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

By

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 institu-
tions, 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 fea-
tures 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 cata-
logued. 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 excep-
tional 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 stud-
ies 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, col-
leagues 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 pre-
pare 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 polar-

ized 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 dendro-
chronological 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 vil-
lage 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 paper-
making 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 coun-
tries. 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 state-
ments but are also artefacts, and their physical properties can be studied using
methods similar to those used to study sculpture, painting, or common mate-
rial 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 archaeol-
ogy 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 crite-
rion 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 vari-
ety 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 physi-
cal 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 flour-
ished 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 cul-
tures, and the availability of materials all conditioned the production of books
written in Tibetan.
Additionally, Tibetan and other great religious languages throughout his-
tory 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 com-
mon medium of communication among Central Asian Buddhists by the end of
the seventeenth century, and was used at the beginning of the twentieth cen-
tury 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 liter-
ary 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 materi-
als 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 out-
side standard.

Statistics in Book History

Viewed as part of world heritage, ancient books are not only vehicles for con-
tent, but also very important sources of information about technology, materi-
als 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 individ-
ual books provides us with valuable information about a particular book and
possibly its provenance. However, from a broader perspective, the examina-
tion of many books and book fragments aids in the reconstruction of the
history of crafts connected with bookmaking, such as papermaking, ink pro-
duction, and the art of scribing. And even more broadly, the examination of
each book facilitates the collecting of evidence on book production from par-
ticular 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 mate-
rial 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 manu-

scripts 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 repre-
sentative of all Tibetan book history. Statistics are nearly as important for his-
torians 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 interest-
ing 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 sin-
gle 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 cre-
ation of blue paper, carpentry, and metalworking on book covers. Interviewing
Tibetan craftsmen was an important component of the reconstruction of prac-
tical work, which, of course, might differ regionally. Independent of regional
differences, the contributions of Tibetan artisans and monks sheds consider-
able 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 select-
ing 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 usu-
ally 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 manu-
script storage became its primary function.3 The almost fifty examined manu-
scripts 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 cal-
ligraphy. 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 docu-
mented 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 frontis-
piece 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 litera-
ture. 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 com-
prise one folio of the Yongle Kanjur preserved in the Special Collections Library,
University of Michigan; fragments of the Wanli Kanjur edition and its supple-
ment and a volume from one of the later editions of the Kangxi Kanjur both
preserved in the Jagiellonian University Library in Cracow, Poland; two vol-
umes 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 associ-
ated with Yonghegong that has not functioned as a library since the Cultural
Revolution. Located in the quarter behind Yonghegong, at present, this build-
ing 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 measure-
ments 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 impor-
tant 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 inter-
disciplinary character of manuscript studies gives rise to many methodological
problems. The lack of obvious or widely agreed-upon methods makes method-
ology 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 under-
take 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, mate-
rial 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.

© koninklijke brill nv, leiden, ���4 | doi 10.1163/9789004275058_003

Methods 13

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, lay-
out of page, and manuscript form. This showed whether all mentioned fea-
tures 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 mate-
rial analyses.
A paleographical approach allowed for comparative analyses of page out-
lines 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 informa-
tion about book origins and sometimes helped with dating. In addition, mate-
rial 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 compo-
nents, ink/paints composition, and wood in covers among other materials.
Because some research requires microsamples, material analyses are con-
ducted 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 geo-
graphical 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 infor-
mation. Nevertheless, technological typology does provide an important hint.
Even though the information will still point to quite a wide region, by compari-
son 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 sur-
face 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 indi-
cation 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 sec-
ondary 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 pig-
ments 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, yel-
low 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 consid-
eration as well. Additionally, there are many books that do not have any chron-
ological 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 mate-
rials 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 possi-
ble. 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 pro-
vide 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 den-
drochronology 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 con-
certina forms and many other methods of Tibetan binding. The differences in
binding styles seem to be more regional than temporal.9 They are also con-
nected 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 charac-
teristics 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 com-
pares 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 tradi-
tional 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 pig-
ments in more contemporary art shows that these rules did not restrict
Tibetans to using only traditional pigments. I interviewed a few thangka paint-
ers 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 mate-
rials from antiquity until the eighteenth century is relatively small, in total
about 25–30 compounds. Many are of mineral origin or synthetic reproduc-
tions 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 materi-
als that may serve as a fingerprint very rarely yields successful results.
Book covers are potential sources of information, but loose-leaf construc-
tion 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 gen-
eral 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 read-
able and continuous series of at least fifty rings. Dendrochronology’s require-
ments 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 chronolo-
gies, 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 nar-
rower 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 combi-
nation 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 inter-
disciplinary 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 terminol-
ogy to describe the paper in works of art.
There are further obstacles. One of them is the common, misleading descrip-
tion 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 technol-
ogy. 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 gen-
erally 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 lay-
ers), 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 papermak-
ing. Since the 1970s, there have been efforts to draw a clear distinction between
the Daphne and Edgeworthia species in ancient papers but without much suc-
cess. 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 simi-
lar 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 dif-
ferences 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 ques-
tion of differences between Daphne, Wikstroemia, and Edgeworthia fibers to
regional characteristics of papers. As was increasingly clear in the examina-
tions 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 maxi-
mum two types of fiber (a mix of Daphne and Edgeworthia fibers, which I clas-
sify 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 spe-
cies 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-mark-
ings, 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 con-
tains 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 tis-
sue 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 transpar-
ency, 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 compo-
nents. 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 longitudi-
nal 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 manu-
facture 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 centu-
ries), 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 usu-
ally envelops the fibers, which is visible on the SEM photo (Figure 7). In histori-
cal 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 mem-
brane depends on processing of pulp during papermaking and the quality of
the raw material. The membrane, or, primary wall, associated with paper mul-
berry 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 mul-
berry; 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 tradi-
tional 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 prob-
ably 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 cli-
mates. 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 disloca-
tions, 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 ves-
sels, 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 cen-
tury, but hemp remained the primary material for paper manufacture. During
the Tang period (618–907), hemp paper is considered as the best for calligra-
phy, 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 (maxi-
mum 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 taper-
ing 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, how-
ever, 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 materi-
als, 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 attribut-
able 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 crite-
ria for paper classification. Moreover, analysis of the fiber in the papers com-
pared 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 stan-
dards: 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, bam-
boo sieve. In Tibet, the woven type of floating mold was used since the begin-
ning 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 classifica-
tion 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, bam-
boo, 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 stan-
dard 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 writ-
ten 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 dis-
cern. 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 incorpo-
rated 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 composi-
tion: 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 xylo-
graph 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 straightfor-
ward (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 sub-
sequently 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 impor-
tant 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 dif-
ficult. 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 sub-
stances 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 differ-
ent 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 meth-
ods—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 poi-
sonous 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 nat-
ural 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 sub-
sequently 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 charac-
ters. 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 sub-
stances, 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 geo-
metric, 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 paint-
ing. 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 absor-
bency, 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 rep-

resenting 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, how-
ever, 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 con-
scious 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 sec-
ond half of the twentieth century, when Tibet had become part of China and
the editing process was automated. The monks reproduced many simple cop-
ies 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 compos-
ing a book and its format reflect its region of origin and cultural affiliation, but

1 Schaeffer 2009: 128–140.

© koninklijke brill nv, leiden, ���4 | doi 10.1163/9789004275058_004

48 chapter 3

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 signifi-
cant 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 man-
uscripts 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 cov-
ers (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 ear-
lier 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 prin-
cipal 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 thou-
sand 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 commu-
nities.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 vir-
tually 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 pos-
sible 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 prove-
nance 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 manu-
scripts 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 evi-
dence 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 cen-
tury, 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 wrap-
ping 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 lev-
els, 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 defi-
nition 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 deco-
rations 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 vol-
umes 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 vol-
umes produced in Beijing.
As mentioned above, the majority of the Tibetan books in all examined col-
lections 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 ori-
gin of a manuscript. Thus far, there have been no general studies of page lay-
outs 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 pres-
ence 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 for-
mats. 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 conif-
erous 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 expedi-
tion, 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 dis-
tance 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 reli-
gious 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 rep-
resentations 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 compris-
ing a variety of texts. An interesting feature of this type of binding is its appar-
ent 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 cen-
tury. 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, increas-
ing the longevity of the books. In contrast to the scrolls’ wooden dowels, con-
certina 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 com-
bination of both methods.22 The two Tibetan concertina manuscripts from
Dunhuang are both currently disjoined, but their original construction is evi-
dent in the narrow strips of paper that were previously used to join the indi-
vidual 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 compara-
ble to Western books. Although the origins of Islamic bookbinding are still
debated, their form seems to be a further development of the protective func-
tions 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 book-
let 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 gather-
ing 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 manu-
scripts 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 pre-
served 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 astonish-
ing 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 con-
firms 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 wood-
block, 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 pos-
sible 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 book-
let 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 col-
lection 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 com-
posed 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 manu-
scripts 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 lon-
ger 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 support-
ing the band of a gathering; the third of ten bi-folia with the same way sup-
ported 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 man-
uscript 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 expen-
sive 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 writ-
ten 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 dur-
ing 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 sur-
vival 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 offer-
ings, 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 edi-
tion 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 fac-
tor 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 elab-
orate 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 pro-
duction 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 func-
tion, 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 man-
uscript 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 bod-
hisattva Dharmodgata (chos ‘phags, Exalted in Dharma) to receive the teach-
ings. 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 teach-
ing 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 aes-
thetic 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 prac-
tices 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 manu-
scripts 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 techno-
logical 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 mate-
rial (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 scrip-
tures 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 under-
stood figuratively. Although, it is possible that royal decrees or letters of signifi-
cant 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 writ-
ten 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 docu-
ment. 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 pre-
pares 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 manu-
scripts 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, includ-
ing 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 complex-
ity 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 min-
iatures 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 cov-
ers, 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 lay-
ered paper joined together with leather strips/clasps and has a green silk ‘cur-
tain’ 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 callig-
raphy. 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 deco-
rated 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 pic-
tures 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, wood-
block-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 man-
uscript 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 cur-
tains, 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 mon-
astery 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 repre-
sentations 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 com-
posed 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 begin-
ning 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 geomet-
ric ornament framing the text of the title. These borders may consist of min-
iatures 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 fif-
teenth 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 composi-
tion 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-cen-
tury 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 con-
tain 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 yel-
low, orange, sometimes red, or purple. Often paints based on arsenic or mer-
cury 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 art-
ists 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 draw-
ings 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 edi-
tion 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 vari-

ety 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 differ-
ent 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 qual-
ity 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 decora-
tive elements, the placement of commentary, if any, in relation to the place-
ment 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 dis-
tance 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 fron-
tispiece. 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 manu-
scripts 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 mar-
gins. Sanskrit words are marked off with red ink while the rest of the manu-
script 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 writ-
ten 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 usu-
ally 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 but-
ter 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 con-
coction 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 pig-
ments. 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 for-
mat 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 con-
cept 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 sup-
ports 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 con-
sidered 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, writ-
ten 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 nar-
row 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 sharp-
ness. 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 combina-
tion 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 pro-
duced 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 sur-
face 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 identifi-
cation 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 com-
posed of Daphne or Edgeworthia plant fibers, this time mixed with paper mul-
berry 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 contain-
ing 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 uni-
formly 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 cen-
tury 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 com-
posed 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 man-
uscript 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 con-
ditions. Jia Sixie (473–554), an agronomist, describes paper mulberry cultiva-
tion 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 major-
ity of Tibetan manuscripts produced in Dunhuang and other Central Asian
sites such as Turfan and Miran are composed of rag paper primarily com-
posed 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 produc-
tion 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 for-
mats 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 hand-
made 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 chamae-
jasme 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 com-
position. 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 (unpub-
lished report). This C14 dating roughly coincides with that based on the people men-
tioned 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 manu-

script from the British Library.
These root fibers create a very specific soft type of paper, which is con-
sidered 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 espe-
cially 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 com-
posed 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 sin-
gular 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 mate-
rials were available to Tibetans during their imperial period and later for cre-
ating 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 main-
tained 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 centu-

ries 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 manu-
scripts 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 tech-
niques 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 mov-
able 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, char-
acters 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 duplica-
tion 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 origi-
nated 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 print-
ing 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 contain-
ing 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 docu-
mented, 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 con-
trast, 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 cul-
tures in Asia.7 Archeological findings suggest that the idea of printing was
inspired somewhere in Central Asia, possibly in a region under Chinese con-
trol. 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 multi-
cultural 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 cen-
tury 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, con-
tain Chinese, Tangut, Mongolian, Tibetan, Old Turkish, and Sanskrit texts, and
most can be dated to the thirteenth and fourteenth centuries, with only singu-
lar 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 informa-
tion 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 wood-
blocks 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 aban-
doned. 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 excel-
lent 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 gener-
ally 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 print-
ing and suggest that this work was executed by Chinese craftsman. However,
judging by the number of booklets preserved, it was not the most popular for-
mat of the book in China, especially not for important administrative or reli-
gious 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; how-
ever, these were achieved using a straight chisel as well. The page layout, qual-
ity 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 prepa-
ration 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 carv-
ers 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 hun-
dred, 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 wood-
block edition of the Tibetan Kanjur was produced in Beijing, all Buddhist
canonical collections were produced and reproduced as handwritten manu-
scripts. The Translation of the Word (Kanjur) and The Translation of Treatises
(Tenjur) undertaken in 1410 by the Yongle Emperor (r. 1402–1424), first pro-
duced in 1410 is known as the first large-scale printed edition and almost noth-
ing 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 (edi-
tion), 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 four-
teenth 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 mas-
ter 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 profes-
sionals 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 informa-
tion 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 pres-
ent 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 compe-
tence). 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 pagina-
tion 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 wood-
block 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 xylo-
graphic 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 preserva-
tion of the library (kun dga’ ra wa) in a monastery, and maintained the wood-
blocks 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 let-
ter 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 distin-
guish 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, exami-
nation of tool traces is even more important than the character and skills of
the craftsman, especially considering that one edition was produced by thou-
sands of carvers. For example, as reported by Corneille Jest, when several carv-
ers 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 proof-
read 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 pos-
sible 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 sub-
sequent 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 medi-
cines. Independently, a group of helpers assisted the highly specialized person-
nel 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 neces-
sary 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 techni-
cal 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, carpen-
ters, and proofreaders.
Since in modern times printing facilities are excellent, work productivity
has greatly improved. In the past, after one had proofread the original manu-
script, 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, secretar-
ies, 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 prepara-
tions 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 work-
shop 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 tech-
niques used for the books’ preparation define a set. According to an explana-
tion 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 out-
line 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 remem-
bered 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, docu-
mentation of all measurements characterizing page outline and material prop-
erties of books are essential. Then identification of woodblocks used in
particular printing workshops is an important benefit of the physical examina-
tion 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 mar-
gins, 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 desir-
able; 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 par-
ticular 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 ‘edi-
tion,’ for which a fresh set of blocks was prepared, and ‘re-printing,’ from
already existing blocks, still causes confusion. Classification becomes compli-
cated when we consider many re-editions, reprints, and only fragmentary cop-
ies 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 fol-
lowed, 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 his-
torical 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 col-
lector 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 follow-
ing the proper sense of the word ‘edition.’ Examining the Pander statement in
the Lamaist Pantheon and books in Cracow, I can explain this. His identifica-
tion 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 accu-
rate 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 transfer-
ring 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 distribu-
tion 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 dis-
tinction 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 proj-
ect 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 consider-
ation 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 par-
ticular 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 dat-
ing 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 partic-
ular 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 pro-
duced 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 pre-
sented 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 wood-
block 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 origi-
nal 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, pro-
duced 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 frag-
ments 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 letter-
ing 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 frag-
ment 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 cov-
ers 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 com-
posed 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 cin-
nabar (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 devel-
oped 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 pro-
cessed 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 cen-
ter of each cover displays a wish-granting vase filled with three gems. Four
meandering stalks emerge from the center to encircle eight auspicious sym-
bols: the wheel of law, victory banner, two fish, treasure vase, parasol, conch
shell, lotus flower, and endless knot (Figure 78). This suggests that this particu-
lar 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 simi-
lar, 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 pro-
duced at a date later than the text, are made from wood purchased elsewhere, or are dam-
aged 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 typi-
cal 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 sym-
bols 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 vol-
umes 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 frag-
ments 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 com-
posed 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 vol-
ume, 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 propor-
tions 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 cho-
sen 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 crafts-
men 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 examina-
tion 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 edi-
tion 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 edi-
tions were executed on laid paper characterized by narrow laid lines and com-
ponents 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.) charac-
terized 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 lay-
ers 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 mul-
berry, 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 hand-
written 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 iden-
tified 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 struc-
ture. 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 com-
posed of three types of cells. Plain paper mulberry fibers can make much bet-
ter 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 pro-
duced in Mongolia. It is assumed that the printed Mongolian Kanjur was mod-
eled 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 manu-
script 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 collec-
tion 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 com-
posed 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 edi-
tion of Kanjur from an approximately similar time at the beginning of the sev-
enteenth 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 inad-
equate 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 approxi-
mately 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 vol-
umes 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 struc-
ture 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 border-
land. 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 typi-
cal 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 cov-
ered 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 con-
tains 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 mea-
sures 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 folia-
tion 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 mea-
surements 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 character-
ized 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 tech-
nique 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 man-
uscripts 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 lay-
ers, 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 origi-
nal paper strings from the Cone Kanjur all show the presence of paper mul-
berry and straw fibers (Figures 101 and 102).
I examined three volumes, 1, 45, and 77, from the set of Derge Kanjur pro-
duced 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 par-
ticular 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 brown-
ish 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 vol-
umes 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 absor-
bent 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 paper-
making 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 bor-
derland 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 pat-
tern 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 vol-
umes and page outline. For example, the leaves of the Cone and Derge Kanjur
sets are visibly smaller and of similar proportions, resembling palm leave pro-
portions (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 edi-
tions, 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 dis-
tance 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 dis-
tance between upper lines of 1.7–1.8 cm. The frame of the printed area mea-
sures 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 pro-
duced 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 dis-
tances 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 paper-
making 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 handwrit-
ten 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 infor-

mation 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 exam-
ined 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 identifica-
tion 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 pro-
duction 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 finger-
printing 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 finger-
printing 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 inher-
itance 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 trans-
fer 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 histori-
cal 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 inven-
tion, 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 west-
ward 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 dis-
covered 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 writ-
ten 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 west-
ward 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 pris-
oners 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 sur-
veys 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 proper-
ties 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; train-
ing 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 mate-
rial 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 gradu-
ally 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 surround-
ing regions are available.5 Perhaps the earliest is the account by the British
official Samuel Turner of papermaking in Bhutan in the mid-eighteenth cen-
tury.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 begin-
ning of papermaking to such a great degree as, for example, in Europe or in
China.
Documentary materials found in remaining archives of Himalayan vil-
lages 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 writ-
ten 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 papermak-
ing. In the Tibetan culture of papermaking, the variations of production meth-
ods, 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 fea-
tures. 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 cli-
mate make the vegetation distinctive from all other areas of Asia. In other
words, the range of plants available for papermaking in Tibet was signifi-
cantly different than in other places. As far as the physical conditions are con-
cerned, 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 west-
ern 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 papermak-
ers 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 materi-
als 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 mate-
rial 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 suppos-
edly 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 clar-
ify 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 con-
sidered to be of lower quality than bark paper. The roots are especially dif-
ficult 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 rela-
tive importance of each of these different fibers in the early history of Tibetan
papermaking. Stellera chamaejasme fibers are distinctive in the fiber exami-
nation of historic papers, and can be clearly differentiated from Daphne and
Edgeworthia despite the fact that all plants belong to the Thymelaeaceae fam-
ily. Species of Oxytropis (Leguminosae) were also used as an addition for mak-
ing 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; how-
ever, 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 dif-
ficult 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—some-
times 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 con-
sidered 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 dur-
ing the papermaking process—allowing for the production of thin sheets. Yet,
Tibetan paper could hardly be described as thin. Quite the contrary—thick-
ness 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: agallo-
cha) 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 tradi-
tional 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 unpro-
cessed form sometimes, similar to palm leaves. Aquilaria trees reach 120 feet in
height with trunks with twelve foot circumferences; they can live up to thou-
sand years. Its wood is pale white in color, light, soft, and porous, and con-
tains 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 phar-
maceutical 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.aloes-
wood.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 mak-
ing 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 manu-
scripts 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 usu-
ally 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 mov-
able 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 cot-
ton, 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 agallo-
cha), 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 east-
ern 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 hand-

made 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 condi-
tions 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 depend-
ing 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, har-
vesting 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 col-
lected 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 con-
tent 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 cook-
ing 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 mak-
ing 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 pro-
long 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 care-
fully 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 cul-
tural 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 monas-
teries 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 some-
times also respected as an object with attributed magical powers. The accu-
mulation 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.

© koninklijke brill nv, leiden, ���4 | doi ��.��63/9789004275058_��8

202 chapter 7

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 loca-
tions 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 transi-
tion.’ 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 genera-
tions 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 pro-
tection. 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 proj-
ects 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 con-
servation with expectations of local communities, which simply desired to
have copies of old paintings produced. Tibetans do not have a tradition of pre-
serving 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, documen-
tation, 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, depend-
ing 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 universi-
ties, 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 profes-
sion 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 impor-
tant 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 termi-
nology 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 treat-
ment 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 (includ-
ing 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 conserva-
tion, archaeology, or archaeometry.
 Preservation: The protection of cultural property through activities that minimize chemi-
cal 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 pur-
pose 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 differ-
ences 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 differ-
ent 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 treat-
ments 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, fra-
gility, 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, some-
times 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 for-
mat, 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 pro-
tected 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 pro-
tected 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 lay-
ers 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 lay-
ers 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 usu-
ally 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 find-
ings that paper in Tibetan books was successfully protected by insect repel-
lent 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 yel-
lowish 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 sub-
stances 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 men-
tioned 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 mate-
rials 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 supple-
mental and the Prajñāpāramitā section volumes are more stained with fox-
ing 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 sug-
gest 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 con-
tains 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 espe-
cially vulnerable to deterioration. During the first half of the nineteenth cen-
tury, 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 destruc-
tion 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 manu-
scripts 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 mois-
ture. 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 stor-
age 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 hydro-
lysis 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 qual-
ity, but the influence of acids may also be observed on good handmade rag
papers. The process of oxidation occurs constantly and slowly, under the influ-
ence 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 manu-
script written on handmade paper, depending on handling and use, has a typi-
cal physical lifetime of two to three centuries. Then it requires a new copy. If
carefully preserved in a more favorable climate, paper manuscripts may sur-
vive 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 deteriora-
tion 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 tempera-
ture 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 envi-
ronment 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 com-
bustion 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 depos-
ited in soot and dust suspended in the air. Additionally, dusts catalyse oxida-
tion 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 practi-
cally 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’ activ-
ity 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 col-
lected 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 iden-
tifying 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 particu-
lar object. Documentation can be produced in electronic or hard copy and pre-
served 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’ web-
site or in ‘Commentaries to the Guidelines for Practice’ on the website of the
American Institute for Conservation of Historic and Artistic Works.3 More gen-
erally, documentation allows for evaluation of treatment methods and mate-
rials, support for education and scholarly research, provision of a record for
current conservation practice, and study of the history of the conservation pro-
fession 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, treat-
ment, 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 analy-
ses 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 con-
servators and owners—regarding expectations, potential benefits, costs, and
risks. The treatment plan should include at least the proposed course of treat-
ment, 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 num-
bered 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 pow-
der, 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 pres-
ervation procedure. Since Tibetan books are very often covered by dust and
dirt, dry cleaning may not give satisfactory results. In such cases, carefully per-
formed 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 deci-
sion 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 usu-
ally 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 sup-
port 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 second-
ary 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 appropri-
ate 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 manu-
scripts 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 flak-
ing. 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 distin-
guished 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.

© koninklijke brill nv, leiden, ���4 | doi ��.��63/9789004275058_�09

226 appendix 1

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 recog-
nize. Large amounts of glue and binding media in the paper structure (stains seem-
ingly 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, paint-
brush 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 nar-
row 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 ana-
lyzed 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 smooth-
ened. 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 rect-
angular frame 25 × 6.7 cm is carved out for 0.3 cm depth. The cover contains a paint-
ing 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 struc-
ture 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 iden-
tified. 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 cen-
trally 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 outer-
most 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%) diago-
nally 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 expo-
sure 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; bot-
tom 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 mar-
gins 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 mar-
gins 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 sug-
gests 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 mar-
gins 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, cen-
trally 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 heav-
ily 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 paper-
maker’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 orna-
ment, a variation of an endless knot. Three curtains are stitched together on the bot-
tom 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 inter-
vals: 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; paper-
making 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

fragment Content production leaves (h × w)

Yongle Kanjur: Special Collections 1410 Beijing 24.2–24.5 cm

the folio 12 of the Library at the [Peking] in height and
Samādhirāja-sūtra University of 68.7–69 cm in
Michigan, Ann length
Arbor, USA

Wanli Kanjur: Jagiellonian 1606 Beijing 23.8–24.5 cm in

the twenty-eight University Library [Peking] height and 68.5 cm
volumes: Nos. 23–28 in Cracow, Poland in length
(rgyud), 38–57 (Biblioteka
(nos. 38–55 = rgyud; Jagiellońska w
56 = mdo; 57 =rgyud), Krakowie)
59=rgyud),
60=dkar chag)

Wanli Kanjur Jagiellonian 1606–1607 Beijing Approximately

Supplement: University Library [Peking] 23.8–24.5 cm in
Pander Pantheon: in Cracow, Poland height and 68.5 cm
volumes 1–22 (Biblioteka in length
Jagiellońska w
Krakowie)
Wanli Kanjur Harvard-Yenching 1606–1607 Beijing Approximately
Supplement: [Peking] 23.8–24.5 cm in
volumes tsa and ku of Library in height and 68.5 cm
the Wanli Supplement Cambridge MA, in length
USA

© koninklijke brill nv, leiden, ���4 | doi ��.��63/9789004275058_��0

 features of paper in SELECTED sets of tibetan kanjur 253

Page outline Ink Raw material Papermaking sieve Construction of

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

8 Red Paper mulberry Laid paper Six or more layers

1.5–1.7 cm (Broussonetia sp.) characterized by glued together; good
Side margins 27–30 laid lines in quality and well
apart 59 cm 3 cm; chain lines preserved paper
from each other not possible to see

8 Black Paper mulberry Laid paper Six or more layers

1.5–1.7 cm (Broussonetia sp.); characterized by glued together;
Side margins within the Wanli 24 laid lines in 3 cm; the paper on which
apart 59–59.5 cm set the best quality some of the papers the laudation text is
from each other (longest) fibers show the presence written has about
were used for of the 15 or 18 laid 4 layers, which
volume 60 lines in 3 cm; chain makes the leaves
(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.

6 Black Paper mulberry Laid paper Six or more layers

1.7–2.2 cm (Broussonetia sp.), glued together
Side margins straw, and bamboo
apart 58.5–59 cm
from each other

6 Black Paper mulberry Laid paper Six or more

1.7–2.2 cm
Side margins (Broussonetia sp.), layers glued together
apart 58.5–59 cm straw, and bamboo
from each other
 254 appendix 2

Table (Continued)

Set Examined Present Location Dating Place of Format Size of

fragment Content production leaves (h × w)

‘Handwritten Kanjur Berlin State 1680 Beijing Approximately

of Berlin’: Library, Germany [Peking] 26.8–29 cm in height
vol. 1, ka; vol. 24 Teil 1 (Staats-bibliothek and 68.9–73.3 cm
and Teil 2, nya; vol. 17 zu Berlin) in length
Teil 1 and Teil 2, ka; vol.
18 Teil 1 and Teil, kha;
vol. 108 Teil 1 and
Teil 2, ja; vol. 85 Teil 1
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

corrected reprints of University Library [Peking] 24.5–25.5 cm in
the Kangxi Kanjur in Cracow, Poland height and 71.5 cm
edition: sher phyin, (Biblioteka in length
tha (vol. 58 from the Jagiellońska w
Pander Collection) Krakowie)

Handwritten Mongolian St Petersburg Not earlier 23.3 cm in height

Kanjur University, Faculty than 1629 and 68 cm in length
First volume of ka of Asian and
tantra section African Studies
collection, Saint
Petersburg, Russia

Printed Mongolian St Petersburg Beginning 22.3 cm in height and

Kanjur from Kangxi University, Faculty of 17th c. 65.9 cm in length
times; tantra of Asian and
volume (E59) African Studies
collection, Saint
Petersburg, Russia
 features of paper in SELECTED sets of tibetan kanjur 255

Page outline Ink Raw material Papermaking sieve Construction of

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

8 Black Paper mulberry Laid paper Three or more layers

1.8–1.9 cm and red (Broussonetia sp.), characterized by glued together;
Full frame bamboo, and jute 27–33 laid lines in highly sized and
18.8×55 cm 3 cm polished; possibly
also covered with
other substances
which increased its
whiteness

8 Red Paper mulberry Laid paper with Three or more layers

1.5–1.7 cm (Broussonetia sp.), hardly visible glued together
Full frame bamboo, and straw structure
19×59 cm

Full frame Black Paper mulberry Laid regular paper Two layers glued
20.5×61.9 cm (Broussonetia sp.), characterized by together
and bamboo 22 laid lines in 3 cm,
chain line intervals
3.5–4 cm

Full frame: Black Bamboo Laid regular paper Two or three layers
17.3×55 cm and red characterized by of paper glued
36 laid lines in 3 cm, together
chain line intervals
not visible
 256 appendix 2

Table (Continued)

Set Examined Present Location Dating Place of Format Size of

fragment Content production leaves (h × w)

Printed edition of St Petersburg 1714 Beijing 23.7 cm in height

Mongolian Kanjur (E2), University, Faculty [Peking] and 67 cm in length
vol. 9, No. 19 in of Asian and
Uspensky Catalog African Studies
Heissing Blockdruke collection, Saint
no. 26 Petersburg, Russia

Cone Kanjur: vol. 1 Library of Congress, 1721–31 Cone Approximately

mdo (sūtra), ka ; vol. 35 Washington, (The 18–18.5 cm in height
’dul-ba (vinaya), ga; DC, USA examined and 56.5–57 cm in
vol. 92, yum, ka; copy was length
vol. 108 dkar chag printed in
(index); vol. 72 rgyud 1926.)
(tantra); Original
Paper Strings from
Cone Kanjur
 features of paper in SELECTED sets of tibetan kanjur 257

Page outline Ink Raw material Papermaking sieve Construction of

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

Black Paper mulberry Laid regular paper Two layers of paper

and red (Broussonetia sp.),
bamboo and jute
characterized by 36 glued together
or even 39 laid lines
in 3 cm; chain line
intervals can be
3.5–4 or 5 cm, but
very hardly visible
due to paper glued
in layers

8 Black Paper mulberry Laid paper Slightly yellowish

0.9–1.3 cm (Broussonetia sp.) characterized by (cream), pasted in
Full frame and straw about 27 laid lines two or possibly three
10.3×47 cm in 3 cm; chain lines layers; absorbent
not visible; a paper and soft; glued using
fragment with a title diluted starch
(label) printed on a paste rather than
lotus flower in the any kind of animal
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.
 258 appendix 2

Table (Continued)

Set Examined Present Location Dating Place of Format Size of

fragment Content production leaves (h × w)

Narthang Kanjur: Berlin State 1730–32 Narthang Approximately

vol. 1 ’dul-ba (vinaya) , Library, Germany 17–18 cm in height
ka, and vol. 2 ’dul-ba (Staats-bibliothek and 61.5–63.5 cm in
(vinaya), kha zu Berlin) length; particular
folios are not evenly
cut

Derge Kanjur: Library of Congress, 1733 Derge Approximately

vol. 1 ’dul-ba (vinaya), Washington, DC, 10–11.5 cm in height
ka; vol. 45, mdo (sūtra) USA and 60.5–62 cm in
volume, ka; vol. 77 length
rgyud (tantra), ka
 features of paper in SELECTED sets of tibetan kanjur 259

Page outline Ink Raw material Papermaking sieve Construction of

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

7 Black Thymelaeaceae Both a woven type Very thin one-

1.5 cm family plants of paper made with layered paper with
Full frame (Daphne or thick textile as a glossy surface and
11.5× 58 cm Edgeworthia sp.) woven sieve and uneven thickness;
possibly a finely the middle part of
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

7 Red Thymelaeaceae Laid paper One-layered, soft

0.9–1 cm family plants characterized by and absorbent paper
Full frame (Daphne or 15 laid lines in 3 cm; characterized by a
7.2–7.4 × Edgeworthia sp.) chain lines in a brownish color and
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
 260 appendix 2

Table (Continued)

Set Examined Present Location Dating Place of Format Size of

fragment Content production leaves (h × w)

Lhasa Kanjur: Berlin State Library, 1934 Lhasa 15.5–17.5 cm in

vol. 1,’dul-ba (vinaya), Germany (Staats- height and 62.8–64
ka and vol. 2, ’dul-ba bibliothek zu cm in length;
(vinaya), kha Berlin) particular leaves are
not evenly cut
 features of paper in SELECTED sets of tibetan kanjur 261

Page outline Ink Raw material Papermaking sieve Construction of

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

7 1.3–1.6 cm Black Thymelaeaceae Both a woven type One-layered, very

Full frame family plants of mold and a laid thin paper with
11.8×57 cm (Daphne or mold characterized visible fiber bundles
Edgeworthia sp.) by about 21 laid in its structure
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 assis-
tance 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 avail-
able at the TBRC in volume 59 (volume letter=ba) here: http://www.tbrc.org/#!rid=W22084.

© koninklijke brill nv, leiden, ���4 | doi ��.��63/9789004275058_011

transcription of the yongle covers inscriptions 263

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 fea-
tures a precious pearl in flames and Tibetan text in the halo.

© koninklijke brill nv, leiden, ���4 | doi ��.��63/9789004275058_012

266 appendix 4

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 some-
thing 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 (dis-
tance 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 writ-
ten 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 (dis-
tance 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 brown-
ish); 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 cor-
rected (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 differ-
ent 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 (dis-
tance 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 cor-
rected (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 (dis-
tance 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 frag-
ment 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 paper-
making 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 thick-
ness, 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 meas-
ures 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 paper-
making 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 print-
ing 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 read-
able 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; absor-
bent, 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 xylo-
graph 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 xylo-
graph 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 folia-
tion 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 dis-
tance 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 measur-
ing 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 infor-
mation 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 com-
pound had been applied against insects. The paper fragments pasted to particular
pages indicating contents of book when looking at fore-edges were added later, prob-
ably 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, prepara-
tion 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 addi-
tional 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 folia-
tion 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 charac-
teristic 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 num-
bers typed on thin tissue paper.
Ink: Characteristic greyish color appears slightly diluted (not the best quality), let-
ters 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 origi-
nal 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 infor-
mation 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 charac-
teristic 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 num-
bers typed on thin tissue paper.
Ink: Characteristic grayish color appears slightly diluted (not the best quality), let-
ters 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 origi-
nal paper strings).
Type of paper: Laid paper (two- or three layers): 27 laid lines in 3 cm; slightly yellow-
ish (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 infor-
mation 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 xylo-
graph 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 dou-
ble-lined frame measuring 11.5 × 47 cm.
Folio 3 and the remaining leaves in the volume: recto—8 lines of text, with a dis-
tance 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 tan-
tra) 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 dam-
age. 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), let-
ters 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 origi-
nal paper strings).
Type of paper: Laid paper (two or three layers): 27 laid lines in 3 cm; slightly yellow-
ish (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 infor-
mation identifying the volume number handwritten in Tibetan letters. Covers: All vol-
umes 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 dis-
tance 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 par-
ticular 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), let-
ters 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 dis-
tance 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 print-
ing 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 exam-
ined; 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 Ö sterreichischen
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 con-
tents: 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 tibe-
tischen 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 liv-
ing 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 availa-
ble 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 pro-
ceedings, 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 manu-
scripts. 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)
methods 20–21
detection of Stellera chamaejasme fibers
in an tenth-century manuscript 111
Aquilaria agallocha (aga ru, eaglewood)
190–191, 194
Arabian book culture
and Chinese book culture 50, 180
and gold on blue manuscripts 80
Islamic bookbinding 62
Islamic-founded cities as centers of
papermaking 51
and Kashmiri book culture 64 and n.27
pens used by 51, 80
See also book cultures; Persian book
culture
‘archaeology of the book,’ defined 2–3
arsenic sulphide
fore-edges painted yellow tested
negatively for 163, 285, 287, 288, 289
as an insecticide 25, 43, 93
Asia and Pacific Museum (Warsaw)
Andrzej Wawrzyniak (director) 1–2
illuminated manuscript in (MAP 4323)
91, 92f48, 100, 211, 224
paper in layers (MAP 14136) 208f 118
’phags pa shes rab kyi pha rol tu phyin pa
rdo rje gcod pa (Diamond Sutra) 1, 91,
92f48, 95–96, 100, 224
restoration of paper at 222–223 and
ff 127–128
Tibetan book (MAP 12385) 215f 121
Tibetan manuscript (MAP 14144)
25f3b, 45f 14, 223f 128
bamboo fibers
added for softness and absorbency 176
characteristics of 32
genera used for papermaking 32
in papermaking in China after the Tang
32
in the paper of the Mongolian Kanjur
(St. Petersburg University, Faculty of
* Page locators in italics indicate figures.
Asian and African Studies collection)
160f92, 161ff94-95, 162
bamboo sieves
with different weaves used
for sections of the Derge Kanjur 170
and dipping methods of papermaking
from China 35–36, 59, 191–192
and 192f 112
laid regular structure of paper from
movable sieves 36, 110, 159
and Tibetan/Tangut manuscript Dx 178
from the Institute of Oriental
Manuscripts 108
See also papermaking tools
Beck manuscript editions. See under
Wanli volumes, Berlin manuscript Beck
copy of the Peking Kanjur
Berlinka collection. See Pander Collection
Berlin Kanjur (manuscript Beck). See under
Wanli volumes, Berlin manuscript Beck
copy of the Peking Kanjur
Berlin manuscript copy of the Wanli volumes.
See under Wanli volumes, Berlin Kanjur
(manuscript Beck)
Berlin State Library
Lhasa Kanjur from 175–176, 177f 108
Narthang Kanjur set from 162, 173,
175–176 and f 107, 264
See also Wanli volumes
Bhutanese book culture
bound book written in gold (MS 13162)
69, 70f29a,b,c, 81–82
laid type of mold used in 172
Samuel Turner’s mid-eighteenth century
account of 182
and Thymelaeaceae-family plants 33
and n.29
See also book cultures
binding. See bookbinding
birch bark
bound book on birch bark in the
Tibetan Museum in Lhasa 63–64, 67
brittle fibers of 211
300
birch bark (cont.)
and Kashmiri book culture 48–49, 63,
84n.64
in Tibetan ink recipes 100
blood
‘nose-blood’ (shangs mtshal) in ink
recipes 101
vermillion blood (mtshal khrag) in ink
recipes 101
Bon tradition, Tibetan gold manuscript
‘Life of Gshen rab mi bo’ 91, 93f49, 94,
224 and f 129
bookbinding
bound book on birch bark in the Tibetan
Museum in Lhasa 63–64, 67
and conservation techniques 53, 55
and hemp 31
Islamic bookbinding 62
linen cord stitching of bound manuscripts
(CC 101, CC 114–115, BL 14727 1–2) 69,
71, 72f30a,b, 73f31a,b,c,d, 74f32a,b,c,d
of palm-leaf pothi books 56, 57f20
regional nature of 18
representative books from the British
Library 7 and n.3, 69, 71
sectional binding of Bhutanese book
written in gold (MS 13162) 69,
70f29a,b,c, 81–82
stitched bindings of deb ther 69,
70f29a,b,c, 81–82
Tibetan manuscript written in dbu med
script (OR 15193) 66f27a,b,c,d, 69, 71
Tibetan stitched book on paper from
Dunhuang (IOL J 401) 63, 64f23a,b
Tibetan stitched book on paper from
Dunhuang (IOL J 530) 63, 65f26a,b
See also codex form (thread booklet)
concertina format
book cultures
and local bookmaking techniques 3–4
material aspects of books as important
historical sources 4–5
See also Arabian book culture; Bhutanese
book culture; Chinese book culture;
European book culture; ‘fingerprints’
of time and origin in materials; Indian
book culture; Japanese book culture;
Korean book culture; Mongolian book
index
culture; Nepalese book culture; Persian
book culture; Tibetan book culture;
Uighur book culture
Broussonetia sp. (genus of the Moracae
family)
about 28–29 and ff6a,b, 109ff59–60
See also paper mulberry
(Broussonetia sp.)
Buddhism
concertina format books used by 60
and the invention of printing 116, 117–118
and the production of books written in
Tibetan 4, 48, 193
Tibetan books as the embodiment of the
Buddha’s words 47
Tibetan Buddhism as a living tradition
201–202 and f 117
Butön (1290–1364) 123
Cai Lun’s biography in the Hou Han Shu 29,
183, 184
calligraphy
dbu can (‘headed’) script 95, 97f52, 97,
98f53
dbu med (‘headless’) script 66f27a, 95,
98
gold script 100
shape and proportions of letters 153
‘spurs’ on letters 153, 155f90, 165–166 and
f98, 167f99
stylus (bamboo pen) used by Tibetans
44
See also gold-color ink; ink; paleography;
raised gold lettering; writing tools
carving tools
about 127, 129 and nn.24–25
chisel marks in letters of booklet from
Khara Khoto 123
chisel marks in letters of the Yongle/Wanli
Kanjurs 136, 155f89, 166, 292
straight type chisel used for the Cone
woodblocks 166
and woodblock identification 16f 1, 127,
129, 133, 135–136, 153, 155f89, 166, 227
Chinese book culture
and the Arab world 50, 180
and block printing 117–119, 121–122
butterfly binding 67–69
index
Chinese pagination 119–121, 122, 154
and fore-edge painting 93, 94 and n.15
and gold on blue manuscripts 80
and hemp cultivation 30–31
and indigo paper dyed by dipping 59,
104, 226
and movable type 116n.2
and Nia documents from the third and
fourth centuries 180
scroll format 48, 59, 74
See also book cultures; Chinese
papermaking
Chinese papermaking
as a writing material 44, 176
and the Cone paper 167, 172–173
dating of oldest paper fragments 31
distinguished from Tibetan
papermaking 176–178
fibers cut during processing of 44
ink blots characteristic of absorbent and
soft types of paper 167
and the migration of Chinese
papermaking communities 179–180
and paper mulberry 26–27, 29–30, 36
plant materials used in 22–23, 31–33
starch paste for gluing layers of paper  conservation science, and restoration
208
and Thymelaeaceae family plants 184
See also Chinese book culture;
papermaking processes
cinnabar
as a coral substitute 100
as a natural insecticide 150
as a symbol of power 100–101
in lacquer 142n.49
and Tibetan Kanjurs 154
codex form (thread booklet)
elements of described 62
and European book culture 3, 53, 62
Tibetan types of 52, 60, 62, 63
codicology
as a tool 3, 12, 13
and the history of Tibetan printing 135
Columbia University, Tibetan gold
manuscript ‘Life of Gshen rab mi bo’ 91,
93f49, 94, 224 and f 129
concertina format
as a synthesis of Chinese and Indian/
Southeast Asian book forms 60
301
books from the Pander collection 60–61
and f23
for government documents 61, 62f24
and Tibetan books produced in China
74–75
Uighur examples from the ninth
century 119
Cone Kanjur (Library of Congress)
dkar chag 163 and n.61, 165, 167, 168f 100,
261–262
Joseph Rock’s acquisition of 164f96, 172
lettering characteristics 165–166 and
f98
paper quality 167, 169 and ff 101–102,
172–173
Prajñāpāramitā volumes (yum) 164–165
and f97, 167f99, 288–289
style and format of pages of 164–169
See also Tibetan book culture; Tibetan
Kanjurs produced in Tibet
conservation
binding elements 53, 55
environmental controls in libraries 213
and handling 206, 212, 214–215
preservation, prevention conservation,
distinguished 205–206
preventive conservation 205 and n.2,
220
Tibetan book storage methods 215–216,
217–219ff 123–125
typical damage to Tibetan books 206
Western conservation attitudes 203–204
See also handling; preservation
conservation science 205–206, 205n.2
coral (byu ru)
cinnabar substituted for 100
and gold manuscripts 80, 100
mounted on front pages as elements of a
circular shape imitating holes in palm
leaf books 86, 87f39, 231
See also organic materials
cotton
features of 32
in Ming Dynasty textiles 31
See also organic materials
covers. See pothi format covers; Yongle Kanjur
covers
Cupa, Adam 15
302
Daphne or Edgeworthia sp. (shog shing or
dung loma in Tibetan)
collection and preparation of 194
and ‘cotton paper’ 184
difficulty in distinguishing two fibers
of 23–24
dug shog (poisonous paper) as a Tibetan
term for Daphne 25, 189
in Dunhuang manuscripts 104–105
Edgeworthia gardineri sp. plant 185f 109
fibers described 24–25, 25f3a,b,c,d
fibers in the paper of the Derge Kanjur
(Library of Congress) 173f 106
fibers in the paper of the Lhasa Kanjur
(Berlin State Library) 175–176, 177f 108
fibers in the paper of the Narthang
Kanjur 176f 107
fibers in the Sel dkar Kanjur 106f57
high starch content of the Daphne plant
39, 47f 13b, 189, 207
known as lokta and kāgate in Nepal
185f 109, 188
and Tibetan papermaking 24, 177
See also Stellera chamaejasme;
Thymelaeaceae family plant fibers
deb ther (Tibetan bound books)
and other formats of Tibetan book 52,
206
and Persia daftar (bound manuscripts) in
Mongolia 63
stitched bindings of 69, 70f29a,b,c, 81–82
See also Tibetan book culture
de La Vallée Poussin, Louis 52
dendrochronology of pothi format covers
19, 20f2, 227
Derge Kanjur (Tibetan Rare Book Collection,
The Library of Congress)
no rounded chisels used for (in contrast
to the Yongle/Wanli) 293
pagination of 170, 291, 292, 293, 294
papermaking molds of various types
used in 170, 171f 104, 172
paper of rgyud volume 77 distinguished
170
paper quality 170–173 and f 104
Thymelaeaceae family plant fibers of
170, 171–172, 173f 106
See also Tibetan book culture; Tibetan
Kanjurs produced in Tibet
index
Derge Printing House (Sde dge par khang)
establishment of 169
papermaking in 35n.33, 171–172, 182
woodcarving methods at 127, 130
Diamond Sutra
illuminated manuscript in the Asia and
Pacific Museum (Warsaw) of 1, 91,
92f48, 95–96, 100, 224
lapis lazuli paper mentioned in colophon
of 103
in the Pomeranian Library in Szczecin,
Poland 100
dug shog (poisonous paper)
as a Tibetan term for Daphne 25, 189
See also Daphne or Edgeworthia sp.
Dunhuang documents (IOL Tib J from the
Stein Collection, the British Library)
231–249
hemp as a component in manuscripts
from 30
laid line characteristics 172
paper types of manuscripts found in
104–105, 110
ramie as a component in manuscripts
from 28f5a,b
root paper of Stellera chamaejasme not
found in 110
See also Dunhuang oasis; Tibetan book
culture
Dunhuang oasis
concertina format manuscripts from 60
and early manuscripts written on
paper 51
oldest preserved Tibetan books found
in 52
paper mulberry fibers in Tibetan/
Tangut manuscripts from 106–108,
109ff59–60, 246
scroll format in 52, 59, 74, 208
wide variety of formats of books in 52
See also Dunhuang documents
eaglewood (Aquilaria agallocha, aga ru)
190–191, 194
Edgeworthia gardineri sp. See Daphne or
Edgeworthia sp.
Thymelaeaceae family plant fibers
edition, re-edition, and reprinting
introduced 132–136
index
technical similarities revealed by
comparative examination of Kanjur
sets 178
Eimer, Helmut
Berlin Wanli Kanjur deemed lost by 10,
139n.41
and the ‘technical identity’ of hundred-
volume sets 132, 135
Euphorbia fisheriana 43, 184, 186, 195, 209
European book culture
and mass printing 77, 211
and movable type 116–117
move from paprus rolls to parchment
codices 3
and printing of images 116n.1
Ramie fibers brought to 27
Roman tablet writing 73 and n.34
sizing of paper for writing 39
woodblock printing in 116 and n.1
See also book cultures; European
medieval manuscripts; Russian
paper
European medieval manuscripts
and curtain coverings 85
gold ink on colored background 80
initial letters 93
patrons as actors in miniature scenes
on 78
rubrics 13 and n.1, 97
Tibetan calligraphy in ancient books
compared with 44
and the valuation of printed books 76
See also book cultures; European book
culture; illuminated manuscripts;
miniature painting
‘fingerprints’ of time and origin in materials
laid- and chain-lines of papermaking
sieves 36, 158, 177–178, 178, 191
overview of 17–18
raw material/fiber composition as 23,
56–57, 104–105, 145f77a,b 177
shape of letters/handwriting style 105,
133, 153
and tool traces 16f 1, 127, 129, 133, 135–136,
153, 155f89, 166, 228
See also book cultures
flax
and other organic materials in Russian
types of paper 23, 24, 35
303
and the term ‘ma’ in old Chinese
documents 31
See also organic materials
fore-edge painting
as a characteristic of Tibetan Peking-
Kanjurs 94 and n.15
as insect repellents (arsenic or mercury
based pigments) 93
ordering of pages in pothi format made
visible by 94
in red on the Derge Kanjur set in the
Library of Congress 170
of Wanli volumes 93–94, 147–149,
148ff80–81, 149f82, 150f83, 15184a,b
yellow tested negatively for arsenic
sulphide 163, 285, 287, 288, 289
See also hand-painted decorations
glues
discoloration caused by 208, 210f 119, 226
glued layers of paper 36, 43–44,
45–46, 156, 157–158, 162, 207, 208
for sizing 39–40, 45f 14
smoothing paper surfaces 43, 44
starch paste 43–44, 115, 132, 167, 206
unripe resin in 100
See also papermaking processes; Tibetan
papermaking
gold-color ink
alternating with silver text on the Lotus
Sutra from Columbia University
Library 97 and f52, 98 and f53
on a manuscript of the Mongolian
Kanjur 159
on black or blue ground in Tibetan book
culture 100
bound book written in gold (MS 13162)
69, 70f29a,b,c, 81–82
and concertina format books 60
deterioration of 211, 223–224 and f 129
Tibetan terms for scriptures written in
81
writing tools for 80
See also gold-color ink; gold manuscripts;
ink; raised gold lettering; Tibetan book
culture; Yongle Kanjur covers
gold manuscripts
Diamond Sutra (in the Asia and Pacific
Museum, Warsaw) 1, 91, 92f48, 95–96,
100, 224
304
gold manuscripts (cont.)
‘Life of Gshen rab mi bo’ (Columbia
University manuscript) 91, 93f49, 94,
224 and f 129
and the stitched book format 81–82
Tibetan Book of the Dead 82,
83–85ff34–38
See also gold-color ink; illuminated
manuscripts; raised gold lettering;
Sel dkar Kanjur
hair
hair ornaments identifying female
figures 78
in paper pulp 30
singed hair (dbu skra’i gzhobs) in ink
recipes 101
handling 206, 212, 214–215
and finger print spots 214
water damage 215f 121
See also conservation; preservation
hand-painted decorations
circles imitating holes after palm-leaf
books 56, 86, 98–99
initial-letter painting 86, 92–93
miniature painting 92–93, 141ff73a,b
sketches and doodles 86, 94, 95f51
See also fore-edge painting; illuminated
manuscripts; miniature painting
Heller, Amy 18
buff-colored sheet of the Tibetan
Prajñāpāramitā provided by 105,
105–106ff55–56
indigo-dyed sheet of paper provided by
114–115
hemp (Cannabis sativa)
common name ‘hemp’ used for other
plants 30
cultivation in China 30–31
in early paper samples from Nepal 110
and other organic materials in Russian
types of paper 23
paper mulberry fibers compared with 28
preferred uses of 31–32
and the term ‘ma’ in old Chinese
documents 31
in Tibetan manuscripts from
Dunhuang 30
index
Holmberg, David 193
Hunter, Dard 191–192, 196n.36
illuminated manuscripts: and the concertina
format 60
double covers used with 58, 82–85,
83–85ff34–38
and woven paper produced with floating
molds 33
xylograph prints on left and right sides
of pages 86, 164–165 and 165f97
See also Diamond Sutra; European
medieval manuscripts; fore-edge
painting; gold manuscripts;
hand-painted decorations; miniature
painting; Tibetan book culture;
Tibetan Book of the Dead
Indian book culture
lithography printing in the 1960s 117,
118f66
pustaka (palm-leaf books) 49-53 and
50f 16, 56, 98–99
raw material for paper making 188
spread of papermaking to 51
writing supports before the twelfth
century in 50
See also book cultures; Kashmiri book
culture
indigo
blue paper imitating indigo dyed paper
77, 78f33, 104
indigo-dyed sheet of paper provided by
Amy Heller 114–115
paper dyed by dipping 59, 104, 226
ink
about 99–101
black and red ink of the Wanli Kanjur
94, 154–156, 155f90
blood in ink recipes 100, 101
cinnabar ink of Tibetan Kanjurs 154
cinnabar/vermillion (mercury sulfide)
as a symbol of power 100–101
milk of a white goat (ra dkar gyi ’o ma)
in ink recipes 101
and paper absorbency 168–169, 170–171
preparation for blue/black paper 
42–43
recipes for 100–101
index
singed hair (dbu skra’i gzhobs) in ink
recipes 101
traditional mineral pigments 15–16, 19
See also gold-color ink; rubrics; writing
tools
Institute of Oriental Manuscripts
(St. Petersburg, Russia)
small prayer book from Khara Khoto  and volume 58 of the Berlin Wanli
67, 68f28, 69, 121–122, 121n.16, 250
See also Mongolian Kanjur from Kangxi
times
Islamic bookbinding 62
See also Arabian book culture;
bookbinding
Jackson, David 12, 18, 123
Jagiellonian University Library
(Biblioteka Jagiellońska) in Cracow,
Poland
fragments of the Berlin Wanli Kanjur
preserved in 8–10, 139–140 and
139nn.41–43
Wanli Supplement volumes 140, 156
and n.58
See also Wanli volumes, Berlin Wanli
Kanjur (Pander Pantheon volumes)
Japanese book culture
absorbency of paper 44
papermaking introduced by the Korean
monk Don-cho 179
and paper mulberry 19
and ramie 26
Thymelaeaceae family plant fiber in
gampi and mitsumata paper 44,
188
See also book cultures
Jest, Corneille 129
Jia Sixie (473–554) 108
jute (Corchorus sp.)
combined with bamboo, straw, and paper
mulberry in Peking sets of Tibetan
Kanjurs 162
Dunhuang and other Central Asian sites
110
in Nepalese and Indian papermaking
188
and the term ‘ma’ in old Chinese
documents 31
305
Kalachakra Tantra printed for funeral
ceremonies of Kublai Khan 123,
124ff68–69
Kangxi Kanjur (1684–92): paper structure
and composition of 156–158, 259
size leaves of 147
‘spurs’ on particular letters 153
Kanjur (Pander Pantheon volumes) 
8, 140
See also Institute of Oriental Manuscripts
Kaochang 51, 180
Kashmiri book culture
and Arabian book culture 64 and n.27
birch bark favored by 48–49, 63, 84n.64
bound book on birch bark 63–64, 67
fine quality of 188
See also book cultures; Indian book
culture
Kawasaki, Kazuhiro 64–65, 67n.28
Khara Khoto
IOL Tib M (58 vol. 1:58) 250
IOL Tib M (60 vol. 1:60) 250
small prayer book 67, 68f28, 69, 121–122,
121n.16, 250
and Thymelaeaceae family plant fibers
in documents from 104–105
Korean book culture: absorbency of paper
44
earliest Buddhist text traced to 117 and
nn.4–5
and metal movable type 116n.2
and paper mulberry 19
and ramie 26
See also book cultures
Koretsky, Elaine 26, 35
Kotan 51, 180
Kusha 51, 180
Lalou, Marcelle 52
lapis lazuli
paper made of mentioned in the
Diamond Sutra 103
in pigments 16
Laufer, Berthold 61–62, 184
Lhasa Kanjur (in the Berlin State Library)
paper quality of 175–176, 177f 108
text lines of 177
306
Library of Congress
Appendix 5 list of studied items from
276–287
Brgyad ston ’pa folio in 10
Narthang Kanjur set from 173–174
Rare Book and Manuscript Collection 7
See also Cone Kanjur; Derge Kanjur;
Narthang Kanjur
Library of Tibetan Works and Archives
conservation survey of 10–11
manuscript LTWA 14459 10n.18, 98
woven paper from document from 
34f9a,b
Linrothe, Rob 70
Lotus Sutra 97 and f52, 98 and f53
Loulan (Xinjiang province) 31, 51, 180
Lozang Trinlé, Dungkar (Dung dkar blo
bzang ’phrin las) 42, 125
McClure, Floyd Alonzo 184
March, Kathryn 193
medieval manuscripts. See European
medieval manuscripts
Meinheit, Susan 10
minerals
as pigments 15–16, 19
sal ammoniac in papermaking 195
and n.33
miniature painting
silk coverings and frames for 85
style of 17–18, 19, 86
on the Tibetan Selkar (sel dkar) Kanjur
88and ff42a,b
on the Wanli Supplement volume’s cover
140, 141ff73a,b
See also European medieval manuscripts;
Tibetan book culture
Mongolian book culture
and Tibetan Buddhism 4
See also book cultures; Mongolian
Kanjurs; Tibetan book culture
Mongolian Kanjurs
paper used for 159
and the Peking Kanjur 158–159, 173
pothi format of printed editions 159
(Berlin State Library), paper quality
of 162
(E2, volume 9) printed in 1714
(Institute of Oriental Manuscripts)
162, 260
index
(handwritten) tantra ‘ka’ section
(from St. Petersburg University,
Faculty of Asian and African Studies)
160ff92–93, 162, 258
tantra volume (E59) from Kangxi times
(Institute of Oriental Manuscripts)
159, 160f92–93, 161ff94–95, 162, 259
See also Institute of Oriental Manuscripts
movable type 116–117 and nn.1–2
myrobalan (a ru ra, Terminalia chebula) 42,
43, 102, 103, 195, 195n34
Narthang Kanjur (1730–1732)
Berlin State Library volumes of 173,
176f 107
paper of 174–175, 176–177, 176f 107,
279–280, 294–295
production of 132, 294
woven mold used for paper of 175
See also Narthang Monastery; Tibetan
book culture; Tibetan Kanjurs
produced in Tibet
Narthang Monastery
carving of woodblock for the Narthang
Kanjur 128ff70–71
Narthang Tibetan Buddhist Canon
production 132
See also Narthang Kanjur
National Museum in Warsaw, Tibetan bound
manuscripts in collection of 94, 95f51
Nepalese book culture
book storage methods 216,
217–219ff 123–125
early palm-leaf manuscripts from 49
Edgeworthia gardineri grown 184, 185f 109
gold on blue paper used for manuscripts
80
painting of inner sides of manuscript
covers 56
pothi format books 54f 17, 56
Prajñāpāramitā manuscript from the
Bicher temple (Dolpo) 105–107
and ff55–56
See also book cultures; Nepalese
papermaking
Nepalese papermaking
boiling process 195
early technology of 51
Edgeworthia gardineri (kāgate) and
Daphne (lokta) used in 185f 109, 188
index
and insect-deterring substances 43
long-distance trading of paper 181
ramie and hemp in early sample from
110
See also papermaking processes
Newark Museum, covers of Yongle volumes
in 142
Nia 180
OR 15190. See Tibetan Book of the Dead
(British Library manuscript OR 15190)
organic materials
acacia catechu (ja do) 195
and the Accelerator Mass Spectrometry
(AMS) 20–21
barley flour (thel chin) 195
black vitriol (nag mtsur) 42, 103
as pigments 16
and radiocarbon dating 19–21
resin of pine or fir 100
thar nu (medicinal plant) 42, 103
See also bamboo fibers; birch bark; coral
(byu ru); cotton; flax; gold-color ink;
hemp; indigo; jute; myrobalan; ramie;
straw; Thymelaeaceae family plant
fibers; wood pulp
pagination
Chinese pagination as an aid to and
indicator of Chinese craftsmen 94
and n.15, 119, 122, 154
Chinese pagination on woodblocks of
non-Chinese texts 119–121, 122
defined 154n.56
and the identification of woodblocks
133, 170
paleography
introduced 13
Tibetan scripts 17
See also calligraphy
Pander Collection (Berlinka collection)
deemed lost by Eimer 10, 139n.41
in the Jagiellonian University Library
8–10 and 9nn.14, 16, 139–140 and
139nn.41–43
volume 58 of the Pander Pantheon
139n.43, 140
See also Wanli volumes, Berlin Wanli
Kanjur (Pander Pantheon volumes)
Pander, Eugene, about 8–9, 134
307
papermaking processes
color 41–42
cotton in machine made paper 32
detoxification of 195, 209
distinguished for handwritten manuscripts
and xylograph prints 176
floating molds and woven paper
production 33–34 and ff9a,b, 35,
191–192, 196–199 and ff 113–116
sizing 39–40, 45f 14, 131
spread of 51
straw added for softness and absorbency
176
See also Chinese papermaking; glue;
Nepalese papermaking; papermaking
tools; Russian paper
papermaking tools
conch shell for smoothening 41
movable sieve/dipping method 35–36,
107–108, 110, 191–192 and 192f 112
shok bre (floating mold) 196–199 and
ff 113–116
for smoothening paper 41
yellow dzi (gzhi) used for smoothening
41
See also bamboo sieves; papermaking
processes
paper mulberry (Broussonetia sp.)
Chinese papermaking 26–27, 29–30, 36
in the Cone Kanjur 169 and ff 101–102
cultivation of 28–29
fibers of 29 and ff6a,b, 109ff59–60
in the paper of the Mongolian Kanjur
160f92
in Tibetan/Tangut manuscripts 106–108,
109ff59–60, 246
Peking editions
bamboo, straw, jute, and paper mulberry
in paper of 162
Chinese paper used for 177
fore-edge painting on early editions of the
Peking Kanjur 94
Mongolian Peking editions 158–159, 173
woodblocks required for 129–130
See also Wanli volumes; Yongle Kanjur
Persian book culture
daftar (bound manuscripts) in Mongolia
63
and hemp cultivation 30
See also Arabia
308
’phags pa shes rab kyi pha rol tu phyin pa rdo
rje gcod pa (Diamond Sutra in the Asia
and Pacific Museum, Warsaw) 1, 91,
92f48, 95–96, 100, 224
pigments
based on arsenic or mercury
(insect poison) 93
traditional mineral pigments 15–16, 19
See also fore-edge painting
pothi format
of Cone Kanjur volumes 164 and f96
dpe ras (cloth wrapping) 53, 54f 18, 94
elements of 53, 54f 17
Indian pustaka (palm-leaf book) origins
of 49-53 and 50f 16, 56, 98–99
Kalachakra Tantra printed for funeral
ceremonies of Kublai Khan 123,
124ff68–69
ordering of pages made visible with
fore-edge painting 94
sizes of pages altered by Tibetan scribes
55–56
textile labels of 53, 94
vulnerability to damage of 206–207
See also Cone Kanjur; Derge Kanjur;
fore-edge painting; Lhasa Kanjur;
Mongolian Kanjurs; Narthang Kanjur;
Peking editions; pothi format covers;
Tibetan book culture; Tibetan Book of
the Dead (British Library manuscript
OR 15190); Wanli volumes; Yongle
Kanjur
pothi format covers
blue paper imitating indigo dyed paper
77, 78f33
decoration of 56, 57f 19
and dendrochronology 19, 20f2, 227
double covers used with illuminated
manuscripts in 58, 82–85,
83–85ff34–38
sewing of 130
with string-holes 56, 57f20
wood used for identified with particular
regions 56, 58, 144
See also Yongle Kanjur covers
Prajñāpāramitā manuscripts
Cone Kanjur (Library of Congress)
volumes 164–165 and 165f97, 288–289
index
Diamond Sutra (in the Asia and Pacific
Museum, Warsaw) 1, 91, 92f48, 95–96,
100, 224
from the early ninth century composed
of Paper mulberry 107
as an element of a ritual ensuring good
harvest 48f 16
volume from the Bicher temple
(Dolpo, Nepal) 105–107 and ff55–56
preservation
and conservation science 205–206,
205n.2
preventive conservation 205 and n.2, 220
of Tibetan books 203–204
See also conservation; handling
printing. See Tibetan block-printing
technology; woodblock printing
radiocarbon dating
as a method 19–21
of the Tibetan Book of the Dead from the
British Library 59, 226–227
raised gold lettering: of front pages of the
London Selkar (Sel dkar) Kanjur 86,
87f40, 231
of front pages of the Wanli Supplement
142f74, 140
of front page titles 85, 92–93, See also
gold-color ink
Ramie (Boehmeria sp.)
and Chinese types of paper 22, 26, 31, 32
features of 28f5a,b
in older paper samples from Nepal 110
and the term ‘ma’ in old Chinese
documents 31
restoration
and conservation science 205–206,
205n.2
loss filling 222–223 and ff 127–128
sacred nature of books defaced by
restoration 201
of Tibetan objects and architecture
203–204
Reynolds, Valrae 198
Rock, Joseph, Cone Kanjur acquired by 172
Roman book culture
tablet writing 73 and n.34
See also European book culture
index
rubrics (red lettering)
on European medieval manuscripts 13
and n.1, 71, 97, 97–98
on Tibetan manuscripts, on the Peking
Kanjur (Berlin State Library) 273, 274,
275, 276, 278, 279
See also ink
Russian paper
deterioration of 211
organic materials in 23, 24, 35
Sumkin factory (Afamovo) paper 36–38
and f 11, 39f 12
See also papermaking processes
St. Petersburg. See Institute of Oriental
Manuscripts
Schaeffer, Kurtis R. 47, 122
Sel dkar Kanjur
and the Brgyad ston ’pa folio in the Library
of Congress 10
decoration of 86, 87ff39–40, 88 and
ff42a,b
paper composition 106–107ff57–58, 107
Shafer, Robert 121
Silk, Jonathan 21, 134
sketches and doodles
linear sketch of a dragon 86, 94, 95f51
painted-edge ornamental sketches  and cotton fibers 31, 32
147–150, 148ff80–81, 149f82, 150f83,
151f84a,b
preparatory sketches (par tsak) of text
on woodblocks 126–127
See also hand-painted decorations
Stein, R.A. 52
Stellera chamaejasme (re lcag pa in Tibetan)
as a component in Tibetan papermaking
only 23, 110–111, 111f61
added to papers for softness and
absorbency 112, 176
collection and preparation of 194
and ‘cotton paper’ 184
features of 26, 27f4a,b, 111f61, 112f53, 113f64,
173f 106, 186–188 and ff 110–111
floating molds used with 194 and n.31
in volumes of the Derge Kanjur 172,
173f 106
See also Daphne or Edgeworthia sp.;
Thymelaeaceae family plant fibers
309
Stoddard, Heather 52–53, 64, 67n.31, 75, 79
straw
added for softness and absorbency 176
fibers in the Cone Kanjur 169 and
ff 101–102
tablets
planks (samta/bka’ sam) for daily notes
71, 73
Roman tablet writing 73 and n.34
Takeuchi, Tsuguhito 180
terminology
and challenges faced by paper historians
12, 21–22
edition, re-edition, and reprinting
132–136
three-level set of criteria for the typology
of paper in Tibetan manuscripts
22–23
Tibetan and Chinese words for ‘printing
block’ 121
Tibetan names for books related to their
format 52–53
Tibetan terms for scriptures written in
gold-color ink 81
textiles
as a writing support in India 50
covers for woodblocks 130
Optical microscopy identification of
raw materials of 16
and papermaking sieves 34fb 172, 175
rag-paper manuscripts 30, 32, 100, 172
ramie textiles 26, 27
silk scrolls 48
and thangka painting 69
See also bookbinding; cotton; hemp; pothi
format covers; ramie
Thymelaeaceae family plant fibers
characteristics of 188–189
and Chinese papermaking 184
in documents from Dunhuang 104–105
eaglewood (Aquilaria agallocha, aga ru)
190–191, 194
and Japanese gampi and mitsumata
paper 44, 188
medicinal properties of some species
of 25, 189
310
Thymelaeaceae family plant fibers (cont.)
See also Daphne or Edgeworthia sp.;
Stellera chamaejasme
Tibetan block-printing technology
blocks used as master copies for
re-editions 123, 125, 132
and Buddhism 116
carving of woodblock for Narthang
Monastery 128f71
and movable type 116–117
non-absorbent paper used with 39,
44, 130, 168
production process 125–132
Tibetans as inventors of 121
and the Turfan oasis 121
and the xylograph of the Tshad ma
rigs pa’i gter 133
See also Cone Kanjur; Derge Kanjur;
Dunhuang documents; Lhasa Kanjur;
Narthang Kanjur; Tibetan book
culture
Tibetan book culture
blue paper imitating indigo dyed paper
42–43, 77, 78f33, 104
Buddhism 50–51, 193, 201
sacred books as the embodiment of the
Buddha’s words 47
sacred nature of books defaced by
restoration 201
and scroll painting (thangka) 85, 144
significance of 47–48, 48f 16
Tibetan terms for books 52–53
See also book cultures; calligraphy;
deb ther (Tibetan bound books);
Dunhuang documents; fore-edge
painting; gold-color ink; hand-
painted decorations; illuminated
manuscripts; miniature painting;
Mongolian book culture; pothi format;
Tibetan block-printing technology;
Tibetan papermaking
Tibetan Book of the Dead (British Library
manuscript OR 15190)
bamboo fibers in Chinese papers used
for 32, 59
double covers of 58, 82, 83ff34–35
and pothi Tibetan manuscripts in gold
81–82
See also illuminated manuscripts
index
Tibetan Kanjurs produced in Tibet
Lithang Kanjur 163
papermaking sieve fingerprinting
information 177–178
similarities of editions produced in
Eastern Tibet 173
See also Cone Kanjur; Derge Kanjur; Lhasa
Kanjur; Narthang Kanjur
Tibetan language, diversity of 3–4
Tibetan papermaking: after the 1950s 192
classes of material for manufacturing
paper 189–190
distinguished from Chinese
papermaking 24, 44, 177–178
floating molds and woven paper
production 33–34 and ff9a,b, 35,
171f 104, 172, 191
highly sized and polished surface of
paper 43, 44, 45f 14, 168
paper mulberry fibers in Tibetan/Tangut
manuscripts from, Dunhuang
oasis 106–108, 109ff59–60, 246
preparation of dark and blue paper
42–43, 103–104
shok bre (floating mold) 196–199 and
ff 113–116
and Stella chamaejasme 23, 110–111, 111f61
thickness of 45–46, 170
See also Dunhuang documents; Dunhuang
oasis; glues; Tibetan book culture
Tibetan woodcarvers 129–130
Turfan oasis
early block prints from 119, 120f67, 180
Tibetan block-printing technology 121
typologies of book features, and local book
culture 3
Uighur book cultures
block print from Turfan in Uighur
language 120f67
concertina printed books from the ninth
century 119
and wooden movable type 116n.2
Uray, Géza 180
van der Kuijp, Leonard 63, 122
van Schaik, Sam 7, 52, 180
Vorobieva-Desiatovskaia, Marguarita
Isifovna 55
index
Wangchuk, Dorji 79–80, 101
Wanli volumes
chisel marks in letters of the Yongle/
Wanli Kanjurs 136, 155f89, 166, 292
laid- and chain-lines of paper 36
painted edges of 93–94, 147–149,
148ff80–81, 149f82, 150f83, 15184a,b
Yongle blocks used a master for 123, 125,
139
See also Peking editions
Berlin Wanli Kanjur (Pander Pantheon
volumes)
Jagiellonian University Library
preservation of 8, 139 and n.43
volume 58 of 139n.43, 140
Wanli Kanjur (Berlin manuscript Beck
copy of the Peking Kanjur)
glued layers of paper of 156, 157–158, 162
handwriting characteristics of 166
painted fore edges of 93–94, 15184a,b
‘spurs’ on particular letters 153, 155f90
text characteristics 150, 153 and f87
See also Berlin State Library; Jagiellonian
University
Wanli Supplement volumes
Chinese origins of papers in 32, 36
cover of two volumes of Harvard-Yenching
Library 140, 141ff73a,b, 142f74
cover preserved in volume 58 from the
Pander Pantheon 140
painted edges of 147–149, 148ff80–81,
149f82, 150f83
six text lines per page of 153, 154f88
two volumes in Harvard-Yenching
Library 8, 147, 149, 150f83
Ważny, Tomasz 16, 140n.48, 145ff77a,b
woodblock printing, European tradition
of 116 and n.1
See also Tibetan block-printing technology
wooden pens/sticks
handwriting characteristics of 166
non-absorbent paper required for 158,
168
See also writing tools
wood pulp
and other organic materials in Russian
types of paper 23, 24, 35, See also
organic materials
311
writing tools
about 101–102
for gold manuscripts 80
pen made of stone (brda rdo’ dmyu gu)
80
Roman styli 73n.34
and ‘spurs’ on particular letters 153,
155f90, 165–166 and f98
stylus (bamboo pen) for Tibetan
calligraphy 44, 168
and tools for executing guidelines 98
tool traces as ‘fingerprints’ 16f 1, 129,
153
wooden pens used with samta (tablets)
71, 73
See also ink; wooden pens/sticks
Xuanzang 49, 191
Yongle Kanjur
book cover from the Arthur Leeper
collection 140, 142–143, 145f77a,b
as the first printed edition of the Tibetan
Kanjur 137
folio preserved in Michigan 55, 137, 138
and f72a,b, 147, 152f86, 154, 156, 157f91
red ink used for 138f72a,b, 152f86, 154
Sera Monastery set 138–139
sets of 137–140
and the Wanli Kanjur 123, 125, 137–138,
139, 140, 154, 157
See also Peking editions; pothi format;
Yongle Kanjur covers
Yongle Kanjur covers
cover from Sakya Monastery set
(the R.R.E. Collection) 85, 143 and
f76
cover from Sera Monastery set
(the R.R.E. Collection) 143 and f75,
146f79, 270–271
Potala set (originally Sakya Monastery)
(the R.R.E. collection) 143 and f76,
271
Sakya type cover from the collection
of Arthur Leeper 140, 142–143,
145f77a,b
transcription of inscriptions on 270–272
See also pothi format covers