paper, with a fibre composition based on paper mulberry, ramie, straw, and possibly bamboo. Those were the most popular raw materials for papermaking in China and also typical components for xuan paper. This type of paper features little space between laid lines. It was made by hand with a movable mold and fine bamboo sieve. A thin layer of paper was dyed indigo by a dipping technique. All these characteristics suggest a Chinese origin of the paper. On one side a 1.5 cm natural-color margin is left—the place for a finger to grip while pulling the paper through the dyeing solution. The book was not dated, but the eighteenth century was deduced from the text. The wooden book cover was analyzed using a non-destructive method of measuring its rings, and a preliminary chronology of the rings suggests the book cover could originate from East Tibet. Radiocarbon dating shows a plateau of the last three hundred years on the calibration line, but with the highest probability of 1680– 1770 or 1800–1940. In general, the scroll format originated in China and is exceptionally rare in Tibet. The Dunhuang collections, however, indicate that some manuscripts that were written in Tibetan and representing this format were used for religious and literary texts, as well as official documents. For the Dunhuang scrolls, we distinguish those that are written with the scroll aligned vertically from those in which the text is arranged in horizontal panels. (In some studies, this form is referred to as a ‘roll’ to distinguish it from the previous type.) When paper became the dominant medium for transcribing texts in the first few 60 chapter 3 centuries of the first millennium in China, replacing wooden strips and silk, the format of the Chinese paper book initially emulated that made of wooden strips tied together to form a continuous text reading top to bottom, right to left, and then rolled for storage. Paper panels, their size dictated by the paper molds, were pasted together to form long scrolls. This was the dominant book form for several centuries in China. In the latter part of the millennium, new forms start to appear that evolved into the codex. In Tibet, the scroll format was adapted for a thangka painting. Since early in Tibet’s history, religious representations have been painted on a specially prepared textile ground, framed with brocade or silk, to which silk curtains were attached. The concertina form was also found among Tibetan manuscripts from Dunhuang. It, however, has not been widely transferred to later times. In Asia the format was in common use for religious books, often compendia comprising a variety of texts. An interesting feature of this type of binding is its apparent synthesis of traditional Chinese and Indian/Southeast Asian book forms. In China concertina books probably evolved from the Chinese scroll format; Buddhists used the concertina format almost exclusively since the ninth century. This transformation, made possible by the new medium of paper, allowed an unfurled scroll to be folded in pleats, much like an accordion or concertina, creating individual pages and thus enabling the reader to flip through the text with ease. This form resembled Indian palm-leaf books and allowed for effort- less navigation through text. Books with such pages did not require a string to be passed through them, and much less damage was done to the paper, increasing the longevity of the books. In contrast to the scrolls’ wooden dowels, concertina covers were wider and flat, which allowed for elaborate decoration. They were often painted and lacquered or inlaid with precious stones and gilded. The paper of concertina manuscripts could be thicker; in Southeast Asia, the paper of concertina manuscripts was often dyed, or colored black, and the text would be inked in gold or silver. These manuscripts were often kept in boxes that were also richly embellished. Manuscripts in the concertina format tend to be made either of individual panels glued together with narrow strips, or of larger, folded panels, or a combination of both methods.22 The two Tibetan concertina manuscripts from Dunhuang are both currently disjoined, but their original construction is evident in the narrow strips of paper that were previously used to join the individual panels of the concertina. Tibetan examples of concertina-style manuscripts were also found in the British Library (MS 13092, MS 12163), and in the Jagiellonian University Library

22 Drége 1984. Tibetan Books: A Craft and Artistry 61

Figure 23 Concertina-format books from the Pander Collection. The Jagiellonian University Library in Cracow (Pander A). within the Pander Collection (part A). MS 12163 is written on thick, glued layers of paper possibly of a Tibetan type; MS 13092 resembles more recent papers type of lesser quality. Both manuscripts are definitely much later than those found among Dunhuang books. The concertinas from the Pander Collection in Cracow were probably produced in China, since the size, bookbinding style, and materials composing these books are typically Chinese (Figure 23). Documents such as contracts and letters were written on single sheets and folded into thin, rectangular packages. Found among Dunhuang manuscripts and constantly used in Tibet, this popular format resembles a flattened scroll or concertina folded first in one direction, then into a square. This format was reserved especially for governmental use (Figure 24). Curiously enough, Berthold Laufer mentions the use of knotted cords in prehistoric China and Tibet for the conveyance of messages, chiefly in the transaction of government business.23 He based his assumption on a mention by the philosopher Laozi, who, in a sentimental yearning for the past, expressed the desire that he might bring his people back to the ancient use of knotted cords. According to Laufer’s

23 Laufer 1931: 6–7. 62 chapter 3

Figure 24 Unfolded Tibetan historical document. Collection of the Library of Tibetan Works and Archives in Dharamsala, India. understanding, Tibetans have a similar tradition, and several native tribes in the south of China availed themselves of this method as late as the twelfth century. To my knowledge, however, no such books have been found. Finally, the codex form (generally stitched on the left side or at the top) also became popular in the tenth century for religious texts, particularly those that were ritual or liturgical in nature. In Medieval Europe, codices eventually replaced scrolls and became the most common format for books by the time of the emergence of the Islamic tradition, which adopted codices as the primary format for sacred texts although in form that was slightly modified from that found in Europe. Islamic manuscripts and books were usually bound with covers and spines comparable to Western books. Although the origins of Islamic bookbinding are still debated, their form seems to be a further development of the protective functions of the book covers and the spine. The front cover was often extended into a flap that secured the book shut. The codex is created from sheets folded in half, called bi-folios, or single folios. Depending on the desired size, the sheets could be folded up to eight times creating sixteen folios, or double-sided pages of the book. Several of these folded sheets are then set inside one another to form a ‘gathering.’ Several gatherings are sewn together to bands of leather, parchment, or textile, thereby binding all the gatherings and creating the spine of the codex. The covers of such medieval books were usually composed of cut-to-size oak or beech boards covered in leather, parchment, or thin sheets of metal. A common view held by many researchers is that very few Tibetan books exist that are bound, and that the few that do exist are thought to be recent. Tibetan Books: A Craft and Artistry 63

Many sewn books, however, written in Tibetan exist over the world and are dated from the ninth century to the present. Jean-Pierre Drége has surveyed the Chinese and Tibetan codex manuscripts (‘cahiers’) in some detail.24 On the other hand, the word deb ther used by Tibetans to label this format does not appear in any early sources. Van der Kuijp suggests a probable Mongol origin of the term.25 A loan-word from the Persian daftar, it probably dates from the period of Mongol dominance in Tibet in the thirteenth and fourteen centuries. Largely made for official, chancellery purposes, the Persian daftar was bound at the top. A variety of kinds of mounting (folding and division of leaves into quires) have been documented in bound books. The thread booklets were made in two ways. In IOL Tib J 401, pothi leaves are folded in half measuring 8 cm in height and 38 cm in length, and stitched with leather thread (Figure 25a, b). This form is not found in any other Tibetan manuscripts from Dunhuang, and palaeographical analysis suggests that it may date to the ninth rather than the tenth century—earlier than other thread booklets.26 In IOL Tib J 530, the booklet is constructed by sewing two sections with silk thread, with each section comprising four bi-folios (Figure 26a, b). This method is also seen in other Dunhuang manuscripts. I found only one example of binding similar to IOL Tib J 401. This is another manuscript from the collection of the British Library written in dbu med script (OR 15193) and composed of seventy-one leaves. It is also made of pothi leaves folded in half measuring 8.5 cm in height and 26 cm in length, and stitched with red thread and additionally reinforced with thicker hemp cord (Figure 27a, b, c, d). This stitched book is composed of two gatherings covered at the front and back side only with a soft facing/lining (one-layer paper lined with red linen fabric) that overlaps the binding along the side by 4 cm. The first gathering contains thirty-four leaves and one side cover and the second gathering contains thirty-seven leaves and cover. There is a very interesting book written on birch bark in the form of a bound book with a leather cover, which resembles the Kashmiri or Arabic style. According to the exhibition label in the Tibetan Museum in Lhasa, this book is dated to the time of Srong btsan sgam po or the period of the Tibetan empire (seventh to ninth century). The book was not available for closer examination when I visited this museum in 2003; however, taking into consideration its almost square form (approximately 16 × 16 cm) and binding style, it must be a later copy of the mentioned text, or the birch bark leaves were bound much

24 Drége 1979. 25 van der Kuijp 2006: 4–5. 26 Helman-Ważny and van Schaik 2013. 64 chapter 3

Figure 25a, b Tibetan stitched book on paper from Dunhuang. The British Library (Stein Collection, IOL Tib J 401). later into the codex form with the leather imprinted cover extended at the front into a flap that secured the book shut—typical of the Islamic tradition. The book is stitched with the cord in four places—a method that is different from other Tibetan bound books. Taking into account all formal elements, such as birch bark pages in a square format, the leather imprinted cover and the bookbinding style, the book was most likely produced in Kashmir, but not earlier than the tenth century.27 Stoddard describes the same book in 2010 with references to Kawasaki’s article that states that colophon of this book records that this manuscript was copied out during the reign of King Anantadeva of

27 The bark of the birch tree is called bhoj patra or burza in Kashmir. Early Kashmiri manuscripts of the pre-Islamic era of Kashmir were written on birch bark, which was the only available, durable writing material in Kashmir at that time. This fact gives credence to the commonly held belief that the use of paper in the valley was introduced during the Sultanate period in the fourteenth century by Sultan Zain-ul Abideen, who sent two natives to Samarkand for this purpose. Tibetan Books: A Craft and Artistry 65

Figure 26a, b Sino-Tibetan stitched book on paper from Dunhuang. The British Library (Stein Collection, IOL Tib J 530). 66 chapter 3

Figure 27a, b, c, d The Tibetan manuscript written in dbu med script, made of pothi leaves folded in half measuring 8.5 cm in height and 26 cm in length, and stitched with red thread and additionally reinforced with thicker hemp cord. The British Library Collection (OR 15193). Tibetan Books: A Craft and Artistry 67

Kashmir, who reigned from 1028 to 1063 CE—a date that corresponds with form of this book.28 The next interesting example of bound book is the small prayer book Kh.- Tib. 67 from the Tangut city of Khara Khoto in western Inner Mongolia preserved at the Institute of Oriental Manuscripts in St. Petersburg (Figure 28).29 According to Shi Jinbo, the front and back covers of Kh.-Tib. 67 are both incomplete.30 This book is especially important, because it has a few astonishing features that are rarely seen together in a single example, making it difficult to interpret. (1) It has a booklet form suggesting an original butterfly-type binding. (2) It is one of the earliest prints in Tibetan language (dated to 1153).31 (3) It has page outline that resembles the usually later dated type of stitched books printed often with movable fonts and called ‘wrapped-back binding.’ (4) It is printed on Chinese-type laid paper made of paper mulberry. The similar form of block-printed and butterfly-bound manuscript is also exemplified by codices Kh.-Tib. 63 and 68 from the same collection of the Institute of Oriental Manuscripts in St. Petersburg. These manuscripts were written during the Xixia dynasty or earlier and were found in an earthen tower outside the city. The manuscripts found in other areas of this city were written during both the Xixia and Yuan periods. This is a double-side-printed, twenty-three-folio booklet (17.3 × 12.7 × 0.6 cm) composed of twelve bi-folios folded in half. Many of the bi-folios are split in the middle along the fold, but those preserved in one piece do not have any sign of needle holes—suggesting that they were not actually sewn. This confirms a butterfly rather than a stitched format of the book. The butterfly binding broke away from the concept of scroll or concertina format and started a new tradition of Chinese binding. A butterfly-bound book was made by folding sheets of paper in half, forming four sides of each for the text. Paste would then be applied to edges along the folds to stick the leaves together and to form the spine of the book. However, this format was usually printed on one side only. When printing two consecutive pages from a woodblock, only one side of each leaf could be printed at a time. This meant that

28 According to Kawasaki, this book contains the text relating to esoteric rituals, manuals, and commentaries, see Kawasaki 2004: 50–52; Stoddard 2010: 371–373. 29 Accession number: Kh-Tib.67. See also chapter 5 ‘Tibetan Woodblock Printing Culture.’ 30 Shi 2005: http://zt.tibet.cn/english/zt/TibetologyMagazine/200312007420165155.htm. 31 This date is suggested by Schaeffer 2009: 9. Stoddard refers to the Renzong reign period when the book was printed. According to Tangut dating, this is 1139–1193 or 1150–1179 according to Chinese. See, Stoddard 2010: 365. 68 chapter 3

Figure 28 A small prayer, printed in Tibetan language and found in Khara Khoto. One of the earliest known documents datable by colophon to 1153. The Institute of Oriental Manuscripts Collection in St. Petersburg (Kh-Tib.67).

every second page of the book would be blank. On the other hand it is printed with black ink on patchy, laid paper that was glued in two layers characterized by twenty-one to twenty-two laid lines in 3 cm. The texture of the paper is soft and absorbent, with brush traces originating from gluing the layers together. The paper resembles the Chinese papers in other books produced at that time that have been found in Dunhuang and other places along the Silk Road. Since the format is small and the leaves are made of two layers of paper, it is still possible that after printing this book from woodblocks that the consecutive pages from the neighboring bi-folia were pasted together. However, if this process was used, there should be a visible split on one side of the preserved bi-folia. Every bi-folio includes six lines of text printed on each page in a full frame 15.8 × 9.6 cm separated by the folding mark in the middle. In this Tibetan book with a butterfly binding, the text lines go right across both pages of one bi-folio. The Song dynasty is the golden age of butterfly-bound manuscripts. However, most Song manuscripts that were originally butterfly-bound were given stitched bindings by later generations. To our surprise, among the ancient manuscripts (with Xixia and Tibetan scripts during a period that corresponds to the Song dynasty) discovered by modern archaeologists, many block-printed Tibetan Books: A Craft and Artistry 69 manuscripts still have an original butterfly binding. The lettering is relatively narrow and in some places not well printed, and some letters were corrected by hand later. This page layout suggests that printing was intended for a booklet format, which is quite exceptional when taking into consideration the early date of its production. Additionally, the page outline resembles those of stitched books printed later, predominantly in the Ming dynasty (1368–1644 CE). Having a text composed on two pages with the folding mark in the middle was typical of those printed with movable metal type—not xylography. However, this example was printed from a woodblock.32 It is hardly possible to find two similar shapes in the same letters in the text, which should be possible if wooden movable types were used. Also, if they had been used to constrain the form for the placement of loose fonts, we would expect the frames to be more continuous and even. When you trace the central lines close to fold in Kh.-Tib. 67 the traces of carving can be observed. Moreover, an upper line of text and heads of letters are cut at a relatively even line, with some ink stains appearing outside of printed areas, which is typical for xylography. The documented forms of Tibetan manuscripts show a relationship between the format associated with the functions that a particular book served and the utility of these books. This can be observed, for example, when one compares the text content and format of loose-leaf pothi books and those that were sewn. The Tibetan Pharmacy from the Ethnographic Museum in Cracow exemplifies this relationship. It consists of two manuscripts—one is a medical treatise in a pothi format, and the second is a register of drugs in the format of a sewn book. Another interesting item is a bound book that is written in gold. This is exemplified by the gold manuscript with a Tibetan religious text from the collection of the British Library (MS 13162). This manuscript, is composed of forty- two leaves measuring 7.5 cm in height and 18.5 cm in width (book is 2 cm thick); it was brought as a gift from Bhutan in 1866. It is a stitched book composed of three sections with the middle band covered with cloth, and the sides covered with rainbow-colored silk (Figure 29a, b, c). The fact that every one of its three sections has its own band supported with silk or other textile is unusual. A band is usually applied to an entire book, and not separately to every gathering. This feature distinguishes almost every deb ther book and is the most typical for Tibetan stitched binding. I found more examples having the same form and binding style. The same or very similar scheme of sewing was seen in other Tibetan religious manuscripts from the British Library collection, such as CC 114–115, BL 14727 1–2, CC 101, OR 15193, or OR 11376. The first three examples are made according to the

32 Shi 2005: http://zt.tibet.cn/english/zt/TibetologyMagazine/200312007420165155.htm. Shi Jinbo 2005 cited by Shen Weirong 2010: 356 and Stoddard 2010: 364. 70 chapter 3

Figure 29a, b, c Bound book written in gold with Tibetan religious text brought as a gift from Bhutan in 1866. The British Library Collection (MS 13162). Tibetan Books: A Craft and Artistry 71 same rule, with every gathering having its own-supported band and stitching at the top of the book with a sewing cord outside of book. The cord is left longer and wrapped over a book. This secures the book closed in the similar way as a flat cord in the cloth attached to pothi binding. The fourth example (OR 15193) is bound at the left side and is formed of two gatherings; this fourth example has a format that is closer to the pothi format in the shape of its leaves. The manuscript with the accession number CC 114–115 has seventy-three leaves measuring 10 cm in height, 30 cm in width (Figure 30a, b). This book is 3 cm thick. Its paper leaves are folded once in a quire (into bi-folia). The book is composed of four gatherings: the first of eight bi-folia and front paper cover backed with textile; the second of eleven bi-folia with a strip of textile supporting the band of a gathering; the third of ten bi-folia with the same way supported band as previous gathering; the fourth of five bi-folia and six folia including cover lined with purple linen. The linen strips guarding over each quire measure approximately 6 cm in width. The book is stitched with linen cord in 3 sewing points in the section. Each leaf is made of three layers of paper. Almost the same construction of book can be observed in the Tibetan manuscript with accession number BL 14727 1–2 (Figure 31a, b, c, d). The third example (CC 101) is a bound manuscript measuring 8.3–8.8 × 14.5 × 2 cm with a religious text written on one-layered paper (Figure 32a, b, c, d). This book has 130 pages on sixty-five leaves and is composed of four gatherings. The leaves are folded into bi-folia: ten folia in the first gathering; nine folia in the second; eight folia in the third; and six folia in the fourth. The book is stitched with linen cord, and every gathering has a separate band secured by a stripe of cloth. Text is written in six lines of dbu can script on each page, and composed in a double frame on the recto pages and a single frame on the verso pages with margins measuring 0.8–1 cm. Red rubrics were observed on a few of the final pages. Tibetans used sets of planks (samta) for daily notes. Paper was too expensive in Tibet and was never been used for daily notes. Instead of paper, Tibetans used small wooden planks varying in size, but normally about 15 cm in height and 35 cm in width with a recessed/hollowed-out, black-inked writing surface. Tibetans learned to write script and calligraphy by writing on such tablets, known either by the name of samta or under the honorific bka’ sam.33 The black surface of the tablet is thinly oiled and covered with a heap of ash before writing. The dust of the ash adheres to the oil and enables the user to write on it with a wooden pen. This is an impermanent notebook; it is easy to efface the

33 Karmay 1998: 16. 72 chapter 3

Figure 30a, b Bound manuscript composed of four gatherings and stitched with linen cord at three sewing points in the section. The British Library Collection (CC 114–115). Tibetan Books: A Craft and Artistry 73

Figure 31a, b, c, d Tibetan bound manuscript composed of four gatherings. The British Library Collection (BL 14727 1–2).

text—similar to writing in sand. And yet, sometimes whole volumes were written on samta prior to being transferred onto paper. Writing on wooden tablets was not solely a Tibetan tradition. The Romans used tablets for their writing and several examples of their tablets have survived until present day.34 To summarize, we can find all the types of described formats of Tibetan books among the Dunhuang manuscripts. It is not yet known whether all these

34 They varied greatly: some were simple pieces of wood used for writing in ink, others resembled frames with soft materials such as lead or wax lining the shallow recess in the wooden leaf. Waxen tablets were used among the Romans for short pieces of writing of almost any kind or genre. Letters were frequently written upon such tablets and secured by being fastened together with pack-thread and sealed with wax. One would write on the surface with a stylus made of either iron, bronze, or bone—the stylus had one pointed end for writing and a flat end used to erase or, more accurately, to smooth out mistakes. For more information see: Smith, William (ed.), A Dictionary of Greek and Roman Antiquities, London, 1890–91. 74 chapter 3

Figure 32a, b, c, d Bound manuscript of four gatherings measuring 8.3–8.8 × 14.5 × 2 cm with religious text written on one-layered paper. The British Library Collection (CC 101).

book formats were in use in Central Tibet are products of an early time, but I am reasonably certain that official letters in the form of folded single sheets, and Buddhist scriptures in pothi and scroll form were brought to Dunhuang from elsewhere in the Tibetan empire in the early ninth century; from then on, these formats become the most commonly used type in Tibetan book culture. The scroll seems to have been an adaptation from China; scrolls were used mostly in Dunhuang, and disappeared after the tenth century. The concertina format was probably rare in early times and later become almost accidental in Tibet; however, this format was used for Tibetan books produced in China. Books having a bound format are probably the most difficult to interpret at the moment. My study of book construction in selected Tibetan bound books sug- Tibetan Books: A Craft and Artistry 75 gest that this format was used, however, on a much smaller scale than pothi, during all periods in Tibet and in Tibetan borderlands. However, ‘bound book’ or ‘stitched book’ as Stoddard call this format is a very general term. In actual- ity, we are talking about at least four types of book construction. Every type may suggest a different region of book origin and cultural influence on the techniques of its production. chapter 4 Indigo, Gold, and Human Blood: Tibetan Illuminated Manuscripts

Manuscripts versus Prints

Tibet produced manuscripts even in the golden age of xylograph printing during the eighteenth century. The coexistence of manuscripts and prints in Tibet until the twenty-first century is significant within Tibetan culture and cannot be easily explained without a thorough study of Tibetan books in the wider social and cultural context. Cheaper and widely disseminated, the value of printed books cannot be easily compared to the value of works distributed in manuscripts, as is often done with European Medieval manuscripts and their later printed copies. Further analysis of the relationship between a book’s form and its utility not only contributes to better understanding of the legal and social roles that manuscripts play in Tibet but also helps explain the survival of Tibetan manuscript culture after the invention of the printing press. Devotional and religious purposes, such as the accumulation of spiritual merit during book production, highly influenced the form and quality of work. The effort put into a book’s creation, however, as well as the type and quality of its materials, often distinguishes manuscripts from prints. Ornate and unique written books were not intended for mass dissemination but rather as offerings, or they were produced as a means of accumulating spiritual merit. Religious and political leaders often sponsored copying of texts by hand, because it was regarded as highly meritorious—even more than by the process of printing. Most manuscripts are believed to contain the Word of the Buddha and are revered and cherished by all Tibetan Buddhist traditions for possessing great sanctity (rtsa chen po). The attitudes and skills of scribes as well as the availability of materials relegate manuscripts to the category of cheap, simple books written by unprofessional scribes for their own purposes, or distinguish them among precious collections of books. The production of handwritten editions of Kanjur and Tenjur or other significant texts was a costly and labori- ous process during which an equally great expenditure of funds and resources was needed for each and every copy. Precious materials, the usually large format, the quality of calligraphy, the lavish decorations, and the complex construction of leaves all distinguish these deluxe editions from simple copies. The use of precious stones, gold and

© koninklijke brill nv, leiden, ��4 | doi ��.��63/9789004275058_�5 Indigo, Gold, and Human Blood 77 silver, and other rare minerals applied with special preparation techniques— especially if more than one copy is considered—make them more expensive than xylograph prints. One comparable example of a deluxe manuscript edition is the copy of the Tibetan Selkar (sel dkar) Kanjur, the so-called ‘London’ Kanjur from the place preserving it. Sometimes, however, when only one simple copy without vast decorations and precious materials was intended, manuscripts could be produced more quickly and cheaply than large projects of woodblock-printed editions. However, rarely also printing in a small scale could imitate the aesthetic of golden manuscripts written in gold on indigo paper, a technique which distinguishes Tibetan manuscript culture (Figure 33). In both manuscript and printed copies usually enormous effort input into book production distinguished Tibetan book culture—especially, the fact that woodblock-printed editions could also be lavishly decorated and painted with precious materials after text was executed. In this sense, the culture of manuscripts and printed book production in Tibet could represent similar attitudes and fulfil needs in much the same way by creating an opportunity to gain spiritual merit. However, the values of the two types of books cannot be measured with the same yardstick. We cannot simply oppose the craft of manuscript production to mass printing culture, as is sometimes done in Europe. Nor can we say that in Tibet printing was simply a more efficient means than handwriting or calligraphy for disseminating information.

Patronage and Ownership

What is evident from cumulative and developing studies of various examples of Tibetan decorated manuscripts is that the most important facilitating factor in Tibetan religious books’ production was the contribution of patrons, who provided resources for artists and often decided on the particular form and quality of a book. This is why there always has been a close link between patronage and the aesthetics of book production. Most extensively ornamented books were either volumes of the Tibetan Buddhist Canon sponsored by wealthy people, or Prajñāparāmitā manuscripts in various versions, the eight thousand and hundred thousand verses. Apart from being a canonical text, the production of this sūtra was often part of elaborate post-mortem rituals associated with amassing great merit on behalf of a teacher/lineage master and the deceased as well. One important factor that might have contributed to the distinctive way these Tibetan Prajñāparāmitā manuscripts were decorated is that their production was largely a means for accumulation of merit by their Tibetan donors. 78 chapter 4

Figure 33 Text printed in gold on blue paper imitating an aesthetic of writing in gold on indigo dyed paper. The R.R.E. Collection, Winterthur, Switzerland.

Even though the eastern Indian manuscripts brought to Tibet shared this function, the Prajñāparāmitā manuscripts produced in Tibet appear to reflect the local Tibetan environment, depicting aspirations of their donors quite literally. This can be understood by examining two decorated parts of the Tibetan manuscript of the Perfection of Wisdom Sūtra (Prajñāpāramitā, Sherabkyi Paröltu Chinpa) from the Herbert Johnson Museum of Art at Cornell University and the R.R.E. Collection. Both testify to the characteristic visual conventions for illustrating this sūtra found in Western Tibet during the fifteenth century. In both cases, the miniatures are extensive and well executed and their subject, the story of the bodhisattva Sadaprarudita (tag tu ngu, ‘Always weeping’), as presented in this manuscript, seems to conflate personages of the story with wishful representations of the donors, who are shown with lavish offerings while accompanying the main character on his journey to the palace of bodhisattva Dharmodgata (chos ‘phags, Exalted in Dharma) to receive the teachings. It is curious that the merchant on the R.R.E. Collection folio is shown as a Tibetan lay religious master wearing a red hat and displaying the teaching mudra, and the merchant’s daughter is shown wearing regional attire and hair ornaments.1 Similar to practices in medieval Europe, when commission- ing these works, wealthy individuals would request that the illuminator place them somewhere in the painting so that they could feel like actors within the narrative space of the story.

1 Pakhoutova and Helman-Ważny 2012. Indigo, Gold, and Human Blood 79

Gold Manuscripts

A distinct practice of writing Tibetan manuscripts in gold on a dark background and embellishing them with other precious stones began during the period of the so-called Tibetan Renaissance (tenth to thirteenth centuries). Tibetans describe the methods of making blue paper as an ancient Tibetan science that suggests a rather early origin. Dorji Wangchuk reports the existence of a deluxe edition of the Corpus of Tantric Scriptures written in gold and commissioned by Mnga’ bdag ’gro mgon dpal for his father’s memorial (dgongs rdzogs) in the year 1192. This way the death of Nyang ral nyi ma ’od zer dates the scriptures to the twelfth century.2 Giuseppe Tucci and Heather Stoddard describe a black-grounded aesthetic as an eighteenth-century technique in Tibetan art and associate it with black-grounded thangkas.3 In his book on the ‘Demonic Divine,’ Rob Linrothe mentions an interesting fact that might point to the possible genesis of the black-grounded aesthetic (nag tang). He describes a process reported by thangka painters according to which the ashes of a holy lama were applied to a canvas, and later a deity was drawn in gold upon the ash-laden canvas.4 This description suggests that this technique may have been favored for religious purposes. However, this may only be a Tibetan myth, as such mythologizing is often the case in discussions of art execution, descriptions of materials used, and explanations of the origins and inventions of Tibetan art. Tucci describes Tibetan paintings as liturgical, both because they originated as ritual elements and because of the evocations that artists must perform before beginning to paint.5 In this sense, the symbolism of this type of painting is richly connected with Buddhist philosophy and grounded in the context of Tibetan religious art. However, it is very difficult to collect evidence related to secret religious practices and the use of gold manuscripts in esoteric arenas. The works edited by Ye shes rgyal mtshan (1713–93) for his pupil and leader, the Eighth Dalai Lama, evince an intention of using those manuscripts “as a visual aid and meditation support.”6

2 All together 113 volumes of various scriptures in gold (gser chos kho na pusti brgya dang bcu gsum) including the largest version of the Prajñāpāramitā, an ‘anthology of [short] sūtras’ (mdo mang), and dhāranīs (gzungs ’dus). See: Wangchuk 2010. 3 Tucci 1949; Stoddard 1998. 4 Linrothe and Watt 2004: 30. 5 Tucci 1949: 288. 6 Karmay 1998. 80 chapter 4

Gold on blue is an attractive match of colors and is not exclusive to the Tibetan culture of manuscript production. Islamic manuscripts written in gold on parchment were produced at least in the ninth century. ‘Blue Qur’an’ pages written using a specially shaped pen on an indigo-dyed parchment or paper are preserved in various collections. We also find early Tang Dynasty Chinese gold sutras, Nepalese manuscripts written in gold on blue paper, and Thai, Burmese, Jewish, and medieval European manuscripts written in gold on purple, indigo, and later also black parchment and paper.7 Tibetans, however, produced them in great numbers, and examples of such manuscripts and individual folios separated from books they once belonged to can be found in many museums, libraries, and private collections. Such manuscripts are distinguished by their special decorative outlook and represent an exceptional form of Tibetan craftsmanship. The processes involved in their production demanded excellent artistic skills that were also considered worthy of gaining spiritual merit. Gold manuscripts are also distinguished from other manuscripts by the particular process needed for the preparation of paper and technique for executing the text by specific writing tools.8 From the technological point of view, these manuscripts are distinguished by the use of silver and gold inks on specially treated paper, the use of glue or particular binding clips that secure both written and painted sides of a folio together, and special treatment of the painted illuminations. As reported by Dorji Wangchuk at the 12th IATS conference in Vancouver in 2010, Bkra shis don drub, in his Mthing shog bzo rim, notes that each of the seven lines on each of the blue-papered folios has been written with a different ink made of a different precious material (rin chen sna bdun). The first line of a text should be written in gold; the second one in turquoise; the third in silver; the fourth in coral; the fifth in iron; the sixth in bronze or copper; and the seventh with ink made of conch shell. Drung chen mkha’ spyod pa Nam mkha’ rgyal mtshan (1370–1433), for example, is said to have commissioned deluxe editions of the large, medium, and small Prajñāpāramitā scriptures in turquoise, gold, and silver, perhaps respectively. However, there were no examples of books like this found among studied materials for this project. It is also supposed to be appropriate to write on blue paper with any sort of pen, as long as the pen is made of stone (brda rdo’ dmyu gu). Possibly, a ‘stone pen’ should be understood figuratively to refer to bamboo that was hard as a stone. Although it is doubtful whether and to what extent stone pens or other precious materials such as turquoise were actually used for writing, we do have accounts mentioning various kinds of materials.

7 Losty 1982; Reynolds et al. 1986. 8 For further information, see Sangs rgyas rgya mtsho 1990. Indigo, Gold, and Human Blood 81

Historical or administrative documents, decrees, certificates, or letters, however, were handwritten and could also be deluxe objects. The terms gser chos, gser yig, gser bris or shog gu bu thing nag were generally used for scriptures written in gold-color ink. Thus gser stands for a precious metal, literary gold, and chos, yig, or bris can mean scripture, inscription, letter, document, decree, or manuscript. As mentioned above, ‘gold’ is sometimes to be understood figuratively. Although, it is possible that royal decrees or letters of significant importance were indeed written with gold, as mentioned in the legend of Srong btsan sgam po’s courting of the Chinese princess.9 There is also a letter of threat from approximately the time of the early Tibetan kings sent to the Chinese emperor, which is described as ‘a letter in Chinese characters written with gold on blue paper’ (rgya nag stong khun gyi yi ge mthing shog la gser gyis bris pa) placed in a ‘casket of the royal commands’ (bka’ shog sgrom bu). It says that ministerial ranks were classified hierarchically according to the six types of insignia or certificates: g.yu yig (‘turquoise insignia/certificate’), gser yig (‘gold insignia/certificate’), dngul yig (‘silver insignia/certificate’), phra men gyi yi ge (perhaps ‘manganese insignia/certificate’), zangs kyi yi ge (‘copper insignia/certificate’), and lcags kyi yi ge (‘iron insignia/certificate’). In addition to the actual document/certificate inscribed in gold and the highest ministerial rank, the term gser yig is refers to a person directly associated with the document. Thus, gser yig (pa) or gser gyi yi ge pa refer to three kinds of persons, namely, (a) a minister who holds a gold insignia or certificate, (b) a messenger who carries the royal letter or decree, and (c) a clerk or calligraphist who prepares such a letter or certificate.10

Format and Book Binding Style

Independent of any particular style, most of the Tibetan manuscripts in gold, such as British Library manuscript OR 15190 or Columbia University manuscripts 1–6 and many others, represent the pothi format, but other formats are also in existence. The stitched book format is exemplified by a gold manuscript, BL 13162, with Tibetan religious text preserved in the British Library, brought as a gift from Bhutan in 1866. This book composed of forty-two leaves, including covers, represents the typical deb ther format (18.5 × 7.5 × 2 cm) specifically

9 Wangchuk 2010 (not published). 10 Wangchuk 2010 (not published). 82 chapter 4 sewn in three sections, where the middle band is covered with cloth, and the sides with silk.11 Independent of format, this aesthetic requires that leaves of books be thick enough to support many layers of paint and allow for heavy processing such as, for example, polishing. This is why they are usually made of at least two layers of thick paper, or up to ten if thin Chinese-type paper was used. This complexity of leaves composed of several layers of paper is common in Tibetan blue- ground manuscripts. Then to firmly secure the thick layered leaf resembling the cardboard and the upper layer, special clips stitch them together. The miniatures are often painted on an additional layer of black paper with a cut-out frame to accommodate the text lines written on the layer below, which is made of thick, black paper or cardboard to which the illuminated layer is attached. Such construction also allows for the division of the label between the painters and scribes, so they can work simultaneously. Finally, all painted and written components of the leaf are combined and mounted together. The front pages are especially elaborate in their construction. These types of complex pages often function as frontispieces or inner covers, which are then lavishly painted and inscribed with calligraphy with golden ink. The appearance of double covers, the outer usually wooden and the inner of paper, is a typical element of the Tibetan pothi format of illuminated manuscripts. The Tibetan Book of the Dead from the British Library, OR 15190 described in the previous chapter, has a wooden cover wrapped with paper dyed indigo, with two silk curtains attached to its verso side. The same book contains internal frontispieces at the beginnings of parts of book also finished with the silk curtains (Figures 34 and 35). Two frontispieces from the Tibetan manuscript of the Perfection of Wisdom Sūtra from the Herbert Johnson Museum of Art at Cornell University and from the R.R.E. Collection in Winterthur, Switzerland were mentioned above. The Johnson Museum manuscript’s inner cover (18–18.5 × 68.7 cm) is made of layered paper joined together with leather strips/clasps and has a green silk ‘curtain’ that is attached to the upper part of the cover drapes and protects the paintings when the manuscript is closed (Figures 36 and 37). The paper sheets were never glued together; they were simply mounted with leather or paper strips (Figure 38). Aside from the cover and frontispiece, all other leaves of the book are missing. However, it is clear that the manuscript was composed of loose leaves and very possibly had an additional wooden cover; such covers were typical of this type of manuscript, which was made as a special offering or gift.

11 This book in the context of its form and binding is also described in chapter 3. Indigo, Gold, and Human Blood 83

Figure 34 Verso side of a wooden cover wrapped with paper dyed indigo on recto side, and with two silk curtains attached to it at the top of verso side. The Tibetan Book of the Dead from the British Library Collection (OR 15190).

Figure 35 The internal frontispiece of a book with the silk curtains attached at the top. The Tibetan Book of the Dead from the British Library Collection (OR 15190). 84 chapter 4

Figure 36 Verso of the inner cover of the Tibetan manuscript of the Prajñāpāramitā, Sherabkyi Paröltu Chinpa written in gold on black-inked paper. Western Tibet, Fifteenth century. Collection of Herbert F. Johnson Museum of Art, Cornell University. Acquired through the George and Mary Rockwell Fund. Prajñāpāramitā Cover and Frontispiece (2006.028a,b).

Figure 37 Illuminated inner cover written in gold on black-inked paper and made of layered paper joined together with leather strips/clasps. A green silk ‘curtain’ that is attached to the upper part of the cover drapes and protects the paintings when the manuscript is closed. Collection of Herbert F. Johnson Museum of Art, Cornell University. Acquired through the George and Mary Rockwell Fund. Prajñāpāramitā Cover and Frontispiece (2006.028a,b). Indigo, Gold, and Human Blood 85

Figure 38 Illuminated inner cover written in gold on black-inked paper. The paper sheets were never glued together; they were simply mounted with leather or paper strips. Collection of Herbert F. Johnson Museum of Art, Cornell University. Acquired through the George and Mary Rockwell Fund. Prajñāpāramitā Cover and Frontispiece (2006.028a,b).

The R.R.E. Collection inner cover is even more elaborate in its golden calligraphy. The large letters that spell the title of the book are decorated with floral motifs and outlines similar to those produced by the ‘raised gold’ technique, which makes the letters appear almost in relief against the surface of the page. This cover does not have a curtain attached, but other elements of the leaves’ construction are very similar to those of the previous one. These Tibetan folios with miniatures are treated in the same way as a Tibetan scroll painting (thangka) would be; they have silk coverings and frames, and are painted on a specially prepared ground. However, the sum total of the decorated folios together with those that contain only text tells more. For example, the curtains distinguish Tibetan manuscripts from the earlier Indian and some Nepalese examples which contain only text and paintings. Interestingly, the only other known examples of books with curtains covering the miniatures, however exceptionally rare, are among European sacred manuscripts.

Illuminations and Decorations

It is well known that books written by hand were often illuminated with pictures and ornaments of different kinds. The word ‘illuminated’ comes from the 86 chapter 4

Latin word illuminare, and in connection with oratory or prose style, it means ‘to adorn.’ In Tibetan book culture both manuscripts and some special, woodblock-printed books were adorned with a variety of hand-painted decorations, such as miniatures, initial paintings, ornamented borders, circles imitating holes after palm-leaf books, and sometimes even doodles. It is notable that in discussions of these objects, the aesthetic appeal of the decorated pages draws the main attention and supplants the notion of them being lesser components of a manuscript. Often the illuminations are considered primarily as examples of painting rather than parts of a book. But when considering the style of manuscript typical for a particular region and time period, we should consider all elements of a book. Single elements can be misleading due to the variety of components and circumstances related to the creation of manuscript. In the making of an illuminated manuscript, first the leaves of paper were cut down to the appropriate size, glued in the appropriate number of layers to make a leaf thick enough, then processed on the surface to make them a good background for writing and painting. After the general layout of the page was planned (e.g., number of text lines, frames/borders, circles, initial signs), the page was lightly ruled with a sharp pointed stick or wooden coloring stick, and the scribe went to work with an ink-pot and reed pen to execute the text. Next, or sometimes simultaneously with the scribe, depending on the construction of the leaves, the painter started to work on decorations. The painter usually used a variety of brushes and mineral or organic-based paints. The main types of decorations are miniatures or small pictures on both sides of the pothi front page, incorporated into the text, often in Tibetan manuscripts representing the Buddhist pantheon. They are usually related to the text or to ownership of the book. Sometimes they were painted directly on the written page and sometimes attached later to the front page and mounted together with many independent elements, such as consecutive layers of paper, sometimes curtains, and later, frames also made of paper. The iconographic program of the miniatures, their layout within the folios, and the special way in which they are treated within the book have many parallels with other generously decorated thangka and wall paintings produced in particular regions of Tibet. The London Selkar (Sel dkar) Kanjur, OR 6724, was copied at Selkar monastery in 1712 by the calligraphers from Snyemo, a place in Central Tibet.12 The verso of the front pages this Kanjur were executed in raised gold on a black background with turquoise and corals mounted on each page within circles imitating holes in palm leaf books and ornamented in gold (Figures 39 and 40). However, every volume does not have precious stones mounted on the front

12 Pagel and Gaffney 1996: 9. Indigo, Gold, and Human Blood 87

Figure 39 The front page of Tibetan Selkar (sel dkar) Kanjur contain the title written in three text lines and two miniatures showing various representations of the Buddhist Pantheon in different volumes within these pages painted with gold on red. The British Library Collection (OR 6724).

Figure 40 Text written in ‘raised gold’ technique on the front page of volume 65 of Tibetan Selkar (sel dkar) Kanjur. The British Library Collection (OR 6724). 88 chapter 4

Figure 41 Front page decorated with ornamented frames in volume 65 of the Tibetan Selkar (sel dkar) Kanjur. The British Library Collection (OR 6724).

Figure 42a, b Miniatures mounted on the front page of volume 65 of Tibetan Selkar (sel dkar) Kanjur. The British Library Collection (OR 6724).

page, some have only circles painted with gold (Figure 41). The pages contain the title written in three text lines and two miniatures showing various representations of the Buddhist Pantheon in different volumes within these pages painted with gold on red in linear style. The only bodies of figures are fully gilded (Figure 42a, b). The upper blue layer of the front page has cut-outs for all elements and serves as a frame for the miniature paintings covered by paper flaps on both side margins. This frame-page is finished with geometric golden ornament resembling a variation of an endless knot (Figure 41). A variety Indigo, Gold, and Human Blood 89

Figure 43 The curtains stitched together on the bottom corners and attached on the top with three leather clips on the verso of the front page of Tibetan Selkar (sel dkar) Kanjur. The British Library Collection (OR 6724).

of such geometric ornaments can be seen on almost all gold manuscripts under study and also a few others written in black ink. A lower, black layer of paper is exposed in the central part of the page. The title is written on the black inked layer there, and above the title three curtains are attached that protect the whole page. The curtains are stitched together on the bottom corners and attached on the top with three leather clips (Figure 43). The rest of the book is written with black ink on raw paper, with four to eight text lines per page composed within a rectangular frame, and side margins sketched with red ink. This also shows that the black-grounded technique was used as a superior aesthetic at least for front pages or the beginnings of texts accompanied by miniatures. Another element reserved for front pages of Tibetan gold manuscripts is the painted border, which usually serves to communicate to the reader the beginning of a significant division in the text, such as a chapter or the start of a new text in a book containing several works. In the Selkar Kanjur it was a geometric ornament framing the text of the title. These borders may consist of miniatures and decorative motifs but sometimes, especially in Western Tibetan manuscripts, can contain a narrative type of painting composed within a border on the front page. Two decorated parts of the Tibetan manuscript of the Perfection of Wisdom Sūtra (Prajñāpāramitā, Sherabkyi Paröltu Chinpa) from the Herbert Johnson Museum of Art at Cornell University testify to such visual conventions for illustrating this Sūtra found in Western Tibet during the fifteenth century (Figures 36 and 44). Numerous female figures are depicted with typical hair ornaments and in Western Tibetan costumes. The color palette and stylistic elements of the paintings, such as the red-line drawings and outlines, 90 chapter 4

proportions of the figures, modes of representing local architecture, chariots with bolts of fabric, horses, and floral ornaments are comparable to the mural paintings in the temples of Tsaparang in Western Tibet. This folio is closely comparable to the published fifteenth-century Western Tibetan manuscript from the R.R.E. Collection (Figures 45 and 46).13 The design and structure of the frontispiece page are very similar to those of the Johnson Museum folio. The upper register of the Johnson Museum frontispiece shows five figures seated on lotuses, the three lines of text occupy the central section of the page, two square groups of the miniatures depicting architecture are positioned on the right and left of the page, and the lower register shows processions on chariots.

Figure 44 Frontispiece of the Tibetan manuscript of the Prajñāpāramitā, Sherabkyi Paröltu Chinpa. Western Tibet, Fifteenth century. Collection of Herbert F. Johnson Museum of Art, Cornell University. Acquired through the George and Mary Rockwell Fund. Prajñāpāramitā Cover and Frontispiece (2006.028a,b).

Figure 45 Verso of the inner cover of the Tibetan manuscript of the Prajñāpāramitā, Sherabkyi Paröltu Chinpa. Western Tibet, Fifteenth century. The R.R.E. Collection, Winterthur (ET77). © Richard Ernst.

13 Pal 2003: 158–159, 290. Indigo, Gold, and Human Blood 91

Figure 46 Frontispiece of the Tibetan manuscript of the Prajñāpāramitā, Sherabkyi Paröltu Chinpa. Western Tibet, Fifteenth century. The R.R.E. Collection, Winterthur (ET77). © Richard Ernst.

The R.R.E. collection folio, even more elaborated, is inscribed and identifies the scenes and the main characters of the illustrations. This story is also set within a generously decorated border. However, the most typical outline of Tibetan frontispieces have two to five lines of text accompanied by two miniatures on both sides and are framed. Sometimes, there are three miniatures painted on the front page, two on the sides and a third in the middle. The early example of two-miniature composition which is not framed is illuminated fourteenth-century manuscript from the Tibetan Rare Book Collection, Asian Division, The Library of Congress, Washington DC (Figure 47). The example of similar but much later dated composition is illuminated manuscript with text of Diamond Sutra (MAP 4323) written in gold ink on a black background from the Museum of Asia and Pacific in Warsaw. This manuscript was probably produced in nineteenth-century Mongolia. The five lines of text and two miniatures on both sides within red frame measuring 6 × 9.2 cm are executed on its front page (Figure 48). On the left side is a representation of Buddha Shakyamuni [Shakya Thub pa] (the historical Buddha), while on the right is a representation of a student listen- ing to his teachings in a slight bow. The rest of the text is executed in seven lines of text centrally in the double-line frame. Surface-colored leaves in black, some leaves have dark-blue margins, sometimes with a touch of grey and brown. Foliation is given on recto pages in Tibetan, on the left side outside the rectangular frame, along the shorter side. At the last page the four Great Kings defenders teachings of the Buddha, the guardians of the world are painted in linear style. Another example of gold and silver calligraphy composed in between two miniatures showing four stupas each in a red frame is Tibetan manuscript 4 from the Columbia University Library belonged to Bon tradition. This shows that this standard pattern for composing front pages is often independent of tradition and type of text (Figure 49). 92 chapter 4

Figure 47 Illuminated fourteenth-century manuscript in gold ink on a black background from Western Tibet.Tibetan Rare Book Collection, Asian Division, The Library of Congress, Washington DC.

Figure 48 Illuminated manuscript ’phags pa shes rab kyi pha rol tu phyin pa rdo rje gcod pa written in gold ink on a black background. The Asia and Pacific Museum in Warsaw, Poland (MAP 4323).

Typical front pages with three miniatures example the manuscript with the same text from the Pomeranian Library in Poland (Figures 50). These are more elaborated with characteristic geometrical ornament described earlier. In Tibetan manuscript culture it is common for initial letters to contain painted miniature scenes, as was seen in Western Tibetan manuscripts. However, such elaborate ornamental decoration are often in other regions replaced by two miniatures and the title written in gold or executed in the raised-gold technique surrounded by borders decorated with a geometric Indigo, Gold, and Human Blood 93

Figure 49 Front page from manuscript describing the life of Gshen rab mi bo the founder of the Bon religious tradition of Tibet. Five lines of text written in gold and silver ink in red rectangular frame and two miniatures representing four stupas on both sides. Rare Book and Manuscript Library, Columbia University in the City of New York (Tibetan manuscript 4).

Figure 50 Illuminated manuscript with text from the Diamond Sutra written in gold ink on a black background from the collection of The Pomeranian Library in Szczecin, Poland (DSB 1).

ornament and sometimes precious stones mounted on the page. Ornamented signs for the beginnings of texts are typical in Tibetan book culture in all regions. They, however, are not elaborated in the same way as initial letters in medieval European manuscripts. Typically, Tibetan books have fore-edges painted with one color, such as yellow, orange, sometimes red, or purple. Often paints based on arsenic or mercury were used to protect the fore-edges. They were allocated to protect the inside of a book from insects. However, a significant group of Tibetan books produced in the Beijing area display lavishly and well executed fore-edges paintings. The Manuscript Kanjur of Berlin acquired in 1889 by V. Brandt and 94 chapter 4 preserved in the Berlin State Library has beautifully painted fore-edges on every volume. These fore edges were painted with lotus flowers and auspicious symbols against a yellow background on the long edges, with a precious pearl in flames and Tibetan text in the halo on the lotus flower on the short edges. Handwritten with black and red ink, this copy was sponsored by Chinese emperors who patronized Tibetan Buddhist teachers. Taking into account rank of the sponsors, it may thus be assumed that highly skilled craftsmen and artists were chosen for this project. Painting of the fore-edges is characteristic of all early editions of Tibetan Peking Kanjur whether written or printed.14 This painting also suggests that these volumes were never overlapped with cotton cloth or attached with textile labels, which are standard elements of Tibetan pothi binding. The fore-edge paintings suggest a Chinese origin. These volumes were probably also stored in the Chinese way, with the titles and numbers of volumes visible directly on the fore-edges.15 Additionally, there is a practical aspect of such a decoration. If the loose leaves of pothi are not in the proper order, the painting’s bad fit is immediately visible on the volume edges. The coloring of the painted fore-edges in all volumes of the same set is quite similar in color and style.16 Both the London Kanjur and Manuscript Kanjur of Berlin, discussed above, represent different regions of Tibet and China, and this is clearly seen in their style of bookmaking and decoration (Appendix 4). Doodle drawings are rare and difficult to find in Tibetan books. Interestingly, however, they appear quite often in Dunhuang manuscripts. Bon manuscripts preserved at Columbia University also bear some. Doodles are often line drawings of mythical animals as we can observe in Tibetan bound manuscript from the collection of the National Museum in Warsaw, Poland (Figure 51).

14 The Berlin Manuscript Kanjur was copied directly from the Yongle/Wanli edition printed in Beijing, and for a long time it has been the only available copy of this oldest printed edition of the Tibetan Kanjur. These printed editions contain Yongle Emperor’s endorsement as a forward to every volume. 15 This type of fore-edge painting can be observed in some Chinese and Mongolian Buddhist books; see for example jīn guāng míng zui shèng wáng jíng shí juàn preserved in the National Library of China in Beijing: (National Library of China) 2006: 105. 16 This refers to any decoration on the fore-edges of the leaves of a book. Indigo, Gold, and Human Blood 95

Figure 51 Linear sketch of a dragon in Tibetan bound manuscript. The National Museum in Warsaw, Poland (SKAZ 1813).

Calligraphy

Writing is an important aspect of Tibetan culture and it comes in a great variety of styles. It is why it has placed a high value on the mastery of calligraphic skills. These styles are often specific to particular social functions, but there is a fundamental distinction between two scripts. On the one hand there is a script with horizontal lines along the tops of many letters, which creates upper line of text, known as dbu can, or ‘headed’, and second there is a script without these lines, known as dbu med, or ‘headless’. In the latter there are numerous different styles including a wide variety of a cursive style for handwriting.17 Despite dbu can script being characteristic in the Tibetan printing heritage and examples of golden manuscripts written in elegant variants of dbu med, among Tibetan scripts dbu can seems preferred for deluxe editions. The quality and techniques of writing differ to a great extent, however, depending on the artistic skills of the scribes. The frontispiece from the Johnson Museum, the frontispiece from the R.R.E. collection, and the Sel dkar Kanjur from the British Library are distinguished by their beautiful calligraphy. Some other books begin with steady writing that later decreases in quality, or sometimes sections of a book are written sloppily. A less skilled scribe may have written

17 For more on Tibetan scripts see: van Schaik 2012. 96 chapter 4 such sections. This can be seen in the Diamond Sutra written in gold from the Asia and Pacific Museum in Warsaw (Figure 48).

Page Layout

In the layout of the pages there is variety as well in the number of text lines, the type and size of frames and margins, the placement of miniatures, the decorative elements, the placement of commentary, if any, in relation to the placement of the main text, and the presence of notes by author, scribe, or readers (called ‘marginalia’ in Western books). In Tibetan book culture text is usually set within a rectangular frame or between side-margins. Those frames or side margins are often drawn with red ink or the same type of ink as that used for writing the text in the corpus of manuscript. Frames in printed books are carved together with text on the same wooden block. General page layout and number of text lines is first of all dependent on the size and format of the manuscript. However, the habits of the scribe and the original text or book used as a master copy also influenced the final result. A good example is the Berlin Kanjur copied from the printed edition of the Yongle/Wanli Kanjur. The text of the Berlin Kanjur is skilfully handwritten and laid out within a 55 × 18.8 cm rectangular frame sketched with a black line 0.3 cm thick inside a thin red line. The size and layout, however, are close to those of the printed version. In the Berlin Kanjur the calligraphy is distinguished by longitudinal proportions and long letters outstretched at the base, such as ka, ta, and na. Distances between upper text lines are exactly 1.8 cm. Scribal text- base lines are sketched with a lead stylus or other slightly sharp tool (one line at the text head or text is inscribed in between two lines). Volumes here, as with Yongle/Wanli printed volumes, are characterized by eight lines of text on a page. Text begins with four lines on the recto side of the first page. The distance between upper text lines is 4 cm, and the first and third lines are written in red. The verso side of the first page bears five lines of text. The distance between the upper lines is 3.2 cm, and the second and fourth lines are written in red. The number of text lines would reach eight on the recto side of the third leaf, with the third and sixth lines in red and 1.8–1.9 cm distances between upper lines of text. In the manuscript frontispiece written with gold ink on black-inked paper from the Johnson Museum, Cornell University, the title is written in two lines on the verso side of the book inner cover; three lines of text are on the recto Indigo, Gold, and Human Blood 97

side of the frontispiece, and eight text lines are on the verso page of the frontispiece. This suggests that the rest of the book may be written in eight lines of text per page if preserved. Turning to another example, the fifteenth century (?) Lotus Sutra, Tibetan ms. 6 from the Columbia University Library, written in golden and silver ink on a black-inked background has six text lines on the first page (recto) and seven text lines on the second page (verso). Text is written in square, relatively wide dbu can script (Figures 52 and 53). In examined manuscripts the number of text lines usually ranges from four to nine, depending on the book format and size of the letters. The rule of increasing the number of verses of the text on the first pages was repeated in most of the studied manuscripts. As in medieval European manuscripts, rubric is also a typical element of Tibetan manuscripts. Red or another color distinguishes sections of text and serves as instructional guides to the reader. Sometimes the highlighting of more important words, lines, or text passages was achieved in Tibetan manuscripts by an alternate use of gold, silver, red, or another color of ink, or by the letters being different in size than those in the rest of book. The Tibetan manuscript Tib. Ms. 6 from Columbia University Library features passages of text marked with gold ink among silver written text. At the beginning, each page contains gold alternately with silver text, with the first lines in gold, and this pattern continues through the book until on the last page there is a golden triangle composed of gold-written words on the silver text background (Figure 53).

Figure 52 This manuscript with text from the Lotus Sutra features passages of text marked with gold ink among silver written text. At the beginning, each page contains gold alternately with silver text, with the first lines in gold, and this pattern continues through the book excepting the last page. The Rare Book and Manuscript Library, Columbia University in the City of New York Columbia University (Tibetan manuscript 6). 98 chapter 4

Figure 53 This manuscript with text from the Lotus Sutra features passages of text marked with gold ink among silver written text. The last page features a golden triangle composed of gold-written words on the silver text background. Rare Book and Manuscript Library, Columbia University in the City of New York (Tibetan manuscript 6).

Red rubrics were usually used within manuscripts written in black ink on raw paper, as in the thirteenth century manuscript Mngon par rtogs pa’i rgyan gyi ‘grel bshad theg pa chen po la ’jug pa shes bya ba, LTWA 14459 from the Library of Tibetan Works and Archives in Dharamsala. This manuscript is written in nine lines of dbu med script per page between clearly sketched-in red side margins. Sanskrit words are marked off with red ink while the rest of the manuscript is in black ink. Text is based on hardly visible scribal guidelines possibly created with the use of a straight and even sandalwood stick, which colors the paper yellow when wet, or with the juice of yellow cypress turmeric (skyer shun gi khuwa) mixed with various fragrances also used in Tibetan medicine, such as cypress or sandalwood, as was typical in Tibetan manuscripts written on raw paper. The tools for executing the guidelines and for writing differ. The presence of those lines distinguishes a manuscript from a printed work, in which they are absent (Figure 54). For manuscripts written with gold ink, a sharp tool would be used to scratch text guidelines into the paper before it was colored. This is preparatory step can be observed in many manuscripts; for example, they are found on the Lotus Sutra (Tib. Ms. 6) from the Columbia University Library and the frontispiece from the Herbert Johnson Museum of Art manuscript. Interestingly, another common element of Tibetan manuscript layout is the shape of circles located exactly in the places where holes would have been Indigo, Gold, and Human Blood 99

Figure 54 Yellow scribal guidelines that distinguish a manuscript from a print are visible on this page fragment from the Tibetan Selkar (sel dkar) Kanjur. The British Library Collection (OR 6724, vol. 65).

pierced for stringing pages together in the Indian pustaka—these circles were primary to the layout of the text. First frames, side margins and circles were sketched, then text was inscribed. These circles were found on some of the Tibetan manuscripts from Dunhuang, but also in many others; for instance, on ms. 6 from Columbia University, there are two circular shapes sketched by a sharp tool. Most of the circles were redrawn with red ink, and a few have an additional silver circle inside.

Ink

Ink remains one of the primary elements in book production. Still, even basic descriptions of ink color appeared useful in classifying books. Most ink is black, but we also find red and occasionally yellow in Tibetan books. Black ink is usually made of carbonized plants or soot with animal glue. It, however, can also contain other organic components that can modify its properties according to needs. Various recipes for its preparation include soot from burnt resinous 100 chapter 4 wood or butter lamps, other pigments, or metal as the ink’s main components with a possible mixture of gum, honey, borax, or, for a special occasion, other unusual and extravagant materials. Blood is such a material. One of the common recipes for Tibetan ink includes vegetable oil (mar nag gi methok), resin of pine or fir tree (sgron shing), birch bark, and finally resin of some fruit trees. Soot collected from burning vegetable oil, instead of butter for religious offering, allows for better ink. Soot can be collected from the interiors of stacked, bottomless, clay pots, usually five to seven, after selected components are burned at the bottom of the column they form. Resin of pine or fir tree should be obtained by peeling green bark from the tree, and collecting from it ripe resin with a consistency similar to melted fat. Unripe resin would be too sticky and could be used only as adhesive, not as an ink medium. All components including soot and resin glue should be heated in a metal vessel over a slow fire and while being stirred from time to time. Then the concoction should be churned in a leather sack for some time with equal force to mix it well. It is said that in this way one person can prepare about four pints (bre) per day. If while mixing the blend makes a sound similar to that of matches striking, it is considered to be of standard quality. The longer we mix it, the better it gets. At last the mixture should be dried to make it long lasting. Precious metals such as gold and silver were also often used as inks and pigments. Use of gold for writing the sacred texts is common in many religious traditions. Calligraphy in gold was well known and appreciated as a skill by itself. Gold script on black or blue ground is a Tibetan convention developed in the culture of sacred books. Gold is for wonder, wealth, and opulence. Gold mining was known in Tibet since early times. Tibetans believed that the soul of the Earth should not be bothered and for this reason only gold sand was collected—bigger pieces were left in the ground.18 Natural silver is usually a component of gold deposits. An illuminated manuscript ’phags pa shes rab kyi pha rol tu phyin pa rdo rje gcod pa (MAP 4323) written in gold ink on a black background in the pothi format provides a good example. A sample taken from the ink on this manuscript contains gold and silver (Au = 68.21% mass Ag = 31.79% by mass). Red is for alarm, power, and resolve. In gold manuscripts red is associated with coral but seems to be made of cinnabar and vermillion rather than real coral. Cinnabar is a natural pigment that is often used for ritual purposes— making certain mandala diagrams and painting. When offered in powder form it is considered a symbol of power, owing to its strong red color. Cinnabar

18 Bareja-Starzyńska 1999. Indigo, Gold, and Human Blood 101 symbolizes the realization of the activities of power (dbang gi las). The concept of power is used in the sense of control over one’s own capacities and command of knowledge and wisdom.19 Blood has also been reported to have been used as red ink in Tibet. For instance, as reported by Dorji Wangchuk, Khri srong lde btsan is said to have commissioned a deluxe edition of the Prajñāpāramita translated into Tibetan by Rlangs Khams pa go cha from the original text that he had memorised (which is thus called Rlang kyi thugs ’gyur ma) written with ink made from a mixture of his own blood (mtshal khrag), the blood of his deceased queen, and the milk of a white goat (ra dkar gyi ’o ma) in commemoration of the queen. There is also said to be an edition of the largest Prajñāpāramita scriptures called the ’Bum dmar (Red [Prajñāpāramitā] in 100,000 Lines) written with the ‘nose-blood’ (shangs mtshal) of Śāntarakṣita, Padmasambhava, and Khri srong lde btsan. There is said to be another special edition of a Prajñāpāramita scripture written with special ink made from Khri srong lde btsan’s singed hair (dbu skra’i gzhobs) and indigo (rams), using the milk of a white goat (ra dkar gyi ’o ma) as the binder.20

Writing Tools

Tools of writing and painting have always depended on materials they were made from. Their shape and properties were also adjusted to writing supports they suppose leave a trace on it. One change in technology makes other changes necessary. Sometimes discovery of a new writing tool revolutionized all other habits in copying manuscripts. The Tibetan art of writing, however, does not seem to have experienced ‘ground revolution’ since very early times. Simply good tools made of high quality materials are essential for the art of writing. Scribes usually learnt how to prepare their tools in the same way as writing or painting. This essential knowledge of making a pen and preparing its tip was necessary to become an expert in writing. A bamboo pen was considered to be the best instrument for writing standard Tibetan. Different types of pens, however, were used depending on the Tibetan region and the period of time. Especially one type of bamboo, recognized thanks to the small hollow space in its stem, was favored. The thick and hard, inner part of the bamboo stem was the best material. Heavy when lifted, it grows thick and tall and is resistant to twisting and crooking. It has a yellow, smooth, and shiny bark and is considered to be the best for writing scripts and documents with golden,

19 Dagyab 1995: 56–57. 20 Wangchuk 2010 (not published). 102 chapter 4 silver, and cinnabar inks. To make a bamboo pen even stronger the bamboo stick was smoked for a long time to make it hard as a stone. Roasting plant material in a pan with old cow’s marrow can also strengthen the material. The pen (a’jang smyug) from eastern Tibet is black and partially hollow. A fountain pen, on the other hand, is good for large-size religious texts. It is also said that bamboo from the south or one that grows in valleys is acceptable for this type of pen, but this information can be found only in the guidelines, written by earlier scholars. These days there are hardly few artists who can work precisely following these rules and the knowledge of different bamboo types is less and less common. A bamboo stick is evenly cut on both sides at one end to make a pen. The length of a pen should not be more than the span between extended thumb and middle finger (tho) and not shorter than a thumb. It should be comfortable to hold, with the tip measuring one-and-a-half times a unit equal to the width of a finger (sor), sharpened from both sides by a knife. Two-thirds of the length of a pen should be for holding and one-third should be prepared as the writing tip. A wider and longer slit should be made to write faster and a small and narrow slit will help to write finer letters. The tip should not be too thin and the finishing cut should be slanted. To write dots used between syllables in Tibetan orthography (tseg), the tip should be cut especially even. To write it spherically, the tip should be cut into two. If you cut it unilaterally, it will influence the quality of writing, and if you cut it bilaterally it will harden the pen’s tip. The blade of the knife for cutting a bamboo should be flexible and hard enough to stay sharp for a long time. It is why the best steel was used to obtain the desired shape and quality. Tibetans used melted butter and honey for sharpening and polishing. The best knife for preparing pens should have a tip similar to a chisel. The knife for cutting paper should look like a rib. It should be curved, and flattened with slightly convex sides. A well-balanced knife can keep its sharpness for a longer time. The fibers of myrobalan or stones from a pond were used for sharpening the pen knife. A knife should be sharpened by moving the whetstone in one direction only, which will give it instant sharpness. If it is sharpened with circular movements, the blade of the knife will not be broken easily.

Paper

Tibetans developed and retained specific skills for creating the blue or black ground on which gold and silver calligraphic texts are written. Good quality Indigo, Gold, and Human Blood 103 materials were used for those special editions, such as specifically prepared paper, ink, and paints. The color in gold manuscripts, the same as in other unique types of Himalayan painting, is used to invoke mood and emotion. Black is for caution, fear, and protection. Regarding to technology blue paper can be prepared by dyeing—dipping the paper into dye, by painting it with a brush, or a combination of both methods. Some manuscripts were also stamped with felt, then ink was applied with a brush only within a central frame as a background for text. This technique can be exampled by manuscripts 1–3 and 5 from the Columbia University Library. The most desirable blue dye, indigo, was, besides Tibetan books, also used for dyeing velum, a medium used in production of sacred books in the west and an exceptional copy of the famous ‘Blue Qur’an’ dated to ninth century. Indigo is a vegetable extract that is relatively stable and used in both oil- and water-based media. This organic colorant could have come from India, where it was first mentioned being used in manuscripts in the fourth century BCE. In many cases indigo was used instead of blue inorganic pigment such as azurite, as in an illuminated Tholing Manuscript “Prajnaparamita with Devotees” in the collection of LACMA (M.81.90.6), believed to have been produced in western Tibet in the first half of the eleventh century. It is doubtful whether and to what extent other precious materials such as turquoise were actually used for writing. The value of lapis lazuli or indigo was the same as the value of gold. Legends say that lapis lazuli covers the ground in the Buddhist Paradise or sometimes that Buddhist Paradise is made of lapis lazuli. It, however, is hardly possible that we will find lapis lazuli in Tibetan books. Maybe only as a legend in written sources, such as in the colophon of the Diamond Sutra from the Pomeranian Library in Szczecin, Poland which says: “Among eighty four thousand collections of Dharma this . . . is written by the power of the virtue of gold ink on good quality lapis lazuli paper to ensure long life to those who support Buddha’s teaching.”21 The blue dye can also be prepared with the use of myrobalan juice, the medicinal plant thar nu, sour wheat beer, purified black vitriol (nag mtsur), and borax (tsa la). Basically it is made by applying to paper an extract of Tibetan wheat beer, ground cardamom (za ti), and burned and ground white cowry shell (mdron bu), and then sprinkling it with myrobalan juice. There are various recipes for producing dark and blue paper. These involve pressing several layers of paper together using a paste made of boiled wheat flour, then applying special ink, paint, or dye to this surface and smoothing it with a stone. Special measures were required to prepare paper for gold

21 Diamentowa Sutry [Diamond Sutra] translated by Marek Mejor 2004. 104 chapter 4 manuscripts. But besides these, the main feature possible to observe was the superior quality that distinguished the paper used. Practically every type of paper could have been used, but different types required specific processing. Chinese soft paper was usually made of paper mulberry, bamboo, straw, or grasses, and, in most cases, Chinese papers were dyed with indigo; by contrast, Tibetan paper was typically made of Daphne/Edgeworthia fibers, and the surface of Tibetan papers were often painted with a brush. The selection of raw materials appears to have no relationship to techniques of gold-manuscript production. Rather raw materials are mainly to be related to the region of their distribution; as such, they provide characteristics that can help in the identification of the regional origin of manuscripts. Based on Tibetan manuscripts preserved in the sealed cave in Dunhuang, Central Asia there are the three main types of paper from the British Library that I could roughly link to geographical regions.22 In the Dunhuang sample, only six manuscripts out of a total of seventy-six that were examined were composed of the Daphne sp. type of paper; including two manuscripts sent to Central Asia from Tibet, and dated to the early ninth century. Among these six samples five were woven papers made on a textile sieve and one represented laid type of paper characterized by eighteen laid lines in 3 cm. It is highly significant that the only manuscript in the sample containing clear textual evidence of having been made in Central Tibet during the Tibetan imperial period (IOL Tib J 1459) was found to have been produced from Daphne or Edgeworthia plant fibers. The only other Dunhuang manuscript made from these fibers is a fragmentary pothi book of the Samdhinirmocana sūtra, written in the archaic ‘square’ style; this too may have been brought from Central Tibet. Conversely, none of the Tibetan manuscripts known to have been made in Dunhuang was constructed from Thymelaeaceae fibers. Another letter, found in Miran (Or. 15000/513) was also found to be composed of Daphne or Edgeworthia plant fibers, this time mixed with paper mulberry fibers. This letter may well have come from Central Tibet, as it contains a seal which is partly effaced but contains the word ‘palace’ (pho brang), an imperial administrative center of which none are known to have existed in Central Asia. In addition, the three twelfth-to-thirteenth-century manuscripts from Khara Khoto are also of the Daphne or Edgeworthia type, and may have been brought to Khara Khoto by Tibetan monks, who are known to have been active at the Tangut court. As far as we know, Thymelaeaceae family plants such as Daphne or Edgeworthia sp. were not used in Central Asia or China for making paper. This supports the view that Tibetans were able to make paper

22 This work was conducted in collaboration with Sam van Schaik. See: Helman-Ważny and van Schaik 2013. Indigo, Gold, and Human Blood 105

Figure 55 Edgeworthia/Daphne sp. fibers stained olive-grey/greenish with Herzberg at 50× magnification. Notice the comparatively great fiber thickness (up to 25 µm), the varying wall thickness and the fiber ends, the many flattened places and the dislocations. The lumen often narrows or wholly disappears. A significant amount of parenchyma cells and fragments of parenchyma tissue present in the pulp is clearly visible in the above photograph. Tibetan Prajñāpāramitā volume preserved in the Bicher temple, Dolpo, Nepal roughly dated to the twelfth and thirteenth centuries by the characteristics of the handwriting style.

at least as early as the first half of the ninth century. The raw material for this technology was the Thymelaeaceae family plants, which widely occur in the Himalayas. Another manuscript from Western Tibet and Nepal containing Daphne or Edgeworthia fibers is the sample from an isolated sheet, from a Tibetan Prajñāpāramitā volume preserved in the Bicher temple, Dolpo, Nepal. The characteristics of the handwriting style of this manuscript date it roughly to the twelfth-to-thirteenth-centuries. A buff-colored sheet of paper was provided by Amy Heller and examined for fiber composition (Figures 55 and 56). Handmade, glued at least of two layers, good quality paper with half- glossy polished surface has been observed.23 The majority of examined books

23 Light microscopy observation in transmitted and polarized light: The sample contains fibers, parenchyma cells and singular sclereid. In Herzberg stain the fibers were uniformly olive-grey/sometimes greenish with slightly pink colored lumen. Features such as 106 chapter 4

Figure 56 Edgeworthia/Daphne sp. fiber olive-grey/greenish with slightly pink colored lumen at 200× magnification observed in polarized light. Notice the comparatively great fiber thickness (up to 25 µm), the varying wall thickness, the many flattened places and the dislocations. The broad portions were not associated with irregular cross-markings, which occurred densely and were often visible as a ‘v’ or zigzag shape. The lumen often narrows or wholly disappears. Tibetan Prajñāpāramitā volume preserved in the Bicher temple, Dolpo, Nepal roughly dated to the twelfth and thirteenth centuries by the characteristics of the handwriting style.

Figure 57 Daphne/Edgeworthia sp. fibers stained olive-grey with Herzberg at 150× magnification observed in polarized light. In the middle the fragment of a long, thick-walled, pointed sclereid fiber stained with yellow. Tibetan Selkar Kanjur. Indigo, Gold, and Human Blood 107

Figure 58 Daphne/Edgeworthia sp. fibers stained olive-grey with Herzberg at 600 × magnification observed in polarized light. Tibetan Selkar Kanjur. from Central and Western Tibet and Nepal and dated after the thirteenth century were composed of Daphne and Edgeworthia sp. The Selkar Kanjur from eighteenth-century Central Tibet is another example (Figures 57 and 58). However, there are also Prajñāpāramitā manuscripts, again from the early ninth century, and produced in the Tibetan cultural area, composed of Broussonetia sp. (Paper mulberry). Thus it seems that both Daphne/Edgeworthia sp. and Broussonetia sp. were both used for papermaking in early Tibet. Given the patterns of plant distribution in Tibet, it is likely that the former was the primary type of paper used in Central (and perhaps Western Tibet) and the latter the primary type in Eastern Tibet. Further research will be needed to confirm this, however. Given the prevalence Broussonetia in Eastern Tibet, it is likely that these manuscripts were created in the eastern regions of the Tibetan empire, where this type of paper is still made. Note that one manuscript (IOL Tib J 570), a tantric ritual probably written in the tenth century, is also composed of Broussonetia paper, but made on a movable sieve rather than a woven

irregularities in fiber walls and lumen, and broad central portions along the fiber length definitely confirm that this paper was made from one of the Thymelaeaceae family plants, presumably species of Edgeworthia or Daphne. The comparatively great fiber thickness (up to 25 µm), the varying wall thickness, and fiber ends with many flattened places and dislocations are noticeable. The lumen often narrows or wholly disappears; varies but is mostly small and narrow and is often indistinct, especially near the fiber ends. The broad portions were not associated with cross-markings, which occurred densely and were often visible as a ‘v’ or zigzag shape in polarized light. Dislocations are numerous, but less sharply delineated. Fibers observed at a lower magnification were lying at a slightly rigid and straight form shape on the slide. 108 chapter 4 mold; this reminds us that paper mulberry was also used in papermaking in China and Central Asia. Another example of paper mulberry can be found in Tibetan/Tangut manuscript Dx 178 from the Institute of Oriental Manuscripts, Russian Academy of Sciences, Sankt Petersburg.24 Vertical scroll composed of eight panels on paper folded in two is not typical for Tibetan book culture. Paper is handmade, very thin and good quality, and soft (not sized) what suggest purpose selection of this type for a particular manuscript. Yellow dye and very good quality of materials used suggest importance of this manuscript. Laid regular structure characterized by twenty-one laid lines in 3 cm indicates that paper was made with a movable type of papermaking mold equipped with a bamboo sieve. This paper contained Broussonetia sp. (Paper mulberry) fibers (Figures 59 and 60). This type of paper, though primarily composed of Broussonetia sp. (Paper mulberry) paper, was sometimes also composed of Morus sp. (Mulberry); both derived from living plants. However, in most cases, these two species are almost impossible to distinguish through microscopy. The majority of early manuscripts on this type of paper were created on a textile sieve or laid patchy. Paper mulberry (Broussonetia papyrifera, in Chinese chushu, goushu, gu) grows in vast areas of southern and northern China; however, it requires specific conditions. Jia Sixie (473–554), an agronomist, describes paper mulberry cultivation for paper-making. He suggests that paper mulberry trees do best when planted in good fields in valleys that have a river or brook nearby. Mulberry has also been cultivated in China since early times. It probably emerged in China in the third century and was used for papermaking since the Wei (220–265) and Jin (265–420) dynasties—before it was used to raise silkworms. The main mulberry species includes Morus alba (zhensang), Morus acidosa (xiaoye- sang), and Morus mongolia (yasang). These plants are considered to be the best material for creating high quality bark paper, which supposedly originated in the Han Dynasty and flourished since the Tang Dynasty. This is why it is not surprising that many Tibetan manuscripts were also produced on this type of paper. It is very possible that since bark paper made of paper mulberry was not produced in the desert area, or some high-altitude regions of western Tibet these manuscripts may have been created in Eastern parts of Tibet and China. This type of paper was known and popular enough to be traded and moved to desert areas from elsewhere. However, at some point paper mulberry was also cultivated in oasis of Dunhuang and Turfan, and could be used for limited paper production as well.

24 For more information about manuscript Dx 178 see: Zorin 2013. Indigo, Gold, and Human Blood 109

Figure 59 Broussonetia sp. fibers stained reddish with Herzberg at 50× magnification. The paper composing the vertical scroll composed of eight panels on paper folded in two. The collection of the Institute for Oriental Manuscripts in St Petersburg, Dx 178.

Figure 60 Broussonetia sp. fibers stained reddish with Herzberg in 200× magnification. The paper composing the vertical scroll composed of eight panels on paper folded in two. The collection of the Institute for Oriental Manuscripts in St Petersburg, Dx 178. 110 chapter 4

On the other hand, it is very clear from the Dunhuang samples that the majority of Tibetan manuscripts produced in Dunhuang and other Central Asian sites such as Turfan and Miran are composed of rag paper primarily composed of Boehmeria nivea (Ramie) and Cannabis sp. (Hemp), sometimes with the addition of other fibers such as jute or Broussonetia sp. (Paper mulberry) (Appendix 1). In the majority of examined samples this type of paper has laid lines, showing that it was made on a movable sieve constructed of bamboo, reed, or grass. It is significant that my research on these rag-paper manuscripts, based on textual, codicological and paleographical evidence, shows that there is no reason to believe that these manuscripts were produced in Tibet. However, because this type of paper is made according to the oldest known technology in China, could also have been produced in other parts of China. The presence of paper-mulberry fibers not previously used for textile production might suggest a shortage of rags in the area and the need for technological innovation in paper production. It is interesting that Trier reported in 1972 that the oldest paper samples that he examined in Nepal are dated to the thirteenth to fifteenth centuries and were made of ramie and hemp, unlike those dated to later times. Additionally, Trier’s group of manuscripts represent a variety of formats, such as scroll, pothi, booklets, and concertina; such a variety of formats suggests the influence of a multi-cultural environment. His preliminary interpretation that the papers originated in India should probably be opened for discussion again in the context of the present research.25 Analysis of the Johnson Museum manuscript indicates the paper was handmade with paper-mulberry fibers. Although examples of such mulberry paper can be found in Eastern Tibet and earlier at Dunhuang, it is usually associated with Chinese paper and with Tibetan books produced in China. This implies that very fine paper was imported for the production of this manuscript. Conversely, the Tibetan manuscript from the R.R.E. Collection was produced on a specific paper made of the Stellera chamaejasme species (re lcag pa) and typical in Central and Western Tibet (Figure 61). This plant grows in relatively dry conditions in Central Asia, Iran, parts of Tibet, and along the Himalayan range and can be used to pointedly mark a manuscript’s origin. On the other hand, we did not find any evidence of root paper made of Stellera chamaejasme species (re lcag pa in Tibetan) in any Dunhuang manuscripts. Though it is possible that further analysis of more samples from early periods may reveal the existence of this paper, it is more likely that it was not in widespread use given the relative ease of paper production from Daphne or Edgeworthia sp. As described in chapter 6, root paper was probably only made where other plant sources were unavailable, such as at particularly high altitudes.

25 Trier 1972: 57. Indigo, Gold, and Human Blood 111

Figure 61 Stellera chamaejasme (re lcag pa) fibers with wide-irregular lumen and narrow cell-walls at 600× magnification observed in polarized light found in the Tibetan manuscript of Prajñāpāramitā, Sherabkyi Paröltu Chinpa from the R.R.E. Collection.

The earliest confirmed manuscript which contains the fibers of this plant is dated to the tenth century.26 This manuscript, discovered by Dr. Pasang Wangdu, originates from Central Tibet (the area of Samye monastery). Furthermore the other manuscript is M81.90.6, an illuminated manuscript from Tholing in the collection of the Istituto Italiano per l’Africa e l’Oriente (IsIAO), dated to the eleventh century. I examined two samples of paper (1329 F and 1329 E) from this manuscript provided by Amy Heller for fiber composition. Both samples show the presence of two types of the Thymelaeaceae family fibers: Edgeworthia sp. or Daphne sp. mixed with Stellera chamaejasme fibers (Figures 62 and 63).27 Another early example of pure Stellera paper

26 It was dated by the Poznań Radiocarbon Laboratory in Poland. See: Goslar 2012 (unpublished report). This C14 dating roughly coincides with that based on the people mentioned in the colophon (Pasang Wangdu forthcoming). 27 Fibers of type 1 were generally narrower and more rigid. The broad portions were not associated with cross-markings, which occurred densely and were often visible as a ‘v’ or zigzag shape in polarized light. Dislocations and irregularities in the fiber thick-walls and narrow-lumens were clearly visible. This type can be species Edgeworthia sp. or Daphne sp. both belong to the Thymelaeaceae family. It was very difficult to tell these two species apart, but I could at least distinguish them from the second type by observing the shape and size of fibers. Both species mentioned above were the basic materials in papers from the foot of the Himalayas. Fibers of type 2 were wider in width, flat, with a very wide lumen, and narrow fiber walls. Ribbon-like fiber placing and very significant irregularities within the fiber length allow us to identify this type as Stellera chamaejasme fibers, which are distinctive in the fiber examination of historic papers, and can be clearly differentiated from Daphne and Edgeworthia despite the fact that all of these plants belong to the Thymelaeaceae family. 112 chapter 4

Figure 62 Edgeworthia/Daphne sp. fibers with Stellera chameajasme fibers olive-grey with wide lumen at 600× magnification observed in polarized light in sample of paper provided by Amy Heller from Tibetan manuscript preserved in the collection of Istituto Italiano per l’Africa e l’Oriente (IsIAO), sample 1329E.

Figure 63 Edgeworthia/Daphne sp. fibers with Stellera chameajasme fibers olive-grey with wide lumen at 600× magnification observed in polarized light in sample of paper provided by Amy Heller from Tibetan manuscript preserved in the collection of Istituto Italiano per l’Africa e l’Oriente (IsIAO), sample 1329F.

that I could confirm with fiber analysis is a thirteenth-century dated manuscript from the British Library. These root fibers create a very specific soft type of paper, which is considered to be of lower quality than bark paper of Daphne or Edgeworthia sp. itself. However, if mixed with Daphne or Edgeworthia sp. fibers, it changes the properties of the paper. Daphne or Edgeworthia sp. fibers give the paper its strength and a gentle, slightly glossy surface. Stellera fibers are responsible for the paper’s characteristic softness and high absorbency in the final stage— features that make this paper more suitable for printing. The roots are especially difficult to harvest, which indicates a serious limitation on the quantity of paper that can be produced from it. This is probably why it was often mixed with other fibers. Indigo, Gold, and Human Blood 113

Figure 64 Daphne/Edgeworthia sp. fibers and Stellera sp. vessels stained olive-grey with Herzberg at 200× magnification. The sample is from one of the prayer pages of the Prajñāpāramitā volumes probably dedicated by local people to the Bicher sanctuary, roughly dated to the fifteenth and sixteenth centuries.

Another example of paper made of Daphne paper mixed with Stellera sp. is the sample from one of the prayer pages of the Prajnaparamita volumes that was probably dedicated by local people to the Bicher sanctuary, and is roughly dated to the fifteenth and sixteenth centuries (Figures 64 and 65).28 This

28 Light microscopy observation in transmitted and polarized light: The sample is composed of fibers and vessels, a large amount of parenchyma, and thick-walled sclereids. In a Herzberg stain: 1) the most thin and straight fibers were uniformly olive-grey/greenish with slightly pink colored lumen; 2) parenchyma is stained with grey-bluish 3) a few singular thick-walled sclereids stained yellow that is often associated with Daphne sp.; and 4) a few much wider vessel elements stained also with olive-grey. Irregularities in fiber walls and lumen, and broad central portions along the fiber length confirm that most of the fibers in the examined paper were derived from one of the Thymelaeaceae-family plants, presumably a species of Daphne or Edgeworthia. In these types of fibers, the lumen often narrows or wholly disappears, the lumen varies but it is mostly small and narrow and often indistinct, especially near the fiber ends. Dislocations are numerous, but less sharply delineated. The broad portions were not associated with cross-markings, which occurred densely and were often visible as a ‘v’ or zigzag shape in polarized light. Large 114 chapter 4

Figure 65 Daphne/Edgeworthia sp. fibers stained olive-grey with Herzberg at 200 × magnification observed in polarized light. The middle the fragment of a long, thick-walled, pointed sclereid fiber stained with yellow. The sample is from one of the prayer pages of the Prajñāpāramitā volumes probably dedicated by local people for the Bicher sanctuary, roughly dated to the fifteenth and sixteenth centuries.

probably can be considered a ‘local’ commission, and this explains the use of ‘local’ paper of the western Himalayas. A blue, indigo-dyed sheet of paper

amount of parenchyma is typical for wildly growing Daphne sp. The fiber-like thick-wall sclereids are also more often associated with Daphne sp. This may suggest that this plant was used in this case. The other type represented by a very small addition to the examined sample also belonged to the Thymelaeaceae family. Cells with a flat, ribbon-like shape and having a very wide lumen and narrow fiber walls of varying thickness were identified as vessels of Stellera sp. This species was used for making paper only in the highest places in the world, where practically nothing else grows. It is very rare to find this material in books, and all samples identified until now were found in manuscripts from Central and Western Tibet. If used as main component, these root fibers create a very specific, soft type of paper, which is considered to be of lower quality than bast paper. The roots are especially difficult to harvest, which indicates a serious limitation on the quantity of paper that can be produced. In this sample Stellera sp. fibers constitute only a small quantity of the overall mix. Indigo, Gold, and Human Blood 115 glued of two layers written in gold was provided by Amy Heller and examined for fiber composition. The paper was most likely glued into layers with starch paste. To conclude, I hope that this study has added complexity to the picture of ‘Tibetan paper’ and its origins, by showing that a variety of methods and materials were available to Tibetans during their imperial period and later for creating manuscripts that were both more and less elaborate. If we can begin to speak of a specifically ‘Tibetan paper’ during the early period, it is of a paper composed of Daphne or Edgeworthia sp., Stellera chameajasme or Broussonetia sp., and paper that is made on a woven mold—a technology that was maintained through nowadays in Tibet. By contrast, paper mulberry is more closely associated with China. chapter 5 Tibetan Woodblock Printing Culture

Origin of New Technology and the Early Tibetan Woodblock Prints

It is generally agreed that woodblock printing developed in Asia several centuries prior to its European use.1 Xylograph printed books have the same form as handwritten books, but the text, its frame, and possible woodcuts are printed from the same engraved wood plates by using black or red ink. This is the type of book that is commonly taken to represent Tibetan publishing (printing) culture. The invention of printing can be directly associated with Buddhism and the need to reproduce religious texts and simple images of the Buddha. This demand greatly influenced the development of printing technology. Text reproduction began with stamping and rubbing, which led directly to block printing. This was achieved through a gradual process, when many manuscripts printed or stamped with short dhāraṇī texts or with the Buddha’s image appeared in Chinese Central Asia. At that time, rich monasteries or wealthy lay persons commissioned books for gaining spiritual merit. Craftsmen had already managed to bring woodblock art in the service of Buddhism to a very high level of artistic development; soon after, they would also perfect the techniques of carving necessary for using movable wooden type, which requires even greater precision than woodblock printing.2 It is significant that Tibetans never adopted movable type, even at the time when Christian missionaries,

1 In Europe the printing of images on cloth advanced to the printing of images on paper using woodcuts in the beginning of the fifteenth century. The use of the same process to print texts of substantial length with images in xylographic books in Europe occurred only after the development of movable type in the 1450s. However, woodcut printing was usually used exclusively for illustrations rather than for printing entire books. 2 Wooden movable types were invented somewhere on the Silk Roads by Uighurs around 1300; ceramic movable type was invented in China by Pi Sheng in 1041–49. Characters engraved into wooden blocks were cut into smaller blocks with single characters; these were finished off with a small knife to ensure a uniform size. These smaller blocks were then fixed in place inside a framed wooden form. The surface of this settled text would then be inked and rubbed into paper in a process similar to that of traditional woodblock printing. Metal movable type sets were first produced by Koreans in the 1370s. The technique of movable metal types printing reached its highest point in Korea in the fifteenth century; at that time, characters were produced in copper on the order of King Sejong (1418–1450).

© koninklijke brill nv, leiden, ��4 | doi 10.1163/9789004275058_006 Tibetan Woodblock Printing Culture 117 trying to be more effective in their missionary activity, brought metal movable type technology into Tibet at the beginning of the eighteenth century. The Capuchin Order, as part of its missionary activity began in 1741 in Lhasa, started to print the polemical texts with the Tibetan mobile printing fonts with the intention of converting Tibetan monks to Christianity. The Tibetan metal fonts were made in Florence. Despite this introduction, this technique of duplication of text in Tibet was ignored by the Tibetans. Also in the nineteenth century, a few sets of Tibetan movable type were cast in metal outside Tibet, for use in a number of books printed in Tibetan. It is certain that these originated outside Tibet. In fact, books printed with movable type in the Tibetan language are rather rare, but they do occur. Also, examples of metal type sets are preserved. Additionally, books printed in the Tibetan language using lithography were noticed in Dharamsala (Figure 66). However, this technique is much more recent than the others mentioned. Both movable type and lithography printing have never been developed in Tibet by Tibetan craftsmen. The earliest known extant printed Buddhist text originates from Korea and has been found at the Pulguk Temple in Kyungju, which was the capital of Tongil-Sinla.3 The text was inside a stupa sealed in 751.4 The earliest well- defined extant blockprint was made in Japan in 764.5 The earliest extant printed book, dated 868 by its colophon and now in the collection of the British Library, was found in Dunhuang, China.6 But, most scholars agree that printing technology appeared before these dates. It is definite that in the ninth century, a variety of book types were printed. But before that advanced level had been

3 Pan 2000: 278–279. 4 It contains text of the Dhāraṇī Sutra and is printed in Korea on domestic paper. By contrast, the story of the dissemination of printing in an eastward direction appears to have been less complicated, and is better documented. When Koryo king Hongjong (982–997) wanted to introduce printing techniques from China to Korea to print Buddhist sutras, he sent his min- isters and monks to Song dynasty emperor Taizong (976–997) and received two sets of printed Tripitaka and two Chinese officials who were in charge of Korean publications. They stayed in Koryo for two months in 993 and recruited forty young Korean students to study printing techniques in China. 5 In 764 the Empress Kōken commissioned one million small wooden pagodas, each containing a small woodblock scroll printed with a Buddhist text (Hyakumanto Dhāraṇī). These were distributed to temples around the country as thanksgiving for the suppression of the Emi Rebellion of 764. These are the earliest examples of woodblock printing known, or documented, from Japan. See: http://www.papermuseum.jp/column/data/001.html. 6 Carter 1955: 54–66; Wood and Barnard 2010: 58–71. 118 chapter 5

Figure 66 Tibetan book printed in the 1960s in India using lithography. Preserved in the Library of Tibetan Works and Archives in Dharamsala, India.

achieved to print such a complicated artwork as a book, craftsmen surely made many attempts to conduct simpler projects. There are quite extensive sources on the history of Chinese printing; by contrast, there is very little written evidence about the subject of Tibetan printing, which is very much inscribed in both in the history of Buddhism and about the nature of the texts that would be selected for printing. In Tibet not long before printing appeared, it had been decided that all Buddhist literature would be inherited and taken from India, not from China. In India texts were written on palm leaves; printing technology was even not considered. This is likely why at that time Tibetans were not yet ready to learn and use a new technology, and most Tibetan early prints originate from areas outside of Tibet. Chinese sources usually locate the beginning of woodblock printing in Northwestern China and are keen to claim this invention, as are many others concerned with book cultures in Asia.7 Archeological findings suggest that the idea of printing was inspired somewhere in Central Asia, possibly in a region under Chinese control. Most early xylograph prints were found in Central Asia along the Silk Road

7 See: Pan 2000: 265–266. Tibetan Woodblock Printing Culture 119 at the beginning of the twentieth century, and those are our primary sources for reconstructing the early history of printing.8 The oases on the Silk Road were inhabited by many different groups, including the Chinese, Uighur, and Tangut; we cannot be sure about the nationality of the first craftsmen, who were able to stamp and later print—this is because we do not have early sources written in languages other than Chinese. The region where the greatest quantity of early block prints was found is the Turfan oasis in what is now Chinese Turkestan. Printed books are a small but significant part of the Turfan collection (Figure 67). This location was first excavated by the expeditions of Grünwedel and von Le Coq in the years 1902 to 1907.9 The peak of Uighur civilization was in the ninth and tenth centuries of our era, but Turfan remained an important cultural center even after its con- quest by the Mongols under Genghis Khan.10 In its golden age, it was a multicultural center of literature and art. The early wood prints discovered there are not only rubbings from stone inscriptions, but also stencils and pounces, printed textiles, seals of different sizes, and many examples of stamped figures of the Buddha.11 Of the printed books which first appeared in the ninth century the most popular is the concertina form printed in Uighur with black ink. These block prints were probably produced mainly in Chinese workshops, despite the fact that they were found in many places in the Turfan oasis, contain Chinese, Tangut, Mongolian, Tibetan, Old Turkish, and Sanskrit texts, and most can be dated to the thirteenth and fourteenth centuries, with only singular examples dated to the ninth.12 It should also be mentioned that many of these sutra translations—regardless of whether they are written in Uighur, Sanskrit, Tibetan, or another language—exhibit Chinese pagination, which inspires doubts that the national origin of the craftsmen who produced these

8 These early manuscripts from the Silk Road, including examples of early prints, are located in more than twenty institutions around the world. Among others, the books can be found in the collections of the British Library in London, the Berlin-Brandenburgische Akademie der Wissenschaften, the Staatsbibliothek in Berlin, the Museum für Asiatische Kunst in Berlin-Dahlem, the Bibliothèque nationale de France, the Musée Guimet, the Institute of Oriental Manuscripts (IOM) and the State Hermitage Museum in St. Petersburg, the National Library in Beijing, and Ryukoku University. For more information about those collections, see: http://idp.bl.uk/pages/collections.a4d and also: Hopkirk 1980. (or its Polish edition: Obce Diabły na jedwabnym szlaku). 9 Carter 1955: 141. 10 For more information about the history of Turfan and its discoveries, see: Hansen 2005. 11 Carter 1955: 39. 12 Berlin-Brandenburg Academy of Sciences and Humanities. Turfan Studies. Berlin 2007. 120 chapter 5

Figure 67 An example of a block print from Turfan in Uighur language executed in six vertical text lines per page on a single-layered, woven type of paper (size 10.5 × 14 cm). Collection of the Depositum der Berlin-Brandenburgischen Akademie der Wissen- Schaften in der Staatsbibliothek zu Berlin—Preussischer Kulturbesitz, Orientabteilung (U 4753). Tibetan Woodblock Printing Culture 121 texts were the same as the linguistic origins of the particular books.13 Interestingly, Chinese pagination can be found in many of the early woodblocks in Tibetan produced outside Tibet proper. Intriguingly, in the 1960s Robert Shafer put forth the hypothesis that Tibetans could have been the inventors of block printing.14 This idea was based mostly on the concept of the formation of the word ‘printing block’ in both the Tibetan and Chinese languages (Tibetan: dpar, spar; Tsangla: par; Chinese: pan). After another half-century of research, this hypothesis seems simplistic without other written and material evidence, and the idea has been abandoned. Even if we discount how developed Chinese book culture was in the seventh century, this fact is unlikely to limit experimentation and invention to the Chinese or Tibetans. The Uighurs and Tanguts were known for their excellent craft skills as well. Tibetan, among other languages, was noticed in early Uighur block prints. These documents, however, are not dated precisely; they could have been printed any time between the ninth and thirteenth centuries. Taking into consideration that just before the time of Uighur domination in Turfan, Tibet expanded its territory greatly to the North, we can assume that this expansion provided one possible way for Tibetans to learn about printing. However, the early history of printing in Tibet still remains obscure. It is generally accepted that Tibetans adopted block-printing technology as early as the twelfth century and continue to carve woodblocks for printing into the twenty- first century.15 This date is estimated as following the occurrence of the earliest known documents printed in the Tibetan language. The earliest datable work in Tibetan language currently known is a small prayer printed in Khara Khoto, a Tangut city in western Inner Mongolia, in 1153 and preserved in the Institute of Oriental Manuscripts in St. Petersburg (Figure 28).16 This is a double-sided, printed, twenty-three-folio booklet (17.3 × 12.7 × 0.6 cm) composed of twelve bi-folios folded in half. Every bi-folio includes six lines of text printed on each page in a full frame 15.8 × 9.6 cm separated by the folding mark in the middle. Letters are relatively narrow and in some places not well printed, and were corrected by hand later. This page layout suggests that printing was intended for a booklet format, which is quite exceptional.

13 Carter 1955. 14 Shafer 1960: 328–329. 15 Harrison 1996: 81; Schaeffer 2009: 9. 16 Accession number: Kh-Tib.67. It is printed with black ink on laid patchy paper glued in two layers characterized by seven twenty-one to twenty-two laid lines in 3 cm. The texture of the paper is soft and absorbent, with brush traces originating from gluing the layers together. 122 chapter 5

Chinese pagination and the folding marks represent the Chinese style of printing and suggest that this work was executed by Chinese craftsman. However, judging by the number of booklets preserved, it was not the most popular format of the book in China, especially not for important administrative or religious texts, which were usually executed as scrolls or concertinas. The shape of the letters does not represent the highest quality of calligraphy serving as a model for preparing these xylograph blocks. The upper line of text formed by parallelogram-shaped heads of the letters is not straight, which suggests care- lessness on the part of the craftsman and that simple and straight chisels were used for its production. At some points rounded strokes are also visible; however, these were achieved using a straight chisel as well. The page layout, quality of work, and type of paper all suggest a Chinese origin and that this early printed work was intended for common use. Leonard van der Kuijp points to thirteenth-century evidence for the preparation of a xylograph with the Tibetan text of the Tshad ma rigs pa’i gter and the autocommentary of Sa skya Paṇḍita Kun dga’ rgyal mtshan (1182–1251), completed sometime in the 1220s.17 One of two different xylographs of this work examined by van der Kuijp in the Tibetan Library of the Cultural Palace of Nationalities in Beijing was probably prepared in the last decades of the thirteenth century. The second block print is dated by colophon to 1339. Unfortunately, the place where the blocks were produced is not mentioned. Interestingly, the pagination in both printed works is given in Tibetan and Chinese on the left of the recto, and only in Chinese on the verso side of the folios. This pattern is repeated in most of the early block prints produced in China and Mongolia. Once again, this suggests that mostly Chinese block carvers and printers were involved in the production and that their work was facili- tated by double-sided pagination to avoid confusion about the woodblock order. There is more evidence on various Mongol-sponsored printing projects propagating the Kalacakra literature first in China and then in Mongolia dated to the last decade of thirteenth century or the first decade of the fourteenth century. These prints are generally known under the name ‘Mongol xylographs’ (hor par ma).18 The other early datable Tibetan work currently known was printed in China under Mongol patronage no later than the 1270s.19 Schaeffer also mentions the very significant role of Pakpa Lödrö Gyeltsen (1235–80) in the dissemination of printing technology in Tibet. In his printing projects, Tibetan and Chinese

17 van der Kuijp 1993: 279–298. 18 Ehrhard 2000: 11. 19 Schaeffer 2009: 9. Tibetan Woodblock Printing Culture 123 craftsmen worked together sponsored by Mongol leaders Yesu Boga and Hayimdu.20 There is also an example of a volume of The Kalachakra Tantra (Tantraraja Shrikalachakratantra or Laghukalacakratantra; dus ’khor rgyud) printed from woodblocks and carved for the funeral ceremonies of the Kublai Khan (1215– 1294) sponsored by Khublai Khan’s wife and son that supports the above- mentioned dates (Figures 68 and 69). This book is in pothi format, with six lines of text printed between two side margins executed with a single line and Chinese pagination on the right side and Tibetan on the left of the recto of the folio. It displays a high-quality printing production with regular and propor- tional letters aligned to the top line of text. The lower part of the head of the letters were cut with a rounded chisel. This feature of the letters’ execution is shared with the first editions of Tibetan Kanjur produced in Beijing (especially Yongle and Wanli). The history of printing in proper Tibet has received some attention in recent years. David Jackson presents the idea that books started to be produced within the Tibetan cultural area by block printing in the beginning of the fifteenth century.21 The most representative works of Tibetan printing culture are the editions of the Tibetan Buddhist Canon, which underwent a final compilation in the fourteenth century by Butön (1290–1364). The Tibetans did not have a formally arranged Mahayana canon and so devised their own scheme, which divided texts into two broad categories of the Tibetan Canon: Kanjur and Tenjur. In general, the combined number of volumes of the two parts is over three hundred, and each side of a single page is printed from an individually carved woodblock.22 It is still believed by many scholars that before 1410, when the first woodblock edition of the Tibetan Kanjur was produced in Beijing, all Buddhist canonical collections were produced and reproduced as handwritten manuscripts. The Translation of the Word (Kanjur) and The Translation of Treatises (Tenjur) undertaken in 1410 by the Yongle Emperor (r. 1402–1424), first produced in 1410 is known as the first large-scale printed edition and almost nothing prior is known with certainty. In Beijing, new impressions continued to be taken from the Yongle blocks and in this way, the Wanli edition printed in black ink in 1606 was produced under the sponsorship of the Wanli emperor (1563– 1620). When the blocks wore out, new blocks were prepared and carved using

20 Schaeffer 2009: 10. 21 Jackson 1983: 107–116; Jackson 1989: 1–18. 22 Harrison 1996: 81. 124 chapter 5

Figure 68 Volume of The Kalachakra Tantra (Tantraraja Shrikalachakratantra or Laghukalacakratantra; dus ’khor rgyud) printed from woodblocks carved for the funeral ceremonies of Kublai Khan (1215–1294) sponsored by Kublai Khan’s wife and son. It provides evidence of the early dates of Tibetan woodblock printing. Ink on paper. The collection of Gene Smith. The Tibetan Buddhist Resource Center, Cambridge, Massachusetts, USA (TBRC W4CZ75).

Figure 69 A fragment of folio from The Kalachakra Tantra printed from old woodblocks that were carved for the funeral ceremonies of the Kublai Khan (1215–1294) under the sponsorship of Kublai Khan’s wife and son. It provides evidence of the early dates of Tibetan woodblock printing. The collection of Gene Smith. The Tibetan Buddhist Resource Center, Cambridge, Massachusetts, USA (TBRC W4CZ75). Tibetan Woodblock Printing Culture 125 prints taken from the old blocks as a master. These are represented by the Qing dynasty re-edition and its reprints. Since the seventeenth century, Kanjur sets have been created, edited, and produced in Tibet and Mongolia. As discussed above, there is a significant distinction between printing a simple diagram or figure, or even a book, and printing an edition of the Tibetan Kanjur, which is even not a single book but a hundred-plus volume set (edition), or the entire Tibetan Canon. Difficulties in the organization of labor for such a large project and differences in the economic situation of particular areas in Tibet might have contributed to the delay in the beginning of printing technology in Tibet. Another reason might be that it is very difficult to trace some fragments of works which were still under the compilation process, even if some of them were printed already. This long and challenging process of editing and compilation of the Tibetan Canon lasted until the end of the fourteenth century. The books that appeared earlier are not yet dated or known. We experience challenges regarding the identification of the origin, date, and provenance of these books or volume sets.

Production of the Tibetan Buddhist Canon

Every project began with a decision of what would be used for the master copy of the planned edition, and required considerable preparatory work. This master work was usually a collection of the original manuscripts or previously printed texts. Scholars worked on text collation and compilation. When a text was accepted, the process of proofreading commenced and lasted until the end of the project. At the same time, wooden supports had to be prepared and paper and ink accumulated or manufactured. Next, fabric covers were stitched for woodblocks, economic plans prepared, and the list of specialized professionals to be involved in the project was determined. Consecutive working stages were well defined and needed to proceed in a systematic order. That is why a high level of understanding on the part of the working personnel was a guarantee of a good-quality print. Simultaneously, many other craftsmen responsible for preparation of the right material were involved. A lot of useful information about the organization of such printing projects can be found in the works of Dung dkar blo bsang ’phrin las (Dungkar Lozang Trinlé).23 Technological knowledge of the process of preparing the wooden blocks, the subsequent stages of production, the number of people involved

23 Blo bsang ’phrin las (Losang Trinlé) 1990: I read an English translation of this article. 126 chapter 5 and their skills, the economic conditions, and all other social and technical circumstances are essential to understanding the information stored in the structure of these books now. I also collected many of these pieces of information through interviews with Tibetans, reports of scholars visiting Derge Printing House, and from my own experience as an artist, which helped me to interpret much of the collected information. This is why before presenting results of my study on particular books and sets of volumes, I will briefly present the character of work, skills, and responsibilities of people involved. The responsibilities of the particular professionals were distinctively shared. The proofreader was a person who was required to have a high level of knowledge in the general sciences, as well as more specialized knowledge. This person usually was experienced in the proofreading of both manuscripts and woodblocks. Proofreaders could be classified in four divisions reflective of their level of competence. A smar zhus pa (reader) compared the sketch of the text prepared for carving into the wood blocks with the original manuscript (master copy). His main task was to compare the syllables, page numbers, chapters, and general content and outline. Then a bskyar zhus pa (repetitive reader) should compare the prepared sketch of text with the woodblock to distinguish and determine the right size of letters and check the depth on the carvings. At the next stage, the yang zhus pa (corrector and proofreader) should check the two previous stages of corrections mentioned above, perform a final proofreading, and insert missing and any extra required syllables. The final word at this stage belonged to a zhus chenpa (scholar with the highest educational background and the most powerful rights and competence). He carefully investigated the three previous corrections to make the final decision of whether the text was correct. He also examined a master copy (original manuscript) to finalize the copy to be transferred to the woodblock. He needed to review the master copy again, as well as the printing plate, after they had been carved. If the material had obtained a desired level of quality, he was obliged to provide a satisfactory footnote. In case of any negative remarks, he was supposed to give his approval after detailed investigation. Text was usually sketched in a full or side-margin-only frame with pagination or foliation located in the left margin perpendicular to the lines of text. A thangka painter was responsible for drawing deities in the empty spaces left by the scribe, usually on the top portion of a couple of the first pages and possibly also on the last pages of the text. In Tibetan books produced from woodblock printing, both text and illustrations were printed from the same carved woodblock and the illustrations were subsequently painted or colored. Tibetan Woodblock Printing Culture 127

The task of the secretary was to write text and attach a standard sketch (par tsak) to the woodblocks to transfer a text designed and written by a scribe. Text accepted by a final proof reader had to be written with ink on very thin Tibetan paper. At the same time, enough wood had to be collected, usually from the closest possible area. For example, in the Derge Printing House they now collect wood from Derge (sde dge), Joda (jo mda) and Pelyül (dpal yul). But in the past, they also collected from Tramok (spra rmog) in Kongpo (kong po), Central Tibet, depending on the economic situation. In general, the selection of wood depended on what grew in the region. The most-often-used materials for xylographic blocks were pear and jojoba trees. In tropical regions shorea, sal, and sandal trees were favored; in Central Asia various fruit trees and jojoba were preferred; and in the northern parts of the Himalayas cedar, juniper, pine, and other coniferous trees were used. A variety of tools were used depending on place and time. This is why, depending on the region and the craftsmen’s pref- erences, chisels and wood used for the printing blocks varied. The woodblocks were manufactured by a carpenter who was also responsible for the preservation of the library (kun dga’ ra wa) in a monastery, and maintained the woodblocks after the printing was completed. Collected wood was cut into boards, and evened. Then it was dried in shade to avoid the formation of cracks, which often appeared when drying wood in direct sun. Then the surface was smoothed. When the blocks were ready, text was transferred from prepared blocks by pasting written paper with wheat flour paste to the wooden block by placing the text downwards on the printing block surface. After the paper was pasted and dried, the text appeared as a mirror-image on the surface of the wooden block. At this stage, the paper could be gently removed by making it damp and slowly taking it out by rubbing (Figure 70). The reverse lettering should show clearly. When the board was dry again, the text could be rubbed lightly with a little mustard oil pad khang snum or pad kha snum, which helped to show letters clearly/sharply. According to patterns left on the surface of the paper, we have some idea how the wooden blocks were carved (Figure 71). First, the straight lines of letter heads, frames, and margins were carved. Then, following the contours, the distance between the characters and their precise shape was cut. This resulted in a relief mirror text. It was common that every carver had his own tools. These were curving knives and chisels (sbug gzong), carpenter’s chisels of two different sizes, five different kinds of chisels to thrust, a mincing knife, also called a cleaver, and a whittling knife. From traces they leave we can distinguish two general types: gzong used for cutting out the straight and curved 128 chapter 5

Figure 70 Fragment of a woodblock for Narthang Monastery being carved by craftsman from Nyemo. After the paper was pasted and dried, the text appeared as a mirror-image on the surface of the wooden block. At this stage, the paper could be gently removed by making it damp and slowly taking it out by rubbing.

Figure 71 Fragment of a woodblock for Narthang Monastery being carved by craftsman from Nyemo. At this stage letters have been already cut out. Tibetan Woodblock Printing Culture 129 outer lines of the letters,24 and che gtsugs connected with the verb ‘btsug’ which means ‘to pierce.’25 The carvers’ tools and skills highly influenced the quality and style of the work. Following these features, the style of a particular edition was formed to some extent. That is why from this point of view, examination of tool traces is even more important than the character and skills of the craftsman, especially considering that one edition was produced by thousands of carvers. For example, as reported by Corneille Jest, when several carvers are working together they often divide their labors over several blocks at once, with one man working with one tool, while another craftsman would use a different tool, etc. When a large project was finished, the tools were often sold to other workmen across the region. This is why some features of printed books, such as the shape of letters originating from particular chisels, can be significant for a particular region or workshop. Otherwise, even taking into consideration that prints are easier to identify than manuscripts, it is still rare for the colophons of Tibetan books to mention the name of a carver or even the name of a patron. The printing projects were usually larger ventures demanding many specialists and good organization. Thus, such ventures were undertaken by people of high social status and by large monasteries. Sometimes when all the blocks of one literary work were finished, the names of carvers were given at the end together with the place of workmanship and the name of the benefactor (patron) who paid for the work.26 This is why it is very difficult to identify the workshop or place of origin of the book. After both proofreading and carving, the finishing grease was applied. One woodblock required approximately two sangs (50 g) of melted butter. After applying grease to all the woodblocks, they were dried in the sun until the grease was absorbed completely. Then the blocks were washed in water. Preparation of printing blocks was the most important and the most labor- intensive step. Each letter had to be very precisely and accurately mapped, and then cut. Woodcarver work required patience and the highest qualifications. Implementation of a medium-sized block lasted about seven to ten days. To print Kanjur, one should prepare approximately forty-two thousand printing blocks engraved on both sides. For example, the Peking Kanjur was printed from 41,522 blocks engraved on both sides. This gives an idea of how long he had to take up this phase of the work, and how many carvers had to take part

24 mig gzon: ‘eye chisel’ used for cutting out the little circles in letters; bug gzon: holing chisel; gas gzon: cleaving chisel. 25 che gtsugs: large piercer; seg gtsugs: oblique piercer. 26 Jest: 83–85. 130 chapter 5 in it. Interestingly, the carvers were free from taxation but could be summoned to Tashihlunpo, Lhasa, or other centers of government work at practically any moment, and it was their duty to report with no option for refusing the commission. After successful completion of all preparations together with proofreading, the actual printing onto paper was usually initiated. As described above, after printing each chapter or part, each woodblock print was compared and proofread four times against the original manuscript by specialists ranging from first readers to scholars with the highest competence. Finally, once the highest edi- tor was done with the proofreading, he placed a note in the printing register and conveyed the corrected material to a supervisor of the woodblocks. This proofreading was supposed to be done straight after obtaining the original to avoid the risk of publishing an incorrect text (ma dag rgyun ’byams), which would later be widely disseminated and used. Even after the initiation of the print itself, proofreading was still vital to ensure the quality of the text. Even at the stage of final proofreading of carved woodblocks, it was still possible to supplement missing syllables and correct mistakes. Also, even after preparation of the woodcut blocks was finished, it was possible to make amendments. Small pieces of text could be removed and replaced with others by pasting a new piece of wood and re-cutting the text. For example, in Derge, ash wood was used for corrections. In the same way, before preparing any subsequent reprint, from a technical point of view, it was possible to enter a text correction. So the next reprints might to some extent vary in terms of content, despite possessing the same formal features of the rest of the edition. However, consecutive reprints never varied to the same extent as editions. Sometimes hundreds of copies of the book were printed from a single set of woodblocks. After taking charge of the woodblock, the supervisor would place it within a cover that had been made earlier, arrange it together with other woodblocks in a woodblock box, and list it in a print register, including a legible reference to facilitate locating the woodblock in the future. Tailors stitched the woodblocks’ covers out of wool. They also patched workers’ clothes and made uniforms, caps, and shoes. The paper manufacturer was responsible for providing the best quality paper for printing or making it, including cutting the edges of the paper at the end and preparing leaves of a suitable size for printing text. Ink and paper for printing differed slightly from the materials produced for manuscript use. Prints achieve better quality on relatively soft paper in comparison to paper used for manuscripts with highly processed surfaces. On the other hand, prints are usually made on paper of worse quality than manuscripts, possibly due to the expense of mass production. Ink had to be thick enough to settle equally Tibetan Woodblock Printing Culture 131 on the paper surface during the short and quick contact between paper and printing block. Paper needed to be less sized than that for manuscripts and soft enough to absorb the ink from the woodblock with one touch. Additionally, doctors were involved with the project from the beginning to the completion of the print, caring for the sick by prescribing required medicines. Independently, a group of helpers assisted the highly specialized personnel in all kinds of matters, as well as doing kitchen work. There were often an insufficient number of highly skilled personnel for enterprises of such a large scale, and the work itself was very time consuming. New apprentices needed to be recruited for each of the tasks involved so that they could learn the necessary skills during the process; in this way, financial losses were avoided. New technicians were summoned to learn the actual work when the printing was being organized. A technical expert (teacher) taught more specialized technical skills such as writing and printing to the newly arrived students, conducted practical courses, and then gave examinations. Once the students obtained the desired level of abilities, the teacher’s responsibility was to shift them to the workshop. After transitioning to the workshop, students studied for a few months under the guidance of the various technical experts; they also assisted them in their work. A few months after beginning the printing work, many of the apprentices were fully skilled and able to work as woodblock carvers, etc.; the same approach was taken in the case of secretaries, blacksmiths, carpenters, and proofreaders. Since in modern times printing facilities are excellent, work productivity has greatly improved. In the past, after one had proofread the original manuscript, one needed to prepare a corrected, handwritten copy of it. This is only one of the reasons why printing texts used to be so time consuming. Printing projects of Tibetan canonical works usually took years and involved numerous specialists of more than ten professions, such as: proofreaders, woodblock carvers, paper manufacturers, carpenters, Tibetan scroll (thangka) painters, blacksmiths, technical experts, and helpers. Also managers, secretaries, and many apprentices were involved.27 Though there was a division of labor between representatives of all mentioned professions, they were all interconnected. Good organization, skillful craftsman, good managers, and the availability of needed materials were the most important factors and were the foundation of every printing project. Therefore, every time enormous preparations were made from the beginning in order to maximize the output of the main workshop. Precise, careful, and systematic preparation was the key to success. Any disturbance that influenced the systematic character/scheme of

27 Blo bsang ’phrin las (Losang Trinlé) 1990. 132 chapter 5 work or the order within its sequence caused serious financial losses to the sponsors. On the other hand, experience gained over time at the same workshop within the same organizational scheme could make a difference in terms of time and quality of work. This was exemplified by the printing of the Narthang Tibetan Buddhist Canon. Printing of the Kanjur took nine years; the subsequent printing of the Tenjur, during which a similar number of wood carvers was involved but a number of volumes were doubled, took only one month and twenty-five days longer. The degree of professionalism and the speed of printing connected with it had increased greatly since the time of publishing the Narthang Kanjur. This established a well-organized enterprise based on experience and skills gathered during the printing of the first project when the Narthang Tenjur was produced.

Edition, Re-edition, and Reprint: The Technical Identity of Hundred-Volume Sets

The general style of a given edition derives from the wooden blocks used to create it. Different sets and editions, however, may be printed from the same wooden blocks, such as in the case of the Yongle, Wanli, and possibly also other editions printed later in Beijing. In this case, the types of materials and techniques used for the books’ preparation define a set. According to an explanation by Helmut Eimer: “For preparing the second set of wooden blocks, printed copies from the first, which served as master copies, were pasted with starch glue on the prepared wooden planks.” So, regarding text content, this could be the same edition despite a new set of wooden blocks having been used. According to art terminology, we could call such an edition a reproduction. Japanese scholars who visited China during World War II brought back the information that two different sets of blocks were generally used for printing the Kanjur in Beijing. According to the terminology above, even two sets characterized by the same page outline measurements do not necessarily need to be reprints; they can also be different editions. It is very difficult to evaluate how re-carving all the blocks would change the final print as compared to the master copy. First of all, this depends on the carvers’ skills. We can assume that measurements of the page outline will be the same. It is rather unlikely, however, that the copies will be identical. At the moment, there is no evidence confirming the total number of woodblocks prepared for the so-called Peking editions. This is why, for my research, identification of different sets was more essential than that of particular editions. Sets can be identified by possessing the same physical Tibetan Woodblock Printing Culture 133 features also besides page outline, such as type of paper and ink and the same style of painting sometimes executed over xylograph print. I am using the term ‘edition’ in this essay for a set of volumes characterized by the same page outline related to one set of woodblocks. It is not possible to know the number of prints struck from one particular set of wooden blocks. It should be remembered that the word ‘edition’ usually relates to more than one set of Tibetan Kanjur volumes. Nevertheless, if we classify canonical volumes into sets or editions, documentation of all measurements characterizing page outline and material properties of books are essential. Then identification of woodblocks used in particular printing workshops is an important benefit of the physical examination of books. The most important features for identification of woodblocks are: (1) the size of the printing area whether full-frame or limited—such as with side margins, the distance between upper lines of text, and the size of the wooden block; (2) whether the curved block is single- or double-sided; (3) traces of ink color on the wooden block (black, red, or both); (4) type of numbering of the leaves (pagination or foliation, and in which language); (5) identification of carving tools, such as both photo and descriptive documentation of chisels and other tools found in the workshop, and documentation of chisel traces on the woodblock (including details of texts and shapes of particular letters, with stress on the upper line of text shape); (6) identification of wood (good-quality photos of the wood pattern on the surface cross section are especially desirable; before taking a photo, the wood should be cleaned and dusted to afford a clear view); and (7) the identification of Tibetan names for different types of tools. As described above, the text of xylograph prints is first written by hand on the paper and transferred onto a wooden block, carving comes next. The shape of each letter derives from the handwriting of the scribe and block carver and the types of tools he used. This is why we can still compare particular letters in books. The main features to compare are proportions of the particular letters and some specific parts of the same letters in different books. For example, the shape of the head part of a letter, which is usually the place where carvers start their work, can be used as a fingerprint for tracing the workshop where a particular book was produced. As mentioned above there are some differences in the use of term ‘edition’ by scholars representing different disciplines. From the viewpoint of philology, the concept of an edition concerns the form and content of the text only. From the viewpoint of a librarian, ‘edition’ is a number of prints struck from one set of wooden blocks, usually at the same time. However, this definition does not 134 chapter 5 provide a clear boundary on this concept in the context of Tibetan printing culture. In his paper, Jonathan Silk refers to the distinction between a re-edition and re-printing.28 Following this way of thinking, the difference between ‘edition,’ for which a fresh set of blocks was prepared, and ‘re-printing,’ from already existing blocks, still causes confusion. Classification becomes complicated when we consider many re-editions, reprints, and only fragmentary copies made from wooden blocks that were sometimes used for hundreds of years. For instance, the Yongle and Wanli were printed from the first set of blocks, which is why they belong to the first edition. They are the first woodblock- printed collections of the Tibetan Buddhist canon sponsored by Chinese emperors who patronized Tibetan Buddhist teachers. For the issues that followed, a new set of blocks was cut in the years 1684–1692 under the Kangxi Emperor and remained in use until the end of the nineteenth century.29 Only the first print of 1410 and the first print prepared during the Qing dynasty between 1684 and 1692 can be called ‘editions’ in the proper sense of this word.30 However, in practice it is impossible to differentiate copies printed from the same woodblocks when there is no additional source information on particular set features. Many aspects of this problem can be resolved with the help of written historical sources, but some can be answered only by penetrating investigation of every volume. For example, the recently rediscovered in 2003 volumes from the Pander Collection in Cracow were not clearly classified within sets as collector Eugen Pander informed us. Within sixty volumes of the Pander Pantheon, there are no volumes from the Yongle Kanjur set. Eugen Pander believed that from the Yonghegong in Beijing, he obtained twenty-seven volumes of the Yongle Kanjur, comprising twenty-four volumes of the tantra and besides the Yongle Kanjur thirty-two other canonical volumes printed during the late Ming Dynasty. However, he could classify the Wanli set as the Yongle edition following the proper sense of the word ‘edition.’ Examining the Pander statement in the Lamaist Pantheon and books in Cracow, I can explain this. His identification of Wanli volumes as the Yongle edition is correct if we know that both the Yongle and Wanli sets were printed from the same wooden blocks. The cap- tions to his iconographical study show clearly that exactly the same pictures as those from the Yongle/Wanli Kanjur I found in volume 46 of the Wanli Kanjur are identified by Eugen Pander as the Yongle Kanjur.31

28 Silk 1996: 155, note 7. 29 On the so-called Peking editions of the Kanjur see Imaeda 1977: 23–51; and Eimer 2007. 30 Silk 1996: 155, Eimer 2007: 40. 31 For further information, see: Pander 1889: 201; Eimer 2000: 27–51; Helman-Ważny 2009. Tibetan Woodblock Printing Culture 135

Thus, the evidence of techniques and materials used within particular books allows us to read distinctive features that collectively may be called a style or technical identity of the edition and set. I found the term ‘technical identity’ used by Helmut Eimer to be very accurate in the context of ‘Kanjur research.’32 However, I should explain that I understand this term slightly differently than it was explained. Eimer views technical identity mostly as a relationship between a master copy and its re- issue, basing his observations mostly on page image, and without considering the materials, such as paper type. He explains that new woodblocks produced for the re-issue of the Kanjur were prepared using the technique of transferring a text from primary woodblocks to new ones by pasting a one-sided print on a new block. This way, the page outline of both master copy and successive issue are very close to each other and the same Tibetan letter/word distribution may be visible in both issues. I would like to propose using this term in the context of complex physical descriptions of the features of particular volumes derived not only from page outlines, but also from the type of paper and ink, and from the technology of printing and papermaking. My present research on successive editions of the Tibetan Kanjur shows that to define a particular set, which can contain more than a hundred volumes, the technical identity must be based on precise physical and technical observations. And drawing the distinction between editions and sets is essential to understanding the Tibetan history of printing. Thus the variations exhibited in different versions of the collection, with regard to style, arrangement, formats, materials, and techniques used in their creation, were the main focus of my study. The comparative examination of different sets discloses the technical similarities between editions within the same set, despite differences occurring between particular volumes within the same edition. The main question, which appeared after examination of many books from distinctive editions, is how closely particular volumes should relate to each other to be called one set. It should be remembered that as described in the previous section of this chapter, the edition of the entire Kanjur set, which comprises more than a hundred volumes, has always been a large project and great challenge for all the craftsmen involved.33 The relationship between a master copy, its reprints, and its re-editions allows for placement of unidentified sets of the Tibetan books in regional and temporal contexts. Materials used in books especially can tell us about the regional origin of a

32 Eimer 2007: 39. 33 A very useful source of information about Tibetan editorial practice in the seventeenth century is provided by Schaeffer 1999. 136 chapter 5 particular edition. The techniques of book production, quality of work, and type of tools sealed in the books can also reveal the identities of the craftsmen producing these books and what skills they had. These pieces of information together can lead us to characterization of particular workshops, which would be essential. Interestingly, in cases when production of printed editions alternated with manuscript editions, the inheritance of particular features was also noticeable. The Kangxi editions/sets, for example, contain text in full frame, exactly as in the manuscript Kanjur of Berlin, which may serve as a master copy for Kangxi production. The Yongle/Wanli edition features text composed between two side margins. Also, the shape of letter heads in Kangxi is very straight, exactly as it is in the handwritten text executed with a wooden stick. It is possible that some features evolved depending on techniques used and available tools. In a manuscript copy, it was not possible to recreate the rounded shapes of letter headings, which could only come from chisel shapes. Taking into consideration inheritance of features of a master copy and its re-edition, in this chapter about printing culture I also include material evidence collected from the Berlin manuscript Kanjur, which was copied by hand from a Wanli print. In the case of consecutive editions of Tibetan Kanjur, we usually know approximate dates of production, as well as the region of origin. This is why the main purpose of my research here is not dating and finding the origin of particular editions, but seeking out and documenting patterns which can then be used as a reference for identification of unknown fragments of Tibetan Kanjur still preserved in many institutions and private collections. For example, on the occasion of an auction of art, it is often possible to see loose covers, title pages, or single volumes which previously were a part of sets. Usually the way in which they are described is insufficient or downright unbelievable, and dating is very rough. When a mass of books is found, the most reasonable course of action is to link particular types, group them, and compare them to those few dated and identified already. The observed patterns characterizing particular sets can link these books to particular places of production, and maybe even workshops in the further perspective, and will allow for viewing Tibetan books in temporal and regional perspectives.

Early Editions of Tibetan Kanjur Produced in Beijing

The same common features are exhibited in a few sets of early editions produced in Beijing. Below I will present these features documented in the Yongle and Wanli sets, the manuscript Kanjur from Berlin copied from the Wanli, and Tibetan Woodblock Printing Culture 137 an example of one of the Kangxi set fragments. The summary of these features in examined fragments and sets of particular Kanjur editions are also presented in table form (Appendix 2). The Yongle edition, produced in 1410 by the Yongle Emperor (r. 1402–1424), is not only the first printed edition of the Tibetan Kanjur, but also one of the first printed Tibetan book collections so far known.34 Unfortunately, I found only one folio of the Yongle available for study.35 It is preserved in the Special Collections Library, University of Michigan in Ann Arbor, USA (Figure 72a, b).36 According to the unpublished hand list prepared in 1986 by Bruce Cameron Hall, “Tibetan Manuscripts and Xylographs in Michigan Collections,” several items were received in 1924 from Edward Barrett, a New York fur trader who trav- eled in China in the 1920s, and as a sideline sold ‘Oriental curiosities,’ mostly books and printing blocks. Among these is the single sheet identified by Hall as belonging to the Yongle Kanjur, catalogued as ‘Central Asian Collection 1.’37 It is not known how many copies were printed at that time from one woodblock set sponsored by the Yongle emperor. However, on the basis of written sources we can assume that there were at least four that were offered to the Mount Wutai monastery, the Karmapa, Phaggru, and Tsongkhapa.38 These were the first prints from a newly made printing block. Although hundreds of copies can be produced from a single woodblock set, this edition is now largely lost. The collection is mentioned only by written sources and brief reports of travelers from Tibet and China who had a chance to see it but not enough time to study it carefully. Both sets having been largely lost, it was not certain whether the Wanli was a reprint or re-edition of the Yongle. Comparative analysis of the page outline and shape of letters recently showed that the Wanli was a faithful copy of the Yongle, which confirms that it was a reprint. It is still not clear, however, if text printed during the Wanli era has been corrected and how far it differs from the

34 Wooden blocks for the Yongle edition were carved before 1410 during the Yongle reign of the Chinese Ming dynasty. 35 In the library card text identified by three Buddhist monks from Nepal (visit of 3/26/81) as a page from a Chinese version of Kanjur translated into Tibetan. Shong Dann Ding Wang Jing—this is first part of Jimma page 12 of Time of Kangxi, 1662–1723 AD. 36 Accession number: Central Asian Collection No.1; other identification number 1229 pasted on the folio with the note: “Prayer sheet from prayer wheel, Urga, Mongolia.” Chinese title: Sheng Chan Diag Wang Jing Yi Juan Xia Shi’er (Shang Shi’er); Sanskrit title: Samādhirāja sutra. 37 Hall 1979: Unpublished typescript revised in 1980 as a PhD thesis for the Sanskrit Department of Harvard University; Silk 1996: 171. 38 Blo bsang ’phrin las (Losang Trinlé) 1990. 138 chapter 5

Figure 72a, b Folio 12 of the Yongle Kanjur volume. A. Recto of folio 12. B. Verso of folio 12. The Special Collections Library, University of Michigan, Ann Arbor, Central Asian Collection (No. 1).

Yongle. As is described in a previous part of this chapter, corrections could have been made on old woodblocks, and some fragments may have been carved again. From a technological point of view, however, such corrections could not be extensive; the wooden block could break if too much of the original text were to be cut and replaced. Independently of possible corrections, the Wanli set from Cracow also includes twenty-two supplemental volumes for which new woodblocks were definitely carved. Only two other sets of the Yongle print could possibly have survived in Lhasa; however, nobody has been able to confirm this in recent years.39 One set was stored in the Potala, the second in Sera Monastery. The set in Sera Monastery was given by Ming Chengzu to Tsongkhapa’s disciple Sakya Yeshe, the founder of Sera Monastery, in 1416. The set in the Potala was originally stored in Sakya Monastery, then moved during the Cultural Revolution to the Potala Palace for safekeeping. Photographs of those Yongle sets included in the

39 Jiacuo et al. 1985: 85–89. Tibetan Woodblock Printing Culture 139 article “Lasa Xiancang di liang bu Yongle ban ‘Ganzhuer’ ” published in Wenwu look like a faithful copy of other fragments I have studied of the Yongle and Wanli edition preserved in Ann Arbor and Cracow.40 The next set printed from Yongle blocks in 1606 was the Wanli Kanjur, produced under the sponsorship of the Wanli emperor (1563–1620). In 2003 the search for Tibetan books in Poland led me to the Pander Collection from the former Prussian State Library (Preußische Staatsbibliothek) and to re- discovery of the Berlin fragment of the Wanli Kanjur, which besides one folio of the Yongle in Ann Arbor, is the earliest large collection recently appeared in the West and is available to study.41 Portions of Kanjur editions printed during the Ming Dynasty have been found among Pander’s books, and the Berlin Wanli Kanjur has been identified among them.42 I examined the fragments of this set and its supplement preserved in the Jagiellonian University Library (Biblioteka Jagiellońska) in Cracow, Poland, then two volumes of the supplement to the Wanli Kanjur located at the Harvard-Yenching Library in Cambridge, Massachusetts, USA, and the Berlin Kanjur (manuscript Beck), which was copied from the Wanli Kanjur and preserved in the Berlin State Library (Staatsbibliothek zu Berlin), Germany. This ‘Handwritten Kanjur of Berlin’ was acquired in 1889 by V. Brandt as a Peking Kanjur.43 At some point new blocks were carved, using a manuscript copy or prints of old blocks as a

40 Jiacuo et al. 1985: 85–89. 41 This collection came to light after having been deemed lost during World War II. For detailed information tracing the history and provenance of the missing volumes of the Berlin Wanli Kanjur, see: Eimer 2000: 27–51; Helman-Ważny 2009; Mejor et al. 2010. 42 The Wanli Kanjur (bka’ ’gyur) has been identified in Cracow during the project “The Lost Fragment of Wanli Kanjur Edition in the Jagiellonian Library? The Value of Authenticity of Tibetan Books from Pander Collection in Poland” initiated by the author of the present essay and funded by the Polish Ministry of Science. The study was done with Prof. Marek Mejor and Dr. Thupten Kunga Chashab from the Faculty of Oriental Studies, University of Warsaw. 43 I examined the twenty-eight volumes of the Wanli Kanjur rediscovered in the Jagiellonian University Library in Cracow within the Pander Collection. These volumes belong to the Pander Collection, and are located in the Pander Pantheon section (Pander Pantheon volumes 23–28, 38–57, 59–60). My codicological and conservation study led me to the classification of the Pander Pantheon collection into four sets. These are twenty eight volumes of the Wanli Kanjur re-printed from the wooden blocks of Yongle Kanjur edition (vol. 23–28, 38–57 and 59–60), twenty two volumes of the Wanli Kanjur supplement (volumes 1–22), nine volumes that are somehow related to the Wanli Kanjur volumes (volumes 29–37), and one volume that was definitely produced later from one of Kangxi sets of wooden blocks, clearly different from the Yongle or Wanli sets (vol. 58). See also: Mejor et al. 2010. 140 chapter 5 master. These are represented by the Qing dynasty re-edition and its reprints. The only representation of this set I had a chance to examine is one volume (58) from the Pander Collection in Cracow. This volume is supposed to belong to one of the several corrected reprints of the Kangxi Kanjur (1684–92).44 Both sets of the Yongle and its Wanli reprint clearly display their affiliated patronage. For example, the Yongle as well as a laudation by the emperor attached to the volumes.45 The miniatures of the supplement volume’s cover from the Harvard-Yenching Library depict figures of the Emperor and the Tibetan master, the Imperial preceptor, in the lower, outer corners (Figure 73a, b). These two volumes of the Wanli Supplement from the Harvard-Yenching Library contain covers made of five elements, a plank and four wooden joints fixed together into a frame mounted on the plank’s verso side. The covers are finished with silk and brocade, with miniature paintings and raised gold lettering in the hollow on the verso side (Figure 74). A very similar type of cover is preserved in volume 58 from the Pander Pantheon, identified as one of the volumes from the Kangxi sets. Its upper cover was made of a plank measuring 24.2 cm in height and 71.5 cm in length, and four strips of wood joined together into a frame on the verso side. The upper cover of this volume is finished with silk, and its verso side contains two miniatures and the title of the volume. Unfortunately, the covers were not preserved together with any other fragment of Yongle or Wanli in the West. The hint suggesting what type of covers were produced for those sets is the discoloration on the first folio of volume 12 identified as the Wanli Supplement from the Jagiellonian University Library.46 This resembles exactly the ornament carved on the verso side of the loose Yongle book covers documented elsewhere. Examined covers from R.R.E. Collection contained a curved seal with affiliation of particular volumes titles.47 For transcription and translation of Tibetan text see appendix 3 (Figures 75 and 76). There is a selection of loose covers identified as Yongle covers preserved in Western collections. They measure 26.5‒27 × 72.5‒73.3 × 3 cm and are made of lacquered sandalwood (Santalum album L.) with incised gold (Figure 77).48 The true lacquer originates from a toxic resin produced by a ‘varnish tree,’ Toxicodendron vernicifluum, indigenous to many areas of China

44 At this time, it is not possible to identify this volume more precisely. 45 Edited and translated in: Silk 1996: 173–178; see also: Mejor et al. 2010: 177–179. 46 Mejor et al. 2010: 24, 75. 47 These are individual scriptures that are associated with the Mahāyāna Sūtra: “The Questions of Brahmadatta.” 48 Prof. Tomasz Ważny, University of Arizona identified wood in covers from Arthur Leeper collection. Tibetan Woodblock Printing Culture 141

Figure 73a, b The miniatures of the Wanli Supplement volume’s cover depict the Emperor, and the Tibetan master (the Imperial preceptor), in the lower outer corners. The collection of Harvard-Yenching Library of the Harvard College Library, Harvard University (TTib 1803.7 13 vol. 2, Da zang jing Ming Yongle ba nian). 142 chapter 5

Figure 74 The verso of the Wanli Supplement volume’s cover. The collection of Harvard-Yenching Library of the Harvard College Library, Harvard University (TTib 1803.7 13 vol. 2, Da zang jing Ming Yongle ba nian).

and Japan.49 One set of these covers is preserved in the Metropolitan Museum of Art in New York.50 Another known set belongs to the Newark Museum. This was bought by the Museum from private art dealer Arthur Leeper, who brought a group of these covers from Lhasa in 1980. He kindly offered for study a few damaged covers from the set the Yongle Emperor commissioned for Kunga Tashi Gyaltsen (1349‒1425), the Tibetan abbot of Sakya Monastery. Although they look as if they were made of one piece of wood, these covers were composed of five elements, a plank and four joints placed into a frame, exactly like

49 A caustic, toxic sap, called urushiol, is tapped from the trunk of the Chinese lacquer tree to produce lacquer. This is done by cutting five to ten horizontal lines on the trunk of a ten-year-old tree, and then collecting the greyish yellow sap that exudes. The sap is then filtered, heat-treated, or colored before applying onto wood or another base material. The resin is naturally dark in color. To obtain such a deep and rich, red coloring—usually cinnabar (mercury sulfide, also known as vermillion) or iron oxide—was added. However, other colors could have been used too. Lacquer can be molded, carved, incised, or used as a fixative for surface decoration. True lacquer objects retain a finish that is resistant to abrasion and corrosive liquids. Since ‘varnish trees’ do not grow in all areas and climates, some lacquer-wares became highly valued trade goods, available only to aristocrats through diplomatic exchange. In areas where varnish trees could not grow, people developed other materials inspired by the elegant look of lacquer. In Tibet, objects were colored red and then coated with a range of oils from a variety of plants to imitate lacquer fin- ishes. In India, stick lac, resin excreted by female scale insects Tachardia lacca, was processed for its naturally rich red color and for its use as clear shellac. Other regions, such as Iran or Burma, developed other materials. 50 Watt and Ford 1991. Watt and Leidy 2005: 55–57. Silk 1996: 153–200. Tibetan Woodblock Printing Culture 143

Figure 75 The fragment of the Yongle cover of Tibetan Kanjur (originally Sera Monastery set) containing a curved seal on its verso side. Additionally as seen here all upper covers feature a summary of contents in Tibetan and Chinese engraved on the verso side. The R.R.E. Collection in Winterthur, Switzerland (ET 107). © Richard Ernst.

the covers of the Wanli Supplement from Harvard-Yenching Library and the Kangxi volume from the Jagiellonian Library in Cracow. I found another cover of the same type in the private R.R.E. collection in Winterthur, Switzerland. This displays the Eight Auspicious Symbols. The center of each cover displays a wish-granting vase filled with three gems. Four meandering stalks emerge from the center to encircle eight auspicious symbols: the wheel of law, victory banner, two fish, treasure vase, parasol, conch shell, lotus flower, and endless knot (Figure 78). This suggests that this particular cover belonged to the Potala set (originally Sakya Monastery set), which has a flaming jewel in the middle flanked by four of eight auspicious symbols, such as the wheel of law, victory banner, two fish, and treasure vase. This is framed with lotus petals and other floral ornaments. The bottom cover looks very similar, but only the four remaining auspicious symbols appear there, such as the parasol, conch shell, endless knot, and lotus flower. Independently of the set, all upper covers feature a summary of contents in Chinese and Tibetan engraved on the verso side. The Sera copy of Yongle Kanjur has a ‘precious pearl in flames’ carved in the middle of the top cover, which size is 72.8–73.3 × 26.4–27.1 cm (Figure 79). 144 chapter 5

Figure 76 The fragment of the Yongle cover of Tibetan Kanjur (originally the Sakya Monastery set preserved later in Potala, Lhasa) containing a curved seal on its verso side. Additionally, as seen here, all upper covers feature a summary of the contents in Tibetan and Chinese engraved and gilded on the verso side. The R.R.E. Collection in Winterthur, Switzerland (ET 87). © Richard Ernst.

The covers of Tibetan Buddhist canonical volumes produced in China were often lacquered and painted. The construction of these covers varies from a single piece of wood to a composite plate with five wooden parts wedged together without nails. Other Tibetan book covers are typically carved, painted, gilded, and sometimes decorated with metal fittings. Tibetan and Nepalese covers were often deeply carved with figural reliefs and gilded. Painted covers frequently display geometric patterns and have some shallow carving. They are usually varnished and appear glossy, although painted with the same mineral pigments as Tibetan scroll paintings. The origin of the wood helps to trace the manuscript’s place of production.51 Each volume usually has more than three hundred leaves, and every leaf is composed of many layers of paper glued together. The size of book leaves

51 Nevertheless, because covers often get separated from manuscripts, occasionally are produced at a date later than the text, are made from wood purchased elsewhere, or are damaged and then renewed, specific manuscript attributions are difficult. Tibetan Woodblock Printing Culture 145

Figure 77a, b Microscopic view of sandal wood in the Sakya type cover of the Yongle Tibetan Kanjur from the collection of Arthur Leeper. a. Radial section; rays heterogeneous with one row of upright and/or square marginal cells; simple perforation plates visible on an image at 100× magnification. b. Tangential section; ray width 1 to 3 cells (dominant 2 cells are visible) at 100× magnification. © Tomasz Ważny. 146 chapter 5

Figure 78 The cover of the Yongle Tibetan Kanjur. This probably belonged to the Potala set (originally Sakya Monastery set) of the Yongle Kanjur, which has a flaming jewel in the middle flanked by four of the eight auspicious symbols, such as the wheel of law, victory banner, two fish, and treasure vase. The R.R.E. Collection in Winterthur, Switzerland (ET 87). © Richard Ernst.

Figure 79 The Sera copy of Yongle Kanjur has a ‘precious pearl in flames’ carved in the middle of the top cover, which size is 72.8–73.3 × 26.4–27.1 cm. The R.R.E. Collection, Winterthur, Switzerland. The R.R.E. Collection in Winterthur, Switzerland (ET 107). © Richard Ernst. Tibetan Woodblock Printing Culture 147 usually differs; it can be another distinctive feature of the technical identity of an edition, but not decisive alone. The examined Yongle leaf from the University of Michigan measuring 24.2–24.5 cm in height and 68.7–69 cm in width was printed in red ink; the same as described in both Lhasa sets. It belongs to a loose leaf book in pothi format. The sizes of leaves of both described Lhasa sets and all fragments of the Yongle, Wanli, and Wanli Supplement are very close, and definitly can be qualified within the same measuring standard of paper size.52 The leaves of the Wanli volumes from Cracow measure 23.8‒24.5 cm in height and 68.5 cm in width. The leaves of all volumes of the Wanli Supplement measure approximately 23.8‒24.5 cm in height and 68.5 cm in width. The leaves of the Berlin manuscript Kanjur are slightly larger than for all other printed sets, and measure approximately 26.8 cm in height and 73.2 cm in width. The leaves of the Kangxi Kanjur fragment measure approximately 24.5‒25.5 cm in height and 71.5 cm in width. The Berlin manuscript Kanjur and then further sets of the Kangxi edition printed from new sets of woodblocks were produced in the second part of the seventeenth century as new projects. The Wanli set is close to the Yongle because it was a reprint. This group of Kanjur sets is distinguished by beautifully painted edges typical only for some of the early Peking editions, such as the Wanli, Kangxi, and the later Peking sets. For example, the twenty-six volumes (i.e., volumes 23–28, 38–46, 48, 50–57, and 60) from the Pander Collection in Cracow are painted on their long edges in the same style, with eleven flowers each containing twelve blue petals and six green leaves (Figure 80). The shorter edge contains the Chinese and Tibetan text completed with a 2.5 cm column ornament along each corner (Figure 81). This ornament depicts Buddhist symbols such as the vase, flowers resembling lotuses, and crossed gems composed in a column form. In the central part of the opposite shorter edge, there is a green halo over the dark blue lotus petals outlined in gold, with a Chinese inscription in gold in the center. The other group, represented by three Pander Pantheon volumes (vols. 47, 49, and 59), shows a kīrtimukha (‘the face of glory’) ornament against a red background on the long edges (Figure 82).53 The short edges were painted in the same style as other Wanli volumes. The long fore-edges of the Wanli Supplement volumes from both the Jagiellonian Library and Harvard-Yenching Library are painted with a flower

52 In fact they differ within a few millimeters in height and one centimeter in length, which I qualified as the same. This little variation cannot be avoided when preparing paper for a hundred-volume set. I observed such differences within the Wanli edition, and volumes of the same set as well. 53 Beer 2003: 78–79. 148 chapter 5

Figure 80 Types of painted long edges in volume 54 of the Pander Collection containing Wanli Kanjur. The same style of painting can be seen in volumes 23–28, 38–46, 48, 50–57 and 60. The Jagiellonian University Library in Cracow (Pander Pantheon).

Figure 81 Types of painted short edges in volume 54 of the Pander Collection containing Wanli Kanjur. The same style of painting can be seen in volumes 23–28, 38–46, 48, 50–57 and 60. The Jagiellonian University Library in Cracow (Pander Pantheon). Tibetan Woodblock Printing Culture 149

Figure 82 Types of painted edges in volume 47 of the Pander Collection containing Wanli Kanjur. The same style of painting can be seen in volumes 49 and 59. The Jagiellonian University Library in Cracow (Pander Pantheon).

ornaments against a red background (Figures 83). There are three flowers, resembling lotuses, each with blue petals and four buds on both sides, and small auspicious symbols in the central area between the flowers. These symbols can be classified into a group that Dagyab has called ‘the Seven Gems’; they were adopted as individual symbols from Chinese art. Certain figures that often recur on Chinese brocades were interpreted by Tibetans as symbols of good fortune and depicted as offerings on painted scrolls and miniature cult pictures for religious use.54 Those depicted on the Wanli Supplement volumes can be described as the Unicorn (bse ru), Earrings of the King and Queen (rgyal po’i rna cha, btsun mo’i rna cha), Crossed Gems (nor bu bskor cha), the Eight- Branched Coral (byu ru yan lag brgyad pa), and Elephant’s Tusks (glang chen mche ba). Additionally, there is a symbol resembling a book and a wheel found on a couple of volumes in this set. The ornament on the corners (1.5–2 cm) is hardly visible; only its green color is noticeable. Chinese characters sketched in a stupa shape or as a Buddha’s halo over the lotus are visible on the short side edge, and Chinese script only on the opposite short edge.

54 Dagyab 1995: 91–93. 150 chapter 5

Figure 83 The long fore-edges of the Wanli Supplement volumes are painted with a flower ornament against a red background. There are three of these flowers, resembling lotuses, each with blue petals and four buds on both sides, and small auspicious symbols in the central area between the flowers. The Jagiellonian University Library in Cracow (Pander Pantheon).

The Berlin manuscript Kanjur edges are painted in a similar style, however in different colors. The edges contain lotus flowers and auspicious symbols on the flowers against a yellow background. Then the short edges feature a precious pearl in flames and Tibetan text in the halo on the lotus flower (Figure 84a, b). Whether they are manuscripts or prints, the edges of Tibetan canonical volumes were usually trimmed and painted in red, yellow, or sometimes even blue. Some scholars argue that this had a very practical purpose because using cinnabar or arsenic sulphide protected the paper from insects. The page image is especially significant—it embraces the readable features from wooden blocks, and allows for grouping and identification. This is why a precisely measured page outline is of vital importance. In general, page outline depends on the size of the paper leaf. All examined Yongle and Wanli set fragments have the same size of leaves and the same outline with text executed within two side margins 59 cm from each other. Within these margins, the eight lines of text aligned to upper lines spaced 1.5–1. 7 cm were carved. On a Yongle/Wanli page, it is possible to observe these eight lines of text in Tibetan dbu can composed to the equal upper line of text (Figures 72a, b and 85). The red ink on a Yongle page however was not thick enough to adhere evenly to the surface of absorbent Chinese paper. This is why it spreads out of the letter outline embossed in the paper in many places (Figure 86). As opposed to the previous sets, the Berlin manuscript Kanjur text is composed in full frame. It, however, is also composed of eight text lines per page with distances of 1.8 cm between lines through all examined volumes (Figure 87). Tibetan Woodblock Printing Culture 151

Figure 84a, b Painted edges of the Berlin Manuscript Kanjur. The edges contain lotus flowers and auspicious symbols on the flowers against a yellow background. a. The short edges feature a Tibetan text in the halo on the lotus flower. b. The short edges feature a precious pearl in flames. The collection of the Staatsbibliothek zu Berlin—Preussischer Kulturbesitz, Orientabteilung, Peking Kanjur, Band 1 (ka, ’dul ba, bzhi). 152 chapter 5

Figure 85 Folio 3 of volume 26 from the Wanli set. Within margins, the eight lines of text in Tibetan dbu can aligned to upper lines in a distance of 1.5–1.7 cm were carved. Collection of the Jagiellonian University Library, Cracow (Pander Pantheon).

Figure 86 A close-up of text in the Yongle folio printed in red ink. The ink has spread from the intended outlines of some of the letters. The slightly rounded parts of letters heads (dashes) forming the bottom of the upper line of letters—in the Yongle/Wanli edition are visible here. The Special Collections Library, University of Michigan, Ann Arbor, Central Asian Collection (No. 1). Tibetan Woodblock Printing Culture 153

Figure 87 The Berlin manuscript Kanjur text composed of eight text lines per page with distances of 1.8 cm between lines in full frame. The collection of the Staatsbibliothek zu Berlin—Preussischer Kulturbesitz, Orientabteilung, Peking Kanjur, Band 1, Blatt 4 recto (ka, ’dul ba, bzhi).

The text starts on folio 1 recto written in four lines of text, then as follows: five lines of text on the verso side of folio 1; six lines of text on the recto, and seven lines of text on the verso side of folio 2; and in folio 3 and the rest of the volume, eight lines of text, with the third and sixth lines written in red. The frame around the written area is sketched with a 0.3 cm black line and a thin red line inside (55 × 18.8 cm). The Kangxi volume features text in a full rectangular frame also printed in red and in similar measurements. There is a variety of shapes of margins and frames in these early editions. The page outline in these volumes does not vary as greatly as in later editions and also in manuscripts of non-canonical texts. All early editions produced in Beijing have eight text lines per page, excepting the Wanli Supplement, which has six (Figure 88). All volumes are printed in dbu can script. However, the shape and proportions of the letters differ regarding the carving style, tool traces, and skills of the carver. I could observe the slightly rounded parts of letters forming the bottom of the upper line of letters—in the Yongle/Wanli edition, the shape of dashes is rounded resembling an upside-down triangle. This is due to the rounded type of chisel used (Figure 89). Then the same parts of headings in the manuscript Kanjur of Berlin and the later Kangxi feature a straight line of text and characteristic ‘spurs’ in the right side of the upper part of letters such as ‘a’ ‘ka’ (Figure 90). The reason for this is obvious for manuscript editions, since a writing stick creates a straight upper line. And this feature was copied very faithfully in later editions of the Kangxi edition, which also display a very straight upper line. However, this kind of fingerprinting can provide information 154 chapter 5

Figure 88 Page image from the Wanli Kanjur Supplement. The collection of Harvard-Yenching Library of the Harvard College Library, Harvard University (TTib 1803.7 13 vol. 1, Da zang jing Ming Yongle ba nian).

about the workshop and area in which the book was made, rather than about the edition—different editions could be printed in the same area or even workshop and not be the same set. The Yongle and Wanli Kanjur sets were sponsored by Emperors of China, and it may thus be assumed that highly skilled craftsmen and artists were chosen for this project. The fact that more technical information in the corpus of the manuscript is written in Chinese than in Tibetan is explained by the books having been printed in Beijing in a Chinese workshop by mostly Chinese craftsmen. Often lacking prior knowledge of Tibetan language and culture, the craftsmen needed more instruction in Chinese to conduct the printing project; for instance, to ensure the proper order of text. There is a Tibetan foliation on the left side margin, only on the recto side of the leaf,55 and Chinese pagination on the right side margin, on the recto and verso side of the leaf.56 The double side- executed leaf was printed from two wooden blocks or one double-side carved block, and without Chinese foliation, Chinese craftsmen would have had a hard time printing the pages in proper order. Previously it was assumed that Tibetan Kanjurs from Yongle to Kangxi were printed in cinnabar, but that the Wanli Kanjur was not. All sets of the Yongle edition mentioned in the literature, and the examined Yongle folio from the University of Michigan, were printed in red ink. There is evidence provided by

55 Folio is a leaf numbered on the recto, or front; foliation is the numeral itself in a foliated book. 56 Pagination is understood as a sequence of figures with which the pages of a book are numbered. These are known individually as page-numerals, collectively as pagination. Tibetan Woodblock Printing Culture 155

Figure 89 Fragment of folio 3 of volume 26 from the Wanli set. Note the slightly rounded parts of letters forming the bottom of the upper line of letters resembling an upside-down triangle visible at some points. Collection of the Jagiellonian University Library, Cracow (Pander Pantheon).

Figure 90 The parts of headings in the manuscript Kanjur of Berlin feature a straight line of text and characteristic ‘spurs’ in the right side of the upper part of letters such as ‘a’ ‘ka.’ The collection of the Staatsbibliothek zu Berlin—Preussischer Kulturbesitz, Orientabteilung, Peking Kanjur, Band 1, Blatt 4 recto (ka, dul ba, bzhi). 156 chapter 5

Helmut Eimer that the Wanli Kanjur was printed in black ink.57 After examination of the Pander Pantheon Collection, I can confirm Eimer’s statement with material evidence. The Berlin Kanjur (manuscript Beck) handwritten with black and red ink is characterized by skilled calligraphy. And the Kangxi edition sets and also later Peking editions are also printed in red. A similar type of paper constituted the Yongle and Wanli editions plus the Berlin manuscript copy of the Wanli edition, and furthermore the Wanli Supplement and fragments from the later Kangxi editions. All mentioned editions were executed on laid paper characterized by narrow laid lines and components typical for China. The preserved leaf of the Yongle set is composed of at least six glued layers of very thin laid type paper made of paper mulberry (Broussonetia sp.) characterized by thirty laid lines in 3 cm (Figure 91). The paper used for volumes of the Wanli Kanjur set is also made of paper mulberry (Broussonetia sp.) fibers, glued together in six or more layers. This means that the paper was done with a dipping technique and by using a papermaking mold with a movable type of sieve. Chain lines were very difficult to see. The paper on which the laudation text is written has about four layers, which makes the leaves thinner. Most of the volumes contain laid paper characterized by twenty-four laid lines in 3 cm. However, some of the papers show the presence of fifteen or eighteen laid lines in 3 cm. Chain lines were hardly visible due to the gluing together of many layers of paper, but were still possible to observe. For example, in volume 28 (eight laid lines in 1 cm) it was also possible to measure the distances between chain lines as follows: 3 cm–2 cm–3 cm–2 cm–2.5 cm–2.5 cm–2.5 cm–2.5 cm. Within the Wanli set, the best quality (longest) fibers were used for volume 60 (dkar chag). The volumes of the rgyud section were produced on medium quality paper. Volumes of the Wanli Supplement are written on paper made of paper mulberry, straw, and bamboo and represent a very thin Chinese type of paper with tiny laid lines visible, glued together in six or more layers.58 Paper of the handwritten Kanjur of Berlin is glued together in two, three, or sometimes more layers. These are characterized by about thirty-three laid lines in 3 cm. The surface of the leaves is highly sized and polished. The examined volume of the Kangxi Kanjur is written on a Chinese type of paper with tiny laid lines visible, glued together in three or more layers. The

57 Eimer 2000: 27–51. 58 I examined paper in the twenty-two volumes from the Pander Collection in Cracow identified as the Wanli Supplement (Pander Pantheon: volumes 1–22), and two volumes of the same supplement found in the Harvard-Yenching Library. Tibetan Woodblock Printing Culture 157

Figure 91 The paper of Yongle folio 12 from the Special Collections Library, University of Michigan, Ann Arbor, Central Asian Collection No.1. This image shows the paper mulberry fibers enveloped by a transparent membrane in 600× magnification.

structure of paper in this volume is hardly visible, but I could identify three types of cells in its components: mulberry paper fibers (cut short), pitted wood/bamboo trachaids or vessels stained blue-gray by Herzberg stain, and narrow straw fibers with pointed ends stained with olive-gray-yellow. However, even within a group of volume sets written on the same type of paper, I could observe differences in the paper’s quality and appearance. For example, the paper of the Yongle and Wanli Kanjur volumes is much whiter and of better quality than that used in the later sets of Kanjur produced in China and characterized by very similar laid lines printed in the paper structure. Within this group, I could distinguish two subtypes. Yongle and Wanli editions belong to the first subtype, and later sets starting from the Wanli Kanjur Supplement volumes represent the second subtype of paper. Leaves in all examined sets from this group were glued in layers. Increasing the thickness of the paper by gluing it in layers was necessary, because thin paper did not allow for creating the thicker leaf of a large format. The leaves of the examined volume of Kangxi, which was not directly modeled on either the Yongle or Wanli editions, are slightly larger than those in the Wanli Kanjur and Wanli Kanjur Supplement, but smaller than in the handwritten copy of the Wanli Kanjur preserved in Berlin. The leaves of the Yongle and Wanli volumes are composed of more layers of paper than those of the Wanli Supplement volumes, the Berlin manuscript copy of Wanli volumes, and the Kangxi Kanjur volume. The pattern here is that the better, much thinner paper in the Yongle and Wanli editions (first subtype) required more layers to prepare a suitable leaf of a large format. A slightly thicker and more absorbent (softer) paper made of mixed components identified in the second subtype did not require so many layers. The Berlin handwritten copy of the Wanli, distinguished by the largest size of leaves compared to all other sets and heavy processing of the 158 chapter 5 surfaces of the leaves, also contained the largest number of layers in single leaves. This was conditioned on different requirements in paper processing for manuscripts. It is why the paper in the Berlin manuscript copy of the Wanli Kanjur is highly sized and polished and after this processing, very heavy. In general, paper prepared for manuscripts needs much more processing, because it needs to be virtually non-absorbent to allow for comfortable writing with a sharp wooden pen. In the paper structure, I could obtain information about the type of sieve attached to the papermaking mold. The chain and laid lines fingerprint pattern suggests that the paper of the Kanjur sets produced in Beijing was produced with a dipping technique and by using a papermaking mold with a movable type of sieve. This type of mold, used in all examined sets of Kanjur from this group and with such fine laid lines (twenty-four to thirty-three laid lines in 3 cm), was not used in Tibet. Its sieve print confirms the Chinese origin of these papers. Chain lines were usually not identifiable due to many layers of paper having been pasted together. I could notice differences between these two subtypes at all levels of my examination. These differences, however, in the quality and appearance of paper derive from the different raw materials used. The Yongle and Wanli Kanjur editions were printed on paper made of plain paper mulberry. Due to extremely long fibers, this plant can constitute very strong and thin paper. The volumes of the Wanli Kanjur Supplement, printed soon after the Wanli edition, already represent a poorer quality of paper. In quality, technology, and raw material, the paper of the Wanli Supplement is similar to the paper found in one examined volume of the Kangxi Kanjur edition described earlier as composed of three types of cells. Plain paper mulberry fibers can make much better quality paper than those mixed with bamboo and straw. The addition of straw and bamboo makes paper softer but less durable. Such paper, however, is especially suitable for printing. The same type of paper when used for writing needs heavy processing, as it is possible to observe in the handwritten copy of the Wanli preserved in Berlin.

Mongolian Kanjurs

Since the seventeenth century, Kanjur sets have been created, edited, and produced in Mongolia. It is assumed that the printed Mongolian Kanjur was modeled after the 1684–92 Peking Tibetan edition. This is why Mongolian Kanjur sets were created soon after the first Peking editions of Tibetan Kanjurs. The Mongolian translation of this vast collection was prepared in 1628–29 by order Tibetan Woodblock Printing Culture 159 of Ligdan Khan (Tib. Legs ldan), the last of the Great Khans of Mongolia.59 There exists strong evidence that the translation of the Kanjur initiated by Liydan Qayan was not the first prepared by Mongolian translators. Six manuscript copies of the Mongolian Kanjur were made at that time, one written in gold, the other five written with black ink. The text of this translation was used for the xylographical edition of the Mongolian Kanjur, which was engraved in Beijing in 1718–1720. Mongolian Kanjur volumes printed from woodblocks are usually in a pothi format. The text is printed on relatively soft paper, by comparison to paper used for manuscripts with highly processed surfaces. Tibetan and Mongolian Kanjur sets share similar size, materials, printing techniques, and layouts despite the fact that the lines of text in Tibetan run horizontally, while the Mongolian text is vertical. For the purpose of this essay I examined three sets of Mongolian Kanjur from the St. Petersburg University Faculty of Asian and African Studies collection and one set from the Berlin State Library, Germany. I started from the first volume of the tantra ka section of the handwritten Mongolian Kanjur, which is supposed to have been created no earlier than 1629. The text is written within a frame measuring 20.5 × 61.9 cm on paper leaves glued in two layers of paper measuring 23.3 × 68 cm. It is laid regular paper characterized by twenty-two laid lines in 3 cm, and with chain line intervals of 3.5–4 cm. The paper is composed of long fibers of paper mulberry and bamboo, which suggests a good- quality paper usually used for printing (Figures 92 and 93). The next examined volume was a tantra volume (E59) from a printed edition of Kanjur from an approximately similar time at the beginning of the seventeenth century (Kangxi times). Text in this set was printed within a frame measuring 17.3 × 55 cm in black and red ink. Two or three layers of paper formed leaves measuring 22.3 cm in height and 65.9 cm in length. These layers were glued with their laid lines perpendicular to each other, which is a specific feature. However, this feature could be a result of accidental, careless, or inadequate skills on the part of the craftsman. A laid regular structure of paper from a bamboo sieve characterized by about thirty-six laid lines in 3cm was identified. Such a fine laid lines sieve is also quite distinctive when compared with Chinese Peking editions of Tibetan Kanjurs. This paper is basically made of bamboo fibers (Figures 94 and 95).

59 Uspensky 1997. 160 chapter 5

Figure 92 Paper mulberry and bamboo fibers composing the paper of the Mongolian Handwritten Kanjur from St. Petersburg University, Faculty of Asian and African Studies collection.

Figure 93 The paper of the Mongolian Handwritten Kanjur from St. Petersburg University, Faculty of Asian and African Studies collection. A large bamboo vessel is visible on an image at 600× magnification. Tibetan Woodblock Printing Culture 161

Figure 94 Bamboo fibers composing the paper of the printed Mongolian Kanjur from St. Petersburg University, Faculty of Asian and African Studies collection (E59). Fibers and large bamboo vessels in polarized light visible on an image at 100× magnification.

Figure 95 Bamboo fibers composing the paper of the printed Mongolian Kanjur from St. Petersburg University, Faculty of Asian and African Studies collection (E59). Fibers visible on an image at 200× magnification. 162 chapter 5

The edition of the Mongolian Kanjur (E2, volume 9) printed in 1714 is approximately one hundred years later than the other two sets.60 Despite the fact that this edition is significantly more recent than that previous tantra volume (E59) from the printed edition of the Kanjur, the calligraphy is very similar in both volumes. The first pages are printed in black and red ink alternately. Both volumes are characterized by the same distances between vertical lines; however, distances between lines on the first pages are different. Also in both volumes, decoration and frame type closely resemble each other. Similarity is also visible in some features of the paper. The texture in both printed volumes is smooth, and its fiber distribution, even. In both volumes, paper is glued in two to a maximum of three layers, and has a laid regular structure characterized by thirty-six or even thirty-nine laid lines in 3 cm. This is an exceptionally narrow/fine laid structure, even as compared to early editions of Peking Tibetan Kanjurs. Thus, this can be a distinctive feature for early Mongolian editions of Tibetan Kanjur. Chain line intervals are 3.5–4 or 5 cm, but are hardly visible due to paper having been glued in layers. This paper is composed of bamboo, straw, jute, and paper mulberry, which is exactly the same composition as in the Berlin manuscript Kanjur produced in 1680. The Mongolian Kanjur from the Berlin State Library has a text on the recto side of folio 1 placed in a doubled frame (every frame is sketched with a double line). Mongolian text is placed vertically in thirty-three lines (from the right side, the first four lines, sixteenth to eighteenth lines, and thirtieth to thirty- third lines are written in red. The outer frame is sketched with a thick (0.6 cm) black line and a thin red line inside and measures 17.4 × 54.6 cm. The type of paper is very similar to that which composes the other volumes described above. Flush-cut leaf edges are in the same dimension of 23.5 × 66.5 cm. In general, the type of paper in Mongolian editions is very similar to that in earlier Chinese Peking sets of Tibetan Kanjurs, which are heavy in weight. However, these are glued in two or three layers, similarly to all examined sets from Kangxi times and from the Berlin manuscript Kanjur. The Yongle and Wanli editions were glued in six or more layers of paper. Paper texture differs between manuscripts and prints. The same distinction between volumes from the Yongle and Wanli and later sets is visible in their fiber composition. First editions were made of paper mulberry; then bamboo, jute, and straws were used along with only small amounts of paper mulberry.

60 No. 19 in Uspensky Catalog; Publisher: Fu Dalai; Heissing Blockdruke No. 26. Tibetan Woodblock Printing Culture 163

Tibetan Kanjurs Produced in Tibet and Tibetan Borderland

Since the seventeenth century, at the same time at which Mongolian editions originated, Kanjur sets have also been produced in Tibet or the Tibetan borderland. A manuscript copy of this line was prepared in the Yarlung valley. This served as a master copy for the edition prepared in ’Jang sa dam (present day Lijiang, Yunnan, China) in the years 1609–14, the first block printed edition published in Tibet. This edition was later moved to the Gelukpa monastery Lithang. That is why this set is commonly known as the Lithang Kanjur. However, I could see this edition only in microform, and I have not had a chance to conduct material analyses for this set. I examined two sets of Kanjur produced in Eastern Tibet. These are the Cone and Derge editions, preserved in the Library of Congress, Washington DC (Figures 96 and 105). The Cone and Derge are located at quite a distance from each other. Additionally, an examined copy of the Cone from the Library of Congress cannot be directly compared to the Derge Kanjur set, since examined volumes were printed much later (in 1926) from original wooden blocks carved in 1721–1731. In all examined volumes, the fore-edges were painted yellow, which is typical for Tibetan book culture and different from early Peking editions. However, an XRF test performed by Adam Cupa at the Nicolaus Copernicus University did not show an arsenic compound having been applied against insects here. The color has been identified as organic dye. The index was added much later, probably recently, with folio numbers typed on thin tissue paper. Text is printed with black ink. A characteristic grayish color appears slightly diluted, letters are not fully and equally printed (the surface is unevenly covered with ink), and the ink is smeared beyond the edges of the letters, which suggests that absorbent paper was used together with ink made with too much water and not mixed well.61 Such a poor effect of printing suggests that the wooden blocks wore out when printing of this set was conducted. The ink in the handwritten folios is different in color from that used for printing. All volumes have wooden covers produced much later (1927) in Shanghai, measuring 56.5–56.7 × 17.8 cm with six longitudinal holes cut in a distance of 1.3–1.5 from the long edges and in a pattern 6.7–18.5–18.5–6.7 cm from each

61 See, for example, folio 2 verso and subsequent pages: There are ink blots outside of the edges of the letters which are clearly embossed into the paper. On the spare sheet attached to the dkar chags, the bottom of the ink pot has left its print, and signs of the blot are visible far into the paper, which is also absorbent with a variety of shades, suggesting uneven concentration of pigment in this ink. 164 chapter 5

Figure 96 Redaction of Cone Kanjur acquired by Joseph Rock in 1928. Among the few extant copies, only this one is accessible. The original woodblocks were destroyed by fire in 1929. Tibetan Rare Book Collection, Asian Division, The Library of Congress, Washington DC.

other, and a plank (cover) thickness of 0.8–0.9 cm, wood cut. Tape is pulled through these holes and around the volume so loose leaves are bound together. The edges of the planks are skilfully worked out, rounded. The surfaces of the planks are protected with varnish. All Cone Kanjur volumes are in a pothi format measuring approximately 18–18.5 cm in height and 56.5–57 cm in length, and have approximately eight text lines per page (excepting the first folios: see particular measurements in appendix 5). Most of the volumes have a spare page for protection, which is also printed, and with additional information on volume number handwritten in Tibetan letters. In all examined volumes on folio 4 and the remaining leaves in the volume, the recto and verso sides of pages contain eight lines of text per page, with distances between upper lines of text of about 1.2–1.3 cm. The frame of the printed area measures 11.4–12 × 47 cm with a 1.5 cm wide frame with foliation on the left side.62

62 Examined volumes repeating the same page outline pattern are: Vol. 1 mdo (sūtra) vol. ka; Vol. 35 ’dul ba (vinaya) vol. ga (3); Vol. 72 rgyud (tantra); Vol. 92 yum, vol. ka (1). Tibetan Woodblock Printing Culture 165

Figure 97 Page outline of the Cone Kanjur folio 1 of volume 92 yum (Prajnaparamita Sutras). Tibetan Rare Book Collection, Asian Division, The Library of Congress, Washington DC.

There is a slightly larger variety within the first 2–3 pages, where text begins on the verso side of folio 1 with 5 lines of text and two xylograph prints (9.5–9.7 × 9.3–9.5 cm) on both the left and right sides of the page (Figure 97).Here two frames were drawn with double lines: the outer frame measures 12 × 47 cm and the inner, 9.3–9.5 × 44.5 cm. Distances between upper lines of text are 1.5–1.7 cm. Then on folio 2, as on the verso side of folio 1, the page outline contains side frames, but without xylograph prints. There are six lines of text with a distance between the upper lines of 1.3–1.4 cm, and a double-lined frame measuring 11.5 × 47 cm. The first pages of Vol. 108 dkar chags (index) differ slightly in page outline from other volumes of the same edition. The recto side of Folio 1 contains a one-line title in a centrally located double-lined frame measuring 29.3 × 2.5 cm, and the verso side features seven lines of text. Text is composed within two frames drawn with double lines in a distance 1–2 mm: the outer frame measures 47.2 × 10.3 cm; the inner, 45.4 × 8.3 cm. Distances between upper lines of text are 0.9–1 cm. Folio 2 contains seven lines of text on the recto and eight lines of text on the verso side, with a distance of 0.9–1.3 cm between the upper lines, and a double- lined frame measuring 47 × 10.3 cm. Folio 3 and subsequent leaves have eight lines of text with a distance of 1 cm between upper lines; the frame of the printed area measures 10.3 × 47 cm, with a frame of 1.7 cm in width with foliation on the left side. There is also some variety in the number of text lines and measurements in the end of the volume (see appendix 5). There is foliation in Tibetan characters on the left margin of the recto sides only. It is composed in a small frame added on the left side perpendicularly to the rest of the text lines. All examined volumes are printed in dbu can script; however, the general character of the letters/script in all Cone volumes is 166 chapter 5

Figure 98 Fragment of text from volume 1 mdo (sutra) of Cone Kanjur. Characteristic ‘spurs’ in the right side of the upper area of letters such as ‘a’ and ‘ba’ in Cone volumes. Tibetan Rare Book Collection, Asian Division, The Library of Congress, Washington DC.

distinguished from script in the other sets of Tibetan Kanjur. I could observe characteristic ‘spurs’ in the right side of the upper area of letters such as ‘a’ and ‘ba’ in Cone volumes. Additionally, the letters in all examined volumes of Cone are characterized by the triangular shape of dashes in the upper line of letters like ‘ca’ and ‘ma’ (Figures 98 and 99). This suggests that a straight type of chisel was used for carving Cone wooden blocks rather than, for example, is the case in the Yongle/Wanli edition, in which the upper area of the letters display round lines/cuts (rounded dashes, often in the shape of a semi-circle). The same area of the letters in one of the Kangxi volumes which I examined in Cracow displayed remnants of the straight line of text. Such straight dashes are characteristic of handwriting with a wooden stick and can be observed also in Beck manuscript editions (the so- called Berlin Kanjur). The edges of the xylograph printing block print were visible on some pages in examined volumes. On the basis of this observation, I could estimate measurements of the Cone wooden blocks as 50.5–54 cm × 12–13 cm. The leaves of the Cone set measure approximately 18–18.5 cm in height and 56.5–57 cm in length (width). The paper of this set is slightly yellowish (cream), Tibetan Woodblock Printing Culture 167

Figure 99 Fragment of text from volume 35 ’dul ba (vinaya) of the Cone Kanjur. Characteristic ‘spurs’ on the right side of the upper area of letters such as ‘a’ ‘ga’ and ‘ka’ in Cone volumes. Additionally, the letters in all examined volumes of the Cone are characterized by the triangular shape of dashes in the upper line of letters like ‘ca’ and ‘ma.’ Tibetan Rare Book Collection, Asian Division, The Library of Congress, Washington DC. pasted in two or possibly three layers. I observed a laid lines pattern characterized by about twenty-seven laid lines in 3 cm.63 Chain lines were hardly visible (Figure 100). A paper fragment with a title (label) printed on a lotus flower in the first volume [ka] of mdo has the same chain and laid lines pattern as that of the paper leaves; here it is possible to observe chain lines in a distance of 3.5 to 4 cm from each other. I could often observe ink blots characteristic of absorbent and soft Chinese types of paper which were not highly sized, with the exception of paper used for handwritten folios. Often the area of the letters printed is not fully and equally printed, and there are ink blots outside of the letters’ edges. Particular layers of paper were glued using diluted starch paste. During examination I found two leaves, folio 244 in volume 35 and folio 289 in volume 92, handwritten on much thicker paper. More layers of paper were

63 Since the laid lines are so tiny that even if we glued two or three layers of paper, we would still see a clear pattern. 168 chapter 5

Figure 100 Cone Kanjur volume 108 dkar chag (index). Paper glued of two or three very thin layers with a laid-lines pattern characterized by about twenty-seven laid lines in 3 cm. Tibetan Rare Book Collection, Asian Division, The Library of Congress, Washington DC.

pasted together to allow for writing with a bamboo or wooden stick. This technique makes paper less absorbent and prevents ink blots. That was usually achieved both by applying more paste and by polishing the surface of the leaf with a stone before writing. I could find traces of polishing on this folio. For prints, such a high content of paste or other glue was not necessary if a balance between thickness of ink and absorbance of paper was maintained. During printing, the pressure usually has an effect for a short time only, exactly when the printer moves a roller down along the wooden block, and then that is it— the page is printed. When writing is executed, the pressure of the wooden pen is stronger at the point, and it takes longer for ink to make a blot if the paper is too absorbent or the ink too diluted. That is why preparation of leaves for manuscripts and prints differed, and this difference is clearly visible in these two volumes, since there was likely a wooden block missing, and printers had to execute folios 244 and 289 as handwritten to retain full textual content in the edition. They were prepared with the same page outline as all other printed Tibetan Woodblock Printing Culture 169

Figure 101 Paper of the Cone Kanjur in volume 1 mdo (sutra) shows the presence of paper mulberry and straw fibers in 60× magnification.

Figure 102 Paper of the Cone Kanjur in volume 1 mdo (sutra). This image shows the paper mulberry fiber enveloped by a transparent membrane in 600× magnification. leaves, and on the same type of paper, but were composed of more glued layers, as described above. The ink is different in color from that used for printing (much blacker). It would not be possible to write with a stick with such diluted ink as that used for printing other volumes. The fiber composition in every one of the examined folios and also in original paper strings from the Cone Kanjur all show the presence of paper mulberry and straw fibers (Figures 101 and 102). I examined three volumes, 1, 45, and 77, from the set of Derge Kanjur produced in 1733 and preserved in the Library of Congress, Washington, DC. This edition contains 100 volumes plus one volume of index. The printing workshop in Derge was built in 1729 for the purpose of printing Buddhist texts, and still is in operation today. The Library of Congress copy was acquired by purchase in the fiscal year 1908/9 through Hon. W.W. Rockhill, American Minister at Beijing. The leaves of this set measure approximately 10–11.5 cm in height and 170 chapter 5

60.5–62 cm in length (width), depending on which part of the Kanjur this particular volume belongs to.64 This set is characterized by five lines of text with distances of 0.9–1 cm between upper lines of text on folio 1. Text is composed centrally within an inner frame measuring 5.4–5.5 × 35.7 cm. There are two xylograph prints on both sides measuring 5.4 × 5.4 cm. A full printing area frame measures 7.4–7.7 × 51.6 cm (Figure 103). After the first two, we observe an increasing number of text lines. On the recto side of folio 2, there are six lines of text with a distance of 0.9cm between upper lines of text and a full frame (outer side) measuring 6.9 × 51.5 cm and drawn with double lines;65 on the verso side of folio 2 and subsequent pages are seven lines of text per page, with a distance of 0.8–1 cm between upper lines of text; this text is composed in a full frame drawn with one line measuring 7.2 × 50.9 cm. Pagination on the verso side is outside of the frame on the left side margin. The Derge has pagination with numbering on the recto and verso sides of leaves in Tibetan letters. The text in this set is printed from wooden blocks with red ink on a Tibetan type of one-layer paper made of Thymelaeaceae family plants (Daphne or Edgeworthia sp.). This strong, good-quality paper is characterized by a brownish color and slightly glossy surface (possibly polished). The thickness of the paper differs in different leaves. This typically Tibetan type contains many fiber bundles and fragments of outer bark in its structure that are visible when examining the paper against light (Figure 104). This feature is caused by an inadequate amount of well-separated fibers. A bamboo-sieve mold characterized by fifteen laid lines in 3 cm and chain lines in a distance of 3–4 cm was used for making the paper in volumes 1 and 45. Chain lines are clearly visible on many leaves within these two volumes. The Tibetan type of paper in rgyud volume 77 is produced by means of a mold with a woven type of sieve. This makes this volume different from the two volumes examined before, which belong to other parts of the Kanjur set. Paper in all three volumes is soft and absorbent, which has resulted in ink showing through both sides of the paper and some blots visible around the letters. The use of such a type of paper also caused the red paint used for the fore-edges to spread along the paper’s edges. I observed a better quality of print in rgyud volumes. This resulted from the use of softer and thicker paper,

64 Examined volumes: Derge Kanjur volume 1 ’dul ba (vinaya) volume ka; Derge Kanjur volume 45 mdo sde (sūtra) part volume ka (1); Derge Kanjur volume 77 rgyud ’bum (tantra) volume ka. 65 On both sides in the frame there are additional areas/inner frames of 1.3 cm in width— there is a numbering of pages on the left side, and the right inner frame is blank. Tibetan Woodblock Printing Culture 171

Figure 103 Page outline of the Derge Kanjur. The Library of Congress, Tibetan Rare Book Collection, Asian Division, Washington DC.

Figure 104 Woven paper made in a floating type of papermaking mold viewed against light. Copy of the Derge Kanjur. The Library of Congress, Tibetan Rare Book Collection, Asian Division, Washington DC.

which was a better match with the ink type. In some areas red ink possibly dried too fast during printing, or the paper surface was not properly adjusted to the ink type. This resulted in many poorly printed letters, their edges often not sharp or with blots around them. Fore-edges were also painted with red paint, which went deep into the paper edge. This again shows that an absorbent paper was used for this edition (Figure 105). Despite the eastern location of the production place, Derge paper belongs to a typically Tibetan type made of Thymelaeaceae family plants such as 172 chapter 5

Figure 105 Copy of the Derge Kanjur. Tibetan Rare Book Collection, Asian Division, The Library of Congress, Washington DC.

Daphne or Edgeworthia sp. with addition of Stellera sp. fibers in some volumes (Figure 106). This paper is much darker (brownish) in color than that of the Cone edition, with many outer bark particles and fiber bundles in its structure. I observed that in creating paper for the Derge Kanjur set, both types of papermaking molds were used. The examined rgyud volume 77 of the Derge Kanjur was printed on paper made by means of a woven type of mold. For the other two examined volumes from other parts of the Kanjur, paper was made on a bamboo sieve characterized by fifteen laid lines in 3 cm. This type of mold, characterized by twelve to fifteen laid lines in 3 cm, was used in Tibetan borderland provinces and in Bhutan. Some of the old Dunhuang rag papers have very similar laid line characteristics. Typically, Tibetan papermaking molds are woven. Paper used for these Cone and Derge editions does not belong to the same type. Cone paper has a typically Chinese characteristic. At the time when Joseph Rock ordered a copy of the Kanjur in the Cone monastery in 1926, it was probably common to order paper from Chinese craftsmen in the region, as it was so closely located to the Chinese border. However, these volume sets differ clearly from the earlier-described Peking Chinese and Mongolian editions. The Tibetan Woodblock Printing Culture 173

Figure 106 Daphne or Edgeworthia sp. with addition of Stellera sp. fibers in polarized light visible on an image at 100× magnification. These fibers compose the paper of volume 77 of the Derge Kanjur. Tibetan Rare Book Collection, Asian Division, The Library of Congress, Washington DC.

size of the leaves is visibly smaller when compared to those of the editions produced in Beijing and described above. Due to their smaller format, the leaves of the Cone set are composed of two to three layers—much fewer layers than the leaves of volumes from the early editions. However, the laid lines pattern in Cone paper is characterized by about twenty-seven laid lines in 3 cm, which is very similar to that in early editions of Tibetan Kanjur sets produced on Chinese paper. However, components of Cone Kanjur paper differ from early Peking editions. However, there are also some similarities between both editions produced in Eastern Tibet, when taking into consideration the style of particular volumes and page outline. For example, the leaves of the Cone and Derge Kanjur sets are visibly smaller and of similar proportions, resembling palm leave proportions (however not that narrow—different from those from the Chinese and Mongolian Peking editions. Seemingly in this region, it was easy to buy paper produced in both China and Tibet. It is also possible that craftsmen were coming from both directions when they were needed to produce the paper for a particular Kanjur printing project. The other sets of Tibetan Kanjur are these produced in the central part of Tibet, in the Lhasa area. These were the Narthang (1730–1732) and Lhasa (1934) Kanjur sets. I examined two volumes from the Narthang Kanjur set preserved in the Berlin State Library and the Library of Congress, Washington DC (Appendices 4 and 5).66 The leaves of this set measure approximately 17–18 cm in height and 61.5–63.5 cm in length. Particular folios are not evenly cut.

66 These were Vol. 1 ’dul ba, ka (1) and Vol. 2 ’dul ba, kha (2). 174 chapter 5

Fore-edges were painted with yellow paint. Since the paper is very absorbent, paint went deep into the paper edge. These volumes were printed from wooden blocks on poor quality paper with uneven thickness. This caused the spread of ink or printed areas without ink. Such poor print quality is caused by very thin and not sized paper and by ink not having been adjusted to such a type of paper. The middle part of the volume was printed on a visibly worse quality of paper, which is why a poor quality of print was observed on many leaves. The frame of the printed area in all examined volumes measures 11–11.5 × 67–68 cm. This edition page outline is characterized by seven lines of text per page, with distances of about 1.5 cm between upper lines of text. This outline begins on folio 5 and continues through the rest of the book. As in other editions, there is variety in measurements and number of text lines within the first four pages. These are related to increasing number of text lines per page, starting from four lines of text on the recto side of folio 1 with 2 cm distance between upper lines of text. This text is composed within a doubled frame which measures as follows: outer frame 11 x 58 cm, inner frame 9.3 × 55.5 cm. The verso side of folio 1 is blank. There are four lines of text on the recto side of folio 2 with a 2.1–2.2 cm distance between upper lines of text. This page includes 3 xylograph prints, and a blank verso. Next, the recto side of folio 3 contains 5 lines of text, with a distance between upper lines of 1.7–1.8 cm. The frame of the printed area measures 11.5 × 58 cm. There are six lines of text per page on the verso side of folio 3 and recto side of folio 4, and seven lines of text per page on the folio 4 verso. This Kanjur was printed with black ink on one very thin layer of paper made of Thymelaeaceae family plants (Daphne or Edgeworthia sp.). The thickness of the paper differs in different leaves. This is a typically Tibetan type of paper with visible fiber bundles and outer bark particles in the structure. Both a woven type of mold (with thick textile as a woven sieve and possibly a finely woven cotton sieve) and a laid mold characterized by about fifteen laid lines in 3 cm were used during the production of this paper. Despite the three different sieve types used, the paper is made of the same raw material. Some of the local papermaking factories used more than one type of mold. The paper surface was possibly polished; but, paper having a glossy surface is a characteristic produced by some Thymelaeaceae plants. Such thin and absorbent (not sized) paper causes many ink blots and a poor quality of printing from wooden blocks. The middle part of the volume was printed on a visibly worse quality paper, and where the paper is printed on both sides, the text is almost impossible to read. Tibetan Woodblock Printing Culture 175

Next I examined two volumes from the Lhasa edition (1934) printed from wooden blocks with black ink and preserved in the Berlin State Library.67 This edition is printed with black ink on Tibetan type paper. The leaves of this set measure 15.5–17.5 cm in height and 62.8–64 cm in length. Particular leaves are not evenly cut, so edges are not even. Fore-edges of volumes are painted with yellow paint. The paint seeped deep into the edges due to the fact that the paper is very absorbent. The frame of the printed area with both text and the little frame with Tibetan foliation on the left recto sides of the folios, depending on volume, measure 11.5–11.8 × 57–57.5 cm. Starting from the verso side of folio 6, seven lines of text are composed within this frame with distances of 1.3–1.6 cm between upper lines of text. The page outline is similar to that of the Narthang Kanjur. On the recto side of folio 1, the title in the frame appears, and the verso side is left blank. Four lines of text are executed on the folio 2 recto with 2–2.5 cm distances between upper lines of text. Between the text, three xylograph prints are incorporated. On the recto side of folio 3, five lines of text with distances of 1.7–1.8 cm between upper lines, and two xylographs were printed within a frame measuring 11.8 × 57 cm. The verso of folio 3 and recto of folio 4 contain six lines of text per page, then seven text lines start on the verso of folio 4 and continue through the rest of the volume. This edition is written on a good-quality, one-layered, very thin paper made of Thymelaeaceae family plants (Daphne or Edgeworthia sp.). This paper is a typically Tibetan type with irregular thickness and visible fiber bundles in the structure. However, quality differs within this volume, as is observable when looking at the color of the paper: some leaves are almost brown, and I could discern more particles of outer bark in their structure. Both a woven type of mold and a laid mold characterized by about twenty-one laid lines in 3 cm were used for producing this paper. Independently of the two types of papermaking sieve used, the paper represents the same Tibetan type, with a slightly glossy surface, as that used for printing Narthang Kanjur. The woven mold had a sieve made of loosely woven textile, which is clearly visible in the paper structure. The paper is also soft and absorbent, and as a result, the yellow paint used was observed to have spread along the paper’s edges. All examined volumes of the Narthang and Lhasa editions were printed on the same, typically Tibetan paper made of Thymelaeaceae family plants (Figures 107 and 108). All volumes from this group were printed on paper of uneven thickness, with many outer bark particles and fiber bundles in its

67 These are Vol. 1 ’dul ba, ka (1) and Vol. 2 ’dul ba, kha (2). 176 chapter 5

Figure 107 Daphne or Edgeworthia sp. fibers in polarized light visible on an image at 600× magnification. These fibers compose the paper of the Narthang Kanjur from the Berlin State Library.

structure. The quality of Narthang Kanjur paper is not as good as that of paper in the Lhasa Kanjur. However, in both editions the quality of printing is worse than in the case of all previously described editions. This is caused by features of Tibetan paper, which is very durable but with a slightly glossy surface that is not as absorbent as that of Chinese papers which, by contrast, are perfect for writing, especially when glued in layers. All sets of the Kanjur in this group were made of one layer of paper. If additional glue were applied to join the paper layers, that would make a perfect material for writing on, but one which is more difficult to print on. This is why sometimes Stellera sp. fibers were added to papers made of Daphne and Edgeworthia sp. Such an addition allowed for making softer paper which absorbs ink better. This kind of softness was more easily achieved in Peking editions by adding straw or bamboo fibers to paper mulberry fibers. In this case, paper mulberry gave strength to the paper and straw, especially, made it softer and more absorbent.

Patterns of Physical Features in Selected Editions of Tibetan Kanjur in a Regional and Temporal Perspective

All examined Kanjur volumes were distinguished by their large size. However, as described above, a variety of sizes and visual appearances of paper were documented. Particular volumes differed in number of paper layers glued together, character of the paper surface, type of ink, page outline, and quality of the materials. I observed that paper was prepared differently for handwritten manuscripts and for xylograph prints. I found that the technology of paper production as determined from the papermaking sieve pattern sealed in the paper structure, and from the kind of plant used for this production and iden- Tibetan Woodblock Printing Culture 177

Figure 108 Daphne or Edgeworthia sp. fibers in polarized light visible on an image at 600× magnification. These fibers compose the paper of the Lhasa Kanjur from the Berlin State Library.

tified during microscopic examination, is essential to know since such information allows for creating an objective typology. Most of the early Chinese Peking editions of Tibetan Kanjur and Cone print have eight text lines per page. Then the editions of Kanjur from Narthang, Derge, and Lhasa have seven text lines per page. I could distinguish two main types of paper used in examined sets of Tibetan Kanjur. The main differences (fingerprinting information) found between these two groups are expressed in fiber composition and traces of the type of papermaking sieve sealed in the paper structure. The identification of fiber composition allows for distinguishing paper produced in Tibet from that produced in China. In Tibet, primarily Thymelaeaceae family plants were used for making paper. This makes Tibetan paper very distinctive from Chinese, which is composed of varieties of plants such as paper mulberry, bamboo, and straw, among many others.68 Regarding regional origin, all Kanjur sets produced in Beijing were executed on typically Chinese paper, and all sets produced in the Lhasa area were printed on the Tibetan type of paper made of Thymelaeaceae family plants. Plants of this family were widely used in Tibet, and they were not used in China for production of paper to be used in books. In Eastern Tibet, both types of paper were available and used. I could observe the same split in paper features when I examined fingerprinting patterns of papermaking sieves used. The majority of Tibetan papers were made with a woven type, and all Chinese papers were characterized by about twenty-four to thirty laid lines in 3 cm. In examined Kanjur sets, Tibetan

68 For more information on Tibetan and Chinese papermaking, see: Meisezahl 1958: 17–28; Trier 1972; Elaine Koretsky 1986; Helman-Ważny 2005: 27–37; Helman-Ważny 2006: 3–9; Hunter 1932; Hunter 1978; Tsien 1973: 510–519; Tsien 1985; McClure 1986. 178 chapter 5 types of paper could be produced by means of both mold types—woven and movable bamboo sieve. However, there is a clear difference between regions when identifying the laid line number in 3 cm. This is why the pattern of the papermaking sieve sealed in the paper structure can provide significant fingerprinting information for Kanjur fragments’ region of origin. Additionally, the comparative examination of different Kanjur sets discloses the technical similarities between different editions, and their reprints and re- editions, when supported by research on paper with examination of other physical features of particular volumes. The relation between a master copy and the subsequent edition of the Kanjur that was directly modeled on this master copy is also characterized by paper features, as was clearly visible in the Yongle edition and its Wanli reprint and consecutive re-editions. In cases when production of printed editions alternated with manuscript editions, the inheritance of particular features is shown in the paper as well. I could observe that raw materials changed with time, and techniques of papermaking also evolved. This research appears to be very promising for identification of newly found fragments of Tibetan Kanjur, which are still being widely discovered. These varieties were closely connected to a wide-ranging geographical area in which these books were produced. This is why, in principle, the results of my paper examination match the regional production of paper. More importantly, however, my documentation of book and paper features in known and date- able sets of volumes allowed me to build a solid preliminary chronological and regional reference for future work. chapter 6 A Survey of Tibetan Paper

History of Paper in Central Asia and Tibet

The historical origins of Tibetan papermaking are difficult to determine. Our knowledge about the invention and transmission of papermaking in Central Asia and Tibet in the first millennium is very selective and fragmentary. Vorobieva-Desiatovskaia dates the beginnings of papermaking in Tibet back to the eighth century,1 while the Tang Annals mention an even earlier date, 648, in the report of the Tibetan emperor Songtsen Gampo’s request of paper, ink, and other writing utensils from the Chinese emperor. Yet, until the middle of the eighth century, most Tibetan official documents were written on wood. The entry for the years 744–745 in the Old Tibetan Annals records the transfer of official records from wooden ‘tallies’ (khram) to paper.2 Thus, it appears that by the time writing appeared in Tibet, the technology of papermaking was already known in the Far East and Central Asia. The invention of paper is traced back to China. The year 105 CE is often cited as the date for the inception of paper technology. According to historical records, the technique of making paper was reported to the Eastern Han Emperor Ho-di by Marquis Cai, an official of the Imperial Court. However, archaeological records contradict this claim, suggesting rather that paper had already been known in China in the second century BCE. Soon after its invention, paper was widely used in China and spread to the rest of world via the Silk Road. In the east, the art of papermaking reached Korea, where paper production began in the fourth century CE. According to traditional accounts, sixty years after Buddhism was introduced in Japan, a Korean monk named Don-cho brought papermaking to Japan, sharing his knowledge at the Imperial Palace in ca. 610 CE. Archaeological records suggest that along the Silk Road, paper was introduced to the Xinjiang area soon after its invention. Crucial for the later dissemination of papermaking to Tibet, Nepal, Bhutan, Burma and Thailand was the migration of Chinese communities that prepared paper for their own needs. The development of papermaking in these countries was spurred by Buddhist monks who copied vast religious literature. The westward spread of papermaking through Chinese Turkestan along the Silk Road

1 Vorobieva-Desiatovskaia 1988: 334. 2 See: Uebach 2008: 57–69; Dotson 2009: 52–53, 124.

© koninklijke brill nv, leiden, ��4 | doi ��.��63/9789004275058_��7 180 chapter 6 has been widely investigated, but its migration to the south towards Himalayas including Tibet and Nepal to India remains unrecognized. Although paper had already been widely known by then, most of the Chinese documents discovered in Nia in Kroraine (dating from the third and fourth centuries) are written on wooden tablets and bamboo sticks and silk, which continued to be used alongside paper for a long time. Manuscripts written on paper found at sites in Kaochang, Loulan, Kusha, Kotan, Dunhuang, and Turfan date as early as the third century. The technology eventually reached Tibet most probably around 650 CE and from there spread to India. Thousands of manuscripts written in Tibetan language from before the tenth century have been discovered at Dunhuang in the Gansu province. Most estimates date these manuscripts to the time of Tibetan occupation of Dunhuang that occurred in c. 781–848 CE, but recent research by Géza Uray, Takeuchi, as well as Dalton, Davis, and van Schaik has dated the large portion of these manuscripts to the tenth century.3 According to written sources, after 751 CE, paper production spread westward to Arabia via Samarkand. There is a story about Chinese papermakers who were captured by the Arabian army as prisoners of war in the battle of Talas and later settled in Samarkand. The Arabs learned the craft from Chinese prisoners and built the first paper industry in Baghdad in 793 CE Yet this account oversimplifies the matter—in fact, papermaking may have been practiced in Samarkand decades before that battle. Around that time, papermaking spread west of the Pamir, but probably not because of one singular event but rather through a gradual transmission by many routes. The craft continued to spread gradually from Islamic Asia to Europe and, from there, around the world. By the seventh century, papermaking was already a highly refined art and it may be asserted beyond doubt that paper had always been a material of great value. For example, in Tibet, paper had never been used for mundane purposes such as writing personal notes. Because of its costliness, it was important to find ways to save it. One of the methods of saving paper and ink was using small wooden planks or tablets (varying in size, but usually about 35 × 15 cm) with a hollowed out black-inked writing surface instead of paper as described in chapter 3. We do not know if these few early dates singled out from written records are truly the turning points of paper history. There exist no written records of papermaking in Central Asia in any of the languages of the region. Yet what we have available are books and documents from that time. This chapter surveys the types of information these sources provide. By expanding the scope of inquiry from the study of their textual content of these documents to an exploration of their material characteristics, we move beyond legends to a

3 See: Uray 1988: 515–528; Takeuchi 2012; Dalton et al. 2007. A Survey of Tibetan Paper 181 more precise understanding of the history of papermaking, its spread, and its defining developments. As a rural and regional tradition, Tibetan papermaking features abundant local plant and fiber resources. Papermakers developed methods that reflected the unique qualities of their raw materials. As a result, the paper products used as writing supports for Tibetan books over centuries have distinctive properties and features; the place of origin of particular papers used as the support for texts is traceable when compared to the distribution of papermaking plants in the Himalayas and Central Asia. The demand for paper usually inspired the creation of workshops; training sessions were most likely dispersed as the demand was not limited to one central location. Although the data is scarce, we may surmise that many small paper workshops existed in the Himalayan region, sometimes limited to one family, sometimes involving an entire village, and, in exceptional cases, extend to an entire region; as a result, particular families, villages, or regions became famous for paper production. Every large project, such as editing the Tibetan Buddhist canon stimulated the paper industry. The largest monasteries involved in book editing must have stimulated the growth of local paper man- ufactures. Unfortunately, there exist very few historical documents that could support this hypothesis about the state of papermaking; moreover, even these few available resources have not been explored with respect to papermaking. Yet, as Holmberg et al. argue, the eighteenth-century villagers of Bomthang, Nepal had an obligation to supply paper to the government administration offices of the Ranas in Kathmandu.4 The village archive contains information about the regions of raw-material gathering, granted officially every few years depending on harvesting needs. This, as many other papermaking reports, shows that paper could be traded for long distances, but harvested raw material has always been exploited locally. This is why the best way to determine the place of origin of a paper or book is to link it to local plant resources. This administrative aspect of paper production ought to be studied more closely in Tibet. It must be noted, however, that in today’s China, finding Tibetan village archives might be an extremely difficult task. However, despite the fact that during and after the Cultural Revolution papermaking gradually died out, the tradition of making paper by hand is still an important part of Tibetan material culture, and it has recently been revived on the basis of local tradition. In the absence of any systematic study and the fact that such a local papermaking tradition constantly disappears, existence of papermaking

4 Holmberg et al. 1999: 44–45. 182 chapter 6 workshops in Himalayas in specific regions of Tibet should be documented as soon as possible. A number of reports on papermaking methods in Tibet and the surrounding regions are available.5 Perhaps the earliest is the account by the British official Samuel Turner of papermaking in Bhutan in the mid-eighteenth century.6 The recently reported large centers of papermaking in central Tibet were near Dakpo, bordering on Kongpo, in the district of Nyemo west of Lhasa, and near Gyantse. In Eastern Tibet they were originally in Kham, specifically near Derge (sde dge), Pelyül (dpal yul) and Garzé (dkar mdze), in Pemakö (pad ma bkod near present-day Loyül and Metok), in Dagyab near Chamdo, and in the Kongpo region.7 Such reports are very informative concerning the methods and technology of papermaking, as those have not changed since the beginning of papermaking to such a great degree as, for example, in Europe or in China. Documentary materials found in remaining archives of Himalayan villages would be an extremely valuable source of knowledge about local paper production—they provide information essential to tracing a past history, clarifying, among other things, the economic factors shaping the production. Nonetheless, even without access to these archives, we can still reach the past through the study of material artefacts such as books and documents written on the paper. Careful attention to paper in microscopic scale reveals the secrets of its production. By characterizing the paper optic and identifying the fiber composition, we may recover the history and geography of papermaking. In the Tibetan culture of papermaking, the variations of production methods, raw materials, and treatment of the paper surface may allow us to recover the time and place of production. The books written in Tibetan language are composed of a variety of materials, not all necessarily from Tibet itself. The results of paper analyses can help us recognize their regional origin, date them or reveal links between groups of objects with the same distinguishable features. However, it is necessary to deepen our current understanding of Tibetan paper technology, going beyond simplistic statements about the origins and category ‘Tibetan paper’ itself, which is too general to cover the vast territory of the Tibetan cultural realm and the variations in materials and technology available to people within it.

5 Tschudin 1958; Sandermann 1968; Rischel 1985; McClure 1986; Koretsky 1986; Helman-Ważny 2001. 6 Turner 1800: 99–100; Schaeffer 2009: 8. 7 Weber 2007: 111. A Survey of Tibetan Paper 183

Himalayan Papermaking Plants and the Localization of Tibetan Paper and Book Origins

Tibetans created a unique type of paper but their technology came from the Chinese. The high altitude of the Tibetan Plateau and the extremes of its climate make the vegetation distinctive from all other areas of Asia. In other words, the range of plants available for papermaking in Tibet was significantly different than in other places. As far as the physical conditions are concerned, the world’s highest mountain is just one of the Tibetan distinctions. The average altitude is about 3,000 m in the south and 4,500 m in the north. Several major rivers—such as the Yangtze, the Yellow River, the Indus River, the Mekong, the Brahmaputra River, the Ganges, the Salween and the Yarlung Tsangpo River—have their source in the Tibetan Plateau (mostly in present- day Qinghai Province). The climate is severely dry nine months of the year, and average annual snowfall is only 18 inches, due to the rain shadow effect whereby mountain ranges prevent moisture from the ocean from reaching the plateaus. Western passes receive small amounts of fresh snow each year but remain traversable all year round. Low temperatures typical throughout these western regions, poor vegetation beyond the size of low bushes, and wind swishing unchecked across vast expanses of arid plain determine western Tibet. The Indian monsoon affects eastern Tibet; northern Tibet is subject to high temperatures in the summer and intense cold in the winter. All these geographical features have influenced greatly all aspects of people’s activity and life, including vegetation and, concomitantly, the production of paper. The distinctive features of Tibetan papers are their texture, thickness and color, which result from the technology of sheet making and writing surface preparation, and, to no lesser degree, from indigenous Tibetan and Himalayan raw material. The latter factor allows us to distinguish between Tibetan and Chinese types of paper through analyzing paper fibers in ancient books. Since the earliest stages in the development of the craft, Chinese papermakers had used all manner of plants suitable for papermaking. They developed most affordable methods and found the most economical materials in the time and area.8 According to written sources, hemp was the earliest known plant known to have been used for papermaking in China before the Christian era; the paper of Cai Lun was in fact distinguished according to the material used as paper made of tree bark, remnants of hemp, rags of cloth, and fishing

8 Tsuen-Hsuin Tsien offers an exhaustive account of the development in usage of raw materials in China. See: Tsien 1973: 510–519. 184 chapter 6 nets.9 This was followed by paper mulberry from the early second century CE. Rattan was especially popular for making the best paper in southeast China from the third to about the twelfth century, when the supply of the raw material was exhausted. Bamboo then gradually replaced both rattan and hemp as the chief material for papermaking since the latter part of the eight century. In the eighth century CE the Chinese were using a variety of raw fibers that included the rags of hemp and flax (ropes), bark of mulberry, a ramie called Chinese grass and several other types of grasses. In addition, other materials were being used in making special kinds of paper in different parts of China, such as rice and wheat straw, the bark of sandal wood and other trees, stalks of hibiscus, seaweed, and certain other plants were also used. It remains unre- solved whether silk or cotton have ever been used. However, Berthold Laufer mentions that before Cai Lun’s time, as early as third century BCE, felt-like paper was being made by the Chinese from ‘silk-waste,’ which Cai Lun supposedly replaced with vegetable fibers.10 Notably, Tsien, Laufer, and others fail to mention plants from the Thymelaeaceae family—among them Daphne, Edgeworthia or Stellera sp.— as significant resources used in China for making paper. Yet Floyd Alonzo McClure writes about a kind of paper called ‘cotton paper’ (Min Chi), which was supposedly made of ‘cotton paper plant,’ Edgeworthia gardneri or Wikstroemia indica (Min Chi Shue).11 It is not clear and rather baffling why paper made from plants of the Thymelaeaceae family was called ‘cotton paper.’ McClure opines that this type of paper was unique, made in the Kiangsi province on the ‘primitive woven screen’ and used principally to make firecracker fuses. The only kind of paper that resembles ‘cotton paper’ is made of Stellera sp. fibers, which, due to the flat and ribbon-like shape of fibers in cross-section, looks very similar to cotton fibers. Nevertheless, this explanation does not clarify the puzzling terminology sufficiently. Originally, Tibetan paper was made from shrubs belonging to the Daphne and Edgeworthia species (shog shing or dung loma in Tibetan); these plants still provide the basic materials for paper made in the Himalayan regions (Figure 109). However, very special Tibetan paper was also made from the roots of both the Stellera chamaejasme species (re lcag pa in Tibetan) and, rarely, Euphorbia fisheriana or Oxytropis sp.12 The Stellera is a small genus of less than ten species, found growing in comparatively dry conditions on sunny and sandy slopes in

9 Laufer 1931: 15. 10 Laufer 1931: 12. 11 McClure 1986: 37, 80–81. 12 Trier 1972: 56; Dawa 1999: 156–159, 320–321. A Survey of Tibetan Paper 185

Figure 109 Edgeworthia gardineri growing in the Khumbu region, Nepal. 186 chapter 6

Figure 110 Stellera chameajasme found in Kyirong, Western Tibet. areas such as Central Asia, Iran, Bhutan, Mongolia, Nepal, Russia, and parts of Tibet.13 It is widely distributed along the Himalayan range, where it is found at altitudes of 2,700 to 4,500 meters (Figure 110). This plant was reported first by Nikolai Przhevalsky in 1873, and subsequently by Hossie in 1910, along with some other ‘plant hunters,’ botanists and geographers.14 The production of paper from these roots is practiced in the highest places in the world, where practically nothing else grows. The altitude range of Daphne reaches 3600 m above sea level so the root paper production is practiced in the highest places in the world. This plant is closely related to Daphne; however, whereas the bark from branches of Daphne is used for papermaking, it is the root bast of Stellera chamaejasme that is the raw material (Figure 111). Although the quality of root paper (especially from Euphorbia) is not as good as the bark paper, the presence of poison in the paper makes it resistant to damage caused

13 The most common places of its occurrence in China according to the Harvard University Herbaria are: Gansu, Hebei, Heilongjiang, Henan, Jilin, Liaoning, Nei Mongol, Ningxia, Qinghai, Shaanxi, Shanxi, Sichuan, Xinjiang, Xizang, Yunnan. See: Flora of China, http:// www.efloras.org/florataxon.aspx?flora_id=2&taxon_id=200014523 (January 15, 2011). 14 Information from Herbarium of L.H. Bailey at Cornell University, Department of Botany, College of Agriculture, Ithaca NY, USA (sample of Stellera chamaejasme collected on June 10, 1922); see also: Kingdon-Ward 1934: 64. A Survey of Tibetan Paper 187

Figure 111 The root of Stellera chamaejasme shown by a former papermaker in Dobe Shang, Western Tibet. 188 chapter 6 by insects and, ultimately, more lasting than other types of paper. The root of the Euphorbia fisheriana is thought to be best when it grows in the soft sandy soil of a black beach. These root bast fibers create a very specific soft type of paper, which is considered to be of lower quality than bark paper. The roots are especially difficult to harvest, which places a serious limitation on quantity of paper that can be produced. They additionally require a longer time for processing, and an extra step in the papermaking process. Thus in general they are only used in papermaking when other sources are not available. On the other hand, the poisonous properties of these plants make the paper resistant to damage caused by insects, meaning that it may ultimately be more durable than other types of paper. In principle, it should not be difficult to determine the relative importance of each of these different fibers in the early history of Tibetan papermaking. Stellera chamaejasme fibers are distinctive in the fiber examination of historic papers, and can be clearly differentiated from Daphne and Edgeworthia despite the fact that all plants belong to the Thymelaeaceae family. Species of Oxytropis (Leguminosae) were also used as an addition for making paper in some regions of Tibet. However, its anatomical features in paper have not been described yet. In Nepal, Daphne sp. (D. cannabina, D. bholue and D. involucrata) and Edgeworthia gardineri were used and mostly known under the local names lokta or kāgate (Figure 109).15 Jute (Corchorus sp.) was also used in Nepal; however, as a raw material for paper making, it is most typical of India, where sunn hemp (Crotalaria juncea), roselle (Hibiscus sabdariffa), and ramie (Boehmeria nivea) were also used. In Kashmir, the waste cocoons of silk were used to make a paper that is smooth, soft, and strong.16 Kashmiri paper was considered the finest among Indian papers and was a prestigious gift exchanged by sultans.17 In Japan, the papers made from Edgeworthia papyrifera and Wikstroemia sikokiana were known under their Japanese names mitsumata, and gampi, respectively. According to written sources, the oldest usage of mitsumata in Japan dates back to 1591 CE.18 Both of these plant species belong to the Thymelaeaceae family and are very similar to those identified in the paper of Tibetan and Nepali books. The Thymelaeaceae family includes shrubs or small trees that are rarely herbal, evergreen, or deciduous. This family of plants is characterized by a hard, fibrous bark. Because the phloem contains very strong

15 Trier 1972: 50–59; Holmberg et al. 1999: 47. 16 Agrawal 1984: 132–137. 17 Field 1995: 24. 18 Bell 1981: 77. A Survey of Tibetan Paper 189 fibers, the bark of many species is suitable for the manufacture of high-quality paper such as that used for banknotes. The stems are extremely supple and difficult to break and are commonly used as a substitute for string. Most species are poisonous; some have medicinal qualities.19 From the Tibetan perspective, there are three different classes of material for manufacturing paper: the best material (shog shing or dung loma), the mid- quality one (a ga ru) and the lowest quality (re lcag pa).20 These plants used in the manufacture of paper of the other classes all belong to the Thymelaeaceae family. Because of that, fibers in ancient papers are extremely difficult—sometimes impossible—to distinguish. The material described by Tibetans as the conch shell tree dung loma (which is supposed to be Daphne sp. [D. bholue?] plant), is considered to be the best material for manufacturing paper in Tibet. The phloem of Daphne and Edgeworthia shrubs is the most common material in the foothills of the Himalayas.21 The bast of Edgeworthia is usually whiter and softer than that of Daphne, but paper made of Daphne generally is considered the best. Due to its length and flexibility, the fibers of Daphne sp. allow for easy defibrilation and for control over the thickness of the paper sheet during the papermaking process—allowing for the production of thin sheets. Yet, Tibetan paper could hardly be described as thin. Quite the contrary—thickness was one of its distinctive features. The high resin and starch content of the inner bark tissue of Thymelaeaceae family plants often yielded a slightly glossy surface on the final paper product. The paper made from Daphne is also known among Tibetans as dug shog, ‘poisonous paper,’ because the bark of this tree contains a substance that repels insects.22 The bark is naturally resistant to insects and deterioration and is even thought to have medicinal properties.23 Among its other uses, Daphne is recommended for fighting high fever. It is also worth noting that Tibetans determined the quality of their paper, among other things, on the basis of the quality of the soil in which the plants grew. Following those standards, for the best quality Daphne paper, the tree should not grow in eroded soil, among juniper trees, in clay, or among boul- ders. The ‘paper tree’ which grows in sand or in red slate is considered to be of slightly lesser quality because it contains hard (i.e., difficult to cook)

19 Flora of China, http://www.efloras.org/florataxon.aspx?flora_id=2&taxon_id=200014523 (accessed January 15, 2011). 20 Tsundru 2010: 54. 21 For a detailed discussion of characteristics of plants used for papermaking in Nepal and other references for papermaking in Central Asia, please see Trier 1972: 50–59. 22 Imaeda 1989: 410. 23 Upreti 2004: 236–243. 190 chapter 6 substances and it can be recognized by its distinctive yellowish color. According to Tibetan papermaking standards, the poorest material for making paper is a tree that has thick bark or is infected with insects. Much more rare paper made of eaglewood bark Aquilaria agallocha (aga ru) also from the Thymelaeaceae family (genus: Aquilaria, species: agallocha) is considered medium quality.24 The main feature of the paper made of agalloch eaglewood bark is its thickness, which is said to be the best for traditional Tibetan scriptures (dpe cha). This plant is very rare; primarily native to the mountains of Silhet and some eastern provinces of Bengal, it later spread to other parts of southeast Asia, such as Vietnam, Malaysia, Thailand, Laos, Indonesia, and parts of India. The bark of this tree was also used in unprocessed form sometimes, similar to palm leaves. Aquilaria trees reach 120 feet in height with trunks with twelve foot circumferences; they can live up to thousand years. Its wood is pale white in color, light, soft, and porous, and contains the precious oleoresin, known in the East as aggur or uggar, which forms dark-colored veins of oleoresin within the wood, in the interior of old trees.25 Aquilaria agallocha or agarwood is considered a pharmacological product formed as the result of a fungus disease resulting from wounds on the trunk or by insect activity.26 The oleoresin is procured by crushing these portions of wood in a mortar, and boiling them in water, which makes the aggur rise to the surface. Traditionally old-growth Aquilaria trees were indiscriminately cut down for the oleoresin, but only a few very old trees actually contained the substance deep inside them. Today in many countries of southeast Asia where the tree was once native, it has become very rare because of intensive harvesting. Oleoresin is extremely valuable because of its significance for a variety of Buddhist and Islamic ceremonies as well for traditional medicine, perfume production, incense, and for the beautiful grain of its wood, which takes a high polish and was used for setting precious stones. Under its pharmaceutical name Lignum Aquilariae, Aquilaria has been also used to enhance cerebral function, balance the mind/body connection and the nervous system. Its anti-asthmatic, antimicrobial, digestive, and stimulant properties are also well documented. Aquilaria has been used in numerous kinds of medication

24 The plant is known under its common names, such as Agar wood, Lign-aloes wood, Eaglewood, Agalloch and also Lignum Aloe, Aguru, chenxiang (沉香), Ch’Ing Kuei Hsiang, Chi Ku Hsiang, Huang Shu Hsiang. 25 See specimen collected in 1927 by J.A. Lörzing preserved in Herbarium of L.H. Bailey at Cornell University, Department of Botany, College of Agriculture, Ithaca, NY, USA. 26 Aloeswood extract paint: file entry for characteristics of aloeswood, http://www.aloeswood.com.cn January 2009. A Survey of Tibetan Paper 191 as well as in insect repellants.27 Works from the seventh century CE such as Xuanzang’s travelogues and the Harshacharita, composed in Northern India, describe the use of agarwood products such as ‘Xasipat’ (writing-material) and ‘aloe-oil’ in ancient Assam (Kamarupa). Today, Aquilaria is not typically mentioned as a raw material for making paper, probably because the tree has become too rare and valuable. However, it is possible that the tradition of making writing materials from Aquilaria bark is still practiced in Assam. As far as Tibetan books are concerned, I have not detected Aquilaria fiber in the manuscripts I have examined.

Papermaking Methods in Tibet Traceable in Books

The actual method of making paper in Tibet seems to have evolved very little over the centuries, with each sheet of paper dried on individual molds (Figure 112). This mold type is called ‘floating’ because it is placed on a water surface such as lake, pond, river or puddle. The other main type of paper mold is usually known as a ‘dipping’ mold, and is thought to have developed subsequent to the floating mold. The dipping mold allows faster paper production because it is possible to remove a wet sheet of paper directly from the sieve just after its shaping. This means that papermakers do not need to wait until the paper has dried before re-using the mold to begin the next sheet. The main difference between the two types of mold is in their construction. The floating mold is built of wooden frame with a woven textile attached to it. In the dipping mold, on the other hand, a movable sieve made from bamboo, reed or another kind of grass is attached to the wooden frame. Modifications to this technology have been reported, including a floating mold resembling a wooden box with movable screen. In this case I would continue to classify this type of paper as laid based on its sieve print. Independent of the particular technique used for sheet formation, every papermaking sieve will make an impression that is specific to the construction of the mold and sieve. This impression is unaffected by most aging processes, and can be read centuries later. The impression of a textile sieve made of cotton, hemp, or flax can be distinguished from the impression produced by a movable sieve from a dipping mold made of bamboo, reed or other grasses. However, to date or identify the origin of a particular paper, this information alone is insufficient. According to Dard Hunter, in the southwestern regions of

27 Raintree Nutrition (Tropical plant database): file entry for Aquilaria (Aquilaria agallocha), http://www.rain-tree.com/aquilaria.htm (accessed January 2009). 192 chapter 6

Figure 112 Drying process during papermaking in the Kullu Valley, Himachal Pradesh, 1940s. The papermaking molds with newly made sheets of paper are left until the sheets are dry. Photo 523/1(154), © The British Library.

China and in the Himalayas, the floating mold was the often used; in more eastern regions, the dipping mold with a bamboo sieve was developed.28 However, both types of molds were used simultaneously during the first millennium for making paper along the Silk Road, and we have to take into consideration that papermaking workshops were not limited to either method and could have employed both types of mold.

Papermaking Tools and Technology

Traditional papermaking in Tibet began to disappear in the 1950s when Tibet was annexed by China and modern paper mills were built in Lhasa. Papermakers lost their customers and production was practically stopped. Additionally, many kinds of paper from China and the West were introduced in Tibet and became widely available. The traditional craft was, nonetheless, popularized by emi- grant papermakers in other countries such as Nepal, Bhutan and India, where traditional papermaking techniques are still preserved. Another factor that has limited hand-made paper manufacture since the 1950s is the development

28 Hunter 1978: 84; Schaeffer 2009: 8. A Survey of Tibetan Paper 193 of machine-made paper technology, which almost entirely superseded handmade paper not only in Tibet, but also in other Himalayan regions. Tibetan papermaking technology had always been very basic, and was not much different from that originally invented in China. The specificity of Tibetan papermaking lies in the properties of native plants, the living conditions of peoples dwelling on the world’s highest plateau, and aspects of Tibetan culture that together create a distinctive craft. The need to make paper arose with the invention of the Tibetan alphabet in the seventh century and with the formation of the administrative apparatus. The demand for paper spurred the creation of workshops, which were most likely dispersed as the demand was not limited to one central location. Although the data is scarce, we may surmise that many small paper workshops existed in the Himalayan region, sometimes limited to one family, sometimes to the village and in exceptional cases to the region, which thus became famous for paper production. Every large project, such as editing the Kanjur or Tanjur (Tibetan Buddhist canon) stimulated the manufacturing of paper. That is the reason why paper manu- factures must have been located close to the largest monasteries involved in book editing. Unfortunately, there exist very few historical documents—which have, moreover, not been explored with respect to papermaking—that could support this hypothesis about the state of papermaking. Yet as Holmberg and March argue, in the eighteenth century, the villagers of Bomthang, Nepal had an obligation to supply paper to the government administration offices of the Ranas in Kathmandu.29 The village archive contains information about the regions of raw-material gathering, granted officially every few years depending on harvesting needs. This administrative aspect of paper production ought to be studied more closely in Tibet. It must be noted, however, that in today’s China, finding Tibetan village archives might be an extremely difficult task. According to standards of Tibetan papermaking, the best quality paper is evenly thick and not torn or pierced. It is free from any impurities even though it is smoothened by a polishing stone; it retains its flexibility and softness and holds the ink well. It is even sometimes believed by Tibetans that ink stays on the best paper despite water soaking in rainy weather. A book written on Tibetan paper does not harm the eyes during long sittings. The thickest type of paper, often laminated or comprised of glued layers was used for writing and printing Tibetan traditional books (glegs bam, dpe cha). In the following section, I describe in detail the stages of the traditional papermaking process with a general distinction between the production of ‘bark paper’ and ‘root paper.’

29 Holmberg et al. 1999: 44–45. 194 chapter 6

Collecting and Preparing Raw Material Traditionally, the production of paper took place in particular villages and involved most of their inhabitants in the harvesting of the raw material, which was the first step of the production. Since the trees were not cultivated, harvesting might require workers to travel sometimes even for days in search of proper plants. Ideal plants were typically two to three years old and measured at least one meter in height. The longest branches were cut at about ten cen- timeters above the ground so that the plant could regenerate.30 The layer of phloem that separated from the wood included a thin layer of bark, which was removed at the initial stage. This step had to be done when the bast was still fresh, as the task becomes harder as the material dries. Sometimes, the collected Daphne/Edgeworthia sp. tree strips of bark (phloem) were dried and stored for a period of time and later washed and rinsed in running water; in this process, the outer bark would be removed and the inner bark would be separated by scraping out the bark in the workshop. The higher the bark content of the wood, the darker the paper, and the weaker its strength. The content of outer bark particles has become the paper’s most recognizable feature and significant value at the perception of paper at the social and cultural level. Next, the raw material used for papermaking is the rte root of Stellera chamaejasne. This is usually harvested in the fall. In order to obtain the right substance for paper manufactured from root material, the lignified root of the Stellera chamaejasme31 must be dug out, and the outer bark and the kernels of the roots removed. The upper part of the plant is cut off, and the root is stripped of its outermost layer for cleaning. Only the middle layer of the root, which resembles tendon, is useful for papermaking. Considering other raw materials, green bark of agalloch eaglewood was the same as Daphne peeled away, the middle layer of the tree is removed and boiled for a while. After boiling, the liquid is drained off leaving the pulp for further processing. The process of boiling, beating, and stirring is the same as for Daphne paper.

Boiling Before starting the preparation of fiber pulp, the phloem was soaked for several hours in order to separate the fibers from the surrounding tissues. Sometimes

30 Trier 1972: 70. 31 The paper made of Stellera sp. was sieved on the river Nanjoon, using a form floating on the surface of the water. The prepared pulp would scatter over the surface of the water and then settle on the fabric as the form was lifted off the surface of the water. This is described in: Koretsky 1986: 3. A Survey of Tibetan Paper 195 the bark was left to soak in water for two hours, sometimes it was left to soak overnight. As recently as twenty years ago in the Nuntala papermaking village in Nepal, I observed a large wooden tank with running water standing in the open air, which was used for this purpose. The soaked fibers swell and their structure loosens up, which results in a higher degree of fiber separation. For a better result, the material prepared in this way was usually cooked for six to eight hours in water filtered through ashes from a fire.32 Boiling performs two functions: it removes the interconnections between fiber bundles and it neutralizes the poisonous substances present in plants used for papermaking (especially in Stellera paper). Unless the bast of the paper tree is boiled well, it is difficult to soften it. Burned acacia catechu (ja do) was added to ice cold water and stirred until the consistency of the liquid is similar to buttermilk. It is ready when it achieves a taste similar to sal ammoniac,33 if cooked longer, the substance will lose its lightness and sharpness. The strips of the inner bark of the paper tree would be boiled in this liquid after it had been cleared of all the sediment. The material for paper was also prepared by boiling in a solution of roasted barley flour (thel chin). The middle layer of the Stellera chamaejasme or Euphorbia fisheriana were boiled additionally in the mentioned thel chin until it became soft and loose as wool. Since this layer of the root contains hard poison, it is necessary to add one dose of myrobalan (Terminalia chebula) to the mixture to neutralize the toxic substances.34 While observing paper preparation at the Nuntala factory, I noticed that this technological stage was omitted. The reason for this omission of the cooking stage could be an effort to quicken the production, with an impairment of paper quality, possibly singular accidental thing. The omission may also constitute a means of cold processing of fibrous material much like in the case of flax production derived from plants.35 I based this hypothesis on my

32 Rischel 1985: 15–16. According to Trier the cooking stage took two to six hours (on the base of reports from five different Nepalese papermaking places) depending on ash or chemicals used during the process: Trier 1972: 69–92. 33 Sal ammoniac is a rare mineral composed of ammonium chloride, NH4Cl. It forms crys- tals in the isometric-hexoctahedral class that are colorless from white to yellow-brown. 34 Myrobalan—so called ‘Buddha’s herb,’ known also as Fructus chebulae (pharmaceutical name) or Terminalia chebula Retz. (botanical name). The fruit from this herb is among the ‘triphala’ (combination of three herbs) of Ayurveda. It is useful in asthma, sore throat, vomiting, eye diseases, heart diseases, hiccup, etc. 35 In flax processing, a procedure called ‘sprinkling’ is used. This is a biochemical process causing gradual disintegration of some of the substances comprising the structure of the 196 chapter 6 observations of large quantities of humid bark stored in the vicinity of tank with running water.

Beating At this stage of paper-pulp preparation it was necessary to separate individual fibers. In the early periods of the papermaking craft, fibers were separated manually: the pulp was beaten by hand and mixed with water. The bark strips were beaten on a stone mortar until they were transformed into pulp similar to dough. The paper would never become smooth if it was not beaten well. If the stone mortar was not placed firmly or cleaned properly, the color of the paper could change. Nowadays, traditional papermaking workshops use a simple mixer of the Hollander type to crush fibers drawn from water. The relative shortness of the fibers observed during the microscopic examination of a few samples taken from the analyzed artefacts is an indication that they were ground to a great extent. On this basis, one might surmise that a grinding mill was used for making the paper; this would suggest that the paper was produced in a fairly recent period. This also raises questions about the precise time when grinding mills replaced simple hand-maceration in the Himalayan region.

Molding Adding water to fibers that have been separated by hand or with the use of a mixer produces a fiber pulp that is ready for paper molding. The paper-pulp stock (the measured quantity of fiber pulp) is poured into a flat mold floating on the surface of a stream, puddle, pond, lake or, in some cases, a vat (Figure 113). The papermaker waits until the fibers are distributed evenly on the sieve and then makes several parallel movements from left to right; he/she then lifts the mold from the water surface, making several perpendicular movements— pushing the mold away and pulling it back until all the remaining water is removed (Figure 114). The mold with the damp layer of pulp is then propped up diagonally until the pulp dries and can be peeled off as a sheet of paper (Figure 115). Tibetan paper, as well as Chinese paper, was produced with floating molds according to the most traditional papermaking technique, which originated in China.36 The Tibetan papermaking mold called shok bre (a tool for shaping

plant. Fiber cells are held together with plant glues (pectins), which are the first to decay under the influence of microorganisms. See: Białousowa et al. 1958: 139–142. 36 The traditional papermaking methods in the Himalaya were studied by many authors during the last century, but Dard Hunter still remains the best authority on papermaking A Survey of Tibetan Paper 197

Figure 113 The process of making paper in the Kullu Valley, Himachal Pradesh, 1940s. Paper pulp prepared by beating upon a stone with a wooden mallet is being poured onto a ‘floating’ type of mold on the surface of a puddle. Photo 523/1(151), © The British Library.

Figure 114 Process of shaping a sheet of paper in the Kullu Valley, Himachal Pradesh, 1940s. The papermaker stirs up a quantity of pulp by moving the frame in the water until it covers the surface of the mold completely and evenly; he then tilts the frame until the water drains off. Photo 523/1(153), © The British Library. 198 chapter 6

Figure 115 Drying process during papermaking in the Kullu Valley, Himachal Pradesh, 1940s. The papermaking molds with newly made sheets of paper are left until the sheets are dry. Photo 523/1(154), © The British Library.

paper) is comprised of a wooden frame made of four pieces of wood, and a piece of cloth stretched on the frame that is thin enough to serve as a screen. This mold fulfilled the function of a sieve—a type of the floating mold called also the ‘woven’ type of mold (because of the ‘woven type of screen’). The size of the paper is determined by the different sizes and shapes of frames (Figure 116). The largest Tibetan papermaking molds were approximately door size. Valrae Reynolds reported seeing a mold measuring one and two-thirds by two-thirds meters—a wooden frame with a woven cotton center, as used at Gyantse, in southern Tibet, during the middle decades of the twentieth century.37 Papermaking molds are frequently constructed out of two pairs of slats joined perpendicularly to form a frame across which a cotton cloth is secured. The pulp is divided into piles of equal size and height, and each pile

in general: Hunter 1943: 77–203; Tschudin 1958: 679–689. More recent studies: Koretsky 1986: 2–6; Rischel 1985: 7–18; Imaeda 1989: 409–414; Dąbrowski and Siniarska-Czaplicka 1991: 30–48. 37 Reynolds 1991: 22. A Survey of Tibetan Paper 199

Figure 116 A large size ‘floating’ mold, constructed with a wooden frame and attached woven textile, placed in water (a stream) in Gyantse, c.1910–1920. Photo 1112/2 (139), © The British Library.

is placed in a container filled with water (vat). In such case, the paper sheets were formed from the entire amount of paper pulp in the tank. The process of shaping the paper sheet using this method requires metic- ulous adherence to technicalities. The thickness of the paper could vary if the shok bre is turned in one direction only while stirring the water. Even if the paper pulp is stirred very conscientiously and the mold is taken out of the water with care, the paper could still tear very easily because it is very fragile until it dries. Allowing the sheets of paper to dry on molds would not only prolong the already very time-consuming papermaking process but also limit the size of the production because the number of sheets of paper a papermaker can produce is determined by how many molds he owns. This is limitation encouraged the popularity of the alternative method of the dipping type of mold that is used in East Asian countries, such as Korea and Japan. 200 chapter 6

Finishing The dried sheets of paper require further processing: the edges of the sheets have to be evened out and the surface of the sheets must be polished. Prior to removing stains and cleaning the paper, tiny hairs from the surface are carefully removed with the use of a snow-white stone. chapter 7 Conservation

It is widely known that the political situation in the twentieth century not only led to the displacement of a large number of Tibetan books among other cultural heritage objects and sites but also that this displacement had a major impact on their conservation status. The presence today of so many Tibetan books in Western collections is not a result of their having been willingly sold or discarded by the previous Tibetan owners; rather, it is directly attributable to the relatively recent Chinese invasion of Tibet, the subsequent destruction of a large percentage of its monasteries, and the death or uprooting of many practicing Tibetan Buddhists. The purposeful laying waste of Tibetan monasteries and devastation of centuries-old collections of books and manuscripts had far-reaching consequences. The fate of the lost books and other textual materials was intertwined with that of the people. In Tibet, the book is an essential tool for Buddhist practice that is sometimes also respected as an object with attributed magical powers. The accumulation of spiritual merit through making or reciting a book composed of precious materials and created through special techniques has always had a meaning for the spiritual life of Tibetans. This fact makes the problems of their conservation very specific. Tibetan beliefs about the sacred nature of books are very strong. Thus, in the opinion of many monks, but also in the view of lay Tibetan Buddhists, the sacred nature of books is defaced if they are restored. For Tibetans, the religious function of a book is prioritized over its educational and archival functions. In my opinion, these feelings should be taken seriously for several reasons. First, Tibetan Buddhism is a living tradition (Figure 117). Moreover, this is not a case in which the present practitioners are far removed in descent from the people who originally made and used the books; there is continuity in the tradition of creating and respecting the books. Therefore, despite mass relocation of books from their places of origin, when Tibetans talk about museum objects, they are talking about a group of objects that were actually used in worship in the monasteries of Tibet. Second, we know from historical and ethnographic sources that books as sacred objects were never discarded in Tibet in the past but remained forever within the monastery walls or in homes. When they became too worn or damaged for ritual use, they were housed in special shrines where they continued to be treated with respect. In special cases, they could be ritually burned.1 Finally, the refugees in India,

1 Snellgrove 1978: 351.

Figure 117 Monks reciting books at Bya mang po in Western Tibet photographed in June 2013.

Nepal, and the West have worked very hard to overcome enormous obstacles and maintain their religion and culture outside of Tibet. We would have to be able to point to great advances in knowledge to be gained by desecrating their sacred objects through conservation treatment in order to even begin to justify that practice. With all due respect to such an attitude, it is astonishing that actions which for Western people mean to ‘preserve’ the object, for Tibetans may mean to ‘desecrate’ or even ‘destroy’ in the sense of their functionality. This disparity does not make discussion of ‘conservation of Tibetan books’ easier. Aspects of conservation are directly related to the materiality and function of a book. However, this sometimes means different things in particular locations and societies. Tibetan books in Western collections are very often objects with no history or even a claim to a particular origin; they are completely dis- placed objects. Despite this fact, they have recently been given center stage in Western collections to represent the Himalayas and Tibet. While becom- ing artworks, objects gain a new power that enables them to impact a new environment. After being handled by Tibetan people, they become actors in the social interactions they have generated, and in turn, can be manipulated. I would say that in this sense, Tibetan books in the West are ‘objects in transition.’ They move from one place to another; they pass through different stages and hands, and change their identity and function. In many cases, they will Conservation 203 already have crossed cultural, ethnic, or caste boundaries on their way from the craftsman to the patron or reader. Objects undergo various transformations from the place where they were originated to collections and institutions that place them in the spotlight before burying them in storage areas to make room for new objects to come. In this way, they are assigned to new roles in new places. By studying these moving objects in different spaces than those they originally came from, we have to consider the new configurations and meanings they take on as chang- ing artefacts, which undergo multiple appropriations and roles, and to assess the concept of authenticity. But these books as material objects become both actors and witnesses in the historical perspective. Additionally, they could also be seen as tools to preserve Tibetan culture outside of Tibet. In my opinion, these books have enormous power in their material form, which is why we should preserve materiality of these ‘witnesses of history’ for future generations of Tibetans who may want to learn original sources of their own tradition and culture. This is especially true because many Tibetans grow up outside of Tibet, and they may have a different view of their roots from that of their parents and grandparents. Undoubtedly, the impact of Tibetan books in the Western world indicates the need for theoretical research and practical work toward their proper protection. Problems of conservation require first and foremost knowing these artefacts—their function, technique, materials, and accurate diagnosis, as well as their conservation status. Thus the problem of preserving Tibetan books is very complex. Whenever we approach books’ fragments or large collections, we should start from an inventory and digitalization work. At the same time, knowledge should be gained through the description and a detailed study of individual objects as case studies, and conservation problems of all collections dependent on their condition, type, and provenance. The ‘biography’ and provenance of every particular object is equally as important as the history of the Tibetan book collections. Both should contribute to available knowledge of the value, trade, and circulation of these manuscripts on the art market. The Tibetan habit of copying the sacred texts and paintings is quite distinct from the production of forgeries for the art market. However, when we deal with unknown books, sometimes it is very difficult to accurately assess the proper value of objects. In many cases, the date attributed to a text is different from the date when the book was made. The discipline of book conservation has not yet developed in Tibet. Over the last twenty years, conservation of Tibetan thangka paintings, sculptures, or elements of architecture has been developed. There have been a few projects on the restoration of architecture and paintings in Tibetan monasteries 204 chapter 7 conducted by Westerners. Some faced the conflict of the Western code of conservation with expectations of local communities, which simply desired to have copies of old paintings produced. Tibetans do not have a tradition of preserving objects including all their material components. In practice, Tibetans understand conservation as the replacement of old artwork with new. In effect, then, the number of original copies is constantly being reduced. Together with original books replaced or wall pictures copied, plenty of historical and scholarly information is lost forever. However, loose-leaf Tibetan books are also often not conserved due to the fact that conservators trained to preserve Western-type books do not feel confident enough to apply the same methods to Tibetan books. However, from the past history of the conservation discipline, we have examples of when Tibetan books were treated with the same attitude as European manuscripts. For example, there is a group of Tibetan pothi books which were transformed during conservation treatment conducted in 1960 in the British Library into bound books. Loose leaves were bound together into gatherings and sewn like those in Western books. Nowadays, Tibetan books are much more often preserved with respect for their original state and meaning. The European and American attitude toward conservation is to retain and conserve the historical material itself. In the most general way, ‘conservation’ can be seen as the profession devoted to the preservation of cultural heritage for future generations. Conservation activities include examination, documentation, treatment, and preventive care, supported by research and education. In fact, it is an interdisciplinary field incorporating material-based practical training and theoretical background from the disciplines, such as theory and history of art, curatorship, chemistry, microbiology, and material science. It is difficult to say what percentage of every discipline appears in the daily life of conservators. I observe significant differences in different countries, depending on educational possibilities available to students, tradition, and economic situation. It is enough to point out that conservation studies programs are sometimes located at fine arts academies and sometimes at technical universities, or they are conducted as a variety of courses and workshops in museums or libraries. There is often a split between people who are trained more on the humanities side and those receiving scientific education, who usually work as conservation scientists. In the historical perspective, the conservation profession evolved from manual repairs in situ done by craftsmen or artists to a vast discipline based on teamwork and involving a variety of people with different backgrounds, whose decisions on the priority of preserving things are important for all scholars of the historical artefacts, since we interpret traditions through describing these objects which are preserved. Conservation 205

Taking all circumstances into account, it is not surprising that the terminology associated with the discipline of conservation can be confusing. Even the distinction between the meaning of the words preservation, restoration, and conservation is misleading, and this confusion is compounded by the fact that some languages use both words to express the same meaning or they use the same words to express a completely different meaning. In some countries, the distinction between conservator and restorer is a matter of semantics, while in others, the profession of conservator corresponds to that of curator in English-speaking countries, or conservator corresponds to collection keeper also conducting necessary treatment, and restorer, to person conducting treatment and reconstruction works which aim to return an object to the original or assumed original state. However, usually conservators agree that nowadays it is most essential to distinguish between preservation, preventive conservation, conservation science, and restoration.2 It is astonishing that all people (including me) working within those disciplines call themselves conservators. But the

2 For more information and definitions, see: Definitions of Conservation Terminology, American Institute for Conservation http://www.conservation-us.org/index.cfm?fuseaction= page.viewPage&PageID=620&E:\ColdFusion9\verity\Data\dummy.txt (viewed January 2013). Conservation science is the interdisciplinary study of cultural heritage conservation through the use of scientific inquiry and analytical equipment. General areas of research include the technology and structure of artistic and historic works and the materials and media from which they are made. Conservation scientists work in the field of art conservation, archaeology, or archaeometry. Preservation: The protection of cultural property through activities that minimize chemical and physical deterioration and damage and that prevent loss of informational content. The primary goal of preservation is to prolong the existence of cultural property. Preventive Conservation: This refers to the mitigation of deterioration and damage to cultural property through the formulation and implementation of policies and procedures for the appropriate environmental conditions; handling and maintenance procedures for storage, exhibition, packing, transport, and use; integrated pest management; emergency preparedness and response; and reformatting/duplication. Preventive conservation is an ongoing process that continues throughout the life of cultural property, and does not end with interventive treatment. Restoration: ‘Treatment procedures intended to return cultural property to a known or assumed state, often through the addition of non-original material.’ The rationale behind a restoration procedure is not necessarily prolonging the life of cultural property. The sole purpose of a restoration treatment might be to improve the aesthetic appearance of an object even when such intervention is not justified by an immediate threat or does not offer a real improvement in the condition of the object. 206 chapter 7 mission of preserving things is definitely what all of them have in common. Taking into consideration all ‘state of art’ problems related to the discipline, the conservation methods applied to Tibetan books should be viewed within the context of the symbolism of this exceptional craftsmanship in connection with Buddhist philosophy, and of Tibetan religious art, and in relation to differences in the character and function of a book in Tibetan and Western culture.

Typical Damage to Tibetan Books

This section describes the most common types of damage to Tibetan books. It should help to identify different types of damage and to recognize its causes. As previously described, Tibetan books are made from a wide variety of different materials, such as paper, birch bark, leather, wood, silk, cloth, animal/yak glue, or starch paste. Book structures range from loose-leaf pothi books to more complex multi-section deb ther codices. Identification and dating of historic and contemporary materials, techniques, and structures is not straightforward. Conservation and preservation methods and materials change and are under constant review. There are different remedial practices and conservation treatments available, and it is important that the right option is chosen and that the work is undertaken by appropriately trained staff or practitioners. The chosen option will also be influenced by the type of book, materials, value, rarity, fragility, and use of the book. Such decisions are always first of all conditioned by evaluation of damage and of factors which have caused it. Within possibilities, these destructive factors should be neutralized and damage caused by them at least stabilized. A book’s condition is determined by internal factors, namely its constituent materials and structure, and external factors such as storage conditions and handling.

Internal Factors Internal factors will determine how well a book ages and how resilient it is to adverse external conditions such as substandard environmental conditions and poor handling. There are the common types of internal weakness, sometimes called ‘natural flaws,’ which come out of materials composing a book or its structure from the moment of its creation. Due to the fact that the most common form of Tibetan book is the pothi format, consisting of loose leaves that you can easily disconnect and mix, the most typical damage observed is missing elements of a book. This is probably why the majority of Tibetan books in Western collections are incomplete. Usually Conservation 207 the missing elements are loose covers, wrapping cloth, or parts of leaves. Many books’ leaves are additionally not arranged in the proper order. Also covers and cloth are often mixed with secondary elements not original to a book. As a result of the loose-leaf form of the books, they were more vulnerable to mechanical damage and the activity of harmful external factors if not protected by wrapping cloth and covers. Tibetan wrapping cloth called pere [dpe ras] isolates leaves from the external environment and thereby protects a book from many factors, such as light and possible harmful external environment. It can be considered that this cloth plays an important role in protecting books, similar to the function of protective conservation boxes. If books are not protected in such a way, the first and last leaves are usually the most damaged. This is why if these elements of binding are missing, the practice is to protect books by making new boxes and wrappings for them composed of chemically stable and safe materials. It is also common that collections of different text, very often also having different sizes, are wrapped together into one cloth and placed between secondary covers. In such cases, larger size leaves are always more exposed to damage, especially if the original size of the leaves is not adjusted to the size of the covers. The leaves of Tibetan traditional pothi books are usually glued in several layers with starch or animal glue. This makes such leaves vulnerable to splitting of these layers when the glue weakens and lose its flexibility. The number of layers usually depends on type of paper and size of a book. Typical Tibetan paper made of Thymelaeaceae family plants, which is relatively thick by nature, does not require as many layers as the Chinese type of paper which is much thinner, softer, and weaker. Thus Tibetan handmade wove or laid paper is usually glued in two to three layers. Machine-made paper usually appeared as a single-layer leaf. However, this type was used only in books produced in the nineteenth century and, later, mostly in Russia and Mongolia. Independently of paper type, animal glue, whether it is mixed with starch or not, is used for gluing layers—it very often causes discolorations on the paper surface when aged (Figure 118). There are two procedures during the production of Tibetan books which may result in the presence of glue. First is the sizing of the paper, which occurs during papermaking when the papermaker adds starch or another plant extract to make fibers more slender and easier to disperse in a water solution. This method, however, is used mainly in China and East Asia, and not in proper Tibet. Additionally, the Daphne plant already contains a high amount of starch and other sticky substances among its bast fibers, which may give paper a slightly glossy surface. This is also why Tibetan papers were not necessarily sized at all during the papermaking process. They were prepared later, when 208 chapter 7

Figure 118 Discoloration caused by animal glue used for joining several layers of paper together. The Asia and Pacific Museum in Warsaw (MAP 14136).

the leaves were cut and layers of paper were glued together, then the surface of paper was processed before being written on. Sizing or gluing layers of paper prevented the penetration of ink into the structure of the leaves and prevented it from showing through to the other side. Animal/yak glue is used in some cases only for gluing the layers of Daphne/Edgeworthia paper together or for gluing scroll panels of rag paper, usually, in manuscripts from Dunhuang. For gluing the Chinese type of paper into layers, starch paste is used. But an animal glue in particular can cause discoloration with age. In general, usually Tibetan papers made of Daphne or Edgeworthia sp. plants are very stable and resistant to chemical and biological damage (insects). It is of interest that a comparatively small number of books within the studied materials were damaged by insects. This supports the author’s earlier findings that paper in Tibetan books was successfully protected by insect repellent or poisonous substances introduced into paper. In some cases, litharge, red lead, or yellow sulfide of arsenic (As2S3) was used, usually as paint applied over the surface of the leaf. These pigments are also very often used to paint the fore-edges of a book block, which also helps to protect the entire book Conservation 209 from insects. Sometimes other insect deterring substances derived from plants were introduced into the paper structure. This procedure also resulted in a yellowish color of the paper. Gibbs and Seddon identified this yellow color as a natural huangbo dye derived from the Phellodendron amurense tree. This dye was made by soaking and repeatedly boiling the dried phellodendron bark in water, pounding it, and then straining it through cloth. However, it should be noted that almost all Thymelaeaceae family plants contain poisonous substances and act as insect repellents by themselves. Some plants, such as Stellera or Euphorbia sp., need to be detoxified even during the papermaking process in order to be safe for people. However, Tibetan books are also produced on a variety of other types of papers, depending on region and time when the books were produced. In such cases, substances were applied to paper before writing. Examples can be seen among Tibetan manuscripts from Dunhuang. All mentioned substances, however, cannot protect paper entirely against fungi and bacteria. Additionally, a higher amount of animal glue in a paper’s structure makes such paper more attractive to insects. As mentioned, Tibetan books are executed on many different types of paper, and depending on this, they differ also in how they age. The Chinese type of paper in general, for example, is originally brighter in color but much more sensitive to color change over time, and less stable chemically. In this group, the best quality and strongest are usually papers made of pure paper mulberry, or of rags composed of recycled ramie and hemps. These raw materials are more resistant to chemical and biological damage. However, many Tibetan books produced in China and Mongolia were executed on paper with mixed components including grasses, straws, bamboo, and a variety of other components, including wood pulp. These are very sensitive to damage by light, humidity, and acids. They often relatively quickly become very brittle, dark yellow in color, or accumulate dark spots called ‘foxing’ often assumed to constitute typical damage to European acidic papers (Figure 119). Examples of such papers are described in the chapter of this book on Tibetan Woodblock Printing Culture. Examined Kanjur volumes of Peking and also Mongolian editions are printed on very thin Chinese papers glued in many layers. The quality of paper in the Wanli Kanjur Supplement volumes (1–22) and from the Prajñāpāramitā section volumes (29–37) is seemingly worse than that in the volumes included in the Wanli Kanjur collection. This is why the supplemental and the Prajñāpāramitā section volumes are more stained with foxing spots—the staining is due to quality of paper as related to raw material used and its processing, but also to the amount of glue used between very thin layers of paper. In addition to microbiological and physical damage, chemical changes were also visible. Breaks in the paper fibers, fragility of paper, and a 210 chapter 7

Figure 119 Discoloration caused by light, but also possibly by glue used for joining several layers of paper together. The Jagiellonian University Library (Pander Pantheon, vol. 26).

brownish color under light revealed the effects of aging. Further, the loss of flexibility in and weakness of the paper, especially in the areas that touch the side margins and down corners and edges, all make the paper very brittle. The color of the Daphne/Edgeworthia sp. papers examined usually ranged from yellowish to dark brown, though no organic substance that could be responsible for the color was detected. Such a dark color of paper could suggest its acidic nature. However, seemingly this is not the case, as my research shows that the paper in the studied books structurally was in good condition due to its physical and chemical stability and strength. The pH of the samples that had deteriorated most was in the range of pH 6–7. Thus, it is not necessary to conduct the deacidification procedures that are often performed in libraries that preserve Western books and other paper materials—especially on acidic paper. Originally, newly made paper of this type has a creamy color and contains small pieces of solid, dark brown pulp impurities. The darker shade/tone of paper appearing during aging can be caused by the presence of a substantial amount of outer bark and animal glue in the structure of paper. Outer bark gets into paper during papermaking due to very simple methods of splitting inner and outer bark from each other after collection of plant material. However, this is a typical feature of paper produced in the Himalayas. Microscopic study clearly indicated that the papers in general contained good quality, rather long fibers with many irregularities in the fiber walls and lumens. Conservation 211

Machine-made wood-pulp paper produced after the nineteenth century and often used in Tibetan books produced in Russia and Mongolia is especially vulnerable to deterioration. During the first half of the nineteenth century, papermaking methods in Europe, then in Asia, began to change into mass production. As the demand for paper increased, handmade rag-pulp paper in Europe, but also some traditional production from bast of trees in Asia, was superseded by machine-made wood-pulp paper. While European rag- pulp paper and traditional bast paper in Asia is very stable, machine-made wood-pulp paper is not; it deteriorates far more rapidly—especially in poor environmental conditions. Birch bark is especially brittle and vulnerable to mechanical damage. Wood as an organic material is also vulnerable to destruction by insects and other biological damage. In many cases, traces of spread ink that result from the natural features of particular types of paper are visible. Such an effect is often possible to observe on prints, where very absorbent and unprocessed paper is used together with diluted ink. These ink spreads appear at the moment of printing. Other types of ink blurs could also have occurred after a book was exposed to moisture or water. However, in this case such damage is an effect of an external factor, and should not be considered a natural flaw. Also, such blurs are distinguishable from the ink spreads that occur during the process of printing. Gold manuscripts are objects that require special care; methods of paper conservation for gold manuscripts should be applied together with procedures applied in the conservation of paintings (polychrome surfaces). These manuscripts require individual treatment. Each leaf is completely covered by one or more layers of paint. The purpose of maintenance is to preserve and consoli- date the layers. Typically, damage of gold manuscripts, such as in the case of the illuminated manuscript from the Museum of Asia and the Pacific (MAP 4323), consists of makeup pigments and a flaking layer of paint. There are white areas caused by abrasion of black ink on the surface of leaves. Because the ink-and- paint binder has been weakened, the painted layer and gold text has started to flake. This process is often intensified by paper deformations caused by moisture. In the above mentioned book, some areas have been weakened due to the stratification of the adhesive used for gluing the paper sheets in layers, as can be seen especially at the corners. Many papers have broken corners, which in some cases led to cavities. There are tabs on the outskirts of a small tear.

External Factors It is important to determine the underlying cause of the damage in order to prevent more damage from occurring in the future; for example, mold may be the direct cause of damage to a book, but mold occurs because the storage environment is poor. The poor storage environment may be caused by 212 chapter 7 building defects, equipment failure, poor air circulation, etc. Brushing mold off a book without attending to the poor storage environment will only lead to more mold, and more time and effort will need to be spent cleaning. In general, paper is an organic, porous, and hygroscopic material sensitive to mechanical, chemical, and biological damage. The natural aging of paper is due mainly to chemical processes which are revealed by yellowing of the paper. The main processes determining the condition of paper are acid hydrolysis and oxidation. At the structural level, acid hydrolysis results primarily in carbon chain shortening, causing a gradual loss of strength properties of the paper, a change in its brightness, and increased susceptibility to mechanical damage. The whole process is dependent on the acidity of the paper and the reaction time. Changes caused by the action of acids are primarily seen in papers of low quality, but the influence of acids may also be observed on good handmade rag papers. The process of oxidation occurs constantly and slowly, under the influence of oxygen from the air and light. It includes in particular the reactions of functional groups, and only in a few cases is interrupted between residues of glycosidic linkages of the cellulose molecule. For this reason, oxidation is more responsible for the yellowing of the paper, and only slightly causes a loss of mechanical properties. Light radiation and the associated increase in temperature usually cause the changes in color and decrease the strength of paper. We cannot really stop those processes, but we can slow them down through proper handling, control of environment, and the application of conservation procedures aimed at the stabilization of deteriorated materials. A range of external factors may damage books if not properly managed. The main ones are:

Environment Sometimes it is said that traditional storage conditions were favorable for books in Tibet, as the climate is dry and cold. However, the problem is much more complex. First of all, the climate in different parts of Tibet can be totally different, and in some areas to the east, quite humid. Additionally, now many Tibetan books have been scattered across the world, where Tibetan minori- ties exist among other cultures. Many Tibetan books found their way to India, where the humid climate is not so favorable for preservation of paper and other organic materials. For example, in a tropical climate, an Indian manuscript written on handmade paper, depending on handling and use, has a typical physical lifetime of two to three centuries. Then it requires a new copy. If carefully preserved in a more favorable climate, paper manuscripts may survive longer. Tibetan manuscripts found in Dunhuang or other oases along the Silk Roads, preserved for more than a millennium in a dry climate, are a good Conservation 213 example. However, inadequate temperature and relative humidity are both usually the most important environmental factors responsible for deterioration of books. The interiors of the Tibetan monasteries were not properly heated, and the buildings were often not sufficiently protected against leakage. Temperatures can dip very low in winter and are high in summer. Such changes in temperature and humidity were unfavorable for preserving books. For this reason, in the European libraries, systems of environmental control are installed which help to maintain a consistent temperature and humidity. Any change in the environment directly stimulates the deterioration processes occurring in paper. In Tibetan monasteries, books were stored in a general room within a monastery, and usually not in a place that is specifically dedicated to this purpose (Figure 120). Invariably, the interiors were poorly lit; fortunately, low exposure to light is a positive factor for paper conservation. Favorably constructed bookshelves are a regular part of the interior. However, they are always placed near an altar, exposing them to butter lamps; the combustion products of butter are deposited on surrounding objects, including books. The presence of such combustion products can significantly influence the air quality in these rooms. Another effect is that the studied books were often covered with a layer of dust and soil or oily dirt that could have the result of burning butter lamps in the monastic environment. In general, the presence of chemically unstable substances in the air brings a risk of damage stimula- tion. Typical air pollutants are the so-called acid gases (nO2, SO2). In particular, sulphur dioxide (SO2) easily absorbs in the paper, then is easily oxidized and dissolved in water to form sulphuric acid, which is exceptionally dangerous to paper. Nitrogen dioxide (nO2) stimulates both oxidation and acid hydrolyses. Ozone is also a factor which may cause oxidation. Fungal spores can be deposited in soot and dust suspended in the air. Additionally, dusts catalyse oxidation reactions as well. Biological factors are easier to control in cold and dry climates. However, in every environment books are exposed to local pests and mold damage. Many examined books evinced signs of moisture and water damage. The damage could have occurred when the books were carried by Tibetans participating in pilgrimage journeys. But damage also may have resulted when rain seeped into storage areas, since buildings were often not well protected against such leaks. The results of such damage are the variety of paper deformations, splitting of paper layers due to the dissolution and distribution of adhesives, brown stains on the outskirts of leaves, colorful spots caused by fungi and bacteria, or cavities caused by both microorganisms and pests. The wet leaves become a preferred environment for the growth of microorganisms and practically all other possible damage (Figure 121). 214 chapter 7

Figure 120 Tibetan books stored in a general room at the Nyingma monastery in Nelung (Nedung).

For example, a few volumes, such as volumes 19 and 22 from the Pander Pantheon, display signs of microbiological attack. The microorganisms’ activity caused decomposition of glues in paints, and resulted in the loss of some colors on the fore-edge paintings. Additionally, fungus caused discoloration of paper and characteristic stains on the paper surface. Particular volumes are damaged to different degrees; however, those with symptoms of biological degradation should be treated first.

Handling The term ‘handling’ describes widely understood human activities such as removal from shelves, support during use, etc. This additionally includes how books are used, including the habits of Tibetan readers. For example, many books feature fingerprint spots in the middle of the top or lateral margins since it is these locations that Tibetans tend to turn over the pages. Regardless of materials used, the books are very often covered with dust and dirt, especially the covers and cloth wrapping of the entire book; a few first and last leaves are usually more contaminated than the rest of the book block, and dirt is more often found on the edges and corners of the paper sheets. These books also exhibit plenty of mechanical damage such as tears, holes, or creases (linked with the object’s use, improper handling, or from poorly Conservation 215

Figure 121 The results of water damage on a book from the collection of the Museum of Asia and Pacific in Warsaw, Poland. Brown stains with fungi visible on the outskirts of cavity. The wet leaves become a preferred environment for the growth of microorganisms and practically all other possible damage. The Asia and Pacific Museum in Warsaw (MAP 12385).

conducted earlier ‘maintenance’). The primary cause of much mechanical damage is usually extensive use of the object, which in this case is related to the religious function of these objects. The above mentioned group of collected texts (the so-called mini-libraries sometimes), which existed as the personal property of both the monastic and lay population, was composed of different materials and is a good example of how every material deteriorates in a different way. Although they may have been carried together, every piece of text is in a different condition depending on the extent of usage and type of materials it was subject to (Figure 122).

Storage The state of preservation of Tibetan books and their general condition is closely associated with shelving, arrangement on shelves, the place where they are stored, and storage methodology used in the monasteries and monastic libraries. Tibetan monastic libraries contain many thousands of old books which are wrapped in cloth and held in place between board covers and piled up to the ceiling. The design of the shelving resembles a pigeon-hole system 216 chapter 7

Figure 122 Collection of texts kept together as mini-library. The Asia and Pacific Museum in Warsaw, Poland. consisting of vertical and horizontal wooden beams with separated spaces for individual books. The books are arranged at a distance from each other (each separately) based only on the rails, and their wooden covers are the basis for the entry card. This is a quite favorable system for storing books (Figures 123 and 124). Such a design promotes good ventilation because books are placed at a distance from each other. In this way, each book has its nest (pigeon hole), which is approximately matched to its format. However, some minor books, especially the copies without covers, were stored in cabinets in which they were piled on top of each other (Figure 125). This storage method could cause deformation of their leaves, and further damage.

Selected Procedures of Conservation Treatment

There are many levels at which we can treat a book with the intention of preserving it for future generations. Tibetan books require treatment that is specific to their form and materials. The loose-leaf form of the Tibetan book causes limitations not only in research, as discussed above (in dating and identifying wood covers as originally having belonged to a particular book) but also during conservation. This loose-leaf form caused many Tibetan books to be scattered in different collections in fragments (Figure 126). Conservation 217

Figure 123 Traditional shelving storage of books at Tengboche monastery, Nepal. 218 chapter 7

Figure 124 Traditional shelving storage of books in Nyingma monastery in Nelung (Nedung), Western Tibet. Conservation 219

Figure 125 Traditional storage of books in Namche Basar Monastery, Nepal.

Figure 126 Tibetan scriptures abandoned in caves near Chingkar, Western Tibet, photographed in June 2013. 220 chapter 7

Documentation Documentation is an integral part of the conservation process, the first very important step which helps to evaluate a book’s present condition, plan its future treatment, and expand our knowledge and understanding of a particular object. Documentation can be produced in electronic or hard copy and preserved together with a book. It usually includes descriptive and photographic components. Exact general requirements and standards are usually described in all conservation recourses, for example on the ‘Conservation OnLine’ website or in ‘Commentaries to the Guidelines for Practice’ on the website of the American Institute for Conservation of Historic and Artistic Works.3 More generally, documentation allows for evaluation of treatment methods and materials, support for education and scholarly research, provision of a record for current conservation practice, and study of the history of the conservation profession and rationales applied to the care of cultural property. Documentation leaves a record even if an object is lost, destroyed, or inaccessible. It has been recommended that procedures for reproducing original works in such unknown collections be planned as a standard procedure. The Tibetan custom of copying Buddhist texts can be considered as transferring books onto other supports. However, the philosophy of copying is different. Generally, digitalization ensures accuracy and facilitates wide dissemination of even the most precious objects. It is important to ensure that all pieces of information about the plan, treatment, and data obtained during the treatment are preserved. This includes procedures and variations from the initial treatment plan, materials added and removed from an object, and all changes introduced to an object, including its state after conservation treatment, in order to collect and make available all results of material and scholarly research to prevent unnecessary future analyses and repetition of treatments, and to serve as a basis for planning future preventive conservation measures and evaluating the safety and efficiency of materials, techniques, and procedures. Documentation provides evidence for every action undertaken by a conservator.

Conservation Research and Treatment Plan The study of an object in terms of its form and materials is necessary in order to make decisions about conservation treatment. Every object is individual, and the most important thing at this stage is to know exactly all the materials which a book is composed of. The conservator first needs to learn as much as possible about an object which he/she is going to treat. This is why study of

3 See http://www.conservation-us.org/. Conservation 221 the type of materials used and their properties, degree of their deterioration, and the construction of book and techniques used for its creation, comprise a good base for sketching a plan of full course treatment. Such study aims is to clarify the goals of treatment and the sequence of steps necessary to reach them. It also serves as a basis of communication and discussion between conservators and owners—regarding expectations, potential benefits, costs, and risks. The treatment plan should include at least the proposed course of treatment, materials to be used, time and cost estimate, objective and limitations of the treatment, benefits and risks, general descriptions of original materials and also these planned to be used, and all possible alternatives to procedures and materials to be used. Treatment begins with descriptive and photographic documentation of the conservation status of a book. Leaves need to be numbered before any cleaning procedures are applied.

Cleaning Procedures Mechanical Cleaning Dry cleaning conducted with soft brushes is usually the basic procedure used to dust off the object. The surface is then gently cleaned by an eraser powder, and then erasers of different hardness may be applied if necessary. At this initial stage, general cleaning and proper packing (in separate boxes made of standard conservation materials) is usually recommended as a common preservation procedure. Since Tibetan books are very often covered by dust and dirt, dry cleaning may not give satisfactory results. In such cases, carefully performed wet cleaning can help to clean stains and dirt difficult to remove by other methods. However, a full-immersing bath can be too invasive due to the fact that it may dissolve the glue between layers of paper and can also wash out organic substances introduced into paper to protect it against biological factors (if they are any).

Washing The process of washing a work of art on paper removes soluble deterioration products such as acids or discoloration from the support. Generally, paper that has been washed has increased folding endurance and tear strength. Washing may not be appropriate for all works on paper, since the washing process may solubilise sensitive media, dyes, and pigments, or wash out fillers or sizing agents, altering the appearance of the work of art. The habit of gluing layers of paper together is very important to take into account when making decision about conservation procedures to be applied, especially washing. For example, wet cleaning used widely in the conservation of flat paper objects and in Western book conservation, especially water or an alcohol bath, is very 222 chapter 7 questionable in the case of Tibetan books. This is why it is sometimes better to accept those discolorations as a witness of time rather than try to remove them by all means. If this is not the case, it is possible to adjust temperature and time or try with a solution of water and alcohol to conduct a special bath during which only the surface of the paper floats on the solution surface. This is only possible when ink is stable and cannot be dissolved by such treatment. This procedure may more often be safely applied to prints, since inks are usually more stable there and paper is softer and not as highly processed as that used for manuscripts.

Repairs and Loss Filling Losses to a paper support may be filled using inserts, pulp, or by lining. Papers used for inserts should be closely matched to the original support. Surface cleaning of the edges of the loss is undertaken, followed by shaping of the repair paper to fit the loss. Edges are bevelled so that the edges of the infilled loss are the right thickness. After alignment and attachment of the fill, the support should be pressed and dried under weights (Figure 127). It may also be

Figure 127 Fragment of a scroll taken out from praying mills during conservation treatment conducted by researcher together with Mirosława Wojtczak at the Department of Paper and Leather Conservation, Nicolaus Copernicus University in Toruń. Paper repairs done with new material pasted into loss of damaged area visible on the photograph. Conservation 223

Figure 128 Leaves of a Tibetan book restored with a leaf-casting method. The Asia and Pacific Museum in Warsaw (MAP 14144).

necessary to tone the fill to match the color of the support. Adhesives used in filling are usually starch pastes or cellulose ethers. In very specific cases when book leaves are extensively damaged, a secondary support can be considered. However, in the case of Tibetan books, it can be applied as an additional layer inside. It should be inserted in the middle of the leaf. This is why the process requires splitting original layers to add it. Lining in general involves the attachment of the entire support to a secondary support which provides structural strength. This technique may be appropriate when the original support is heavily deteriorated and requires additional structural stability. Otherwise, it should be avoided. The most common lining papers chosen are Japanese tissues, as they provide excellent strength despite being thin and flexible. The adhesives used in lining are generally the same as those used in other paper conservation treatments, namely starch pastes or cellulose ethers. Paper pulp may be prepared in a blender, though a leaf casting apparatus may be more appropriate for large losses. This method is slightly more difficult in the case of several layers glued together, since it needs to be done in water immersion. This is why leaf casting definitely cannot be applied to gold manuscripts or any other illuminated surfaces of Tibetan books (Figures 128).

Consolidation Friable or flaking media requires the application of some form of fixative or consolidant. Desiccation or deterioration of the binding medium may be 224 chapter 7 causing delamination or migration of the paint; Tibetan gold manuscripts often suffer from this problem (Figure 129). Movement of the support may result in the formation of cracks within the paint layer, and may lead to flaking. For example, the black paper to be applied to fill losses in the leaves of the Diamond Sutra from the Museum of Asia and the Pacific was prepared by painting it twice with Chinese ink and allowing it to dry. All leaves of this book were treated as painted surfaces. Therefore, it was not possible to perform any operations with water. A series of tests used methylcellulose (MC) as a gel applied to the surface of the card. Finally, some additional cleaning cards with 3% methylcellulose (MC) were applied to the surface; this process provided a good cleaning result and encouraged the paper layers to reunite. Small holes and tears were supplemented with appropriate paper that was pasted into the object with modified starch. All leaves were sized with MC (1%) in the end by applying it to the surface and allowing it to dry. The same procedure was repeated on the verso side of the leaves.

Figure 129 Deterioration of Tibetan gold manuscript ‘Life of Gshen rab mi bo’ caused by migration of and chemical reaction between the paints. Rare Book and Manuscript Library, Columbia University in the City of New York (Tibetan manuscript 4). Appendix 1 Selected Items from the Tibetan Collections of the British Library

Tibetan Books from the Tibetan Collections, The British Library

1 British Library 15193; Manuscript Raw material: Bark paper: Daphne or Edgeworthia sp. Type of paper: Woven paper (two or three layers), paper handmade by means of a woven type of mold, some layers are split/unglued, glossy surface. Find site/provenance: Tibet. Dating: n/a. Format: Stitched book; 2 sections covered on the front and back sides only with a soft facing/lining (a single layer of paper lined with red linen/cotton fabric), which overlaps the binding in 4 cm on the side; first section is composed of 34 leaves and one-side cover. Book sewn with red string and reinforced by a raw-color thicker string (robe); 71 leaves and 2 covers (2 front leaves and last leaf are missing deduced from construction of the book). Size (h × w): 8.5 × 26 × 2.5 cm. Language/script: Tibetan (headless, dbu med). Page layout: 5 lines / page, side margins (1.5–2.5 cm) clearly sketched in a red double line (the bound close margins usually narrower), the same layout on both recto and verso sides; two sizes of lettering, base lines clearly visible on some pages. Content: incomplete gter ma (treasure) text. Ink: text in black ink, margins and particular words in the second section are distinguished in red ink. Condition of the book: The book is covered with dust and heavily soiled with stains. Binding: damaged, but construction of the book is stable; textile has white spots and stains on the surface. Book block: a minimum of three leaves are missing; glue is dried in many areas, which caused the splitting of paper layers within the leaf, small cavities, cracks, and tears. Water activity: brown stains on side corners and margins. Insects: slight insect activity on a few front leaves (bottom, side corners). Micro-organisms: white fur on the paper surface. Cavities in the paper: many tears and creases.

2 British Library 13162 (Previous Collection of British Museum); Gold Manuscript Raw material: not examined. Type of paper: not visible under the black ink. Find site/provenance: Brought as a gift from Bhutan in 1866 (note from 14.08.1934 by J.S. Zua [. . .]?). Dating: n/a. Format: Stitched book; 3 sections; middle band covered with cloth, sides with silk; 42 leaves (including covers). Size (h × w): 7.5 × 18.5 × 2 cm. Language/script: Tibetan (headed, dbu can); front page with framed title. Page layout: 3–5 lines / page, margins 1 cm framed clearly sketched in a gold double line on side margins, the same layout on both recto and verso sides. Content: bka’ ‘gyur bsdus pa’i gzungs (a collection of dharaṇis from the Buddhist canon) Ink: text in gold ink.

3 British Library or 15190; Gold Manuscript; Golden and Silver Text on Black/ Dark Blue Background Raw material: Many fiber components (paper mulberry, jute, bamboo, straw, and ramie possible). Fibers are damaged and their anatomical features, difficult to recognize. Large amounts of glue and binding media in the paper structure (stains seemingly of animal glue, other ingredients also possible). Type of paper: Laid paper (three layers): the two outside layers have been dyed with treatment that is similar to the treatment of fabric—dipping (in the Chinese way), inside layer not dyed. Paper was dyed by dipping and with black ink placed centrally, then the surface was polished. Paper is indigo, dyed by dipping. On one side, a natural- color margin remains (1.5 cm)—the place where a finger would grip while pulling the paper through the dyeing solution. Although the paper was dyed by dipping, paintbrush marks are visible on the surface (perhaps the glue to bind the three layers was covered with a paintbrush, or a brush was used to apply a protective substance before dipping). A leaf prepared in this way was covered with black ink in the area delimited by the frame. Next, a lined base was created, followed by gold and silver text. At the end of the process, the frame was marked with a double red line 12.3‒12.5 × 38.5‒39 cm marking off the margins (in this way, the black central area with text was separated from the blue margins). Find site/provenance: Tibet. Dating: Probably 18th century, deduced from text. Radiocarbon dating shows the last three hundred years plateau across the calibration line, but indicate the highest TIBETAN COLLECTIONS OF The british library 227 probability as 1680–1770 or 1800–1940. A Tibetan object produced on Chinese paper (deduced from fiber composition of paper mulberry, jute, bamboo, straw, and ramie— these were the most popular raw materials for papermaking); type of paper with narrow wire marks—paper handmade with movable type of mold with a fine (narrow distance) bamboo sieve; a thin layer of paper use—all suggest a Chinese origin of this paper; and dyeing technique—indigo paper dyed by dipping. On one side, a natural- color margin is left (1.5 cm)—the place where a finger would grip while pulling the paper through the dyeing solution. Dendrochronology: The book was not dated, but the eighteenth century was deduced from the text. The wooden book cover was analyzed using the non-destructive method of measuring the rings—preliminary results of dendrochronological analyses suggest that the book cover could have originated in East Tibet due to its similarity with chronology from that region. Format: Loose leaves pothi and upper cover. Upper cover: a wooden plank wrapped in paper (dyed indigo) glued to the cover with a top layer of silk (possibly dyed indigo as well); the cover is carved and painted on the inside with two silk respect curtains attached to the top. The plank is made of softwood cut radially; the surface is smoothened. The inner side of the cover has the shape of a frame carved out: side edges—1.3 cm; upper and bottom edges 1.5–1.8 cm; depth—0.2 cm. A centrally-composed rectangular frame 25 × 6.7 cm is carved out for 0.3 cm depth. The cover contains a painting the style of which corresponds with the technique of executing the book; on the side margins, two larger figural paintings (the one on the right is accompanied by two smaller paintings on each side of its bottom). On the upper margin there are 13 small paintings; on the bottom margin, 12 paintings; on the upper edge, two (yellow and green) respect curtains installed (silk with a delicate, geometrical pattern in the structure of the fabric). The inside of the cover is painted; the surface is not ideally smooth, visible marks of a flat chisel; in the places where paint has peeled off, marks of priming is visible. Main colors used in painting include: red, black, and gold. Pigments were not identified. The surface appears to have been covered with half-transparent lac or another varnish. The title was executed in gold letters on red background, and is placed centrally in the more deeply-set rectangular frame. Stratigraphy: (1) drawing of cloud ornaments in dark ink on the inside set; (2) layer of red paint on the whole surface; (3) gold ornaments—auspicious symbols, and on the left margin, gold stupa with a black contour; (4) title in the form of large Sanskrit letters, gilt, the contour of which are emphasized with gold relief letters; (5) the outermost layer consists of varnish (possibly lac). Size (h × w): Leaves: 20.5 × 49.5 cm; cover: 21 × 50 × 1.5 cm. Language/script: Tibetan (headed, dbu can). 228 appendix 1

Page layout: 1–6 lines / page. Guidelines scraped by a sharp tool (no coloring involved). A sharp tool was also used to impress vertical lines corresponding to side margins; the frame defined by margins has a regular rectangular shape and held all text lines. Margins clearly marked off with double red line: side margins—title pages 5.5 cm; upper and bottom 4.5 cm. Recto pages contain additional frames on side margins with title pages and volume number. Content: The Tibetan Book of the Dead, dam tshig thams cad kyi nyams chag skong ba’i lung lnga | bshags pa thams cad kyi rgyud dri ma med pa’i rgyal po zhes bya ba, a tantra of the Mahāyoga class. Miniatures: book is decorated with miniatures. Ink: silver and gold ink; fragments of text marked with gold ink (each page contains gold alternately with silver text fragments). Condition of the book: Silk binding lined with paper preserved in 70%; blisters on successive layers; numerous cavities in the plank; the fabric coming off (50%) diagonally with holes and scrappy edges. The fabric is creased and threadbare in many areas, with color changes particularly visible. On the inside of the plaque, the remains of a silk fabric lining are visible, which indicates the silk was wrapped on the inside surface. Book block: wavy leaves. Moisture and water traces: perhaps the leaves are wavy because of previous exposure to higher humidity or waviness may be the result of gluing layers of paper: with a natural layer covered by two dyed layers additionally partly covered with ink; due to thinness of paper and outer ink layer having been superimposed only centrally (partly)—paper size is unstable. At the moment, the block is in good condition. Other stains: white fur; Cavities in the paper: numerous small losses and creases on leaf edges and tears; in the central part of the block, only a few cavities resulting from insect activity; several pages heavily creased, major losses on page 112—a simple repair is visible (green painted paper).

4 British Library Tib CC 74 (Previous Collection of India Office); Gold Manuscript Raw material: Bark paper: Daphne or Edgeworthia sp. Type of paper: Woven paper (two or three layers): paper handmade by means of a woven type of mold (technology is visible only on a small area since leaves are painted with black ink and some indigo remains on the margins). Find site: n/a. Dating/place of origin: Darjeeling (?): on the bottom cover sticker (India-Bengal Collection No. 6194), it is written that the book was produced in Darjeeling. Format: Loose leaves pothi with two handmade wooden covers (chisel traces are visible on the wood surface), upper cover carved in geometrical pattern; leaves seem TIBETAN COLLECTIONS OF The british library 229 to be cut to the text frame size, only small parts of indigo-painted margin edges are left; 80 leaves. Size (h × w): leaves block: 8.5‒9 × 33.5 × 3 cm, upper cover: 35 × 7.8‒8 × 1.8 cm; bottom cover: 8 × 35 × 1.2 cm. Language/script: Tibetan (headed dbu can). Page layout: 3–5 lines / page, no guidelines; text frame—about 29 × 7.5 cm; side margins decorated with gold patterns on margins of leaves 1‒5, later leaves have numbers on recto sides only, margins (0.5‒1 cm) framed clearly by a gold double line, the same layout on both recto and verso sides; in some areas, text is cut due to secondary margins having been cut; skilled and beautiful calligraphy. Paper decoration: golden line drawing on the side margins of first 5 folios—geometrical pattern composed of dots and lines resembling the lotus petals. Content: Collected minor works (ngag thor bu) of the 2nd Paṇ chen bla ma Blo bzang ye shes (1663–1737). Ink identification: gold and silver ink. Condition of the book: Binding: book covers preserved well; however, their size does not fit the size of the leaves, which probably were much larger than they are (due to margin cuts)—this suggests that the covers may originate from a different object. Book block: extensive destruction—paper in bad condition (cavities, cracks, and tears); black ink layer damaged—wear through, inner layers of paper have a natural color which suggests that they were not dyed with indigo before—possibly only margins or surface were painted with blue paint before central part was covered with black ink; Paper heavily soiled with stains. Water activity is not visible. Cavities in the paper: many tears and creases.

5 British Library Tib CC 101 (Previous Collection of India Office); Manuscript Raw material: not tested. Type of paper: Woven paper (one layer). Find site/provenance: Tibet. Dating: n/a. Format: Stitched book; one-layer paper leaf folded once in a quire; 4 quires (1st–10 folios; 2nd–9 folios; 3rd–8 folios; 4th–6 folios); loose cloth lined with a paper guard over each quire; book stitched with linen string; 65 leaves. Size (h × w): 8.3–8.8 × 14.5 × 2 cm. Language/script: Tibetan (headed, dbu can). Page layout: 6 lines / page, no guidelines, text in double frame on recto pages and single frame on verso sides; margins (0,8–1 cm) framed clearly sketched in red, centrally situated, marking text fragments on a few of the last pages with red ink. 230 appendix 1

Content: O rgyan gyi gu ru padma ‘byung gnas kyi rdo rje’i gsungs med pa’i gsol ‘debs le’u bdun ma, Prayer to Padmasambhava. Ink: black (text) and red ink (a few lines on last pages). Condition of the book: Last leaf is missing; leaves close to the covers are in very poor condition; extensive destruction; cracks and torn quire folds edges; many tears and creases in paper leaves; paper almost brown, especially cover; object dusted and heavily soiled with stains. Water activity is not visible. Fur is visible on the paper surface.

6 British Library TIB CC 114–115 (Previous Collection of India Office); Manuscript Raw material: Not analyzed. Type of paper: Woven paper (three layers): inserted folio—single layer of paper folded once, thicker than others, with rough surface. Find site/provenance: Tibet. Dating: n/a. Format: Stitched book: Three layers of paper leaves folded once in a quire; 4 quires + 1 leave with cloth (1st–8 folios+ front textile with paper cover; 2nd–11th folios covering (supporting) the bind of a quire; 3rd–10th folios; 4th–5th folio + 6th folios + 1st folio covered partly with purple linen); cloth (linen stripe around 6 cm wide) guard over each quire; book stitched with linen string; 3 sewing points in the section; 73 leaves. Size (h × w): 10 × 24 × 3 cm. Language/script: Tibetan (headed dbu can). Page layout: 3–5 lines / page, margins (0.8–1 cm) framed clearly, sketched in red double line, red control lines; the same layout on both recto and verso sides. Content: Tibetan religious book. Ink identification: black (text) and red ink (some words in last quire; frames). Condition: Binding damaged, strings (sewing) eaten by rodents; textile in very poor condition; Book leaves extensively damaged; cavities, cracks, and tears. Paper heavily soiled with stains; almost brown, especially margins on edges; object dusted and soiled. High insect activity observed. Extensive white fur is visible on the paper surface. Many tears and creases on paper.

7 The London Sel Dkar Kanjur or 6724; Manuscript/Gold Manuscript (Folio 1–2 Written in Gold on Blue/Black Paper, Rest of the Book Written with Black Ink on Raw, Polished Paper) Raw material: Bark paper: Daphne or Edgeworthia sp. Type of paper: Collection contains both types of paper (two or three layers)—laid paper with 4 laid lines in 1 cm, and woven paper. Woven sieve print—for example, folio 3 in vol. 56, and a bamboo sieve print (characterized by 4 laid lines in 1 cm and TIBETAN COLLECTIONS OF The british library 231 chain lines in distance about 4.5‒5‒5 cm) was observed on leaves 4 and 5 of volume 65. Surface of paper is glossy, polishing tools’ traces are clearly visible, which suggests that paper was sized with starch paste and surface was later polished with a stone. Find site/provenance: Tibet, Sel dkar Monastery, brought to London by Younghusband expedition. Usually on the first, natural-color paper folio there is an oval seal (2.3 × 2 cm) printed with bright yellowish ink; the text of this seal is hardly visible, impossible to identify—this is probably the owner’s seal, rather than the papermaker’s mark, because it is not repeated in all other volumes of the Sel dkar. Dating: 1712. Format: Loose leaves pothi. Most volumes contain one or two covers. Upper cover: a wooden plank; carved with a rectangular, geometrical pattern; red lacquered; painted and gilded (floral decoration and title written in Lanza script). Size (h × w): 24.5‒25 × 65‒66.5 cm; wooden cover 26.5 × 68.5 × 3.5 cm. Language/script: Tibetan (headed, dbu can). Page layout: 3–8 lines / page. The front page features three text lines of the title within a rectangular frame, then succeeding pages contain up to eight text lines per page limited by side margins sketched with red ink. Text outlined in frames sketched with red ink. Guidelines scraped by a sharp tool (no coloring involved) on the front, black inked leaves sketched with yellow dyes on the remainder of the book. The title page usually features two miniatures painted with gold on red, covered with paper flaps and a 3-layered curtain. The miniatures are mounted under the upper layer framed with a geometrical ornament, a variation of an endless knot. Three curtains are stitched together on the bottom corners and attached on the top with three leather clips. In this case, a sharp tool was also used to create vertical and horizontal lines corresponding to side margins, and to design the space for a later addition of miniatures. The verso of some front pages are executed in raised gold on a black background with turquoise and corals mounted on the page as elements of a circular shape imitating holes in palm leaf books. Content: Tibetan Buddhist Canon, London Kanjur. Ink: gold, red, and black ink.

Tibetan Books from the Stein Collection, British Library (in Collaboration with the International Dunhuang Project)

8 IOL Tib J 99 vol. 14:1 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: Laid paper: 12 laid lines in 3 cm. Find site: Dunhuang Cave 17. 232 appendix 1

Dating: 9th c. Format: Originally concertina, now two loose leaves. Size (h × w): 7.5 × 27 cm. Language/script: dbu can (recto: square style with archaic features; verso: sutra style). Page layout: 5 lines / page, no margins or guidelines. Content: Tibetan Buddhist text: The Cycle of Birth and Death.

9 IOL Tib J 229 vol. 69:33 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: Laid paper: 15‒18 laid lines in 3 cm; chain lines not visible. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll fragment (horizontal). Size (h × w): 26 × 20 cm. Language/script: recto: Tibetan (sutra style), verso: Chinese. Page layout: 14 lines / section, margins and guidelines in black. Content: recto: Pañcaviṃśatisāhasrikā-prajñāpāramitā Patches on verso: Confucian text.

10 IOL Tib J 354 vol. 69:51 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: Laid paper: 12 laid lines in 3 cm (one-layered paper); chain lines intervals: 6‒7 cm. Find site: Dunhuang Cave 17. Dating: 9th c. Format: Scroll (horizontal). Size (h × w): 19.5 × 27.7 cm. Language/script: Tibetan (sutra style). Page layout: 20–22 lines / section, no margins or guidelines. Content: Uṣṇīṣa-sitātapatra-dhāraṇī.

11 IOL Tib J 401 vol. 76:2 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: Laid paper (one- or two layers): 12‒15 laid lines in 3 cm, some of the folios might be reused. Find site: Dunhuang Cave 17. Dating: 9th c. Format: Thread booklet, pothi loose leaves (8 × 38 cm) folded in half and stitched with leather thread. Size (h × w): 8 × 19 cm. TIBETAN COLLECTIONS OF The british library 233

Language/script: Tibetan (official styles, headed and headless). Page layout: 6–8 lines / page, no margins or guidelines. Content: Collection of rituals owned by monk.

12 IOL Tib J 401 vol. 76:6 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: Laid paper (one- or two layers): 12‒15 laid lines in 3 cm; some of the folios might be reused. Find site: Dunhuang Cave 17. Dating: 9th c. Format: Thread booklet, pothi loose leaves (8 × 38 cm) folded in half and stitched with leather thread. Size (h × w): 8 × 19 cm. Language/script: Tibetan (official styles, headed and headless). Page layout: 6–8 lines / page, no margins or guidelines. Content: Collection of rituals owned by monk.

13 IOL Tib J 401 vol. 76:7 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: Laid paper (one- or two layers): 12‒15 laid lines in 3 cm; some of the folios might be reused. Find site: Dunhuang Cave 17. Dating: 9th c. Format: Thread booklet, pothi loose leaves (8 × 38 cm) folded in half and stitched with leather thread. Size (h × w): 8 × 19 cm. Language/script: Tibetan (official styles, headed and headless). Page layout: 6–8 lines / page, no margins or guidelines. Content: Collection of rituals owned by monk.

14 IOL Tib J 438 vol. 33:15 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: Laid paper: 12 laid lines in 3 cm, chain line distance: 7.5‒8 cm. Find site: Dunhuang Cave 17. Dating: 9th–10th c. Format: Pothi with string holes. Size (h × w): 9 × 46.5 cm. Language/script: Tibetan (sutra style, with interlinear commentary). Page layout: 5 lines / page, margins and guidelines in black. Content: Guhyasamaja tantra, fine copy. 234 appendix 1

15 IOL Tib J 454 (upper part) Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: Laid paper (one or two layers), dye in the structure possible. Find site: Dunhuang Cave 17. Dating: 10th c. Format: Scroll (vertical). Size (h × w): 30.5 × 430 cm. Language/script: Tibetan (sutra style). Page layout: ~22 lines / panel, no margins, guidelines in black. Content: Tantric treatise.

16 IOL Tib J 530 vol. 79:2 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: Laid paper glued of layers: 12 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: 10th c. Format: Thread booklet, 2 sections, 4 bi-folio in each, sewn with silk thread (7 folios in first section). Size (h × w): 15.8‒16 × 11.5 cm. Language/script: Chinese, Tibetan (post-imperial Buddhist style, headed). Page layout: 6 lines / page (Tibetan text), margins and guidelines in black. Content: Chinese Buddhist texts, Tibetan mantras and transcriptions of Chinese in Tibetan.

17 IOL Tib J 530 vol. 79:3 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: Laid paper glued in layers: 12 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: 10th c. Format: Thread booklet, 2 sections, 4 bi-folio in each, sewn with silk thread (7 folios in first section). Size (h × w): 15.8 × 11.5 cm. Language/script: Chinese, Tibetan (post-imperial Buddhist style, headed). Page layout: 6 lines / page (Tibetan text), margins and guidelines in black. Content: Chinese Buddhist texts, Tibetan mantras and transcriptions of Chinese in Tibetan.

18 IOL Tib J 612 vol. 68:77 Raw material: Rag paper: Boehmeria nivea sp. (Ramie) + Jute possible. Type of paper: Laid lines: 15 laid lines in 3 cm. Find site: Dunhuang Cave 17. TIBETAN COLLECTIONS OF The british library 235

Dating: 9th–10th c. Format: Scroll (vertical). Size (h × w): 27 × 27.5 cm. Language/script: Tibetan (monastic style). Page layout: 15, 16, 22 lines / panel, light margins, guidelines in black. Content: (i) Pañca-skandha (ii) Text on the bodhisattva path.

19 IOL Tib J 1126 vol. 55:55 (right bottom corner) Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: Laid paper: 12‒15 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Folded document with a square vermilion seal at the bottom right 6 × 6 cm. Size (h × w): 20‒23 × 30.5‒31 cm. Language/script: Tibetan (official style, headless). Page layout: 9 lines, no margins or guidelines, square red seal and cross-hatching at bottom. Content: Letter from Dunhuang, Tibetan imperial period.

20 IOL Tib J 1254 vol. 56:73 (right upper margin) Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: Laid paper: 12 laid lines in 3 cm, thick. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Sheet, folded. Size (h × w): 26.5 × 41.5 cm. Language/script: Tibetan (official style, headed, dbu can). Page layout: 26 lines on sheet, no margins or guidelines. Content: Letters regarding the copying of Śatasāhasrikā prajñāpāramitā manuscripts.

21 IOL Tib J 1357(B) vol. 69:46 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: Laid paper: 12 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll fragment (vertical). Size (h × w): 52.5 × 30 cm. Language/script: Tibetan (official style, headed, dbu can). Page layout: 27 lines / panel, margins and guidelines in black. Content: Official document: list of income for monasteries. 236 appendix 1

22 IOL Tib J 107.1 vol. 57:119 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: Laid paper glued in layers, 12 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Pothi, two holes, with circles. Size (h × w): 20.5 × 72.5 cm. Language/script: Tibetan (sutra style). Page layout: 12 lines, light guidelines, margins not visible. Content: Śatasāhasrikā prajñāpāramitā.

23 IOL Tib J 107.1 vol. 57:137 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: Laid paper glued in layers, 12 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Pothi, two holes, with circles. Size (h × w): 20.5 × 72.5 cm. Language/script: Tibetan (sutra style). Page layout: 12 lines, light margins, guidelines in black. Content: Śatasāhasrikā prajñāpāramitā.

24 IOL Tib J 107.1 vol. 57:138 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: Laid paper glued in layers: 12 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Pothi, two holes, with circles. Size (h × w): 20.5 × 72.5 cm. Language/script: Tibetan (sutra style). Page layout: 12 lines, light margins, guidelines in black. Content: Śatasāhasrikā prajñāpāramitā.

25 IOL Tib J 612 vol. 68:67 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: One layer of laid paper with irregular structure: 12‒15 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: 9th–10th c. TIBETAN COLLECTIONS OF The british library 237

Format: Scroll (vertical). Size (h × w): 43.5 × 27.8‒29 cm. Language/script: Tibetan (monastic style). Page layout: 15, 16, 22 lines / panel, light margins, guidelines in black. Content: (i) Pañca-skandha, (ii) Text on the bodhisattva path.

26 IOL Tib J 612 vol. 68:69 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: One layer of laid paper with irregular structure: 12‒15 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: 9th–10th c. Format: Scroll (vertical). Size (h × w): 43.5‒43.7 × 27.8 cm. Language/script: Tibetan (monastic style). Page layout: 15, 16, 22 lines / panel, light margins, guidelines in black. Content: (i) Pañca-skandha, (ii) Text on the bodhisattva path.

27 A IOL Tib J 612 vol. 68:70 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: One layer of laid paper with irregular structure: 12‒15 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: 9th–10th c. Format: Scroll (vertical). Size (h × w): 43.5 × 27.9 cm. Language/script: Tibetan (monastic style). Page layout: 15, 16, 22 lines / panel, light margins, guidelines in black. Content: (i) Pañca-skandha, (ii) Text on the bodhisattva path.

28 IOL Tib J 612 vol. 68:72 Raw material: Rag paper: Boehmeria nivea sp. (Ramie). Type of paper: One layer of laid paper with irregular structure: 12‒15 laid lines in 3 cm; chain lines intervals distance: 5‒9 cm. Find site: Dunhuang Cave 17. Dating: 9th–10th c. Format: Scroll (vertical). Size (h × w): 41.8 × 26.5‒27 cm. Language/script: Tibetan (monastic style). Page layout: 15, 16, 22 lines / panel, light margins, guidelines in black. Content: (i) Pañca-skandha, (ii) Text on the bodhisattva path. 238 appendix 1

29 Or.8210/S.9498 (A): f. 1A bottom margin, close to missing part) Raw material: Rag paper: Rag paper: Boehmeria nivea sp. (Ramie) and Cannabis sativa sp. (Hemp). Type of paper: Laid paper glued of 2 layers: 12 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: 9th–10th c. Format: Fragment of pothi or concertina. Size (h × w): 9.5 × 24 cm (upper right corner is missing). Language/script: Tibetan (sutra style). Page layout: 6 lines, guidelines in red. Content: Historical fragment related to the Dba’ bzhed.

30 IOL Tib J 1404 vol. 72 f.72 (bottom left edge) Raw material: Rag paper: Boehmeria nivea sp. (Ramie) and Cannabis sativa sp. (Hemp). Type of paper: Laid paper: 12 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: 9th–10th c. Format: Scroll fragment (vertical). Size (h × w): 61 × 31 cm. Language/script: Tibetan (official style) and Chinese. Page layout: 29 lines on fragment, no margins or guidelines. Content: Recto: Tibetan transliteration of Chinese Sukhāvatīvyūha sūtra. Verso: various memoranda concerning provisions.

31 OR 15000/496: Left Bottom Corner Close to the Seal Raw material: Rag paper: Boehmeria nivea sp. (Ramie), Cannabis sativa sp. (Hemp) and unidentified addition. Type of paper: Laid paper: 12 laid lines in 3 cm. Find site: Miran. Dating: mid-8th to mid-9th c. Format: Sheet. Size (h × w): 25 × 7.6 cm. Language/script: Tibetan (official style, headed). Page layout: 8 lines, and seal with design of horse and rider, and the syllables phyag rgya. Content: Official dispatch from an assembly convoked by Blon Mtsho bzher. TIBETAN COLLECTIONS OF The british library 239

32 IOL Tib 687 vol. 33:109–111 Raw material: Rag paper: Boehmeria nivea sp. (Ramie) and Cannabis sativa sp. (Hemp). Type of paper: Laid paper: 15 laid lines in 3 cm. Find site: n/a. Dating: early to mid-9th c. Format: Pothi, one hole, with red circle. Size (h × w): 45 × 8.5 cm. Language/script: Tibetan (monastic style). Page layout: 6 lines / page, margins in red. Content: Compilation from the scriptures, composed by ’Go Chos grub (Facheng) and possibly written in his hand.

33 IOL Tib J 321 vol. 23:13 Raw material: Rag paper: Boehmeria nivea sp. (Ramie), addition of Broussonetia sp. (Paper mulberry)/ Cannabis sativa sativa sp. (Hemp) possible. Type of paper: Laid paper: 15 laid lines in 3 cm, chain lines hardly visible. Find site: Dunhuang Cave 17. Dating: 10th c. Format: Pothi, two holes, with circles. Size (h × w): 9.8‒10 × 31 cm. Language/script: Tibetan (post-imperial Buddhist style, headless). Page layout: 5 lines / page, no margins or guidelines. Content: Upayapaśa tantra.

34 IOL Tib J 321 vol. 23:15 Raw material: Rag paper: Boehmeria nivea sp. (Ramie), addition of Broussonetia sp. (Paper mulberry) and Cannabis sativa sp. (Hemp) possible. Type of paper: Laid paper: 15 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: 10th c. Format: Pothi, two holes, with circles. Size (h × w): 9.8‒10 × 31 cm. Language/script: Tibetan (post-imperial Buddhist style, headless). Page layout: 5 lines / page, no margins or guidelines. Content: Upayapaśa tantra.

35 IOL Tib J 321 vol. 23:100 (verso, bottom right corner) Raw material: Rag paper: Boehmeria nivea sp. (Ramie), addition of Broussonetia sp. (Paper mulberry) and Cannabis sativa sp. (Hemp) possible. 240 appendix 1

Type of paper: Laid paper (thin one-layer): 18 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: 10th c. Format: Pothi, two holes, with circles. Size (h × w): 9.8‒10 × 31 cm. Language/script: Tibetan (post-imperial Buddhist style, headless, dbu med). Page layout: 5 lines / page, no margins or guidelines. Content: Upayapaśa tantra.

36 IOL Tib J 470 (upper part) Raw material: Rag paper: Boehmeria nivea sp. (Ramie), addition of Cannabis sativa sp. (Hemp) and Broussonetia sp. (Paper mulberry) possible. Type of paper: Laid paper: 12 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: 10th c. Format: Scroll (vertical). Size (h × w): 30.5 × 399 cm. Language/script: Tibetan (sutra style, with interlinear notes in a post-imperial style). Page layout: ~22 lines/panel, with interlinear notes, margins, and guidelines in black. Content: Tantric treatise: Rdo rje sems dpa’i zhus lan. Scribe is Chinese official.

37 IOL Tib J 105 vol. 5:2 (bottom right edge from damaged area near corner) Raw material: Rag paper: Boehmeria nivea sp. (Ramie), addition of Cannabis sp. (Hemp) and Broussonetia sp. (Paper mulberry) possible. Type of paper: Laid paper glued in layers: 12 laid lines in 3 cm, chain lines interval in distance 6‒7.5 cm when visible. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Pothi with holes for a string. Size (h × w): 20.4 × 73.2 cm. Language/script: Tibetan (sutra style). Page layout: 12 lines / page, margins and guidelines in black. Content: Śatasāhasrikā prajñāpāramitā.

38 IOL Tib J 109.21 vol. 87:8, bi-f. 2 (bottom right hand corner) Raw material: Rag paper: Boehmeria nivea sp. (Ramie), Cannabis sativa sp. (Hemp) and Broussonetia sp. (Paper mulberry) possible. Type of paper: Laid paper: 12 laid lines in 3 cm; chain lines not visible. Find site: Dunhuang Cave 17. TIBETAN COLLECTIONS OF The british library 241

Dating: early to mid-9th c. Format: Scroll (horizontal); presently 3 separated scroll sections: 26.5‒28.2 × 48.2 cm in bi-folio format. Size (h × w): 26.5‒28.2 × 48.2 cm. Language/script: Tibetan (sutra style). Page layout: 16 lines / section, margins and guidelines in black. Content: Śatasāhasrikā prajñāpāramitā.

39 IOL Tib J 1587 vol. 87:16, bi-f. 1 (top edge near corner); 2 (bottom edge close to left corner); 5 (left bottom corner) Raw material: Rag paper: Boehmeria nivea sp. (Ramie), Cannabis sativa sp. (Hemp) and Broussonetia sp. (Paper mulberry) possible. Type of paper: Laid paper: 15 laid lines in 3 cm with hardly visible laid structure of paper in some places. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll (horizontal). Size (h × w): Presently 5 separated scroll sections 26 × 38.5‒39.5 cm in bi-folio format. Language/script: Tibetan (sutra style). Page layout: 15 lines / section, no margins or guidelines. Content: Śatasāhasrikā prajñāpāramitā.

40 IOL Tib J 1718 vol. 120:2, bi-f. 1 (right edge near top corner) Raw material: Rag paper: Boehmeria nivea sp. (Ramie), Cannabis sativa sp. (Hemp) and Broussonetia sp. (Paper mulberry) possible. Type of paper: Laid paper: 15‒18 laid lines in 3 cm (bi-f. 1) and 18‒21 laid lines in 3 cm (bi-f. 2). Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll (horizontal). Size (h × w): Presently 2 separated scroll sections: 26.5‒28 × 42‒49 cm in bi-folio format. Language/script: Tibetan (sutra style). Page layout: 15 lines / section, margins and guidelines in black. Content: Śatasāhasrikā prajñāpāramitā.

41 IOL Tib J 82: vol. 22 F. 2–67: f. 11 and 51 Raw material: Rag paper: Boehmeria nivea sp. (Ramie), Cannabis sativa sp. (Hemp) and addition of Broussonetia sp. (Paper mulberry) in folio 11. Type of paper: Laid paper: 15 laid lines in 3 cm. Find site: Dunhuang Cave 17. 242 appendix 1

Dating: 9th–10th c. Format: Pothi. Size (h × w): 32 × 11.5 cm. Language/script: Tibetan (post-imperial Buddhist style, headless, dbu med). Page layout: 8–11 lines / page, no margins or guidelines. Content: Daśabhūmika sūtra.

42 IOL Tib J 587 vol. 68:79 Raw material: Rag paper: Boehmeria nivea sp. (Ramie), Cannabis sativa sp. (Hemp), and Broussonetia sp. (Paper mulberry) possible. Type of paper: One layer of laid paper with irregular structure: 12‒15 laid lines in 3 cm (reed/grass papermaking sieve). Find site: Dunhuang Cave 17. Dating: 9th–10th c. Format: Originally concertina, now six loose leaves. Size (h × w): 6.5 × 13.8 cm. Language/script: Tibetan (sutra style). Page layout: 5 lines / panel, no margins or guidelines. Content: Buddhist prayer.

43 IOL Tib J 1726 vol. 120:10, bi-f. 1 & 6 (top right edge) Raw material: Rag paper: Boehmeria nivea sp. (Ramie), Cannabis sativa sp. (Hemp) and Broussonetia sp. (Paper mulberry) possible. Type of paper: Laid paper: 12‒15 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll (horizontal). Size (h × w): Presently 6 separated scroll sections: 29.5‒30 × 41.5‒43 cm in bi-folio format. Language/script: Tibetan (sutra style). Page layout: 15 lines on section, margins and guidelines in black. Content: Vajracchedika-prajñāpāramitā.

44 IOL Tib J 1354 vol. 69:32 Raw material: Rag paper: Boehmeria nivea sp. (Ramie)/ Broussonetia sp. (Paper mulberry). Type of paper: Laid paper: 12 laid lines in 3 cm (grass/reed papermaking sieve). Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll fragment (horizontal). TIBETAN COLLECTIONS OF The british library 243

Size (h × w): n/a. Language/script: Tibetan (sutra style). Page layout: 16 lines on fragment, margins and guidelines in black. Content: Śatasāhasrikā prajñāpāramitā.

45 IOL Tib J 1355 vol. 69:32 Raw material: Rag paper: Boehmeria nivea sp. (Ramie)/ Broussonetia sp. (Paper mulberry). Type of paper: Structure hardly visible (grass/reed papermaking sieve). Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll fragment (horizontal). Size (h × w): n/a. Language/script: Tibetan (sutra style). Page layout: 16 lines on fragment, margins and guidelines in black. Content: Śatasāhasrikā prajñāpāramitā.

46 IOL Tib J 1356 vol. 69:32 Raw material: Rag paper: Boehmeria nivea sp. (Ramie) /Broussonetia sp. (Paper mulberry). Type of paper: Laid paper: 9‒12 laid lines in 3 cm (grass/reed papermaking sieve). Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll fragment (horizontal). Size (h × w): n/a. Language/script: Tibetan (sutra style). Page layout: 16 lines on fragment, margins and guidelines in black. Content: Śatasāhasrikā prajñāpāramitā.

47 IOL Tib J 232(a) vol. 69:16 Raw material: Bark paper: Broussonetia sp. (Paper mulberry) + Boehmeria nivea sp. (Ramie) possible. Type of paper: Irregular laid paper: 12 laid lines in 3 cm; chain lines not visible; papermaking sieve made of grass/reed. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll fragment (horizontal). Size (h × w): 25 × 27 cm. Language/script: Tibetan (sutra style). 244 appendix 1

Page layout: 16 lines on section, margins and guidelines in black. Content: Pañcaviṃśatisāhasrikā-prajñāpāramitā.

48 IOL Tib J 1581 vol.87:10, bi-f. 1 (top edge toward the centre) Raw material: Bark paper: Broussonetia sp. (Paper mulberry) + Cannabis sativa sp. (Hemp) possible. Type of paper: Woven paper. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll (horizontal). Size (h × w): Presently 4 separated scroll sections: 28.5 × 43 cm in bi-folio format. Language/script: Tibetan (sutra style). Page layout: 17 lines on section, margins and guidelines in black. Content: Śatasāhasrikā prajñāpāramitā.

49 IOL Tib J 1586 vol. 87:15, bi-f. 1 (bottom edge near corner) Raw material: Bark paper: Broussonetia sp. (Paper mulberry), Cannabis sativa sp. (Hemp) possible. Type of paper: Both types: woven paper and laid paper characterized by 18 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll (horizontal). Size (h × w): Presently 7 separated scroll sections of different size 26.5‒30 × 28‒45 cm. Language/script: Tibetan (sutra style). Page layout: 14, 17, 18, and 19 lines of text on particular fragments respectively, margins and guidelines in black (faint). Content: Śatasāhasrikā prajñāpāramitā.

50 IOL Tib J 241 vol. 69:9 Raw material: Bark paper: Broussonetia sp. (Paper mulberry) and Cannabis sativa sp. (Hemp) possible. Type of paper: Unclear structure suggesting that papermaking sieve is made of grass/ reed; in fragments 9‒12 laid lines in 3 cm visible. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll fragment (horizontal). Size (h × w): 20 × 28 cm. Language/script: Tibetan (sutra style). Page layout: 18 lines in section, no margins or guidelines. Content: Sūtra fragment: Prajñāpāramitā. TIBETAN COLLECTIONS OF The british library 245

51 IOL Tib J 1560 vol. 85:2, bi-f. 2 (bottom edge, toward the middle of the edge) Raw material: Bark paper: Broussonetia sp. (Paper mulberry) and Cannabis sativa sp. (Hemp) possible. Type of paper: Larger fragment on woven paper and smaller fragment on laid paper characterized by 15 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll (horizontal). Size (h × w): Presently 2 separated fragments (approximately 26.5 × 27 cm). Language/script: Tibetan (sutra style). Page layout: 13 and 16 lines on two fragments of scroll, no margins or guidelines. Content: Śatasāhasrikā prajñāpāramitā.

52 IOL Tib J 1561 vol.85: 4, bi-f. 1 (top right corner near damaged area) Raw material: Bark paper: Broussonetia sp. (Paper mulberry)/ Cannabis sativa sp. (Hemp) possible. Type of paper: Woven paper. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll (horizontal). Size (h × w): presently one bi-folio (fragment 28.5 × 38 cm). Language/script: Tibetan (sutra style). Page layout: 17 lines on section, no margins or guidelines. Content: Śatasāhasrikā prajñāpāramitā.

53 IOL Tib J 1577 vol. 87:3, bi-f. 4 (bottom of page right hand side of the corner) Raw material: Bark paper: Broussonetia sp. (Paper mulberry)/ Cannabis sativa sp. (Hemp) possible. Type of paper: Laid paper glued of layers in 3 bi-folios: 18 laid lines in 3 cm, Woven paper in 1 bi-folio; 2 bi-folios lined with backing paper during conservation. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll (horizontal). Size (h × w): Presently 4 separated scroll sections: 27.8‒28 × 44.3 cm in bi-folio format. Language/script: Tibetan (sutra style). Page layout: 18 lines on every scroll section, no margins or guidelines. Content: Śatasāhasrikā prajñāpāramitā. 246 appendix 1

54 IOL Tib J 1578 vol. 87:4, bi-f. 3 (bottom of page right hand side of the corner) Raw material: Bark paper: Broussonetia sp. (Paper mulberry), Cannabis sativa sp. (Hemp) possible. Type of paper: One-layered paper of two types: woven paper in 2 sections and laid paper characterized by 18 laid lines in 3 cm in section 13 of the original scroll. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll (horizontal). Size (h × w): Presently 4 separated scroll sections: 27.8‒28.2 × 44 cm in bi-folio format. Language/script: Tibetan (sutra style). Page layout: 17 lines on section, no margins or guidelines. Content: Śatasāhasrikā prajñāpāramitā.

55 IOL Tib J 570 vol. 13:79 Raw material: Bark paper: Broussonetia sp. (Paper mulberry). Type of paper: Laid paper glued of two layers: 18 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: 10th c. Format: Pothi, one string hole, no string. Size (h × w): 9 × 26.5 cm. Language/script: Tibetan (post-imperial Buddhist style, headed, dbu can, and headless, dbu med). Page layout: 6 lines / page, margins and guidelines in red. Content: Gtor ma offering ritual.

56 IOL Tib J 754 (A): vol. 72 f.74 (left side margin close to upper corner) Raw material: Bark paper: Broussonetia sp. (Paper mulberry). Type of paper: Laid paper: 18 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: ca. 968. Format: Scroll (vertical). Size (h × w): 63 × 24.5 cm. Language/script: Tibetan (post-imperial epistolary style) and Chinese. Page layout: Scattered notes, no margins or guidelines. Content: Tibetan letters of passage for a Chinese pilgrim. TIBETAN COLLECTIONS OF The british library 247

57 IOL Tib J 754 (A): vol. 72 f.77 (central part of right side margin edge) Raw material: Bark paper: Broussonetia sp. (Paper mulberry). Type of paper: Laid paper: 18 laid lines in 3 cm; chain lines interval in distance 5–5, 3–5–5–5–5–4–1–4–4.5–1–5.5 cm. Find site: Dunhuang Cave 17. Dating: ca. 968. Format: Scroll (vertical). Size (h × w): 63 × 25.5 cm. Language/script: No text. Page layout: n/a. Content: Tibetan letters of passage for a Chinese pilgrim.

58 IOL Tib J 754 (C): vol. 72 f.78 (central part of bottom edge) Raw material: Bark paper: Broussonetia sp. (Paper mulberry). Type of paper: Laid paper: 18 laid lines in 3 cm. Find site: Dunhuang Cave 17. Dating: 968. Format: Sheet. Size (h × w): 25.5‒26 × 34–35 cm. Language/script: Chinese. Page layout: 16 columns on sheet. Content: Chinese copy of a stele inscription from Liangzhou.

59 IOL Tib J 754 (B): vol. 72 f.79 (middle part of right side margin) Raw material: Bark paper: Broussonetia sp. (Paper mulberry). Type of paper: Woven paper, two layers glued together. Find site: Dunhuang Cave 17. Dating: ca. 968. Format: Scroll (vertical). Size (h × w): 69 × 25 cm. Language/script: Chinese and Tibetan (post-imperial style). Page layout: Recto: 28 columns / panel Verso: 41 lines, no margins or guidelines. Content: Recto: Baoenjing (Chinese) Verso: Tibetan tantric texts.

60 IOL Tib J 754(B): vol. 72 f.83 Raw material: Bark paper: Broussonetia sp. (Paper mulberry). Type of paper: One layer of paper with unclear laid structure. Find site: Dunhuang Cave 17. Dating: ca. 968. 248 appendix 1

Format: Scroll (vertical). Size (h × w): 53.5 × 25 cm. Language/script: Chinese and Tibetan (post-imperial style). Page layout: Recto: 28 columns / panel Verso: 32 lines, no margins or guidelines. Content: Recto: Baoenjing (Chinese) Verso: Tibetan tantric texts.

61 IOL Tib J 750 Raw material: Bark paper: Broussonetia sp. (Paper mulberry). Type of paper: Woven paper (?) Document after conservation—paper is sandwiched in between silk gauze (very difficult access to original material). Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll (vertical). Size (h × w): n/a. Language/script: Tibetan and Chinese. Page layout: Recto: 28 columns/panel; Verso: ~33 lines/panel, faint margins and guidelines. Content: Recto: Saddharmapuṇḍarīka sūtra (Chinese); Verso: Old Tibetan Annals Version I.

62 Or.8212/187 Raw material: Bark paper: Broussonetia sp. (Paper mulberry). Type of paper: Woven paper. Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll (vertical). Size (h × w): 27 × 135 cm. Language/script: Tibetan and Chinese. Page layout: Recto: 28 columns/panel; Verso: variable, no margins or guidelines. Content: Recto: Bhaiṣajyaguru sūtra (Chinese); Verso: Old Tibetan Annals Version II.

63 IOL Tib J 1732 vol. 120:16, bi-f. 1 (bottom edge) Raw material: Bark paper: Broussonetia sp. (Paper mulberry). Type of paper: Unclear patchy structure of paper: laid structure of paper visible in some places (15‒18 laid lines in 3 cm). Find site: Dunhuang Cave 17. Dating: early to mid-9th c. Format: Scroll (horizontal). TIBETAN COLLECTIONS OF The british library 249

Size (h × w): Presently 6 separated scroll sections: 28.5 × 45.5 cm in bi-folio format. Language/script: Tibetan (sutra style). Page layout: 15 lines on section, no margins or guidelines. Content: Śatasāhasrikā prajñāpāramitā.

64 IOL Tib J 194 vol.13:26 Raw material: Bark paper: Daphne or Edgeworthia sp. Type of paper: Woven paper (?) with a surface print of fabric on which the paper was dried. Find site: Dunhuang Cave 17. Dating: late 8th to early 9th c. Format: Pothi: One hole, no circle. Size (h × w): 8 × 30 cm. Language/script: Tibetan (square style with archaic features). Page layout: 5 lines / page, margins and guidelines in black. Content: Samdhinirmocana sūtra.

65 IOL Tib J 194 vol. 13:30 Raw material: Bark paper: Daphne or Edgeworthia sp. Type of paper: Woven paper (?) with a surface print of fabric on which the paper was dried. Find site: Dunhuang Cave 17. Dating: late 8th to early 9th c. Format: Pothi, one hole, no circle. Size (h × w): 8 × 30 cm. Language/script: Tibetan (square style). Page layout: 5 lines / page, margins and guidelines in black. Content: Samdhinirmocana sūtra.

66 IOL Tib J 1459 vol. 73:37 Raw material: Bark paper: Daphne or Edgeworthia sp. Type of paper: Woven paper with a surface print similar to laid lines. Find site: Dunhuang Cave 17. Dating: late 8th to early 9th c. Format: Folded document. Size (h × w): 16 × 16 cm (fold distance 3 cm). Language/script: Tibetan (official style, headed and headless). Page layout: 11 lines, no margins or guidelines. Content: Letter from Central Tibet, Tibetan imperial period. 250 appendix 1

67 IOL Tib M 58 vol. 1:58 Raw material: Bark paper: Daphne or Edgeworthia sp. Type of paper: Woven paper, two layers glued together. Find site: Khara Khoto. Dating: 12th–13th c. Format: Pothi. Size (h × w): 10 × 30‒32 cm. Language/script: Tibetan (post-imperial Buddhist style, headless). Page layout: 8 lines / page, faint black margins, no guidelines. Content: Commentary on a Buddhist treatise on service to the teacher.

68 IOL Tib M 59 vol. 1:59 Raw material: Bark paper: Daphne or Edgeworthia sp. Type of paper: Woven paper, two layers glued together. Find site: Khara Khoto. Dating: 12th–13th c. Format: Pothi. Size (h × w): 10 × 30‒34 cm. Language/script: Tibetan dbu med (post-imperial Buddhist style, headless). Page layout: 5 lines /page, with interlinear notes, red margins, no guidelines. Content: Commentary on tantric practice.

69 IOL Tib M 60 vol. 1:60 Raw material: Bark paper: Daphne or Edgeworthia sp. Type of paper: Woven paper glued in layers. Find site: Khara Khoto. Dating: 12th–13th c. Format: Pothi. Size (h × w): 9 × 32‒32.5 cm. Language/script: Tibetan dbu med (post-imperial Buddhist style, headless). Page layout: 9 lines / page, red margins, no guidelines. Content: Tantric ritual text.

70 OR 15000/513: left side margin Raw material: Mixed components: Daphne or Edgeworthia sp. + Broussonetia sp. (Paper mulberry) and straw possible. Type of paper: Woven paper glued of 2 layers with the surface print similar to laid lines. Find site: Miran. Dating: late 8th to early 9th c. TIBETAN COLLECTIONS OF The british library 251

Format: Folded document. Size (h × w): 12 × 2.5 cm. Language/script: Tibetan, dbu med (official style, headed). Page layout: 4 lines (recto), no margins or guidelines, verso contains seal with bird design, and the syllables pho brang. Content: Letter to Jo co Lha ’brul, perhaps from Central Tibet. Appendix 2 Features of Paper in Selected Sets of Tibetan Kanjur

Set Examined Present Location Dating Place of Format Size of Page outline Ink Raw material Papermaking sieve Construction of fragment Content production leaves (h × w) No. of verses // Fiber composition print (number of laid the leaf and visual Distances between lines in 3 cm) properties of its upper lines of text // surface Type of frame

Yongle Kanjur: Special Collections 1410 Beijing 24.2–24.5 cm 8 Red Paper mulberry Laid paper Six or more layers the folio 12 of the Library at the [Peking] in height and 1.5–1.7 cm (Broussonetia sp.) characterized by glued together; good Samādhirāja-sūtra University of 68.7–69 cm in Side margins 27–30 laid lines in quality and well Michigan, Ann length apart 59 cm 3 cm; chain lines preserved paper Arbor, USA from each other not possible to see

Wanli Kanjur: Jagiellonian 1606 Beijing 23.8–24.5 cm in 8 Black Paper mulberry Laid paper Six or more layers the twenty-eight University Library [Peking] height and 68.5 cm 1.5–1.7 cm (Broussonetia sp.); characterized by glued together; volumes: Nos. 23–28 in Cracow, Poland in length Side margins within the Wanli 24 laid lines in 3 cm; the paper on which (rgyud), 38–57 (Biblioteka apart 59–59.5 cm set the best quality some of the papers the laudation text is (nos. 38–55 = rgyud; Jagiellońska w from each other (longest) fibers show the presence written has about 56 = mdo; 57 =rgyud), Krakowie) were used for of the 15 or 18 laid 4 layers, which 59=rgyud), volume 60 lines in 3 cm; chain makes the leaves 60=dkar chag) (dkar chag); the lines visible in thinner; good quality volumes of the volume 28: distances and well preserved rgyud (Tantra) between them as paper section were follows: 3 cm–2 cm– produced on 3 cm–2 cm–2.5 cm– medium quality 2,5 cm–2.5 cm– raw material 2.5 cm paper.

Wanli Kanjur Jagiellonian 1606–1607 Beijing Approximately 6 Black Paper mulberry Laid paper Six or more layers Supplement: University Library [Peking] 23.8–24.5 cm in 1.7–2.2 cm (Broussonetia sp.), glued together Pander Pantheon: in Cracow, Poland height and 68.5 cm Side margins straw, and bamboo volumes 1–22 (Biblioteka in length apart 58.5–59 cm Jagiellońska w from each other Krakowie) Wanli Kanjur Harvard-Yenching 1606–1607 Beijing Approximately 6 Black Paper mulberry Laid paper Six or more Supplement: [Peking] 23.8–24.5 cm in 1.7–2.2 cm volumes tsa and ku of Library in height and 68.5 cm Side margins (Broussonetia sp.), layers glued together the Wanli Supplement Cambridge MA, in length apart 58.5–59 cm straw, and bamboo USA from each other

Table (Continued)

‘Handwritten Kanjur Berlin State 1680 Beijing Approximately 8 Black Paper mulberry Laid paper Three or more layers of Berlin’: Library, Germany [Peking] 26.8–29 cm in height 1.8–1.9 cm and red (Broussonetia sp.), characterized by glued together; vol. 1, ka; vol. 24 Teil 1 (Staats-bibliothek and 68.9–73.3 cm Full frame bamboo, and jute 27–33 laid lines in highly sized and and Teil 2, nya; vol. 17 zu Berlin) in length 18.8×55 cm 3 cm polished; possibly Teil 1 and Teil 2, ka; vol. also covered with 18 Teil 1 and Teil, kha; other substances vol. 108 Teil 1 and which increased its Teil 2, ja; vol. 85 Teil 1 whiteness and Teil 2, ka plus 64 folios of dkar chag attached to the top of this volume)

One of the several Jagiellonian 1684–92 Beijing Approximately 8 Red Paper mulberry Laid paper with Three or more layers corrected reprints of University Library [Peking] 24.5–25.5 cm in 1.5–1.7 cm (Broussonetia sp.), hardly visible glued together the Kangxi Kanjur in Cracow, Poland height and 71.5 cm Full frame bamboo, and straw structure edition: sher phyin, (Biblioteka in length 19×59 cm tha (vol. 58 from the Jagiellońska w Pander Collection) Krakowie)

Handwritten Mongolian St Petersburg Not earlier 23.3 cm in height Full frame Black Paper mulberry Laid regular paper Two layers glued Kanjur University, Faculty than 1629 and 68 cm in length 20.5×61.9 cm (Broussonetia sp.), characterized by together First volume of ka of Asian and and bamboo 22 laid lines in 3 cm, tantra section African Studies chain line intervals collection, Saint 3.5–4 cm Petersburg, Russia

Printed Mongolian St Petersburg Beginning 22.3 cm in height and Full frame: Black Bamboo Laid regular paper Two or three layers Kanjur from Kangxi University, Faculty of 17th c. 65.9 cm in length 17.3×55 cm and red characterized by of paper glued times; tantra of Asian and 36 laid lines in 3 cm, together volume (E59) African Studies chain line intervals collection, Saint not visible Petersburg, Russia features of paper in SELECTED sets of tibetan kanjur 255

Table (Continued)

Printed edition of St Petersburg 1714 Beijing 23.7 cm in height Black Paper mulberry Laid regular paper Two layers of paper Mongolian Kanjur (E2), University, Faculty [Peking] and 67 cm in length and red (Broussonetia sp.), characterized by 36 glued together vol. 9, No. 19 in of Asian and bamboo and jute or even 39 laid lines Uspensky Catalog African Studies in 3 cm; chain line Heissing Blockdruke collection, Saint intervals can be no. 26 Petersburg, Russia 3.5–4 or 5 cm, but very hardly visible due to paper glued in layers

Cone Kanjur: vol. 1 Library of Congress, 1721–31 Cone Approximately 8 Black Paper mulberry Laid paper Slightly yellowish mdo (sūtra), ka ; vol. 35 Washington, (The 18–18.5 cm in height 0.9–1.3 cm (Broussonetia sp.) characterized by (cream), pasted in ’dul-ba (vinaya), ga; DC, USA examined and 56.5–57 cm in Full frame and straw about 27 laid lines two or possibly three vol. 92, yum, ka; copy was length 10.3×47 cm in 3 cm; chain lines layers; absorbent vol. 108 dkar chag printed in not visible; a paper and soft; glued using (index); vol. 72 rgyud 1926.) fragment with a title diluted starch (tantra); Original (label) printed on a paste rather than Paper Strings from lotus flower in the any kind of animal Cone Kanjur first volume [ka] of glue; two leaves mdo (sūtra) has (ms. folio 244 in the same chain and vol. 35 and ms. folio laid lines pattern as 289 in vol. 92) paper leaves; here handwritten on it is possible to much thicker paper observe chain lines in which more layers in a distance of were pasted together 3.5–4 cm from each to allow for writing other with a bamboo or wooden stick. The surface of the leaf was polished with a stone before writing. features of paper in SELECTED sets of tibetan kanjur 257

Table (Continued)

Narthang Kanjur: Berlin State 1730–32 Narthang Approximately 7 Black Thymelaeaceae Both a woven type Very thin one- vol. 1 ’dul-ba (vinaya) , Library, Germany 17–18 cm in height 1.5 cm family plants of paper made with layered paper with ka, and vol. 2 ’dul-ba (Staats-bibliothek and 61.5–63.5 cm in Full frame (Daphne or thick textile as a glossy surface and (vinaya), kha zu Berlin) length; particular 11.5× 58 cm Edgeworthia sp.) woven sieve and uneven thickness; folios are not evenly possibly a finely the middle part of cut woven cotton sieve, the volume was and a laid paper printed on visibly characterized by worse quality paper; about 15 laid lines fiber bundles and in 3 cm were used outer bark particles during production in the structure of of this paper paper

Derge Kanjur: Library of Congress, 1733 Derge Approximately 7 Red Thymelaeaceae Laid paper One-layered, soft vol. 1 ’dul-ba (vinaya), Washington, DC, 10–11.5 cm in height 0.9–1 cm family plants characterized by and absorbent paper ka; vol. 45, mdo (sūtra) USA and 60.5–62 cm in Full frame (Daphne or 15 laid lines in 3 cm; characterized by a volume, ka; vol. 77 length 7.2–7.4 × Edgeworthia sp.) chain lines in a brownish color and rgyud (tantra), ka 50.9–52 cm distance 3–4 cm in slightly glossy volumes 1 and 45; surface (possibly paper in rgyud polished); the volume 77 produced thickness of the by means of a mold paper differs in with a woven type different leaves; of sieve many fiber bundles and fragments of outer bark in its structure caused by an inadequate amount of well- separated fibers features of paper in SELECTED sets of tibetan kanjur 259

Table (Continued)

Lhasa Kanjur: Berlin State Library, 1934 Lhasa 15.5–17.5 cm in 7 1.3–1.6 cm Black Thymelaeaceae Both a woven type One-layered, very vol. 1,’dul-ba (vinaya), Germany (Staats- height and 62.8–64 Full frame family plants of mold and a laid thin paper with ka and vol. 2, ’dul-ba bibliothek zu cm in length; 11.8×57 cm (Daphne or mold characterized visible fiber bundles (vinaya), kha Berlin) particular leaves are Edgeworthia sp.) by about 21 laid in its structure not evenly cut lines in 3 cm; the and a slightly woven mold had a glossy surface; the sieve made of loosely thickness of the woven textile, which paper and its quality is clearly visible in varies: some leaves the paper structure are almost brown with more particles of outer bark in their structure features of paper in SELECTED sets of tibetan kanjur 261

Table (Continued)

Lhasa Kanjur: Berlin State Library, 1934 Lhasa 15.5–17.5 cm in 7 1.3–1.6 cm Black Thymelaeaceae Both a woven type One-layered, very vol. 1,’dul-ba (vinaya), Germany (Staats- height and 62.8–64 Full frame family plants of mold and a laid thin paper with ka and vol. 2, ’dul-ba bibliothek zu cm in length; 11.8×57 cm (Daphne or mold characterized visible fiber bundles (vinaya), kha Berlin) particular leaves are Edgeworthia sp.) by about 21 laid in its structure not evenly cut lines in 3 cm; the and a slightly woven mold had a glossy surface; the sieve made of loosely thickness of the woven textile, which paper and its quality is clearly visible in varies: some leaves the paper structure are almost brown with more particles of outer bark in their structure Appendix 3 Transcription of the Yongle Covers Inscriptions

Transcription of Figures 75 and 76. The fragment of verso side of two Yongle covers of Tibetan Kanjur (75. originally Sera Monastery set, and 76. Sakya Monastery set) containing an engraved and gilded seals on its verso sides. Both covers are from the R.R.E. Collection in Winterthur, Switzerland (ET 107 and ET 87). As seen here all upper covers feature a summary of contents in Tibetan and Chinese engraved on the verso side. This transcription and translation was prepared by Dr. Cynthia Col with assistance of Dr. Yudru Tsomu (the Center for Tibetan Studies of Sichuan Province), Bugang Chashingtsang, and Steven Weinberger (Tibetan & Himalayan Library).

Verso of Yongle Cover (ET 107) Inscription (Sera set type Showed on Figure 75)

||mdo sna tshogs kyi ba dum la| ‘Miscellaneous Sutras’ in the ‘ba’ section. This statement is followed by a list of thirteen titles in Tibetan and Chinese.

1 ‘phags pa tshangs pas zhus pa zhes bya ba theg pa chen po’i mdo1 大梵天王请问经 (Dafan tianwang qingwen jing) 一卷内 (yi juan jing) “one volume” 2 ‘phags pa tshangs pa byin gyis zhus pa zhes bya ba theg pa chen po’i mdo 聖施大梵天請問大乘经 (sheng shi da fantian qingwen dacheng jing) 3 ‘phags pa tshangs pa khyad par sems kyis zhus pa zhes bya ba theg pa chen po’i mdo| 勝思惟梵天請問大乘经 (sheng siwei fantian qingwen dacheng jing) 4 ‘phags pa lha yi bu rab rtsal gyi sems kyi zhus pa zhes bya ba theg pa chen po’i mdo| 聖勇心天子請問大乘经 (shengyong xin tianzi qingwen da cheng jing) 5 ‘phags pa dpal dbyig gis zhus pa zhes bya ba theg pa chen po’i mdo | 聖者寶德請問大乘经 (shengzhe baode qingwen dacheng jing)

1 The Sanskrit title of this sutra is: Brahma-paripṛcchā-nāma-mahāyāna-sūtra: “Questions of Brahmā Mahāyāna Sūtra,” Thanks to Steven Weinberger of the Tibetan & Himalayan Library for identifying this text and for pointing out that it is found in the Derge Kanjur that is available at the TBRC in volume 59 (volume letter=ba) here: http://www.tbrc.org/#!rid=W22084.

6 ‘phags pa rin chen dra ba can gyi zhus pa zhes bya ba theg pa chen po’i mdo | 聖者寶網請問大乘经 (shengzhe baowang qingwen dacheng Jinga 7 ‘phags pa rin chen zla bas zhus pas zhes bya ba theg pa chen po’i mdo 聖寶月請問大乘经 ( shengbao yue qingwen dacheng jing) 8 ‘phags pa bde byed gyis zhus pa zhes bya ba theg pa chen po’i mdo| 聖者安隐大乘经 (shengzhe an yin dacheng jing) 9 ‘phags pa yul ‘khor skyob gis zhus ba zhes bya ba theg pa chen po’i mdo | 聖者持國天王請問大乘经 (shengzhe chiguo tianwang qingwen dacheng jing) 10 ‘phags pa rnam par ‘phrul pa’i rgyal pos zhus pa zhes bya ba thegs chen po’i mdo| 聖者能化現王請問大乘经 (shengzhe neng hua xian wang qingwen dacheng jing) 11 ‘phags pa thegs chen po’i man ngag ces bya ba theg pa theg pa chen po’i mdo | 聖大乘密意大乘经 (sheng dacheng mi yi dacheng jing) 12 ‘phags ba bram ze mo dpal ldan mas zhus pa zhes bya ba theg pa chen po‘i mdo| 聖具吉祥婆羅門女請問大乘经 (sheng ju jixiang poluomen nu qingwen dacheng jing) 13 ‘phags pa bgres mo mo zhus pa zhes bya ba theg ba chen po‘i mdo rnams bzhugs sto|| 聖者老姥請問大乘经 (shengzhe lao lao qingwen dacheng jing)

Verso of Yongle Cover (Et 87) Inscription (Originally the Sakya Monastery Set Preserved Later in Potala, Lhasa Shown on Figure 76)

||mdo sna tshogs kyi na dum la| ‘Miscellaneous Sutras’ in the ‘na’ section This statement is followed by a list of seven titles in Tibetan and Chinese.2

1 ‘phags pa da ltar gyis sangs rgyas mngon sum du bzhugs pa’i ting nge ‘dzin ces bya ba theg pa chen po’i mdo| 现在諸佛禅定经一卷内 (xian zai zhu fo chanding yi juan nei) 2 ‘phags pa bsod nams thams cad bsdus pa’i ting nge ‘dzin ces bya ba theg pa chen po’i mdo|| 聖妙集福禅定大乘经 (sheng miao jifu chanding dacheng jing) 3 ‘phags pa khye’u bzhi’i ting nge ‘dzin ces bya ba theg pa chen po’i mdo|| 聖者四孺童禅定大乘经 (sheng zhe si rutong dacheng jing)

2 In the Derge Kanjur, this is volume ‘na’ (volume 56) has nine texts. 264 appendix 3

4 ‘phags ba ting nge ‚dzin mchog dam pa|| 聖微妙勝上禅定经 (sheng weimiao shengshang chanding jing) 5 ‘phags pa ‘dus pa chen po rin po che’i tog gi gzungs zhes bya ba theg pa chen po’i mdo|| 聖者大聚宝宝顶陀罗尼大乘经 (sheng zhe da jubao baoding tuo luo ni dacheng jing) 6 rdo rje snying po zhes bya ba’i gzhungs| 金刚藏陀罗尼大乘经 (jingang zang tuo luo ni dacheng jing) 7 ‘phags pa sgo mtha’ yas pa sgrub pa zhes bya ba’i gzungs rnams bzhugs to|| 聖修习无量法门陀罗尼 (sheng xiuxi wuliang famen tuo luo ni) Appendix 4 Selected Items from the Tibetan Collection from the Staatsbibliothek zu Berlin—Preußischer Kulturbesitz (the Berlin State Library), Orientabteilung

‘Handwritten Kanjur of Berlin’ Acquired in 1889 by v. Brandt (Peking Kanjur) Handwritten with Black and Red Ink

1 Volume Peking Kanjur 1 (Preserved in 2 Boxes) ‘ka’ Raw material: Bamboo, straw, jute, and paper mulberry. Type of paper: Laid paper (two or three layers): 33 laid lines in 3 cm; paper highly sized and polished; possibly also covered with other substances, such as starch glue, which make it whiter. Find site: Beijing [Peking]. Dating: 1680. Format: Loose leaves pothi. Size (h × w): 26.8 × 73.2 cm. Language/script: Tibetan (dbu can, block-print, ‘headed letters’). Calligraphy: handwritten—very skilled calligraphy characterized by longitudinal proportions of letters (letters with ‘long legs:’ ‘ka,’ ‘ta,’ and ‘na,’ etc. Perfectly balanced composition of the letters and the light between the letters). Page layout: 4–8 lines / page. Folio 1: recto—4 lines of text, first and third lines in red (distance between upper lines is 4 cm), verso 5 lines of text, second and fourth lines in red (distance between upper lines is 3.2 cm). Folio 1: missing. Folio 2: recto—6 lines of text; first, third, and fifth lines in red (distance between upper lines of text is 2.6 cm), verso 7 lines of text with third and fifth written in red (distance between upper lines of text is 2.2–2.3 cm). Folio 3 and the rest of the volume: 8 lines of text: third and sixth lines written in red (rubrics); distance between upper lines is 1.8–1.9 cm. Distance between letterhead lines in the rest of the book: 1.8 cm (very fine), text- base lines are sketched with a lead stylus or other slightly sharp tool at the upper part of the text (just one line at the text head); A frame around the written area was traced with a 0.3 cm black line that surrounds a thin red slightly inside it (18.8 × 55 cm). Content: Tibetan Buddhist Canon; Kanjur. Decorations: Fore-edges are painted with lotus flowers and auspicious symbols against a yellow background, on the short edges—there is the lotus flower that features a precious pearl in flames and Tibetan text in the halo.

2 Volume Peking Kanjur 24 (Preserved in 2 Boxes), ‘nya’ same text as in the volume from Pander Collection: 47 / 1108, Kanjur W, 30. sher phyin (Prajñāpāramitā), nya (8), 1108–25/47 NYA. Raw material: Bamboo, straw, jute, and paper mulberry. Type of paper: Laid paper (two- or three layers): 27–33 laid lines in 3 cm; the paper is highly sized and polished; possibly also covered with other substances, such as starch glue, which make it whiter; on the first folio, the writing area was covered with something and has remained white as compared to the margins, which are much darker (yellowish to brownish); this paper is very heavy with glue; traces of brush marks with glue or other substances are clearly visible on the paper surface. Find site: Beijing [Peking]. Dating: 1680. Format: Loose leaves pothi. Size (h × w): 26.8–27 × 68.9–73.2 cm. Language/script: Tibetan (dbu can, block-print, ‘headed letters’). Calligraphy: Very skilled calligraphy characterized by longitudinal proportions of letters (letters with ‘long legs:’ ‘ka,’ ‘ta,’ and ‘na., etc. Text fits very well into the frame. Page layout: 4–8 lines / page. Folio 1: recto—4 lines of text, first and third lines in red (distance between upper lines is 3.8‒4 cm), verso 5 lines of text, second and fourth lines in red (distance between upper lines is 3‒3.2 cm). Folio 2: recto—6 lines of text; second and fifth lines in red (distance between upper lines of text is 2.3‒2.7 cm), verso 7 lines of text with third and fifth written in red (distance between upper lines of text is 2.2‒2.3 cm). Folio 3 and rest of the volume: 8 lines of text per page: third and sixth lines are written in red (rubrics); distance between upper lines is 1.9‒2 cm). Distance between lines: 1.8‒2 cm, text-base lines are sketched with lead stylus or other slightly sharp tool at the upper part of the text (just one line at the text head); this volume is slightly less carefully written than vol. 1. A frame around the written area was traced with a 0.3 cm black line that surrounds a thin red slightly inside it (18.8 × 55 cm). Content: Tibetan Buddhist Canon; Kanjur. Decorations: Fore-edges are painted with lotus flowers and auspicious symbols on the flowers against a yellow background, on the short edges—there is a precious pearl in flames and Tibetan text in the halo on the lotus flower. Damage in the area of the green paint is visible in the form of brownish fragments and paper damage. tibetan collection of the berlin state library 267

3 Volume Peking Kanjur 17 (Preserved in 2 boxes) ‘ka’ same text as in the volume from the Pander Collection: 36 / 1097, Kanjur, 24. sher phyin (Prajñāpāramitā), ka (1), 1097–14/36 KA. Raw material: Bamboo, straw, jute, and paper mulberry. Type of paper: Laid paper (two or three layers): 27‒33 laid lines in 3 cm; paper highly sized and polished; possibly also covered with other substances, such as starch glue, which make it whiter; on the first folio, the writing area is covered with something and remained white as compared to the margins, which are much darker (yellowish to brownish); possibly, the margins are covered with something which deteriorated more quickly, since the rest of the book is much whiter in color. This paper is very heavy with glue; traces of brush with glue or other substances are clearly visible on the paper surface. It is also possible that the middle layer of paper was composed of a decidedly worse quality of paper, which caused acidity and damage in other materials. Find site: Beijing [Peking]. Dating: 1680. Format: Loose leaves pothi. Size (h × w): 26.8 × 73 cm. Language/script: Tibetan (dbu can, block-print, ‘headed letters’). Calligraphy: Skilled calligraphy characterized letters with ‘long legs’ such as ka, ta, na etc. text fits very well into frame); on the first 2–3 folios where the letters are larger, it is possible to observe that a stick (stylus) was used to form the letters, and that a brush was used to finish them. Page layout: 4–8 lines / page. Folio 1: recto—4 lines of text, first and third lines in red (distance between upper lines is 3.9–4 cm), verso 5 lines of text, second and fourth lines in red (distance between upper lines is 3–3.1 cm). Folio 2: recto—6 lines of text; second and fifth lines in red (distance between upper lines of text is 2.5–2.7 cm), verso contains 7 lines of text with the third and fifth written in red (distance between upper lines of text is 2.3–2.4 cm). Folio 3 and rest of the volume feature 8 lines of text per page: third and sixth lines written in red (rubrics); distance between upper lines is 1.9 cm). Distance between lines: 1.8–2 cm, text-base lines are sketched with a lead stylus or other slightly sharp tool at the upper part of the text (just one line at the text head). A frame around the written area was traced with a 0.3 cm black line that surrounds a thin red slightly inside it (18.3 × 54.9 cm), foliation only in Tibetan on the left side margin, outside of the frame. Content: Tibetan Buddhist Canon; Kanjur. Decorations: Fore-edges are painted with lotus flowers with auspicious symbols against a yellow background, on the short edges—there is a precious pearl in flames 268 appendix 4 and Tibetan text in the halo on a lotus flower. Damage in the area of the green paint is visible in the form of brownish fragments and paper damage.

4 Volume Peking Kanjur 18 (preserved in 2 boxes) ‘kha’ same text as in the volume from the Pander Collection: 37 / 1098 Kanjur 25. sher phyin (Prajñāpāramitā), kha (2) 1098–15/37 KHA. Raw material: Bamboo, straw, jute, and paper mulberry. Type of paper: Laid paper (two or three layers): 27‒33 laid lines in 3 cm; the paper was highly sized and polished and possibly also covered with other substances, such as starch glue, which have made it whiter. On the first folio, the surface of the writing area of the paper is white in contrast to the margins, which are much darker (yellowish to brownish). Throughout the book the margins are darker than the rest of the book. Thus the writing area maybe prepared with a substance that made it whiter or possibly the margins were covered with something that deteriorated faster. This paper is very heavy with glue; traces of brush with glue or other substances are clearly visible on the paper surface. It is also possible that the middle layer of paper was composed of a decidedly worse quality of paper which caused acidity and damage in other materials. Find site/provenance: Beijing [Peking]. Dating: 1680. Format: Loose leaves pothi. Size (h × w): 26.9 × 73.3 cm. Language/script: Tibetan (headed dbu can). Calligraphy: Skilled calligraphy characterized by the ‘long legs’ of letters such as: ‘ka,’ ‘ta,’ ‘na,’ etc. text fits very well into frame; calligraphy is quite unified within all volumes; on the first 2–3 folios where the letters are larger, it is possible to observe that a stick (stylus) was used to inscribe letters that were further finished with a brush. Page layout: 4–8 lines / page. Folio 1: recto—4 lines of text, first and third lines in red (distance between upper lines is 3.9–4 cm), verso 5 lines of text, second and fourth lines in red (distance between upper lines is 3.2–3.4 cm). Folio 1: missing. Folio 2: recto—6 lines of text; second and fifth lines in red (distance between upper lines of text is 2.7–2.8 cm), verso 7 lines of text with third and fifth written in red (distance between upper lines of text is 2.3–2.4 cm). Folio 3 and rest of the volume contains 8 lines of text: third and sixth lines written in red (rubrics); distance between upper lines is 2 cm). Distance between lines: 1.8–2 cm, guidelines are sketched with lead stylus or other slightly sharp tool at the upper part of the text (just one line at the text head). A frame around the written area was traced with a 0.3 cm black line that surrounds a thin red tibetan collection of the berlin state library 269 slightly inside it (18.5 × 55 cm), foliation only in Tibetan on the left side margin, outside of the frame. Content: Tibetan Buddhist Canon; Kanjur. Decorations: Fore-edges are painted with lotus flowers and auspicious symbols on the flowers against a yellow background, on the short edges—there is a precious pearl in flames and Tibetan text in the halo on the lotus flower. Damage in the area of green paint is visible in the form of brownish fragments and paper damage.

5 Volume Peking Kanjur 108 (Preserved in 2 Boxes) ‘ja’ Raw material: Bamboo, straw, jute, and paper mulberry. Type of paper: Laid paper (two or three layers): 27‒33 laid lines in 3 cm; paper highly sized and polished; possibly also covered with other substances, such as starch glue, which make it whiter; on the first folio, the writing area is covered with something and remained white as compared to margins which are much darker (yellowish to brownish); possibly, margins are covered with something which deteriorated faster, since the rest of the book is much whiter in color. This paper is very heavy with glue; traces of brush with glue or other substances are clearly visible on the paper surface. It is also possible that the middle layer of paper was composed of a decidedly worse quality of paper, which caused acidity and damage in other materials. Find site/provenance: Beijing [Peking]. Dating: 1680. Format: Loose leaves pothi. Size (h × w): 26.8–29 × 73.3 cm. Language/script: Tibetan (dbu can, block-print, ‘headed letters’). Calligraphy: Skilled calligraphy, quite unified within all volumes, sometimes corrected (finished) with brush; on the first 2‒3 folios where the letters are larger, it is possible to observe that the letters were written by stick (stylus), and further finished with brush. Distance between lines: 1.9–2 cm, guidelines are sketched with lead stylus or other slightly sharp tool at the upper part of the text (just one line at the text head). The letters are calligraphed slightly differently in this volume (perhaps because they are done by a different scribe, differently cut writing stick, or with an ink with a different consistency). Page layout: 4–8 lines / page. Folio 1: recto—4 lines of text, first and third lines in red (distance between upper lines is 3.9–4 cm), verso 5 lines of text, second and fourth lines in red (distance between upper lines is 3.2–3.3 cm). Folio 2: recto—6 lines of text; second and fifth lines in red (distance between upper lines of text is 2.6–2.7 cm), verso 7 lines of text with third and fifth written in red (distance between upper lines of text is 2.2–2.4 cm). 270 appendix 4

Folio 3 and rest of the volume contain 8 lines of text per page: third and sixth lines written in red (rubrics); distance between upper lines is 1.9–2 cm. A frame around the written area was traced with a 0.3 cm black line that surrounds a thin red slightly inside it (18.7 × 55.1 cm), foliation only in Tibetan on the left side margin, outside of the frame. Content: Tibetan Buddhist Canon; Kanjur. Decorations: Fore-edges are painted with lotus flowers and auspicious symbols on the flowers against a yellow background, on the short edges—there is a precious pearl in flames and Tibetan text in the halo on the lotus flower. Damage in the area of green paint is visible in the form of brownish fragments and paper damage.

6 Volume Peking Kanjur 85 (Preserved in 2 Boxes) ‘ka’ same text as in the volume from Pander Collection: 23 / 1084 Kanjur W 1. rgyud (tantra), ka (1) 1084–1/23 KA + dkar chag (index). Raw material: Bamboo, straw, jute, and paper mulberry. Type of paper: Laid paper (two or three layers): 27‒33 laid lines in 3 cm; paper highly sized and polished; possibly also covered with other substances, such as starch glue, which make it whiter; on first folio, the writing area is covered with something and remained white as compared to the margins which are much darker (yellowish to brownish); possibly, margins are covered with something which deteriorated more quickly, since the rest of the book is much whiter in color. This paper is very heavy with glue; traces of brush with glue or other substances are clearly visible on the paper surface. It is also possible that the middle layer of paper was composed of a decidedly worse quality of paper, which caused acidity and damage in other materials. Find site/provenance: Beijing [Peking]. Dating: 1680. Format: Loose leaves pothi. Size (h × w): 26.8 × 73 cm. Language/script: Tibetan (dbu can, block-print, ‘headed letters’). Calligraphy: Skilled calligraphy, unified within the entire volume, sometimes corrected (finished) with a brush, especially on the first 2‒3 folios where the letters are larger. Page layout: 4–8 lines / page. Folio 1: recto—4 lines of text, first and third lines in red (distance between upper lines is 3.9–4 cm), verso 5 lines of text, second and fourth lines in red (distance between upper lines is 3.2–3.3 cm). Folio 2: recto—6 lines of text; second and fifth lines in red (distance between upper lines of text is 2.6–2.7 cm), verso 7 lines of text with third and fifth written in red (distance between upper lines of text is 2.2–2.4 cm). tibetan collection of the berlin state library 271

Folio 3 and rest of the volume contain 8 lines of text per page: third and sixth lines written in red (rubrics); distance between upper lines is 1.9–2 cm. Distance between lines: 1.9–2 cm, text-base lines are sketched with lead stylus or other slightly sharp tool at the upper part of the text (just one line at the text head). A frame around the written area was traced with a 0.3 cm black line that surrounds a thin red slightly inside it (18.8 × 55.3 cm), foliation only in Tibetan on the left side margin, outside of the frame. Content: Tibetan Buddhist Canon; Kanjur. Decorations: Fore-edges are painted with lotus flowers and auspicious symbols on the flowers against a yellow background, on the short edges—there is a precious pearl in flames and Tibetan text in the halo on the lotus flower. Damage in the area of green paint is visible in the form of brownish fragments and paper damage. dkar chag 64 folios attached to the top of the volume 85. Folio 1: recto—blank; verso: all dkar chag has 8 lines of text per page with distances between upper lines of text measuring 2 cm, and third and sixth lines written in red (rubrics); Paper looks exactly the same type as all other volumes examined, not better quality as it is possible to observe in Wanli Kanjur edition. This dkar chag is an integral part of this volume, which can also be confirmed by fore-edges paintings in which this fragment fits as an upper part of volume 85.

1 Narthang Kanjur

Vol. 1 ’dul ba (vinaya), ka (1) printed from wooden blocks. Raw material: Bark paper: Daphne or Edgeworthia sp. Type of paper: Mostly woven paper (one layer) with both a thick woven textile as a sieve and a fine one—visible against light. However, some leaves are executed on laid paper (one layer): 15 laid lines in 3 cm (for example, folio 71). Independently of papermaking sieve print, both papers represent the same type, with visible fibers bound in the structure, absorbent, with thickness differing in different leaves. The quality of Narthang Kanjur paper is not as good as Lhasa Kanjur paper; paper surface possibly polished; gloss of paper surface. Paper is not sufficiently sized and is uneven in thickness, which is why I observed a poor quality of woodblock printing (middle part of the volume printed on a visibly worse quality of paper). Find site/provenance: Narthang. Dating: 1730–32. Format: Loose leaves pothi. Size (h × w): 17–18 × 61.5–63 cm. Language/script: Tibetan (dbu can, block-print, ‘headed letters’). 272 appendix 4

Page layout: 4–7 lines / page, full frame. Folio 1: missing. Folio 2: recto—4 lines of text, 2.1–2.2 cm distance between upper lines of text, 3 xylograph prints, verso—blank. Folio 3: recto—5 lines of text, distance between upper lines is 1.7–1.8 cm, frame of the printed area 11.5 x 58 cm; verso—6 lines of text on page. Folio 4: recto—6 lines, verso 7 lines of text per page. Folio 5 and rest of the book: 7 lines of text per page. Distance between upper lines of text about 1.5 cm. Frame of the printed area measures 11.5 × 67 cm. Content: Tibetan Buddhist Canon; Kanjur. Decorations: Fore-edges painted with yellow paint—paper very absorbent, and paint penetrated its edges. Vol. 2 ’dul ba (vinaya), kha (2) printed from wooden blocks. Raw material: Bark paper: Daphne or Edgeworthia sp. Type of paper: Both a woven type of mold (with thick textile as a woven sieve and possibly a finely woven cotton sieve), and a laid mold characterized by about 15 laid lines in 3 cm were used during production of this paper. Despite the three different sieve types used, the paper is made of the same raw material. Some of the local papermaking factories used more than one type of mold. The paper surface was possibly polished; gloss of paper is a surface characteristic for some Thymelaeaceae plants. The thickness of the paper differs in different leaves. This is a typically Tibetan type of paper with visible fiber bundles and outer bark particles in the structure. Such thin and absorbent (not sized) paper causes many ink blots and a poor quality of printing from wooden blocks. The middle part of the volume was printed on visibly worse quality paper, and when paper is printed on both sides, the text is almost impossible to read. Find site: Narthang. Dating: 1730–32. Format: Loose leaves pothi. Size (h × w): 17–18 × 63–63.5 cm (particular folios are not evenly cut; this is why it is not possible to paint on the edges). Language/script: Tibetan (dbu can, block-print, ‘headed letters’). Page layout: 4–7 lines / page, full frame. Folio 1: recto—4 lines of text, 2 cm distance between upper lines of text, double frame with printed area is: outer 11 × 58 cm, inner 9.3 × 55.5 cm; verso—blank. Folio 2 contains text from folio 1 verso (paper is too thin, and text would not be readable when printed on verso side). tibetan collection of the berlin state library 273

Folio 3 (numbered as folio 2 according to the text): recto—5 lines of text, distance between upper lines is 1.7–1.8 cm, frame of the printed area is 11.5 × 58 cm; verso—6 lines of text per page. Folio 4 (numbered as folio 3, according to the text): recto—6 lines, verso, 7 lines of text per page. Folio 5 (numbered as folio 4 according to the text) and the rest of the book: 7 lines of text per page. Distance between upper lines of text is about 1.5 cm. Frame of the printed area printed together with text is 11–11.5 × 67–68 cm. Content: Tibetan Buddhist Canon; Kanjur. Decorations: Fore-edges painted with yellow paint—paper is very absorbent, and paint penetrated the paper’s edges.

2 Lhasa Kanjur

Volume 1 ’dul ba (vinaya), ka (1) printed from wooden blocks. Raw material: Bark paper: Daphne or Edgeworthia sp. Type of paper: Both woven and laid paper (one layer): 21 laid lines in 3 cm; absorbent, very thin, typically Tibetan type with visible fibers bound in the structure, even thickness, good quality, polished (slightly glossy surface). Find site/provenance: Lhasa. Dating: 1934. Format: Loose leaves pothi. Size (h × w): 15.5 × 62.8 cm. Language/script: Tibetan (dbu can, block-print, ‘headed letters’). Page layout: 4–7 lines / page, full frame. Folio 1: recto—title in the frame, verso—blank. Folio 2: recto—4 lines of text, 2–2.5 cm distance between upper lines of text, 3 xylograph prints a) left side (?) b) Siakjamuni c) Tsonkhapa; frame of the printed area is 11.8 × 57 cm. Folio 3: recto—5 lines of text, distance between upper lines is 1.7–1.8 cm, and 2 xylograph prints, frame of the printed area is 11.8 × 57 cm; verso—6 lines of text per page. Folio 4: recto—6 lines, verso 7 lines of text per page. Folio 5 and rest of the book: 7 lines of text per page. Distance between upper lines of text is 1.3—1.6 cm. Frame of the printed area traced with a black 0.2 cm thick line measures 11.8 × 57 cm. Content: Tibetan Buddhist Canon; Kanjur. 274 appendix 4

Decorations: Fore-edges painted with yellow paint—paper very absorbent, and paint penetrated the paper’s edges. Volume 1 ’dul ba (vinaya), ka (1) printed from wooden blocks with black ink. Raw material: Bark paper: Daphne or Edgeworthia sp. Type of paper: Woven paper (one layer): absorbent and soft, typically Tibetan type with visible fibers bound in the structure, irregular thickness, polished (slightly glossy surface); good quality in general; however, quality differs within this volume, which is evident when looking at the color of the paper; some leaves are almost brown, and more particles of outer bark are visible in the paper structure; the woven mold had a sieve made of loosely woven textile, which we can see against light. Find site/provenance: Lhasa. Dating: 1934. Format: Loose leaves pothi. Size (h × w): 17 × 64 cm. Language/script: Tibetan (dbu can, block-print, ‘headed letters’). Page layout: 4–7 lines / page, full frame. Folio 1: recto—title in the frame (doubled frame, drawn with double black lines; its outer outline measures 4.9 × 22.7 cm), verso—blank. Folio 2: recto—title page repeated, as on folio 1; verso—4 lines of text, 2–2.5 cm distance between upper lines of text, 3 xylograph prints; frame of printed area—12 × 57 cm. Folio 3: recto—5 lines of text, distance between upper lines is 1.6–1.8 cm, and 2 xylograph prints, frame of printed area 11.7 × 57 cm; verso–6 lines of text per page, distance between upper lines of text measures 1.7–1.8 cm, no xylograph prints, just margin frames 11.7 × 5.7 cm. Folio 4: recto—6 lines (1.7–1.8 cm distance between upper lines of text), also margin frames 11.7 × 5.7 cm with foliation in Tibetan located on the left one, verso—7 lines of text per page. Folio 5 and the rest of the book: 7 lines of text per page. Distance between upper lines of text is 1.3–1.6 cm. Frame of printed area printed together with text + a small frame with Tibetan foliation on the left (11.5–11.8 × 57 cm). Content: Tibetan Buddhist Canon; Kanjur. Fore-edges painted with yellow. Particular leaves are not evenly cut, so edges are not aligned (also because paper is soft and rough). Vol. 2 ’dul ba (vinaya), kha (2) printed from wooden blocks with black ink. Raw material: Bark paper: Daphne or Edgeworthia sp. Type of paper: Woven paper (one layer): absorbent and soft, typically Tibetan type with visible fibers bound in the structure, irregular thickness, polished (slightly glossy tibetan collection of the berlin state library 275 surface); good quality in general; however, quality differs within this volume, which is evident when looking at the color of the paper—some leaves are almost brown and more particles of outer bark are visible in the paper structure; the woven mold had a sieve made of loosely woven textile, which we can see against light. Find site: Lhasa. Dating: 1934. Format: Loose leaves pothi. Size (h × w): 17.5 × 64 cm. Language/script: Tibetan (dbu can, block-print, ‘headed letters’). Page layout: 4–7 lines / page, full frame. Folio 1: recto—title in the frame (doubled frame drawn with double black lines, its outer outline measures 4.2 × 27.2 cm), slip of paper attached to the left margin (11.2 × 12.5 cm) with the number of the volume, kha and title of section, ’dul ba; verso—blank. Folio 2: recto—title page repeated as on folio 1; verso—4 lines of text, 1.9–2 cm distance between upper lines of text, frame of printed area—11 × 57 cm with no xylograph prints, just margin frames measuring 5.7 × 9.5 cm in inner outline, Tibetan foliation in the left margin frame. Folio 3: recto—5 lines of text, distance between upper lines is 1.6–1.7 cm, frame of printed area 11.7 × 57.2 cm; verso—6 lines of text per page, distance between upper lines of text measures 1.7–1.8 cm, no xylograph prints, just margin frames 5.7 × 11.7 cm. Folio 4: recto—6 lines (1.7–1.8 cm distance between upper lines of text), also margin frames 5.7 × 11 with foliation in Tibetan located on the left one; verso, 6 lines of text per page, the same page outline as recto side. Folio 5: recto has the same page outline as folio 4; verso has 7 lines of text per page with a distance between upper lines of 1.3–1.5 cm and text in a singular frame measuring 11.2 × 57.5 cm; The rest of the book has the same outline as folio 5 verso, with 7 lines of text per page. The distance between upper lines of text is 1.3–1.7 cm. The frame of the printed area together with the text + smaller frame with Tibetan foliation on the left of recto sides of folios is: 11.5–11.8 × 57–57.5 cm. Content: Tibetan Buddhist Canon; Kanjur. Fore-edges painted with yellow. The edges are not aligned because particular leaves are not evenly cut and also the paper is soft and rough. Appendix 5 Selected Items from the Tibetan Collection of the Library of Congress, Washington, DC: The Asian Reading Room

1 Cone Kanjur

Vol. 1 mdo (sutra) ‘ka’ Raw material: Paper mulberry and straw fibers. Type of paper: Laid paper (two or three layers): 27 laid lines in 3 cm; layers pasted together with starch; slightly yellowish (cream) in color. A paper fragment with a title (label) printed on a lotus flower in the first volume [ka] of mdo has the same chain and laid lines pattern as that of the paper leaves; here it is possible to observe chain lines in a distance of 3.5 to 4 cm from each other. Find site/Provenance: This set was purchased for the Library of Congress in 1926 at the Cone Monastery in Gansu Province, China, by botanist Joseph F. Rock. Dating: The xylograph blocks for this set were produced in the area of Cone Monastery at the beginning of the eighteenth century. Because the examined copy was printed in 1926, the paper in the Cone edition should be associated with the twentieth century—not the eighteenth. Format: Loose leaves pothi with wooden covers. Originally, volumes were tied with paper strings. Features a spare page for protection, also printed, with additional information on volume number handwritten in Tibetan letters. Covers: All volumes have wooden covers produced much later (1927) in Shanghai, measuring 17.8 × 56.5–56.7 cm with 6 longitudinal holes cut in a distance of 1.3–1.5 from the long edges and in a pattern 6.7‒18.5‒18.5‒6.7 cm from each other, plank (cover) thickness 0.8‒0.9 cm, wood cut. Tape is pulled through these holes and around the volume so loose leaves are bound together. The edges of the planks have been skillfully worked and rounded. The surfaces of the planks are protected with varnish. Size (h × w × d): 18.3 × 57 × 11 cm. Language/script: Tibetan (dbu can, block-print, ‘headed letters’). Characteristic ‘spurs’ in the right side of the upper area of letters such as ‘a’ and ‘ka.’ Triangular dashes in the upper line of letters like ‘ca’ and ‘ma.’ Page layout: 5–8 lines / page, full frame. There is foliation in Tibetan characters on the left margin of the recto sides only. It is composed in a small frame added on the left side perpendicularly to the rest of the text lines. Folio 1: recto—one-line title in a centrally located, double-lined frame measuring 23 × 2.2 cm; verso—5 verses of text and two xylograph prints (8.5 × 9.5 cm) on both sides representing the Buddhist Pantheon. Text is composed in two frames drawn with

© koninklijke brill nv, leiden, ��4 | doi ��.��63/9789004275058_�13 TIBETAN COLLECTION OF THE LIBRARY OF CONGRESS 277 double lines: the outer frame measures 11.4 × 47–47.3 cm; the inner, 9.7 × 45.4 cm. Distances between upper lines of text measure 2.3 cm. Folio 2: recto—page outline is the same as in the verso side of folio 1; verso—6 lines of text with a 1.5–1.8 cm distance between the upper lines of the text. Text is composed in a double-lined frame measuring 12 × 47 cm. Folio 3: recto—7 lines of text, with 1.3–1.5 cm distance between the upper lines; the frame of the printed area measures 47 × 10.9 with a 1.5 cm wide frame with foliation on the left side; verso—8 lines of text with a distance of 1.2–1.3 cm between upper lines. Folio 4 and the remaining leaves in the volume: recto and verso contain 8 lines of text per page, with distances between upper lines of text of about 1.2–1.3 cm; the frame of the printed area measures 11.4 × 47 cm. Content: Cone [Choné] Kanjur, composed of 108 volumes. Vol. 1 mdo (sutra) vol. Ka Fore-edges: painted organic yellow. An XRF test did not show that an arsenic compound had been applied against insects. The paper fragments pasted to particular pages indicating contents of book when looking at fore-edges were added later, probably recently, with folio numbers typed on thin tissue paper. Ink: Characteristic greyish color appears slightly diluted (not the best quality), ink is smeared beyond the edges of the letters.

Vol. 35 ’dul-ba (vinaya) ‘ga’ (3) Raw material: Paper mulberry and straw fibers (all examined folios and also in original paper strings). Type of paper: Laid paper (two or three layers): 27 laid lines in 3 cm. The paper of this set is slightly yellowish (cream). A paper fragment with a title (label) printed on a lotus flower on the first volume [ka] of mdo has the same chain and laid lines pattern as that of the paper leaves; here it is possible to observe chain lines at a distance of 3.5 to 4 cm from each other. With the exception of paper used for handwritten folios, I often observed ink blots characteristic of absorbent and soft Chinese types of paper that were not highly sized. Often the area of the letters printed is not fully and equally printed, and there are ink blots outside of the letters’ edges. Particular layers of paper were glued using diluted starch paste rather than animal glue. During examination I found one leaf (folio 244 in vol. 35) handwritten on much thicker paper. More layers of paper were pasted together to accommodate writing with a bamboo or wooden stick. This technique makes paper less absorbent and prevents ink blots. This was usually achieved both by applying more paste and polishing the surface of the leaf with stone. I found traces of polishing on this folio. For prints, such a high content of paste or use of another glue was not necessary if a balance between the thickness of ink and absorbance of paper was maintained. During printing, 278 appendix 5 pressure usually a factor for a brief time only (in the moment when the printer moves a roller down along the wooden block). When writing is executed, the pressure of the wooden stick/pen is concentrated at the point, and it takes longer for ink to make a blot if the paper is too absorbent or the ink is too diluted. For these reasons, preparation of leaves for manuscripts and prints differed; this difference is clearly visible in these two volumes, since there was likely a wooden block missing, and printers had to execute handwritten copies of these folios (244) in order to retain the full textual content in the edition. Find site/Provenance: This set was purchased for the Library of Congress in 1926 at Cone Monastery in Gansu Province, China, by botanist Joseph F. Rock. Dating: The xylograph blocks for this set were produced in the area of Cone Monastery at the beginning of the eighteenth century. Because the examined copy was printed in 1926, the paper in the Cone kanjur should be associated with the twentieth century—not the eighteenth century. Format: Loose leaves pothi with wooden covers. Originally, the volumes were tied with paper strings. Featured is a spare page for protection, also printed, with additional information on the volume number handwritten in Tibetan script. Covers: All volumes have wooden covers produced much later (1927) in Shanghai, measuring 17.8 × 56.5–56.7 cm with 6 longitudinal holes cut in a distance of 1.3–1.5 from the long edges and in a pattern 6.7‒18.5‒18.5‒6.7 cm from each other, with a plank (cover) thickness of 0.8-0.9 cm, wood cut. Tape is pulled through these holes and around the volume so loose leaves are bound together. The edges of the planks have been skillfully worked and rounded. The surfaces of the planks are protected with varnish. Size (h × w × d): 18.3 × 56.5 × 10 cm. Language/script: Tibetan (dbu can, block-print, ‘headed letters’). Characteristic ‘spurs’ in the right side of the upper area of letters such as ‘a’ and ‘ka.’ Triangular shape of dashes in the upper line of letters such as ‘ca’ and ‘ma.’ Page layout: 5–8 lines / page, full frame. Foliation is written in Tibetan script on the left margin of the recto sides only. The foliation is composed in a small frame added on the left side perpendicular to the rest of the text lines. Folio 1: recto—blank; verso—5 verses of text and two xylograph prints (8.5 × 7.8 cm) on both sides. Two frames drawn with double lines: the outer frame measures 11.5 × 47 cm; the inner, 8.5–8.8 × 44.5 cm. Distances between upper lines of text are 1.4–1.5 cm. Folio 2: recto—page outline as in verso of folio 1; verso—6 lines of text with a 1.5–1.8 cm distance between upper lines and a double-lined frame measuring 12 × 47 cm. Folio 3: recto—7 lines of text with a distance between upper lines measuring 1.3–1.5 cm; the frame of the printed area is 10.9 × 47 cm with a 1.5 cm wide frame with foliation on the left side; verso—8 lines of text with a distance between upper lines of 1.2–1.3 cm. TIBETAN COLLECTION OF THE LIBRARY OF CONGRESS 279

Folio 4 and the remaining leaves in the volume: recto and verso have 8 lines of text per page, with distances between upper lines of about 1.2–1.3 cm; the frame of the printed area measures 11.4 × 47 cm. Content: Cone [Choné] Kanjur, composed of 108 volumes; Vol. 35 ’dul-ba (vinaya) ‘ga’ (3). Fore-edges: In all examined volumes, the fore-edges were painted yellow, a characteristic color. An XRF test did not show that an arsenic compound had been applied against insects. The paper fragments pasted to particular pages indicating contents of book when looking at fore-edges were added later, probably recently, with folio numbers typed on thin tissue paper. Ink: Characteristic greyish color appears slightly diluted (not the best quality), letters are not fully and equally printed (surface unevenly covered with ink), and ink is smeared beyond the edges of the letters. Such poor quality printing suggests that the wooden blocks were worn out when the printing of this set was conducted. The ink in the handwritten folios is different in color from that used for the printing.

Vol. 72 rgyud (tantra) Raw material: Paper mulberry and straw fibers (in all examined folios and also in original paper strings). Type of paper: Laid paper (two or three layers): 27 laid lines in 3 cm. The paper of this set is slightly yellowish (cream). Find site/Provenance: This set was purchased for the Library of Congress in 1926 at the Cone Monastery in Gansu Province, China, by botanist Joseph F. Rock. Dating: The xylograph blocks for this set were produced in the area of Cone Monastery at the beginning of the eighteenth century. Because the examined copy was printed in 1926, the paper in the Cone edition should be associated with the twentieth century—not the eighteenth century. Format: Loose leaves pothi with wooden covers. Originally, volumes were tied with paper strings. Features a spare page for protection, also printed, with additional information on volume number handwritten in Tibetan letters. Covers: All volumes have wooden covers produced much later (1927) in Shanghai, measuring 17.8 × 56.5–56.7 cm with 6 longitudinal holes cut in a distance of 1.3–1.5 from the long edges and in a pattern 6.7‒18.5‒18.5‒6.7 cm from each other, with a plank (cover) thickness of 0.8‒0.9 cm, wood cut. Tape is pulled through these holes and around the volume so loose leaves are bound together. The edges of the planks have been skillfully worked and rounded. The surfaces of the planks are protected with varnish. Size (h × w × d): 18 × 57 × 9 cm. Language/script: Tibetan (dbu can, block-print, ‘headed letters’). Characteristic ‘spurs’ in the right side of the upper area of letters such as ‘a’ and ‘ka.’ Triangular shape of dashes in the upper line of letters like ‘ca’ and ‘ma.’ 280 appendix 5

Page layout: 5–8 lines / page, full frame. There is foliation in Tibetan characters on the left margin of the recto sides only. It is composed in a small frame added on the left side perpendicularly to the rest of the text lines. No title page; the first page is a folio with two xylographs and 5 lines of text, with a distance of 1.4–1.5 cm between upper lines of text on the recto; verso is blank. Other features are the same as in previous volumes. Also, a spare leaf was added on the top with some information on volume number, a black ink blot, and handwritten Tibetan letters. Content: Cone [Choné] Kanjur, composed of 108 volumes; Vol. 72 rgyud (tantra). Fore-edges: In all examined volumes, the fore-edges were painted yellow, a characteristic color. An XRF test did not show that an arsenic compound had been applied against insects. The paper fragments pasted to particular pages indicating contents of book when looking at fore-edges were added later, probably recently, with folio numbers typed on thin tissue paper. Ink: Characteristic grayish color appears slightly diluted (not the best quality), letters are not fully and equally printed (surface unevenly covered with ink).

Vol. 92 yum (Mother tantra) ‘ka’ (1) Raw material: Paper mulberry and straw fibers (in all examined folios and also in original paper strings). Type of paper: Laid paper (two- or three layers): 27 laid lines in 3 cm; slightly yellowish (cream) color of paper; ink blots characteristic of absorbent and soft Chinese types of paper. Layers of paper glued together with diluted starch paste, polished surface. The handwritten folio 289 uses much thicker paper. More layers of paper were pasted together to allow for writing with a bamboo or wooden stick. Find site/Provenance: This set was purchased for the Library of Congress in 1926 at the Cone Monastery in Gansu Province, China, by botanist Joseph F. Rock. Dating: The xylograph blocks for this set were produced in the area of Cone Monastery at the beginning of the eighteenth century. Because the examined copy was printed in 1926, the paper in the Cone edition should be associated with the twentieth century—not the eighteenth century. Format: Loose leaves pothi with wooden covers. Originally, volumes were tied with paper strings. Features a spare page for protection, also printed, with additional information on volume number handwritten in Tibetan letters. Covers: All volumes have wooden covers produced much later (1927) in Shanghai, measuring 17.8 × 56.5–56.7 cm with 6 longitudinal holes cut in a distance of 1.3–1.5 from the long edges and in a pattern 6.7‒18.5‒18.5‒6.7 cm from each other, with a plank (cover) thickness of 0.8‒0.9 cm, wood cut. Tape is pulled through these holes and around the volume so loose leaves are bound together. The edges of the planks have been skillfully worked and rounded. The surfaces of the planks are protected with varnish. TIBETAN COLLECTION OF THE LIBRARY OF CONGRESS 281

Size (h × w × d): 18.5 × 56.5 × 8 cm. Language/script: Tibetan (dbu can, block-print, ‘headed letters’). Characteristic ‘spurs’ in the right side of the upper area of letters such as ‘a’ and ‘ka.’ Triangular shape of dashes in the upper line of letters like ‘ca’ and ‘ma.’ Page layout: 5–8 lines / page, full frame. There is foliation in Tibetan characters on the left margin of the recto sides only. It is composed in a small frame added on the left side perpendicularly to the rest of the text lines. Folio 1: recto—blank; verso—5 verses of text and two xylograph prints (9.3–9.5 × 9.5–9.7 cm) on both sides. Two frames were drawn with double lines: the outer frame measures 12 × 47 cm; the inner, 9.3–9.5 × 44.5 cm. Distances between upper lines of text are 1.5–1.7 cm. Folio 2: recto—page outline as in verso of folio 1 with side frames, but without xylograph prints, containing 6 verses of text with a distance between upper lines of 1.3–1.4 cm; verso—6 lines of text with a 1.3–1.4 cm distance between upper lines, and a double-lined frame measuring 11.5 × 47 cm. Folio 3 and the remaining leaves in the volume: recto—8 lines of text, with a distance of 1.3 cm between the upper lines; the frame of the printed area measures 11.8–12 × 47 cm with a 1.5 cm wide frame with foliation on the left side. Verso—8 lines of text, a distance of 1.2–1.3 cm between the upper lines; the frame of the printed area together with text measures 11.8–12 × 47 cm. Content: Cone [Choné] Kanjur, composed of 108 volumes; Vol. 92 yum (Mother tantra) vol. ka (1). Fore-edges: In all examined volumes, the fore-edges were painted yellow. An XRF test did not show that an arsenic compound had been applied to combat insect damage. The paper fragments pasted to particular pages indicating contents of book when looking at fore-edges were added later, probably recently, with folio numbers typed on thin tissue paper. Ink: Characteristic greyish color appears slightly diluted (not the best quality), letters are not fully and equally printed (surface unevenly covered with ink), and ink is smeared beyond the edges of the letters. The ink in the handwritten folios is different in color from that used for printing. The edges of xylograph printing blocks print were visible on some pages in examined volumes. On the basis of this observation, I could estimate measurements of the Cone wooden blocks as 12–13 × 50.5–54 cm.

Vol. 108 dkar chag (Index) Raw material: Paper mulberry and straw fiber s (in all examined folios and also in original paper strings). Type of paper: Laid paper (two or three layers): 27 laid lines in 3 cm; slightly yellowish (cream) in color. 282 appendix 5

Find site/Provenance: This set was purchased for the Library of Congress in 1926 at the Cone Monastery in Gansu Province, China, by botanist Joseph F. Rock. Dating: The xylograph blocks for this set were produced in the area of Cone Monastery at the beginning of the eighteenth century. Because the examined copy was printed in 1926, the paper in the Cone edition should be associated with the twentieth century—not the eighteenth century. Format: Loose leaves pothi with wooden covers. Originally, volumes were tied with paper strings. Features a spare page for protection, also printed, with additional information identifying the volume number handwritten in Tibetan letters. Covers: All volumes have wooden covers produced much later (1927) in Shanghai, measuring 17.8 × 56.5–56.7 cm with 6 longitudinal holes cut in a distance of 1.3–1.5 from the long edges and in a pattern 6.7‒18.5‒18.5‒6.7 cm from each other, with a plank (cover) thickness of 0.8‒0.9 cm, wood cut. Tape is pulled through these holes and around the volume so loose leaves are bound together. The edges of the planks have been skillfully worked and rounded. The surfaces of the planks are protected with varnish. Size (h × w × d): 18 × 57 × 1.5 cm. Language/script: Tibetan (dbu can, block-print, ‘headed letters’). Characteristic ‘spurs’ in the right side of the upper area of letters such as ‘a’ and ‘ka.’ Triangular dashes in the upper line of letters like ‘ca’ and ‘ma.’ Page layout: 7–8 lines / page, full frame. There is foliation in Tibetan characters on the left margin of the recto sides only. It is composed in a small frame added on the left side perpendicularly to the rest of the text lines. Folio 1: recto—one-line title in a centrally located double-lined frame 2.5 × 29.3 cm; verso—7 verses of text. Text is composed within two frames drawn with double lines in a distance of 1–2 mm: the outer frame measures 47.2 × 10.3 cm; the inner, 45.4 × 8.3 cm. Distances between upper lines of text are 0.9–1 cm. Folio 2: recto—page outline as in verso of folio 1; verso—8 lines of text, with a distance of 0.9‒1.3 cm between the upper lines, and a double-lined frame measuring 10.3 × 47 cm. Folio 3 and subsequent leaves: recto—8 lines of text with a distance of 1 cm between upper lines; the frame of the printed area measures 10.3 × 47 cm with a frame 1.7 cm in width with foliation on the left side; verso—8 lines of text with a distance of 1–1.1 cm between the upper lines. Second-to-last Folio: recto—7 lines of text with a distance of 1 cm between upper lines (printing frame 9.3 × 47.3 cm); verso—6 lines of text with the upper lines in a distance of 1–1.2 cm (frame 9 × 47.3 cm). Last Folio: recto has two lines of text and the distance between the upper lines of these two lines is 1.2 cm. Content: Cone [Choné] Kanjur, composed of 108 volumes. Vol. 108 dkar chag (index). TIBETAN COLLECTION OF THE LIBRARY OF CONGRESS 283

Fore-edges: Painted with organic yellow dye. The paper fragments pasted to particular pages indicating contents of book when looking at fore-edges were added later, probably recently, with folio numbers typed on thin tissue paper. Ink: Characteristic greyish color appears slightly diluted (not the best quality), letters are not fully and equally printed (surface unevenly covered with ink), and ink is smeared beyond the edges of the letters, which suggests that absorbent paper was used with ink that was too diluted and not mixed well. Ink blots outside of the edges of the letters are clearly embossed into the paper. On the spare sheet attached to the dkar chag (index), the bottom of the ink pot left its print, and signs of the blot are visible far into the paper, which is also absorbent with a variety of shades, suggesting uneven concentration of pigment in this ink. Such poor quality printing suggests that the wooden blocks wore out when printing of this set was conducted.

2 Derge Kanjur

Volume 45 mdo-sde (sutra) ‘ka’ printed from Wooden Blocks with Red Ink Raw material: Bark paper of Daphne or Edgeworthia sp. Type of paper: Laid paper (one layer): 15 laid lines in 3 cm; thickness slightly differs in different leaves, typically Tibetan type with visible fiber bundles in the structure, brownish color, good quality, strong paper. Find site/provenance: The Derge edition. 100 v. & 1 v. of the dkar chag (index) acquired by purchase in the fiscal year 1908/9 through Hon. W.W. Rockhill, American Minister at Beijing. Order no. 121986. Dating: 1733. Format: Loose leaves pothi. Size (h × w × d): 10.5–11.5 × 60.3 × 7.5 cm. Language/script: Tibetan (dbu can, block-print, ‘headed letters’). Page layout: 5–7 lines / page, full frame. Folio 1: recto—blank; verso—5 lines of text in inner frame measuring 5.4–5.5 × 35.7 cm; there are two xylograph prints on both sides measure 5.4 × 5.4 cm, and full printing area frame measuring 7.4–7.7 × 51.6 cm; distance between upper lines of text is 0.9–1 cm. Folio 2: recto—6 lines of text; distance between upper lines of text is 0.9 cm; full frame (outer side) 6.9 × 51.5 cm is drawn with double lines; on both sides in the frame there are additional areas/inner frames 1.3 cm in width—pages are numbered on the left side, and the right inner frame is blank. verso–7 lines of text per page; distance between upper lines of text is 0.9–1 cm; text is composed in a full frame drawn with one line measuring 7.2 × 50.9 cm; ‘pagination’ on verso side is outside of frame on the left side margin. 284 appendix 5

Folio 3 and following leaves: recto—7 lines of text, distance between upper lines of 0.8–1 cm; text composed in a full frame drawn with a single line, measuring 7.2–7.4 × 52 cm; width of printing block may be 55.5 cm × (not readable). verso—7 lines of text per page, and other features as above. Derge has ‘pagination’ (numbering of pages on recto and verso sides of leaves). However, in fact there is foliation since consecutive numbers are only on recto sides together with volume number (in Tibetan letters). Content: Tibetan Buddhist Canon, Derge Kanjur Volume 45 mDo-sde (Sūtra) part vol. ka (1). Ink: Red. During printing possibly the ink dried too fast, or the surface of the paper was not properly adjusted to accommodate this type of ink—this resulted in many poorly printed letters, edges of letters often do not appear sharp, or there are blots around the letters. Fore-edges: painted with red paint, which penetrated the paper’s edge.

Derge Kanjur Volume 1 ’dul-ba (vinaya) ‘ka’ Printed from Wooden Blocks with Red Ink on Tibetan Type of Paper Raw material: Bark paper of Daphne or Edgeworthia sp. Type of paper: Laid paper (one layer): 15 laid lines in 3 cm and chain lines in a distance 3–4 cm; thickness differs in different leaves; typically Tibetan type of paper with visible fiber bundles in the structure, many fragments of outer bark in the paper, brownish color, good quality, strong paper, slightly glossy surface (possibly polished), fibers mixed to a higher degree, which makes this paper more transparent and sized. Find site/provenance: The Derge edition. 100 v. & 1 v. of the dkar chag (index) acquired by purchase in the fiscal year 1908/9 through Hon. W.W. Rockhill, American Minister at Beijing. Order no. 121986. Dating: 1733. Format: Loose leaves pothi. Size (h × w × d): 10–11.5 × 60.7–61.5 × 8.5 cm. Language/script: Tibetan (dbu can block-print, ‘headed letters’); regular and propor- tional letters with double, ‘long legs;’ even/straight upper line of text in thickness, no rounded chisels have been used as in Yongle/Wanli editions. Page layout: 5–7 lines / page, full frame. Folio 1: recto—blank; verso—5 lines of text in inner frame measuring 5.4 × 34 cm; there are two xylograph prints on both sides measuring 5.4 × 5.2 cm, and a full printing area frame sketched with a double line measuring 7.6–7.7 × 49 cm; the distance between the upper lines of text is 0.9 cm. Folio 2: recto—6 lines of text; the distance between the upper lines of text is 0.9 cm; full frame (outer side) measuring 6.5 × 49 cm is drawn with double lines; on both sides in the frame, there are additional areas/inner frames 1.3 cm in width—pages are numbered on the left side, the right inner frame is blank; verso—7 lines of text per TIBETAN COLLECTION OF THE LIBRARY OF CONGRESS 285 page; the distance between the upper lines of text is 0.9–1 cm; text is composed in a full frame drawn with a single line measuring 7.3 × 49.5 cm; ‘pagination’ on verso side is outside of frame on the left side margin. Folio 3 and subsequent leaves: recto—7 lines of text, distance between upper lines of text is 0.9–1 cm; text composed in a full frame drawn with a single line measuring 7.5 × 50 cm; width of printing block may be 54.5 cm in length, pagination outside of this frame (no side frames in subsequent recto sides of leaves); verso—7 lines of text per page, and other features as above. The Derge edition has ‘pagination’ in this volume (numbering of pages on recto and verso sides of leaves). However, on recto sides, the volume number additionally appears (in Tibetan letters). Content: Tibetan Buddhist Canon, Derge Kanjur Volume 1 ’dul-ba (vinaya) ‘ka’. Ink: Red. Sometimes the letters are difficult to read due to the thinness of the paper, which allows ink to show through both sides of the paper. Fore-edges: painted with red/pink paint—paint penetrated the paper’s edge.

Derge Kanjur Volume 77 rgyud ’bum (The Hundred Thousand Tantras of the Nyingma School) ‘ka’ Printed from Wooden Blocks with Red Ink on Tibetan Type of Paper Raw material: Bark paper of Daphne or Edgeworthia sp. Type of paper: Woven paper (one layer); the woven type of mold makes the paper in this volume different from that found in the two other volumes that were examined; the thickness of the paper differs in different leaves; typically Tibetan-type paper with visible fiber bundles in the structure; many fragments of outer bark in the paper, as well as fibers that are not well-separated; brownish color, good quality, very strong paper, soft and absorbent. Find site/provenance: The Derge edition. 100 v. & 1 v. of dkar chag (index) acquired by purchase in the fiscal year 1908/9 through Hon. W.W. Rockhill, American Minister at Beijing. Order no. 121986. Dating: 1733. Format: Loose leaves pothi. Size (h × w): 10.2–10.5 × 61.7‒62 × 16 cm. Language/script: Tibetan (dbu can block-print, ‘headed letters’). Regular and pro- portional letters with double, ‘long legs,’ even/straight upper lines of text in thickness, no rounded chisels have been used as in Yongle/Wanli. Page layout: 5–7 lines / page, full frame. Derge has ‘foliation’ in this volume (page numbering is on the recto sides of leaves only) + volume number (in Tibetan letters). However, on all verso sides the title of part of the Kanjur is printed (rgyud ‘bum (The Hundred Thousand Tantras of the Nyingma School), in this case). 286 appendix 5

Folio 1: recto—blank; verso—5 lines of text in an inner frame measuring 5.4 × 34.7 cm; there are two xylograph prints on both sides measuring 5.4 × 5.4 cm, and a full printing area frame drawn with a double line measuring 7.7 × 50.3 cm; distance between upper lines of text is 0.9 cm. Folio 2: recto—6 lines of text; distance between the upper lines of text is 0.9 cm; full frame (outer side) 6.7 × 49.9 cm is drawn with double lines; on both sides of the frame, there are additional areas/inner frames 1.3 cm in width—pages are numbered on the left side, and the right inner frame is blank; printing block may be 9.5 × 54 cm, verso—7 lines of text per page; distance between upper lines of text is 0.9–1 cm; text is composed in a full frame drawn with a single line measuring 7.3 × 49.5 cm; ‘pagination’ on verso side is outside of frame on the left side margin. Folio 3 and subsequent leaves: recto—7 lines of text, distance between upper lines of text is 0.9–1 cm; text composed in a full frame drawn with a single line measuring 7.2 × 50 cm; foliation outside of this frame (no side frames in subsequent recto sides of leaves); verso—7 lines of text per page, and other features as above. Content: Tibetan Buddhist Canon, Derge Kanjur, Volume 77 rgyud-’bum (The Hundred Thousand Tantras of the Nyingma School) ‘ka’. Ink: Red. This is a better quality of print, softer and thicker paper makes for a better match with the ink type. Fore-edges of volume: painted with red/pink paint—paint penetrated the paper’s edge.

3 Narthang Kanjur

Vol. 1 ’dul-ba (vinaya), ‘ka’ (1) Raw material: Bark paper of Daphne or Edgeworthia sp. Type of paper: Both a woven type of mold (with thick textile as a woven sieve and possibly a finely woven cotton sieve) and a laid mold, characterized by about 15 laid lines in 3 cm, were used during production of this paper. The paper surface was possibly polished; gloss of paper is a surface characteristic for some Thymelaeaceae plants. The thickness of the paper differs in different leaves. This is a typically Tibetan type of paper with visible fiber bundles and outer bark particles in the structure. Find site/provenance: Narthang. Dating: 1730‒32. The printing of Narthang Kanjur began in the autumn of 1730 and was completed in the first month of the Tibetan calendar in 1732. Its print took alto- gether one year and six months. Format: Loose leaves pothi; particular folios are not evenly cut. Size (h × w): 17–18 × 61.5–63 cm. TIBETAN COLLECTION OF THE LIBRARY OF CONGRESS 287

Language/script: Tibetan (dbu can block-print, ‘headed letters’). Page layout: 4–7 lines / page, full frame. Frame of printed area: 11.5 × 67 cm. Folio 1: missing. Folio 2: recto—4 lines of text, 2.1–2.2 cm distance between upper lines of text, 3 xylograph prints, and verso side is blank. Folio 3: recto—5 lines of text, distance between upper lines is 1.7–1.8 cm, frame of printed area measures 11.5 × 58 cm; verso—6 lines of text per page. Folio 4: recto—6 lines, verso 7 lines of text per page. Folio 5 and rest of the book: 7 lines of text per page, the distance between upper lines of text is about 1.5 cm. Content: Tibetan Buddhist Canon, Kanjur. Fore-edges painting: yellow. Ink: Printed from wooden blocks (poor quality of printing was caused by use of very thin and unsized paper—the middle part of the volume was printed on a visibly worse quality of paper). Bibliography

Turfan Studies. 2007. Berlin-Brandenburg Academy of Sciences and Humanities. Berlin.

Agrawal, Om Prakash. 1984. Conservation of Manuscripts and Paintings of South- East Asia. (Butterworth-Heinemann series in Conservation and Museology). London /Boston /Durban /Singapore /Sydney /Toronto /Wellington: Butterworth- Heinemann in association with The International Institute for Conservation of Historic and Artistic Works. Agrawal, Yashodhara. 2003. Silk Brocades. New Delhi: Roli & Janssen BV. Bareja-Starzyńska, Agata. 1999. Złoto w Tybecie. Złoto ze skarbca Muzeum Narodowego w Warszawie. Warszawa: Muzeum Narodowe w Warszawie. Beer, Robert. 2003. The Handbook of Tibetan Buddhist Symbols. Boston: Serindia Publications. Bell, Lilian A. 1981. Plant Fibers for Papermaking. McMinnville: Liliacaea Press. ———. 1983. Papyrus, tapa, amate and rice paper: papermaking in Africa, the Pacific, Latin America and Southeast Asia. McMinnville: Liliacaea Press. Białousowa, J.A., Bartosik, M. Kurhański, A. Nagórski, and B. Tumalewicz. 1958. Konopie, Rośliny włókniste [Fibrous Plants]. Warszawa. Bloom, Jonathan M. 2001. Paper before Print. The History and Impact of Paper in the Islamic World. New Haven and London: Yale University Press. Brock, Fiona, Thomas Higham, Peter Ditchfield, Christopher Bronk Ramsey. 2010. ‘Current pretreatment methods for AMS radiocarbon dating at the Oxford Radiocarbon Accelerator Unit (ORAU).’ Radiocarbon 52.1, 103–112. Browning, Bertie Lee. 1977. Analysis of Paper. New York/ Basel: Marcel Dekker. Carter, Thomas Francis. 1955. The Invention of Printing in China and its Spread Westward. New York: Ronald Press. Carter, John, and Nicolas Barker. 2004. ABC for Book Collectors. London: Oak Knoll Press/the British Library. Catling, Dorothy and John Grayson. 1982. Identification of Vegetable Fibres. London: Archetype Books. Chayet, Anne, Cristina Scherrer-Schaub, Françoise Robin, and Jean-Luc Achard, eds. 2010. Edition, éditions: l’écrit au Tibet, évolution et devenir. (Collectanea Himalyica, 3) Munich: Indus Verlag. Chiodo, E., and K. Sagaster. 1995. “The Mongolian and Tibetan Manuscripts on Birch bark from Xarbuxyn Balgas: A preliminary description.’ In Zentralasiatische Studien des Seminars für Sprach- und Kulturwissenschaft Zentralasiens der Universität Bonn 25 edited by W. Heissig, G. Verhufen, and M. Weiers, pp. 28–41. Wiesbaden: Harrassowitz Verlag. Bibliography 289

Collings, Thomas, and Derek Milner. 1978. “The identification of oriental papermaking fibers.” The Paper Conservator 3: 51–79. Cüppers, Christoph. 1989. “On the Manufacture of Ink.” Ancient Nepal 113: 8–9. Dagyab, Loden Sherap. 1977. Tibetan Religious Art. Wiesbaden: O. Harrassowitz. ———. 1995. Buddhist symbols in Tibetan culture: an investigation of the nine best-known groups of symbols. Translated from the German by Maurice Walshe. Somerville, MA: Wisdom Publications. Dalton, Jacob, Tom Davis, and Sam van Schaik. 2007. “Beyond Anonymity: Paleographic Analyses of the Dunhuang Manuscripts.” Journal of the International Association of Tibetan Studies, 3, 1–23. http://www.thlib.org?tid=T3106 (accessed February 2014). Dąbrowski, Józef, Jadwiga Siniarska-Czaplicka. 1991. Rękodzieło papiernicze. Warszawa: Wydawnictwo Czasopism i Książek Technicznych Sigma NOT Sp. Z o.o. Dawa, Doctor. 1999. A Clear Mirror of Tibetan Medicinal Plants. Vol. 1. Rome: Tibet Domani Associazione. Diamentowa Sutra [Diamond Sutra]. 2004. Translated into Polish by Marek Mejor from a nineteenth-century gold manuscript preserved in the Pomeranian Library in Szczecin. Bibliotheca Buddhica Polonica. Szczecin: Pomeranian Library. Dotson, Brandon. 2009. The Old Tibetan Annals: An Annotated Translation of Tibet’s First History, With an Annotated Cartographical Documentation by Guntram Hazod. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Drège, Jean-Pierre. 1979. “Les cahiers des manuscrits de Touen-houang.” In Contributions aux études de Touen-houang. Vol. 1. Publications de l’École Française d’Extrême- Orient, vol. 10, edited by Michel Soymié, 17–28. Genève-Paris: Librarie Droz. ———. 1981. “Papiers de Dunhuang, essai d’analyse morphologique des manuscrits chinois datés” T’oung Pao 67: 305–360. ———. 1984. “Les accordéons de Dunhuang.” In Contributions aux études de Touen- houang. Vol. 3, Publications de l’École Française d’Extrême-Orient, vol. 135, 195–211. Paris: École française d’Extrême-Orient. ———. 2002. “Dunhuang papers: Preliminary Morphological Analysis of Dated Chinese Manuscripts.” In Dunhuang Manuscript Forgeries edited by Susan Whitfield, 115–179. London: The British Library. Ehrhard, Franz-Karl. 2000. Early Buddhist Block Prints from Mang-yul Gung-thang. Vol. 1 (Lumbini International Research Institute Monograph Series 2) Lumbini: Lumbini International Research Institute. Eimer, Helmut. 2000. “Spurensicherung: Das verschollene Berliner Fragment des Wanli-Kanjur.” Zentralasiatische Studien 30: 27–51. ———. 2007. “The Tibetan Kanjur Printed in China.” Zentralasiatische Studien 36: 35–60. Ernst, Richard R. 2001. “Arts and Sciences. A Personal Perspective of Tibetan Painting.” Chimia 55 (11): 900–914. 290 Bibliography

Field, Dorothy. 1995. “Traditional Values in Indian Handmade Paper.” Hand Papermaking 10 (2): 24–28. Gaur, Albertine. 1979. Writing Materials of the East. London: The British Library. Gega, Lama. 1983. Principles of Tibetan Art: Illustrations and Explanations of Buddhist Iconography and Iconometry according to the Karma Gardri school. Vol. 1–2. Darjeeling: Jamyang Singe. Gibbs, Peter J., and Kenneth R. Seddon. 1998. Berberine and Huangbo: Ancient Colorants and Dyes. The British Library Studies in Conservation Science v. 2. London: The British Library. Grönbold, Günther. 1982. “Tibet: Literature und Buchkunst.” In Das Buch im Orient: Handschrift en und Kostbare Drucke aus Zwei Jahrtausenden, edited by J. Assfalg et al., 253–264 Wiesbaden: Ludwig Reichert Verlag. ———. 1991. Tibetische Buchdeckel. München: Bayerische Staatsbibliothek. Hall, Bruce Cameron. 1979. A descriptive list of the Tibetan (Tib.) Collection in the Rare Books of the Harvard-Yenching Library. Cambridge. Unpublished typescript revised in 1980 as a PhD thesis at the Sanskrit Department of Harvard University. Hansen, Valerie. 2005. “The Impact of the Silk Road Trade on a Local Community: The Turfan Oasis, 500–800.” In Les Sogdiens en Chine, edited by Étienne de la Vaissière and Eric Trombert, 283–310. Paris: École Française d’Extrême-Orient. Harrison, Paul. 1996. “A Brief History of the Tibetan Kanjur.” In Tibetan Literature. Studies in Genre, edited by José Ignacio Cabezón and Roger R. Jackson, 70–94. Ithaca, New York: Snow Lion. Heller, Amy. 1999. Tibetan Art. Tracing the Development of Spiritual Ideals and Art in Tibet, 600–2000 AD. Milan: Jaca Books. ———. 2009. Hidden Treasures of the Himalayas: Tibetan Manuscripts, Paintings and Sculptures of Dolpo. Chicago: Serindia Publications. Helman-Ważny, Agnieszka. 2001. “Traditional Papermaking in Mountains in Central Asia.” Conservator-Restorers’ Bulletin 12 (2): 50–55. ———. 2004. “Tibetan Historic Manuscripts as a Source of Information on Past Papermaking in Inner Asia.” In IPH Congress Book 15/2004, edited by Bożena Schweizer-Makowska and Maciej Szymczyk, 27–37. Duszniki Zdrój: International Association of Paper Historians and Duszniki Paper Museum. ———. 2006. “Asian Paper in Works of Art: a Comparative Fiber Analysis.” Handmade Paper 21 (2): 3–9. ———. 2007. “A Chronology of Tibetan Bookbinding.” International Dunhuang Project News Issue 29: 4–5. ———. 2007. “Tibetan manuscripts: scientific examination and conservation approaches.” In Edinburgh Conference Papers 2006: Proceedings from the Fifth International Conference of the Institute of Paper Conservation and First International Bibliography 291

Conference of the Institute of Conservation, Book and Paper Group, edited by S. Jaques, 247–256. London: Institute of Conservation (ICON). ———. 2008. Examination of the Tibetan illuminated book cover and frontispiece of Eight Thousand Lines Perfection of Wisdom Sūtra (Acc. Number: 2006.028 a,b). Report for H. Johnson Museum of Art at Cornell University. ———. 2009. “Recovering a Lost Literary Heritage: Preliminary Research on the Wanli Kanjur from Berlin.” Journal of the International Association of Tibetan Studies 5: 1–27. ———. 2009. “Local papermaking, paper trade and book-culture promotion in Inner Asia.” In IPH Congress Book 17/2008, edited by J.-E. Levlin, 173–186. Stockholm: International Association of Paper Historians. ———. 2009. Sztuka tybetańskich ksiąg klasztornych [Art of Tibetan Traditional Books], Warsaw: TRIO Publications. ———. 2010. “Historical, Art-Historical, and Natural Science Approaches to Dating Tibetan Books—Possibilities and Limitations.” In Tibetan Studies: An Anthology. PIATS 2006: Tibetan Studies; proceedings of the eleventh seminar of the International Association for Tibetan Studies, Königswinter 2006, edited by S. Arslan and P. Schwieger, 279–295. Halle: International Institute for Tibetan and Buddhist Studies. Helman-Ważny, Agnieszka and Sam Van Schaik. 2013. “Witnesses for Tibetan Craftsmanship: Bringing Together Paper Analysis, Palaeography and Codicology in the Examination of the Earliest Tibetan Manuscripts.” Archaeometry 55 (4): 707–741 .doi: 10.1111/j.1475-4754.2012.00687.x Holmberg David, Kathryn March, Suryaman Tamang and Tamang Elders. 1999. “Local Production/Local Knowledge: Forced Labour from Below.” Studies in Nepali History and Society 4.1, 5–64. Hopkirk, Peter. 1980. Foreign Devils of the Silk Road: The Search for the Lost Cities and Treasures of Chinese Central Asia. London: John Murray. Hunter, Dard. 1932. Old Papermaking in China and Japan. Ohio: Mountain House Press. ———. 1943. Papermaking. The History and Technique of an Ancient Craft. New York: Alfred A. Knopf. ———. 1978. Papermaking. The History and Technique of an Ancient Craft. New York: Dover Publications. Ilvessalo-Pfäffli, Marja-Sisko. 1995. Fiber Atlas, Identification of Papermaking Fibers. Berlin Heidelberg: Springer-Verlag. Imaeda, Yoshiro. 1977. “Mise au point concernant les éditions chinoises du Kanjur et du Tanjur tibétains.” A. Macdonald, Y. Imaeda. Eds. Essais sur l’art du Tibet. Paris: 23–51. ———. 1989. ‘Papermaking in Bhutan.’ Acta Orientalia Academiae Scientiarum Hung XLIII.2–3, 409–414. ———. 2008. “The Provenance and Character of the Dunhuang Documents.” Memoirs of the Toyo Bunko 66, 81–102. 292 Bibliography

Jackson, David. 1983. “The Earliest Printing of Tsong-kha-pa’s Works: The Old Dga’— ldan Editions.” In Reflections on Tibetan Culture. Essays in Memory of Turrel V. Wylie. Edited by L. Epstein and R. Sherburne. Lewiston: E. Mellen Press, 1990. ———. 1989. “More on the Old Dga’ ldan and Gong dkar ba Xylographic Editions.” Studies in Central and East Asian Religions 2: 1–18. ———. 1996. A History of Tibetan Painting. Wien: Verlag der Österreichischen Akademie der Wissenschaften. Jackson, David, and Janice Jackson. 1994. Tibetan Thangka Painting. Methods and Materials. London: Serindia Publications. Jaglarz, Monika. 2008. “Zbiór rękopisów i druków orientalnych z byłej Pruskiej Biblioteki Państwowej w Berlinie przechowywanych w Bibliotece Jagiellońskiej.” [The collection of oriental manuscripts and prints from the former Prussian State Library in Berlin, preserved at present at the Jagiellonian Library]. In Torun Studies on Oriental Art 3 edited by Jerzy Malinowski and Mirosława Wojtczak, 49–52. Toruń: Wydawnictwo Naukowe Uniwersytetu Mikołaja Kopernika. Jest, Corneille. 1961. “A Technical Note on the Tibetan Method of Block-Carving.” Man 61, 83–85. Jiacuo et al. 1985. “Lasa Xiancang di liang bu Yongle ban ‘Ganzhuer’.” [The two Yongle Versions of Kanjur extant in Lhasa.’] Wenwu 9, 85–89. Kara, György. 2005. Books of the Mongolian Nomads. More than Eight Centuries of Writing Mongolian. Bloomington: Indiana University Uralic and Altaic Series 171. Karabacek, Joseph von. 2001. Arab Paper. London: Archetype Publications. Karmay, Samten Gyaltsen. 1998. Secret Visions of the Fifth Dalai Lama. London: Serindia Publications. Kawasaki, Kazuhiro. 2004. “On a Birch-bark Sanskrit Manuscript Preserved in the Tibet Museum.” Journal of Indian and Buddhist Studies 52 (2): 50–52. Kingdon-Ward, Frank. 1934. A Plant Hunter in Tibet. London: Jonathan Cape. Koretsky, Elaine. 1986. “Papermaking Today in Tibet and China.” Hand Papermaking 1: 2–6. van der Kuijp, Leonard W.J. 1993. “Two Mongol Xylographs (Hor Par Ma) of the Tibetan Text of Sa Skya Paṇḍita’s Work on Buddhist Logic and Epistemology.” The Journal of the International Association of Buddhist Studies 16 (2): 279–298. ———. 2006. “On the Composition and Printings of the Deb gter sngon po by ’Gos lo tsā ba gzhon nu dpal (1392–1481).” Journal of the International Association of Tibetan Studies 2: 1–46. Lalou, Marcelle. 1939–1961. Inventaire des Manuscrits tibétains de Touen-houang con- servés à la Bibliothèque Nationale (Fonds Pelliot tibétain) [3 vols]. Paris: Bibliothèque Nationale de France. Laufer, Berthold. 1931. Paper and Printing in Ancient China. Chicago. Bibliography 293

Lenhart, J.M. 1950. “Capuchins Introduce Printing into Tibet in 1741.” Franciscan Studies 10: 62–72. Linrothe, Rob, and Jeff Watt. 2004. Demonic Divine. Himalayan Art and Beyond. New York/Chicago: Rubin Museum of Art and Serindia Publications. Lohia, Sushama. 1994. Lalitavajra’s Manual of Buddhist Iconography. New Delhi: International Academy of Indian Culture and Aditya Prakashan. Blo Bsang ’Phrin Las, Dung dkar [Dungkar Losang Trinlé]. 1989. “Printing of Ancient Tibetan Texts and Discussion of a few Related Matters.” The Tibetan Academy of Social Sciences 4.32. ———. 1990. “Printing of Ancient Tibetan Text and Discussion of Some Related Matters Regarding Wages and Rewards for Various Professions Connected with Printing.” Tibetan Studies 1. Losty, Jeremiah. 1982. The Art of the Book in India. London: The British Library. Macdonell, Arthur A. 1968. A History of Sanskrit Literature. New York: Haskell House Publishers. Mcclure, Floyd Alonzo. 1986. Chinese Handmade Paper. Newtown, PA: Bird and Bull Press. Meinheit, Susan. 2009. “Notes on the Co ne Bka’ ’gyur and Bstan ’gyur in the Library of Congress, Washington, D.C.” Journal of the International Association of Tibetan Studies, 5 (December): 1–14. http://www.thlib.org?tid=T5697 (accessed March 3, 2013). Meisezahl, Richard Othon. 1958. “Bemerkungen zu Tibetischen Handschriften des 17.–19. Jahrhunderts im Institut für Cellulosechemie der Technischen Hochschule Darmstad, Forschungsstelle. Papier geschichte. Herausgegeben vom Verein der zell- stoff- und Papier-Chemiker und –Ingenieure Jahrgang 8 (2): 17–28. ———. 1960. “Über den Derge Tanjur der ehemaligen Preussischen Staatsbibliothek, ergänzt durch eine mikroskopische Papieruntersuchung.” Libri 10 (4): 292–306. ———. 1961. “Alt-tibetische Handschriften im Reiss-Museum, Mannheim, ergänzt durch eine mikroskopische Papieruntersuchung von M. Harders-Steinhäuser und G. Jayme.” Libri 11(1): 1–48. ———. 1981. “Die tibetischen Handschriften und Blockdrucke im Antwerpener Ethnographie Museum, ergänzt durch eine mikroskopische Papieruntersuchung.” Oriens 27, 490–521. Mejor, Marek, Agnieszka Helman-Ważny, and Thupten Kunga Chashab, eds. 2010. A Preliminary Report on the Wanli Kanjur Kept at the Jagiellonian Library, Kraków. Warsaw: Research Centre of Buddhist Studies, Faculty of Oriental Studies, University of Warsaw. Mook, W.G., and H.T. Waterbolk. 1985. Handbook for archaeologists. No. 3 Radiocarbon dating. Strasbourg: European Science Foundation. 294 Bibliography

Müller, Claudius C., and Walter Raunig eds. 1982. Der Weg zum Dach der Welt. Innsbruck/ Frankfurt-Main: Pinguin/Umschau. Neate, S., and D. Howell. 2011. “Conservation Issues and Research Questions: The Role of Analysis in Book and Manuscript Conservation.” In The Technological Study of Books and Manuscripts as Artefacts edited by Neate S. et al., 9–14 Oxford: Archaeopress. Nebesky-Wojkowitz, Rene. 1949. Schriftwesen, Papierherstellung und Buchdruck bei den Tibetern. Vienna: Wiener philos. Dissertation [Manuscript]. Olson, Eleanor. 1971. “Writing and Printing Equipment.” In Catalogue of the Tibetan Collection and other Lamaist Material in the Newark Museum, vol. 3 edited by E. Olson et al., 104–127. Newark: The Newark Museum. Pagel, Ulrich, and Séan Gaffney. 1996. Location List to the Texts in the Microfiche Edition of the Śel dkar (London) Manuscript bKa’ ’gyur (Or. 6724). Catalogus Codicum Tibetanorum I. London: The British Library. Pakhoutova, Elena, and Agnieszka Helman-Ważny. 2012. “Tools of Persuasion: The Art of Sacred Books.” Orientations Magazine 43 (2): 124–129. Pal, Pratapaditya, and Julia Meech-Pekarik. 1988. Buddhist Book Illuminations. Hong Kong/New York/Paris/New Delhi: Ravi Kumar Publisher. ———, ed. 2003. Himalayas: An Aesthetic Adventure. Chicago: Art Institute of Chicago/ University of California Press/Mapin Publications. Pal, Pratapaditya. 2001. Desire and Devotion. Art from India, Nepal, and Tibet in the John and Berthe Ford Collection. Baltimore: Philip Wilson Pub Ltd. Pan, Jixing. 1981. “On the Origin of Papermaking in the Light of Newest Archeological Discoveries.” International Association of Paper Historians (IPH) Information 15 (2): 38–48. ———. 2000. A History of Movable Metal-Type Printing Technique in China. Liaoning: Science and Technology Publishing House. Pander, Eugen. 1889. “Geschichte des Lamaismus.” In Verhandlungen der Berliner Gesellschaft für Anthropologie, Ethnologie und Urgeschichte, Jahrgang 1889. Sitzung vom 16. Februar 1889: 199–210 [Zeitschrift für Ethnologie 21 (1889).] ———. 1890. Das Pantheon des Tschangtscha Hutuktu. Ein Beitrag zur Iconographie des Lamaismus. Herausgegeben und mit Inhaltsverzeichnissen versehen von Albert Grünwedel. Berlin: Kgl. Museum für Völkerkunde. Pietrzyk, Zdzisław. 2005. “Book Collections from the Former Preussische Staats- bibliothek in the Jagiellonian Library.” Polish Libraries Today: Foreign Collections in Polish Libraries. A Historical Overview 6, 81–87. ———. 2008. “Zbiory z byłej Pruskiej Biblioteki Państwowej w Bibliotece Jagiellońskiej.” Alma Mater 100: 15–19. Premchand, Neeta. 1995. Of the Deckle Edge. A Paper-making Journey through India. Bombay: The Ankur Project. Bibliography 295

Reedy, Chandra L. 1991. “The opening of consecrated Tibetan bronzes with interior contents: scholarly, conservation, and ethical considerations.” Journal of the American Institute for Conservation 30 (1): 13–34. Reynolds, Valrae, Amy Heller, Janet Gyatso. 1986. Catalogue of the Tibetan Collection of the Newark Museum 3. Newark: The Newark Museum. Reynolds, Valrae. 1991. “Discoveries about Tibetan Manuscripts in the Newark Museum.” Hand Papermaking 6 (2): 20–23. Rischel, Anna-Grethe. 1985. “Traditional Paper Making in the Far East.” Annual Newsletter of the Scandinavian Institute of Asian Studies 19: 7–18. ———. 2004. “Analysis of Papermaker’s Choice of Fibrous Materials and Technology Along the Paper Road.” In Paper as a Medium of Cultural Heritage. Archaeology and Conservation. 26th Congress—International Association of Paper Historians edited by R. Graziaplena, 202–208. Rome: International Association of Paper Historians. Rong Xinjiang. 2000. “The Nature of the Dunhuang Library Cave and the Reasons for its Sealing.” Cahiers d’Extrême-Asie 11: 247–275. Sandermann, W. 1968. Alte Techniken der Papierherstellung in Südostasien und den Himalaya Ländern.’ Papiergeschichte 18: 29–39. Sandermann, W., and H. Funke. 1970. Chemische Untersuchung eines schwarzen tibetischen Papiers.ʼ Papiergeschichte 20: 41–13. Sangs Rgyas Rgya Mtsho, Sde-srid [Desi Sangye Gyatso]. 1990. Mchod sdong ‚dzam gling rgyan gcig gi dkar chag, Edited by Blo rdor [Some Remarks on Bka’ ’gyur Productioning 17th-century Tibet]. Lhasa: Tibet National publishing house (Bod ljongs mi dmngs dpe skrun khang). Савицкий, Л.С. [Savitzkiy, L.S.] 1991. Описание тибетских свитков из Дуньхуана в собрании Института востоковедения АН СССР. Москва: Наука, ГРВЛ. van Schaik, Sam and Imre Galambos. 2012. Manuscripts and Travellers: The Sino-Tibetan Documents of a Tenth-Century Buddhist Pilgrim. Berlin: de Gruyter. van Schaik, Sam. 2012. “The Origin of the Headless Style (dbu med) in Tibet.” In Medieval Tibeto-Burman Languages IV edited by Nathan Hill, 411–446. Leiden: Brill. Scherrer-Schaub, Cristina A. 1999. “Towards a Methodology for the Study of Old Tibetan Manuscripts: Dunhuang and Tabo.” In Tabo Studies 2: Manuscripts, Texts, Inscriptions and the Arts edited by Cristina Scherrer-Schaub and Ernst Steinkellner, 3–36. Rome: Istituto Italiano per l’Africa e l’Oriente. Scherrer-Schaub, Cristina and George Bonani. 2002. “Establishing a typology of the old Tibetan manuscripts: a multidisciplinary approach.” In Dunhuang Manuscript Forgeries edited by Susan Whitfield, Susan, 184–215. London: The British Library. Schmidt, B., T. Ważny, K. Malla, E. Höfs, and M. Khalessi 1999. “Chronologies for Historical Dating in High Asia, Nepal.” In Tree-Ring Analysis. Biological, Metho dological and Environmental Aspects edited by R. Wimmer and R.E. Vetter,205–211. Wallingford, New York: CABI Publishing. 296 Bibliography

Schaeffer, Kurtis R. 1999. “Printing the Words of the Master. Tibetan Editorial Practice in the Collected Works of ’Jam dbyangs bzhad pa’i rdo rje I (1648–1721).” Acta Orientalia 60: 159–177. ———. 2009. The Culture of the Book in Tibet. New York: Columbia University Press. Shafer, Robert. 1960. “Words for ‘Printing Block’ and the Origin of Printing.” Journal of the American Oriental Society 80 (4): 328–329. Sheppard, P.R., P.E. Tarasov, L.J. Graumlich, K.U. Heussner, M. Wagner, H. Österle, and L.G. Thompson. 2004. “Annual precipitation since 515 BC reconstructed from living and fossil juniper growth of northeastern Qinghai Province, China.” Climate Dynamics 23: 869–881. Shen, Weirong. 2010. “Reconstructing the History of Buddhism in Central Eurasia (11th–14th Centuries): An Indesciplinary and Multilingual Approach to the Khara Khoto texts.” In Edition, éditions: l’écrit au Tibet, évolution et devenir edited by Anne Chayet et al. 337–362. Munich: Indus Verlag. Shi, Jinbo. 2005. “A study of the Earliest Tibetan Woodcut Copies.” In China Tibetology Magazine (Chinese Edition) 4. Translated by Zhang Xiaomei available at the China Tibetology Information Center Website: zt.tibet.cn/english/zt/ TibetologyMagazine/200312007420165155.htm (accessed 13.02.2013). Shrestha, Bharat P. 1989. Forest Plants of Nepal. Lalitpur: Educational Enterprise. Silk, Jonathan A. 1996. “Notes on the History of the Yongle Kanjur.” In Suhrllekhah: Festgabe für Helmut Eimer (Indica et Tibetica 28). Monographien zu den Sprachen und Literaturen des indo-tibetischen Kulturraumes, edited by Michael Hahn, Jensj- Uwe Hartmann, and Roland Steiner, 153–200. Swisttal-Odendorf: Indica et Tibetica Verlag. Silva, Kapila D., and Neel Kamal Chapagain, eds. 2013. Asian Heritage Management: Contexts, Concerns, and Prospects. London: Routledge. Slavin, John, ed. 2001. Looking at paper: evidence and interpretation. Symposium proceedings, Toronto 1999. Ottawa: Canadian Conservation Institute. Smith, William, ed. 1890–91. A Dictionary of Greek and Roman antiquities. London: John Murray. Snellgrove David, and Hugh Richardson. A Cultural History of Tibet. London 1968. Soteriou, Alexandra. 1999. Gift of conquerors. Hand papermaking in India. New Delhi: Mappin Publ. van Staalduinen, Mark, Jan C.A. van der Lubbe, Eric Backer, Pavel Paclik. 2006. “Paper Retrieval Based on Specific Paper Features: Chain and Laid Lines.” In Multimedia Content Representation, Classification and Security edited by Bilge Gunsel et al., 346–353. Berlin/Heidelberg: Springer. Stein, Rolf A. 2010. Rolf Stein’s Tibetica Antiqua. Translated and edited by Arthur McKeown. Leiden: Brill. Bibliography 297

Steinkellner, Ernst. 2004. “A Tale of Leaves. On Sanskrit Manuscripts in Tibet, their Past and their Future.” 2003 Gonda Lecture. Amsterdam: Royal Netherlands Academy of Arts and Sciences. [“http://www.oeaw.ac.at/ias/Mat/steinkellner_leaves.pdf”] (20.12.2013). Stoddard, Heather. 1998. “The Style and Artistic Context.” In Secret Visions of the Fifth Dalai Lama, ed. S.G. Karmay, 10–13. London: Serindia Publications. ———. 2010. “Stitched Books from the Tibetan World.” In Edition, éditions: l’écrit au Tibet, évolution et devenir edited by Anne Chayet et al., 363–379. Munich: Indus Verlag. Storrs, A., J. Storrs. 1998. Trees and Shrubs of Nepal and the Himalayas. New Delhi: Book Faith India. Takeuchi, Tsuguhito. 2012. “Old Tibetan Buddhist Texts from the Post-Tibetan Imperial Period (mid-9 c. to late 10 c.).” In Old Tibetan Studies: Proceedings of the 10th Seminar of the IATS 2003, edited by Christina Scherrer-Schaub, 205–216. Leiden: Brill. Trier, Jasper. 1972. Ancient Paper of Nepal. Result of ethno-technological field work on its manufacture, uses and history—with technical analyses of bast, paper and manuscripts. Copenhagen: Jutland Archaeological Society Publications. Tschudin, Walter Friedrich. 1958. Älteste Papierbereitungsverfahren im Fernen Osten.’ Textil-Rundschau 13: 679–689. Tschudin, P.F. 2004. “Non-Destructive Optical Investigation of Paper.” In Paper as a Medium of Cultural Heritage. Archaeology and Conservation. 26th Congress of the International Association of Paper Historians, edited by Rosella Graziaplena, 135–141. Roma: International Association of Paper Historians. Tsien, Tsuen-Hsuin. 1973. “Raw Materials for Old Papermaking in China.” Journal of the American Oriental Society 93(4): 510–519. ———. 2004. Written on Bamboo and Silk. The beginnings of Chinese Books and Inscriptions. Chicago/London: The University of Chicago Press. Tsien, Tsuen-Hsuin and Joseph Needham. 1985. Science and Civilisation in China, v. 5: Chemistry and Chemical Technology; Paper and Printing (Pt. 1). Cambridge: Cambridge University Press. Tsundru, Jampa. 2010. Preserve and Development of the Manufacturing Technology of Traditional Tibetan Paper. Lhasa: The Ethnic Publishing House. Tucci, Giuseppe. 1949. Tibetan painted scrolls. Roma: Libreria dello Stato. Turner, Samuel. 1800. An Account of an Embassy to the Court of the Teshoo Lama in Tibet. London: G. and W. Nicol. Uebach, Helga. 2008. “From Red Tally to Yellow Paper: The official introduction of paper in the Tibetan administration in 744/745.” Revue d’Études Tibétaines 14: 57–69. Участкина, Зoя Вacильевна [Uchastkina, Zoya Vacilyevna]. 1972. Развитие бумажного производства в России. Москва. 298 Bibliography

Upreti, Manohar. 2004. “Lokta, King of Nepalese Paper.” In Spirit of paper, edited by A. Westerhof, 236–243. Leiden: Uitgeverij Compres. Uray, Géza. 1988. “New Contributions to Tibetan Documents from the post-Tibetan Tun-huang.” In Tibetan Studies: Proceedings of the Fourth Seminar of the International Association for Tibetan Studies, edited by Helga Uebach and Jampa L. Panglung, 515–528. Munich: Kommission für Zentralasiatische Studien, Bayerische Akademie der Wissenschaften Uspensky, Vladimir. 1997. “The Tibetan Equivalents to the Titles of the Texts in the St. Petersburg Manuscript of the Mongolian Kanjur: A Reconstructed Catalogue.” In Proceedings of the 7th Seminar of the International Association of Tibetan Studies. Graz 1995, Vol. 3, edited by Ernst Steinkellner, 113–176. Wien: Verlag der Osterreichis chen Akademie der Wissenschaften. De La Vallée Poussin, Louis. 1962. Catalogue of the Tibetan Manuscripts from Tun-Huang in the India Office Library. Oxford: Oxford University Press. Воробьева-Десятовская, Маргарита Иосифовна. [Vorobieva-Desiatovskaia, Mar guarita Iosifovna] 1988. Рукописная книга в культуре народов Востока, v. 1&2, Москва: ГРВЛ Наука. Wangchuk, Dorji. 2010. “Legendary and Historical Accounts of Manuscripts Written with Precious Materials.” Paper read at the 12th Seminar of the International Association for Tibetan Studies Sunday August 15th–Saturday August 21st, 2010 hosted by Institute of Asian Research and University of British Columbia in Vancouver, Canada. Watt, James C.Y., and Barbara Brennan Ford. 1991. East Asian Lacquer. The Florence and Herbert Irving Collection. New York: The Metropolitan Museum of Art. Watt, James C.Y., and Denise Patry Leidy. 2005. Defining Yongle: Imperial Art in Early Fifteenth-Century China. New York: Metropolitan Museum of Ar. Weber, Therese. 2007. The Language of Paper. A History of 2000 Years. Bangkok: Orchid Press. Winter, John. 1985. “Paints and supports in Far Eastern Pictorial Art.” The Paper Conservator 9.1: 24–31. ———. 2008. East Asian paintings: materials, structures and deterioration mechanisms. London: Archetype Publications. Wood, Frances, and Mark Barnard. 2010. The Diamond Sutra. The Story of the World’s Earliest Dated Printed Book. London: The British Library. Wujastyk, Dominik. Forthcoming. “Indian manuscripts.” In Manuscript Cultures: Mapping the Field. Studies in Manuscript Cultures 1, edited by Jörg Quenzer and Jan- Ulrich Sobisch. Berlin: De Gruyter. Zorin, Alexander. 2013. “A Collection of Tantric Ritual Texts from an Ancient Tibetan Scroll Kept at the Institute of Oriental Manuscripts of the Russian Academy of Sciences.” Journal of the International College for Postgraduate Buddhist Studies 17: 98–146 (2013-03-31). Index*

Accelerator Mass Spectrometry (AMS) Asian and African Studies collection) methods 20–21 160f92, 161ff94-95, 162 detection of Stellera chamaejasme fibers bamboo sieves in an tenth-century manuscript 111 with different weaves used Aquilaria agallocha (aga ru, eaglewood) for sections of the Derge Kanjur 170 190–191, 194 and dipping methods of papermaking Arabian book culture from China 35–36, 59, 191–192 and Chinese book culture 50, 180 and 192f 112 and gold on blue manuscripts 80 laid regular structure of paper from Islamic bookbinding 62 movable sieves 36, 110, 159 Islamic-founded cities as centers of and Tibetan/Tangut manuscript Dx 178 papermaking 51 from the Institute of Oriental and Kashmiri book culture 64 and n.27 Manuscripts 108 pens used by 51, 80 See also papermaking tools See also book cultures; Persian book Beck manuscript editions. See under culture Wanli volumes, Berlin manuscript Beck ‘archaeology of the book,’ defined 2–3 copy of the Peking Kanjur arsenic sulphide Berlinka collection. See Pander Collection fore-edges painted yellow tested Berlin Kanjur (manuscript Beck). See under negatively for 163, 285, 287, 288, 289 Wanli volumes, Berlin manuscript Beck as an insecticide 25, 43, 93 copy of the Peking Kanjur Asia and Pacific Museum (Warsaw) Berlin manuscript copy of the Wanli volumes. Andrzej Wawrzyniak (director) 1–2 See under Wanli volumes, Berlin Kanjur illuminated manuscript in (MAP 4323) (manuscript Beck) 91, 92f48, 100, 211, 224 Berlin State Library paper in layers (MAP 14136) 208f 118 Lhasa Kanjur from 175–176, 177f 108 ’phags pa shes rab kyi pha rol tu phyin pa Narthang Kanjur set from 162, 173, rdo rje gcod pa (Diamond Sutra) 1, 91, 175–176 and f 107, 264 92f48, 95–96, 100, 224 See also Wanli volumes restoration of paper at 222–223 and Bhutanese book culture ff 127–128 bound book written in gold (MS 13162) Tibetan book (MAP 12385) 215f 121 69, 70f29a,b,c, 81–82 Tibetan manuscript (MAP 14144) laid type of mold used in 172 25f3b, 45f 14, 223f 128 Samuel Turner’s mid-eighteenth century account of 182 bamboo fibers and Thymelaeaceae-family plants 33 added for softness and absorbency 176 and n.29 characteristics of 32 See also book cultures genera used for papermaking 32 binding. See bookbinding in papermaking in China after the Tang birch bark 32 bound book on birch bark in the in the paper of the Mongolian Kanjur Tibetan Museum in Lhasa 63–64, 67 (St. Petersburg University, Faculty of brittle fibers of 211

* Page locators in italics indicate figures. 300 index birch bark (cont.) culture; Nepalese book culture; Persian and Kashmiri book culture 48–49, 63, book culture; Tibetan book culture; 84n.64 Uighur book culture in Tibetan ink recipes 100 Broussonetia sp. (genus of the Moracae blood family) ‘nose-blood’ (shangs mtshal) in ink about 28–29 and ff6a,b, 109ff59–60 recipes 101 See also paper mulberry vermillion blood (mtshal khrag) in ink (Broussonetia sp.) recipes 101 Buddhism Bon tradition, Tibetan gold manuscript concertina format books used by 60 ‘Life of Gshen rab mi bo’ 91, 93f49, 94, and the invention of printing 116, 117–118 224 and f 129 and the production of books written in bookbinding Tibetan 4, 48, 193 bound book on birch bark in the Tibetan Tibetan books as the embodiment of the Museum in Lhasa 63–64, 67 Buddha’s words 47 and conservation techniques 53, 55 Tibetan Buddhism as a living tradition and hemp 31 201–202 and f 117 Islamic bookbinding 62 Butön (1290–1364) 123 linen cord stitching of bound manuscripts (CC 101, CC 114–115, BL 14727 1–2) 69, Cai Lun’s biography in the Hou Han Shu 29, 71, 72f30a,b, 73f31a,b,c,d, 74f32a,b,c,d 183, 184 of palm-leaf pothi books 56, 57f20 calligraphy regional nature of 18 dbu can (‘headed’) script 95, 97f52, 97, representative books from the British 98f53 Library 7 and n.3, 69, 71 dbu med (‘headless’) script 66f27a, 95, sectional binding of Bhutanese book 98 written in gold (MS 13162) 69, gold script 100 70f29a,b,c, 81–82 shape and proportions of letters 153 stitched bindings of deb ther 69, ‘spurs’ on letters 153, 155f90, 165–166 and 70f29a,b,c, 81–82 f98, 167f99 Tibetan manuscript written in dbu med stylus (bamboo pen) used by Tibetans script (OR 15193) 66f27a,b,c,d, 69, 71 44 Tibetan stitched book on paper from See also gold-color ink; ink; paleography; Dunhuang (IOL J 401) 63, 64f23a,b raised gold lettering; writing tools Tibetan stitched book on paper from carving tools Dunhuang (IOL J 530) 63, 65f26a,b about 127, 129 and nn.24–25 See also codex form (thread booklet) chisel marks in letters of booklet from concertina format Khara Khoto 123 book cultures chisel marks in letters of the Yongle/Wanli and local bookmaking techniques 3–4 Kanjurs 136, 155f89, 166, 292 material aspects of books as important straight type chisel used for the Cone historical sources 4–5 woodblocks 166 See also Arabian book culture; Bhutanese and woodblock identification 16f 1, 127, book culture; Chinese book culture; 129, 133, 135–136, 153, 155f89, 166, 227 European book culture; ‘fingerprints’ Chinese book culture of time and origin in materials; Indian and the Arab world 50, 180 book culture; Japanese book culture; and block printing 117–119, 121–122 Korean book culture; Mongolian book butterfly binding 67–69 index 301

Chinese pagination 119–121, 122, 154 books from the Pander collection 60–61 and fore-edge painting 93, 94 and n.15 and f23 and gold on blue manuscripts 80 for government documents 61, 62f24 and hemp cultivation 30–31 and Tibetan books produced in China and indigo paper dyed by dipping 59, 74–75 104, 226 Uighur examples from the ninth and movable type 116n.2 century 119 and Nia documents from the third and Cone Kanjur (Library of Congress) fourth centuries 180 dkar chag 163 and n.61, 165, 167, 168f 100, scroll format 48, 59, 74 261–262 See also book cultures; Chinese Joseph Rock’s acquisition of 164f96, 172 papermaking lettering characteristics 165–166 and Chinese papermaking f98 as a writing material 44, 176 paper quality 167, 169 and ff 101–102, and the Cone paper 167, 172–173 172–173 dating of oldest paper fragments 31 Prajñāpāramitā volumes (yum) 164–165 distinguished from Tibetan and f97, 167f99, 288–289 papermaking 176–178 style and format of pages of 164–169 fibers cut during processing of 44 See also Tibetan book culture; Tibetan ink blots characteristic of absorbent and Kanjurs produced in Tibet soft types of paper 167 conservation and the migration of Chinese binding elements 53, 55 papermaking communities 179–180 environmental controls in libraries 213 and paper mulberry 26–27, 29–30, 36 and handling 206, 212, 214–215 plant materials used in 22–23, 31–33 preservation, prevention conservation, starch paste for gluing layers of paper conservation science, and restoration 208 distinguished 205–206 and Thymelaeaceae family plants 184 preventive conservation 205 and n.2, See also Chinese book culture; 220 papermaking processes Tibetan book storage methods 215–216, cinnabar 217–219ff 123–125 as a coral substitute 100 typical damage to Tibetan books 206 as a natural insecticide 150 Western conservation attitudes 203–204 as a symbol of power 100–101 See also handling; preservation in lacquer 142n.49 conservation science 205–206, 205n.2 and Tibetan Kanjurs 154 coral (byu ru) codex form (thread booklet) cinnabar substituted for 100 elements of described 62 and gold manuscripts 80, 100 and European book culture 3, 53, 62 mounted on front pages as elements of a Tibetan types of 52, 60, 62, 63 circular shape imitating holes in palm codicology leaf books 86, 87f39, 231 as a tool 3, 12, 13 See also organic materials and the history of Tibetan printing 135 cotton Columbia University, Tibetan gold features of 32 manuscript ‘Life of Gshen rab mi bo’ 91, in Ming Dynasty textiles 31 93f49, 94, 224 and f 129 See also organic materials concertina format covers. See pothi format covers; Yongle Kanjur as a synthesis of Chinese and Indian/ covers Southeast Asian book forms 60 Cupa, Adam 15 302 index

Daphne or Edgeworthia sp. (shog shing or Derge Printing House (Sde dge par khang) dung loma in Tibetan) establishment of 169 collection and preparation of 194 papermaking in 35n.33, 171–172, 182 and ‘cotton paper’ 184 woodcarving methods at 127, 130 difficulty in distinguishing two fibers Diamond Sutra of 23–24 illuminated manuscript in the Asia and dug shog (poisonous paper) as a Tibetan Pacific Museum (Warsaw) of 1, 91, term for Daphne 25, 189 92f48, 95–96, 100, 224 in Dunhuang manuscripts 104–105 lapis lazuli paper mentioned in colophon Edgeworthia gardineri sp. plant 185f 109 of 103 fibers described 24–25, 25f3a,b,c,d in the Pomeranian Library in Szczecin, fibers in the paper of the Derge Kanjur Poland 100 (Library of Congress) 173f 106 dug shog (poisonous paper) fibers in the paper of the Lhasa Kanjur as a Tibetan term for Daphne 25, 189 (Berlin State Library) 175–176, 177f 108 See also Daphne or Edgeworthia sp. fibers in the paper of the Narthang Dunhuang documents (IOL Tib J from the Kanjur 176f 107 Stein Collection, the British Library) fibers in the Sel dkar Kanjur 106f57 231–249 high starch content of the Daphne plant hemp as a component in manuscripts 39, 47f 13b, 189, 207 from 30 known as lokta and kāgate in Nepal laid line characteristics 172 185f 109, 188 paper types of manuscripts found in and Tibetan papermaking 24, 177 104–105, 110 See also Stellera chamaejasme; ramie as a component in manuscripts Thymelaeaceae family plant fibers from 28f5a,b deb ther (Tibetan bound books) root paper of Stellera chamaejasme not and other formats of Tibetan book 52, found in 110 206 See also Dunhuang oasis; Tibetan book and Persia daftar (bound manuscripts) in culture Mongolia 63 Dunhuang oasis stitched bindings of 69, 70f29a,b,c, 81–82 concertina format manuscripts from 60 See also Tibetan book culture and early manuscripts written on de La Vallée Poussin, Louis 52 paper 51 dendrochronology of pothi format covers oldest preserved Tibetan books found 19, 20f2, 227 in 52 Derge Kanjur (Tibetan Rare Book Collection, paper mulberry fibers in Tibetan/ The Library of Congress) Tangut manuscripts from 106–108, no rounded chisels used for (in contrast 109ff59–60, 246 to the Yongle/Wanli) 293 scroll format in 52, 59, 74, 208 pagination of 170, 291, 292, 293, 294 wide variety of formats of books in 52 papermaking molds of various types See also Dunhuang documents used in 170, 171f 104, 172 paper of rgyud volume 77 distinguished eaglewood (Aquilaria agallocha, aga ru) 170 190–191, 194 paper quality 170–173 and f 104 Edgeworthia gardineri sp. See Daphne or Thymelaeaceae family plant fibers of Edgeworthia sp. 170, 171–172, 173f 106 Thymelaeaceae family plant fibers See also Tibetan book culture; Tibetan edition, re-edition, and reprinting Kanjurs produced in Tibet introduced 132–136 index 303

technical similarities revealed by and the term ‘ma’ in old Chinese comparative examination of Kanjur documents 31 sets 178 See also organic materials Eimer, Helmut fore-edge painting Berlin Wanli Kanjur deemed lost by 10, as a characteristic of Tibetan Peking- 139n.41 Kanjurs 94 and n.15 and the ‘technical identity’ of hundred- as insect repellents (arsenic or mercury volume sets 132, 135 based pigments) 93 Euphorbia fisheriana 43, 184, 186, 195, 209 ordering of pages in pothi format made European book culture visible by 94 and mass printing 77, 211 in red on the Derge Kanjur set in the and movable type 116–117 Library of Congress 170 move from paprus rolls to parchment of Wanli volumes 93–94, 147–149, codices 3 148ff80–81, 149f82, 150f83, 15184a,b and printing of images 116n.1 yellow tested negatively for arsenic Ramie fibers brought to 27 sulphide 163, 285, 287, 288, 289 Roman tablet writing 73 and n.34 See also hand-painted decorations sizing of paper for writing 39 woodblock printing in 116 and n.1 glues See also book cultures; European discoloration caused by 208, 210f 119, 226 medieval manuscripts; Russian glued layers of paper 36, 43–44, paper 45–46, 156, 157–158, 162, 207, 208 European medieval manuscripts for sizing 39–40, 45f 14 and curtain coverings 85 smoothing paper surfaces 43, 44 gold ink on colored background 80 starch paste 43–44, 115, 132, 167, 206 initial letters 93 unripe resin in 100 patrons as actors in miniature scenes See also papermaking processes; Tibetan on 78 papermaking rubrics 13 and n.1, 97 gold-color ink Tibetan calligraphy in ancient books alternating with silver text on the Lotus compared with 44 Sutra from Columbia University and the valuation of printed books 76 Library 97 and f52, 98 and f53 See also book cultures; European book on a manuscript of the Mongolian culture; illuminated manuscripts; Kanjur 159 miniature painting on black or blue ground in Tibetan book culture 100 ‘fingerprints’ of time and origin in materials bound book written in gold (MS 13162) laid- and chain-lines of papermaking 69, 70f29a,b,c, 81–82 sieves 36, 158, 177–178, 178, 191 and concertina format books 60 overview of 17–18 deterioration of 211, 223–224 and f 129 raw material/fiber composition as 23, Tibetan terms for scriptures written in 56–57, 104–105, 145f77a,b 177 81 shape of letters/handwriting style 105, writing tools for 80 133, 153 See also gold-color ink; gold manuscripts; and tool traces 16f 1, 127, 129, 133, 135–136, ink; raised gold lettering; Tibetan book 153, 155f89, 166, 228 culture; Yongle Kanjur covers See also book cultures gold manuscripts flax Diamond Sutra (in the Asia and Pacific and other organic materials in Russian Museum, Warsaw) 1, 91, 92f48, 95–96, types of paper 23, 24, 35 100, 224 304 index gold manuscripts (cont.) Holmberg, David 193 ‘Life of Gshen rab mi bo’ (Columbia Hunter, Dard 191–192, 196n.36 University manuscript) 91, 93f49, 94, 224 and f 129 illuminated manuscripts: and the concertina and the stitched book format 81–82 format 60 Tibetan Book of the Dead 82, double covers used with 58, 82–85, 83–85ff34–38 83–85ff34–38 See also gold-color ink; illuminated and woven paper produced with floating manuscripts; raised gold lettering; molds 33 Sel dkar Kanjur xylograph prints on left and right sides of pages 86, 164–165 and 165f97 hair See also Diamond Sutra; European hair ornaments identifying female medieval manuscripts; fore-edge figures 78 painting; gold manuscripts; in paper pulp 30 hand-painted decorations; miniature singed hair (dbu skra’i gzhobs) in ink painting; Tibetan book culture; recipes 101 Tibetan Book of the Dead handling 206, 212, 214–215 Indian book culture and finger print spots 214 lithography printing in the 1960s 117, water damage 215f 121 118f66 See also conservation; preservation pustaka (palm-leaf books) 49-53 and hand-painted decorations 50f 16, 56, 98–99 circles imitating holes after palm-leaf raw material for paper making 188 books 56, 86, 98–99 spread of papermaking to 51 initial-letter painting 86, 92–93 writing supports before the twelfth miniature painting 92–93, 141ff73a,b century in 50 sketches and doodles 86, 94, 95f51 See also book cultures; Kashmiri book See also fore-edge painting; illuminated culture manuscripts; miniature painting indigo Heller, Amy 18 blue paper imitating indigo dyed paper buff-colored sheet of the Tibetan 77, 78f33, 104 Prajñāpāramitā provided by 105, indigo-dyed sheet of paper provided by 105–106ff55–56 Amy Heller 114–115 indigo-dyed sheet of paper provided by paper dyed by dipping 59, 104, 226 114–115 ink hemp (Cannabis sativa) about 99–101 common name ‘hemp’ used for other black and red ink of the Wanli Kanjur plants 30 94, 154–156, 155f90 cultivation in China 30–31 blood in ink recipes 100, 101 in early paper samples from Nepal 110 cinnabar ink of Tibetan Kanjurs 154 and other organic materials in Russian cinnabar/vermillion (mercury sulfide) types of paper 23 as a symbol of power 100–101 paper mulberry fibers compared with 28 milk of a white goat (ra dkar gyi ’o ma) preferred uses of 31–32 in ink recipes 101 and the term ‘ma’ in old Chinese and paper absorbency 168–169, 170–171 documents 31 preparation for blue/black paper in Tibetan manuscripts from 42–43 Dunhuang 30 recipes for 100–101 index 305

singed hair (dbu skra’i gzhobs) in ink Kalachakra Tantra printed for funeral recipes 101 ceremonies of Kublai Khan 123, traditional mineral pigments 15–16, 19 124ff68–69 See also gold-color ink; rubrics; writing Kangxi Kanjur (1684–92): paper structure tools and composition of 156–158, 259 Institute of Oriental Manuscripts size leaves of 147 (St. Petersburg, Russia) ‘spurs’ on particular letters 153 small prayer book from Khara Khoto and volume 58 of the Berlin Wanli 67, 68f28, 69, 121–122, 121n.16, 250 Kanjur (Pander Pantheon volumes) See also Mongolian Kanjur from Kangxi 8, 140 times See also Institute of Oriental Manuscripts Islamic bookbinding 62 Kaochang 51, 180 See also Arabian book culture; Kashmiri book culture bookbinding and Arabian book culture 64 and n.27 birch bark favored by 48–49, 63, 84n.64 Jackson, David 12, 18, 123 bound book on birch bark 63–64, 67 Jagiellonian University Library fine quality of 188 (Biblioteka Jagiellońska) in Cracow, See also book cultures; Indian book Poland culture fragments of the Berlin Wanli Kanjur Kawasaki, Kazuhiro 64–65, 67n.28 preserved in 8–10, 139–140 and Khara Khoto 139nn.41–43 IOL Tib M (58 vol. 1:58) 250 Wanli Supplement volumes 140, 156 IOL Tib M (60 vol. 1:60) 250 and n.58 small prayer book 67, 68f28, 69, 121–122, See also Wanli volumes, Berlin Wanli 121n.16, 250 Kanjur (Pander Pantheon volumes) and Thymelaeaceae family plant fibers Japanese book culture in documents from 104–105 absorbency of paper 44 Korean book culture: absorbency of paper papermaking introduced by the Korean 44 monk Don-cho 179 earliest Buddhist text traced to 117 and and paper mulberry 19 nn.4–5 and ramie 26 and metal movable type 116n.2 Thymelaeaceae family plant fiber in and paper mulberry 19 gampi and mitsumata paper 44, and ramie 26 188 See also book cultures See also book cultures Koretsky, Elaine 26, 35 Jest, Corneille 129 Kotan 51, 180 Jia Sixie (473–554) 108 Kusha 51, 180 jute (Corchorus sp.) combined with bamboo, straw, and paper Lalou, Marcelle 52 mulberry in Peking sets of Tibetan lapis lazuli Kanjurs 162 paper made of mentioned in the Dunhuang and other Central Asian sites Diamond Sutra 103 110 in pigments 16 in Nepalese and Indian papermaking Laufer, Berthold 61–62, 184 188 Lhasa Kanjur (in the Berlin State Library) and the term ‘ma’ in old Chinese paper quality of 175–176, 177f 108 documents 31 text lines of 177 306 index

Library of Congress (handwritten) tantra ‘ka’ section Appendix 5 list of studied items from (from St. Petersburg University, 276–287 Faculty of Asian and African Studies) Brgyad ston ’pa folio in 10 160ff92–93, 162, 258 Narthang Kanjur set from 173–174 tantra volume (E59) from Kangxi times Rare Book and Manuscript Collection 7 (Institute of Oriental Manuscripts) See also Cone Kanjur; Derge Kanjur; 159, 160f92–93, 161ff94–95, 162, 259 Narthang Kanjur See also Institute of Oriental Manuscripts Library of Tibetan Works and Archives movable type 116–117 and nn.1–2 conservation survey of 10–11 myrobalan (a ru ra, Terminalia chebula) 42, manuscript LTWA 14459 10n.18, 98 43, 102, 103, 195, 195n34 woven paper from document from 34f9a,b Narthang Kanjur (1730–1732) Linrothe, Rob 70 Berlin State Library volumes of 173, Lotus Sutra 97 and f52, 98 and f53 176f 107 Loulan (Xinjiang province) 31, 51, 180 paper of 174–175, 176–177, 176f 107, Lozang Trinlé, Dungkar (Dung dkar blo 279–280, 294–295 bzang ’phrin las) 42, 125 production of 132, 294 woven mold used for paper of 175 McClure, Floyd Alonzo 184 See also Narthang Monastery; Tibetan March, Kathryn 193 book culture; Tibetan Kanjurs medieval manuscripts. See European produced in Tibet medieval manuscripts Narthang Monastery Meinheit, Susan 10 carving of woodblock for the Narthang minerals Kanjur 128ff70–71 as pigments 15–16, 19 Narthang Tibetan Buddhist Canon sal ammoniac in papermaking 195 production 132 and n.33 See also Narthang Kanjur miniature painting National Museum in Warsaw, Tibetan bound silk coverings and frames for 85 manuscripts in collection of 94, 95f51 style of 17–18, 19, 86 Nepalese book culture on the Tibetan Selkar (sel dkar) Kanjur book storage methods 216, 88and ff42a,b 217–219ff 123–125 on the Wanli Supplement volume’s cover early palm-leaf manuscripts from 49 140, 141ff73a,b Edgeworthia gardineri grown 184, 185f 109 See also European medieval manuscripts; gold on blue paper used for manuscripts Tibetan book culture 80 Mongolian book culture painting of inner sides of manuscript and Tibetan Buddhism 4 covers 56 See also book cultures; Mongolian pothi format books 54f 17, 56 Kanjurs; Tibetan book culture Prajñāpāramitā manuscript from the Mongolian Kanjurs Bicher temple (Dolpo) 105–107 paper used for 159 and ff55–56 and the Peking Kanjur 158–159, 173 See also book cultures; Nepalese pothi format of printed editions 159 papermaking (Berlin State Library), paper quality Nepalese papermaking of 162 boiling process 195 (E2, volume 9) printed in 1714 early technology of 51 (Institute of Oriental Manuscripts) Edgeworthia gardineri (kāgate) and 162, 260 Daphne (lokta) used in 185f 109, 188 index 307

and insect-deterring substances 43 papermaking processes long-distance trading of paper 181 color 41–42 ramie and hemp in early sample from cotton in machine made paper 32 110 detoxification of 195, 209 See also papermaking processes distinguished for handwritten manuscripts Newark Museum, covers of Yongle volumes and xylograph prints 176 in 142 floating molds and woven paper Nia 180 production 33–34 and ff9a,b, 35, 191–192, 196–199 and ff 113–116 OR 15190. See Tibetan Book of the Dead sizing 39–40, 45f 14, 131 (British Library manuscript OR 15190) spread of 51 organic materials straw added for softness and absorbency acacia catechu (ja do) 195 176 and the Accelerator Mass Spectrometry See also Chinese papermaking; glue; (AMS) 20–21 Nepalese papermaking; papermaking barley flour (thel chin) 195 tools; Russian paper black vitriol (nag mtsur) 42, 103 papermaking tools as pigments 16 conch shell for smoothening 41 and radiocarbon dating 19–21 movable sieve/dipping method 35–36, resin of pine or fir 100 107–108, 110, 191–192 and 192f 112 thar nu (medicinal plant) 42, 103 shok bre (floating mold) 196–199 and See also bamboo fibers; birch bark; coral ff 113–116 (byu ru); cotton; flax; gold-color ink; for smoothening paper 41 hemp; indigo; jute; myrobalan; ramie; yellow dzi (gzhi) used for smoothening straw; Thymelaeaceae family plant 41 fibers; wood pulp See also bamboo sieves; papermaking processes pagination paper mulberry (Broussonetia sp.) Chinese pagination as an aid to and Chinese papermaking 26–27, 29–30, 36 indicator of Chinese craftsmen 94 in the Cone Kanjur 169 and ff 101–102 and n.15, 119, 122, 154 cultivation of 28–29 Chinese pagination on woodblocks of fibers of 29 and ff6a,b, 109ff59–60 non-Chinese texts 119–121, 122 in the paper of the Mongolian Kanjur defined 154n.56 160f92 and the identification of woodblocks in Tibetan/Tangut manuscripts 106–108, 133, 170 109ff59–60, 246 paleography Peking editions introduced 13 bamboo, straw, jute, and paper mulberry Tibetan scripts 17 in paper of 162 See also calligraphy Chinese paper used for 177 Pander Collection (Berlinka collection) fore-edge painting on early editions of the deemed lost by Eimer 10, 139n.41 Peking Kanjur 94 in the Jagiellonian University Library Mongolian Peking editions 158–159, 173 8–10 and 9nn.14, 16, 139–140 and woodblocks required for 129–130 139nn.41–43 See also Wanli volumes; Yongle Kanjur volume 58 of the Pander Pantheon Persian book culture 139n.43, 140 daftar (bound manuscripts) in Mongolia See also Wanli volumes, Berlin Wanli 63 Kanjur (Pander Pantheon volumes) and hemp cultivation 30 Pander, Eugene, about 8–9, 134 See also Arabia 308 index

’phags pa shes rab kyi pha rol tu phyin pa rdo Diamond Sutra (in the Asia and Pacific rje gcod pa (Diamond Sutra in the Asia Museum, Warsaw) 1, 91, 92f48, 95–96, and Pacific Museum, Warsaw) 1, 91, 100, 224 92f48, 95–96, 100, 224 from the early ninth century composed pigments of Paper mulberry 107 based on arsenic or mercury as an element of a ritual ensuring good (insect poison) 93 harvest 48f 16 traditional mineral pigments 15–16, 19 volume from the Bicher temple See also fore-edge painting (Dolpo, Nepal) 105–107 and ff55–56 pothi format preservation of Cone Kanjur volumes 164 and f96 and conservation science 205–206, dpe ras (cloth wrapping) 53, 54f 18, 94 205n.2 elements of 53, 54f 17 preventive conservation 205 and n.2, 220 Indian pustaka (palm-leaf book) origins of Tibetan books 203–204 of 49-53 and 50f 16, 56, 98–99 See also conservation; handling Kalachakra Tantra printed for funeral printing. See Tibetan block-printing ceremonies of Kublai Khan 123, technology; woodblock printing 124ff68–69 ordering of pages made visible with radiocarbon dating fore-edge painting 94 as a method 19–21 sizes of pages altered by Tibetan scribes of the Tibetan Book of the Dead from the 55–56 British Library 59, 226–227 textile labels of 53, 94 raised gold lettering: of front pages of the vulnerability to damage of 206–207 London Selkar (Sel dkar) Kanjur 86, See also Cone Kanjur; Derge Kanjur; 87f40, 231 fore-edge painting; Lhasa Kanjur; of front pages of the Wanli Supplement Mongolian Kanjurs; Narthang Kanjur; 142f74, 140 Peking editions; pothi format covers; of front page titles 85, 92–93, See also Tibetan book culture; Tibetan Book of gold-color ink the Dead (British Library manuscript Ramie (Boehmeria sp.) OR 15190); Wanli volumes; Yongle and Chinese types of paper 22, 26, 31, 32 Kanjur features of 28f5a,b pothi format covers in older paper samples from Nepal 110 blue paper imitating indigo dyed paper and the term ‘ma’ in old Chinese 77, 78f33 documents 31 decoration of 56, 57f 19 restoration and dendrochronology 19, 20f2, 227 and conservation science 205–206, double covers used with illuminated 205n.2 manuscripts in 58, 82–85, loss filling 222–223 and ff 127–128 83–85ff34–38 sacred nature of books defaced by sewing of 130 restoration 201 with string-holes 56, 57f20 of Tibetan objects and architecture wood used for identified with particular 203–204 regions 56, 58, 144 Reynolds, Valrae 198 See also Yongle Kanjur covers Rock, Joseph, Cone Kanjur acquired by 172 Prajñāpāramitā manuscripts Roman book culture Cone Kanjur (Library of Congress) tablet writing 73 and n.34 volumes 164–165 and 165f97, 288–289 See also European book culture index 309 rubrics (red lettering) Stoddard, Heather 52–53, 64, 67n.31, 75, 79 on European medieval manuscripts 13 straw and n.1, 71, 97, 97–98 added for softness and absorbency 176 on Tibetan manuscripts, on the Peking fibers in the Cone Kanjur 169 and Kanjur (Berlin State Library) 273, 274, ff 101–102 275, 276, 278, 279 See also ink tablets Russian paper planks (samta/bka’ sam) for daily notes deterioration of 211 71, 73 organic materials in 23, 24, 35 Roman tablet writing 73 and n.34 Sumkin factory (Afamovo) paper 36–38 Takeuchi, Tsuguhito 180 and f 11, 39f 12 terminology See also papermaking processes and challenges faced by paper historians 12, 21–22 St. Petersburg. See Institute of Oriental edition, re-edition, and reprinting Manuscripts 132–136 Schaeffer, Kurtis R. 47, 122 three-level set of criteria for the typology Sel dkar Kanjur of paper in Tibetan manuscripts and the Brgyad ston ’pa folio in the Library 22–23 of Congress 10 Tibetan and Chinese words for ‘printing decoration of 86, 87ff39–40, 88 and block’ 121 ff42a,b Tibetan names for books related to their paper composition 106–107ff57–58, 107 format 52–53 Shafer, Robert 121 Tibetan terms for scriptures written in Silk, Jonathan 21, 134 gold-color ink 81 sketches and doodles textiles linear sketch of a dragon 86, 94, 95f51 as a writing support in India 50 painted-edge ornamental sketches and cotton fibers 31, 32 147–150, 148ff80–81, 149f82, 150f83, covers for woodblocks 130 151f84a,b Optical microscopy identification of preparatory sketches (par tsak) of text raw materials of 16 on woodblocks 126–127 and papermaking sieves 34fb 172, 175 See also hand-painted decorations rag-paper manuscripts 30, 32, 100, 172 Stein, R.A. 52 ramie textiles 26, 27 Stellera chamaejasme (re lcag pa in Tibetan) silk scrolls 48 as a component in Tibetan papermaking and thangka painting 69 only 23, 110–111, 111f61 See also bookbinding; cotton; hemp; pothi added to papers for softness and format covers; ramie absorbency 112, 176 Thymelaeaceae family plant fibers collection and preparation of 194 characteristics of 188–189 and ‘cotton paper’ 184 and Chinese papermaking 184 features of 26, 27f4a,b, 111f61, 112f53, 113f64, in documents from Dunhuang 104–105 173f 106, 186–188 and ff 110–111 eaglewood (Aquilaria agallocha, aga ru) floating molds used with 194 and n.31 190–191, 194 in volumes of the Derge Kanjur 172, and Japanese gampi and mitsumata 173f 106 paper 44, 188 See also Daphne or Edgeworthia sp.; medicinal properties of some species Thymelaeaceae family plant fibers of 25, 189 310 index

Thymelaeaceae family plant fibers (cont.) Tibetan Kanjurs produced in Tibet See also Daphne or Edgeworthia sp.; Lithang Kanjur 163 Stellera chamaejasme papermaking sieve fingerprinting Tibetan block-printing technology information 177–178 blocks used as master copies for similarities of editions produced in re-editions 123, 125, 132 Eastern Tibet 173 and Buddhism 116 See also Cone Kanjur; Derge Kanjur; Lhasa carving of woodblock for Narthang Kanjur; Narthang Kanjur Monastery 128f71 Tibetan language, diversity of 3–4 and movable type 116–117 Tibetan papermaking: after the 1950s 192 non-absorbent paper used with 39, classes of material for manufacturing 44, 130, 168 paper 189–190 production process 125–132 distinguished from Chinese Tibetans as inventors of 121 papermaking 24, 44, 177–178 and the Turfan oasis 121 floating molds and woven paper and the xylograph of the Tshad ma production 33–34 and ff9a,b, 35, rigs pa’i gter 133 171f 104, 172, 191 See also Cone Kanjur; Derge Kanjur; highly sized and polished surface of Dunhuang documents; Lhasa Kanjur; paper 43, 44, 45f 14, 168 Narthang Kanjur; Tibetan book paper mulberry fibers in Tibetan/Tangut culture manuscripts from, Dunhuang Tibetan book culture oasis 106–108, 109ff59–60, 246 blue paper imitating indigo dyed paper preparation of dark and blue paper 42–43, 77, 78f33, 104 42–43, 103–104 Buddhism 50–51, 193, 201 shok bre (floating mold) 196–199 and sacred books as the embodiment of the ff 113–116 Buddha’s words 47 and Stella chamaejasme 23, 110–111, 111f61 sacred nature of books defaced by thickness of 45–46, 170 restoration 201 See also Dunhuang documents; Dunhuang and scroll painting (thangka) 85, 144 oasis; glues; Tibetan book culture significance of 47–48, 48f 16 Tibetan woodcarvers 129–130 Tibetan terms for books 52–53 Turfan oasis See also book cultures; calligraphy; early block prints from 119, 120f67, 180 deb ther (Tibetan bound books); Tibetan block-printing technology 121 Dunhuang documents; fore-edge typologies of book features, and local book painting; gold-color ink; hand- culture 3 painted decorations; illuminated manuscripts; miniature painting; Uighur book cultures Mongolian book culture; pothi format; block print from Turfan in Uighur Tibetan block-printing technology; language 120f67 Tibetan papermaking concertina printed books from the ninth Tibetan Book of the Dead (British Library century 119 manuscript OR 15190) and wooden movable type 116n.2 bamboo fibers in Chinese papers used Uray, Géza 180 for 32, 59 double covers of 58, 82, 83ff34–35 van der Kuijp, Leonard 63, 122 and pothi Tibetan manuscripts in gold van Schaik, Sam 7, 52, 180 81–82 Vorobieva-Desiatovskaia, Marguarita See also illuminated manuscripts Isifovna 55 index 311

Wangchuk, Dorji 79–80, 101 writing tools Wanli volumes about 101–102 chisel marks in letters of the Yongle/ for gold manuscripts 80 Wanli Kanjurs 136, 155f89, 166, 292 pen made of stone (brda rdo’ dmyu gu) laid- and chain-lines of paper 36 80 painted edges of 93–94, 147–149, Roman styli 73n.34 148ff80–81, 149f82, 150f83, 15184a,b and ‘spurs’ on particular letters 153, Yongle blocks used a master for 123, 125, 155f90, 165–166 and f98 139 stylus (bamboo pen) for Tibetan See also Peking editions calligraphy 44, 168 Berlin Wanli Kanjur (Pander Pantheon and tools for executing guidelines 98 volumes) tool traces as ‘fingerprints’ 16f 1, 129, Jagiellonian University Library 153 preservation of 8, 139 and n.43 wooden pens used with samta (tablets) volume 58 of 139n.43, 140 71, 73 Wanli Kanjur (Berlin manuscript Beck See also ink; wooden pens/sticks copy of the Peking Kanjur) glued layers of paper of 156, 157–158, 162 Xuanzang 49, 191 handwriting characteristics of 166 painted fore edges of 93–94, 15184a,b Yongle Kanjur ‘spurs’ on particular letters 153, 155f90 book cover from the Arthur Leeper text characteristics 150, 153 and f87 collection 140, 142–143, 145f77a,b See also Berlin State Library; Jagiellonian as the first printed edition of the Tibetan University Kanjur 137 Wanli Supplement volumes folio preserved in Michigan 55, 137, 138 Chinese origins of papers in 32, 36 and f72a,b, 147, 152f86, 154, 156, 157f91 cover of two volumes of Harvard-Yenching red ink used for 138f72a,b, 152f86, 154 Library 140, 141ff73a,b, 142f74 Sera Monastery set 138–139 cover preserved in volume 58 from the sets of 137–140 Pander Pantheon 140 and the Wanli Kanjur 123, 125, 137–138, painted edges of 147–149, 148ff80–81, 139, 140, 154, 157 149f82, 150f83 See also Peking editions; pothi format; six text lines per page of 153, 154f88 Yongle Kanjur covers two volumes in Harvard-Yenching Yongle Kanjur covers Library 8, 147, 149, 150f83 cover from Sakya Monastery set Ważny, Tomasz 16, 140n.48, 145ff77a,b (the R.R.E. Collection) 85, 143 and woodblock printing, European tradition f76 of 116 and n.1 cover from Sera Monastery set See also Tibetan block-printing technology (the R.R.E. Collection) 143 and f75, wooden pens/sticks 146f79, 270–271 handwriting characteristics of 166 Potala set (originally Sakya Monastery) non-absorbent paper required for 158, (the R.R.E. collection) 143 and f76, 168 271 See also writing tools Sakya type cover from the collection wood pulp of Arthur Leeper 140, 142–143, and other organic materials in Russian 145f77a,b types of paper 23, 24, 35, See also transcription of inscriptions on 270–272 organic materials See also pothi format covers